Diamond Magnetometry for sensing in biological environment

University of Groningen

Diamond magnetometry for sensing in biological environmentPerona Martinez, Felipe

DOI:10.33612/diss.111974782

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2020

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Perona Martinez, F. (2020). Diamond magnetometry for sensing in biological environment. RijksuniversiteitGroningen. https://doi.org/10.33612/diss.111974782

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license.More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-amendment.

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 17-03-2022

https://doi.org/10.33612/diss.111974782

https://research.rug.nl/en/publications/269d62d9-4c9d-44ba-8b1f-70817e2d418b

https://doi.org/10.33612/diss.111974782

Diamond Magnetometryfor sensing in biological

environment

Felipe Perona Martınez

The research presented in this thesis was carried out in the Bioimagingand Bioanalysis group at the Department of Biomedical Engineering ofthe University of Groningen. More information about this group can befound in the web site bioanalysisgroup.com.Felipe Perona Martınez was supported from the Chilean government via aCONICYT scholarship (Grant No. 72160222).

Copyright c© 2020 by Felipe Patricio Perona Martınez. All rightsreserved. No part of this publication may be reproduced, stored in aretrieval system or transmitted in any form or by any means, withoutpermission of the author and, when appropriate, the publisher holding thecopyrights of the published articles.

Cover : Felipe Perona MartınezISBN : 978-94-034-2258-9ISBN digital : 978-94-034-2257-2

This book was made in LATEXPrinted in the Netherlands by Ipskamp Printing.

Diamond Magnetometryfor sensing in biological

environment

Proefschrift

ter verkrijging van de graad van doctor aan deRijksuniversiteit Groningen

op gezag van derector magnificus prof. dr. C. Wijmenga

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

woensdag 22 januari 2020 om 14:30 uur

door

Felipe Patricio Perona Martınez

geboren op 29 september 1979te Recoleta, Chili

PromotorProf. dr. R. Schirhagl

CopromotorDr. C.G.M. Mignon

BeoordelingscommissieProf. dr. Y. RenProf. dr. P. RudolfProf. dr. K. Haenen

ParanimfenKiran van der LaanAlina Sigaeva

To Mariella

Contents

1 Introduction 1

2 Improving the cellular uptake 5

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.2 Results and discussion . . . . . . . . . . . . . . . . . . . . . 9

2.2.1 Characterization of coated nanodiamonds with AFM 9

2.2.2 Colloidal stability of protein-based polymer coatednanodiamonds . . . . . . . . . . . . . . . . . . . . . 10

2.2.3 Binding strength of protein polymers to nanodiamonds 12

2.2.4 Photoluminescence properties of protein polymer coa-ted nanodiamonds . . . . . . . . . . . . . . . . . . . 13

2.2.5 Cytotoxicity of protein polymer coated nanodiamonds 14

2.2.6 Cell uptake of bare and protein polymer coated nan-odiamonds . . . . . . . . . . . . . . . . . . . . . . . 15

2.3 Experimental Details . . . . . . . . . . . . . . . . . . . . . . 18

2.3.1 Materials . . . . . . . . . . . . . . . . . . . . . . . . 18

2.3.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . 19

2.4 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . 21

2.5 Supporting Information . . . . . . . . . . . . . . . . . . . . 23

2.5.1 Binding of the C4 protein polymer . . . . . . . . . . 23

2.5.2 Clusters of nanodiamonds . . . . . . . . . . . . . . . 23

2.5.3 Magneto-optical properties . . . . . . . . . . . . . . 24

2.5.4 Protein polymer binding stability at different pH . . 25

3 Sample Preparation in Proteomics 33

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

3.2 Materials and Methods . . . . . . . . . . . . . . . . . . . . . 36

3.2.1 Materials . . . . . . . . . . . . . . . . . . . . . . . . 36

3.2.2 Sample preparation . . . . . . . . . . . . . . . . . . 37

3.2.3 Sample preparation with carbon black . . . . . . . . 37

3.2.4 Protein analysis . . . . . . . . . . . . . . . . . . . . 37

vii

CONTENTS

3.2.5 Data processing . . . . . . . . . . . . . . . . . . . . . 383.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

4 Detecting free radicals 534.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 554.2 Results and discussion . . . . . . . . . . . . . . . . . . . . . 56

4.2.1 Dependence of the size of the nanodiamonds on therelaxation constant T1 . . . . . . . . . . . . . . . . . 58

4.2.2 Performance of the NVs in biologically relevant con-ditions . . . . . . . . . . . . . . . . . . . . . . . . . . 59

4.2.3 Measuring free radicals in situ . . . . . . . . . . . . 624.3 Experimental Details . . . . . . . . . . . . . . . . . . . . . . 65

4.3.1 Nanodiamonds . . . . . . . . . . . . . . . . . . . . . 654.3.2 T1 measurements . . . . . . . . . . . . . . . . . . . . 654.3.3 GdCl3 sensitivity in different conditions . . . . . . . 654.3.4 Hydrogen Peroxide and UV light . . . . . . . . . . . 664.3.5 Measuring the hydroxyl radical by HPF . . . . . . . 664.3.6 Measuring the concentration of hydroxyl radical by

HTA . . . . . . . . . . . . . . . . . . . . . . . . . . . 664.3.7 Measuring the Fenton reaction by diamond magne-

tometry . . . . . . . . . . . . . . . . . . . . . . . . . 674.4 Concluding remarks . . . . . . . . . . . . . . . . . . . . . . 674.5 Supporting Information . . . . . . . . . . . . . . . . . . . . 68

4.5.1 Measuring free radicals in-situ . . . . . . . . . . . . 684.5.2 Measuring the concentration of hydroxyl radical by

HTA . . . . . . . . . . . . . . . . . . . . . . . . . . . 684.5.3 Size distribution of the nanodiamonds . . . . . . . . 70

5 Discussion 75

6 Summary — Samenvatting 816.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816.2 Samenvatting . . . . . . . . . . . . . . . . . . . . . . . . . . 84

Acknowledgments 89

About the author 93

viii

1Introduction

Their crystal clear appearance, surrounded by a glare of colourful raysof light have captured the attention of people since thousands of years

ago. Today, diamonds offer much more than their aesthetic beauty. Theyare a crucial element in our technology. Because of their extensive list ofphysical properties, diamonds are found in a wide range of devices, fromdrill bits to sophisticated sensors.

Biomedicine has been taking advantages of the diamond’s propertiesfor different purposes. The hardness of the diamond has been used toextend the lifetime of implants[1]. Its biocompatibility has been exploitedto build substrates for growing diverse cell types such as fibroblasts[2], HeLacells and osteoblasts[3]. Its stable chemical surface has been used to carrymolecules in drug delivery devices[4].

Although the highly organised structure of the diamond lattice is re-sponsible for most of its physical properties, the defects found in the crys-tal enable it with additional optical attributes. One of this defect is theNitrogen-Vacancy center (NV center). The NV center is a point defectoccurring in the crystal lattice of diamonds, it consists of a vacancy withone atom of nitrogen substituting carbon in two adjacent lattice sites. Thisconfiguration presents exceptional optical properties. First, it is photolu-minescent, it emits one red photon after being exited by a green one. Thismeans that it is possible to measure a small magnetic field by measuring alarge optical signal that is related to it. Since optical signals are easier tomeasure and can be measured more sensitively, we achieve many orders ofmagnitude better sensitivities[5, 6] than with conventional MRI. Moreover,the intensity of photoluminescence is modulated by a magnetic momentpresent at the same location of the NV center. This attribute is the basicconcept behind Diamond Magnetometry (DM), i.e. measuring magneticfield through NV centers in diamonds.

The project conducted throughout the prosecution of the PhD aimed to

1

INTRODUCTION

research the implementation of Diamond Magnetometry to detect (oxygen)free radical inside living cells.

The main hypothesis that sustains the idea of detecting of (oxygen) freeradicals by diamond magnetometry is the fact that the missing electron ofthe radical grants a magnetic moment to the molecule, which interacts withthe NV centers allowing its detection. The research presented in this bookhas proven that this hypothesis is true, but also it has yielded additionalknowledge about the interaction of nanodiamonds with biomolecules andcells.

The intention of focusing the application of diamond magnetometry todetect (oxygen) free radicals is to overcome the limitations present in thecurrent measuring techniques.

The cell produces Oxygen Free Radicals (OFR) in a regulated physio-logical process, but the loss of balance between its production and clear-ance is associated with oxidative stress, which triggers the development ofpathologies[7]. The research of that relationship has linked the altered reg-ulation of OFR with the origin of cancer, cardiovascular diseases, diabetesand neurological disorders[8].

Considering the key role that OFR play, it is of great importance to im-prove the pool of resources available for their study. Diamond magnetome-try offers the possibility to fill the gaps that other techniques cannot resolve.The use of nanodiamonds implanted with NV centers allows measuring atnanometric scale, at a diffraction-limited resolution. This attribute mightallow determining the oxidative stress of a cell at specific places. Moreover,the NV center also offers to improve the time resolution, allowing real-timemeasuring.

The works collected in this volume can be seen as landmarks over thepath that has been built during the prosecution of the PhD project. Un-der that point of view, this book reflects how the Diamond magnetometrytechnique has been developed in our laboratories. Starting from funda-mental questions, such as: “How to place the nanodiamonds inside thecells?” or “Is the functionalization of the nanodiamonds interfering withtheir magneto-optical properties?”, we progressed to actually use the NVcenters hosted in the nanodiamonds as sensors of magnetic noise, which wecan use to detect and measure chemical reactions in inert samples or evenin living cells.

Each chapter of this book reports important scientific contributionsto the field of Diamond Magnetometry (DM) applied to cell biology, butalso they show implicitly the state of the technical development of thistechnique in our facilities. To be able to operate over the NV center we have

2

INTRODUCTION

designed and built the magnetometers, we have programmed the controlsystem that coordinates the instrumentation and all the software needed tooperate the new device. Also, the analysis of the photoluminescent signalwas completely built in house as part of my work as PhD student. Albeitthe similitude of our system with the standard magnetometer used in thecommunity, the biological nature of our samples, and the use of ensemblesof NV centers in nanodiamonds, instead of bulk diamonds, set particularrestrictions and considerations that make our system unique.

At the time when this book was printed, two of its chapters were alreadypublished in a high impact journal, and a third one was under peer-reviewevaluation in another important journal. Chapter 2 explores the use ofrecombinant proteins as coatings for nanodiamonds. The study is centredon the effect of the coatings in the colloidal stability of the nanoparticles,and the influence of these coatings in the particle uptake of two differentcell lines. In Chapter 3 an innovative application for nanodiamonds isproposed. The investigations started by determining the proteins that at-tach stronger to the surface of the untreated nanodiamond, then we havedetermined that several of these proteins match with biomarkers used todetect diseases. These results suggest the use of nanodiamonds to depletehigh abundant proteins from a sample. Chapter 4 presents the first re-sults of using Diamond Magnetometry to detect a free radical produced inreal-time under physiological conditions. This is an important landmarktowards the measurement of free radical inside cells. Finally, Chapter 5discusses the implications of these investigations, how are they related andhow this research enables further developments in diamond magnetometry.

3

INTRODUCTION

References

[1] M. D. Fries and Y. K. Vohra, “Nanostructured diamond film deposi-tion on curved surfaces of metallic temporomandibular joint implant,”Journal of Physics D: Applied Physics, vol. 35, no. 20, 2002.

[2] B. Shi, Q. Jin, L. Chen, and O. Auciello, “Fundamentals of ultra-nanocrystalline diamond (UNCD) thin films as biomaterials for develop-mental biology: Embryonic fibroblasts growth on the surface of (UNCD)films,” Diamond and Related Materials, vol. 18, no. 2-3, pp. 596–600,2009.

[3] P. Bajaj, D. Akin, A. Gupta, D. Sherman, B. Shi, O. Auciello, andR. Bashir, “Ultrananocrystalline diamond film as an optimal cell inter-face for biomedical applications,” Biomedical Microdevices, vol. 9, no. 6,pp. 787–794, 2007.

[4] R. A. Shimkunas, E. Robinson, R. Lam, S. Lu, X. Xu, X. Q. Zhang,H. Huang, E. Osawa, and D. Ho, “Nanodiamond-insulin complexes aspH-dependent protein delivery vehicles,” Biomaterials, vol. 30, no. 29,pp. 5720–5728, 2009.

[5] L. Ciobanu, D. A. Seeber, and C. H. Pennington, “3D MR microscopywith resolution 3.7 µm by 3.3 µm by 3.3 µm,” Journal of MagneticResonance, vol. 158, no. 1-2, pp. 178–182, 2002.

[6] P. Glover and S. P. Mansfield, “Limits to magnetic resonance mi-croscopy,” Reports on Progress in Physics, vol. 65, no. 10, pp. 1489–1511, 2002.

[7] W. Droge, “Free radicals in the physiological control of cell function,”Physiological Reviews, vol. 82, no. 1, pp. 47–95, 2002.

[8] M. Valko, D. Leibfritz, J. Moncol, M. T. Cronin, M. Mazur, andJ. Telser, “Free radicals and antioxidants in normal physiological func-tions and human disease,” International Journal of Biochemistry andCell Biology, vol. 39, no. 1, pp. 44–84, 2007.

4

2Recombinant protein polymers for colloidal stabilization

and improvement of cellular uptake of diamondnanosensors

Tingting Zhenga,c,⊥, Felipe Perona Martınezb,⊥, Ingeborg Maria Storma, WolfRomboutsa, Joris Sprakela, Romana Schirhaglb, and Renko de Vriesa

a Physical Chemistry and Soft Matter, Wageningen University & Research, Stippeneng4, 6708 WE Wageningen, The Netherlands.b Department of Biomedical Engineering, University Medical Center Groningen, Gronin-gen University, Antonius Deusinglaan 1, 9713 AW Groningen, The Netherlands.c Shenzhen Key Laboratory for Drug Addiction and Medication Safety, Departmentof Ultrasound, Peking University Shenzhen Hospital & Biomedical Research Institute,Shenzhen-PKU-HKUST Medical Center, 518036 Shenzhen, China.⊥These two authors contributed equally.

Analytical Chemistry 89:23 12805-12811 (2017)

IMPROVING THE CELLULAR UPTAKE

Abstract

Fluorescent nanodiamonds are gaining increasing attention as fluorescentlabels in biology in view of the fact that they are essentially non-toxic, donot bleach, and can be used as nanoscale sensors for various physical andchemical properties. To fully realize the nanosensing potential of nanodia-monds in biological applications, two problems need to be addressed: theirlimited colloidal stability, especially in presence of salts, and their limitedability to be taken up by cells. We show that the physical adsorption of asuitably designed recombinant polypeptide can address both the colloidalstability problem and the problem of the limited uptake of nanodiamondsby cells in a very straightforward way, while preserving both their spectro-scopic properties and their excellent biocompatibility.

6

INTRODUCTION

2.1 Introduction

In recent years fluorescent nanodiamonds (FNDs) have gained a lot of at-tention, in view of their very stable fluorescence (that allows for long termtracking), and in view of their unique magneto-optical behavior, whichallows them to be used as nanosensors that probe their immediate environ-ment [1]. Nanodiamond sensing can be done at ambient conditions, and canshow very high sensitivities. Promising sensing applications include usingnanodiamonds as nanoscale temperature sensors (with accuracies down to 1mK[2]), and using them as nanoscale magnetic resonance sensors (with sin-gle electron spin sensitivity[3]) sensing the chemical nature of compoundsin the immediate surroundings of the nanodiamonds. To fully realize thepotential of nanodiamonds for nanosensing applications in biology however,the particles need to be stable in biological fluids, and they need to be ableto enter cells.

Mammalian HeLa cells[4, 5, 6], as well as some other types of cells[7, 8,9], readily ingest diamond nanoparticles without any surface modification.By treating cells with chemicals such as NaN3, sucrose or filipin[10, 11],which inhibit certain uptake routes, the uptake mechanism has found tobe clathrin mediated endocytosis. However, these cases are more excep-tions than the rule. In addition, typical types of nanodiamonds, such asacid treated fluorescent nanodiamonds, quickly aggregate in common cellculture media[12]. The aggregation reduces uptake and is undesired fornanosensing applications, where the nanodiamonds are desired to be com-pletely dispersed.

In order to bypass the requirement for the nanodiamonds to be col-loidally stable in biological fluids, and the requirement for them to be takenup spontaneously by cells, nanodiamonds have been directly injected intohuman embryonic fibroblast WS1 cells, using a silicon nanowire[13]. Al-ternatively, Tzeng et al. used electroporation, in combination with BovineSerum Albumin (BSA) coating, to facilitate uptake by HeLa cells[14]. Otherapproaches to make cell walls more permeable have also been explored[15],but were found to be problematic since they decreased cell viability. Com-plex chemistries for attaching specific molecules to the surfaces of nanodi-amonds have also been done[16]. One of the most prominent approachesis to attach polyethylene glycol as this molecule is known for its ability torepel other proteins[17]. Another advantage of this approach is that meth-ods have been developed to further functionalize these so-called PEGylateddiamond.[18, 19, 20]

Generally, chemical methods are quite involved and not necessarily lead

7


to much improved control over colloidal stability and cellular uptake thansimple physical coating with polymers[21, 22], or serum proteins such asBSA[23]. A coating procedure for the nanodiamonds that combines thesimplicity of physical adsorption, but with much more control over thebiochemistry of the surface would therefore be very welcome.

Figure 2.1: Expected mode of bind-ing of protein polymers C4 and C4-K12

(in green) to the surface of (negativelycharged) nanodiamonds (in pink). a) Hy-drophilic C4 block weakly adsorbs to thenanodiamond surface via hydrogen bondsand electrostatically via a small numberof basic residues in the C4 block. b) TheC4-K12 binds more strongly and predom-inantly electrostatically via its dodecaly-sine binding block.

A major class of ligands thatcan be used to control cellular up-take are protein- or peptide- do-mains. While short peptides canbe produced synthetically and thenchemically attached, another ap-proach is to develop dedicated re-combinant polypeptides for coat-ing surfaces. Previously, wehave designed various recombi-nant polypeptides, or “protein-based polymers”[24] (or, for short,“protein polymers”) for coating in-dividual DNA molecules[25, 26].One of them is a diblock co-polypeptide denoted C4-K12 thatcoats individual DNA moleculeswith a hydrophilic polypeptidebrush layer[25]. The dodecalysine domain K12 provides electrostatic an-choring to the negatively charged DNA, while the C4 domain forms a hy-drophilic polymer brush around the DNA. The C4 domain is a tetramer ofa 98 amino acid long, hydrophilic random coil polypeptide C, which is richin glycines, prolines and other hydrophilic amino acids.[27]

As illustrated in Figure 2.1a, we expect that the hydrophilic C4 do-main by itself should bind weakly to the surface of the (negatively charged)nanodiamonds, via hydrogen bonds and electrostatically via the 12 lysineresidues spread regularly throughout the sequence of the domain. In anal-ogy with its interaction with negatively charged DNA[26], the C4-K12 di-block is expected to bind more strongly, and predominantly electrostat-ically, via its dodecalysine binding domain, forming a polymer brush asillustrated in Figure 2.1b.

As we will show, the C4-K12 polymer indeed rapidly and strongly ad-sorbs on the nanodiamonds providing excellent colloidal stability. It pro-motes the uptake of the nanodiamonds in a well dispersed state in cells,while leaving the optical properties and excellent biocompatibility of thenanodiamonds unaffected.

8

RESULTS AND DISCUSSION

2.2 Results and discussion

2.2.1 Characterization of coated nanodiamonds with AFM

Nanodiamonds are produced by grinding and some crystal planes breakmore likely than in others. This leads to a flake like structure[28]. Theyhave a relatively bright red photoluminescence signal from nitrogen-vacancy(NV) centers.

The nanodiamonds that we use (see Materials and Methods for details)have a hydrodynamic diameter DH ≈ 120 nm, and are negatively charged,with a zeta potential ≈ -50 mV.

Figure 2.2: AFM analysis of sizes ofC4-K12 coated nanodiamonds (0.85 mgmL−1 nanodiamonds dispersed in 5.7 mgmL−1 C4-K12 in MilliQ) adsorbed onfreshly cleaved mica and dried. a) rep-resentative AFM image b) zoom showingindividual nanodiamonds c) distributionof heights h and d) distribution of max-imum widths at half maximum heightsWFWHM derived by analysing all 73particles of the 1X1 µm area shown inb).

Using AFM we have analysedsize distributions of the nanodi-amonds by adsorbing them onfreshly cleaved mica and imagingthem in a dried state. We foundC4-K12 coated nanodiamonds ad-sorbed predominantly as singlydispersed particles, whereas forboth uncoated nanodiamonds, andnanodiamonds coated with C4,many aggregated particles were alsopresent. Nanodiamonds were coa-ted with the protein polymers bysimply mixing nanodiamond dis-persions with excess protein poly-mer (see Materials and Methods fordetails). The fact that the C4-K12

coated nanodiamonds are completedispersed in the AFM experiment,is already a first indication that thisprotein polymer provides the nan-odiamonds with a stabilizing coat-ing.

