The Bone Marrow Microenvironment is Essential for the Function of
Connective Tissue Growth Factor (CTGF/CCN2) in Childhood
Acute Lymphoblastic Leukaemia
Julia Wells, B.A, B.Sc (Hons)
This thesis is presented for the degree of Doctor of PhilosophyThe University of Western Australia
School of Paediatrics and Child HealthTelethon Institute for Child Health Research
Faculty of Medicine and Dentistry-2014-
i
Abstract
Acute lymphoblastic leukaemia (ALL) is the most commonly diagnosed cancer
in children. Successive chemotherapy treatment protocols have been
successful in increasing overall survival rates to 90%. However, further
improvement is achievable only through a better understanding of the
pathogenesis of ALL and the mechanisms of drug resistance. Connective tissue
growth factor (CTGF/CCN2) has previously been reported as the most
deregulated gene in childhood B-cell precursor (BCP) ALL, existing in
approximately 75% of cases (Boag et al., 2007) and has been associated with a
poor prognosis in both children and adults (Kang et al., 2010, Sala-Torra et al.,
2005). A functional role for CTGF has yet to be determined in ALL, though in
other cancers it’s deregulation contributes positively to tumour progression and
survival.
In this thesis the functional role of CTGF in childhood BCP ALL was explored in
three ways. Firstly, two BCP ALL cell lines (PER-278 and PER-371, established
from primary patient specimens) where engineered to stably express human
CTGF and a GFP reporter coding sequence. Overexpression of CTGF in PER-
278 and PER-371 was confirmed by qRT-PCR (2.6 and 3.4 fold, respectively)
and by western blot, for both cellular protein and conditioned media. In culture
conditions, high levels of CTGF did not alter proliferation, drug sensitivity (to
cytarabine and vincristine) or the gene expression profile of PER-278 or PER-
371. A marked difference was seen between the gene expression profiles of the
two BCP ALL cell lines, resulting in different genes being co-expressed with
CTGF and foreshadowing the cell line-specific interactions seen in the mouse
bone marrow microenvironment.
Secondly, in culture conditions recombinant human CTGF was found to
significantly alter gene expression in a human bone marrow stromal cell line
(HS5) after 8, 16 and 48 hours in culture. Microarray analyses revealed the
greatest number of genes were deregulated at 8 hours (787 upregulated and
622 downregulated) determined by LIMMA significance and a greater than 2-
fold change compared with controls. A large number were associated with
ii
secreted proteins functioning in extracellular matrix, lipid metabolism, adhesion
and cytokine signalling.
Lastly, xenograft models of leukaemia were established using non-obese
diabetic/ severe combined immunodeficiency (NOD/SCID) mice injected via the
tail-vein with either 2x105 PER-278 or 1x106 PER-371 CTGF-High and Low
cells. Within the murine microenvironment, a divergent response was observed
due to CTGF overexpression in these cell lines, attributable to their distinct
gene expression profiles. PER-371 xenografts showed reduced survival in
CTGF High compared with Low groups (median survival 62 versus 71 days
post-injection, p=0.01). This difference was attributed to increased engraftment
of PER-371 High cells in the bone marrow, and associated with an increased
deposition of reticulin fibres in the surrounding extracellular matrix (Ag+ stain for
reticulin). No difference in survival was observed between PER-278 High and
PER-278 Low xenografted mice, with both groups having a median survival of
49 days post-injection. However, increased resistance to low dose cytarabine, a
chemotherapeutic drug, occurred exclusively in PER-278 xenografted mice due
to CTGF overexpression, though a clear mechanism of action was not
distinguished. Both cell lines responded to high dose cytarabine in vivo, with
differing levels of drug toxicity. This indicates that CTGF can contribute to
cytarabine resistance in some subtypes of BCP ALL, but the drug overcomes
these effects at increased doses, that are more clinically relevant.
The results presented in this thesis suggest a complex role for CTGF in BCP
ALL, with functions being sub-type and context specific. New understanding of
the role of CTGF in the disease emphasised the requirement for interaction with
stromal elements from the bone marrow to accelerate tumour growth or
contribute to drug resistance. This thesis outlines an extracellular matrix-
dependent molecular mechanism by which CTGF confers a growth advantage
to BCP ALL cells. Contributing to the list of known genotype changes that lead
to phenotypic differences in BCP ALL cells, with the goal of providing new
treatment strategies for children with the disease.
iii
Table of Contents Abstract i Table of Contents iii Figure List vi Table List ix
Abbreviations x
Acknowledgments xii Publications arising xiii Presentations related to this thesis xiv
Statement of Candidature Contribution xvii
Chapter 1 Introduction 1
1.1 ⏐ Acute Lymphoblastic leukaemia 3
1.2 ⏐ The bone marrow microenvironment 10
1.3 ⏐ Connective tissue growth factor (CTGF) 20
1.4 ⏐ Aims and outline of the thesis 37
Chapter 2 Materials and Methods 39
2.1⏐ Cell lines and tissue culture 41
2.1.1 Suspension cell lines
2.1.2 Adherent cell lines
2.2⏐ Cell transfections 42
2.2.1.1 Plasmids used in cell transfection
2.2.2 Standard electroporation of B-cell precursor ALL cells
2.2.3 Neon electroporation of B-cell precursor ALL cells
2.2.4 Selection of stable cells transfected with pcDNA3.1 vector
2.2.5 Lentiviral production
2.2.6 Lentiviral infection
2.2.7 selection of stable cells onfected with pCDH vector
2.3⏐ RNA analyses 49
2.3.1 RNA extraction
2.3.2 cDNA synthesis
2.3.3 Quantitation of CTGF gene expression
2.3.4 Global gene expression to determine effect of CTGF on stable CTGF-
iv
overexpressing cell lines
2.3.5 Global gene expression to determine effect of recombinant human CTGF
on stromal cells
2.4⏐ Protein Analyses 53
2.4.1 Cellular protein extraction
2.4.2 Secreted protein purification
2.4.3 Western blot
2.4.4 Enzyme linked immunosorbent assay
2.4.5 Protein array
2.5⏐ Proliferation assays 58
2.6⏐ Drug sensitivity assays 59
2.6.1 Optimisation of antibiotics for stable cell line selection
2.6.2 Sensitivity to chemotherapy drugs in CTGF-High and CTGF-Low cell lines
2.6.3 Nucleoside transporter activity in PER-278 High or Low cells
2.7⏐ NOD/SCID mouse xenografts 62
2.7.1 Mice
2.7.2 Generation of NOD/SCID mice
2.7.3 Chemotherapy regimen
2.7.4 Mouse dissection and tissue preparation
2.7.5 Tissue analysis by flow cytometry
2.7.6 Histochemical and immunohistochemical (IHC) staining
2.8⏐ Statistical analyses 72
Chapter 3 Stable overexpression of CTGF in B-cell precursor ALL cell lines
does not alter function 73
3.1⏐ Abstract 75
3.2⏐ Introduction 76
3.3⏐ Results 76
3.4⏐ Discussion 94
Chapter 4 Recombinant human CTGF has the capacity to alter the gene
expression profile of human BM stromal cells 99
4.1⏐ Abstract 101
4.2⏐ Introduction 101
4.3⏐ Results 102
4.4⏐ Discussion 114
v
Chapter 5 CTGF contributes to drug sensitivity in PER-278
B-cell precursor ALL
5.1⏐ Abstract
5.2⏐ Introduction
5.3⏐ Results
5.4⏐ Discussion
Chapter 6 CTGF confers a growth advantage to PER-371 B-cell
precursor ALL cells in the BM microenvironment
6.1⏐ Abstract
6.2⏐ Introduction
6.3⏐ Results
6.4⏐ Discussion
Chapter 7 General Discussion
7.1⏐ Introduction
7.2⏐ Limitations of the current research
7.3⏐ Critical findings
7.4⏐ Proposed mechanism of CTGF action
7.5⏐ Future directions
7.4⏐ Final conclusions
Chapter 8 References Appendix: Supplementary Figures
117 119 120 121 134
137 139 140 141 174 179
181
183
184
189
189 191 193 217
vi
Figure List No. Figure Title Page Number
1.1 1.2 1.3 1.4 2.1 2.2 2.3 2.4 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 4.1 4.2
B-precursor acute lymphoblastic leukaemia (ALL) originates in the bone marrow . Mesenchymal stem cells (MSC) have the potential to differentiate into divergent cell lineages Human long bone as a model for the bone marrow niche Conserved structure of the CCN family members Leukaemia engraftment was determined by GFP+ cell detection using flow cytometry Immunohistochemical staining was scored by the Nuance FX multispectral imaging system and InForm software (Perkin Elmer) Reticulin stain scoring Sample size required per treatment group (x axis) to detect a difference in population means (y axis) Transfection of B-precursor ALL cell lines by electroporation was not successful in producing stable clones Transduction efficiency of B-precursor ALL cell lines by lentiviral infection Verification of stable overexpression of CTGF mRNA in B-precursor ALL cell lines PER-278 and PER-371 Overexpression of CTGF cellular and secreted protein in B-precursor ALL cell lines PER-278 and PER-371 PER-278 and PER-371 CTGF overexpression did not effect proliferation PER-278 and PER-371 representative drug sensitivity curves for vincristine (VCR) and cytarabine (ARAC) PER-278 and PER-371 CTGF overexpression did not affect sensitivity to chemotherapeutic drugs RNA integrity analysis for PER-278 and PER-371 cell line sample CTGF RNA and protein was overexpressed in the PER-278 and PER-371 cell line samples collected for microarray analysis PER-278 and PER-371 CTGF High cells did not affect global gene expression patterns compared to CTGF Low cells Changes to global gene expression in BM stromal cells (HS5) caused by treatment with rhCTGF compared to control buffer CTGF deregulates HS5 genes involved in the IGF1 signalling pathway
5 13 15 23 67 69 71 71 79 81 83 85 87 89 90 92 93 95 103 107
vii
No. Figure Title Page Number
4.3 4.4 4.5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9
CTGF deregulates HS5 genes involved in the PI3K/AKT signalling pathway CTGF deregulates HS5 genes involved in TGFβ networks. Validation of two ECM-associated genes upregulated in HS5 cells after treatment with CTGF CTGF protein expression levels for PER-278 High and Low NOD/SCID xenografts PER-278 NOD/SCID xenografts showed no difference in survival or BM engraftment between High and Low injected cell lines Rapid engraftment of leukaemia at end stage disease for PER-278 NOD/SCID xenografts Kaplan Meier survival curve of PER-278 High and Low engrafted NOD/SCID xenografts treated with single agent chemotherapy Snapshot of proliferation in PER-278 NOD/SCID xenografts 24 hours post ARAC dose Snapshot of apoptosis in PER-278 NOD/SCID xenografts 24 hours post ARAC injection Reduced drug sensitivity to ARAC by PER-278 High cells was not replicated in presence of secreted proteins from stromal-leukaemia cell interactions CTGF secreted protein expression levels for PER-371 High and Low cells intended for NOD/SCID inoculation Kaplan-Meier survival curve of PER-371 High and Low engrafted NOD/SCID xenografts treated with chemotherapy drugs or saline PER-371 NOD/SCID xenografts showed increased BM engraftment in CTGF High compared to Low cells at late stage disease Rapid engraftment of leukaemia cells at end stage disease for PER-371 NOD/SCID xenografts Snapshot of proliferation and apoptosis in PER-371 cells at late stage disease Leukemic-burden in diaphysis of PER-371 engrafted bone marrow Leukemic-burden in epiphysis/ metaphysisof PER-371 engrafted bone marrow Proliferating cells in diaphysis of PER-371 leukaemia-burdened bone marrow Proliferating cells in depiphysis/ metaphysis of PER-371 leukaemia-burdened bone marrow
108 109 113 122 124 125 127 129 131 133 142 144 146 147 149 153 155 157 159
viii
No. Figure Title Page Number
6.10 6.11 6.12 6.13 6.14
Megakaryocyte count in diaphysis of PER-371 leukaemia-burdened bone marrow Reticulin connective tissue density in diaphysis of PER-371 leukaemia-burdened bone marrow Reticulin connective tissue density in epiphysis/ metaphysis Invading Leukaemia front depicts immunostaining associations significant by Pearson’s correlation coefficient of PER-371 leukaemia-burdened bone marrow Decreased survival of PER-371 High xenografts was not replicated in vitro
161 164 165 167 169
ix
Table List No. Table Title Page Number
1.1 1.2 2.1 3.1 3.2 3.3 3.4 4.1 4.2 5.1 6.1 6.2 6.3
A topography of CTGF-associated neoplasms in humans Pattern of CTGF deregulation compared to normal tissue and its association with clinical outcome and cellular mechanism of action List of primary, secondary and tertiary antibodies Patient details for derived cell lines PER-278 and PER-371 Patient-derived B-precursor ALL cell lines maintain patient DNA profile RNA integrity number for PER-278 and PER-371 cell line samples Fold change of CTGF High or Intermediate compared to Low for RNA and cellular protein samples Ingenuity Pathway Analysis (IPA) for differentially regulated genes between stromal cells (HS5) treated with rhCTGF (100ng/ml) or control buffer for 8 hours Top functions of secreted HS5 genes differentially regulated after treatment with rhCTGF (100ng/ml) or control buffer for 8 hours. Fold change difference between PER-278 High and Low cells before tail-vein injection into NOD/SCID mice Fold change difference between PER-371 High and Low cells before injection of 1x106 cells into the tail-vein of NOD/SCID mice Leukaemia cells alter secreted proteins from HS5 stromal cells after co-culture CTGF alters secreted proteins from stromal-leukaemia cell co-cultures
25 29 55 77 77 92 93 105 111 122 142 171 173
x
Abbreviations ABI3BP - ABI gene family member 3 binding protein ADAMTS1 - A disintegrin and metalloproteinase with thrombospondin motifs 1 ALL - Acute lymphoid leukaemia ARAC - Cytarabine ASNS - Asparagine synthetase ATCC - American Type Culture Collection BCA - Bicinchoninic acid BCR - B-cell receptor BM - Bone marrow BSA - Bovine serum albumin CAR - CXCL12-abundant reticular CMV - Cytomegalovirus CNT - Cumulative nucleoside transporters COG - Childrenís Oncology Group COL3A1 - Collagen, type 3, alpha 1 CSF3/GCSF - Colony stimulating factor 3 CTGF - Connective tissue growth factor CXCL1 - Chemokine (C-X-C motif) ligand 1 CXCL12 - CXC-chemokine ligand 12 CXCL6 - Chemokine (C-X-C motif) ligand 6 CXCR4 - C-X-C chemokine receptor type 4 DMEM - Dulbeccoís Modified Eagleís Medium ECM - Extracellular matrix EDTA - Ethylenediaminetetraacetic acid EF1 - Elongation factor 1 EGTA - Ethylene glycol tetraacetic acid ELISA - Enzyme-linked immunosorbent assay ENT - Equilibrative nucleoside transporters ERK1/2 - Extracellular signal-regulating kinases FCS - Fetal calf serum FN - Fibronectin
FSC vs SSC - Forward and side scatter plot GFP - Green fluorescent protein HIV-1 - Human immunodeficiency virus HSC - Haematopoietic stem cell IGF - Insulin-like growth factor IHC - Immunohistochemical IL11 - Interleukin 11 IL6 - Interleukin 6 IL7 - Interleukin 7 IL8 - Interleukin 8 ILK - Integrin-linked kinase IPA - Ingenuity Pathway Analysis IRF4 - Interferon regulatory factor 4 ITGAV - Integrin, alpha V IVC - Individually ventilated cages LOX - Lysyl oxidase MHCI - Major histocompatability complex class I MLL - mixed lineage leukaemia MMN/CD10 - Membrane metallo-endopeptidase MMP1 - Matrix metalloproteinase 1 MSC - Mesenchymal stem cells MSCV - Murine stem cell virus NFKB - Nuclear factor kappa-B NMBPR - Nitrobenzylmercaptopurine ribonucleoside NOD/SCID - Non-obese diabetic/ severe-combined immunodeficient NSG - NOD scid gamma OPN - Osteopontin PCA - Principle component analysis PCR - Polymerase chain reaction PI3K/AKT - Phosphoinositide 3-kinase/protein kinase C PP2A - Protein phosphatase 2 PSMB4 - Proteasome subunit beta, type 4 PURO - Puromycin resistance gene RANKL - Receptor activator of nuclear factor-?B ligand SCF - Stem cell factor SPDI - Secreted protein database initiative SSC-A vs SSC-W – Side scatter area and width plot STR - Short tandem repeat TE - Tris-EDTA TGFβ - Transforming growth factor beta
xi
TMB - Tetramethylbenzidine TMSB10 - Thymosin beta 10 TRIM22 - Tripartite motif containing 22 U - Units VCAM1 - Vascular cell adhesion molecule 1 VCR - Vincristine XIAP - X-linked inhibitor of apoptosis protein
xii
Acknowledgements
Apparently it takes a pit crew to drive a race car and it takes a village to
raise a child... and I now know these sums mean nothing compared to what it
takes to earn a doctorate. My thanks go out to all who helped me on my way.
To my supervisors Prof. Ursula Kees, Prof. Cathy Cole, Dr Amy Samuels
and Dr Meegan Howlett, the league of extraordinary ladies, who have each at
particular moments throughout my candidature been beacons of skill, logic and
hope. Thanks for everything you amazing women!
To my scholarship funding bodies, I consider myself fortunate indeed to
be a recipient of your monetary support, the Richard Walter Gibbon Medical
Research Scholarship, Stan and Jean Perron Top-Up Scholarship, TICHR
Completion Scholarship, Children’s Leukaemia and Cancer Research
Foundation Travel Award, Friends of the Institute Travel Award, UWA Graduate
Research Centre Travel Grant and UWA School of Paediatrics and Child Health
Travel Grant.
To our collaborators Prof. David Brigstock, Dr Makiko Fujii and Prof.
Richard Lock, who shared resources and essential knowledge of the field.
To the histology department at PMH, especially Dr Adrian Charles,
pathologist to all needy scientists, and Maxine Crook, who generously saved
the day by teaching me some old school stains.
To my biostatistician and bioinformatician friends, Denise Anderson, Dr.
Kim Carter and Richard Francis, who always helped me with a smile.
To everyone in LCR and BTRP for the great friendships and lab
expertise. To Jasmin, who was my Virgil through the Inferno for the first two
years. To Ursula, Meegan, Jasmin, Jette, Heloise, Laurence, Sajzla, Patrizja,
Mat and Jo who have all been team CTGF, there is a reward waiting for us in
the afterlife for working with this promiscuous and tricky protein.
To my family and friends, even in your bafflement of the mysterious world
of academia you never wavered in your support. Putting up with my lame
excuses, cooking me meals and buying me drinks. (Except Matt, you told me
from the start that I should instead get a job, then helped me the most looking
for typos and abbreviations at 4AM.)
To Mubby, because no matter what we’re doing, so long as it’s together it
will turn out to be the good old days.
xiii
Publications arising from this thesis Wells, J.E., Cole, C.H., Kees, U.R., Deregulated expression of connective tissue growth factor (CTGF/CCN2) is linked to poor outcome in human cancer, Int J Cancer, 2015;137(3):504-11 (Review Article)
Wells, J.E., Howlett, M, Cheung, LC, Kees, UR, The role of CCN family genes in haematological malignancies, J Cell Commun Signal., 2015; E-pub DOI: 10.1007/s12079-015-0296-4 (Review Article)
Wells, JE, Howlett, M, Halse, H, Heng, J, Ford, J, Cheung, LC, Samuels, AL, Crook, M, Charles, AK, Cole, CH, Kees, UR, High expression of connective tissue growth factor (CTGF) accelerates dissemination of leukaemia (Experimental article, for submission)
xiv
Presentations arising from this thesis
Presentations (Oral) Wells, J.E. How Connective Tissue Growth Factor Mediates Communication between Pre-B Acute Lymphoblastic Leukaemia and the Bone Marrow Microenvironment, Telethon Institute for Child Health Research 4th Annual Postgraduate Student Forum, TICHR Seminar Room, Perth, Western Australia, August, 2010. Wells, J.E. Role of Microenvironment Interactions in Childhood Leukaemia, Child and Adolescent Health Research Symposium, Princess Margaret Hospital, Perth, Western Australia, October, 2010. Wells, J.E. Role of Connective Tissue Growth Factor in in vitro Models of Childhood Leukaemia, 7th State Cancer Conference, Perth Convention and Exhibition Centre, Perth, Western Australia, March, 2011. Wells, J.E., Ford, J., Samuels, A.L., Cole, C.H., Kees, U.R., Determining the role of connective tissue growth factor in childhood acute lymphoblastic leukaemia, Australian Society for Medical Research Scientific Symposium, Perth, Western Australia, June 2011. Wells, J.E., Microenvironment Interactions in Childhood Leukaemia, UWA Three Minute Thesis Semi-Finals, Perth, Western Australia, July 2011. Wells, J.E., Ford, J., Samuels, A.L., Cole, C.H., Kees, U.R., Connective tissue growth factor contributes to microenvironment interactions in childhood leukaemia, Telethon Institute for Child Health Research 5th Annual Postgraduate Student Forum, Perth, Western Australia, August 2011. Wells, J.E., Ford, J., Samuels, A.L., Cole, C.H., Kees, U.R., How connective tissue growth factor is involved in childhood leukaemia and the bone marrow microenvironment, Princess Margaret Hospital Child and Adolescent Health Research Symposium, Perth, Western Australia, October 2011. Wells, JE, Heng, J, Ford, J, Howlett, M, Brigstock, DR, Samuels, AL, Cole, CH, Kees, UR, In vivo model to determine the role of connective tissue growth factor (CTGF) in childhood leukaemia, Telethon Institute for Child Health Research 6th Annual Postgraduate Student Forum, Perth, Western Australia, August 2012. Wells, JE, Heng, J, Ford, J, Howlett, M, Brigstock, DR, Samuels, AL, Cole, CH, Kees, UR, Love thy neighbour: an in vivo model to determine the role of the bone marrow microenvironment in childhood leukaemia, Perth Cancer Club, Perth, Western Australia, September 2012. Wells, JE, Heng, J, Ford, J, Howlett, M, Brigstock, DR, Samuels, AL, Cole, CH, Kees, UR, Neighbours from hell: leukaemia cells and the bone marrow microenvironment, TICHR Child Health Research Seminar, Perth, Western Australia, September 2012.
xv
Wells, JE, Heng, J, Ford, J, Howlett, M, Brigstock, DR, Samuels, AL, Cole, CH, Kees, UR, The role of connective tissue growth factor (CTGF) in an in vivo model of childhood leukaemia, Child and Adolescent Health Research Symposium, Perth, Western Australia, October 2012. Wells, JE, Heng, J, Ford, J, Howlett, M, Brigstock, DR, Samuels, AL, Cole, CH, Kees, UR, Connective tissue growth factor (CTGF) affects microenvironment interactions in a mouse xenograft model of acute lymphoblastic leukaemia, Australian Society for Medical Research Scientific Symposium, Perth, Western Australia, June 2013. Wells, J.E., Finding the genetic differences that matter in childhood leukaemia, UWA Three Minute Thesis Semi-Finals, Perth, Western Australia, July 2013. Wells, JE, Heng, J, Ford, J, Halse, H, Samuels, AL, Howlett, M, Cole, CH, Kees, UR, Connective tissue growth factor (ctgf/ccn2) alters progression of leukaemia in a xenograft mouse model, Child and Adolescent Health Research Symposium, Perth, Western Australia, October 2013. Wells, JE, Howlett, M, Halse, HM, Ford, J, Heng, J, Cheung LC, Kees, UR, Connective tissue growth factor is increased in acute lymphoblastic leukaemia and confers a growth advantage within the bone marrow niche, 8th State Cancer Conference, Perth Convention and Exhibition Centre, Perth, Western Australia, October 2013.
Wells, JE, Howlett, M, Cheung, LC Halse, H, Heng, J, Cole, CH, Kees, UR, Targeting the microenvironment in childhood acute lymphoblastic leukaemia: Identifying key players in the spread of disease, Australian and New Zealand Children's Haematology/Oncology Group (ANZCHOG) Annual Scientific Meeting, Perth, Western Australia, June 2015.
Presentations (Poster) Wells, J.E., Ford, J., Samuels, A.L., Cole, C., Kees, U.R., Role of Microenvironment Interactions in Childhood Leukaemia, 23rd Lorne Cancer Conference, Mantra Erskine, Lorne, Victoria, February, 2011. Wells, J.E., Ford, J., Samuels, A.L., Cole, C.H., Kees, U.R., The role of connective tissue growth factor in microenvironment interactions of childhood leukaemia, Students in Health and Medical Research Conference, Perth, Western Australia, August 2011. Wells, J.E., Ford, J., Samuels, A.L., Cole, C.H., Kees, U.R., Connective tissue growth factor (CTGF) contributes to microenvironment interactions in childhood leukaemia, AACR Tumor microenvironment complexity: Emerging roles in cancer therapy, Orlando, Florida, November 2011. Wells, JE, Heng, J, Ford, J, Howlett, M, Brigstock, DR, Samuels, AL, Cole, CH, Kees, UR, In vivo model to determine the role of connective tissue growth factor (CTGF) in childhood leukaemia, Cold Spring Harbor Asia- International Cancer Microenvironment Society Joint International Conference, Suzhou, China, November 2012.
xvi
Wells, JE, Howlett, M, Halse, M, Ford, J, Heng, J, Samuels, AL, Cole, CH, Kees, UR, Connective tissue growth factor (CTGF) is increased in pre-B acute lymphoblastic leukaemia and confers a growth advantage within the bone marrow niche, AACR conference on cellular heterogeneity in the tumor microenvironment, San Diego, USA, February 2014. Wells, JE, Howlett, M, Halse, M, Ford, J, Heng, J, Samuels, AL, Cole, CH, Kees, UR, Connective tissue growth factor (CTGF) is increased in pre-B acute lymphoblastic leukaemia and confers a growth advantage within the bone marrow niche, GTC4th International Conference on Tumor Progression and Therapeutic Resistance, Boston, USA, March 2014.
xvii
Statement of Candidate Contribution
The experiments detailed in this thesis have been conducted at the Telethon
Institute for Child Health Research, Perth, Western Australia. All work presented
was performed solely by the author under the supervision of Prof Ursula Kees,
Dr Amy Samuels and Dr Meegan Howlett, except were contributions from
others have been acknowledged.
This thesis contains no material, which has previously been submitted for any
other degree or diploma.
In instances were work has been performed by other parties, permission has
been granted from these parties to include this work in the thesis.
………………….. …………………….
Julia E. Wells Prof Ursula R Kees
July 2015
111111111Introduction
There's no limit to how complicated things can get, on account of one thing always
leading to another.
E.B. White
2
3
Chapter 1 Introduction
Acute lymphoblastic leukaemia (ALL) is a significant disease in childhood. To
improve therapy and cure rates for patients, it is vital to understand the tumour
biology. This will increase our chances of finding novel selective agents that are
less toxic to normal cells and more effective than currently used cytotoxic
agents. At present, there is much we don’t know about the development and
propagation of ALL. This literature review will first cover our current knowledge
of ALL tumour biology and then expand on the role of the surrounding cells in
the bone marrow microenvironment. The third section will discuss the
significance of a deregulated protein, connective tissue growth factor (CTGF),
and the evidence supporting a possible functional role of this protein in ALL.
Finally, the aims and outline of this thesis will be addressed.
1.1 Acute lymphoblastic leukaemia
Cancer is the second most common cause of death, after accidents, in
Australian children under 15 years of age. The most prevalent form of childhood
cancer, accounting for 25.7 percent of total cases, is acute lymphoid leukaemia
(ALL) (Baade, 2010). Australian figures reflect world trends in cancer statistics
with ALL being the most prominent form of cancer in children (Jemal, 2008;
Kaatsch, et al., 2006; Pritchard-Jones, et al., 2006). Due to modern multimodal
chemotherapy regimes treatment of ALL has been a major success story with
overall event-free survival rates reaching approximately 90 percent (Inaba, et
al.; Pui, et al., 2010).
Leukaemias are malignant disorders of the haematopoietic system and are
characterised by alterations in aberrant lymphoid and myeloid progenitor cell. .
In ALL, lymphoid progenitor cells that normally differentiate into mature B, T and
natural killer cells, instead undergo unlimited self-renewal and clonal expansion.
ALL comprises several subtypes, broadly differentiated into B or T cell origin.
Leukaemogenesis involves the induction, development and progression of the
disease and is associated with changes to cellular proliferation, differentiation
4
and apoptosis (cell death). The successive transformative steps required for
malignancy are not fully understood. However, leukemia induction in many
childhood cases begins in-utero and requires other “promotional” exposures for
emergence of disease (Inaba, et al., 2013). An exception to this is infant ALL
with a rearrangement of the mixed lineage leukaemia (MLL) gene, where the
disease is established by birth (Greaves, et al., 2003). After a precursor
lymphoid cell becomes malignant, accumulation of these clonal or oligoclonal
malignant cells continues unabated. The high numbers of malignant cells
overwhelm the immunocompetent cells of haematopoietic origin and affect
normal function. Childhood leukaemia is often not clinically diagnosed until
normal physiological function is affected caused by bone marrow failure,
presenting as neutropenia, thrombocytopenia, anemia or bone pain.
B-cell ALL Normal B-cell lymphopoiesis in humans initially occurs in the fetal liver and then
shifts to the bone marrow from mid-gestation and until death of the individual.
Haematopoietic stem cells commit to a lymphoid progenitor cell fate and further
differentiate into T-cell progenitors, which migrate to the thymus to before
differentiation into mature cells, and B-cell progenitors, which differentiate
entirely within the bone marrow in close association with stromal cells.
Immature-B cells migrate into the peripheral blood to reach spleen and
lymphoid organs where they differentiate into mature-B and plasma cells.
Normal mammalian B-cell development is highly ordered, comprises many
stages, involves sequential immunoglobulin gene recombination and expression
of transcription factors specific to the B-cell lineage (Clark, et al., 2014; Reya &
Rudolf, 1998). B-cells lymphopoiesis begins with the committed lymphoid
progenitor cells and progresses sequentially from early-B, pro-B, pre-B,
immature-B, mature-B, memory-B to plasma cells (Figure 1.1).
Human precursor B-cell ALLs similarly originate in the bone marrow and are the
most common form of acute leukaemia. They have a better overall prognosis
than T-cell ALLs, but due to their high incidence account for substantial
mortality (Armstrong, 2005). Human B-cell ALLs are classified by tumour
phenotype, which involves characterisation of antigen surface markers, nuclear
tumourigenesis
Figure 1.1. B-precursor acute lymphoblastic leukaemia (ALL) originates in the bone marrow. B-precursor ALL occurs after initial genetic mutations in early B-lineage progenitor cells receive secondary ‘hits’ creating malignant cells with undeterred proliferation. B-cell leukaemia subtypes, such as B-precursor ALL, arise from similar transformations taking place along the various stages of the normal B cell maturation pathway. Haematopoietic stem cells (HSCs) become more lineage specific as they differentiate into immature B-cells in the bone marrow, which escape into peripheral blood vessels for subsequent maturation at external sites.
Early-B cell
B-precursor ALL
6
gene expression and immunoglobulin gene rearrangements. The findings from
these tests identify at which stage of B-cell lineage-commitment the tumour is
formed and any chromosomal rearrangements they contain. Assigning B-
lineage ALL subtypes by stage of lineage-commitment does not necessarily
identify the cell of origin, but it can point to features of the cancer stem cell
(defined as a tumour cell that can proliferate and differentiate into daughter cells
that recapitulate the heterogeneity of the disease). It has been found that
experimentally-modified expression of transcription factors essential to normal
B-cell development can reverse differentiation all the way back to a common
lymphoid progenitor state (Cobaleda, et al., 2007). This high plasticity in B-cell
development leaves identifying the cell of origin within the B-cell lineage difficult
to ascertain. The cell where the first genetic lesion occurred may have been an
earlier stage of B-cell maturation, and differentiated further before being
arrested. Similarly the cell where the first genetic lesion occurred may have
been a later stage of B-cell development, after which it dedifferentiated
(Cobaleda & Sánchez-García, 2009). We must consider this limitation when
discussing specific B-lineage ALL subtypes such as precursor-B ALL, with the
knowledge that the origin of the disease may have been at one of many stages
of B-cell maturation (Figure 1.1).
Risk stratification of B-cell ALL
B-lineage ALL can be risk stratified to ascertain the most suitable treatment
protocol for the patient. The main risk factors include initial white blood cell
count, gender, rapidity of initial treatment response and cytogenetics. Recurrent
chromosomal alterations (cytogenetics) include changes to chromosome copy
number; where hyperdiploidy (greater than 50 chromosomes) predicts a good
prognosis and is found in approximately 24% of B-cell ALL cases in children.
Hypodiploidy (less than 44 chromosomes) is associated with a poor prognosis
and is rare, found in 1% of cases. Cytogenetic translocations, which create
fusion genes, are also prognostic in childhood B-lineage ALL. The most
common genetic lesion, named ETV6-RUNX1, occurs in 26% of cases and is
identified by a t(12;21)(p13;q22) chromosomal translocation. ETV6-RUNX1 has
been found to be a prenatal translocation and denotes a favourable outcome for
patients. The E2A-PBX1 (also known as TCF3-PBX1) fusion gene present at
7
t(1;19)(q23;p13) is found in 5% of cases but has no association with outcome in
children receiving contemporary therapy. The BCR-ABL1 translocation at
t(9;22)(q34;q11), also called the Philadelphia chromosome, is found in 5% of B-
lineage cases. This translocation has been implicated in various downstream
pathways and is associated with a poor response to treatment. The MLL gene
forms translocations at 11q23 with several partner genes and forms many
distinct fusions. MLL translocations are found in 7% of cases and are
associated with a poor prognosis, particularly in infants (Armstrong, 2005;
Inaba, et al., 2013; O'Brien & Lacayo, 2008; Teitell & Pandolfi, 2009).
Consequently, B-lineage ALL can be separated into several distinct subtypes,
which are associated with distinct disease progression and response to
treatment. However, many B-lineage ALL cases have no gross chromosomal
alterations and no cytogenetic alterations to predict risk of relapse. For these
patients risk of relapse is determined by response to first phase treatment. This
suggests that many other genetic structural changes and mutations are
necessary for leukaemogenesis.
Treatment regimen
The large success in overall survival of paediatric patients with B-lineage ALL,
now reaching 90% event-free survival, can largely be attributed to the
successive clinical trials carried out by international study groups such as the
Children’s Oncology Group (COG), Berlin-Frankfurt-Munster and St Jude
Children’s Research Hospital. Half a century of clinical trials has tailored
treatment protocols for drug combinations, concentrations, administration
methods and timing. These modifications were made possible by the
concurrent advancements in supportive care, allowing patients to better tolerate
highly intensive treatment regimens (Pui, 2006, 2009). The response to
treatment is a consequence of four facets; the regimen administered and its
effectiveness, treatment adherence by the oncologists, nurses and patient, the
pharmacogenomics of the patient and the leukaemic cell’s biological features
(Pui, 2008). The facets that cannot be controlled is the response to
administered drugs due to the patient’s own genetic variation
(pharmacogenomics) and the leukaemic cell’s genomic profile.
8
Currently, ten different chemotherapeutic drugs are used in a two to three year
treatment period comprising four phases (Inaba, et al., 2013). Many of the
drugs currently in use were developed before 1970. The first phase, an
induction stage, is a high-intensity therapy administered over 4-6 weeks at near
maximal tolerated doses. During induction, the leukaemic cell burden is
reduced by at least 2-5 logs and normal haematopoiesis is restored.
Chemotherapies used include a glucocorticoid (prednisone or dexamethasone),
vincristine and L-asparaginase, with or without anthracycline. The second
phase is a consolidation period; during this phase different drugs are used with
the aim of destroying residual leukaemic cells that did not respond to induction
phase. Chemotherapies include high-dose methotrexate, mercaptopurine and
asparaginase with frequent pulses of vincristine and glucocorticoids.
Consolidation can last 20-30 weeks and culminates in a reinduction therapy,
which repeats induction phase drugs. To prevent central nervous system
metastasis a third phase involves administration of drugs intrathecally. One
improvement to treatment protocols is the retirement of radiation in frontline
therapy, and the avoidance of its many adverse neurological effects. Currently
cranial irradiation may be used in salvage treatment for patients who relapse.
Finally, a maintenance phase, lasting 2 years or longer, comprises
uninterrupted daily mercaptopurine and weekly methotrexate, with or without
monthly pulses of vincristine and dexamethasone. Allogeneic haematopoietic
stem cell transplantation is used for very high-risk or relapsed patients. The
benefits of the procedure are more pronounced as the patients risk profile
increases (Balduzzi, et al., 2005).
The use of combination chemotherapy reduces the potential development of
drug resistance to individual agents and is more likely to elicit a response.
However, current conventional anticancer drugs have non-selective
mechanisms of action. They target important macromolecules or metabolic
pathways that are crucial to both malignant and normal cells. Therefore,
treatment of ALL is toxic and can damage or interfere with normal organ
function. Long-term effects of treatment remain a problem for social, cognitive
and physical development in children that continue into adulthood (Hudson,
2003; Schwartz, 1999). Specifically, chemotherapies result in early mortality,
cardiac complications, obesity, muscle weakness, neurosensory impairment,
9
neurocognitive deficits or endocrine and metabolic disorders (Ness, et al.,
2011). The need for improving the long-term health of paediatric leukaemia
survivors means reducing dosage and administration schedules of existing
drugs, as is currently being adopted for low-risk patients (Kamps, et al., 2002;
Vora, et al., 2013). These adverse outcomes highlight the need for newer more-
effective therapies with fewer long-term effects.
Drug resistance
Fortunately, with modern therapy most children diagnosed with leukaemia are
cured. However there is a group of patients that remain impossible to cure with
current chemotherapy approaches. To effectively prevent relapse, therapy must
suppress leukaemic growth and provide leukaemic cell reduction while not
permitting the emergence of drug resistant clones. In the past two decades
there has been limited progress in chemotherapy regimens for relapsed
patients. Relapses that occur early in relation to the time of diagnosis, and are
located in the bone marrow, are the most aggressive and have the worst
survival outcomes (Bhojwani & Pui, 2013). In the continued effort to achieve a
cure for all patients, the significant scientific challenge of counteracting or
preventing drug resistance must be overcome. We know that drug resistance
results from the intrinsic drug sensitivity of leukaemic cells themselves by clonal
selection or after sequential genetic changes or by protection from their
immediate environment (Meads, et al., 2008). Unravelling the molecular
pathways underlying drug resistance in B-lineage ALL will lead to better
knowledge of the disease and the potential discovery of new genes or proteins
suitable for targeted therapy.
Targeted therapies in childhood leukaemia are appearing with the push for a
“genomic medicine” approach to treating disease. A prime example is the
success in targeting of the BCR-ABL1 fusion with an ABL tyrosine kinase
inhibitor (Dasatinib/ Imatinib) (Evans, et al., 2013). The discovery of genetic
alterations that link to prognostic outcomes or targeted treatment options will
expand in future. As more genotype-phenotype associations are made the
efficacy for cancer and germline sequencing of patients increases (Bhojwani &
Pui, 2013). Using genome variation, which includes DNA sequence, epigenetic
10
changes, messenger RNA and micro RNA expression as a prognostic factor
and treatment indicator will lead to the greater individualisation of cancer
treatment. To increase the available number of targeted therapies, further
investigation of the biology of ALL is needed, and is the rationale for the work
presented in this thesis.
1.2 The bone marrow microenvironment The formation of cells into tissues, organs and systems is made possible
through the microenvironment. As every one of the trillions of cells within a
human has the same DNA sequence, there must be a dominant force that
drives differences in gene expression. Without the reciprocal communication of
cells with the microenvironment, they would remain largely plastic with no
determined phenotype. Early embryonic events such as vascular
morphogenesis, heart and endoderm-derived organ development and
innervation of tissues by neural cells and blood vessels, all rely on
microenvironment cues (Red-Horse, et al., 2007).
In tumourigenesis, the microenvironment plays a role in both tumour
suppression and promotion. Premalignant tumours may be more common than
currently recognised, but in the right context the microenvironment can
constrain tumours to a dormant state and likewise, aberrant signals from the
microenvironment can promote malignancy (Bissell & Hines, 2011; Potter,
2007). Although studies investigating the role of the bone marrow (BM)
microenvironment in haematological malignancies have been published for
decades, it is the solid-tumour microenvironment that has been most abundant
in the literature and therefore better defined. Across the gamut of solid-
tumours, stromal cells of the microenvironment (including fibroblasts, vascular
cells and immune cells) contribute to 7 of the 8 hallmarks of cancer, a
conceptual framework that unifies the genotypes and phenotypes of diverse
cancer types (Hanahan & Coussens, 2012). As stated by Douglas Hanahan
and Lisa Coussens, “Cancer cells do not manifest the disease alone, but rather
conscript resident and recruited normal cell types to serve as contributing
members to the outlaw society of cells.”
11
Structure and composition of the BM microenvironment The most prominent common denominator for B-lineage ALL and normal B-cell
development is the BM origin of both. The normal BM microenvironment is a
complex organization of a heterogenous assembly of cells, including lineages
from haematopoietic stem cells, stromal cells and extracellular matrix (ECM)
(Travlos, 2006). Stromal cells are derived from mesenchymal stem cells (MSC)
(Seshi, 2000). However, the term MSC is currently under scrutiny as the
majority of cells identified as mesenchymal stem cells are more accurately
heterogenous cells that show a large degree of plasticity (Bianco, et al., 2008;
Lindner, et al., 2010; Wagner, et al., 2008). What is certain is that MSCs
differentiate into committed progenitors, which become osteoblasts,
chondrocytes, myoblasts, tenocytes, adipocytes, fibroblasts and reticular cells
(Figure 1.2). The umbrella term of stromal cells in relation to the BM,
incorporates mesenchymal progenitors, endothelial cells, fibroblasts, reticular
cells, adipocytes and osteoblasts (Kiel & Morrison, 2008). In addition to stromal
cells, regulatory factors including the ECM, cytokines, chemokines, growth
factors and neural peptides control important pathways and cell interactions
(Janowska-Wieczorek, 2001; Lichtman, 1981; Mayani, 1992).
Vascular endothelial cells contribute to tumour formation by delivering nutrients
through blood vessels and also producing pro-tumorigenic factors (Butler, et al.,
2010). Immune cells are also important contributors to tumour progression.
Immune cell effects are complex and initiated by cells previously present in the
tumour microenvironment or attracted by inflammatory chemotactic signals.
