ETH Z...Diss. ETH. No. 23749 Activity and controlled delivery of epigallocatechin 3-gallate in the...

Research Collection

Doctoral Thesis

Activity and controlled delivery of epigallocatechin 3-gallate inthe treatment of degenerative disc disease

Author(s): Krupkova, Olga

Publication Date: 2016

Permanent Link: https://doi.org/10.3929/ethz-a-010738702

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.

ETH Library

https://doi.org/10.3929/ethz-a-010738702

http://rightsstatements.org/page/InC-NC/1.0/

https://www.research-collection.ethz.ch

https://www.research-collection.ethz.ch/terms-of-use

Diss. ETH. No. 23749

Activity and controlled delivery of

epigallocatechin 3-gallate in the treatment

of degenerative disc disease

A thesis submitted to attain the degree of

DOCTOR OF SCIENCES of ETH ZURICH

(Dr. sc. ETH Zurich)

presented by

Olga Krupkova

MSc in Molecular Biology and Genetics, Masaryk University

Born on 01.09.1986

Citizen of Czech Republic

Accepted on the recommendation of

Prof. Dr. Stephen Ferguson

Prof. Dr. Karin Wuertz-Kozak

Prof. Dr. Benjamin Gantenbein

2016

3

Contents

Zusammenfassung ...................................................................................................................................... 6

Abstract ......................................................................................................................................................... 8

1. Introduction .............................................................................................................................................. 11

1.1 Motivation ................................................................................................................................... 13

1.2 Goal, Aims and Hypotheses ......................................................................................................... 15

1.3 Thesis outline............................................................................................................................... 18

2. Background ............................................................................................................................................. 19

2.1 Anatomy of the spine .................................................................................................................. 21

2.2 Intervertebral disc ....................................................................................................................... 21

2.2.1 Functional anatomy .............................................................................................................. 21

2.2.2 Nutrition and Innervation ..................................................................................................... 23

2.3 Intervertebral disc degeneration ................................................................................................ 24

2.3.1 Extracellular matrix decline .................................................................................................. 25

2.3.2 Reduced nutrition ................................................................................................................. 26

2.3.3 Cell reactions ........................................................................................................................ 26

2.3.4 Gene polymorphisms ........................................................................................................... 28

2.4 Pathologies of the intervertebral disc ......................................................................................... 29

2.4.1 Degenerative disc disease .................................................................................................... 30

2.4.2 Disc herniations .................................................................................................................... 31

2.4.3 Diagnostics of disc pathologies ............................................................................................ 32

2.5 Treatment of low back pain in clinical practice ........................................................................... 33

2.5.1 Physiotherapy and exercise .................................................................................................. 33

2.5.2 Pharmacotherapy ................................................................................................................. 34

2.5.3 Surgical treatments .............................................................................................................. 34

2.6 Therapeutic strategies in preclinical development ..................................................................... 35

2.6.1 Molecular therapy ................................................................................................................ 36

2.6.2 Cell-based regeneration ....................................................................................................... 38

2.6.3 Gene therapy ........................................................................................................................ 39

2.6.4 Bio-enhanced materials ........................................................................................................ 40

2.6.5 Future perspectives .............................................................................................................. 41

4

3. Epigallocatechin 3-gallate suppresses interleukin-1β-induced inflammatory responses in

intervertebral disc cells in vitro and reduces radiculopathic pain in rats ........................................ 43

3.1 Abstract ....................................................................................................................................... 45

3.2 Introduction ................................................................................................................................. 46

3.3 Methods ...................................................................................................................................... 48

3.4 Results ......................................................................................................................................... 54

3.5 Discussion .................................................................................................................................... 59

3.6 Conclusion ................................................................................................................................... 63

3.7 Author’s contributions ................................................................................................................ 63

3.8 Acknowledgement ....................................................................................................................... 63

4. The natural polyphenol epigallocatechin gallate protects intervertebral disc cells from

oxidative stress .......................................................................................................................................... 65

4.1 Abstract ....................................................................................................................................... 67

4.2 Introduction ................................................................................................................................. 68

4.3 Methods ...................................................................................................................................... 70

4.4 Results ......................................................................................................................................... 77

4.5 Discussion .................................................................................................................................... 85

4.6 Conclusions .................................................................................................................................. 88

4.7 Authors’ contributions ................................................................................................................ 88

4.8 Acknowledgement ....................................................................................................................... 89

4.9 Supplement ................................................................................................................................. 89

5. An inflammatory nucleus pulposus tissue culture model to test molecular regenerative

therapies: Validation with epigallocatechin 3-gallate ........................................................................ 93

5.1 Abstract ....................................................................................................................................... 95

5.2 Introduction ................................................................................................................................. 96

5.3 Methods ...................................................................................................................................... 98

5.4 Results ....................................................................................................................................... 103

5.5 Discussion .................................................................................................................................. 108

5.6 Conclusions ................................................................................................................................ 112

5.7 Author’s contributions .............................................................................................................. 112

5.8 Acknowledgement ..................................................................................................................... 112

5.9 Supplement ............................................................................................................................... 113

6. Stability of (–)-epigallocatechin gallate and its activity in liquid formulations and delivery

systems ....................................................................................................................................................... 115

6.1 Abstract ..................................................................................................................................... 117

6.2 Biological effects of (–)-epigallocatechin gallate ....................................................................... 118

6.3 Degradation of (–)-epigallocatechin gallate .............................................................................. 122

6.4 Stability of (–)-epigallocatechin gallate in solution ................................................................... 124

6.5 Stabilization of (–)-epigallocatechin gallate for clinical application .......................................... 127

5

6.6 Conclusions ................................................................................................................................ 134

6.7 Acknowledgement ..................................................................................................................... 134

6.8 Supplement ............................................................................................................................... 135

7. Towards controlled delivery of epigallocatechin 3-gallate as an anti-inflammatory treatment

for degenerative disc disease................................................................................................................. 139

7.1 Abstract ..................................................................................................................................... 141

7.2 Introduction ............................................................................................................................... 142

7.3 Methods .................................................................................................................................... 144

7.4 Results ....................................................................................................................................... 145

7.5 Discussion .................................................................................................................................. 149

8. Discussion .............................................................................................................................................. 153

8.1 Summary.................................................................................................................................... 155

8.1.1 Clinical relevance ................................................................................................................ 155

8.1.2 Clinical applicability ............................................................................................................ 158

8.2 Limitations ................................................................................................................................. 160

8.3 Future steps ............................................................................................................................... 161

8.4 Conclusion ................................................................................................................................. 162

8.6 References ................................................................................................................................. 163

Appendix .................................................................................................................................................... 191

List of abbreviations ........................................................................................................................ 193

Acknowledgement ........................................................................................................................... 196

Curriculum Vitae .............................................................................................................................. 197

6

Zusammenfassung

Bei der degenerativen Bandscheibenerkrankung (im Englischen degenerative disc

disease = DDD) handelt es sich um ein progressives, multifaktorielles Krankheitsbild, das eine

entscheidende Rolle bei der Entstehung von Rückenschmerzen spielt. Ein typisches Merkmal

der schmerzhaften DDD ist die gesteigerte Expression von Zytokinen und Schmerzmediatoren,

die zur Reizung von nozizeptiven Nerven sowie zur neuronale Schädigung beitragen können.

Die klinische Behandlung von DDD zielt zwar auf eine Bekämpfung der Schmerzen ab, diese ist

jedoch rein symptomatisch. Da die angewendeten Methoden nicht die verursachenden

biologischen Vorgänge, wie etwa den Abbau der extrazellulären Matrix oder das Vorkommen

von oxidativem Stress, Entzündungen, Seneszenz und Zelltod bekämpfen, kommt es meist zu

einer Verschlimmerung der DDD. Neue therapeutische Ansätze sind also von Nöten, um die

Behandlung zu verbessern. Vielsprechend als Therapeutika sind jene Moleküle, die die während

der DDD abnormal aktivierten Signalwege regulieren können. Präklinische Studien der

vergangenen Jahre haben vor allem Wirkstoffe mit anaboler Wirkung untersucht, während

anti-entzündliche und anti-oxidative Stoffe erst seit kurzem im Forschungs-Fokus sind, so dass

es bisher kaum Daten zu deren Effektivität gibt. Deswegen hatte die hier vorliegende

Doktorarbeit das Ziel, neue molekulare Therapeutika zu identifizieren und zu testen. Im Detail

sollen solche Therapeutika untersucht werden, die die während der schmerzhaften DDD

vorliegenden molekularpathologischen Veränderungen (Entzündung, Katabolismus, oxidativer

Stress) positiv beeinflussen können. Die durchgeführten Versuche zielten darauf ab, die

Wirksamkeit des natürlichen Polyphenols Epigallocatechin 3-gallat (EGCG) bei DDD zu testen.

Der therapeutische Einfluss von EGCG auf Entzündung, Katabolismus und oxidativen Stress

wurde an primären Bandscheibenzellen untersucht. Die Zellantworten wurden mittels Gen-

und Proteinexpression (RT-qPCR, WB, ELISA) und mittels spezifischen Assay (NF-kB Bindungs-

Assay, Seneszenz-assoziierte β-Galactosidase Färbung, Zellviabilität, mitochondriales

Membranpotential) ermittelt. Um den Einfluss von EGCG auf komplettes Bandscheibengewebe

untersuchen zu können, wurde ein degeneratives Organkulturmodel entwickelt, das aus

bovinem Nucleus Pulposus Gewebe in einem artifiziellen Annulus Fibrosus besteht. Um die

Induktion degenerativer Prozesse und deren Kontrolle durch EGCG nachweisen zu können,

wurde die Gewebezusammensetzung (Proteoglykane, Kollagene, Wasser und DNA) sowie die

Genexpression analysiert. Des Weiteren wurden die analgetischen Eigenschaften von EGCG in

einem Radikulopathie-Rattenmodel mittels von Frey Test überprüft. Um die Wirksamkeit von

EGCG in der klinischen Anwendung sicher zu stellen, wird zurzeit eine Methode zur

intradiskalen Applikation mit retardierter Freisetzung entwickelt. Das Freisetzungssystem

besteht aus EGCG-Gelatine-Mikropartikeln, die mit einem injizierbarem HA-pNIPAM Hydrogel

kombiniert werden. Die Mikropartikel werden durch Electrospraying hergestellt und mittels

7

Rasterelektronen-Mikroskopie analysiert. Zur Ermittlung der Konzentration und Stabilität des

freigesetzten EGCG werden der Ferrous Tartarate Assay und der DPPH Assay verwendet. Nach

Abschluss der Entwicklungsarbeiten kann das Freisetzungssystem an Bandscheibenzellen und

im NP Organkulturmodel getestet werden.

In primären Bandscheibenzellen konnte EGCG die durch IL-1β-induzierte

Entzündungskaskade therapeutisch beeinflussen. Die bei Entzündungen relevanten Pathways

NF-kB, JNK and p38 wurden gehemmt und die Gen- und Proteinexpression von entzündlichen

und katabolen Markern reduziert. Bei Ratten mit künstlich induziertem Bandscheibenvorfall

konnte durch lokale Applikation der radikulopathische Schmerz verringert werden, wodurch die

klinische Relevanz von EGCG weiter bestätigt wird. Zusätzlich zu einer anti-entzündlichen

Wirkung vermag EGCG des Weiteren, den durch oxidativen Stress ausgelösten Zelltod durch

eine Hochregulierung des PI3K/Akt Pathway zu verringern. Im NP Organkulturmodel konnte

EGCG ebenfalls die Expression entzündlicher und kataboler Gene hemmen. Da EGCG also auf

viele der dysregulierten, molekularpathologischen Prozesse bei DDD einen positiven Einfluss

nimmt, hat es hohes therapeutisches Potential für die Behandlung von degenerativen

Bandscheibenerkrankungen. Nachdem die vorhandene Literatur über die Stabilität von EGCG in

Flüssigformulierungen und Freisetzungssystemen umfassend untersucht wurde, konnte eine

geeignete intradiskale Applikationsmethode mit retardierter Freisetzung von EGCG entworfen

werden. Die Herstellung biokompatibler EGCG-Gelatine-Mikropartikel mittels Elektrospraying

konnte bereits erfolgreich umgesetzt werden.

In dieser Doktorarbeit konnte gezeigt werden, dass es sich bei EGCG um einen

vielversprechenden therapeutischen Kandidaten für die Behandlung von DDD handelt. EGCG

konnte nicht nur die Zellen der Bandscheibe vor Entzündungen und oxidativem Stress

schützen, sondern konnte zusätzlich auch das Schmerzempfinden in vivo reduzieren. Um die

klinische Anwendbarkeit von EGCG zu erhöhen, wird aktuell ein minimal-invasiv applizierbares

Freisetzungssystem für EGCG entwickelt. Nach weiteren präklinischen und möglicherweise

klinischen Tests wird sich zeigen, ob die in dieser Doktorarbeit entwickelte Strategie erfolgreich

umsetzbar ist.

8

Abstract

Degenerative disc disease (DDD) is progressive multifactorial disorder of the

intervertebral disc (IVD) and one of the major factors in the development of low back pain.

Painful DDD is typically accompanied by elevated levels of cytokines and pain-mediating

chemicals that irritate nociceptive nerves and promote neuronal damage. Clinical treatments

of DDD are aimed at resolution of pain and do not target the biological hallmarks of disc

degeneration, such as extracellular matrix decline, oxidative stress, inflammation, senescence,

and cell death. As a result, DDD can advance. Novel molecules targeting signaling pathways

involved in DDD are promising therapeutic candidates. Although progress in the use of ECM

anabolics has been made in preclinical research, anti-inflammatory and antioxidant

approaches have not yet been thoroughly explored. Therefore, the overall goal of the thesis was

to identify and test novel therapeutic candidates that can counteract inflammation,

catabolism, and oxidative stress involved in painful DDD.

The experiments conducted in this thesis were designed to test the effect of the

compound epigallocatechin 3-gallate (EGCG) on processes involved in DDD. The effects of EGCG

on inflammation, catabolism, and oxidative stress were investigated in primary IVD cell

cultures. Cellular responses were measured by gene and protein expression (RT-qPCR, WB,

ELISA) and by specific assays (NF-kB binding, senescence associated β-galactosidase, cell

viability, mitochondrial membrane potential). To evaluate the effects of EGCG on the tissue, a

degenerative organ culture model of bovine nucleus pulposus (NP) in an artificial annulus

fibrosus was developed and tissue composition (proteoglycan, collagen, water, DNA) and gene

expression were analyzed. Furthermore, the analgesic properties of EGCG were determined in a

rat model of radicular pain, using von Frey test. To promote clinical translation of EGCG-based

therapy, a system with the ability to provide sustained delivery in the IVD has been designed.

The delivery system consists of EGCG-gelatin microparticles, which are dispersed in injectable

HA-pNIPAM. The EGCG microparticles were produced by electrospraying and analyzed by

scanning electron microscopy. The concentration of EGCG released from the HA-pNIPAM was

measured by ferrous tartarate assay.

In primary IVD cell cultures, EGCG interfered with the IL-1β-induced inflammatory

cascade, inhibiting the key inflammatory mediators NF-kB, JNK and p38 and reducing the

expression of inflammatory and catabolic genes/proteins. Furthermore, local application of

EGCG in a rat model of IVD herniation attenuated radiculopathy, hence supporting the clinical

relevance of this compound. In addition to its anti-inflammatory effects, EGCG also possessed

9

the ability to reduce oxidative stress-induced cell death of IVD cells, via up-regulation of the

PI3K/Akt pathway. This suggests that EGCG constitutes a promising therapeutic option as it

targets multiple deregulated processes in the degenerative IVD. In the NP culture model, EGCG

reduced RNA expression of inflammatory and catabolic genes. After reviewing the stability of

EGCG in liquid formulations and delivery systems, an intradiscal sustained release system for

EGCG, composed of gelatin microparticles and HA-pNIPAM, was designed. Electrospraying of

biocompatible gelatin-based microparticles for the encapsulation of EGCG has been optimized.

In this thesis, evidence was provided that EGCG is a promising therapeutic candidate for

the treatment of DDD, as it can reduce inflammation and oxidative stress in vitro as well as

pain intensity in vivo. To increase the clinical applicability of our findings, a minimally invasive

delivery system for EGCG is currently under development. The question whether our strategy is

optimal for the clinical use will hence be answered in the future, after additional preclinical and

possibly also clinical testing.

.

11

Chapter 1

Introduction

13

1.1 Motivation

Degenerative disc disease (DDD) is a common progressive disorder with both genetic

and environmental components. DDD has been defined as accelerated ageing of intervertebral

disc (IVD), accompanied by inflammation, cell senescence and cell death [1-3], resulting in a

structural failure [4]. It is now established that DDD causes nerve irritation and compression,

largely contributing to the development of low back pain (LBP) [5-7]. Up to 80 % of the

population suffers from LBP at least once in their life, most frequently between the ages of 35

to 55. As LBP constitutes the 2nd leading cause for doctors’ visit and 3rd most common cause for

surgery in the economically active population, LBP plays a crucial role in work absence [8].

Moreover, 5% of all low back pain cases lead to chronic disability [9].

Despite increasing prevalence of discogenic back pain and the consequently high

socioeconomic burden [9], only symptomatic therapies are currently available [10, 11].

Conservative treatment is comprised of analgesics, with side effects such as gastric and

cardiovascular damage or drug dependence [12, 13]. Surgical treatment entails complications

arising from implant displacement, reduced spinal flexibility, and accelerated degeneration of

adjacent IVDs [14]. As current therapies have only limited success, the need for advancement in

long-term therapeutic strategies is apparent.

Therefore, research focus has shifted from symptom-driven medicine towards

regenerative medicine and tissue engineering of the IVD. Nowadays, the objective is to restore

IVD homeostasis and function via inhibiting inflammation, preventing premature aging, and

delaying cell death, ideally in a minimally invasive manner. This can be achieved by intradiscal

injection of biologics (cells, molecules) embedded in biocompatible materials. Although

injected cells can support the existing cell population and “rejuvenate” the aging disc [15, 16],

they are at risk of apoptosis due to the surrounding degenerative and inflammatory

environment [15, 17]. Injectable molecular therapeutics such as anti-inflammatory compounds

[18, 19] or growth factors [20, 21], can reduce inflammation and catabolism, thus enhancing

survival and function of the resident disc cells. Therefore, identifying and testing of novel

compounds that are able to interfere with degenerative processes within the IVD is crucial for

therapeutic advancement. Phytochemicals are good therapeutic candidates for the treatment

of DDD, as they exhibit multiple effects via modulation of several signal transduction and

apoptotic cascades, often with minimal adverse effects [22]. In the future, novel compounds

could be used as single treatments or in combination with conservative therapies or cells.

Drug development process consists of several consecutive phases (Figure 1). An initial

phase of preclinical testing takes place in in vitro models, where biological activities of

candidate compounds are evaluated. Due to the multifactorial nature of DDD, it is challenging

to develop a clinically relevant and reliable in vitro model. It requires thorough understanding

14

of the pathophysiological processes in the IVD tissue and the ability to perfectly represent these

processes in vitro. Despite the existing challenges, a number of cell and organ culture models

have been developed over the past years [23-29]. Although whole-disc organ cultures in

bioreactors appropriately mimic the in vivo status, they do not allow for high throughput

experiments, due to space-related demands. On the other hand, space-saving cell cultures lack

native extracellular matrix (ECM). Recently, a new organ culture model, comprised of a bovine

nucleus pulposus in an artificial annulus fibrosus, has been developed [30]. This model

overcomes space-related constrains, allowing for long-term high throughput experiments.

However, a degenerative phenotype has not yet been induced in this model. Activation of

appropriate degenerative changes in this model will be crucial for advancement in preclinical

testing of compounds that interfere with DDD. Compounds and formulations with successful

outcomes in cell and organ culture studies can further progress towards animal studies and

clinical trials (Figure 1).

Understanding compounds’ degradation properties is essential for successful clinical

translation. The disadvantage of some pharmacological compounds is their low stability,

diminishing their therapeutic effects in vivo and resulting in unsuccessful drug development

process. Encapsulation into protective biomaterials can overcome this drawback by improving a

compounds’ chemical stability and prolonging its activity. In fact, appropriately designed drug

delivery systems are able to improve the clinical outcome and provide long-term therapeutic

effects [31].

Figure 1. Drug development process. Candidate compounds are identified and tested in appropriate cell and organ culture tests. Following successful in vitro tests, the activity of compounds is verified in a suitable animal model before clinical trials are initiated. Then, clinical trials follow. The clinical trials comprise of sequential evaluation of toleration (Phase I), efficacy and dose range (Phase II), followed by trials in a large number of patients to develop a broad database of efficacy and safety (Phase III). After successful clinical trials, the candidate compound can receive regulatory approval, e.g. through the FDA. The approved compound formulation can be applied in the treatment of the disease for which it was designed [32].

15

1.2 Goal, Aims and Hypotheses

The overall goal of the thesis was to identify and test novel molecular therapeutic(s)

with multiple beneficial effects against DDD. Upon identification of a promising candidate

(epigallocatechin 3-gallate, EGCG), four specific aims were formulated to investigate its

properties (Figure 2). Each of these aims constitutes an independent study with objectives and

hypotheses outlined below and in.

Figure 2. Aims of the thesis. Biological effects of epigallocatechin gallate (EGCG) were analyzed in vitro (Aim 1 and 2). A degenerative nucleus pulposus model was developed and biological effects of EGCG were evaluated in this organ culture model (Aim 3). Further, a drug delivery system for intradiscal delivery of EGCG was designed (Aim 4).

Aim 1

The first study aimed to identify the effects of EGCG on the inflammatory and catabolic

responses of IVD cells. This study, published in 2014 in European cells & Materials, forms

Chapter 3: Epigallocatechin 3-gallate suppresses interleukin-1β-induced inflammatory

responses in intervertebral disc cells in vitro and reduces radiculopathic pain in rats.

Inflammatory cytokines IL-1β and TNF-α play a major role in the pathogenesis of DDD

and discogenic back pain [2]. Therefore, compounds interfering with cytokine-induced signaling

pathways can inhibit degenerative process in the IVD. Recently, EGCG showed anti-

inflammatory effects in osteoarthritic (OA) chondrocytes [33], which bear similarities to IVD

16

cells. EGCG reduced an IL-1β-induced response in chondrocytes by interacting with c-Jun N-

terminal protein kinase (JNK), nuclear factor kappa B (NF-κB) and p38 MAPK [34-36]. In another

study, EGCG inhibited the gene expression of cyclooxygenase 2 (COX-2) and prostaglandin E

synthase (PGES), molecules involved in the transduction and sensation of pain [37]. Based on

the promising effects of EGCG on OA chondrocytes, we hypothesized that EGCG can possess

anti-inflammatory and anti-catabolic effects in IVD cells and thus have a potential for the

treatment of inflammation and pain in DDD. To test this hypothesis, we examined the effects

of EGCG on gene and protein expression in IVD cells with simulated inflammation and the

effects of EGCG on IVD-related radiculopathy in rats.

Aim 2

The second study aimed to identify the effects of EGCG on IVD cells exposed to

oxidative stress in vitro. This study, published in 2016 in Oxidative Medicine and Cellular

Longevity, forms Chapter 4: The natural polyphenol epigallocatechin gallate protects

intervertebral disc cells from oxidative stress.

Oxidative stress in tissues is caused by extensive formation of reactive oxygen species

(ROS) and insufficient cellular anti-oxidant capacity [38]. Oxidative stress in the IVD contributes

to local inflammation, enhances expression of matrix degrading enzymes [39], causes cellular

damage and senescence, and activates cell death [40]. Moreover, oxidative damage positively

correlates with age and with the degree of disc degeneration [3, 40-47]. EGCG was shown to

directly interact with ROS or operate in an indirect manner by inhibiting ROS-generating

enzymes, by chelating pro-oxidant metal ions [31], or by enhancing the activity of pro-survival

pathways [48]. Therefore, we hypothesized that EGCG can protect IVD cells against oxidative

stress. To test this hypothesis, we examined the effects of EGCG on senescence and cell death

in ROS-treated IVD cells in vitro and we investigated the involved molecular mechanisms.

Aim 3

Organ culture models are valid alternatives for screening and feasibility testing of novel

therapeutic strategies for the treatment of DDD [23]. However, currently available organ

culture models have several limitations such as size, cost, and durability [23, 27]. The main

objective of the third study was to develop a cost-efficient, degenerative organ culture model

for preclinical testing of molecular therapeutics. An additional aim was to validate this model

with EGCG. This study, under review in the International Journal of Molecular Sciences, forms

Chapter 5: A novel tissue culture model simulating inflammatory degenerative disc disease to

test molecular regenerative therapies: Validation with epigallocatechin 3-gallate.

We hypothesized that a previously described bovine artificial annulus fibrosus system

may constitute a suitable base to develop a degeneration model, either by (1) inducing ECM

17

degradation through injection of a biologically relevant enzyme [49] or by (2) activating

degenerative responses through diffusion of cytokines from the culture media [50]. To test this

hypothesis, ECM composition and the expression of inflammatory and catabolic genes was

examined upon the above mentioned treatments.

Aim 4

The fourth aim focused on understanding and improving the stability of EGCG in order

to enhance its clinical applicability [51]. The fourth aim comprises two studies.

A first study, published in The Journal of Nutritional Biochemistry in 2016, forms

Chapter 6: Stability of (–)-epigallocatechin gallate and its activity in liquid formulations and

delivery systems. The objective of this review was to provide detailed information on prior

findings about EGCG stability in solution and biological systems as well as to evaluate available

encapsulation methods to improve the stability of EGCG.

It has been reported that the molecular structure of EGCG allows fast degradation [52,

53] and limits the efficiency of common encapsulation techniques [31, 51]. Therefore, clinical

translation of EGCG-based therapies is challenging. The second study, presented as a report of

work in progress, forms Chapter 7: Towards controlled delivery of epigallocatechin 3-gallate as

an anti-inflammatory treatment for degenerative disc disease. In this study the intradiscal

delivery system for EGCG was designed. We hypothesized that EGCG can be efficiently

encapsulated using a gentle electrospraying process [54], ultimately prolonging stability and

providing sustained release upon injection. To test this hypothesis, release and activity tests are

being performed.

18

1.3 Thesis outline

Chapter 2 provides relevant background information on the anatomy and function of

the IVD and describes the main biological processes involved in IVD degeneration and DDD.

Further, current clinical treatments for DDD are summarized. The chapter ends with the

discussion of novel therapeutic strategies in preclinical research. Chapters 3 to 7 comprise the

main body of work of the thesis. Chapter 3 to 6 represents unmodified manuscripts, published

or under review in peer-reviewed journals. Chapter 7 includes report on the currently ongoing

study, providing a summary of preliminary data and perspectives for the future research. In

Chapter 8, the findings and limitations are discussed in a general context and the thesis closes

with conclusion.

19

Chapter 2

Background

21

2.1 Anatomy of the spine

The spine has three primary functions: to support the upper body, to allow movement

of the trunk, and to protect the spinal cord and spinal nerve roots. The bony vertebrae and the

surrounding soft tissues including intervertebral discs, muscles and ligaments, perform these

functions [55]. A human spine is composed of 34 vertebrae: 7 cervical, 12 thoracic, 5 lumbar, 5

sacral, and 4 coccygeal vertebrae. In the cervical, thoracic and lumbar spine region, the

vertebrae are connected with 24 intervertebral discs, forming a flexible column that supports

the weight of the trunk and head. The sacral and coccygeal vertebrae are fused. 31 pairs of

spinal nerves exit the spine to transmit motor, sensory, and autonomic signals between the

spinal cord and the body. 8 nerve pairs exit at the cervical region, 12 pairs at the thoracic region,

5 pairs at the lumbar region, 5 pairs at the sacrum region, and 1 pair exits at the coccygeal

region [55]. Muscles and ligaments provide further support to the spinal integrity, and prevent

damages to the spinal column by restricting excessive movement.

2.2 Intervertebral disc

The intervertebral discs (IVD) are cartilaginous joints connecting two adjacent

vertebrae, from the second segment of cervical spine to the first sacral segment. The major

functions of the IVD are to distribute the applied loads to the vertebrae, and to provide spinal

flexibility within the limited range [55].

2.2.1 Functional anatomy

The IVD is anatomically divided into 3 distinct parts with substantially different

structural and mechanical properties: outer annulus fibrosus (AF), inner nucleus pulposus (NP),

and two cartilaginous endplates (CEPs) [1, 56]. The AF is composed of circumferential layers of

lamellae formed by closely arranged fibers of collagen type I. The AF has significant load-

bearing function, provides tensile resistance to the applied loads and aids the maintenance of

adequate NP pressure. The NP, located inside the AF, is predominantly comprised of a loose

network of collagen type II and highly hydrated proteoglycans. This gel-like structure equally

distributes the compressional and torsional stresses applied to the spinal column [57]. The

main differences between the composition of AF and NP are shown in Table 1. The CEPs, that

connect the IVDs with the vertebrae, are composed of hyaline cartilage located against the

vertebral body, and the fibrocartilage adjacent to the IVD. The CEPs are porous and allow fluid

exchange between the vertebrae and the IVD [55, 58].

22

Table 1. Composition of nucleus pulposus (NP) and annulus fibrosus (AF) in the lumbar spine [3, 55, 59]. †Percentage of wet weight of IVD; ‡Percentage of dry weight of IVD.

Water Collagens Proteoglycans Elastin Other proteins

Cells Enzymes

NP 70-90%† 15-20%‡ mainly type II

65%‡ 20-45%‡ Minor

AF 60-70%† 50-60‡ mainly type I

20%‡ 10%‡ Minor

Function Hydrostatic pressure

Tensile strength

Osmotic pressure

Support lamellae and cells

Homeostasis

IVD contains ~1% of cells, responsible for the maintenance of IVD homeostasis [60]. The

cells of the NP are chondrocyte-like with oval shape, whereas the cells of the AF, particularly in

the outer region, are elongated and aligned parallel to the collagen fibers [59, 61]. As the

cellular activities govern the organization and properties of the extracellular matrix (ECM),

healthy IVD cells are essential for normal IVD function. To ensure proper turnover of the ECM,

cells balance the synthesis of macromolecules by the expression of proteinases such as matrix

metalloproteinases (MMPs), aggrecanases, and their inhibitors such as tissue inhibitors of

matrix metalloproteinases (TIMPs) [58, 59]. The main macromolecules synthesized by the disc

cells are outlined below.

Aggrecan

The most abundant proteoglycan synthesized by IVD cells is aggrecan (>2,500 kDa).

Aggrecan is formed by a core protein with ~100 covalently attached glycaosaminoglycan (GAG)

chains of highly sulfated chondroitin sulfate and keratan sulfate. The core protein allows

aggrecan to aggregate with hyaluronic acid (HA). Resulting large-sized complexes have limited

diffusion, hence maintaining their location within the disc. The GAG chains provide aggrecan

with its osmotic properties, and therefore, the IVDs possess the ability to swell and resist

against the compressive loads. The high concentration of aggrecan also protects the IVDs from

neovascularization and innervation [62]. The turnover time for proteoglycans in the disc is

between 3 to 5 years [55].

Collagens

To date, 19 different types of collagens have been identified (collagen I – XIX). Types I to

V are the most abundant, whereas types VI to XIX occur only in small quantities. The IVD is

mainly comprised of the fibers of collagen I and II, with minor amounts of other collagen types

(III, VI, IX, X, XI, XII, and XIV) [55, 61]. The collagen synthesis starts with formation of triple helix

pro-collagen molecules inside a disc cell. Pro-collagen is then secreted into the ECM, where final

collagen assembly takes place. Flexible collagen fibers are spontaneously formed from a bundle

23

of threadlike pro-collagen fibrils and are cross-linked by lysyl oxidases to gain additional

strength [55, 63]. The principal amino acids of the collagen are glycine (35%), proline (12%),

hydroxyproline (10%), and alanine (11%) [55]. Type I collagen fibers form lamellae of the AF and

are found to have a diameter between 1 to 12 μm. Type II collagen fibers in the NP are smaller,

with average diameter of 20 nm [55]. Under the physiological conditions, collagens are stable

throughout the life [64].

2.2.2 Nutrition and Innervation

An adult IVD is an avascular structure, except for the most peripheral region of the AF.

The IVD cells receive nutrients mainly by diffusion from vertebrae, as vertebral capillaries

penetrate through the bone marrow spaces and terminate in the vicinity of the CEPs (Figure 1)

[58, 59]. The majority of fluids received by the IVD are absorbed by the NP, due to the presence

of negatively charged proteoglycans. Their fixed charges are electrically balanced by positive

cations (mainly potassium and sodium) in the interstitial fluid. Upon application of a load, the

NP loses water but not the ions, while removal of the applied load causes rapid rehydration due

to the osmotic gradient in the NP [55, 59, 60]. Therefore, spinal loading, proteoglycan content

and structure of the CEPs play major roles in the nutrition of the IVD cells.

Lumbar IVDs are innervated by a branching network of afferent and efferent fibers that

arise from the spinal nerves. While the NP is aneural, the outer third of a healthy AF is

innervated by sensory fibers with nociceptive and proprioceptive functions (Figure 2) [65, 66].

The AF also contains vasomotor nerve fibers associated with the small blood vessels on the

surface of the AF [55]. The CEPs are also innervated, especially at their center [67].

24

Figure 1. Nutrition (left) and innervation (right) of the IVD. Blood vessels terminate in the vicinity of CEP and nutrients are delivered to the NP via diffusion. Outer part of the AF is innervated via nerve branches arising from dorsal root ganglion (DRG) and spinal nerves (cauda equina). AF = annulus fibrosus, NP = nucleus pulposus, DRG = dorsal root ganglion, CE = cauda equina. Reprinted from [58] with permission from Macmillan Publishers Ltd. Reprinted from [55] with permission from Elsevier. http://www.sciencedirect.com/science/book/9780323079549 http://www.nature.com/nrrheum/journal/v10/n9/full/nrrheum.2014.91.html

2.3 Intervertebral disc degeneration

The lumbar IVDs are the most susceptible to degeneration, as they are too large for

efficient diffusion and carry the highest body load. The ability of human IVDs to withstand the

applied compressive loads gradually decreases throughout lifetime, with the onset of

degeneration observed from the second life decade. IVD degeneration is caused by a

combination of decreased tissue cellularity and molecular, structural, and mechanical

alterations of the ECM (Figure 2) [3]. Age-related disc degeneration starts in the NP and

progressively changes the load distribution and reduces the spinal flexibility [60]. A structural

disorganization of the CEPs affects the transport of nutrients, oxygen and lactic acid across the

disc [1, 61]. As a consequence, activity of the IVD cells is altered, leading to an impairment of IVD

repair [3]. It is difficult to distinguish between the normal IVD ageing and the pathologies of

the IVD due to the similarities in the mechanisms that are involved [60]. Clinically relevant

features associated with lumbar IVD degeneration are described below.

http://www.sciencedirect.com/science/book/9780323079549

http://www.nature.com/nrrheum/journal/v10/n9/full/nrrheum.2014.91.html

25

Figure 2. Early signs of the IVD degeneration appear during the second life decade (left). Tears in the AF start to emerge and begin to spread. The CEP shows early signs of cracking and plugging. Advanced signs of degeneration can be found later in life (right). The IVD dehydrates and bulges in several locations. The border between the AF and the NP becomes indistinctive and multiple circumferential and radial tears form in the AF, allowing the NP to move through the AF. The CEPs become thinner, dry and brittle [55]. AF = annulus fibrosus, NP = nucleus pulposus, CEP = cartilaginous endplate. Reprinted from [55] with permission from Elsevier. http://www.sciencedirect.com/science/book/9780323079549

2.3.1 Extracellular matrix decline

The most significant biochemical change of a degenerating NP is the age-related loss of

proteoglycans. For normal IVD function it is essential that the aggrecan content, charge and

size remain as large as possible. Aggrecan can be enzymatically cleaved by proteinases such as

the MMPs and aggrecanases [68]. The proteinases are produced by the IVD cells to regulate

the ECM turnover, however, in a degenerated IVD, the expression of these proteinases

increases [69]. Although IVD proteinases degrade collagens (MMP1, 2, 3, 7, 8, 9, 12, 13), they also

actively degrade aggrecan, especially MMP3, 7 and 12. The most active aggrecanases (a

disintegrin and metalloproteinase with thrombospondin motifs, ADAMTS) are ADAMTS4 and 5

[68-70]. HA can be degraded by hyaluronidases (especially HYAL2) and non-enzymatically by

free radicals [71]. In addition, aggrecan and collagens are negatively affected by irreversible

cross-linking, formed by non-enzymatic glycation due to the high blood glucose. The resulting

advanced glycation end products (AGEs) accumulate in long-lived proteins of the IVD,

preventing their repair and turnover [72, 73].

ECM turnover can be influenced by mechanical loads. Physiological loading can restore

proteoglycans in the IVD [74]. On the other hand, excessive dynamic and static compressive

loading activates MMP synthesis, leading to decreased aggrecan content and cell viability [75,

76]. Similarly, excessive torsional and multi-axial loads promote lamellar disorganization of the

AF, leading to the formation of fissures [55, 61]. The formation of fissures alters the stress

distribution in the AF and result in the loss of disc height, increasing the risk of annular injuries


26

and load-induced structural failure (herniation) [60, 77]. The loss of aggrecan also promotes the

neural and vascular ingrowth into the IVD [62].

2.3.2 Reduced nutrition

The IVD is the largest avascular organ in human body. As mentioned before, the

majority of the IVD cells receive their nutrients by diffusion through the CEPs from vertebral

capillaries. Throughout the lifetime, the CEPs calcify, become thinner, brittle and lose their

permeability for nutrients [55, 78]. The malnutrition of IVD cells can also occur with

pathological changes of the vertebrae such as trabecular fractures, inflammation, or

obstruction of capillaries [58]. Once nutrient supply is restricted, oxygen and glucose levels in

the center of the NP drop and at the same time lactic acid, which is produced by the IVD cells at

high rates, accumulates and reduces tissue pH [78]. Inefficient transport of solutes have

multiple negative consequences such as accelerated cell death, reduced ECM production, and

increased ECM degradation [78].

Cells in degenerative discs are able to activate the production of vascular endothelial

growth factor (VEGF) [79]. This suggests that attenuated nutrient supply can promote vascular

ingrowth, possibly as an attempt to enhance the availability of nutrients. However, an

abnormal blood supply in an originally avascular tissue can increase oxidative stress and

promote cytokine release. Moreover, in severely degenerated discs, newly formed blood vessels

are frequently accompanied by nociceptive nerves, mediating pain sensation [1].

2.3.3 Cell reactions

Inflammatory phenotype

Inflammatory cytokines are key molecules involved in IVD degeneration (Figure 3) [80].

The event that promotes cytokine production by IVD cells, especially in the absence of trauma

or herniation, is usually chronic stress such as smoking, IVD overload or spinal infection [2, 81].

Increased levels of interleukin-1 beta (IL-1β) and tumor necrosis factor alpha (TNF-α) are

frequently found in the degenerative IVDs. The presence of IL-1β and TNF-α correlates with the

degradation of ECM [28, 82-84], as cytokines can induce the expression of collagenases (MMPs)

and aggrecanases (ADAMTSs) [85-87]. Mechanistically, IL-1β binds to a cell surface receptor IL-

1R1, activating nuclear factor κB (NF-κB), c-jun N-terminal protein kinase (JNK), and p38

mitogen-activated protein kinase (MAPK). The activation of these signaling molecules results in

the transcription of inflammatory and catabolic genes such as interleukins (IL-1α, IL-1β, IL-6, IL-

8), matrix metalloproteinases (MMP1, MMP3 and MMP13), and molecules that promote pain

(COX-2, NGF, iNOS) [2, 86]. Similarly, the interaction of TNF-α with its receptor TNFR1 leads to an

27

activation of NF-κB and MAPK pathways through tumor necrosis factor receptor-associated

factors (TRAFs) [2, 88]. In addition to the catabolic and pro-inflammatory effects, IL-1β and TNF-α

can influence cell senescence, autophagy and expression of genes involved in proliferation [80].

Inflammatory cytokines play important role in the neuronal sensitization.

Figure 3. Signaling pathways activated in IVD cells by inflammatory cytokines IL-1β (left) and TNF-α (right). After the interaction with their specific receptors, cytokines activate key inflammatory mediators (p38, JNK, NF-kB), which results in the expression of inflammatory and catabolic genes. This further promotes tissue inflammation and pain [2, 18]. IL-1R = interleukin 1 receptor, IRAK-1 = interleukin 1 receptor-associated kinase 1, TNRR1 = TNF receptor, TRAF = TNF receptor-associated factor.

Cell senescence and cell death

Although cells form only ~1% of the IVD, they are extremely important for IVD

maintenance [61]. Environmental factors such as non-physiological loading, oxidative stress, or

acidic environment can alter the cellular function and enhance premature senescence and cell

death in the IVD (Figure 4). Cell senescence is defined as permanent, irreversible replication

arrest, which occurs after a certain number of cell divisions [89], or when the cell encounters

sub-lethal stress. Senescent cells with reduced functionality accumulate in aged tissues, secrete

pro-inflammatory cytokines and promote tissue fibrosis [90, 91]. Stress-induced death of the

IVD cells, executed either through apoptosis or necrosis, is frequently found during IVD

degeneration [39, 61, 92-96]. The senescence and death of IVD cells do not only positively

correlate with age, but also with the degree of disc degeneration [3, 40-47].

Oxidative damage of macromolecules, mediated by reactive oxygen species (ROS)

including superoxide anion radical (O2−), hydroxyl radicals (OH·) or singlet oxygen (1O2), is one of

28

the most common causes of senescence and cell death. Cellular antioxidants such as

superoxide dismutase, catalase and glutathione peroxidase counteract the formation of ROS,

but their activity decreases with age [97]. Additional radical scavengers namely vitamins C and

E, carotenoids, or flavonoids, can be obtained through dietary sources, but the delivery of the

antioxidants into the disc is dependent on healthy blood supply [98].

Figure 4. Signaling pathways activated in IVD cells by oxidative stress. Sub-lethal damage (left) leads to the activation of p53-p21 and/or p16 pathways, causing irreversible cell cycle arrest via the inhibition of protein RB. Lethal stress (right) commonly induces intrinsic apoptotic pathway. Depolarized mitochondria release cytochrome C, which is implicated in the activation of caspase 3 or 7. The activation of caspase 3 or 7 executes apoptosis. Anti-apoptotic Bcl proteins (such as Bcl-xL or Bcl-2) and pro-apoptotic BH3 proteins (such as PUMA or NOXA) play major roles in the regulation of this pathway [40, 44, 99]. 2.3.4 Gene polymorphisms

Although the pathogenesis of IVD degeneration is not completely understood, it was

found to be associated with several genetic polymorphisms. Polymorphic genes are

represented by more than one allele within a population, and this allele commonly causes

abnormal gene expression and protein production. A polymorphism can alter promoter activity,

influence transcription factor binding, may result in shorter transcript, or faster product decay.

Investigating the polymorphisms associated with IVD degeneration could provide an insight

into the disease mechanism, reveal therapeutic targets, and provide useful information for the

identification of patients at risk [100, 101]. The important gene polymorphisms are outlined in

Table 2.

29

Table 2. The effects of selected polymorphisms on gene and protein level, and their participation in the

degenerative process of IVD [100]. TF = transcription factor, GDF = Growth Differentiation Factor 5, VDR

= vitamin D receptor, COL = collagen, MMP = matrix metalloproteinase, IL = interleukin, COX =

cyclooxygenase, PGE = prostaglandin E, uk = unknown.

Gene Function Functional change GENE

Functional change PROTEIN

GDF5 Growth and repair ↓ transcription ↓ growth and repair

VDR Sulfate concentration ↓TF binding , ↑ decay ↓ amount of sulfate in proteoglycans

Aggrecan Structural Shorter allele ↓ functional proteoglycan

COL1 Structural ↑ TF binding Impaired COL1 protein

COL9 Structural Addition of Tryptophan ↓ interactions with matrix molecules

Fibronectin Structural Fragments ↓ aggrecan, ↑ MMP

MMP2 Catabolic ↑ TF binding ↑ MMP

MMP3 Catabolic ↑ promoter activity ↑ MMP

IL1 Inflammation ↑ transcription ↑ IL1 and inflammation

IL6 Inflammation uk ↑ IL6 and inflammation

COX-2 Pain sensation uk ↑ COX2, PGE, pain and inflammation

2.4 Pathologies of the intervertebral disc

A pathological IVD degeneration is defined as accelerated ageing process of a disc,

accompanied with pain and structural failure [4]. Two types of pain are commonly associated

with IVD degeneration: a nociceptive pain, caused by a neuronal sensitization due to a tissue

damage, and a radicular pain, resulting from neuronal damage and inflammation (Table 3) [55].

Therefore, a degenerated IVD invaded with newly formed nerve fibers is the primary source of

nociceptive back pain [61], while IVD-related compression of nerve roots and spinal nerves

causes radiculopathy and muscle weakness [55, 66]. Both sensitization and damage of the

nerves can occur simultaneously. Although there can be various reasons for low back pain, an

IVD-related problem is generally accepted to be the most common cause. Two of the major

problems associated with the IVD are the DDD and the IVD herniation. These painful disc

pathologies have a strong impact on the quality of life [9].

30

Table 3. Classification and causes of IVD-related back pain. Pain arising from the IVD can be caused by internal disc disruption, injury or herniation. IVD-related pain have both nociceptive and neuropathic component (radicular pain). AF = annulus fibrosus, CEP = cartilaginous endplate.

IVD-related pain

Nociceptive Radicular

Internal disc disruption Internal disc disruption

Torsion AF injuries Herniations

Compression CEP injuries Loss of disc height

↓ ↓

Nerve sensitization Nerve damage

2.4.1 Degenerative disc disease

Both, the outer part of the AF and the central part of CEP are innervated by sensory

nerves. These nerves can be sensitized by a variety of chemicals directly released from disc cells,

among which cytokines are the most prominent. Both TNF-α and IL-1β have been implicated in

nerve sensitization and nerve ingrowth in a non-herniated degenerative IVD [66]. They act as

nociceptive triggers and also induce the expression of other potentially nociceptive cytokines

(e.g. IL-6, IL-8, IL-12, IL-17) and direct pain mediators (nitric oxide, prostaglandin E) [66, 80, 102-

104]. Neuropeptides such as substance P are located in the sensory nerve endings [105] and also

produced by the IVD cells [106]. By-products of disc cell metabolism such as lactic acid can also

be noxious [58, 107]. Additionally, IVD cells express neurotrophins such as nerve growth factor

(NGF) and brain-derived neurotrophic factor (BDNF), which contribute to neo-innervation of the

disc [106, 108]. Molecules implicated in nociception are listed in Table 4.

Early degenerative disc disease

It has been reported that the molecules larger than 400 Da have difficulties to

penetrate across a healthy IVD [66, 109]. Early degenerative IVDs still contain a dense

proteoglycan network, which does not allow for free diffusion of large cytokines such as TNF-α

and IL-1β, reducing their ability to reach the nerves in the AF and CEP. However, cytokines

released in the center of NP disturb local homeostasis and enhance degradation of the ECM,

hence facilitating a progression of the inflammatory process from the NP towards the AF [110].

On the other hand, smaller pain-mediating molecules that are able to diffuse freely can be

implicated in the pain sensation at earlier stages of DDD.

Advanced degenerative disc disease

A moderate or advanced IVD disruption leads to a direct chemical stimulation of

nociceptors by cytokines. Pain develops after torsional injuries in the AF (collagen disruption),

with neural ingrowth into the annular tears worsening the outcome, or after compression

31

injuries, causing nerve irritation [55]. Compressive loading was shown to cause CEP fractures

and internal disc disruption [26, 111], with invading leukocytes possibly worsening the process

[112]. An exposure of NP to the blood circulation breaks the immune privilege, further

promoting infiltration of nerves and blood vessels and resulting in immune reactions within

the IVD. It has been suggested that not only the proteoglycans, but also the IVD cells contribute

to the immune privilege of the IVD. Fas ligand (Fas-L) presence in the membranes of healthy

human NP cells can prevent angiogenesis in the IVD by inducing Fas-mediated apoptosis of

vascular endothelial cells and white blood cells [113]. However, Fas-L expression and the ability

of NP cells to induce Fas-mediated apoptosis of leukocytes decrease with advancement of IVD

degeneration [114]. Compression and torsion injuries in some cases progress to IVD herniations

and spinal stenosis [55].

Table 4. Molecules released from IVD cells and implicated in nociception. NGF = nerve growth factor, BDNF = brain-derive neurotrophic factor, CGRP = calcitonin gene-related peptide. M = membrane, S = soluble. Molecule Size Role References

TNF-α 26 kDa (M) 17 kDa (S)

Pro-inflammatory cytokine Neuronal sensitization

[80, 88]

IL-1β 17 kDa Pro-inflammatory cytokine Neuronal sensitization

[2, 88]

Nitric oxide 30 Da Inflammatory mediator Neuronal sensitization

[102, 115]

Lactic acid 90 Da Glycolysis end product Neuronal sensitization

[116, 117]

COX-2 69 kDa Inflammatory mediator Converts arachidonic acid to prostaglandins

[18, 39]

Prostaglandin E 352 Da Inflammatory mediator Transmits pain signal within the disc

[118, 119]

Substance P 1.35 kDa Pro-inflammatory Transmits pain signal within the disc

[105, 106]

CGRP 15 kDa Transmits pain signal in neurons [108, 120]

NGF 26 kDa Nerve growth Neuronal sensitization

[106, 121]

BDNF 14 kDa Nerve growth Neuronal sensitization

[106, 121]

2.4.2 Disc herniations

The lumbar IVDs are in close contact with dorsal nerve roots. Therefore, lumbar IVD

herniations are considered to be a major cause of radicular pain (Figure 5). Prolapsed IVD can

apply pressure on the spinal cord, dorsal root or DRG, inducing nerve damage and pain. A

32

radicular pain occurs along the nerves branching from the affected dorsal root. Tissues

surrounding the affected nerves are sensitized to motions and pressures that normally do not

cause pain [55]. In addition, pain-mediating chemicals released from the herniated NP activate

inflammation of the DRG and contribute to functional and structural nerve damage [122]. Apart

from the herniations, loss of disc height also causes mechanical deformation of DRGs by

compression, which is associated with radiculopathy [123].

Figure 5. Two major types of IVD herniation. In IVD protrusion (left), the NP has herniated through several lamellae of the AF and pushes on the spinal nerves, but has not penetrated through the most external lamella. In IVD extrusion (right), the NP has passed through the most external lamella of the AF and comes in contact with the spinal nerves or DRG. NP = nucleus pulposus, AF = annulus fibrosus, DRG = dorsal root ganglion, CE = cauda equina. Reprinted from [1] with permission from Elsevier. http://www.sciencedirect.com/science/book/9780323079549

2.4.3 Diagnostics of disc pathologies

Most spinal structures are innervated by nerves from multiple vertebral levels, which

poses difficulties in the diagnostics of back pain [55]. Nevertheless, axial back pain is often

diagnosed as DDD. Diagnostic methods used in patients with back pain include lumbar

radiographs, computed tomography (CT) scans, magnetic resonance imaging (MRI) and

provocative discography [124]. The lumbar radiographs are commonly used to assess the

anatomy of vertebrae and exclude other diagnoses such as displacement (spondylolisthesis),

fractures or scoliosis of the spine. The 3D CT scans are an alternative diagnostic method to

better assess the spinal canal, the anatomy of vertebrae, the appearance of osteophytes, and

endplate sclerosis. Although CT scan does not directly visualize the soft tissues, it allows

diagnostics of disc herniations [124, 125]. With MRI images, soft tissues are visualized, hence

permitting direct assessment of the nerves and discs. A darker MRI signal in the center of the

disc is associated with a loss of proteoglycans, whereas brighter irregular signals suggest the

presence of inflammation or infection [124]. Other common structural changes visualized using


33

MRI include radial fissures, radial bulging of the annulus, reduced disc height, osteophyte

formation and endplate defects. The major problem of MRI as a diagnostic method for DDD is

the fact that scans of symptomatic patients sometimes do not show any morphological signs

of IVD degeneration and vice versa, the degenerated discs are not always painful. It has been

reported that IVDs of up to 30% of asymptomatic volunteers contained protrusions and

herniations as well as reduced proteoglycan content [124]. Issues in diagnostics cause

difficulties to determine the clinically relevant structural hallmarks of DDD [100]. Nowadays it

is becoming widely accepted that persistent inflammation and innervation, rather than other

morphological changes, are the key factors distinguishing symptomatic and asymptomatic

discs [126].

In symptomatic patients, lumbar discography can help determining which disc is

painful. An injection of radiopaque dye followed by a CT scan enables the evaluation of disc

morphology and its defects, as the dye can leak through the disrupted layers of the AF. The

discography also reproduces the patient’s symptoms, since flux caused by injection of the dye

irritates the nerve endings. The discography is the only currently available diagnostic method to

evaluate the actual pain response of a patient [124]. However, it has been reported that needle

puncture of the AF can accelerate degeneration, especially in healthy IVDs that are used as

controls for reliable evaluation of painful disc [127].

2.5 Treatment of low back pain in clinical practice

Low back pain can be treated conservatively by physiotherapy and pharmacotherapy.

However, the success of these methods is not always guaranteed. Especially in patients with

chronic pain, complete eradication of pain can almost never be achieved. The long-term goals,

when treating the patients with chronic pain, are moderation of pain and decreased healthcare

utilization [128]. Upon failure of conservative treatment, a disc surgery is commonly performed.

The treatment strategies in the clinical practice are described below.

2.5.1 Physiotherapy and exercise

It is accepted that rehabilitation and exercise can reduce chronic low back pain and

disability [129, 130]. Physical therapies possibly promote healing of the disc by stimulating cells,

boosting metabolite transport, and preventing re-injury by strengthening back muscles [131].

However, excessive loading can have deleterious effects on both, tissues and cells, and further

promote the degenerative cascade [132]. Recently it has been suggested that the use of isolated

lumbar extension (ILEX) may be the most effective exercise to reduce LBP, thus potentially

improving the disc condition [130]. ILEX trains isolated muscles that extend the lower back and

34

represents a relatively high loading on the IVD at a low frequency [129]. This can promote fluid

flow across the disc, which in turn also enhances transport of larger molecules into the NP

[109]. However, in many cases, physiotherapy and exercise do not lead to major improvement

[133] and it is still unknown whether specific loading modes applied through exercise can heal

or regenerate the IVDs.

2.5.2 Pharmacotherapy

The most common analgesics for discogenic pain are paracetamol, non-steroid anti-

inflammatory drugs (NSAIDs) and opioids. Paracetamol is generally safe, but does not have

significant anti-inflammatory effects and likely does not ameliorate the neuropathic

component of pain [134].

Traditional NSAIDs (e.g. ibuprofen, aspirin) inhibit both COX-1 and COX-2, enzymes

involved in the synthesis of prostaglandins. More advanced COX-2 NSAIDs (e.g. celecoxib)

selectively inhibit COX-2-mediated synthesis of pro-inflammatory prostaglandins. NSAIDs are

effective for short-term symptomatic relief in patients with acute low back pain, but are

ineffective in the management of chronic pain and pain arising from neural damage [12, 135,

136]. Traditional NSAIDs can cause gastrointestinal complications in elderly, while specific COX-

2 inhibitors have less such side effects [137]. In addition, the long-term use of NSAIDs is

associated with serious cardiovascular events [12, 138, 139].

Opioids (e.g. morphine, oxycodone) are effective in treating chronic back pain with a

neuropathic component [13, 140]. They interact with neuronal opioid receptors in the spinal

cord and brain, alleviating pain signals. However, their use poses the risk for analgesic tolerance

over time [12, 141]. Complications of opioid use further include addiction and overdose-related

mortality, which have risen in parallel with the prescription rates [142]. Despite these

limitations, opioids are considered both effective and safe if used appropriately [143].

The majority of patients can be successfully treated by conservative therapies. However,

especially in the case of chronic back pain, these therapies can fail. None of the available dugs

can be used on the long term basis without side effects. Therefore, a surgical treatment is

required for those patients suffering from prolonged, life-limiting pain.

2.5.3 Surgical treatments

Injection therapy is usually the first line of more invasive treatment offered to the

patient. A needle is used to deliver drugs (local anesthetics, NSAIDs, opioids) into and around

the IVD and the nerve roots. This allows local delivery of higher doses, but inflammation and

pain is suppressed only temporarily. As the degenerative process progresses, disc surgery

usually follows [128].

35

Spinal fusion is the gold standard surgical treatment for disc removal. It alleviates pain

by removing a painful disc and preventing movement between affected spinal segments.