A representative AFM image of C4-K12 coated nanodiamonds adsorbedon freshly cleaved mica and dried is shown in Figure 2.2a. For the 73particles observed on a 1× 1 µm area (Figure 2.2b, which is a zoom of thearea enclosed by the white square in Figure 2.2a) we determined the heighth and maximum width w (at half maximum height). The resulting sizedistributions for height h and width w are shown in Figures 2.2c, d. Theaverage values are w = 54 ± 11 nm, and h = 15 ± 4 nm. This confirms

9


the flake-like geometry of the nanodiamonds although the anisotropy isnot very large. Additionally, AFM generally overestimates the widths ofparticles as the width values are always a convolution of the tip radius andthe actual particle size. In aqueous solutions, the hydrodynamic diameterof the highly hydrated C4-K12 protein polymer is[25] DH ≈ 10 nm, butwhen adsorbed and dried, its layer thickness should be much less, suchthat the sizes found for the coated particles should be close to those for theuncoated nanodiamonds.

2.2.2 Colloidal stability of protein-based polymer coated nan-odiamonds

As mentioned, the bare nanodiamonds have a very limited colloidal stabil-ity. This is illustrated in Figure 2.3a, which shows the time-dependence ofthe hydrodynamic size of bare nanodiamonds at a concentration C = 0.85mg mL−1. The nanodiamonds are initially dissolved in MilliQ water, inwhich they are colloidally stable due to their negative surface charge.

At t = 0 min, the salt concentration is increased to 0.05 M NaCl. Figure2.3a shows that within 10 minutes, the nanodiamonds aggregate to formlarge clusters. Figure 2.3b shows the hydrodynamic size of the bare nanodi-amonds at a concentration C = 0.85 mg mL−1 in MilliQ, 2 minutes after anincrease of the concentration of NaCl to the indicated values. Clearly, thenanodiamonds are stable against aggregation at low ionic strengths, but notat physiological ionic strengths. Figure 2.3b also shows the correspondingplots for the nanodiamonds coated with the protein polymers C4 (5.5 mgmL−1) and C4-K12 (5.7 mg mL−1). We find that both polymers stabilizethe nanodiamonds against aggregation, up to very high concentrations ofadded NaCl (at least 1M).

There is some increase in the particle size due to polymer adsorption.The magnitude of the increase is not really consistent with the additionallayer thickness due to the protein polymers, which is only expected to beon the order of the hydrodynamic diameter of the C4 and C4-K12 proteinpolymers, which is around 10 nm. Instead, the increase probably reflects avery limited degree of polymer-bridging induced nanodiamond aggregationthat occurs during the coating process itself.

As mentioned, we believe the driving forces for the adsorption of thehydrophilic protein polymers to the negatively charged and somewhat hy-drophobic nanodiamond surfaces is a combination of the formation of hy-drogen and ionic bonds.

At neutral pH, the C4 block is weakly negatively charged but it hasa small number of basic residues (uniformly distributed over its contour

10


length) that may still interact electrostatically with the negatively chargedsurfaces of the nanodiamonds. The fact that other forces such as hydrogenbonds presumably also play a role is evident from the observation that eventhe C4 domain alone stabilizes the nanodiamonds against aggregation upto very high ionic strengths at which these weak electrostatic interactionsmust be completely screened.

Since the C4 domain is long, (400 amino acids), only a weak affinityof a fraction of the amino acids to the nanodiamond surface suffices tocause adsorption of the polymer as whole. This is the classic limit of ho-mopolymer adsorption, for which case one expects[29] a dense adsorbedinner layer dominated by loops and a dilute outer layer dominated by tails,as illustrated in Figure 2.1a.

For the C4-K12 diblock, there will be a very strong preference of theK12 anchoring blocks to electrostatically interact with negatively chargedsurface of the nanodiamonds. We expect this will result in a dense polymerbrush-like layer, as illustrated in Figure 2.1b. Hence, the diblock is expectedto bind at higher densities, and is expected to give rise to a thicker polymerlayer, and a stronger stabilizing effect than just the C4 domain.

Figure 2.3: Hydrodynamic diameter (DH) of nanodiamond particles both in thepresence and absence of protein polymers C4 and C4-K12 as determined using dy-namic light scattering.Nanodiamond concentration is 0.85 mg mL−1, while proteinpolymer concentrations are 5.5 mg mL−1 (C4) and 5.7 mg mL−1 (C4-K12). a) Hy-drodynamic diameter for bare nanodiamonds in MilliQ as a function of the timeafter an increase of the salt concentration to 0.05 M NaCl. b) Hydrodynamic di-ameter of nanodiamonds dispersed in MilliQ, 2 min. after the NaCl concentrationis increased to the value indicated on the x-axis. Black symbols: bare nanodia-monds. Red symbols: C4–coated nanodiamonds. Blue symbols: C4-K12 -coatednanodiamonds.

11


2.2.3 Binding strength of protein polymers to nanodiamonds

We have used isothermal titration calorimetry (ITC) to further quantify thebinding of the C4-K12 diblock polymer to the nanodiamonds. For convert-ing the nanodiamond weight concentration CND in mg mL−1 to a molarconcentration cND, we assume an average particle volume of V ≈ 34000nm3, as estimated from AFM. Using a density of = 3.52 g cm−3, this trans-lates into a molar mass of MND ≈ 7.2× 104 kg mol−1 that is used for theconversion to a molar concentration of nanodiamonds. The ITC raw data(heat flow as a function of time during the injections) for the titration of aCND = 1 mg mL−1, or cND = 0.1 µM suspension of nanodiamonds titratedwith a 1 mM solution of C4-K12 is shown in Figure 2.4a. The interactionis purely exothermic and the heat released per mole of injectant, versusthe molar ratio of protein to nanodiamonds can be very well fitted witha simple one-site binding model (Figure 2.4b). The analysis results in anequilibrium dissociation constant Kd = 1.5 × 105 M−1 and an enthalpychange 4H = -22 kcal mol−1 of adsorbed C4-K12. The estimated numberof adsorbed C4-K12 molecules is around 4 C4-K12 molecules per 100 nm2

of nanodiamond surface. This is quite reasonable, given the hydrodynamicradius of the isolated C4 domains of DH ≈ 10 nm. There appear to besmall deviations between the experimental curve and the fit to the one-sitebinding model at high concentrations. However, at these high concentra-tions the surface is nearly saturated and the ITC signal is weak, such thatwe believe these small deviations do not warrant an extension of the fit tomore complicated models.

Figure 2.4: Characterization of bind-ing of C4-K12 protein polymer to nanodi-amonds using ITC. Nanodiamonds sus-pended in MilliQ water at a concentra-tion of CND = 1 mg mL−1, were titrated(at 30◦C) with a 1 mM C4-K12 in MQwater. a) Heat flow versus time duringthe injections. b) Heat released per moleof added C4-K12 versus the molar proteinto nanodiamond ratio.

A similar ITC experiment wasperformed for the C4 polymer, butin this case, we found a non-monotonic dependence of the heatreleased per mole of injectant, ver-sus the molar ratio of protein tonanodiamonds (Supplementary in-formation Figure 2.9), suggestingthat in this case at least twoprocesses contribute to the heatevolved during the binding reac-tion. Possibly adsorption of the C4

polymer chains involves the break-ing of some weak intramolecularbonds. This results in an endother-mic contribution to the adsorption

12


enthalpy at low protein to nanoparticle ratios. Since the focus here ismainly on the C4-K12 diblock that is expected to be the better stabilizer,we refrain from a further quantitative interpretation of the ITC data forthe C4 polymer.

Additionally, the stability of the coatings to changes in pH was assessedby measuring the hydrodynamic diameter of the particles suspended inalkaline and acidic media. The results proved that the protein polymersremains attached to the nanodiamonds surface when they are exposed to arange of pH from 4.5 to 8.9. Detail of this experiment and its results areshown in the supplementary information.

2.2.4 Photoluminescence properties of protein polymer coa-ted nanodiamonds

The nitrogen vacancy (NV) center in nanodiamonds, is a point defectformed in the diamond lattice by one substitutional nitrogen atom andan adjacent vacancy[30, 31]. One of its most explored properties is pho-toluminescence, which can be detected from an individual NV center[32].

Figure 2.5: Normalized photolumines-cence spectrum of bare nanodiamonds(blue), and spectra of C4 coated (red)and C4-K12 coated (green) nanodia-monds at 250 µg mL−1 in MilliQ water,C4 protein at 16 mg mL−1 and C4-K12 at17 mg mL−1 in MilliQ water. The decre-ment in the PL intensity registered inthe coated diamonds, relative to the non-coated case, doesn’t suppose an obstaclefor sensing purposes and can be compen-sated by, either, increasing the excitationpower or by increasing the signal integra-tion time or by sensing a wider segmentof wavelengths.

There are two charge states of thisdefect, one is neutral NV0 (excita-tion (Ex) wavelength at 490 nm,emission (Em) wavelength of thezero phonon line at 575 nm), theother is negative NV- (Ex at 490nm, Em of the zero phonon line at637 nm). An important propertyof luminescence from individual NVcenter is that it has very highphotostability, with essentially nobleaching being observed at roomtemperature[33, 34].

As shown in Figure 2.5 , theprotein polymer coating only leadsto minor changes in the photolumi-nescence of the nanodiamonds. Thepeak of the spectra remains closeto the 692 nm, and the bandwidths(FWHM) also do not vary signifi-cantly, being 122 nm for the barenanodiamonds, 118 nm for C4-K12

coated- and 114 nm for C4 coated-

13


nanodiamonds. The main effect is a slight decrease of the fluorescence inten-sity of the protein polymer coated nanodiamonds as compared to the barenanodiamonds, which is 19% for C4-K12 coated- and 14% for C4 coated-nanodiamonds.

Performing optically detected magnetic resonance measurements withsimilar equipment as previously described[35, 36] (see supplementary ma-terial) reveals that also the magneto-optical properties of the NV centerare not perturbed. Therefore, the protein polymer coating does not atall preclude the use of the nanodiamonds for sensing applications. Thesesensing applications include measurements of temperature[2], or particleorientation[6], which can be obtained from such magnetic resonance mea-surements [3, 37]. There are also a number of interesting quantum sensingprotocols, which measure small magnetic or electric fields or their fluctu-ations. These require close proximity between the defect and the cellularenvironment. The protein polymer coating increases the distance betweenthe defects and the analyte molecules. This is critical for sensitivity andthus this parameter will definitely increase the detection limit. As all kindsof molecules in a cellular environment attach to bare nanodiamond surfacesthis distance is, however, in any case difficult to avoid.

2.2.5 Cytotoxicity of protein polymer coated nanodiamonds

It has been shown by several researches that bare nanodiamonds are gen-erally non-cytotoxic[7, 38, 39, 40]. To make sure that the protein poly-mer coating does not affect this positive quality of the nanodiamonds,we have tested whether the protein polymer coating leads to any changesin the cytotoxicity of the nanodiamonds. To this end a series of MTT(3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assays wereconducted. The study consisted of a 3 hours incubation of HeLa and HT29GFP-EpCAM (HT29*) cells with protein polymer coated and bare nan-odiamonds at different concentrations (1 µg mL−1, 5 µg mL−1 and 10 µgmL−1), followed by the use of MTT to assay cell viability. As a refer-ence, we use cells incubated with growth medium only. As is shown inFigure 2.6a, for HeLa cells, and Figure 2.6b, for HT29* cells, to within themargin of error of the experiments, the viability of both types of cells isnot affected by incubation with nanodiamonds, either bare or coated withthe protein polymers, confirming that like the bare diamonds, the proteinpolymer coated nanodiamonds are completely harmless for the cells.

14


Figure 2.6: Relative cell viability (com-pared to cell incubated without diamonds)as deduced from MTT assay, versus con-centration CND of nanodiamonds in thegrowth medium. In the first group of bars,the growth medium contained 120 nm nan-odiamonds at 1 µg mL−1 and C4 polymerprotein at 63 µg mL−1 or C4-K12 polymerprotein at 66.24 µg mL−1. The growthmedium used in the second groups of barscontained 120 nm nanodiamonds at 5 µgmL−1 and C4 polymer protein at 315 µgmL−1 or C4-K12 polymer protein at 331.2µg mL−1. In the last group, the growthmedium contained 120 nm nanodiamondsat 10 µg mL−1 and C4 polymer protein at630 µg mL−1 or C4-K12 polymer protein at662.4 µg mL−1. Cells were incubated by 3hour in these mediums. White bars: barenanodiamonds. Red bars: C4 coated nan-odiamonds. Blue bars: C4-K12 coated nan-odiamonds. Error bars show the standarddeviation. a) HeLa cells, b) HT29* cells.

2.2.6 Cell uptake of bare and protein polymer coated nan-odiamonds

Finally, we qualitatively evaluated the uptake of the bare and protein poly-mer coated nanodiamonds by cells using confocal scanning light microscopy(CSLM). We first imaged bare and protein polymer coated nanodiamondsadsorbed to the glass microscope slides, from a 10 µg mL−1 solution of nan-odiamonds in growth medium (DMEM complete). The images (Figure 2.10in the supplementary material) clearly shows the strong effect the proteinpolymers have on the state of aggregation of the nanodiamonds. For barenanodiamonds the mean area of particles that had visibly adhered was 1.7µm2, for C4 coated nanodiamonds it was 0.9 µm2 and finally, for C4-K12

coated nanodiamonds it was 0.3 µm2. This shows that especially C4-K12 isvery effective at preventing nanodiamond aggregation, even though it doesnot completely prevent it. It should be noted that these results do notimply that most nanodiamonds are in an aggregated state since the highbrightness of the nanodiamonds clusters hinders the visualization of singlenanodiamond in the confocal images.

Next, both HeLa and HT29* cells were incubated for two hours with

15


either bare or protein polymer coated nanodiamonds at a nanodiamondconcentration CND = 5 µg mL−1. For HeLa cells, already without a pro-tein polymer coating, most cells have taken up at least some nanodiamonds(Figure 2.7a). However, the number of fluorescent objects that could bedetected using the CSLM was higher for the nanodiamonds coated withprotein polymers, than for the bare nanodiamonds, with a higher num-ber of fluorescent objects being detected for C4-K12 coated nanodiamonds(Figure 2.7c), as compared to the case of C4 coated nanodiamonds (Figure2.7b). Also, the size of the fluorescent objects detected inside the HeLacells was significantly smaller for the coated than for the bare nanodia-monds, especially for the C4-K12 coated nanodiamonds. For HT29* cellsthe situation was quite different. As shown Figure 2.8a, using CSLM, fewfluorescent objects could be detected inside the HT29* cells for the barenanodiamonds; fluorescent object were detected in only 30% of the cells.In contrast, for protein polymer coated nanodiamonds, fluorescent objectswere detected inside 90% of the cells for C4 coated nanodiamonds and in-side 100% of the cells for C4-K12 coated nanodiamonds (Figures 2.8b andc).

16


Figure 2.7: Confocal images of HeLacells incubated for 2 hours with nan-odiamonds. Cells were incubated witha 5 µg mL−1 dispersion of (a) barenanodiamonds, (b) nanodiamonds coa-ted with C4 protein polymer (315 µgmL−1) and (c) nanodiamonds coatedwith C4-K12 protein polymer (331.2 µgmL−1).

Figure 2.8: Confocal images ofHT29* incubated for 2 hours with nan-odiamonds. Cells were incubated witha 5 µg mL−1 dispersion of (a) barenanodiamonds, (b) nanodiamonds coa-ted with C4 protein polymer (315 µgmL−1) and (c) nanodiamonds coatedwith C4-K12 protein polymer (331.2 µgmL−1).

17


2.3 Experimental Details

2.3.1 Materials

Fluorescence nanodiamonds

ND-NV-120nm with a reported diameter of 120 nanometers were purchasedfrom Adamas Nanothechnologies. The stock suspension contains nanodia-monds at a concentration of 1 mg mL−1 in water. All other reagents andsolvents were obtained from Sigma-Aldrich.

Protein polymers

The C4 (molar mass 37,285 g mol−1) and C4-K12 (molar mass 38,408 gmol−1) protein polymers were obtained via secreted expression by genet-ically engineered Pichia pastoris strains, in a methanol fed-batch fermen-tation, as described previously[25, 27]. After fermentation, the protein-containing supernatant was separated from the P. pastoris cells by centrifu-gation (30 min at 16000 g) and subsequent microfiltration. Next, proteinpolymers were further purified as described previously[25, 27]. In short, C4

and C4-K12 proteins were selectively precipitated using ammonium sulfate(at 45% saturation). For the C4 protein this was followed by two acetoneprecipitation steps (at, respectively, 40% (v/v) and 80% (v/v) saturation).The acetone precipitation was omitted for the C4-K12 protein polymer.The C4 protein was desalted using extensive dialysis against MilliQ water,whereas C4-K12 was dialyzed against 50 mM formic acid. Finally, the bothproteins were lyophilized and the protein polymer powders were stored atroom temperature.

Protein polymer stocks

For the C4 protein polymer, stock solutions were prepared by dissolving3.5 mg of C4 powder in 500 µl of MilliQ water. For the C4-K12 proteinpolymer, 3.7 mg was dissolved in 500 µl of MilliQ water. After mixing, thesolutions were filtrated with a 220 nm centrifugal filter for 5 minutes at13.2 krpm. All the concentrations reported In this article were measuredbefore filtration.

18

EXPERIMENTAL DETAILS

2.3.2 Methods

Isothermal titration calorimetry

Isothermal titration calorimetry experiments were performed using a VP-ITC Micro-Calorimeter from Malvern Instrument Ltd. U.K. 1.43 mL of de-gassed nanodiamond dispersion (0.1 mg mL−1 in MilliQ water) was loadedin the cell as analyte, while 250 µL degassed solution of protein polymer(1 mM in MilliQ water, corresponding to 37.3 mg mL−1 for C4 and 38.4mg mL−1 for C4-K12) was loaded in the auto-pipette as titrant. One initialinjection of 2 µL, for saturating the titration cell walls was followed by 20injections of 10 µL each. The time between subsequent injections was 250s. All measurements were conducted at 30◦C. The reference cell was filledwith degassed MilliQ water. During the experiment, the sample cell wasalways stirred continuously at 329 rpm.

The heat of protein polymer dilution in MilliQ water was subtractedfrom the titration data for each experiment. For analysing the data weused the Origin software that came with the VP-ITC Micro-Calorimeter.The reported parameters deduced by fitting the experimental data are anaverage of a duplicate experiment.

Coating of nanodiamonds with the polymer proteins

The nanoparticles were coated with the C4 and C4-K12 protein polymersby mixing the stock solution of nanodiamonds (1 mg mL−1 in water), withthe protein polymer stocks and let them incubate for 30 minutes. Thesubsequent dilutions were prepared by diluting these solutions in water orgrowth medium, depending on the type of experiment.

Photoluminescence spectroscopy

Fluorescence measurements were performed using a ThermoFisher Var-ioskan instrument. Emission spectra were measured from 550 nm to 840 nmin 1 nm steps, with a measurement time of 300 ms per step. The excitationwavelength was set at 533 nm and the temperature at 30◦C. To decreasethe effect of noise, the raw data was processed by a median filter. Spectrawere acquired for dispersions of 250 µg mL−1 nanodiamonds in MilliQ wa-ter. For the C4 -coated nanodiamonds the protein polymer concentrationwas 16 µg mL−1, for the C4-K12 coated nanodiamonds the protein polymerconcentration was 17 µg mL−1.

19


Atomic-force microscopy

For Atomic Force Microscopy (AFM) imaging, we used a NanoScope V Mul-timode Scanning Probe Microscope, with an ultra-sharp silicon cantilever(NT-MDT CSCS11) in the scan-assist imaging mode. In measurement ofC4-K12 coated nanodiamonds, dispersions of 0.85 mg mL−1 of nanodia-monds in MiliQ were used, and C4-K12 concentration was 5.7 mg mL−1.For the C4 -coated nanodiamonds the protein polymer concentration was0.65 mg mL−1 while nanodiamond concentration was 0.1 mg mL−1, forbare nanodiamond measurement,the nanodiamond concentration was 0.1mg mL−1. A volume of 10 µL of bare or coated nanodiamond dispersionwas deposited on a freshly cleaved mica surface and incubated for 1 min.to allow for nanodiamond adsorption. Next, 2 mL of filtered MilliQ waterused to wash the mica sample surface. Finally, the sample surface was driedusing nitrogen gas. AFM images were analyzed using NanoScope Analysissoftware.

Dynamic light scattering

Hydrodynamic diameters of bare nanodiamonds and of nanodiamonds coa-ted with the protein polymers were measured at a temperature of 30◦Cusingdynamic light scattering. For the scattering experiments, a Malvern Zeta-sizer Nano ZS ZEN3500 was used, equipped with a cell with peltier temper-ature control. The laser wavelength was 633 nm and the scattering anglewas 173◦. For each sample, 2 minutes of equilibration was followed by 10measurements, where each measurements consisted of 14 scattering runs of10 s. Dispersions of 0.85 mg mL−1 of nanodiamonds in MiliQ were used.For the C4 -coated nanodiamonds the protein polymer concentration was5.5 mg mL−1, for the C4-K12 coated nanodiamonds the protein polymerconcentration was 5.7mg mL−1. The reported hydrodynamic diameter isthe dominant peak from a distribution analysis performed on the raw databy the Zetasizer Nano software v7.11.