Adaptive immune cells are tumour-destructive and directly involved in tumour
cell cytotoxicity and complement-mediated lysis, whereas innate immune cells
are tumour-supportive promoting fibroblast activation, angiogenesis, tissue
remodelling, and suppression of adaptive immune responses (de Visser, et al.,
2006; Lin & Karin, 2007). The collective immune presence in tumours evolves
dependent on tumour stage and location, indicating an organised response (Lin
& Karin, 2007). Specific roles of tumour cells in B-precursor ALL have not been
elaborated.
12
Haematopoietic and stromal cells exist within an abundant ECM, a three-
dimensional latticework of collagens, proteoglycans and glycoproteins. Stromal
cells within the BM synthesise ECM, which is not an inert scaffold. On the
contrary, it plays a role in cell adhesion and migration, cytokine presentation
and proliferation within the BM (Klein, 1995). Heparan sulfate proteoglycans are
known to avidly bind many growth factors and when bound attach to the ECM,
to act as a localised reservoir for the protein. In this way, the ECM assists in
establishing stable release of soluble growth factors. These same
proteoglycans can also bind growth factors to function as a solid-state ligand
which docks onto integrin adhesive receptors and initiates cell signalling. The
regulation of transforming growth factor-β (TGFβ) signalling by ECM proteins is
one of the better-understood examples. Large, inactive TGFβ complexes are
secreted from cells and upon binding with ECM components, such as fibrillins,
collagens and fibronectins, are held until activation. Activation of TGFβ can
occur by proteolysis or by mechanical strain (Hynes, 2009). This example
illustrates how ECM can regulate cell signalling through growth factor
presentation. The versatility of ECM molecules allows juxtaposition of growth
factors, integrins and growth factor receptors in a confined spatial context that
can encourage or inhibit cell signalling (Brizzi, et al., 2012). The precise
mechanisms of ECM interactions are not well defined. The hundreds of ECM
proteins in humans are predominately large, having several distinct domains,
highly-conserved and complex (Hynes, 2009). The conserved and complex
nature of ECM proteins suggests there would be many more biologically
important functions, with some components being common among all
echinoderms and chordates (which includes all vertebrates).
Haematopoietic stem cell niche (Leask & Abraham, 2006)
Three decades ago it was postulated that a localized niche in the BM was
necessary for haematopoietic stem cell (HSC) reconstitution (Schofield, 1978).
The niche in question was localised to the endosteal bone surface, the interface
of bone and BM (Gong, 1978). Even recent studies, that brought in more
complexity, followed the niche hypothesis by stipulating that differentiation of
dormant HSCs into progenitor cells is initiated by migration from the endosteal
niche to a perivascular niche (Wilson, 2006).
Figure 1.3. Human long bone as a model for the bone marrow niche. a. The long bone is divided into 3 anatomical sections: the epiphysis at the end of the bones and the metaphysis below the endosteal line, which are filled with spongy/ trabecular bone and the diaphysis along the middle shaft filled with compact bone. The bones are covered in periosteum and the bone joints with articular cartilage. Red bone marrow, where haematopoiesis occurs, is found within the trabecular bone for the lifetime in addition to the medullary cavity in childhood. b. Within the bone marrow spaces of the spongy bone a wide variety of vascular, endothelial and perivascular cells known as stromal cells contribute to haematopoietic stem cell (HSC) maintenance and differentiation (image adapted from Kiel and Morrison, 2008) c. A scanning electron microscopy image illustrates the persavive extracellular matrix present in all aspects of BM function. Depicts co-culture of mesenchymal stem cells (MSC) and endothelial progenitor cells (EPC) in an artificial collagen, type 1 scaffold. (adapted from O’loughlin and O’brien, 2011).
Megakaryocyte
HSC
Endosteal cell
Sinusoidalendothelium
reticular cell
Osteoclast
b.a.
c.
MSC
EPC
Co
Sca
14
However, the use of improved HSC markers in vivo and three-dimensional
imaging techniques have found that the anatomical location of approximately
two-thirds of HSCs is the BM space are found within trabecular, spongy bone
throughout the metaphysis, with the rest located throughout the diaphysis
(Figure 1.3). While HSCs are distributed throughout this space, a minority are
found near the endosteum (roughly 20%) and the vast majority are adjacent to
or near a sinusoid (Kiel, et al., 2005; Morrison & Scadden, 2014). Sinusoids are
the thin-walled blood vessels that allow easy passage of cells between the
circulation and haematopoietic tissue. Endosteal BM niches within the
trabeculae are associated with a rich network of blood vessels including
sinusoids very near the endosteal surfaces. While HSCs may not be in direct
contact with the endosteum, the two niches are capable of influencing each
other (Lassailly, et al., 2013). However, perivascular-associated mesenchymal
stromal cells and endothelial cells appear to be the main stromal cell types
involved in HSC maintenance and localisation, with a possibly more complex
and indirect role for mature osteo-lineage cells at the endosteal niche (Morrison
& Scadden, 2014).
Location of B-cell lymphopoiesis
Interactions within the BM play a critical role in the development of normal B-
cells (Kurosaka, et al., 1999; Nagasawa, 2006). Essential microenvironmental
components for B-cell development show some overlap with HSC requirements
and are in addition to essential transcriptional factors expressed by B-lineage
cells themselves. Essential components are released by endothelial and
mesenchymal stromal cells and include CXC-chemokine ligand 12 (CXCL12),
FLT3 ligand (FLT3), interleukin 7 (IL7), stem cell factor (SCF) and receptor
activator of nuclear factor-κB ligand (RANKL) (Nagasawa, 2006). Efforts to
identify cells responsible for expression of these essential microenvironment
components have been hindered by a lack of unique markers among stromal
cells and the redundancy in expression across different cell types. This has left
the source of some essential B-cell components well defined, while others are
not (Nagasawa, et al., 2011).
Figure 1.2. Mesenchymal Stem Cells (MSC) have the potential to differentiate into divergent cell lineages. MSCs have a high capacity for self renewal before differentiating into a range of mesenchymal cell types. Mature cells form tissues such as bone, cartilage, muscle and tendons. MSCs are also the source of BM stromal cells including adipocytes, fibroblast and reticular cells (adapted from Caplan and Bruder, 2001).
MSC
Progenitor
Maturecell
Terminal differentiation
(Fibroblasts,reticular cells)
TendonTenocytes
Adipocytes
Stromal cells (fibroblasts, reticular cells)
MuscleMyoblasts
CartilageChondrocytes
BoneOsteocytes
16
One well-defined stromal cell is the CXCL12-abundant reticular (CAR) cell that
is found throughout the BM and plays various roles. They have been found to
be essential for HSC maintenance, homing, B-cell and erythroid progenitor cell
maturation. CAR cells express not only CXCL12 but also SCF and vascular cell
adhesion molecule 1 (VCAM1) (Sugiyama, et al., 2006). It has been postulated
that the network created by CAR cells and their processes provides the
substructure for movement of HSCs and their progeny (Nagasawa, et al., 2011).
However, the redundancy of expression of CXCL12 in the BM, which is also
expressed by osteoblasts and endothelial cells, can serve to define niches
providing different functional roles in haematopoiesis.
Conditional deletion of CXCL12 from osteoprogenitors, mature osteoblasts,
osteoclasts and reticular cells (all expressing osterix), resulted in a loss of B
lymphoid progenitors, while HSC function was normal. Conditional deletion of
CXCL12 from limb bud-derived mesenchymal progenitors (expressing Prx1)
which differentiate into the majority of CXCL12-expressing cells in the BM with
the exception of vascular endothelial cells, was associated with a loss of HSCs,
HSC quiescence and common lymphoid progenitor cells (Greenbaum, et al.,
2013). This and other cumulative evidence suggest that early B-lymphoid
progenitors occupy endosteal niches in association with CXCL12-positive
osteo-lineage cells, which is an anatomically distinct site from the perivascular
niche where HSC cells and common lymphoid progenitors are associated with
CAR cells and CXCL12-positive endothelial cells (Ding & Morrison, 2013).
The movement of B-cell progenitors within the BM is driven by expression of
membrane receptors. Expression of the CXCL12 receptor (CXCR4) occurs at
all stages of development from HSC to immature B-cell and CXCR4 expression
drives the association with CAR cells. Lymphoid progenitor cells move from
their perivascular niche toward osteo-lineage CAR cells located at the
endosteum. From early-B to the pre-B cell stage there is also expression of the
IL7 receptor, and this drives pro- and pre-B cell association with IL7-expressing
stromal cells at perivascular sites and plays a crucial role in driving proliferation
at these stages. Pre-B cells then upregulate expression of CXCR4 in response
to complex rearrangements resulting in expression of interferon regulatory
factor 4 (IRF4) and migrate to CAR cells in a distinct niche away from HSCs.
The final move occurs after attenuation of IL7 signalling and exit from the cell
17
cycle, which allows final immunoglobulin light chain rearrangements and
expression of an antigen-specific B-cell receptor (BCR). These last steps are
necessary for selection into the immature-B cell pool, where cells move toward
vascular endothelial cells before migrating into the blood stream to reach the
spleen for further processing. This outlines a three niche model for B-cell
differentiation involving a CXCL12-abundant endosteal niche, IL7-abundant
perivascular niche and CXCL12-abundant perivascular or reticular niche (Clark,
et al., 2014; Nagasawa, 2006).
B-lineage leukaemia niche
The microenvironment of B-lineage ALL has been found to hijack normal HSC
niches leading to the suppression of normal haematopoiesis (Colmone, et al.,
2008). In a xenograft model in immune-compromised mice, a B-precursor ALL
cell line (Nalm6) homed to CXCL12-positive vascular niches within the BM and
then disrupted chemokine CXCL12 expression. Consequently, CD34-positive
haematopoietic progenitor cells were found to be re-directed to malignant
leukaemic niches and to exhibit abnormal behaviour at these tumour-associated
sites. The vascular leukaemic niche of ALL is in contrast to that of myeloid
leukaemias, where the majority of leukaemic blasts home to the endosteal
niche (Schepers, et al., 2013). A role for the endosteal niche in ALL has been
recently suggested with a study by Boyerinas et al. (2013) finding that dormant
B-precursor ALL blast cells from a human patient preferentially engrafted to
osteopontin (OPN)-expressing cells in the endosteum of NOD scid gamma
(NSG) mice. The endosteal niche in B-precursor ALL engrafted mice showed
evidence of a positive feedback loop between stromal and leukaemic cells for
OPN expression, suggesting maintenance of quiescence at endosteal sites.
The BM microenvironment affects leukaemic cell survival
From early studies in ALL, the BM microenvironment was cited as providing
support to leukaemia cell survival. Manabe et al. (1992) were first to
demonstrate BM stromal cells supported leukaemic cell survival in vitro in
comparison to serum-free tissue culture medium. B-precursor ALL was the
leukaemic subtype that had the greatest benefit to survival when adhered to a
18
BM stromal cell layer in vitro, further distinguishing the heterogeneity that exists
between the leukaemic lineages in regards to stromal dependency (Manabe, et
al., 1994). This beneficial in vitro interaction between B-precursor ALL cells and
BM stromal cells was most obviously demonstrated by direct cell-contact and to
a lesser extent by soluble factors alone (Ashley, et al., 1994). It was found that
within 30 minutes of contact with BM fibroblasts, B-precursor ALL cells
developed complex pseudopodia at sites of cell-cell contact and microvillus
projections, undergoing major shape changes that allowed penetration of the
fibroblast adherent layers (Hewson, 1996). Other leukaemic lineages, while
undergoing some shape changes, displayed no signs of migration within the
fibroblast layer. However, the direction of communication in the
microenvironment is not only BM stromal cells conferring supportive changes
onto ALL cells. ALL cells can impart a pro-angiogenic potential to BM
endothelial cells and produce an array of soluble factors that promote
neovascularisation (Veiga, et al., 2006). The exact role of the different growth
factors and cytokines in the leukaemic BM is not well defined, most likely they
form complex and dynamic networks where leukaemic blasts are in cross-talk
with stromal cells and facilitate complementary expression of soluble factors
(Ayala, et al., 2009; Wu, et al., 2005).
The BM microenvironment affects leukemic cell drug resistance
Environment–mediated drug resistance can be due to transcription of soluble
factors such as cytokines, chemokines and growth factors or the resistance can
be due to non-transcriptional mechanisms of cell adhesion (e.g. when tumour
cell integrins bind to fibroblasts or components of the ECM). Environment-
mediated drug resistance can precede minimal residual disease (the presence
of surviving tumour cells immediately after therapy) and the development of
acquired resistance. In leukaemia, the understanding that environment-
mediated drug resistance is largely caused by BM stromal cells is well
established.
Protection from the cytotoxic effects of chemotherapeutic drugs has been found
to occur after direct BM-leukaemia cell contact (Mudry, et al., 2000). Leukaemia
cell adhesion to the BM stromal cells has been shown to result in decreased
apoptosis through initiation of integrin signalling cascades, including
19
phosphorylation of extracellular signal-regulating kinases (ERK1/2) and Akt,
that provide protection from cytotoxic cell death (Hazlehurst, 2003; Shain, 2000;
Tabe, et al., 2007). In T-cell ALL, it has been shown that gap junction coupling
between BM stroma and leukaemic lymphoblasts diminishes proliferation,
keeps cells in a quiescent state and protects them from drug-induced apoptosis
(Paraguassu-Braga & Paraguassu, 2003).
Resistance to (L-)asparaginase cytotoxicity in B-precursor ALL, was found in
co-culture experiments with BM-derived mesenchymal stromal cells
(immortalised hTERT and primary cells)(Iwamoto, et al., 2007). Stromal cells
were found to have 20-fold greater expression of asparagine synthetase
(ASNS) expression (which catalyses the reaction that forms asparagine),
compared to ALL cells (3 cell lines and primary blasts). It was postulated that
stromal cells provided excess asparagine for tumour cells, which are incapable
of producing their own supply and in this way stromal cells provided a protective
effect, resulting in viable leukaemic cells capable of proliferation. BM-derived
stromal cells (immortalised HS5) also protected B-precursor ALL cells from
cytotoxic effects of vincristine and cytarabine when in co-culture, with survival
40% and 60% higher, respectively, in two B-precursor ALL cell lines compared
to the absence of stroma (Tesfai, et al., 2012). However, no protective effect
was found for dexamethasone or L-asparaginase in this study. The opposing
findings on (L-)asparaginase sensitivity in B-precursor ALL cells co-cultured
with BM-derived stromal cells can be attributed to the heterogenous expression
of ASNS in BM stromal cell lines, even among sub-clones of the same cell line,
and the variable response of B-precursor ALL cells cultured with BM stromal
cells, as discussed by Iwamoto et al. (2007).
The response of B-lineage ALL, T-lineage ALL or myelogenous leukaemia cells
in the presence of BM stromal cells have can be very different and result in
unique adhesion-based protective mechanisms or survival capabilities (Ayala,
et al., 2009; Hall & Gibson, 2004). Although all leukaemias exist in the same
BM microenvironment, biological differences in the tumour cell itself appear to
exert distinctive control over the BM response, reaffirming that context is always
vital in microenvironment response.
20
1.3 Connective tissue growth factor (CTGF) Relevance of CTGF in B-precursor ALL
Global gene-expression profiling using microarrays encompasses a subset of
over 20,000 genes of potential importance and is therefore well suited to
identifying new genes of interest. ALL, a relatively homogenous cancer that
allows easy isolation of malignant cells, is an ideal model system and many
investigators have used microarray to elucidate therapeutic targets and
diagnostic markers (Cheok, 2006). Using microarray technology, previous
studies conducted within our laboratory at the Telethon Institute for Child Health
Research have found that CTGF is a highly expressed gene in B-precursor ALL
(Boag, et al., 2007). We found CTGF to be upregulated 37-fold compared with
normal B-precursor cells from umbilical cord blood and 19-fold compared with
haematopoietic progenitor cells from healthy BM. The analysis revealed that
high expression levels of CTGF were restricted to B-lineage and not T-lineage
ALL. It was found overall that CTGF was highly expressed in 75% of B-
precursor ALL cases. This percentage was largely comprised of only four
different cytogenetic subgroups; hyperdiploidy, ETV6-RUNX1, MLL and BCR-
ABL. E2A-PBX1 translocations and other more infrequent aberrations did not
show significant CTGF expression across the independent cohorts. Protein
studies by western blot analysis confirmed the presence of CTGF in ALL cell-
conditioned media, confirming protein translation and secretion. Boag et al.
concluded that CTGF expression from B-precursor ALL cells was heterogenous
and could play a role in the pathophysiology of the disease. Moreover, other
microarray studies of adult and childhood B-precursor ALL samples found
CTGF was highly expressed compared to T-cell ALL, myeloid leukaemias and
control patient BM and peripheral blood samples (Sala-Torra, et al., 2005;
Vorwerk, et al., 2000).
Importantly CTGF expression is correlated with a poor prognosis in B-precursor
ALL. A gene identifier set correlated to a high-risk group of paediatric ALL
patients with the poorest clinical outcome, included CTGF expression (Kang, et
al., 2010). Similarly, in adult ALL high CTGF expression correlated with poor
overall survival of patients in a mortality hazard rate model (Sala-Torra, et al.,
21
2005). Increased CTGF expression having an association with poor prognosis
suggests the protein may confer a cellular phenotype that benefits the
leukaemia cell. In an in vitro functional study, CTGF adsorption by human BM
stromal cells promoted adhesion to synthetic surfaces (Ono, et al., 2007). If
CTGF were proven to alter adhesion of BM microenvironment components to
leukaemia cells in vivo, this would support the possibility of it playing a role in
environment-mediated drug resistance. Normal function of CTGF
It has been over 20 years since the discovery of CTGF from human umbilical
vein endothelial cells (Bradham, 1991). The CTGF gene has been
independently identified and named several times since then, as IGFBP-
8/IGFBP-r2/HCS24/NOV2, in consequence of its multi-faceted nature. CTGF is
an ECM-associated protein of 36-38kDa and a member of the CCN family of
proteins. The family is named after the three founding proteins, CYR61/CCN1
(Cysteine rich 61), CTGF/CCN2 and NOV/CCN3 (Nephroblastoma
overexpressed), and now comprises six members, with the addition of Wnt-
induced secreted proteins Wisp-1/CCN4, Wisp-2/CCN5 and Wisp-3/CCN6.
Each member consists of a highly conserved four-module structure that allows it
to regulate a multitude of cellular responses (Figure 1.4). The four motifs bear
resemblance to other unrelated extracellular proteins and include an N-terminal
region with an insulin-like growth factor (IGF)-binding domain, followed by a von
Willebrand type C domain, a thrombospondin-1 domain, and a cysteine knot at
the C-terminal end (Bork, 1993). CTGF is strongly induced by TGF-β and other
growth factors and is tagged for secretion to the ECM, the protein has been
highly researched for its role in various fibrotic and desmoplastic diseases,
including atherosclerosis, organ fibrosis and fibrotic skin disorders (Jun & Lau,
2011; Leask, 2002; Shi-Wen, et al., 2008; Sonnylal, et al., 2010). CTGF is also
involved in many normal biological processes, such as wound repair, bone
formation, ECM remodelling and angiogenesis (Arnott, et al., 2011; Babic, et al.,
1999; De Winter, et al., 2008; Hall-Glenn, et al., 2012; Kondo, et al., 2002;
Leask & Abraham, 2006; Safadi, et al., 2003). These normal functions of CTGF
play an essential role in development of epithelial tissues, as seen in
endochondral bone formation where CTGF provides differentiation and survival
22
signals to growth plate chondrocytes through ECM interactions (Hall-Glenn &
Lyons, 2011; Hall-Glenn, et al., 2013).
Known receptors for CTGF are cell type- and context-specific, allowing
remarkable functional versatility throughout the body. The known receptors
include heparan sulfate proteoglycans, the low-density lipoprotein receptor-
related proteins, TrkA and 6 integrins: α6β1, αvβ3, αvβ5, αIIbβ3, αMβ2,, α5β1 (Jun &
Lau, 2011). A specific receptor has not been found, though there is evidence
CTGF may also function as an adaptor molecule facilitating binding of
extracellular proteins to their cell surface receptors (Shi-Wen, et al., 2008). The
known ligands of CTGF include: fibronectin, decorin, vitronectin, TGFβ, VEGF,
FGF2 and BMP (Jun & Lau, 2011). Canonical pathways activated by CTGF
include Notch, TGFβ, WNT and PI3K/AKT signalling pathways, though there is
no definitive signalling pathway for the protein. These known interacting
proteins, and potentially other undiscovered molecules, allow CTGF to act in an
autocrine and paracrine manner to effect functional changes in normal and
neoplastic cells. In only a limited number of studies has there been functional
insight into how CTGF contributes to disease pathology, these will be discussed
below.
Role of CTGF in human cancers CTGF has long been associated with human cancers. The role it plays in these
neoplasms is diverse and tumour-specific. Recurring patterns in clinical
outcome, histological desmoplasia and mechanisms of action have been found.
When CTGF is overexpressed compared with low–expressing normal tissue or
is underexpressed compared with high-expressing normal tissue, the functional
outcome favours tumour survival and disease progression. CTGF acts by
altering proliferation, drug resistance, angiogenesis, adhesion and migration,
contributing to metastasis. The pattern of CTGF expression and tumour
response helps to clarify the role of this matricellular protein across a multitude
of human cancers.
Across the gamut of human cancers with deregulated CTGF gene expression,
there are similarities to be drawn that can help us better understand the role
CTGF plays in cancer. The review of the literature aims to accomplish this by
Figure 1.4. Conserved structure of CCN family members. CCN family members have a conserved four domain structure. A secretory signal peptide (SP) is followed by Module I, an Insulin-like growth factor binding domain (IGF-binding), then module II, a Von willebrand type C domain (VWC), this is linked with a protease susceptible hinge region to module III, a thrombospondin-1 domain (TSP-1) and lastly module IV, a cysteine knot domain (adapted from Leask and Abraham, 2006).
CCN1/ CYR61CCN2 / CTGFCCN3/ NOVCCN4/ WISP1CCN5/ WISP3
CCN5/ WISP2
24
clarifying the common characteristics among distinct human cancers and CTGF
expression, exploring clinical outcome in relation to CTGF expression and
examining the location of CTGF expression in cancer and surrounding cells.
The molecular mechanisms are summarised according to CTGF functions in
human cancers, including autocrine control of proliferation and drug resistance
and paracrine control of angiogenesis, invasion and metastasis.
Deregulated CTGF expression in cancer CTGF was first associated with human breast and pancreatic cancers in the late
1990’s (Frazier & Grotendorst, 1997; Hishikawa, et al., 1999). To date, 27
human cancers have been linked to aberrant CTGF expression (Table 1.1).
They include cancers affecting patients from very young children to older adults
and arising from many tissue types and systems of the body. Identifying the role
CTGF plays in these cancers is complicated by the tissue-dependent and
seemingly complex functions of the protein. However, as a secreted protein,
many functions take place in the ECM and tumour microenvironment, and their
role in cancer progression and survival has been well established (Mbeunkui &
Johann, 2009; Mueller & Fusenig, 2004). Studies investigating CTGF in human
cancers report both up- and downregulated expression with varying survival
outcomes. The reports comprise varying numbers of patient samples (Table
1.1). Methods of CTGF analysis also differ. Despite this, consistent trends are
apparent between CTGF expression and clinical outcome.
When interpreting the role of CTGF in human cancer, it is important to establish
when that deregulation occurs. Expression of CTGF may be global for a cancer
type or specific to a sub-type, differentiation grade (poorly or well differentiated)
or tumour-stage (such as the TNM classification). Specificity of CTGF
expression is not only evident in tumour cells but also normal cells (Safadi, et
al., 2003). Of the cancers with known deregulated CTGF expression, 44%
(12/27) were found to have expression levels dependent on the degree of
tumour differentiation or TNM stage, as assessed by univariate or multivariate
analysis (Table 1.1).
Location Corresponding normal tissue
Malignancy Patient samples (total)
References
Nervous system
GLIA Glioma (glioblastoma multiforme)
66^ Xie et al., 2004
Head and neck mucosa
ORAL MUCOSA, SINUS, TONGUE
Head and neck SCC 163^ Yang et al., 2011, Chang et al.,
2013 Skin MELANOCYTE Melanoma 10 Kubo et al., 1998,
Braig et al., 2011 Respiratory system
LUNG Lung Cancer (NSCLC and adenocarcinoma)
150^ McDoniels et al., 2002, Chang et
al., 2004,Chien et al., 2006
Digestive system
OESOPHAGUS Oesophageal cancer (SCC and adenocarcinoma)
113* Koliopanos et al., 2002, Zhou et al.,
2009 STOMACH Gastric cancer 195^ Liu et al., 2008,
Mao et al., 2011 COLON/ RECTUM
Colorectal cancer 255^ Lin et al., 2005, Lin et al., 2011
LIVER Intrahepatic cholangiocarcinoma
55* Gardini et al., 2005
Hepatocellular carcinoma
52^ Zeng et al., 2004, Mazzocca et al.,
2010 GALLBLADDER Gallbladder cancer 238* Alzarez et al.,
2008, Garcia et al., 2013
PANCREAS Pancreatic cancer 80^ Wenger et al., 1999, Dornhofer
et al., 2006, Bennewith et al.,
2009 Reproductive system
BREAST Breast cancera 55^ Xie et al., 2001
CERVIX Cervical cancer 57^ Wong et al., 2006, Xie et al.,
2012 OVARY Ovarian cancer 173^ Gery et al., 2005,
Kikuchi et al., 2007
PROSTATE Prostate cancer 704 Gorlov et al., 2010
Urinary system
KIDNEY Nephroblastoma (Wilm’s tumour)
67 Zirn et al., 2006
Table 1.1. A topography of CTGF-associated neoplasms in humans. CTGF-associated neoplasms are found throughout the body. The comprehensive number of patient samples that have been studied in their association with CTGF is also featured. Significant correlations between CTGF expression and the degree of tumour cell differentiation is indicated by: ^, associated or *, not associated with tumour differentiation grade or TNM staging. SCC: squamous cell carcinoma, NSCLC: non-small cell lung carcinoma. a The mammary gland is part of the integumentary system more closely related to skin and subcutaneous tissue but grouped here to highlight gender differences. b Currently classified as Undifferentiated Pleiomorphic Sarcomas c Ileal carcinoids are derived from enterochromaffin cells and gastric carcinoids are derived from enterochromaffin-like cells, both enteroendocrine cells derived from the gut epithelial cell lineage with enteric nervous system properties.
Connective and soft tissue
CHONDROCYTE Chondrosarcoma 18^ Shakunaga et al., 2000
FIBROUS TISSUE
Fibrosarcomab and Fibrous histiocytomasc
42 Igarashi et al., 1998
Myofibroblastic tumour
12 Kasaragod et al., 2001
MESOTHELIUM Mesothelioma 24 Fujii et al., 2012
SKELETAL MUSCLE
Rhabdomyosarcoma 0d Astolfi et al., 2001, Croci et al.,
2004 Endocrine system
ENTEROCHROM-AFFIN CELL
Ileal carcinoidc , Gastric carcinoidc
179* Kidd et al., 2007, Cunningham et
al., 2010 Haematologi-cal system
LYMPH NODE Hodgkin lymphoma 44 Birgesdotter et al., 2010
Large B-cell lymphoma
240 Rosenwald et al., 2010
THYROID Thyroid Cancer 20^ Cui et al., 2011 LEUKOCYTE B-precursor
acute lymphoblastic leukaemia
666 Vorwerk et al., 2007, Boag et al., 2007, Sala-Torra et al., 2007, Kang
et al., 2010 PLASMA CELL Multiple myeloma 57 Munemasa et el.,
2007
27
CTGF expression and clinical outcome Divergent CTGF expression in cancer cells compared to normal tissue can be
correlated with clinical outcome (Table 1.2). Normal tissue expression was
defined by specimens analysed in the relevant studies or data from GeneCards
microarray gene expression for normal tissues ("Genecards,"). No studies to
date have correlated CTGF expression with clinical outcome in connective
tissue neoplasms or neuroendocrine tumours, both of which have proven to be
affected by CTGF in experimental conditions, as discussed in section 5. (Croci, et al., 2004; Kang, et al., 2010; Kidd, et al., 2007; Rosenwald, et al., 2002; Wenger, et al., 1999) Higher CTGF expression in tumour compared to corresponding normal cells
Tumours overexpressing CTGF above the levels of their normal cell counterpart
correlate with poor clinical outcome. This is the case for B-cell acute
lymphoblastic leukaemia (Boag, et al., 2007; Sala-Torra, et al., 2007; Vorwerk,
et al., 2000), breast cancer (Wang, et al., 2009; Xie, et al., 2001), gastric cancer
(Liu, et al., 2008; Mao, et al., 2011), glioma (Xie, et al., 2004), hepatocellular
carcinoma (Mazzocca, et al., 2010; Xiang, et al., 2012), melanoma (Braig, et al.,
2011), oesophageal cancer (Deng, et al., 2007; Li, et al., 2010; Zhou, et al.,
2009), pancreatic cancer (Bennewith, et al., 2009; Dornhofer, et al., 2006),
prostate cancer (Gorlov, et al., 2010) and thyroid cancer (Cui, et al., 2011). In
these cancers CTGF expression in biopsy samples is associated with worse
survival outcomes and clinico-pathological findings, including advanced stage,
larger tumour size, increased metastasis, and poorer chemotherapy response.
In mesothelioma CTGF protein expression was only detectable in the
sarcomatoid subtype, which is the most aggressive and has the poorest
outcome (Fujii, et al., 2012). In multiple myeloma a cleaved CTGF fragment is
increased in plasma levels and associated with bone disease (Munemasa, et
al., 2007). In chondrosarcoma and pancreatic cancer CTGF is transiently
increased compared to normal skeletal muscle cells during early tumour stages,
suggesting a role in initial tumour development (Hartel, et al., 2004; Shakunaga,
et al., 2000).
In contrast, in gallbladder cancer higher expression of CTGF in tumour cells
leads to better patient outcomes, as CTGF is overexpressed compared to
28
normal tissue and increases with advancing tumour stage (Garcia, et al.,
2013a)
and high CTGF expression correlates with a better survival outcome in patients
with advanced gallbladder cancer (Alvarez, et al., 2008). Functional studies
have revealed that CTGF increases proliferation of gallbladder cancer cells in
vitro, suggesting the microenvironment may be responsible for host-protective
functions (Garcia, et al., 2013b).
Lower CTGF expression in tumour compared to corresponding normal cells
Cancer cells with low expression of CTGF compared to normal high-expressing
cells correlate with poorer clinical outcomes. In colorectal cancer (Lin, et al.,
2005; Lin, et al., 2011), Wilm’s tumour (Li, et al., 2008; Zirn, et al., 2005; Zirn, et
al., 2006), intrahepatic cholangiocarcinoma (Gardini, et al., 2005) and non-small
cell lung cancer and lung adenocarcinoma (Chang, et al., 2004; Chien, et al.,
2006; McDoniels-Silvers, et al., 2002), reduced CTGF expression compared
with highly-expressing normal tissue was associated with worse survival
outcomes and increased tumour recurrence, advanced tumour stage and lymph
node metastases.
CTGF expression is tumour-stage dependent in both ovarian and cervical
cancer. In ovarian cancer CTGF expression is lost in early stages of the tumour
then regained (Gery, et al., 2005; Kikuchi, et al., 2007). In cervical cancer CTGF
is downregulated compared to matched-normal controls and upregulated in late
compared to early tumour stages (Wong, et al., 2006; Xie, et al., 2012). This
suggests that in ovarian and cervical cancer, CTGF plays a role in tumour
initiation. In head and neck squamous cell carcinoma, CTGF-expression is lost
in late stages of the tumour, suggesting a role in tumour progression (Yang, et
al., 2011).
Taken together, there is a consistent pattern between poor clinical outcome and
divergent CTGF expression compared with normal tissue. As seen in Table 1.2,
deregulated CTGF expression is most often associated with poor clinical
outcome, whether the gene is over expressed when CTGF expression is low in
normal tissue or under expressed when CTGF is found to be high in normal
tissue. The exceptions to this pattern, lymphoma and gallbladder cancer are the
Table 1.2. Pattern of CTGF deregulation compared to normal tissue and its association with clinical outcome and cellular mechanisms of action. Deregulated CTGF is associated with a mainly poor prognosis. CTGF is associated with an increased (·) or decreased (‚) activity of several cellular mechanisms and its deregulation in cancer leads to increased tumour survival. DLBCL: diffuse large B-cell lymphoma, SCC: squamous cell carcinoma, AC: adenocarcinoma, NSCLC: non-small cell lung carcinoma. *Drug resistance was increased when CTGF was deregulated in the relevant tumour types.
Cancer type
Clin
ical
ou
tcom
e
Prol
ifera
tion
Drug
re
sist
ance
*
Angi
ogen
esis
Adhe
sion
M
igra
tion
Met
asta
sis
Higher CTGF Expression B-precursor acute lymphoblastic leukaemia
Poor Brain cancer Poor ? ? ? Breast cancer Poor ? ? ? ? Chondrosarcoma Poor Gastric cancer Poor ? ? Hepatocellular carcinoma
Poor ? ? ? Ileal carcinoid Poor Melanoma Poor ? Mesothelioma Poor ? Multiple myeloma Poor Oesophageal cancer Poor ? Pancreatic cancer Poor ? ? ? ? Prostate cancer Poor ? Thyroid cancer Poor ? ? Gallbladder cancer Good ? Lymphoma Good Lower CTGF Expression Cervical cancer Poor ? Colorectal cancer Poor ? ? Head and Neck SCC Poor ? ? ? Intrahepatic cholangiocarcinoma
Poor Lung cancer Poor ? ? ? ? Nephroblastoma Poor Ovarian cancer Poor ? ? ?
30
only cancers where deregulated CTGF expression in tumour cells is associated
with better patient outcomes.
Location of CTGF expression and desmoplasia
Desmoplasia is a dense accumulation of fibrotic tissue surrounding a tumour.
As mentioned previously, CTGF has been shown to be actively involved in
fibrosis, whereby excess fibrous connective tissue is deposited. It is not
surprising that CTGF expression from both tumour and stromal cells has
associations with desmoplasia. Whether desmoplasia provides a protective
environment for the tumour or supports tumour elimination remains to be
determined for many cancer types. Therefore, there is no easily predictable
relationship between CTGF-associated desmoplasia and patient outcome.
CTGF expression in stromal cells
Elevated CTGF expression in stromal fibroblasts is implicated in the
desmoplastic response in breast cancer, colorectal cancer, desmoplastic
malignant melanoma and various connective tissue neoplasms including infant
myofibrosis, fibrosarcoma, malignant fibrous histiocytoma and malignant
hemangiopericytoma (Frazier & Grotendorst, 1997; Igarashi, et al., 1998;
Kasaragod, et al., 2001; Knowles, et al., 2011; Kubo, et al., 1998). Conversely,
high expression of stromal-CTGF in hepatocellular carcinoma was linked to a
lack of fibrous capsule around the tumour (Wang, et al., 2010; Xiang, et al.,
2012).
Aside from a role in desmoplasia, CTGF expression in stromal cells can
advance tumour growth or promote invasion in glioma, pancreatic and prostate
cancer. In prostate cancer, stromal cells expressing CTGF demonstrated
increased angiogenesis and tumour growth when co-injected with pancreatic
cancer cells (Yang, 2005). In glioma, stromal astrocytes at the tumour front
secrete high levels of CTGF and create a permissive microenvironment for
invasion (Edwards, et al., 2011). In pancreatic cancer, stromal-CTGF
expression contributed to increased angiogenesis and tumour-invasiveness
(Eguchi, et al., 2012; Ijichi, et al., 2011; Kwon, et al., 2007).
CTGF expression in tumour cells
31
High tumour-CTGF expression was associated with an increased desmoplasia
or close proximity to fibrovascular stroma in hepatocellular carcinoma,
oesophageal cancer, Hodgkin’s lymphoma and ileal carcinoids (Mazzocca, et
al., 2010). Similarly, in diffuse large B-cell lymphoma, increased CTGF
expression was associated with a more stromal subtype (germinal centre-like)
and present in gene signatures associated with increased ECM deposition and
monocytic invasion (Blenk, et al., 2007).
As discussed previously, lymphoma and gallbladder cancer are the only human
cancers studied to date with CTGF-deregulation being linked to a good
prognosis. That there is an increased stromal component in lymphoma,
suggests CTGF-microenvironment interactions may be responsible for
improved patient outcomes. The composition of the microenvironment when
CTGF is deregulated has not been reported for gallbladder cancer. However, a
microenvironment-mediated role for CTGF in gallbladder cancer is supported by
evidence that TGFβ was found to produce anti-angiogenesis and anti-
proliferation factors, which maintained distant metastatic tumour cells in a state
of dormancy in vivo (Gohongi, et al., 1999). As CTGF is a known downstream
mediator of TGFβ, it is plausible that it could function through the same
signalling pathways in gallbladder cancer to mediate microenvironment
interactions. CTGF operating in this capacity could certainly be the means to
improved clinical outcome, as the presence of distant metastases is an
important independent factor predicting survival in gallbladder cancer (Alvarez,
et al., 2008).
Mechanisms of CTGF action in cancer
CTGF has been found to change tumour cell proliferation and drug resistance.
Further, CTGF changes the tumour microenvironment by altering angiogenesis
and invasion of new tissues, achieved through changes in adhesion and
migration and finally the formation of metastatic colonies. Following the pattern
seen in clinical outcomes, the effect of CTGF on these mechanisms is
dependent on whether CTGF is over- or underexpressed compared with normal
cell counterparts. In tumours where CTGF is overexpressed functional changes
serve to promote tumour cell survival or metastasis. In contrast, in tumours
32
where CTGF is underexpressed compared with normal the presence of CTGF
reduces the aggressiveness of a tumour (Table 1.2).
Proliferation
CTGF has a proliferative role in gallbladder cancer, hepatocellular carcinoma,
malignant mesothelioma, oesophageal cancer, rhabdomyosarcoma, pancreatic
cancer and thyroid cancer (Fujii, et al., 2012; Garcia, et al., 2013b). In
hepatocellular carcinoma, increased proliferation was attributed to changes in
the cell cycle with increased numbers of cells in S-phase (Xiu, et al., 2012). In a
pancreatic mouse flank-tumour model, a CTGF monoclonal antibody (FG-3019,
Fibrogen Inc) reduced tumour growth (Aikawa, et al., 2006). In humans, a
phase 1 clinical trial of FG-3019 was used in locally advanced or metastatic
pancreatic cancer patients and was well tolerated (Heestand, et al., 2011).
Recently in pancreatic cancer, Neesse et al. found in vivo that gemcitabine
delivered alongside FG-3019 increased tumour cell death without altering
gemcitabine delivery, reported to be due to FG-3019 disrupting the survival
signal x-linked inhibitor of apoptosis protein (XIAP) (Neesse, et al., 2013).
CTGF has repeatedly been found to have no effect on in vitro proliferation of
breast cancer cells (Chen, et al., 2007; Pandey, et al., 2009; Shimo, et al.,
2006). However, the findings of Capparelli et al. (2012) have shown tumour
growth was inhibited when MDA-MB-231 breast tumour cells overexpressed
CTGF in vivo, as tumour-autophagy caused tumour lysis. This is an indirect
method of limiting proliferation of tumour cells. But tumour growth increased
when co-injected fibroblasts overexpressed CTGF, as fibroblast-autophagy
created nutrients for the developing tumour.
In cancers with reduced CTGF expression compared with normal cells an anti-
proliferative mechanism was found. CTGF overexpression inhibited tumour
proliferation in lung adenocarcinoma and non-small cell lung cancer, head and
neck cancer and ovarian cancer (Chang, et al., 2006; Chien, et al., 2006;
Kikuchi, et al., 2007; Moritani, et al., 2003). In ovarian cancer and lung cancer,
CTGF alters the cell cycle with more cells in G1-G0 and causing G1-S
checkpoint arrest. In these neoplasms the loss of CTGF would effectively
increase proliferation.
33
Drug resistance
CTGF has been associated with drug resistance in glioblastoma multiforme,
breast cancer and thyroid cancer. CTGF was involved in modulating apoptosis
genes, which caused resistance to several apoptosis-inducing drugs. In
glioblastoma multiforme CTGF was associated with expression of the anti-
apoptotic genes, survivin and FLIP, and a decrease in pro-apoptotic PARP
expression (Yin, et al., 2010). In breast cancer, CTGF modulated expression of
anti-apoptotic genes Bcl-xL and cIAP1 through an integrin αvβ3/FAK/ERK1/2
pathway. Furthermore, breast cancer patients with high CTGF expression
correlated with a poor chemotherapeutic response (Wang, et al., 2009). In
thyroid cancer, CTGF overexpression was linked to drug resistance, with
knockdown studies indicating cleaved caspase-3 and -9, members of the cell-
apoptosis pathway, are modified (Cui, et al., 2011).
Angiogenesis
CTGF deregulation acts to increase tumour angiogenesis in tumours over- and
underexpressing CTGF compared with normal cell counterparts. CTGF
increased tumour angiogenesis in vivo in breast cancer, glioblastoma
multiforme and pancreatic cancer (Aikawa, et al., 2006; Kondo, et al., 2002; Yin,
et al., 2010). In all these cancer types CTGF is over expressed compared with
normal cells. In a bone metastasis model of breast cancer, treatment with an
anti-CTGF antibody reduced angiogenesis, revealing a role for CTGF at a
secondary site (Shimo, et al., 2006).
In lung adenocarcinoma, where CTGF is underexpressed compared with
normal cells, forced CTGF expression reduced angiogenesis in vivo. Further,
patient samples with high CTGF correlated with less angiogenesis (Chang, et
al., 2006). Again, the ultimate outcome of deregulated CTGF expression
increased angiogenesis, maintaining tumour progression.
Adhesion
Adhesion is linked to the invasive abilities of cancers, with the need to reduce
adhesion to extravasate from the primary tumour site and then increase
adhesion to colonise distant sites of metastasis. CTGF plays a role in
extravasation and colonisation. In ovarian cancer and colorectal cancer, CTGF
expression is reduced compared with normal tissue, correspondingly CTGF
34
works to disadvantage the tumour. When recombinant CTGF is present, ovarian
cancer cells adhere more strongly to collagen 1, a major building block of the
ECM (Barbolina, et al., 2009). The low level of CTGF allows for reduced
adhesion and increased capacity for extravasation. Inducing CTGF expression
in colorectal cancer cells resulted in reduced peritoneal tumour dissemination.
CTGF inhibited the ex vivo adhesion abilities of colorectal cancer cells to
matrigel and excised peritoneum, acting at least partially through activation of
integrin α5 (Lin, et al., 2011). The reduced CTGF expression in colorectal
cancer may lead to an increase in adhesion and the ability of tumour cells to
seed the abdominal cavity, the precursor to peritoneal tumour dissemination.