Overall, lumbar fusion mostly results in significant reduction in back and leg pain intensities

[128]. However, fusion of a spinal segment may have detrimental effects on the normal

physiological and biomechanical function of the rest of the spine, causing adjacent segment

disease [128].

Total disc replacement (TDR) may be able to prevent adjacent segment disease, as it

allows motion to be preserved at the “degenerated” level. Pain relief is thought to be due to a

combination of removal of the painful disc and improvement of load transfer. Short-term trials

of TDR indicate quicker recovery and better clinical outcome compared to fusion surgery.

However, 5-years follow ups show less significant differences between these 2 groups and the

long term effectiveness and safety of TDR is unclear. Controlled trials with relevant control

group and long-term follow-up are needed to evaluate the efficacy and safety of TDR [128, 144].

Pain arising from herniated lumbar discs can be treated with conservative therapy and

normally improves in 6 weeks. Surgical removal of herniated tissue (discectomy) is performed

when the nerve root is compressed and at the same time no improvement was observed after 6

weeks of conservative therapy. Discectomy have high success rates, but disturbed nerves may

become a source of chronic pain later in the life [145]. Discectomy is also one of the most

common risk factors for recurrent disc herniation [146].

2.6 Therapeutic strategies in preclinical development

Biological hallmarks of DDD, including inflammation, catabolism, senescence, and ECM

decline, can be specifically targeted. In the future, strategies such as in vivo molecular and cell

therapies may achieve inhibition of catabolism and activation of ECM production in the disc. In

more advanced stages of the DDD, degenerated discs could be replaced with tissue-engineered

implants.

Currently, molecular mechanisms of novel therapeutics are being evaluated in cell and

organ cultures, derived from human or animal discs, as well as in animal models. Organ

cultures provide valuable information about the ECM, but they possess limitations related to

explant size and culture durability. Although there are large differences between the tissue

composition, anatomy or physiology of animal species and human, large animal models are the

last and most important part of the complex preclinical testing. Biological treatments with

promising results in research and preclinical development are outlined in the following

subchapters.

36

2.6.1 Molecular therapy

Molecules promoting ECM regeneration

Growth factors are polypeptides involved in the maintenance of the disc ECM integrity

[147]. Their application as ECM stimulating agents is promising, as growth factors, namely bone

morphogenetic proteins (BMP-2, BMP-7), insulin growth factor (IGF-1), tumor growth factor

(TGF-β3), epidermal growth factor (EGF), basic fibroblast growth factor (bFGF) and growth

differentiation factor-5 (GDF-5), induce the formation of new disc ECM via interaction with

their surface receptors on the disc cells [147]. The drawbacks of exogenous growth factors are

their costs and their short half-life (hours to days), resulting in only transient effects and the

need for repeated injections. Another complication is their large size, which can impair their

diffusion through the NP [109]. Moreover, optimal doses of growth factors for clinical use still

remain to be elucidated.

Link protein is found in the disc ECM, stabilizing the interaction between

aggrecan/hyaluronan aggregates and blocking the access for hyaluronidases. The N-Terminus

of Link Protein (LinkN) is a peptide cleaved by stromelysin from the Link protein, acting similarly

as growth factors hence stimulating ECM synthesis [21, 147].

Notochordal cells are present in young NP tissue. They play an important role in the

maintenance of a healthy IVD by secreting growth and pro-survival factors. However, in

humans these cells disappear during the first life decade, which is thought to contribute to the

early onset of disc degeneration. Notochordal cell-conditioned medium (NCCM), obtained from

either isolated notochordal cells encapsulated in alginate beads, or directly from the

notochordal cell-rich NP tissue, is capable of stimulating the ECM production in the disc organ

cultures [148]. The drawback of the use of NCCM is the uncertainty in the composition of

NCCM. Therefore, the identification of bioactive factors and their active concentrations in

NCCM is crucial for further development of this technique [149].

Similarly, blood plasma that has been enriched with platelets (platelet rich plasma, PRP)

can be used to stimulate healing of bone and soft tissues. Activated platelets release various

growth factors (e.g. PDGF, CTGF, bFGF, EGF, IGF1) and healing-promoting cytokines. This

mixture maintains the intervertebral disc homeostasis by shifting the cellular catabolism

towards the anabolic state, as shown in the cell and organ culture studies [20, 150]. However,

the composition of PRP is variable and requires standardization for the clinical use.

Other molecules, such as anti-catabolic enzymes (e.g. TIMPs), intracellular regulators

(e.g. Sox-2) [147] and inorganic polyphosphates [151] also have promising anabolic effects on the

ECM.

37

Molecular antagonists of degenerative pathways

Activation of inflammatory pathways is caused by the imbalance between the

inflammatory cytokines and their antagonists in the disc tissue, commonly observed during

human disc degeneration [152]. Interleukin-1 receptor antagonist (IL-1Ra) was shown to inhibit

ECM degradation in an intact degenerative human IVD [153]. Similarly, TNF-α receptor

antagonist inhibits inflammatory responses in the human IVD cells [154]. TNF-α receptor

antagonist (etanercept) or antibodies (adalimumab) that attenuate TNF-α-mediated effects

and reduce inflammation and pain are currently under clinical investigation [155].

As mitogen-activated protein kinases (MAPKs) are directly affected by pro-inflammatory

cytokines and regulate progression of inflammation, they are promising targets for the

treatment of DDD. Several specific small molecule MAPK inhibitors of p38, JNK and ERK are

available (e.g. SB202580, SP600125, PD0325901). The role of p38 in the expression and activity of

COX-2 is well established [156] and therefore p38 inhibition reduces cytokine-activated

prostaglandin E2 (PGE2) accumulation in the IVD [157]. However, gene expression of MMPs and

cytokines in the disc cells is more complex, commonly regulated by p38, JNK and ERK together

[18, 158, 159]. Therefore, a simultaneous inhibition of multiple MAPKs can synergistically block

the inflammatory and pain reactions arising from the IVD.

NF-κB is another direct target of pro-inflammatory cytokines. Therefore, NF-κB

inhibition can become a valuable therapeutic strategy for DDD. In healthy cells, an inactive NF-

κB complex binds to an IκB inhibitor located in the cell cytoplasm. In contrast, IκB degradation

allows nuclear translocation of NF-κB subunit p65 during inflammation. Its transcription

activity correlates with the degenerative changes of the IVD [160]. IL-1β-activated NF-κB

increases the expression of matrix degrading enzymes, promoting ECM degradation in the IVD

[160]. In addition, the expression of VEGF in IVD cells is regulated by TNF-α-activated NF-κB

[161]. Small molecule inhibitors of NF-κB such as MG132, BAY11-7082 and IMD0354 are available.

Although these inhibitors can be used as therapeutic agents, they have limitations. MG132 is a

proteasome inhibitor, preventing IκB phosphorylation, ubiquitination and degradation.

However, MG132 is not NF-κB-specific and can thus interfere with normal protein turnover and

activate JNK-mediated apoptosis [162]. BAY11-7082 and IMD0354 block specifically IκB

phosphorylation [163], however, their overall effects on the cells are not yet well understood

[164]. Another therapeutic target associated with NF-κB is receptor activator of NF-κB (RANK), a

member of the TNFR molecular sub-family, as the expression of its ligand (RANKL) is elevated in

DRG during disc-related pain. RANKL antibodies suppress the expression of TNF-α and IL-6 in

injured lumbar IVDs as well as the expression of CGRP in DRG neurons. This indicates the

possible use of RANKL as a pain control agent in the patients with lumbar IVD degeneration

[165].

38

Oxidative stress in the IVD is caused by the imbalance between the production of ROS

and the activity of the antioxidant protection systems. As cellular capacity to repair oxidative

damage decreases with age, antioxidants can constitute a therapeutic strategy to counteract

IVD degeneration [98]. It has been shown that antioxidants such as N-acetyl cytosine (NAC) or

ferulic acid alleviate the detrimental effects of oxidative stress in the IVD cells [166, 167].

Oxidative stress is also induced by high glucose (e.g. in diabetic patients), causing accumulation

of AGEs [168] and premature senescence of IVD cells in vitro and in vivo [19]. A combination of

pyridoxamine (AGE inhibitor) and pentosan polysulphate (anti-inflammatory) was shown to

protect diabetic mice from high glucose-induced IVD degeneration [19].

In addition to the inflammatory and nociceptive factors, degenerative and painful

human IVDs release NGF and BDNF, promoting pathological nerve ingrowth [108]. An

intradiscal administration of anti-NGF antibody reduces nerve sprouting and suppresses the

expression of CGRP in the rat DRG, located next to the injured IVD [169].

Due to the deregulation of diverse signaling pathways in a degenerated IVD,

compounds with multiple beneficial effects are the most interesting option. Compounds

derived from traditional medicinal plants (phytochemicals) exhibit antioxidant, anti-

inflammatory and pro-survival activities in the treatment of various age-related diseases.

Phytochemicals act on several signal transduction and apoptotic cascades with minimal

adverse effects [22]. Naturals compounds such as resveratrol [170], curcumin [171], osthole [172],

ferulic acid [167], or triptolide [173] demonstrated beneficial activity in reducing inflammation,

catabolism and oxidative stress in IVD-related studies. Although further research is required to

fully understand the effects of phytochemicals on the integrity of the IVD, it has been

suggested that they can delay or revert the degeneration of the NP. Polyphenols can also

enhance the disc nutrition by improving vascular health [174], and relieve pain [170].

2.6.2 Cell-based regeneration

Transplantation of autologous NP cells, which is currently under evaluation in the

clinical trials, restores the proteoglycan content in the IVD [175]. However, poor expansion rates

and the loss of native phenotypic features of the NP cells during the expansion in monolayer

cultures are major drawbacks [176]. Autologous NP cells for transplantation can be obtained

from surgical specimens, removed during discectomy. Nevertheless, the numbers of obtained

cells are low, and the cells are likely to be affected by the degeneration [17].

Stem cells possess the capacity of self-renewal, maintaining their undifferentiated

phenotype in multiple subcultures. Upon application of appropriate stimuli, stem cells undergo

differentiation. Various types of adult stem cells were tested for IVD regeneration [176, 177] such

as bone marrow mesenchymal stromal/stem cells (MSCs) [178, 179], adipose tissue-derived

stem cells [180] and muscle-derived stem cells [181]. For example MSCs cultured under

39

chondrogenic conditions are capable of differentiating towards the NP cells [182]. Stem cells

can be injected into the disc as free undifferentiated cells or encapsulated in hydrogels. In

animal models of IVD degeneration, implantation of MSCs has resulted in restoration of the

IVD height, IVD-like phenotype expression, ECM synthesis, and improvement in MRI signals [17,

56, 177]. Although promising outcomes were found in pilot clinical applications (phase I/III

clinical trials [155]), no fully conclusive data exist to date [16, 183].

The major drawback of cell therapies is the poor survival rate of implanted cells due to

the low oxygen content, low glucose, acidic pH, and complex mechanical loading [16, 58]. In

addition, cells can undergo apoptosis as a result of limited nutrition caused by the calcification

of CEPs or age-related vascular changes in the vertebrae [1]. Implanted cells can also escape the

disc and induce the formation of osteophytes [184]. Nevertheless, the disc microenvironment

may be able to aid the differentiation into NP phenotype in some cases [185]. As the

microenvironment plays likely a crucial role in the clinical outcome of stem cell therapies,

careful patient selection is required [186]. In addition, future therapeutic strategies can use

genetically modified cells, with unfavorable polymorphisms being corrected, using an ex vivo

gene therapy strategy [107].

2.6.3 Gene therapy

A gene therapy is a technique used to deliver exogenous genetic material (RNA or DNA)

into cells in order to provide prolonged protein synthesis. Gene therapy using TGF-β, BMP-2,

SOX-2 or BMP-7 was tested in several in vitro and in vivo IVD studies [187, 188]. These showed

improvements in disc height and ECM synthesis and demonstrated the ability to modulate the

biological processes of IVD cells. However, severe drawbacks of in vivo gene delivery still exist.

Viral vectors can be immunogenic and non-viral vectors exhibit low transfection efficiency. In

addition, a misdirected injection may lead to the expression of transgenic growth factors

outside the IVD, leading to toxicity with detrimental consequences [189]. For this reason,

inducible gene expression systems were developed to reduce these problems [176]. Ex vivo gene

delivery has been tested as well. IVD cells were infected with adenoviral IL-1Ra in vitro and

injected into degenerative IVD tissue explants, which resulted in increased IL-1Ra production

[190]. However, the NP cells are not the ideal target for genetic alterations, due to their scarcity

in the adult IVD, inaccessibility, and slow proliferation rate with dedifferentiation in vitro. On

the other hand, stem cells (such as MSCs or induced pluripotent stem cells) modified by

targeting genome editing systems (such as CRISPR/Cas9) could provide better results in the

future. Gene silencing using RNA interference in the treatment of DDD is currently also being

investigated in vitro and ex vivo [191].

40

2.6.4 Bio-enhanced materials

Biocompatible materials mixed with biologics are commonly used for tissue

engineering (biomaterials enhanced with cells) and drug delivery (biomaterials enhanced with

drugs). Natural materials such as polysaccharides (e.g. chitosan) or proteins (e.g. collagen)

mimic the in vivo environment, as they bear similarities with the native ECM [150]. Widely used

synthetic biodegradable polymers such as poly(lactide) (PLA), poly(glycolide) (PGA), poly(ε-

caprolactone) (PCL) and their copolymers have tunable properties and low immunogenicity.

Biomaterials from both natural sources and synthetic sources were tested and are used in

clinical applications [192, 193].

Biomaterial-based scaffolds are ideal for AF tissue engineering. Scaffolds can be created

by electrospinning, which allows fiber alignment resembling the AF structure. Electrospinning

generates solid structures in a one-step process, without the need of employing high

temperatures or toxic solvents. The scaffolds generated by electrospinning can retain cells at a

desired location, provide optimal mechanical properties, and facilitate tissue ingrowth. The

fiber diameter and the stiffness of a scaffold influence the cell function and proliferation [194].

Growth factors and small molecules can be incorporated into electro-spun scaffolds to further

enhance their regenerative function [195].

Hydrogels are extremely useful biomaterials for the regeneration of the NP. They

consist of highly-hydrated networks of hydrophilic polymers, capable of absorbing a large

amount of water. Due to their similar composition, hydrogels are used to deliver drugs and cells

into the NP. Upon the injection of environmentally-responsive hydrogels into the NP, the

hydrogels polymerize in situ to form a depot for drug release or cell encapsulation [155].

Thermo-reversible hydrogels are one of the most promising in situ polymerizing hydrogels.

Poly(N-isopropylacrylamide) (pNIPAM) is a frequently investigated thermo-reversible

polymer[196], which undergoes sol-gel transition upon change in temperature via intra- and

inter-molecular hydrophobic interactions between its isopropyl groups (gel point ~32°C).

pNIPAM’s ability to solidify at human body temperature is extremely attractive for in vivo

applications (e.g. injections) [155]. A modified pNIPAM was recently used as a vehicle for siRNA

and celecoxib in experimental IVD degeneration studies [191, 197]. Hyaluronan-grafted pNIPAM

(HA-pNIPAM) has been used to deliver growth factors and cells into degenerated IVDs [198-

200]. Even though hydrogels have mechanical properties similar to the native NP tissue, they

are generally weak and their extrusion through an injection site can occur [155]. For this reason

simultaneous regeneration or repair of the AF is preferred.

It is well documented that the efficiency of pharmaceutical agents can be improved by

controlled delivery using biomaterial carriers. For example, PLGA microparticles, used to

encapsulate IL1-Ra or TFG-β, exhibit superior effects on the formation of ECM compared to the

injection of free agents [155]. Another example is a sustained local delivery of encapsulated

41

COX-2 inhibitor, which showed better results in the reduction of inflammation than its free

form [201]. Biologics-loaded particles can be created by modification of electrospinning called

electrospraying. During electrospraying, the solution jet breaks down into fine droplets which

acquire spherical shapes due to the surface tension, subsequently producing solid

microparticles upon solvent evaporation [54]. These biologics-loaded microparticles can be

administered into the disc by injection, which is less invasive than the total disc replacement

surgery.

2.6.5 Future perspectives

DDD is a multifactorial disorder accompanied by nerve compressions and irritations,

which strongly contributes to life-limiting low back pain. Traditional clinical treatment

strategies are mainly focused on the resolution of pain, rather than targeting the biological

causes of IVD degeneration. Nowadays, the main focus in research has shifted towards

regenerative therapies and tissue engineering of the IVD. The aim of these novel approaches is

to decelerate or reverse IVD degeneration and to restore the IVD for long-term prevention of

pain. Several molecular therapeutics and cell-based strategies have been tested in preclinical

and clinical settings and showed promising results. However, at present, there is no consensus

on the optimal method to resolve DDD using these strategies. It seems clear that the best

results will be achieved by the combination of multiple approaches in a patient-specific

manner. A suitable biomaterial will facilitate the molecular and cellular delivery into the disc,

improve the IVD microenvironment, release substances, reduce pain and enhance the

mechanical properties of the degenerative disc.

43

Chapter 3

Epigallocatechin 3-gallate suppresses

interleukin-1β-induced inflammatory responses

in intervertebral disc cells in vitro and reduces

radiculopathic pain in rats

Olga Krupkova, Miho Sekiguchi, Juergen Klasen, Oliver Hausmann,

Shinischi Konno, Stephen John Ferguson, Karin Wuertz-Kozak

European cells & Materials, 2014, 28: 372-86.

45

3.1 Abstract

Intervertebral disc (IVD) disease, which is characterized by are-related changes in the

adult disc, is the most common cause of disc failure and low back pain. The purpose of this

study was to analyze the potential of the biologically active polyphenol Epigallocatechin 3-

gallate (EGCG) for the treatment of painful IVD disease by identifying and explaining its anti-

inflammatory and anti-catabolic activity. Human IVD cells were isolated from patients

undergoing surgery due to degenerative disc disease (n=34) and cultured in 2D or 3D. An

inflammatory response was activated by IL-1β, EGCG was added, and the expression/activity of

inflammatory mediators and pathways was measured by qPCR (mRNA), Western blotting,

ELISA, Immunofluorescence and Transcription Factor assay. The small molecule inhibitor

SB203580 was used to investigate the involvement of the p38 pathway in the observed effects.

The analgesic properties of EGCG were analyzed by von Frey filament test in Sprague-Dawley

rats (n=60). EGCG significantly inhibited the expression of pro-inflammatory mediators and

matrix metalloproteinases in vitro, as well as radiculpathic pain in vivo, most probably via

modulation of the activity of IRAK-1 and its downstream effectors p38, JNK and NF-κB.

Key words

Epigallocatechin 3-gallate; intervertebral disc degeneration; low back pain; IRAK-1; p38; NF-κB;

inflammation, radiculopathy

46

3.2 Introduction

Intervertebral disc (IVD) degeneration is a normal part of the aging process that starts in

the late second or early third decade [202]. Although often asymptomatic, specific histological,

biochemical and functional changes have been observed during IVD degeneration. These

changes are consistent with pain generation and disc degeneration has been suggested as the

most common cause of low back pain in adults [202]. Costs associated with health services for

spinal problems have been estimated at 85.9 billion USD in 2005 in the United States [203, 204]

and many studies have shown that indirect costs related to reduced productivity and

subsequent impact on national economy are also significant [202].

Degeneration of the IVD occurs due to an imbalance between anabolic and catabolic

processes, leading to a loss in collagen and proteoglycan and a concomitant reduction in water

content by enhanced exposure to matrix metalloproteinases (MMPs) and a disintegrin and

metalloproteinases with thrombospondin motifs (ADAMTs) [3, 86, 205-207]. These

degenerative processes lead to mechanical dysfunction and altered stress distribution in the

tissue, hence increasing the risk for load-induced structural failure in the annulus fibrosus (AF),

so-called clefts, tears or fissures. Under these circumstances, pain sensation can be provoked by

leakage of nucleus pulposus (NP) material through the AF (i.e. disc herniation) and consequent

irritation of spinal nerves (radiculopathy) and/or nerve infiltration into the compromised disc

(nociception).

However, even without a disc prolapse, the degenerated IVD can be painful, especially if

high levels of pro-inflammatory mediators are secreted. Cytokines, e.g. from the interleukin-1

(IL-1) superfamily, not only irritate nerve endings in the AF [56], but also stimulate the

production of matrix-degrading enzymes, hence further worsening the degenerative processes

[82, 110, 208]. The healthy IVD is excluded from the development of immunologic tolerance as

an immune-privileged organ with no access to systemic circulation [209]. In contrast, diseased

human discs are heavily invaded by blood vessels and nociceptive nerve fibers [210], so that

anti-angiogenic and anti-neurogenic factors are of therapeutic interest in terms of nociceptive

pain [209]. Once NP extrudes from the IVD to the systemic circulation, the immune system

recognizes it as a foreign body, leading to autoimmune reactions conducted by activated B and

cytotoxic T lymphocytes, which may further enhance the destruction of NP tissue inside the

disc [211]. On the other hand, some immune cells (macrophages, mast cells) might be beneficial

to the regrowth of the IVD wounds and immune privilege maintenance at early stage of the

degeneration [212] as well as in the absorption of herniated NP [209, 213, 214]. When

conservative treatment (physical therapy, pharmacological treatment) fails, patients with disc

herniation or painful degenerative disc disease are likely to undergo surgical interventions,

discectomy in case of herniation, spinal fusion or disc replacement in case of disc disease.

47

Imaging techniques such as MRI are a valuable tool to identify the source of pain in disc

herniation patients, however in patients with degenerative disc disease, identification of the

painful disc is error-prone and removal of the degenerated tissue has a negative impact on disc

height or load-bearing capacity [194]. In case of spinal fusion, a consequent risk for adjacent

segment degeneration, with the need of additional surgeries, is also significant [14, 56, 194].

Currently used procedures for the treatment of degenerative disc disease are therefore

not optimal and new, less invasive but more targeted strategies are being developed [56]: NP

replacement, e.g. by the injection of biocompatible hydrogels with or without cells may have

the potential to restore normal disc height and load distribution, as well as limit degenerative

changes in adjacent discs [56]. Recently, tissue-engineered IVDs that formed a functional

motion segment were used to replace the degenerated discs in rodent caudal spine [215] and

chemical crosslinking was applied to stabilise the AF tissue in vitro [216]. Anti-inflammatory and

anti-catabolic substances that target the metabolism and inflammatory signaling within the

IVD also represent a new, interesting treatment option [170, 171, 173, 217].

Epigallocatechin 3-gallate (EGCG) is a biologically active polyphenolic catechin present

in green tea. EGCG forms 40-60 % of all green tea catechins and is reported to be responsible

for the major health benefits of green tea due to its anti-oxidant, anti-aging and anti-

inflammatory properties, which are exhibited by direct or indirect interaction with many

molecules and signaling transduction pathways in cells [218, 219]. EGCG was shown to have

beneficial effects for a number of clinical conditions including cancer, obesity, atherosclerosis,

diabetes, liver and neurodegenerative diseases [220-222], with cell-type specific effects and

modes of action [218, 221]. Although EGCG has chemopreventive and anticancer effects and acts

synergistically with several anticancer drugs [223], it has also been shown that EGCG can block

the function of boronic acid proteasome inhibitors [224] and reduce the bioavailability of the

novel oral multi-target tyrosine kinase inhibitor sunitinib [225]. Moreover, the consumption of

EGCG by pregnant women could increase a risk of innate acute myeloid leukemia in children,

because EGCG inhibits topoisomerase II activity not only in cancer cells, but also in the foetus.

This can lead to the translocation at chromosome 11q23 involving the mixed-lineage leukemia

(MLL) gene [226, 227]. Therefore, EGCG is possibly contraindicated in particular cases.

Beneficial effects of EGCG and green tea extract were described in osteoarthritic (OA)

chondrocytes [33], which bear similarities to IVD cells. It has recently been demonstrated that

EGCG inhibits the general IL-1β-induced response, but does not have significant anabolic effects

in chondrocytes. The anti-inflammatory effect of EGCG in human chondrocytes is mainly

mediated by inhibition of c-Jun N-terminal protein kinase (JNK) and nuclear factor kappa-light-

chain-enhancer of activated B cells (NF-κB) activity [34, 35]. EGCG was shown to inhibit NF-κB

activation by suppressing the degradation of its inhibitory protein in the cytoplasm [36]. EGCG

also reduces the activity of activator protein 1 (AP-1) transcription factor [34], p38 mitogen-

48

activated protein kinase (MAPK) [36] and blocks interleukin-1 receptor-associated kinase 1 (IRAK-

1) degradation [35] in OA chondrocytes. EGCG is able to inhibit the expression of cyclooxygenase

2 (COX-2) and prostaglandin E synthase (PGES) [37], which are both involved in the development

and transduction of inflammation and pain. EGCG also potently blocks the toll-like receptor 4

(TLR4) signaling pathway, which may play an important role in the occurrence and

development of neuropathic pain in rats [228]. Based on the promising findings of EGCG effects

in cartilage, we hypothesize that EGCG may also possess anti-inflammatory and anti-catabolic

effects in the IVD and thus have a potential for the treatment of pain and inflammation in

degenerative disc disease.

3.3 Methods

Human IVD cell culture preparation

The study was approved by ethic committees Kantonale Ethikkommission Zürich EK-

16/2005 and Ethikkommission des Kantons Luzern 1007IVD. Human NP tissue was removed

from patients undergoing spinal surgery for degenerative disc disease or disc herniation after

informed consent was granted. The details about the donors used in this study are listed in

Table 1. Tissue was enzymatically digested using a mixture of 0.2% collagenase NB4 (17454,

Serva) and 0.3% dispase II (04942078001, Roche) for 4–8 h at 37°C and isolated primary cells

were seeded in Dulbecco’s Modified Eagle’s Medium (DMEM/F12, D8437, Sigma) supplemented

with 10% fetal calf serum (F7524, Sigma), penicillin (50 units/ml), streptomycin (50 μg/ml) and

ampicillin (125 ng/ml, 15240-062, Gibco) and sub-cultured up to passage 3 using 1.5% trypsin

(15090-046, Gibco). Cells in passage 1-3 that were cultured either adherently (2D) or in alginate

beads (3D) were used for experiments.

Alginate beads preparation

Alginate was prepared as a solution of 1.2 % alginic acid sodium salt (180947, Sigma-

Aldrich) in 0.9% sodium chloride, stirred overnight and sterile filtered. Human IVD cells in

passage 1-3 were harvested with trypsin and gently mixed with alginate (4 x 106 cells per 1 ml of

alginate). Alginate-cell suspension was slowly pushed through a sterile syringe with a 21G

needle and dropped into a 102 mM calcium chloride solution with constant speed. Beads were

formed after 5 minutes of stirring in calcium chloride solution and then washed with 0.9%

NaCl and PBS. Beads were spread into 6-well plates with cell culture medium containing 50

µg/ml Vitamin C (A4403, Sigma) and cultured for 7 days with one media exchange.

49

Table 1. Demographic data of all donors used for the experiments. STEN = discogenic stenosis, DH = disc herniation, DDD = degenerative disc disease, SP = spondylolisthesis. E = ELISA, G = gene expression, W = Western blotting, V = Viability, TF = transcription factor assay, IC = immunocytochemistry, uk = unknown.

Donor Sex Age Level Pathology Grade Experiment

1 F 43 L4/5 DH IV E(3D), G(3D), W

2 F 62 L4/5 SP V E(3D), G(3D), W

3 F 49 L5/S1 protrusion IV E(3D), G(3D)

4 M 42 C6/7 DH III E(3D), G(3D)

5 M 47 L2/3 DH IV E(3D), G(3D)

6 F 67 L4/5 DH IV E(3D), G(3D)

7 uk uk uk uk uk E(2D,3D), G(2D,3D)

8 M 50 L5/S1 SP III E(3D), G(3D)

9 uk uk uk uk uk E(3D), G(3D)

10 M 41 L4/5 DH IV E(2D), G(2D), G(SB), W

11 M 52 L1/2 DH V E(2D), G(2D), TF

12 F 53 C6/7 DH IV E(2D), G(2D), IS

13 F 46 L5/S1 DH V E(2D), G(2D), IS

14 M 42 L4/5 DH IV G(2D),W

15 M 39 L4/5 DH IV G(2D,3D), G(SB)

16 F 54 L3/4 SP V G(2D), TF

17 F 38 L5/S1 DH IV G(2D), G(SB), TF

18 M 59 L4/5 DH IV G(3D)

19 uk uk uk uk uk G(3D), W

20 M 59 L3/4 DH V G(2D), W

21 M 57 L5/S1 DH V G(2D), W

22 uk uk uk uk uk G(2D), G(SB)

23 M 48 L5/S1 DH IV G(2D), G(SB), W

24 uk uk uk uk uk G(3D), W

25 M 54 L4/5 DH III G(2D),W

26 F 49 C4/5 DH II V, G(2D)

27 M 62 C5/6-C6/7 STEN III V, G(2D)

28 uk uk uk uk uk V, G(2D)


30 F 57 L3/4 SP III V, G(2D)

31 M 56 L5/S1 DH III V, G(2D)

32 M 63 L2/3 DH IV V, G(2D)


34 F 25 L5/S1 DDD III G(2D)

Viability measurement

Non-toxic concentrations of EGCG (E4243, Sigma) were defined using the MTT assay in

2D cell culture. Cells were seeded in 12-well plates (1 x 105 cells/well) and treated with different

concentrations of EGCG (0.1 - 50 μM) in medium with 10% FCS or without FCS. After 24 and 48

hours, fresh MTT solution (3-[4,5-Dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide,

M5655, Sigma) in media (0.5 mg/ml) was added and kept for 4 hours in 37°C. MTT was

50

discarded, cells were lysed in DMSO (D8418, Sigma) and absorbance was measured at 565 nm

relative to untreated controls. The non-toxic concentration 10 μM was chosen for subsequent

experiments. Data not shown.

Gene expression analysis

The effect of EGCG on the expression of inflammatory mediators and matrix degrading

enzymes was studied in 2D and 3D cell culture models that were pre-treated with IL-1β (211-11,

Peprotech) in order to induce the expression of pro-inflammatory and catabolic genes and

hence simulate the changes seen during inflammation-related disc degeneration [82, 229]. IVD

cells were seeded in 6-well plates (3 x 105 cells/well) (2D) or alginate beads (3D), serum-starved

for 2 hours and then exposed to 10 ng/ml IL-1β for 2 hours before treatment with 10 μM EGCG.

The involvement of the p38 pathway was studied using the small molecule inhibitor of p38

MAPK SB203580 (SB, S8307, Sigma), which was added 2 hours after IL-1β pre-stimulation in the

concentration of 10 µM to the cells seeded in 2D (n = 3 for MMPs and 4 for iNOS and COX-2).

After 18 hours, RNA was extracted with the TRIzol/chloroform method according to the

manufacturer instructions (15596-018, Invitrogen AG) and 1 µg was reverse transcribed to

complementary DNA (cDNA) using a reverse transcription kit (4374966, Applied Biosystems).

cDNA was then mixed with primers and master mix (4352042, Applied Biosystems) and gene

expression was measured using real-time PCR. Data was analysed by the comparative Cq

method (2ΔΔCq method, housekeeping gene Tata Box binding protein = TBP). Primer details are

given in Table 2. Results are presented as a gene expression relative to IL-1β pre-stimulation (100

%).

Table 2. Primers used for real-time RT-PCR (TaqMan Gene Expression Assays, Applied Biosystems).

Gene Primer Sequence Number Base Pairs TATA box binding protein (TBP) Hs00427620_m1 91 Interleukin-6 (IL-6) Hs00174131_m1 95 Interleukin-8 (IL-8) Hs00174103_m1 101 Matrix metalloproteinase-1 (MMP1) Hs00233958_m1 133 Matrix metalloproteinase-3 (MMP3) Hs00968308_m1 98 Matrixmetalloproteinase-13 (MMP13) Hs00233992_m1 91 Toll-like receptor 2 (TLR2) Hs00152932_m1 80 Cyclooxygenase-2 (COX-2) Hs00153133_m1 75 Inducible nitric oxide synthase (iNOS) Hs01075521_m1 82 Nerve growth factor (NGF) Hs00171458_m1 102

Western blotting

IVD cells seeded in 12-well plates (1 x 105 cells/well) were serum starved for 2 hours and

then exposed to 10 ng/ml IL-1β for 2 hours before treatment with 10 μM EGCG or 10 μM SB, 15

minutes for p-p38 (n = 9), p-JNK (n = 3) and IRAK-1 (n = 5), 6 hours for COX-2 (n = 1). For total

protein analysis, cells were lysed in 0.1% SDS buffer, mixed with Laemmli buffer (S3401, Sigma),

51

heated up (99°C, 5 minutes) and lysates were loaded onto 10% SDS-polyacrylamide gels.

Separated proteins were transferred to PVDF membranes (RPN303F, GE Healthcare) and

membranes were blocked in 5% non-fat milk in TBS-T for 1 hour/RT. Primary antibodies were

applied overnight at 4 °C. After washing in 1% non-fat milk in TBS-T (3×10 minutes), membranes

were incubated with secondary antibody conjugated to horseradish peroxidase (HRP) for 1

hour/RT. Visualization was performed on medical X-ray film (28906836, GE Healthcare) using a

chemiluminescence kit West Dura (34076, Thermo Scientific), films were scanned and scans

were processed by GIMP2. Tubulin was used as a loading control. Antibodies and dilutions used:

p38 MAPK, 1:1000 (9212, Cell Signaling); phospho-SAPK/JNK, 1:1000 (9251, Cell Signaling);

phospho-p38MAPK, 1:1000 (9211, Cell Signaling); α-Tubulin, 1:1000 (2144, Cell Signaling); IRAK-1,

1:1000 (4359, Cell Signaling); COX-2, 1:1000 (4842, Cell Signaling) and mouse anti-rabbit IgG

HRP, 1:5000 (211-032-171, Jackson Immuno Research).

Immunocytochemistry

IVD cells (n = 2) seeded onto cover slips in Petri dishes were serum-starved for 2 hours

and then exposed to 10 ng/ml IL-1β alone or in combination with 10 μM EGCG. After 1 hour of

simultaneous treatment, cells were washed with PBS (3×5 minutes), fixed with ice-cold

methanol (10 minutes on ice), washed with PBS again (3×5 minutes) and blocked with 2% BSA

(A4503, Sigma) with the addition of 5% goat serum (G9023, Sigma) (1 hour/RT). Primary

antibody p65 (RelA, the large subunit of NF-κB) was applied overnight at 4 °C. The next day,

cells were washed with PBS (3×5 minutes) and secondary antibody (CY2 goat anti-rabbit IgG)

was applied for 1 hour at RT. After washing in PBS (3×5 minutes), DAPI (D9542, Sigma) was

applied (5 minutes), cells were washed and embedded in Mowiol 4-88 (81381, Sigma). Analysis

of p65 translocation was performed with a fluorescent microscope (Olympus IX51) and

micrographs were processed in ImageJ. Cells without primary antibody were used as a non-

specific binding control, untreated cells were used as a negative control and cells treated with

10 ng/ml IL-1β were used as a positive control for p65 translocation. Antibodies and dilutions

used: NF-κB p65, 1:200 (Santa Cruz, sc-372) and CY2 goat anti-rabbit IgG, 1:400 (Jackson Immuno

Research, 111-165-144).

Nuclear extraction and Transcription factor assay

Nuclear extracts were prepared using a Nuclear Extraction kit (10009277, Cayman). Cells

(n = 3) were seeded in T75 flasks in the density 5 x 106 cells/flask and pre-cultured 24 hours. Cells

were serum-starved for 2 hours and then exposed to 10 ng/ml IL-1β alone or in combination

with 10 μM EGCG. After 1 hour of simultaneous treatment, nuclear extraction was performed

according to the producer’s protocol. Briefly, cells were collected into 15 ml pre-chilled tubes,

centrifuged 2 times with kit PBS/Phosphatase inhibitor solution and then re-suspended in ice-

52

cold kit hypotonic buffer. After 15 minutes of incubation, 10% Nonidet P-40 was added, cells

were centrifuged for 1 minute/140×g and supernatants containing the cytoplasmic fraction

were collected. Pellets were further re-suspended in kit complete nuclear extraction buffer by

vortexing (15 seconds), shaking (15 minutes on ice) and vortexing (15 seconds). Supernatants

containing the nuclear fraction were collected into new pre-chilled tubes after 10 minutes of

centrifugation (16000×g/4 °C). Protein concentration was determined using BCA assay kit

(23227, Thermo Scientific). NF-κB (p65) Transcription Factor Assay Kit (10007889, Cayman) was

used for the assessment of NF-κB (p65) DNA-binding activity in nuclear extracts according to

the producer’s protocol on a 96-well plate. An equal amount of protein (15 μg/well) was loaded

in duplicates for each sample. Provided non-specific binding control, competitor dsDNA control

and positive control were used to monitor appropriate assay function. The 96-well plate was

incubated 2 hours at RT, washed with kit wash buffer and primary antibody was added. After

one hour, the plate was washed, incubated with secondary antibody, washed again and

incubated with kit developing solution for 30 minutes. Finally, kit stop solution was applied and

absorbance at 450 nm was measured immediately. The result is presented as NF-κB (p65) DNA-

binding activity relative to IL-1β pre-stimulation (100 %).

Enzyme-linked immunosorbent assay (ELISA)

To detect secreted proteins, IVD cells in passage 1-3 were seeded on 6-well plates (3 x 105

cells/well) (n = 5) or alginate beads (n = 9), serum starved for 2 hours and then exposed to 10

ng/ml IL-1β for 2 hours before treatment with 10 μM EGCG. Cell culture media was collected

after 18 hours and the level of IL-6 protein expression was analysed by ELISA according to the

producer’s protocol (Human IL-6 ELISA set, 555220 with Reagent set B, 550534, BD Biosciences).

Briefly, 96-well plates were coated by kit capture antibody at 4 °C overnight, washed and

blocked in kit assay diluent. After washing, samples with controls and standards were

incubated on the plates for 2 hours at RT. Then plates were washed, kit detection antibody,

streptavidin-HRP and substrate solution were applied according to the producer’s protocol.

Absorbance was measured within 30 minutes at 450 nm with 570 nm correction after the

addition of a kit stop solution. Results are presented as protein expression relative to IL-1β pre-

stimulation (100 %).

In vivo study on pain behaviour

All animal experiments were carried out under the control of the Animal Care and Use

Committee in accordance with local guidelines for animal experiments and government law

concerning the protection and control of animals. Female Sprague-Dawley rats (n = 60, 200–

250 g) (Japan SLC, Shizuoka, Japan) were used in this study. A combination of 0.3 ml

medetomidine hydrochloride (1.0 mg/ml), 0.8 ml midazolam (5.0 mg/ml), and 1.0 ml

53

butophanol tartrate (5.0 mg/ml) was prepared as an anesthetic. Animals were anesthetized by

intraperitoneal injection of 0.1 ml/100g of BW of the above described mixed anesthetic.

Animals were placed in a prone position and the surgical intervention was performed with a

stereo operating microscope and microsurgical instruments. An incision was made to the

spinal midline, then fascia and multifidus muscle were resected, and the left L5 nerve root and

dorsal root ganglion (DRG) were exposed to L5-L6 facetectomy on the left side, with great care

taken to avoid trauma to the tissue. In the NP application group (NP group, n=48), autologous

NP was harvested from the tail and applied to the DRG [230-232]. In the sham-operated group

(Sham group, n = 12), the left L5 nerve root and DRG were exposed to L5-L6 facetectomy, but no

other procedures were performed. Animals from the NP application group were divided into

four groups (n=12 in each): NP + EGCG 10 µM (0.1 ml of 10 µM EGCG), NP + EGCG 100 µM (0.1 ml

of 100 µM EGCG), NP + water (0.1 ml of water as a vehicle), and NP (non-treatment) group.

Animals in the treatment groups were injected with 0.1 ml of the designed treatment solution

into the underlayer of epineurium, just distal to the NP, before closing the incisions [170, 232].

Sensitivity to non-noxious mechanical stimuli was tested using the von Frey Filament test.

Baseline testing was performed before the start of the experiment to accommodate animals

with normal responses. Hind paw withdrawal response to von Frey hair (North Coast Medical,

Morgan Hill, CA, USA) stimulation of the plantar surface of the footpads was determined at day

2, 7, 14, 21, and 28 after surgery. Individual rats were placed in an acrylic cage with a mesh floor

and allowed to acclimate for 15 minutes or until cage exploration and major grooming activities

ceased. The lateral plantar surface of the operated hind paw, innervated by the L5 nerve, was

stimulated with nine von Frey filaments (1.02, 1.4, 2.0, 4.1, 6.1, 8.0, 10.6, 15.4, and 26.0 g) threaded

under the mesh floor. The gram ratings for von Frey hairs were based on the ratings supplied by

the manufacturer. The filaments were sequentially applied to the paw surface just until the

filament bent and was held for 3 seconds. The response was considered positive if the rat lifted

the foot in combination with either licking or shaking of the foot as an escape response [170,

232].

Statistical Analysis

Statistical significance between IL-1β and IL-1β+EGCG groups (Figure 1, 2) as well as

between IL-1β and IL-1β+SB groups (Figure 3) was analyzed using Student’s T-test. Asterisks

represent a significance level P ˂ 0.05. For the animal study, behavioural data between the

groups over the experimental period (28 days) was statistically evaluated by ANOVA with

Bonferroni post-hoc testing to correct for multiple comparisons. Asterisks represent a

significance level P ˂ 0.01.

54

3.4 Results

EGCG causes a significant decrease in the expression of inflammatory mediators and matrix

metalloproteinases in IVD cells cultured in vitro

The effect of EGCG on the expression of inflammatory mediators and matrix

metalloproteinases was determined in adherent 2D cell culture and in 3D alginate beads on the

mRNA level after 18 hours. A significant inhibition of mRNA expression of interleukins (IL-6, IL-

8), matrix metalloproteinases (MMP1, MMP3, MMP13), toll-like receptor 2 (TLR2), nerve growth

factor (NGF), inducible nitric oxide synthase (iNOS) and cyclooxygenase 2 (COX-2) was detected

upon EGCG treatment in both, 2D and 3D (Figure 1a, 1b). The exact values for each gene,

including p-values, are listed in Table 3. To confirm these observations on the protein level, IL-6

was chosen to be measured in cell culture media by ELISA, due to its strong regulation on the

mRNA level and pathological relevance. EGCG treatment significantly decreased IL-6 secretion

to the media, compared to the IL-1β pre-stimulated cells both in 2D (31.26 %, p = 0.0007) and 3D

(64.45 %, p = 0.0012) (Fig. 1c). Displaying the data relative to the IL-1β pre-stimulation (set as

100%) minimizes the inter-donor variation and shows clearly the effect of studied compound

on the acute inflammatory response.

Figure 1. EGCG exhibits anti-inflammatory and anti-catabolic effects in 2D and 3D IVD cell culture. Cells were pre-stimulated with IL-1β for 2 hours before treatment with EGCG and compared to cells treated with IL-1β and EGCG separately. After 18 hours, EGCG significantly reduced mRNA expression of IL-6, IL-8, MMP1, MMP3, MMP13, TLR2, NGF, iNOS and COX-2, both in 2D adherent cell culture (a) and in 3D alginate beads (b). IL-6 protein secretion in cell culture media was significantly decreased as well (c). Data is

55

presented as mRNA or protein expression relative to IL-1β pre-stimulation. Asterisks indicate statistical significance (p ˂ 0.05, Student´s T-test). Table 3. Percentage of mRNA expression in samples treated with IL-1β + EGCG, relative to IL-1β pre-stimulation, showed in Figure 1.

Gene 2D 3D Expression, P-Value Expression, P-Value

IL-6 22.72 % P ˂ 0.0001 32.79 % P = 0.0002 IL-8 4.99 % P ˂ 0.0001 25.69 % P = 0.0002 MMP1 30.40 % P ˂ 0.0001 41.40 % P = 0.0013 MMP3 19.73 % P ˂ 0.0001 32.55 % P = 0.0020 MMP13 28.93 % P ˂ 0.0001 54.08 % P = 0.0096 TLR2 10.43 % P ˂ 0.0001 3.25 % P = 0.0001 NGF 10.67 % P = 0.0002 46.79 % P = 0.0003 iNOS 5.91 % P ˂ 0.0001 45.44 % P = 0.0096 COX-2 20.68 % P ˂ 0.0001 39.62 % P = 0.0067

EGCG inhibits IL-1β-induced IRAK-1 degradation and partially blocks NF-κB, p38 and JNK activity

IRAK-1 was reported to be important for the transduction of inflammatory signals (IL-1β,

LPS), which cause its phosphorylation and degradation, hence enabling the activation of stress-

related signaling molecules NF-κB, p38 and JNK [233]. IL-1β pre-stimulated IVD cells were treated

with EGCG and the level of IRAK-1, together with the activity of its downstream effectors p38

and JNK, were determined by Western blotting. IRAK-1 was degraded in the cells stimulated

with IL-1β, but the level of IRAK-1 degradation was reduced in cells treated with IL-1β + EGCG

(Figure 2a). Phosphorylation of p38 and JNK was also decreased in IL-1β + EGCG-treated cells,

compared to IL-1β stimulated cells, i.e. EGCG reduced p38 and JNK activity (Figure 2b). The effect

of EGCG on the activity of the transcription factor NF-κB was studied by immunofluorescence,

visualising the translocation of NF-κB subunit p65 (RelA) to the nucleus and by NF-κB (p65)

Transcription Factor Assay. The effect of EGCG + IL-1β studied by immunocytochemistry

prevented p65 translocation in approximately 50% of cells. Cells without primary antibody,

untreated cells and cells treated with IL-1β were used as controls. As nuclear NF- κB was still

detectable in the EGCG + IL-1β-treated cells (Figure 2c), a Transcription Factor Assay was

performed to quantify the nuclear localization of NF-κB in EGCG + IL-1β-treated cells via DNA

binding activity. Results of the Transcription Factor Assay showed 50.83 % (p = 0.0327) decrease

in NF-κB DNA-binding activity upon EGCG treatment (Figure 2d). Results are presented as NF-κB

DNA binding activity relative to IL-1β stimulated cells (100%). Suggested mechanism of EGCG

action in IVD cells is shown in Figure 5.

56

Figure 2. EGCG inhibits IRAK-1 - NF-κB/p38 /JNK signaling in IVD cells. Cells were pre-stimulated with IL-1β for 2 hours before EGCG was added. After 15 minutes, cells were lysed and Western blotting was performed. IRAK-1 degradation (a), p38 and JNK phosphorylation (b) were inhibited upon EGCG treatment of IL-1β stimulated cells. Combination of IL-1β with EGCG partially blocks NF-κB (p65) translocation to the nucleus (c) and NF-κB (p65) DNA-binding activity (d) after 1 hour. Asterisk indicates statistical significance (p ˂ 0.05, Student´s T-test).

EGCG inhibits p38 MAPK-dependent expression of COX-2, iNOS, MMP1 and MMP13

As shown in Fig 2, EGCG modulates the IRAK-1 - p38 signaling pathway in IVD cells. In

order to reveal which inflammation-related molecules are regulated by p38 MAPK on the gene

expression level, a small molecule inhibitor of p38 MAPK SB203580 (SB) was used. IL-1β pre-

stimulated cells were treated with either EGCG or SB and mRNA expression was measured after

18 hours. Both, EGCG and SB treatment significantly inhibits COX-2 (SB: 26.79 %, p = 0.0151),

iNOS (SB: 19.39 %, p = 0.0042), MMP1 (SB: 44.54 %, p = 0.0494) and MMP13 (SB: 6.66 %, p =

0.0042) expression compared to IL-1β pre-stimulated cells (100 %) (p ˂ 0.05) (Figure 3a). The

ability of SB to inhibit p38 activity was confirmed by Western blotting (Figure 3b). The

regulation of COX-2 expression by EGCG and SB was proved also on protein level by Western

blotting (Figure 3c). The expression of other studied genes (IL-6, IL-8, MMP3, TLR2, NGF) is only

partially p38 dependent or p38 independent (data not shown).

57

Figure 3. EGCG supresses p38 MAPK-dependent MMP1, MMP13, iNOS and COX-2 expression in IL-1β pre-stimulated IVD cells. Cells were pre-stimulated with IL-1β for 2 hours before treatments with EGCG and SB203580 (SB). Control cells were either untreated or treated with IL-1β, EGCG and SB separately. After 18 hours, EGCG and SB203580 both reduced mRNA expression of MMP1, MMP13, iNOS and COX-2. Data are presented as mRNA expression relative to IL-1β pre-stimulation. Asterisks indicate statistical significance (p ˂ 0.05, Student´s T-test) (a). The function of p38 MAPK inhibitor SB203580 was confirmed by Western blotting (b). The addition of EGCG or SB to IL-1β-stimulated cells inhibits p38-dependent COX-2 protein expression (c).

EGCG inhibits pain behaviour in vivo

Animal behaviour in response to mechanical stimulation by von Frey filaments was

compared between the NP + EGCG treatment groups (NP + EGCG 10 µM or NP + EGCG 100 µM),

the NP group (NP + no treatment), the water group (NP + water) and the sham group (only

facetectomy). Furthermore, the NP + EGCG treatment groups were compared amongst each

other. In the NP groups without EGCG treatment, the mechanical withdrawal thresholds were

significantly decreased for 28 days compared with the sham group, indicating that pain was

evoked by application of NP tissue to the DRG. Over the experimental period, mechanical

withdrawal thresholds upon EGCG treatment were significantly higher than in the NP group

and NP + water group and reached levels measured in the sham group. The mean hind paw

withdrawal thresholds for each group at each time point are listed in Table 4. The exact P-

Values for each comparison are listed in Table 5. Results indicate that EGCG treatment reduced

pain perception and this effect was independent of the applied concentration of EGCG (NP +

EGCG 10 µM vs NP + EGCG 100 µM = n.s.). Decrease in the mechanical withdrawal threshold

observed in the sham group arises from facetectomy-related distress (Figure 4).

58

Table 4. The mean hind paw withdrawal thresholds for each group and time point.

Baseline day 2 day 7 day 14 day 21 sham 26±0 23.5±8.0 22.8±7.2 15.6±9.5 16.6±9.2 NP 26±0 10.2±6.3 9.5±6.3 5.7±4.7 9.1±6.2 NP + water 26±0 10.9±5.9 7±1.8 7.8±3.9 9.9±6.2 NP + EGCG 10 µM 26±0 15±7.3 11.5±7.5 15.4±8.5 17.8±8.9 NP + EGCG 100 µM 26±0 16.9±8.7 11.9±7.0 17.5±10.6 19.6±8.4 Table 5. Statistical analysis of von Frey filament test between the groups. ns = non-significant. sham NP NP

+ water NP + EGCG 10

NP + EGCG 100

sham x P = 0.0001 P < 0.0001 ns ns NP P = 0.0001 x ns P = 0.0098 P = 0.0004 NP + water P < 0.0001 ns x P = 0.0026 P < 0.0001 NP + EGCG 10 µM ns P = 0.0098 P = 0.0026 x ns NP + EGCG 100 µM ns P = 0.0004 P < 0.0001 ns x

Figure 4. EGCG inhibits pain behaviour in vivo. Pain sensitivity was measured by von Frey filament test for 28 days. When NP tissue was placed on the DRG without therapeutic treatment (NP group and NP + water group), the mechanical withdrawal thresholds were significantly decreased. EGCG treatment (NP + EGCG 10 µM, NP + EGCG 100 µM) prevented the decrease in the mechanical withdrawal threshold observed in the NP group and NP + water group and restored the threshold level observed in the sham group.

59

Figure 5. Suggested mechanism of EGCG action in IVD cells. Upon IL-1β stimulation, IRAK-1 is phosphorylated and degraded, which enables the formation of multiprotein complex (TRAF6/TAB1/TAKs) and activation of key stress-related effectors NF-κB, p38 and JNK, which are further involved in the activation of inflammation, matrix breakdown and pain (A). EGCG blocks IRAK-1 degradation, formation of multiprotein complex and hence subsequent activation of NF-κB, p38 and JNK (B). TRAF6: TNF receptor-associated factor 6, TAK: TGF-beta activated kinase 1, TABs - TGF-beta activated kinase 1 binding proteins (schematic).

3.5 Discussion

The herein presented study clearly demonstrates the anti-inflammatory and anti-

catabolic properties of EGCG in human IVD cells in vitro, hence highlights its potential for the

treatment of IVD-related back pain. The development of back pain involves both nociceptive

and neuropathic pathophysiological mechanisms [12, 234] with IVD degeneration being one of

the major underlying causes. IVD degeneration is induced by phenotypic changes in the NP

cells, leading to tissue degradation [86], as well as to promotion of neoinnervation of the disc

and hence the transmission of pain in vivo [235, 236]. The evidence that IVD degeneration and

low back pain are correlated with increased levels of pro-inflammatory cytokines is increasing

[110]. Inflammatory processes are regulated, for example, by toll-like receptors (TLRs), which are

primarily expressed on the immune cells as a first line of host defence, but have also been

detected in the other cell types, including IVD [237, 238]. Recently, our group confirmed basal

expression of TLRs in IVD cells and showed that expression of TLR2, 4 and 6 is increased during

60

IVD degeneration [239] thus further underlying the clinical relevance of inhibiting IVD

inflammation. In this study, we demonstrated that 10 μM EGCG can significantly inhibit the

mRNA expression of inflammatory and pain-related mediators IL-6, IL-8, TLR2, NGF, iNOS, COX-2

and matrix metalloproteinases MMP1, MMP3, MMP13 in 2D and 3D IL-1β- stimulated IVD cell

culture models. All of these markers have been described to be of relevance during IVD

degeneration and inflammation and hence pain development [240, 241]. Although all cells used

in this study were isolated from the degenerated discs, the difference in the expression of

studied inflammation-related genes between untreated control group and EGCG-treated group

is not significant, because basal expression of these genes decreases when cells are transferred

to in vitro conditions (data not shown).

While this is the first IVD-related study, protective effects of EGCG and green tea extract

have been reported in inflammatory arthritis [33], where EGCG was shown to block IL-1β-

induced expression of IL-6, IL-7, IL-8, IL-1β, TNF-α [35], MMP1 and MMP13 [34], iNOS [242], COX-2

[243] and these effects were linked mostly to NF-κB and JNK signalling pathways [35-37, 242,

244]. In order to determine the underlying mechanisms in the IVD, potential pathways were

identified from the literature [35-37, 242, 244] and the effects of EGCG on IRAK-1 degradation as

well as NF-kB and MAPK activity in IVD cells were analyzed. We propose that EGCG acts as an IL-

1 receptor antagonist in IL-1β stimulated IVD cells. Our suggested mechanism of action, which is

illustrated in Figure 5, is through the inhibition of IRAK-1 degradation. As IL-1 receptors do not

possess intrinsic kinase activity, they rely on the recruitment and activation of intrinsic kinases

(IRAKs) [233]. Upon phosphorylation, IRAK-1 is ubiquitinated and degraded, which initiates a

formation of cytosolic multi-protein complex. The IRAK-1-activated protein complex is involved

in the phosphorylation of the inhibitor of κB kinase (IKK) as well as of MAPK kinases 3/4/6,

leading to the activation of NF-κB and JNK/p38 MAPKs, respectively [233, 245]. In this study, we

demonstrate inhibition of IRAK-1 degradation in IVD cells, similar to IL-1β-stimulated OA

chondrocytes [35]. As a result, activity of NF-κB was decreased in IL-1β-stimulated EGCG-treated

IVD cells. This effect was observed after 1 hour, since NF-κB is a protein complex that functions

as a “rapid-acting” transcription factor and as such regulates the transcription of more than 150

genes related to stress responses, including MMPs [110, 229, 246, 247].