Cell cultures

Nanodiamond cell uptake experiments were performed for HeLa and HT29*cell lines. Cells were seeded at low density and incubated for 20 hours inglass bottom petri dishes until they achieved 25% of confluency. DMEM-HG Complete was used as growth medium during this time. To assessthe particle internalization, cells were incubated in nanodiamond enrichedgrowth medium (DMEM-HG) for a period of 120 minutes. Every Petry dishwas filled with 1 ml of the 5 µg mL−1 nanodiamond enriched medium, which

20

CONCLUDING REMARKS

was prepared by dispersion of 50 µl of protein coated nanodiamonds (100µg mL−1 nanodiamonds, for the C4 -coated nanodiamonds additionally315 mg mL−1 and for the C4-K12 coated nanodiamonds additionally 331.2mg mL−1 of protein polymer) in 950 µl of DMEM-HG. Bare nanodiamondswere used as control. To prepare 1 ml of a suspension of bare nanodiamondsat 5 µg mL−1, 5 µl of the nanodiamonds stock slurry (1 mg mL−1) wasdispersed in 995 µl of DMEM-HG. After the two hours incubation, the cellswere washed and fixated with 3.7% paraformaldehyde. Additionally, HeLacells were stained with phalloidin FITC. The HT29* cells have fluorescentCAM lipoprotein, so no additional staining was performed to visualize thecells in this case. The samples were conserved in 1% PFA in PBS at 4◦C.

Confocal imaging

The samples, described in the previous paragraph, were imaged in differentregions using a confocal microscope (Zeiss 780). The fluorochromes wereexcited with lasers at wavelengths of 488 nm (FITC and GFP) and 561 nm(NV-centers). The NV center’s fluorescence was collected at wavelengthfrom 597 nm to 694 nm.

Cell uptake experiments

The detection of nanodiamonds in the cells was made by visual inspection.The background light in the images was removed by filtering the pixels withvalue less than 80. Next, the images were analysed using the software ZENBlack 2.1 (Carl Zeiss).

2.4 Concluding remarks

Coating with de novo designed recombinant protein polymers offers a con-venient way to disperse nanoparticles such as nanodiamonds, and to providecolloidal stability to such particles in biological fluids such as serum, or cellculture medium. It also offers a highly engineerable approach to modulatethe interactions of such nanoparticles with living cells. Via peptide syn-thesis it is possible to change the peptide sequence as we demonstrated byusing C4 and C4-K12. Once functional groups are inserted via changing thepeptide sequence it is also possible to further react these groups (beforeor after adhesion to the diamond surface). However, there are likely alsogroups, which might compromise the binding. For instance a large numberof negatively charged proteins is expected to repel the diamond surface. Forthe diblock polypeptide C4-K12 we observe that simple physical adsorption

21


to nanodiamonds leads to an excellent stability of the nanodiamonds up tovery high salt concentrations, and to a much greater degree of dispersion ofthe nanoparticles, not only in buffers and biological fluids, but also whentaken up by living cells such as HeLa cells. For the specific case of HT29*cells, we also find that the protein polymer coating improves the uptake ofthe nanodiamonds.

The protein polymer coating does not affect the photoluminescence northe magneto-optical properties of the nanodiamonds, nor their excellentbiocompatibility. Hence the protein polymer coated nanodiamonds are apromising candidate to be used as intracellular magneto-sensitive nanosen-sors.

We have used a protein polymer that was designed to coat DNA molecules[27]. As such, it does not yet contain any specific motifs that one mightwant to include for a dedicated protein polymer for stabilizing nanodia-mond sensors and for promoting their uptake by various cell lines. Addi-tional motifs that could be included are, for example, cell-binding motifs,cell-penetrating peptide motifs, as well as intracellular targeting motifs.

22

SUPPORTING INFORMATION

2.5 Supporting Information

2.5.1 Binding of the C4 protein polymer

Figure 2.9: Characterization of binding of C4 protein polymer to nanodiamondsusing ITC. Nanodiamonds suspended in MilliQ water at a concentration of CND= 1 mg mL−1, were titrated (at 30◦C) with a 1 mM C4 in MQ water. a) Heat flowversus time during the injections. b) Heat released per mole of added C4 versusthe molar protein to nanodiamond ratio.

2.5.2 Clusters of nanodiamonds

Figure 2.10: During the uptake experiments clusters of nanodiamonds depositedover the coverslip. Those clusters were imaged with a confocal microscope. (a)The cluster formed by bare nanodiamonds have bigger size (average area: 1.7µm2) in comparison with both coated cases, (b) C4 (average area: 0.9 µm2) and(c) C4-K12 (average area: 0.3 µm2) conjugated nanodiamonds. In all the cases,bare and conjugated nanodiamods were mixed with growth medium (DMEM-HGcomplete), at concentration of 10 µg mL−1 of nanodiamonds, 630 µg mL−1 of C4

and 662.4 µg mL−1 of C4-K12 polymer protein, and let it incubate for two hourswith HT29* cells.

23


2.5.3 Magneto-optical properties

Method: For magnetic resonance measurements, 10 µL of dispersions ofcoated and uncoated nanodiamonds were applied on clean microscope slidesand dried. The dispersions were prepared by suspending 10 µL of the 120nm nanodiamond stock solution in 90 µl of water, C4 or C4-K12 proteinpolymers at concentrations of, respectively, 7 -1 -1mg mL and 7.36 mg mL .A home built diamond magnetometer was used, similar to those previouslyused by others[35, 36]. The magnetometer is essentially a confocal micro-scope with built in microwave electronics. A laser power of 1 mW was usedfor illumination at a wavelength of 532 nm.

Figure 2.11: The magneto-opticalproperties of the nanodiamonds (nD) re-main almost unalterable after being coa-ted with the C4 and C4-K12 proteinpolymers, as it is shown by ESR mea-surements of bare nanodiamonds (blue -1line), C4 (red line) and C4-K12 (greenline) coated nanodiamonds. Nanodia-monds at 100 µg mL in MilliQ water,C4 protein at 6.3 mg mL−1 and C4-K12 at 6.624 mg mL−1 in MilliQ water.The fluorescence intensity (y axis) of theNV center drops when it is excited withan external electromagnetic field at fre-quency near to 2.87 [GHz] (x-axis).

After scanning the sample withadsorbed diamond nanoparticles,we focused on individual diamondparticles and recorded an opti-cally detected magnetic resonance.A frequency sweep was performedfor the microwave, for frequenciesaround the expected resonance fre-quency of the NV center at 2,87GHz. This microwave signal wasproduced with a microwave syn-thesizer (Hittite HMC-T2100) thatsent its signal (power of 27 dBm) toa homemade antenna (a short cir-cuit of a copper wire at the end of acoaxial cable) wish was located fewmicrometer from the sample. Si-multaneously with the electromag-netic irradiation, the intensity ofthe fluorescence was collected usingan Olympus UPLSAP40x NA=0.95objective and an Avalanche pho-todiode (SPCM-AQRF- 15-FC) insingle photon counting mode.

Results: We characterize the NV center’s magneto-optical propertiesusing Electron Spin Resonance ESR for the bare and the protein-polymercoated nanodiamonds. Results for the ESR experiment are shown in Fig-ure 2.11. As expected, for bare nanodiamonds, the NV centers in Figure2.11 show a decrease of their fluorescence intensity when exposed to anexternal electromagnetic field at frequency near 2.87 GHz. As for the pho-

24


toluminescence, we find that the protein polymer coating hardly affects themagneto-optical properties of the nanodiamonds. The magnitude of theESR signal for bare nanodiamonds and protein polymer coated nanodia-monds is almost the same.

2.5.4 Protein polymer binding stability at different pH

Considering the fact that the conjugation of the protein polymers and thesurface of the nanodiamonds is made merely by physical absorption, wewere interested in to evaluate the robustness of these bonds when the coatedparticles are exposed to an alkaline or acidic environment. The experimentsconsisted in measuring the hydrodynamic diameter of the coated particlesafter dispersing them in media at pH 4.5, 5.5, 6.8, 7.9 and 8.9. It wasassumed that the desorption of protein polymers from the nanodiamondssurface would be reflected in a reduction of the hydrodynamic diameter(HD) of the particles when they are measured by DLS. Method: Media atpH 4.5 and 5.5 was prepared by diluting hydrogen chloride (HCl) in MilliQwater (pH 5.31) until the desired pH values were reached. The media atpH 6.8, 7.9 and 8.9 were prepared similarly but by adding sodium hydrox-ide (NaOH) instead. The samples were prepared by dispersing 1 µL of C4

coated- or C4-K12 coated nanodiamonds (100 µg L−1), as appropriate, in999 µL of the media previously made. The hydrodynamic diameter wasmeasured and analysed following the same procedure explained previouslyin the method section of the main article. Results: The measures of hydro-dynamic diameter showed low variability between different pH conditions.Especially, the C4 coated nanodiamonds reported more consistent resultsacross different samples. On the other hand, the average of the HD of theC4-K12 coated particles shows a small increase as the pH turns more basic.Neither of these situations suggests the occurrence of desorption of the pro-tein polymers from the nanodiamonds surface. On the contrary, the slightincrement in size could be an indication of the increasing in the thicknessof the ions layer surrounding the particles, or very slight aggregation.

25


Figure 2.12: At pH 6.8, the average HD is 150.3 and 180.9 of the C4- (blue) andC4-K12- (red) coated nanodiamonds (nD) respectively. The comparison of thisvalue with the one from samples in alkaline and acidic medium, and consideringthe wide distribution of the results, doesn’t suggest a considerable reduction of theparticle’s size that could be attributable to the desorption of the protein polymers.Instead, the small changes of size can be explained by a change in the thicknessof the electric dipole layer that surrounds the nanoparticles.

26

REFERENCES

References

[1] R. Schirhagl, K. Chang, M. Loretz, and C. L. Degen, “Nitrogen-Vacancy Centers in Diamond: Nanoscale Sensors for Physics and Bi-ology,” Annu. Rev. Phys. Chem., vol. 65, no. 1, pp. 83–105, 2014.

[2] P. Neumann, I. Jakobi, F. Dolde, C. Burk, R. Reuter, G. Waldherr,J. Honert, T. Wolf, A. Brunner, J. H. Shim, D. Suter, H. Sumiya,J. Isoya, and J. Wrachtrup, “High-precision nanoscale temperaturesensing using single defects in diamond,” Nano Lett., vol. 13, no. 6,pp. 2738–2742, 2013.

[3] M. S. Grinolds, S. Hong, P. Maletinsky, L. Luan, M. Lukin, R. L.Walsworth, and A. Yacoby, “Nanoscale magnetic imaging of a singleelectron spin under ambient conditions,” Nat. Phys., vol. 9, no. 4,pp. 215–219, 2013.

[4] O. Faklaris, D. Garrot, W. Joshi, F. Druon, J. P. Boudou, T. Sauvage,P. Georges, P. A. Curmi, and F. Treussart, “Detection of single pho-toluminescent diamond nanoparticles in cells and study of the inter-nalization pathway,” Small, vol. 4, pp. 2236–2239, dec 2008.

[5] I. Pope, L. Payne, G. Zoriniants, E. Thomas, O. Williams, P. Wat-son, W. Langbein, and P. Borri, “Coherent anti-Stokes Raman scat-tering microscopy of single nanodiamonds,” Nat. Nanotechnol., vol. 9,pp. 940–946, nov 2014.

[6] L. P. McGuinness, Y. Yan, A. Stacey, D. A. Simpson, L. T. Hall,D. Maclaurin, S. Prawer, P. Mulvaney, J. Wrachtrup, F. Caruso, R. E.Scholten, and L. C. L. Hollenberg, “Quantum measurement and orien-tation tracking of fluorescent nanodiamonds inside living cells,” Nat.Nanotechnol., vol. 6, no. 6, pp. 358–363, 2011.

[7] S. J. Yu, M. W. Kang, H. C. Chang, K. M. Chen, and Y. C. Yu, “Brightfluorescent nanodiamonds: No photobleaching and low cytotoxicity,”J. Am. Chem. Soc., vol. 127, pp. 17604–17605, dec 2005.

[8] C. Y. Fang, V. Vaijayanthimala, C. A. Cheng, S. H. Yeh, C. F. Chang,C. L. Li, and H. C. Chang, “The exocytosis of fluorescent nanodiamondand its use as a long-term cell tracker,” Small, vol. 7, pp. 3363–3370,dec 2011.

[9] Z. Chu, S. Zhang, B. Zhang, C. Zhang, C. Y. Fang, I. Rehor, P. Cigler,H. C. Chang, G. Lin, R. Liu, and Q. Li, “Unambiguous observation of

27


shape effects on cellular fate of nanoparticles,” Sci. Rep., vol. 4, mar2014.

[10] O. Faklaris, V. Joshi, T. Irinopoulou, P. Tauc, M. Sennour, H. Girard,C. Gesset, J. C. Arnault, A. Thorel, J. P. Boudou, P. A. Curmi, andF. Treussart, “Photoluminescent diamond nanoparticles for cell label-ing: Study of the uptake mechanism in mammalian cells,” ACS Nano,vol. 3, pp. 3955–3962, dec 2009.

[11] E. Perevedentseva, S. F. Hong, K. J. Huang, I. T. Chiang, C. Y. Lee,Y. T. Tseng, and C. L. Cheng, “Nanodiamond internalization in cellsand the cell uptake mechanism,” J. Nanoparticle Res., vol. 15, no. 8,2013.

[12] S. R. Hemelaar, A. Nagl, F. Bigot, M. M. Rodrıguez-Garcıa, M. P.de Vries, M. Chipaux, and R. Schirhagl, “The interaction of fluorescentnanodiamond probes with cellular media,” Microchim. Acta, vol. 184,no. 4, pp. 1001–1009, 2017.

[13] G. Kucsko, P. C. Maurer, N. Y. Yao, M. Kubo, H. J. Noh, P. K. Lo,H. Park, and M. D. Lukin, “Nanometre-scale thermometry in a livingcell,” Nature, vol. 500, no. 7460, pp. 54–58, 2013.

[14] Y. K. Tzeng, O. Faklaris, B. M. Chang, Y. Kuo, J. H. Hsu, andH. C. Chang, “Superresolution imaging of albumin-conjugated fluores-cent nanodiamonds in cells by stimulated emission depletion,” Angew.Chemie - Int. Ed., vol. 50, pp. 2262–2265, mar 2011.

[15] S. R. Hemelaar, K. J. Van Der Laan, S. R. Hinterding, M. V. Koot,E. Ellermann, F. Perona Martinez, D. Roig, S. Hommelet, D. Nova-rina, H. Takahashi, M. Chang, and R. Schirhagl, “Generally ApplicableTransformation Protocols for Fluorescent Nanodiamond Internaliza-tion into Cells,” Sci. Rep., vol. 7, dec 2017.

[16] A. Nagl, S. R. Hemelaar, and R. Schirhagl, “Improving surfaceand defect center chemistry of fluorescent nanodiamonds for imagingpurposes-a review,” Anal. Bioanal. Chem., vol. 407, no. 25, pp. 7521–7536, 2015.

[17] T. Takimoto, T. Chano, S. Shimizu, H. Okabe, M. Ito, M. Morita,T. Kimura, T. Inubushi, and N. Komatsu, “Preparation of fluorescentdiamond nanoparticles stably dispersed under a physiological envi-ronment through multistep organic transformations,” Chem. Mater.,vol. 22, pp. 3462–3471, jun 2010.

28

REFERENCES

[18] L. Zhao, Y. H. Xu, T. Akasaka, S. Abe, N. Komatsu, F. Watari,and X. Chen, “Polyglycerol-coated nanodiamond as a macrophage-evading platform for selective drug delivery in cancer cells,” Biomate-rials, vol. 35, no. 20, pp. 5393–5406, 2014.

[19] L. Zhao, A. Shiino, H. Qin, T. Kimura, and N. Komatsu, “Synthesis,Characterization, and Magnetic Resonance Evaluation of Polyglycerol-Functionalized Detonation Nanodiamond Conjugated with Gadolin-ium(III) Complex,” J. Nanosci. Nanotechnol., vol. 15, pp. 1076–1082,sep 2014.

[20] L. Zhao, Y. H. Xu, H. Qin, S. Abe, T. Akasaka, T. Chano, F. Watari,T. Kimura, N. Komatsu, and X. Chen, “Platinum on nanodiamond: Apromising prodrug conjugated with stealth polyglycerol, targeting pep-tide and acid-responsive antitumor drug,” Adv. Funct. Mater., vol. 24,pp. 5348–5357, sep 2014.

[21] J. Slegerova, M. Hajek, I. Rehor, F. Sedlak, J. Stursa, M. Hruby,and P. Cigler, “Designing the nanobiointerface of fluorescent nanodia-monds: Highly selective targeting of glioma cancer cells,” Nanoscale,vol. 7, pp. 415–420, jan 2015.

[22] I. Rehor, H. Mackova, S. K. Filippov, J. Kucka, V. Proks, J. Slegerova,S. Turner, G. Van Tendeloo, M. Ledvina, M. Hruby, and P. Cigler,“Fluorescent nanodiamonds with bioorthogonally reactive protein-resistant polymeric coatings,” Chempluschem, vol. 79, pp. 21–24, jan2014.

[23] J. W. Lee, S. Lee, S. Jang, K. Y. Han, Y. Kim, J. Hyun, S. K.Kim, and Y. Lee, “Preparation of non-aggregated fluorescent nan-odiamonds (FNDs) by non-covalent coating with a block copolymerand proteins for enhancement of intracellular uptake,” Mol. Biosyst.,vol. 9, pp. 1004–1011, may 2013.

[24] O. S. Rabotyagova, P. Cebe, and D. L. Kaplan, “Protein-based blockcopolymers,” Biomacromolecules, vol. 12, pp. 269–289, feb 2011.

[25] A. Hernandez-Garcia, M. W. Werten, M. C. Stuart, F. A. De Wolf,and R. De Vries, “Coating of single DNA molecules by geneticallyengineered protein diblock copolymers,” Small, vol. 8, pp. 3491–3501,nov 2012.

[26] A. Hernandez-Garcia, D. J. Kraft, A. F. Janssen, P. H. Bomans, N. A.Sommerdijk, D. M. Thies-Weesie, M. E. Favretto, R. Brock, F. A. De

29


Wolf, M. W. Werten, P. Van Der Schoot, M. C. Stuart, and R. DeVries, “Design and self-assembly of simple coat proteins for artificialviruses,” Nat. Nanotechnol., vol. 9, no. 9, pp. 698–702, 2014.

[27] M. W. T. Werten, W. H. Wisselink, T. J. Jansen-van den Bosch,E. C. de Bruin, and F. A. de Wolf, “Secreted production of a custom-designed, highly hydrophilic gelatin in Pichia pastoris,” Protein Eng.Des. Sel., vol. 14, pp. 447–454, jul 2002.

[28] S. Y. Ong, M. Chipaux, A. Nagl, and R. Schirhagl, “Shape and crys-tallographic orientation of nanodiamonds for quantum sensing,” Phys.Chem. Chem. Phys., vol. 19, no. 17, pp. 10748–10752, 2017.

[29] G. Fleer, M. Cohen Stuart, J. Scheutjens, T. Cosgrove, and B. Vincent,Polymers at Interfaces. London: Chapman and Hall, first edit ed.,1993.

[30] M. V. Gurudev Dutt, L. Childress, L. Jiang, E. Togan, J. Maze,F. Jelezko, A. S. Zibrov, P. R. Hemmer, and M. D. Lukin, “Quan-tum register based on individual electronic and nuclear spin qubits indiamond,” Science (80-. )., vol. 316, pp. 1312–1316, jun 2007.

[31] G. D. Fuchs, G. Burkard, P. V. Klimov, and D. D. Awschalom, “Aquantum memory intrinsic to single nitrogen-vacancy centres in dia-mond,” Nat. Phys., vol. 7, pp. 789–793, oct 2011.

[32] J. Tisler, R. Reuter, A. Lammle, F. Jelezko, G. Balasubramanian, P. R.Hemmer, F. Reinhard, and J. Wrachtrup, “Highly efficient FRET froma single nitrogen-vacancy center in nanodiamonds to a single organicmolecule,” ACS Nano, vol. 5, pp. 7893–7898, oct 2011.

[33] A. Gruber, A. Drabenstedt, C. Tietz, L. Fleury, J. Wrachtrup, andC. Von Borczyskowski, “Scanning confocal optical microscopy andmagnetic resonance on single defect centers,” Science (80-. )., vol. 276,pp. 2012–2014, jun 1997.

[34] S. Kuhn, C. Hettich, C. Schmitt, J. P. Poizat, and V. Sandoghdar,“Diamond colour centres as a nanoscopic light source for scanningnear-field optical microscopy,” J. Microsc., vol. 202, no. 1, pp. 2–6,2001.

[35] M. Loretz, S. Pezzagna, J. Meijer, and C. L. Degen, “Nanoscale nu-clear magnetic resonance with a 1.9-nm-deep nitrogen-vacancy sensor,”Appl. Phys. Lett., vol. 104, jan 2014.

30

REFERENCES

[36] B. K. Ofori-Okai, S. Pezzagna, K. Chang, M. Loretz, R. Schirhagl,Y. Tao, B. A. Moores, K. Groot-Berning, J. Meijer, and C. L. Degen,“Spin properties of very shallow nitrogen vacancy defects in diamond,”Phys. Rev. B - Condens. Matter Mater. Phys., vol. 86, aug 2012.

[37] S. Kaufmann, D. A. Simpson, L. T. Hall, V. Perunicic, P. Senn,S. Steinert, L. P. McGuinness, B. C. Johnson, T. Ohshima, F. Caruso,J. Wrachtrup, R. E. Scholten, P. Mulvaney, and L. Hollenberg, “De-tection of atomic spin labels in a lipid bilayer using a single-spin nan-odiamond probe,” Proc. Natl. Acad. Sci., vol. 110, pp. 10894–10898,jul 2013.