Conversely, in gastric cancer CTGF is over expressed compared with normal
tissue. Increased CTGF promotes epithelial to mesenchymal transition in
peritoneal mesothelial cells. This provides a favourable environment for gastric
cancer cells to adhere to the peritoneum, an initial step in peritoneal tumour
dissemination (Jiang, et al., 2012).
Migration
Migration is used by cancer cells to spread within a tissue space and can occur
in many forms. It is required for the invasion of tissue barriers that occurs as
part of metastasis (Friedl & Wolf, 2003). In cancers that have higher expression
of CTGF compared with normal tissue, CTGF increases migration. This is the
case in glioma, gastric cancer, hepatocellular carcinoma, malignant melanoma
and pancreatic cancer (Mazzocca, et al., 2010). In glioma CTGF expression
was found to increase in intensity from the centre of the tumour to the infiltrative
front (Yin, et al., 2010). In glioma and gastric cancer activated nuclear factor
kappa-B (NFKB) suppressed E-cadherin and migration (Edwards, et al., 2011;
Mao, et al., 2011). In breast cancer activation of the integrin αvβ3/ERK1/2
pathway along with S100A4 was responsible for cell motility (Chen, et al.,
2007). CTGF was upregulated in invasive compared to non-invasive uveal
melanoma cell lines as determined by microarray and RT-PCR (Seftor, et al.,
2002).
In cancers where CTGF is under expressed compared with normal tissue,
CTGF inhibits migration and thereby promotes a favourable outcome for the
tumour when underexpressed. This is the case in cervical, colorectal, head and
35
neck, lung and ovarian cancer (Barbolina, et al., 2009; Chang, et al., 2004;
Chang, et al., 2013; Chuang, et al., 2011; Lin, et al., 2005; Xie, et al., 2012;
Yang, et al., 2011). In head and neck cancer, loss of CTGF promotes migration
through a switch to N-cadherin expression and epithelial mesenchymal
transition (Chang, et al., 2013). This is the opposing mechanism to glioma and
gastric cancer.
Metastasis
Studies using in vivo metastasis models show differences in metastasisng
colonies due to alterations in adhesion, migration, invasion and anoikis. In
tumours with higher CTGF expression than in normal tissue metastasis was
increased, shown in breast cancer, gastric cancer, hepatocellular carcinoma,
pancreatic and prostate cancer (Aikawa, et al., 2006; Dornhofer, et al., 2006;
Gorlov, et al., 2010; Kang, et al., 2003; Mao, et al., 2011; Mazzocca, et al.,
2010; Minn, et al., 2005). The models ranged from mouse cardiac injections,
chicken embryo metastases and meta-analyses of human tissue microarrays
from primary and metastatic tumours. In breast cancer, CTGF was associated
with bone-metastasis gene-sets; it was not sufficient alone to induce metastasis
to bone, though CTGF enhanced metastasis in vivo when overexpressed with
Osteopontin (OPN) and Interleukin 11 (IL11). CTGF was upregulated in clinical
breast cancer bone metastasis samples compared with normal breast tissue.
However, the expression of CTGF was not different between bone metastasis
samples from other solid tumours. This suggests the involvement of CTGF in
bone metastasis may not be specific to breast cancer. The expression of CTGF
in clinical bone metastasis samples significantly correlated with interleukin 11
(IL11), a disintegrin and metalloproteinase with thrombospondin motifs 1
(ADAMTS1) and C-X-C chemokine receptor type 4 (CXCR4). (Casimiro, et al.,
2012).
CTGF monoclonal antibodies have led to reduced metastases in breast and
pancreatic cancer. A breast cancer metastasis model treated with murine CTGF
monoclonal antibody (JMab 31, JT Central Pharmaceutical, Japan) resulted in
significantly less osteolytic lesions (Shimo, et al., 2006). In pancreatic cancer,
the combined treatment of gemcitabine and a CTGF monoclonal antibody (FG-
3019) reduced liver metastases (Neesse, et al., 2013).
36
In cancers with reduced CTGF expression compared with normal tissue, CTGF
acts to inhibit metastasis and therefore its underexpression leads to increased
metastasis. This is seen in mouse models for colorectal cancer, head and neck
cancer and lung cancer (Chang, et al., 2004; Chang, et al., 2006; Lin, et al.,
2005; Yang, et al., 2011). In colorectal cancer patients, low CTGF expression
was associated with an increase in peritoneal dissemination recurrence (Lin, et
al., 2011). In lung adenocarcinoma, cells with increased CTGF expression had
increased apoptosis after cell detachment, known as anoikis (Chang, et al.,
2012). Lung adenocarcinoma cells, with reduced expression, can escape
anoikis and undergo metastasis.
Conclusions CTGF activity has been investigated and reported in 27 human cancers, with
many more possibly still unrecognised. A common feature among these cancer
types is a pattern of deregulated CTGF expression and worse clinical outcome.
Stromal-expression of CTGF was also found to affect tumour survival through a
desmoplastic response. CTGF was found to regulate proliferation, drug
resistance and interactions with the stromal microenvironment.
Microenvironment changes due to CTGF gave an advantage to tumours by
modifying angiogenesis, adhesion, migration and metastasis. CTGF was found
to increase these mechanisms of action when it was over expressed in tumours
compared with normal tissue. In contrast, CTGF was found to decrease these
mechanisms of action in tumours where it was under expressed compared with
normal tissue. The ultimate outcome is that deregulated CTGF leads to an
increase in cellular mechanisms of action that support tumour cell survival.
Exceptions to this pattern include Hodgkin’s lymphoma, diffuse large B-cell
lymphoma and gallbladder cancer, where increased CTGF expression from
these tumours is associated with better prognosis. This is likely due to the
tumour-stroma interactions eliciting a protective fibrotic response, hindering
tumour progression or metastasis. Deregulated CTGF being associated with a
host-protective microenvironment has only been reported in these three cancer
types, reasons for this should be explored in future studies.
37
The role of CTGF in human cancer is not straightforward. Each cancer has
proven to have modifications attributable to CTGF that are unique due to the
matricellular nature of the signalling protein. However, it is clear that when
CTGF is deregulated in cancer compared with its normal counterpart, it effects
changes that advantage the tumour. Early evidence of CTGF antibody use in
pancreatic and breast cancer mouse models is a promising model that targeting
specific proteins in the microenvironment can lead to improved patient
outcomes.
1.4 Aims and outline of the thesis
This thesis is guided by an overarching question: Is there a specific functional
role for CTGF in B-precursor ALL?
The question is of importance as in childhood B-precursor ALL, CTGF is the
most deregulated gene and has been associated with poor outcome and high-
risk stratification. Yet the functional role CTGF plays in B-precursor ALL,
contributing to the phenotype of the leukaemia cell, remains to be identified.
The hypothesis for this thesis is that secretion of the CTGF matricellular protein
initiates a cascade of events, mediated through the microenvironment, that
contribute to leukaemogenesis and/or tumour cell survival. This thesis
contributes new knowledge of childhood leukaemias, verifying the existence of
a functional role for CTGF and the nature of microenvironment interactions
between B-precursor ALL and BM stromal cells.
The specific Aims of this thesis were to:
1. Engineer stable CTGF expression in established human B-precursor ALL cell
lines
(PER-278 and PER-371) and characterise their ability to contribute to the
leukaemia cell phenotype through intracellular or autocrine mechanisms.
38
2. Determine the effect of ectopic CTGF on an immortalised human BM stromal
cell line HS5, thereby verifying the potential of CTGF to be an intermediary in
B-precursor ALL cross-talk with components of the BM microenvironment.
3. Delineate the functional outcomes of CTGF-overexpression in B-precursor
ALL within the context of a murine BM microenvironment. This was
accomplished by establishing a xenograft model using immune-compromised
mice and engineered cell lines from Aim 1.
Chapter 2 provides a technical description of all materials and methods used to
establish experimental results for this thesis.
Chapter 3 covers the production of CTGF-overexpressing B-precursor ALL cell
lines and their characterisation (Aim 1).
Chapter 4 explores the changes to human BM stromal cell gene expression
due to treatment with ectopic CTGF (Aim 2).
Chapters 5 and 6 describe the PER-278 and PER-371 mouse xenograft
studies, respectively, due to differences between these cell lines in the
functional outcomes of xenografted mice (Aim 3).
Chapter 6 provides a summary and general discussion of the results that were
obtained as part of this PhD research and explores future research directions.
222222222
Materials and Methods
I keep six honest serving-men,They taught me all I knew;
There names are What and Why and When,And How and Where and Who.
Rudyard Kipling
40
41
Chapter 2 Materials and Methods
2.1 Cell lines and tissue culture
All cultured cell lines were verified to be free of mycoplasma contamination by
polymerase chain reaction (Tang, et al., 2000). Authentication of cell lines was
performed by short tandem repeat (STR) profiling on DNA using the GenePrint
10 System (Promega, Madison, WI, USA, #B9510) at the Johns Hopkins
genetic resources core facility (Baltimore, MD, USA).
2.1.1 Suspension cell lines ALL leukaemia cell lines used in this study, PER-255, PER-278, PER-371 and
PER-377, were generated at the Telethon Institute for Child Health Research in
the laboratory of Prof Ursula Kees according to previously published methods
(Kees, et al., 1987). Culture medium used for all suspension cell lines was
RPMI-1640 (MP Biomedicals, LLC) supplemented with 1mM sodium pyruvate
(MP Biomedicals, LLC), non-essential amino acids (MP Biomedicals, LLC), 2nM
2-mercapto-ethanol, and 2mM L-glutamine. Culture medium for suspension cell
lines was supplemented with 10% (v/v) fetal calf serum (FCS) (Invitrogen, Life
Technologies, Carlsbad, CA, USA). Cells were passaged twice weekly (split
1:2) and maintained in 24-well plates (Nunc, ThermoFisher Scientific, Waltham,
MA, USA) at 5% CO2 and 37°C. Suspension cell lines were quantitated using
the Vicell XR cell viability analyzer (Beckman Coulter).
2.1.2 Adherent cell lines The immortalised human bone marrow stromal cell line HS5 (CRL-11882)
(Roecklein & Torok-Storb, 1995b) and human embryonic kidney cell line 293T
(CRL-3216) were obtained from the American Type Culture Collection (ATCC,
Manassas, VA, USA). HS5 cells were maintained in RPMI-1640 supplemented
with 1mM sodium pyruvate, non-essential amino acids, 2nM 2-mercapto-
42
ethanol, 2mM L-glutamine and 10% (v/v) FCS. Cells were passaged twice
weekly on reaching 80-90% confluence (split 1:10) and maintained in 75cm2
filter-top culture flasks (Nunc, ThermoFisher Scientific, Waltham, MA, USA).
293T cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM,
Gibco, Life Technologies, Carlsbad, CA, USA) with 10% (v/v) FCS, 2mM L-
glutamine and 1% Penicillin/Streptomycin (Life Technologies, Carlsbad, CA,
USA). Cells were passaged thrice weekly on reaching 80-90% confluence (1:5)
and maintained in 100x20mm tissue culture dishes (BD Falcon, Becton
Dickinson, Franklin Lakes, NJ, USA). All adherent lines were cultured at 5%
CO2 and 37°C.
For passaging, cells were washed with phosphate buffered saline [PBS;
1.84mM KH2PO4, 2.68mM KCl, 136.87mM NaCl, 7.87mM Na2HPO4.2H2O] and
Trypsin-EDTA (Invitrogen, Life Technologies, Carlsbad, CA, USA) was used to
detach adherent cell lines from the culture dish or flask. Adherent cells were
quantitated using the trypan blue exclusion technique. Briefly, cells were mixed
1:1 with 0.1% Trypan Blue (Sigma Aldrich, St. Louis, MO, USA) in PBS and
pipetted into a coverslipped haemocytometer. The concentration of cells per ml
was found by the multiplication of the average viable cell count per square, the
cell dilution factor and 10,000.
2.2 Cell Transfections 2.2.1 Plasmids used in cell transfection To overexpress CTGF from leukaemia cell lines PER-278 and PER-371 two
techniques were attempted. The mammalian expression plasmid used for
electroporation, pcDNA3.1-CTGF, and its empty vector, pcDNA3.1-eGFP,
contained a cytomegalovirus (CMV) promoter and green fluorescent protein
(GFP) marker. Both were kindly donated by Prof David. R. Brigstock (Children’s
Research Institute, Columbus, OH, USA). Both vectors contained a neomycin
resistance gene. The mammalian expression plasmid used for lentiviral
infection, pCDH-MSCV-MCS-EF1-GFP-PURO (empty vector) was purchased
from SBI (System Biosciences, Mountain View, CA, USA, #CD713B-1) and
featured a murine stem cell virus (MSCV) promoter to drive expression of
CTGF, production of the CTGF plasmid is detailed in section 2.2.1.1, and an
43
elongation factor 1α (EF1) promoter to drive expression of GFP. Both pCDH-
MSCV-MCS-EF1-GFP-PURO and pCDH-MSCV-CTGF-EF1-GFP-PURO
contained a puromycin resistance gene (PURO).
Lentiviral packaging plasmids were kindly donated by St. Jude Children’s
Hospital (Memphis, TN, USA). They include the major proteins of human
immunodeficiency virus (HIV-1): gag and pol (in plasmid pCDH-KGP1-1R), rev
and tat (in plasmid pCDH-RTR2) and envelope (in plasmid pCAG-VSVG).
2.2.1.1 Ligation of human CTGF insert into pCDH vector Double restriction digests using restriction enzymes BamHI (Promega, Madison,
WI, USA, #R6021) and NotI (Promega, Madison, WI, USA, #R6431) and
10xBuffer D (Promega, Madison, WI, USA) were performed individually for both
pcDNA3.1-CTGF, to obtain the full-length human CTGF insert, and pCDH-
MSCV-MCS-EF1-GFP-PURO, to produce sticky ends for insertion of CTGF.
Briefly, 4µg of plasmid DNA was added to 5 units (U) each of NotI and BamHI
and made up to 40µl with 1xBuffer D and 0.1% bovine serum albumin (BSA,
Sigma Aldrich, St. Louis, MO, USA). The solution was incubated at 37°C for 90
min. Following incubation, the restriction digest was deactivated by incubating
at 65°C for 10 min.
Digested plasmids were resolved using agarose gel electrophoresis. Molecular
biology grade agarose (Amresco, Solon, OH, USA) was melted in 1x TAE buffer
[40mM Tris, 5.7% glacial acetic acid, 1mM ethylenediaminetetraacetic acid
(EDTA)] and supplemented with SYBR Safe DNA stain (Invitrogen, Life
Technologies, Carlsbad, CA, USA). Samples were combined with 10xloading
buffer (4.4mM Orange G, 40% glycerol) and run across 3 lanes alongside a 1Kb
DNA ladder (Invitrogen, Life Technologies, Carlsbad, CA, USA). Samples were
run for 80 min at 80V. The linear pCDH-MSCV-MCS-EF1-GFP-PURO band and
CTGF insert from pcDNA3.1-CTGF were isolated under ultra-violet (UV) light
with a clean scalpel. Plasmid DNA was purified with a QIAquick gel extraction
kit for cleanup of DNA (Qiagen, Venlo, Limburg, Netherlands, #28704). Purified
DNA was quantitated using the Nano drop ND-1000 spectrophotometer
(ThermoFisher Scientific, Waltham, MA, USA). The linear pCDH vector and
sticky-ended CTGF insert were ligated together with LigaFast rapid DNA
44
ligation system (Promega, Madison, WI, USA, #M8221), according to modified
manufacturer’s instructions, using only 20ng of both vector and insert, and
incubating for 24 hours at room temperature. Ligated pCDH-MSCV-CTGF-EF1-
GFP-PURO plasmid DNA was cloned through bacterial transformation,
explained in 2.2.1.2 and confirmed through DNA sequencing, explained in
2.2.1.3.
2.2.1.3 Transformation of chemically competent cells for DNA cloning Competent Top10 E.coli cells [in 75mM Cacl2/25% glycerol] were thawed on ice
for 15 min. After thawing, a 50µl aliquot was added to 10ng of plasmid DNA
(pCDH-MSCV-CTGF-E2A-GFP-PURO) in nuclease-free water, followed by a
further 30 min incubation on ice. Cells were then subjected to heat shock
transformation by incubation in a 42°C heat block for 30 sec, then returned to
ice and incubated for 2 min. Cells were added to 450µl of SOC medium [10mM
NaCl, 2.5mM KCl, 10mM MgCl2, 10mM MgSO4, 20mM glucose] over flame and
tubes were incubated for 60 min at 37°C with moderate shaking. At the
completion of the 60 min incubation, transformed cells were centrifuged for 5
min at 800x g and then resuspended in 100µl SOC. 50µl of transformed cells
were spread on prepared Luria broth [LB: 0.1% Bacto-tryptone (BD
Biosciences, San Jose, CA, USA), 0.05% Yeast extract (BD Biosciences, San
Jose, CA, USA), 170mM NaCl in distilled water (dH20)] agar plates containing
100µg/ml ampicillin (Sigma Aldrich, St.Louis, MO, USA). Agar plates were
incubated overnight at 37°C. The following day, single colonies were used to
inoculate 3ml of LB supplemented with 100µg/ml ampicillin. These cultures
were grown overnight at 37°C with vigorous shaking (250RPM) or for 6 hours
then diluted 1:500 into a conical flask with 50ml or 250ml LB supplemented with
100µg/ml ampicillin. The following day, glycerol stocks were prepared from
each culture by combining 100µl of culture with 100µl of 80% glycerol. These
stocks were frozen immediately at -80°C. The remainder of the overnight
culture was centrifuged in preparation for plasmid DNA extraction using the
QIAprep Spin Miniprep kit (Qiagen, Venlo, Limburg, Netherlands, #27104) or
Midi/Maxiprep HiSpeed plasmid kit (Qiagen, Venlo, Limburg, Netherlands,
#12643/12663) depending on volume and according to the manufacturer’s
instructions. Purified plasmid DNA was eluted in 30µl or 300µl of Tris-EDTA
45
(TE) buffer and stored at -20°C. Miniprep DNA was not used for lentivirus
production.
2.2.1.3 DNA sequencing to verify CTGF insertion into pCDH plasmid Sequencing was carried out using BigDye version 3.1 (Applied Biosystems, Life
Technologies, Carlsbad, CA, USA). Firstly, pCDH-MSCV-CTGF-E2A-GFP-
PURO plasmid DNA was amplified by polymerase chain reaction (PCR)
performed on a C1000 Thermal Cycler (Bio-Rad Laboratories, Hercules, CA,
USA). All PCR reactions were assembled on ice in sterile 0.2 ml thin-walled
tubes (Bio-Rad Laboratories, Hercules, CA, USA). Briefly, 75ng of plasmid
DNA was combined with 3.5µl of 5x BigDye reaction buffer, 9.3µl of ultrapure
water (Baxter, Deerfield, IL, USA), 3.2µl of 1 µM sequencing primer and 1µl of
BigDye v3.1. Four primers spanning the CTGF insert were used to sequence
the plasmid as indicated:
CD_1F (5'-3', AGTGCGACTCCACCCTCCA)
CD_1R (5'-3', GCTGGCAGCAGCTGGAG)
CD_3F (5'-3', GCTCCCTGCATCTTCGGT)
CD_3R (5'-3', CTCTTCCAGGTCAGCTTC)
Samples were denatured at 96°C for 1 min, followed by 25 cycles of 96°C for 10
s, 50°C for 5 s and 60°C for 4 min. Samples were precipitated in 1.5mL
Eppendorf tubes (Eppendorf, Hamburg, Germany) by adding 2µl sodium
acetate (3M) and 50µl 100% ethanol (EtOH). The solution was incubated at
room temperature for 30 min in the dark. After incubation, DNA was centrifuged
at 13kRPM for 20 min at 4°C. Supernatant was aspirated and 200µl of 70%
EtOH was pipetted on top of DNA pellet, pellet was centrifuged a further 10 min
at 13kRPM at 4°C. Supernatant was aspirated and the plasmid DNA pellet was
left to air dry before being resuspended in 10µl of Hi-Di injection buffer. The
solution was vortexed and spun briefly before being transferred to a 96-well
plate suitable for the 3130x Genetic Analyzer sequencing platform (Applied
Biosystems, Life Technologies, Carlsbad, CA, USA). The plate was then loaded
onto a 3130x Genetic Analyzer (Applied Biosystems, Life Technologies,
Carlsbad, CA, USA) for capillary sequencing. Sequencing data generated from
BigDye sequencing was analysed using MacVector software (MacVector, Cary,
NC, USA).
46
2.2.1 Linearisation of the pcDNA3.1 vector for electroporation The pcDNA3.1/eGFP vector was linearised with a BglII restriction enzyme
(Promega, Madison, WI, USA, #R6081) that cut distal to the CMV promoter
region. Digests were performed overnight at 37°C with Buffer D (Promega,
Madison, WI, USA). Digested DNA was cleaned up using the QIAquick PCR
purification kit (Qiagen, Venlo, Limburg, Netherlands, #28104) according to
manufacturer’s instructions. Plasmid DNA was then precipitated with EtOH.
Briefly, 2.5x 100% EtOH and 1.5x Sodium acetate (3M) were added to the
linearised pcDNA3.1/eGFP sample and incubated for 2 hours at -20°C.
Samples were then centrifuged at 13kRPM for 30 min at 4°C. Supernatant was
removed and 3 washes were performed in 500µl of 70% EtOH followed by
centrifugation at 13kRPM for 10 min at 4°C. The pellet was then left to air-dry,
resuspended in 30µl endotoxin-free TE buffer and quantitated before being
stored at -80°C.
2.2.2 Standard electroporation of B-precursor ALL cells Standard electroporation was performed with linear and supercoiled plasmid
DNA using the Bio-Rad gene pulser II (Bio-Rad Laboratories, Hercules, CA,
USA). Briefly, a cuvette was filled with 2x106 PER-278 cells and 5µg of linear or
supercoiled pcDNA3.1-eGFP plasmid DNA in 400µl of RPMI and left to stand
for 10 min at room temperature. The cells were then electroporated at high
resistance, 0.975mF capacitance and 0.3kV voltage and left to stand for 10 min
at room temperature. Then, foam residue was removed from the top of the
cuvette and cells were transferred to 5ml of normal media, seeded into 5-wells
of a 24-well plate and incubated at 37°C and 5%CO2.
2.2.3 Neon electroporation of B-precursor ALL cells Linearised or supercoiled pcDNA3.1-eGFP plasmid DNA was used in Neon
electroporation transfections (Neon transfection system, Life Technologies,
Carlsbad, CA, USA). Electroporation of 2x105 PER-278 cells with 1µg of
plasmid DNA was performed following manufacturer’s instructions. The
47
suspension of electroporated PER-278 cells was added to 500µl of pre-warmed
culture media in a 24-well plate and incubated at 37°C and 5%CO2.
2.2.4. Selection of stable cells transfected with pcDNA3.1 vector PER-278 cells were selected after electroporation, with 1.1µg/ml G418 (Sigma
Aldrich, St. Louis, MO, USA, #A1720). Optimal G418 concentration for PER-278
cell selection was determined by using serial dilutions of G418 and measuring
cell viability in an MTT assay, outlined in 2.6.1. For selection of cells that had
incorporated the neomycin resistance gene, G418 was added 48 hours after
electroporation, 250µl of culture media was replaced with 2.2µg/ml G418 in
normal media for a final concentration of 1.1µg/ml G418. Media was replaced
again at day 5 and 9 post-electroporation with 1.1µg/ml G418 in normal media
and selection was performed for a total of 21 days.
2.2.5 Lentiviral production Lentiviral particles were produced by transiently transfecting 293T cells to
express the HIV-1 plasmids. Briefly, 3x106 293T cells were plated in a 100mm
culture dish in 10mL normal media. Two ratios of lentiviral plasmids were used
to transfect 293Ts: Ratio1 [12µg plasmid DNA, 6µg gag-pol, 2µg rev-tat and
2µg VSV-G envelope] or Ratio2 [20µg plasmid DNA, 10µg gag-pol, 6µg rev-tat
and 5µg envelope]. These two ratios of plasmids were used to transfect 293Ts
by two methods, either using CaCl2 or Fugene (Fugene6, Roche, Basel,
Switzerland). The CaCl2 method involved adding 50µl of 2.5M CaCl2 to the
vector and packaging plasmid DNA solution with enough 0.1xTE buffer to make
the solution up to 500µl. To this solution, 500µl of 2xHBS was added [50mM
HEPES, 1.5mMNa2HPO4 and 140mM NaCl, pH7.5] and the solution was
vortexed and incubated at room temperature for 1 min. The DNA and CaCl2
solution was then added dropwise to 293T cells that had been covered with 9ml
of fresh normal media. For the Fugene method, vector and packaging plasmid
DNA was made up to 400µl with normal media and added to 32µl Fugene made
up to 200µl in normal media. The combined solutions were incubated at room
temperature for 15 min. The DNA and Fugene solution was then added
dropwise to 293T cells that had been covered with 8ml of fresh normal media.
48
For both methods of transfection, 293T cells were incubated at 37oC and
5%CO2 overnight and the next day media replaced with 5ml of fresh media that
was specific for target cell line intended for transduction. Media was collected
48 hours later and filtered through a 0.45µm filter (Sarstedt, Nümbrecht,
Germany) then frozen in aliquots at -80oC.
Lentiviral particles were titred on 293T cells. In 100mm culture dishes, 3x105
293T cells were plated with 5 ml of normal media in triplicate for each virus to
be titred. One spare plate was included, to determine cell count at transduction.
The next day, transduction was performed by addition of 3, 10 or 30µl of
lentiviral particles in 3ml normal media and 8µg/ml polybrene (Sigma Aldrich,
St. Louis, MO, USA) to 293T cells. After 3 hours at 37oC and 5%CO2, an extra
6ml of normal media was added per plate. The next day, media was replaced
with 9ml of fresh normal media. Transduction efficiency was measured 72 hours
after infection on trypsinised 293T cells by detection of the GFP marker present
on the pCDH plasmid (%GFP+ cells) using a LSRII flow cytometer (Becton
Dickinson, Franklin Lakes, NJ, USA). Lentiviral titre was calculated as an
average from all three volumes of viral particles using the following formula:
(cell count at transduction x %GFP+ cells) = Infective Units (IFU)/ml
(volume of lentiviral particles added (ml))
2.2.6 Lentiviral infection As lentiviral titres were below 1x107 IFU/ml, lentiviral particles were not diluted
for leukaemia cell line infection. To transduce PER-278 and PER-371 cells,
5x106 cells were resuspended in 1ml of undiluted lentiviral particles with 8µg/ml
polybrene. Cells were plated in one well of a 24-well plate and spun for 90 min
at 800x g at room temperature then incubated at 37oC and 5%CO2 for 90 min.
After incubation, cells were pelleted and resuspended in 5ml normal media to
be plated across 5 wells of a 24-well plate. Infected cells were then incubated at
37oC and 5%CO2 until selection with antibiotics, detailed in 2.2.7.
2.2.7. Selection of stable cells infected with pCDH vector Puromycin (Sigma Aldrich, St. Louis, MO, USA, #P8833) was used to select
infected PER-278 and PER-371 cells. Optimal puromycin concentrations were
49
determined by using serial dilutions of puromycin and measuring cell viability in
an MTT assay, outlined in 2.6.1. Following 48 hours after lentiviral infection,
250µl of culture media was replaced with fresh media containing 1.6µg/ml
puromycin for PER-278 (in normal media for a final concentration of 0.8µg/ml)
and 0.6µg/ml puromycin for PER-371 (in normal media for a final concentration
of 0.3µg/ml). Cells were then split (1:2) twice weekly with puromycin in normal
media. Successful production of stable cell lines was determined by detection of
GFP using a LSRII flow cytometer, when cells reached >95% GFP+ cells, they
were frozen down with 10%DMSO (v/v) in normal media, to be stored in liquid
nitrogen.
2.3 RNA analyses 2.3.1 RNA extraction RNA extraction was performed using a phenol-chloroform based protocol and
involved the addition of 1mL of TRIzol (Invitrogen, Life Technologies, Carlsbad,
CA, USA) to 1x107 pelleted cells. Samples were vortexed and incubated at
room temperature for 5 min. After the addition of 0.2ml of chloroform (AnalaR,
Merck) samples were mixed and separated by centrifugation at 11,000x g for 15
min at 4°C. The upper aqueous phase was collected and adjusted to 35% EtOH
by the addition of 100% EtOH. The resulting solution was purified using an
RNeasy Mini Kit (Qiagen) following the manufacturer’s instructions. Samples
were eluted in 30µl of nuclease-free water. After isolation all samples were
quantitated using the Nano Drop ND-1000 spectrophotometer (ThermoFisher
Scientific, Waltham, MA, USA) and stored immediately at -80°C.
2.3.2 cDNA synthesis First strand cDNA synthesis was performed using the Moloney murine
leukaemia virus (M-MLV) reverse transcriptase (Promega, Madison, WI, USA,
#M3682) according to the manufacturer’s instructions. RNA was first denatured
at 65°C for 5 min to reduce secondary structure and then snap chilled on ice for
a further 5 min. Reactions were assembled on ice and consisted of 1µg of total
RNA, 4µl of 5 x reaction buffer (Promega, Madison, WI, USA), 500µM dNTPs
50
(ThermoFisher Scientific, Waltham, MA, USA), 1µM OligodT primer, 10U of
RNase inhibitor, and 200U of MMLV RT enzyme (Promega, Madison, WI, USA,
#M3682). Reactions were adjusted to 20µl total volume with nuclease free H2O
mixed thoroughly and incubated at 37°C for 1 hour. After extension, specimens
were diluted 1 in 5 with nuclease free H2O and stored at -80°C until required.
2.3.3 Quantitation of CTGF gene expression Quantitative real-time PCR (qRT-PCR) was performed to measure expression
of CTGF mRNA in cell lines. CTGF expression was measured using the on-
demand assay Hs00170014_m1 (Applied Biosystems, Life Technologies,
Carlsbad, CA, USA). Reactions were performed in 384-well plates on an ABI
7900HT Fast Real-Time PCR System (Applied Biosystems, Life Technologies,
Carlsbad, CA, USA) with the following thermal profile, 50°C for 2 min then 95°C
for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. CTGF
reactions were prepared by combining the following components per well, 1µl of
CTGF assay, 10µl of universal PCR master mix, 8µl of dH2O and 1µl of first
strand cDNA. CTGF gene expression was normalised using an endogenous
control GAPDH assay (Applied Biosystems, Life Technologies, Carlsbad, CA,
USA) comprising 10µl of universal PCR master mix, 1µl GAPDH assay, 8µl of
dH2O and 1µl of cDNA. For the relative standard curve method of expression,
standards were prepared from cDNA generated from the PER-377 cell line. This
cDNA was not diluted after cDNA synthesis and thus was approximately 5 times
more concentrated than test specimens, which were diluted 1 in 5 immediately
after cDNA synthesis. Standards were prepared by serial dilution and assayed
in duplicate to control for pipette error, while test specimens were assayed in
triplicate. The relative expression for each assay was calculated using the
standard curves. For relative quantitation using the comparative Ct (∆∆Ct)
method, no standard curve was prepared. Briefly, the Ct values obtained from
High CTGF and Low CTGF cells were directly normalized to GAPDH and then
compared to each other. The fold change expression difference between the
two samples is calculated by 2^(-ΔΔCt).
51
2.3.4 Global gene expression to determine effect of CTGF on stable CTGF-overexpressing leukaemia cell lines RNA was analysed for suitable purity for downstream microarray analysis using
the Agilent 2100 Bioanalyzer (Agilent Technologies). RNA samples from
lentiviral-infected cell lines were analysed using the RNA 6000 Nano kit (Agilent
Technologies, #5067-1511), following the manufacturer’s instructions.
Electropherogram and RNA integrity number were used for quality control.
Samples were then sent for processing on GeneChip Human Gene 1.0 ST
Arrays (Affymetrix, Santa Clara, CA, USA) by the Ramaciotti Centre for Gene
Function Analysis (The University of New South Wales, Sydney, Australia).
Analysis of microarray data was performed by Denise Anderson (Biostatistician,
Telethon Institute for Child Health Research, Perth, WA, Australia), including
robust multi-array average (RMA) normalization and linear models for
microarray data (LIMMA) analysis using the R platform (www.r-project.org).
2.3.5 Global gene expression to determine effect of recombinant human CTGF on stromal cells
Recombinant human (rh)CTGF was kindly donated by Prof David. R. Brigstock
(Children’s Research Institute, Colombus, OH, USA). All technical steps
required for stromal cell gene expression microarrays were performed by Jette
Ford in the laboratory of Prof Ursula Kees, Telethon Institute for Child Health
Research, Perth, Western Australia. HS5 cells were incubated with 100ng/ml
rhCTGF or vehicle control (control buffer) for 8,16 or 48 hours. There were 3-4
experimental and control replicates at 8 and 16 hours, and 2 at 48 hours. For
hybridization to the HG-U133A microarrays (Affymetrix, Santa Clara, CA, USA)
cRNA was prepared following an adapted Affymetrix protocol as described
previously (Boag, et al., 2006). Briefly, RNA was reverse transcribed into double
stranded DNA and was then used to synthesize biotinylated cRNA, with the use
of a Bioarray High Yield RNA Transcript Labelling Kit (Enzo Life Sciences,
Farmingdale, NY, USA). Biotinylated cRNA was fragmented and hybridized to
the microarray and then washed and stained before scanning (GeneArray
Scanner, Agilent Technologies, Santa Clara, CA, USA) in accordance with
52
Affymetrix protocols. Scanned images met stringent quality controls, having
scaling factors of <20, noise values below three and the signal ratios of the
housekeeping genes ACTB and GAPDH were <3 for all arrays. Principle
component analysis (PCA) was used to determine similarity of gene expression
patterns between the CTGF-treated and control HS5 samples.
Microarray analysis was performed by Julia Wells. Data were normalized by
RMA using GeneSpring (Agilent Technologies, Santa Clara, CA, USA).
Statistical analysis was performed using LIMMA and p-values were adjusted for
multiple testing using the Benjamini and Hochberg method for false discovery
rate (Benjamini & Hochberg, 1995; Smyth, 2004, 2005) on the R platform
(www.r-project.com). Data were analysed using Ingenuity Pathway Analysis
(IPA, Ingenuity, Redwood City, CA, USA). IPA was employed to gain
information about networks, canonical signalling pathways and biological
functions in CTGF-treated HS5 cells. The input dataset comprised genes from
the 8 hour time point showing a ≥2-fold difference compared to control cells,
and statistical difference in expression as determined by LIMMA p-value. A list
of secreted genes from the IPA input dataset were found by comparing the
dataset to that of the Secreted Protein Discovery Initiative (Clark, et al., 2003)).
The extracellular function of these secreted genes was determined manually
using GeneCards (Rebhan, et al., 1998)).
Validation of four genes from the microarray analysis was performed using qRT-
PCR, as described in section 2.3.3. Briefly, RNA was extracted from HS5 cells
after incubation in 100ng/ml rhCTGF for 8 hours as per the protocol described
in section 2.3.1, cDNA was then synthesized as per the protocol described in
section 2.3.2. Lastly, qRT-PCR was performed using the relevant Taqman
assays (Applied Biosystems, Life Technologies, Carlsbad, CA, USA) for
Collagen, Type 3, alpha 1 (COL3A1, #Hs00943809_m1), Lumican (Dornhofer,
et al., Hs00929860_m1), Lysyl Oxidise (LOX, #Hs00942480_m1) and Integrin,
alpha V (ITGAV, #Hs00233808_m1) all normalized to GAPDH. Gene
expression levels were quantified using the comparative Ct method as
described in section 2.3.3.
53
2.4 Protein analyses Protein expression of CTGF was detected with two antibodies. The L-20 anti-
CTGF antibody is an affinity purified goat polyclonal antibody raised against a
peptide mapping within an internal region of CTGF of human origin (Santa Cruz
Biotechnology, Santa Cruz, CA, USA, #sc-14939). The H-55 anti-CTGF
antibody is a rabbit polyclonal antibody raised against amino acids 295-349 of
CTGF of human origin (Santa Cruz Biotechnology, #sc-25440). A summary of
primary, secondary and tertiary antibodies used in protein analyses are listed in
Table 2.1.
2.4.1 Cellular protein extraction Total protein was isolated from cell lines for analysis by western blotting. Cells
were pelleted by centrifugation and washed twice with ice-cold PBS. Cell pellets
were resuspended in one pellet volume of ice-cold PBS and then lysed using
100µl/1x107 cells of Evan’s protein lyses buffer [1mM EDTA, 1mM ethylene
glycol tetraacetic acid (EGTA), 1%NP-40, 50mM Tris-HCl, 120nM NaCl and
peptide inhibitors], and incubated on ice for 30 min. Samples were centrifuged
for 10 min at 13kRPM at 4°C and supernatants were transferred to fresh 1.5 ml
tubes. Protein samples were stored at -20°C or used immediately in
downstream assays.
Protein concentration in cell line extracts was determined using a bicinchoninic
acid (BCA) protein assay (Pierce, ThermoFisher Scientific, Waltham, MA, USA)
according to manufacturer’s instructions. Samples were diluted 1 in 10 and
together with standards [10µg/ml to 5mg/ml serial dilution of BSA], were added
to a 96-well flat-bottomed plate at 10µl per well. Standards were assayed in
duplicate, while protein samples were assayed in triplicate. After addition of
samples and standards, the plate was incubated at 37oC for 30 min, and the
absorbance was measured at 562nm on a Synergy plate reader (BioTek,
Winooski, VT, USA). Protein concentration in cell line extracts was calculated
using the standard curve.
54
2.4.2 Secreted protein purification
Heparin-affinity enrichment of conditioned culture medium was necessary to
detect secreted CTGF. Cells were seeded in fresh culture medium at a density
of 1x106 cells per ml in 24-well plates (Nunc, Thermo Fisher Scientific, Waltham,
MA, USA), and grown for 24 hours prior to isolation of conditioned medium
(CM). Cells and CM were collected and centrifuged at 800x g for
5 min and 5ml of supernatants comprising the CM were then transferred to
fresh tubes containing 5ml of ice-cold PBS together with 50µl of heparin
agarose beads (Sigma Aldrich, St. Louis, MO, USA, #H6508). The solution was
then incubated for 24 h at 4°C on a fly-wheel. After 24 hours, beads were
collected by centrifugation at 800x g for 5 min at 4°C, then supernatant was
carefully removed by vacuum aspiration. Beads were washed twice with ice-
cold PBS and centrifugation at 800x g for 5 min at 4°C. After the second wash,
beads were resuspended in 1ml ice-cold PBS and transferred to 1.5 ml tubes.
Beads were collected by centrifugation at 800x g for 3 min at 4°C and the
supernatant was carefully aspirated, leaving a 50µl bead slurry. Bound protein
was prepared for western blot analysis by adding 19.25µl of 4x LDS sample
loading buffer (Invitrogen, Life Technologies, Carlsbad, CA, USA) and stored at
-80oC. After thawing, 7.7µl of 10x Denaturing agent (Invitrogen, Life
Technologies, Carlsbad, CA, USA) was added to the sample, which was
denatured at 70oC for 10 min. Samples were then quenched on ice for 2 min
and used for western blot.
2.4.3 Western blot Electrophoresis of NuPAGE Novex BisTris mini gels using the XCell surelock
Mini-Cell was carried out under manufacturer’s instructions (Invitrogen, Life
Technologies, Carlsbad, CA, USA). Electrophoresis using the mini-cell was
conducted using 4-12% BisTris pre-cast gels and MES running buffer with
antioxidant. Prior to electrophoresis, quantitated cellular protein samples were
combined with 4x LDS sample buffer and 10x Reducing agent and heated at
70oC for 10 min. Samples were then quenched on ice for 2 min then loaded
onto the pre-cast 4-12% Bis-Tris gel with a Precision Plus Protein Kaleidoscope
standards ladder (Bio-Rad Laboratories, Hercules, CA, USA). Gels were run at
anti-Annexin V-PE
Primary Antibody Dilution Supplier
monoclonal mouse anti-Actin pan Ab-5
anti-BrdU-APC
mouse anti-CD19-APC
polyclonal rabbit anti-CTGF (H-55)
monoclonal mouse anti-Ki67
monoclonal rabbit anti-Pax5
1:50 FC BD (#556421)
1:5000 WB
1:50 FC
1:20 FC
Santa Cruz (#sc-25440)
1:800 WB1:100 E
1:100 IH Cell Signaling (#12709)
1:100 IH BD (#550609)
Table 2.1 List of primary, secondary and tertiary antibodies . Summary of the antibody name, dilution and supplier (with catalogue number). A dash indicates conjugation of the antibody to a fluorescent dye for flow cytometry or enzyme label for peroxidase staining. FC, flow cytometry, WB, western blot, E, elisa, IH, immunohistochemistry.
polyclonal rabbit anti-VwF 1:400 IH DAKO (#A0082)
Santa Cruz (#sc-14939) polyclonal goat anti-CTGF (L-20)
BD (#552598)
BD (#555415)ThermoFisher Scientific (#MS-1295)
goat anti-rabbit-Biotin
1:5000 E
1:10000 IH Vector Labs (#PK-6101)
polyclonal rabbit anti-goat-Biotin 1:5000 WB DAKO (#E0466)
Secondary/Teriary Antibody Dilution Supplier
GE Healthcare (#NA934)
polyclonal donkey anti-rabbit-HRP
HRP-streptavidin 1:3000 WB DAKO (#P0397)
1:800 WB1:100 E
1:10000 WB GE Healthcare (#NA9310)
polyclonal donkey anti-mouse-HRP
56
145V for 60 min. Gels were transferred to mini nitrocellulose membranes (Bio-
Rad Laboratories, Hercules, CA, USA, #170-4158) using a semi-dry transblot
turbo (Bio-Rad Laboratories, Hercules, CA, USA).
Membranes were then washed in TBST [1xTris buffered saline, TBS [50mM
Tris-Cl and 150mM NaCl; pH7.5 with 0.1% Tween 20] for 3x 5 min, before being
blocked at room temperature for 1 hour with 5% skim milk powder in TBST, with
gentle agitation. After blocking, membranes were washed in TBST for 3x 5 min
and then incubated in 10mL primary antibody (anti-CTGF L-20, 1:800 dilution)
in diluent [TBST with 1% skim milk powder] overnight at 4oC with gentle
agitation. The next day, membranes were washed in TBST for 3x5 min and then
incubated in secondary antibody (rabbit anti-goat, 1:5000) in diluent for 1 hour
at room temperature with gentle agitation. Membranes were washed in TBST
for 3x 5 min and then incubated in HRP-Straptavidin (1:3000) in diluent for 1
hour at room temperature with gentle rocking. Membranes were washed in
TBST for 3x 5 min and then bound proteins were detected using an Immuno-
Star WesternC chemiluminescence HRP detection kit (Bio-Rad Laboratories,
Hercules, CA, USA), according to manufacturers instructions. Nitrocellulose
membranes were imaged on a Chemidoc MP imaging system (Bio-Rad
Laboratories, Hercules, CA, USA) at intervals until detected bands were
saturated.