The inhibition of IRAK-1 degradation is accompanied by a decrease in the activity of its

downstream effectors p38 and JNK. p38 kinase is a MAPK family member that is activated in

various stress conditions like UV irradiation, hypoxia, oxidative stress and inflammation and

regulates the expression of many stress-related genes, leading either to restoration of

homeostasis or to alteration of cellular functions and apoptosis. Recently it was published that

blockage of p38 MAPK activity in rabbit [248] and human [156] NP cells reduces IL-1β-induced

NO and prostaglandin E-2 accumulation [248, 249] and partially restores proteoglycan

synthesis [248]. Using the small-molecule inhibitor SB203580, we confirmed that not only iNOS

61

and COX-2, but also MMP1 and MMP13 are p38 MAPK target genes in human IVD cells (similar

to bovine IVD cells [250]), hence indicating that the p38 pathway may play an essential role in

ECM degradation and nociception in the disc. Importantly, the concomitant regulation of MMP1

and MMP13 suggests that EGCG may slow down MMP-induced collagen loss in all stages of IVD

degeneration as MMP1 has been shown to be significantly increased in severely degenerated

NPs, whereas MMP13 is increased in earlier stages of NP degeneration [69, 86]. The expression

of other studied genes (IL-6, IL-8, MMP3, TLR2, NGF) is regulated differently in IVD cells,

although p38 may co-operate too. In the context of nociception, the observed decrease in the

prostaglandin H synthase COX-2 can be specifically relevant as it catalyses the initial step in the

conversion of arachidonic acid to prostaglandins, which have a variety of physiological effects

including sensitisation of spinal neurons to pain [248]. Similarly, NOS plays an important role in

the processing of pain [251] so that the reduction of iNOS expression by EGCG should be

highlighted. Interestingly, it was recently reported that intrathecal administration of EGCG

could produce an antiallodynic effect against spinal nerve ligation-induced neuropathic pain,

mediated by blockade of neuronal NOS protein expression and inhibition of nitric oxide (NO)

[252]. As described above, our in vitro results suggest that EGCG-driven p38-mediated inhibition

of COX-2 and iNOS expression may be helpful for nociceptive as well as neuropathic pain

reduction. In order to monitor changes in pain behaviour upon EGCG treatment in vivo, we used

an animal model in which NP-mediated pain is simulated by the application of NP tissue to the

DRG, thus representing typical radiculopathic pain [253-255]. EGCG treatment was able to

prevent the threshold reduction and thus the pain-related behavior. The full recovery to the

sham threshold level was observed on day 14 in both EGCG treatment groups. Our results

correspond to recent findings using the rat model of neuropathic pain (sciatic nerve

constriction) where intrathecal injection of EGCG markedly improved pain behaviour in rats

and decreased the expression of TLR4, NF-κB, high-mobility group protein B1 (HMGB1), tumor

necrosis factor alpha (TNF-α) and IL-1β in the spinal cord [228]. In humans, back pain is however

caused by complex interactions of biological, psychological and social factors and can be

considered as a syndrome with both nociceptive and neuropathic components (mixed pain)

[256]. Nociceptive nerves are a major transmitter of discogenic back pain [257], therefore

alteration of the nerve ingrowth into the disc by EGCG-mediated inhibition of neurotrophins

can modulate the development of acute and chronic nociceptive pain. On the other hand,

discogenic radiculopathy occurs when functional, vascular and morphological changes of the

nerve root are activated, e.g. by disc-related inflammation, which leads to the intraradicular

fibrosis and nerve fiber atrophy [257]. From this point of view and according to our in vivo

results, EGCG-mediated inhibition of the inflammatory response of NP tissue may be beneficial

also against the development of disc-related radiculopathy. Based on these results, we

hypothesize that EGCG may reduce pain behaviour in vivo via inhibiting the expression of

62

cytokines and pain-related inflammatory mediators, similar to the mechanism observed in our

in vitro cell culture study.

This study however has certain limitations; firstly, the mechanism of EGCG action is not

entirely clear. During disc degeneration, the regulation and function of IL-1-mediated pathway

is altered: IL-1α and IL-1β isoforms as well as IL-1 receptor are synthesized in bigger quantities

compared to healthy tissue, whereas the expression of IL-1 receptor antagonist is unchanged

[84]. This imbalance activates the IL-1 pathway in NP tissue, which leads to the induction of the

expression of MMPs, ADAMTs, as well as to the enhanced angiogenesis and neurogenesis [257].

Although our study does not address the exact mechanism of how EGCG influences the IL-

mediated pathway, it provides evidence that EGCG can act as an IL-1 pathway antagonist,

therefore can possibly help to restore the tissue homeostasis in vivo. The second limitation is

that NP cells were isolated from donors with different medical conditions due to the limited

number of available biopsies. In addition, exact separation of NP and AF tissue (which was

performed in the surgical as well as the laboratory setting) can be challenging in degenerated

material so that certain impurities are possible. Nevertheless, all donors respond very similar.

Finally, the effect of EGCG on the tissue level (matrix degradation) was not analyzed in this

study.

Considering the anti-inflammatory, anti-catabolic and analgesic properties of EGCG

presented in this study, we suggest that EGCG could in the future supplement or even

substitute drugs which are used currently for the treatment of low back pain and disc

degeneration. However, the best mode of application is still a matter of debate. While oral

application has been considered for diseases such as obesity or diabetes, large in vitro/in vivo

discrepancy was observed [258] due to the low bioavailability of EGCG, with an elimination

half-life in plasma of around 3.5 hours [259]. For the treatment of back pain, oral application of

EGCG is certainly not a promising approach because the inflammatory processes within the IVD

need to be targeted directly. Therefore, local application (e.g. intradiscal or epidural injection)

will be required, which could be done in conjunction with other surgical or regenerative

interventions or independently, i.e. as an early, non-invasive treatment. However, it is likely that

the effects of non-modified (free) EGCG will be less pronounced in vivo compared to in vitro

because of its low stability, which can be further influenced by extrinsic conditions such as type

of storage and intrinsic factors like Ca2+ and Mg2+ ions or antioxidant concentration [258].

Encapsulation of EGCG in polymeric carriers represents a promising strategy to increase its

bioavailability and release period. In fact, different types of polymeric carriers administered

orally or intravenously were demonstrated to significantly increase the stability and efficiency

of EGCG in cancer therapy [260, 261], showing their potential to be used also inside the IVD.

However, the disc-specific environment (high amount of proteoglycans, lower pH in

degenerated disc, low amount of cells) should be taken in account when choosing a carrier and

63

delivery method for EGCG. The efficacy of these respective EGCG slow-release systems will have

to be determined in vitro using cell or organ culture studies (by detecting inflammation) as well

as in appropriate in vivo models (by detecting pain sensation). Despite the remaining

challenges, EGCG offers several potential clinical advantages: it is globally available, it is

inexpensive to isolate, it was reported to be safe, well tolerated and clinically active

concentrations can be reached by its oral administration or by its modification [218, 219].

3.6 Conclusion

In this study, we examined the effects of the bioactive polyphenol from green tea,

Epigallocatechin 3-gallate, on IVD cells cultured in 2D and 3D in vitro and on IVD-related

radiculopathic pain in vivo. We have described its anti-inflammatory and anti-catabolic effects

and the benefit of EGCG in the reduction of both nociceptive and neuropathic disc-related pain.

We identified IRAK-1 - NF-κB/p38/JNK signaling pathways to be modulated by EGCG and we

provide evidence that this contributes to the observed effects. Although precise mechanisms of

EGCG action and its direct targets in IVD cells still need to be elucidated, we showed promising

therapeutic potential of EGCG in the treatment of disc-related inflammation and back pain in

degenerative disc disease.

3.7 Author’s contributions

OK performed cell culture experiments, statistical analysis and helped to draft the

manuscript. MS performed animal experiments, statistical analysis and helped to draft the

manuscript. JK participated in study design, provided clinical samples and medical scientific

input and helped to draft the manuscript. OH participated in study design, provided clinical

samples and medical scientific input and helped to draft the manuscript. SK helped with study

design and coordination and helped to draft the manuscript. SJF conceived funding of the

study, helped with study design and coordination and helped to draft the manuscript. KW

designed and coordinated the study and helped to draft the manuscript. All authors approved a

final version of manuscript. Authors have no competing interests.

3.8 Acknowledgement

Funding for this research project was provided by the European Union through a Marie

Curie action (FPT7-PITN-GA-2009-238690-SPINEFX) and Theodor und Ida Herzog-Egli

Foundation. The authors would like to thank Ms. Greutert for technical assistance.

65

Chapter 4

The natural polyphenol epigallocatechin gallate

protects intervertebral disc cells from oxidative

stress

Olga Krupkova*, Junichi Handa*, Marian Hlavna, Juergen Klasen, Caroline

Ospelt, Stephen John Ferguson, Karin Wuertz-Kozak

* these authors contributed equally

Oxidative medicine and cellular longevity, 2016, Jan(9): 1-17.

67

4.1 Abstract

Oxidative stress-related phenotypic changes and a decline in the number of viable cells

are crucial contributors to intervertebral disc degeneration. The polyphenol epigallocatechin 3-

gallate (EGCG) can interfere with painful disc degeneration by reducing inflammation,

catabolism and pain. In this study we hypothesized that EGCG furthermore protects against

senescence and/or cell death, induced by oxidative stress.

Sub-lethal and lethal oxidative stress was induced in primary human intervertebral disc

cells with H2O2 (total n = 36). Under sub-lethal conditions, the effects of EGCG on p53-p21

activation, proliferative capacity and accumulation of senescence-associated β-galactosidase

were tested. Further, the effects of EGCG on mitochondria depolarization and cell viability were

analyzed in lethal oxidative stress. The inhibitor LY249002 was applied to investigate the

PI3K/Akt pathway.

EGCG inhibited accumulation of senescence-associated β-galactosidase, but did not

affect the loss of proliferative capacity, suggesting that EGCG did not fully neutralize

exogenous radicals. Furthermore, EGCG increased the survival of IVD cells in lethal oxidative

stress via activation of pro-survival PI3K/Akt and protection of mitochondria.

We demonstrated that EGCG not only inhibits inflammation, but also can enhance the

survival of disc cells in oxidative stress, which makes it a suitable candidate for the

development of novel therapies targeting disc degeneration.

Key words

EGCG, epigallocatechin 3-gallate, oxidative stress, intervertebral disc degeneration, senescence,

apoptosis, H2O2, PI3K/Akt pathway

68

4.2 Introduction

Degenerative disc disease, characterized by spinal micro-instability and lower back pain,

is a result of multiple events like age-related degradation of extracellular matrix (ECM),

inflammation or trauma [262, 263]. During degeneration, the intervertebral disc (IVD)

undergoes morphological and functional changes such as calcification of endplates, disruption

of annulus fibrosus (AF) and loss of water in the nucleus pulposus (NP), which not only impairs

disc function, but also decreases nutrient supply and causes accumulation of cellular waste

products [99, 264, 265]. In such situation cells can compensate for impaired homeostasis e.g. by

activation of neovascularization factors, to enhance the availability of nutrients. However, a

catabolic environment and abnormal blood supply in the originally avascular tissue can

increase reactive oxygen species (ROS) production and oxidative stress [94, 266-269].

ROS are highly reactive molecules with unpaired electrons formed externally or in

mitochondria as a normal part of the aerobic metabolism. To ensure a properly maintained

redox balance, organisms require a complex, coordinated network of antioxidants from various

sources that control ROS generation [270]. Cellular responses to ROS depend on the

concentration and duration of ROS exposure as well as the cell type. As signaling molecules,

ROS modulate a variety of physiological events, including proliferation, differentiation, host

defense and wound healing [271]. On the other hand, extensive ROS exposure and/or

insufficient cellular anti-oxidant capacity cause deleterious damage, senescence and cell death

[38].

Cellular senescence is described as an irreversible stress-induced cell cycle arrest [272].

Senescence followed by tissue remodeling is beneficial in healthy tissue, where it contributes to

the elimination of damaged cells. Furthermore, it is a major protective mechanism against

carcinogenesis. However, at the same time, persistent senescence impairs tissue function and

leads to inflammation and premature degeneration [90]. Internal DNA damage (telomere

shortening) as well as multiple stress signals, including ROS, aberrant oncogene activation and

chemotherapeutic drugs, are able to induce premature senescence via activation of the p53-p21

and/or p16 Ink4a [273]. Although cell and tissue aging have been extensively investigated, the

precise mechanisms underlying the development of a senescence phenotype are still unknown

[272].

Extensive oxidative stress disrupts the integrity of mitochondria, which can further

elicit cell death through three major mechanisms: During apoptosis (1), cytochrome c is

released into the cytosol where it activates caspase 9, which allows executioner caspases 3, 6

and 7 to cleave their substrates [274]. Caspase-independent cell death mechanisms, such as

regulated necrosis (2) or autophagy (3), are equally important in various pathologies, including

IVD degeneration [96, 275]. Similarly to senescence, the biological purpose of apoptosis is to

69

eliminate damaged or sub-optimal cells [90], although both processes are regulated differently

[272]. During apoptosis, a variety of triggers can quickly converge to the executioner effectors

through a common mechanism. In contrast, senescence is typically a delayed progressive

process, during which various effector mechanisms can each contribute and collectively define

the phenotype [272].

The functionality of human IVDs as spinal shock-absorbers gradually decreases

throughout lifetime. Senescence of IVD cells, which not only positively correlates with age, but

also with the degree of disc degeneration, can be linked to oxidative stress [3, 40-47]. Oxidative

stress-induced senescence and death of IVD cells contributes to the local tissue inflammation

and the expression of matrix degrading enzymes, hence accelerating loss of proteoglycans [39].

Extensive production of mitochondrial ROS during IVD cell death may further impair disc tissue

homeostasis [92-96]. Apart from oxidative stress, also other factors, like non-physiological

mechanical loading, an acidic environment or auto-immune reactions, can alter the cellular

phenotype and enhance premature senescence and cell death in the IVD [276-278].

Despite the fact that IVD degeneration and back pain represents an economic burden,

only symptomatic therapies, like analgesics or invasive surgeries, are currently available. Such

therapies do not target the biological causes of disc degeneration and either offer only

temporary relief or entail risks and complications. The focus of current research is to restore the

disc function via prevention of premature aging, inhibition of cell death, and enhancement of

ECM maintenance with an ultimate goal to prevent low back pain [56, 265]. One approach to

compensate the loss of functional disc cells is to apply stem cells embedded in diverse

biomaterials mimicking disc tissue, which may support the old disc population and

“rejuvenate” the aging disc. However, the implanted cells can in some cases undergo incorrect

differentiation, can induce inflammatory reactions and may not survive due to prior catabolic

environment [17, 279]. Another approach is to enhance survival of the existing intrinsic disc cells

using various pharmacological agents such as growth factors or inhibitors of inflammatory

pathways [265, 280].

We have previously shown that epigallocatechin 3-gallate (EGCG), a polyphenol

naturally occurring in green tea, inhibits inflammatory responses in IVD cells in vitro and

exhibit analgesic activity against disc-related radiculopathy in experimental animals [18]. The

aim of this in vitro study was to (1) further investigate whether EGCG exhibits protective effects

in mild and high oxidative stress and to (2) investigate the molecular mechanism involved.

70

4.3 Methods

DPPH radical scavenging activity assay

Antioxidant activity of EGCG was determined by measuring its scavenging capacity of

the 2, 2’-diphyenyl-1-picrylhydrazyl (DPPH) radical as previously described [281]. 100 μl of 10-300

μM EGCG was incubated with 500 μl of DPPH (250 μM in ethanol; D9132 Sigma, Buchs,

Switzerland) for 1 hour at room temperature (n = 3). DPPH radical scavenging activity, which

manifests as a decrease in absorbance, was measured at 517 nm with a spectrophotometer

(Infinite M200 PRO, TECAN Group AG, Männedorf, Switzerland). L-Ascorbic acid (A4403, Sigma)

and ethanol (02860, Sigma) in equal amounts were used as positive and negative control,

respectively. DPPH radical scavenging activity (%) was calculated as [negative control optical

density (OD) – sample OD]*100 / negative control OD.

Cell isolation and cell culture

The study was approved by the cantonal ethic committee (Kantonale Ethikkommission

Zürich EK-16/2005). After informed consent was granted, human NP tissue (grade III-V) was

removed from donors undergoing spinal surgeries for degenerative disc disease or disc

herniation (n = 36). Details about the donors for this study are listed in Table 1. The tissue was

enzymatically digested using a mixture of 0.2 % collagenase NB4 (17454, Serva, Heidelberg,

Germany) and 0.3 % dispase II (04942078001, Roche, Basel, Switzerland) for 4-8 hours at 37 °C

and isolated primary cells were seeded in Dulbecco’s Modified Eagle’s Medium (DMEM/F12,

D8437, Sigma, St. Louis, MO, USA), supplemented with 10 % fetal calf serum (FCS, F7524, Sigma),

penicillin (50 units/mL), streptomycin (50 μg/mL) and ampicillin (125 ng/mL, 15240-062, Gibco,

Carlsbad, CA USA) and sub-cultured up to passage 3 using 1.5 % trypsin (15090- 046, Gibco).

Adherent cells in passage 1-3 were used for experiments. Due to the small biopsy size, low

proliferation rate and dedifferentiation of primary IVD cells in monolayer, a single donor cannot

be used for all required experiments. Therefore, different donors were randomly assigned to

the specific experiments. Cells were seeded on 6-well plates at a density of 1 × 105 cells per well

for all senescence experiments. Cells were seeded on 12-well plates at a density of 1 × 105 cells

per well for all lethal oxidative stress experiments, except JC-1 staining assay, which was

performed on 6-well plates. After seeding, cells were left to adhere overnight.

71

Table 1. Intervertebral disc pathologies used for primary cell cultures. DDD = degenerative disc disease, PI/A = Propidium iodide/Annexin staining, MTT = MTT assay, MTT (LY) = MTT assay with LY, MTT (I) = MTT assay with insulin, uk = unknown.

Donor information Experiments

Number Gender, Age Level Pathology, Grade

Sub-lethal oxidative stress Lethal oxidative stress

1 M, 35 L4/L5 Herniation IV

H2O2 sensitivity study, p21 time course, senescence validation 10 days: additive effect

H2O2 sensitivity study PI/A

2 F, 77 L4/L5 DDD III

p21 time course 10 days: additive effect

JC-1, MTT

3 F, 53 L4/L5 DDD IV

H2O2 sensitivity study, p21 time course, senescence validation 10 days: additive effect

H2O2 sensitivity study PI/A, MTT

4 F, 39 L5/S1 Herniation V

H2O2 sensitivity study senescence validation 10 days: antioxidant

H2O2 sensitivity study PI/A, MTT)

5 F, 31 L4/L5 Herniation uk

H2O2 sensitivity study senescence validation

H2O2 sensitivity study JC-1, PI/A, MTT


H2O2 sensitivity study senescence validation

H2O2 sensitivity study JC-1, PI/A, MTT

7 F, 53 L5/S1 Herniation IV

- JC-1

8 M, 50 L4/5 DDD III

- JC-1

9 F, 54 L5/S1 Herniation V

10 days: recovery , RT qPCR -

10 M, 64 L5/S1 Herniation III


11 M, 44 L5/S1 Herniation IV

10 days: antioxidant, RT qPCR -

12 M, 56 L4/L5 Herniation III


13 F, 51 C4/5 Herniation III

10 days: antioxidant -

14 F, 42 L3/L4 Herniation III


15 F, 35 L5/S1 DDD, Herniation IV

10 days: recovery -

16 M, 41 L4/L5 DDD III

10 days: additive effect -

17 M, 40 L5/S2 DDD, Herniation V


18 M, 50 L4/L5 DDD, Herniation III

10 days: additive effect -

19 F, 42 L3/L4 Herniation III

10 days: recovery -

20 M, 43 L5/S1 DDD, uk

- JC-1, bright field, MTT, MTT (I)

21 F, 51 L4/L5 DDD IV

- JC-1, bright field, MTT, MTT (LY), MTT (I)

72

Experimental design and treatments

Oxidative stress was induced with H2O2 (V800211, Sigma), which initiates the generation

of ROS that are able to attack all types of biomolecules [282, 283]. The active concentration of

H2O2 was selected based on the results of a preliminary H2O2 sensitivity study (n = 5). Increasing

concentrations of H2O2 (10-200 µM) were applied for 24 hours and cellular responses were

tested by MTT assay. The lethal concentration of H2O2 (100 µM and 200 µM for 24 hours) as well

as the sub-lethal concentration of H2O2, which can induce senescent changes without

significant cell death (50 µM H2O2 applied for 2 hours, followed by recovery period), were

identified. FCS-free media was used during all oxidative stress treatments to rule out the

possible interaction of radicals with FCS components. Previously reported non-toxic EGCG

concentrations (5 µM and 10 µM) were applied in this study [18].

22 M, 47 L4/L5 DDD, Herniation III

- JC-1, bright field, MTT



24 F, 43 L5/S1 DDD, Herniation III


25 M, 47 L4/5 DDD, Herniation III

- Lethal: MTT (LY), MTT (I)

26 M, 39 L4/5 DDD, Herniation IV




28 F, 43 L4/5 DDD, Herniation IV



- Lethal: WB

30 F, 39 L5/S1 DDD, Herniation III

- Lethal: MTT (LY), WB, MTT (I)

31 F, 39 L4/5 DDD, Herniation IV

- Lethal: MTT (LY), WB, MTT (I)

32 uk

uk - Lethal: MTT (LY), WB

33 M, 55 L3/4 DDD III

- Lethal: WB

34 F, 56 L3/4 Herniation IV

- Lethal: MTT (LY), WB

35 M, uk L4/5 Herniation uk

- Lethal: MTT (LY)

36 uk uk - Lethal: MTT (LY)

73

The effect of EGCG (E4143, Sigma) on premature senescence was tested in three

different experimental setups to identify mechanism of EGCG’s action (Table 2): (1) to test its

antioxidant activity, 10 µM EGCG was applied in the stress phase simultaneously with H2O2, (2)

to test its effects on post-stress processes, 10 µM EGCG was added only in the recovery phase

and (3) to combine its potential benefits in both phases, 5 µM EGCG was added in the stress

phase and later again in the recovery phase.

The effect of 10 µM EGCG on cell death was tested directly in the stress phase, without

recovery period (Table 2). To further test the involved mechanism, LY294002 – an inhibitor of

the PI3K/Protein kinase B (Akt) pathway (LY, L9908, Sigma) – was applied at a concentration of

10 µM, along with EGCG (n = 10). PI3K/Akt activator insulin (I9278, Sigma) was used at a

concentration of 0.5 µl/mL to confirm functionality and specificity of LY in our experiments (n =

10).

Table 2. Experimental design

Sub-lethal oxidative stress – induction of premature senescence

Experimental setup Stress phase (2 hours)

Recovery phase (up to 15 days)

Tested EGCG effect

1. Antioxidant 50 µM H2O2 + 10 µM EGCG

- ROS neutralization

2. Recovery 50 µM H2O2

+ 10 µM EGCG interaction with post-stress signaling (e.g. changes in gene expression)

3. Combined 50 µM H2O2 + 5 µM EGCG

+ 5 µM EGCG ROS neutralization interaction with post-stress signaling

Lethal oxidative stress – cell death induction

Experimental setup Stress phase (24 hours)

Tested EGCG effect

1. Survival 100 and 200 µM H2O2

+ 10 µM EGCG pro-survival antioxidant

In vitro model system of premature senescence

Premature senescence was induced with sub-lethal H2O2. After seeding, cells were

starved in FCS-free medium for 2 hours before applying 50 µM H2O2 for another 2 hours. As a

next step, media was replaced by complete media (+ 10% FCS) to let the cells recover from

oxidative stress. Premature senescence was monitored for 15 days with media exchange on day

5 and 10. To verify senescence-associated changes, various tests were performed: Senescence-

associated β-galactosidase activity was tested on day 1, 5, 10 and 15 post-stress (n = 5). The

number of viable cells was determined by Trypan blue exclusion test on day 8 and 15 post-stress

(n = 5). Cellular metabolic activity was measured by MTT assay and expression/activity of

senescence-associated proteins p21 and p53 was analyzed by immunoblotting on day 15 post-

stress (n = 5). In separate experiment on day 8, trypsin-detached cells were reseeded again to

check their ability to adhere, which reflects general cellular fitness.

74

Senescence-associated SA β-galactosidase assay

Senescence-associated β-galactosidase (SA β-gal), which accumulates in aging cells,

serves as a reliable marker of senescence in vitro [284]. Senescence was induced with 50 µM

H2O2, followed by EGCG treatments as described above (n = 5 for each experimental setup). For

SA β-gal analysis, staining solution containing 40 mM Na3citrate x 2H2O + 5 mM K4Fe(CN)6 + 5

mM K3Fe(CN)6 + 0.15 M NaCl + 2 mM MgCl2 in H2O was prepared and its pH was adjusted to 6 by

adding 0.5 M NaH2PO4 (all from Sigma). X-gal (11680293, Roche) was dissolved in

dimethylformamide (33120, Sigma) to prepare a 20 mg/mL stock solution. Adherent cells were

washed twice with PBS and fixed with 3% paraformaldehyde (30525894, Acros, Geel, Belgium)

in PBS for 5 minutes at RT. After the fixation, cells were washed twice with PBS and mixture of

X-gal + staining solution in a 1:20 ratio was added. After 18 hours, the staining solution was

removed, cells were washed two times in distilled water and passed through graded ethanol

(75%, 95%, 99%), 1 minute each. After air drying, micrographs were taken from 3 random spots

of each well using light microscopy. The number of all cells (minimum 300) and SA β-gal

positive cells was counted by processing images with GIMP (http://www.gimp.org/) and

ImageJ (http://imagej.nih.gov/ij/). The percentage of SA β-gal positive cells in each treatment

group was calculated as the number of SA β-gal positive cells/number of all cells*100. SA β-gal

positive cells in the treatment groups were displayed relative to the control groups, due to the

high inter-donor variability in basal SA β-gal staining.

Trypan blue exclusion test

Trypan blue staining, which distinguishes viable and dying cells, was performed to

analyze the total number of viable cells at the end of each experiment. Senescence was induced

with 50 µM H2O2 and EGCG treatments were performed as described above (n = 5 for each

experimental setup). Cells were harvested using 1.5% trypsin into the complete media, an

aliquot was mixed 1:1 with 0.4% Trypan blue dye (93595, Fluka) and the cell suspension was

immediately analyzed on the grids of the hemacytometer (DHC-N01, DigitalBio). The absolute

number of non-stained (viable) cells was determined in each group to make a comparison with

the total number of seeded cells (1 × 105 cells per well). Non-viable cells, which took up Trypan

blue dye, were excluded from the analysis.

Metabolic activity measurement

Metabolic activity, which reflects cellular viability, was determined using the MTT assay.

Following seeding, sub-lethal or lethal oxidative stress was applied and the treatments were

performed as described above (n = 5 for sub-lethal oxidative stress experiments, n = 10 for

lethal oxidative stress experiment, n = 10 for inhibition experiments). After 24 hours or 10 days,

fresh MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide, M5655, Sigma)

75

solution in media (0.5 mg/mL) was added and kept for 3 hours at 37°C. MTT was discarded, cells

were lysed in DMSO (D8418, Sigma) and the absorbance was measured at 565 nm. Metabolic

activity was calculated relative to the untreated control (100 %).

Immunoblotting

Treatments were performed as described above. Cells were harvested after 15 min

treatment for lethal oxidative stress experiments (n = 5) or after 10 days and 15 days for sub-

lethal oxidative stress experiments (n = 5). Whole cell lysates were prepared in RIPA buffer

(89900, Thermo Scientific, Waltham, MA, USA) according to producer’s instructions, mixed

with Laemmli buffer (S3401, Sigma), heated (99°C, 5 minutes), and then loaded onto 12% SDS

polyacrylamide gels. Separated proteins were transferred onto polyvinylidene difluoride (PVDF)

membranes (RPN303F, GE Healthcare, Little Chalfont, UK) and membranes were blocked in 5%

non-fat milk in Tris-buffered saline-Tween (TBS-T) for 1 hour at room temperature. Primary

antibodies were applied overnight at 4°C. After washing in 1% non-fat milk in TBS-T (3 × 10 min),

membranes were incubated with a secondary antibody conjugated to horseradish peroxidase

(HRP) for 1 hour at room temperature and washed in 1% non-fat milk in TBS-T (3 × 10 min).

Visualization was performed on medical X-ray film (28906836, GE Healthcare), using a

chemiluminescence kit West Dura (34076, Thermo Scientific) and films were scanned and

processed by GIMP. Tubulin was used as a loading control. Antibodies and dilutions: phospho-

p53 (Ser15) 1:1000 (9284, Cell Signaling, Danvers, MA, USA); p21 1:1000 (2947, Cell Signaling);

phospho-Akt (Ser473) 1:1000 (9271, Cell Signaling); Akt 1:1000 (9272, Cell Signaling); p16 Ink4a

(p16) 1:1000 (ab108349, Abcam); α-Tubulin 1:1000 (2144, Cell Signaling); mouse anti-rabbit IgG

HRP 1:5000 (7074, Cell Signaling).

Analysis of mitochondrial transmembrane potential (JC-1 staining)

JC-1 fluorochrome (5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolcarbocyanine

iodide) allows to monitor changes in mitochondrial transmembrane potential, which precedes

cell death. In physiological conditions, JC-1 accumulates in the intact mitochondrial membrane,

where it forms pronounced red aggregates. Once mitochondrial membrane potential is

disturbed, red fluorescence decreases, which is accompanied by an increase in green

fluorescence of the free JC-1 form. Substantial increase in the green fluorescence (= reduced

red/green ratio) suggests a breakdown of mitochondrial potential and activation of cell death.

Lethal oxidative stress was applied and EGCG treatment was performed as described

above (n = 10). After 24 hours, 1 µg/mL of JC-1 (T0046, Chemodex) diluted in FCS-free media was

added to the cells and kept at 37°C in the dark for 30 minutes. Then, cells were washed (3 × PBS)

and images of 3 random spots per well were taken immediately with a fluorescence microscope

76

(Olympus IX51, Volketswil, Switzerland). The ratio of red/green fluorescence was calculated

using ImageJ analysis (http://imagej.nih.gov/ij/).

Propidium iodide/Annexin V-FITC staining

The membrane of living cells is impermeable for Propidium iodide (PI), a red fluorescent

dye which intercalates into DNA. Annexin V-FITC (A) binds membrane phosphatidylserine,

which is exposed to the extracellular environment as one of the early apoptotic events. Flow

cytometry analysis of PI/A staining allows distinguishing viable cells (PI-negative/A-negative),

cells in early apoptosis (PI-negative/A-positive) and dead cells (PI-positive/A-positive) [285].

Lethal oxidative stress was applied and EGCG treatment was performed as described

above (n = 5). After 24 hours, supernatants were collected into pre-chilled tubes and placed on

ice. Cells were washed with PBS, detached from cell culture wells with 1.5% trypsin and

transferred into the corresponding pre-chilled tubes with matching supernatants. All the

further steps were performed on ice. Cell-supernatant mixtures were centrifuged (250 × g/5

min, 4°C), supernatants were discarded and 400 µL of Annexin Binding Buffer (556454, BD

Biosciences, Franklin Lakes, NJ, US) was added to the pellets and mixed. After the addition of 1

µL of PI (556463, BD Biosciences) and 2.5 µL of Annexin (556420, BD Biosciences), tubes were

kept in the dark for 15 minutes. Flow cytometry was performed immediately using the

cytometer FACS ARIA III (BD Biosciences). Results were analyzed with FlowJO

(http://www.flowjo.com/) and are displayed as percentage of viable cells (PI-/A-) in each

treatment group. Unstained cells, cells stained with Annexin V-FITC alone and cells stained

with PI alone were used as controls to set up quadrants.

Gene expression analysis (qRT-PCR)

The effect of sub-lethal oxidative stress on the activation of senescence-associated

secretory phenotype (SASP) was tested by means of gene expression of interleukin 6 (IL-6),

interleukin 8 (IL-8), matrix metalloproteinase 1 (MMP1), matrix metalloproteinase 3 (MMP3), and

matrix metalloproteinase 13 (MMP13). These genes are typically expressed in a pro-

inflammatory and catabolic environment during inflammation-related disc degeneration [82,

229]. IVD cells were seeded on 6-well plates and senescence was induced as described above (n

= 4). After 24 hours, RNA was extracted with the Trizol/chloroform method according to the

manufacturer instructions (15596-018, Invitrogen, Carlsbad, CA, USA) and 1 μg was reverse

transcribed to cDNA using a reverse transcription kit (4374966, Applied Biosystems). cDNA was

then mixed with primers and master mix (4352042, Applied Biosystems) and gene expression

was measured using real-time PCR. The following primers were used: TATA box binding protein

(TBP) Hs00427620_m1, IL-6 Hs00174131_m1, IL-8 Hs00174103_m1, MMP1 Hs00233958_m1, MMP3

Hs00968308_m1 and MMP13 Hs00233992_m1. Data was analyzed with the comparative Cq

77

method (2-ΔΔCq, housekeeping gene TBP). Results are presented as a fold change relative to the

untreated control group.

Statistical analysis

A power calculation (80% power) determined the required sample size (n = 5). Data

normality was tested by Schapiro-Wilk test. Statistical significance between two groups was

analyzed using Student’s t-test. Statistical significance between more than two groups was

evaluated by ANOVA with Tukey post-hoc test. The number of donors used for each experiment

is listed in the methods section. Mean and SEM are displayed in graphs and asterisks represent

a significance level of p ˂ 0.05.

4.4 Results

Validation of the in vitro model system of premature senescence of IVD cells

Based on the sensitivity study, 50 µM H2O2 applied for 2 hours and followed by a

recovery period up to 15 days was chosen to induce premature senescence of IVD cells without

significant cell death (Supplementary figure 1). Analysis of senescence-associated changes was

performed on day 1, 5, 8, 10 and 15 post-stress. In the H2O2 treatment group, the number of SA β-

gal positive cells increased significantly from day 1 (12.67±2.05%) up to day 15 (49.53±4.58%)

post-stress, compared with the untreated controls (Figure 1A). Representative images of SA β-

gal staining on day 8 and reseeded cells on day 9 are shown (Figure 1B). Reseeded cells in the

H2O2 treatment group were able to adhere, which confirmed a good general cellular fitness;

however, altered cellular morphology suggested an aging cell population.

Phosphorylation of the stress-induced protein p53 (Ser15) and enhanced expression of

the cell cycle inhibitor p21 on day 15 in the H2O2 treatment group indicated oxidative damage

(Figure 1C). In order to compare the replicative potential of stressed and control cells, the living

cells were counted on day 8 and 15 by Trypan blue exclusion test. Cell number on day 8 in the

H2O2 group (1.53±0.16 × 105 cells/well) was significantly lower than in the control group

(2.63±0.18 × 105 cells/well), confirming oxidative stress-induced loss of proliferative capacity.

Cell death in the H2O2 group on day 8 did not occur, as the number of cells was higher than the

cell seeding density on day 0 (1 × 105 cells/well, marked as red line) (Figure 1D). However, on day

15, the number of cells in the H2O2 treatment group decreased below the seeding density

(0.68±0.13 × 105 cells/well versus 1 × 105 cells/well, marked as red line), indicating that the cells

slowly started to die from the ROS exposure (Figure 1E). As cell death may hamper data

interpretation, all subsequent experiments were performed for a maximum 10 days.

78

The expression of inflammatory genes in the H2O2 treatment group, measured on day 1,

was not significantly elevated (Supplementary figure 1).

Figure 1. In vitro model system of stress-induced premature senescence. Sub-lethal oxidative stress (50 µM H2O2) with subsequent recovery period activated premature senescence of IVD cells in vitro. (A) Percentage of SA β-gal-positive cells in the H2O2 treatment group gradually increased during 15 days post-stress (n = 5). (B) Upper part: representative images of SA β-gal staining of the untreated (ctrl) and the H2O2-treated cells on day 8 post-stress, showing senescent (blue) cells. (B) Lower part: representative images of reseeded cells on day 9, confirming general cellular fitness. (C) Phosphorylation of p53 (Ser15) and expression of p21 in the H2O2 treatment group on day 15 post-stress indicated cellular senescence. (D, E) Proliferative capacity, displayed as number of cells on day 8 and 15 post-stress, was reduced in the H2O2 groups. On day 15, the number of cells in the H2O2 treatment group decreased below the seeding number (1 × 105 cells per well, depicted as red line), suggesting ongoing cell death. Asterisks indicate statistical significance at p < 0.05 (ANOVA, Tukey post-hoc, Student’s t-test).

79

Figure 2. As an antioxidant, EGCG inhibited senescence-associated β-galactosidase accumulation. (A) EGCG exhibited increasing 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging activity between 10 and 100 µM, which confirmed its antioxidant properties. Ascorbic acid in the same concentration was used as positive control (n = 3). (B-D) Oxidative stress was induced with 50 µM H2O2 for 2 hours and cellular senescence was measured during following 10 days. 10 µM EGCG inhibited SA β-gal accumulation when added to the oxidative stress phase, when its antioxidant activity was confirmed (n = 5). (C) 10 µM EGCG added to the recovery phase did not influence SA β-gal accumulation compared to the H2O2-only group (n = 5). (D) EGCG combined in both phases (5+5 µM) did not significantly inhibit SA β-gal accumulation, although a trend is visible (n = 5). Asterisks indicate statistical significance at p < 0.05 (ANOVA, Tukey post-hoc).

EGCG inhibited senescence-associated β-galactosidase accumulation in sub-lethal oxidative

stress

Previously reported antioxidant properties of EGCG were confirmed using 2,2-diphenyl-

1-picrylhydrazyl (DPPH) radical assay. EGCG exhibited dose-dependent radical scavenging

activity between 10 and 100 µM. As common antioxidant, ascorbic acid at the same

concentrations was used as positive control (Figure 2A) [286]. Apart from its antioxidant effects,

EGCG may also possess other beneficial functions. Therefore, the effect of EGCG on the

accumulation of SA β-gal was tested in 3 different experimental setups displayed in Table 2,

namely (1) antioxidant activity by adding EGCG in the stress phase, (2) activity on post-stress

signaling by adding EGCG in the recovery phase and (3) potential combined effect by adding

EGCG in both phases.

The percentage of SA β-gal positive cells was evaluated on day 1, 5, and 10 post-stress.

EGCG significantly inhibited the accumulation of SA β-gal on day 5 and 10 when added during

the stress phase, indicating that an antioxidant mechanism is involved (Figure 2B). EGCG did

80

not influence the rate of SA β-gal accumulation during the recovery phase, suggesting that the

interaction of EGCG with post-stress signaling is less relevant (Figure 2C). The presence of EGCG

in both phases did not provide significant protection against oxidative stress, which is most

likely caused by insufficient ROS neutralization, as lower concentrations of EGCG (5 µM) were

used in the stress phase of this setup. However, an apparent trend towards statistical

significance was detectable (Figure 2D).

EGCG did not protect IVD cells from loss of proliferative capacity in sub-lethal oxidative stress

The effects of EGCG on proliferation and expression of the senescence markers p53 and

p21 were also tested in each experimental setup: (1) EGCG as an antioxidant, (2) EGCG in

recovery and (3) combined effects. In order to compare the proliferative capacity of H2O2-

trerated and H2O2+EGCG-treated cell populations, the number of viable cells on day 10 of

recovery period was determined in each treatment group. Untreated and EGCG-treated cells

proliferated, but all stress groups showed significantly lower cell numbers, regardless of the

presence of EGCG (Figure 3A-C). On day 10 post-stress, p53 was slightly phosphorylated in all

stress groups. Expression of p21 was induced by H2O2, but also not affected by the presence of

EGCG, hence indicating that exogenous ROS were not entirely neutralized by 10 µM EGCG

(Figure 3D-F).

As both non-senescent and senescent cells are metabolically active, detected metabolic

activity on day 10 corresponds to the number of cells in each treatment group: stress groups

showed significantly lower metabolic activity than untreated controls. No significant difference

between the H2O2 and H2O2+EGCG treatment groups was detected in either experimental

setup, confirming that EGCG did not protect IVD cells from loss of proliferative capacity (Figure

3G-I). Metabolic activity of cells on day 10 is displayed relative to the control group (100%).

81

Figure 3. EGCG did not influence reactive oxygen species-induced loss of proliferative capacity. Senescence of IVD cells was induced with 50 µM H2O2 for 2 hours and proliferative capacity was evaluated on day 10 post-stress. (A-C) Over a period of 10 days the H2O2-treated cells proliferated significantly less than the cells in control groups. The number of cells in the H2O2 and H2O2+10 µM EGCG treatment groups did not significantly differ in either experimental setup (n = 5). (D-F) The activity of p53 and the expression of p21 was not significantly affected by EGCG in either experimental setup (n = 5). (G-I) Metabolic activity in the H2O2 and the H2O2+EGCG treatment groups did not significantly differ, indicating reduced proliferative capacity in all stress groups (n = 5). Asterisks indicate statistical significance at p < 0.05 vs. control group (ANOVA, Tukey post-hoc).

EGCG protected IVD cells from lethal oxidative stress via inhibition of mitochondrial

membrane depolarization

Based on the sensitivity study, 100 µM and 200 µM H2O2 applied for 24 hours was

identified to trigger significant cell death (Supplementary figure 1). As such ROS doses may

induce excessive mitochondria membrane depolarization and unrepairable damage, the

general pro-survival effect of EGCG can be tested.

Extensive oxidative stress (100 µM H2O2) significantly decreased the metabolic activity

of IVD cells (51.49±5.64%), compared with the control group (100%), while the addition of EGCG

reversed this effect (80.58±4.31%) (Figure 4A). More specific PI/A staining confirmed

82

significantly higher number of living cells in the H2O2+EGCG treatment group (52.91±8.02%)

compared to the H2O2-only group (28.31±5.83%) (Figure 4B).

During lethal oxidative stress mitochondria loses its transmembrane potential, which

leads to the release of mitochondrial ROS and serves as a pro-apoptotic signal. EGCG may

possibly interact with ROS on the mitochondrial membrane and prevent further ROS leakage.

To test this hypothesis, the JC-1 dye, which forms red aggregates in intact mitochondria, was

used. H2O2 in a concentration of 100 µM (Figure 5A) and 200 µM (Figure 5B) caused extensive

mitochondrial membrane depolarization, while addition of EGCG prevented this effect. The

red/green fluorescence ratio of the untreated (10.19±4.54) and the EGCG-treated cells

(8.15±3.29) was significantly higher than the ratio of both H2O2 treatment groups (100 µM:

1.85±0.96, 200 µM: 0.09±0.07). Addition of EGCG significantly increased the ratio in H2O2

treatment groups to 6.05±1.14 (100 µM) and 0.63±0.2 (200 µM) (Figure 5D). Membrane blebbing

and nuclear shrinkage was detected in the H2O2+EGCG treatment group, confirming that the

cells were not completely protected from deleterious ROS and cell death was possibly only

delayed (Figure 5C).

Figure 4. EGCG significantly inhibited reactive oxygen species-induced cell death. Cell death was induced by lethal concentration of H2O2 (100 µM) applied for 24 hours. (A) 10 µM EGCG significantly reversed the detrimental effects of H2O2 on metabolic activity, measured by MTT assay (n = 10). (B) 10 µM EGCG also significantly increased cell viability in oxidative stress, as measured by Propidium Iodide/Annexin staining (n = 5). (C) Visualization of representative Propidium Iodide/Annexin V staining measured by flow cytometry. Asterisks indicate statistical significance at p < 0.05 (Student’s t-test).

83

Figure 5. EGCG inhibited the loss of mitochondrial membrane potential. Cell death was activated by lethal concentrations of H2O2 (100 µM, 200 µM) applied for 24 hours. (A) 10 µM EGCG significantly inhibited mitochondrial membrane depolarization induced with 100 µM H2O2 (n = 10). (B) 10 µM EGCG significantly inhibited mitochondrial membrane depolarization induced with 200 µM H2O2 (n = 10). (C) Cellular morphology of the H2O2+EGCG treatment group was different from the control group: signs of membrane blebbing and nuclear condensation indicated that cell death was not completely inhibited (n = 5). (D) Ratio of red/green fluorescence showing the loss of mitochondrial membrane potential in the H2O2 treatment groups. Asterisks indicate statistical significance at p < 0.05 (ANOVA, Tukey post-hoc).

EGCG increased survival of IVD cells in lethal oxidative stress via PI3K/Protein kinase B (Akt)

activation

To further explore the mechanism underlying action of EGCG, activity of the pro-

survival PI3K/Akt pathway was studied. As shown by immunoblotting, EGCG activated Akt

84

under oxidative stress conditions (Figure 6A). The biological relevance of this finding was tested

using the PI3K/Akt-specific inhibitor LY294002 (LY). Depending on the type of stimulus, Akt can

activate various downstream signaling pathways inducing proliferation, glucose metabolism or

survival. Akt activator insulin, which stimulates cell proliferation, was used as positive control

to confirm the activity and specificity of LY. As expected, insulin modestly increased the

proliferation of IVD cells (117.79±2.91% versus 100% in control), whereas the addition of LY

abolished this effect (108.16±2.86%), hence confirming the activity of LY for further PI3K/Akt

pathway tests (Supplementary figure 1).

PI3K/Akt inhibition with LY did not significantly alter the pro-survival effect of EGCG at

24 hours, as measured by MTT assay (Figure 6B). However, at longer time point (48 hours), the

difference in the metabolic activity between the H2O2+EGCG (80.01±8.93%) and the

H2O2+EGCG+LY treatment group (34.81±4.69%) became significant (Figure 6C), indicating that

EGCG-based activation of the PI3K/Akt pathway plays indeed an important role in the survival

of IVD cells under the oxidative stress.

Figure 6. PI3K/Akt is important for the protective function of EGCG under the lethal oxidative stress. Cell death was activated by 200 µM H2O2. (A) After 15 minutes of co-treatment, 10 µM EGCG activated PI3K/Akt (n = 5). (B) After 24 hours, no significant difference in metabolic activity between the H2O2+EGCG and the H2O2+EGCG+LY groups was detected (n = 5). (C) After 48 hours, LY completely abolished the protective effects of EGCG, underlying the importance of the PI3K/Akt pathway in the survival of IVD cells

(n = 5). Asterisks indicate statistical significance at p < 0.05 (ANOVA, Tukey post-hoc).

85

4.5 Discussion

A stress-related decrease in the number of viable cells and phenotypic changes of living

cells are two major events in the process of IVD degeneration that warrant therapeutic

targeting [276]. Oxidative stress-induced senescence and cell death in vitro can represent

mechanisms involved in human disc aging in vivo [3, 39, 41]. We hypothesized that the natural

polyphenol EGCG can inhibit oxidative stress-induced senescence and/or cell death of IVD cells

and we investigated which mechanisms may be involved in its action.

In vitro as well as in vivo, EGCG can directly interact with ROS or operate in an indirect

manner by inhibiting ROS generating enzymes or by chelating potentially pro-oxidant metal

ions [31]. Nevertheless, antioxidant activity is not the only biological effect of EGCG. In fact,

many of its health-beneficial functions are independent of redox reactions. Depending on the

cell type, EGCG can interact with proteins and phospholipids in the plasma membrane, regulate

signal transduction and gene expression, DNA methylation, modulate mitochondrial function,

and autophagy [48].

Results of the radical scavenging assay confirmed the previously reported antioxidant

activity of 10 µM EGCG, which can neutralize 32.89±6.39% of 250 µM DPPH radicals. Although

higher concentrations of EGCG exhibited better ROS trapping, they were not used in this study

to avoid possible unfavorable interactions with cell culture media components. Dose-

dependent oxidation of polyphenols and subsequent ROS generation was reported as common

artifact of in vitro cell culture [287, 288]. The thus induced toxicity has no significance in vivo,

because of the presence of plasma antioxidants, but can interfere with research hypotheses via

increasing unspecific stress in cell culture. From this point of view, 10 µM EGCG was considered

active, yet experimentally safe.

We developed an in vitro model system of premature senescence of IVD cells, induced

by sub-lethal oxidative stress. Due to the vastly heterogeneous nature of the senescence

phenotype, the use of various markers is required to precisely identify senescence in a given cell

population [272]. The expression of certain markers, such as p21, can also increase during in

vitro cell culture as an artifact (Supplementary figure 1). Senescent cells commonly undergo cell

cycle arrest, start expressing SA β-gal, exhibit a senescence-associated secretory phenotype

(SASP) and activate p53-p21 and/or p16-RB pathways [40, 90]. Morphology of senescent cells

can differ from their young counterparts: cells become bigger, with prominent nucleoli and

numerous cytoplasmic vacuoles [289]. Sub-lethal doses of the widely used oxidative stress

inducer H2O2 inhibited proliferation and activated p53-p21 pathway and SA β-gal accumulation

in IVD cells, proving the suitability of the system to test potential anti-senescence compounds.

EGCG did not influence the progress of senescence when applied during the recovery

phase, suggesting no interaction with post-stress signaling. During recovery after oxidative

86

stress, cells can undergo several changes, one of which can be the activation of gene expression

towards the senescence-associated secretory phenotype (SASP). The SASP is mediated e.g. by

the transcription factor nuclear factor κB (NF‑κB) and many other redox-sensitive transcription

factors, which can induce the expression of pro-inflammatory cytokines, proteases and

catabolic enzymes [90, 270, 280]. We have shown before that EGCG inhibited IL-1β-activated

inflammatory responses in the IVD [18], implying that EGCG may act via similar mechanisms

also in the ROS-activated SASP. However, sub-lethal oxidative stress in our settings did not

increase the expression of inflammatory markers (IL-6, IL-8) and catabolic enzymes (MMP1,

MMP3, MMP13). Therefore, the effect of EGCG on SASP could not be evaluated (Supplementary

figure 1). Previously, stress-induced SASP was activated in IVD cells under different culture

conditions [40], in our setup either longer time point or higher ROS concentration will be

needed.

As an antioxidant, 10 µM EGCG inhibited accumulation of SA β-gal, but did not protect

IVD cells from loss of proliferative capacity, suggesting that the exogenous radicals were not

entirely neutralized. The increase in SA β-gal expression/activity results from the need to

compensate for the accumulation of damaged macromolecules and organelles in lysosomes. As

the expression of SA β-gal is not required for senescence, we hypothesize that EGCG in

antioxidant setup can ameliorate the increase in lysosome dysfunction, rather than inhibiting a

complex senescence phenotype. In accordance, EGCG has been shown to promote

macroautophagy and lysosome recycling of accumulated intracellular materials [290, 291].

Nevertheless, the role of EGCG in enhancing cellular waste-disposal mechanisms is still

controversial and presumably cell-type specific.

Since EGCG inhibited SA β-gal accumulation in antioxidant setup and not in recovery

setup, one could expect similar responses when combining EGCG in both stress and recovery

phase. However, in the third setup, SA β-gal assay showed only a trend towards the statistical

significance, possibly because lower EGCG concentration was applied in the stress phase, in

order to avoid potential toxic effects mentioned above [292].

As p16 is another important senescence marker, its expression was initially tested in IVD

cells treated with EGCG in antioxidant settings, on day 10 (n = 4). Because of the fact that p16

expression in H2O2-treatment groups was detected only in 1 out of 4 donors, the experiments

on p16 expression were not continued (Supplementary figure 2). In fact, the relative

contributions of the p53-p21 and p16-RB pathways to the senescence growth arrest may vary,

depending on the type and level of stress [91]. Moreover, in different cell types the transition of

temporal to stable cell-cycle arrest can involve either an activation of p21, or p16, or an

activation of both [91]. Human IVD cells are generally slowly proliferating and responding cell

type, in which ROS can possibly induce p16 with delayed kinetics. p16, if activated, subsequently

provides a second barrier to stop cell proliferation and plays a role in maintaining senescence,

87

rather than in its initiation, as suggested before [293, 294], which may explain its low

expression in our study.

To test whether EGCG can influence the expression of p16, SAOS-2 cells, with

constitutive expression of p16, were treated with 10 µM EGCG for 24 hours. This experiment

showed no influence of EGCG on the expression of p16 in SAOS-2 (Supplementary figure 2).

Lethal oxidative stress causes unrepairable damage on mitochondrial membranes,

leading to extensive ROS leakage and quick activation of death signals [295] (Supplementary

figure 3). Two independent viability assays revealed that EGCG can significantly increase

survival of IVD cells in lethal oxidative stress. As a primary target of oxidative stress,

mitochondria quickly lose their membrane integrity. We demonstrated that EGCG can delay

the onset of mitochondrial membrane depolarization, hence protecting mitochondria from ROS

leakage. Similar antioxidant effects of EGCG were previously described, e.g. in the mice liver-

injury in vivo model [296] or in the rat retinal primary cells in vitro [297].

As an exogenous oxidizing agent, H2O2 produces hydroxyl radicals, which can efficiently

react with all cellular components. On the other hand, damaged mitochondria generate mainly

superoxide, followed by lower amounts of hydrogen peroxide and hydroxyl radicals [298, 299].

Therefore, EGCG can affect mitochondrial ROS leakage by being a good superoxide anion

radical scavenger [299], but it may neutralize other ROS species less efficiently (Supplementary

figure 3). As previously reported, EGCG can also activate superoxide dismutase expression,

another mechanism to counteract exclusively superoxide anion radicals and to protect

mitochondria [300, 301].

Interestingly, EGCG activated PI3K/Akt under lethal oxidative stress, an ability that was

important for IVD cell survival. We suggest that the reason why EGCG alone did not activate

PI3K/Akt pathway can be an absence of the stress signal (H2O2). In non-stressed cells, the

activity of pro-survival pathways (such as PI3K/Akt) is negatively regulated by their inhibitors

(such as PTEN). In stressed cells, the activity of intracellular kinase inhibitors can be reduced

[302] in order to facilitate activation of the pro-survival kinases. However, the pro-survival

kinases can fully function only upon external activation [303]. Interestingly, it has been shown

that EGCG interacts with various membrane receptors [48, 304]. We therefore hypothesize that

EGCG-mediated interaction upstream of Akt, in combination with stress-mediated reduction in

the activity of intracellular kinase inhibitors, can lead to complete PI3K/Akt activation, hence its

pro-survival effects in H2O2-treated cells.

Previously, Akt gene expression has been correlated with increased disc cell proliferation

in vitro [305] and lumbar disc herniation in vivo [306]. More importantly, active PI3K/Akt

signaling enhanced pro-survival capacity of disc cells in vitro [307] and antagonized disc

degeneration in vivo [308]. The PI3K/Akt pathway can therefore represent another attractive

therapeutic goal in disc degeneration, which can be targeted by EGCG.

88

In this study, 2D cell culture was employed to ensure applicability of cell analysis assays.

Adherent in vitro system, in contrast to 3D, reduced cell loss during analyses, which is

particularly important for stress experiments. The 2D system also allowed quick and precise

staining of organelles in living cells. While results are very clear, it remains open whether cells

react similarly in their natural 3D environment. This study is furthermore limited by the inter-

donor variability of disc tissue biopsies, originating from differences in age and degeneration

grades [309]. Nevertheless, the use of primary cells during first three passages represents a

system more close to human tissue than a stabilized cell line. The inter-donor variability issue

can be corrected by displaying results relative to a control group, instead of using absolute

numbers. The main third limitation is that the effect of EGCG during the recovery phase of

senescence was tested in complete media, which could possibly influence the activity of EGCG.

4.6 Conclusions

This study showed that EGCG can counteract oxidative stress-induced changes of IVD

cells in vitro via protection of the mitochondrial membrane from depolarization. In oxidative

stress, EGCG also activated PI3K/Akt as an important pro-survival mechanism of IVD cells. As

this study did not provide direct evidence whether and how the EGCG-activated Akt and

mitochondria protection are linked, investigation of this coupling in IVD cells will be an

objective of future research.

Together with previously reported anti-inflammatory, anti-catabolic and analgesic

effects of EGCG our findings can be further used for the development of novel therapies

targeting consequences of oxidative stress and neovascularization in degenerative disc disease.

4.7 Authors’ contributions

OK carried out sub-lethal stress experiments, participated in lethal stress experiments,

contributed to study design and drafted the manuscript. JH participated in lethal stress

experiments, performed pathway analysis and statistical analysis. MH performed gene

expression experiments and statistical analysis. JK participated in study design, provided

clinical samples and medical scientific input. CO conceived funding, participated in study

design and provided scientific input. SJF conceived funding, helped with study design,

coordination and helped to draft the manuscript. KW conceived funding, designed and

coordinated the study and helped to draft the manuscript. All authors approved a final version

of the manuscript.

89

4.8 Acknowledgement

The authors would like to thank the Herzog-Egli Foundation and the Mäxi Foundation

for financial support, Sonia Rossi for preliminary H2O2 tests and Helen Greutert and ETH FACS

core facility for technical support. The authors have no conflict of interests.