[38] O. Shenderova, N. Nunn, T. Oeckinghaus, M. Torelli, G. McGuire,K. Smith, E. Danilov, R. Reuter, J. Wrachtrup, A. Shames,D. Filonova, and A. Kinev, “Commercial quantities of ultrasmall flu-orescent nanodiamonds containing color centers,” in Adv. PhotonicsQuantum Comput. Mem. Commun. X, vol. 10118, p. 1011803, SPIE,feb 2017.

[39] Z. Y. Lien, T. C. Hsu, K. K. Liu, W. S. Liao, K. C. Hwang, andJ. I. Chao, “Cancer cell labeling and tracking using fluorescent andmagnetic nanodiamond,” Biomaterials, vol. 33, pp. 6172–6185, sep2012.

[40] V. Paget, J. A. Sergent, R. Grall, S. Altmeyer-Morel, H. A. Girard,T. Petit, C. Gesset, M. Mermoux, P. Bergonzo, J. C. Arnault, andS. Chevillard, “Carboxylated nanodiamonds are neither cytotoxic norgenotoxic on liver, kidney, intestine and lung human cell lines,” Nan-otoxicology, vol. 8, no. SUPPL. 1, pp. 46–56, 2014.

31

3Nanodiamond for Sample Preparation in Proteomics

Felipe Perona Martınez, Andreas Nagl, Sona Guluzade, Romana Schirhagl

Department of Biomedical Engineering, University Medical Center Groningen, GroningenUniversity, Antonius Deusinglaan 1, 9713 AW Groningen, The Netherlands.

Analytical Chemistry 91:15 9800-9805 (2019)

SAMPLE PREPARATION IN PROTEOMICS

Abstract

Protein analysis of potential disease markers in blood is complicated bythe fact that proteins in plasma show very different abundances. As a re-sult, highly abundant proteins dominate the analysis, which often renderanalysis of low abundance proteins impossible. Depleting highly abundantproteins is one strategy to solve this problem. Here we present for the firsttime a very simple approach based on selective binding of serum proteinsto the surface of nanodiamonds. In our first proof of principle experimentswe were able to detect on average 8 proteins that are below a ng/ml (in-stead of 0.5 in the control without sample preparation). Remarkably, wedetect proteins down to a concentration of 400 pg/ml after only one simpledepletion step. Among the proteins we could analyse are also numerousdisease biomarkers including markers for multiple cancer forms, cardiovas-cular diseases or Alzheimer’s disease. Remarkably, many of the biomarkerswe find could also not be detected with a state-of-the-art UHPLC column(which depletes the 64 most abundant serum proteins).

34

INTRODUCTION

3.1 Introduction

It is believed that the majority of disease markers are still unidentified sincethey are among the low abundance proteins in plasma[1]. However, in thepast years several methods have been developed to deplete high abundantproteins from serum and thus allow analysis of low abundance proteins:For instance, there are commercially available HPLC (high pressure liquidchromatography) columns, which contain antibodies against high abundantproteins and thus retain them in the column[2, 3, 4]. While initially onlya few proteins were depleted, now columns are available which depleteseveral tens of proteins simultaneously. An alternative is extraction withan organic solvent[5]. Another approach is to use nanoparticles, which bindto certain proteins. Liu et al for instance used several steps of precipitationwith PEG for this purpose followed by depletion with one of the above-mentioned antibody columns[6]. Large amounts of proteins have also beenidentified. However, the authors used more complex multi-step protocols[7].An alternative where you do not need specific antibodies for are molecularlyimprinted polymer particles[8]. To produce these one needs to imprint apolymer with the proteins that need to be depleted. However, in order toachieve this one needs know the proteins that should be depleted and havethem available. This issue was solved elegantly by Yang et al.;[9, 10] theauthors imprinted with the full bovine serum. By varying the concentrationthat was used for imprinting, they could tune the amount of proteins thatare adsorbed.

An alternative approach for protein enrichment is combinatorial pep-tide ligand libraries (CPLL)[11]. To produce the library, beads are coatedwith many different covalently attached peptides[12]. These bind differentproteins in the serum, which are thus removed from the sample. The re-maining serum is strongly depleted of all kinds of proteins including themost abundant ones. This approach does not require specific antibodiesor prior knowledge and has already been successfully applied to severaldifferent samples with a complex proteome[13, 14, 15].

However, despite these efforts depletion of high abundance proteins stillremains an issue[16]. Here we show a simple, fast and cost effective methodto achieve high abundant protein depletion. To achieve protein depletion,we use the fact that only some proteins bind to nanodiamonds. Our ap-proach works similarly to CPLLs in the sense that there are particles thatbind to a lot of different proteins. However, we have the advantage thatour particles are slightly simpler and since there is no biomolecules attachedthey are likely more durable. A disadvantage is probably that the surfacechemistry is less complex and thus probably binds less proteins than the

35


complex surface of CPLLs.

The nanodiamonds in our experiments have traditionally been used asabrasive and are thus readily available commercially. They also recentlygained popularity for their magneto-optical properties[17] their use as long-term fluorescent label[18, 19], as well as their use in drug delivery[20]. How-ever, their application in depleting high abundant proteins from plasma isentirely new.

3.2 Materials and Methods

To eliminate high abundance proteins nanodiamonds and NaCl were addedto the serum. As a result, aggregates precipitate. Since several of the highlyabundant proteins bind poorly to the diamond surface one can deplete themby removing the supernatant. When the protein corona on the diamondsurface is analysed with mass spectrometry we find an increased numberof low abundance proteins. For a schematic representation of the proteindepletion see Figure 3.1.

Figure 3.1: Schematic representation of the experiment: First nanodiamondsand salts are mixed with the serum samples. Certain proteins (mostly proteinswho’s biological function is binding to negatively charged molecules) adhere to thediamonds. Analysing proteins on the diamond particles reveals that high abundantproteins were successfully depleted. At this point loosely binding proteins, the so-called soft corona is still adhered. These proteins can be removed by an additionalwashing step, which further depletes some proteins.

3.2.1 Materials

Throughout this article we used nanodiamonds with a hydrodynamic di-ameter of 25 nm from Microdiamant and a flake like structure[21]. Theyare produced by the manufacturer via grinding high pressure high temper-ature diamonds. Since the diamonds are acid cleaned their surface containsoxygen groups[22]. As a result, mostly proteins with positive domains or

36

MATERIALS AND METHODS

proteins, which in nature bind to negatively charged molecules adhere tothe particles. Human plasma was donated to us from the Bischoff groupand stored at −80◦C in aliquots until use.

3.2.2 Sample preparation

To achieve binding, we added nanodiamonds (25 nm diameter from Micro-diamant) and NaCl, which were previously identified to facilitate diamondaggregation, to the serum[23]. After aggregation the samples were cen-trifuged (13200 rpm for 21 minutes) and the supernatant was removed.These aggregates contain also loosely bound proteins, the so-called softcorona (which was also found on other nanoparticles[24, 25, 26]). The sam-ples were then either analysed immediately or washed. To wash the par-ticles the pellets were resuspended in distilled water once and centrifugedagain. Subsequently, the supernatant was removed leaving only the tightlybound proteins behind in the pellet followed by freeze-drying. The controlsample was the pure serum. To prepare the samples for mass spectrome-try, they were subjected to the digesting protocol published in [27]. Smallamounts of the freeze-dried sample (and a few microliters of the control,respectively) were first treated with 20 µL freshly prepared 10 mM dithio-threitol (DTT) in 100 mM NH4HCO3 to reduce the protein. This wasfollowed by an incubation step at 55-60◦C for 30 minutes. The alkylationof the cysteines was achieved by adding 10 µL iodoacetamide in 100 mMNH4HCO3 (incubation for 45 minutes). Subsequently a second treatmentwith DTT followed for 30 minutes (to remove unreacted iodoacetamide). Atrypsin digest followed by adding 20 µL solution with 10 ng/µL trypsin (se-quencing grade, Promega, Madison, United States). An overnight incuba-tion followed at 37◦C. A clean-up using SPE with C-18 cartridges followedusing a 70/30/0.1 acetonitrile/water/formic acid mixture for elution.

3.2.3 Sample preparation with carbon black

Next we answered whether the protein depletion is specific for diamondnanoparticles. To this end, we prepared our samples in the exact same wayas with FND except replacing the FNDs with carbon black.

3.2.4 Protein analysis

The samples were analysed by nanoLC–MS/MS on an Ultimate 3000 sys-tem (Dionex, Amsterdam, The Netherlands) interfaced on-line with a Q-ExactivePlus (Orbitrap) mass spectrometer (Thermo Fisher Scientific Inc.,Waltham, Massachusetts, United States). Peptide mixtures were loaded

37


onto a 5 mm × 300 µm i.d. trapping micro column packed with C18PepMAP100 5 µm particles (Dionex) in 2% acetonitrile in 0.1% formic acidat the flow rate of 20 µL/min. After loading and washing for 3 minutes,peptides were back-flush eluted onto a 15 cm × 75 µm i.d. nanocolumn,packed with C18 PepMAP100 1.8 µm particles (Dionex). The following mo-bile phase gradient (total run time: 75 minutes) was delivered at the flowrate of 300 nL/min: 2–50% of solvent B in 60 min; 50–90% B in 1 min; 90%B during 13 min, and back to 2% B in 1 min (held for 15 minutes). SolventA was 100:0 H2O/acetonitrile (v/v) with 0.1% formic acid and solvent Bwas 0:100 H2O/acetonitrile (v/v) with 0.1% formic acid. Peptides wereinfused into the mass spectrometer via dynamic nanospray probe (ThermoFisher Scientific Inc.) with a stainless steel emitter (Thermo Fisher Scien-tific Inc.). Typical spray voltage was 1.8 kV with no sheath and auxiliarygas flow; ion transfer tube temperature was 275◦C. Mass spectrometer wasoperated in data-dependent mode. DDA cycle consisted of the survey scanwithin m/z 300–1650 at the Orbitrap analyser with target mass resolutionof 70,000 (FWHM, full width at half maximum at m/z 200) followed byMS/MS fragmentations of the top 10 precursor ions. Singly charged ionswere excluded from MS/MS experiments and m/z of fragmented precursorions were dynamically excluded for further 20 s.

3.2.5 Data processing

The software PEAKS Studio version 7 (Bioinformatics Solutions Inc., Wa-terloo, Canada) was applied to the spectra generated by the Q-exactive plusmass spectrometer to search against either the protein sequence databaseUniProtKB/Trembl of the UniProt Knowledgebase (UniProtKB), limitedto protein sequences of homo sapiens (a search including the whole databasewas performed as well in order to rule out the relevance of possible con-tamination). Searching for the fixed modification carbamidomethylation ofcysteine and the variable post-translational modifications oxidation of me-thionine was done with a maximum of 5 post-translational modificationsper peptide at a parent mass error tolerance of 10 ppm and a fragmentmass tolerance of 0.02 Da. False discovery rate was set at 0.1%.

From the mass spectrometry one obtains spectral counts. These reflecthow often protein fragments are found that can be attributed to a cer-tain protein. However, larger proteins naturally lead to more fragments.To compensate this fact one needs to calculate normalised spectral counts.These give a semi-quantitative measure for the (relative) concentration ofa certain protein in the sample. The normalized spectral counts are calcu-

38

MATERIALS AND METHODS

lated by using the following equation[28, 29, 30].

NpSpC k = 100 ∗(

(SpC/MW )k∑ni=1(SpC/MW )i

)(3.1)

Where NpSpCk is the normalized percentage of spectral count (whichis the number of spectra associated to a protein) for protein k, SpC is thespectral count identified and Mw is the molecular weight (in daltons) ofthe protein k.

Waterfall plots were created by comparing the protein lists with thehuman proteome project database (HPP-DB). The concentrations in thedatabase reflect the current knowledge from selected references.

39


3.3 Results

When we precipitate proteins together with nanodiamonds in a salt-contain-ing medium, we find that some of the most abundant serum proteins bindpoorly to the nanodiamonds surface. We then used liquid chromatogra-phy coupled with mass spectrometry (LCMS) analysis to determine whichproteins can be found on the diamond surface. We typically find severalhundred proteins on our diamond surface. Figure 3.2 summarizes the de-pletion that we find for different media.

To generate the figure, we added the normalized spectral count val-ues (which give a rough estimate for the concentration) for the five most-abundant proteins. The first bar (shown in green in Figure 3.2) representsthe control, where the serum was analysed without our method. We in-vestigated the depletion after adding Dulbecco Modified Eagle Medium(DMEM), since this is one of the most common cell culture media. Inaddition, we had first found a similar depletion effect for bovine serum pro-teins, which are routinely used in mammalian cell cultures[16]. However,as we can see here, mainly the salt component of the medium is responsiblefor the precipitation.

To determine the optimal conditions where most low-abundance pro-teins bind to the surface while high-abundance proteins remain in the su-pernatant, we tested different salt concentrations. The concentrations thatwe chose were near the physiological concentration of 6.9 mg/mL NaCl. Inaddition to varying the salt concentration, we also investigated the effect ofwashing in order to differentiate between the hard and soft protein corona.The soft corona (before washing) contains loosely and strongly binding pro-teins. The hard corona is what remains after washing and only containsstrongly binding proteins. For most cases, we do see a small decrease inhigh-abundance proteins after washing. In addition to quantifying the mostabundant proteins, we were also interested in the composition of the pro-tein corona. Figure 3.3 shows which categories of proteins we find on whichsample.

The categories are chosen based on their biological function. To makethis classification, we ranked the proteins from the highest concentration tothe lowest concentration. We took into account all proteins in the top 50%.We chose to use the top 50% here, since, for lower-abundance proteins,these classifications are scarce or not available at all. The groups, based onbiological functions that we could distinguish, are apolipoproteins (APO),complement factors (COM), other (O), immunoglubulins (IG), acute phaseproteins (ACP), and coagulation factors (COA). We found large differencesin the corona composition. Whereas, in the control, the top 50% of the

40

RESULTS

Figure 3.2: Depletion of high-abundance proteins with nanodiamonds. Com-pared to the control (serum without any treatment), shown in green, the amountof high-abundance proteins that is found by mass spectrometry is significantly re-duced when these were previously depleted with nanodiamonds. Different mediaare used to precipitate protein-coated diamonds, and the depletion is compared.Error bars are generated from three different independent experiments and repre-sent the standard error of the mean.

41


Figure 3.3: Analysing the proteins that are found on the diamonds. Dependingon the sample (panels (a)–(i)), the most prominent 50% can be assigned to differentgroups of proteins with different functions. [Legend: APO, apolipoproteins; COM,complement factors; O, other; IG, immunoglubulins; ACP, acute phase proteins;and COA, coagulation factors.

corona consists of apolipoproteins, the diamond samples are more diverse.Most likely binding to the diamond surface occurs via electronegative oxy-gen groups on the diamond surface, which can interact with electropositivegroups within proteins. While we could not establish a clear relationshipbetween, for instance, binding and the isoelectric point of the proteins, wedo often see proteins binding whose function in biology is to bind to elec-tronegative structures. What we observe is similar to CPLLs, which offera rich surface chemistry, to which proteins can bind. Similar to CPLLs,we also do not target a specific protein or a number of protein (as an anti-body column does) but rather deplete anything that does not bind. Next,we compared the samples based on their ability to detect low-abundanceproteins. To this end, we used so-called “waterfall plots”. To constructa waterfall plot, the protein lists are compared with the database. Figure3.4 shows one of these waterfall plots, which we obtained for the best con-dition (serum + 6.9 mg/mL NaCl + FND). The proteins in the databaseare plotted in order of decreasing concentrations. Every protein that isidentified in the sample receives a blue dot. To illustrate the improvement,a dotted line is used to indicate the lowest concentrated protein that wecould detect with the control. The proteins below that dotted line (marked

42

RESULTS

Figure 3.4: Graphical depiction of the waterfall plot: the waterfall plot lists allproteins, starting with the most-concentrated ones down to the least concentratedones. Each blue dot indictes that the protein was found in the sample. Thewaterfall plot shown here is from the condition with serum + 6.9 mg/mL NaCl +nanodiamonds. The dotted green line shows the detection limit for the control.All the proteins in the red square are only accessible with our sample preparationmethod.

with a rectangle) are only accessible with the diamond sample preparationstep.

Most interesting for proteomics are proteins with concentrations of <1ng/mL. These are challenging to analyse without specialized sample prepa-ration. In Figure 3.5, we compare how many of these low-abundance pro-teins one can find with each sample preparation method. The conditionwith serum + 6.9 mg/mL NaCl + FND, which can reveal eight proteins,on average, gives the best results. For instance, the control only gives 0.5proteins, on average.

As a final assessment of usefulness of our method, we compared the pro-teins that we could identify with proteins that are already used as biomark-ers in the literature. Table 3.1 gives a few examples, which seemed to bemost interesting to us.

43


Figure 3.5: Low-abundance proteins: To demonstrate the abilities of our method,we compare the amount of proteins that were found in the samples that are below1 ng/mL in the original plasma sample. Error bars are generated from threedifferent independent experiments and represent the standard error of the mean.

44

RESULTS

Protein Clinical relevance Ref.

von Willebrandfactor

Willebrand disease, the most common inheritedbleeding disorder.

[31]

TetranectinMarker for disease activity in patients withrheumatoid arthritis.

[32]

Proteoglycan 4Diagnostic biomarker for COPD (chronicobstructive pulmonary disease).

[33]

Vitamin D- bindingprotein

Risk factor for colorectal cancer. [34]

Fibulin-1Cardiovascular risk markers in chronic kidneydisease and diabetes.

[35]

Hornerin Aberrantly expressed in breast cancer. [36]

Hepatocyte growthfactor activator

Diagnostic value for numerous diseases as well asage and pregnancy.

[37]

Apolipoprotein MSuspected to be a biomarker for certain diabetestypes.

[38]

EndostatinDiagnosing malignant pleural effusions, antiangiogenic agent.

[39]

Suprabasin Tumor endothelial cell marker. [40]

AngiogeninUsed in prediction of failure on long-term treatmentresponse and for poor overall survival innon-Hodgkin lymphoma (a certain cancer type).

[41]

Desmoplakin Biomarker for Creutzfeldt-Jakob disease. [42]

Ribonuclease 4 Diagnosis of pancreatic cancer. [43]

Table 3.1: Examples of Proteins Identified in the Best Sample (Serum + 6.9mg/mL NaCl + FND) That Could Be Detected Neither in the Reference norwith a State-of-the-Art Depletion Column with 64 Antibodies and Their ClinicalRelevance

45


Sample preparation with carbon black: Finally, we wanted to de-termine if the depletion effect that we see is specific for diamond. Whendiamond is replaced in the above-mentioned experiments, as shown in Fig-ure 3.6, we do not observe any depletion effects under any conditions. Thisfinding indicates that the depletion of high-abundance serum proteins isindeed a peculiarity of diamond nanoparticles (or particles that resemblethem). The main difference between carbon black and HPHT diamond isthe content of SP2 vs SP3. While carbon black contains large amountsof SP2 (carbon black is actually more similar to graphite than it is to dia-mond), HPHT diamond is almost exclusively SP3 carbon. The consequenceis that carbon black can interact with proteins via π–π interactions (whichare not available in diamond). If such groups are exposed on the proteinsurface, they will interact more with carbon black. Oxygen-containing po-lar groups, on the other hand, are more prominent on the diamond surface.Since graphitic layers are (apart from defects) saturated and give less op-portunities for oxygen-containing groups.

Figure 3.6: Comparison with carbon black. Compared with the control (green,just serum), we do not observe any significant depletion for any conditions usingcarbon black (blue). Also, the washing step did not improve the situation (red).We added the FND samples for comparison and for a positive control.

46

CONCLUSIONS

3.4 Conclusions

While antibody-based depletion columns are generally quite expensive, nan-odiamonds are surprisingly inexpensive, since they are commercially avail-able mass products, which are used as abrasives. In addition, the deple-tion process is just one fast and straightforward step. While antibodiesbind very specifically to a predefined target, here, we use a less-specificapproach. We believe that proteins bind to specific groups on the dia-mond. Diamond particles provide a rich surface chemistry, which provideall types of oxygen-containing groups that (similar to a CPLL) can interactwith different proteins. During our experiments, we were able to depletehigh-abundance proteins significantly. As a result, we have access to low-abundance proteins for analysis, which would otherwise be undetectable.With this simple method, we were able to detect proteins down to thepg/mL range. The best results (the highest number of low-abundance pro-teins) that we can achieve were found when salt was added in physiologicalconcentrations. With this approach, we are able to detect several diseasebiomarkers, including, among others, markers for several cancer types, car-diovascular diseases, or kidney function.

47


References

[1] V. Polaskova, A. Kapur, A. Khan, M. P. Molloy, and M. S. Baker,“High-abundance protein depletion: Comparison of methods for hu-man plasma biomarker discovery,” Electrophoresis, vol. 31, pp. 471–482, feb 2010.

[2] L. A. Echan, H. Y. Tang, N. Ali-Khan, K. B. Lee, and D. W. Spe-icher, “Depletion of multiple high-abundance proteins improves proteinprofiling capacities of human serum and plasma,” Proteomics, vol. 5,pp. 3292–3303, aug 2005.