Where it was necessary, membranes were stripped in stripping buffer [50 mM
Tris (pH 6.8), 2% SDS, 100 mM 2-mercaptoethanol] at 50°C for 30 min with
gently agitation, then washed three times in TBST for 5 min at room
temperature. Membranes were blocked and re-probed for anti-pan-Actin as a
loading control (1:5000, ThermoFisher Scientific, Waltham, MA, USA).
Secondary antibody was goat anti-mouse (1:10,000) and proteins detected as
previously described.
2.4.4 Enzyme-linked immunosorbent assay PER-278 Low, PER-278 High, PER-371 Low and PER-371 High were seeded
at 1.5x106 cells per well in fresh normal medium and incubated for 72 hours at
37oC and 5%CO2. Conditioned media was collected and cleared of cell debris
by centrifugation at 800x g for 5 min. After centrifugation, 5 ml of conditioned
medium was added to 5ml of 1xPBS with 50µl of Heparin-agarose beads, and
57
the solution was incubated overnight on a fly-wheel at 4oC. The next day, the
solution was centrifuged at 800x g for 10 min at 4oC. Pellets were then washed
twice with 10ml Tris-HCl (10mM). Pellets were transferred to 1.5ml tubes and
washed with 1ml Tris-HCl (10mM). Protein was eluted from beads with 100µl
buffer [10mM Tris-HCl; 1.5M NaCl] and centrifuged at 1000x g for 3 min.
Supernatants were transferred to fresh 1.5ml tubes to be stored at -20oC for the
enzyme-linked immunosorbent assay (ELISA).
For ELISA, 96-well flat-bottom plates (Nunc, ThermoFisher Scientific, Waltham,
MA, USA) were coated with 2g/ml L-20 CTGF antibody. Plates were then
incubated for 24 hours at 4oC, flick-rinsed and patted dry on absorbent towel.
After coating plates with the CTGF capture antibody, wells were filled with 200µl
of blocking solution [1xPBS with 1% BSA and 5% sucrose]. Incubation with
blocking solution was performed for 2 hours at room temperature with gentle
agitation. Wells were then washed twice with PBST. In triplicate wells, 50µl
sample diluent [1xPBS with 1% BSA] and 50µl conditioned media sample was
added and in duplicate wells standards were added (0 to 0.5µg/ml rhCTGF
serially diluted across 12 standards). Plates were then incubated overnight at
4oC with gentle agitation. The next day, plates were washed four times with
PBST and 100µl of primary H-55 CTGF antibody was added to wells in diluent
solution to a concentration of 2µg/ml. Plates were then incubated at room
temperature for 2 hours with gentle agitation. After incubation, plates were
washed four times with PBST and then 100µl of secondary detection antibody
was added to wells (donkey anti-rabbit-HRP, 1:5000 in diluent solution). Plates
were then incubated at room temperature for 1 hour with gentle agitation. Plates
were then washed four times with PBST and 100µl of 3,3’,5,5’-
tetramethylbenzidine (TMB) reagent (Sigma Aldrich, St. Louis, MO, USA) was
added to each well. Plates were incubated at 37oC for 15 min then colour
development was stopped by addition of 100µl HCl (1M). Plate absorbency was
read at 450nm on a Synergy plate reader. Protein concentration in each
conditioned media extract was calculated using the standard curve.
Experiments were performed in triplicate.
2.4.5 Protein Array For detection of differential protein secretion from PER-371 High and PER-371
58
Low cells, a label-based L-series human antibody array with membranes L-507
and L-493 was used (Ray Biotech, Norcross, GA, USA, #AAH-BLM-1000-4),
according to manufacturer’s instructions. HS5 cells were plated at 0.25x106
cells/ml in 7ml of normal media and left for 48 hours at 37oC and 5%CO2. After
incubation, either 1x106 cells/ml of PER-371 Low or PER-371 High cells, was
added to HS5 cells in 7ml of low-serum media (0.2%FCS). A control flask of
HS5 cells alone was incubated with low-serum media. HS5 and leukaemia cells
were co-incubated for 48 hours, then conditioned media was filtered through a
0.2µm nylon filter and 2x 3ml samples were added to dialysis vials. Dialysis
vials were placed in a foam holder inside a beaker containing 500ml of cold
PBS and left overnight at 4oC with gentle stirring. The next morning PBS was
replaced and dialysis vials were incubated a further 3 hours. Dialysed samples
were transferred to tubes and centrifuged at 1000x g for 15 min, and then
supernatants were combined from the two dialysed samples. The primary
amines of the proteins in the samples were then labelled with biotin. The
membrane arrays were incubated with 7.2µl/mg of 1x Labelling Reagent for 30
min at room temperature then filtered according to instructions. Biotinylated
samples were incubated with the membrane array to allow for interaction
between the sample proteins and the pre-printed capture antibodies for 1 hour
at room temperature with gentle agitation. Membranes were then incubated with
HRP-conjugated streptavidin and visualized by chemiluminescence, similar to
western blots (2.4.3).
Densitometry readings for duplicate protein dots per membrane were taken
using Bio-Rad Image Lab software (Bio-Rad Laboratories, Hercules, CA, USA).
Readings were discarded if signal intensity for dots was below background
levels. Normalistaion of intensity values above background were made in ratio
to positive controls within the array. Normalised signal intensities were then
compared across the three experimental conditions for each detectable protein,
PER-371 High+HS5, PER-371 Low+HS5 and HS5 alone.
2.5 Proliferation assays Proliferation of cells was determined in the presence of increasing
concentrations of IL7 (0-1ng/ml) in standard or low serum culture media (0.2%
FCS). The four lentiviral-infected B-precursor ALL cell lines (PER-278 High/Low
59
and PER-371 High/Low) were plated at 1x105 cells per well for each IL7
experimental condition and 0.5x105 cells per well in low serum, in triplicate, in a
96-well round bottom plate (Nunc, ThermoFisher Scientific, Waltham, MA,
USA). Cells were incubated in the presence of low serum or IL7 for 48 hours,
0.5µCi of Thymidine, [methyl-3H] (PerkinElmer, Waltham, MA, USA, #NET027A)
was then added and cells were incubated at 37oC and 5%CO2 for 3 hours in low
serum or 24 hours with IL7. Cells were then lysed by freezing at -80oC. For
processing, cells were thawed and DNA was harvested onto a glass fibre
filtermat (PerkinElmer, Waltham, MA, USA, #1450-421) using a FilterMate
vacuum harvester (PerkinElmer, Waltham, MA, USA). The filtermat was dried at
60oC for 60 min, then sealed in a sample bag (PerkinElmer, Waltham, MA,
USA) with liquid scintillation cocktail fluid (PerkinElmer, Waltham, MA, USA,
#0013320). Radioactivity was assessed by counts per minute using a 1450
MicroBeta TriLux (PerkinElmer, Waltham, MA, USA). Experiments were
performed in triplicate.
Additional PER-371 proliferation assays were performed as above, but with the
exception of cells being suspended in conditioned media from HS5-leukaemia
cell co-cultures. Briefly, 2x106 HS5 cells in 2ml of normal media were plated on
a 6-well flat-bottomed plate (Nunc, Life Technologies, Carlsbad, CA, USA).
Cells were then left to adhere overnight. The next day, media was replaced with
either 2ml fresh normal media, fresh normal media with 100ng/ml rhCTGF,
2x106 PER-371 Low cells in normal media or 2x106 PER-371 High cells in
normal media. Plates were incubated for 72 hours and then conditioned media
was extracted, filter-sterilised and stored at -20oC until needed.
2.6 Drug sensitivity assays 2.6.1 Optimisation of antibiotics for stable cell line selection Optimisation of antibiotic concentration of G418 for the pcDNA3.1 vector and
puromycin for the pCDH vector were performed using MTT assays on a 96-well
flat bottom plate (Nunc, ThermoFisher Scientific, Waltham, MA, USA). Briefly,
the G418 titration curve contained 12 dilutions, each in triplicate, ranging from
0.5 to 1.6µg/ml in 50µl of normal media. The puromycin titration curve contained
60
15 dilutions, each in triplicate, ranging from 0.05 to 1.5µg/ml in 50µl of normal
media. To each drug dilution, 1.5x105 PER-371 or PER-278 cells were added in
50µl of normal media. In addition, each plate contained 3 wells of 100µl normal
media as a negative control and leukaemia cells with no antibiotic as a positive
control. Plates were incubated for 3 days at 37oC and 5%CO2. On the third
day, 10µl of filtered 5mg/ml (3-(4 5-dimethylthiazol-2-yl)-2 5-diphenyltetrazolium
bromide (MTT, Sigma Aldrich, St. Louis, MO, USA) in 1xPBS was transferred to
each well. MTT was incubated for 6 hours at 37oC and 5%CO2, after which
100µl of 4% isopropanol was added to each well and resuspended to ensure all
formazon crystals were solubilised. The plate was then read at 495nm
absorbance on a Synergy plate reader (BioTek, Winooski, VT, USA). Drug
concentrations for selection of stable cell lines were determined as the
concentration required to kill 95% of cells in 3 days.
2.6.2 Sensitivity to chemotherapy drugs in CTGF-High and CTGF-Low cell lines Sensitivity to the cytotoxic drugs vincristine (VCR, Pfizer, New York City, NY,
USA) and cytarabine (1-β-d-arabinofuranosylcytosine, ARAC, Pfizer, New York
City, NY, USA) was determined in vitro using the Alamar Blue assay (Life
Technologies, Carlsbad, CA, USA), according to manufacturer’s instructions.
Production of HS5-leukaemia cell conditioned media was performed as before,
see 2.5. Alamar Blue assays were set up using 96-well flat-bottomed plates
(Nunc, ThermoFisher Scientific, Waltham, MA, USA). First, serial dilutions of
ARAC (2µg/ml to 0.5ng/ml) or VCR (0.2µg/ml to 0.05ng/ml) were plated in
normal media, in triplicate wells. Cells were added at 1.5x105 cells in 50µl of
normal media or conditioned media from HS5-leukaemia cell co-cultures. Plates
were then incubated for 72 hours before addition of 10µl Alamar Blue. After a 3-
hour incubation, absorbance was read at 570nm and 600nm on a Synergy plate
reader (BioTek, Winooski, VT, USA). The EC50 (drug concentration that inhibits
cell growth by 50%) was determined by calculating the percentage difference
between treated and untreated cells for each drug concentration.
61
2.6.3 Nucleoside transporter activity in PER-278 High or Low cells ARAC acts as a nucleoside analogue and inhibits DNA polymerase, primarily
killing cells undergoing DNA synthesis (S phase). Nucleosides must be taken
up by cells through one of two distinct families of human nucleoside
transporters: cumulative nucleoside transporters (CNTs) which are sodium-
dependent or equilibrative nucleoside transporters (ENTs) which are sodium-
independent. Uridine, a nucleoside, is taken up by cells through both
transporters. ARAC uses predominantly ENT1 in human leukaemic blast cells
(Hubeek, et al.). By measuring uptake of tritiated uridine (5,6-3H-uridine) into
cells, we obtained an approximate measure of nucleoside transporter activity.
The uridine uptake assay was based on methods previously published by
Macanas-Pirard et al. , with the exception of using nitrobenzylmercaptopurine
ribonucleoside (NMBPR, ENT1 inhibitor) PER-278 Low cells were plated at
4x105 cells per well in 50µl of normal media, in a 96-well conical-bottom plate
(Nunc, ThermoFisher Scientific, Waltham, MA, USA). 8µCi of 5,6-3H-uridine
(PerkinElmer, Waltham, MA, USA, #NET367001MC) was added to wells with
50µl of normal media, normal media with rhCTGF (100ng/ml in normal media)
or conditioned media from co-culture of HS5 cells with lentiviral-infected PER-
278 cell lines. After 15 min, cells were washed with 100µL of wash buffer
containing 1xPBS with 1mM uridine (Sigma Aldrich, St. Louis, MO, USA,
#U3003-5G). The plate was centrifuged at 1200RPM for 5 min at room
temperature to pellet cells, and washed 4 times with wash buffer. Cells were
resuspended in 50µl of wash buffer and then frozen at -80°C. 3H-uridine
incorporation was measured using a TRI-CARB 1500 liquid scintillation counter
(Packard Instrument Company, Canberra Industries, Meriden, CT, USA),
whereby cells were added to a scintillation vial with 2ml of Ultima Gold liquid
scintillation counter cocktail (PerkinElmer, Waltham, MA, USA, #0013320). 3H-
uridine incorporation was measured in counts per minute.
62
2.7 NOD/SCID mouse xenografts
2.7.1 Mice Non-obese diabetic/ severe-combined immunodeficient (NOD/SCID) mice were
obtained from the Animal Resources Centre (Perth, WA, Australia) and
randomly assigned to experimental groups. The age of the mice at injection of
leukaemia cells ranged from 6 to 8 weeks of age. Animals were housed in
individually ventilated cages (IVC), on a 12-hour light/dark cycle and with
unlimited access to water and food including regular mushy chow. All animal
experimentation was approved by the Animal Ethics Committee of the Telethon
Institute for Child Health Research and conformed to NHMRC guidelines, n=
approx. 7 animals per group were analysed, as detailed in each figure legend.
2.7.2 Generation of NOD/SCID xenografts The ALL xenograft model has previously been established for use as a
preclinical model for therapy evaluation. NOD/SCID mice proved to have a
higher efficiency of engraftment and provide an accurate mirroring of human
paediatric ALL with the immunophenotype corresponding to the primary patient
sample (Liem, et al., 2004; Lock, et al., 2002). RNA and protein samples were
taken from PER-278 and PER-371 cell lines on the day of injection for
quantification of CTGF expression in RNA, protein and conditioned media by
qRT-PCR or western blot.
Xenograft studies were carried out using female NOD/SCID mice, however due
to availability in some instances male mice were used. Only groups of the same
sex were compared to each other. Female mice were sub-lethally irradiated
(250cGy) on the same day as injection, while male mice were injected without
irradiation. Mice were kept on 375mg Enrofloxacin (Bayer, Leverkusen,
Germany) and 165mg Amoxicillin (Sandoz, Princeton, NJ, USA) per 500ml
drinking water, for 4 weeks post-irradiation.
For PER-278 xenografts, female mice were used for homing, early stage, end
stage and chemotherapy experiments. For the snapshot of disease 24 hours
63
post-ARAC injection male mice were used. All mice were tail-vein injected with
2x105 PER-278 Low or PER-278 High cells. End-stage data were analysed
from three experimental groups of mice (n=7). Homing, early stage,
chemotherapy response and post-ARAC time point data were analysed from
one experimental group of mice (n=7).
For PER-371 xenografts, female mice were used for end stage and
chemotherapy experiments. Male mice were used for homing, early and late
stage disease time points. Mice were injected with 1x106 PER-371 Low or PER-
371 High cells. End-stage data were analysed from two experimental groups of
mice (n=7). Homing, early stage and late stage time point data were analysed
from one experimental group of mice (n=7).
2.7.3 Chemotherapy regimen An induction phase drug regimen was administered to determine if CTGF
affects drug efficacy. Blood samples were used to determine levels of
leukaemic cells in the blood. When a mouse in the group had 1% leukaemic
engraftment in the blood, treatment was initiated for the whole group.
Chemotherapy was delivered as a single-agent using either vincristine (VCR,
Pfizer, New York, NY, USA) or cytarabine (ARAC, Pfizer, New York, NY, USA).
Drug concentration and regimen for maximally tolerated doses were determined
by Prof. Richard Lock (via correspondence, Childrens Cancer Institute Australia,
Sydney, NSW, Australia). Administration of VCR consisted of intraperitoneal
injections (0.5mg/kg) weekly for 4 weeks. Administration of ARAC consisted of
intraperitoneal injections (either 0.5mg/kg or 100mg/kg) daily for 5 days, 2
weeks off, then repeat of cycle. The higher dose for cytarabine represented the
clinically-relevant maximum tolerated dose. The lower dose facilitated detection
of drug resistance in the cells. Control mice were intraperitoneally injected using
the same regimen but with sodium chloride solution (0.9% sterile, Pfizer, New
York, NY, USA). For the PER-278 mouse cohort culled 24 hours post-ARAC
injection, this was done 24 hours after the first ARAC injection in the second
round of treatment.
64
2.7.4 Mouse dissection and tissue preparation Mice were culled at specified time points or when they presented with two or
more identifiable leukaemic symptoms, including hind leg paralysis, weight loss,
ruffled fur, decline in activity, rapid breathing and hunched posture. At these
times, mice were anaesthetized with methoxyflurane (dosage to effect, Medical
Developments International, Springvale, VIC, Australia) and underwent a
cardiac puncture before being euthanised by cervical dislocation. Harvested
organs were halved and allocated to 4% Paraformaldehyde for immunostaining
or to 5%FCS in PBS (FACS buffer) for analysis by flow cytometry. Organs
extracted were hind-leg femurs and tibias, spleen, liver and lymph nodes. Blood
was collected in microtainer tubes with K2EDTA (BD Biosciences, San Jose,
California, USA).
2.7.5 Tissue analysis by flow cytometry A summary of primary and secondary antibodies used in flow cytometry are
listed in Table 2.1.
Bone marrows were flushed with a 19-gauge needle and ends of bones were
crushed. Flushed-bone marrow, spleens, livers and lymph nodes were filtered
through 100µm nylon cell strainers (Falcon, Corning Inc, Corning, NY, USA)
with ample FACS buffer to obtain single cell suspensions. Single cell
suspensions were then centrifuged at 800x g for 5 min at 4oC. Supernatant was
aspirated and pellets were resuspended in 5ml of RBC lysis buffer
(eBioscience, Affymetrix, San Diego, CA, USA). Mouse blood was also lysed,
by adding 80µl of blood to 1ml of red blood cell (RBC) lysis buffer (eBioscience,
San Diego, CA, USA). Cell suspensions were incubated at room temperature
for 5 min with occasional shaking. The reaction was stopped with 20ml (tissue)
or 10ml (blood) of 1x ice-cold PBS. Cells were then centrifuged at 800x g and
4°C. Cell pellets were resuspended in FACS buffer and analysed by flow
cytometry or quantitated using the Vicell for use in proliferation or apoptosis
assays.
Engraftment of GFP+ leukaemic cells was then determined using a LSRII flow
65
cytometer (Becton Dickinson, Franklin Lakes, NJ, USA) running FACSDiva
software (Becton Dickinson, Franklin Lakes, NJ, USA) and using the blue laser
(488nm excitation, 530/30 band pass filter). Analysis was performed using
FlowJo (Tree Star Inc, Ashland, OR, USA). Cells were gated in hierarchy, first
gating live cells with a forward and side scatter plot (FSC vs SSC), then gating
out doublets using a side scatter area and width plot (SSC-A vs SSC-W), then
lastly gating GFP+ cells (GFP vs APC), as shown in Figure 2.1. As a control
representative bone marrow cells were stained for anti-human CD19 (BD
Biosciences, San Jose, CA, USA) by incubating 1x106 cells (after RBC lysis) in
5µL anti-CD19 and 100µL FACS buffer for 30 minutes in the dark on ice. Cells
were then washed in 1ml of 1x PBS and centrifuged at 800x g for 5min at 4°C.
Pellets were resuspended in 500µl FACS buffer for analysis by flow cytometry
using the red laser (633nm excitation, 660/20 band pass filter). GFP+ cells were
shown to overlap with CD19+ stained cells, verifying that the GFP fluorescence
was exclusive to injected leukaemia cells (data not shown).
Proliferation and apoptosis levels at time points of late stage disease were
determined using the BrdU:APC Flow Kit (BD Biosciences, San Jose, CA, USA,
#557892) or the Annexin V: PE apoptosis detection kit I (BD Biosciences, San
Jose, CA, USA, #559763), both used according to manufacturer’s instructions
and detected with the red laser (633nm excitation, 660/20 band pass filter).
For BrdU staining, mice were injected with 100µl of 10mg/ml BrdU (1mg)
intraperitoneally, 24 hours prior to sacrifice. For intra-cellular staining, 2x106
bone marrow cells were resuspended in 100µl BD Cytofix/Cytoperm buffer and
incubated at room temperature for 15 min. Cells were then washed with 1ml
Perm/Wash buffer and centrifuged 300x g for 5 min. Pellets were resuspended
in 100µl BD Cytoperm Plus buffer and incubated on ice for 10 min. Cells were
then washed in Perm/Wash buffer and resuspended in 100µl of BD
Cytofix/Cytoperm buffer for 5 min at room temperature. Cells were again
washed in Perm/Wash buffer, then resuspended in 50µl DNase (300µg/ml) and
incubated for 1 hour at 37oC. Cells were washed in Perm/Wash buffer and then
split into two aliquots, one resuspended in 50µl of anti-BrdU-APC (1:50 in
Perm/Wash buffer) or 50µl of Perm/Wash buffer only. Cells were then incubated
for 20 min in the dark at room temperature. After antibody incubation cells were
washed in Perm/Wash buffer and resuspended in 300µl FACS buffer to be
66
analysed by flow cytometry. For Annexin V staining, bone marrow cells were washed twice with cold PBS
and then resuspended in 1ml of 1xBinding buffer at a concentration of 1x106
cells/ml. An aliquot of 100µl of the solution was added to 5µl of PE Annexin V,
cells were gently vortexed and incubated at room temperature for 15 min in the
dark. Cells were washed with 400µl of 1x Binding buffer, resuspended in 300µl
FACS buffer and analyzed by flow cytometry within the hour.
2.7.6 Histochemical and immunohistochemical (IHC) staining A summary of primary and secondary antibodies used in immunostaining are
listed in Table 2.1.
2.7.6.1 Tissue Processing Hind leg femurs and tibias were fixed in freshly thawed aliquots of 4%
paraformaldehyde (Sigma Aldrich, St. Louis, MO, USA, #P6148) in 1x PBS.
Tibias and femurs were immersed immediately in 5ml of 4% Paraformaldehyde
and then stored at 4oC for 24 to 48 hours, before being washed twice in 1x PBS
and stored at 4oC until decalcification. Bones were decalcified in 10%
ethylenediaminetetraacetic acid (EDTA, Sigma Aldrich, St. Louis, MO, USA,
#E9884) at 4oC for 8 days with gentle agitation. After decalcification bones were
dehydrated through a series of graded alcohols to xylene and infiltrated with
paraffin using a Leica ASP200 S tissue processor (Leica microsystems,
Wetzlar, Germany).
Paraffin-infiltrated bones were embedded into wax blocks. Tissue blocks were
stored at room temperature until sectioned with a Leica RM2135 microtome
(Leica microsystems, Wetzlar, Germany). Tissue was cut into 4µm sections and
adhered to SuperFrost Plus slides (ThermoFisher Scientific, Waltham, MA,
USA). Individual slides contained two sections from one mouse block, that
encompassed a longitudinal section through mid-bone to obtain the highest
amount of bone marrow. Slides were left to dry overnight at room temperature
and stored at room temperature until ready to use. For use in
immunohistochemistry and histology stains, slides were first melted at 60oC for
15-30 min.
+Figure 2.1. Leukaemia engraftment was determined by GFP cell detection using flow cytometry. Gating (depicted by red line) was performed in hierarchy for the following populations: a. Live cells on a forward scatter and side scatter plot. b. Single cells on a side scatter-area and side scatter-width
+plot. c. GFP cell on a FITC by APC plot.
b.
a.
c.
F
ITC
-A
APC-A
SSC-A
FSC-A
SS
C-A
SS
C-W
68
2.7.6.2 Immunohistochemistry
Immunohistochemistry was used to detect leukaemia cells (anti-Pax5)
proliferating cells (anti-Ki67) or megakaryocytes (anti-VwF) with antibodies and
concentrations given in Table 2.1.
Melted slides were deparaffinised using a Leica Autostainer XL (Leica
microsystems, Wetzlar, Germany). The protocol involved incubating 3x 3 min in
xylene (Hurst Scientific, Perth, WA, Australia), 2x 2 min in 100% EtOH, 1 min
95% EtOH, 1 min 70% EtOH, 1 min 40% EtOH and 5 min running tap water.
After rehydration slides were kept from drying out. Rehydrated slides were
rinsed 3x 5 min at room temperature in PBS with gentle agitation in between
each of the steps. All other incubations were performed in a humidified chamber
moistened with TBS.
Bone sections required low temperature antigen retrieval. Therefore, slides
were immersed in citrate buffer [10mM citric acid, 0.05% Tween 20; pH6.0] and
incubated overnight (approx. 16 hours) at 70oC. Slides were then cooled and
washed in TBS for 10 min. Slides were quenched for endogenous peroxidase
activity by incubating in 3%H2O2 in TBST for 10 min at room temperature with
gentle agitation. Slides were then washed 3x 5 min in TBST 0.01% [1x TBS with
0.01% Tween 20]. A border was drawn around tissue on the slides using a wax
pen (Dako, Glostrup, Denmark) then slides were blocked with 10% normal goat
serum (Vector Labs, Burlingame, CA, USA, #PK6101) in TBST 0.01% (diluent
solution) for 1 hour at room temperature in a humidified chamber. Blocking
solution was then tipped off slides to be replaced with 300-500µl of primary
antibody in diluent solution (anti-Pax5, anti-Ki67 or anti-VwF, dilutions in Table
2.1). Slides were incubated overnight at 4oC. The next day, slides were washed
3x 5 min in TBST 0.01%, before 300-500µl of secondary antibodies in diluent
solution were applied (anti-rabbit-biotin, anti-mouse-biotin and anti-rabbit-biotin,
respectively). Slides were incubated for 1 hour at room temperature. Slides
were then washed 3x 5 min and Elite ABC solution in TBST 0.01% was added.
Slides were then incubated at room temperature for 30 min. Slides were then
washed 3x 5 min.
Staining was visualized with NovaRed (Vector Labs, Burlingame, CA, USA,
#SK-4800) applied for 2-3 minutes depending on antibody. Slides were then
Figure 2.2. Immunohistochemical staining was scored by the Nuance FX multispectral imaging system and inForm software (Perkin Elmer). multispectral cube was processed with a spectral library unmixing Nova Red and Haemtaoxylin. b. Sections were first segmented into bone marrow (red) or bone (green) tissue sections. c. Individual nuclei were then detected (green). d. Lastly unmixed Nova Red staining (desaturated of haematoxylin) was scored 0-3+, with 0-1 (blue), 1-2 (yellow), 2-3 (orange) and 3+ (maroon). Positive staining was the sum of 1-3+.
a. The
b.a.
c. d.
70
rinsed in dH20 and washed twice in dH20. Slides were then counterstained with
Gill’s Formula Haematoxylin (Vector Labs, Burlingame, CA, USA, #H-3401) for
30 sec followed by 10 dips in 2% acetic acid in dH20 and 60 sec in Bluing
Reagent (Richard Allen Scientific, ThermoFisher Scientific, St. Louis, MO, USA,
#7301). Slides were dehydrated using a Leica Autostainer XL (Leica
microsystems, Wetzlar, Germany). The protocol involved incubating for 1 min in
40% EtOH, 1 min 70% EtOH, 1 min 95% EtOH, 2x 2 min in 100% EtOH and 3x
3 min in xylene. Slides were then coverslipped (Corning, New York, NY, USA)
using Permount mounting medium (ThermoFisher Scientific, St. Louis, MO,
USA).
2.7.6.3 Quantitative morphometry Quantitative morphometry was carried out using the Nuance FX multispectral
imaging system and inForm software (PerkinElmer, Waltham, MA, USA),
according to manufacturer’s instructions for acquisition of brightfield cubes and
4-bin scoring.
Multispectral cubes were obtained for 3 sections of both diaphysis and
epiphysis/metaphysis for each slide. A spectral library was created to unmix
Nova Red and Haematoxylin staining. For Pax5 and Ki67 stained slides, inForm
was trained to segment bone marrow tissue from bone. Using only bone
marrow segments, nuclei were detected for Haematoxylin or Nova Red stained
cells above 0.2% threshold and with 0.25µm roundness. Scoring was then done
for Nova Red stained nuclei using the 4-bin method, whereby, each nucleus
was given a score of 0-1 (no staining), 1-2 (slight staining), 2-3 (moderate
staining) or 3-4 (heavy staining), see Figure 2.2. Statistical analysis was
performed using the percentage of stained cells (% of nuclei with a score of 1-
4). For VwF stained slides, bone marrow segments were detected for
megakaryocytes stained with Nova Red above a threshold of 0.2% and 600
pixels, object count was determined for each image.
Figure 2.3. Reticulin stain scoring. Scoring was performed on diaphysis and epiphysis/ metaphysis by two blinded observers based on pre-established levels of reticulin intensity: a. Score 0 with focal, fine reticulin (normal). b. Score 1 with diffuse fine reticulin (abnormal). c. Score 2 with diffuse fine reticulin and focal coarse reticulin (abnormal). d. Score 3 with diffuse coarse reticulin (fibrosis). Bone stains brown, nuclei stain red and reticulin fibres stain black. Scale bar represents 200mm.
b.a.
c. d.
0
2
4
6
8
10
5 10 15 20 25 30Difference in Population Means
Figure 2.4. Sample size required per treatment group (x axis) to detect a difference in population means (y axis). Calculated using a standard deviation of 6 based on a pilot study where the response within each mouse group was approximately normally distributed. If the true difference in the CTGF High and CTGF Low engraftment is 10% (as detected in the pilot study), we will need to study 7 mice per group to be able to reject the null hypothesis that the population means of CTGF High and CTGF Low are equal.
Expe
rimen
tal S
ampl
e Si
ze
72
2.7.5.3 Silver stain Gordon and Sweets’ modified reticulin silver stain was performed on bone
sections (Gordon & Sweets Jr, 1936). Slides were deparaffinised in xylene and
rehydrated through graded alcohol to dH20. Slides were then exposed to
acidified permanganate solution for 8 min (1% potassium permanganate and
3% sulfuric acid) then rinsed in tap water followed by dH20. Slides were then
bleached in 5% oxalic acid for 2 min then rinsed twice in dH20. Slides were then
set with a 2% iron alum mordant for 8 min to produce sensitized sites for silver
deposition, then rinsed well in dH20. Slides were then treated with silver solution
for 3.5 min (5% silver nitrate, 3% sodium hydroxide and 3 drops concentrated
ammonia in 70% EtOH in 40 ml). Slides were then immediately placed in
formalin developer and agitated for 2 min (5% concentrated formalin in 70%
EtOH). Slides were then washed in dH20 and treated with 5% sodium
thiosulphate to remove unprecipitated silver and washed in dH20. Slides were
then counterstained with Kernechtrot for 5 min and rinsed in dH20. Slides were
then dehydrated through graded alcohol, cleared with xylene and coverslipped
with Ultramount mounting medium (Dako, Glostrup, Denmark). Reticulin
staining was scored manually on a scale of 0-3 from the average score given by
two blinded-observers (Figure 2.3).
2.8. Statistical analyses
Graphing of all data was performed using Prism 6.0 (GraphPad Software, De La
Playa, CA, USA). Survival curves were created by Kaplan Meier method.
Statistical comparison of means was performed by student t-test or log-rank
test. Results were considered statistically significant at p<0.05. Unless
otherwise indicated, average values were expressed as mean±SEM. For mouse
studies, group sizes of 7 were chosen to detect a 10% difference in population
means with a standard deviation of 6 (data from unpublished pilot experiment)
(Figure 2.4).
333333333
Stable overexpression of CTGF in B-precursor ALL cell lines does not
alter cell function
If you want a wise answer,ask a reasonable question.
Johann Wolfgang Von Goethe
74
75
Chapter 3 Stable overexpression of CTGF in
B-cell precursor ALL cell lines does not alter cell function
3.1 Abstract The CTGF gene encodes for a matricellular protein found throughout the body
and associated with diverse biological functions. Deregulated CTGF has been
associated with 27 human cancers acting on various functional mechanisms
within the cancer cell and in association with surrounding stromal cells. In
childhood B-cell precursor (BCP) ALL, CTGF has been found upregulated in
75% of patient specimens compared to normal CD34+ haematopoietic
progenitor cells and to be associated with worse clinical outcomes. A functional
role for CTGF in BCP ALL has yet to be discerned.
In this study BCP ALL cell lines (PER-278 and PER-371) were successfully
transduced using lentiviral technology for stable overexpression of CTGF. RNA
overexpression was 3-fold higher on average in CTGF High transduced cells
compared to CTGF Low using qRT-PCR. CTGF overexpression was similar for
cellular and secreted protein levels. Quantification of CTGF in conditioned
culture media had a high correlation with CTGF mRNA expression levels
(Pearson correlation coefficient= 0.97).
PER-278 and PER-371 CTGF High cells were found to have no observable
differences in vitro in proliferation or sensitivity to two chemotherapy drugs
(vincristine and cytarabine) compared to CTGF Low cells. Following microarray
analysis, significant differences in global gene expression were seen between
PER-278 and PER-371 cell lines and between lentiviral-infected and uninfected
cell lines. These large differences in expression most likely overwhelm smaller
differences in expression due to higher CTGF expression that are common to
both cell lines. The lack of any perceptible response due to increased CTGF
expression in BCP ALL cells in vitro indicates any functional role for the protein
in this disease is likely to be mediated through microenvironment
intermediaries.
76
3.2 Introduction Most cancers reach a stage of autonomous control independent of normal cell
signalling. At this stage intracellular production of essential proteins by the
tumour cell replaces extracellular mitogenic cues for proliferation and survival
(Martin, 2003; Ravandi, et al., 2003). This is in contrast to normal cells that are
able to identify upregulated mitogen stimulation and initiate cell-cycle arrest or
apoptosis. CTGF is a gene that is overexpressed in 75% of paediatric BCP ALL
patient specimens, occurring across the majority of chromosomal
rearrangements (Boag, et al., 2007). It has also been associated with worse
outcome in childhood BCP ALL patients with high-risk stratification (Kang, et al.,
2010). Importantly, in other human cancers it acts to alter proliferation through
various cell specific signalling pathways.
It is unknown if CTGF has an impact on intracrine or autocrine mechanisms in
BCP ALL. Therefore in this study two BCP ALL cell lines, derived from primary
patient specimens, were engineered to create stable populations with
upregulated CTGF expression. BCP ALL cell lines with high CTGF expression
were then assessed for changes in cell function. The potential for CTGF to
stimulate cell proliferation by itself or as a co-factor with the known B-cell
mitogen interleukin 7 (IL7) was measured by tritiated [3H]thymidine
incorporation. Altered sensitivity to essential chemotherapy drugs, vincristine
and cytarabine, was assessed by Alamar Blue viability assays. Lastly, potential
intracellular effects were determined by characterising global gene expression
of the BCP ALL cell lines using Affymetrix Human Gene 1.0 ST Arrays.
3.3 Results PER-278 and PER-371 cell lines were chosen for their low but detectable levels
of CTGF mRNA expression compared to a panel of ALL cell lines and primary
BCP ALL specimens, as measured by qRT-PCR (Welch, et al., 2013). Both cell
lines are derived from primary specimens obtained from patients whose details
are outlined in Table 3.1. Cell lines were authenticated by STR
Table 3.2. Patient-derived B-precursor ALL cell lines maintain patient DNA profile. Single Tandem Repeat (STR) analysis of PER-278 and PER-371 cell lines were found to have 100% allelic concordance with their original patient specimens.
AMOLOGENIN
Locus Cell line PER-278 Cell line PER-371
CSF1PO
D13S317
D16S539
D5S818
D7S820
Th01
TPOX
vWA
X,Y X,Y
11,12 12,13
8,9 10,12
13,13 11,13
12,13 11,12
11,12 8,10
6,8 6,6
14,19 15,16
8,11 10,11
Feature Cell line PER-278 Cell line PER-371
1987Isolated 1988
malemaleGender
Specimen diagnosisdiagnosis
Phenotype B-precursor ALLB-precursor ALL
Genetic Feature E2A-PBX expression E2A-PBX expression
46,XY,der(16)t(1;16)(q2?1;p13),der(19)t(1;19)(q?13;p13)/46,X,_
Y,_?der(1)t(Y;1)(q12;?q21), add(11)(q21),der(19)t(1;19)
46,XY,der(9)t(1;9)(q23;p13),
der(19)t(1;19)(q23;p13)Karyotype
Table 3.1. Patient details for derived cell lines PER-278 and PER-371. Characteristics of two cell lines derived from primary patient specimens at Princess Margaret Hospital, Perth, Western Australia (Kees et al., 1990, Kees et al., 2003).
78
analysis and found to have 100% accordance with the DNA profile of their
associated primary patient DNA profile (Table 3.2). (Kees, et al., 1990; Kees, et
al., 2003)
3.3.1 Optimisation of transfection in BCP ALL cell lines As haematopoietic cells are particularly hard to transfect, optimisation was
required for the transfection of the BCP ALL cell lines. Two electroporation
techniques were used and one lentiviral infection method.
3.3.1.1 Transfection by electroporation The pcDNA3.1 plasmid containing the full human CTGF coding region or the
pcDNA3.1-eGFP control vector was used for all electroporation attempts
(Figure 3.1a). Plasmid DNA was used in a supercoiled state or was digested
with a BglII enzyme cutting distal to the important coding sequences to ensure
successful chromosomal insertion.
Standard electroporation was performed at 300 V or using the Neon
transfection system at 1400 V (optimised from a 900 to 1700 V voltage range).
The PER-278 cell line together with pcDNA3.1-eGFP plasmid DNA was used
for optimisation and transfection efficiency was measured from 24 hours post-
electroporation by intracellular GFP expression. Cells were selected for viable
stable tranfectants from 48 hours with G418 antibiotic. Electroporation was
slightly more efficient at transfecting BCP ALL cells when supercoiled plasmid
DNA was used compared to linearised DNA (Figure 3.1b). However, within two
weeks no surviving transfected cells were detected. Low transfection efficiency
from 24 hours post-electroporation suggested the PER-278 BCP ALL cell line
did not successfully integrate plasmid DNA using the electroporation technique.
Figure 3.1.Transfection of B-precursor ALL cell lines by electroporation was not successful in producing stable clones. a. The pcDNA3.1 plasmid with a human CTGF sequence inserted after the CMV promoter sequence. The empty vector plasmid containing reporter enhanced (e)GFP was used for optimisation. b. Transfection efficiency of plasmid DNA incroporation into PER-278 after electroporation with respect to time (days). The plasmid was transfected in supercoiled (blue line) or linear form (orange line). Transfection was performed by standard electroporation (solid line) or the Neon
+transfection system (dashed line). Data are expressed as the % GFP cells per total population of live and dead cells (n=1). Grey vertical bars represent addition of G418 to tissue culture media for selection of stable clones.
b.
a.
80
3.3.1.2 Transduction by lentiviral infection To overcome low transfection efficiency, the use of lentiviruses was adopted.
Lentiviral ability to transduce dividing and non-dividing cells efficiently made
them an appealing candidate to establish stable transgene expression in our
BCP ALL cell lines. Manufacture of a lentiviral plasmid was performed using the
full human CTGF coding sequence excised from the pcDNA3.1-CTGF plasmid.
The CTGF sequence was then inserted into the empty vector pCDH-MSCV-
MCS-EF1-GFP-T2A-PURO (pCDH-EV) backbone (Figure 3.2a). Optimisation of
lentiviral transduction utilised two methods of lentiviral particle production,
namely by transiently transfecting 293T cells using CaCl2 or Fugene.
Optimisation of the ratio of pCDH vector and three packaging plasmids was
performed and preparations with the highest three viral titres were used to
transduce both PER-278 and PER-371. All lentiviral particles were used
undiluted and included CaCl2 (Ratio 1) with a titre of 1.0x106 IFU/mL, CaCl2
(Ratio 2) with a titre of 5.0x106 IFU/mL and Fugene (Ratio 1) with a titre of
4.3x106 IFU/mL. Transduction efficiency was measured after 48 hours by GFP
expression; subsequently lentiviral transduction using particles produced with
Fugene (Ratio 1) and CaCl2 (Ratio 2) methods proved most efficient in our BCP
ALL cell lines (Figure 3.2b).
Figure 3.2. Transduction efficency of B-precursor ALL cell lines by lentiviral infection. a. The pCDH1-MSCV-CTGF-EF1-GFP (pCDH-CTGF) plasmid used in the described lentiviral infections, the empty vector plasmid omitted CTGF (pCDH-EV). b. Lentiviral transduction efficiency was expressed
+as the % GFP cells 48 hours after infection and measured by flow cytometry. Infections were performed on PER-371 (orange) and PER-278 (blue) cell lines. Lentiviral particles were generated using 293Ts with CaCl or Fugene 2transfection methods and two virus concentrations (Ratio1 or Ratio2).
b.
a.
82
3.3.2.1 Verification of CTGF overexpression in stably transduced cell lines After selection of stably transduced cells with puromycin, chromosomal insertion
of the pCDH vector was verified by detection of GFP expression. Flow
cytometry analysis of transduced PER-278 and PER-371 cell lines showed GFP
expression in over 95% of cells. Mean fluorescence intensity differed between
pCDH-EV and pCDH-CTGF expression with simultaneous expression of the
transgene and GFP reducing intensity levels by a factor of one logarithm
(Figure 3.3a).
After verifying successful chromosomal insertion of plasmid DNA by GFP
expression, CTGF mRNA levels were then measured by qRT-PCR.
Approximately 3-fold overexpression of CTGF was detected using both CaCl2
and Fugene-produced lentiviral particles in pCDH-CTGF compared to pCDH-EV
PER-278 and PER-371 transduced cell lines (Figure 3.3b).
Figure 3.3. Verification of stable CTGF mRNA in + precursor ALL cell lines PER-278 and PER-371. a. GFP expression
intensity was measured by flow cytometry for pCDH-empty vector (pCDH-EV), blue line, pCDH-CTGF, red line and uninfected stable cell lines, black. b. CTGF mRNA expression levels detected by qRT-PCR. Fold changes for pCDH-CTGF compared to pCDH-EV are displayed for both CaCl and Fugene 2production methods. Data are expressed as ratio of CTGF expression to housekeeping gene GAPDH. FC=fold change.
overexpression of B-
pCDH-CTGF
UninfectedpCDH-EV
b.
a.