4.9 Supplement

Supplementary figure 1. H2O2 sensitivity study, p21 expression during cell culture period, the effect of H2O2 on senescence-associated secretory phenotype and confirmation of the LY294002 activity/specificity. IVD cells were treated with different concentrations of H2O2 to determine the cellular metabolic activity by MTT assay. (A) After 24 hours 50, 100 and 200 µM H2O2 caused a significant decrease in the metabolic activity. Thus concentrations 100 and 200 µM were selected for lethal oxidative stress induction (n = 5). (B) Metabolic activity of IVD cells treated with 50 µM H2O2 for 2 hours did not change within 3 days, therefore this concentration was chosen for sub-lethal oxidative stress induction (n = 5). (C) The expression of p21 on day 1, 5, 10 and 15 was tested by immunoblotting, with

90

etoposide (10 µM, 72 hours) as a positive control. p21 expression increased during the cell culture period as an artifact of the in vitro environment (n = 3). (D) Sub-lethal oxidative stress (50 µM H2O2, 2 hours) did not increase the expression of inflammatory markers (IL-6, IL-8) and catabolic enzymes (MMP1, MMP3, MMP13) on day 1 post-stress (n = 4). (E) Akt activator insulin at 0.5 µl/mL increased proliferation of IVD cells whereas addition of 10 µM LY294002 (LY) abolished this effect, confirming functionality of LY for further tests involving Akt. Asterisks indicate statistical significance p < 0.05 (ANOVA, Tukey post-hoc).

Supplementary figure 2. The expression of p16 Ink4a in oxidative stress-induced senescence of IVD cells. Senescence was induced with 50 µM H2O2 for 2 hours and 10 µM EGCG was added directly to the stress phase (antioxidant experimental setup). The expression of p16 was measured on day 10 post-stress with SAOS-2 cells as positive control. (A) The expression of p16 is constitutively active in SAOS-2 cells. The expression of p16 was not induced by H2O2 in 3 out of 4 donors, therefore the effects of EGCG could not be studied (n = 4). (B) The constitutive expression of p16 in SAOS-2 cells is not affected by 10 µM EGCG.

91

Supplementary figure 3. The role of reactive oxygen species and mitochondria in senescence and apoptosis of IVD cells (schematic). (A, C) Sub-lethal oxidative stress causes damage of cellular macromolecules, which activates the p53-p21 pathway and growth arrest, without significant mitochondrial depolarization. During growth arrest, cells can repair the damage or irreversibly enter senescence, depending on various factors, such as activity of cellular antioxidants and inflammatory pathways. (B, D) Lethal oxidative stress leads to extensive mitochondrial membrane depolarization and subsequent mitochondrial ROS leakage, which induces apoptosis. Based on our data, EGCG can protect mitochondrial membrane from ROS leakage and prevent subsequent activation of cell death cascade. (C, D) The ratio of red/green fluorescence, with a loss of mitochondrial membrane potential in the H2O2

treatment groups is shown. Data presented in (D) are replicated from Figure 5D to emphasize the difference between the mitochondrial membrane potential in senescence and apoptosis. Asterisks indicate statistical significance at p < 0.05 (ANOVA, Tukey post-hoc) (n = 5).

93

Chapter 5

An inflammatory nucleus pulposus tissue

culture model to test molecular regenerative

therapies: Validation with epigallocatechin 3-

gallate

Olga Krupkova, Marian Hlavna, Julie Amir Tahmasseb, Joel Zvick, Dominik

Kunz, Keita Ito, Stephen J. Ferguson, Karin Wuertz-Kozak

Under review in the International Journal of Molecular Sciences

95

5.1 Abstract

Organ cultures are practical tools to investigate regenerative strategies for the

intervertebral disc. However, most existing organ culture systems induce severe tissue

degradation with only limited representation of the in vivo processes. The objective of this

study was to develop a space- and cost-efficient tissue culture model, which represents

degenerative processes of the nucleus pulposus (NP). Intact bovine NPs were cultured in a

previously developed system using Dyneema jackets. Degenerative changes in the NP tissue

were induced either by the direct injection of chondroitinase ABC (1–20U/mL) or by the

diffusion of IL-1β and TNF-α (both 100ng/mL) from the culture media. Extracellular matrix

composition (collagens, proteoglycans, water, DNA) and the expression of inflammatory and

catabolic genes were analyzed. The anti-inflammatory and anti-catabolic compound

epigallocatechin 3-galalte (EGCG, 10µM) was employed to assess the relevance of the

degenerative NP model. Although a single injection of chondroitinase ABC reduced the

proteoglycan content in the NPs, it did not activate cellular responses. On the other hand, IL-1β

and TNF-α significantly increased the mRNA expression of inflammatory mediators (IL-6, IL-8,

iNOS, PTGS2) and matrix metalloproteinases (MMP1, MMP3, MMP13). The cytokine-induced

gene expression in the NPs was ameliorated with EGCG. This study provides a proof of concept

that inflammatory NP cultures, with appropriate containment, can be useful for the discovery

and evaluation of molecular therapeutic strategies against early degenerative disc disease.

Keywords

nucleus pulposus, organ culture model, degenerative disc disease, cytokines, chondroitinase

ABC, epigallocatechin gallate, inflammation, extracellular matrix

96

5.2 Introduction

The intervertebral disc (IVD) consists of three structurally and functionally different

tissues: the outer annulus fibrosus (AF), the inner nucleus pulposus (NP), and two cartilaginous

endplates that connect the disc with the adjacent vertebrae [1, 56]. IVD degeneration is caused

by a combination of many events throughout a lifetime, including non-physiological loading,

endplate calcification, inflammation, immune reactions, cell senescence, and cell death [1-3],

and occurs initially and predominantly in the NP.

The inflammatory cytokines interleukin-1 beta (IL-1β) [82, 84] and tumor necrosis factor

alpha (TNF-α) [28] are key molecules involved in IVD degeneration. The events that promote

their production by NP and AF cells, especially in the absence of trauma or herniation, can be

chronic stress, spinal infection or free swelling, i.e. during herniation [2, 81, 310]. IL-1β and TNF-α

bind to their cell surface receptors (e.g. IL-1R1, TNFR1), which activate the signaling mediators

NF-κB, JNK, and p38 MAPK, resulting in the transcription of inflammatory and catabolic genes

such as interleukins (IL-1α, IL-1β, IL-6, IL-8), matrix metalloproteinases (MMP1, MMP3 and

MMP13), aggrecanases (ADAMTs) and other relevant molecules that can promote inflammation

and pain (PTGS2, NGF, iNOS) [2, 18, 86, 88]. In addition to the catabolic and pro-inflammatory

effects, IL-1β and TNF-α can influence disc cell senescence, autophagy and the expression of

genes involved in proliferation [80]. MMPs and aggrecanases further promote degradation of

the extracellular matrix (ECM) [85-87]. Changes of the NP ultimately result in a loss of disc

hydration, disc height and load-bearing capacity at later stages, when the AF is also affected [4,

60]. These accelerated pathological processes with structural failure and the development of

low back pain, e.g. due to nerve compression and/or chemical irritation, are together termed

degenerative disc disease, a distinct entity to the process of physiological ageing [4-7].

Despite an increasing prevalence of discogenic back pain and the consequently high

economic burden [9], only therapies targeting symptoms are currently available [10, 11]. Most of

the current research on novel treatments focuses on restoring the function and homeostasis of

degenerated discs via inhibition of inflammation, prevention of premature ageing, and

augmentation of the ECM content. Cell therapies (e.g. mesenchymal stem cells) can support

the existing cell population and “rejuvenate” the aging disc [15, 16]. However, newly implanted

cells are at risk of undergoing apoptosis, due to the surrounding inflammatory and catabolic

environment [17]. Inflammation, catabolism and matrix degradation in the disc can be

controlled using molecular therapeutics, such as anti-inflammatory compounds [18, 19], growth

factors [20, 150] and N-terminus of Link protein (LinkN) [21]. Application of growth factors like

bone morphogenetic proteins (BMP-2, BMP-7), insulin growth factor (IGF-1) or tumor growth

factor (TGF-β3) can induce the formation of new matrix and improve the mechanical properties

of the disc [147]. However, drawbacks of exogenous growth factors are their short half-life

97

(hours to days) and possibly only transient effects. Another complication is their large size,

which can limit their diffusion through the NP [109].

Diverse signaling pathways are deregulated in disc degeneration. Therefore, small

molecules with multiple effects, targeting both survival and function of resident disc cells, are

promising therapeutic candidates. Compounds derived from traditional medicinal plants act on

several signal transduction and apoptotic cascades at the same time, while having minimal

side effects [22]. We have shown before that polyphenols such as resveratrol [170], curcumin

[171], triptolide [173] and epigallocatechin 3-gallate (EGCG) [18, 311] exhibit anti-inflammatory,

anti-catabolic and antioxidant activity in disc cells. Although more research is needed to fully

understand the effects of polyphenols on the integrity of the IVD, it has been suggested that

they can delay degeneration of the NP or even reverse it. Polyphenols can also enhance disc

nutrition via improving vascular health [174] and relieve radicular pain [18, 170].

A number of organ culture models have been developed over the past years to facilitate

testing of novel regenerative treatments. Unlike cell cultures, organ culture models retain

native ECM and allow for mechanical loading [23]. Induction of degeneration in these models

commonly is achieved by injection of matrix degrading enzymes such as trypsin [24, 25] and

papain [312, 313]. Enzymatic tissue damage allows evaluation of the efficiency of cell therapies,

but is less efficient for testing molecular therapeutics. Cells injected or implanted within a

supportive scaffold can fill in the missing ECM, proliferate, produce new matrix and restore

tissue properties. However, while ECM degradation is often extensive, and thus higher than

clinically observed, no or limited cellular inflammatory and catabolic reactions have been found

in these enzymatic models. This suggests that in vivo disc degeneration is not fully mimicked.

Mechanical damage such as endplate fracture [26] or stab incision in animal models [314] can

activate cellular responses in the disc more reliably than non-physiological proteolytic enzymes.

Recently, a stab injury model, based on motion segments isolated from transgenic animals

with an NF-κB-luc reporter, was developed. Longitudinal monitoring of NF-κB activity in this

model can provide insights into the mechanisms that regulate IVD inflammatory responses to

stab injury [315]. At the same time, physiologically relevant cytokine-induced inflammatory

organ culture models, without involvement of mechanical damage, have emerged [27-29].

Whole disc organ cultures with biologically relevant ECM degradation and inflammatory

responses are certainly most suitable for testing cell therapies and regenerative compounds.

However, although whole-disc organ cultures have many advantages, they typically do not

allow for high throughput experiments due to space-related demands and sample

heterogeneity. For this reason, a nucleus pulposus (NP) culture in Dyneema jackets has been

recently developed [30], (further referred to as “NP culture”). The Dyneema jackets (also called

“artificial AF”) restrain swelling, while maintaining diffusion of nutrients from culture media.

This bovine NP culture was previously shown to preserve a glycosaminoglycan (GAG) content

98

and gene expression pattern similar to native tissue over sustained periods, which makes it

suitable for further use [30]. The objective of this work was to adapt the existing NP culture as a

novel platform for testing regenerative treatment strategies.

We hypothesized that (1) physiological degradation of the ECM can be induced by injection

of the biologically relevant enzyme chondroitinase ABC and that (2) degenerative responses can

be activated by diffusion of IL-1β and TNF-α from culture media, hence allowing for the creation

of a suitable testing platform for novel disc therapeutics. (3) Furthermore, we hypothesized that

the anti-inflammatory and anti-catabolic compound EGCG will ameliorate this degenerative

response. Using bovine caudal NPs for preclinical testing is not only cost-effective, but also in

accordance with the 3R animal welfare recommendations.

5.3 Methods

Bovine nucleus pulposus tissue culture

NP cultures were prepared as previously described, with minor changes [30]. Bovine tails

were obtained from the local slaughterhouse (18-24 month old cows), muscles and fat were

removed and vertebrae were cut to obtain individual motion segments of the first 4 – 5 intact

proximal discs. The motion segments were sterilized with betadine for 20 minutes and further

washed under sterile conditions with 70% ethanol. Intervertebral discs were cut transversally,

close to the endplate, and nucleus pulposus tissue was isolated using an 8-mm biopsy punch

(8437995, Polymed) and scalpel (Scheme 1A). NP biopsies were placed in 1.5 mL tubes and

weighed.

For NP pre-shrinkage, dialysis membranes (Spectrapor 15 kDa MWCO/18 mm, 734-0507,

VWR) were washed in PBS, opened with tweezers and one NP each was pushed inside using a

custom-made cylindrical metal device. The Spectrapore membranes with NPs were then placed

into membrane clips, closed and allowed to shrink in a 6-well plate with 5 ml of 30% w/v

polyethylene glycol (PEG) solution in PBS for 100 minutes (Scheme 1B). After that, the

membrane clips were opened, the membranes were cut open and the NP biopsies were

weighed again. NP biopsies were placed on a regenerated cellulose membrane (97010-104

RC58, 0.2 µm pore size, GE) and folded into the membrane. Then NPs were inserted into the

custom-made Dyneema jacket (UHMWPE fiber, Dyneema, DSM, The Netherlands). The open

end of the jacket was twisted once and turned inside-out again over the sample. This step was

repeated once more to create three layers of the jacket, preventing excessive swelling. A sterile

needle and thread was used to suture the jacket closed (Scheme 1C).

The encapsulated NP cultures were placed into 6-well plates with 10 mL of DMEM/F-12

(D8437, Sigma, St. Louis, MO, USA) supplemented with 10% FCS (F7524, Sigma), penicillin (100

99

units/mL), streptomycin (100 μg/mL) and ampicillin (250 ng/mL, 15240-062, Gibco). 2% FCS was

used for experiments with IL-1β and TNF-α. NP tissues were cultured for up to 21 days. At each

time point (day 3, 7, 14, 21), NP cultures were harvested for analysis, weighed, immediately

frozen in liquid nitrogen and stored at -80°C.

Scheme 1. Preparation of bovine NP cultures. (A) Bovine tails were dissected and NP tissue was isolated using an 8-mm biopsy punch. (B) NP tissue was placed in 30% w/v PEG solution (100 minutes) to reduce water content. (C) NPs packed in regenerated cellulose membranes and Dyneema jackets were sutured and placed in culture media.

Induction of nucleus pulposus degeneration

To induce early degenerative changes of NP tissue, two approaches were tested (1) enzyme

treatment and (2) cytokine treatment. In the first approach, NPs were injected with

chondroitinase ABC (1 - 20 U/mL), which selectively removes chondroitin-4-sulphate, chondroitin-

6-sulphate and dermatan sulphate, the major components of proteoglycans [316].

Chondroitinase ABC is a known inducer of disc degeneration in vitro and in vivo [316, 317].

However, it is not known whether chondroitinase ABC can activate inflammatory responses in NP

tissues. Lyophilized chondroitinase ABC from Proteus vulgaris (120 kDa, C3667, Sigma) was diluted

according to the producer’s instructions in PBS and 1% BSA to 20 U/mL and stored as a stock

solution at -20ºC until use. 20 µl of 1, 5, 10 or 20 U/mL chondroitinase ABC was injected with a 27G

needle in the center of the NPs before packing in Dyneema jackets. In the second approach,

human recombinant TNF-α (17 kDa, 100 ng/mL, PHC3016, Gibco) and bovine recombinant IL-1β (31

kDa, 100 ng/mL, RBOIL1BI, Pierce) were added to the culture media on day 0 and 7. Re-swelling of

the NP in the Dyneema jacket to its initial weight is thought to allow the media with IL-1β and

TNF-α to actively enter the NP tissue. Samples were collected for analyses of degenerative

responses and ECM composition on days 0 (native), 3, 7, and 14. Epigallocatechin 3-gallate (EGCG,

100

10 µM), which inhibits inflammatory and catabolic responses in IVD cells in vitro [18], was selected

to test the relevance and applicability of the degenerative NP culture model for therapeutic

testing. The experimental setup is depicted in Scheme 2. The number of NP cultures used for each

experimental group, treatment, and type of analysis are listed in Table 1.

Scheme 2. Experimental setup. (A) NP tissues were cultured up to 21 days to confirm the system’s stability. (B) Chondroitinase ABC was injected into NP cultures. (C) TNF-α and IL-1β were added into the culture media. Gene expression and extracellular matrix (ECM) composition were analyzed on days 0, 3, 7, 14, and/or 21. Histological analysis was performed to confirm GAG depletion in the enzyme treatment group. Epigallocatechin gallate was used to validate cytokine treatment group.

101

Table 1. Number of NP cultures used for each experimental group, treatment, and type of analysis. ECM = extracellular matrix.

Experimental groups Analyses Treatments Gene expression ECM composition Native tissue n = 8

n = 5

Stable NP culture Day 3: n = 8 Day 7: n = 8 Day 14: n = 8 Day 21: n = 8

Day 3: n = 4 Day 7: n = 4 Day 14: n = 4 Day 21: n = 4

Enzyme treatment 1. PBS 2. Chondroitinase ABC 1 U/mL 3. Chondroitinase ABC 5 U/mL 4. Chondroitinase ABC 10 U/mL 5. Chondroitinase ABC 20 U/mL

Day 3: n = 4 / treatment Day 7: n = 4 / treatment Day 14: n = 4 / treatment


Cytokine treatment 1. Control 2. TNF-α + IL-1β (100 ng/mL) 3. EGCG (10 µM) 4. TNF-α + IL-1β + EGCG



Total n = 321 NPs

RNA isolation

RNA was extracted with the classical TRIzol/chloroform method, but optimized for bovine

discs. Each bovine NP sample was quickly ground and pulverized in liquid nitrogen using

custom-made grinders. The NP powder was transferred into 1 mL of TRIzol (15596018, Thermo

Scientific) and the mixture was further homogenized with a polytron 4 times/20 seconds

(POLYTRON® PT 10/35 GT). During homogenization, samples were kept on ice. Homogenized

samples were transferred to 15 mL tubes with 9 mL of fresh TRIzol, mixed and left at room

temperature for 5 minutes. Then 2 mL of chloroform (C7559, Sigma) was added, samples were

vortexed and left 10 minutes on a shaker. After shaking, samples were centrifuged (4000g/10

minutes) and the upper phase (5 mL) was transferred into new 15 mL tubes before adding 10 mL

of 2-propanol (I9516, Sigma), followed by mixing, vortexing and centrifugation after 5 minutes

at room temperature (4000g/10 minutes). Isopropanol was discarded, 1 mL of 70% ethanol

(51976, Sigma) was added, and samples were transferred into 2 mL tubes and centrifuged

(7500g/5 minutes). If the pellet contained white proteoglycans, samples were thoroughly

vortexed for 2 minutes and a second TRIzol/chloroform isolation in smaller volume (2 mL tube)

was performed for additional purification. Purified samples were centrifuged (7500g /5

minutes), ethanol was aspirated and the pellets were dried for 5 minutes at room temperature.

Dry pellets were then mixed with 30 µL of RNase-free water and immediately frozen at -80°C

until further analysis.

102

Gene expression (RT-qPCR)

As RNA yields were low, due to the low cellularity of the tissue and the high amount of

proteoglycans [318], the whole RNA solution (30 µL) was reverse transcribed to cDNA using a

reverse transcription kit (4374966, Applied Biosystems). Resulting cDNA was diluted 3-5 times

in RNase free water, mixed with primers and master mix (4352042, Applied Biosystems) and

gene expression was measured by real-time PCR (CFX96 Touch™ Detection System, Biorad).

Bovine TaqMan primers for interleukin-6 (IL-6), interleukin-8 (IL-8), interleukin-1β (IL-1β), tissue

inhibitor of matrix metalloproteinases (TIMP1), matrix metalloproteinase 1 (MMP1), matrix

metalloproteinase 3 (MMP3), matrix metalloproteinase 13 (MMP13), aggrecanase 1 (ADAMTS4),

prostaglandin synthase 2 (PTGS2), inducible nitric oxid synthase (iNOS), collagen 1, collagen 2

and aggrecan were employed to study the inflammatory and catabolic phenotype. S18, Actin

and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were used as housekeeping genes

(Table 2). Obtained Cq values were analyzed by the comparative method (2-ΔΔCq) and displayed

as fold change to control or relative to IL-1β and TNF-α stimulation.

Table 2. TaqMan primers and sequence numbers used for qPCR analysis.

Gene Primer Seq. N. Gene Primer Seq. N. S18 Bt03225193_m1 ADAMTS4 Bt03224693_m1 GAPDH Bt03210919_g1 MMP1 Bt04259497_m1 Actin-β Bt03279174_g1 MMP3 Bt04259495_m1 Interleukin-6 Bt03211904_m1 MMP13 Bt03214052_m1 Interleukin-8 Bt03211908_m1 TIMP1 Bt03223720_m1 Interleukin-1β Bt03212741_m1 Collagen 1 Bt03214883_m1 iNOS Bt03249584_m1 Collagen 2 Bt03251861_m1 PTGS2 Bt03214492_m1 Aggrecan Bt03212186_m1

Extracellular matrix composition (glycosaminoglycans, hydroxyproline, DNA, water)

NP samples were dried at 60°C overnight and weighed to determine dry weight. The dried

samples were then digested in papain buffer (55mM Na-Citrate, 150mM NaCl, 5mM EDTA, pH

6) with 5 mM L-cystein and 125 µg/mL papain from papaya latex (P-3125, Sigma) at 60°C

overnight. The digested samples were used to quantify the content of sulfated

glycosaminoglycans (GAG), collagen (hydroxyproline) and DNA. GAGs were measured using

dimethylmethylene blue (DMMB) assay with chondroitin sulfate as standard (C6737, Sigma)

[319]. Samples were mixed with DMMB stock solution (10.5 mg DMMB, 2.5 ml absolute ethanol,

1.0 g sodium formate in 500 ml dH2O, pH 1.5) in a 96-well plate and the absorbance was

measured at 595 nm. The collagen content was analyzed with the hydroxyproline (HYP) assay

kit (MAK008, Sigma) and DNA was determined using the Quant-iT PicoGreen dsDNA assay kit

(P11496, Thermo Fischer) according to the manufacturer’s recommendations. The amounts of

GAG, HYP, DNA and water were expressed per mg dry or wet weight of the tissue (for stable NP

culture) or relative to the control (for degeneration groups) [30].

103

Histology

After harvesting, NPs were fixed in 4% buffered paraformaldehyde (48 hours, 5

mL/sample). After fixation, specimens were washed, dehydrated and stored in 75% ethanol

[320]. Before paraffin embedding, samples were cut transversally into two halves. After

embedding, 3 µm sections were prepared with a microtome (CryoStar NX70, ThermoScientific)

and stained with Alcian blue/Picosirius red and Safranin O. Images were taken with a CCD

camera mounted on an Olympus IX51 microscope (Volketswil, Switzerland).

Statistical analysis

All data sets were tested for normal distribution with the Shapiro-Wilk test and found to

be normally distributed. A Student’s t-test (2 groups) and a one-way ANOVA with Tukey post-

hoc test (more than 2 groups) were performed to detect statistical significance of differences at

p < 0.05.

5.4 Results

Nucleus pulposus tissue in Dyneema jackets is stable over a period of 21 days

A stable amount of GAG, collagen, DNA and water reflects healthy extracellular matrix.

These parameters were determined in the NP cultures on day 3, 7, 14 and 21 and compared with

native tissue (day 0). The content of GAG, collagen, DNA and water in the NP cultures did not

significantly differ from the native tissue, confirming stability of the ECM [30] (Figure 1A-D).

Gene expression of MMP3 and TIMP1 increased on days 3 and day 7, which may reflect a tissue

response to culture conditions. Their expression was downregulated at later time points (Figure

1E). Genes for other matrix degrading enzymes (MMP1, MMP13) and interleukins (IL-6, IL-8, IL-1β)

were not detectable. In comparison with the native tissue, gene expression of aggrecan and

collagens decreased between day 0 and day 3, which can be explained by the absence of

physiological loading [30]. After this initial down-regulation, the expression of ECM genes

remained stable (Figure 1E). Average Cq ± SD values for genes expressed in the stable NP

cultures are listed in Table 3.

104

Figure 1. Bovine NP culture preserves extracellular matrix over a period of 21 days. (A) Glycosaminoglycan, (B) collagen, (C) DNA and (D) water content on day 3, 7, 14 and 21 did not differ from the native values (day 0). Graphs show the percentage of dry or wet weight relative to the native tissue (mean ± SEM, n = 4 per group). (E) Gene expression of MMP3 and TIMP1 increased and subsequently returned close to the native levels. Gene expression of aggrecan, collagen 1 and collagen 2 was down-regulated after NPs were placed in culture, but remained stable thereafter. Graph shows fold change relative to the native tissue (mean ± SEM, n = 4-8 per group). *p < 0.05 vs. native tissue, #p < 0.05 vs. each other (ANOVA). Table 3. Average Cq ± SD values for all expressed genes in stable NP cultures (day 3).

Gene mean Cq ± SD Gene mean Cq ± SD

S18 28.24 ± 5.22 Collagen 2 26.92 ± 2.85

GAPDH 29.69 ± 1.78 Aggrecan 27.13 ± 3.33

Actin-β 30.02 ± 2.98 MMP3 33.65 ± 3.55

Collagen 1 35.52 ± 3.14 TIMP1 27.78 ± 2.49

Chondroitinase ABC reduces proteoglycan content in the NP

As chondroitinase ABC removes chondroitin sulfate from proteoglycans, a dose- and/or

time-dependent decrease in sulfated GAGs was expected. Compared with the PBS injection, the

injection of 20 U/mL chondroitinase ABC significantly reduced GAG content at all tested time

points (Figure 2A). No major alterations in collagen, DNA and water content were detected over

a period of 14 days (Figure 2B-D). Alcian blue/Picrosirius red and Safranin O staining confirmed

that 20 U/mL of chondroitinase ABC depleted proteoglycans, whereas 1 U/mL chondroitinase

ABC retained a macromolecular composition closer to the native tissue (Figure 2E-F). However,

RT q-PCR revealed no profound changes in the expression of MMP3 TIMP1 and ECM

105

macromolecules. Chondroitinase ABC downregulated the expression of collagen 2 and

aggrecan, but the expression of collagen 1 remained unchanged (Figure 3A-C). Genes for other

matrix degrading enzymes (MMP1, MMP13) and interleukins (IL-6, IL-8, IL-1β) were not

detectable.

Figure 2. Chondroitinase ABC reduces proteoglycan content in the NP. (A) As expected, the glycosaminoglycan content significantly decreased in the NPs injected with 20 U/mL Chondroitinase ABC compared with the NPs injected with PBS (set as 1). (B) Collagens, (C) DNA content and (D) water content remained unchanged. Graphs show ECM composition relative to the PBS injection (mean ± SEM, n = 4 per group). (E) Alcian blue/Picrosirius red and (F) Safranin O staining confirmed that 1 U/mL chondroitinase ABC did not have major proteoglycan-degrading effects, whereas 20 U/mL reduced GAG content at all tested time points. Scale bar 100 µm. *p < 0.05 vs. PBS injection, #p < 0.05 vs. each other (ANOVA).

106

Figure 3. Chondroitinase ABC does not activate cell reactions in the NP. (A-C) Injection of 1 - 20 U/mL chondroitinase ABC reduced gene expression of aggrecan and collagen 2 in the NP cultures on day 3, 7 and 14. Gene expression of MMP3 increased on day 3, but was downregulated at later time points. Gene expression of TIMP1 and collagen 1 remained close to the control levels. Inflammatory and catabolic markers were not detectable. Graphs show fold change relative to the PBS injection (mean ± SEM, n = 4). *p < 0.05 vs. PBS injection, #p < 0.05 vs. each other (ANOVA).

IL-1β and TNF-α induce gene expression of inflammatory and pain-related markers in the NP

Cytokines IL-1β and TNF-α are known activators of inflammatory and catabolic cascades in

the IVD and their expression may affect ECM composition at later stages. However, during 14

days of culture, IL-1β and TNF-α did not influence GAG, collagen, DNA and water content in the

NPs (Figure 4A-D). On day 3 and 7, cytokine-treated NPs expressed low levels of MMP3, TIMP1

107

and ECM macromolecules, similarly to the stable and enzyme-treated NP cultures (Figure 4E).

Cytokines significantly induced the expression of pro-inflammatory interleukins (IL-6, IL-8),

collagenases (MMP1, MMP3, MMP13), their inhibitor TIMP1, and pain-related mediators (iNOS,

PTGS2) on day 14 (Figure 4F). These results indicate that a catabolic/inflammatory shift towards

a degenerative phenotype developed between day 7 and day 14. Changes in the expression of

the ECM macromolecules may suggest remodeling attempts in response to the inflammatory

stimuli. Average Cq ± SD values for all genes in all treatment groups are shown in

Supplementary table 1.

Figure 4. IL-1β and TNF-α induce an inflammatory and catabolic phenotype in the NPs. (A) Glycosaminoglycan, (B) collagen, (C) DNA and (D) water content in cytokine-treated NPs remained unchanged over a period of 14 days, when compared with the untreated group (set as 1). Graphs show ECM composition relative to untreated samples (mean ± SEM from n = 4). (E) Inflammatory and catabolic gene expression was not detectable on day 3 and 7, and the expression of MMP3, TIMP1 and ECM macromolecules were close to the control values (fold change). (F) On day 14, IL-1β and TNF-α significantly induced gene expression of interleukins (IL-6, IL-8), collagenases (MMP1, MMP3, MMP13) and their tissue inhibitor (TIMP1), aggrecanase 1 (ADAMTS4) and pain mediators (PTGS2, iNOS). Graphs show fold change relative to the untreated samples (mean ± SEM from n = 4-7). *p < 0.05 vs. untreated control, #p < 0.05 vs. each other (ANOVA).

Epigallocatechin 3-gallate ameliorates the inflammatory and catabolic phenotype in the NP

As a significant up-regulation of inflammatory and catabolic genes was detected in the IL-

1β and TNF-α-treated NP cultures on day 14, this experimental group was subjected to EGCG

treatment. Culture media was supplemented with EGCG (10 µM) on day 0 and day 7. Neither

EGCG alone nor in combination with IL-1β and TNF-α significantly influenced the composition of

the ECM (Figure 5A-D) or the gene expression of ECM macromolecules, but EGCG inhibited the

expression of inflammatory markers (IL-6, IL-8), catabolic enzymes (MMP1, MMP3), and pain

108

mediators (iNOS, PTGS2). This not only provided evidence on the activity of EGCG in the NP

tissue, but also proved the suitability of the inflammatory NP cultures for evaluating the effects

of potential molecular therapeutics (Figure 5E).

Figure 5. Epigallocatechin 3-gallate ameliorates inflammatory and catabolic responses in the NPs. (A) Glycosaminoglycan, (B) collagen, (C) DNA and (D) water content in the NPs treated with EGCG, IL-1β and TNF-α, and their combination remained unchanged over a period of 14 days. Graphs show ECM composition relative to the untreated samples (mean ± SEM from n = 4). (E) On day 14, EGCG significantly reduced the expression of interleukins (IL-6, IL-8), collagenases (MMP1, MMP3) and pain mediators (PTGS2, iNOS) in cytokine-treated NPs. Gene expression of ECM components remained close to the control values. Graphs show mRNA expression relative to the cytokine-stimulated samples (mean ± SEM from n = 4-7). *p < 0.05 vs. IL-1β and TNF-α treatment (t-test).

5.5 Discussion

Throughout a lifetime, lumbar intervertebral discs undergo a morphological and functional

decline. Early signs of disc degeneration, caused by impaired exchange of solutes and activated

catabolic reactions, occur in the NP [110]. In some individuals, NP degeneration leads to internal

disc disruption and annulus fibrosus damage, which can result in low back pain. Although

discogenic back pain constitutes a rising socioeconomic burden for western countries,

therapies targeting its biological causes are lacking. The aim of this study was to develop an

organ culture model, which will allow preclinical screening of novel molecular therapeutics for

the degenerative nucleus pulposus.

109

A previously validated system that employs bovine caudal NPs cultured in Dyneema jackets

(“artificial AF”) was used [30]. Although bovine caudal discs are not weight bearing, they have

been proposed as a suitable biological model for human disc degeneration, due to their close

size and similar in vivo pressure (0.1–0.3 MPa) [321]. The bovine NP culture potentially represents

a space-efficient and relatively inexpensive approach to screen for molecular therapeutics that

can be used in human medicine. Bovine and human NP cultures were used for investigating the

effects of ECM anabolics such as LinkN, notochordal cell-conditioned medium [148] and BMP-7

[322]. However, the stable bovine NP culture exhibits neither an inflammatory phenotype nor

ECM degradation. Inducing biologically relevant degenerative changes in this system will

provide a platform for testing compounds that not only induce ECM production, but also

interfere with inflammation, catabolism and pain on the cellular level.

Although validated, NP cultures may not completely preserve the native tissue

characteristics, as a decrease in GAG content and an increase in gene expression of MMPs have

been described, especially at later time points [30]. Therefore, our initial experiments aimed to

test NP culture stability over a period of 21 days, to evaluate its relevance for development

towards a degenerative phenotype. Similarly to the previously published study, we observed an

initial downregulation of ECM genes and a transient increase in the expression of MMP3, likely

caused by a lack of mechanical loading. As stabilization of these parameters followed, the NP

culture was considered suitable for optimization towards a degenerative model.

ECM degrading enzymes that are commonly used to induce proteoglycan depletion in the

NP are the protein hydrolases trypsin and papain. However, as there is no naturally occurring

proteolysis in the disc, these enzymes do not activate cell reactions such as the expression of

interleukins and MMPs, but induce excessive matrix degradation, with a low physiological

fidelity. On the other hand, enzymes that degrade extracellular matrix in a biologically relevant

manner can possibly induce cellular responses that are found during disc degeneration.

Recently MMP3, ADAMTS4, and HTRA1 were injected into bovine discs and cultured for 8 days.

Even though these enzymes did not affect matrix degradation and gene expression, they

reduced cellular metabolic activity. HTRA1 additionally caused a loss of disc height and was able

to induce MMP expression in vitro [323, 324]. In our study, chondroitinase ABC with lyase

activity was employed to facilitate GAG depletion. Chondroitinase ABC degrades

polysaccharides, reproducing age-related removal of chondroitin sulfate from aggrecan [49].

However, although injection of chondroitinase ABC accelerated degradation of GAGs in a

biologically relevant manner, it was not sufficient for activation of inflammation and

catabolism in the NP tissue. This suggests that ECM degradation is a consequence, rather than

a cause of degenerative disc disease and confirms previous observations that injection of the

ECM degrading enzyme alone does not entirely mimic the degenerative process. Therefore,

110

chondroitinase ABC-injection into a NP culture is not entirely suitable when the aim is to create

a tissue culture system for testing bio-active compounds.

The involvement of pro-inflammatory cytokines in painful degenerative disc disease is well

established [2]. TNF-α and IL-1β can act as nociceptive triggers as well as induce the expression of

other potentially nociceptive interleukins (such as IL-6, IL-8, IL-12, IL-17) and direct pain mediators

(nitric oxide, prostaglandin E, substance P) [66, 80, 102-104, 106]. Degenerated discs with newly

formed nociceptive nerve fibers can thus be a primary source of back pain [61]. Moreover, disc

degeneration also can result in compression of nerve roots and spinal nerves. IL-1β, TNF-α and

other pain-mediating chemicals released from the NP can irritate dorsal root ganglions,

contributing to functional and structural nerve damage, radiculopathy and muscle weakness in

vivo [55, 66, 122]. Replicating such an inflammatory phenotype in organ cultures is therefore

needed for a successful drug discovery process.

In the study of Purmessur et al. (2013), TNF-α (200 ng/mL) activated the expression of

inflammatory and catabolic genes and accelerated degradation of aggrecan in bovine explants

[28]. Similarly, in another study, IL-1β up-regulated pro-inflammatory markers (IL-6, IL-8,

prostaglandin E2) and MMPs in punctured bovine discs, compared with disc organ cultures

treated with needle puncture alone [29]. In our study, a combination of IL-1β and TNF-α (100

ng/mL each) strongly induced the expression of interleukins (IL-6, IL-8), collagenases (MMP1,

MMP3, MMP13), aggrecanase (ADAMTS4) and molecules associated with pain sensation (iNOS,

PTGS2). The inflammatory and catabolic phenotype in our NP cultures fully developed on day 14,

possibly due to the dense proteoglycan network in initially healthy bovine NP tissue with lower

diffusivity, which can slow down cytokine penetration [109]. In smaller rat lumbar discs (± 0.5

cm in diameter), TNF-α and IL-1β activated the expression of MMPs and ECM changes already

after 3 days [50], most probably due to shorter diffusion distances [321]. Mechanical loading of

the bovine NP cultures in future studies may aid to promote cytokine penetration and activate

responses at earlier time points.

IL-1β and TNF-α-induced inflammation did not affect composition of the ECM over 14 days.

It has been shown before that healthy disc cells respond to stress by upregualtion of matrix

metalloproteinase inhibitors (TIMPs) [325], which inhibit the activity of MMPs in a 1:1

stoichiometry. In particular, TIMP-1 forms a complex with the catalytic domain of MMP-3,

reducing its activity in the NP tissue [323]. Therefore, the inflammation-related loss of ECM in

our model is expected to occur at later time points. Although ECM degradation has not been

observed in the cytokine-treated group, the current inflammatory bovine NP model can be used

to screen therapeutics against early stages of disc degeneration, before the loss of ECM occurs.

Another advantage of our bovine NP culture model is that molecular diffusion distances are

relevant to human NPs [321]. An interesting option would be to combine chondroitinase ABC

injection with cytokine treatment. Biologically relevant proteoglycan depletion, accompanied

111

by inflammatory and catabolic cell reactions, could accurately reproduce more advanced stages

of NP degeneration. Possibilities to modulate various stages of degeneration in the NP cultures

are currently being investigated in our group.

In the past, we have identified the polyphenol epigallocatechin gallate (EGCG) as a

promising compound for the treatment of painful degenerative disc disease. EGCG inhibited

the expression of inflammatory markers (IL-6, IL-8, PTGS2, TLR2, iNOS), collagenases (MMP1,

MMP3, MMP13) and NGF in human IVD cells cultured adherently and in alginate beads [18].

EGCG also protected IVD cells from oxidative stress caused by exogenous radicals [311] and

moreover, locally delivered EGCG reduced radicular pain in a rat model of NP herniation [18].

Therefore, we selected this compound for validation of the new inflammation-induced NP

culture model of degeneration. EGCG treatment of NP cultures exposed to IL-1β and TNF-α

reduced the expression of interleukins, MMPs, iNOS and PTGS2, which confirmed our results

from previous in vitro tests. In addition to in vitro studies, the NP culture system revealed no

negative effects of EGCG on the extracellular matrix, further supporting future research on this

compound. As EGCG ameliorated the expression of inflammatory markers in the tissue, the NP

culture system proved suitable for preclinical testing of bio-active compounds and their

combinations as well as delivery systems. Although the effects of EGCG have been shown

before in cell culture studies, it is necessary to include testing on the tissue level. While cell

cultures are suitable to evaluate mechanism of action, they are less relevant to determine

overall efficacy of compounds, due to the possible effects on the extracellular matrix. As human

NP contains less than 1% cells [60], an organ culture study should be performed in order to

assess the clinical relevance of EGCG. To our knowledge, this is the first study presenting the

effects of EGCG on inflammatory intervertebral disc tissue.

In comparison with animal models, organ culture systems represent a cost-effective

alternative for discovery and testing of novel therapeutic compounds. Organ cultures can

bridge basic science with translational medicine and allow in vitro experiments while

maintaining cells in their native tissue [23]. NP cultures confined in Dyneema jackets exhibit

rather stable characteristics over sustained periods of time. Moreover, accessibility and low

costs of bovine tissue, as well as the possibility to modulate the phenotype from early to

advanced NP degeneration, make this model attractive. Although more research is needed, we

conclude that these advanced NP cultures can be suitable for the screening of novel

regenerative compounds and for revealing their mechanisms of action in the NP tissue.

112

5.6 Conclusions

Organ culture models are valuable tools for studying therapeutic strategies against

degenerative disc disease. However, it is challenging to develop a cost-effective organ culture

model that contains all main hallmarks of degenerative disc disease and is at the same time

biologically relevant. In this study, an inflammatory NP culture model was established and

validated with the anti-inflammatory compound EGCG. This NP culture model expresses

inflammatory, catabolic, and pain-related genes that are involved in the progression of human

degenerative disc disease. As composition and molecular diffusion distances of the bovine NP

are comparable to the human NP, obtained results are applicable to human. Moreover, this

model is cost-effective and allows high throughput tests for molecular therapeutics. ECM

anabolic agents and compound combinations can be tested in the NP cultures, in which TNF-α

and IL-1β-induced inflammation will be combined with Chondroitinase ABC-induced loss of

GAG.

5.7 Author’s contributions

OK performed experiments (NP cultures, ECM analyses, RNA isolations, RT-qPCRs), data

analysis, helped with the study design, and drafted the manuscript. MH performed

experiments (NP cultures, ECM analyses, RNA isolations and RT-qPCRs) and helped with the

study design. JAT performed NP culture experiments and RNA isolations. JZ and DK performed

NP culture experiments and ECM analyses. SJF conceived funding, helped to draft the

manuscript and coordinated the study. KW conceived funding, designed and coordinated the

study and helped to draft the manuscript. KI introduced the NP culture systems and helped to

conceive the study aims. All authors approved a final version of manuscript.

5.8 Acknowledgement

Funding for this research project was provided by Sciex (Project 14.145) and the Mäxi

Foundation. The authors would like to thank Ms. Ladina Ettinger-Ferguson for histological

analysis and to acknowledge Simon Zollinger for the images of the bovine NP cultures. The

authors wish to confirm that there are no known conflicts of interest associated with this

publication.

113

5.9 Supplement Supplementary table 1. Average Cq ± SD values for all expressed genes in inflammation group.

Condition Gene mean Cq ± SD Condition Gene mean Cq ± SD Control S18 27.49 ± 5.12 IL-1β S18 27.61 ± 5.19 GAPDH 27.74 ± 0.95 GAPDH 27.83 ± 1.17 Collagen 1 43.46 ± 3.43 Collagen 1 43.38 ± 3.43

Collagen 2 34.13 ± 3.29 Collagen 2 38.01 ± 3.55

Aggrecan 36.14 ± 3.53 Aggrecan 35.59 ± 4.35 MMP1 41.37 ± 8.87 MMP1 32.51 ± 12.11 MMP3 43.61 ± 2.91 MMP3 32.47 ± 8.21 MMP13 37.50 ± 10.18 MMP13 33.92 ± 5.44 TIMP1 35.71 ± 9.91 TIMP1 31.41 ± 2.94 IL-6 45 ± 0 IL-6 36.35 ± 4.49 IL-8 45 ± 0 IL-8 38.76 ± 4.12 PTGS2 43.13 ± 3.23 PTGS2 37.21 ± 2.21 ADAMPTS4 45 ± 0 ADAMPTS4 43.57 ± 2.85 iNOS 45 ± 0 iNOS 35.31 ± 1.54 Condition Gene mean Cq ± SD Condition Gene mean Cq ± SD EGCG S18 27.69 ± 4.49 IL-1β + EGCG S18 27.61 ± 5.25 GAPDH 27.91 ± 0.45 GAPDH 27.84 ± 1.14 Collagen 1 45 ± 0 Collagen 1 45 ± 0 Collagen 2 36.62 ± 5.01 Collagen 2 37.86 ± 4.09 Aggrecan 36.31 ± 4.43 Aggrecan 37.01 ± 3.71 MMP1 39.28 ± 11.15 MMP1 37.95 ± 8.34 MMP3 40.79 ± 5.05 MMP3 35.82 ± 7.13 MMP13 37.84 ± 8.86 MMP13 35.23 ± 8.58 TIMP1 36.31 ±7.25 TIMP1 32.83 ± 5.05 IL-6 41.62 ± 4.65 IL-6 38.06 ± 3.95 IL-8 44.41 ± 1.45 IL-8 42.25 ± 4.35 PTGS2 41.67 ± 3.83 PTGS2 38.94 ± 2.61 ADAMPTS4 42.76 ± 4.46 ADAMPTS4 43.33 ± 3.32 iNOS 42.77 ± 4.44 iNOS 38.31 ± 4.57

115

Chapter 6

Stability of (–)-epigallocatechin gallate and its

activity in liquid formulations and delivery

systems

Olga Krupkova, Stephen John Ferguson, Karin Wuertz-Kozak

The Journal of Nutritional Biochemistry, 2016, Nov(37): 1–12.

117

6.1 Abstract

(–)-Epigallocatechin gallate (EGCG) has become a popular disease-preventive

supplement worldwide because it may aid in slowing down the onset of age-related diseases

such as cancer, diabetes, and tissue degeneration. As largely demonstrated in cell culture

studies, EGCG possesses antioxidant properties and exhibits favorable effects on gene

expression, signal transduction and other cell functions. However, only limited effects have

been observed in experimental animals and human epidemiological studies. The inconsistency

between the biological activity of EGCG in cell cultures and in vivo can be attributed to its low

stability, which not only decreases its bioavailability, but also leads to the formation of

degradation products and pro-oxidant molecules with possible side-effects.

Understanding EGCG degradation kinetics in solution and in vivo is crucial for its

successful clinical application. Ambient conditions (pH, temperature, oxygen) can either

enhance or decrease the stability of EGCG, thus influencing its biological activity. Usage of

stabilizers and/or encapsulation of EGCG into particulate systems such as nanoparticles or

microparticles can significantly increase its stability. In this review, the effects of ambient

conditions, stabilizers and encapsulation systems on EGCG stability, activity and degradation

rate are illustrated.

Graphical abstract

Key words

Epigallocatechin gallate, stability, activity, degradation, encapsulation

118

6.2 Biological effects of (–)-epigallocatechin gallate

Natural polyphenols, including catechins, have received considerable attention as

potential therapeutics against lifestyle diseases. In the past decade, a substantial contribution

to understanding various biological activities of (–)-epigallocatechin gallate (EGCG) in cell

cultures and in vivo has been made [218, 326].

EGCG (Mw = 458.372 g/mol) belongs to the family of catechins, which are polyphenols

found in tea (Camelia sinensis) and other plants as secondary metabolites. The principal

epicatechins in fresh tea leaves apart from EGCG are (–)-epicatechin (EC), (–)-epigallocatechin

(EGC), and (–)-epicatechin gallate (ECG). Their less active trans isoforms catechin, gallocatechin,

catechin gallate and gallocatechin gallate are present in lower amounts. Dry green tea contains

approximately 10% catechins, i.e. about 8-15 g per 100 g of dry leaves [327]. Classically, 4-5 g of

dry tea leaves is used for the preparation of a tea cup infusion, hence 400-500 mg catechins

can be ingested per cup of tea (300-400 mL) [327, 328]. However, only 0.1-1.1% of the orally

administered dose reaches the systemic circulation [329, 330]. EGCG is the most abundant and

active tea catechin. Structurally, EGCG has hydroxyl groups at carbons 3', 4', and 5' of the B ring

(3'4'5'-OH), and a gallate moiety esterified at carbon 3 of the C ring [326, 331], which enables

EGCG to exhibit antioxidant properties. Nevertheless, the chemical structure of EGCG makes it

susceptible to degradation via auto-oxidation [31].

EGCG has been traditionally regarded as beneficial because of its antioxidant effects,

which may aid in a number of clinical conditions such as cancer, obesity, atherosclerosis,

diabetes and neurodegeneration. These age-related diseases are collectively characterized by

increased production of reactive oxygen species (ROS) and/or insufficient cellular anti-oxidant

capacity [218, 332-336]. Although balanced cellular ROS modulate a variety of physiological

events including proliferation, differentiation, host defense and wound healing [271], they are

highly reactive. Extensive ROS exposure causes deleterious damage of proteins, lipids and DNA,

which leads to premature senescence and cell death [38]. The antioxidant activity of EGCG does

not only involve direct trapping of ROS [326, 337-342], but also inhibition of ROS production via

interaction with anti- and pro-oxidant proteins [296, 343-345], and chelation of potentially pro-

oxidant metal ions [31, 326, 346-349].

Several structures with the capability of donating hydrogen atoms and electrons appear

to be important in radical scavenging activity of EGCG: the trihydroxyl group in the B ring

(3'4'5'-OH) participates in electron delocalization and stabilization [339-341], the gallate moiety

at the 3 position in the C ring is associated with increased ROS and reactive nitrogen species

(RNS) scavenging properties [326] and the A ring was identified as an antioxidant site as well

[337, 338, 342].

119

EGCG also has the ability to counteract the formation of ROS by inhibiting ROS-

generating enzymes (e.g. xanthine oxidase, cyclooxygenase and lipoxygenase [296, 343-345])

and to affect the production of nitric oxide via an interaction with nitric oxide synthase (NOS)

[350]. Importantly, EGCG may be able to activate superoxide dismutase expression, hence

providing yet another mechanism to inhibit ROS and to protect mitochondria [300, 301].

The described ability of catechins to chelate metal ions may contribute to their

antioxidant activity, specifically by preventing redox-active transition metals from catalyzing

free radical formation, e.g. in degenerated nerve tissue [346]. Chelating groups in EGCG are the

3,4-dihydroxyl groups on the B ring as well the gallate group, both of which are able to bind

several metals bearing strong positive charges (Fe3+, Al3+ and Cu2+) [326, 347, 348]. EGCG can

inactivate iron ions, therefore suppressing the superoxide-driven Fenton reaction, inhibiting

H2O2 formation and lipid peroxidation, which is believed to be a further source of harmful ROS

in degenerative diseases [349].

The ability of EGCG to inhibit free radicals and other reactive species in vitro is well

established. However, it is also known that its molecular structure allows fast degradation,

which complicates EGCG’s clinical application [52, 53]. Encapsulation methods, usually used to

increase stability of natural polyphenols [31, 51], can protect EGCG only for a short period of

time. Understanding conditions influencing EGCG stability and degradation can increase the

potential success of its novel therapeutic formulations. Detailed information on findings about

EGCG stability in biological systems and solution is provided in this review.

Pharmacokinetics

The most common route of administration is oral, as tea infusions or pills, but it is

known that orally ingested EGCG undergoes unfavorable metabolic changes in the

gastrointestinal (GI) tract and liver. Enzymatic transformation reactions start already in the

saliva, where EGCG is hydrolyzed by esterases [326]. As catechins are metabolized

predominantly in the intestine and liver, where glucuronidation and sulfation of the hydroxyl

groups and O-methylation of the catechol groups through UDP-glucuronosyltransferase (UGT),

phenolsulfotransferase (SULT) and catechol-O-methyltransferase (COMT) takes place [330, 351],

the free catechins as well as O-methylated and both O-methylated and glucuronidated

conjugates can be found in plasma, urine, and bile [352-354]. Conjugates still contain intact

catechol and gallate moieties and can have similar biological activity as their parent

compounds [334]. Catechins that are not absorbed in the small intestine and conjugated

catechins excreted in bile reach the colon where they can be metabolized by colonic bacteria

and reabsorbed [334, 354].

Interestingly, although EGCG seems not to be significantly affected by P450

oxidases/reductases, it belongs to the group of compounds that can inhibit or activate P450

120

enzymes hence modulating pharmacokinetics of co-administered drugs. For this reason,

caution is needed especially regarding the co-administration of drugs with small therapeutic

windows and drugs with considerable side-effects [224, 225, 355].

Cellular uptake of catechins occurs mainly by passive transport. The amount of catechin

incorporated is related to its affinity for biomembranes, which increases with catechin

hydrophobicity (i.e. uptake of EC/EGC < ECG/EGCG). Therefore the interaction of catechins with

lipid bilayers may partly govern their biological activity [356]. Intracellular amount of EGCG and

its metabolites is regulated by ATP-dependent efflux pumps with a broad substrate specificity,

by multidrug resistance-associated protein (MRP) efflux pumps [357], and by P-glycoprotein (P-

gp) [358], which actively pump EGCG to the extracellular or intestinal space and so decrease its

bioavailability.

Despite the limited bioavailability of catechins, which is attributed to their low stability

[331] and poor absorption within the GI tract [330, 359], elevated plasma levels of catechins

have consistently been found in humans following the consumption of green tea. Plasma

catechin concentration peaks 1 to 4 hours after ingestion of green tea or catechin supplements

and gradually decreases back to baseline within 24 hours [259, 329, 334, 352, 360]. After a single

dose of green tea or green tea extract, the concentration of individual catechins in rodent or

human plasma increase up to 1 μM [330, 361], but EGCG levels can be much higher in the

tissues that are in direct contact, such as oral cavity [362] and GI tract [334, 363, 364]. Catechin

metabolism is extensively reviewed in [326, 351].

Biological activity in cell cultures

EGCG not only possesses antioxidant activity as described above, but also exhibits

multiple other beneficial functions, which are independent from the antioxidant reactions and

have been demonstrated in cell culture studies.

As an active anti-inflammatory compound, EGCG inhibits phosphorylation of p38, JNK

and nuclear translocation of NF-kB, and therefore negatively regulates the expression of pro-

inflammatory genes and proteins in various cell types [34, 365-367]. Furthermore, EGCG reduces

accumulation of malfunctioning and abnormal proteins and other waste products during

cellular aging [290, 291, 368]. EGCG may also alleviate the detrimental effects of diabetes, as it

was shown to improve insulin secretory function and viability of β-cells under glucotoxicity, and

insulin resistance in skeletal muscle cells [369, 370]. Diverse EGCG activities are reviewed in

[371-374] and summarized in Figure 1.

Cells with features of genetic instability are affected by EGCG in a different way. It has

been shown that EGCG exhibits diverse toxic effects in cancer cell lines with deregulated

signaling, and therefore can contribute to the treatment of e.g. skin, breast and lung cancers or

tumors of GI tract, either alone or in combination with other anti-cancer compounds [375-377].

121

EGCG also interacts with proteins and phospholipids in the plasma membrane, influencing

membrane receptor activity and consequent cellular responses, which can be particularly

useful in cancer chemoprevention [48, 304, 378].

Although bioactivity of EGCG has been widely linked to its antioxidant effects and

interactions with other molecules, increasing evidence suggests that EGCG can also function as

a pro-oxidant, hence promoting oxidative stress [326, 334, 338, 364, 379-382]. After contact with

components in cell culture media, EGCG can be oxidized to form semiquinone radicals,

superoxide, and hydrogen peroxide [342, 364, 380, 381, 383]. H2O2 generation, followed by

effects that can be abolished by antioxidant enzymes, was described particularly at higher

EGCG concentrations in vitro (˃ 50 µM) [384]. EGCG-mediated H2O2 generation is lower in the

presence of cells [288, 364, 380, 385], due to the action of cellular glutathione peroxidase and

catalase [288, 381].

Figure 1. Reported biological activity of EGCG in cell cultures. Oxidation of EGCG molecule leads to generation of various pro-oxidants with unfavorable effects. EGCG can exhibit antioxidant activity via several mechanisms, such as neutralization of reactive oxygen species (ROS), chelation of metals or modulation of ROS-regulating enzymes. EGCG also possesses redox-independent activity on multiple stages of cellular signal transduction and gene expression.

122

Biological activity in vivo

Depending on the cell type, the therapeutically active, non-toxic concentration of EGCG

in vitro lies between 1 and 100 µM, but certain toxic effects may start to occur already at 50 µM.

Whether EGCG induces oxidative stress in vivo is however controversial as plasma proteins and

antioxidants can enhance its stability [386]. The anti-inflammatory, antioxidant and other

health-protective effects of EGCG have been continuously reported in animal models [387-390],

but in vivo studies describing its toxicity emerged as well [53].

The discrepancy can be explained by the fact that plasma antioxidants may not be

sufficient to balance the pro-oxidative effect of EGCG when applied in high doses. Cases of

hepatotoxicity and nephrotoxicity upon consumption of high doses of green tea-containing

dietary supplements and preparations (10-29 mg/kg/day) have been reported in experimental

animals and humans [53, 381, 391-393]. A single oral dose of 1500 mg/kg EGCG in CF-1 mice

increased plasma alanine aminotransferase and reduced survival by 85%. High-dose EGCG-

activated hepatotoxicity was associated with oxidative stress, including increased hepatic lipid

peroxidation, plasma 8-isoprostane, hepatic metallothionein and γ-histone 2AX protein

expression. EGCG also increased plasma interleukin-6 and monocyte chemoattractant protein

1. Long-term consumption of 1% EGCG in a diet for 6 weeks evoked elevated splenocyte and

macrophage inflammatory markers such as tumor necrosis factor-α, interleukin-6, interleukin-

1β, and prostaglandin E2, and disturbed immune cell populations in mice [394]. A single

intraperitoneal (IP) administration of 100 mg/kg EGCG into mice has generated hepatotoxicity,

whereas 150 mg/kg resulted in 100% mortality within 24 hours [392].

Animal studies unfortunately do not allow full extrapolation of a safe EGCG dose and

mode of application to humans. Differences in pharmacokinetics and bioavailability of EGCG

caused by different enzyme activities have been described between species, namely mice, rats,

and humans. As a result, different metabolites are produced: in mouse plasma, mostly

conjugates were identified, while in human plasma mainly free catechins were present after

oral administration. Results of epidemiological and animal studies involving EGCG are reviewed

in [326, 395].