[3] R. Millioni, S. Tolin, L. Puricelli, S. Sbrignadello, G. P. Fadini, P. Tes-sari, and G. Arrigoni, “High abundance proteins depletion vs lowabundance proteins enrichment: Comparison of methods to reducethe plasma proteome complexity,” PLoS One, vol. 6, no. 5, 2011.

[4] N. I. Govorukhina, A. Keizer-Gunnink, A. G. Van Der Zee, S. DeJong, H. W. De Bruijn, and R. Bischoff, “Sample preparation of hu-man serum for the analysis of tumor markers: Comparison of differentapproaches for albumin and γ-globulin depletion,” J. Chromatogr. A,vol. 1009, no. 1-2, pp. 171–178, 2003.

[5] O. Chertov, A. Biragyn, L. W. Kwak, J. T. Simpson, T. Boronina,V. M. Hoang, D. R. A. Prieto, T. P. Conrads, T. D. Veenstra, andR. J. Fisher, “Organic solvent extraction of proteins and peptidesfrom serum as an effective sample preparation for detection and iden-tification of biomarkers by mass spectrometry,” Proteomics, vol. 4,pp. 1195–1203, apr 2004.

[6] Z. Liu, S. Fan, H. Liu, J. Yu, R. Qiao, M. Zhou, Y. Yang, J. Zhou, andP. Xie, “Enhanced detection of low-abundance human plasma proteinsby integrating polyethylene glycol fractionation and immunoaffinitydepletion,” PLoS One, vol. 11, nov 2016.

[7] H. Ye, L. Sun, X. Huang, P. Zhang, and X. Zhao, “A proteomic ap-proach for plasma biomarker discovery with 8-plex iTRAQ labelingand SCX-LC-MS/MS,” Mol. Cell. Biochem., vol. 343, pp. 91–99, oct2010.

[8] R. Gao, S. Zhao, Y. Hao, L. Zhang, X. Cui, D. Liu, M. Zhang, andY. Tang, “Synthesis of magnetic dual-template molecularly imprintednanoparticles for the specific removal of two high-abundance proteins

48

REFERENCES

simultaneously in blood plasma,” J. Sep. Sci., vol. 38, pp. 3914–3920,nov 2015.

[9] C. Yang, Y. R. Liu, Y. Zhang, J. Wang, L. L. Tian, Y. N. Yan, W. Q.Cao, and Y. Y. Wang, “Depletion of abundant human serum pro-teins by per se imprinted cryogels based on sample heterogeneity,”Proteomics, vol. 17, no. 9, pp. 1–8, 2017.

[10] C. Yang, Y. Zhang, W. Q. Cao, X. F. Ji, J. Wang, Y. N. Yan, T. L.Zhong, and Y. Wang, “Synthesis of molecularly imprinted cryogelsto deplete abundant proteins from bovine serum,” Polymers (Basel).,vol. 10, no. 1, 2018.

[11] E. Boschetti, A. D’Amato, G. Candiano, and P. G. Righetti, “Proteinbiomarkers for early detection of diseases: The decisive contributionof combinatorial peptide ligand libraries,” sep 2018.

[12] G. Candiano, L. Santucci, A. Petretto, C. Lavarello, E. Inglese, M. Br-uschi, G. M. Ghiggeri, E. Boschetti, and P. G. Righetti, “Widening anddiversifying the proteome capture by combinatorial peptide ligand li-braries via alcian blue dye binding,” Anal. Chem., vol. 87, pp. 4814–4820, may 2015.

[13] E. Gonzalez-Garcıa, M. L. Marina, M. C. Garcıa, P. G. Righetti, andE. Fasoli, “Identification of plum and peach seed proteins by nLC-MS/MS via combinatorial peptide ligand libraries,” J. Proteomics,vol. 148, pp. 105–112, oct 2016.

[14] M. Van Vaerenbergh, G. Debyser, G. Smagghe, B. Devreese, and D. C.De Graaf, “Unraveling the venom proteome of the bumblebee (Bom-bus terrestris) by integrating a combinatorial peptide ligand libraryapproach with FT-ICR MS,” Toxicon, vol. 102, pp. 81–88, jul 2015.

[15] S. Pisanu, G. Biosa, L. Carcangiu, S. Uzzau, and D. Pagnozzi, “Com-parative evaluation of seven commercial products for human serumenrichment/depletion by shotgun proteomics,” Talanta, vol. 185,pp. 213–220, aug 2018.

[16] C. Tu, P. A. Rudnick, M. Y. Martinez, K. L. Cheek, S. E. Stein, R. J.Slebos, and D. C. Liebler, “Depletion of abundant plasma proteins andlimitations of plasma proteomics,” J. Proteome Res., vol. 9, pp. 4982–4991, oct 2010.

49



[18] Y. K. Tzeng, O. Faklaris, B. M. Chang, Y. Kuo, J. H. Hsu, andH. C. Chang, “Superresolution imaging of albumin-conjugated fluores-cent nanodiamonds in cells by stimulated emission depletion,” Angew.Chemie - Int. Ed., vol. 50, pp. 2262–2265, mar 2011.

[19] T. D. Merson, S. Castelletto, I. Aharonovich, A. Turbic, T. J. Kil-patrick, and A. M. Turnley, “Nanodiamonds with silicon vacancy de-fects for nontoxic photostable fluorescent labeling of neural precursorcells,” Opt. Lett., vol. 38, p. 4170, oct 2013.

[20] K. K. Llu, W. W. Zheng, C. C. Wang, Y. C. CMu, C. L. Cheng,Y. S. Lo, C. Chen, and J. I. Chao, “Covalent linkage of nanodiamond-paclitaxel for drug delivery and cancer therapy,” Nanotechnology,vol. 21, jul 2010.

[21] S. Y. Ong, M. Chipaux, A. Nagl, and R. Schirhagl, “Shape and crys-tallographic orientation of nanodiamonds for quantum sensing,” Phys.Chem. Chem. Phys., vol. 19, no. 17, pp. 10748–10752, 2017.

[22] A. Nagl, S. R. Hemelaar, and R. Schirhagl, “Improving surfaceand defect center chemistry of fluorescent nanodiamonds for imagingpurposes-a review,” Anal. Bioanal. Chem., vol. 407, no. 25, pp. 7521–7536, 2015.

[23] S. R. Hemelaar, A. Nagl, F. Bigot, M. M. Rodrıguez-Garcıa, M. P.de Vries, M. Chipaux, and R. Schirhagl, “The interaction of fluorescentnanodiamond probes with cellular media,” Microchim. Acta, vol. 184,pp. 1001–1009, apr 2017.

[24] S. Winzen, S. Schoettler, G. Baier, C. Rosenauer, V. Mailaender,K. Landfester, and K. Mohr, “Complementary analysis of the hardand soft protein corona: Sample preparation critically effects coronacomposition,” Nanoscale, vol. 7, pp. 2992–3001, feb 2015.

[25] S. Tenzer, D. Docter, J. Kuharev, A. Musyanovych, V. Fetz, R. Hecht,F. Schlenk, D. Fischer, K. Kiouptsi, C. Reinhardt, K. Landfester,H. Schild, M. Maskos, S. K. Knauer, and R. H. Stauber, “Rapid for-mation of plasma protein corona critically affects nanoparticle patho-physiology,” Nat. Nanotechnol., vol. 8, no. 10, pp. 772–781, 2013.

50

REFERENCES

[26] A. Salvati, A. S. Pitek, M. P. Monopoli, K. Prapainop, F. B. Bombelli,D. R. Hristov, P. M. Kelly, C. Aberg, E. Mahon, and K. A. Dawson,“Transferrin-functionalized nanoparticles lose their targeting capabil-ities when a biomolecule corona adsorbs on the surface,” Nat. Nan-otechnol., vol. 8, no. 2, pp. 137–143, 2013.

[27] N. I. Govorukhina, T. H. Reijmers, S. O. Nyangoma, A. G. van derZee, R. C. Jansen, and R. Bischoff, “Analysis of human serum byliquid chromatography-mass spectrometry: Improved sample prepara-tion and data analysis,” J. Chromatogr. A, vol. 1120, pp. 142–150, jul2006.

[28] U. Sakulkhu, L. Maurizi, M. Mahmoudi, M. Motazacker, M. Vries,A. Gramoun, M. G. Ollivier Beuzelin, J. P. Vallee, F. Rezaee, andH. Hofmann, “Ex situ evaluation of the composition of protein coronaof intravenously injected superparamagnetic nanoparticles in rats,”Nanoscale, vol. 6, pp. 11439–11450, oct 2014.

[29] U. Sakulkhu, M. Mahmoudi, L. Maurizi, G. Coullerez, M. Hofmann-Amtenbrink, M. Vries, M. Motazacker, F. Rezaee, and H. Hofmann,“Significance of surface charge and shell material of superparamagneticiron oxide nanoparticle (SPION) based core/shell nanoparticles on thecomposition of the protein corona,” Biomater. Sci., vol. 3, pp. 265–278,feb 2015.

[30] W. Zhu, J. W. Smith, and C. M. Huang, “Mass spectrometry-basedlabel-free quantitative proteomics,” 2010.

[31] B. P. Yawn, W. L. Nichols, and M. E. Rick, “Diagnosis and man-agement of von willebrand disease: Guidelines for primary care,” dec2009.

[32] E. F. Kamper, L. T. Kopeikina, V. Koutsoukos, and J. Stavridis,“Plasma tetranectin levels and disease activity in patients withrheumatoid arthritis,” J. Rheumatol., vol. 24, pp. 262–268, feb 1997.

[33] K. Y. Lee, H. C. Chuang, T. T. Chen, W. T. Liu, C. L. Su, P. H. Feng,L. L. Chiang, M. Y. Bien, and S. C. Ho, “Proteoglycan 4 is a diagnosticbiomarker for COPD,” Int. J. COPD, vol. 10, pp. 1999–2007, sep 2015.

[34] H. Q. Ying, H. L. Sun, B. S. He, Y. Q. Pan, F. Wang, Q. W. Deng,J. Chen, X. Liu, and S. K. Wang, “Circulating Vitamin D bindingprotein, total, free and bioavailable 25-hydroxyVitamin D and risk ofcolorectal cancer,” Sci. Rep., vol. 5, jan 2015.

51


[35] A. Scholze, E. M. Bladbjerg, J. J. Sidelmann, A. C. Diederichsen,H. Mickley, M. Nybo, W. S. Argraves, P. Marckmann, and L. M. Ras-mussen, “Plasma concentrations of extracellular matrix protein fibulin-1 are related to cardiovascular risk markers in chronic kidney diseaseand diabetes,” Cardiovasc. Diabetol., vol. 12, jan 2013.

[36] J. M. Fleming, E. Ginsburg, S. D. Oliver, P. Goldsmith, and B. K.Vonderhaar, “Hornerin, an S100 family protein, is functional in breastcells and aberrantly expressed in breast cancer,” BMC Cancer, 2012.

[37] H. Funakoshi and T. Nakamura, “Hepatocyte growth factor: Fromdiagnosis to clinical applications,” 2003.

[38] S. A. Mughal, R. Park, N. Nowak, A. L. Gloyn, F. Karpe, H. Matile,M. T. Malecki, M. I. Mccarthy, M. Stoffel, and K. R. Owen,“Apolipoprotein M can discriminate HNF1A-MODY from Type 1 di-abetes,” Diabet. Med., 2013.

[39] W. B. Zhou, M. Bai, and Y. Jin, “Diagnostic value of vascular en-dothelial growth factor and endostatin in malignant pleural effusions,”Int. J. Tuberc. Lung Dis., 2009.

[40] M. T. Alam, H. Nagao-Kitamoto, N. Ohga, K. Akiyama, N. Maishi,T. Kawamoto, N. Shinohara, A. Taketomi, M. Shindoh, Y. Hida, andK. Hida, “Suprabasin as a novel tumor endothelial cell marker,” Can-cer Sci., 2014.

[41] S. Fang, H. Repo, H. Joensuu, A. Orpana, and P. Salven, “High serumangiogenin at diagnosis predicts for failure on long-term treatmentresponse and for poor overall survival in non-Hodgkin lymphoma,”Eur. J. Cancer, 2011.

[42] J. Gawinecka, B. Ciesielczyk, P. Sanchez-Juan, M. Schmitz, U. Heine-mann, and I. Zerr, “Desmoplakin as a potential candidate for cere-brospinal fluid marker to rule out 14-3-3 false positive rates in sporadicCreutzfeldt-Jakob disease differential diagnosis,” Neurodegener. Dis.,2012.

[43] Y. Wang and s. Yu, “Serum ribonuclease and its isoenzymes for diag-nosis of pancreatic cancer,” Chin. Med. J. (Engl)., 1989.

52

4Observing chemical reactions in situ with diamond based

magnetometry

Felipe Perona Martınez⊥, Anggrek Citra Nusantara⊥, Mayeul Chipaux, SandeepKumar Padamati, Romana Schirhagl

Department of Biomedical Engineering, University Medical Center Groningen, GroningenUniversity, Antonius Deusinglaan 1, 9713 AW Groningen, The Netherlands.⊥These two authors contributed equally.

Submitted article (2019)

DETECTING FREE RADICALS

Abstract

The novel field of diamond magnetometry recently lead to powerful quan-tum sensing methods that allow detecting magnetic resonances with nano-scale resolution. For instance, T1 relaxation measurements, inspired fromequivalent concepts in MRI, provide a signal which is equivalent to T1 inconventional MRI but from a nanoscale environment. Here we provide thefirst real-time measurements of free radicals while they are generated in achemical reaction. More specifically, we follow photolysis of H2O2 as well asthe so-called Haber-Weiss reaction. Both of these processes are importantreactions biological environments. Unlike with other fluorescent probes weare able to detect both increase and decrease in real time. We also inves-tigate different diamond probes and their ability to sense gadolinium spinlabels. While this was so far only done in a clean environment, we take intoaccount the effect of salts and proteins that are present in a biological en-vironment. Furthermore, we conduct our experiments with nanodiamonds,which are compatible with intracellular measurements. Surprisingly, wefind that in contrast to single defect measurements, smaller nanodiamondhave better coherence times. This is an important step towards label-freenano-MRI and quantifying signals in a biological environment.

54

INTRODUCTION

4.1 Introduction

Color defects in diamonds have been studied intensively as novel and pow-erful (quantum) magnetometers[1]. This field has recently gained attentionin other disciplines including biology[2] and geoscience[3], or for industrialapplications[4, 5, 6]. Due to its nearly infinite photo-stability, the nitrogen-vacancy (NV) defect in diamond has been presented as an attractive labelfor cellular structures[7, 8]. Additionally, nanodiamonds show excellentbiocompatibility in all kinds of cell types[9] and organisms[10].

Arguably more remarkable, however, is the NV center’s ability to sensemagnetic resonances optically. It does so by changing its brightness basedon the magnetic surrounding. The technique is so sensitive that the faintmagnetic resonance of a single electron[11] or even a few nuclear spins[12,13] can be detected experimentally. This effect has already been utilised forseveral different applications including characterizing magnetic vortices[14],hard drives[15, 16], nanoparticles[17], single proteins[18] or different chem-icals as vacuum oil[19].

One particularly interesting sensing scheme are the so-called relaxation(or T1) measurements, that are sensitive to spin fluctuations. This puls-ing scheme, which only requires optical pulsing, has already been usedsuccessfully, to detect spin labels[20] copper ions[21], temperature[22] orconductivity[23]. Sushkov et al. demonstrated single-molecule sensitivitywhen detecting gadolinium-containing molecules attached to a diamondsurface[24]. Kaufmann et al. have achieved gadolinium detection in lipidbilayers[25]. Besides detecting spin labels in water Steinert et al. havedemonstrated the first nano-MRI with this technique from a fixed slice of acell embedded in a polymer[20]. The cell was prepared similarly to samplesfor electron microscopy but stained with gadolinium.

For the first time, here we detect free radicals in-situ during a chemicalreaction. We were able to measure a low concentration of *OH radicals (2micro molar) as naturally present in living cells. We generate them froman H2O2 precursor either by UV radiation or by using iron ions with theHaber-Weiss reaction, which can be used to generate radicals. The hydroxylradical, like other free radicals, can easily react with important componentsof the cell, affecting its function and contributing to the development ofdiseases[26]. For example, oxidative alterations of DNA are linked withthe causes and development of cancer[27]. In Alzheimer’s disease, reactiveoxygen species (ROS), and the hydroxyl radical is one of them, play a rolein the origin and propagation of this neuropathology[28]. Hydroxyl radicalsalso play a major role in fighting bacterial or viral infections by the immunesystem, in apoptosis (programmed cell death) as well as the natural ageing

55


process.

There are a couple of indicators for the hydroxyl radical[29], whichcan be used directly in cells or in solutions, but diamond magnetometryoffers additional advantages. The fluorescent labels (some of which areused here for direct comparison) suffer from bleaching and thus continuousmeasurements are not possible. Additionally, the label is consumed inthe reaction and thus it is only possible to detect an increase in radicalformation. Diamond magnetometry allows for better spatial and temporalresolution as well as the ability to repeat the measurement over time withneither destructing the probe nor killing the cell. In addition, we haveshown that it also offers improved sensitivity when quantifying free radicalconcentrations.

Additionally, we perform calibration measurements with known concen-trations of gadolinium to directly compare different sensing conditions. Forthe first time we take the presence of salts and proteins into account. Weinvestigate different effects by comparing the T1 responses to gadoliniumunder various conditions. Moreover, we have also considered the effect ofthe size of the nanodiamonds in the T1 relaxation constant.

4.2 Results and discussion

In this article, we measure and compare the spin-lattice relaxation times(T1) under biologically relevant conditions. To perform a T1 there is aspecific pulse sequence required which is shown in Figure 4.1a. To do themeasurement we first prepare the defects in the ground state. This is doneby a laser pulse. Then we measure again after specific times to see whetherthe NV centers are still in this state or not. Since the states differ inbrightness (the ground state is brighter) we can observe the process byrecording the change in fluorescence. When there are flipping spins (inthis case from gadolinium or free radicals) in the surrounding, the NVcenters will lose this state faster. Thus the time that is required to lose theprepared state gives a quantitative measure for the concentration for theconcentration of these species. Two typical curves taken in presence andabsence of gadolinium are shown in Figure 4.1b.

Unlike many other groups in the field, we use NV center ensembleshosted in nanodiamonds. Compared to single NV center measurements thishas the advantage, that each particle is brighter (and thus easier to findeven if there is background fluorescence). Additionally, each measurementalready combines multiple NV centers. As a result, the variability betweenparticles is smaller.

56


Figure 4.1: T1 measurements: (a)shows the pulsing sequence to gener-ate a T1 curve. The green rectanglesindicate when the laser is on while bluerectangles indicate when we read out.The green bars are separated by a darktime τ . This dark time is systemati-cally increased. Plotting the differentdark times against the fluorescence in-tensity that we record during the bluewindows results in curves shown in (b).In the presence of flipping spins, thisdecay is faster. This is for examplethe case when adding gadolinium. Theblue curve results from adding 0.5 mMGd3+

It has to be noted, that the form of the curve we obtain for ensemblesof NV centers is slightly different from single NVs. While for a single NVthe T1 relaxation can be described by a two exponential model[30], here weobserve multiple of these decays with different relaxation times at the sametime. The difference in relaxation times within a particle likely results fromdifferent distances to the particle surface as well as their respective nano-scale environment. The result is a decay curve with a shoulder (see figure4.1b (water)). We also do not observe the effect of the relaxation throughthe metastable state, which is related to an initial build-up of the relax-ation curve. Instead, we obtained monotonically decreasing functions. Weexplain this difference by two reasons. First, in our experiments, the short-est dark-times are not small enough to sample the effect of the metastablestate. Second, because the signal is emitted from hundreds of NV centers,following different dynamics, the initial small build-up of the relaxationcurve is covered by the dispersion of the photoluminescent signals emittedby the NV centers in the excited ensemble.

Due to the fact that we have an ensemble rather than a single center weuse a different model to fit the data and calculate the relaxation constant.For analysis, we approximate that the relaxation of the ensemble consists oftwo components[31], one from NV centers that are very close to the surfaceand another one from NV centers that are deeper in the crystal and thusless influenced by the surface. The model used to fit the data is:

PL(τ) = Iinf + Cae−τ/Ta + Cbe

−τ/Tb

T1 = max(Ta, Tb)(4.1)

57


From equation 4.1, we obtain the constants Ta and Tb. Longer T1 timesare more sensitive to changes in magnetic noise. Generally, the longer theT1 before adding the analytes, the more it can still be reduced which leadsto higher sensitivity. Thus we used the slower one of these relaxation con-stants for analysis and quantification throughout this article. We performedmeasurements of known concentrations of gadolinium under different con-ditions. The aim of these measurements is to optimise the measurementconditions and understand what influence different factors have on the sens-ing capabilities.

4.2.1 Dependence of the size of the nanodiamonds on therelaxation constant T1

The size of the nanodiamonds used in the experiment is an important pa-rameter to set in diamond magnetometry experiments. At first, particles ofdifferent sizes host a different amount of defects. Larger particles containmore NV centers and thus are brighter and easier to detect. On the otherhand, when using the NV centers for sensing, having more NVs in the par-ticles could deteriorate the signal if some of them are not exposed to theexternal magnetic noise. If the NV centers are too close to one another theywill also start to sense each other rather than the external environment.This process decreases sensitivity. In this case, defects in the core of theparticle (too far away to sense the external spins) would only add noise tothe signal.