FC=3.08 FC=2.59 FC=2.46 FC=3.37
84
In addition to mRNA levels, CTGF protein levels in cell extracts and secreted
into culture supernatants was determined by western blot. Cellular protein
levels of CTGF in pCDH-EV transduced cell lines were just detectable and
overexpression of CTGF in pCDH-CTGF transduced cell lines indicated greater
expression in cells from Fugene compared to CaCl2-produced lentiviral particles
(Figure 3.4.a). As the CTGF protein is known to be matricellular, secreted levels
of the protein were detected from purified conditioned media samples separated
with heparin-agarose beads. Again CTGF expression was barely detectable in
pCDH-EV cell lines and protein upregulation in pCDH-CTGF transduced cell
lines was observed by western blot (Figure 3.4b).
Stratification of the transduced cell lines according to CTGF expression level
was done using cellular protein expression, as mRNA expression levels were
indistinguishable between CaCl2 and Fugene-produced lentiviral particles and
secreted protein levels have no loading control due to their purification
methods. In this manner, Fugene-produced pCDH-EV transduced cells were
termed CTGF Low (or Low), CaCl2-produced pCDH-CTGF transduced cells
were termed CTGF Intermediate (or Intermediate) and Fugene-produced
pCDH-CTGF transduced cells were termed CTGF High (or High).
Quantification of secreted CTGF in PER-278 Low and High and PER-371 Low
and High cell lines was determined by ELISA (Figure 3.4c). PER-278 High
conditioned media contained an average level of CTGF protein 3 times greater
than PER-278 Low (49.3 ng/ml compared to 16.5 ng/ml). Similarly, PER-371
High conditioned media contained significantly higher levels of CTGF protein
3.4 fold greater than PER-371 Low (39.5 ng/ml compared to 11.5 ng/ml,
p=0.04). These levels correlate well with the previously mentioned CTGF
mRNA expression levels of these cell lines (Pearson correlation coefficient=
0.97).
Figure 3.4. Oprecursor ALL cell lines PER-278 and PER-371. a. CTGF cellular protein expression at 38 kDa measured by western blot. pCDH-CTGF and pCDH-EV stable cell lines were detected for both CaCl and Fugene production 2methods. Positive control was high-CTGF expressing cell line PER-377 and loading control was a pan-Actin antibody. b. Secreted CTGF protein expression from Heparin-purified conditioned media standardised by collection after 24 hours in culture. Positive control was PER-377. c. Fugene EV is designated CTGF Low and Fugene CTGF is designated CTGF High for PER-278 and PER-371. Concentration of secreted CTGF was measured by ELISA. Data are expressed as the mean protein concentration in ng/ml (n=3). Vertical bars show standard error of the mean.
verexpression of CTGF cellular and secreted protein in B-
pan-Actin
CTGF
PER-
371
EVPE
R-37
1 CT
GF
Posit
ive co
ntro
l
37KDa
37KDa
PER-
371
EVPE
R-37
1 CT
GF
PER-
278
EVPE
R-27
8 CT
GFPE
R-27
8 EV
PER-
278
CTGF
PER-
371
uninf
ecte
dPE
R-37
1 EV
PER-
371
EV
PER-
371
CTGF
PER-
371
CTGF
Posit
ive co
ntro
l
Fugene CaCl 2
PER-
278
uninf
ecte
d
37KDa CTGF
PER-
278
EVPE
R-27
8 CT
GF
Fugene CaCl 2
PER-
278
EVPE
R-27
8 CT
GF
b.
a.Fugene CaCl 2 Fugene CaCl 2
c.
PER-278 Low PER-371 Low PER-278 High PER-371 High
Cell line
86
3.3.3 Characterisation of stable CTGF overexpressing cell lines After validation of 3-fold overexpression of CTGF in PER-278 and PER-371 cell
lines at the mRNA and protein level, the effect of increased levels of CTGF on
cell function was determined.
3.3.3.1 Proliferation Potential differences in the proliferation rate of both PER-278 and PER-371 cell
lines due to CTGF overexpression were investigated by tritiated [3H]thymidine
incorporation into the cells. Cells plated in low serum (0.2% FCS) revealed no
significant difference in proliferation between CTGF Low and High (Figure 3.5a).
To determine if CTGF could act as a co-factor for stimulation in these cell lines,
proliferation was measured in the presence of various concentrations IL7 (0.01
to 1 ng/ml). CTGF also had no effect on the proliferation of CTGF High cells in
the presence of IL7 (Figure 3.5b). However, our BCP ALL cell lines appeared
non-responsive to increasing concentrations of IL7 indicating no mitogenic
activation was produced by the protein. In fact lower counts of [3H]thymidine
incorporation were detected in all cells used in the IL7 proliferation assay
including media positive controls, although detected levels were still greater
than negative controls. This could be due to differences in the
seeding density of cells between assays (50,000 in low serum assay compared
to 100,000 in IL7 assay) and cells nearing confluence at the time of
[3H]thymidine incorporation in the IL7 assay.
Figure 3.5. did not affect proliferation in vitro. Proliferation was measured by tritiated thymidine incorporation (counts/min) for PER-278 and PER-371 Low, Intermediate or High levels. a. Proliferation in low serum after 48hours (0.2% FCS) b. Proliferation after 48 hours incubation with IL7 at varying concentrations 0-1ng/ml (black to pale grey bars). Data are expressed as the mean isotope count per minute (n=3). Vertical bars show standard error of the mean.
PER-278 and PER-371 CTGF overexpression3[ H]
b.
a.
3 [H]
88
3.3.3.2 Drug sensitivity
The effective dose capable of killing 50% of cells (ED50) was determined for
two chemotherapy drugs vincristine (VCR) and cytarabine (ARAC). Both drugs
have been used before in vitro and in vivo as effective agents against BCP ALL
primary cells and cell lines (Liem, et al., 2004; Tesfai, et al., 2012).
The viability of PER-278 and PER-371 cells was determined after treatment
with a serial dilution of a single chemotherapy drug. Representative kill curves
(expressed as the % difference in reduction of Alamar Blue between treated
and untreated cells) depict a plateau at higher drug concentrations of
approximately 30% viability for both VCR and ARAC (Figure 3.6). Every cell line
was successfully reduced in viability to determine ED50.
There was no effect on drug sensitivity due to increased CTGF expression in
either PER-278 or PER-371 cells (Figure 3.7). Marked differences in ED50
between cell lines did exist, with PER-371 showing reduced sensitivity to VCR
and PER-278 showing reduced sensitivity to ARAC compared to the other.
VCR
Figure 3.6. for vincristine (VCR) and cytarabine (ARAC). Representative cell viability curves are depicted for High and Low cells to determine the effective dose to kill 50% of the cell population, ED50, after 72 hours incubation with two chemotherapy drugs: VCR with a. PER-278 and b. PER-371. ARAC with c. PER-278 and d. PER-371. Data are expressed as % difference in Almar Blue Reduction in drug treated compared to untreated cells.
PER-278 and PER-371 representative drug sensitivity curves
ARACc.
a.
b.
d.
371
371
Figure 3.7. sensitivity to chemotherapeutic drugs in vitro. The effective dose to kill 50% of the cell population, ED50, was determined for PER-278 and PER-371cell lines using Alamar Blue after 72 hours incubation with the chemotherapy drugs: a. vincristine (VCR) and b. cytarabine (ARAC). Data are expressed as the mean ED50 concentration in µg/ml (n=3). Vertical bars show standard error of the mean.
PER-278 and PER-371 CTGF overexpression did not affect
VCRa.
ARACb.
91
3.3.3.3 Microarray profiling With no observable differences due to increased CTGF expression in either
proliferation or drug sensitivity, changes in cellular gene expression patterns
were investigated. Microarray profiling of PER-278 and PER-371 cell lines with
CTGF at Low, Intermediate or High levels was performed in order to explore
potential intracrine changes taking place inside the cell.
RNA integrity analysis was performed on PER-278 and PER-371 mRNA
samples and all were determined to be of suitable quality for microarray
applications, as evidenced by 28S/17S ratio and RNA integrity number (Figure
3.8 and Table 3.3). Expression levels of CTGF mRNA and cellular protein were
determined by qRT-PCR and western blot and identified continued
overexpression of CTGF in CTGF High cells compared to CTGF Low (Figure
3.9). Secreted protein levels could not be determined after an 8 hour incubation
period. Differences between mRNA and cellular protein expression were
evident for CTGF High compared to Low and CTGF Intermediate compared to
Low for both PER-278 and PER-371 cells (Table 3.4). These RNA samples
were then run on an Affymetrix Human 1.0 ST array platform (conducted by the
Ramaciotti Centre for Gene Function Analysis, Sydney, Australia).
Normalisation of microarrays and LIMMA analysis was performed by
biostatistician Denise Anderson (TICHR, Perth, Australia). Plotting the data by
the variances of the first two principle components indicated the largest
differences in expression were between PER-278 and PER-371 cell lines and
then between uninfected and lentiviral-infected cells (Figure 3.10a). PER-371
Intermediate cells had an expression profile more closely associated with PER-
371 uninfected cells than their lentiviral-infected counterparts and was therefore
not used in further downstream analysis. LIMMA analysis revealed a large
number of significantly different genes that were altered by greater than 2-fold
between PER-278 and PER-371 cell lines (Figure 3.10b). Comparison of
CTGF-Low lentiviral-infected cell lines with uninfected parental cell lines again
revealed a multitude of genes with significantly different expression of greater
than 2-fold. However, no differences were found due to CTGF expression that
were common across the two cell lines, after comparing CTGF High to Low
Table 3.3. RNA integrity number for PER-278 and PER-371 cell line samples. RNA integrity numbers (RIN) for all 8 samples used in the Affymetrix microarray. A RIN>7 represents RNA that is not degraded and suitable for use in microarray processing.
RNA sample RINPER-278 uninfected
PER-278 Low 1010
PER-278 Intermediate 9.9PER-278 High
PER-371 uninfectedPER-371 Low
PER-371 IntermediatePER-371 High
9.310
9.89.9
9.8
Figure 3.8. RNA integrity analysis for PER-278 and PER-371 cell line samples. Representative electropherogram with a suitable 28S/17S ratio for samples used in the Affymetrix Human 1.0 ST arrays.
Table 3.4. Fold change of CTGF Hfor RNA and cellular protein expression. PER-278 and PER-371 cell line samples were collected after 8 hours incubation in normal media. RNA expression levels for CTGF were measured by qRT-PCR and cell protein CTGF expression levels were measured by western blot. Data presented as fold change for PER-278 and PER-371 Intermediate to Low and High to Low.
igh or Intermediate compared to Low
PER-278 Int/Low
PER-371 Int/LowPER-278 High/Low
PER-371 High/Low4.83.5
5.62.1
RNA (qRT-PCR)
4.4
4.16.4
2.6
Cell protein Fold Change
PER-
278
Low
PER-
278
Int
CTGF (CP)pan-Actin
PER-
278
Low
PER-
278
High
PER-
371
Low
PER-
371
Int
PER-
371
Low
PER-
371
High
b.
Figure 3.9. CTGF was overexpressed in the PER-278 and PER-371 cell line samples collected for microarray analysis. Samples were collected after 8 hours incubation in normal media. a. CTGF RNA expression levels as measured by qRT-PCR. Data are expressed as ratio of CTGF expression to housekeeping gene GAPDH. b. Cellular protein (CP) CTGF expression levels as measured by western blot. A pan-Actin antibody was used as a loading control.
RNA and protein
a.
94
cells. Only CTGF itself was found to be significantly upregulated in unadjusted,
but not adjusted, analysis (unadjusted p-value= 0.01).
3.4 Discussion
The present study lentiviral transfection proved to be the optimal method of
transfecting BCP ALL cell lines. Previously published studies have identified the
Neon transfection system as capable of achieving high efficiency in difficult to
transfect cells, including stable transfection of human embryonic stem cells,
transient transfection of diffuse large B-cell lymphoma cell lines and even
transient siRNA transfection of B-cell ALL cell lines (Hagiya, et al., 2011; Jiang,
et al., 2011; Pham, et al., 2010). However, Neon and standard electroporation
failed to efficiently transfect BCP ALL cell lines with a CTGF transgene. As less
than 1% of cells showed incorporation of plasmid DNA, the method was not
sufficient for stable transfection.
As stable transfection was a requirement for in vivo studies, use of lentiviral
technology was adopted. Third generation lentiviral plasmids including the tat
gene, were transfected with a VSV-G envelope plasmid into 293T cells together
with the transfer plasmid containing the CTGF sequence. Transient transfection
of 293Ts was equally successful using both CaCl2 and Fugene methods so long
as the ratio of plasmids had been individually optimised. The VSV-G envelope
proved to have sufficient tropism for BCP ALL cells and incorporation of the
transfer plasmid into the chromosomal DNA proved successful. BCP ALL
transduction efficiency at 48 hours was 30% and puromycin selection enabled
greater than 95% purity. Lentiviral transduction was the preferred mechanism
for genetically engineering expression of a transgene in BCP ALL cell lines.
The BCP ALL engineered cell lines were found to have biologically relevant
levels of CTGF overexpression. A 3-fold upregulation in CTGF mRNA
expression was achieved in both PER-278 and PER-371 High cells compared
to empty-vector infected Low cells when measured by qRT-PCR. Further,
CTGF protein levels were 3- to 6-fold greater in CTGF High cells when
measured by densitometry of western blot bands. This fold change appears
small compared to the original microarray deregulation of 19.45-fold between
BCP ALL patient specimens and CD34+ haematopoietic progenitor cells found
Figure 3.10. PER-278 and PER-371 CTGF High cells did not affect global gene expression patterns compared to CTGF Low cells. a. Principal component analysis (PCA) plot of the microarray data. The plot displays the variances of the first 2 principle components. Each dot represents a sample and each colour the type of sample. For PER-278 uninfected (green) to CTGF High cells (yellow) and PER-371 uninfected (purple) to CTGF High cells (pink). b. Volcano plots displays LIMMA statistical significance against fold change for three analyses: the comparison of cell lines (PER-278 vs PER-371), lentiviral infection (Low vs Uninfected) and CTGF expression (High vs Low).
PER-278 uninfectedPER-278 LowPER-278 IntermediatePER-278 HighPER-371 uninfectedPER-371 LowPER-371 IntermediatePER-371 High
b.
a.
High vs Lowlog fold change2log fold change2
log fold change2
-log
P-v
alue
10
-log
adj
uste
d P-
valu
e10
-log
P-v
alue
10
Low vs Uninfected PER-278 vs PER-371
96
by Boag and colleagues (2007). However, in the same study validation of
microarray results by qRT-PCR found only 2.4 fold difference between average
CTGF expression in 18 BCP ALL patient specimens and 4 CD34+ umbilical cord
specimens. In additional studies a wide range of CTGF expression levels
existed among 14 primary BCP ALL patient specimens with the highest CTGF
level having 10-fold greater expression than non-detectable specimens by qRT-
PCR (Welch, et al., 2013). Similarly, CTGF expression levels in adult B-cell ALL
specimens spanned more than 70-fold from highest to lowest (Sala-Torra, et al.,
2007). Thus the CTGF levels in PER-278 and PER-371 High and Low cell lines
places them within the range recorded in primary samples, with Low CTGF-
expressing cells clearly toward the lower end of the spectrum and High CTGF-
expressing cells in the middle of the spectrum and they’re therefore acceptable
candidates for further study.
As there is large heterogeneity in CTGF expression among primary patient
samples, identifying an expression level where CTGF initially affects cell
function is close to impossible. It is more likely each tumour cell has its own
innate propensity to respond to the cues created by CTGF overexpression and
therefore each tumour may possess a unique threshold where CTGF
expression contributes to disease outcome.
In the present study, no autocrine or intracrine function for CTGF was detected
in BCP ALL cell lines as biologically relevant upregulation of CTGF expression
in BCP ALL cell lines did not elicit any observable changes in vitro. Neither
proliferation nor drug sensitivity was affected due to increased CTGF
expression. However, there was a very distinct response pattern to VCR and
ARAC in the two cell lines with PER-371 cells showing a 3.7-fold reduced
sensitivity to VCR compared to PER-278. Even more disparate was the
response to ARAC with PER-278 cells 25-fold less sensitive to the drug
compared with PER-371 cells. Global gene expression analysis found no
commonly deregulated genes due to altered CTGF in the cell lines. Lack of
common deregulated genes due to CTGF is not surprising considering the
significant gene expression differences between PER-278 and PER-371 cell
lines at the RNA level. With 2 biological replicates with a high variance in
97
expression, the ability to determine deregulated gene patterns due to CTGF
becomes challenging.
Determining a functional role for CTGF in BCP ALL cells was investigated by
successfully engineering the overexpression of CTGF in two BCP ALL cell lines,
PER-278 and PER-371. In culture, CTGF overexpression was found to have no
effect on proliferation of cells or sensitivity to the chemotherapy drugs VCR or
ARAC. Further, global gene expression arrays, while detecting significant
changes in gene expression profiles due to cell line or lentiviral infection, found
no common alteration in gene expression between PER-278 and PER-371 High
cells due to upregulated CTGF.
The absence of an observable response in functional assays due to increased
CTGF expression could be attributed to the tests being performed in vitro. It is
plausible that as a secreted matricellular protein, any functional role of CTGF
involves interaction with the microenvironment. With the necessary intermediary
cells being absent in laboratory culture conditions.
98
444444444Recombinant human CTGF has the
capacity to alter the gene expression profile of human BM
stromal cells
All life is an experiment.The more experiments you make the better.
Ralph Waldo Emerson
100
101
Chapter 4 Recombinant human CTGF has the capacity to
alter the gene expression profile of human BM stromal cells
4.1 Abstract CTGF belongs to the CCN family of matricellular proteins. The CCN family are
well known to interact with ECM components throughout normal life and in
disease. A unique role for CTGF may exist in B-cell precursor (BCP) ALL, the
most common cancer diagnosed in children. CTGF is significantly upregulated
in childhood BCP ALL patient specimens and is associated with poor outcome.
CTGF, as a secreted protein with known ECM functions, may act in childhood
BCP ALL to alter cellular interactions and structure of the extracellular space in
the BM microenvironment, the haematopoietic niche where the disease
originates.
In this study the effect of recombinant human (rh)CTGF on the gene expression
profile of a human BM stromal cell line, HS5, is explored at three incubation
times: 8, 16 and 48 hours. Significant differences in HS5 gene expression due
to rhCTGF supports a role for the protein in insulin-like growth factor 1 (IGF1)
and phosphoinositide 3-kinase/protein kinase C (PI3K/AKT) signalling
pathways. The potential involvement of CTGF in ECM remodelling and
abnormal wound response through its effects on HS5 cells is also put forward.
This microarray study presents strong evidence for a functional role of CTGF in
human BCP ALL through BM microenvironment interactions.
4.2 Introduction The CCN family of matricellular proteins are defined as such due to their role as
ECM-related signalling proteins. Together, the family are involved in a large
number of normal functions from embryonic chondrogenesis and bone
formation through to roles in adult wound healing (Jun & Lau, 2011). Though
the CCN family members are involved in similar activities they respond to
stimuli in their own unique fashion. CTGF/CCN2 is known to respond to various
102
growth factors and environmental cues, largely by enacting a response at the
transcriptional level, with some instances of post-transcriptional regulation
(Leask & Abraham, 2006). The response to CTGF is not only specific compared
to other family members but also to the type and location of interacting cells.
Like all CCN family members, CTGF is deregulated in many human diseases
including cancers. This study investigates the role of CTGF in childhood BCP
ALL. The CTGF gene is known to be upregulated in the majority of childhood
BCP ALL cases and is associated with high-risk stratification in children and
worse overall survival in adults (Boag, et al., 2007; Kang, et al., 2010; Sala-
Torra, et al., 2005). A specific functional role for CTGF in BCP ALL has yet to
be recognised but as CTGF has associations with the ECM and is tagged for
secretion from the cell, the secreted protein may act through the
microenvironment in BCP ALL.
This study explores the ability of rhCTGF to modify the BCP ALL
microenvironment, consisting of stromal cells and regulatory proteins in the BM
niche. Global gene expression was assessed in an immortalised human BM
stromal cell line (HS5) after treatment with rhCTGF or control buffer for 8, 16
and 48 hours.
4.3 Results HS5 is a faithful representative of bone marrow stroma that supports
proliferation of hematopoietic progenitor cells when cocultured in serum-
deprived media with no exogenous factors (Roecklein & Torok-Storb, 1995a).
HS5 cells were incubated with 100ng/ml rhCTGF and RNA extracted at 8, 16
and 48 hours, hybridisation of RNA to Affymetrix HG-U133A arrays was
performed by research officer Jette Ford (TICHR, Perth, Australia) as described
in the materials and methods. After normalisation, principle component analysis
(PCA) of the first three principal components showed clear separation of the 48-
hour incubation data compared to 8 and 16 hours (Fig 4.1). To define the
accuracy of the expression profiles, data from the three untreated control times
were analysed in comparison to the originally published features of the HS5 cell
line (Roecklein & Torok-Storb, 1995a). At all three time points, control HS5
samples showed high expression of genes identified as showing strong positive
immunostaining or strong protein expression in the original immortalised cell
Extracellular matrix
Lipid metabolism
Adhesion Cytokine Signaling
GENE FC GENE FC GENE FC GENE FC
COL3A1 15.0 NPTX2 3.7 TGFB2 15.3 CCL2 3.6
LOX 5.4 MAN2A1 3.6 EFEMP1 5.4 CCL7 2.8 LUM 4.8 GALNT11 3.5 DPP4 4.4 CCL8 2.5
COL1A2 4.2 PTX3 3.1 FN1 4.0 IL1R1 2.1
FN1 4.0 ASAH1 2.7 LAMB1 3.6 CXCL1 -8.8 LAMB1 3.7 COLEC12 2.7 NID2 3.6 CXCL6 -5.8
TGFBR3 3.6 HMMR 2.4 TGFBR3 3.5 IL32 -2.8 ITGAV 3.4 SERPINE2 2.3 SEMA3C 3.2 CXCL5 -2.6 COL5A1 3.3 LGALS8 2.2 TNFAIP6 3.0 CCL20 -2.5
FBN2 3.3 GALNT7 2.2 WNT5A 2.5 CCL3 -2.2
NID1 3.1 FADS1 2.1 ITGBL1 2.3 TUFT1 2.7
MATN3 2.5
FBN1 2.3 TFPI2 2.0
ECM1 -2.1
Figure 4.1. Ccaused by treatment with rhCTGF compared to control buffer. HS5 cells were treated with rhCTGF (100ng/ml) or control buffer for 8, 16 and 48 hours before collection of RNA for gene expression analysis a. Principal component analysis (PCA) plot of the microarray data. The plot displays the variances of the first 3 principle components. Each dot represents a sample and each colour the type of sample. 8 hour incubations are green, 16 hour are pink and 48 hour are blue. CTGF treatments are a lighter shade to control buffer. b. Venn diagram displaying the number of significantly altered probe sets for all three time points between CTGF-treated and untreated HS5 cells. Seven probe sets were deregulated at all three time points. Significance was determined by LIMMA analysis (p<0.05; R platform).
hanges to global gene expression in BM stromal cells (HS5)
16hr CTGF16hr control
8hr CTGF8hr control
48hr CTGF48hr control
b.
a.
104
line including: interleukin 6 (IL6) and interleukin 8 (IL8) in the top 1% of
microarray probe sets and major histocompatability complex class I
(MHCI),colony stimulating factor 3 (CSF3/GCSF) and membrane metallo-
endopeptidase (MMN/CD10) in the top 10% of genes.
4.3.1 LIMMA analysis LIMMA analysis was performed to enable identification of statistically significant
differences from the small sample size with p<0.05. Treatment with rhCTGF
resulted in differential genes at all 3 times (Figure 4.1b). The greatest number of
significantly altered probe sets occurred 8 hours post rhCTGF treatment,
followed by 16 and then 48 hours and overlapping probes were present across
all three times. Only seven probe sets identifying 6 genes were significantly
different at all 3 times, and none showed consistent up- or downregulation
across the 48-hour period. Three patterns of deregulation were present. The
first involved ABI gene family, member 3 binding protein (ABI3BP) and tripartite
motif containing 22 (TRIM22) which were upregulated following 8 and 16 hours
treatment followed by downregulation at 48 hours post treatment compared to
their time-matched controls. The second pattern involved thymosin beta 10
(TMSB10) and proteasome subunit beta, type 4 (PSMB4) which showed
downregulation at 8 and 48 hours and upregulation at 16 hours post treatment
compared to controls. Lastly, the genes chemokine (C-X-C motif) ligand 1
(CXCL1) and chemokine (C-X-C motif) ligand 6 (CXCL6) were downregulated
at 8 hours then upregulated at 16 and 48 hours post treatment compared to
controls. Due to the small number of genes with continuous deregulation and
none with a consistent pattern of expression, further analysis focused solely on
genes deregulated 8 hours after rhCTGF treatment.
4.3.2 Ingenuity Pathway Analysis As the broadest range of deregulated HS5 genes was observed 8 hours post
treatment with rhCTGF, this time was chosen to perform Ingenuity Pathway
Analysis (IPA). The list of significant probe sets was filtered to include only
those with fold changes greater than 2 in the treated compared to control group.
This probe set list was used for IPA analysis to highlight top associated
networks, canonical pathways and biological functions associated with
Tab
le 4
.1. In
gen
uit
y p
ath
way a
naly
sis
(IP
A)
for
dif
fere
nti
ally r
eg
ula
ted
gen
es b
etw
een
str
om
al cells (
HS
5)
treate
d w
ith
rh
CT
GF
(1
00n
g/m
l) o
r co
ntr
ol b
uff
er
for
8 h
ou
rs . T
he IP
A g
ene li
st c
om
prise
s genes
identif
ied s
ignifi
cantly
diff
ere
nt by
LIM
MA
and w
ith a
n
exp
ress
ion d
iffere
nce
gre
ate
r th
an tw
o-f
old
com
pare
d to u
ntr
eate
d c
ontr
ol c
ells
(B
enja
min
i and H
och
berg
-adju
sted p
-valu
e <
0.0
5, fo
ld
change >
2 a
nd <
2).
Num
ber
of upre
gula
ted (á
) and d
ow
nre
gula
ted (â
) genes
indic
ate
d w
ere
use
d to g
enera
te IP
A s
ignifi
cant netw
ork
, ca
nonic
al p
ath
way
and k
now
n b
iolo
gic
al f
unct
ion a
sso
ciatio
ns.
Cell
Cyc
le
Cellu
lar
Deve
lopm
ent
IGF
-1 S
ignalin
g
Andro
gen S
ignalin
g
1.C
ell
Cyc
le, D
NA
Replic
atio
n, R
eco
mbin
atio
n,
and R
epa
ir, G
ene E
xpre
ssio
n
2. C
arb
ohyd
rate
Meta
bolis
m, S
mall
Mole
cule
B
ioch
em
istr
y, C
onnect
ive T
issu
e D
isord
ers
á787
â622
Cell
Death
Ary
l Hyd
roca
rbon R
ece
pto
r S
ignalin
g
3. C
onnect
ive T
issu
e D
eve
lopm
ent and F
unct
ion,
Tis
sue M
orp
holo
gy,
R
NA
Post
Tra
nsc
riptio
nal M
odifi
catio
n
Cellu
lar
Gro
wth
and P
rolif
era
tion
Neure
glin
Sig
nalin
g
Cellu
lar
Move
ment
RA
N S
ignalin
g
IPA
Netw
ork
# G
en
es
To
p B
iolo
gic
al F
un
cti
on
sIP
A C
an
on
ical
106
deregulation of HS5 cells due to rhCTGF (Table 4.1). IPA analyses for
deregulated HS5 genes 16 and 48 hours post CTGF treatment is provided as
supplementary material (Table S1 and S2). At 8 hours, the top IPA network and
biological function associated with rhCTGF was cell cycle and the top canonical
pathway was insulin-like growth factor 1 (IGF1) signalling (Figure 4.2). In
addition, the activation of phosphoinositide 3-kinase/ protein kinase B
(PI3K/AKT) signalling was associated with deregulated HS5 genes (Figure 4.3).
Many of the genes associated with the TGFβ network were found to be both
significantly up- and downregulated in CTGF-treated HS5 cells (Figure 4.4). IPA
analysis highlighted intracellular signalling cascades associated with
significantly deregulated HS5 genes after rhCTGF treatment.
4.3.3 Functions associated with secreted proteins HS5 cells may affect leukaemia cell activity via the secretion of proteins, which
was investigated with the use of a publically available protein database. The
IPA gene list comprised significantly deregulated HS5 genes after 8 hours
rhCTGF treatment compared to untreated control cells, p<0.05 and fold change
≥2. This list was further filtered to identify significantly differentiated expression
of genes associated with the cellular membrane or extracellular secretion.
Overlap between the list of deregulated HS5 genes and genes listed in the
secreted protein database initiative (SPDI) found 78 secreted genes were
significantly upregulated and 66 genes were significantly downregulated in HS5
cells post-CTGF treatment (Supplementary Tables S3 and S4).
Figure 4.2. CTGF dpathway. HS5 cells were treated with rhCTGF (100ng/ml) or control buffer for 8 hours and significantly altered probe sets due to CTGF treatment were determined by LIMMA (Benjamini and Hochberg-adjusted p-value <0.05, fold change >2 and <2). IPA analysis of the filtered gene list determined Insulin-like growth factor 1 (IGF1) signalling pathway to be the top canonical pathway associated with deregulated HS5 genes. Red indicates overexprerssion and green indicates underexpression of the gene.
eregulates HS5 genes involved in the IGF1 signalling
Figure 4.3. CTGF deregulates genes involved in the PI3K/AKT signaling pathway. HS5 cells were treated with rhCTGF (100ng/ml) or control buffer for 8 hours and significantly altered probe sets due to CTGF treatment were detremined by LIMMA (Benjamini and Hochberg-adjusted p-value <0.05, fold change >2 and <2). IPA analysis of the filtered gene listdetermined the phosphoinositide 3-kinase/ protein kinase B (PI3K/AKT) canonical signalling pathway to be associated with deregulated HS5 genes. Red indicates overexpression and green indicates underexpressionof the gene.
Figure 4.4. eregulates HS5 genes involved in TGFb networksHS5 cells4were treated with rhCTGF (100ng/ml) or control buffer for 8 hours and significantly altered probe sets due to CTGF treatment were determined by LIMMA (Benjamini and Hochberg-adjusted p-value <0.05, fold change >2 and <2). IPA analysis of the filtered gene list determined Transforming growth factor b (TGFb) networks to be the top associated IPA network with deregulated HS5 cells. TGFb network functions were associated with cell cycle, DNA replication, recombination and repair and gene expression.
CTGF d .
Red indicates overexprerssion and green indicates underexpression of the gene.
110
The list of differentially expressed secreted HS5 genes was analysed for the
most frequently associated extracellular functions (Table 4.2). The greatest
number of significantly deregulated secreted genes was associated with ECM
interactions, followed by lipid metabolism, adhesion and cytokine signalling.
4.3.4 Verification of deregulated genes by RT-PCR RNA expression by RT-PCR was performed to validate the deregulation of 4
secreted HS5 genes. Three were the top upregulated genes associated with the
ECM; collagen, type 3, alpha 1 (COL3A1), lysyl oxidase (LOX) and lumican
(Dornhofer, et al.), and one was the upregulated cell membrane binding protein
integrin, alpha V (ITGAV). Expression of HS5 mRNA extracted from cells after
treatment with rhCTGF for 4, 8 and 16 hours were compared to untreated
control cells. These results were confirmed if genes showed greater than 1.5-
fold upregulation due to rhCTGF. Validating the deregulation of LOX at 4 and 8
hours and ITGAV at 8 hours post CTGF treatment (Figure 4.5). Failure to verify
upregulation of COL3A1 and LUM due to CTGF treatment could be attributed to
the biological activity of rhCTGF being reduced between microarray studies and
RT-PCR validation several years later. However, those CTGF-deregulated
genes associated with secretion that were confirmed by qRT-PCR support the
findings of the microarray analysis.
SYM
BOL
CCL2
FC 3.6
GEN
Ech
emok
ine
(C-C
mot
if)
ligan
d 2
SYM
BOL
TGFB
2
FC 15.3
GEN
ETr
ansf
orm
ing
grow
th
fact
or, b
eta
2
SYM
BOL
NPTX
2
FC 3.7
GEN
E
neur
onal
pen
traxin
2
SYM
BOL
COL3
A1
FC 15.0
GEN
Eco
llage
n, ty
pe 3
, al
pha
1
CCL7
2.8
chem
okin
e (C
-C m
otif)
lig
and
7EF
EMP1
5.4
EGF
cont
aini
ng fi
bulin
-like
EC
M p
rote
in 1
MAN
2A1
3.6
man
nosid
ase,
alp
ha, c
lass
2A
, mem
ber 1
LOX
5.4
lysyl
oxid
ase
CCL8
2.5
chem
okin
e (C
-C m
otif)
lig
and
8DP
P44.
4di
pept
idyl-
pept
idas
e 4
GAL
NT11
3.5
polyp
eptid
e G
alNA
c Tr
ansf
eras
e 11
LUM
4.8
lum
ican
IL1R
12.
1in
terle
ukin
1 re
cept
or,
type
IFN
14.
0fib
rone
ctin
1PT
X33.
1pe
ntra
xin 3
, lon
gCO
L1A2
4.2
colla
gen,
type
1,
alph
a 2
CCL3
-2.2
chem
okin
e (C
-C m
otif)
lig
and
3LA
MB1
3.6
lam
inin
, bet
a 1
ASAH
12.
7N-
acyls
phin
gosin
e am
idoh
ydro
lase
1FN
14.
0fib
rone
ctin
1
CCL2
0-2
.5ch
emok
ine
(C-C
mot
if)
ligan
d 20
NID2
3.6
nido
gen
2CO
LEC1
22.
7co
llect
in s
ub-fa
mily
m
embe
r 12
LAM
B13.
7la
min
in, b
eta
1
CXCL
5-2
.6Ch
emok
ine
(C-X
-C
Mot
if) L
igan
d 5
TGFB
R33.
5Tr
ansf
orm
ing
grow
th
fact
or, b
eta
rece
ptor
3HM
MR
2.4
hyal
uron
an-m
edia
ted
mot
ility
rece
ptor
TGFB
R33.
6tra
nsfo
rmin
g gr
owth
fa
ctor
, bet
a re
cept
or 3
Il32
-2.8
inte
rleuk
in 3
2SE
MA3
C3.
2se
map
horin
3C
SERP
INE2
2.3
serp
in p
eptid
ase
inhi
bito
r, cla
de E
, mem
ber 2
ITG
AV3.
4in
tegr
in, a
lpha
v
Lipi
d M
etab
olis
mEx
trace
llula
r Mat
rixCy
toki
ne s
igna
lling
Adhe
sion
CXCL
6-5
.8Ch
emok
ine
(C-X
-C
Mot
if) L
igan
d 6
TNFA
IP6
3.0
tum
or n
ecro
sis fa
ctor
, al
pha-
indu
ced
prot
ein
6LG
ALS8
2.2
lect
in, g
alac
tosid
e-bi
ndin
g,
solu
ble,
8CO
L5A1
3.3
colla
gen,
type
5,
alph
a 1
CXCL
1-8
.8Ch
emok
ine
(C-X
-C
Mot
if) L
igan
d 1
WNT
5A2.
5wi
ngle
ss-ty
pe M
MTV
in
tegr
atio
n sit
e fa
mily
, m
embe
r 5A
GAL
NT7
2.2
polyp
eptid
e G
alNA
c Tr
ansf
eras
e 7
FBN2
3.3
fibril
lin 2
ITG
BL1
2.3
inte
grin
, bet
a-lik
e 1
FADS
12.
1fa
tty a
cid d
esat
uras
e 1
NID1
3.1
nido
gen
1
TUFT
12.
7tu
ftelin
1
Tabl
e 4.
2. T
op fu
nctio
ns o
f sec
rete
d HS
5 ge
nes
dbu
ffer f
or 8
hou
rs. T
he ta
ble
com
prise
s se
cret
ed g
enes
iden
tifie
d by
LIM
MA
as g
reat
er th
an 2
-fold
and
sig
nific
antly
diff
eren
t due
to
rhCT
GF
com
pare
d to
con
trols
(Ben
jam
ini a
nd H
ochb
erg-
adju
sted
p-v
alue
<0.
05, f
old
chan
ge >
2 an
d <2
). Th
e filt
ered
gen
e lis
t for
se
cret
ed p
rote
ins
were
gro
uped
into
the
mos
t com
mon
ly as
socia
ted
extra
cellu
lar b
iolo
gica
l fun
ctio
ns: e
xtra
cellu
lar m
atrix
func
tions
, lip
id
Met
abol
ism, a
dhes
ion
and
cyto
kine
signa
ling.
FC,
fold
cha
nge
betw
een
CTG
F-tre
ated
com
pare
d to
unt
reat
ed H
S5 B
M s
trom
al c
ells.
iffer
entia
lly re
gula
ted
afte
r tre
atm
ent w
ith rh
CTG
F (1
00ng
/ml)
or c
ontro
l
SYM
BOL
FCG
ENE
SYM
BOL
FCG
ENE
SYM
BOL
FCG
ENE
SYM
BOL
MAT
N3
FC 2.5
GEN
E
mat
rilin
3
FBN1
2.3
fibril
lin 1
TFPI
22.
0tis
sue
fact
or p
athw
ay
inhi
bito
r 2
PLAU
R-2
.1 p
lasm
inog
en a
ctiva
tor,
urok
inas
e re
cept
or
MM
P12
-2.1
mat
rix m
etal
lope
ptid
ase
12
ECM
1-2
.1ex
trace
llula
r mat
rix
prot
ein
1
TIM
P2-2
.3TI
MP
Met
allo
pept
idas
e In
hibi
tor 2
TIM
P1TI
MP
Met
allo
pept
idas
e In
hibi
tor 1
Lipi
d M
etab
olis
mEx
trace
llula
r Mat
rixCy
toki
ne s
igna
lling
Adhe
sion
PLAU
-5.7
plas
min
ogen
act
ivato
r, ur
okin
ase
-2.8
Figure 4.5. Validation of two genes upregulated in HS5 cells after treatment with CTGF. RNA expression of stromal cells (HS5) treated with rhCTGF (100ng/ml) or control buffer for 8 hours was measured by RT-PCR for Collagen, type 3, alpha1 (COL3A1), lysyl oxidase (LOX), lumican (LUM) and integrin, alpha V (ITGAV). Data are expressed as a comparative ratio of HS5 cells normalised to GAPDH then comparing treatment with rhCTGF (HS5+rhCTGF) to untreated control buffer (HS5). A greater than 1.5 fold change was required to validate upregulation of the tested RNA, confirming LOX and ITGAV to be upregulated due to rhCTGF. (n=1).
ECM-associated
114
4.4 Discussion The longitudinal effects of CTGF were assessed by gene expression profiling of
HS5 cells at 8, 16 and 48 hours post rhCTGF treatment. It was shown that
CTGF altered intracellular signalling to influence gene expression at every
assessed time. CTGF is a growth-factor inducible immediate-early gene in
normal cell function (Bradham, 1991). Given its transcription occurs minutes
after mitogenic stimulation, the decision was made to focus on HS5 genes
deregulated after the shortest incubation time, being 8 hours with rhCTGF.
Microarray analyses indicated that in response to CTGF treatment stromal HS5
cells deregulated genes associated with cell cycle, IGF1 and PI3K/AKT
signalling pathways. PI3K/AKT pathway activation can induce cell cycle
progression and macrophage survival. PI3K/AKT signalling is known to regulate
CTGF expression in a cell-dependent manner, with smooth muscle cells
increasing CTGF expression and vascular endothelial cells decreasing CTGF
expression via PI3K/AKT activation (Samarin, et al., 2009). In a non-small cell
lung carcinoma cell line IGF1-dependent AKT phosphorylation was suppressed
by CTGF overexpression (Chien, et al., 2006). In this study, activation of MEK
and AKT in HS5 cells was likely an IGF1-independent phenomenon, as in
stromal cells the IGF1 receptor (IGF1R) was downregulated in response to
CTGF. IPA analyses also suggested a direct link between CTGF and activation
of the PI3K/AKT signalling pathway via binding of an integrin-receptor and
subsequent activation of integrin-linked kinase (ILK), protein phosphatase 2
(PP2A) and AKT. Upregulation of ITGAV suggests it could function as an
integrin-linked transmembrane receptor for CTGF in HS5 cells. In osteoarthritic
chondrocytes, TGFβ, a mediator of CTGF action, was found to upregulate
ITGAV and in turn increase phosphorylation of PI3K and AKT (Kostopoulou, et
al., 2012). TGFβ2 was a gene significantly upregulated in HS5 cells due to
CTGF treatment. IPA analyses indicated CTGF modulated HS5 genes
associated with survival and proliferation networks and pathways.
Many of the secreted genes differentially expressed by HS5 cells due to CTGF
treatment are involved in ECM remodelling. LOX and ITGAV were validated as
upregulated by rhCTGF in HS5 cells by qRT-PCR. CTGF has been found to
115
induce LOX synthesis in cultured cardiac fibroblasts, in association with atrial
fibrillation (Adam, et al., 2011; Lavall, et al., 2013). ITGAV, an integrin
heterodimer alpha chain, when co-located with a beta subunit forms an integrin
receptor that has been shown to promote migration and survival when
expressed in cancer cells (Desgrosellier & Cheresh, 2010; Jandova, et al.,
2012; Uhm, et al., 1999). Primary BM-derived mesenchymal stromal cells are
also known to express ITGAV (Shur, et al., 2002). These proteins are also
targetable and inhibitors of both ITGAV and LOX have been previously
proposed as anti-cancer agents (Barker, et al., 2012; Nemeth, et al., 2007; van
der Horst, et al., 2011).
In this study, a subset of deregulated HS5 genes involved in ECM remodelling
are associated with the wound healing response which suggests BCP ALL cells
may hijack a normal fibrotic response in the BM through CTGF secretion.
Stromal cell expression of CTGF is known to play an important role in organ
fibrosis with secretion of CTGF by fibroblasts causing the overstimulation of
genes associated with the wound healing response (Leask, 2002; Sonnylal, et
al., 2010). A fibrotic stroma in solid cancers has also been associated with
misappropriated wound healing cues (Dvorak, 1986). In primary mesenchymal
stromal cells, knockdown of CTGF expression was shown to reduced
proliferation and promote differentiation into adipocytes but not fibroblasts
(Battula, et al., 2013). In leukemia CTGF may act in a similar fashion to promote
fibrosis.