6.3 Degradation of (–)-epigallocatechin gallate

Catechins can undergo significant degradation, not only in biological fluids, but also

during tea processing and storage, leading to the loss of their health-promoting effects.

However, some biological effects have been directly attributed to EGCG degradation products,

rather than to EGCG itself [380]. The instability of EGCG is caused by two major reactions,

epimerization and auto-oxidation [31, 396]. Rates of these two reactions are affected by factors

123

like the presence of oxygen, antioxidants and EGCG concentration. Therefore, the stability and

biological activity of EGCG can be controlled by the physicochemical conditions of ambient

media.

Figure 2. Scheme of EGCG degradation pathways in various ambient conditions. In aqueous media EGCG can undergo epimerization, yielding (–)-gallocatechin gallate (GCG) with similar biological activities, or degrade via auto-oxidation, yielding dimers such as thenasinensin A that can further oligomerize. Epimerization is favored in mild acidic pH, above 40°C and in the presence of antioxidants. Auto-oxidation, which occurs at mild alkaline pH and temperatures below 40°C, is unfavorable as various radical intermediates can be formed.

Auto-oxidation

In solution, catechins rapidly degrade via auto-oxidation, involving a loss of hydrogen

atoms, generation of semiquinone radical intermediates, superoxide and formation of quinone

oxidized products, which may possibly have damaging consequences for cells and tissues [338,

351, 379, 397-400]. Subsequently, dimers theasinensin A and compound P2 can be identified as

major products of B-ring auto-oxidation in mild alkaline fluids with micromolar EGCG

concentrations, which is typical for cell culture. EGCG degradation is usually accompanied by

the color change of its aqueous solution from transparent to brown, indicating the presence of

structures with larger molecular weight, via polymerization of EGCG dimers [396]. In contrast

to auto-oxidation in aqueous media, which involves the B-ring, oxidation of EGCG in organic

solvents rather takes place at the D-ring [401].

124

Epimerization

In millimolar EGCG concentrations and when auto-oxidation is prevented by

antioxidants, epimerization of EGCG to its trans epimer (–)-gallocatechin gallate (GCG) is

preferred [327, 380, 385, 396, 402, 403] (Figure 2). The epimerization of EGCG to GCG is more

favorable at pH below 5.5, at high concentrations (mM solution) or at temperatures above 50°C,

e.g. during autoclaving [396, 402, 403]. The conversion of EGCG to GCG occurs also under

anaerobic conditions and once EGCG is stabilized by the presence of antioxidants [380].

Catechin trans-epimers do not have toxic effects, but rather possess similar biological activities

to their cis- counterparts [339, 340, 404, 405]. Catechin epimerization can be reversible [402,

403].

Table 1. The stability of EGCG in aqueous solution. The initial concentrations of EGCG, its mean half-life, and its degradation products in various solvents are listed. GTC = green tea catechins, RT = room temperature, na = information not available. EGCG, 37 °C [328, 380, 385, 396] solvent additive initial concentration mean half-life (t1/2)

degradation product

McCoy’s 5a - 10 µM 30 minutes theasinensin A product 2

McCoy’s 5a HT29 cells 10 µM 130 minutes theasinensin A product 2

Ham’s F12 - 10 µM ˃ 30 minutes theasinensin A product 2

HBSS - 10 µM 30 minutes theasinensin A product 2

Na2HPO4 (50 mM)

- 20 µM 30 minutes theasinensin A product 2

Ham’s F12:RPMI (1:1)

- 20 µM 20 minutes theasinensin A product 2


KYSE 150 cells 20 µM 1 hour theasinensin A H2O2

PBS (0.05 M) - 218 µM ˃ 40 minutes na water (100 °C) - 1.09 mM ˃ 4 hours na PBS (60 mM, pH 7.4)

- 1.09 mM 1.5 hours na

EGCG, 50 °C [406] solvent additive initial concentration t = concentration

decreased to 90% degradation product

Glycerin - 10.9 mM 24.4 days na Transcutol P - 10.9 mM 22.8 days na

6.4 Stability of (–)-epigallocatechin gallate in solution

Since EGCG is prone to fast degradation, knowledge of the degradation kinetics is

necessary to predict its stability during thermal processing, storage as well as during

preparation and application of therapeutic formulations and food supplements [407] (Table 1).

Stability of EGCG is affected by parameters such as oxygen concentration [396, 407], pH [342,

125

407-409], temperature [327, 408, 409], and ionic strength [409]. Stability of EGCG in aqueous

media depends also on its initial concentration [396, 409] and the presence of antioxidants,

which can prevent its auto-oxidation [385, 396, 409]. EGCG stability can be accurately

measured by high-performance liquid chromatography (HPLC) [409] and by various assays

using spectrophotometry [410].

Effect of concentration

The stability of EGCG in aqueous solution depends on the initial concentration [396,

400, 403, 409], with a clear difference between the micromolar and millimolar range

(Supplementary Table 1). The half-life of EGCG in water slightly increased from 30 minutes at 20

µM to 150 minutes at 96 µM [396]. After 2 days of storage at room temperature (RT), 54.5 µM

EGCG completely degraded, whereas 1.97 mM EGCG degraded only to 80% of its initial

concentration [409]. After 7 days of storage, 0.5 mM EGCG decreased to 80%, while 7 mM EGCG

solution remained stable [396].

Major degradation products of EGCG in micromolar concentrations are EGCG dimers. In

higher millimolar concentrations, however, dimers do not form, but EGCG epimerizes to GCG

[396, 403], suggesting that the catechin molecules are able to protect each other from auto-

oxidation [400].

The safe and active in vitro concentration of EGCG varies with cell type, but generally it

is less than 100 µM. Oral route of administration is highly affected by the composition of diet,

fasting conditions, and enzyme activities, therefore the bioavailability of EGCG can differ

between individuals. The maximum EGCG concentration in human plasma after oral

consumption was reported as 1 µM, but higher concentration can be achieved via other

administration routes or encapsulation [411].

Effect of pH

Upon oral administration, EGCG is exposed to variable pH levels, ranging from 1.5 in the

stomach up to 8.5 in the duodenum, where most of the absorption takes place. EGCG is rather

stable at lower pH of 2.0−5.5. Its instability particularly in the alkaline environment of the GI

tract has been reported as one of the reasons for its poor bioavailability [327, 328, 385, 400, 407-

409]. It was shown that 109 mM EGCG solution at pH 9 completely degraded after 1 day,

whereas a solution at pH 3 degraded only to 90% of its initial concentration at day 28 [406].

The degradation rate of EGCG is comparable to that of green tea catechin (GTC) extract, which

showed a pH-dependent rate of degradation as well [328, 397]. Long-term stability of GTC (4.4

mM) was examined at pH 4−5 for 6 months, which is the time desired for a storage of drinks

and liquid formulations, but already after one month, 30−78% of GTC was destroyed. The pH of

126

solution also influenced the degradation rate of GTC during autoclaving: at pH 3, 95% of GTC

remained stable, whereas at pH 5 45% of GTC degraded [328] (Supplementary Table 2).

The acidic environment enhanced EGCG stability, but neutral and alkaline pH caused

auto-oxidation on the B-ring and produced theasinensin A and product P2 [385, 401].

Susceptibility to oxidation and formation of free radicals is caused by increased proton-

donating potential of EGCG in alkaline fluids [337, 406]. Aside from the stomach (pH ± 2.5),

acidic pH occurs also in various tumor tissues due to the cellular production of lactic acid and

ATP hydrolysis in hypoxic regions of tumors [412]. In contrast to the very low pH in the stomach,

the mildly acidic pH of tumors allows to prolong EGCG’s half-life, which may help to target

subpopulations of tumor cells that frequently escape conventional therapy [413]. A mildly acidic

pH is also present in degenerative tissues like cartilage [414] or intervertebral disc [206].

Effect of temperature

Temperature is a significant parameter governing EGCG stability and the mode of its

degradation (Supplementary Table 3). Low temperature can preserve the stability of EGCG: the

half-life of 109 mM EGCG solution at 4°C was 7 days, but decreased to 2 days when kept at RT

[406]. Similarly, upon storage at RT, 54.5 µM EGCG solution completely degraded by day 2,

while the same solution at 4°C degraded only to 60% of its initial concentration [409]. At 37°C,

10 µM EGCG solution degraded completely in 1 hour [385, 396], which was the most relevant

finding for cell culture experiments.

At temperatures below 50°C, EGCG underwent degradation via auto-oxidation while at

higher temperatures, epimerization was more profound [327, 396, 403, 406, 408, 409]. The

temperature 44°C was identified as crucial in aqueous system: below 44°C oxidation prevailed,

whereas above 44°C the epimerization from EGCG to GCG became prominent [415]. High

temperatures can be used to sterilize EGCG-containing formulations for long term storage.

After 20 minutes of autoclaving at 120°C, 75% of GTC with initial concentration 10.9 mM

remained in the solution and less than 12% of GCG was formed [327].

Effect of oxygen partial pressure and ionic strength

Oxygen partial pressure may have significant consequences for EGCG stability [327, 396,

407], as standard cell culture experiments are performed under the atmospheric pressure of

oxygen, which is 160 mmHg (21.1%). In the tissues, however, oxygen partial pressure drops to ˂

20 mmHg inside cells or within upper skin layer. Maximum tissue oxygen pressure can rise to

110 mmHg in alveoli and to 100 mmHg in arterial blood [416]. This shows that even the highest

values of tissue oxygen pressure are still lower than atmospheric oxygen pressure present

under the cell culture conditions, suggesting the degradation kinetics of EGCG may differ

between cell culture and in vivo situation. The majority of all the mentioned EGCG stability

127

studies were performed under atmospheric oxygen partial pressure. Apart from that, the effect

of oxygen on the stability of EGCG was tested in phosphate buffer (pH 7.4) in the absence or

presence of nitrogen. In air-saturated buffer, EGCG underwent auto-oxidation; oxygen was

consumed and H2O2 was formed [380, 398]. Under low oxygen partial pressure, generated by

extensive flushing of buffer solution with nitrogen gas before adding EGCG, EGCG was greatly

stabilized, with 95% remaining after 6 hours. No dimers were detected, but GCG epimers were

formed [327, 396]. Indeed, EGCG was shown to be rather stable in blood, where no EGCG dimers

were present, which may reflect a combined effect of low oxygen partial pressure and presence

of plasma antioxidants [380].

The activity of EGCG is also likely to be dependent on the salt concentration of ambient

medium [356]. As degradation and transformation of EGCG increased with rising ionic strength

of the solvent, EGCG aqueous solution in the lowest buffer capacity possible may be a way to

avoid accelerated degradation [328, 406, 409]. On the other hand, the incorporation of EGCG

into the lipid bilayers was accelerated as the salt concentration increased [356] (Supplementary

table 4).

6.5 Stabilization of (–)-epigallocatechin gallate for clinical

application

Antioxidants

The presence of reducing agent is essential for the stability of EGCG in aqueous solution

[409]. In buffer or cell culture media, such as McCoy’s 5A, Ham’s H-12 or RPMI, the half-life of

micromolar EGCG is about 30 minutes [380, 385, 396], but addition of antioxidants can

significantly stabilize EGCG and increase its half-life to more than 24 hours. Antioxidants

inhibit formation of EGCG dimers, so epimerization of EGCG to GCG prevails [380].

In the presence of superoxide dismutase (SOD), which catalyzes the reduction of

superoxide anions to hydrogen peroxide, no superoxide-mediated auto-oxidation took place,

but GCG was formed instead [396]. Ascorbic acid (AA) was shown to stabilize EGCG at various

conditions in aqueous buffers preventing auto-oxidation, because it was preferentially oxidized

in place of the catechin [327, 400, 409, 417]. However, in the presence of sucrose, AA greatly

destabilized GTC [328] (Table 2).

Amphiphilic compounds and other stabilizers

Amphiphilic compounds are used to partition the phases in an immiscible biphasic

system consisting of aqueous and organic solvents. Knowledge of the degradation kinetics of

EGCG in the presence of amphiphilic compounds is important e.g. for transdermal drug delivery

128

formulations, where organic permeation enhancers are used to promote drug migration

through the skin.

EGCG (10.9 mM) was tested in solutions of permeation enhancers propylene glycol,

polyethylene glycol 200, glycerin, and Transcutol P [406]. EGCG exhibited low stability in

propylene glycol and polyethylene glycol 200, whereas a higher stability was observed in

glycerin and Transcutol P, supported by the lack of brown discoloration for a period of one

month. Accelerated degradation studies conducted at 50°C revealed only a 10% decrease of the

total EGCG amount at day 23-24, showing a much higher stability in Transcutol P and glycerin

compared with aqueous media [406].

The effect of antioxidants propyl gallate (Pgal), AA, and butylated dihydroxytoluen (BHT)

on the stability of EGCG in glycerin and Transcutol P was also tested. In the presence of

amphiphilic compounds, AA did not have any significant stabilizing effect. Both Pgal and BHT

increased the stability of EGCG in glycerol, but decreased its stability in Transcutol P.

Combination of BHT and Pgal however significantly destabilized EGCG, most likely due to their

molecular interactions [406].

Additives and stabilizers in food, drinks and oral formulations can also influence the

stability of EGCG (Table 2). Several critical structural features of proteins for binding to

flavonoids, like the amino-acid composition and conformation, have been highlighted in the

literature [418]. For example milk, which is commonly added to the tea to reduce an astringent

taste, is able to complex flavonoids. Authors of several clinical studies on tea ingestions

concluded that milk might adversely affect tea catechin absorption [326]. Nevertheless,

interaction with milk proteins did not seem to impair catechin stability [419].

β-casein and gelatin were suggested as the most promising carriers for EGCG due to

their high EGCG affinity [418], but the stability of EGCG in the presence of β-casein was

improved only mildly [417]. A slight improvement in EGCG stability was reported also for

proline, aspartic acid and ovalbumin, whereas glutamic acid, β-cyclodextrin, trehalose, pectin

and alginic acid had no effect on EGCG stabilization [417]. In another study, a GTC mixture (1.09

mM) was added either to a sucrose solution (0.15 g/mL) or to a sucrose-citric acid solution (0.15

g sucrose/mL, 2 mg citric acid/mL), both of which are common additives in tea drinks. The

solution was autoclaved at 120°C for 20 minutes and analyzed monthly. A 6-months

assessment revealed that addition of sucrose had little effect on the stability of GTC, whereas

addition of citric acid destabilized GTC, which could explain the very low content of GTC in the

commercial tea drinks [328]. The stability of EGCG also increased in the presence of Na2EDTA (14

mM), which can scavenge metal ions involved in catalyzing EGCG auto-oxidation [396, 402]

129

Table 2. The effect of antioxidants and stabilizers on the stability of EGCG. The initial concentrations of EGCG, its mean half-life, and its degradation products in respective solvents with additives are listed. GTC = green tea catechins, GCG = (–)-gallocatechin gallate, SOD = superoxide dismutase, AA = ascorbic acid, Pgal = propyl gallate, BHT = butylated dihydroxytoluen, hl = half-life, na = information not available. EGCG, 37 °C, pH 7.4 [396] solvent additive initial concentration mean half-life (t1/2)

% at t = 6 hours (6h) degradation product

Na2HPO4 (50 mM)

- 20 µM t1/2: 30 minutes theasinensin A product 2

Na2HPO4 (50 mM)

saturated nitrogen

20 µM 6h: 100% GCG


- 20 µM t1/2: 20 minutes theasinensin product 2


SOD (5U/mL) 20 µM 6h: 80% GCG

GTC, RT, pH 7.4 [328] solvent additive initial concentration mean half-life (t1/2) degradation product water - 1.09 mM 6 months na water sucrose

(0.15 mg/mL) 1.09 mM 3 months na

water sucrose+citric acid (2 mg/mL)

1.09 mM 1 month na

water sucrose+AA (11 mg/mL)

1.09 mM 0.5 month na

EGCG, RT, pH 5.4 [396] solvent additive initial concentration mean half-life (t1/2) degradation product Na2HPO4 (50 mM)

- 20 µM 30 minutes na

Na2HPO4 (50 mM)

Na2EDTA (14 mM)

20 µM 4 hours na

EGCG 37 °C, pH 6.5 [417] solvent additive initial concentration concentration at t =

5 hours degradation product

PBS (0.05 M) - 218 µM 106.9 µM na PBS (0.05 M) AA (0.1 mg/mL) 218 µM 213.8 µM na PBS (0.05 M) proline (0.1

mg/mL) 218 µM 122.2 µM na

PBS (0.05 M) cysteine (0.1 mg/mL)

218 µM 202.9 µM na

PBS (0.05 M) aspartic acid (0.1 mg/mL)

218 µM 226.5 µM na

PBS (0.05 M) glutamic acid (0.1 mg/mL)

218 µM 220.0 µM na

PBS (0.05 M) casein (0.625 mg/mL)

218 µM 124.3 µM na

PBS (0.05 M) ovalbumin (0.625 mg/mL)

218 µM 133.1 µM na

PBS (0.05 M) b-cyclodextrin (1.25 mg/mL)

218 µM 111.3 µM na

PBS (0.05 M) trehalose (2.0 mg/mL)

218 µM 102.5 µM na

PBS (0.05 M) pectin (0.1 mg/ml) 218 µM 113.4 µM na PBS (0.05 M) alginic acid

(0.5 mg/mL) 218 µM 120.0 µM na

EGCG 37 °C, pH 6.5, t = 10h [417] solvent additive initial concentration concentration at t =

10 hours degradation product

PBS (0.05 M) - 218 µM 19.6 µM na PBS (0.05 M) AA (0.1 mg/mL) 218 µM 43.6 µM na

130

EGCG 37 °C, cell culture conditions [380] solvent additive initial concentration mean half-life (t1/2) degradation product RPMI: Ham’s F12 (1:1)

KYSE 150 cells 20 µM 1 hour theasinensin A H2O2

RPMI: Ham’s F12 (1:1)

KYSE 150 cells SOD (5 U/mL)

20 µM 24 hours GCG dimer

EGCG, 50 °C, dark [406] solvent additive initial concentration t = concentration

decreased to 90% degradation product

Glycerin - 10.9 mM 24.4 days na Glycerin AA (0.1% w/w) 10.9 mM 25.7 days na Glycerin Pgal (0.1% w/w) 10.9 mM 29.3 days na Glycerin BHT (0.1% w/w) 10.9 mM 76.1 days na Glycerin BHT+Pgal

(0.05%+0.05%) 10.9 mM 7.4 days na

Transcutol P - 10.9 mM 22.8 days na Transcutol P AA (0.1% w/w) 10.9 mM 24.7 days na Transcutol P Pgal (0.1% w/w) 10.9 mM 3.5 days na Transcutol P BHT (0.1% w/w) 10.9 mM 8.6 days na Transcutol P BHT+Pgal

(0.05%+0.05%) 10.9 mM 7.0 days na

Oral delivery systems

Oral drug delivery, although not very efficient, is the most convenient and well accepted

route of administration. Encapsulation can protect EGCG from extreme pH changes in the

digestive system as well as increase the GI retention time after oral consumption [420].

Biodegradable and biocompatible cationic polysaccharides, such as chitosan (CS) are suitable

carriers for oral drug delivery. Chitosan adheres to the mucosa through ionic interactions

between its positively charged amine groups and the negatively charged groups present on

intestinal surfaces. Furthermore, interaction between chitosan molecules and polyphenolic

compounds could be a useful strategy to increase their encapsulation efficiency [421]. The

summary of stability and biological activity of encapsulated EGCG compared to free EGCG in

oral delivery formulations is listed in Table 3.

In the absence of food, intestinal mucosa is regularly washed and therefore a relatively

rapid release/absorption of EGCG is required. Diet type and food quantity greatly influence GI

retention time, which should be taken into consideration while developing oral delivery

systems. Significant increases in stability and/or activity of EGCG with sufficient GI retention

time have already been achieved in experimental animals [420]. Stabilization of EGCG and its

optimized release in the GI tract over a period of 20 hours should be adequate for humans.

Oral drug delivery systems demonstrated increased intestinal retention time, stability

and adsorption of EGCG, and consequently enhanced its plasma concentration. However, with

oral administration side effects in non-target tissues can occur. Despite the fact that these

systems are being established in experimental animals, their translation to humans still

requires some effort.

131

Intravenous and local delivery systems

Intravenous administration overcomes the issue of poor intestinal absorption, but can

cause discomfort, rapid particle clearance and can activate host defense mechanisms.

Intravenously administered nanoparticles are passively targeted into tumors due to an

enhanced permeability and retention (EPR) effect. In addition to EPR effect, selective, receptor-

mediated drug delivery of EGCG is another attractive approach to tumor targeting [422].

Biocompatible polymers with a tunable degradation rate and release are often used (e.g. PLA,

PLGA) [423, 424].

The advantage of particles and liposomes – compared to macromolecules – lies in their

reduced kidney clearance, which is related to their larger size. Importantly, surface properties

significantly influence the fate of a particle, e.g. PEG coating protects particles from recognition

by the mononuclear phagocyte system (opsonisation). Properties of particulate systems can be

tuned by integration of ligands, coating, or stimuli-responsive components, to trigger e.g. pH-

catalyzed hydrolysis. The stability and biological effect of encapsulated EGCG compared to free

EGCG in intravenous delivery formulations is listed in Table 4.

Local delivery methods include non-invasive (e.g. topical, ocular) and invasive (e.g. intra-

tumoral) routes of administration. Local delivery avoids rapid kidney clearance of particles and

significantly reduces the risk of systemic side-effects. The stability and biological effect of

encapsulated EGCG compared to free EGCG in local delivery formulations is listed in Table 4.

Modification of (–)-epigallocatechin gallate molecule

Due to high solubility of EGCG in water, it is difficult to obtain effective encapsulation.

One approach to reduce the water solubility of EGCG is to hide hydroxyl groups. Acetylation of

EGCG masked the hydrophilic OH-groups, which increased the antioxidant properties and

solubility of EGCG in fat [425]. In a comparable approach, EGCG was treated with acetic

anhydride and pyridine to create peracetate protection of OH-groups and further encapsulated

into lipid nanoparticles. Despite a partial deacetylation (50% of initially acetylated EGCG), the

amount of EGCG in nanoparticles remained constant over 25 days, suggesting improved

stability [331].

To increase its lipophilicity and bioavailability, the EGCG molecule was esterified with

fatty acids [426]. EGCG lipophilised by docosapentaenoic acid (DPA) exhibited an enhanced

anti-inflammatory effect compared with free EGCG [427].

Cocrystallization can also be used to generate novel forms of EGCG with modulated

dissolution and pharmacokinetic profiles. As demonstrated by Smith et al., cocrystallization can

reduce water solubility of EGCG and slightly increase its bioavailability in mice after oral

administration. Although the in vivo effect compared with free EGCG was very modest in this

study, cocrystallization offers a tool to further modulate its pharmacokinetics [428].

132

Table 3. EGCG stability in oral drug delivery formulations. Physicochemical characteristics [EE = encapsulation efficiency (%), APS = average particle size, ZP = zeta potential (mV), PI = polydispersity index], release properties [IC = initial concentration of EGCG, RM = release media, %(t) =% of EGCG released in time t] as well as biological effect are listed. CS = chitosan, PAA = polyaspartic acid, CPP = caseinophosphopeptide, KBR = Krebs Bicarbonate-buffered Ringers solution, SGF = simulated gastric fluid, SIF = simulated intestinal fluid, * = significant difference in the stability or effect of encapsulated EGCG compared to free EGCG.

Chitosan (CS) microparticles aim: stability increase [421] delivery route

physicochemical characteristics release properties significance of encapsulation*

biological effect

na EE: 47.12±2.01% APS: 1.79±0.38 µm

ZP: na PI: na

IC: 218 mM catechin extract RM: PBS pH 5.5 PBS pH 7.4

%(t): 80% in 4 hours (60% in 1st hour)

na na

Nanoparticles CS aim: oral delivery [330, 429]

delivery route


biological effect

oral EE: na LE: 3.7±0.1 µg/mg APS: 440±37 nm

ZP: +25±1 PI: na

IC: 1.09 mM RM: KBR, pH 6.0

%(t): 60% in 2 hours

p < 0.05 in vitro (organ culture) in vivo

Nanoparticles CS-PAA aim: oral delivery [420]

delivery route


biological effect

oral EE: 25.0±2.1% APS: 102.4±5.6 nm

ZP: 33.3±0.2 PI: 0.224

IC: 4.4 mM RM: SGF, SIF

%(t): 100% in 5 hours

p < 0.05 in vivo (rabbit)

Nanoparticles CS-CPP aim: chemoprevention [430]

delivery route


biological effect

not specified

EE: 81.7±4.1% APS: 143.7±6.7 nm

ZP: 30.8±4.6 PI: 0.08-0.13

IC: 5.5 mM RM: PBS pH 6.2

%(t): 48% in 20 hours

p < 0.01 in vitro

Nanoliposomes aim: tumor delivery [431]

delivery route


biological effect

not specified

EE: 85.79±1.65% APS: 180±4 nm

ZP: na PI: na

IC: 10.6 mM RM: SGF, SIF

%(t): SGF: 21% in 4 hours SIF: 35% in 4 hours

p < 0.05 in higher concentrations

in vitro

Nanoparticles Gum Arabic-Maltodextrin aim: chemoprevention [432]

delivery route


biological effect

not specified

EE: > 80% APS: 120±28 nm

ZP: -12.3±0.8 PI: 0.45±0.23

IC: 109 mM RM: PBS pH 7.4

%(t): >40% in 10 minutes

p < 0.05 in lower concentrations

in vitro

133

Table 4. EGCG stability in intravenous and local drug delivery formulations. Physicochemical characteristics [EE = encapsulation efficiency, APS = average particle size, ZP = zeta potential (mV), PI = polydispersity index], release properties [IC = initial concentration of EGCG, RM = release media, %(t) =% of EGCG released in time t] as well as biological effect are listed. PLA = polylactic acid, PEG = polyethylene glycol, EtOH = ethanol, CTAB-LN = cetyltrimethylammonium bromide lipid nanoparticles, DDAB-LN = dimethyldioctadecylammonium bromide lipid nanoparticles, PLGA = poly(lactic-co-glycolic acid), CP = citrate-phosphate, * = significant difference in the stability or effect of encapsulated EGCG compared to free EGCG.

Nanoparticles PLA-PEG aim: chemoprevention [423]

delivery route


biological effect

systemic injection

EE: na APS: 260 nm

ZP: -7.92 PI: 0.07

IC: 218 mM RM: blood

%(t): 22-65 µM during 4 hours

p < 0.05 in vitro in vivo

Niosomes aim: tumor targeting [422]

delivery route


biological effect

systemic injection

EE: 30.3±2.8% APS: 294 nm

ZP: -36 PI: 0.414

IC: 2.2 mM RM: PBS

%(t): 80% in 24 hours

p < 0.05 in vitro in vivo

Lipid nanocapsules aim: stability increase [331]

delivery route


biological effect

not specified

EE: 80% APS: 57 nm

ZP: na PI: < 0.15

IC: 32.8 mM acetyated EGCG RM: water

%(t): 60% in 8 days

- -

Liposomes (-EtOH) aim: intra-tumor retention [433]

delivery route


biological effect

intra-tumor injection, topical

EE: 36.3-89.7% APS: na

ZP: na PI: na

IC: 10.9 mM RM: na

%(t): na in most tests not found

in vivo

Liposomes (+EtOH) aim: intra-tumor retention [424]

delivery route


biological effect

intra-tumor injection

EE: >98% APS: 378.2±10.9

ZP: -26±0.1 PI: na

IC: 17.2 mM RM: CP buffer pH 7.4

%(t): 0% in 12 hours

p < 0.05 for concentrations below 50 µM

in vitro in vivo

Lipid nanoparticles CTAB-LN, DDAB-LN

aim: ocular delivery [434]

delivery route


biological effect

not specified

EE: 98.99±0.19% 96.86±1.88% APS: 149.1±1.78 143.7±0.45

ZP: 20.8±0.89 25.7±1.42 PI: 0.240±0.008 0.160±0.015

IC: 26.2 mM RM: na

%(t): na na na

Nanoparticles PLGA aim: chemoprevention [411]

delivery route


biological effect

topical EE: 26% LE: 5.76% APS: 127.2±12 nm

ZP: -24.5 ± 1.89 PI: 0.189

IC: 87 mM RM: PBS, pH 7.4

%(t): 50% in 24 hours

p < 0.05 in vivo

134

6.6 Conclusions

Diverse beneficial activities of EGCG were demonstrated in cell culture models, among

which its antioxidant, anti-inflammatory and anti-cancer effects are most pronounced. This

indicates that EGCG is indeed a promising compound for clinical translation. However, EGCG is

less effective in corresponding animal experiments and moreover, only limited effects are

observed in human epidemiological studies.

The inconsistency between the activity of EGCG in the cell cultures and in the whole

organism can be attributed to the insufficient stability of the EGCG molecule: uncontrolled

EGCG degradation does not only lead to inaccurate conclusions of in vitro studies, but can also

limit EGCG’s bioavailability in vivo, hence decreasing its beneficial effects. For this reason,

accurate translation of in vitro and preclinical research to humans has not yet been successful.

Various encapsulation techniques for plant polyphenols have been developed, however,

they have several disadvantages. Classical methods include the use of extreme conditions (pH,

temperature), which causes a low encapsulation efficiency of EGCG, and the use of toxic

solvents with the need for time-consuming post-treatments. Moreover, the hydrophilic nature

of EGCG further reduces its encapsulation efficiency, as EGCG tends to leak between the

organic and aqueous phases during the standard encapsulation process.

Bearing in mind the clinical translation, stabilization method should reflect the final

EGCG application: targeted delivery with rapid release of a toxic dose is desired in cancer

treatment, whereas slow release of low EGCG doses over longer periods of time may be needed

to treat age-related diseases like atherosclerosis or neurodegeneration. Reducing agents and

appropriate encapsulation strategies, which sufficiently improve the stability of EGCG during

the delivery process, in combination with in vivo stabilizing mechanisms such as antioxidant

enzymes and lower oxygen partial pressure, will allow for full manifestation of EGCG’s

beneficial activities in vivo.

6.7 Acknowledgement

The authors have no conflict of interest. The authors would like to thank the Mäxi

Foundation for financial support and Prof. Dr. Paola Luciani and Prof. Dr. Bruno Gander for their

valuable advice.

135

6.8 Supplement Supplementary Table 1. The effect of concentration on the stability of EGCG. The initial concentrations of EGCG, its mean half-life, and its degradation products in respective solvents are listed. GCG = (–)-gallocatechin gallate, RT = room temperature, AA = ascorbic acid, hl = half-life, na = information not available.

EGCG, RT, pH 5.4 [396] solvent additive Initial concentration % at t = 7 days (7d)

mean half-life (t1/2) degradation product

water - 0.51 mM 7d: 80% GCG water - 2.06 mM 7d: 80% GCG water - 6.99 mM 7d: 100% GCG water - 96 µM t1/2: 150 minutes theasinensin A water - 20 µM t1/2: 30 minutes theasinensin A EGCG, RT, pH 7.4 [409] solvent additive initial concentration % at t = 2 days

degradation product

HEPES - 0.55 µM 0% na HEPES - 1.09 µM 0% na HEPES - 54.5 µM 0% na HEPES - 218 µM 18% na HEPES - 654 µM 58% na HEPES - 1964 µM 80% na HEPES AA 0.25% (w/v) 0.55 µM 0% na HEPES AA 0.25% (w/v) 1.09 µM 0% na HEPES AA 0.25% (w/v) 54.5 µM 90% na HEPES AA 0.25% (w/v) 218 µM 98% na HEPES AA 0.25% (w/v) 654 µM 98% na HEPES AA 0.25% (w/v) 1964 µM 100% na EGCG, 37 °C, pH 7.4 [400] solvent additive initial concentration mean half-life (t1/2)

degradation product

K2HPO4 buffer (50 mM)

- 3.27 µM 40 minutes na





EGCG, various temperatures (-20 °C to +25 °C) [409] solvent additive initial concentration % at t = 7 days

degradation product

Transcutol P (20%, v/v)

AA (0.25%, w/v) ethanol (24%, v/v)

1964 µM ±100% na


AA (0.25%, w/v) ethanol (24%, v/v)

982 µM ±100% na


AA (0.25%, w/v) ethanol (24%, v/v)

654 µM ±100% na


AA (0.25%, w/v) ethanol (24%, v/v)

218 µM ±100% na


AA (0.25%, w/v) ethanol (24%, v/v)

54.5 µM ±100% na

136

Supplementary Table 2. The effect of pH on the stability of EGCG. The initial concentrations of EGCG, its mean half-life, and its degradation products in respective pH are listed. GTC = green tea catechins, GCG = (–)-gallocatechin gallate, hl = half-life, na = information not available.

EGCG, 4 °C [406] pH solvent Initial concentration % at t = 28 days (28d)

mean half-life (t1/2) degradation product

3 citrate buffer (0.05 M)

109 mM 28d: 90% na


109 mM 28d: 90% na

7 phosphate buffer (0.05 M)

109 mM t1/2: 7 days na

9 boric acid-borax buffer (0.05 M)


EGCG, 21 °C [406] pH solvent initial concentration % at t = 28 days (28d)

mean half-life degradation product


109 mM 28d: 90% na


109 mM 28d: 70% na




109 mM t1/2: 1 day na

EGCG, 50 °C [406] pH solvent initial concentration mean half-life (t1/2) degradation

product 3 citrate buffer

(0.05 M) 109 mM 28 days na


109 mM 4.5 days na


109 mM < 10 hours na


109 mM < 10 hours na

EGCG, RT [409] pH solvent Initial concentration % at t = 2 days degradation

product 3.5 HEPES 54.5 µM 40% na 7.4 HEPES 54.5 µM 0% na 3.5 HEPES 1.09 µM 0% na 7.4 HEPES 1.09 µM 0% na GTC mix [328] pH solvent initial concentration % at t = 18 hours (18h),

6 months (6m) degradation product

- water 1.09 mM 6m: 50% na 4 sodium

phosphate buffer (60 mM)

1.09 mM 6m: 30% na

5 sodium phosphate buffer (60 mM)

1.09 mM 6m: 10% na


1.09 mM 18h: 100% na


1.09 mM 18h: 80% na

137

Supplementary Table 3. The effect of temperature on the stability of EGCG. The initial concentrations of EGCG, its mean half-life, and its degradation products in respective temperatures are listed. GTC = green tea catechins, GCG = (–)-gallocatechin gallate, na = information not available. EGCG, pH 7, PBS [406] temperature solvent initial concentration mean half-life (t1/2) degradation product 4 °C PBS 109 mM 7 days na 21 °C PBS 109 mM 2 days na 55 °C PBS 109 mM less than 10 hours na EGCG, pH 7.4 [409] temperature solvent initial concentration % at t = 2 days degradation product 4 °C HEPES 54.5 µM 60% na 25 °C HEPES 54.5 µM 0% na 4 °C HEPES 1.09 µM 43% na 25 °C HEPES 1.09 µM 0% na GTC mix [328] temperature solvent Initial concentration % at t = 3 hours degradation product 100 °C deionized water 1.09 mM 75% na 70 °C deionized water 1.09 mM 80% na 24 °C deionized water 1.09 mM 100% na GTC mix [327] temperature solvent initial concentration % at t = 7 hours degradation product 98 °C distilled water 10.9 mM 80% GCG 37 °C distilled water 10.9 mM 100% na

6.5 sodium phosphate buffer (60 mM)

1.09 mM 18h: 50% na


1.09 mM 18h: 30% na

7.4 sodium phosphate buffer (60 mM)

1.09 mM 18h: 17% na

EGCG [328] pH solvent initial concentration % at t = 6 days degradation

product 7.4 PBS (60 mM) 1.09 mM 0% na EGCG, 120 °C [327] pH solvent initial concentration % at t = 20 minutes degradation

product 3 citric acid-sodium

citrate buffer (0.1 M)

1.09 mM 95% GCG 1%

4 citric acid-sodium citrate buffer (0.1 M)

1.09 mM 90% GCG 5%

5 NaH2PO4-Na2HPO4 buffer (0.1 M)

1.09 mM 55% GCG 19%

6 NaH2PO4-Na2HPO4 buffer (0.1 M)

1.09 mM 20% GCG 8%

- distilled water 1.09 mM 75% GCG 13%

138

Supplementary Table 4. The effect of ionic strength on the stability of EGCG. The initial concentrations of EGCG, its mean half-life, and its degradation products in respective ionic strengths are listed. na = information not available. EGCG, pH 5, t=50 °C, acetate buffer (0.05 M) + NaCl to adjust ionic strength

[406]

ionic strength

solvent initial concentration t = concentration decreased to 90%

degradation product

0.1 M acetate buffer 109 mM 7.35 hours na 0.2 M acetate buffer 109 mM 5.51 hours na 0.3 M acetate buffer 109 mM 5.16 hours na 0.4 M acetate buffer 109 mM 4.52 hours na 0.5 M acetate buffer 109 mM 3.21 hours na

139

Chapter 7

Towards controlled delivery of epigallocatechin

3-gallate as an anti-inflammatory treatment for

degenerative disc disease

Olga Krupkova, Aikaterini Zafeiroupoulou, Oddny Björgvinsdóttir, David

Eglin, Stephen J Ferguson, Karin Wuertz-Kozak

Report of work in progress

141

7.1 Abstract

We have previously identified the polyphenol epigallocatechin 3-gallate (EGCG) as a

promising therapeutic candidate for the treatment of degenerative disc disease. In human

intervertebral disc (IVD) cells, EGCG not only inhibits inflammatory responses, but also reduces

the detrimental effects of oxidative stress. In addition, EGCG ameliorates IVD-related

radiculopathy in experimental animals, which underlines its clinical significance. However, the

EGCG molecule is prone to degradation, which complicates clinical translation of EGCG-based

treatments. Therefore, a delivery system that protects EGCG and provides its stable release has

to be developed to ensure efficient long-lasting treatment of degenerative disc disease.

The overall goal of this project is to develop and characterize an intradiscal delivery

system, stabilizing EGCG and providing sustained release. The delivery system consists of

EGCG-containing gelatin microparticles dispersed in injectable thermo-reversible hydrogel (HA-

pNIPAM). Microparticles were created by electrospraying and their biocompatibility was

verified on human IVD cells in vitro using MTT assay. The release test was designed and the

release of free EGCG from HA-pNIPAM was measured, to establish a baseline release. This

report outlines methodology, current progress, and future steps of the project.

142

7.2 Introduction

Degeneration and aging of the intervertebral disc (IVD) includes morphological and

functional changes such as calcification of endplates, disruption of the annulus fibrosus (AF)

and loss of extracellular matrix (ECM) in the nucleus pulposus (NP). The degenerative process

involves a combination of multiple events including inflammation, auto-immune reactions,

senescence and cell death [43, 240]. Although IVD degeneration is not symptomatic in all cases,

it is known to be one of the major factors in the development of low back pain [128]. Despite

the high prevalence of discogenic low back pain and the consequently high economic burden,

only symptomatic treatments (e.g. analgesics or spinal surgeries) are available [435-437]. Novel

therapeutic strategies such as molecular and cell therapies could restore the disc function in a

biologically relevant manner. Injected or implanted cells can support the old disc population,

produce ECM and “rejuvenate” the aging disc. However, as degenerative IVDs are often

catabolic, inflammatory, and acidic, the implanted cells are at risk of apoptosis [59, 177].

Inflammation and catabolism in the IVD can be regulated by molecular therapeutics, which can

prevent premature aging, reduce cell death, enhance ECM turnover and inhibit expression of

pain-mediators, thus enhancing survival and functionality of the resident cells while

counteracting nociception. Several pharmacological molecules promoting IVD regeneration

have been tested to date, with encouraging results [16, 176, 438].

Recently our group has shown that phytochemicals can serve as promising compounds

to reduce inflammation, catabolism and pain arising from the IVD [18, 170, 171, 173].

Epigallocatechin gallate (EGCG), a polyphenol found in tea (Camelia sinensis), interferes with

the IL-1β-induced inflammatory cascade, inhibiting the key inflammatory mediators NF-kB, JNK

and p38 and reducing the expression of inflammatory and catabolic genes/proteins (IL-6, IL-8,

MMP1, MMP3, MMP13, TLR2, COX-2, NGF, iNOS). Moreover, local application of EGCG in a rat

model of IVD herniation attenuates radiculopathy, hence supporting the clinical relevance of

this compound. In addition to its anti-inflammatory effects, EGCG also possesses the ability to

inhibit free radical-mediated damage [439]. We have shown that EGCG does not only

ameliorate inflammation and pain, but also reduces oxidative stress-induced cell death of IVD

cells, via up-regulation of PI3K/Akt pathway [311]. This suggests that EGCG constitutes a

promising therapeutic option, as it targets multiple deregulated processes in the degenerative

IVD.

Oral drug delivery of pharmaceutical compounds has the highest compliance rates.

However, after the oral ingestion of green tea or EGCG supplements, only 0.1-1.1% of the orally

administered dose reaches the systemic circulation [329]. On the other hand, when high doses

of EGCG-containing preparations (10-29 mg/kg/day) are ingested, cases of hepatotoxicity and

nephrotoxicity can occur [391]. Importantly, the avascular disc has limited blood supply.

143

Therefore, EGCG delivered orally is most likely not appropriate for IVD applications, as rapid

degradation or side effects in non-target tissues can occur. Local delivery (e.g. intradiscal

injection) overcomes the poor intestinal absorption and reduced blood supply to the IVD and

significantly lowers the risk of systemic side-effects [31]. To promote the successful clinical

translation of safe and therapeutically active EGCG doses, delivery system with the ability to

enhance EGCG’s stability, to provide sustained delivery, and to account for specific local tissue

requirements is necessary.

Although various encapsulation techniques for plant polyphenols have been developed,

these techniques are not suitable for the encapsulation of EGCG [51]. Unlike most of other used

compounds, EGCG is hydrophilic and thus tends to leak out during encapsulation, causing low

encapsulation efficiency. Moreover, the EGCG molecule is prone to degradation in extreme

conditions (high temperature and pH), common to traditional encapsulation methods [31, 439].

Bearing in mind future clinical application, the use of toxic solvents and encapsulation-

enhancing additives is not appropriate.

Electrospraying is an electrohydrodynamic encapsulation method that overcomes these

limits. Electrospraying can be used to generate solid particles in a one-step process without the

need of employing high temperatures or toxic solvents. During electrospraying, the solution jet

breaks down into fine droplets, which acquire spherical shapes due to the surface tension,

subsequently producing solid microparticles upon solvent evaporation [54]. EGCG-loaded

particles, created by electrospraying, exhibit high encapsulation efficiency [54]. However, once

injected, particles can escape the IVD space, or be damaged by mechanical loading.

To reduce particle displacement and damage, microparticles can be dispersed in in situ-

forming biomaterials, which provide protection and in the ideal case also improve the load-

bearing function of the NP. Environmentally-responsive hydrogels solidify after their injection

into the IVD tissue. Thermo-reversible poly(N-isopropylacrylamide) (pNIPAM) solidifies via intra-

and inter-molecular hydrophobic interactions between isopropyl groups of pNIPAM at body

temperature (gel point ~32°C). High molecular weight (Mw) pNIPAM is non-biodegradable, but

shorter pNIPAM chains (up to Mw 10.000) can undergo renal excretion [200]. A modified

pNIPAM has been recently used as a vehicle for siRNA and celecoxib in IVD degeneration

studies in vitro and in vivo [191, 197]. As hyaluronic acid (HA) is the most abundant water-

absorbing molecule in the NP tissue, HA-based biomaterials are suitable for the use in IVD

repair and regeneration [155]. HA-grafted pNIPAM (HA-pNIPAM) has been successfully tested

for delivery of cells and growth factors into the degenerative IVD [198, 199, 440, 441].

The overall goal of this project is to achieve controlled delivery of EGCG as an anti-

inflammatory therapeutic method in the treatment of degenerative disc disease. We

hypothesize that electrosprayed EGCG microparticles dispersed in HA-pNIPAM can serve as an

intradiscal delivery system with enhanced stability and sustained release of EGCG, compared to

144

direct incorporation of EGCG in HA-pNIPAM or compared to pure EGCG microparticles without

a carrier. The specific aims of the project are (1) to prepare stable biocompatible EGCG

microparticles for dispersion in HA-pNIPAM and (2) to characterize the release of EGCG from

this delivery system.

7.3 Methods

Electrospraying of gelatin microparticles

The biocompatible natural material gelatin was used for the encapsulation of EGCG.

Gelatin A (G1890, Sigma) was dissolved in 20% (v/v) acetic acid (A6283, Sigma), stirred at 40°C

for 6 hours and cooled down to a room temperature for electrospraying [54, 442]. As gelatin

tends to swell and dissolve rapidly at 37°C, the cross-linking agent glutaraldehyde (GA, 5% w/v),

added at different concentrations (37.5, 62.5 and 87.5 μg/mL), was tested to extend the lifetime

of the gelatin microparticles [443]. 6% (w/v) gelatin was used for preparation of non-cross-

linked microparticles, whereas 4% (w/v) gelatin was used for crosslinking [54, 442].

Electrospraying was conducted using a previously described custom-made setup

connected to high voltage supply and situated within a climate chamber for environmental

control, with the temperature and humidity being fixed to 24°C and 40%, respectively [444].

Spherical particles can be formed by varying electrospraying parameters, namely flow rate,

voltage and concentration [445]. Gelatin was introduced in a 5 mL plastic syringe, pumped at a

steady flow-rate and voltage (2 and 4 mL/min, 20-22 kV) through a 20G stainless steel needle

and microparticles were collected on an aluminum sheet placed at a distance of 10 cm from the

spray needle [443, 446]. Samples of particles were sputter-coated with gold under vacuum (8

µm) and analyzed by scanning electron microscopy (SEM, FEI Quanta 200F) at an accelerating

voltage of 10 kV and a working distance of 10 mm. Particle shape and size distribution was

calculated from 3 independent images analyzed by Image J (>100 particles in total).

Biocompatibility of gelatin microparticles

Biocompatibility of gelatin microparticles without EGCG was tested by MTT assay, to

exclude toxicity of the solvent acetic acid and of the crosslinking agent glutaraldehyde. Human

NP tissue (grade III-V), isolated from donors undergoing spinal surgeries for degenerative disc

disease or disc herniation (approved by Kantonale Ethikkommission Zürich EK-16/2005), was

used. The tissue was enzymatically digested and primary cells were isolated [18]. Primary cells

were seeded in Dulbecco’s Modified Eagle’s Medium (DMEM/F12, D8437, Sigma, St. Louis, MO,

USA), supplemented with 10 % fetal calf serum (FCS, F7524, Sigma), penicillin (50 units/mL),

streptomycin (50 μg/mL) and ampicillin (125 ng/mL, 15240-062, Gibco, Carlsbad, CA USA) and

sub-cultured using 1.5% trypsin (15090- 046, Gibco). Microparticles were sterilized by UV-C (3

145

hours) and placed in the upper part of a trans-well system (ThinCert™ Tissue Culture Inserts,

82051-570, VWR), with human primary IVD cells seeded on the bottom. On day 4, fresh MTT (3-

[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide, M5655, Sigma) solution in media

(0.5 mg/mL) were added and maintained for 3 hours at 37°C. MTT was discarded, cells were

lysed in DMSO (D8418, Sigma) and the absorbance was measured at 565 nm [311]. The result

was expressed as percentage of control cells, without exposure to microparticles.

The release of EGCG from HA-pNIPAM

The release test of free EGCG from HA-pNIPAM was performed at 37°C in 10 mL of 0.9%

NaCl as release media. 10 mL of the release media was collected and exchanged at 1h, 24h and

on days 3, 5 and 7. The release of EGCG was tested by ferrous tartarate assay, specific for

catechins. EGCG dyeing solution [1 g ferrous sulfate (F8633, Sigma) + 5 g potassium sodium

tartrate tetrahydrate (S2377, Sigma) in 1000 mL of distilled water] was mixed with 0.067 M

potassium phosphate buffer (pH 7.5) as 1:4. This mixture was used to dilute EGCG samples for

colorimetric spectroscopy (540 nm) in a 1:1 ratio [410]. UV-VIS spectroscopy (275 nm) was

employed as a complementary method to ensure accuracy of values obtained by ferrous

tartarate assay [54].

7.4 Results

Electrospraying of gelatin microparticles

Well-shaped microparticles with an average diameter of 351.52 ± 62.56 nm can be

produced by electrospraying of 6% non-cross-linked gelatin in 20% acetic acid under 21 kV and 2

µL/min. 5 mM EGCG (E4143, Sigma) was mixed with 6% non-cross-linked gelatin in 20% acetic

acid and sprayed under the same conditions. As a result, spherical microparticles with diameter

496.67 ± 151.66 nm were formed (Figure 1).

As non-cross-linked gelatin exhibits quick degradation at 37°C, crosslinking is required.

Glutaraldehyde (GA) crosslinking increases the stability of gelatin, reduces its solubility, and

provides slower release of encapsulated compounds [443]. Therefore, 37.5 µg/mL, 62.5 µg/mL,

and 87.5 µg/mL of GA was added into plain 6% gelatin and sprayed under the flow rate of 2 and

4 µL/min, based on the initial results described above. However, the addition of glutaraldehyde

resulted in formation of fibers between particles, especially at the higher flow rate (Figure 2).

To reduce the fiber formation, the concentration of gelatin for GA cross-linking was

lowered to 4%, while the flow rate was unchanged. As a result, smaller fibers were formed. 2

µL/min flow rate further reduced the fiber formation, especially in the group cross-linked with

37.5 µg/mL GA (Figure 3). Gelatin cross-linked with 37.5 µg/mL GA and sprayed under the 2

146

µL/min yielded best results, with an average particle size of 907.5 ± 241.49 nm and was thus

selected for further experiments. EGCG encapsulation tests are currently ongoing.

Figure 1. SEM images of electrosprayed non-crosslinked 6% gelatin and gelatin-EGCG. The most suitable parameters for electrospraying of plain gelatin have been determined as 6 % (w/v), 2 µl/min and 21 kV. EGCG was encapsulated and sprayed under the same conditions, forming spherical particles with diameter 496.67 ± 151.66 nm.

Figure 2. Scanning electron microscopy images of electrosprayed cross-linked 6 % gelatin. Glutaraldehyde (GA) was used as a crosslinking agent. 37.5 µg/mL, 62.5 µg/mL, and 87.5 µg/mL of GA in 6% gelatin solution caused fiber formation, especially at flow rate of 4 µL/min.

147

Figure 3. Scanning electron microscopy images of electrosprayed cross-linked 4 % gelatin. Glutaraldehyde (GA) was used as a crosslinking agent. 37.5 µg/mL, 62.5 µg/mL, and 87.5 µg/mL of GA in 4% gelatin solution reduced fiber formation, especially at the flow rate of 2 µL/min.

Biocompatibility of gelatin microparticles

While the non-toxicity of HA-pNIPAM has previously been demonstrated [441], the

biocompatibility of the cross-linked gelatin microparticles was examined within this study.

Biocompatibility of cross-linked gelatin microparticles was tested on human IVD cells (n = 3

independent repeats of one donor). 15, 30, 60 mg of particles were tested to span the amount

planned to be used in the proposed delivery system (= 42.3 mg). Particles were added into 2 mL

of complete media and kept at 37°C for 4 days. No major changes in cell viability were detected

in the groups treated with 15 and 30 mg of particles cross-linked with 62.5 µg/mL GA compared

to untreated IVD cells. However, 60 mg of particles reduced cell viability, indicating toxicity of

higher quantities of microparticles, most likely due to the increased concentration of GA (Figure

4). With the lower GA concentration, the designed delivery system is thus expected to be non-

toxic for IVD cells. Experiments with pure gelatin (without crosslinking) as well as with lower

GA concentrations (37.5 µg/mL GA ) are currently conducted to confirm this notion.

148

Figure 4. Biocompatibility test of electrosprayed cross-linked gelatin particles. 15 and 30 mg of gelatin particles, cross-linked with 62.5 µg/mL glutaraldehyde, did not activate cell death in IVD cells during 4 days in culture, while a drop in viability was observed at the highest concentration. Graph shows mean ± SD (n = 3).

Design of EGCG release tests

To determine the release of EGCG from the delivery system, appropriate control systems

are included (Table 1). The delivery system is composed of EGCG, encapsulated in 4% gelatin

particles cross-linked with 37.5 µg/mL glutaraldehyde, and thermo-sensitive injectable HA-

pNIPAM, provided by the AO Foundation Davos (System 1). Control systems contain free 5 mM

EGCG in HA-pNIPAM (System 2) and pure EGCG-microparticles without any carrier (System 3).

Additional controls include free 5 mM EGCG (System 4), and HA-pNIPAM without particles

(System 5).

The release of EGCG from HA-pNIPAM

The data on the release of free EGCG from HA-pNIPAM (System 2) is presented in this

report. 5 mM EGCG was mixed with 1 mL of 10% and 15% HA-pNIPAM to establish a baseline

release of the free compound over a period of 7 days (n = 5) (Figure 5A). Cumulative release was

calculated from the obtained data (Figure 5B). Free EGCG, incorporated in HA-pNIPAM, showed

high-dose burst release during the first 24 hours. EGCG was completely released before the day

5. This demonstrated that improvement of EGCG retention in the HA-pNIPAM through

encapsulation is needed. No major difference between EGCG release from 10% and 15% HA-

pNIPAM was detected.

149

Table 1. Experimental design of the release tests: EGCG delivery system and control systems. EGCG particles, formed by electrospraying of 4% gelatin, will be mixed with 1 mL of HA-pNIPAM for the release tests (System 1). Appropriate control systems will be used (System 2 - 5). A total concentration of 5 mM (2.293 mg/mL) EGCG was chosen, aiming for long-term sustained release of 5 - 15 µM EGCG. 100% encapsulation efficiency of EGCG in gelatin particles is assumed. Incorporation of 5 mM EGCG into 1 mL of HA-pNIPAM for the release tests

Free EGCG

(mg) % of EGCG in particles

EGCG particles (mg)

HA-pNIPAM 10% / 15% (mg)

1 EGCG particles in HA-pNIPAM

- 5.42% 42.3 100 / 150

2 Free EGCG in HA-pNIPAM

2.293 - - 100 / 150

3 EGCG particles - 5.42% 42.3 -

4 Free EGCG 2.293 - - -

5 HA-pNIPAM - - - 100 / 150

Figure 5. EGCG release (A) and cumulative EGCG release (B) from 10% and 15% HA-pNIPAM measured by the ferrous tartarate assay. Burst release of free EGCG was detected during first 24 hours. EGCG was completely released before the day 5. Graph shows mean ± SD (n=5).

7.5 Discussion

The molecular structure of EGCG is prone to rapid degradation, which causes issues in

clinical translation of EGCG-based therapeutics, despite their high potential. Therefore, a

suitable delivery system, including an effective means to protect EGCG from degradation, has

to be developed for each specific application. To stabilize EGCG and to ensure its sustained

release in the degenerative IVD, EGCG was incorporated into gelatin microparticles, which will

be then combined with the injectable carrier HA-pNIPAM. The anticipated benefits from this

system are: (1) Minimal-invasive applicability, (2) tissue retention, (3) sustained release, (4)

degradation protection and consequently (5) prolonged activity of EGCG.

Biocompatible polymers such as PLGA, pNIPAM, chitosan, or gelatin, have been tested as

drug carriers for the IVD [155, 447-449]. Gelatin is a collagen derivate, hence closely resembles

150

one of the main extracellular matrix components of the IVD. Gelatin was previously shown to

be appropriate for intradiscal delivery: The injection of pure gelatin-based microparticles

without addition of cells or drugs already positively affected the NP microenvironment and

decreased the rate of apoptosis in a rabbit IVD degeneration model [150, 450]. Gelatin

microparticles were successfully used to deliver platelet rich plasma (PRP) into the NP in rabbits,

with sustained release through diffusion via the pores of the slowly degrading gelatin network

[192, 193].