Figure 4.2: Spin-Lattice (T1) relax-ation constant of three sizes of nanodi-amonds. Under the same conditions, theT1 constant changes depending on thesize of the nanodiamond. The sensitivityto Gd3+ (100 nM) also depends on thesize of the nanoparticles. The error barsshow the standard error of the mean.

To investigate which size issuited best, nanodiamonds of differ-ent sizes were diluted in water andexposed to a fixed concentration ofgadolinium (100 nM). The resultsof these measurements are shown inFigure 4.2. This experiment revealstwo main conclusions. The value ofthe relaxation constant depends onthe size of the nanodiamond host-ing the defects. The smaller theparticle the longer is the relaxationtime. This is unexpected from par-ticles with a single defect where thesize dependency is reversed. Alsoin bulk diamonds coherence timesdecrease with the distance to the

58


surface[25]. In our case, we can explain this behaviour considering thatlarger particles contain not just more NVs but also a higher density. Theinteraction between many NV centers promotes the relaxation of the en-semble, decreasing the value of the T1 constant.

A second observation over this result is made considering the change ofthe T1 constant after adding gadolinium to the sample. The average changein the T1 constant for the different particle size is 209.38 µs (57.3%) for the40 nm, 64.832 µs (45.50%) for the 70 nm and 19.478 µs (46.00%) for the120 nm. This means that the 40 nm particles are more sensitive than thebigger ones. However, as the error bars (standard error) show, the 40 nmparticle show highest dispersion of values compared with the particles of 70and 120 nm. Thus, the accuracy is decreased. The reason is that smallerparticle contain fewer defects. Thus, a lower number of defects leads to agreater spread. On the contrary, on big particles, having a high amountof NVs, the differences are compensated in the average photoluminescentsignal. Taking this result into account, we suggest using bigger particles inapplications where absolute accuracy is crucial and smaller particles whensensitivity is a key and when a high dynamic range is expected.

To reduce the variability of values of T1 while retaining higher coherencetimes and thus sensitivity, the following experiments were performed usingthe 70 nm nanodiamonds.

4.2.2 Performance of the NVs in biologically relevant con-ditions

When sensing spins in a biological environment, the molecules of interestcoexist with other chemicals. Salts and proteins available in the mediummight cover the surface of the particle hosting the NVs[32].To quantifychemicals accurately in such an environment, it is important to know howthese factors influence the signal from the NV centers.

The cell culture medium is composed of salts, proteins, glucose andwater. We have tested the effect of adding these components to a solutionof Gd3+ (a common spin label). Then we compared with the performance ofthe NV centers in water. Figure 4.3 summarises the investigated conditions.

GdCl3 in water

Gadolinium is one of the most common contrast agents in conventionalmagnetic resonance measurements. Thus, measuring gadolinium ions pro-vides a convenient way to compare conditions and determine the influenceof different factors. Measurements in water were conducted as a reference.

59


The concentration of GdCl3 was increased in steps of one order of magni-tude from 1 nM to 100 mM. As expected, there was an inverse relationshipbetween the relaxation constant T1 and the concentration of GdCl3 (reddots in Figure 4.4). The signal saturates around 10 µM. The lowest con-centration we were able to detect was 1 nM.

The presence of salts

Phosphate-buffered saline (PBS) is the most common buffer for biologicalexperiments. The most abundant salts present in the medium are sodiumchloride (NaCl) and disodium phosphate (Na2HPO4). PBS is used to main-tain the pH of the growth medium constant at 7.4 and it is required for cellsto prevent osmotic stress. To investigate how these salts influence the sen-sitivity of quantum sensing we used 70 nm nanodiamonds with ensemblesof NV-centers. We measured different concentrations of gadolinium trichlo-ride (GdCl3) diluted in PBS. Figure 4.4 (green dots) shows the result of themeasurements. The measured T1 constants are close to the control sample(water), indicating that PBS doesn’t influence the sensing performance.

Figure 4.3: Overview of the samples in this article. The experiments in the firstline aim to measure (a) GdCl3 in water, (b) GdCl3 in PBS and (c) GdCl3 in cellculture medium. The goal of these experiments is to determine the influence ofsalts and cell medium proteins on the sensing ability of NV center ensembles innanodiamonds. The lower half of the figure represents measurements of naturallyoccurring species which give a magnetic resonance signal. In (d) Fe(OCl4)2 ismeasured in water and in (e) H2O2 is added to Fe(OCl4)2, which leads to thegeneration of *OH radicals. In each experiment, the compound which causes ismeasured is circled in green dotted lines.

60


Nanodiamonds covered with a protein corona

In a biological medium, nanoparticles never just stand by themselves. In-stead, a corona of proteins and salts surrounds them. We have also ob-served this for diamond particles in our group[32]. There are some ongoingefforts to avoid corona formation on nanodiamonds[33], however they re-quire elaborate chemistry and the coating that is required might influencesensing performance as well. Here we investigate how the presence of thecorona influences sensing performance. The blue points in Figure 4.4 showhow the T1 values change as the concentration of Gd3+ increases. In thepresence of proteins, the change in T1 is shifted towards higher concentra-tions. This is likely due to a shielding effect, in which the proteins attachto the surface of the nanodiamonds and thus hinder Gd3+ to approach thesurface. Thus the decrease in T1 is considerably slower. Additionally, weobserve a larger spread of data in presence of the corona. Although the dis-persion of the measurements we clearly see the reduction of the relaxationconstant as the concentration of gadolinium increases. The larger spreadin T1 values in the samples with protein corona is likely due to differencesin the corona that is formed around each particle. Thickness and exactcomposition (the medium contains a mixture of thousands of proteins) ofthe protein corona can vary significantly.

Figure 4.4: T1 value at sev-eral concentrations of Gd3+.The gadolinium salt was di-luted in three different media,Water, PBS and cell growthmedium (DMEM complete).The error bars represent thestandard error of the meanfrom 4 particles.

61


4.2.3 Measuring free radicals in situ

Photolysis of H2O2

In two different experiments, we show the detection of the hydroxyl radi-cal produced in situ. In the first experiment, we have produced *OH byphotolysis of hydrogen peroxide (H2O2) as indicated in equation 4.2.

H2O2 + hν → 2 *OH (4.2)

To this end, the T1 constant was first measured in H2O2 (30%) in thedark and then with a UV light on (λ = 275 nm). Finally, we also performedanother measurement in the dark (after the radicals had reacted) to testthe reversibility of the measurement.

To estimate the concentration of *OH during the reaction we performeda quantification with disodium terepthalic acid[34]. Na2TH reacts withthe hydroxyl radical, resulting in the formation of the fluorescent moleculeHTA (2-hydroxy terepthalic acid) in a 1:1 proportion. Based on the HTAcalibration curve (see supplementary information Figures 4.8,4.9,4.10) weestimate a concentration of *OH radicals is 0.9 µM.

As Figure 4.5 shows, after initiating the photolysis reaction, the relax-ation of the NV centers find the thermal equilibrium about 30% faster. Thespeedup in the relaxation (and the reduction of T1) in this case, only canbe explained by the generation of the hydroxyl radicals. Moreover, afterstopping the photolysis, we observe a recovery of the initial value indicat-ing a reduction in the concentration of radicals in the NV center’s nearenvironment. As a control we performed the same experiment in absenceof H2O2. This is an important control to rule out any effects the UV ir-radiation might have on the NV centers themselves (as for instance chargeconversion which might occur during irradiation.). In absence of H2O2

we observed the relaxation time to be unperturbed by the UV light (seesupplementary Figure 4.7).

Figure 4.5: Detection of *OHproduced by photolysis of hydro-gen peroxide. The change in theT1 relaxation time after turning onthe UV light is explained by theemergence of 0.9 µM of hydroxylradicals. Error bars show the stan-dard error of the mean obtainedfrom 7 different experiments withdifferent diamonds.

62


The ability to sense both an increase and decrease in the radical con-centration in real-time is specific for the NV center. To the best of ourknowledge this cannot be achieved with any other fluorescent probe todate.

Haber-Weiss reaction

The second experiment performed to detect *OH, mimics the Haber-Weissreaction. This is one of the pathways that cells use to produce *OH. One keystep in this process is the well-known Fenton reaction[35], which describesthe oxidation of Iron II to Iron III by the action of hydrogen peroxide(equation 4.3) producing one hydroxyl radical and one hydroxide ion.

Fe2+ +H2O2 → Fe3+ + *OH +OH− (4.3)

Cellular iron is present mostly linked to other molecules, such as pro-teins or chelated in the labile iron pool (LIP). Free iron (Fe(II) and Fe(III))available in the cell can catalyse a Haber-Weiss reaction, generating oxygen,hydroxide, and the highly reactive hydroxyl radicals[36].

We have reproduced this reaction and measured the radicals created.To benchmark the performance of diamond magnetometry we have runtwo experiments in parallel. One using NV centers and another using thereactive oxygen species probe hydroxyphenyl fluorescein (HPF). While T1

measurements are sensitive to spin noise (giving the overall concentration ofradicals or paramagnetic chemicals), HPF is sensitive to *OH. The resultscan be seen in Figure 4.6. In both cases, we have used ultrapure water asthe negative control, we do not expect to find a measurable trace of *OH inthis sample. The second sample consisted of only the salt which providesthe iron II (Iron II Perchlorate) in ultrapure water. Finally, we start theFenton reaction by incorporating hydrogen peroxide to the sample.

In the HPF measurement the fluorescence intensity of HPF is propor-tional to the amount of *OH in the sample. In Figure 4.6a, it is clearlyvisible that the production of the *OH starts only after the addition of H2O2

to the sample. Also here we determined the concentration of *OH usingNa2TH. The amount of *OH radicals during the reaction is approximately1.96 µM.

When using diamond magnetometry the T1 relaxation time already re-sponds to iron (II) perchlorate. Then, the relaxation constant T1 is reducedeven more after starting the generation of *OH. The first drop is associ-ated with the presence of a paramagnetic form of iron in the sample. Thesecond drop in the T1 value is explained by the generation of the radicalin the sample and it proves that the sensitivity of diamond magnetometry

63


is sufficient to detect the radical, even when it coexists with other sourcesof magnetism. This measurement also demonstrates a useful feature of di-amond magnetometry: the ability to perform a measurement before andafter a change. Having the ability to measure at the exact same spot onthe exact same particle, before and during the reaction gives a powerfulcontrol measurement. It is also important to point out that using the HPFprobe took 14 hours of incubation before having a measurable fluorescentsignal from the sample. The experiment using NV centers required only 20minutes of acquisition.

In this article, we have investigated several aspects of diamond mag-netometry. We have shown that the NV center’s sensitivity to fluctuatingspins depends on the size of the particle which hosts it. The study of theperformance of the NVs under different conditions suggests that a pro-tein corona interferes with the measurement while salts do not alter theoutcome significantly. Despite this perturbation, it is remarkable that di-amond magnetometry detects a concentration of gadolinium as small as 1nM. The detection of hydroxyl radicals, in situ, by means of NV centers,demonstrates the potential of the technique as a sensor. While conventionalprobes suffer from bleaching (and thus can often only be measured in oneshot) NV centers are almost infinitely stable and thus allow real time detec-tion over long times. We also showed that we can follow radical generationduring a chemical reaction and that the detection is fully reversible.

Figure 4.6: Detection of *OH and iron: (a) shows the results of a conventional kitto detect *OH (HPF). An increase in the *OH concentration leads to an increase influorescence (b) shows the same measurements using NV centers in nanodiamondsand T1 relaxation time. The errors bars show the standard error generated from3 independent measurements on different particles.

64

EXPERIMENTAL DETAILS

4.3 Experimental Details

4.3.1 Nanodiamonds

The nanodiamonds used in the experiments were ground HPHT diamondsof different sizes, which are commercially available from Adamas Nanotech-nologies. As a last step of the manufacturing process, these are cleaned withan oxidizing acid and thus have an oxygen-terminated surface. They arealso irradiated by the manufacturer and contain NV centers in a propor-tion of 2 ppm for 40 nm diamonds, 2.5 ppm for 70 nm diamonds and 3ppm for 120 nm diamonds. To produce a homogeneous distribution of nan-odiamonds on the bottom of a glass-bottom cell-culture dish, 100 µl of asuspension of fluorescent nanodiamonds (0.1 µg/ml) was poured into thedish and then the medium was removed. 200 microliters of new media (wa-ter, PBS, cell medium, gadolinium chloride solution or iron(II) perchloratesolutions) were added.

4.3.2 T1 measurements

Using a home-made diamond magnetometer (a confocal microscope withthe ability for sensitive detection with an avalanche photodiode and puls-ing), nanodiamonds were identified. For our measurements we excludedobvious aggregates. We also chose particles with counts between 106 and107 counts per seconds, and relaxation times between 90 and 300 µs. Ex-treme values were excluded since they were either from dirt particles on thesurface, aggregates or background or exceptionally large or small particles.T1 relaxation measurements were conducted. The same set of parameterswas used in each T1 relaxometry experiment. The NV centers were polar-ized by a train of laser pulses with variable dark times (from 0.2 µs to 10ms). The laser (532 nm, CNI, Changchun, China) was attenuated to 100µW at the location of the sample. To ensure the polarization of the NVcenters, the pulse length was set to 5 µs. The photoluminescent signal (PL)was quantified in a detection window of 0.6 µs. The T1 relaxation curvewas fitted with a bi-exponential function, the reported T1 constant is thehigher time constant yield by the fitting (equation 4.1).

4.3.3 GdCl3 sensitivity in different conditions

To prepare a stock solution gadolinium (III) chloride (Aldrich 439770-5G)was dissolved in MQ water, PBS or cell medium to a concentration of1 M. Starting with the control sample (100 µL of solvent). Before themeasurement the GdCl3 solution was added gradually to give the desired

65


concentrations (0.001 µM, 0.01 µM, 0.1 µM, 1 µM, 10 µM, 100 µM, 1 mM 10mM and 100 mM,). T1 measurements were recorded at each concentrationaccording to the procedure described above.

4.3.4 Hydrogen Peroxide and UV light

The hydrogen peroxide 30% was purchased from Sigma-Aldrich. *OH rad-icals are generated by photolysis of hydrogen peroxide stimulated by UVlight (275 nm, 23.7 mW cm-2). The relaxation constant T1 was recordedwith and without exposure to UV light. The particles were exposed to theUV radiation during the complete acquisition time, which was about 17minutes.

4.3.5 Measuring the hydroxyl radical by HPF

To validate our results we repeated the same experiments that we did withT1 measurements with HPF. The hydroxyl radical and hydroxyphenyl flu-orescein (HPF) were purchased from ThermoFisher (H36004). The exper-iments were conducted by following the procedure proposed by the manu-facturer. The samples were prepared in a 96 well-plate in sextuplet. Theconcentration of HPF in each well was 10 µM. Iron (II) perchlorate wasadded to a concentration of 100 µM, and hydrogen peroxide at 1 mM. Thesamples were excited with light at 485 nm and the signal was quantified at520 nm using a FLUOstar OPTIMA plate reader (BMG LABTECH). Thesamples were measured at intervals of one hour for 14 hours.

4.3.6 Measuring the concentration of hydroxyl radical byHTA

This assay was used to validate our T1 data and to determine which con-centration of *OH was present. 2-hydroxy terepthalic acid (HTA, sigma-aldrich used without further purification) acts as a standard chemical trapfor hydroxyl radicals and is a standard hydroxyl dosimeter. To determinethe concentrations of *OH in our reaction and to validate our T1 results,we used iron (II) perchlorate (10 µM), H2O2 (1000 µM) and Na2TH (200µM). A calibration curve has been established with different concentrationsof HTA (vs) Intensity using fluorimeter (Edinburgh instruments (modulesc-20), λExcition = 330 nm and λEmission = 420 nm). From the calibrationcurve the concentration of hydroxyl radicals is determined[34].

66

CONCLUDING REMARKS

4.3.7 Measuring the Fenton reaction by diamond magne-tometry

70 nm nanodiamonds were fixated in a glass bottom culture dish accordingto the procedure described above (section “Nanodiamonds”). After findinga particle containing the ensemble of NVs, 180 µl of ultrapure water wasadded to the plate and the first set of measurements was performed (inwater only). Iron (II) perchlorate was added until a concentration of 10 µMwas reached. Another set of T1 relaxometry measurements was conductedover the same ensemble of NV center used in the previous case. To finish,H2O2 was incorporated into the previous solution and a new set of measureswere taken.

4.4 Concluding remarks

While so far chemicals have only been detected by diamond magnetometryin water we take into account the presence of salts, glucose and proteinsin the medium. We find, that the formation of a protein corona in nan-odiamonds influence sensing performance while no differences are observedin presence of salts. These experiments are essential for understandingand quantifying measurements in biological environments. Here we alsodemonstrate diamond magnetometry measurements of free radicals whichare generated in situ in a chemical reaction for the first time. Comparedto measurements with the conventional probe HPF, our T1 measurementsprovide real-time data rather than accumulating over the entire incubationperiod. While HPF required 14 hours to reveal the concentration of *OH,we obtain a signal from an equivalent sample within a few minutes. More-over, our experiments have proven that, unlike chemical probes, the NVcenter can detect decreases and increases in the concentration of hydroxylradical. In addition, we found that surprisingly smaller nanodiamonds showhigher coherence times for particles containing dense ensembles.

67


4.5 Supporting Information

4.5.1 Measuring free radicals in-situ

Photolysis of H2O2

Figure 4.7: Spin-lattice re-laxation time of a 70 nm par-ticle immersed in water. TheT1 constant for the particle inabsence of UV light was 976.5µs, while the value of T1 afterturning the UV light on was1055 µs. The relaxation con-stant increases 8% after turn-ing the UV lamp on. Thisshifting is considered negligi-ble and part of the experimen-tal error.

4.5.2 Measuring the concentration of hydroxyl radical byHTA

Figure 4.8: The formation of HTA by Fenton reaction of Fe(ClO4)2 (10 µM)with H2O2 (1000 µM) and Na2TH (200 µM) over 20 min.

68


Figure 4.9: The spectra measured after the formation of HTA by photolysis ofH2O2 (9.7 M) with Na2TH (200 µM) using UV lamp (λ=275 nm) for 20 min.

Figure 4.10: Calibration of HTA with intensities plotted against concentrations.

69


4.5.3 Size distribution of the nanodiamonds

Figure 4.11: The size dis-tribution of the nanodiamondsused in the experiments. Insection 4.3, the reported size ofthe nanodiamonds is the nom-inal value given by the man-ufacturer (Adamas Nanotech-nologies). This figure showsthe distribution of sizes, ofa sample of 70 nm nanodia-monds, measured by dynamiclight scattering (DLS). Themode of the particle’s size is68.06 nm and the standard de-viation is 21.89 nm.

70

REFERENCES

References

[1] L. Rondin, J. P. Tetienne, T. Hingant, J. F. Roch, P. Maletinsky,and V. Jacques, “Magnetometry with nitrogen-vacancy defects in di-amond,” Reports Prog. Phys., vol. 77, no. 5, 2014.


[3] D. R. Glenn, R. R. Fu, P. Kehayias, D. Le Sage, E. A. Lima, B. P.Weiss, and R. L. Walsworth, “Micrometer-scale magnetic imaging ofgeological samples using a quantum diamond microscope,” Geochem-istry, Geophys. Geosystems, vol. 18, pp. 3254–3267, aug 2017.

[4] F. M. Sturner, A. Brenneis, J. Kassel, U. Wostradowski, R. Rolver,T. Fuchs, K. Nakamura, H. Sumiya, S. Onoda, J. Isoya, and F. Jelezko,“Compact integrated magnetometer based on nitrogen-vacancy centresin diamond,” Diam. Relat. Mater., vol. 93, pp. 59–65, mar 2019.

[5] M. Chipaux, L. Toraille, C. Larat, L. Morvan, S. Pezzagna, J. Meijer,and T. Debuisschert, “Wide bandwidth instantaneous radio frequencyspectrum analyzer based on nitrogen vacancy centers in diamond,”Appl. Phys. Lett., vol. 107, dec 2015.

[6] A. Nowodzinski, M. Chipaux, L. Toraille, V. Jacques, J. F. Roch, andT. Debuisschert, “Nitrogen-Vacancy centers in diamond for currentimaging at the redistributive layer level of Integrated Circuits,” Mi-croelectron. Reliab., vol. 55, pp. 1549–1553, aug 2015.

[7] S. R. Hemelaar, P. De Boer, M. Chipaux, W. Zuidema, T. Hamoh,F. Perona Martinez, A. Nagl, J. P. Hoogenboom, B. N. Giepmans, andR. Schirhagl, “Nanodiamonds as multi-purpose labels for microscopy,”Sci. Rep., vol. 7, no. 1, pp. 1–9, 2017.

[8] S. Nagarajan, C. Pioche-Durieu, L. H. Tizei, C. Y. Fang, J. R.Bertrand, E. Le Cam, H. C. Chang, F. Treussart, and M. Kociak, “Si-multaneous cathodoluminescence and electron microscopy cytometryof cellular vesicles labeled with fluorescent nanodiamonds,” Nanoscale,vol. 8, pp. 11588–11594, jun 2016.

[9] M. Chipaux, K. van der Laan, S. R. Hemelaar, M. Hasani, T. Zheng,and R. Schirhagl, “Nanodiamonds and Their Applications in Cells,”jun 2018.

71


[10] K. van der Laan, M. Hasani, T. Zheng, and R. Schirhagl, “Nanodia-monds for In Vivo Applications,” may 2018.