In this study, rhCTGF was able to significantly deregulate the gene expression
profile of the human BM stromal cell line HS5. IPA analyses identified cell cycle
regulation as the top network and biological function associated with CTGF
incubation for 8 hours. Significantly altered canonical pathways included the
IGF1 signalling pathway, which appeared to be activated independently of the
IGF1 gene. Deregulated genes were associated with PI3K/AKT pathway
activation, and ITGAV, may act as the transmembrane receptor activating the
pathway. Secreted genes deregulated in response to 8 hour CTGF treatment
revealed strong associations with ECM remodelling and similarities with the
normal wound healing response. Follow-up studies are needed to validate more
genes of interest and determine protein expression and phosphorylation levels
116
in human stromal cells after exposure to CTGF protein. The presented
microarray analyses provide plentiful evidence that CTGF secreted from BCP
ALL cells activates intracellular signalling pathways in human BM stromal cells.
555555555CTGF contributes to drug sensitivity in PER-278 B-precursor ALL cells in
the BM microenvironment
The body is a big sagacity,a plurality with one sense,
a war and a peace,a flock and a shepherd.
Friedrich Nietzsche
118
119
Chapter 5 CTGF contributes to drug sensitivity in PER-278 B-cell precursor ALL
5.1 Abstract In childhood B-cell precursor (BCP) ALL patients who relapse are faced with
reduced treatment options and worse survival outcomes. Early relapses are
more likely to derive from sub-populations within the diagnostic leukaemic
mass. The CTGF gene is upregulated in 75% of primary cases of childhood
BCP ALL and has been associated with worse clinical outcome in high-risk
patients, suggesting a potential role in leukaemia cell drug resistance. A role for
CTGF in BCP ALL has not been functionally defined, however associations
between CTGF and many human cancers provide evidence of a tumour
promoting function for the protein.
To assess the effects of CTGF on leukaemia cell outgrowth and drug resistance
in BCP ALL, NOD/SCID xenografts were generated using the BCP ALL patient-
derived cell line PER-278 engineered to overexpress CTGF. Comparison
between these PER-278 CTGF High and Low xenografts found no difference in
overall survival or organ engraftment at 24 hours and 3 weeks post-injection to
represent survival and proliferation of leukaemia cells at homing and early stage
disease. Organ engraftment at end stage disease revealed an increased
percentage of leukaemic cells in PER-278 High engrafted lymph nodes
compared with PER-278 Low cells (67.2% versus 43%, p=0.04). In PER-278
Low xenografts treatment with low dose ARAC significantly increased survival
compared with saline-treated controls (p=0.003). This is in contrast to PER-278
High xenografts that did not show a significant difference between low dose
ARAC and saline treated mice groups. The PER-278 xenograft model found
differences due to CTGF overexpression in end stage lymph node engraftment
and reduced drug sensitivity to ARAC, establishing that CTGF can affect
disease outcomes in the context of a leukaemia-infiltrated BM
microenvironment.
120
5.2 Introduction Improving relapse rates and treatment options in BCP ALL remains a major
hurdle to improving disease outcomes (Bailey, et al., 2008). A common cause
of relapse is sub-clones from within the diagnostic leukaemic cell population, in
these cases resistance is inherent in the primary disease and not an acquired
event after exposure to chemotherapy drugs (Choi, et al., 2007). In 94% of
diagnosis-relapse sample pairs, DNA modifications found in sub-clones to be
responsible for relapse were associated with an ancestral leukaemic cell from
one of the various levels of hierarchy since time of diagnosis (Mullighan, et al.,
2008). Increased expression of CTGF is found in 75% of diagnosis specimens
of childhood BCP ALL and has also been identified in a gene-set with
associations to high-risk paediatric patients that have the poorest clinical
outcomes (Boag, et al., 2007; Kang, et al., 2010).
In order to investigate BCP ALL biology, mouse models are used that provide
excellent platforms to introduce more complex microenvironment interactions
contributing to tumourigenesis and drug response. For leukaemia studies the
NOD/SCID xenograft mouse model has been proven to establish disease with
similar features to the human patient (Lock, et al., 2002). The NOD/SCID model
does however have limitations in mimicking the clinical conditions found in
patients, most importantly differences in mouse and human genomes potentially
omit important xenograft-host interactions (Meyer & Debatin, 2011). The
importance of replicating an immune response as in human tumours is in direct
opposition to the necessity of using immune-compromised mice that allow
propagation of foreign human cells. The use of the NOD/SCID strain is a
suitable compromise that maintains some residual immune functioning, with
reduced NK cell activity and small populations of functioning mature
macrophages (Shultz, et al., 1995b). This improves the biology of the leukaemia
model by providing partial immunological selective pressure that would be
present in greater abundance in human patients.
In this study, a functional role for CTGF was investigated using NOD/SCID
xenografts. The BCP ALL cell line PER-278 was found to be characteristically
similar to the primary patient specimen it was derived from (see section 3.3).
PER-278 cells were previously engineered to stably overexpress CTGF and
GFP producing PER-278 High cells or expressed GFP alone producing PER-
121
278 Low cells (see section 2.2.5-7). PER-278 High and Low cells were then
used to create xenografts by injecting 2x105 cells into the tail-vein of NOD/SCID
mice, under both irradiated and non-irradiated conditions (see section 2.7.2).
The effect of CTGF overexpression on leukaemia cell growth in the BM was
assessed at times specific for homing, early and end stage disease and
included the response to single-agent chemotherapy drugs ARAC and VCR.
5.3 Results 5.3.1 CTGF expression level at time of injection CTGF expression levels were measured for PER-278 High and Low cells before
inoculating NOD/SCID mice. Three experimental times were examined: homing
at 24 hours post-injection, early stage disease at 3 weeks post-injection and
end stage at presentation of clinical signs of morbidity. Cellular and secreted
protein was measured by western blot (Figure 5.1). CTGF overexpression in
PER-278 High compared with PER-278 Low cells was verified at the cellular
protein and secreted protein level (Table 5.1).
5.3.2 Survival outcomes and organ engraftment throughout spread of leukaemia cells Flow cytometry was used to detect GFP-expressing leukaemia cells in organs
and peripheral blood and showed good correlation with CD19-expressing cells
(data not shown). However, detection of GFP+ cells in the peripheral blood was
not a good indicator of disease progression with sustained low levels of
detection throughout disease progression and maximal detection of 20% in
PER-278 High and 11% in PER-278 Low xenografts (Figure 5.3e). For this
reason, survival was determined by observable symptoms in the mice, which
were sacrificed when they presented with the first signs of a moribund
phenotype. Survival of PER-278 xenografts was not affected by CTGF
overexpression with both PER-278 High and PER-278 Low xenografts having a
median survival time of 49 days (p=0.34; Fig 5.2a). This shows CTGF had no
effect on survival outcomes of PER-278 xenografted mice.
PER-278 High/LowHoming (24hrs)
PER-278 High/LowEarly disease (Wk3)
PER-278 High/LowEnd stage disease
2.93
9.686.40
2.93
Cell protein (western)
17.06
Fold Change Secreted protein (western)
17.06
Table 5.1. Fold change difference between PER-278 High and Low cells before tail-vein injection into NOD/SCID mice. Cellular and secreted protein CTGF expression was measured by densitometry from western blot bands. Overexpression of CTGF in PER-278 High compared to Low cells was analysed for 3 experimental times (homing, early and end stage disease).
PER-
278
Low
CTGF (CP)pan-actin
CTGF (CM)PE
R-27
8 Hi
gh
PER-
278
Low
PER-
278
High
Hom
ing
& ea
rly s
tage
End
sta
ge
Figure 5.1. CTGF protein expression levels for PER-278 High and Low NOD/SCID xenografts. Cellular (CP) and secreted protein (CM) levels of CTGF were detected by western blot for cells used in tail-vein injection of NOD/SCID mice. Data represents CTGF expression for 3 xenograft experiments (homing, early and end stage disease). Anti-pan-actin was used as a loading control for cellular protein.
123
The ability of human leukaemia cells to home to the BM, the primary site of BCP
ALL, was not affected by CTGF overexpression with PER-278 High having
0.006% and PER-278 Low cells having 0.004% infiltration of GFP+ cells (Figure
5.2b). Within 24 hours, PER-278 High and Low cells were also capable of
homing to spleen with 0.063% versus 0.021% infiltration, liver with 0.04%
versus 0.11% infiltration and lymph node with 0.12% versus 0.04% infiltration.
No discernible difference was observed between BM engraftment of PER-278
cells with High and Low CTGF expression at early or end stage disease (Fig
5.2c and 5.2d). A gradual increase in BM engraftment was discernible between
homing and end stage disease with the mean percentage of leukaemia cells
increasing from 0.005% at 24 hours to 0.3% at 3 weeks and 75% at end stage
disease (mean of 7 weeks). This rapid expansion of leukaemia engraftment
toward end stage disease reflects the progress of disease in patients and was
also noted in xenografted mice for the spleen, liver, lymph nodes and peripheral
blood (Figure 5.3b-e). End stage leukaemia engraftment was lower in the
extramedullary sites of haematopoiesis, reaching approximately 50% mean
engraftment in the spleen and 70% in the liver. In the lymph nodes differences
in leukaemia engraftment at end stage disease were detected with PER-278
High having 67.2% engraftment compared with 43.0% in PER-278 Low
xenografts (p=0.04; Figure 5.3d). This finding was an exception with CTGF
expression resulting in uniform engraftment in all other detected organs.
Figure 5.2. PER-278 NOD/SCID xenografts showed no difference in survival or BM engraftment between High and Low injected cell lines.
5Mice were injected via the tail-vein with 2x10 PER-278 High or Low cells. a. Kaplan-Meier survival curve for PER-278 Low (dashed line) and PER-278 High xenografted mice (solid line). Data are expressed as the percent survival of mice against time (n=15). BM engraftment was determined for 3 times: b. homing at 24 hours post injection, c. early stage disease at 3 weeks and d.
+end stage disease. Data are expressed as the mean % GFP cells detected by flow cytometry (n=6-7, homing and early stage times, n=15, end stage from two separate experiments. Each dot represents an individual mouse). Vertical bars show standard error of the mean.
b. Homing (bone marrow)
c. Early stage (bone marrow) d. End stage (bone marrow)
a.
Figure 5.3. Rapid engraftment of leukameia at end stage disease for PER-278 NOD/SCID xenografts. Leukaemia engraftment was analysed at
5three times (homing, early and end stage disease) after injection of 2x10PER-278 High (triangle) or Low cells (circle). a. Engraftment was determined in the in BM and in the extramedullary organs; b. spleen, c. liver and d. lymph nodes. e. Leukaemia burden in peripheral blood was also determined. Data
+are expressed as the mean % GFP cells detected by flow cytometry (n=6-7, homing and early stage times, n=15, end stage, from two separate experiments; student t-test: *, p<0.05). Vertical bars show standard error of the mean.
Spleenb. Liverc.
Bloode.Lymph nodesd.p=0.04
Bone marrowa.
126
5.3.3 Response to single agent chemotherapy Experiments were performed to determine if CTGF overexpression in BCP ALL
cells played a role in drug sensitivity to single dose chemotherapy agents VCR
or ARAC. Chemotherapy treatment or saline were administered to PER-278
xenografts from day 25 post-injection, determined by the first mouse to display
1% engraftment in peripheral blood (Figure 5.4a). Kaplan-Meier survival plots
were constructed for both VCR and low dose ARAC chemotherapy protocols
(Figure 5.4b and 5.4c). Survival outcomes due to VCR treatment were
significantly higher compared with saline controls for both PER-278 Low (79
days versus 49 days survival, p=0.0003) and PER-278 High (77.5 days versus
49 days survival, p=0.004). Survival outcomes due to low dose ARAC were
significant for PER-278 Low xenografts, with 63 days versus 49 days survival in
control groups (p=0.003; Figure 5.3d). However, PER-278 High xenografts
while displaying a trend for prolonged survival due to drug treatment compared
with saline controls, did not reach significance with 59.5 days versus 49 days
survival (p=0.23; Figure 5.3e). Both PER-278 High and Low xenografts were
more sensitive to VCR, with longer survival times after the last chemotherapy
dose, than to low dose ARAC.
Treatment of PER-278 xenografted mice with a maximally tolerated dose of
ARAC (100mg/kg) found PER-278 Low cells responded with high sensitivity
and with minimal numbers succumbing to drug toxicity (1/7 mice). PER-278
High xenografts responded to treatment with reduced leukaemia engraftment
seen in the BM, however due to drug toxicity all mice (6/6) were unable to finish
two rounds of treatment (Supplementary Figure S1.1). This prevented a proper
comparison of high dose ARAC response between PER-278 High and Low cell
lines, though low BM engraftment of leukaemia cells at time of sacrifice
indicated they were sensitive to higher ARAC doses.
Figure 5.4. Kaplan-Meier survival curve of PER-278 High and Low engrafted NOD/SCID xenografts treated with single agent chemotherapy
5Mice were injected via the tail-vein with 2x10 PER-278 High or Low cells. a. Percentage leukaemic engraftment in the peripheral blood determined treatment initiation at day 25 post-injection. b. PER-278 Low or c. PER-278 High xenografted mice were treated with saline (dashed line) or with 0.5 mg/kg VCR once a week for 4 weeks. d. PER-278 Low or e. PER-278 High xenografted mice were treated with saline (dashed line) or with 0.5 mg/kg ARAC, five times a week then left for 2 weeks before another 5 days of treatment. Grey arrows indicate first and last drug injection. Data are expressed as the percent survival of mice against time. (Saline treated, n=15 from two separate experiments. Chemotherapy treated, n= 5-6). (Log-Rank test: **, p<0.01, ***, p<0.001).
b. c.
d. e.
a.
128
5.3.5 Snapshot of disease 24 hours post ARAC dose To further characterise the reduced sensitivity to low dose ARAC seen in PER-
278 High xenografts, a snapshot of ARAC response was taken. Prior to
establishing xenografts, CTGF expression in secreted protein was confirmed to
be upregulated in PER-278 High cells with a 3.14-fold greater expression than
PER-278 Low cells (Figure 5.5a). Mice were sacrificed 24 hours after a single
ARAC dose, administered within the ARAC protocol on the first day of the
second round of treatment. In addition to ARAC, BrdU (1mg) was delivered
intraperitoneally for detection of proliferation.
BM engraftment levels for both PER-278 High and Low xenografts were
reduced in the ARAC snapshot compared with end stage disease, confirming
suitable timing for measuring the effect of ARAC on an earlier stage of disease
progression. A trend was evident for increased engraftment in PER-278 Low
BM compared with PER-278 High, with 57.6% and 32.0% engraftment
(p=0.052; Figure 5.5b). The reduced mean BM engraftment seen in PER-278
High cells was partly a result of increased variation in engraftment between
mice, with the data suggesting only a sub-population of PER-278 High cells
contained reduced numbers of leukaemia cells in the BM.
Gating cells by flow cytometry that were positive for both BrdU and GFP
antibodies allowed detection of proliferating leukaemia cells within the whole
BM sample. In the BM population, PER-278 Low cells showed a trend for more
proliferating leukaemia cells compared with PER-278 High (p=0.13; Figure
5.5c). To account for the slight increase in the numbers of leukaemia cells
present in PER-278 Low compared with PER-278 High BM, proliferation was
assessed as a percentage of the GFP+ leukaemia cell sub-population within the
BM, consequently PER-278 High and Low xenografts were found to have a
similar mean percentage of proliferating cells (Figure 5.5d). Proliferation of
stromal cells was also comparable between PER-278 High and Low
xenografted mice, as defined by the BrdU+ GFP- population in the BM (Figure
5.5d). In summary, ARAC treatment produced no observable differences in the
proliferation of PER-278 xenografts.
PER-
278
Low
PER-
278
High
CTGF (CM)
a. Bone marrowb.
Proliferating leukaemic cells
+ +(BRDU GFP )24hrs post ARAC dose
c.
Proliferating non-leukaemic cells -+(BRDU GFP )
24hrs post ARAC dose
d.
Figure 5.5.24 hours post ARAC dose.
1 mg BromodeoxyUridine (BrdU) and a. CTGF secreted protein levels in PER-278 Low and High cells before xenograft injection. b. Leukemia engraftment in the BM 24 hours post-ARAC dose. c. Percent of proliferating
+ +leukaemic (BRDU GFP ) cells in the BM and d. percent of proliferating cells + +(BRDU ) in the leukaemic sub-population (GFP ). e. Proliferation of non-
-+leukaemic cells (BRDU GFP ) in the BM population. Data are expressed as + +the mean % GFP or BRDU cells detected by flow cytometry (n=6-7, each
dot represents an individual mouse). Vertical bars show standard error of the mean.
Snapshot of proliferation in PER-278 NOD/SCID xenografts 5Mice were injected via the tail-vein with 2x10
PER-278 High or Low cells. When peripheral blood reached 1% engraftment mice were treated with ARAC, at 0.5 mg/kg for 5 days. After 2 weeks recovery mice received one dose of ARAC in addition to
were sacrificed 24 hours later.
Proliferating cells ) +(BRDUin leukaemic sub-population
e.
% B
one
mar
row
cells
% B
one
mar
row
cells
130
Apoptotic leukaemia cells in the BM were measured by Annexin V+ GFP+ flow
cytometry. PER-278 Low xenografts were found to have a significantly greater
mean percentage of apoptotic leukaemic cells compared with PER-278 High
(p=0.039, Figure 5.6a). Again to account for the slight increase in leukaemia
cells in the BM of PER-278 Low xenografted mice, apoptosis was counted as a
percentage of the GFP+ leukaemic sub-population. A reverse in the trend for
apoptotic leukaemia cells was observed, with 6.0% of PER-278 High leukaemia
cells undergoing apoptosis compared with 3.9% in PER-278 Low leukaemia
cells (p=0.16; Figure 5.6b). The data suggests the existence of two separate
cell populations within PER-278 High leukaemia cells, with only a subset of cells
in PER-278 undergoing increased apoptosis (3/ 7 mice).
As no significant differences where observed in apoptosis 24 hours after ARAC
treatment, control mice were used to compare apoptosis levels. BM cells from
untreated PER-278 xenografts at end stage disease were used as a baseline
measure of apoptosis in leukaemia-engrafted mice. No significant difference
was found due to CTGF overexpression in the percentage of apoptotic
leukaemic cells or apoptotic stromal cells in the BM of PER-278 xenografts
(Figure 5.6c and 5.6d). In comparison to control mice, the ARAC snapshot
contained a higher population of stromal cells undergoing apoptosis than
leukaemic cells (6.6% leukaemic versus 0.44% stromal apoptotic cells,
p=0.0001; Figure 5.6). This reflects the non-specific cell targeting of ARAC,
further evidenced by stromal cells in the ARAC snapshot having higher levels of
apoptosis compared with untreated controls (6.6% ARAC versus 3.3% controls;
p= 0.02).
Apoptotic cells in leukaemic sub-population
+(Annexin V )
Figure 5.6. 5 Mice were injected via the tail-vein with 2x10
PER-278 High or Low cells. When peripheral blood reached 1% engraftment mice were treated with ARAC, at 0.5 mg/kg for 5 days. After 2 weeks recovery mice received one dose of ARAC in addition to
were sacrificed 24 hours later.poptotic cells in the
leukaemic sub-population ( within the BM. c. Percent of a d. To evaluate post-ARAC
apoptosis levels, untreated PER-278 High and Low xenografted mice received saline and were sacrificed at end stage disease. Percent of apoptotic leukaemic cells apoptotic non-leukaemic cells Data are
++ expressed as the mean % GFP or Annexin cells detected by flow cytometry (n=6-7, ARAC treated mice, n=3-6, saline treated mice. Each dot represents an individual mouse). Vertical bars show standard error of the mean. (student t-test: *, p<0.05)
Snapshot of apoptosis in PER-278 NOD/SCID xenografts 24 hours post ARAC injection.
1 mg BromodeoxyUridine (BrdU) and a. Percent of apoptotic leukaemic
+ +cells in the BM (Annexin V GFP ). b. Percent of a+GFP ) poptotic non-
+ -leukaemic cells in the BM (Annexin V GFP ).
+ +(Annexin V GFP ) was detected and e. percent of + -(Annexin V GFP ) in the BM.
poptotic leukaemia cells+ + untreated (Annexin V GFP )
Ad. e.
a. Apoptotic leukaemia cells+ +ARAC snapshot (Annexin V GFP )
b.
p=0.039
Apoptotic non-leukaemic cells+ -ARAC snapshot (Annexin V GFP )
c.
% le
ukae
mia
cel
ls
Apoptotic non-leukaemic cells + - untreated (Annexin V GFP )
% B
one
mar
row
cells
% B
one
mar
row
cells
% B
one
mar
row
cells
% B
one
mar
row
cells
% L
euka
emia
cel
ls
132
5.3.6 PER-278 Low cells could not be sensitized to ARAC by secreted proteins from BM co-cultures Experiments were performed in culture to ascertain if secreted products from
co-cultures of PER-278 and HS5 BM stromal cells could alter PER-278 Low cell
viability to ARAC. PER-278 Low cells were exposed to serial concentrations of
ARAC in the presence of conditioned media from either HS5 co-incubations
with PER-278 Low cells or co-incubations with PER-278 High cells. Conditioned
media was not successful in modifying PER-278 Low response to ARAC
(Figure 5.7a).
For ARAC to be activated it must cross the cellular membrane. To discover if
transfer of ARAC across the cell membrane could be altered in PER-278 High
xenografts, uptake of tritiated [3H]uridine was measured. Uptake of [3H] uridine
is measured over a much smaller time period than cell viability, 15 minutes
compared with 72 hours, and is therefore capable of detecting differences
missed after long incubation times. Uridine uses the same equilibrative
nucleoside transporter (ENT1) as ARAC, in addition to two other concentrative
nucleoside transporters. Therefore, changes seen in the uptake of [3H]uridine
indicate potential differences in ARAC uptake. However, no difference was
found in [3H]uridine uptake by PER-278 Low cells in the presence of secreted
proteins from stromal (HS5) cells co-incubated with rhCTGF or PER-278 High
cells compared with HS5 cells alone (Figure 5.7b).
Reduced sensitivity to low dose ARAC in PER-278 High xenografts was not
reproduced in PER-278 Low cells via the activity of secreted proteins produced
by stromal cells exposed to high CTGF levels.
Figure 5.7. Reduced drug sensitivity to ARAC by PER-278 High cells was not replicated in presence of secreted proteins from stromal-leukaemia cell interactions. Secreted products in conditioned media were collected from HS5 stromal cells cultured with either PER-278 High cells, PER-278 Low cells or rhCTGF (100ng/ml) for 48 hours. a. Cell viability of PER-278 Low cells treated with ARAC was measured in the presence of HS5+PER-278 Low (open circle) or HS5+PER-278 High (closed circle) conditioned media (CM). Data are expressed as the mean % difference (treated vs untreated) in reduction of Alamar Blue (n=3). b. Nucleoside transporter activity, to determine uptake of ARAC, was measured in PER-278 Low cells after 15
3mins in the presence of [ H]Uridine with either normal media, media+rhCTGF, CM from HS5, HS5+CTGF or HS5+PER-278 High co-cultures. Data are expressed as the mean isotope count per minute (n=3). Vertical bars show standard error of the mean. (student t-test: *, p<0.05)
a.
b.
134
5.4 Discussion The use of a maximum tolerated dose of ARAC in xenografted mice was
successful in prolonging survival regardless of CTGF expression levels,
therefore a low dose concentration of ARAC was used to increase the likelihood
of acquired drug resistance. The insights gleaned from this low dose resistance
then served to highlight the capability of CTGF to influence response to
chemotherapy drugs. Treatment with a low dose ARAC found increased
survival in PER-278 High compared with PER-278 Low xenografts. Follow up
experiments failed to find the responsible mechanism. Proliferation and
apoptosis were measured in the BM of PER-278 xenografted mice 24 hours
post ARAC treatment, PER-278 High xenografts revealed a trend for reduced
levels of apoptosis compared with PER-278 Low xenografted mice. Peak ARAC
activity is possibly not captured 24 hours post drug administration, with studies
in humans indicating that the plasma levels of ARAC after bolus injection peak
after 15 minutes and were undetectable by 6 hours after injection (Spriggs, et
al., 1985). Additionally, the use of male and female mice for time points and
initial survival experiments might present differences in proliferation and
apoptosis that are unrelated to leukaemia engraftment, although no differences
in survival were found between untreated PER-278 High and Low xenografts in
either mouse gender. Another factor that could mask CTGF effects is the timing
of ARAC delivery, with the possibility that CTGF has not established drug-
protective modifications to the microenvironment at the initiation of treatment.
Further experimental work is essential to clarify the significance of drug
resistance due to CTGF in a leukaemia xenograft model.
A significant finding that resulted form CTGF overexpression in PER-278
xenografted mice, involved the increased engraftment of leukaemia cells in the
lymph nodes of PER-278 High compared with PER-278 Low xenografts.
Increased lymph node metastases due to CTGF have been previously reported
in gastric, breast and pancreatic cancer (Dornhofer, et al., 2006; Liu, et al.,
2008; Xie, et al., 2001). The current NOD/SCID xenograft model fails to address
whether leukaemia cells in the lymph node are a population expanded from
time of tail-vein injection or a case of true metastasis from BM-originating
135
leukaemia cells. The clinical impact of higher lymph node engraftment in BCP
ALL due to CTGF expression requires further validation.
The homing process is a response to long distance adhesion and migration
cues usually delivered by circulating cytokines. In the xenograft mouse model,
systemic delivery of leukaemic cells via the tail-vein of NOD/SCID mice allows
for leukaemia-driven positioning within the BM in response to interactions with
the relevant cytokines and chemotactic signals. PER-278 High and Low
NOD/SCID xenografts were found to have equivalent levels of homing to the
BM, spleen and liver after 24 hours, indicating that CTGF played no role in
influencing leukaemia cell systemic migration. In the present study, homing was
assessed by flow cytometry, as the technique allows a higher sample rate and
robust antibodies compared with immunohistochemical methods. However, the
geography of leukaemia cells is not assessed and the leukaemia cell niche
within the BM has been indicated as important for perturbing normal HSC
function (Schepers, et al., 2013a). Progression of leukaemia cells from homing
to early and end stage disease was comparable between PER-278 High and
Low xenografted mice. While this indicates CTGF does not play a role in
expansion of leukaemia cells in PER-278 cells, it also provides an unbiased
baseline to test for differences in response to chemotherapy drugs. Survival and
drug response findings highlight that PER-278 mouse xenografts responded in
a unique fashion to CTGF overexpression compared with PER-371 xenografted
mice, which are discussed in Chapter 6.
PER-278 xenografts showed no difference in homing, disease progression or
overall survival between CTGF High and Low expressing cell lines.
Overexpression of CTGF in PER-278 xenografts resulted in significantly greater
peripheral lymph node engraftment at end stage disease, though this did not
modify survival outcomes. Inreased CTGF reduced sensitivity to low dose
chemotherapy in PER-278 xenografts. The mechanism of drug resistance
remains to be determined with repeated in vivo modelling needed. These
results provide new evidence in BCP ALL that CTGF exists as functionally
active protein within the context of the BM microenvironment and could
influence drug resistance in the clinical setting.
666666666CTGF confers a growth advantage
to PER-371 B-precursor ALL cells in the BM microenvironment
We shall not cease from explorationAnd the end of all our exploring
Will be to arrive where we startedAnd know the place for the first time.
T.S. Eliot
138
139
Chapter 6 CTGF confers a growth advantage to
PER-371 B cell-precursor ALL cells in the BM microenvironment
6.1 Abstract B cell-precursor (BCP) ALL is a cancer of genetic mutations. Initial
chromosomal rearrangements in the disease have been well defined however
the additional mutations required to initiate malignancy are still not clearly
understood. Recent studies indicate that at diagnosis the primary leukaemia
contains a multitude of sub-clones with distinct DNA mutations. In BCP ALL, the
most deregulated gene was identified as CTGF through microarray studies of
primary patient specimens. Overexpression of CTGF in BCP ALL was
associated with poor outcome in patients identified as high-risk at diagnosis.
Functionally defining the role of CTGF in BCP ALL is vital in order to expand on
and discover mechanisms of action relevant to patient outcomes in the disease.
To define a functional role for CTGF in BCP ALL, NOD/SCID xenograft mouse
models were established using a primary patient-derived cell line, PER-371,
engineered to overexpress CTGF. Leukaemia cell expansion, survival and
response to ARAC chemotherapy were analysed with respect to High and Low
CTGF expression. CTGF was not found to alter drug response to ARAC or
affect early leukaemia engraftment in xenografts with this cell line. However,
increased BM engraftment at late stage disease was found in PER-371 CTGF
High xenografted mice with significant difference in the percentage of
leukaemia cells compared with CTGF Low controls (43.7% versus 6.3%,
p=0.01). This differential engraftment at late stage correlated with a more
aggressive disease, as a significant reduction in overall survival was seen in
PER-371 CTGF High compared with CTGF Low xenografted mice (89 versus
70 days post-injection, p=0.01). The mechanism by which CTGF
overexpression led to more aggressive disease in PER-371 xenografts were
due to the influence of CTGF on stromal-leukaemia cell interactions. High
CTGF caused increased stromal cell proliferation at early stage disease,
increased reticulin fibre deposition at late stage disease and altered secretion of
proteins that are associated with ECM remodelling. This study defines a new
140
functional role for CTGF in BCP ALL, working indirectly through BM
microenvironment interactions to enable enhanced disease progression.
6.2 Introduction Childhood BCP ALL results from two or more cytogenetic ‘hits’ that occur in
normal B-cell progenitors, resulting in arrested development at the BCP cell
state. The initial genetic mutations have been well defined for the major sub-
types and appear to originate largely from a pre-natal period. However, the
additional essential genetic alterations required for the pre-leukaemic cell to
become malignant are not well defined. Genomic profiling of patients at
diagnosis has revealed a multitude of sub-clone populations within the ‘bulk’
leukaemia cell mass (Mullighan, 2013). This indicates that early leukaemic
progenitors frequently produce genetically divergent daughter cells.
In BCP ALL, the CTGF gene was 19-fold overexpressed compared with CD45+
haematopoietic progenitor cells (Boag, et al., 2007). Overexpression of CTGF
was found in 75% of BCP ALL paediatric cases and associated with a gene
subset indicative of poor outcome in patients identified as high-risk at diagnosis
(Kang, et al., 2010). Identification of a functional mechanism, whereby CTGF
could influence disease outcome in BCP ALL, has not yet occurred.
The current study investigates a functional role for CTGF in BCP ALL using the
cell line PER-371, which has previously been found to strongly resemble the
disease characteristics of the primary patient sample (see section 3.3). For the
purposes of detecting differences attributable to higher CTGF expression,
lentiviral technology was used to establish two PER-371 cell populations, CTGF
High and Low (see section 2.2.5-7). In chapter 5, a potential role for the PER-
278 cell line in reduced sensitivity to low-dose ARAC was explored. Despite
many common characteristics between PER-278 and PER-371, mouse
xenograft models revealed pronounced differences in phenotype. For this
reason the two cell lines have been presented separately to highlight their
distinct outcomes, however a consistent approach in methodology and
experimental outline were followed between the two chapters.
NOD/SCID mice were inoculated with 1x106 PER-371 High or Low cells via tail-
vein injection (see section 2.7.2). The effect of high CTGF expression in PER-
371 xenografts was assessed at four times; at homing, early and late
141
tumourigenesis and end stage disease, which included the leukaemic cell
response to low-dose ARAC. Differences in survival outcomes of PER-371
xenografts due to increased CTGF expression were investigated further to
decipher essential mechanisms. Mechanistic studies included proliferation
assays and apoptosis assays in vivo. Complementary immunostaining of BM
sections to characterise megakaryocytes, reticulin fibre deposition, leukaemia
and proliferating cells was also carried out. Lastly, identification of modified
secreted proteins was performed using conditioned media from co-culture of
PER-371 CTGF High cells with HS5 stromal cells.
6.3 Results 6.3.1 CTGF overexpression in cells used for PER-371 xenografts Overexpression of CTGF in conditioned media of PER-371 High cells was
confirmed by western blot before tail-vein injection of NOD/SCID mice (Figure
6.1). Fold change differences in CTGF expression between PER-371 High and
PER-371 Low cells were interpreted for RNA and conditioned media (Table
6.1). Conditioned media overexpression was between more than 11.8 and 26.5-
fold higher in PER-371 High cells compared with Low, which is greater than the
originally detected levels after lentiviral transduction. Differences in secreted
media for CTGF expression are likely due to the scant and therefore variable
detection of PER-371 Low CTGF in western bands from conditioned medium.
PER-371 High/LowHoming (24hrs)
PER-371 High/LowLate disease (Wk11)
PER-371 High/LowEarly disease (Wk3)
PER-371 High/LowEnd stage disease
1.06
1.59
1.06
1.66
RNA (qRT-PCR)
11.81
26.46
26.46
Fold Change Secreted protein (western)
13.47
Table 6.1. 6before injection of 1x10 cells into the tail vein of NOD/SCID mice. CTGF
RNA expression was detected by qRT-PCR and secreted protein levels by densitometry on western blot bands. Leukaemia engraftment was analysed at 4 times (homing, early, late and end stage disease).
Fold change difference between PER-371 High and Low cells
PER-
371
Low
CTGF (CM)
PER-
371
Low
PER-
371
High
PER-
371
High
PER-
371
Low
PER-
371
High
Hom
ing
at
e st
age
Early
&
l E
nd s
tage
Figure 6.1. and Low cells intended for NOD/SCID inoculation. Secreted protein from PER-371 High and Low conditioned media (CM) was detected by western blot
6prior to injection of 1x10 cells into the tail vein of NOD/SCID mice. Leukaemia engraftment was analysed at 4 times (homing, early , late and end stage disease).
CTGF secreted protein expression levels for PER-371 High
143
6.3.2 Survival outcomes and response to single agent chemotherapy Survival studies of PER-371 xenografts found PER-371 High cells resulted in a
more aggressive disease with a significantly reduced median survival, 89 days
post-injection, compared with 70 days in PER-371 Low xenografts (p=0.01;
Figure 6.2a). This shows CTGF was capable of accelerating leukaemic growth
in PER-371 xenografted mice.
To characterise chemotherapy response, PER-371 xenografted mice were
started on a low dose ARAC treatment protocol from day 40, after the first
mouse reached 1% engraftment in peripheral blood. Low-dose ARAC treatment
significantly extended survival in PER-371 High xenografted mice, to 84.5 days
compared with 70 days survival in saline controls (p=0.004; Figure 6.2b).
Surprisingly, PER-371 Low xenografted mice treated with low-dose ARAC
showed no difference in survival compared with controls, with 82 days for ARAC
treatment versus 89 days survival for saline controls (p=0.11; Figure 6.2c). As
survival curves for both groups of ARAC treated xenograft mice found they
succumbed to disease in a similar time frame post ARAC treatment, significant
differences in survival between treated and untreated groups is a consequence
of differential survival in the untreated mice. As reduced median survival in
untreated PER-371 High xenografted mice showed a distinct separation in
timing with the last ARAC injection and the increased median survival in
untreated PER-371 Low xenografted mice overlapped with the end of ARAC
treatment. Additionally, both PER-371 High and Low xenografted mice
responded with high sensitivity to high-dose ARAC with little toxicity
(Supplementary Figure S1). Collectively, PER-371 cells were found to have no
difference in response to ARAC due to CTGF expression.
Figure 6.3. PER-371 NOD/SCID xenografts showed increased BM engraftment in CTGF High compared to Low cells at late stage disease.
6Mice were injected via the tail-vein with 1x10 PER-371 High and Low cells. Bone marrow engraftment was determined at 4 times: a. homing at 16 hours post injection, b. early stage disease at 8 weeks, c. late stage disease at 10
+weeks and d. end stage disease. Data are expressed as the mean % GFP cells detected by flow cytometry (n=6-7, homing, early and late stage times, n=11-12, end stage. Each dot represents an individual mouse). Vertical bars show standard error of the mean. (student t-test: *, p<0.05).
b.a.
c. d.
Homing (bone marrow) Early stage (bone marrow)
Late stage (bone marrow) End stage (bone marrow)
p=0.01
145
6.3.3 Organ engraftment throughout spread of leukaemia The percentage of engrafted leukaemia cells in the BM and extramedullary
organs was assessed at four times after injection. Homing of cells to the BM
was assessed at 16 hours, early stage disease at 3 weeks, late stage disease
at 11 weeks and end stage disease, when mice were sacrificed due to clinical
presentation of leukaemia.
In the BM, the primary site of leukaemogenesis, CTGF had no effect on
homing, with no difference in infiltration of leukaemia cells between CTGF High
and Low groups (Figure 6.3a). At early stage disease a trend for higher BM
engraftment was seen for PER-371 High compared with Low xenografted mice
and by late stage disease mean engraftment between the two groups was
significantly different, with 43.7% engraftment in High and 7.6% engraftment in
Low xenografted mice (p=0.01; Figure 6.3b and 6.3c). At late stage disease a
similar trend for increased engraftment of PER-371 High cells compared with
PER-371 Low is evident in the spleen, liver, lymph nodes and blood (Figure
6.4). Leukaemia engraftment in the BM at end stage disease, which is an
independent time for each group, returned to comparable levels of engraftment,
with 65.6% in PER-371 High and 76.7% in PER-371 Low xenografted mice
(Figure 6.3d). In the BM, increased leukaemia engraftment due to CTGF were
found from early stage disease with the separation between High and Low
CTGF-expressing cells most significant at late stage disease.
PER-371 cells were capable of early homing to the BM, with both PER-371
High and Low cells detectable at 16 hours post tail-vein injection, with a mean
of 0.005% and 0.002% GFP+ cells, respectively (Figure 6.3a). The only other
organ with detectable infiltration of GFP+ cells at 16 hours was the spleen, with
0.003% in both CTGF High and Low xenografted mice (Figure 6.4b). For all
examined organs, the overall percentage of leukaemia cells detected at early
stage disease was increased compared with 16 hours post injection regardless
of CTGF expression (Figure 6.4). The overall percentage of leukaemic cells
were found to increase again at late stage disease and reached a peak at end
stage disease engraftment. Across both xenograft groups slower early
leukaemic growth was followed by rapid expansion near end stage disease.
Figure 6.2. engrafted NOD/SCID xenografts treated with chemotherapy drugs or saline. a. PER-371 Low (dashed line) and PER-371 High xenografted mice (solid line) were untreated and injected only with 0.9% saline on the same days as chemotherapy treated mice. Percentage leukaemic engraftment in the peripheral blood determined treatment initiation with ARAC at day 40 post-injection with low dose ARAC (0.5 mg/kg), five times a week then 2 weeks recovery before another 5 days of injections or treated with saline (dashed line) for b. PER-371 High and c. PER-371 Low xenografted mice. Grey arrows indicate first and last ARAC injection. Data are expressed as the percent survival of mice against time. (saline treated, n=11-12, chemotherapy treated treated, n= 5-6). (Log-Rank test: **, p<0.01, ***, p<0.001).
Kaplan-Meier survival curve of PER-371 High and Low
c.
a.
b.
a.
50
Figure 6.4. PER-371 NOD/SCID xenografts. Leukaemia engraftment was analysed at 4
6 times (homing, early, late and end stage disease) after injection of 1x10PER-371 High (triangle) or Low cells (circle). a. Engraftment in BM. Engraftment in the extramedullary organs: b. spleen, c. liver and d. lymph nodes. e. Leukaemia burden in peripheral blood was also determined. Data
+are expressed as the mean % GFP cells detected by flow cytometry (n=6-7, homing, early and late stage times, n=11-12, end stage. Each dot represents an individual mouse). Vertical bars show standard error of the mean. (student t-test: *, p<0.05).
Rapid engraftment of leukaemia at end stage disease for
BloodLymph nodes
Bone marrow Spleen
Liver
b.a.
c.
d. e.
p=0.04
148
6.3.4 Snapshot of proliferation and apoptosis at late stage disease To understand differences seen in BM engraftment at late stage disease
proliferation (detected 24 hours after BrdU injection) and apoptosis (detected by
Annexin V labelling) were measured in xenograft BM cells by flow cytometry.
When assessed as a proportion of total BM, there was a trend for higher
proliferation of leukaemic cells (mean BrdU+ GFP+) in PER-371 High
xenografted mice (2.3%) compared with Low (0.22%), although this did not
reach statistical significance (p=0.10; Figure 6.5a). As there were more PER-
371 High cells overall in the BM, we assessed the number of proliferating cells
as a proportion of leukaemic cells. As a percentage of leukaemic cells, the trend
for higher levels of proliferation in PER-371 High xenografted mice is less
pronounced (Figure 6.5b). The number of proliferating non-leukaemic cells in
the BM was found to be comparable between CTGF High and Low xenografted
mice, with 2.9% and 2.2% mean BrdU+ GFP- cells respectively (Figure 6.5c).
Overall no discernible differences in proliferation were found due to CTGF
expression from PER-371 cells.
Apoptosis was then measured at late stage disease in the whole BM of
xenografted mice. When measured as a proportion of total BM cells, there were
significantly more apoptotic leukaemia cells in PER-371 High xenografted mice
with a mean of 0.67% compared with PER-371 Low xenografted mice, with a
mean of 0.3% of BM cells (p=0.04; Figure 6.5d). To account for the differences
in overall leukaemic cell numbers, apoptotic cells were measured as a
proportion of the leukaemic cell sub-population in the BM. When analysed in
this way, there was a trend for decreased apoptosis in the leukaemic
subpopulation of PER-371 High xenografted mice compared with Low, with a
mean of 16% and 37% apoptotic cells respectively (p=0.12; Figure 6.5e). The
percentage of Annexin+ GFP- non-leukaemic apoptotic cells in the BM, was
similar between groups (Figure 6.5f). Overall a small trend for decreased
apoptosis in the BM of PER-371 High compared with PER-371 Low xenografted
mice was found at late stage disease.
a.b.
Prol
ifera
ting
leuk
aem
ic ce
lls+
+ in
the
BM (
BRDU
GFP
)
Apo
ptot
ic le
ukae
mic
cells
++
in th
e BM
(Ann
exin
GFP
)Ap
opto
tic n
on-le
ukae
mic
cells -
+
(Ann
exin
GFP
)in
the
BM
Pro
lifera
ting
non-
leuk
aem
ic ce
lls-
+
(BRD
U G
FP)
in th
e BM
c.
d.
Pr
olife
ratin
g ce
lls
in le
ukae
mic
sub-
popu
latio
n
+(B
rdU
)
Apop
totic
cel
ls in
leuk
aem
ic su
b-po
pula
tion
+ (A
nnex
in)
p=0.