Aside from gelatin-EGCG-microparticles, a critical component of the designed delivery

system is the carrier. HA-pNIPAM hydrogel was selected as the carrier, as it has previously been

shown to be suitable for intradiscal delivery [440]. A major challenge of injectable hydrogel-

based therapies in degenerative disc disease is to replicate the mechanical properties of the

native NP. Although hydrogel-based materials are presumed suitable for mimicking the NP

environment [451], their mechanical properties such as the Poisson’s ratio, bulk and shear

modulus, can differ from the native tissue. Poisson’s ratio of the NP determines the change of

lateral expansion or shrinkage under compressive or tensile loading. The bulk modulus

indicates the stiffness of the NP under compression or tension and the shear modulus

determines the torsional stiffness of the NP. For instance, the complex shear modulus (G*) of a

human NP is in the range of 7–21 kPa [451, 452], while G* of HA-pNIPAM (at 37°C) has been

determined as 4.6 kPa [441], suggesting lower mechanical stability of the HA-pNIPAM. The

influence of the hydrogel on the mechanical properties of the NP can be investigated by

measuring the change in the aforementioned parameters, before and after injection, using e.g.

whole organ cultures in suitable bioreactors. Based on the results, the mechanical properties of

HA-pNIPAM can be tuned by varying the concentration and ratio of hydrogel constituents.

It has been suggested that slowly biodegradable materials are favorable for IVD repair.

Although high Mw pNIPAM is not biodegradable and does not undergo renal clearance [155], it

has been shown that short pNIPAM chains (Mw < 100.000) undergo renal excretion in vivo

[453]. Moreover, HA-pNIPAM can be broken down in vivo due to the degradation of HA. A

modified HA-pNIPAM with hydrolytically degradable linkers is currently under development

(AO Foundation, Davos). The linkers will enable generation of small Mw degradation products

for improved renal excretion, hence increasing clinical applicability of HA-pNIPAM [200].

To date, we have optimized the conditions for electrospraying of EGCG-gelatin

microparticles and verified biocompatibility of particles following crosslinking with

glutaraldehyde. We observed that formation of beaded fibers, which occurred specifically under

increased flow rate and with gelatin crosslinking, could be compensated by lowering the

gelatin concentration. We have shown that electrospraying of 4 % gelatin, cross-linked with

37.5 µg/mL glutaraldehyde, resulted in formation of well-shaped and biocompatible

microparticles. We have also shown that free EGCG mixed with HA-pNIPAM (System 2)

151

exhibited an unfavorable high-dose release, indicating the need to include an EGCG

encapsulation in the system.

The aim 1 was achieved, while the aim 2 will be completed in the future. The release of

EGCG from the delivery system (System 1) will be measured in solution and compared with

control systems (System 2, 3, 4 and 5). The release of EGCG will be determined by ferrous

tartarate assay, as described above. In addition, gelatin degradation in the delivery system will

be measured, as the release of EGCG from the microparticles is expected to occur due to the

gelatin degradation. Gelatin degradation will be measured by the presence of hydroxyproline

(HYP) in the release media, using the HYP assay kit (MAK008, Sigma) [30]. Gelatin degradation

will be expressed relative to control HA-pNIPAM without particles.

To verify whether encapsulation influenced the functionality of the EGCG molecule, the

antioxidant activity of EGCG in release media will be measured using 2, 2’-diphyenyl-1-

picrylhydrazyl (DPPH) radical scavenging assay (D9132, Sigma) at 517 nm (n = 5). 100 μl of EGCG

solution will be incubated with 500 μl of DPPH (250 μM in ethanol; D9132 Sigma) for 1 hour at

RT. DPPH radical scavenging, measured at 517 nm, manifests as a decrease in absorbance. L-

Ascorbic acid (A4403, Sigma) and ethanol (O2860, Sigma) in equal amounts will be used as

assay controls [454]. In the future, the long-term anti-inflammatory activity of optimized EGCG

delivery system will be characterized in IVD cell and organ cultures treated with inflammatory

cytokines. The anti-inflammatory activity will be measured by RT-qPCR, as described before [18].

It has been reported that 1-3 mL of material can be injected into the center of human

lumbar degenerative NP [455, 456]. Therefore, we incorporate EGCG microparticles in 1 ml of

HA-pNIPAM. Our aim is to achieve a stable release of 5-15 µM EGCG over a period of one month

[457]. Given the fact that diurnal fluid loss in the human NP is 10-20 % [458, 459] and burst

release always occurs [31], our currently ongoing release tests are designed to encapsulate a

high-dose (5 mM) of EGCG. As the relationship between the concentration of encapsulated

EGCG, the quantity of HA-pNIPAM, and the EGCG diffusion rate is not yet known, the amounts

of these components may need to be adjusted, to achieve optimum EGCG activity. Once

optimized and fully characterized, the HA-pNIPAM-based delivery system can be compatible

also with other small-molecular therapeutics such as kinase inhibitors and other

phytochemicals, hence offering great clinical versatility/applicability.

153

Chapter 8

Discussion

155

8.1 Summary

Molecules targeting deregulated signaling pathways involved in painful degenerative

disc disease (DDD) are promising therapeutic candidates. Natural compounds can exhibit

antioxidant, anti-inflammatory and pro-survival activities via modulation of multiple signal

transduction and apoptotic cascades, with minimal adverse effects. The overall goal of this

thesis was to evaluate polyphenol epigallocatechin gallate (EGCG) as therapeutic candidate

against DDD. In this thesis, the anti-inflammatory (Aim 1) and the antioxidant (Aim 2)

properties of EGCG were investigated, mainly in primary cell cultures. Subsequently, a

degenerative nucleus pulposus (NP) organ culture model was developed in order to assess the

effects of EGCG in the NP tissue (Aim 3). An intradiscal sustained release system for EGCG was

designed and is currently under development (Aim 4).

The model systems used in our studies were chosen to adequately represent

inflammation and oxidative stress that occur in diseases. During DDD, regulation and function

of the IL-1 pathway are altered. Synthesis of IL-1α, IL-1β and IL-1R increases, whereas the

expression of IL-1R antagonist remains unchanged [82]. Cytokine-stimulated cell cultures

appropriately reproduce deregulated inflammatory signaling, representing a good system for

the evaluation of anti-inflammatory treatments. Similarly, hydroxyl radical, released from H2O2,

induces oxidative stress in a biologically relevant manner, allowing to test antioxidant

molecules in cell cultures [460]. However, IL-1R antagonists and natural antioxidant enzymes

are commonly extracellular (e.g. diffused in the ECM). Therefore, organ cultures are essential

for investigating the complex effects of the therapeutic compounds, as organ cultures contain

both, the cells and the native ECM. In this thesis, cell cultures were used to reveal the

mechanisms involved in EGCG action, whereas organ cultures were employed to investigate

the effects of EGCG on the whole NP. Importantly, clear evidence that EGCG interferes with the

inflammatory and the oxidative stress responses in the IVD and exhibits analgesic effects in

vivo could be provided.

8.1.1 Clinical relevance

Low back pain, often caused by complex interactions of biological and psychological

factors, is comprised of both nociceptive and neuropathic component [461].

Molecules that play a major role in the neuronal sensitisation are prostaglandins.

Biosynthesis of prostaglandin E2, mediated by cyclooxygenase-2 (COX-2), significantly increases

in degenerative and injured tissues, magnifying nociceptive and inflammatory responses [462].

The acute nociceptive IVD-related pain can be reduced by analgesics such as non-steroidal anti-

inflammatory drugs (NSAIDs) that target the production of prostaglandins. However,

156

prolonged PGE2-mediated sensitization of the nociceptor neurons of dorsal root ganglion (DRG)

contributes to the transition from acute to chronic pain [463]. NSAIDs are not effective in

chronic pain conditions, hence being unable to prevent the DDD progression. As EGCG does not

only reduce the expression of COX-2, but also inhibits other pain-related mediators and

inflammation, it is considered a good therapeutic candidate for the management of chronic

pain.

Aside from these effects, EGCG was furthermore able to reduce nerve growth factor

(NGF) expression. NGF has been recognized as an important mediator of chronic pain and as a

marker of pathological inflammation [464, 465]. NGF binds to two different receptors on the

nociceptor neurons: (1) To the low-affinity 75 kDa neurotrophin receptor (p75NTR), which binds

all neurotrophins, and (2) to the tropomyosin-related tyrosin kinase A (TrkA), which binds NGF

more selectively than the other neurotrophins. NGF-activated p75NTR regulates a wide range of

functions, including axonal growth, apoptosis and myelination [466]. NGF signaling through

TrkA mediates the nociceptive functions of the sensory neurons through enhanced expression

of substance P, calcitonin gene-related peptide (CGRP) and transient receptor potential

vanilloid 1 (TRPV1) [464, 467]. Therefore, EGCG-mediated inhibition of NGF expression in the IVD

may reduce both, nerve ingrowth and nerve sensitization.

Neuropathic pain, arising from neuronal damage, often fails to respond to ordinary

medication. Therefore, therapeutics with the ability to attenuate the IVD-related radicular pain,

are desired. In order to determine the effectiveness of EGCG against neuropathic pain, in vivo

testing is required. As currently no model of DDD neuropathy exists, a well-established rat

model of IVD-induced radicular pain, coupled with the von Frey test, was employed in our

study. The von Frey test is widely used to investigate pain-related behavior that is caused by

radiculopathy with limb symptoms, typical for IVD herniations [170, 230, 232]. In our 28-day

animal experiment, EGCG significantly improved the hind paw withdrawal threshold,

compared to the NP-only groups. We hypothesize that EGCG attenuated disc-related radicular

pain by inhibiting the inflammatory responses of the DRG, induced through the herniated NP

material. Recently, it has been reported that EGCG and its derivatives can reduce the intensity

of neuropathic pain after chronic nerve constriction injury in mice through an interaction with

NF-κB and the fatty acid synthase (FASN) in the dorsal horn of the spinal cord [468].

Furthermore, nerve damage and resulting neuropathic pain are closely linked with the

increased levels of the reactive oxygen or the nitrogen species [469]. As nitric oxide synthase

(iNOS in IVD cells, nNOS in nervous tissue) is involved in the generation of reactive nitric oxide

(NO) [470], elevated presence of NOS in DRGs and IVDs can intensify the pain sensation [471,

472]. Although we did not specifically test the DRG tissue, we demonstrated that EGCG reduced

the expression of iNOS in IVD cells. This possibly contributes to the amelioration of pain in vivo.

Moreover, EGCG enhanced survival of the IVD cells in the presence of the oxidative stress.

157

Recently, EGCG was shown to exhibit significant anti-apoptotic effects in skeletal muscles after

induction of sciatic nerve crush injury in mice [473], supporting our hypothesis on the protective

role of EGCG against oxidative stress-induced neuronal damage.

Interestingly, the improvement in the pain behavior was not immediate. This could be

caused by interaction between the extruded NP tissue and the immune system. It is well

known that macrophages are recruited to herniated NP material, where they can initiate

resorption through the release of MMPs and stimulate the IVD cells to express e.g. MMP1 and

MMP3 [474, 475]. On the other hand, macrophages also contribute to the increased cytokine

levels, hence promoting DRG damage [476]. Based on literature and our data, we suggest that

although EGCG reduced radiculopathy by inhibiting inflammation, it may have also affected

clearance of the displaced NP tissue in this model. Additional tests on the effects of EGCG on

macrophage-induced expression of MMPs will be necessary to determine whether resorption

of the herniated material may be slowed down upon treatment. If so, EGCG may not be

suitable for treating pain in patients with NP extrusions. To closely investigate the effects of

EGCG on radicular pain, an animal model where NP is not directly exposed to the DRG (e.g.

nerve constriction injury model) should be used.

We showed that EGCG did not only inhibit the expression of molecules involved in

nociception (cytokines, COX-2, NGF), but also reduced neuropathic pain markers (oxidative

stress, iNOS) and limb symptoms in experimental animals. Therefore, EGCG may be superior to

current clinical strategies, as it can attenuate both, the inflammation and pain, providing a

possible solution for the management of chronic and radicular pain in affected patients.

As 10 µM EGCG exhibited promising effects in the cell and organ cultures, we believe

that it possesses the ability to balance the inflammatory and catabolic changes also in human

IVDs in vivo. However, the exact dose and the duration of an EGCG intervention for DDD

therapy are yet to be established. The low stability of the EGCG molecule reduces its

bioavailability and limits its activity in biological systems. Upon oral ingestion, only 0.01 - 0.1%

of EGCG is absorbed, leading to maximum plasma concentration of 1 µM with a half-life of no

more than 12 hours. Oral intake of higher doses of EGCG can cause side-effects, affecting the

liver and the kidney [477, 478]. Therefore, encapsulation of EGCG into microparticles is desired

to increase its stability, bioavailability, and at the same time to protect the patient from

unspecific toxicity. Nevertheless, oral delivery is not practical for targeting the human IVD, due

to its avascular nature. Moreover, DDD is further limiting the diffusion of the solutes from

blood into the center of IVD. Therefore, encapsulated EGCG should be delivered locally in a

minimally-invasive manner.

An ideal NP delivery system would be composed of a biocompatible material that is

liquid before application and thus injectable through a needle, but rapidly solidifies after the

158

injection [200]. A biomaterial with similar mechanical properties as the native NP is needed for

normal IVD function. However, the mechanical properties of the NP are complex and vary with

the stage of degeneration, thus difficult to replicate [452]. We are exploring and option of using

an injectable thermo-reversible hydrogel HA-pNIPAM to deliver EGCG encapsulated in gelatin

microparticles. Previously, HA-pNIPAM was proven suitable for stable drug release in the IVD,

exhibiting rapid gelation and good mechanical stability [200]. Although our delivery system has

not yet been fully tested, we provide evidence that electrosprayed gelatin microparticles are

biocompatible for IVD cells. We believe that once established, this treatment option can

constitute a clear therapeutic advantage over the current treatments of DDD.

8.1.2 Clinical applicability

The therapeutic activity of EGCG in the IVD is yet to be established in clinical trials.

Nevertheless, several phase I-III clinical trials have been performed on the effects of EGCG in

neurodegenerative diseases [479, 480], cancer chemoprevention [481-483] and metabolic

disorders such as insulin resistance and obesity [484-486], with promising results. In addition, it

has been proposed that the stabilization of EGCG via encapsulation can facilitate the clinical

translation of EGCG. Hydrogel-based injectable systems for the delivery of the therapeutic

agents and cells into the IVD are currently being tested in phase I-III clinical trials [487]. In our

approach, an injectable in-situ forming hydrogel was chosen to deliver EGCG-loaded

microparticles into the IVD. The ultimate goal is to reduce inflammation, catabolism, and pain

occurring due to the degenerated IVD.

Once the preclinical phase is successfully completed, clinical trials will determine to

which extend our experimental results reflect the treatment outcomes in clinical situation.

Careful patient selection is needed to evaluate treatment suitability. We believe that our

strategy can be useful in the treatment of early and moderate painful DDD with no major

defects in the annulus fibrosus (AF) structure. In early and moderate DDD, homeostasis in the

NP has shifted towards the catabolism. In such condition, cytokines and pain-mediating

chemicals are released from IVD cells, irritating existing and newly formed nerve fibers in the

AF. At the same time, the proteoglycan content in the NP is still sufficient and the disc height is

preserved. According to the diagnostics using T2-weighted MRI images, our strategy is intended

for homogeneous (grade I) or inhomogeneous (grade II) painful discs with bright hyperintense

white signal intensity and normal disc height as well as for inhomogeneous discs with

intermittent gray MRI signal intensity, without clear distinction between the NP and the AF,

and normal or slightly decreased height (grade III) [488]. Therapeutic interventions using anti-

inflammatory and anti-catabolic compounds are thought to restore homeostasis and

functionality of the resident disc cells. On the other hand, the proposed therapeutic strategy

may not be suitable as a single treatment for the advanced stages of DDD, characterized by

159

hypointense dark gray MRI signal and decreased disc height (grade IV), or by hypointense black

MRI signal with collapsed disc space (grade V) [488]. In advanced DDD, proteoglycans of the NP

are degraded, disc cells are dysfunctional or disappearing, and the AF is affected by non-

physiological load distribution. The total disc implants or the tissue-engineered materials are

more appropriate for the treatment of advanced DDD. Whether a combination of tissue-

engineered materials with EGCG microparticles could provide any benefits will have to be

tested in future experiments.

The majority of novel treatment strategies for DDD, including our approach, involve

injection of therapeutic agents or cells into the IVD. However, it has been reported that the

needle puncture of the AF can cause micromechanical alterations and accelerate disc

degeneration [489]. It has been reported that the provocative discography (using

predominantly 21G and 26G needles [490]) can result in higher incidence of disc degeneration

over a 10-year period in some patients [127, 491]. Nevertheless, it was shown that most discs,

subjected to discography, d0 not degenerate [127]. Importantly, the puncture outcome is likely

determined by the needle size, as demonstrated in experiments in rabbits [127].

Despite the concerns, intradiscal injections are being used in clinical practice not only

for provocative discography, but also to deliver conventional analgesics. Serious complications

of the current spinal injection therapy (e.g. cauda equina syndrome, septic facet joint arthritis,

puncture-related disc damage, or paraplegia) have been reported, but only in rare cases, hence

making it an overall low-risk intervention compared to classical surgical treatments.

Consequently, there has been no strong evidence for or against the use of any type of injection

therapy in individuals with chronic low back pain [492]. Based on the available data, we

conclude that employing intradiscal injections in our therapeutic approach can be safe. We aim

to use needle size, no bigger than 22G, with which the delivery of cells in the HA-pNIPAM has

been previously validated [441].

Apart from the intradiscal injections, free or encapsulated EGCG can be injected into the

vicinity of the painful IVD, to avoid further damage to the IVD as a result of needle puncture. As

an example of similar approach, a sustained release curcumin depot was recently implanted

near the sciatic nerve in mice and tested for the treatment of IVD-related neuroinflammation

[493]. Topical delivery of EGCG, e.g. via lumbar patches, may also be possible. However,

transdermal delivery systems of EGCG are not yet well developed, due to the hydrophilic nature

of EGCG molecule and the need for using potentially toxic skin penetration enhancers [494].

160

8.2 Limitations

The main limitation of primary IVD cell cultures is their tendency to dedifferentiate in

vitro, thus losing the IVD phenotype. Dedifferentiation can be partially avoided using primary

cells in low passages and limiting cell expansion. However, the main disadvantage of this

approach is the low amount of primary cells available for the experiments. As primary IVD cells

are isolated from multiple donors with different genetic background and medical condition,

inter-donor variability occurs, leading to high variations in the cellular responses.

All our in vitro experiments have been performed in normoxia (21 % O2). However, O2

concentration in the center of the healthy human IVD can decrease below 4 % [58]. Low oxygen

concentration in the NP is considered physiological, as it is involved in maintaining the

phenotype of healthy NP cells and ECM. Therefore, our experimental conditions may not

exactly replicate the in vivo situation. On the other hand, it has been reported that normoxia in

vitro correlates with the non-physiological oxidative stress in the degenerative IVD [40, 92].

Therefore, normoxia most likely does not interfere with either of our hypotheses.

Although our studies provided evidence that EGCG functions as an IL-1 pathway

antagonist and PI3K/Akt activator, these studies did not address the exact mechanisms of

action. Previously, EGCG has been shown to interact with surface receptors such as TLR4,

laminin receptor or RTK [378, 495, 496]. Although we did not test this alternative, we

hypothesize that the interaction between EGCG and the surface receptors may be involved in

the effects observed in our studies. Mechanism of EGCG-mediated activation of PI3K/Akt and

its link to protection of mitochondria should be investigated in more detail.

The natural level of the inflammation-related gene expression and ROS was measured

neither in obtained biopsy tissue, nor in healthy donors. Therefore, the anti-inflammatory and

anti-oxidant effects of EGCG on non-stimulated IVD tissue could not be tested. In the animal

study, neither inflammatory responses of the DRG were evaluated, nor were the effects of

EGCG on the resorption of herniated NP tissue. Future studies should address these points.

A clear limitation of our approach is the low stability of the EGCG molecule, which may

impair the drug development process. If this will be the case, using EGCG derivatives with

chemically increased stability such as lipophilic EGCG [426] instead of natural EGCG may solve

this problem.

161

As our proposed therapeutic strategy involves an injection, it is not completely non-

invasive. In order to increase patient compliance, the EGCG sustained release system must

exhibit no side effects and possess high efficacy with therapeutic advantage over current

injection strategies. These aspects can be thoroughly evaluated in animal studies. Single

injection approach is preferred.

8.3 Future steps

Although we showed the promising effects of 10 µM EGCG in vitro, the appropriate dose

and duration off its application in the treatment of DDD in vivo is not known. In DDD, the

expression of MMPs significantly increases, shifting the metabolism towards catabolism and

reducing ECM turnover. Similarly, regulation and function of the IL-1 pathway and ROS is

altered, activating an inflammatory cascade [82]. On the other hand, it is well documented that

balanced levels of cytokines, MMPs, and ROS play an important role in IVD homeostasis, [209,

497], which can be disturbed by common therapeutics. As an example, corticosteroids, which

are classical anti-inflammatory agents often used in clinical practice, directly inhibit the

synthesis and release of TNF-α, IL-1, arachidonic acid, and its metabolites [498], and can thus

significantly reduce the activity of cytokine-regulated pathways in vivo. Consequent complete

cytokine inhibition, as well as the negative effects of corticosteroids on proliferation

morphology, exocytosis, phagocytosis, or cell motility [499, 500], suggest potential damaging

consequences of their use in the IVD. Therefore, molecules that are able to preserve naturally

low levels of cytokines, MMPs, and ROS in vivo, such as the herein tested substance EGCG, can

be better therapeutic candidates. Our studies suggest that EGCG does not pose a risk to the

non-degenerated IVD. However, the mechanisms of EGCG action in healthy and degenerative

NP in vivo are unknown and thus should be carefully investigated in the future.

We hypothesize that the increased EGCG stability and its sustained release will be

linked to more profound effects in the biological systems. However, the release properties of

EGCG from our delivery system are yet to be evaluated. It has been shown that the release of

EGCG in fully controlled in vitro experiments can differ from its release in the tissues. We

expect that the release properties in vivo will depend on the absolute amount of the material

injected, the EGCG dose, the stage of NP degeneration and the applied mechanical loading.

To answer some of these questions, long-term study on the activity of free and

encapsulated EGCG will be performed in the near future, using the degenerative NP culture

model, developed within this thesis. The NP culture model will allow us to evaluate the

therapeutic potential of EGCG and to fine-tune the dose before designing animal studies.

162

In the future, thorough release-activity studies performed in vitro, in organ cultures,

under mechanical loading, and in experimental animals are needed to fully evaluate the

behavior of our delivery system. In addition, the effects of EGCG on adjacent tissues (e.g.

vertebrae and endplates) must also be determined in vivo.

The evidence that phytochemicals can possess beneficial effects in the treatment of the

painful DDD has been provided. Phytochemicals are promising candidates to inhibit

inflammation, catabolism and oxidative stress in the degenerative tissues, while restoring

homeostasis. As different molecular therapeutics function via different mechanisms,

appropriate combinations of these molecules can produce additive and synergistic effects. In

the future, EGCG will be combined with other phytochemicals, to screen for additive and

synergistic effects. In addition, our EGCG-based approach will be tested together with

molecular therapeutics such as growth factors or LinkN, and cell therapies, to investigate full

potential of EGCG in the regeneration of IVD. Our NP culture model will serve for high

throughput preclinical tests of these treatment combinations.

8.4 Conclusion

Management of chronic pain conditions that arise from the degenerative IVD is

challenging due to the multifactorial nature of DDD. Degradation of the ECM, inflammation

and oxidative stress are common hallmarks of painful DDD. However, these hallmarks are not

yet being therapeutically targeted in a specific way. Unspecific analgesics and invasive

surgeries are used nowadays as standard treatments, with limited success. We believe that

minimally invasive procedures, delivering regenerative and analgesic substances and/or cells

will be superior to current standard treatments. Although some progress in the use of ECM

anabolics has been made in preclinical research, anti-inflammatory and antioxidant

approaches have not yet been thoroughly explored. The goal of this thesis was to identify and

test natural compounds which can interfere with the multiple aspects of the DDD, and to

evaluate their potential therapeutic use. In this thesis, evidence could be provided that EGCG is

a promising therapeutic candidate, as it reduces inflammation and oxidative stress in vitro as

well as pain intensity in vivo. To increase the clinical applicability of our findings, a delivery

system for EGCG has been designed and is currently under development. The question whether

our strategy is optimal for the clinical use will hence be answered in the future, after additional

preclinical and possibly also clinical testing.

163

8.6 References

[1] J.P.G. Urban, S. Smith, J.C.T. Fairbank, Nutrition of the intervertebral disc, Spine, 29 (2004) 2700-2709. [2] Z.I. Johnson, Z.R. Schoepflin, H. Choi, I.M. Shapiro, M.V. Risbud, DISC IN FLAMES: ROLES OF TNF-alpha AND IL-1 beta IN INTERVERTEBRAL DISC DEGENERATION, European cells & materials, 30 (2015) 104-117. [3] C.L. Le Maitre, A.J. Freemont, J.A. Hoyland, Accelerated cellular senescence in degenerate intervertebral discs: a possible role in the pathogenesis of intervertebral disc degeneration, Arthritis research & therapy, 9 (2007) R45. [4] M.A. Adams, P.J. Roughley, What is intervertebral disc degeneration, and what causes it?, Spine, 31 (2006) 2151-2161. [5] S.D. Kuslich, C.L. Ulstrom, C.J. Michael, The tissue origin of low back pain and sciatica: a report of pain response to tissue stimulation during operations on the lumbar spine using local anesthesia, The Orthopedic clinics of North America, 22 (1991) 181-187. [6] A.J. Freemont, T.E. Peacock, P. Goupille, J.A. Hoyland, J. OBrien, M.I.V. Jayson, Nerve ingrowth into diseased intervertebral disc in chronic back pain, Lancet, 350 (1997) 178-181. [7] Y. Wang, T. Videman, M.C. Battie, ISSLS Prize Winner: Lumbar Vertebral Endplate Lesions Associations With Disc Degeneration and Back Pain History, Spine, 37 (2012) 1490-1496. [8] G. Wynne-Jones, J. Cowen, J.L. Jordan, O. Uthman, C.J. Main, N. Glozier, D. van der Windt, Absence from work and return to work in people with back pain: a systematic review and meta-analysis, Occupational and environmental medicine, 71 (2014) 448-456. [9] S. Pai, L.J. Sundaram, Low back pain: an economic assessment in the United States, The Orthopedic clinics of North America, 35 (2004) 1-5. [10] W.C. Jacobs, S.M. Rubinstein, B. Koes, M.W. van Tulder, W.C. Peul, Evidence for surgery in degenerative lumbar spine disorders, Best practice & research. Clinical rheumatology, 27 (2013) 673-684. [11] K. Verma, S.D. Gandhi, M. Maltenfort, T.J. Albert, A.S. Hilibrand, A.R. Vaccaro, K.E. Radcliff, Rate of adjacent segment disease in cervical disc arthroplasty versus single-level fusion: meta-analysis of prospective studies, Spine, 38 (2013) 2253-2257. [12] B. Morlion, Pharmacotherapy of low back pain: targeting nociceptive and neuropathic pain components, Current medical research and opinion, 27 (2011) 11-33. [13] A.D. Furlan, J.A. Sandoval, A. Mailis-Gagnon, E. Tunks, Opioids for chronic noncancer pain: a meta-analysis of effectiveness and side effects, CMAJ : Canadian Medical Association journal = journal de l'Association medicale canadienne, 174 (2006) 1589-1594. [14] A.C. Disch, W. Schmoelz, G. Matziolis, S.V. Schneider, C. Knop, M. Putzier, Higher risk of adjacent segment degeneration after floating fusions: long-term outcome after low lumbar spine fusions, Journal of spinal disorders & techniques, 21 (2008) 79-85. [15] S.C. Chan, B. Gantenbein-Ritter, Intervertebral disc regeneration or repair with biomaterials and stem cell therapy--feasible or fiction?, Swiss medical weekly, 142 (2012) w13598. [16] D. Sakai, S. Grad, Advancing the cellular and molecular therapy for intervertebral disc disease, Advanced drug delivery reviews, 84 (2015) 159-171. [17] D. Sakai, G.B. Andersson, Stem cell therapy for intervertebral disc regeneration: obstacles and solutions, Nature reviews. Rheumatology, 11 (2015) 243-256. [18] O. Krupkova, M. Sekiguchi, J. Klasen, O. Hausmann, S. Konno, S.J. Ferguson, K. Wuertz-Kozak, Epigallocatechin 3-gallate suppresses interleukin-1beta-induced inflammatory responses in intervertebral disc cells in vitro and reduces radiculopathic pain in rats, European cells & materials, 28 (2014) 372-386. [19] S. Illien-Junger, F. Grosjean, D.M. Laudier, H. Vlassara, G.E. Striker, J.C. Iatridis, Combined Anti-Inflammatory and Anti-AGE Drug Treatments Have a Protective Effect on Intervertebral Discs in Mice with Diabetes, PloS one, 8 (2013). [20] S.Z. Wang, Y.F. Rui, Q. Tan, C. Wang, Enhancing intervertebral disc repair and regeneration through biology: platelet-rich plasma as an alternative strategy, Arthritis research & therapy, 15 (2013).

164

[21] R. Gawri, J. Antoniou, J. Ouellet, W. Awwad, T. Steffen, P. Roughley, L. Haglund, F. Mwale, Best Paper Nass 2013: Link-N Can Stimulate Proteoglycan Synthesis in the Degenerated Human Intervertebral Discs, European cells & materials, 26 (2013) 107-119. [22] M.K. Shanmugam, R. Kannaiyan, G. Sethi, Targeting Cell Signaling and Apoptotic Pathways by Dietary Agents: Role in the Prevention and Treatment of Cancer, Nutr Cancer, 63 (2011) 161-173. [23] B. Gantenbein, S. Illien-Junger, S.C.W. Chan, J. Walser, L. Haglund, S.J. Ferguson, J.C. Iatridis, S. Grad, Organ Culture Bioreactors - Platforms to Study Human Intervertebral Disc Degeneration and Regenerative Therapy, Current stem cell research & therapy, 10 (2015) 339-352. [24] R. Gawri, F. Mwale, J. Ouellet, P.J. Roughley, T. Steffen, J. Antoniou, L. Haglund, Development of an Organ Culture System for Long-Term Survival of the Intact Human Intervertebral Disc, Spine, 36 (2011) 1835-1842. [25] B. Jim, T. Steffen, J. Moir, P. Roughley, L. Haglund, Development of an intact intervertebral disc organ culture system in which degeneration can be induced as a prelude to studying repair potential, European Spine Journal, 20 (2011) 1244-1254. [26] S. Dudli, D. Haschtmann, S.J. Ferguson, Fracture of the vertebral endplates, but not equienergetic impact load, promotes disc degeneration in vitro, Journal of Orthopaedic Research, 30 (2012) 809-816. [27] D.Z. Markova, C.K. Kepler, S. Addya, H.B. Murray, A.R. Vaccaro, I.M. Shapiro, D.G. Anderson, T.J. Albert, M.V. Risbud, An organ culture system to model early degenerative changes of the intervertebral disc II: profiling global gene expression changes, Arthritis research & therapy, 15 (2013). [28] D. Purmessur, B.A. Walter, P.J. Roughley, D.M. Laudier, A.C. Hecht, J. Iatridis, A role for TNF alpha in intervertebral disc degeneration: A non-recoverable catabolic shift, Biochemical and biophysical research communications, 433 (2013) 151-156. [29] G.Q. Teixeira, A. Boldt, I. Nagl, C.L. Pereira, K. Benz, H.J. Wilke, A. Ignatius, M.A. Barbosa, R.M. Goncalves, C. Neidlinger-Wilke, A Degenerative/Proinflammatory Intervertebral Disc Organ Culture: An Ex Vivo Model for Anti-inflammatory Drug and Cell Therapy, Tissue Eng Part C-Me, 22 (2016) 8-19. [30] B.G. van Dijk, E. Potier, K. Ito, Long-term culture of bovine nucleus pulposus explants in a native environment, The spine journal : official journal of the North American Spine Society, 13 (2013) 454-463. [31] A. Munin, F. Edwards-Levy, Encapsulation of natural polyphenolic compounds; a review, Pharmaceutics, 3 (2011) 793-829. [32] J.G. Lombardino, J.A. Lowe, The role of the medicinal chemist in drug discovery - Then and now, Nature Reviews Drug Discovery, 3 (2004) 853-862. [33] C.L. Shen, B.J. Smith, D.F. Lo, M.C. Chyu, D.M. Dunn, C.H. Chen, I.S. Kwun, Dietary polyphenols and mechanisms of osteoarthritis, The Journal of nutritional biochemistry, 23 (2012) 1367-1377. [34] S. Ahmed, N. Wang, M. Lalonde, V.M. Goldberg, T.M. Haqqi, Green tea polyphenol epigallocatechin-3-gallate (EGCG) differentially inhibits interleukin-1 beta-induced expression of matrix metalloproteinase-1 and -13 in human chondrocytes, The Journal of pharmacology and experimental therapeutics, 308 (2004) 767-773. [35] N. Akhtar, T.M. Haqqi, Epigallocatechin-3-gallate suppresses the global interleukin-1beta-induced inflammatory response in human chondrocytes, Arthritis research & therapy, 13 (2011) R93. [36] Z. Rasheed, A.N. Anbazhagan, N. Akhtar, S. Ramamurthy, F.R. Voss, T.M. Haqqi, Green tea polyphenol epigallocatechin-3-gallate inhibits advanced glycation end product-induced expression of tumor necrosis factor-alpha and matrix metalloproteinase-13 in human chondrocytes, Arthritis research & therapy, 11 (2009) R71. [37] S. Ahmed, A. Rahman, A. Hasnain, M. Lalonde, V.M. Goldberg, T.M. Haqqi, Green tea polyphenol epigallocatechin-3-gallate inhibits the IL-1 beta-induced activity and expression of cyclooxygenase-2 and nitric oxide synthase-2 in human chondrocytes, Free radical biology & medicine, 33 (2002) 1097-1105. [38] H. Zou, E. Stoppani, D. Volonte, F. Galbiati, Caveolin-1, cellular senescence and age-related diseases, Mechanisms of ageing and development, 132 (2011) 533-542.

165

[39] S. Suzuki, N. Fujita, N. Hosogane, K. Watanabe, K. Ishii, Y. Toyama, K. Takubo, K. Horiuchi, T. Miyamoto, M. Nakamura, M. Matsumoto, Excessive reactive oxygen species are therapeutic targets for intervertebral disc degeneration, Arthritis research & therapy, 17 (2015) 316. [40] A. Dimozi, E. Mavrogonatou, A. Sklirou, D. Kletsas, Oxidative stress inhibits the proliferation, induces premature senescence and promotes a catabolic phenotype in human nucleus pulposus intervertebral disc cells, European cells & materials, 30 (2015) 89-102; discussion 103. [41] S.W. Jeong, J.S. Lee, K.W. Kim, In vitro lifespan and senescence mechanisms of human nucleus pulposus chondrocytes, The spine journal : official journal of the North American Spine Society, 14 (2014) 499-504. [42] H.E. Gruber, J.A. Ingram, D.E. Davis, E.N. Hanley, Jr., Increased cell senescence is associated with decreased cell proliferation in vivo in the degenerating human annulus, The spine journal : official journal of the North American Spine Society, 9 (2009) 210-215. [43] K.W. Kim, H.N. Chung, K.Y. Ha, J.S. Lee, Y.Y. Kim, Senescence mechanisms of nucleus pulposus chondrocytes in human intervertebral discs, The spine journal : official journal of the North American Spine Society, 9 (2009) 658-666. [44] L. Jiang, X. Zhang, X. Zheng, A. Ru, X. Ni, Y. Wu, N. Tian, Y. Huang, E. Xue, X. Wang, H. Xu, Apoptosis, senescence, and autophagy in rat nucleus pulposus cells: Implications for diabetic intervertebral disc degeneration, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, 31 (2013) 692-702. [45] G. Hou, H. Lu, M. Chen, H. Yao, H. Zhao, Oxidative stress participates in age-related changes in rat lumbar intervertebral discs, Archives of gerontology and geriatrics, 59 (2014) 665-669. [46] J.S. Park, J.B. Park, I.J. Park, E.Y. Park, Accelerated premature stress-induced senescence of young annulus fibrosus cells of rats by high glucose-induced oxidative stress, International orthopaedics, 38 (2014) 1311-1320. [47] S. Roberts, E.H. Evans, D. Kletsas, D.C. Jaffray, S.M. Eisenstein, Senescence in human intervertebral discs, European Spine Journal, 15 (2006) S312-S316. [48] H.S. Kim, M.J. Quon, J.A. Kim, New insights into the mechanisms of polyphenols beyond antioxidant properties; lessons from the green tea polyphenol, epigallocatechin 3-gallate, Redox biology, 2 (2014) 187-195. [49] J.S. Park, J.I. Ahn, The Effect of Chondroitinase Abc on Rabbit Intervertebral Disc - Radiological, Histological and Electron-Microscopic Finding, International orthopaedics, 19 (1995) 103-109. [50] R.K. Ponnappan, D.Z. Markova, P.J.D. Antonio, H.B. Murray, A.R. Vaccaro, I.M. Shapiro, D.G. Anderson, T.J. Albert, M.V. Risbud, An organ culture system to model early degenerative changes of the intervertebral disc, Arthritis research & therapy, 13 (2011). [51] C.F. Rodrigues, K. Ascencao, F.A. Silva, B. Sarmento, M.B. Oliveira, J.C. Andrade, Drug-delivery systems of green tea catechins for improved stability and bioavailability, Current medicinal chemistry, 20 (2013) 4744-4757. [52] C. Huo, S.B. Wan, W.H. Lam, L. Li, Z. Wang, K.R. Landis-Piwowar, D. Chen, Q.P. Dou, T.H. Chan, The challenge of developing green tea polyphenols as therapeutic agents, Inflammopharmacology, 16 (2008) 248-252. [53] R.A. Isbrucker, J.A. Edwards, E. Wolz, A. Davidovich, J. Bausch, Safety studies on epigallocatechin gallate (EGCG) preparations. Part 2: Dermal, acute and short-term toxicity studies, Food and Chemical Toxicology, 44 (2006) 636-650. [54] L.G. Gomez-Mascaraque, A. Lopez-Rubio, Protein-based emulsion electrosprayed micro- and submicroparticles for the encapsulation and stabilization of thermosensitive hydrophobic bioactives, Journal of colloid and interface science, 465 (2016) 259-270. [55] G.D. Cramer, S. Darby, Basic and clinical anatomy of the spine, spinal cord, and ANS, Journal of manipulative and physiological therapeutics, 20 (1997) 294-294. [56] S.M. Richardson, A. Mobasheri, A.J. Freemont, J.A. Hoyland, Intervertebral disc biology, degeneration and novel tissue engineering and regenerative medicine therapies, Histology and histopathology, 22 (2007) 1033-1041. [57] J.A. Iorio, A.M. Jakoi, A. Singla, Biomechanics of Degenerative Spinal Disorders, Asian spine journal, 10 (2016) 377-384.

166

[58] Y.C. Huang, J.P.G. Urban, K.D.K. Luk, OPINION Intervertebral disc regeneration: do nutrients lead the way?, Nature Reviews Rheumatology, 10 (2014) 561-566. [59] J.P.G. Urban, S. Roberts, Degeneration of the intervertebral disc, Arthritis research & therapy, 5 (2003) 120-130. [60] F. Galbusera, M. van Rijsbergen, K. Ito, J.M. Huyghe, M. Brayda-Bruno, H.J. Wilke, Ageing and degenerative changes of the intervertebral disc and their impact on spinal flexibility, European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society, 23 Suppl 3 (2014) S324-332. [61] S. Roberts, H. Evans, J. Trivedi, J. Menage, Histology and pathology of the human intervertebral disc, The Journal of bone and joint surgery. American volume, 88 Suppl 2 (2006) 10-14. [62] S.S. Sivan, E. Wachtel, P. Roughley, Structure, function, aging and turnover of aggrecan in the intervertebral disc, Bba-Gen Subjects, 1840 (2014) 3181-3189. [63] D.M. Gilkes, G.L. Semenza, D. Wirtz, Hypoxia and the extracellular matrix: drivers of tumour metastasis, Nature reviews. Cancer, 14 (2014) 430-439. [64] S.S. Sivan, E. Wachtel, E. Tsitron, N. Sakkee, F. van der Ham, J. Degroot, S. Roberts, A. Maroudas, Collagen turnover in normal and degenerate human intervertebral discs as determined by the racemization of aspartic acid, The Journal of biological chemistry, 283 (2008) 8796-8801. [65] N. Bogduk, W. Tynan, A.S. Wilson, The nerve supply to the human lumbar intervertebral discs, J Anat, 132 (1981) 39-56. [66] K. Ito, L. Creemers, Mechanisms of intervertebral disk degeneration/injury and pain: a review, Global Spine J, 3 (2013) 145-152. [67] A.J. Fields, E.C. Liebenberg, J.C. Lotz, Innervation of pathologies in the lumbar vertebral end plate and intervertebral disc, Spine Journal, 14 (2014) 513-521. [68] R. Sztrolovics, M. Alini, P.J. Roughley, J.S. Mort, Aggrecan degradation in human intervertebral disc and articular cartilage, Biochemical Journal, 326 (1997) 235-241. [69] C.L. Le Maitre, A.J. Freemont, J.A. Hoyland, Localization of degradative enzymes and their inhibitors in the degenerate human intervertebral disc, The Journal of pathology, 204 (2004) 47-54. [70] M. Molinos, C.R. Almeida, J. Caldeira, C. Cunha, R.M. Goncalves, M.A. Barbosa, Inflammation in intervertebral disc degeneration and regeneration, J R Soc Interface, 12 (2015). [71] R. Stern, G. Kogan, M.J. Jedrzejas, L. Soltes, The many ways to cleave hyaluronan, Biotechnol Adv, 25 (2007) 537-557. [72] S.S. Sivan, E. Tsitron, E. Wachtel, P. Roughley, N. Sakkee, F. van der Ham, J. Degroot, A. Maroudas, Age-related accumulation of pentosidine in aggrecan and collagen from normal and degenerate human intervertebral discs, The Biochemical journal, 399 (2006) 29-35. [73] S. Illien-Junger, F. Grosjean, D.M. Laudier, H. Vlassara, G.E. Striker, J.C. Iatridis, Combined anti-inflammatory and anti-AGE drug treatments have a protective effect on intervertebral discs in mice with diabetes, PloS one, 8 (2013) e64302. [74] R. Gawri, J. Moir, J. Ouellet, L. Beckman, T. Steffen, P. Roughley, L. Haglund, Physiological Loading Can Restore the Proteoglycan Content in a Model of Early IVD Degeneration, PloS one, 9 (2014). [75] K. Wuertz, K. Godburn, J.J. MacLean, A. Barbir, J.S. Donnelly, P.J. Roughley, M. Alini, J.C. Iatridis, In vivo remodeling of intervertebral discs in response to short- and long-term dynamic compression, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, 27 (2009) 1235-1242. [76] S. Illien-Junger, B. Gantenbein-Ritter, S. Grad, P. Lezuo, S.J. Ferguson, M. Alini, K. Ito, The combined effects of limited nutrition and high-frequency loading on intervertebral discs with endplates, Spine, 35 (2010) 1744-1752. [77] A.J. Michalek, J.C. Iatridis, Height and torsional stiffness are most sensitive to annular injury in large animal intervertebral discs, Spine Journal, 12 (2012) 425-432. [78] J.P. Urban, S. Smith, J.C. Fairbank, Nutrition of the intervertebral disc, Spine, 29 (2004) 2700-2709.

167

[79] A.L. Binch, A.A. Cole, L.M. Breakwell, A.L. Michael, N. Chiverton, A.K. Cross, C.L. Le Maitre, Expression and regulation of neurotrophic and angiogenic factors during human intervertebral disc degeneration, Arthritis research & therapy, 16 (2014) 416. [80] M.V. Risbud, I.M. Shapiro, Role of cytokines in intervertebral disc degeneration: pain and disc content, Nature reviews. Rheumatology, 10 (2014) 44-56. [81] H. Oda, H. Matsuzaki, Y. Tokuhashi, K. Wakabayashi, Y. Uematsu, M. Iwahashi, Degeneration of intervertebral discs due to smoking: experimental assessment in a rat-smoking model, Journal of Orthopaedic Science, 9 (2004) 135-141. [82] C.L. Le Maitre, A.J. Freemont, J.A. Hoyland, The role of interleukin-1 in the pathogenesis of human intervertebral disc degeneration, Arthritis research & therapy, 7 (2005) R732-745. [83] C.L. Le Maitre, J.A. Hoyland, A.J. Freemont, Catabolic cytokine expression in degenerate and herniated human intervertebral discs: IL-1 beta and TNF alpha expression profile, Arthritis research & therapy, 9 (2007). [84] C.L. Le Maitre, J.A. Hoyland, A.J. Freemont, Catabolic cytokine expression in degenerate and herniated human intervertebral discs: IL-1beta and TNFalpha expression profile, Arthritis research & therapy, 9 (2007) R77. [85] A.J. Freemont, J.A. Hoyland, A. Rajpura, R.J. Byers, C. Bartley, M. Jeziorska, M. Knight, R. Ross, J. O'Brien, J. Sutcliffe, C. LeMaitre, Matrix degradation by chondrocytes in the intervertebral disc is mediated by interleukin-1, Rheumatology, 40 (2001) 75-75. [86] N.V. Vo, R.A. Hartman, T. Yurube, L.J. Jacobs, G.A. Sowa, J.D. Kang, Expression and regulation of metalloproteinases and their inhibitors in intervertebral disc aging and degeneration, The spine journal : official journal of the North American Spine Society, 13 (2013) 331-341. [87] W.J. Wang, X.H. Yu, C. Wang, W. Yang, W.S. He, S.J. Zhang, Y.G. Yan, J. Zhang, MMPs and ADAMTSs in intervertebral disc degeneration, Clinica chimica acta; international journal of clinical chemistry, 448 (2015) 238-246. [88] L.H. Karin Wuertz, Inflammatory Mediators in Intervertebral Disk Degeneration and Discogenic Pain, Global Spine Journal, 2013. [89] L. Hayflick, P.S. Moorhead, The serial cultivation of human diploid cell strains, Experimental cell research, 25 (1961) 585-621. [90] D. Munoz-Espin, M. Serrano, Cellular senescence: from physiology to pathology, Nat Rev Mol Cell Bio, 15 (2014) 482-+. [91] J.M. van Deursen, The role of senescent cells in ageing, Nature, 509 (2014) 439-446. [92] L.A. Nasto, A.R. Robinson, K. Ngo, C.L. Clauson, Q. Dong, C. St Croix, G. Sowa, E. Pola, P.D. Robbins, J. Kang, L.J. Niedernhofer, P. Wipf, N.V. Vo, Mitochondrial-derived reactive oxygen species (ROS) play a causal role in aging-related intervertebral disc degeneration, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, 31 (2013) 1150-1157. [93] L.A. Nasto, D. Wang, A.R. Robinson, C.L. Clauson, K. Ngo, Q. Dong, P. Roughley, M. Epperly, S.M. Huq, E. Pola, G. Sowa, P.D. Robbins, J. Kang, L.J. Niedernhofer, N.V. Vo, Genotoxic stress accelerates age-associated degenerative changes in intervertebral discs, Mechanisms of ageing and development, 134 (2013) 35-42. [94] L. Poveda, M. Hottiger, N. Boos, K. Wuertz, Peroxynitrite induces gene expression in intervertebral disc cells, Spine, 34 (2009) 1127-1133. [95] Y.J. Kuo, L.C. Wu, J.S. Sun, M.H. Chen, M.G. Sun, Y.H. Tsuang, Mechanical stress-induced apoptosis of nucleus pulposus cells: an in vitro and in vivo rat model, Journal of orthopaedic science : official journal of the Japanese Orthopaedic Association, 19 (2014) 313-322. [96] F. Ding, Z.W. Shao, L.M. Xiong, Cell death in intervertebral disc degeneration, Apoptosis : an international journal on programmed cell death, 18 (2013) 777-785. [97] S.E. Espinoza, H.F. Guo, N. Fedarko, A. DeZern, L.P. Fried, Q.L. Xue, S. Leng, B. Beamer, J.D. Walston, Glutathione peroxidase enzyme activity in aging, J Gerontol a-Biol, 63 (2008) 505-509. [98] V.H. Smith, Vitamin C deficiency is an under-diagnosed contributor to degenerative disc disease in the elderly, Medical hypotheses, 74 (2010) 695-697. [99] W. Wu, X.L. Zhang, X.Q. Hu, X.Y. Wang, L.J. Sun, X.H. Zheng, L.B. Jiang, X. Ni, C. Xu, N.F. Tian, S.P. Zhu, H.Z. Xu, Lactate Down-Regulates Matrix Systhesis and Promotes Apoptosis and Autophagy in Rat Nucleus Pulposus Cells, Journal of Orthopaedic Research, 32 (2014) 253-261.

168

[100] J.E. Mayer, J.C. Iatridis, D. Chan, S.A. Qureshi, O. Gottesman, A.C. Hecht, Genetic polymorphisms associated with intervertebral disc degeneration, Spine Journal, 13 (2013) 299-317. [101] Y.C. Li, J.J. Zhu, C.H. Gao, B.G. Peng, Vitamin D Receptor (VDR) Genetic Polymorphisms Associated with Intervertebral Disc Degeneration, J Genet Genomics, 42 (2015) 135-140. [102] J.D. Kang, H.I. Georgescu, L. McIntyre-Larkin, M. Stefanovic-Racic, W.F. Donaldson, 3rd, C.H. Evans, Herniated lumbar intervertebral discs spontaneously produce matrix metalloproteinases, nitric oxide, interleukin-6, and prostaglandin E2, Spine, 21 (1996) 271-277. [103] J.D. Kang, M. Stefanovic-Racic, L.A. McIntyre, H.I. Georgescu, C.H. Evans, Toward a biochemical understanding of human intervertebral disc degeneration and herniation. Contributions of nitric oxide, interleukins, prostaglandin E2, and matrix metalloproteinases, Spine, 22 (1997) 1065-1073. [104] S.M. Richardson, P. Doyle, B.M. Minogue, K. Gnanalingham, J.A. Hoyland, Increased expression of matrix metalloproteinase-10, nerve growth factor and substance P in the painful degenerate intervertebral disc, Arthritis research & therapy, 11 (2009) R126. [105] C.K. Kepler, D.Z. Markova, A.S. Hilibrand, A.R. Vaccaro, M.V. Risbud, T.J. Albert, D.G. Anderson, Substance P stimulates production of inflammatory cytokines in human disc cells, Spine, 38 (2013) E1291-1299. [106] D. Purmessur, A.J. Freemont, J.A. Hoyland, Expression and regulation of neurotrophins in the nondegenerate and degenerate human intervertebral disc, Arthritis research & therapy, 10 (2008). [107] M.K. Jha, I.K. Lee, K. Suk, Metabolic reprogramming by the pyruvate dehydrogenase kinase-lactic acid axis: Linking metabolism and diverse neuropathophysiologies, Neuroscience and biobehavioral reviews, 68 (2016) 1-19. [108] E. Krock, D.H. Rosenzweig, A.J. Chabot-Dore, P. Jarzem, M.H. Weber, J.A. Ouellet, L.S. Stone, L. Haglund, Painful, degenerating intervertebral discs up-regulate neurite sprouting and CGRP through nociceptive factors, Journal of cellular and molecular medicine, 18 (2014) 1213-1225. [109] S.J. Ferguson, K. Ito, L.P. Nolte, Fluid flow and convective transport of solutes within the intervertebral disc, Journal of biomechanics, 37 (2004) 213-221. [110] K. Wuertz, N. Vo, D. Kletsas, N. Boos, Inflammatory and catabolic signalling in intervertebral discs: the roles of NF-kappaB and MAP kinases, European cells & materials, 23 (2012) 103-119; discussion 119-120. [111] D. Haschtmann, J.V. Stoyanov, P. Gedet, S.J. Ferguson, Vertebral endplate trauma induces disc cell apoptosis and promotes organ degeneration in vitro, European Spine Journal, 17 (2008) 289-299. [112] S. Dudli, D.B. Boffa, S.J. Ferguson, D. Haschtmann, Leukocytes Enhance Inflammatory and Catabolic Degenerative Changes in the Intervertebral Disc After Endplate Fracture In Vitro Without Infiltrating the Disc, Spine, 40 (2015) 1799-1806. [113] Z. Sun, Z.Y. Wan, Y.S. Guo, H.Q. Wang, Z.J. Luo, FasL on human nucleus pulposus cells prevents angiogenesis in the disc by inducing Fas-mediated apoptosis of vascular endothelial cells, International journal of clinical and experimental pathology, 6 (2013) 2376-2385. [114] Z.H. Liu, Z. Sun, H.Q. Wang, J. Ge, T.S. Jiang, Y.F. Chen, Y. Ma, C. Wang, S. Hu, D. Samartzis, Z.J. Luo, FasL Expression on Human Nucleus Pulposus Cells Contributes to the Immune Privilege of Intervertebral Disc by Interacting with Immunocytes, Int J Med Sci, 10 (2013) 1053-1060. [115] J.N. Sharma, A. Al-Omran, S.S. Parvathy, Role of nitric oxide in inflammatory diseases, Inflammopharmacology, 15 (2007) 252-259. [116] K.R. Keshari, J.C. Lotz, T.M. Link, S. Hu, S. Majumdar, J. Kurhanewicz, Lactic acid and Proteoglycans as metabolic markers for discogenic back pain, Spine, 33 (2008) 312-317. [117] M.H. Rahman, M.K. Jha, J.H. Kim, Y. Nam, M.G. Lee, Y. Go, R.A. Harris, D.H. Park, H. Kook, I.K. Lee, K. Suk, Pyruvate Dehydrogenase Kinase-mediated Glycolytic Metabolic Shift in the Dorsal Root Ganglion Drives Painful Diabetic Neuropathy, Journal of Biological Chemistry, 291 (2016) 6011-6025. [118] A. Kawabata, Prostaglandin E2 and pain--an update, Biological & pharmaceutical bulletin, 34 (2011) 1170-1173. [119] A. Hiyama, K. Yokoyama, T. Nukaga, D. Sakai, J. Mochida, Response to tumor necrosis factor-alpha mediated inflammation involving activation of prostaglandin E2 and Wnt

169

signaling in nucleus pulposus cells, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, 33 (2015) 1756-1768. [120] S. Kimura, Y. Sakuma, M. Suzuki, S. Orita, K. Yamauchi, G. Inoue, Y. Aoki, T. Ishikawa, M. Miyagi, H. Kamoda, G. Kubota, Y. Oikawa, K. Inage, T. Sainoh, J. Sato, J. Nakamura, T. Toyone, K. Takahashi, S. Ohtori, Evaluation of Pain Behavior and Calcitonin Gene-Related Peptide Immunoreactive Sensory Nerve Fibers in the Spinal Dorsal Horn After Sciatic Nerve Compression and Application of Nucleus Pulposus in Rats, Spine, 39 (2014) 455-462. [121] E. Krock, J.B. Currie, M.H. Weber, J.A. Ouellet, L.S. Stone, D.H. Rosenzweig, L. Haglund, Nerve Growth Factor Is Regulated by Toll-Like Receptor 2 in Human Intervertebral Discs, Journal of Biological Chemistry, 291 (2016) 3541-3551. [122] D. Mulleman, S. Mammou, I. Griffoul, H. Watier, P. Goupille, Pathophysiology of disk-related sciatica. I.--Evidence supporting a chemical component, Joint, bone, spine : revue du rhumatisme, 73 (2006) 151-158. [123] S.Y. Lee, T.H. Kim, J.K. Oh, S.J. Lee, M.S. Park, Lumbar Stenosis: A Recent Update by Review of Literature, Asian spine journal, 9 (2015) 818-828. [124] M.W. Hasz, Diagnostic testing for degenerative disc disease, Advances in orthopedics, 2012 (2012) 413913. [125] D.S. Kreiner, S.W. Hwang, J.E. Easa, D.K. Resnick, J.L. Baisden, S. Bess, C.H. Cho, M.J. DePalma, P. Dougherty, 2nd, R. Fernand, G. Ghiselli, A.S. Hanna, T. Lamer, A.J. Lisi, D.J. Mazanec, R.J. Meagher, R.C. Nucci, R.D. Patel, J.N. Sembrano, A.K. Sharma, J.T. Summers, C.K. Taleghani, W.L. Tontz, Jr., J.F. Toton, S. North American Spine, An evidence-based clinical guideline for the diagnosis and treatment of lumbar disc herniation with radiculopathy, The spine journal : official journal of the North American Spine Society, 14 (2014) 180-191. [126] Y. Peng, F.J. Lv, Symptomatic versus Asymptomatic Intervertebral Disc Degeneration: Is Inflammation the Key?, Crit Rev Eukar Gene, 25 (2015) 13-21. [127] J.D. Kang, Does a needle puncture into the annulus fibrosus cause disc degeneration?, Spine Journal, 10 (2010) 1106-1107. [128] S. Baliga, K. Treon, N.J. Craig, Low Back Pain: Current Surgical Approaches, Asian spine journal, 9 (2015) 645-657. [129] J. Steele, S. Bruce-Low, D. Smith, N. Osborne, A. Thorkeldsen, Can specific loading through exercise impart healing or regeneration of the intervertebral disc?, Spine Journal, 15 (2015) 2117-2121. [130] J.A. Hayden, M.V. van Tulder, A. Malmivaara, B.W. Koes, Exercise therapy for treatment of non-specific low back pain, Cochrane Db Syst Rev, (2005). [131] M.A. Adams, M. Stefanakis, P. Dolan, Healing of a painful intervertebral disc should not be confused with reversing disc degeneration: Implications for physical therapies for discogenic back pain, Clin Biomech, 25 (2010) 961-971. [132] B.A. Walter, C.L. Korecki, D. Purmessur, P.J. Roughley, A.J. Michalek, J.C. Iatridis, Complex loading affects intervertebral disc mechanics and biology, Osteoarthr Cartilage, 19 (2011) 1011-1018. [133] F. Zaina, C. Tomkins-Lane, E. Carragee, S. Negrini, Surgical versus non-surgical treatment for lumbar spinal stenosis, Cochrane Db Syst Rev, (2016). [134] R. Baron, A. Binder, N. Attal, R. Casale, A.H. Dickenson, R.D. Treede, Neuropathic low back pain in clinical practice, European journal of pain, 20 (2016) 861-873. [135] N. Katz, D.B. Rodgers, D. Krupa, A. Reicin, Onset of pain relief with rofecoxib in chronic low back pain: results of two four-week, randomized, placebo-controlled trials, Current medical research and opinion, 20 (2004) 651-658. [136] C.L. Romano, D. Romano, C. Bonora, G. Mineo, Pregabalin, celecoxib, and their combination for treatment of chronic low-back pain, Journal of orthopaedics and traumatology : official journal of the Italian Society of Orthopaedics and Traumatology, 10 (2009) 185-191. [137] P.D. Roelofs, R.A. Deyo, B.W. Koes, R.J. Scholten, M.W. van Tulder, Nonsteroidal anti-inflammatory drugs for low back pain: an updated Cochrane review, Spine, 33 (2008) 1766-1774. [138] P.M. Kearney, C. Baigent, J. Godwin, H. Halls, J.R. Emberson, C. Patrono, Do selective cyclo-oxygenase-2 inhibitors and traditional non-steroidal anti-inflammatory drugs increase the risk of atherothrombosis? Meta-analysis of randomised trials, Bmj, 332 (2006) 1302-1308.