[11] M. S. Grinolds, S. Hong, P. Maletinsky, L. Luan, M. Lukin, R. L.Walsworth, and A. Yacoby, “Nanoscale magnetic imaging of a singleelectron spin under ambient conditions,” Nat. Phys., vol. 9, no. 4,pp. 215–219, 2013.

[12] J. Zopes, K. S. Cujia, K. Sasaki, J. M. Boss, K. M. Itoh, and C. L.Degen, “Three-dimensional localization spectroscopy of individual nu-clear spins with sub-Angstrom resolution,” Nat. Commun., vol. 9, dec2018.

[13] H. J. Mamin, M. Kim, M. H. Sherwood, C. T. Rettner, K. Ohno,D. D. Awschalom, and D. Rugar, “Nanoscale nuclear magnetic reso-nance with a nitrogen-vacancy spin sensor,” Science (80-. )., vol. 339,pp. 557–560, feb 2013.

[14] L. Rondin, J. P. Tetienne, S. Rohart, A. Thiaville, T. Hingant, P. Spini-celli, J. F. Roch, and V. Jacques, “Stray-field imaging of magneticvortices with a single diamond spin,” Nat. Commun., vol. 4, 2013.

[15] M. Pelliccione, A. Jenkins, P. Ovartchaiyapong, C. Reetz, E. Em-manouilidou, N. Ni, and A. C. Bleszynski Jayich, “Scanned probeimaging of nanoscale magnetism at cryogenic temperatures witha single-spin quantum sensor,” Nat. Nanotechnol., vol. 11, no. 8,pp. 700–705, 2016.

[16] L. Rondin, J. P. Tetienne, P. Spinicelli, C. Dal Savio, K. Karrai,G. Dantelle, A. Thiaville, S. Rohart, J. F. Roch, and V. Jacques,“Nanoscale magnetic field mapping with a single spin scanning probemagnetometer,” Appl. Phys. Lett., vol. 100, no. 15, 2012.

[17] J. P. Tetienne, A. Lombard, D. A. Simpson, C. Ritchie, J. Lu, P. Mul-vaney, and L. C. Hollenberg, “Scanning Nanospin Ensemble Micro-scope for Nanoscale Magnetic and Thermal Imaging,” Nano Lett.,vol. 16, no. 1, pp. 326–333, 2016.

[18] I. Lovchinsky, A. O. Sushkov, E. Urbach, N. P. De Leon, S. Choi, K. DeGreve, R. Evans, R. Gertner, E. Bersin, C. Muller, L. P. McGuin-ness, F. Jelezko, R. L. Walsworth, H. Park, and M. Lukin, “Nuclearmagnetic resonance detection and spectroscopy of single proteins usingquantum logic,” Science (80-. )., vol. 351, no. 6275, pp. 836–841, 2016.

72

REFERENCES

[19] S. J. Devience, L. M. Pham, I. Lovchinsky, A. O. Sushkov, N. Bar-Gill,C. Belthangady, F. Casola, M. Corbett, H. Zhang, M. Lukin, H. Park,A. Yacoby, and R. L. Walsworth, “Nanoscale NMR spectroscopy andimaging of multiple nuclear species,” Nat. Nanotechnol., vol. 10, no. 2,pp. 129–134, 2015.

[20] S. Steinert, F. Ziem, L. T. Hall, A. Zappe, M. Schweikert, N. Gotz,A. Aird, G. Balasubramanian, L. C. Hollenberg, and J. Wrachtrup,“Magnetic spin imaging under ambient conditions with sub-cellularresolution,” Nat. Commun., vol. 4, 2013.

[21] D. A. Simpson, R. G. Ryan, L. T. Hall, E. Panchenko, S. C. Drew,S. Petrou, P. S. Donnelly, P. Mulvaney, and L. C. Hollenberg, “Elec-tron paramagnetic resonance microscopy using spins in diamond underambient conditions,” Nat. Commun., vol. 8, dec 2017.

[22] M. Mrozek, D. Rudnicki, P. Kehayias, A. Jarmola, D. Budker, andW. Gawlik, “Longitudinal spin relaxation in nitrogen-vacancy ensem-bles in diamond,” EPJ Quantum Technol., vol. 2, no. 1, 2015.

[23] A. Ariyaratne, D. Bluvstein, B. A. Myers, and A. C. Jayich, “Nano-scale electrical conductivity imaging using a nitrogen-vacancy centerin diamond,” Nat. Commun., vol. 9, dec 2018.

[24] A. O. Sushkov, N. Chisholm, I. Lovchinsky, M. Kubo, P. K. Lo,S. D. Bennett, D. Hunger, A. Akimov, R. L. Walsworth, H. Park,and M. Lukin, “All-optical sensing of a single-molecule electron spin,”Nano Lett., vol. 14, no. 11, pp. 6443–6448, 2014.

[25] S. Kaufmann, D. A. Simpson, L. T. Hall, V. Perunicic, P. Senn,S. Steinert, L. P. McGuinness, B. C. Johnson, T. Ohshima, F. Caruso,J. Wrachtrup, R. E. Scholten, P. Mulvaney, and L. C. Hollenberg, “De-tection of atomic spin labels in a lipid bilayer using a single-spin nan-odiamond probe,” Proc. Natl. Acad. Sci., vol. 110, no. 27, pp. 10894–10898, 2013.

[26] W. Droge, “Free radicals in the physiological control of cell function,”Physiol. Rev., vol. 82, no. 1, pp. 47–95, 2002.

[27] M. Valko, D. Leibfritz, J. Moncol, M. T. Cronin, M. Mazur, andJ. Telser, “Free radicals and antioxidants in normal physiological func-tions and human disease,” Int. J. Biochem. Cell Biol., vol. 39, no. 1,pp. 44–84, 2007.

73


[28] W.-J. Huang, X. Zhang, and W.-W. Chen, “Role of oxidative stress inAlzheimer’s disease (Review),” Biomed. Reports, vol. 4, no. 5, pp. 519–522, 2016.

[29] X. Chen, X. Tian, I. Shin, and J. Yoon, “Fluorescent and luminescentprobes for detection of reactive oxygen and nitrogen species.,” Chem.Soc. rev., vol. 40, no. 9, pp. 4783–804, 2011.

[30] J. P. Tetienne, T. Hingant, L. Rondin, A. Cavailles, L. Mayer, G. Dan-telle, T. Gacoin, J. Wrachtrup, J. F. Roch, and V. Jacques, “Spin re-laxometry of single nitrogen-vacancy defects in diamond nanocrystalsfor magnetic noise sensing,” Phys. Rev. B - Condens. Matter Mater.Phys., vol. 87, no. 23, pp. 1–12, 2013.

[31] J. A. Rioux, I. R. Levesque, and B. K. Rutt, “Biexponential longitu-dinal relaxation in white matter: Characterization and impact on T1mapping with IR-FSE and MP2RAGE,” Magn. Reson. Med., vol. 75,no. 6, pp. 2265–2277, 2016.


[33] V. Merz, J. Lenhart, Y. Vonhausen, M. E. Ortiz-Soto, J. Seibel,and A. Krueger, “Zwitterion-Functionalized Detonation Nanodiamondwith Superior Protein Repulsion and Colloidal Stability in Physiolog-ical Media,” Small, vol. 1901551, 2019.

[34] W. Freinbichler, M. A. Colivicchi, M. Fattori, C. Ballini, K. F. Tipton,W. Linert, and L. Della Corte, “Validation of a robust and sensitivemethod for detecting hydroxyl radical formation together with evokedneurotransmitter release in brain microdialysis,” J. Neurochem., 2008.

[35] M. Valko, C. J. Rhodes, J. Moncol, M. Izakovic, and M. Mazur, “Freeradicals, metals and antioxidants in oxidative stress-induced cancer,”Chem. Biol. Interact., vol. 160, no. 1, pp. 1–40, 2006.

[36] M. Kruszewski, “Labile iron pool: The main determinant of cellularresponse to oxidative stress,” Mutat. Res. - Fundam. Mol. Mech. Mu-tagen., vol. 531, no. 1-2, pp. 81–92, 2003.

74

5Discussion

Progressively, the research on applying the Nitrogen-Vacancy center (NVcenter) in nanodiamonds to biomedical challenges has unveiled a long seriesof opportunities to exploit. In parallel of the main focus of this project,applying diamond magnetometry to sense oxidative stress, this research hasyielded significant knowledge about the interaction of the nanodiamondswith the physiological medium and the cell. Chapter 3, is a clear exampleof this.

In commercially available high-pressure high-temperature (HPHT) nan-odiamonds, the diamond’s surface is covered by oxygen-containing groups[1].In chapter 3 we investigate which proteins available in human serum aremore likely to attach to the diamond’s surface. As a result, we found thatcertain low abundance proteins are more likely to bind to the nanodia-monds than the more plentiful ones. Although the mechanism of affinityis not completely understood, this fact directly suggests the use of nanodi-amonds to deplete high abundance proteins from a biological sample, ormore precisely, in this case, to seize low abundance proteins. Moreover, wealso have determined that part of the recovered proteins play an importantrole as biomarkers for several diseases. We expect that the discovery of thisaffinity will promote a deeper study of the mechanism to determine moreprecisely which proteins are willing to be recuperated with nanodiamonds.Meanwhile, the current performance of depleting using nanodiamonds canbe compared with other nonspecific depletion techniques such as combina-torial peptide ligand libraries (CPLLs).

A closer investigation about the interaction of nanodiamonds with thecell and its environment is presented in chapter 2. Sensing the cellularoxidative stress from the interior of cells critically depends on ingestion ofparticles by the cells. From previous experiments, we learned that severalfactors are involved in the number of particles that are taken up effectively.The particle size, the chemical properties of the particle’s surface and the

75

DISCUSSION

cell type are relevant factors which determine the degree of engulfmentof nanodiamonds[2]. In that chapter, we present a successful method toimprove the particle uptake of a colon cancer cell line (HT29). The im-provement in uptake was addressed by controlling two variables, effectiveparticle size and surface chemistry of the nanodiamonds. By using the re-combinant proteins C4 and C4-K12 we have gained in colloidal stability ofthe nanoparticles, resulting in the reduction of the flocculation of the par-ticles. At the same time, and partially as a consequence of the reductionof the effective particle size, the number of cells having taken up nanodia-monds increased, particularly in the case of the HT29 cells. Initially, afterexposing the HT29 to nanodiamonds without any special treatment, only30% of the cells were found with at least one particle in the cytoplasm.The situation changed drastically after coating the particles with the C4-K12 protein. In that case, all the assessed HT29 cells contain at least oneparticle on the inside. As mention before, this finding can be explained bytwo factors. First, the reduction of the particle size facilitates the engulf-ment of the nanodiamonds by the cells. Second, the hydrophilic characterof the polypeptide brush layer enhances the interaction with the hydrophilicexternal surface of the cell membrane, promoting the internalization.

It is known that a similar effect can be observed by producing the coronafrom the proteins available in the cellular growth medium[3]. Making useof the engineered proteins C4 and C4-K12 also enable the potential addi-tion of supplementary motifs which serve for modulating the interactionsbetween the nanodiamonds and the cell. Additionally, taking into accountthe main aim of this work, this research has shown that the sensing prop-erties of the NV centers remain practically unaltered after coating the hostnanodiamonds with a C4 or C4-K12 protein corona. This point is par-ticularly relevant because it is known that the NV centers bleach by theaction of near charges (due to the conversion from the NV− to the NV0

states)[4], which is the case when coating with the C4-K12 protein. Thefact that the protein corona is not banishing the NV center’s photolumi-nescent signal might be explained as a consequence of using ensembles ofNV centers. The total signal produced by several hundred of, randomlylocalised, NV centers is more stable to perturbations than one producedby a small amount of defects, in which the weight of each one in the totalsignal is higher.

Finally I would like to recall the main aim of this PhD project, “researchthe implementation of Diamond Magnetometry to detect free radical insideliving cells”. Here we focus on building a measuring technique for thedetection of free radicals in a biological environment. Because of its sim-

76

DISCUSSION

plicity, we have implemented the measuring of the spin-lattice relaxationtime (T1) of an ensemble of NV centers. The T1 experiment allows sensingthe strength of random magnetic noise[5] employing a simple laser pulsesequence, avoiding the use of additional microwaves and magnets whichmight disturb the biological samples and complicate experiments.

Chapter 4 presents the first results of using the T1 relaxation experi-ments for detecting a relevant biological free radical, the hydroxyl radical.The success of actually measuring the radical while it is produced by twodifferent reactions has a consequence on validating the main hypothesiswhich originated the present study. Additionally, we demonstrate the ad-vantages and potential of using diamond magnetometry over the standardchemicals indicator. The sensitivity of diamond magnetometry allows mea-suring a small amount of the radical (1 µM) in a short period (about 20minutes) compared with the performance of a chemical indicator (hydrox-yphenyl fluorescein, HPF). Besides, the research has shown that diamondmagnetometry allows the direct measurement of increase and decrease inthe concentration of the radical. This capacity is not present in chemicalindicators and in combination with the reduced time of sensing, it mightallow the monitoring of the dynamics of a reaction at a time resolution inthe order of few minutes.

Another important output of this research is the comparison betweenthree different sizes of nanodiamonds containing ensembles of NV centers.As was discussed earlier, the size of the nanodiamonds plays a role in theparticle uptake by the cell. A smaller particle is in many cases preferredover a big one. We also found that having more NV centers in an ensemblecontributes to the stability of the photoluminescent signal. This situa-tion presents the problem of finding a nanodiamond’s size which is smallenough to facilitate uptake but big enough to contain a number of NV cen-ters which enable for robust sensing. The experiment measured the changein the T1 relaxation time when the concentration of a spin-label (Gd3+) isincreased from zero to 100 nM. The experiment showed an inverse relation-ship between the nanodiamond size and the T1 relaxation time, a biggerparticle manifests a shorter T1. This counter-intuitive result suggests thatas more NV centers are in the particle, the interaction between closer spinsturns more important than the relaxation propelled by an external sourceof magnetic noise. Having tested nanodiamonds with sizes of 40, 70 and120 nm, the biggest change in the T1 value, which represents the dynamicrange of the sensor, was expressed by the 40 nm particle (57.3%). The 70nm and the 120 nm particle were very close to each other (45.5% and 46%respectively). Due to its reduced signal intensity, the 40 nm particle also

77

DISCUSSION

showed the highest standard deviation of the measured T1 values. Takingthose two results into account, we propose the 70 nm nanodiamond offersa good compromise between sensitivity and signal intensity.

An additional assay was performed to determine the effect of the pro-teins present in the serum (which is a component of the cell growth medium)on the performance of the T1 relaxation experiment. When measuring T1

on a serum coated nanodiamond, the operation of the NV centers is af-fected. The thick layer of proteins seems to act as a shield that reduces thesensitivity of the ensemble and increases the uncertainty of the measure-ment. This is a weakness which should be studied in more details. Oneapproach that might help to reduce the effect of the serum in measuringthe T1 constant is to previously coat the nanodiamonds with the proteinC4-K12 as was presented in chapter 2. Doing this, there will be less freedomains for the serum proteins to bind, reducing the diversity of proteinsanchored in the diamond’s surface. Replacing the serum protein coronawith one that has been proven to be innocuous, should improve the sensingperformance of the NV centers immersed in cell growth media.

The next step in this investigation is to finally measure changes in theconcentration of free radicals inside a living cell. Under this condition,the complexity of the problem increases substantially, because the level ofcontrol over the environment is very limited. Although there is knowledgeof the fate of the nanodiamonds in the cell[6, 7], there are still severalquestions regarding how the nanoparticles are internalised and released inthe cytoplasm. It is also an open question how the cell reacts against thisexogenous element and where the particles are conducted after the uptake.Also, it is expected to find gradients or accumulation of radicals in certainlocations of the cell. Gaining control of the location of the nanoparticlesin the cell is an important requirement for a systematic study of the cell’sstress.

Although the core of the magnetometer is already complete, a couple ofupgrades are proposed to enable faster and more sensitive measurements.The time resolution of a T1 experiment can be improved by simplifyingthe pulse sequence used during the experiment (choosing a smaller set ofdark times to measure). This requires more knowledge of the behaviour ofthe ensemble of NV centers in the biological sample. Defining clear criteriaabout which nanodiamonds are better to use for sensing might increase theconsistency of the results. In addition, improving the post-acquisition dataanalysis might improve the signal to noise ratio of the measured signal andthus the sensitivity. Such an improvement would also shorten the acquisi-tion time. Implementing the measurement of the spin-spin relaxation time

78

DISCUSSION

(T2) and the transverse relaxation time constant (T2*) might also improvethe sensitivity of the system, at expense of increasing the complexity of theexperiment and the operation of the system. Even more complex pulsingmodalities as for instance double electron electron resonance (DEER mea-surements) would potentially offer a way to differentiate between radicals.

At this point, the research presented in this thesis has proven the viabil-ity of applying diamond magnetometry to detect free radicals in a biologicalenvironment. This is the first prove that sustain the idea of measuring thisquantity using diamond magnetometry.

79

DISCUSSION

References

[1] A. Nagl, S. R. Hemelaar, and R. Schirhagl, “Improving surface and de-fect center chemistry of fluorescent nanodiamonds for imaging purposes-a review,” Anal. Bioanal. Chem., vol. 407, no. 25, pp. 7521–7536, 2015.

[2] M. Chipaux, K. van der Laan, S. R. Hemelaar, M. Hasani, T. Zheng,and R. Schirhagl, “Nanodiamonds and Their Applications in Cells,” jun2018.


[4] M. Kaviani, P. Deak, B. Aradi, T. Frauenheim, J.-P. Chou, and A. Gali,“Proper surface termination for luminescent near-surface NV centers indiamond,” Nano Lett., vol. 14, no. 8, pp. 4772–4777, 2014.

[5] R. Schirhagl, K. Chang, M. Loretz, and C. L. Degen, “Nitrogen-VacancyCenters in Diamond: Nanoscale Sensors for Physics and Biology,”Annu. Rev. Phys. Chem., vol. 65, no. 1, pp. 83–105, 2014.

[6] S. R. Hemelaar, B. Saspaanithy, S. R. L’Hommelet, F. Perona Martinez,K. J. van der Laan, and R. Schirhagl, “The response of HeLa cells tofluorescent nanodiamond uptake,” Sensors (Switzerland), vol. 18, no. 2,2018.

[7] E. Perevedentseva, S.-F. Hong, K.-J. Huang, I.-T. Chiang, C.-Y. Lee,Y.-T. Tseng, and C.-L. Cheng, “Nanodiamond internalization in cellsand the cell uptake mechanism,” J. Nanoparticle Res., vol. 15, no. 8,p. 1834, 2013.

80

6Summary — Samenvatting

6.1 Summary

Several diseases are related to the overproduction of free radicals in theorganism. Cancer, Alzheimer’s, diabetes mellitus to mention only a few.Also, the natural and unavoidable process of ageing has been, partially,explained by the action of free radicals. This relation has been broadlyadvertised even outside the scientific community. For example, nowadaysit is common to read about the benefits of antioxidant products and anti-ageing creams. In both cases, the promoted underlying effect of thoseproducts is counteracting free radicals to keep your body young and healthy.Leaving aside the questions about the efficacy of that products, there isscientific support that relates the development of diseases[1, 2, 3, 4, 5]and the declination of the cellular function over time[6, 7] with the loss ofbalance between the generation and clearance of free radicals in our cells.Considering the high impact of free radicals in health, the importance ofhaving relevant tools available to facilitate its research is obvious.

One characteristic of the molecules classified as free radicals is the ab-sence of one electron in their last orbital. This characteristic is responsiblefor the high reactivity of the free radicals and gives a magnetic momentto the molecule. Therefore, measuring the magnetic moment in a samplemight give information about the amount of free radicals in the sample.

The equipment used to measure the magnetic field is called Magne-tometer. Several principles allow the measurement of magnetism. In thisproject we took the challenge of implementing a very sensitive magnetome-ter which is able to detect a small amount of radicals in a very small volume.Making use of nanodiamonds as a magnetic sensor allows us to measure athigh spatial and temporal resolution, at room temperature and physiolog-ical conditions.T he key element of these diamonds are specific impuritiescalled Nitrogen-Vacancy centers (NV centers).

81

SUMMARY — SAMENVATTING

In our application the NV center is a quantum sensor. The signal weread out is equivalent to conventional magnetic resonance signals. But, inour case, all the manipulation and reading of the NV center is done withlight (in contrast with the use of radio waves). Thereby the method is calledOptically Detected Magnetic Resonance (ODMR). When the NV center isexcited with a green LASER, it emits a red fluorescence. The intensitycan be processed and translated to the strength of an external magneticfield at the location of the nanodiamond. Measuring that magnetic fieldwe might be able to know the amount of radicals at that location. Thatidea sets the main hypothesis of this research. To prove it, we have built amagnetometer and conducted an investigation to get knowledge about howour sensor (the nanodiamond containing NV centers) interacts with the celland the cell’s environment.