044
e.f.
% Bone marrow cells % Bone marrow cells
% Bone marrow cells
% Bone marrow cells% Bone marrow cells
Figure 6.5. Snapshot of proliferation and apoptosis in PER-371 cells at late stage disease.
6 disease, 11 weeks after injection of 1x10 PER-371 High (square) or Low cells (circle) onto NOD/SCID mice. a.
d.
f. Data are expressed as the mean %
+ + + GFP , Annexin or BRDU cells detected by flow cytometry (n=6-7, each dot represents an individual mouse). Vertical bars show standard error of the mean. (student t-test: *, p<0.05)
Leukaemia engraftment was analysed at late stage
Percentage of proliferating leukaemic + +cells (BRDU GFP ) in the BM and b. percentage of proliferating cells
+ +(BRDU ) in the leukaemic sub-population (GFP ) within the BM. c. + -Percentage of proliferating non-leukaemic cells (BRDU GFP ) in the BM.
+ +Percentage of apoptotic leukaemic cells (Annexin GFP ) in the BM and e. +percentage of apoptotic cells (Annexin ) in the leukaemic sub-population
+(GFP ) within the BM. Percentage of apoptotic non-leukaemic cells + -(BRDU GFP ) in the bone marrow.
151
6.3.5 Immunohistochemistry on BM sections of PER-371 xenografted mice Bone marrow sections from PER-371 Low and High xenografted mice were
immunohistochemically stained for detection of leukaemia cells, with an anti-
Pax5 antibody, proliferating cells, with an anti-Ki67 antibody and
megakaryocytes, with an anti-VwF antibody. Additionally, an Ag+ stain was
performed to identify reticulin fibres within the BM. Multispectral imaging of both
the diaphysis and epiphysis/metaphysis of the mouse femur and tibia was
obtained before analyses of positive staining by calculating the percentage of
antibody-stained cells in the BM compared with haematoxylin-stained nuclei.
Pax5+ immunostaining detecting leukaemia cells
Leukaemia cell detection by IHC allowed analysis of the location and proximity
of tumour cells. In the diaphysis, leukaemia cells were more commonly found as
a mass of cells than dispersed freely (Figure 6.6b). Whereas, in the epiphysis
and metaphysis leukaemia cells were found as small populations of cells that
were more dispersed throughout the space (Figure 6.7b). At early stage
disease, the first detectable groups of Pax5+ cells were more commonly found
proximal to the epiphyseal plate in the metaphysis and sometimes epiphysis.
Differences in leukaemic engraftment due to CTGF were found at late stage
disease with a greater percentage of Pax5+ cells found in the diaphysis of PER-
371 High compared with PER-371 Low BM sections, with a mean of 56% and
10% Pax5+ cells, respectively (p=0.008; Figure 6.6a). Whereas, at early and
end stage disease in the diaphysis there was no significant difference between
the mean percentage of Pax5+ cells in PER-371 High compared with PER-371
Low xenografted mice. The epiphysis and metaphysis revealed similar findings
as the diaphysis in the BM with an increased mean percentage of Pax5+ cells
found in PER-371 High (with 60%) compared with PER-371 Low xenografted
mice (with 24%) at late stage disease (p=0.01; Figure 6.71a). A trend was found
at early stage disease in the epiphysis and metaphysis, with double the average
number of Pax5+ cells in PER-371 High compared with Low xenografted mice
which did not reach significance. At both early and end stage disease was there
no significant difference in the percentage of Pax5+ cells between the two
groups, therefore CTGF was found to significantly influence the engraftment
level of Pax5+ leukaemic cells specifically at late stage disease in the BM.
152
Ki67+ immunostaining detecting proliferating cells
To determine if leukaemia cells were the predominant cell type undergoing
proliferation, Ki67+ staining was performed. No pattern was evident for Ki67+
staining in the BM, instead large numbers of Ki67+ proliferating cells were found
throughout the diaphysis, epiphysis and metaphysis at all stages of disease
progression (Figure 6.8b and 6.9b). At early stage disease, a greater
percentage of Ki67+ cells were found in PER-371 High compared with PER-371
Low BM sections in the diaphysis (p=0.02) and epiphysis/metaphysis (p=0.03;
Figure 6.8a and 6.9a). Whereas at late and end stage disease there there was
a similar population of Ki67+ cells between the two groups. Increased Ki67+ cells
in PER-371 High compared with PER-371 Low xenografted mice at early stage
disease preceded the difference found in Pax5+ cells at late stage disease.
VwF+ immunostaining detecting megakaryocytes
To investigate the influence of the immune presence in NOD/SCID mice, VwF+
staining was performed detecting megakaryocytes. Megakaryocytes were
present throughout the BM with distinctive VwF+ staining, blood vessels were
also detected with the VwF antibody, but the specificity for endothelial cells was
not sufficient for use in quantification (Figure 6.10b). Megakaryocytes were
primarily associated with a vascular morphology. The average number of
megakaryocytes was assessed across the entire BM space and not segregated
into diaphysis and epiphysis/metaphysis. PER-371 High and PER-371 Low
xenografted mice contained comparable counts of megakaryocytes in the BM
(Figure 6.10a). The highest number of megakaryocytes was found at early
stage disease and decreasing populations were detected at late and end stage
disease. The lowest number of megakaryocytes at end stage resulted in a 13-
fold reduced cell count for PER-371 High and a 3-fold reduced cell count for
PER-371 Low xenografted mice when compared with early stage disease. At
late stage disease, differential populations of megakaryocytes were found, with
a higher count in PER-371 Low compared with PER-371 High BM sections (p=
0.004; Figure 6.10a). Overall, megakaryocytes were found to decrease in
number across the BM as the leukaemia stage progressed, with high CTGF
expression from PER-371 cells resulting in reduced megakaryocyte numbers at
late stage disease.
b.
a.
Pax5PER-371 HighPER-371 Low
Early stage disease
PER-371 Low PER-371 HighLate stage disease
PER-371 Low PER-371 HighEnd stage disease
Diaphysis
Figure 6.66marrow. Mice were injected via the tail-vein with 1x10 PER-371 High and
Low cells. At 10 weeks mice were sacrificed and bone marrows were fixed in paraformaldehyde for 48 hours and later decalcified in 10% EDTA for 7 days and embedded in paraffin wax for immunohistochemistry. a. Leukaemic cells in the diaphysis, detected by immunohistochemistry with a paired box 5 antibody (Pax5, 1:100 dilution). Data are expressed as % positively stained nuclei (n=6-7, early and late stage. n=5, end stage.) Vertical bars represent standard error of the mean. b. Representative diaphysis sections of PER-371 Low and High at early, late and end stage time points. Scale bar represents 500mm.
. Leukemic-burden in diaphysis of PER-371 engrafted bone
b.
a.
Pax5PER-371 HighPER-371 Low
Early stage disease
PER-371 Low PER-371 HighLate stage disease
PER-371 Low PER-371 HighEnd stage disease
Epiphysis/ Metaphysis
Figure 6.7. Leukemic-burden in epiphysis/metaphysis of PER-371 6engrafted bone marrow. Mice were injected via the tail-vein with 1x10 PER-
371 High and Low cells. At 10 weeks mice were sacrificed and bone marrows were fixed in paraformaldehyde for 48 hours and later decalcified in 10% EDTA for 7 days and embedded in paraffin wax for immunohistochemistry. a. Leukaemic cells in the epiphysis/metaphysis, detected by immunohistochemistry with a paired box 5 antibody (Pax5, 1:100 dilution). Data are expressed as % positively stained nuclei (n=6-7, early and late stage. n=5, end stage.) Vertical bars represent standard error of the mean. b. Representative epiphysis/metaphysis sections of PER-371 Low and High at early, late and end stage time points. Scale bar represents 500mm.
b.
a.
Ki67
PER-371 HighPER-371 LowEarly stage disease
PER-371 Low PER-371 HighLate stage disease
PER-371 Low PER-371 HighEnd stage disease
Diaphysis
Figure 6.86burdened bone marrow. Mice were injected via the tail-vein with 1x10 PER-
371 High and Low cells. At 10 weeks mice were sacrificed and bone marrows were fixed in paraformaldehyde for 48 hours and later decalcified in 10% EDTA for 7 days and embedded in paraffin wax for immunohistochemistry. a. Proliferating cells in the diaphysis, detected by immunohistochemistry with a Ki67 antibody (1:100 dilution). Data are expressed as % positively stained nuclei (n=6-7, early and late stage. n=5, end stage.) Vertical bars represent standard error of the mean. b. Representative diaphysis sections of PER-371 Low and High at early, late and end stage time points. Scale bar represents 500mm.
. Proliferating cells in diaphysis of PER-371 leukaemia-
b.
a.
Ki67PER-371 High
PER-371 Low PER-371 High
PER-371 Low
Late stage disease
Early stage disease
PER-371 Low PER-371 HighEnd stage disease
Epiphysis/ Metaphysis
Figure 6.9leukaemia-burdened bone marrow. Mice were injected via the tail-vein with
61x10 PER-371 High and Low cells. At 10 weeks mice were sacrificed and bone marrows were fixed in paraformaldehyde for 48 hours and later decalcified in 10% EDTA for 7 days and embedded in paraffin wax for immunohistochemistry. a. Proliferating cells in the epiphysis/metaphysis, detected by immunohistochemistry with a Ki67 antibody (1:100 dilution). Data are expressed as % positively stained nuclei (n=6-7, early and late stage. n=5, end stage.) Vertical bars represent standard error of the mean. b. Representative epiphysis/metaphysis sections of PER-371 Low and High at early, late and end stage time points. Scale bar represents 500mm.
. Proliferating cells in epiphysis/metaphysis of PER-371
a.
VwFPER-371 HighPER-371 Low
Early stage diseaseb.
PER-371 Low PER-371 HighLate stage disease
PER-371 Low PER-371 HighEnd stage disease
Figure 6.10. Megakaryocyte count in6burdened bone marrow. Mice were injected via the tail-vein with 1x10 PER-
371 High and Low cells. At 10 weeks mice were sacrificed and bone marrows were fixed in paraformaldehyde for 48 hours and later decalcified in 10% EDTA for 7 days and embedded in paraffin wax for immunohistochemistry. a. Megakaryocyte count, detected by immunohistochemistry with a von Willebrand factor antibody (VwF, 1:400 dilution) in the BM. Data are expressed as average number megakaryocytes per BM (sum for 6 images). (n=6-7, early and late stage. n=5, end stage.) Vertical bars represent standard error of the mean. b. Representative BM sections of PER-371 Low and High at early, late and end stage time points. Scale bar represents 500mm. VwF is a positive stain for endothelial cells (open arrow) and megakaryocytes (filled arrow).
diaphysis of PER-371 leukaemia-
163
Ag+ stain for reticulin fibres
In response to the findings of chapter 4, where human HS5 cells were found to
secrete ECM remodelling genes due to CTGF treatment, the location and
density of reticulin connective tissue was analysed by Ag+ staining. Reticulin
fibre deposition was scored on early and late stage BM sections. Fine reticular
fibres in normal BM sections were associated with blood vessels and dense
reticulin fibre networks were most often associated with the endosteum (Figure
6.11b and 6.12b). At early stage disease, increased reticulin deposition was
found in PER-371 High BM compared with PER-371 Low for both diaphysis
(Figure 6.11a) and epiphysis/metaphysis (Figure 6.12a) although this did not
reach statistical significance. In late stage disease, the same trend remained
with the average reticulin score found to be significantly higher in PER-371 High
BM compared with PER-371 Low (p=0.04; Figure 6.11a). Reticulin connective
tissue was found to be more abundant and with thicker fibre density in PER-371
High compared with PER-371 Low xenografted mice at late stage disease.
a.
Ag stain
PER-371 HighPER-371 LowEarly stage disease
b.
PER-371 Low PER-371 HighLate stage disease
Diaphysis
Figure 6.11. Reticulin connective tissue density in diaphysis of PER-371 leukaemia-burdened bone marrow. Mice were injected via the tail-vein with
61x10 PER-371 High and Low cells. At 10 weeks mice were sacrificed and bone marrows were fixed in paraformaldehyde for 48 hours and later decalcified in 10% EDTA for 7 days and embedded in paraffin wax for immunohistochemistry. a. Reticulin connective tissue density, scored by an Ag stain of the diaphysis. Data represents average reticulin score per mouse (n=6-7, early and late stage. n=4-5, end stage.) Vertical bars represent standard error of the mean. b. Representative diaphysis sections Ag stained for PER-371 Low and High at early and late stage time points. Ag detects reticulin fibres (open arrow). Scale bar represents 500mm.
a.
Ag stain
PER-371 HighPER-371 LowEarly stage disease
b.
PER-371 Low PER-371 HighLate stage disease
Epiphysis/ Metaphysis
Figure 6.12. Reticulin connective tissue density in epiphysis/ metaphysis of PER-371 leukaemia-burdened bone marrow. Mice were
6injected via the tail-vein with 1x10 PER-371 High and Low cells. At 10 weeks mice were sacrificed and bone marrows were fixed in paraformaldehyde for 48 hours and later decalcified in 10% EDTA for 7 days and embedded in paraffin wax for immunohistochemistry. a. Reticulin connective tissue density, scored by an Ag stain of the epiphysis/ metaphysis. Data represents average reticulin score per mouse (n=6-7, early and late stage. n=4-5, end stage.) Vertical bars represent standard error of the mean. b. Representative epiphysis/ metaphysis sections Ag stained for PER-371 Low and High at early and late stage time points. Ag detects reticulin fibres (open arrow). Scale bar represents 500mm.
166
Correlation between the immunohistochemical and histochemical detection of
cells within the BM
An invading leukaemia front present in a PER-371 High BM diaphysis at early
stage disease illustrates the various BM cell components in relation to each
other (Figure 6.13a). Clear overlap was seen in Ki67+ and Pax5+ staining and
an inverse relationship existed between the location of Pax5+ staining and VwF+
megakaryocytes. Reticulin connective tissue was densely deposited with thick
processes at the site of increased Pax5+ leukaemic cells and abundant fibres
also present in the BM space adjacent to the leukaemia cell front. To determine
if the patterns between Pax5 expression and Ki67, VwF expression and Ag
staining were not due to chance the Pearson’s correlation coefficient was
calculated for the immunohistochemical outcomes presented in Figures 6.6 to
6.12.
The Pearson’s correlation coefficient was used to determine if a relationship
existed among the immunohistochemical stains for PER-371 High and Low
xenografted mice across early and late stage leukaemia engraftment (Figure
6.13b). Confidence in the correlation coefficient (r value) was determined from
the r value and sample size at two alpha levels, p<0.05 and p<0.01. Across all
investigated histochemical staining relationships, PER-371 Low cells were more
commonly linked to the expression of other cell types at early stage disease,
where the opposite was true for PER371 High cells, which showed greater
significant relationships at late stage disease. Pax5+ leukemia cells were found
to have a strong linear relationship with Ki67+ proliferating cells at late
compared with early stage disease for both PER-371 High and Low xenograft
mice, with the exception of PER-371 Low engrafted cells in the diaphysis. An
inverse relationship existed between Pax5+ leukemia cells and VwF+
megakaryocytes at both stages of disease progression. Surprisingly, the
inverse relationship between Pax5+ and VwF+ expression levels was significant
for PER-371 Low xenografted mice at early stage disease (r= -0.8 in the
diaphysis and -0.74 in the epiphysis/ metaphysis) and for PER-371 High
xenografted mice at late stage disease (r= -0.81 in the diaphysis and -0.85 in
the epiphysis/ metaphysis). CTGF was found to influence the relationship
between reticulin score and percentage of Pax5+ leukemic cells at late stage
disease in the diaphysis. In PER-371 High BM sections there was a strong
Figure 6.13associations significant by Pearson's correlation coefficient.
6a. Representative images from a mouse injected via the tail-vein with 1x10 PER-371 High cells and sacrificed after 3 weeks (early stage disease). BM femur and tibia were fixed in paraformaldehyde for 48 hours and later decalcified in 10% EDTA for 7 days before embedding in paraffin wax for immunostaining. Sections 4 mm thick, from the diaphysis were stained for leukaemia cells (Pax5 antibody, 1:100 dilution), proliferating cells (Ki67 antibody, 1:100 dilution), megakaryocytes (VwF antibody, 1:400 dilution) and reticulin fibers (Ag stain). Sections are at equivalent locations within the BM, with no more than 40 mm between images in the z-stack orientation. Scale bar represents 500mm. b. Pearson's correlation coefficient between Pax5 and Ki67, VwF and Ag staining immunostaining for PER-371 High and Low xenografts in the diaphysis (DIA) and epyphysis/ metaphysis (EP/ME) for both early and late stage disease. Data is expressed as correlation coefficient values (r) with red text indicating p<0.01 and blue text indicating p<0.05.
. Invading leukaemia front depicts immunostaining
Pax5
Ag stain
Ki67
VwF
Early stage disease
Late stage disease
Pearson's correlation coefficient
+ +Pax 5 and Ag0.91
PER-371 cell line
CTGF High
+ +Pax 5 and Ki67
+ +Pax 5 and VwF
0.32
-0.45
CTGF High
CTGF High
DIA
0.75
0.39
-0.52
EP/ME
0.92
0.89
-0.81
DIA
0.53
0.93
-0.85
EP/ME
-0.76 CTGF Low-0.47 -0.13 0.97
-0.80 CTGF Low-0.74 -0.65 -0.25
0.82 CTGF Low0.95 + no Agscore 0.29
a.
b.
168
linear correlation between reticulin score and Pax5+ staining (r=0.92), in
contrast to PER-371 Low BM sections where no coefficient could be measured
as none scored above zero for reticulin deposition in the diaphysis, despite a
mouse from the group demonstrating 64% Pax5+ leukemic cells. Results from
the Pearson corellation coefficients supported many of the relationships
observed among the antibodies and histochemical stains tested in PER-371
engrafted BM sections.
sections found evidence for a causal relationship between CTGF and reticulin
deposition in the diaphysis at late stage disease.
6.3.6 BCP ALL proliferation in the presence of conditioned media from stromal-leukaemia cell cultures To determine if BM secreted proteins contained the necessary elements
responsible for the differences in leukaemia cell engraftment in PER-371 High
and Low xenografted mice, conditioned media from HS5 and PER-371 cell line
co-cultures were used to assess changes to leukaemia cell proliferation in vitro.
Proliferation of BCP ALL cell lines was measured by tritiated [3H]thymidine
incorporation after a 48 hour incubation in the presence of conditioned media
from HS5 BM stromal cells co-cultured with PER-371 Low or High cells.
Secreted proteins produced by HS5 and PER-371 High co-cultures did not alter
proliferation in PER-371 High or Low cells (Figure 6.14a). The conditioned
media from HS5 and PER-371 co-cultures also had no effect on proliferation in
PER-278 High or Low cells (Figure 6.14b). In vitro proliferation assays found
that secreted proteins from BM stromal-leukaemia cell interactions were unable
to replicate the difference in leukaemia cell growth due to CTGF in PER-371
High compared with PER-371 Low xenografted mice.
Fireplicated in vitro. Pre-B ALL cell lines were cultured for 48 hours in presence of conditioned media and proliferation was measured by tritiated 3[ H]thymidine uptake. a. Proliferation of PER-371 Low cells in the presence of conditioned media from stromal cells (HS5) co-cultured with PER-371 High or Low cells. b. Proliferation of PER-278 Low cells in the presence of conditioned media from stromal cells (HS5) co-cultured with PER-371 High or Low cells. Data are expressed as the mean isotope count per minute (n=3). Vertical bars show standard error of the mean.
gure 6.14. Decreased survival of PER-371 High xenografts was not
a.
b.
Cell line
170
6.3.7 Changes to secreted protein expression from BM cell co-culture with PER-371 High compared with PER-371 Low cells To investigate the mechanism by which CTGF was increased leukaemia cell
progression in PER-371 xenografted mice, a protein array was performed to
detect proteins being secreted as a result of stromal and leukaemia cell
interaction. HS5 stromal cells were cultured in low serum media (HS5) or in
culture with PER-371 High (HS5:PER-371 High) or PER-371 Low cells
(HS5:PER-371 Low) for 48 hours. Conditioned media was filtered to remove
cells and used in a commercial protein array to detect the expression of 1,000
human proteins (a full protein list is described in Supplementary Table S3). Raw
intensity signals were normalised to background and then to internal loading/
positive controls on the membrane to create adjusted intensity signals (raw
signals from array membrane exposures are presented in Supplementary
Figure S2). Secreted proteins present in the top 1% of adjusted intensity signals
from HS5 control membranes found high expression of common stromal-
associated proteins including; Interleukin 8 (IL8), Interleukin 6 (IL6), Matrix
metalloproteinase 1 (MMP1) and Fibronectin (FN) (Roecklein & Torok-Storb,
1995a). Presence of these proteins confirmed protein array results fitted with
previously reported secreted proteins in HS5 cells.
Altered secreted protein expression due to leukaemia cells
Changes in secreted proteins due to stromal cell interaction with leukaemia
cells was assessed with PER-371 leukaemia cells and HS5 protein secretion.
To achieve this the effect of was evaluated as follows. All HS5 adjusted
intensity signals were procured that had a greater than 2-fold change difference
in expression compared with HS5:PER-371 High and HS5:PER-371 Low
adjusted intensity signals. Differences in protein expression were included as a
positive or negative fold change if the protein intensity was increased/
decreased in the mean fold change (of FC1 and FC2) greater than 2 –fold.
Seventeen secreted proteins were found differentially expressed and the
majority of were underexpressed due to leukaemia cell interaction compared
with stromal cells alone (Table 6.2).
Tableafter co-culture. Continued next page.
6.2. Leukaemia cells alter secreted proteins from HS5 stromal cells
FADD
CA19-9
fas-associated protein with death domain
cancer antigen 19-90.27
0.17
HS5
b.t.
b.t.
HS5:PER-371 Low
- FC
Adj. Signal .
Mean
FC
- FC
Protein
IGFBP6
CXCL1
insulin-like growth factor, binding protein 6
chemokine (C-X-C motif) ligand 1
0.50
0.36 b.t.
b.t. - FC
- FC
CA15-3cancer antigen 15-3
0.36 b.t. - FC
CXCL2chemokine (C-X-C
motif) ligand 20.54 0.23 - 3.4
CXCL5
GPC5
chemokine (C-X-C motif) ligand 5
glypican 50.15
0.55 0.15
b.t. - FC
- 2.9
SFRP4secreted frizzled-related protein 4
0.80 0.32 - 2.1
SPARCsecreted protein,
acidic, cysteine rich0.97 0.57 - 2.0
- FC
- FC
FC 1
- FC
- FC
- FC
- 2.35
- 3.7
- FC
- 2.5
- 1.7
1.16 0.28 - 2.7TBSP1
thrombospondin 1- 4.1
1.11 0.39 - 2.1IGFBP7
insulin-like growth factor, binding protein 7
- 2.9
EDARectodysplasin A
receptor1.02 b.t. - FC- FC
b.t.
b.t.
HS5:PER-371 High
b.t.
b.t.
b.t.
0.12
0.27
b.t.
0.48
0.41
0.96
0.86
b.t.
- FC
- FC
FC 2
- FC
- FC
- FC
- 4.5
- 2.0
- FC
- 1.7
- 2.4
-1.3
-1.3
- FC
TPM1tropomyosin 1
b.t.
HS5
0.11
HS5:PER-371 Low
+ FC
Adj. Signal .
Mean
FCProtein
TGFB5
SMAC/ DIABLO
transforming growth factor, beta 5
IAP-binding mitochondrial protein
b.t.
b.t. 0.19
0.13 + FC
+ FC
ITLN1omentin
b.t. 0.17 + FC
+ FC
FC 1
+ FC
+ FC
+ FC
PLGplasminogen
b.t. 0.41 + FC+ FC
0.17
HS5:PER-371 High
0.12
0.18
0.13
0.38
+ FC
FC 2
+ FC
+ FC
+ FC
+ FC
Tableleukaemia-stromal cell co-culture. Low-serum conditioned media from HS5 stromal cells or from HS5 cells co-cultured with PER-371 High or Low cells for 48 hours was used on a human protein array membrane detecting 1000 proteins (Human Array L-1000, Ray Biotech). Raw signal intensity data were normalised to background and then to internal positive control values (adj. signal). Altered expression was analysed as the fold change between HS5:PER-371 Low compared with HS5 adjusted signals (FC1) and HS5:PER-371 High compared with HS5 adjusted signals (FC2). The mean fold change was calculated as the average between FC1 and FC2, and included all secrted proteins with a mean FC≥2. Data are represented as the average adj. signal intensity for the duplicated antibody dots. FC, fold change, b.t., adjusted signal <0.01 and below threshold cut-off.
6.2. Leukaemia cells alter expression of secreted proteins after
CXCL2chemokine (C-X-C
motif) ligand 20.54
HS5
0.23
HS5:PER-371 Low
Adj. Signal .
Mean
FC
- 3.2
Protein
CCR2
TF
chemokine (C-C motif) receptor 2
transferrinb.t.
b.t. b.t.
b.t. + FC
+ FC
CCR3chemokine (C-C motif)
receptor 3b.t. b.t. + FC
activin-a/ INHBA
inhibin, beta A0.51 0.42 - 3.1
TIMP1
VEGF
TIMP metalloprotease inhibitor 1
vascular endothelial growth factor
0.46
0.94 0.89
0.48 - 3.0
- 3.8
TIMP2TIMP metalloprotease
inhibitor 20.53 0.71 - FC
TIMP3TIMP metalloprotease
inhibitor 30.24 0.35 - FC
- 4.5
FC 1
+ FC
+ FC
+ FC
- 3.4
- 3.9
- 2.9
- FC
- FC
CXCR2chemokine (C-X-C motif) receptor 2
b.t. b.t. + FC+ FC
0.12
HS5:PER-371 High
0.11
0.20
0.12
0.15
0.24
0.16
b.t.
b.t.
0.37
- 1.9
FC 2
+ FC
+ FC
+ FC
- 2.8
- 3.7
- 3.0
- FC
- FC
+ FC
Tablecultures. Low-serum conditioned media from HS5 stromal cells or from stromal cells co-cultured for 48 hours with PER-371 High or Low cells was used on a human protein array membrane detecting 1000 proteins (Human Array L-1000, Ray Biotech). Raw signal intensity data were normalised to background and then to positive control values (adj. signal). Altered expression was analysed as the fold change between HS5:PER-371 High compared with HS5 adjusted signals (FC1) and HS5:PER-371 High compared with HS5:PER-371 Low adjusted signals (FC2). The mean fold change was calculated as the average between FC1 and FC2, and included all secreted proteins with a mean FC≥2. Data are represented as the average adj. signal intensity for the duplicated antibody dots. FC, fold change, b.t., adjusted
6.3. CTGF alters secreted proteins from stromal-leukaemia cell co-
174
Altered secreted protein expression due to CTGF
The differential expression of secreted proteins after leukaemia-stromal
interaction was assessed due to increased CTGF expression. The adjusted
signal intensity for PER-371 High cells cultured with HS5 cells was compared
with HS5:PER-371 Low and HS5 adjusted intensity signals using the same
parameters outlined for detecting differences due to leukaemia cell interaction
with HS5 cells. Mainly, the final calculated fold change was the average of the
individual fold changes between HS5:PER-371 High adjusted signal intensities
and HS5:PER-371 Low and HS5 adjusted signal intensitites. Eleven secreted
proteins were deregulated by more than 2-fold in HS5:PER-371 High cultured
medium compared with in HS5:PER-371 Low or HS5 media (Table 6.3).
6.4 Discussion Overexpression of CTGF in PER-371 xenografted mice resulted in a more
aggressive leukaemia expansion. Reduced survival in PER-371 High compared
with PER-371 Low xenografted mice is evidence of a functional role for CTGF in
disease outcome. Differences in leukaemia cell engraftment in the BM at late
stage disease were confirmed by multiple methods. Increased engraftment of
PER-371 High compared with PER-371 Low cells was found by flow cytometry
and immunohistochemical methods of leukaemia cell detection in the BM
space. The effects of CTGF appear to be either not relevant until a significant
population of leukaemia cells has amounted or require a substantial amount of
time to establish an influence over tumour progression. With late stage disease
the first time where differences in leukaemia cell behaviour are observed in
response to increased CTGF.
Mechanisms of CTGF action were investigated in PER-371 xenografted mice.
Functional experiments were performed and were successful in outlining a
mechanism of action for CTGF within BCP ALL. Differences in engraftment
between PER-371 High and Low xenografted mice at late stage disease
occurred independently to any overt changes in leukaemia cell proliferation.
Similarly, conditioned media from leukaemia and BM stromal cell interactions
was insufficient in altering proliferation in leukaemia cells in vitro.
175
However, a trend for reduced apoptosis in PER-371 High compared with PER-
371 Low leukaemia cells could explain one potential pathway by which PER-
371 High xenografted mice increase engraftment in the BM.
Evidence that CTGF influences leukaemia cells indirectly through the
microenvironment include increased Ki67+ proliferation in PER-371 High
xenografted mice at early stage disease, before differences in leukaemia
engraftment can be detected and in a larger population of cells than Pax5+
staining accounts for, these proliferating cells are therefore most likely stromal
cells. CTGF has been shown to increase proliferation of mesenchymal stem
cells and contribute to both to osteogenic and fibroblastic differentiation (Lee, et
al., 2010; Wang, et al., 2009a). In PER-371 cells there is evidence that CTGF
requires a complex microenvironment to indirectly facilitate leukaemia cell
progression.
Immunostaining of BM sections supported a relationship between reticulin fibre
deposition and PER-371 High leukaemia engraftment. CTGF is known to be
involved in inducing deposition of collagen fibres in wound healing and in organ
repair in humans (Leask, 2002; Sonnylal, et al., 2010). Increased reticulin
deposition seen in PER-371 High BM sections, at late stage disease in
particular, could be induced directly by CTGF binding to ECM or binding to
stromal cells with known associations with reticulin deposition such as CAR
cells or fibroblasts within the BM space. Reticulin fibre density at time of
diagnosis in human cases has been associated with worse disease outcome in
one study of BCP ALL (Noren-Nystrom, et al., 2007) and BM fibrosis has also
been shown to correlate with worse overall survival and worse progression-free
survival after autologous haematopoietic stem cell transplantation (Suyani, et
al., 2010). However, in other retrospective studies of childhood and adult ALL,
BM fibrosis had no correlation to disease outcome (Bharos, et al., 2010;
Thomas, et al., 2002). Whether BM fibrosis influences BCP ALL clinical
outcome, is not yet clearly defined. Use of the PER-371 NOD/SCID xenografted
model could prove valuable in the study of the effects of BM reticulin fibrosis.
Differences in secreted proteins from stromal-leukaemia cell interactions
indicated several potential important secreted proteins supporting leukaemia
176
infiltration or survival. A larger list of deregulated proteins were decreased in
expression due to PER-371 Low and HS5 incubation. Many of these proteins
have associations as tumour suppressors including: TSP1, IGFBP6, IGFBP7,
GPC5, SFRP4 and SPARC (Bach, et al., 2013; Giefing, et al., 2008; Jacob, et
al., 2012; Mok, et al., 1996; Rodriguez-Manzaneque, et al., 2001; Said, et al.,
2013; Wajapeyee, et al., 2008; Yang, et al., 2013). Many inhibited proteins have
also been identified as involved in ECM remodelling. SPARC has been shown
to act in similar ways to CTGF, by inducing COL3A1 and COL1A2 expression
and driving a fibrotic response in cells (Wang, et al., 2010b). Therefore
suppression of SPARC in PER 371 High with HS5 indicates different
mechanism are involved in BM fibrosis. Downregulated chemokines CXCL1,
CXCL2 and CXCL5 attract neutrophils to tumour sites and their suppression is
associated with primary tumour cell metastasis (Toh, et al., 2011). Additionally,
increased expression of these chemokines has been associated with fibrosis
and tissue remodelling in myeloid neoplasia and liver fibrosis (Jen Chin, 2005).
PER-371 High or Low incubations with HS5 cells increased expression of a
smaller number but diverse range of proteins including; an enzyme precursors
for ECM degradation (PLG), positive apoptotic regulator (SMAC/ DIABLO),
adipokine (ITLN1), cytokine (TGFB5) and an actin-binding contractile protein
(TPM1). No relationship between upregulated-secreted proteins was evident in
the literature, though all genes have been associated with cancer in various
ways. Leukaemia cells were found to have a significant effect on protein
secretion from BM stromal cells and these findings concur with the reported
literature that the microenvironment response facilitates leukaemia cell survival
and progression.
Protein arrays also found reduced secreted proteins due to CTGF expression
by PER-371 cells. Reduced expression of tissue remodelling proteins included
activin-a, VEGF and TIMP1,2 and 3. Deregulated secreted proteins from
stromal-leukaemia cell interactions due to CTGF expression suggest a role in
ECM modification and tie in with immunohistochemical findings reported in this
chapter and previously discussed findings of altered HS5 gene expression due
to rhCTGF in Chapter 4, this is discussed in further detail in the General
Discussion (Chapter 7).
177
The NOD/SCID xenograft model allowed us to investigate some fundamental
biological aspects of leukaemia survival and expansion in the BM. Both
monocytic and lymphoblastic leukaemias are clinically associated with a
gradual eviction of normal haematopietic cells from the BM. The specific
movement of megakaryocytes have not been addressed. In PER-371
xenografted mice, megakaryocytes were found diminished in number as
leukaemia engraftment was increased. Additionally, megakaryocytes were
present at the invading tumour edge in PER-371 xenografted mice, which
supports a possible role for megakaryocytes as tumour modulators.
Megakaryocytes have been reported as functional immune cell types in
NOD/SCID mice with evidence that they phagocytose human erythroid cells
and platelets (Psaila, et al., 2012). Psaila et al. indicate an important role for
megakaryocytes in modifying angiogenesis and osteogenesis during tumour
development by secretion of integral proteins. Follow-up studies will evaluate
spleen megakaryocyte numbers in PER-371 xenografts, to determine if the
spleen compensates for the loss of megakaryocytes in the bone marrow.
CTGF was found to not contribute to ARAC drug sensitivity in PER-371
xenografted mice. PER-371 High xenografted mice were sensitive to low-dose
ARAC, while survival analysis suggests PER-371 Low xenografted mice did not
respond to low-dose ARAC. However, as the survival curves of PER-371 High
and Low ARAC treated mice are not different this suggests the drug behaves
similarly on both cell lines. Timing of ARAC delivery in treated mice had the last
ARAC dose coincide with end stage disease in untreated PER-371 Low
xenografted mice. If the data reveals anything it is that low-dose ARAC is only
effective during the course of treatment and expansion of leukaemia resumes
immediately after the drug is withdrawn. The need for high-dose ARAC regimes
to achieve complete remission has been noted in human AML and refractory
ALL adult clinical trials (Kern & Estey, 2006; Willemze, et al., 1983). The
response found to low and high-dose ARAC in the PER-371 NOD/SCID
xenograft model supports the need for high-dose ARAC for drug efficacy.
However, the two doses used in the current study leave a large middle ground
and it is likely optimisation of ARAC dose would be required to determine if
higher doses of ARAC would improve leukaemic blast reduction beyond the
treatment phase in the NOD/SCID model.
178
The various roles of CTGF throughout the body have supported a multitude of
mechanistic pathways. Despite broad CTGF overexpression across a range of
BCP ALL subtypes and associations with worse clinical outcome, a functional
role for CTGF in BCP ALL has been lacking.
This study outlines a role for CTGF using the PER-371 NOD/SCID xenograft
model, with an internalised CTGF transgene responsible for differential
expression. High CTGF expression was found to decrease survival in PER-371
xenografted mice compared with low-expressing control groups. Observable
changes in engraftment were only evident from late stage disease in the BM,
and suggested a time-dependent component of CTGF function. Investigations
into leukaemia cell interactions with the BM microenvironment and in particular
the ECM, confirmed increased secretion of ECM remodelling proteins due to
CTGF overexpression in PER-371 High cells. This data together with observed
increases in stromal cell proliferation at early stage disease and increased
reticulin fibre density at late stage disease. Taken together, CTGF is postulated
to indirectly enhance the progression of PER-371 cells by modifications to the
ECM.
777777777
General Discussion
Intellectually we stand on an island in the midst of an illimitable ocean of inexplicability. Our business in every generation is to reclaim
a little more land.
Thomas Huxley
180
181
Chapter 7 General Discussion
7.1 Introduction The overall Aim of this thesis was to discover potential mechanisms of action of
CTGF in childhood B-precursor ALL. Previous studies have highlighted its
significance as an overexpressed gene in the majority of childhood B-precursor
ALL sub-types and its association with poor clinical outcome (Boag, et al., 2007;
Kang, et al., 2010). In order to clarify the importance of CTGF in driving disease
outcome, it was necessary to define any mechanism by which it could alter B-
precursor ALL cell function. The overall Aim was achieved through three smaller
Aims:
Firstly, engineering two B-precursor ALL cell lines to overexpress CTGF 3-fold
above a baseline detectable level was found to have minimal impact on the
behaviour of the cell in vitro. In particular, proliferation and sensitivity to ARAC
and VCR were found to be similar between CTGF High and Low cell lines in
both PER-278 and PER-371. Though no differences were found due to CTGF
expression, PER-278 cells exhibited a 25-fold reduced ARAC sensitivity
compared with PER-371 cells. Similarly, though no genes were commonly
deregulated due to CTGF overexpression by microarray, PER-278 and PER-
371 show distinct global gene expression profiles though primary material was
from patients with common ages, genders and E2A-PBX1 translocations. The
overall findings from Aim 1 indicate mechanisms of action of CTGF are
consequent to its role as a matricellular protein and not due to direct autocrine
or intracrine changes to the cell.
Secondly, the role of CTGF in extracellular signalling and communication was
explored. The presence of recombinant CTGF was found capable of changing
gene expression in the HS5 human BM stromal cell line, in a pronounced and
time dependent manner. Following 8 hours CTGF exposure, genes involved in
the cell cycle and activation of PI3K/AKT and IGF1 pathway signalling were
altered. Among altered secreted genes, the greatest number were involved in
182
ECM remodelling. The overall findings of Aim 2 verified CTGF is capable of
communication with human BM stromal cells and affects genes involved in
important downstream functions.
Thirdly, mechanisms of CTGF action in B-precursor ALL were explored within
the context of the murine BM microenvironment using NOD/SCID xenograft
models that closely intimate the human disease. In PER-278 xenografted mice,
CTGF overexpression had no effect on leukaemia growth but was found to
reduce sensitivity to low dose ARAC, with no effect on VCR senisitivity. In PER-
371 xenografted mice CTGF overexpression caused a more aggressive
expansion of leukaemia resulting in reduced survival, while sensitivity to ARAC
was unaltered. The different responses of the two cell lines due to CTGF were
considered separately.
Mechanisms by which CTGF altered PER-278 ARAC sensitivity were not fully
determined. Compared to PER-371 cells, reduced ARAC sensitivity in vitro was
found for both PER-278 High and Low cells suggesting that PER-278 cell lines
would be expected to be less sensitive to ARAC in vivo. However in PER-278
Low mice the low ARAC dose resulted in a small but significantly prolonged
survival and only PER-278 High mice proved to be insensitive to low dose
ARAC. Mechanisms by which CTGF altered PER-278 ARAC sensitivity in the
BM were not determined and remain a potential avenue of exploration.
In PER-371 xenografts, CTGF was found to reduce survival by inciting a more
aggressive progression of leukaemic cells in the BM and to a lesser extent in
the extramedullary organs. CTGF mechanisms of action included increasing
stromal cell proliferation and reticulin connective tissue deposition indirectly
affecting leukaemia cell expansion in the BM. The overall findings of Aim 3
conclusively proved a functional role for CTGF in B-precursor ALL, which was
dependent on the expression of cooperating genes by the leukaemia cell and
interaction with the microenvironment.
183
7.2 Limitations of the current research
Immune-compromised mouse model and importance of immune response The NOD/SCID mouse model has a severely depleted immune response. While
there is residual NK cell activity and macrophage function, it is a subset of the
full range of immune elements that would be relevant in human BM (Shultz, et
al., 1995a). While the other cells types in the BM stromal compartment are not
affected by the strain, this limitation is important to consider when defining
responses within the microenvironment, no model system of leukaemia is
currently available that allows engraftment of a large number of foreign cells
while maintaining a complete innate and adaptive immune response. The
NOD/SCID mouse was the most suitable strain for the presented studies, which
focused on growth of leukaemia cells within the BM, as this strain has a greater
immune response compared with the other immune-compromised mouse
strains and therefore a more biologically similar microenvironment to human
BM.
CTGF affected ARAC drug sensitivity only in restricted conditions
CTGF-induced ARAC resistance in PER-278 cells was found only at low dose
in female xenografted mice. The response of PER-278 High cells to ARAC at
higher doses and at 24 hours after a single dose was administered was no
different to PER-278 Low cells. The ARAC snapshot did not highlight
differences in proliferation or apoptosis due to CTGF with resistance to ARAC
not evident due to the 24-hour window missing the effects of drug activity or the
use of larger male mice with differences in the onset of 1% engraftment and
timing of drug administration. The restricted conditions under which CTGF was
found to effect ARAC sensitivity in PER-278 cells, particularly re-sensitisation at
higher doses, diminishes its likely importance in the clinical setting, where high
dose regimens of ARAC are more commonly used. The finding does provide
further insight into the variety of responses enabled by CTGF activity in B-
precursor ALL.
184
7.3 Critical findings
The microenvironment is essential for CTGF function in B-precursor ALL The inability to recapitulate in vitro the responses observed due to CTGF in vivo
in B-precursor ALL cells, highlighted the importance of the BM in CTGF
function. Neither CTGF overexpression alone nor secreted proteins from co-
incubations of leukaemia cells with stromal cells could replicate in vitro the
resistance of PER-278 cells to ARAC or the increased proliferation of PER-371
cells seen in xenograft mice. These results indicate that CTGF does not act
directly on B-precursor ALL cells and physical interaction with components of
the ECM is required. Tumour-associated fibroblasts are known to influence
tumour invasiveness and survival (Egeblad, et al., 2010). Further, in other
human cancers CTGF expression is commonly associated with increased
fibrosis in the surrounding tumour parenchyma (Birgersdotter, et al., 2010;
Cunningham, et al., 2010). In B-precursor ALL it is possible the effects of CTGF
are due solely to changes induced in BM stromal cells resulting in ECM
restructuring and increased BM fibrosis. Whether CTGF-mediated crosstalk with
the BM modifies leukaemia cell gene expression in turn, remains to be
determined.