170

[139] P. American Geriatrics Society Panel on Pharmacological Management of Persistent Pain in Older, Pharmacological management of persistent pain in older persons, Journal of the American Geriatrics Society, 57 (2009) 1331-1346. [140] L.E. Chaparro, A.D. Furlan, A. Deshpande, A. Mailis-Gagnon, S. Atlas, D.C. Turk, Opioids compared to placebo or other treatments for chronic low-back pain, The Cochrane database of systematic reviews, (2013) CD004959. [141] J. Jage, Opioid tolerance and dependence -- do they matter?, European journal of pain, 9 (2005) 157-162. [142] R.A. Deyo, M. Von Korff, D. Duhrkoop, Opioids for low back pain, Bmj, 350 (2015) g6380. [143] P.G. Fine, G. Mahajan, M.L. McPherson, Long-acting opioids and short-acting opioids: appropriate use in chronic pain management, Pain medicine, 10 Suppl 2 (2009) S79-88. [144] K.D. van den Eerenbeemt, R.W. Ostelo, B.J. van Royen, W.C. Peul, M.W. van Tulder, Total disc replacement surgery for symptomatic degenerative lumbar disc disease: a systematic review of the literature, European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society, 19 (2010) 1262-1280. [145] R.A. Deyo, J.D. Loeser, S.J. Bigos, Herniated lumbar intervertebral disk, Annals of internal medicine, 112 (1990) 598-603. [146] B.J. Shin, Risk factors for recurrent lumbar disc herniations, Asian spine journal, 8 (2014) 211-215. [147] F. Mwale, Molecular therapy for disk degeneration and pain, Global Spine J, 3 (2013) 185-192. [148] S.A. de Vries, M. van Doeselaar, B.P. Meij, M.A. Tryfonidou, K. Ito, The Stimulatory Effect of Notochordal Cell-Conditioned Medium in a Nucleus Pulposus Explant Culture, Tissue engineering. Part A, 22 (2016) 103-110. [149] F.C. Bach, S.A. de Vries, A. Krouwels, L.B. Creemers, K. Ito, B.P. Meij, M.A. Tryfonidou, The species-specific regenerative effects of notochordal cell-conditioned medium on chondrocyte-like cells derived from degenerated human intervertebral discs, European cells & materials, 30 (2015) 132-146; discussion 146-137. [150] K. Sawamura, T. Ikeda, M. Nagae, S. Okamoto, Y. Mikami, H. Hase, K. Ikoma, T. Yamada, H. Sakamoto, K. Matsuda, Y. Tabata, M. Kawata, T. Kubo, Characterization of in vivo effects of platelet-rich plasma and biodegradable gelatin hydrogel microspheres on degenerated intervertebral discs, Tissue engineering. Part A, 15 (2009) 3719-3727. [151] R. Gawri, T. Shiba, R. Pilliar, R. Kandel, Inorganic polyphosphates enhances nucleus pulposus tissue formation in vitro, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, (2016). [152] K.L.E. Phillips, N. Jordan-Mahy, M.J.H. Nicklin, C.L. Le Maitre, Interleukin-1 receptor antagonist deficient mice provide insights into pathogenesis of human intervertebral disc degeneration, Annals of the rheumatic diseases, 72 (2013) 1860-1867. [153] C.L. Le Maitre, J.A. Hoyland, A.J. Freemont, Interleukin-1 receptor antagonist delivered directly and by gene therapy inhibits matrix degradation in the intact degenerate human intervertebral disc: an in situ zymographic and gene therapy study, Arthritis research & therapy, 9 (2007). [154] S.M. Sinclair, M.F. Shamji, J. Chen, L.F. Jing, W.J. Richardson, C.R. Brown, R.D. Fitch, L.A. Setton, Attenuation of Inflammatory Events in Human Intervertebral Disc Cells With a Tumor Necrosis Factor Antagonist, Spine, 36 (2011) 1190-1196. [155] S.B. Blanquer, D.W. Grijpma, A.A. Poot, Delivery systems for the treatment of degenerated intervertebral discs, Advanced drug delivery reviews, 84 (2015) 172-187. [156] R.K. Studer, A.M. Aboka, L.G. Gilbertson, H. Georgescu, G. Sowa, N. Vo, J.D. Kang, p38 MAPK inhibition in nucleus pulposus cells: a potential target for treating intervertebral disc degeneration, Spine, 32 (2007) 2827-2833. [157] R.K. Studer, A.M. Aboka, L.G. Gilbertson, H. Georgescu, G. Sowa, N. Vo, J.D. Kang, P38 MAPK inhibition in nucleus pulposus cells - A potential target for treating intervertebral disc degeneration, Spine, 32 (2007) 2827-2833.

171

[158] J.J. Park, H.J. Moon, J.H. Park, T.H. Kwon, Y.K. Park, J.H. Kim, Induction of proinflammatory cytokine production in intervertebral disc cells by macrophage-like THP-1 cells requires mitogen-activated protein kinase activity, J Neurosurg-Spine, 24 (2016) 167-175. [159] W. Yuan, Z.H. Wu, G.X. Qiu, B. Yu, J. Chen, Y.P. Wang, Extracellular Signal-Regulated Kinase Inhibition Modulates Rat Annulus Fibrosus Cell Response to Interleukin-1, Spine, 38 (2013) E1075-E1081. [160] S. Zhongyi, Z. Sai, L. Chao, T. Jiwei, Effects of nuclear factor kappa B signaling pathway in human intervertebral disc degeneration, Spine, 40 (2015) 224-232. [161] T. Ohba, H. Haro, T. Ando, M. Wako, F. Suenaga, Y. Aso, K. Koyama, Y. Hamada, A. Nakao, TNF-alpha-induced NF-kappaB signaling reverses age-related declines in VEGF induction and angiogenic activity in intervertebral disc tissues, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, 27 (2009) 229-235. [162] N. Guo, Z. Peng, MG132, a proteasome inhibitor, induces apoptosis in tumor cells, Asia-Pacific journal of clinical oncology, 9 (2013) 6-11. [163] J. Lee, M.H. Rhee, E. Kim, J.Y. Cho, BAY 11-7082 Is a Broad-Spectrum Inhibitor with Anti-Inflammatory Activity against Multiple Targets, Mediators of inflammation, (2012). [164] S. Strickson, D.G. Campbell, C.H. Emmerich, A. Knebel, L. Plater, M.S. Ritorto, N. Shpiro, P. Cohen, The anti-inflammatory drug BAY 11-7082 suppresses the MyD88-dependent signalling network by targeting the ubiquitin system, The Biochemical journal, 451 (2013) 427-437. [165] M. Sato, K. Inage, Y. Sakuma, J. Sato, S. Orita, K. Yamauchi, Y. Eguchi, N. Ochiai, K. Kuniyoshi, Y. Aoki, J. Nakamura, M. Miyagi, M. Suzuki, G. Kubota, T. Sainoh, K. Fujimoto, Y. Shiga, K. Abe, H. Kanamoto, G. Inoue, K. Takahashi, S. Ohtori, Anti-RANKL antibodies decrease CGRP expression in dorsal root ganglion neurons innervating injured lumbar intervertebral discs in rats, European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society, 24 (2015) 2017-2022. [166] S. Suzuki, N. Fujita, N. Hosogane, K. Watanabe, K. Ishii, Y. Toyama, K. Takubo, K. Horiuchi, T. Miyamoto, M. Nakamura, M. Matsumoto, Excessive reactive oxygen species are therapeutic targets for intervertebral disc degeneration, Arthritis research & therapy, 17 (2015). [167] Y.H. Cheng, S.H. Yang, K.C. Yang, M.P. Chen, F.H. Lin, The effects of ferulic acid on nucleus pulposus cells under hydrogen peroxide-induced oxidative stress, Process Biochem, 46 (2011) 1670-1677. [168] T.T. Tsai, N.Y. Ho, Y.T. Lin, P.L. Lai, T.S. Fu, C.C. Niu, L.H. Chen, W.J. Chen, J.H. Pang, Advanced glycation end products in degenerative nucleus pulposus with diabetes, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, 32 (2014) 238-244. [169] T. Sainoh, Y. Sakuma, M. Miyagi, S. Orita, K. Yamauchi, G. Inoue, H. Kamoda, T. Ishikawa, M. Suzuki, G. Kubota, Y. Oikawa, K. Inage, J. Sato, J. Nakamura, Y. Aoki, M. Takaso, T. Toyone, K. Takahashi, S. Ohtori, Efficacy of Anti-Nerve Growth Factor Therapy for Discogenic Neck Pain in Rats, Spine, 39 (2014) E757-E762. [170] K. Wuertz, L. Quero, M. Sekiguchi, M. Klawitter, A. Nerlich, S. Konno, S. Kikuchi, N. Boos, The red wine polyphenol resveratrol shows promising potential for the treatment of nucleus pulposus-mediated pain in vitro and in vivo, Spine, 36 (2011) E1373-1384. [171] M. Klawitter, L. Quero, J. Klasen, A.N. Gloess, B. Klopprogge, O. Hausmann, N. Boos, K. Wuertz, Curcuma DMSO extracts and curcumin exhibit an anti-inflammatory and anti-catabolic effect on human intervertebral disc cells, possibly by influencing TLR2 expression and JNK activity, J Inflamm (Lond), 9 (2012) 29. [172] Q.L. He, Y. Chen, J. Qin, S.L. Mo, M. Wei, J.J. Zhang, M.N. Li, X.N. Zou, S.F. Zhou, X.W. Chen, L.B. Sun, Osthole, a herbal compound, alleviates nucleus pulposus-evoked nociceptive responses through the suppression of overexpression of acid-sensing ion channel 3 (ASIC3) in rat dorsal root ganglion, Medical science monitor : international medical journal of experimental and clinical research, 18 (2012) BR229-236. [173] M. Klawitter, L. Quero, J. Klasen, T. Liebscher, A. Nerlich, N. Boos, K. Wuertz, Triptolide exhibits anti-inflammatory, anti-catabolic as well as anabolic effects and suppresses TLR expression and MAPK activity in IL-1beta treated human intervertebral disc cells, European spine journal : official publication of the European Spine Society, the European Spinal Deformity

172

Society, and the European Section of the Cervical Spine Research Society, 21 Suppl 6 (2012) S850-859. [174] J.A. Vita, Polyphenols and cardiovascular disease: effects on endothelial and platelet function, The American journal of clinical nutrition, 81 (2005) 292S-297S. [175] H.J. Meisel, T. Ganey, W.C. Hutton, J. Libera, Y. Minkus, O. Alasevic, Clinical experience in cell-based therapeutics: intervention and outcome, European Spine Journal, 15 (2006) S397-S405. [176] G. Vadala, F. Russo, A. Di Martino, V. Denaro, Intervertebral disc regeneration: from the degenerative cascade to molecular therapy and tissue engineering, Journal of tissue engineering and regenerative medicine, 9 (2015) 679-690. [177] D. Sakai, S. Grad, Advancing the cellular and molecular therapy for intervertebral disc disease, Advanced drug delivery reviews, (2014). [178] D. Sakai, J. Mochida, Y. Yamamoto, T. Nomura, M. Okuma, K. Nishimura, T. Nakai, K. Ando, T. Hotta, Transplantation of mesenchymal stem cells embedded in Atelocollagen gel to the intervertebral disc: a potential therapeutic model for disc degeneration, Biomaterials, 24 (2003) 3531-3541. [179] A. Bertolo, T. Thiede, N. Aebli, M. Baur, S.J. Ferguson, J.V. Stoyanov, Human mesenchymal stem cell co-culture modulates the immunological properties of human intervertebral disc tissue fragments in vitro, European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society, 20 (2011) 592-603. [180] X. Li, J.P. Lee, G. Balian, D. Greg Anderson, Modulation of chondrocytic properties of fat-derived mesenchymal cells in co-cultures with nucleus pulposus, Connective tissue research, 46 (2005) 75-82. [181] G. Vadala, S. Sobajima, J.Y. Lee, J. Huard, V. Denaro, J.D. Kang, L.G. Gilbertson, In vitro interaction between muscle-derived stem cells and nucleus pulposus cells, Spine Journal, 8 (2008) 804-809. [182] M.V. Risbud, T.J. Albert, A. Guttapalli, E.J. Vresilovic, A.S. Hillibrand, A.R. Vaccaro, I.M. Shapiro, Differentiation of mesenchymal stem cells towards a nucleus pulposus-like phenotype in vitro: implications for cell-based transplantation therapy, Spine, 29 (2004) 2627-2632. [183] S. Vedicherla, C.T. Buckley, Cell-based therapies for intervertebral disc and cartilage regeneration - current concepts, parallels and perspectives, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, (2016). [184] G. Vadala, G. Sowa, M. Hubert, L.G. Gilbertson, V. Denaro, J.D. Kang, Mesenchymal stem cells injection in degenerated intervertebral disc: cell leakage may induce osteophyte formation, Journal of tissue engineering and regenerative medicine, 6 (2012) 348-355. [185] J.V. Stoyanov, B. Gantenbein-Ritter, A. Bertolo, N. Aebli, M. Baur, M. Alini, S. Grad, Role of hypoxia and growth and differentiation factor-5 on differentiation of human mesenchymal stem cells towards intervertebral nucleus pulposus-like cells, European cells & materials, 21 (2011) 533-547. [186] J. Zeckser, M. Wolff, J. Tucker, J. Goodwin, Multipotent Mesenchymal Stem Cell Treatment for Discogenic Low Back Pain and Disc Degeneration, Stem Cells Int, 2016 (2016) 3908389. [187] S.H. Moon, K. Nishida, L.G. Gilbertson, H.M. Lee, H. Kim, R.A. Hall, P.D. Robbins, J.D. Kang, Biologic response of human intervertebral disc cells to gene therapy cocktail, Spine, 33 (2008) 1850-1855. [188] S. Ren, Y. Liu, J. Ma, Y. Liu, Z. Diao, D. Yang, X. Zhang, Y. Xi, Y. Hu, Treatment of rabbit intervertebral disc degeneration with co-transfection by adeno-associated virus-mediated SOX9 and osteogenic protein-1 double genes in vivo, International journal of molecular medicine, 32 (2013) 1063-1068. [189] G. Fontana, E. See, A. Pandit, Current trends in biologics delivery to restore intervertebral disc anabolism, Advanced drug delivery reviews, 84 (2015) 146-158. [190] C.L. Le Maitre, A.J. Freemont, J.A. Hoyland, A preliminary in vitro study into the use of IL-1Ra gene therapy for the inhibition of intervertebral disc degeneration, Int J Exp Pathol, 87 (2006) 17-28. [191] H.Y. Yang, R.J. van Ee, K. Timmer, E.G.M. Craenmehr, J.H. Huang, F.C. Oner, W.J.A. Dhert, A.H.M. Kragten, N. Willems, G.C.M. Grinwis, M.A. Tryfonidou, N.E. Papen-Botterhuis, L.B.

173

Creemers, A novel injectable thermoresponsive and cytocompatible gel of poly(N-isopropylacrylamide) with layered double hydroxides facilitates siRNA delivery into chondrocytes in 3D culture, Acta biomaterialia, 23 (2015) 214-228. [192] M. Nagae, T. Ikeda, Y. Mikami, H. Hase, H. Ozawa, K. Matsuda, H. Sakamoto, Y. Tabata, M. Kawata, T. Kubo, Intervertebral disc regeneration using platelet-rich plasma and biodegradable gelatin hydrogel microspheres, Tissue engineering, 13 (2007) 147-158. [193] Y.H. Cheng, S.H. Yang, C.C. Liu, A. Gefen, F.H. Lin, Thermosensitive hydrogel made of ferulic acid-gelatin and chitosan glycerophosphate, Carbohydrate polymers, 92 (2013) 1512-1519. [194] R. Kandel, S. Roberts, J.P. Urban, Tissue engineering and the intervertebral disc: the challenges, European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society, 17 Suppl 4 (2008) 480-491. [195] G. Vadala, P. Mozetic, A. Rainer, M. Centola, M. Loppini, M. Trombetta, V. Denaro, Bioactive electrospun scaffold for annulus fibrosus repair and regeneration, European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society, 21 Suppl 1 (2012) S20-26. [196] M. Peroglio, D. Eglin, L.M. Benneker, M. Alini, S. Grad, Thermoreversible hyaluronan-based hydrogel supports in vitro and ex vivo disc-like differentiation of human mesenchymal stem cells, The spine journal : official journal of the North American Spine Society, 13 (2013) 1627-1639. [197] N. Willems, H.Y. Yang, M.L. Langelaan, A.R. Tellegen, G.C. Grinwis, H.J. Kranenburg, F.M. Riemers, S.G. Plomp, E.G. Craenmehr, W.J. Dhert, N.E. Papen-Botterhuis, B.P. Meij, L.B. Creemers, M.A. Tryfonidou, Biocompatibility and intradiscal application of a thermoreversible celecoxib-loaded poly-N-isopropylacrylamide MgFe-layered double hydroxide hydrogel in a canine model, Arthritis research & therapy, 17 (2015) 214. [198] R.J. Seelbach, P. Fransen, D. Pulido, M. D'Este, F. Duttenhoefer, S. Sauerbier, T. Freiman, P. Niemeyer, F. Albericio, M. Alini, M. Royo, A. Mata, D. Eglin, Injectable Hyaluronan Hydrogels with Peptide-Binding Dendrimers Modulate the Controlled Release of BMP-2 and TGF-beta 1, Macromol Biosci, 15 (2015) 1035-1044. [199] M. Peroglio, D. Eglin, L.M. Benneker, M. Alini, S. Grad, Thermoreversible hyaluronan-based hydrogel supports in vitro and ex vivo disc-like differentiation of human mesenchymal stem cells, Spine Journal, 13 (2013) 1627-1639. [200] D. Mortisen, M. Peroglio, M. Alini, D. Eglin, Tailoring Thermoreversible Hyaluronan Hydrogels by "Click" Chemistry and RAFT Polymerization for Cell and Drug Therapy, Biomacromolecules, 11 (2010) 1261-1272. [201] B. van Dijk, E. Potier, M. van Dijk, M. Langelaan, N. Papen-Botterhuis, K. Ito, Reduced tonicity stimulates an inflammatory response in nucleus pulposus tissue that can be limited by a COX-2-specific inhibitor, Journal of Orthopaedic Research, 33 (2015) 1724-1731. [202] S. Dagenais, J. Caro, S. Haldeman, A systematic review of low back pain cost of illness studies in the United States and internationally, The spine journal : official journal of the North American Spine Society, 8 (2008) 8-20. [203] S. Dagenais, S. Haldeman, Commentary: Laboring to understand the economic impact of spinal disorders, The spine journal : official journal of the North American Spine Society, 12 (2012) 1119-1121. [204] B.I. Martin, R.A. Deyo, S.K. Mirza, J.A. Turner, B.A. Comstock, W. Hollingworth, S.D. Sullivan, Expenditures and health status among adults with back and neck problems, JAMA : the journal of the American Medical Association, 299 (2008) 656-664. [205] A.J. Pockert, S.M. Richardson, C.L. Le Maitre, M. Lyon, J.A. Deakin, D.J. Buttle, A.J. Freemont, J.A. Hoyland, Modified expression of the ADAMTS enzymes and tissue inhibitor of metalloproteinases 3 during human intervertebral disc degeneration, Arthritis and rheumatism, 60 (2009) 482-491. [206] S.R. Bibby, D.A. Jones, R.M. Ripley, J.P. Urban, Metabolism of the intervertebral disc: effects of low levels of oxygen, glucose, and pH on rates of energy metabolism of bovine nucleus pulposus cells, Spine, 30 (2005) 487-496. [207] S. Roberts, E.H. Evans, D. Kletsas, D.C. Jaffray, S.M. Eisenstein, Senescence in human intervertebral discs, European spine journal : official publication of the European Spine Society,

174

the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society, 15 Suppl 3 (2006) S312-316. [208] M.F. Shamji, L.A. Setton, W. Jarvis, S. So, J. Chen, L. Jing, R. Bullock, R.E. Isaacs, C. Brown, W.J. Richardson, Proinflammatory cytokine expression profile in degenerated and herniated human intervertebral disc tissues, Arthritis and rheumatism, 62 (2010) 1974-1982. [209] Z. Sun, M. Zhang, X.H. Zhao, Z.H. Liu, Y. Gao, D. Samartzis, H.Q. Wang, Z.J. Luo, Immune cascades in human intervertebral disc: the pros and cons, International journal of clinical and experimental pathology, 6 (2013) 1009-1014. [210] H. Brisby, Pathology and possible mechanisms of nervous system response to disc degeneration, The Journal of bone and joint surgery. American volume, 88 Suppl 2 (2006) 68-71. [211] A. Geiss, K. Larsson, B. Rydevik, I. Takahashi, K. Olmarker, Autoimmune properties of nucleus pulposus: an experimental study in pigs, Spine, 32 (2007) 168-173. [212] B. Peng, J. Hao, S. Hou, W. Wu, D. Jiang, X. Fu, Y. Yang, Possible pathogenesis of painful intervertebral disc degeneration, Spine, 31 (2006) 560-566. [213] M. Doita, T. Kanatani, T. Ozaki, N. Matsui, M. Kurosaka, S. Yoshiya, Influence of macrophage infiltration of herniated disc tissue on the production of matrix metalloproteinases leading to disc resorption, Spine, 26 (2001) 1522-1527. [214] M. Gronblad, J. Virri, J. Tolonen, S. Seitsalo, E. Kaapa, J. Kankare, P. Myllynen, E.O. Karaharju, A controlled immunohistochemical study of inflammatory cells in disc herniation tissue, Spine, 19 (1994) 2744-2751. [215] R.D. Bowles, H.H. Gebhard, R. Hartl, L.J. Bonassar, Tissue-engineered intervertebral discs produce new matrix, maintain disc height, and restore biomechanical function to the rodent spine, Proceedings of the National Academy of Sciences of the United States of America, 108 (2011) 13106-13111. [216] T.P. Hedman, H. Saito, C. Vo, S.Y. Chuang, Exogenous cross-linking increases the stability of spinal motion segments, Spine, 31 (2006) E480-485. [217] S.M. Sinclair, M.F. Shamji, J. Chen, L. Jing, W.J. Richardson, C.R. Brown, R.D. Fitch, L.A. Setton, Attenuation of inflammatory events in human intervertebral disc cells with a tumor necrosis factor antagonist, Spine, 36 (2011) 1190-1196. [218] B.N. Singh, S. Shankar, R.K. Srivastava, Green tea catechin, epigallocatechin-3-gallate (EGCG): mechanisms, perspectives and clinical applications, Biochemical pharmacology, 82 (2011) 1807-1821. [219] A.M. Bode, Z. Dong, Epigallocatechin 3-gallate and green tea catechins: United they work, divided they fail, Cancer Prev Res (Phila), 2 (2009) 514-517. [220] S.D. Smid, J.L. Maag, I.F. Musgrave, Dietary polyphenol-derived protection against neurotoxic beta-amyloid protein: from molecular to clinical, Food & function, 3 (2012) 1242-1250. [221] Y. Suzuki, N. Miyoshi, M. Isemura, Health-promoting effects of green tea, Proceedings of the Japan Academy. Series B, Physical and biological sciences, 88 (2012) 88-101. [222] B.B. Aggarwal, S. Shishodia, Molecular targets of dietary agents for prevention and therapy of cancer, Biochemical pharmacology, 71 (2006) 1397-1421. [223] N. Khan, H. Mukhtar, Tea polyphenols for health promotion, Life sciences, 81 (2007) 519-533. [224] E.B. Golden, P.Y. Lam, A. Kardosh, K.J. Gaffney, E. Cadenas, S.G. Louie, N.A. Petasis, T.C. Chen, A.H. Schonthal, Green tea polyphenols block the anticancer effects of bortezomib and other boronic acid-based proteasome inhibitors, Blood, 113 (2009) 5927-5937. [225] J. Ge, B.X. Tan, Y. Chen, L. Yang, X.C. Peng, H.Z. Li, H.J. Lin, Y. Zhao, M. Wei, K. Cheng, L.H. Li, H. Dong, F. Gao, J.P. He, Y. Wu, M. Qiu, Y.L. Zhao, J.M. Su, J.M. Hou, J.Y. Liu, Interaction of green tea polyphenol epigallocatechin-3-gallate with sunitinib: potential risk of diminished sunitinib bioavailability, Journal of molecular medicine, 89 (2011) 595-602. [226] J.D. Lambert, S. Sang, C.S. Yang, Possible controversy over dietary polyphenols: benefits vs risks, Chemical research in toxicology, 20 (2007) 583-585. [227] R. Strick, P.L. Strissel, S. Borgers, S.L. Smith, J.D. Rowley, Dietary bioflavonoids induce cleavage in the MLL gene and may contribute to infant leukemia, Proceedings of the National Academy of Sciences of the United States of America, 97 (2000) 4790-4795.

175

[228] X. Kuang, Y. Huang, H.F. Gu, X.Y. Zu, W.Y. Zou, Z.B. Song, Q.L. Guo, Effects of intrathecal epigallocatechin gallate, an inhibitor of Toll-like receptor 4, on chronic neuropathic pain in rats, European journal of pharmacology, 676 (2012) 51-56. [229] J.A. Mengshol, M.P. Vincenti, C.I. Coon, A. Barchowsky, C.E. Brinckerhoff, Interleukin-1 induction of collagenase 3 (matrix metalloproteinase 13) gene expression in chondrocytes requires p38, c-Jun N-terminal kinase, and nuclear factor kappaB: differential regulation of collagenase 1 and collagenase 3, Arthritis and rheumatism, 43 (2000) 801-811. [230] K. Otoshi, S. Kikuchi, S. Konno, M. Sekiguchi, The reactions of glial cells and endoneurial macrophages in the dorsal root ganglion and their contribution to pain-related behavior after application of nucleus pulposus onto the nerve root in rats, Spine, 35 (2010) 264-271. [231] M. Sekiguchi, K. Otoshi, S. Kikuchi, S. Konno, Analgesic effects of prostaglandin E2 receptor subtype EP1 receptor antagonist: experimental study of application of nucleus pulposus, Spine, 36 (2011) 1829-1834. [232] H. Kobayashi, S. Kikuchi, S. Konno, K. Kato, M. Sekiguchi, Interaction of 5-hydroxytryptamine and tumor necrosis factor-alpha to pain-related behavior by nucleus pulposus applied on the nerve root in rats, Spine, 36 (2011) 210-218. [233] S. Janssens, R. Beyaert, Functional diversity and regulation of different interleukin-1 receptor-associated kinase (IRAK) family members, Molecular cell, 11 (2003) 293-302. [234] M. Forster, F. Mahn, U. Gockel, M. Brosz, R. Freynhagen, T.R. Tolle, R. Baron, Axial low back pain: one painful area - many perceptions and mechanisms, PloS one, 8 (2013) e68273. [235] S.M. Richardson, D. Purmessur, P. Baird, B. Probyn, A.J. Freemont, J.A. Hoyland, Degenerate human nucleus pulposus cells promote neurite outgrowth in neural cells, PloS one, 7 (2012) e47735. [236] W. Alimasi, Y. Sawaji, K. Endo, M. Yorifuji, H. Suzuki, T. Kosaka, T. Shishido, K. Yamamoto, Regulation of nerve growth factor by anti-inflammatory drugs, a steroid, and a selective cyclooxygenase 2 inhibitor in human intervertebral disc cells stimulated with interleukin-1, Spine, 38 (2013) 1466-1472. [237] H.J. Yoon, S.B. Jeon, I.H. Kim, E.J. Park, Regulation of TLR2 expression by prostaglandins in brain glia, J Immunol, 180 (2008) 8400-8409. [238] M.B. Ellman, J.S. Kim, H.S. An, D. Chen, R. Kc, J. An, T. Dittakavi, A.J. van Wijnen, G. Cs-Szabo, X. Li, G. Xiao, S. An, S.G. Kim, H.J. Im, Toll-like receptor adaptor signaling molecule MyD88 on intervertebral disk homeostasis: in vitro, ex vivo studies, Gene, 505 (2012) 283-290. [239] M. Klawitter, M. Hakozaki, H. Kobayashi, O. Krupkova, L. Quero, C. Ospelt, S. Gay, O. Hausmann, T. Liebscher, U. Meier, M. Sekiguchi, S. Konno, N. Boos, S.J. Ferguson, K. Wuertz, Expression and regulation of toll-like receptors (TLRs) in human intervertebral disc cells, European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society, 23 (2014) 1878-1891. [240] K. Wuertz, L. Haglund, Inflammatory Mediators in Intervertebral Disk Degeneration and Discogenic Pain, Global Spine Journal, (2013) 175-184. [241] M.A. Gabr, L. Jing, A.R. Helbling, S.M. Sinclair, K.D. Allen, M.F. Shamji, W.J. Richardson, R.D. Fitch, L.A. Setton, J. Chen, Interleukin-17 synergizes with IFNgamma or TNFalpha to promote inflammatory mediator release and intercellular adhesion molecule-1 (ICAM-1) expression in human intervertebral disc cells, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, 29 (2011) 1-7. [242] R. Singh, S. Ahmed, N. Islam, V.M. Goldberg, T.M. Haqqi, Epigallocatechin-3-gallate inhibits interleukin-1beta-induced expression of nitric oxide synthase and production of nitric oxide in human chondrocytes: suppression of nuclear factor kappaB activation by degradation of the inhibitor of nuclear factor kappaB, Arthritis and rheumatism, 46 (2002) 2079-2086. [243] L.F. Heinecke, M.W. Grzanna, A.Y. Au, C.A. Mochal, A. Rashmir-Raven, C.G. Frondoza, Inhibition of cyclooxygenase-2 expression and prostaglandin E2 production in chondrocytes by avocado soybean unsaponifiables and epigallocatechin gallate, Osteoarthritis and cartilage / OARS, Osteoarthritis Research Society, 18 (2010) 220-227. [244] R. Singh, S. Ahmed, C.J. Malemud, V.M. Goldberg, T.M. Haqqi, Epigallocatechin-3-gallate selectively inhibits interleukin-1beta-induced activation of mitogen activated protein kinase

176

subgroup c-Jun N-terminal kinase in human osteoarthritis chondrocytes, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, 21 (2003) 102-109. [245] K. Burns, S. Janssens, B. Brissoni, N. Olivos, R. Beyaert, J. Tschopp, Inhibition of interleukin 1 receptor/Toll-like receptor signaling through the alternatively spliced, short form of MyD88 is due to its failure to recruit IRAK-4, The Journal of experimental medicine, 197 (2003) 263-268. [246] A. Liacini, J. Sylvester, W.Q. Li, W. Huang, F. Dehnade, M. Ahmad, M. Zafarullah, Induction of matrix metalloproteinase-13 gene expression by TNF-alpha is mediated by MAP kinases, AP-1, and NF-kappaB transcription factors in articular chondrocytes, Experimental cell research, 288 (2003) 208-217. [247] S.F. Elliott, C.I. Coon, E. Hays, T.A. Stadheim, M.P. Vincenti, Bcl-3 is an interleukin-1-responsive gene in chondrocytes and synovial fibroblasts that activates transcription of the matrix metalloproteinase 1 gene, Arthritis and rheumatism, 46 (2002) 3230-3239. [248] R.K. Studer, L.G. Gilbertson, H. Georgescu, G. Sowa, N. Vo, J.D. Kang, p38 MAPK inhibition modulates rabbit nucleus pulposus cell response to IL-1, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, 26 (2008) 991-998. [249] J.S. Kim, M.B. Ellman, H.S. An, D. Yan, A.J. van Wijnen, G. Murphy, D.W. Hoskin, H.J. Im, Lactoferricin mediates anabolic and anti-catabolic effects in the intervertebral disc, Journal of cellular physiology, 227 (2012) 1512-1520. [250] C.A. Seguin, M. Bojarski, R.M. Pilliar, P.J. Roughley, R.A. Kandel, Differential regulation of matrix degrading enzymes in a TNFalpha-induced model of nucleus pulposus tissue degeneration, Matrix biology : journal of the International Society for Matrix Biology, 25 (2006) 409-418. [251] S.T. Meller, G.F. Gebhart, Nitric oxide (NO) and nociceptive processing in the spinal cord, Pain, 52 (1993) 127-136. [252] J.I. Choi, W.M. Kim, H.G. Lee, Y.O. Kim, M.H. Yoon, Role of neuronal nitric oxide synthase in the antiallodynic effects of intrathecal EGCG in a neuropathic pain rat model, Neuroscience letters, 510 (2012) 53-57. [253] H. Tachihara, M. Sekiguchi, S. Kikuchi, S. Konno, Do corticosteroids produce additional benefit in nerve root infiltration for lumbar disc herniation?, Spine, 33 (2008) 743-747. [254] K. Kato, S. Kikuchi, S. Konno, M. Sekiguchi, Participation of 5-hydroxytryptamine in pain-related behavior induced by nucleus pulposus applied on the nerve root in rats, Spine, 33 (2008) 1330-1336. [255] N. Sasaki, S. Kikuchi, S. Konno, M. Sekiguchi, K. Watanabe, Anti-TNF-alpha antibody reduces pain-behavioral changes induced by epidural application of nucleus pulposus in a rat model depending on the timing of administration, Spine, 32 (2007) 413-416. [256] M. Forster, F. Mahn, U. Gockel, M. Brosz, R. Freynhagen, T.R. Tolle, R. Baron, Axial low back pain: one painful area--many perceptions and mechanisms, PloS one, 8 (2013) e68273. [257] A.J. Freemont, The cellular pathobiology of the degenerate intervertebral disc and discogenic back pain, Rheumatology, 48 (2009) 5-10. [258] D. Mereles, W. Hunstein, Epigallocatechin-3-gallate (EGCG) for Clinical Trials: More Pitfalls than Promises?, International journal of molecular sciences, 12 (2011) 5592-5603. [259] M.J. Lee, P. Maliakal, L. Chen, X. Meng, F.Y. Bondoc, S. Prabhu, G. Lambert, S. Mohr, C.S. Yang, Pharmacokinetics of tea catechins after ingestion of green tea and (-)-epigallocatechin-3-gallate by humans: formation of different metabolites and individual variability, Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology, 11 (2002) 1025-1032. [260] S. Wang, J. Zhang, M. Chen, Y. Wang, Delivering flavonoids into solid tumors using nanotechnologies, Expert opinion on drug delivery, 10 (2013) 1411-1428. [261] D. Wang, E.W. Taylor, Y. Wang, X. Wan, J. Zhang, Encapsulated nanoepigallocatechin-3-gallate and elemental selenium nanoparticles as paradigms for nanochemoprevention, International journal of nanomedicine, 7 (2012) 1711-1721. [262] B.A. Walter, O.M. Torre, D. Laudier, T.P. Naidich, A.C. Hecht, J.C. Iatridis, Form and function of the intervertebral disc in health and disease: a morphological and stain comparison study, J Anat, (2014). [263] K. Wuertz, L. Haglund, Inflammatory mediators in intervertebral disk degeneration and discogenic pain, Global Spine J, 3 (2013) 175-184.

177

[264] S.S. Sivan, E. Wachtel, P. Roughley, Structure, function, aging and turnover of aggrecan in the intervertebral disc, Biochimica et biophysica acta, 1840 (2014) 3181-3189. [265] G. Fontana, E. See, A. Pandit, Current trends in biologics delivery to restore intervertebral disc anabolism, Advanced drug delivery reviews, (2014). [266] E.D. Schleicher, E. Wagner, A.G. Nerlich, Increased accumulation of the glycoxidation product N-epsilon(carboxymethyl)lysine in human tissues in diabetes and aging, J Clin Invest, 99 (1997) 457-468. [267] S.S. Sivan, E. Tsitron, E. Wachtel, P. Roughley, N. Sakkee, F. Van Der Ham, J. DeGroot, A. Maroudas, Age-related accumulation of pentosidine in aggrecan and collagen from normal and degenerate human intervertebral discs, Biochemical Journal, 399 (2006) 29-35. [268] A.G. Nerlich, B.E. Bachmeier, E. Schleicher, H. Rohrbach, G. Paesold, N. Boos, Immunomorphological analysis of RAGE receptor expression and NF-kappa B activation in tissue samples from normal and degenerated intervertebral discs of various ages, Ann Ny Acad Sci, 1096 (2007) 239-248. [269] K.W. Kim, H.N. Chung, K.Y. Ha, J.S. Lee, Y.Y. Kim, Senescence mechanisms of nucleus pulposus chondrocytes in human intervertebral discs, Spine Journal, 9 (2009) 658-666. [270] H.Y. Chung, B. Sung, K.J. Jung, Y. Zou, B.P. Yu, The molecular inflammatory process in aging, Antioxid Redox Signal, 8 (2006) 572-581. [271] H. Sies, Role of metabolic H2O2 generation: redox signaling and oxidative stress, The Journal of biological chemistry, 289 (2014) 8735-8741. [272] A.R. Young, M. Narita, M. Narita, Cell senescence as both a dynamic and a static phenotype, Methods in molecular biology, 965 (2013) 1-13. [273] Y. Qian, X. Chen, Senescence regulation by the p53 protein family, Methods in molecular biology, 965 (2013) 37-61. [274] C.M. Walsh, Grand challenges in cell death and survival: apoptosis vs. necroptosis, Frontiers in cell and developmental biology, 2 (2014) 3. [275] J. Karch, J.D. Molkentin, Regulated Necrotic Cell Death: The Passive Aggressive Side of Bax and Bak, Circulation research, 116 (2015) 1800-1809. [276] C.Q. Zhao, L.M. Wang, L.S. Jiang, L.Y. Dai, The cell biology of intervertebral disc aging and degeneration, Ageing Res Rev, 6 (2007) 247-261. [277] S.Z. Wang, Y.F. Rui, J. Lu, C. Wang, Cell and molecular biology of intervertebral disc degeneration: current understanding and implications for potential therapeutic strategies, Cell proliferation, 47 (2014) 381-390. [278] J.S. Park, J.B. Park, I.J. Park, E.Y. Park, Accelerated premature stress-induced senescence of young annulus fibrosus cells of rats by high glucose-induced oxidative stress, International orthopaedics, 38 (2014) 1311-1320. [279] Z. Li, M. Peroglio, M. Alini, S. Grad, Potential and limitations of intervertebral disc endogenous repair, Current stem cell research & therapy, 10 (2015) 329-338. [280] E. Sikora, Rejuvenation of senescent cells-The road to postponing human aging and age-related disease?, Exp Gerontol, 48 (2013) 661-666. [281] L.L. Mensor, F.S. Menezes, G.G. Leitao, A.S. Reis, T.C. dos Santos, C.S. Coube, S.G. Leitao, Screening of Brazilian plant extracts for antioxidant activity by the use of DPPH free radical method, Phytotherapy Research, 15 (2001) 127-130. [282] A. Brandl, A. Hartmann, V. Bechmann, B. Graf, M. Nerlich, P. Angele, Oxidative stress induces senescence in chondrocytes, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, 29 (2011) 1114-1120. [283] Y. Ido, A. Duranton, F. Lan, J.M. Cacicedo, T.C. Chen, L. Breton, N.B. Ruderman, Acute Activation of AMP-Activated Protein Kinase Prevents H2O2-Induced Premature Senescence in Primary Human Keratinocytes, PloS one, 7 (2012). [284] D.J. Kurz, S. Decary, Y. Hong, J.D. Erusalimsky, Senescence-associated beta-galactosidase reflects an increase in lysosomal mass during replicative ageing of human endothelial cells, Journal of cell science, 113 (2000) 3613-3622. [285] D. Wlodkowic, J. Skommer, Z. Darzynkiewicz, Flow Cytometry-Based Apoptosis Detection, Apoptosis: Methods and Protocols, Second Edition, 559 (2009) 19-32. [286] W. Brandwilliams, M.E. Cuvelier, C. Berset, Use of a Free-Radical Method to Evaluate Antioxidant Activity, Food Sci Technol-Leb, 28 (1995) 25-30.

178

[287] P. Bellion, M. Olk, F. Will, H. Dietrich, M. Baum, G. Eisenbrand, C. Janzowski, Formation of hydrogen peroxide in cell culture media by apple polyphenols and its effect on antioxidant biomarkers in the colon cell line HT-29, Mol Nutr Food Res, 53 (2009) 1226-1236. [288] L.H. Long, M.V. Clement, B. Halliwell, Artifacts in cell culture: Rapid generation of hydrogen peroxide on addition of (-)-epigallocatechin, (-)-epigallocatechin gallate, (+)-catechin, and quercetin to commonly used cell culture media, Biochemical and biophysical research communications, 273 (2000) 50-53. [289] Q.M. Chen, V.C. Tu, J. Catania, M. Burton, O. Toussaint, T. Dilley, Involvement of Rb family proteins, focal adhesion proteins and protein synthesis in senescent morphogenesis induced by hydrogen peroxide, Journal of cell science, 113 ( Pt 22) (2000) 4087-4097. [290] H.S. Kim, V. Montana, H.J. Jang, V. Parpura, J.A. Kim, Epigallocatechin gallate (EGCG) stimulates autophagy in vascular endothelial cells: a potential role for reducing lipid accumulation, The Journal of biological chemistry, 288 (2013) 22693-22705. [291] J. Zhou, B.L. Farah, R.A. Sinha, Y. Wu, B.K. Singh, B.H. Bay, C.S. Yang, P.M. Yen, Epigallocatechin-3-gallate (EGCG), a green tea polyphenol, stimulates hepatic autophagy and lipid clearance, PloS one, 9 (2014) e87161. [292] N.C. Yang, C.H. Lee, T.Y. Song, Evaluation of resveratrol oxidation in vitro and the crucial role of bicarbonate ions, Bioscience, biotechnology, and biochemistry, 74 (2010) 63-68. [293] J. Campisi, F.D. di Fagagna, Cellular senescence: when bad things happen to good cells, Nat Rev Mol Cell Bio, 8 (2007) 729-740. [294] G.H. Stein, L.F. Drullinger, A. Soulard, V. Dulic, Differential roles for cyclin-dependent kinase inhibitors p21 and p16 in the mechanisms of senescence and differentiation in human fibroblasts, Molecular and cellular biology, 19 (1999) 2109-2117. [295] W.L. Yen, D.J. Klionsky, How to Live Long and Prosper: Autophagy, Mitochondria, and Aging, Physiology, 23 (2008) 248-262. [296] G.L. Tipoe, T.M. Leung, E.C. Liong, T.Y. Lau, M.L. Fung, A.A. Nanji, Epigallocatechin-3-gallate (EGCG) reduces liver inflammation, oxidative stress and fibrosis in carbon tetrachloride (CCl4)-induced liver injury in mice, Toxicology, 273 (2010) 45-52. [297] D. Cia, J. Vergnaud-Gauduchon, N. Jacquemot, M. Doly, Epigallocatechin gallate (EGCG) prevents H2O2-induced oxidative stress in primary rat retinal pigment epithelial cells, Current eye research, 39 (2014) 944-952. [298] E. Cadenas, K.J. Davies, Mitochondrial free radical generation, oxidative stress, and aging, Free radical biology & medicine, 29 (2000) 222-230. [299] T. Nakagawa, T. Yokozawa, Direct scavenging of nitric oxide and superoxide by green tea, Food and chemical toxicology : an international journal published for the British Industrial Biological Research Association, 40 (2002) 1745-1750. [300] Y.M. Li, H.Y. Chan, Y. Huang, Z.Y. Chen, Green tea catechins upregulate superoxide dismutase and catalase in fruit flies, Mol Nutr Food Res, 51 (2007) 546-554. [301] Y.V. Simos, Verginadis, II, I.K. Toliopoulos, A.P. Velalopoulou, I.V. Karagounis, S.C. Karkabounas, A.M. Evangelou, Effects of catechin and epicatechin on superoxide dismutase and glutathione peroxidase activity, in vivo, Redox report : communications in free radical research, 17 (2012) 181-186. [302] H. Maccario, N.M. Perera, A. Gray, C.P. Downes, N.R. Leslie, Ubiquitination of PTEN (Phosphatase and Tensin Homolog) Inhibits Phosphatase Activity and Is Enhanced by Membrane Targeting and Hyperosmotic Stress, Journal of Biological Chemistry, 285 (2010) 12620-12628. [303] M. Osaki, M. Oshimura, H. Ito, PI3K-Akt pathway: its functions and alterations in human cancer, Apoptosis : an international journal on programmed cell death, 9 (2004) 667-676. [304] H. Tachibana, K. Koga, Y. Fujimura, K. Yamada, A receptor for green tea polyphenol EGCG, Nature structural & molecular biology, 11 (2004) 380-381. [305] H. Pratsinis, D. Kletsas, PDGF, bFGF and IGF-I stimulate the proliferation of intervertebral disc cells in vitro via the activation of the ERK and Akt signaling pathways, European Spine Journal, 16 (2007) 1858-1866. [306] D. Pasku, G. Soufla, P. Katonis, A. Tsarouhas, A. Vakis, D.A. Spandidos, Akt/PKB isoforms expression in the human lumbar herniated disc: correlation with clinical and MRI findings, European spine journal : official publication of the European Spine Society, the European Spinal

179

Deformity Society, and the European Section of the Cervical Spine Research Society, 20 (2011) 1676-1683. [307] Z. Liu, K. Zhou, W. Fu, H. Zhang, Insulin-Like Growth Factor 1 Activates PI3k/Akt Signaling to Antagonize Lumbar Disc Degeneration, Cellular physiology and biochemistry : international journal of experimental cellular physiology, biochemistry, and pharmacology, 37 (2015) 225-232. [308] D.W. Wang, Z.M. Hu, J. Hao, B. He, Q. Gan, X.M. Zhong, X.J. Zhang, J.L. Shen, J. Fang, W. Jiang, SIRT1 inhibits apoptosis of degenerative human disc nucleus pulposus cells through activation of Akt pathway, Age, 35 (2013) 1741-1753. [309] G. Pattappa, Z. Li, M. Peroglio, N. Wismer, M. Alini, S. Grad, Diversity of intervertebral disc cells: phenotype and function, J Anat, 221 (2012) 480-496. [310] B. van Dijk, E. Potier, D.M. van, M. Langelaan, N. Papen-Botterhuis, K. Ito, Reduced tonicity stimulates an inflammatory response in nucleus pulposus tissue that can be limited by a COX-2-specific inhibitor, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, 33 (2015) 1724-1731. [311] O. Krupkova, J. Handa, M. Hlavna, J. Klasen, C. Ospelt, S.J. Ferguson, K. Wuertz-Kozak, The Natural Polyphenol Epigallocatechin Gallate Protects Intervertebral Disc Cells from Oxidative Stress, Oxidative medicine and cellular longevity, 2016 (2016) 7031397. [312] S.C.W. Chan, A. Burki, H.M. Bonel, L.M. Benneker, B. Gantenbein-Ritter, Papain-induced in vitro disc degeneration model for the study of injectable nucleus pulposus therapy, Spine Journal, 13 (2013) 273-283. [313] C. Bucher, A. Gazdhar, L.M. Benneker, T. Geiser, B. Gantenbein-Ritter, Nonviral Gene Delivery of Growth and Differentiation Factor 5 to Human Mesenchymal Stem Cells Injected into a 3D Bovine Intervertebral Disc Organ Culture System, Stem Cells Int, (2013). [314] M.A.A. Rousseau, J.A. Ulrich, E.C. Bass, A.G. Rodriguez, J.J. Liu, J.C. Lotz, Stab incision for inducing intervertebral disc degeneration in the rat, Spine, 32 (2007) 17-24. [315] A.C. Abraham, J.W. Liu, S.Y. Tang, Longitudinal changes in the structure and inflammatory response of the intervertebral disc due to stab injury in a murine organ culture model, Journal of orthopaedic research : official publication of the Orthopaedic Research Society, (2016). [316] R.J. Hoogendoorn, P.I. Wuisman, T.H. Smit, V.E. Everts, M.N. Helder, Experimental intervertebral disc degeneration induced by chondroitinase ABC in the goat, Spine, 32 (2007) 1816-1825. [317] K. Chiba, K. Masuda, G.B. Andersson, S. Momohara, E.J. Thonar, Matrix replenishment by intervertebral disc cells after chemonucleolysis in vitro with chondroitinase ABC and chymopapain, The spine journal : official journal of the North American Spine Society, 7 (2007) 694-700. [318] K.M.C.C. Juliana T. Y. Lee, Victor Y. L. Leung, Extraction of RNA from tough tissues with high proteoglycan content by cryosection, second phase separation and high salt precipitation, J Biol Methods 2(2015). [319] R.W. Farndale, C.A. Sayers, A.J. Barrett, A direct spectrophotometric microassay for sulfated glycosaminoglycans in cartilage cultures, Connective tissue research, 9 (1982) 247-248. [320] S. Dudli, D. Haschtmann, S.J. Ferguson, Persistent degenerative changes in the intervertebral disc after burst fracture in an in vitro model mimicking physiological post-traumatic conditions, European Spine Journal, 24 (2015) 1901-1908. [321] M. Alini, S.M. Eisenstein, K. Ito, C. Little, A.A. Kettler, K. Masuda, J. Melrose, J. Ralphs, I. Stokes, H.J. Wilke, Are animal models useful for studying human disc disorders/degeneration?, European Spine Journal, 17 (2008) 2-19. [322] B.G. van Dijk, E. Potier, M. van Dijk, L.B. Creemers, K. Ito, Osteogenic protein 1 does not stimulate a regenerative effect in cultured human degenerated nucleus pulposus tissue, Journal of tissue engineering and regenerative medicine, (2015). [323] T. Furtwangler, S.C.W. Chan, G. Bahrenberg, P.J. Richards, B. Gantenbein-Ritter, Assessment of the Matrix Degenerative Effects of MMP-3, ADAMTS-4, and HTRA1, Injected Into a Bovine Intervertebral Disc Organ Culture Model, Spine, 38 (2013) E1377-E1387. [324] A.N. Tiaden, M. Klawitter, V. Lux, A. Mirsaidi, G. Bahrenberg, S. Glanz, L. Quero, T. Liebscher, K. Wuertz, M. Ehrmann, P.J. Richards, Detrimental Role for Human High Temperature Requirement Serine Protease A1 (HTRA1) in the Pathogenesis of Intervertebral Disc (IVD) Degeneration, Journal of Biological Chemistry, 287 (2012) 21335-21345.

180

[325] G.A. Sowa, J.P. Coelho, N.V. Vo, C. Pacek, E. Westrick, J.D. Kang, Cells from degenerative intervertebral discs demonstrate unfavorable responses to mechanical and inflammatory stimuli: a pilot study, American journal of physical medicine & rehabilitation / Association of Academic Physiatrists, 91 (2012) 846-855. [326] J.V. Higdon, B. Frei, Tea catechins and polyphenols: health effects, metabolism, and antioxidant functions, Critical reviews in food science and nutrition, 43 (2003) 89-143. [327] Z. Chen, Q.Y. Zhu, D. Tsang, Y. Huang, Degradation of green tea catechins in tea drinks, Journal of agricultural and food chemistry, 49 (2001) 477-482. [328] Y.L. Su, L.K. Leung, Y. Huang, Z.Y. Chen, Stability of tea theaflavins and catechins, Food chemistry, 83 (2003) 189-195. [329] S.M. Henning, Y. Niu, N.H. Lee, G.D. Thames, R.R. Minutti, H. Wang, V.L. Go, D. Heber, Bioavailability and antioxidant activity of tea flavanols after consumption of green tea, black tea, or a green tea extract supplement, The American journal of clinical nutrition, 80 (2004) 1558-1564. [330] A. Dube, J.A. Nicolazzo, I. Larson, Chitosan nanoparticles enhance the plasma exposure of (-)-epigallocatechin gallate in mice through an enhancement in intestinal stability, European Journal of Pharmaceutical Sciences, 44 (2011) 422-426. [331] A. Barras, A. Mezzetti, A. Richard, S. Lazzaroni, S. Roux, P. Melnyk, D. Betbeder, N. Monfilliette-Dupont, Formulation and characterization of polyphenol-loaded lipid nanocapsules, International journal of pharmaceutics, 379 (2009) 270-277. [332] S. Wang, N. Moustaid-Moussa, L. Chen, H. Mo, A. Shastri, R. Su, P. Bapat, I. Kwun, C.L. Shen, Novel insights of dietary polyphenols and obesity, The Journal of nutritional biochemistry, 25 (2014) 1-18. [333] S. Khurana, K. Venkataraman, A. Hollingsworth, M. Piche, T.C. Tai, Polyphenols: benefits to the cardiovascular system in health and in aging, Nutrients, 5 (2013) 3779-3827. [334] A. Rietveld, S. Wiseman, Antioxidant effects of tea: evidence from human clinical trials, The Journal of nutrition, 133 (2003) 3285S-3292S. [335] Y. Teshima, N. Takahashi, S. Nishio, S. Saito, H. Kondo, A. Fukui, K. Aoki, K. Yufu, M. Nakagawa, T. Saikawa, Production of reactive oxygen species in the diabetic heart. Roles of mitochondria and NADPH oxidase, Circulation journal : official journal of the Japanese Circulation Society, 78 (2014) 300-306. [336] B. Uttara, A.V. Singh, P. Zamboni, R.T. Mahajan, Oxidative stress and neurodegenerative diseases: a review of upstream and downstream antioxidant therapeutic options, Current neuropharmacology, 7 (2009) 65-74. [337] J.D. Lambert, R.J. Elias, The antioxidant and pro-oxidant activities of green tea polyphenols: A role in cancer prevention, Archives of biochemistry and biophysics, 501 (2010) 65-72. [338] K.F. Pirker, J.F. Severino, T.G. Reichenauer, B.A. Goodman, Free radical processes in green tea polyphenols (GTP) investigated by electron paramagnetic resonance (EPR) spectroscopy, Biotechnol Ann Rev, 14 (2008) 349-401. [339] T. Unno, A. Sugimoto, T. Kakuda, Scavenging effect of tea catechins and their epimers on superoxide anion radicals generated by a hypoxanthine anal xanthine oxidase system, J Sci Food Agr, 80 (2000) 601-606. [340] Q. Guo, B. Zhao, S. Shen, J. Hou, J. Hu, W. Xin, ESR study on the structure-antioxidant activity relationship of tea catechins and their epimers, Biochimica et biophysica acta, 1427 (1999) 13-23. [341] Z. An, Y.M. Qi, D.J. Huang, X.Y. Gu, Y.H. Tian, P. Li, H. Li, Y.M. Zhang, EGCG inhibits Cd2+-induced apoptosis through scavenging ROS rather than chelating Cd2+ in HL-7702 cells, Toxicol Mech Method, 24 (2014) 259-267. [342] N. Zhu, T.C. Huang, Y. Yu, E.J. LaVoie, C.S. Yang, C.T. Ho, Identification of oxidation products of (-)-epigallocatechin gallate and (-)-epigallocatechin with H(2)O(2), Journal of agricultural and food chemistry, 48 (2000) 979-981. [343] J.K. Lin, P.C. Chen, C.T. Ho, S.Y. Lin-Shiau, Inhibition of xanthine oxidase and suppression of intracellular reactive oxygen species in HL-60 cells by theaflavin-3,3'-digallate, (-)-epigallocatechin-3-gallate, and propyl gallate, Journal of agricultural and food chemistry, 48 (2000) 2736-2743.