Across the course of the investigation, we found several problems, whichneeded our attention. To be able to measure from the inside of a cellfirst it is needed to cross the cell membrane and place the sensor in thecytoplasm. We found that the nanodiamond size, the chemistry of thesurface of the particle and the cell type are key variables that set the degreeof ingestion of nanodiamonds by cells. In general terms, it is more likelyfor a small particle to be engulfed than for a big one. Also, Macrophages(a type of white blood cell) are more willing to ingest external particlesthan other cells such as HeLa or HT29 (cells from cervical and colon cancerrespectively). In our experiments (chapter 2) we were able to increase thenumber of cells ingesting nanodiamonds by wrapping the nanodiamondswith artificial proteins. Using this method, we succeed in avoiding theformation of clusters of nanodiamonds, reducing the effective particle sizeof the sensor. Moreover, the chemical properties of the proteins coveringthe nanodiamonds improved the interaction between the sensor and the cellmembrane, promoting the voluntary engulfment of the nanodiamonds. Inthe same investigation, we also have demonstrated that the magneto-opticalproperties of the NV centers are not compromised after the modification ofthe nanodiamonds’ surface.

A closer look at the interaction between the nanodiamonds and the cellgrowth medium yielded an astonishing result (Chapter 3). Knowing thefavourable affinity of the nanodiamond’s surface for certain proteins, wehave mixed uncoated nanodiamond with human blood plasma and analysedthe proteins that attached strongest to the diamonds. As a result, wefound that most of the proteins anchored to the nanodiamond’s surfacewere low abundant proteins, many of which also can be used as indicatorsfor certain diseases. This outcome presents a new biomedical application

82

SUMMARY

for nanodiamonds, the capture of low abundance proteins from a plasmasample.

After having gained experience and knowledge about the interaction ofnanodiamonds with biological samples, we focused the research to detectthe generation of free radicals in a biological environment (Chapter 4). Inthis set of experiments, we have produced the free radicals by two meth-ods, first by the photolysis of hydrogen peroxide (H2O2) and by mimickinga biological reaction named Fenton reaction. In both cases, we have usednanodiamonds containing NV centers to perform relaxometry experiments.In the first case, the results showed that using our technique we can detectthe increment of the quantity of free radicals in the sample. Moreover, afterstopping the reaction that generates the radicals we are able to detect thereduction of the concentration of radicals. This last result distinguishes ourtechnique from other chemical methods of measuring free radicals, whichare not able to directly detect the decay in the concentration of radicals.Additionally, our method proved to be faster than other chemical indica-tors. Analogously, the experiments using the Fenton reaction shown thatDiamond magnetometry can detect the generation of a free radical, butthis time the results also suggested the detection of intermediate productsof the reaction. The Fenton reaction is a more complex process, involvinghydrogen peroxide and ferrous iron, that creates additional products. Herethe origin of the signal is not completely clear yet. Although this resultbrought new questions to our investigation, it is an important input to ourknowledge as it represents the situations we will find better when measuringinside a living cell.

The research conducted in this project was able to confirm the viabilityof using Diamond Magnetometry as a tool for detecting free radicals ina biological environment. The next step is to improve it for allowing thesensing inside living cells.

83


6.2 Samenvatting

Verschillende ziektes zijn gerelateerd aan een overproductie van zuurstofrad-icalen die in een organisme plaats vindt. Kanker, Alzheimer en diabetesmellitus zijn enkele voorbeelden hiervan. Ook het natuurlijke en onafwend-bare verouderingsproces is deels toe te wijzen aan de acties van deze vrijeradicale deeltjes. Dit verband wordt ook buiten de wetenschappelijke ge-meenschap breed uitgedragen. Producten met antioxidanten of anti-verou-dering cremes baseren hun reclame op dit principe. In beide gevallen wordtgeadverteerd met de werking tegen oxidatieve radicalen, om je lichaamjong en gezond te houden. Buiten de vraag over de efficientie van deze pro-ducten, is er ook wetenschappelijk bewijs dat een verband legt tussen hetontstaan van ziektes [1, 2, 3, 4, 5] en het verval van cellulaire functie metde tijd enerzijds [6, 7] met de verstoring van de balans tussen de productieen het opruimen van vrije radicalen in onze cellen anderzijds. De groteimpact van vrije radicalen op de gezondheid toont aan hoe belangrijk hetis om de juiste onderzoeksmiddelen te hebben om het onderzoek hiernaaruit te voeren.

Een belangrijk karakteristiek van moleculen geclassificeerd als vrije rad-icalen is de afwezigheid van een elektron in de buitenste schil van een atoom.Deze eigenschap is de reden van de hoge reactiviteit van de vrije radicalenen geeft de moleculen een magnetisch moment. Daardoor kan het metenvan een magnetisch moment in een specimen informatie geven over de ho-eveelheid van vrije radicalen.

De apparatuur die wordt gebruikt voor het meten van een magnetischveld heet een magnetometer. Verschillende principes liggen ten grondslagaan het meten van magnetisme. De uitdaging van dit project was om eenhoog sensitieve magnetometer, welke een kleine hoeveelheid radicalen kanmeten in een zeer klein volume, te implementeren in een onderzoeksopzet.Met behulp van nanodiamanten welke dienen als een magnetische sensorkunne we met een zeer hoge ruimtelijke en temporale resolutie metingenverichten, bij kamertemperatuur en onder fysiologische condities. De sleutelhierbij zijn kleine imperfecties in de diamant die Nitrogen Vacancy centers(NV centers) worden genoemd.

In onze opzet dient het NV center als een kwantum sensor. Het sig-naal dat we uitlezen is equivalent aan dat van conventionele magnetis-che resonantie. Maar in ons geval, wordt zowel de manipulatie, als hetuitlezen van het signaal gedaan met licht (in tegenstelling tot radiogolven).Daarom wordt deze techniek ook wel Optically Detected Magnetic Reso-nance (ODMR) genoemd. Wanneer het NV center wordt geexciteerd meteen groene LASER, reageert het door een rode fluorescentie te verspreiden.

84

SAMENVATTING

De intensiteit van deze fluorescentie van worden vertaald naar de sterktevan het externe magnetische veld in de nabijheid van de nanodiamant. Hetmeten van dit magnetische veld kan leiden tot wetenschap van de hoeveel-heid radicalen op die locatie. Dit idee is de overkoepelende hypothese vandit onderzoek. Om het te bewijzen, hebben we een magnetometer gebouwden onderzocht hoe onze sensor (de nanodiamant met NV centers) reageertmet een cel en de cellulaire omgeving.

Gedurende het onderzoeksproces liepen we tegen verscheidene prob-lemen aan die onze aandacht vereisten. Om binnen in een cel te kun-nen meten, is het vereist om eerst het celmembraan te passeren, zodat desensor in het cytoplasma terecht kan komen. Wij vonden dat de groottevan de nanodiamant, de scheikundige samenstelling van het oppervlak vandeze deeltjes en het celtype een sleutelrol spelen in de hoeveelheid vannanodiamanten die worden opgenomen in cellen. Algemeen beschouwdis het waarschijnlijker dat een klein deeltje wordt geaccepteerd dan eengroot deeltje. Macrofagen (een type witte bloedcel) zijn sneller geneigdom externe deeltjes op te nemen dan andere cellen zoals HeLa of HT29cellen (cellen van cervixcarcinoom en coloncarcinoom, respectievelijk). Inonze experimenten (hoofdstuk 2) hebben we aangetoond dat we de opnamekonden verhogen door de nanodiamanten een schil van kunstmatige eiwit-ten te geven. Met deze methode konden we met succes het vormen vannanodiamant-clusters tegengaan, waardoor we de effectieve deeltjesgroottevan de sensor verminderden. Daarnaast leverde de chemische eigenschap-pen van de eiwitten die de nanodiamanten bedekten een verbetering op vande interactie van het celmembraan met de sensor, wat leidde tot een ver-hoogde inname van nanodiamanten. In hetzelfde onderzoek hebben we ookgedemonstreerd dat de magnetisch-optische kenmerken van de NV centersniet aangetast worden door de aanpassingen op het nanodiamant opper-vlak.

Een gedetailleerd onderzoek naar de interacties tussen nanodiamantenen het medium waar cellen in groeien leverde verrassende resultaten op(Hoofdstuk 3). In de wetenschap dat het oppervlak van de nanodiamanteneen zekere affiniteit heeft voor een bepaald soort eiwitten, hebben we deongecoate nanodiamanten met menselijk bloedplasma gemengd en geanal-yseerd welke eiwitten zicht het sterkst aan de diamanten hechtten. We von-den dat de eiwitten die geankerd waren op het nanodiamantoppervlak, juistde eiwitten waren die relatief weinig voorkomen in het bloedplasma. Velenvan die eiwitten zijn ook indicatoren voor specifieke ziektebeelden. Dezeuitkomst presenteerde een nieuwe biomedische applicatie voor nanodiaman-ten; het aantonen van de relatief zeldzamere eiwitten uit het bloedplasma.

85


Met de ervaringen en kennis opgedaan met het onderzoek naar de in-teractie van nanodiamanten met een biologisch specimen, hebben we onsgefocust op het detecteren van vrije radicalen in een biologische omgev-ing (Hoofstuk 4). In deze set experimenten hebben op twee verschillendemanieren zuurstof radicalen geproduceerd: eerst door fotolyse van water-stofperoxide (H2O2) en ten tweede door het imiteren van een biologischeproces genaamd de Fenton reactie. In beide gevallen hebben we nanodia-manten met NV centers gebruikt om relaxometrie metingen uit te voeren.In de eerste opzet lieten de resultaten zien dat onze techniek gebruikt kanworden om een stapsgewijs verhoogde hoeveelheid vrije radicalen in hetspecimen te detecteren. Daarnaast kan gedetecteerd worden dat er na hetstoppen van de fotolyse een vermindering van het aantal vrije radicalenplaatsvindt. Dit laatste resultaat onderscheid onze techniek van anderechemische methodes om vrije radicalen te meten. Deze andere methodeskunnen namelijk niet het direct de vermindering van de radicalenconcen-tratie meten. Daarnaast is onze methode sneller dan andere chemischeindicatoren. In aanvulling hierop lieten de experimenten met de Fentonreactie zien dat diamant magnetometrie het ontstaan van vrije radicalenkon detecteren, maar suggereerde het ook dat we intermediaire productenvan de reactie konden analyseren. De Fenton reactie is een complex pro-ces, waar met waterstof peroxide en geoxideerd ijzer meerdere productengegenereerd worden. De origine van het magnetische signaal is hier nogniet eenduidig. Ondanks dat dit resultaat ons nieuwe vragen bracht overons onderzoek, achtten wij deze kennis zeer belangrijk omdat het de daad-werkelijke omstandigheden in een levende cel beter reflecteert.

Dit onderzoek heeft het gebruik van diamantmagnetometrie gevalideerdals onderzoeksmiddel om vrije radicalen te detecteren in een biologischeomgeving. De volgende stap in dit onderzoek zal het verbeteren van metenbinnenin levende cellen zijn.

86

REFERENCES

References

[1] U. Das, “A radical approach to cancer,” Med. Sci. Monit., vol. 8, no. 4,pp. RA79–92, 2002.

[2] J. Emerit, M. Edeas, and F. Bricaire, “Neurodegenerative diseases andoxidative stress,” Biomed. Pharmacother., vol. 58, no. 1, pp. 39–46,2004.

[3] M. Valko, D. Leibfritz, J. Moncol, M. T. Cronin, M. Mazur, andJ. Telser, “Free radicals and antioxidants in normal physiological func-tions and human disease,” Int. J. Biochem. Cell Biol., vol. 39, no. 1,pp. 44–84, 2007.

[4] D. Dreher and a. F. Junod, “Role of oxygen free radicals in cancerdevelopment.,” Eur. J. Cancer, vol. 32A, no. 1, pp. 30–8, 1996.

[5] W.-J. Huang, X. Zhang, and W.-W. Chen, “Role of oxidative stress inAlzheimer’s disease (Review),” Biomed. Reports, vol. 4, no. 5, pp. 519–522, 2016.

[6] W. Droge, “Free radicals in the physiological control of cell function,”Physiol. Rev., vol. 82, no. 1, pp. 47–95, 2002.

[7] D. Harman, “Forum Mini Review,” Antioxid. Redox Signal., vol. 5,no. 5, pp. 557–561, 2003.

87

Acknowledgments

The last five years I have seen the Bioanalysis and Bioimaging groupgrowing quickly, from two members (Romana and a summer student

called Francois) to twenty-two researchers today. As the group grew, thediversity of people grew as well, different backgrounds, nationalities, expe-riences and personalities joined the team making it richer and every daymore interesting. Although this explosive growing, it is amazing that theworking (and non-working) atmosphere always has been open, supportiveand friendly. This is not a random situation of course, so, I want to startrecognizing the dedication and care that Romana puts in building her teamwith brilliant people who also bring a positive attitude to contribute andshare their experience with respect and appreciation for their colleague’swork. I think this is a special situation in the research environment, so Iwanted to highlight it and thanks to all of you for making every day a greatday to work.

I had the good luck of sharing the laboratory with Mayeul. I appreciatehis devotion to teach, from which I learned so much about NV-centers andphysics in general. Although it is not explicit in all the chapters presented inthis book, there is much time behind the scenes, discussions in the hallwayor in the laboratory, which contribute to bringing all these works to reality.Thanks for your time and patience.

As Mayeul was my reference for physics, Simon was for biology, sincemy precarious skills in biology found strong support in him. From pipettingto data analysis. When there were cells involved, he was my first source ofknowledge and experience. But not only while working in the lab, I alsohighly appreciated his support when I started supervising students andwhen I had to deal with the dutch bureaucracy. Of course, many thanksfor translating the summary of this thesis.

Spending so many hours a day in your working place needs to be sur-rounded by great and nice people and I have been lucky to be surrounded bythree exceptional persons, Kiran, Aryan and Linyan. Thank you for thepatience, advice, friendship and all the cookies and cakes which sweetened

89

ACKNOWLEDGMENTS

the days.

In our business, it is invaluable to have a trustful and honest opinion,it helps us to improve and to keep our feet on the ground. During thistime, I profited from the sharp and objective points of view of Alina andYori. I always enjoyed our talks, which sooner or later turned into one ofthose twisted and tasty discussions about philosophy, science fiction, theunambiguous meaning of life, and so on. These are always enlightening.

Also in those four years, I discovered how motivating and gratifying thesupervision of students is. I have great memories working in the lab withthem, and particularly with Citra. She came hungry to learn it all withdedication and persistence, now that we are colleagues I get to know thestrong and brave person she is.

I met Thamir just when I came to Groningen and we have been work-ing together since then. My profound gratitude to this wise man for hisfriendship and continuous support. I also sincerely appreciate the time withThea, and all her commitment, ideas, questioning and energy expended topush our limits to the next level. The conversations I shared with Aldonawere very inspiring, her motivation and clear vision flood the corridors andlabs.

Aside from being excellent scientists, I found very talented people inthe group. Yuchen impressed everybody with his acting skills, thank youfor all your contributions and ideas. Neda can prepare the tastiest andwonderfully decorated Iranian food, I enjoyed every conversation we had,many thanks for your kindness. Viraj plays the sitar beautifully, I appre-ciate your attitude of sharing your experience and knowledge. At dancing,Kaiqi is the best, and his positive and cheerful attitude is always welcome.Claudia can make the best designed and beautiful presentations I haveever seen in science.

For the chapter 4 I received the support of Sandeep. His contributionwas very important to make the article stronger, adding new experimentsand expertise with the chemistry involved.

Many thanks to Charles for taking the responsibility of being my co-supervisor and for quickly assessing this thesis. Also, I deeply appreciatehis support in optics and physics. Also many thanks to Ari for his input,discussions and comments about magnetometry.

In this last period so many people have joined our group, unfortunately,the time is limited, and I didn’t have the opportunity to spend more timewith my new colleagues, Yue, Rokshana and Runrun. I keep my bestimpression of you, and I can only see a successful time coming for you as aPhD student.

90

ACKNOWLEDGMENTS

I couldn’t have done the experiments correctly without the resourcesand support of the Biomedical Engineering department and its staff. Manythanks for constantly improving the laboratories and keeping them orga-nized. Also, for being a permanent support, especially when things are notgoing as expected, Ranier, Gesinda, Joop always had an answer to myquestions. Ina, Willy, Betsy, Wya many thanks for your help in deal-ing with all the administrative matters. And Ed of course, a brilliant andcreative man, who is always ready to give a hand, to read some code or torepair whatever you have spoiled.

A substantial part of my work as a PhD involved very technical andpractical aspects of microscopy, at the UMIC I found Klaas Sjollema,who was my first source of help when I had doubts or troubles in this area,I’m very grateful for all the pieces of advice I got from him.

Chapter 2 is the result of the collaboration with Renko de Vries’group from the Wageningen University & Research. It was a big pleasureto work on empirical sections together with Tingting Zheng, many thanksfor sharing your knowledge, materials and credits of the results of our work.

My profound gratitude to the reading committee Prof. dr. PetraRudolf, Prof. dr. Yijin Ren and Prof. dr. Ken Haenen for findinga space in their tight agendas to assess this work.

I have the fortune of sharing my life with an exceptional woman, Mariella,her permanent support, advice and affection have been fundamental duringall this time. Thank you for clarifying my view when it’s fuzzy and to bemy partner in the celebration.

To close this work I want to give a special mention to my parents whomade so much effort to give me the chance to develop my education andprofession and from whom I learned what hard work means.

91

About the author

Felipe Perona graduated as an Electronic Engineer at theUniversidad Tecnica Federico Santa Marıa, Chile (2007).Afterwards, he worked two years as a Development Engi-neer of Embedded Systems and four years managing tech-nological projects as PMO in the telecommunication in-dustry. In 2013, he moved to the Netherlands in orderto continue his studies at the University of Groningen,

where he graduated as M.Sc. in Biomedical Engineering with specialisa-tion of Clinical Physics in 2015. The same year Felipe was awarded witha scholarship to study a Ph.D. at the University of Groningen. In Novem-ber, he started his training in the Bioimaging and Bioanalysis group leadby Romana Schirhagl. His research was focused on applications of Dia-mond Magnetometry concerning biomedical problems, specifically, the useof nitrogen-vacancy centers in diamonds to sense the interior of living cellsand to map the distribution of reactive oxygen species in the cytoplasm.Felipe defended the PhD degree on January 2020 and continued his researchon diamond magnetometry as a post-doc associated to the Humboldt Uni-versity of Berlin and the Ferdinand-Braun-Institut in Germany.

Publications

1. F. Perona Martinez, A. Nagl, S. Guluzade, and R. Schirhagl, “Nan-odiamond for Sample Preparation in Proteomics”, Anal. Chem. 91,15, 9800-9805, 2019.

2. A. Sigaeva, F. Perona Martinez, A. C. Nusantara, N. Mougios,and R. Schirhagl. “Paving the way for intracellular detection of freeradicals with nanodiamond magnetometry”, Free Radical Biology andMedicine, 139(S1), 2019

3. A. Morita, F. Perona Martinez, M. Chipaux, N. Jamot, S. Heme-laar, K. van der Laan, and R. Schirhagl. Cell Uptake of Lipid-Coated

93

ABOUT THE AUTHOR

Diamond. Part. Part. Syst. Charact. 1900116, 2019.

4. S. R. Hemelaar, B. Saspaanithy, S. R. M. L’Hommelet, F. PeronaMartinez, K. J. van der Laan, and R. Schirhagl, “The response ofHeLa cells to fluorescent nanodiamond uptake”, Sensors (Switzer-land), vol. 18, no. 2, 2018.

5. K. J. Van Der Laan, J. Naulleau, V. G. Damle, A. Sigaeva, N. Jamot,F. Perona Martinez, M. Chipaux, and R. Schirhagl, “Toward UsingFluorescent Nanodiamonds to Study Chronological Aging in Saccha-romyces cerevisiae”, Anal. Chem., vol. 90, no. 22, pp. 13506-13513,2018.

6. F. Perona Martinez⊥, T. Zheng⊥, I. M. Storm, W. Rombouts, J.Sprakel, R. Schirhagl, and R. de Vries, “Recombinant protein poly-mers for colloidal stabilization and improvement of cellular uptake ofdiamond nanosensors”, Anal. Chem., vol. 89, pp. 12812-12820, 2017.

7. M. W. H. Garming, I. G. C. Weppelman, P. de Boer, F. PeronaMartinez, R. Schirhagl, J. P. Hoogenboom, and R. J. Moerland,“Nanoparticle discrimination based on wavelength and lifetime-multiplexedcathodoluminescence microscopy”, Nanoscale, vol. 9, no. 34, pp.12727-12734, 2017.

8. S. R. Hemelaar, K. J. van der Laan, S. R. Hinterding, M. V. Koot,E. Ellermann, F. Perona Martinez, D. Roig, S. Hommelet, D. No-varina, H. Takahashi, M. Chang, and R. Schirhagl, “Generally Appli-cable Transformation Protocols for Fluorescent Nanodiamond Inter-nalization into Cells”, Sci. Rep., vol. 7, 2017.

9. S. R. Hemelaar, P. De Boer, M. Chipaux, W. Zuidema, T. Hamoh,F. Perona Martinez, A. Nagl, J. P. Hoogenboom, B. N. G. Giep-mans, and R. Schirhagl, “Nanodiamonds as multi-purpose labels formicroscopy”, Sci. Rep., vol. 7, no. 1, pp. 1-9, 2017.

Patents

• Schirhagl, R., Perona Martinez, F. P., & Chipaux, M. S. (2017).“Method and detector for microscopic measurement by means of acolour center” (Patent No. WO2017217847.)

94

Diamond Magnetometry for sensing in biological environment

Documents

Transcript of Diamond Magnetometry for sensing in biological environment