181

[344] A. Koeberle, J. Bauer, M. Verhoff, M. Hoffmann, H. Northoff, O. Werz, Green tea epigallocatechin-3-gallate inhibits microsomal prostaglandin E(2) synthase-1, Biochemical and biophysical research communications, 388 (2009) 350-354. [345] J. Hong, T.J. Smith, C.T. Ho, D.A. August, C.S. Yang, Effects of purified green and black tea polyphenols on cyclooxygenase- and lipoxygenase-dependent metabolism of arachidonic acid in human colon mucosa and colon tumor tissues, Biochemical pharmacology, 62 (2001) 1175-1183. [346] S.A. Mandel, Y. Avramovich-Tirosh, L. Reznichenko, H.L. Zheng, O. Weinreb, T. Amit, M.B.H. Youdim, Multifunctional activities of green tea catechins in neuroprotection - Modulation of cell survival genes, iron-dependent oxidative stress and PKC signaling pathway, Neurosignals, 14 (2005) 46-60. [347] E.Y. Kim, S.K. Ham, M.K. Shigenaga, O. Han, Bioactive dietary polyphenolic compounds reduce nonheme iron transport across human intestinal cell monolayers, The Journal of nutrition, 138 (2008) 1647-1651. [348] O. Weinreb, T. Amit, S. Mandel, M.B. Youdim, Neuroprotective molecular mechanisms of (-)-epigallocatechin-3-gallate: a reflective outcome of its antioxidant, iron chelating and neuritogenic properties, Genes & nutrition, 4 (2009) 283-296. [349] S. Mandel, T. Amit, O. Bar-Am, M.B. Youdim, Iron dysregulation in Alzheimer's disease: multimodal brain permeable iron chelating drugs, possessing neuroprotective-neurorescue and amyloid precursor protein-processing regulatory activities as therapeutic agents, Progress in neurobiology, 82 (2007) 348-360. [350] Z.M. Wang, C.L. Tidrick, M. Haque, D.J. Stuehr, Green Tea Polyphenols Decrease Enzyme Activity of Nitric Oxide Synthase, Faseb Journal, 27 (2013). [351] J.D. Lambert, S. Sang, C.S. Yang, Biotransformation of green tea polyphenols and the biological activities of those metabolites, Molecular pharmaceutics, 4 (2007) 819-825. [352] M.J. Lee, P. Maliakal, L.S. Chen, X.F. Meng, F.Y. Bondoc, S. Prabhu, G. Lambert, S. Mohr, C.S. Yang, Pharmacokinetics of tea catechins after ingestion of green tea and (-)-epigallocatechin-3-gallate by humans: formation of different metabolites and individual variability, Cancer Epidem Biomar, 11 (2002) 1025-1032. [353] C. Li, X. Meng, B. Winnik, M.J. Lee, H. Lu, S. Sheng, B. Buckley, C.S. Yang, Analysis of urinary metabolites of tea catechins by liquid chromatography/electrospray ionization mass spectrometry, Chemical research in toxicology, 14 (2001) 702-707. [354] H.H.S. Chow, I.A. Hakim, Pharmacokinetic and chemoprevention studies on tea in humans, Pharmacol Res, 64 (2011) 105-112. [355] S. Misaka, K. Kawabe, S. Onoue, J.P. Werba, M. Giroli, S. Tamaki, T. Kan, J. Kimura, H. Watanabe, S. Yamada, Effects of Green Tea Catechins on Cytochrome P450 2B6, 2C8, 2C19, 2D6 and 3A Activities in Human Liver and Intestinal Microsomes, Drug Metab Pharmacok, 28 (2013) 244-249. [356] T. Nakayama, K. Kajiya, S. Kumazawa, Interaction of Plant Polyphenols with Liposomes, Adv Planar Lip Bilay, 4 (2006) 107-133. [357] J. Hong, J.D. Lambert, S.H. Lee, P.J. Sinko, C.S. Yang, Involvement of multidrug resistance-associated proteins in regulating cellular levels of (-)-epigallocatechin-3-gallate and its methyl metabolites, Biochemical and biophysical research communications, 310 (2003) 222-227. [358] J. Jodoin, M. Demeule, R. Beliveau, Inhibition of the multidrug resistance P-glycoprotein activity by green tea polyphenols, Bba-Mol Cell Res, 1542 (2002) 149-159. [359] M. Kadowaki, N. Sugihara, T. Tagashira, K. Terao, K. Furuno, Presence or absence of a gallate moiety on catechins affects their cellular transport, Journal of Pharmacy and Pharmacology, 60 (2008) 1189-1195. [360] L. Mata-Bilbao Mde, C. Andres-Lacueva, E. Roura, O. Jauregui, E. Escribano, C. Torre, R.M. Lamuela-Raventos, Absorption and pharmacokinetics of green tea catechins in beagles, The British journal of nutrition, 100 (2008) 496-502. [361] J.D. Lambert, S.M. Sang, C.S. Yang, Biotransformation of green tea polyphenols and the biological activities of those metabolites, Molecular pharmaceutics, 4 (2007) 819-825. [362] C.S. Yang, M.J. Lee, L.S. Chen, Human salivary tea catechin levels and catechin esterase activities: Implication in human cancer prevention studies, Cancer Epidem Biomar, 8 (1999) 83-89.

182

[363] D.A. August, J. Landau, D. Caputo, J. Hong, M.J. Lee, C.S. Yang, Ingestion of green tea rapidly decreases prostaglandin E2 levels in rectal mucosa in humans, Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology, 8 (1999) 709-713. [364] L. Elbling, R.M. Weiss, O. Teufelhofer, M. Uhl, S. Knasmueller, R. Schulte-Hermann, W. Berger, M. Micksche, Green tea extract and (-)-epigallocatechin-3-gallate, the major tea catechin, exert oxidant but lack antioxidant activities, FASEB journal : official publication of the Federation of American Societies for Experimental Biology, 19 (2005) 807-809. [365] M.E. Cavet, K.L. Harrington, T.R. Vollmer, K.W. Ward, J.Z. Zhang, Anti-inflammatory and anti-oxidative effects of the green tea polyphenol epigallocatechin gallate in human corneal epithelial cells, Molecular vision, 17 (2011) 533-542. [366] O. Krupkova, M. Sekiguchi, J. Klasen, O. Hausmann, S. Konno, S.J. Ferguson, K. Wuertz-Kozak, EPIGALLOCATECHIN 3-GALLATE SUPPRESSES INTERLEUKIN-1 beta-INDUCED INFLAMMATORY RESPONSES IN INTERVERTEBRAL DISC CELLS IN VITRO AND REDUCES RADICULOPATHIC PAIN IN RATS, European cells & materials, 28 (2014) 372-386. [367] F. Oliviero, P. Sfriso, A. Scanu, U. Fiocco, P. Spinella, L. Punzi, Epigallocatechin-3-gallate reduces inflammation induced by calcium pyrophosphate crystals in vitro, Frontiers in pharmacology, 4 (2013) 51. [368] H.J. Wobst, A. Sharma, M.I. Diamond, E.E. Wanker, J. Bieschke, The green tea polyphenol (-)-epigallocatechin gallate prevents the aggregation of tau protein into toxic oligomers at substoichiometric ratios, FEBS letters, 589 (2015) 77-83. [369] Y.T. Deng, T.W. Chang, M.S. Lee, J.K. Lin, Suppression of free fatty acid-induced insulin resistance by phytopolyphenols in C2C12 mouse skeletal muscle cells, Journal of agricultural and food chemistry, 60 (2012) 1059-1066. [370] E.P. Cai, J.K. Lin, Epigallocatechin Gallate (EGCG) and Rutin Suppress the Glucotoxicity through Activating IRS2 and AMPK Signaling in Rat Pancreatic beta Cells, Journal of agricultural and food chemistry, 57 (2009) 9817-9827. [371] H.S. Kim, M.J. Quon, J.A. Kim, New insights into the mechanisms of polyphenols beyond antioxidant properties; lessons from the green tea polyphenol, epigallocatechin 3-gallate, Redox biology, 2 (2014) 187-195. [372] J. Duarte, V. Franciscobc, F. Perez-Vizcainod, Modulation of nitric oxide by flavonoids, Food & function, 5 (2014) 1653-1668. [373] R. Johnson, S. Bryant, A.L. Huntley, Green tea and green tea catechin extracts: An overview of the clinical evidence, Maturitas, 73 (2012) 280-287. [374] P.V. Babu, D. Liu, E.R. Gilbert, Recent advances in understanding the anti-diabetic actions of dietary flavonoids, The Journal of nutritional biochemistry, 24 (2013) 1777-1789. [375] D. Chen, S.B. Wan, H.J. Yang, J. Yuan, T.H. Chan, Q.P. Dou, Egcg, Green Tea Polyphenols and Their Synthetic Analogs and Prodrugs for Human Cancer Prevention and Treatment, Adv Clin Chem, 53 (2011) 155-177. [376] G.J. Du, Z.Y. Zhang, X.D. Wen, C.H. Yu, T. Calway, C.S. Yuan, C.Z. Wang, Epigallocatechin Gallate (EGCG) Is the Most Effective Cancer Chemopreventive Polyphenol in Green Tea, Nutrients, 4 (2012) 1679-1691. [377] A. Papi, M. Govoni, C. Ciavarella, E. Spisni, M. Orlandi, F. Farabegoli, Epigallocatechin-3-gallate increases RXRg-mediated pro-apoptotic and anti-invasive effects in gastrointestinal cancer cell lines, Current cancer drug targets, (2015). [378] S. Adachi, T. Nagao, S. To, A.K. Joe, M. Shimizu, R. Matsushima-Nishiwaki, O. Kozawa, H. Moriwaki, F.R. Maxfield, I.B. Weinstein, (-)-Epigallocatechin gallate causes internalization of the epidermal growth factor receptor in human colon cancer cells, Carcinogenesis, 29 (2008) 1986-1993. [379] S. Song, Y.W. Huang, Y. Tian, X.J. Wang, J. Sheng, Mechanism of action of (-)-epigallocatechin-3-gallate: auto-oxidation-dependent activation of extracellular signal-regulated kinase 1/2 in Jurkat cells, Chin J Nat Medicines, 12 (2014) 654-662. [380] Z. Hou, S.M. Sang, H. You, M.J. Lee, J. Hong, K.V. Chin, C.S. Yang, Mechanism of action of (-)-epigallocatechin-3-gallate: Auto-oxidation-dependent inactivation of epidermal growth factor receptor and direct effects on growth inhibition in human esophageal cancer KYSE 150 cells, Cancer research, 65 (2005) 8049-8056.

183

[381] J.D. Lambert, M.J. Kennett, S. Sang, K.R. Reuhl, J. Ju, C.S. Yang, Hepatotoxicity of high oral dose (-)-epigallocatechin-3-gallate in mice, Food and chemical toxicology : an international journal published for the British Industrial Biological Research Association, 48 (2010) 409-416. [382] G.Y. Yang, J. Liao, K. Kim, E.J. Yurkow, C.S. Yang, Inhibition of growth and induction of apoptosis in human cancer cell lines by tea polyphenols, Carcinogenesis, 19 (1998) 611-616. [383] M. Schmidt, H.J. Schmitz, A. Baumgart, H. Kreuter, M. Netsch, C. Schmidlin, D. Schrenk, Toxicity of green tea extracts and their constituents in rat hepatocytes in primary culture, N-S Arch Pharmacol, 369 (2004) R143-R143. [384] D.Y. Wu, Z.Y. Guo, Z.H. Ren, W.M. Guo, S.N. Meydani, Green tea EGCG suppresses T cell proliferation through impairment of IL-2/IL-2 receptor signaling, Free Radical Bio Med, 47 (2009) 636-643. [385] J. Hong, H. Lu, X. Meng, J.H. Ryu, Y. Hara, C.S. Yang, Stability, cellular uptake, biotransformation, and efflux of tea polyphenol (-)-epigallocatechin-3-gallate in HT-29 human colon adenocarcinoma cells, Cancer research, 62 (2002) 7241-7246. [386] A. Zinellu, S. Sotgia, B. Scanu, M. Forteschi, R. Giordo, A. Cossu, A.M. Posadino, C. Carru, G. Pintus, Human Serum Albumin Increases the Stability of Green Tea Catechins in Aqueous Physiological Conditions, PloS one, 10 (2015) e0134690. [387] S. Thangapandiyan, S. Miltonprabu, Epigallocatechin gallate effectively ameliorates fluoride-induced oxidative stress and DNA damage in the liver of rats, Can J Physiol Pharm, 91 (2013) 528-537. [388] P.V.A. Babu, H.W. Si, D.M. Liu, Epigallocatechin gallate reduces vascular inflammation in db/db mice possibly through an NF-kappa B-mediated mechanism, Mol Nutr Food Res, 56 (2012) 1424-1432. [389] D.J. Leong, M. Choudhury, R. Hanstein, D.M. Hirsh, S.J. Kim, R.J. Majeska, M.B. Schaffler, J.A. Hardin, D.C. Spray, M.B. Goldring, N.J. Cobelli, H.B. Sun, Green tea polyphenol treatment is chondroprotective, anti-inflammatory and palliative in a mouse post-traumatic osteoarthritis model, Arthritis research & therapy, 16 (2014) 508. [390] H. Ortsater, N. Grankvist, S. Wolfram, N. Kuehn, A. Sjoholm, Diet supplementation with green tea extract epigallocatechin gallate prevents progression to glucose intolerance in db/db mice, Nutrition & metabolism, 9 (2012) 11. [391] G. Mazzanti, F. Menniti-Ippolito, P.A. Moro, F. Cassetti, R. Raschetti, C. Santuccio, S. Mastrangelo, Hepatotoxicity from green tea: a review of the literature and two unpublished cases, European journal of clinical pharmacology, 65 (2009) 331-341. [392] G. Galati, A. Lin, A.M. Sultan, P.J. O'Brien, Cellular and in vivo hepatotoxicity caused by green tea phenolic acids and catechins, Free Radical Bio Med, 40 (2006) 570-580. [393] H. Inoue, M. Maeda-Yamamoto, A. Nesumi, T. Tanaka, A. Murakami, Low and medium but not high doses of green tea polyphenols ameliorated dextran sodium sulfate-induced hepatotoxicity and nephrotoxicity, Bioscience, biotechnology, and biochemistry, 77 (2013) 1223-1228. [394] M. Pae, Z. Ren, M. Meydani, F. Shang, D. Smith, S.N. Meydani, D. Wu, Dietary supplementation with high dose of epigallocatechin-3-gallate promotes inflammatory response in mice, The Journal of nutritional biochemistry, 23 (2012) 526-531. [395] Y.C. Patry, P.H. Nachman, M.A.P. Audrain, R.J. Falk, K. Meflah, V.L.M. Esnault, Difference in antigenic determinant profiles between human and rat myeloperoxidase, Clin Exp Immunol, 132 (2003) 505-508. [396] S. Sang, M.J. Lee, Z. Hou, C.T. Ho, C.S. Yang, Stability of tea polyphenol (-)-epigallocatechin-3-gallate and formation of dimers and epimers under common experimental conditions, Journal of agricultural and food chemistry, 53 (2005) 9478-9484. [397] P. Janeiro, A.M.O. Brett, Catechin electrochemical oxidation mechanisms, Analytica chimica acta, 518 (2004) 109-115. [398] M. Mochizuki, S. Yamazaki, K. Kano, T. Ikeda, Kinetic analysis and mechanistic aspects of autoxidation of catechins, Biochimica et biophysica acta, 1569 (2002) 35-44. [399] K. Kondo, M. Kurihara, N. Miyata, T. Suzuki, M. Toyoda, Scavenging mechanisms of (-)-epigallocatechin gallate and (-)-epicatechin gallate on peroxyl radicals and formation of superoxide during the inhibitory action, Free Radical Bio Med, 27 (1999) 855-863.

184

[400] A. Dube, K. Ng, J.A. Nicolazzo, I. Larson, Effective use of reducing agents and nanoparticle encapsulation in stabilizing catechins in alkaline solution, Food chemistry, 122 (2010) 662-667. [401] J.F. Severino, B.A. Goodman, C.W.M. Kay, K. Stolze, D. Tunega, T.G. Reichenauer, K.F. Pirker, Free radicals generated during oxidation of green tea polyphenols: Electron paramagnetic resonance spectroscopy combined with density functional theory calculations, Free Radical Bio Med, 46 (2009) 1076-1088. [402] M. Suzuki, M. Sano, R. Yoshida, M. Degawa, T. Miyase, M. Maeda-Yamamoto, Epimerization of tea catechins and O-methylated derivatives of (-)-epigallocatechin-3-O-gallate: relationship between epimerization and chemical structure, Journal of agricultural and food chemistry, 51 (2003) 510-514. [403] R. Wang, W. Zhou, X. Jiang, Reaction kinetics of degradation and epimerization of epigallocatechin gallate (EGCG) in aqueous system over a wide temperature range, Journal of agricultural and food chemistry, 56 (2008) 2694-2701. [404] M. Muzolf-Panek, A. Gliszczynska-Swiglo, H. Szymusiak, B. Tyrakowska, The influence of stereochemistry on the antioxidant properties of catechin epimers, Eur Food Res Technol, 235 (2012) 1001-1009. [405] M.A. Timmel, J.A.W. Byl, N. Osheroff, Epimerization of Green Tea Catechins during Brewing Does Not Affect the Ability to Poison Human Type II Topoisomerases, Chemical research in toxicology, 26 (2013) 622-628. [406] S. Proniuk, B.M. Liederer, J. Blanchard, Preformulation study of epigallocatechin gallate, a promising antioxidant for topical skin cancer prevention, Journal of pharmaceutical sciences, 91 (2002) 111-116. [407] J. Zimeri, C.H. Tong, Degradation kinetics of (-)-epigallocatechin gallate as a function of pH and dissolved oxygen in a liquid model system, J Food Sci, 64 (1999) 753-758. [408] Y. Komatsu, S. Suematsu, Y. Hisanobu, H. Saigo, R. Matsuda, K. Hara, Studies on Preservation of Constituents in Canned Drinks .2. Effects of Ph and Temperature on Reaction-Kinetics of Catechins in Green Tea Infusion, Biosci Biotech Bioch, 57 (1993) 907-910. [409] J.F. Fangueiro, A. Parra, A.M. Silva, M.A. Egea, E.B. Souto, M.L. Garcia, A.C. Calpena, Validation of a high performance liquid chromatography method for the stabilization of epigallocatechin gallate, International journal of pharmaceutics, 475 (2014) 181-190. [410] N. Turkmen, F. Sari, Y.S. Velioglu, Effects of extraction solvents on concentration and antioxidant activity of black and black mate tea polyphenols determined by ferrous tartrate and Folin-Ciocalteu methods, Food chemistry, 99 (2006) 835-841. [411] A.K. Srivastava, P. Bhatnagar, M. Singh, S. Mishra, P. Kumar, Y. Shukla, K.C. Gupta, Synthesis of PLGA nanoparticles of tea polyphenols and their strong in vivo protective effect against chemically induced DNA damage, International journal of nanomedicine, 8 (2013) 1451-1462. [412] I.F. Tannock, D. Rotin, Acid pH in tumors and its potential for therapeutic exploitation, Cancer research, 49 (1989) 4373-4384. [413] N.W. Lutz, Y. Le Fur, J. Chiche, J. Pouyssegur, P.J. Cozzone, Quantitative in vivo characterization of intracellular and extracellular pH profiles in heterogeneous tumors: a novel method enabling multiparametric pH analysis, Cancer research, 73 (2013) 4616-4628. [414] Y.T. Konttinen, J. Mandelin, T.F. Li, J. Salo, J. Lassus, M. Liljestrom, M. Hukkanen, M. Takagi, I. Virtanen, S. Santavirta, Acidic cysteine endoproteinase cathepsin K in the degeneration of the superficial articular hyaline cartilage in osteoarthritis, Arthritis and rheumatism, 46 (2002) 953-960. [415] R. Wang, W.B. Zhou, X.H. Jiang, Reaction kinetics of degradation and epimerization of epigallocatechin gallate (EGCG) in aqueous system over a wide temperature range, Journal of agricultural and food chemistry, 56 (2008) 2694-2701. [416] A. Carreau, B. El Hafny-Rahbi, A. Matejuk, C. Grillon, C. Kieda, Why is the partial oxygen pressure of human tissues a crucial parameter? Small molecules and hypoxia, Journal of cellular and molecular medicine, 15 (2011) 1239-1253. [417] T. Hatano, M. Tsugawa, M. Kusuda, S. Taniguchi, T. Yoshida, S. Shiota, T. Tsuchiya, Enhancement of antibacterial effects of epigallocatechin gallate, using ascorbic acid, Phytochemistry, 69 (2008) 3111-3116. [418] M.C. Bohin, J.P. Vincken, H.T.W.M. van der Hijden, H. Gruppen, Efficacy of Food Proteins as Carriers for Flavonoids, Journal of agricultural and food chemistry, 60 (2012) 4136-4143.

185

[419] K.H. van het Hof, G.A. Kivits, J.A. Weststrate, L.B. Tijburg, Bioavailability of catechins from tea: the effect of milk, European journal of clinical nutrition, 52 (1998) 356-359. [420] Z. Hong, Y. Xu, J.F. Yin, J. Jin, Y. Jiang, Q. Du, Improving the effectiveness of (-)-epigallocatechin gallate (EGCG) against rabbit atherosclerosis by EGCG-loaded nanoparticles prepared from chitosan and polyaspartic acid, Journal of agricultural and food chemistry, 62 (2014) 12603-12609. [421] W. Wisuitiprot, A. Somsiri, K. Ingkaninan, N. Waranuch, A novel technique for chitosan microparticle preparation using a water/silicone emulsion: green tea model, International journal of cosmetic science, 33 (2011) 351-358. [422] F. Lemarie, C.W. Chang, D.R. Blatchford, R. Amor, G. Norris, L. Tetley, G. McConnell, C. Dufes, Antitumor activity of the tea polyphenol epigallocatechin-3-gallate encapsulated in targeted vesicles after intravenous administration, Nanomedicine, 8 (2013) 181-192. [423] I.A. Siddiqui, V.M. Adhami, D.J. Bharali, B.B. Hafeez, M. Asim, S.I. Khwaja, N. Ahmad, H. Cui, S.A. Mousa, H. Mukhtar, Introducing nanochemoprevention as a novel approach for cancer control: proof of principle with green tea polyphenol epigallocatechin-3-gallate, Cancer research, 69 (2009) 1712-1716. [424] J.Y. Fang, W.R. Lee, S.C. Shen, Y.L. Huang, Effect of liposome encapsulation of tea catechins on their accumulation in basal cell carcinomas, Journal of dermatological science, 42 (2006) 101-109. [425] S. Zhu, Y. Li, Z. Li, C.Y. Ma, Z.X. Lou, W. Yokoyama, H.X. Wang, Lipase-catalyzed synthesis of acetylated EGCG and antioxidant properties of the acetylated derivatives, Food Res Int, 56 (2014) 279-286. [426] Y. Zhong, Y.S. Chiou, M.H. Pan, F. Shahidi, Anti-inflammatory activity of lipophilic epigallocatechin gallate (EGCG) derivatives in LPS-stimulated murine macrophages, Food chemistry, 134 (2012) 742-748. [427] Y. Zhong, F. Shahidi, Lipophilised epigallocatechin gallate (EGCG) derivatives and their antioxidant potential in food and biological systems, Food chemistry, 131 (2012) 22-30. [428] A.J. Smith, P. Kavuru, K.K. Arora, S. Kesani, J. Tan, M.J. Zaworotko, R.D. Shytle, Crystal engineering of green tea epigallocatechin-3-gallate (EGCg) cocrystals and pharmacokinetic modulation in rats, Molecular pharmaceutics, 10 (2013) 2948-2961. [429] A. Dube, J.A. Nicolazzo, I. Larson, Chitosan nanoparticles enhance the intestinal absorption of the green tea catechins (+)-catechin and (-)-epigallocatechin gallate, European journal of pharmaceutical sciences : official journal of the European Federation for Pharmaceutical Sciences, 41 (2010) 219-225. [430] B. Hu, Y. Ting, X. Zeng, Q. Huang, Bioactive peptides/chitosan nanoparticles enhance cellular antioxidant activity of (-)-epigallocatechin-3-gallate, Journal of agricultural and food chemistry, 61 (2013) 875-881. [431] X. Luo, R. Guan, X. Chen, M. Tao, J. Ma, J. Zhao, Optimization on condition of epigallocatechin-3-gallate (EGCG) nanoliposomes by response surface methodology and cellular uptake studies in Caco-2 cells, Nanoscale research letters, 9 (2014) 291. [432] S. Rocha, R. Generalov, C. Pereira Mdo, I. Peres, P. Juzenas, M.A. Coelho, Epigallocatechin gallate-loaded polysaccharide nanoparticles for prostate cancer chemoprevention, Nanomedicine, 6 (2011) 79-87. [433] J.Y. Fang, C.F. Hung, T.L. Hwang, Y.L. Huang, Physicochemical characteristics and in vivo deposition of liposome-encapsulated tea catechins by topical and intratumor administrations, Journal of drug targeting, 13 (2005) 19-27. [434] J.F. Fangueiro, T. Andreani, L. Fernandes, M.L. Garcia, M.A. Egea, A.M. Silva, E.B. Souto, Physicochemical characterization of epigallocatechin gallate lipid nanoparticles (EGCG-LNs) for ocular instillation, Colloids and surfaces. B, Biointerfaces, 123 (2014) 452-460. [435] M.W. van Tulder, R.J. Scholten, B.W. Koes, R.A. Deyo, Nonsteroidal anti-inflammatory drugs for low back pain: a systematic review within the framework of the Cochrane Collaboration Back Review Group, Spine, 25 (2000) 2501-2513. [436] M.W. van Tulder, B.W. Koes, L.M. Bouter, Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions, Spine, 22 (1997) 2128-2156.

186

[437] P.C. Willems, J.B. Staal, G.H. Walenkamp, R.A. de Bie, Spinal fusion for chronic low back pain: systematic review on the accuracy of tests for patient selection, The spine journal : official journal of the North American Spine Society, 13 (2013) 99-109. [438] K. Masuda, Biological repair of the degenerated intervertebral disc by the injection of growth factors, European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society, 17 Suppl 4 (2008) 441-451. [439] O. Krupkova, S.J. Ferguson, K. Wuertz-Kozak, Stability of (−)-epigallocatechin gallate and its activity in liquid formulations and delivery systems, The Journal of nutritional biochemistry, 37 (2016) 1-12. [440] M. D'Este, M. Alini, D. Eglin, Single step synthesis and characterization of thermoresponsive hyaluronan hydrogels, Carbohydrate polymers, 90 (2012) 1378-1385. [441] M. Peroglio, S. Grad, D. Mortisen, C.M. Sprecher, S. Illien-Junger, M. Alini, D. Eglin, Injectable thermoreversible hyaluronan-based hydrogels for nucleus pulposus cell encapsulation, European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society, 21 Suppl 6 (2012) S839-849. [442] B. Dhandayuthapani, U.M. Krishnan, S. Sethuraman, Fabrication and characterization of chitosan-gelatin blend nanofibers for skin tissue engineering, Journal of biomedical materials research. Part B, Applied biomaterials, 94 (2010) 264-272. [443] L.T. Nguyen TH, Fabrication and Characterization of cross-linked gelatin electro-spun nanofibers, J. biomedical science and Engineering, 2 (2010) 1117-1124. [444] J. Walser, S.J. Ferguson, Oriented nanofibrous membranes for tissue engineering applications: Electrospinning with secondary field control, Journal of the mechanical behavior of biomedical materials, 58 (2016) 188-198. [445] M. Zamani, M.P. Prabhakaran, S. Ramakrishna, Advances in drug delivery via electrospun and electrosprayed nanomaterials, International journal of nanomedicine, 8 (2013) 2997-3017. [446] N. Okutan, P. Terzi, F. Altay, Affecting parameters on electrospinning process and characterization of electrospun gelatin nanofibers, Food Hydrocolloid, 39 (2014) 19-26. [447] D.R. Pereira, J. Silva-Correia, J.M. Oliveira, R.L. Reis, Hydrogels in acellular and cellular strategies for intervertebral disc regeneration, Journal of tissue engineering and regenerative medicine, 7 (2013) 85-98. [448] D.J. Gorth, R.L. Mauck, J.A. Chiaro, B. Mohanraj, N.M. Hebela, G.R. Dodge, D.M. Elliott, L.J. Smith, IL-1ra delivered from poly(lactic-co-glycolic acid) microspheres attenuates IL-1 beta-mediated degradation of nucleus pulposus in vitro, Arthritis research & therapy, 14 (2012). [449] C.Z. Liang, H. Li, Y.Q. Tao, L.H. Peng, J.Q. Gao, J.J. Wu, F.C. Li, J.M. Hua, Q.X. Chen, Dual release of dexamethasone and TGF-beta 3 from polymeric microspheres for stem cell matrix accumulation in a rat disc degeneration model, Acta biomaterialia, 9 (2013) 9423-9433. [450] E.M. Schutgens, M.A. Tryfonidou, T.H. Smit, F.C. Oner, A. Krouwels, K. Ito, L.B. Creemers, Biomaterials for Intervertebral Disc Regeneration: Past Performance and Possible Future Strategies, European cells & materials, 30 (2015) 210-231. [451] J.M. Cloyd, N.R. Malhotra, L. Weng, W. Chen, R.L. Mauck, D.M. Elliott, Material properties in unconfined compression of human nucleus pulposus, injectable hyaluronic acid-based hydrogels and tissue engineering scaffolds, European Spine Journal, 16 (2007) 1892-1898. [452] J.C. Iatridis, L.A. Setton, M. Weidenbaum, V.C. Mow, The viscoelastic behavior of the non-degenerate human lumbar nucleus pulposus in shear, Journal of biomechanics, 30 (1997) 1005-1013. [453] F. Kohori, K. Sakai, T. Aoyagi, M. Yokoyama, Y. Sakurai, T. Okano, Preparation and characterization of thermally responsive block copolymer micelles comprising poly(N-isopropylacrylamide-b-DL-lactide), Journal of Controlled Release, 55 (1998) 87-98. [454] O. Krupkova, J. Handa, M. Hlavna, J. Klasen, C. Ospelt, S.J. Ferguson, K. Wuertz-Kozak, The Natural Polyphenol Epigallocatechin Gallate Protects Intervertebral Disc Cells from Oxidative Stress, Oxidative medicine and cellular longevity, (2016). [455] K. Akeda, T. Imanishi, K. Ohishi, K. Masuda, A. Uchida, T. Sakakibara, Y. Kasai, A. Sudo, INTRADISCAL INJECTION OF AUTOLOGOUS SERUM ISOLATED FROM PLATELET‐RICH‐PLASMA

187

FOR THE TREATMENT OF DISCOGENIC LOW BACK PAIN: PRELIMINARY PROSPECTIVE CLINICAL TRIAL: GP141, Spine Journal Meeting Abstracts, (2011). [456] M. Gallucci, N. Limbucci, L. Zugaro, A. Barile, E. Stavroulis, A. Ricci, R. Galzio, C. Masciocchi, Sciatica: Treatment with intradiscal and intraforaminal injections of steroid and oxygen-ozone versus steroid only, Radiology, 242 (2007) 907-913. [457] N. Willems, H.Y. Yang, M.L.P. Langelaan, A.R. Tellegen, G.C.M. Grinwis, H.J.C. Kranenburg, F.M. Riemers, S.G.M. Plomp, E.G.M. Craenmehr, W.J.A. Dhert, N.E. Papen-Botterhuis, B.P. Meij, L.B. Creemers, M.A. Tryfonidou, Biocompatibility and intradiscal application of a thermoreversible celecoxib-loaded poly-N-isopropylacrylamide MgFe-layered double hydroxide hydrogel in a canine model, Arthritis research & therapy, 17 (2015). [458] F. Galbusera, H. Schmidt, C. Neidlinger-Wilke, A. Gottschalk, H.J. Wilke, The mechanical response of the lumbar spine to different combinations of disc degenerative changes investigated using randomized poroelastic finite element models, European Spine Journal, 20 (2011) 563-571. [459] J. Kraemer, D. Kolditz, R. Gowin, Water and Electrolyte Content of Human Intervertebral Disks under Variable Load, Spine, 10 (1985) 69-71. [460] M.J. Sullivan-Gunn, P.A. Lewandowski, Elevated hydrogen peroxide and decreased catalase and glutathione peroxidase protection are associated with aging sarcopenia, BMC geriatrics, 13 (2013). [461] M. Forster, F. Mahn, U. Gockel, M. Brosz, R. Freynhagen, T.R. Tolle, R. Baron, Axial Low Back Pain: One Painful Area - Many Perceptions and Mechanisms, PloS one, 8 (2013). [462] E. Ricciotti, G.A. FitzGerald, Prostaglandins and Inflammation, Arterioscl Throm Vas, 31 (2011) 986-1000. [463] B. St-Jacques, W.Y. Ma, Peripheral prostaglandin E2 prolongs the sensitization of nociceptive dorsal root ganglion neurons possibly by facilitating the synthesis and anterograde axonal trafficking of EP4 receptors, Exp Neurol, 261 (2014) 354-366. [464] V. Kumar, B.A. Mahal, NGF - the TrkA to successful pain treatment, Journal of pain research, 5 (2012) 279-287. [465] V. Freund, F. Pons, V. Joly, E. Mathieu, N. Martinet, N. Frossard, Upregulation of nerve growth factor expression by human airway smooth muscle cells in inflammatory conditions, The European respiratory journal, 20 (2002) 458-463. [466] B.R. Kraemer, S.O. Yoon, B.D. Carter, The biological functions and signaling mechanisms of the p75 neurotrophin receptor, Handbook of experimental pharmacology, 220 (2014) 121-164. [467] A.M. Skoff, J.E. Adler, Nerve growth factor regulates substance P in adult sensory neurons through both TrkA and p75 receptors, Exp Neurol, 197 (2006) 430-436. [468] X. Xifro, L. Vidal-Sancho, P. Boadas-Vaello, C. Turrado, J. Alberch, T. Puig, E. Verdu, Novel Epigallocatechin-3-Gallate (EGCG) Derivative as a New Therapeutic Strategy for Reducing Neuropathic Pain after Chronic Constriction Nerve Injury in Mice, PloS one, 10 (2015). [469] A. Areti, V.G. Yerra, V. Naidu, A. Kumar, Oxidative stress and nerve damage: role in chemotherapy induced peripheral neuropathy, Redox biology, 2 (2014) 289-295. [470] D. Cizkova, N. Lukacova, M. Marsala, J. Marsala, Neuropathic pain is associated with alterations of nitric oxide synthase immunoreactivity and catalytic activity in dorsal root ganglia and spinal dorsal horn, Brain research bulletin, 58 (2002) 161-171. [471] D. Levy, D.W. Zochodne, NO pain: potential roles of nitric oxide in neuropathic pain, Pain practice : the official journal of World Institute of Pain, 4 (2004) 11-18. [472] R.A. Floyd, Antioxidants, oxidative stress, and degenerative neurological disorders, Proceedings of the Society for Experimental Biology and Medicine. Society for Experimental Biology and Medicine, 222 (1999) 236-245. [473] W.M. Renno, M. Al-Maghrebi, A. Al-Banaw, (-)-Epigallocatechin-3-gallate (EGCG) attenuates functional deficits and morphological alterations by diminishing apoptotic gene overexpression in skeletal muscles after sciatic nerve crush injury, Naunyn-Schmiedeberg's archives of pharmacology, 385 (2012) 807-822. [474] M. Doita, T. Kanatani, T. Ozaki, N. Matsui, M. Kurosaka, S. Yoshiya, Influence of macrophage infiltration of herniated disc tissue on the production of matrix metalloproteinases leading to disc resorption, Spine, 26 (2001) 1522-1527.

188

[475] T. Ikeda, T. Nakamura, T. Kikuchi, S. Umeda, H. Senda, K. Takagi, Pathomechanism of spontaneous regression of the herniated lumbar disc: histologic and immunohistochemical study, Journal of spinal disorders, 9 (1996) 136-140. [476] N. Fujiwara, K. Kobayashi, Macrophages in inflammation, Current drug targets. Inflammation and allergy, 4 (2005) 281-286. [477] M. Schmidt, H. Schmitz, A. Baumgart, D. Guedon, M.I. Netsch, M.H. Kreuter, C.B. Schmidlin, D. Schrenk, Toxicity of green tea extracts and their constituents in rat hepatocytes in primary culture, Food and Chemical Toxicology, 43 (2005) 307-314. [478] A. Murakami, Dose-dependent functionality and toxicity of green tea polyphenols in experimental rodents, Archives of biochemistry and biophysics, 557 (2014) 3-10. [479] R. de la Torre, S. de Sola, G. Hernandez, M. Farre, J. Pujol, J. Rodriguez, J.M. Espadaler, K. Langohr, A. Cuenca-Royo, A. Principe, L. Xicota, N. Janel, S. Catuara-Solarz, G. Sanchez-Benavides, H. Blehaut, I. Duenas-Espin, L. Del Hoyo, B. Benejam, L. Blanco-Hinojo, S. Videla, M. Fito, J.M. Delabar, M. Dierssen, T.s. group, Safety and efficacy of cognitive training plus epigallocatechin-3-gallate in young adults with Down's syndrome (TESDAD): a double-blind, randomised, placebo-controlled, phase 2 trial, The Lancet. Neurology, 15 (2016) 801-810. [480] A. Mahler, S. Mandel, M. Lorenz, U. Ruegg, E.E. Wanker, M. Boschmann, F. Paul, Epigallocatechin-3-gallate: a useful, effective and safe clinical approach for targeted prevention and individualised treatment of neurological diseases?, The EPMA journal, 4 (2013) 5. [481] C.A. Elmets, D. Singh, K. Tubesing, M. Matsui, S. Katiyar, H. Mukhtar, Cutaneous photoprotection from ultraviolet injury by green tea polyphenols, Journal of the American Academy of Dermatology, 44 (2001) 425-432. [482] T.D. Shanafelt, Y.K. Lee, T.G. Call, G.S. Nowakowski, D. Dingli, C.S. Zent, N.E. Kay, Clinical effects of oral green tea extracts in four patients with low grade B-cell malignancies, Leukemia research, 30 (2006) 707-712. [483] T.D. Shanafelt, T.G. Call, C.S. Zent, B. LaPlant, D.A. Bowen, M. Roos, C.R. Secreto, A.K. Ghosh, B.F. Kabat, M.J. Lee, C.S. Yang, D.F. Jelinek, C. Erlichman, N.E. Kay, Phase I trial of daily oral Polyphenon E in patients with asymptomatic Rai stage 0 to II chronic lymphocytic leukemia, Journal of clinical oncology : official journal of the American Society of Clinical Oncology, 27 (2009) 3808-3814. [484] A.L. Brown, J. Lane, J. Coverly, J. Stocks, S. Jackson, A. Stephen, L. Bluck, A. Coward, H. Hendrickx, Effects of dietary supplementation with the green tea polyphenol epigallocatechin-3-gallate on insulin resistance and associated metabolic risk factors: randomized controlled trial, The British journal of nutrition, 101 (2009) 886-894. [485] K.C. Maki, M.S. Reeves, M. Farmer, K. Yasunaga, N. Matsuo, Y. Katsuragi, M. Komikado, I. Tokimitsu, D. Wilder, F. Jones, J.B. Blumberg, Y. Cartwright, Green tea catechin consumption enhances exercise-induced abdominal fat loss in overweight and obese adults, The Journal of nutrition, 139 (2009) 264-270. [486] C.H. Hsu, T.H. Tsai, Y.H. Kao, K.C. Hwang, T.Y. Tseng, P. Chou, Effect of green tea extract on obese women: a randomized, double-blind, placebo-controlled clinical trial, Clinical nutrition, 27 (2008) 363-370. [487] D. Oehme, T. Goldschlager, P. Ghosh, J.V. Rosenfeld, G. Jenkin, Cell-Based Therapies Used to Treat Lumbar Degenerative Disc Disease: A Systematic Review of Animal Studies and Human Clinical Trials, Stem Cells Int, (2015). [488] C.W.A. Pfirrmann, A. Metzdorf, M. Zanetti, J. Hodler, N. Boos, Magnetic resonance classification of lumbar intervertebral disc degeneration, Spine, 26 (2001) 1873-1878. [489] A.J. Michalek, M.R. Buckley, L.J. Bonassar, I. Cohen, J.C. Iatridis, The effects of needle puncture injury on microscale shear strain in the intervertebral disc annulus fibrosus, Spine Journal, 10 (2010) 1098-1105. [490] H.G. Deen, D.S. Fenton, T.J. Lamer, Minimally invasive procedures for disorders of the lumbar spine, Mayo Clinic proceedings, 78 (2003) 1249-1256. [491] E.J. Carragee, A.S. Don, E.L. Hurwitz, J.M. Cuellar, J. Carrino, R. Herzog, 2009 ISSLS Prize Winner: Does Discography Cause Accelerated Progression of Degeneration Changes in the Lumbar Disc A Ten-Year Matched Cohort Study, Spine, 34 (2009) 2338-2345. [492] J.B. Staal, R.A. de Bie, H.C. de Vet, J. Hildebrandt, P. Nelemans, Injection therapy for subacute and chronic low back pain: an updated Cochrane review, Spine, 34 (2009) 49-59.

189

[493] S.M. Sinclair, J. Bhattacharyya, J.R. McDaniel, D.M. Gooden, R. Gopalaswamy, A. Chilkoti, L.A. Setton, A genetically engineered thermally responsive sustained release curcumin depot to treat neuroinflammation, Journal of Controlled Release, 171 (2013) 38-47. [494] J.D. Lambert, D.H. Kim, R. Zheng, C.S. Yang, Transdermal delivery of (-)-epigallocatechin-3-gallate, a green tea polyphenol, in mice, The Journal of pharmacy and pharmacology, 58 (2006) 599-604. [495] E. Hong Byun, Y. Fujimura, K. Yamada, H. Tachibana, TLR4 signaling inhibitory pathway induced by green tea polyphenol epigallocatechin-3-gallate through 67-kDa laminin receptor, J Immunol, 185 (2010) 33-45. [496] H. Tachibana, Green tea polyphenol EGCG signaling pathway through the 67-kDa laminin receptor, Nihon yakurigaku zasshi. Folia pharmacologica Japonica, 132 (2008) 145-149. [497] M. Schieber, N.S. Chandel, ROS function in redox signaling and oxidative stress, Current biology : CB, 24 (2014) R453-462. [498] R.F. McLain, L. Kapural, N.A. Mekhail, Epidural steroids for back and leg pain: Mechanism of action and efficacy, Clev Clin J Med, 71 (2004) 961-970. [499] B.F. Cullen, R.H. Haschke, Local-Anesthetic Inhibition of Phagocytosis and Metabolism of Human Leukocytes, Anesthesiology, 40 (1974) 142-146. [500] C. Eder, A. Pinsger, S. Schildboeck, E. Falkner, P. Becker, M. Ogon, Influence of intradiscal medication on nucleus pulposus cells, Spine Journal, 13 (2013) 1556-1562.

191

Appendix

193

List of abbreviations

A/PI Annexin V-FITC/Propidium Iodide

AA Ascorbic acid

ADAMTs A disintegrin and metalloproteinases with thrombospondin motifs

AF Annulus fibrosus

Akt Protein kinase B

AP-1 Activator protein 1

APS Average particle size

BHT Butylated dihydroxytoluen

BDNF Brain-derived neurotrophic factor

BMP Bone morfogenic protein

COMT Catechol-O-methyltransferase

COX Clooxygenase

CPP Caseinophosphopeptide

CS Chitosan

CTAB Cetyltrimethylammonium bromide

DA Deoxycholic acid

DDAB Dimethyldioctadecylammonium bromide

DDD Degenerative disc disease

DMBA Dimethylbenz(a)anthracene

DMEM Dulbecco’s Modified Eagle’s Medium

DMMB Dimethylmethylene blue

DP Docosapentaenoic acid

DRG Dorsal root ganglion

EE Encapsulation efficiency

EGCG Epigallocatechin 3-gallate

ELISA Enzyme-linked immunosorbent assay

EPR Electron paramagnetic resonance

FGF Fibroblast growth factor

GA Glutaraldehyde

194

GAG Glycosaminoglycan

GCG (–)-gallocatechin gallate

GI Gastrointestinal

GTC Green tea catechins

HMGB1 High-mobility group protein B1

HPLC High-performance liquid chromatography

HYP Hydroxyproline

IGF Insuline growth factor

IKK Inhibitor of κB kinase

IL Interleukin

ILEX Isolated lumbar extension

IP Intraperitoneal

IRAK-1 Interleukin1 receptor-associated kinase 1

IVD Intervertebral disc

JC-1 5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolcarbocyanine iodide

JNK c-Jun N-terminal protein kinase

LBP Low back pain

LE Loading efficiency

LinkN N-terminus of Link protein

LNs Lipid nanoparticles

LY LY294002

MAPK Mitogen-activated protein kinase

MLL Mixed-lineage leukemia

MMP Matrix metalloproteinase

MRP Multidrug resistance-associated protein

MTT 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide

NF‑κB Nuclear factor kappa-light-chain-enhancer of activated B cells

NGF Nerve growth factor

NO Nitric oxide

iNOS Inducible nitric oxide synthase

NP Nucleus pulposus

195

OA Osteoarthritis

P-gp P-glycoprotein

PAA Polyaspartic acid

PCL Polycaprolactone

PEG Polyethylene glycol

Pgal Propyl gallate

PGES Prostaglandin E synthase (COX-2)

PI Polydispersity index

PLA Polylactic acid (PEG)

PLGA Poly(lactide-co-glycolic acid)

PTGS2 Prostaglandin endoperoxide synthase 2 (COX-2)

ROS Reactive oxygen species

RT Room temperature

SA β-gal Senescence-associated β-galactosidase

SASP Senescence-activated secretory phenotype

SEM Scanning electron microscopy

SGF Simulated gastric fluid

SIF Simulated intestinal fluid

SOD Superoxide dismutase

SULT Phenolsulfotransferases

TBP Tata box-binding protein

TIMP Tissue inhibitor of matrix metalloproteinase

TLRs Toll-like receptors

TNF-α Tumor necrosis factor alpha

UGT UDP-glucuronosyltransferases

ZP zeta potential

196

Acknowledgement

First and foremost, I would like to thank my doctoral supervisors, Prof. Dr. Karin Wuertz-Kozak

and Prof. Dr. Stephen Ferguson for the opportunity to work under their guidance. I would like to

thank them for their understanding, support, and encouragement throughout the project.

My acknowledgement also belongs to Dr. Junichi Handa and Dr. Marian Hlavna for their work

and discussions that contributed to this thesis. I also would like to thank my collaborators Dr.

Juergen Klasen (Spital Zollikerberg), Dr. Miho Sekiguchi (Fukushima Medical University), Prof.

Keita Ito (Technical University Eindhoven), and Dr. David Eglin (AO Research Institute Davos), for

their inputs to the studies, revisions of manuscripts, and provided materials. I would like to

thank Helen Greutert for technical support.

I appreciate the students Sophie Chabloz, Aikaterini Zafeiroupoulpu, Julie Amir Tahmasseb,

Sonia Rossi, Seraina Domenig, Joel Zvick and Dominik Kunz, for their work within the

framework of this thesis. Their thoughtful remarks were very helpful, and some results of their

work are adopted in this thesis. In addition, I would like to thank Dr .Karin Wuertz-Kozak, Dr.

Ilsoo Koh and Dr. Hadi Seyed Hosseini for proofreading my thesis.

A special acknowledgement extends to the funding agencies: European Union/Marie Curie

action (FPT7-PITN-GA-2009-238690-SPINEFX), the Herzog-Egli foundation, and the Mäxi

Foundation, for providing the financial support of the studies.

I would like to sincerely thank my friends Dominika Ignasiak, Ilsoo Koh, Patrick Vogt, and Yabin

Wu for listening, advices, and valuable discussions about work and daily life throughout the

years of my PhD.

Last but not least, I would like to thank Hadi and my family for their patience, understanding

and endless support.

197

Curriculum Vitae

Olga Krupkova

born 1.9. 1986 in Brno, Czech Republic.

EDUCATION

2013-2016 PhD candidate, Institute for Biomechanics, ETH Zurich, Switzerland

Supervisors: Prof. Dr. Stephen John Ferguson, Prof. Dr. Karin Wuertz-Kozak

Thesis: Activity and controlled delivery of epigallocatechin 3-gallate in the

treatment of degenerative disc disease

2012-2013 SpineFX Marie Curie training fellow, Institute for Biomechanics, ETH Zurich,

Switzerland

Supervisors: Prof. Dr. Stephen John Ferguson, Prof. Dr. Karin Wuertz-Kozak

Project: Testing the anti-inflammatory effects of epigallocatechin 3-gallate in

intervertebral disc

2010-2011 Research assistant, Faculty of Medicine, Masaryk University, Brno, Czech

Republic

Supervisor: Dr. Stjepan Uldrijan

Project: Therapeutic potential of p38 MAPK inhibition in melanoma

2005-2010 BSc and MSc in Molecular Biology and Genetics, Faculty of Science, Masaryk

University, Brno, Czech Republic

Supervisor: Prof. Dr. Renata Veselská

Thesis: Expression and subcellular localization of nestin in tumor cells

HONORS AND AWARDS

2015 Travel award for participation at winter school of Electrospin paramagnetic

resonance, Ascona, Switzerland.

2014 2nd best student oral presentation award at the 15th eCM conference: Cartilage &

Disc: Repair and Regeneration, Davos, Switzerland.

2010 Special Rector´s research award for the best diploma thesis in the study

program Molecular Biology and Genetics, Masaryk University, Brno.

2010 Dean´s Award for excellent organization of The Student Scientific Conference on

Genetically Modified Organisms, Masaryk University, Brno.

198

PUBLICATIONS

Published

2016 Krupkova O, Ferguson SJ, Wuertz-Kozak K. Stability of (–)-epigallocatechin gallate

and its activity in liquid formulations and delivery systems – a review. J Nutr

Biochem. 2016 Nov(37):1–12.

2016 Handa J, Sekiguchi M, Krupkova O, Konno S. The effect of serotonin noradrenaline

reuptake inhibitor duloxetine on the intervertebral disc-related radiculopathy in

rats. Eur Spine J. 2016 Mar(3):877-87.

2016 Krupkova O, Handa J, Hlavna M, Klasen J, Ospelt C, Ferguson SJ, Wuertz-Kozak K.

The natural polyphenol epigallocatechin gallate protects intervertebral disc cells

from oxidative stress. Oxid Med Cell Longev. 2016 Jan (9):1-17.

2014 Krupkova O, Sekiguchi M, Klasen J, Hausmann O, Konno S, Ferguson SJ, Wuertz

K. Epigallocatechin 3-gallate supresses Interleukin-1β-induced inflammatory

responses of intervertebral disc cells in vitro and reduces radiculopathic pain in

rats. Eur Cell Materials. 2014. 28: 372-86.

2014 Klawitter M, Hakozaki M, Kobayashi H Krupkova O et al., Expression and

regulation of toll-like receptors (TLRs) in human intervertebral disc cells. Eur Spine

J. 2014. 23 (9) 1878-91.

2011 Krupkova O, Loja T, Redova M, Neradil J, Zitterbart K, Sterba J, Veselska R. Analysis

of nuclear nestin localization in cell lines derived from neurogenic tumors. Tumour

Biol. 2011. Aug(4):631-9.

2010 Krupkova O, Loja T, Zambo I, Veselska R. Nestin expression in human tumors and

tumor cell lines. Neoplasma. 2010. 57(4): 291-8.

Under review

2016 Krupkova O, Hlavna M, Amir Tahmasseb J, Zvick J, Kunz D, Ferguson SJ, Wuertz-

Kozak K. An inflammatory nucleus pulposus tissue culture model to test molecular

regenerative therapies: Validation with epigallocatechin 3-gallate.

PRESENTATIONS

Conferences

2015 Krupkova O, Sekiguchi M, Handa J, Klasen J, Konno S, Wuertz-Kozak K, Ferguson

SJ. Polyphenol Epigallocatechin 3-gallate Significantly Decreases Inflammation

and Oxidative Stress in Human Intervertebral Disc Cells. Podium presentation at

the 3rd International ORS Spine Symposium, Philadelphia, USA.

199

2015 Krupkova O, Handa J, Hlavna M, Klasen J, Wuertz-Kozak K, Ferguson SJ.

Combating oxidative stress-induced premature senescence in the intervertebral

disc. Podium presentation at the Eurospine, Copenhagen, Denmark.

2014 Krupkova O, Sekiguchi M, Klasen J, Konno S, Wuertz-Kozak K, Ferguson SJ.

Epigallocatechin 3-gallate suppresses interleukin-1β-induced inflammatory

responses in intervertebral disc cells in vitro and reduces radiculopathic pain in

vivo. Podium presentation at the 15th Cartilage & Disc: Repair and Regeneration,

Davos, Switzerland.

2013 Krupkova O, Sekiguchi M, Klasen J, Konno S, Wuertz-Kozak K, Ferguson SJ. The

potential of EGCG for the treatment of painful disc degeneration. Podium

presentation at the Swiss-Japanese Symposium on Musculoskeletal Research,

ETH Zurich, Switzerland.

2013 Krupkova O, Klasen J, Wuertz-Kozak K, Ferguson SJ. The potential of

Epigallocatechin 3-gallate for the treatment of painful intervertebral disc

degeneration. Podium presentation at the Eurospine, Liverpool, UK.

Guest lectures

2016 Targeting degenerative disc disease with bio-active compounds. Center for

Applied Biotechnology and Molecular Medicine (CABMM), University of Zurich

STUDENT SUPERVISION

Master Theses

2016 Aikaterini Zafeiroupoulou (TU Delft/ETH Zurich). Controlled delivery of anti-

inflammatory compounds as a therapeutic strategy for degenerative disc disease.

2015 Sophie Chabloz (ETH Zurich). Investigating the anti-aging effects of resveratrol on

intervertebral disc cells.

Semester Projects

2016 Julie Amir Tahmasseb (ETH Zurich). Anti-inflammatory activity of

epigallocatechin gallate in bovine disc organ culture.

2013 Ion Tripa (ETH Zurich). Anti-oxidant and DNA-protective activity of resveratrol in

human intervertebral disc cells.

Internships

2016 Joel Zvick (ETH Zurich). Analysis of glycosaminoglycan composition in

degenerative nucleus pulposus culture in artificial annulus fibrosus.

200

2016 Dominik Kunz (Zurich University of Applied Sciences). Analysis of collagens and

water content in degenerative nucleus pulposus culture in artificial annulus

fibrosus 2016)

2016 Seraina Domenig (ETH Zurich). Effect of curcumin on apoptotic and survival

pathways in intervertebral disc cells.

2014 Sonia Rossi (Kungsholmen gymnasium Stockholm, Sweden). The effects of

natural compounds on intervertebral disc degeneration.

ETH Z...Diss. ETH. No. 23749 Activity and controlled delivery of epigallocatechin 3-gallate in the...

Documents

Transcript of ETH Z...Diss. ETH. No. 23749 Activity and controlled delivery of epigallocatechin 3-gallate in the...