The individual responses observed in PER-278 and PER-371 are dependent on
intrinsic differences in gene expression. The distinct responses highlight that
CTGF acts in cooperation with other genes expressed by leukaemia cells in
microenvironment interactions. It is also possible that the effect of CTGF on BM
stromal cells is threshold dependent and the slightly higher level of baseline
CTGF in PER-278 Low cells may diminish the response to lentiviral-induced
CTGF overexpression. It is possible a greater fold-change difference in CTGF
expression may also have led to a greater survival differential in PER-371
CTGF Low and High xenograft mice. However, the expression level caused by
lentiviral transfection in PER-278 and PER-371 fell within the range seen in
patient samples by qRT-PCR (Welch, et al., 2013) and therefore corresponds
with real observations made in the human disease.
CTGF mediates communication between B-precursor ALL cells and BM stromal cells
185
The role of CTGF in leukaemia and BM stromal cell communication was
outlined in two ways: firstly by changes to HS5 gene expression due to
recombinant CTGF and secondly changes to secreted proteins in co-cultures of
HS5 cells with PER-371 High and Low. Several genes and their corresponding
proteins were deregulated in both the microarray and protein array. These
included the chemokines; CCL2, CCL7 and CCL8, with upregulated gene
expression in HS5 cells. The receptor for all 3 ligands, CCR2, was upregulated
in the secreted protein from HS5 and PER-371 High. Similarly, increased gene
expression of CXCL1 and CXCL5 accompanied increased protein levels of
CXCR2, their known chemokine receptor. Whereas, downregulation of VEGF,
TIMP1 and TIMP2 were found concurrently at the gene expression and
secreted protein level due to overexpression of CTGF. The potential effects of
these genes are varied. Chemokines are responsible for immune cell trafficking,
cell survival and proliferation (Proudfoot, 2002). The receptors CCR2, CCR3
and CXCL2 have been associated with inflammation in disease and
chemokines have shown involvement in tumour growth, immune evasion,
metastasis and angiogenesis (Bendall, 2005). The TIMP proteins and VEGF
are associated with ECM remodelling, both capable of altering HSC niches
within the BM (Shen, et al., 2010). Beyond their MMP-inhibitory activity, the
TIMPS also act in cooperation or opposition to VEGF in establishing neo-
angiogenesis (Qi, et al., 2003; Stetler-Stevenson & Seo, 2005; Yamada, et al.,
2001). As all four proteins are downregulated they could function separately
from their known role in angiogenesis in the B-precursor ALL BM. These
deregulated genes are evidence of CTGF mediating communication between B-
precursor ALL cells and BM stromal cells, as HS5 and PER-371 High
interactions have increased CTGF protein due to PER-371 High cells and
secreted protein deregulation from HS5 cells.
Additionally, CTGF increased proliferation of stromal cells at early stages of
leukaemia cell engraftment, as evidenced by increased Ki67-positive
immunostaining in PER-371 CTGF High xenografted mice unaccompanied by
an increased percentage of Pax-5-positive leukaemia cells in the BM. Increased
stromal cell proliferation in BM sections is consistent with the changes in HS5
gene expression after 8 hours in the presence of rhCTGF, as the most altered
genes were associated with the cell cycle. While the stromal cell type
186
undergoing proliferation has not been qualified, it is clear that through CTGF,
PER-371 leukaemia cells directly communicate with components of the BM
microenvironment.
CTGF enables a more aggressive expansion of PER-371 cells in xenograft mice In PER-371 xenografted mice, overexpression of CTGF resulted in reduced
survival compared with PER-371 Low mice. Reduced survival was the result of
a faster spread of disease with a greater percentage of engraftment at late
stage disease in the BM, spleen, liver and lymph node in PER-371 High
compared with PER-371 Low cells. PER-371 High and Low xenografted mice
had similar levels of proliferation and apoptosis in the BM at late stage disease
indicating CTGF does not exert a sustained effect on leukaemia cell growth
throughout disease progression. Evidence from microarray studies found CTGF
affected a significant human BM stromal cell response and supports the notion
that a probable mechanism of CTGF action in PER-371 xenografts is through
microenvironment interactions. As reduced survival was specific to PER-371
xenografted mice, CTGF must interact with components of the BM
microenvironment in cooperation with genes expressed specifically in PER-371
cells and not PER-278.
CTGF alters drug sensitivity in PER-278 cells In PER-278 xenografts, overexpression of CTGF reduced the sensitivity of
leukaemia cells to low dose ARAC. The resistance was peculiar to the low
ARAC concentration and proximal to the time of ARAC injection as no
differences due to CTGF were found using a maximum-tolerated ARAC dose or
24 hours after a low dose ARAC bolus injection. The cellular changes enacted
by CTGF that confer tumour cell resistance were not identified; though they
appear dependent on the physical interaction with the microenvironment, as
secreted proteins from leukaemia cell-stroma interactions could not replicate
ARAC resistance in vitro. Potentially CTGF could provide protection from ARAC
in PER-278 cells by activating signalling pathways by adhesion of leukaemia
187
and stromal cells or by creating a physical barrier through increased fibrotic
deposition around the leukaemia cells. Stromal cell adhesion has previously
been shown to suppress apoptosis in B-lineage ALL cell lines after exposure to
ARAC and etoposide, acting via expression of the vascular cell adhesion
molecule 1 (VCAM1) (Mudry, et al., 2000). In these experiments soluble factors
were unable to attain the same level of drug resistance as adhesion to stromal
cells. Nor was adhesion to fibronectin alone, in contrast to chronic myelogenous
leukaemia were adhesion to fibronectin was sufficient to induce drug resistance
(Damiano, et al., 2001). In multiple cancer types tumour cell adhesion to stroma
via integrins or cell adhesion molecules has proven to be a potent mediator of
multidrug resistance (Aoudjit & Vuori, 2012; Correia & Bissell, 2012). However,
mechanisms of stromal-induced drug resistance are not completely understood,
in colon cancer adhesion to ECM components increased expression of anti-
apoptotic proteins, bcl-2 and bcl-x(L), but their expression did not correlate to
the protective effect induced by ECM contact suggesting differences in other
apoptotic or unidentified mechanisms were involved (Kouniavsky, et al., 2002).
The ECM can also affect drug resistance indirectly; collagen deposition in
ovarian cancer created increased interstitial pressure inside the tumour mass
and thereby reduced the accessibility of antibodies after intraperitoneal injection
(Choi, et al., 2006). Defining a mechanism by which CTGF alters cell sensitivity
to nucleoside analogues such as ARAC would extend our knowledge of drug
resistance mechanisms in B-precursor ALL.
In B-precursor ALL, the overall phenotype that results from the complex
regulation of cells by CTGF appears to depend on tumour genotype. In
oesophageal cancer CTGF has differential effects between squamous and
adenocarcinoma histological subtypes. Increased CTGF was associated with
better survival in squamous cell cancers, correlated with increased formation of
fibrotic capsules surrounding the tumour, and worse survival with
adenocarcinoma with no change to fibrosis and therefore postulated to directly
mediate apoptosis of tumour cells (Koliopanos, et al., 2002).
CTGF increases deposition of reticular connective tissue in B-precursor ALL BM microenvironment
188
Overexpression of CTGF in PER-371 xenografts was associated with increased
reticulin fibre deposition at early and late stage disease, reaching a significant
difference in mean reticulin score at late stage disease in the diaphysis.
Reticulin fibres are mainly composed of type 3 collagen and associated with
adventitial reticular cells in the BM (Cattoretti, et al., 1993; Gelse, et al., 2003;
Ushiki, 2002). ECM-associated genes COL3A1 and COL1A2 were found
upregulated in HS5 cells after an 8 hour exposure to rhCTGF, along with lysyl
oxidase (LOX), integrin, alpha V (ITGAV), fibronectin 1 (FN1), pentraxin 3
(PTX3), transforming growth factor beta 2 (TGFB2) and the proteoglycan
decorin (DCN). These genes are capable of contributing to the observed ECM
modifications and possibly play other associated roles in promoting leukaemia
cell growth via the microenvironment. Two ECM related genes were verified by
qRT-PCR as being upregulated by CTGF in HS5 cells: LOX (upregulated 5.4-
fold) and ITGAV (upregulated 1.8-fold). ITGAV may act as the transmembrane
receptor for CTGF when binding to stromal cells and LOX is implicated in cross-
linking collagen fibres. LOX has been postulated as a target for therapy in
many solid tumours (Barker, et al., 2012). In breast cancer, collagen cross-
linking by LOX was essential for collagen fibrosis and the resulting increased
tumour stiffness and focal adhesion signalling led to increased tumour invasion
into surrounding epithelium (Levental, et al., 2009).
With evidence that acute leukaemia exists as a closely associated population of
cells and are not dispersed throughout the BM space, has led to the recent
appreciation of leukaemia and stromal cell interactions in the disease. Fibrosis
accounts for a third of natural death globally and includes a chronic loss of
organ function in BM, heart, intestine, kidney, liver, lung, skin and more solid
tumours (Zeisberg & Kalluri, 2013). TGFβ overexpression is common among all
fibrotic tissues and has been extensively studied in bone marrow fibrosis, where
it is largely expressed by megakaryocytes and platelets (Kuter, et al., 2007). B-
precursor ALL cell expression of CTGF, may activate the same pathways and
cellular responses as megakaryocyte-expressed TGFβ.
189
7.4 Proposed mechanism of CTGF action
ECM remodelling by CTGF, was reported in this thesis to indirectly support
leukaemic growth and expansion in a NOD/SCID xenograft mouse model of B-
precursor ALL. It is unknown how CTGF modifies the deregulated ECM-
interacting genes presented in Chapter 4 and also how they specifically
implement the observed phenotype. We hypothesise, that in the early stages of
disease development CTGF acts to increase proliferation of adventitial reticular
cells. These cells increase production of COL3A1, LOX and other important
ECM-interacting genes to set up new niches tailored for expansion of leukaemia
cells. Production of reticulin connective tissue may modify the number of normal
cells, such as megakaryocytes, within the newly created space and can also
control the sequester and release of certain growth factors. The fibrotic reticular
connective tissue may act as a physical barrier to chemotherapy drugs,
however reticulin fibre density would be unique to each patient. Plasma CTGF
or LOX levels may act as a biomarker for fibrotic connective tissue and guide
concomitant delivery of ECM-degrading drugs.
7.5 Future directions
Further study will be essential to validate the proposed mechanism of action by
which CTGF alters disease progression in childhood B-precursor ALL.
Delineation of the mechanism by which CTGF increases fibrotic reticulin
connective tissue in the BM will increase our understanding of the clinical
impact of CTGF. Studies using B-precursor ALL cell lines with downregulation
of CTGF by shRNAs have already been initiated in our laboratory and will act to
validate the mechanisms presented in this thesis.
Use of the NOD/SCID xenograft model to test further cell lines with lentiviral-
induced CTGF overexpression or primary patient cell samples, would contribute
toward understanding how prevalent the effects of CTGF might be among
childhood B-precursor ALL cases. While this thesis successfully identifies a
mechanism of action for CTGF in B-precursor ALL, extrapolation of the findings
would help define how prolific or specific this CTGF-initiated phenotype is
among B-precursor ALL patients. Further, identifying the differences in gene
expression between PER-278 and PER-371 responsible for the dissimilar
190
outcomes in vivo will uncover important genetic traits currently overlooked in the
clinical profile of patients.
In order to confirm a causative role for CTGF in deposition of reticulin
connective tissue, an ECM inhibitor can be used, such as β–aminopropionitrile
(BAPN) an irreversible LOX inhibitor (Levental, et al., 2009). If CTGF is
causative for inducing reticulin deposition in the PER-371 NOD/SCID xenograft
model, and in turn increasing progression of leukaemia engraftment, then
preventing deposition of ECM should reverse the observed differences in
survival between PER-371 High and Low xenografted mice.
Further experimental opportunities include determining the effect of CTGF on
neo-angiogenesis, with possible links between CTGF as a proangiogenic
modulator and increased angiogenesis being associated with acute leukaemias
(Hussong, et al., 2000; Perez-Atayde, et al., 1997; Shimo, et al., 1999; Shimo,
et al., 2001). Studies to identify adventitial reticular cells as the stromal cell
population undergoing proliferation in early stage disease would provide further
substance to the argument that CTGF plays a direct role in reticulin deposition.
Lastly, elucidating undiscovered CTGF-induced changes to drug sensitivity by
testing of further chemotherapies could be performed in combination with
different ECM inhibitors. A recent study using a NOD/SCID xenograft model of
B-cell ALL found a monoclonal antibody against CTGF increased survival when
mice were co-treated with conventional chemotherapy drugs: vincristine, L-
asparaginase and dexamethasone (Lu, et al., 2013). Comparatively, in
pancreatic ductal adenocarcinoma CTGF was largely secreted from cancer-
associated fibroblasts and a CTGF monoclonal antibody given in combination
with single agent chemotherapy induced tumour cell apoptosis. Apoptosis was
induced without altering microenvironment composition, angiogenesis or
chemotherapy drug uptake but instead by down-regulating expression of
prosurvival genes in the tumour cells (Neesse, et al., 2013). Defining the
mechanism by which CTGF acts would lead to a better understanding of ECM
deposition in B-precursor ALL and allow the identification of the best ECM
inhibitor to counteract increased reticulin deposition and maximize synergy with
standard chemotherapeutic drugs. The results of which could potentially lead to
a personalised course of therapy for children with high CTGF expression,
191
preventing the worse clinical outcomes associated with CTGF expression in
childhood B-precursor ALL.
7.6 Final conclusions
This thesis provides the first evidence of a mechanism of action for CTGF in B-
precursor ALL. In PER-371 cells CTGF acts on BM stromal cells to increase the
deposition of reticular fibres forming fibrotic reticular connective tissue that
potentially provides space for B-precursor ALL cells to grow and expand within
the BM, contributing to the deregulation of normal haematological and immune
cell functions. Further understanding of other important elements in the fibrotic
response will aid in establishing drug targets and expanding current knowledge
of the networks of protein interactions that contribute to leukaemia progression
and survival. This thesis has provided strong evidence for CTGF initiating a
series of interactions that influence disease outcome in childhood B-precursor
ALL.
888888888References
What we call the beginning is often the endAnd to make and end is to make a beginning.
The end is where we start from.
T.S. Eliot
195
Adam, O., Theobald, K., Lavall, D., Grube, M., Kroemer, H. K., Ameling, S., Schafers, H. J., Bohm, M., & Laufs, U. (2011). Increased lysyl oxidase expression and collagen cross-linking during atrial fibrillation. J Mol Cell Cardiol, 50(4), 678-685.
Aikawa, T., Gunn, J., Spong, S. M., Klaus, S. J., & Korc, M. (2006). Connective tissue growth factor–specific antibody attenuates tumor growth, metastasis, and angiogenesis in an orthotopic mouse model of pancreatic cancer. Molecular Cancer Therapeutics, 5(5), 1108-1116.
Alvarez, H., Corvalan, A., Roa, J. C., Argani, P., Murillo, F., Edwards, J., Beaty, R., Feldmann, G., Hong, S.-M., Mullendore, M., Roa, I., Ibanez, L., Pimentel, F., Diaz, A., Riggins, G. J., & Maitra, A. (2008). Serial analysis of gene expression identifies connective tissue growth factor expression as a prognostic biomarker in gallbladder cancer. Clinical Cancer Research, 14(9), 2631-2638.
Armstrong, S. A. (2005). Molecular genetics of acute lymphoblastic leukemia. Journal of clinical oncology, 23(26), 6306.
Arnott, J. A., Lambi, A. G., Mundy, C. M., Hendesi, H., Pixley, R. A., Owen, T. A., Safadi, F. F., & Popoff, S. N. (2011). The role of connective tissue growth factor (ctgf/ccn2) in skeletogenesis. Critical Reviews in Eukaryotic Gene Expression, 21(1), 43.
Ashley, D. M., Bol, S. J., & Kannourakis, G. (1994). Human bone marrow stromal cell contact and soluble factors have different effects on the survival and proliferation of paediatric b-lineage acute lymphoblastic leukaemic blasts. Leukemia Research, 18(5), 337-346.
Ayala, F., Dewar, R., Kieran, M., & Kalluri, R. (2009). Contribution of bone microenvironment to leukemogenesis and leukemia progression. Leukemia, 23(12), 2233-2241.
Baade, P. D. (2010). Trends in incidence of childhood cancer in australia, 1983 - 2006. British Journal of Cancer, 102(3), 620.
Babic, A. M., Chen, C.-C., & Lau, L. F. (1999). Fisp12/mouse connective tissue growth factor mediates endothelial cell adhesion and migration through integrin alpha vbeta 3, promotes endothelial cell survival, and induces angiogenesis in vivo. Mol. Cell. Biol., 19(4), 2958-2966.
Bach, L. A., Fu, P., & Yang, Z. (2013). Insulin-like growth factor-binding protein-6 and cancer. Clinical Science, 124(4), 215-229.
Bailey, L. C., Lange, B. J., Rheingold, S. R., & Bunin, N. J. (2008). Bone-marrow relapse in paediatric acute lymphoblastic leukaemia. The Lancet Oncology, 9(9), 873-883.
Balduzzi, A., Valsecchi, M. G., Uderzo, C., De Lorenzo, P., Klingebiel, T., Peters, C., Stary, J., Felice, M. S., Magyarosy, E., & Conter, V. (2005). Chemotherapy versus allogeneic transplantation for very-high-risk childhood acute lymphoblastic leukaemia in first complete remission: Comparison by genetic randomisation in an international prospective study. The Lancet, 366(9486), 635-642.
196
Barbolina, M. V., Adley, B. P., Kelly, D. L., Shepard, J., Fought, A. J., Scholtens, D., Penzes, P., Shea, L. D., & Stack, M. S. (2009). Downregulation of connective tissue growth factor by three-dimensional matrix enhances ovarian carcinoma cell invasion. International Journal of Cancer, 125(4), 816-825.
Barker, H. E., Cox, T. R., & Erler, J. T. (2012). The rationale for targeting the lox family in cancer. Nat Rev Cancer, 12(8), 540-552.
Battula, V. L., Chen, Y., Cabreira, M. d. G., Ruvolo, V., Wang, Z., Ma, W., Konoplev, S., Shpall, E., Lyons, K., Strunk, D., Bueso-Ramos, C., Davis, R. E., Konopleva, M., & Andreeff, M. (2013). Connective tissue growth factor regulates adipocyte differentiation of mesenchymal stromal cells and facilitates leukemia bone marrow engraftment. Blood, 122(3), 357-366.
Bendall, L. (2005). Chemokines and their receptors in disease. Histology and Histopathology, 20(3), 907-926.
Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society. Series B (Methodological), 289-300.
Bennewith, K. L., Huang, X., Ham, C. M., Graves, E. E., Erler, J. T., Kambham, N., Feazell, J., Yang, G. P., Koong, A., & Giaccia, A. J. (2009). The role of tumor cell-derived connective tissue growth factor (ctgf/ccn2) in pancreatic tumor growth. Cancer Res, 69(3), 775-784.
Bharos, A., de Jong, A. J., Manton, N., Venn, N., Story, C., Hodge, G., Nath, S., Suppiah, R., Tapp, H., Sutton, R., & Revesz, T. (2010). Bone marrow fibrosis and vascular density lack prognostic significance in childhood acute lymphoblastic leukaemia. Leukemia, 24(8), 1537-1538.
Bhojwani, D., & Pui, C.-H. (2013). Relapsed childhood acute lymphoblastic leukaemia. The Lancet Oncology, 14(6), e205-e217.
Bianco, P., Robey, P. G., & Simmons, P. J. (2008). Mesenchymal stem cells: Revisiting history, concepts, and assays. Cell Stem Cell, 2(4), 313-319.
Birgersdotter, A., Baumforth, K. R. N., Wei, W., Murray, P. G., Sjöberg, J., Björkholm, M., Porwit, A., & Ernberg, I. (2010). Connective tissue growth factor is expressed in malignant cells of hodgkin lymphoma but not in other mature b-cell lymphomas. American Journal of Clinical Pathology, 133(2), 271-280.
Bissell, M. J., & Hines, W. C. (2011). Why don't we get more cancer? A proposed role of the microenvironment in restraining cancer progression. Nat Med, 17(3), 320-329.
Blenk, S., Engelmann, J., Weniger, M., Schultz, J., Dittrich, M., Rosenwald, A., Müller-Hermelink, H., Müller, T., & Dandekar, T. (2007). Germinal center b cell-like (gcb) and activated b cell-like (abc) type of diffuse large b cell lymphoma (dlbcl): Analysis of molecular predictors, signatures, cell cycle state and patient survival. Cancer Informatics, 3, 399.
197
Boag, J. M., Beesley, A. H., Firth, M. J., Freitas, J. R., Ford, J., Hoffmann, K., Cummings, A. J., De Klerk, N. H., & Kees, U. R. (2006). Altered glucose metabolism in childhood pre-b acute lymphoblastic leukaemia. Leukemia, 20(10), 1731-1737.
Boag, J. M., Beesley, A. H., Firth, M. J., Freitas, J. R., Ford, J., Brigstock, D. R., de Klerk, N. H., & Kees, U. R. (2007). High expression of connective tissue growth factor in pre-b acute lymphoblastic leukaemia. British Journal of Haematology, 138(6), 740-748.
Bork, P. (1993). The modular architecture of a new family of growth regulators related to connective tissue growth factor. FEBS Letters, 327(2), 125.
Boyerinas, B., Zafrir, M., Yesilkanal, A. E., Price, T. T., Hyjek, E. M., & Sipkins, D. A. (2013). Adhesion to osteopontin in the bone marrow niche regulates lymphoblastic leukemia cell dormancy. Blood, 121(24), 4821-4831.
Bradham, D. M. (1991). Connective tissue growth factor: A cysteine-rich mitogen secreted by human vascular endothelial cells is related to the src-induced immediate early gene product cef-10. The journal of cell biology, 114(6), 1285.
Braig, S., Wallner, S., Junglas, B., Fuchshofer, R., & Bosserhoff, A. (2011). Ctgf is overexpressed in malignant melanoma and promotes cell invasion and migration. British journal of cancer, 105(2), 231-238.
Brizzi, M. F., Tarone, G., & Defilippi, P. (2012). Extracellular matrix, integrins, and growth factors as tailors of the stem cell niche. Current Opinion in Cell Biology, 24(5), 645-651.
Butler, J. M., Kobayashi, H., Rafii, S., (2010). Instructive role of the vascular
niche in promoting tumour growth and tissue repair by angiocrine factors. Nature reviews cancer. 10(2). 138-146.
Caplan, A. I., & Bruder, S. P. (2001). Mesenchymal stem cells: Building blocks for molecular medicine in the 21st century. Trends in Molecular Medicine, 7(6), 259-264.
Capparelli, C., Whitaker-Menezes, D., Guido, C., Balliet, R., Pestell, T. G.,
Howell, A., Sneddon, S., Pestell, R. G., Martinez-Outschoorn, U., & Lisanti, M. P. (2012). Ctgf drives autophagy, glycolysis, and senescence in cancer associated fibroblasts via hif1 activation, metabolically promoting tumor growth. Cell Cycle, 11(12), 2272-2284.
Casimiro, S., Luis, I., Fernandes, A., Pires, R., Pinto, A., Gouveia, A. G.,
Francisco, A. F., Portela, J., Correia, L., & Costa, L. (2012). Analysis of a bone metastasis gene expression signature in patients with bone metastasis from solid tumors. Clinical and Experimental Metastasis, 29(2). 155-164.
Cattoretti, G., Schiro, R., Orazi, A., Soligo, D., & Colombo, M. P. (1993). Bone
marrow stroma in humans: Anti-nerve growth factor receptor antibodies selectively stain reticular cells in vivo and in vitro. Blood, 81(7), 1726-1738.
198
Chang, C., Yang, M., Lin, B., Chen, S., Pan, S., Hsiao, M., Lai, T., Lin, S., Jeng, Y., & Chu, C. (2012). Ccn2 inhibits lung cancer metastasis through promoting dapk-dependent anoikis and inducing egfr degradation. Cell Death & Differentiation. 20(3). 443-455.
Chang, C.-C., Shih, J.-Y., Jeng, Y.-M., Su, J.-L., Lin, B.-Z., Chen, S.-T., Chau, Y.-P., Yang, P.-C., & Kuo, M.-L. (2004). Connective tissue growth factor and its role in lung adenocarcinoma invasion and metastasis. Journal of the National Cancer Institute, 96(5), 364-375.
Chang, C.-C., Hsu, W.-H., Wang, C.-C., Chou, C.-H., Kuo, M. Y.-P., Lin, B.-R., Chen, S.-T., Tai, S.-K., Kuo, M.-L., & Yang, M.-H. (2013). Connective tissue growth factor activates pluripotency genes and mesenchymal–epithelial transition in head and neck cancer cells. Cancer Research, 73(13), 4147-4157.
Chang, C. C., Lin, M. T., Lin, B. R., Jeng, Y. M., Chen, S. T., Chu, C. Y., Chen, R. J., Chang, K. J., Yang, P. C., & Kuo, M. L. (2006). Effect of connective tissue growth factor on hypoxia-inducible factor 1β degradation and tumor angiogenesis. Journal of the National Cancer Institute, 98(14), 984-995.
Chen, P.-S., Wang, M.-Y., Wu, S.-N., Su, J.-L., Hong, C.-C., Chuang, S.-E., Chen, M.-W., Hua, K.-T., Wu, Y.-L., & Cha, S.-T. (2007). Ctgf enhances the motility of breast cancer cells via an integrin-αvβ3–erk1/2-dependent s100a4-upregulated pathway. Journal of cell science, 120(12), 2053-2065.
Cheok, M. H. (2006). Acute lymphoblastic leukaemia: A model for the pharmacogenomics of cancer therapy. Nature Reviews. Cancer, 6(2), 117.
Chien, W., Yin, D., Gui, D., Mori, A., Frank, J. M., Said, J., Kusuanco, D., Marchevsky, A., McKenna, R., & Koeffler, H. P. (2006). Suppression of cell proliferation and signaling transduction by connective tissue growth factor in non‚äìsmall cell lung cancer cells. Molecular Cancer Research, 4(8), 591-598.
Choi, J., Credit, K., Henderson, K., Deverkadra, R., He, Z., Wiig, H., Vanpelt, H., & Flessner, M. F. (2006). Intraperitoneal immunotherapy for metastatic ovarian carcinoma: Resistance of intratumoral collagen to antibody penetration. Clinical Cancer Research, 12(6), 1906-1912.
Choi, S., Henderson, M. J., Kwan, E., Beesley, A. H., Sutton, R., Bahar, A. Y., Giles, J., Venn, N. C., Pozza, L. D., Baker, D. L., Marshall, G. M., Kees, U. R., Haber, M., & Norris, M. D. (2007). Relapse in children with acute lymphoblastic leukemia involving selection of a preexisting drug-resistant subclone. Blood, 110(2), 632-639.
Chuang, J.-Y., Yang, W.-Y., Lai, C.-H., Lin, C.-D., Tsai, M.-H., & Tang, C.-H. (2011). Ctgf inhibits cell motility and cox-2 expression in oral cancer cells. International immunopharmacology, 11(8), 948-954.
Clark, H. F., Gurney, A. L., Abaya, E., Baker, K., Baldwin, D., Brush, J., Chen, J., Chow, B., Chui, C., & Crowley, C. (2003). The secreted protein discovery initiative (spdi), a large-scale effort to identify novel human secreted and transmembrane proteins: A bioinformatics assessment. Genome research, 13(10), 2265-2270.
199
Clark, M. R., Mandal, M., Ochiai, K., & Singh, H. (2014). Orchestrating b cell lymphopoiesis through interplay of il-7 receptor and pre-b cell receptor signalling. Nat Rev Immunol, 14(2), 69-80.
Cobaleda, C., Jochum, W., & Busslinger, M. (2007). Conversion of mature b cells into t cells by dedifferentiation to uncommitted progenitors. Nature, 449(7161), 473-477.
Cobaleda, C., & Sánchez-García, I. (2009). B-cell acute lymphoblastic leukaemia: Towards understanding its cellular origin. BioEssays, 31(6), 600-609.
Colmone, A., Amorim, M., Pontier, A. L., Wang, S., Jablonski, E., & Sipkins, D. A. (2008). Leukemic cells create bone marrow niches that disrupt the behavior of normal hematopoietic progenitor cells. Science, 322(5909), 1861-1865.
Croci, S., Landuzzi, L., Astolfi, A., Nicoletti, G., Rosolen, A., Sartori, F., Follo, M. Y., Oliver, N., De Giovanni, C., Nanni, P., & Lollini, P.-L. (2004). Inhibition of connective tissue growth factor (ctgf/ccn2) expression decreases the survival and myogenic differentiation of human rhabdomyosarcoma cells. Cancer Research, 64(5), 1730-1736.
Cui, L., Zhang, Q., Mao, Z., Chen, J., Wang, X., Qu, J., Zhang, J., & Jin, D. (2011). Ctgf is overexpressed in papillary thyroid carcinoma and promotes the growth of papillary thyroid cancer cells. Tumor Biology, 32(4), 721-728.
Cunningham, J. L., Tsolakis, A. V., Jacobson, A., & Janson, E. T. (2010). Connective tissue growth factor expression in endocrine tumors is associated with high stromal expression of {alpha}-smooth muscle actin. Eur J Endocrinol, 163(4), 691-697.
Damiano, J. S., Hazlehurst, L. A., & Dalton, W. S. (2001). Cell adhesion-mediated drug resistance (cam-dr) protects the k562 chronic myelogenous leukemia cell line from apoptosis induced by bcr/abl inhibition, cytotoxic drugs, and gamma-irradiation. Leukemia 15(8), 1232-1239.
de Visser, K. E., Eichten, A., & Coussens, L. M. (2006). Paradoxical roles of the immune system during cancer development. [10.1038/nrc1782]. Nat Rev Cancer, 6(1), 24-37.
De Winter, P., Leoni, P., & Abraham, D. (2008). Connective tissue growth factor: Structure-function relationships of a mosaic, multifunctional protein. [Article]. Growth Factors, 26(2), 80-91.
Deng, Y.-Z., Chen, P.-P., Wang, Y., Yin, D., Koeffler, H. P., Li, B., Tong, X.-J., & Xie, D. (2007). Connective tissue growth factor is overexpressed in esophageal squamous cell carcinoma and promotes tumorigenicity through b-catenin-t-cell factor/lef signaling. Journal of Biological Chemistry, 282(50), 36571-36581.
Desgrosellier, J. S., & Cheresh, D. A. (2010). Integrins in cancer: Biological implications and therapeutic opportunities. Nat Rev Cancer, 10(1), 9-22.
200
Ding, L., & Morrison, S. J. (2013). Haematopoietic stem cells and early lymphoid progenitors occupy distinct bone marrow niches. Nature, 495(7440), 231-235.
Dornhofer, N., Spong, S., Bennewith, K., Salim, A., Klaus, S., Kambham, N., Wong, C., Kaper, F., Sutphin, P., Nacalumi, R., Hockel, M., Le, Q., Longaker, M., Yang, G., Koong, A., & Giaccia, A. (2006). Connective tissue growth factor-specific monoclonal antibody therapy inhibits pancreatic tumor growth and metastasis. Cancer Res, 66(11), 5816-5827.
Dvorak, H. F. (1986). Tumors: Wounds that do not heal: Similarities between tumor stroma generation and wound healing. The New England journal of medicine, 315(26), 1650-1659.
Edwards, L. A., Woolard, K., Son, M. J., Li, A., Lee, J., Ene, C., Mantey, S. A., Maric, D., Song, H., & Belova, G. (2011). Effect of brain-and tumor-derived connective tissue growth factor on glioma invasion. Journal of the National Cancer Institute, 103(15), 1162-1178.
Egeblad, M., Rasch, M. G., & Weaver, V. M. (2010). Dynamic interplay between the collagen scaffold and tumor evolution. Current Opinion in Cell Biology, 22(5), 697-706.
Eguchi, D., Ikenaga, N., Ohuchida, K., Kozono, S., Cui, L., Fujiwara, K., Fujino, M., Ohtsuka, T., Mizumoto, K., & Tanaka, M. (2012). Hypoxia enhances the interaction between pancreatic stellate cells and cancer cells via increased secretion of connective tissue growth factor. Journal of Surgical Research.
Evans, W. E., Crews, K. R., & Pui, C. H. (2013). A health-care system perspective on implementing genomic medicine: Pediatric acute lymphoblastic leukemia as a paradigm. [State of the Art]. Clin Pharmacol Ther, 94(2), 224-229.
Frazier, K. S., & Grotendorst, G. R. (1997). Expression of connective tissue growth factor mrna in the fibrous stroma of mammary tumors. The International Journal of Biochemistry & Cell Biology, 29(1), 153-161.
Friedl, P., & Wolf, K. (2003). Tumour-cell invasion and migration: Diversity and escape mechanisms. Nat Rev Cancer, 3(5), 362-374.
Fujii, M., Nakanishi, H., Toyoda, T., Tanaka, I., Kondo, Y., Osada, H., & Sekido, Y. (2012). Convergent signaling in the regulation of connective tissue growth factor in malignant mesothelioma: Tgfβ signaling and defects in the hippo signaling cascade. Cell Cycle, 11(18), 3373-3379.
Garcia, P., Leal, P., Alvarez, H., Brebi, P., Ili, C., Tapia, O., & Roa, J. (2013a). Connective tissue growth factor immunohistochemical expression is associated with gallbladder cancer progression. Archives of pathology & laboratory medicine, 137(2), 245-250.
Garcia, P., Leal, P., Ili, C., Brebi, P., Alvarez, H., & Roa, J. (2013b). Inhibition of connective tissue growth factor (ctgf/ccn2) in gallbladder cancer cells leads to decreased growth in vitro. International Journal of Experimental Pathology, 94(3), 195-202.
201
Gardini, A., Corti, B., Fiorentino, M., Altimari, A., Ercolani, G., Grazi, G. L., Pinna, A. D., Grigioni, W. F., & D'Errico Grigioni, A. (2005). Expression of connective tissue growth factor is a prognostic marker for patients with intrahepatic cholangiocarcinoma. Digestive and Liver Disease, 37(4), 269-274.
Gelse, K., Poschl, E., & Aigner, T. (2003). Collagens - structure, function, and biosynthesis. Advanced Drug Delivery Reviews, 55(12), 1531-1546.
Genecards. CTGF expression in normal human tissues, 2013, from http://www.genecards.org/cgi-bin/carddisp.pl?gene=CTGF
Gery, S., Xie, D., Yin, D., Gabra, H., Miller, C., Wang, H., Scott, D., Yi, W. S., Popoviciu, M. L., Said, J. W., & Koeffler, H. P. (2005). Ovarian carcinomas: Ccn genes are aberrantly expressed and ccn1 promotes proliferation of these cells. Clinical Cancer Research, 11(20), 7243-7254.
Giefing, M., Arnemann, J., Martin-Subero, I., Nielander, I., Bug, S., Hartmann, S., Arnold, N., Tiacci, E., Frank, M., & Hansmann, M. (2008). Identification of candidate tumour suppressor gene loci for hodgkin and reed‚äêsternberg cells by characterisation of homozygous deletions in classical hodgkin lymphoma cell lines. British Journal of Haematology, 142(6), 916-924.
Gohongi, T., Fukumura, D., Boucher, Y., Yun, C.-O., Soff, G. A., Compton, C., Todoroki, T., & Jain, R. K. (1999). Tumor-host interactions in the gallbladder suppress distal angiogenesis and tumor growth: Involvement of transforming growth factor β 1. Nature medicine, 5(10), 1203-1208.
Gong, J. K. (1978). Endosteal marrow: A rich source of hematopoietic stem cells. Science, 199(4336), 1443.
Gordon, H., & Sweets Jr, H. H. (1936). A simple method for the silver impregnation of reticulum. The American journal of pathology, 12(4), 545.
Gorlov, I., Sircar, K., Zhao, H., Maity, S., Navone, N., Gorlova, O., Troncoso, P., Pettaway, C., Byun, J., & Logothetis, C. (2010). Prioritizing genes associated with prostate cancer development. BMC Cancer, 10(1), 599.
Greaves, M. F., Maia, A. T., Wiemels, J. L., & Ford, A. M. (2003). Leukemia in twins: Lessons in natural history. Blood, 102(7), 2321-2333.
Greenbaum, A., Hsu, Y.-M. S., Day, R. B., Schuettpelz, L. G., Christopher, M. J., Borgerding, J. N., Nagasawa, T., & Link, D. C. (2013). Cxcl12 in early mesenchymal progenitors is required for haematopoietic stem-cell maintenance. Nature, 495(7440), 227-230.
Hagiya, K., Yasunaga, J.-i., Satou, Y., Ohshima, K., & Matsuoka, M. (2011). Atf3, an htlv-1 bzip factor binding protein, promotes proliferation of adult t-cell leukemia cells. Retrovirology, 8(1), 19.
Hall, B. M., & Gibson, L. F. (2004). Regulation of lymphoid and myeloid leukemic cell survival: Role of stromal cell adhesion molecules. Leukemia and Lymphoma, 45(1), 35-48.
Hall-Glenn, F., & Lyons, K. (2011). Roles for ccn2 in normal physiological processes. Cellular and Molecular Life Sciences, 68(19), 3209-3217.
202
Hall-Glenn, F., De Young, R. A., Huang, B.-L., van Handel, B., Hofmann, J. J., Chen, T. T., Choi, A., Ong, J. R., Benya, P. D., Mikkola, H., Iruela-Arispe, M. L., & Lyons, K. M. (2012). Ccn2/connective tissue growth factor is essential for pericyte adhesion and endothelial basement membrane formation during angiogenesis. PLoS ONE, 7(2), e30562.
Hall-Glenn, F., Aivazi, A., Akopyan, L., Ong, J. R., Baxter, R. R., Benya, P. D., Goldschmeding, R., van Nieuwenhoven, F. A., Hunziker, E. B., & Lyons, K. M. (2013). Ccn2/ctgf is required for matrix organization and to protect growth plate chondrocytes from cellular stress. Journal of Cell Communication and Signaling, 1-12.
Hanahan, D., & Coussens, L. M. (2012). Accessories to the crime: Functions of cells recruited to the tumor microenvironment. Cancer Cell, 21(3), 309-322.
Hartel, M., di Mola, F. F., Gardini, A., Zimmermann, A., Di Sebastiano, P., Guweidhi, A., Innocenti, P., Giese, T., Giese, N., Büchler, M. W., & Friess, H. (2004). Desmoplastic reaction influences pancreatic cancer growth behavior. World Journal of Surgery, 28(8), 818-825.
Hazlehurst, L. A. (2003). Role of the tumor microenvironment in mediating de novo resistance to drugs and physiological mediators of cell death. Oncogene, 22(47), 7396.
Heestand, G., Pipas, J., Valone, F., McMullen, A., Gadea, P., Williams, D., & Zhong, M. (2011). A phase i trial of the monoclonal antibody fg-3019 to connective tissue growth factor (ctgf) in locally advanced or metastatic pancreatic cancer. J Clin Oncol, 29(4), 269.
Hewson, J. (1996). Ultrastructural changes during adhesion and migration of pre-b lymphoid leukaemia cells within bone marrow stroma. British Journal of Haematology, 92(1), 77.
Hishikawa, K., Oemar, B. S., Tanner, F. C., Nakaki, T., Loscher, T. F., & Fujii, T. (1999). Connective tissue growth factor induces apoptosis in human breast cancer cell line mcf-7. Journal of Biological Chemistry, 274(52), 37461-37466.
Hubeek, I., Stam, R. W., Peters, G. J., Broekhuizen, R., Meijerink, J. P. P., Wering, E. R. v., Gibson, B. E. S., Creutzig, U., Zwaan, C. M., Cloos, J., Kuik, D. J., Pieters, R., & Kaspers, G. J. L. (2005). The human equilibrative nucleoside transporter 1 mediates in vitro cytarabine sensitivity in childhood acute myeloid leukaemia. Br J Cancer, 93(12), 1388-1394.
Hudson, M. M. (2003). Health status of adult long-term survivors of childhood cancer: A report from the childhood cancer survivor study. JAMA, 290(12), 1583.
Hussong, J. W., Rodgers, G. M., & Shami, P. J. (2000). Evidence of increased angiogenesis in patients with acute myeloid leukemia. Blood, 95(1), 309-313.
Hynes, R. O. (2009). The extracellular matrix: Not just pretty fibrils. Science, 326(5957), 1216-1219.
203
Igarashi, A., Segoshi, K., Sakai, Y., Pan, H., Kanawa, M., Higashi, Y., Sugiyama, M., Nakamura, K., Kurihara, H., Yamaguchi, S., Tsuji, K., Kawamoto, T., & Kato, Y. (2007). Selection of common markers for bone marrow stromal cells from various bones using real-time rt-pcr: Effects of passage number and donor age. Tissue Engineering, 13(10), 2405-2417.
Ijichi, H., Chytil, A., Gorska, A. E., Aakre, M. E., Bierie, B., Tada, M., Mohri, D., Miyabayashi, K., Asaoka, Y., & Maeda, S. (2011). Inhibiting cxcr2 disrupts tumor-stromal interactions and improves survival in a mouse model of pancreatic ductal adenocarcinoma. The Journal of Clinical Investigation, 121(10), 4106.
Inaba, H., Greaves, M., & Mullighan, C. G. (2013). Acute lymphoblastic leukaemia. The Lancet, 381(9881), 1943-1955.
Iwamoto, S., Mihara, K., Downing, J. R., Ching-Hon, P., & Campana, D. (2007). Mesenchymal cells regulate the response of acute lymphoblastic leukemia cells to asparaginase. Journal of Clinical Investigation, 117(4), 1049-1057.
Jacob, F., Ukegjini, K., Nixdorf, S., Ford, C. E., Olivier, J., Caduff, R., Scurry, J. P., Guertler, R., Hornung, D., & Mueller, R. (2012). Loss of secreted frizzled-related protein 4 correlates with an aggressive phenotype and predicts poor outcome in ovarian cancer patients. PloS one, 7(2), e31885.
Jandova, J., Beyer, T. E., Meuillet, E. J., & Watts, G. S. (2012). The matrix protein ccn1/cyr61 is required for αvβ5-mediated cancer cell migration. Cell Biochemistry and Function, 30(8), 687-695.
Janowska-Wieczorek, A. (2001). Autocrine/paracrine mechanisms in human hematopoiesis. Stem Cells, 19(2), 99.
Jemal, A. (2008). Cancer statistics, 2008. CA, 58(2), 71.
Jen Chin, W. (Writer) (2005). Importance of plasma matrix metalloproteinases (mmp) and tissue inhibitors of metalloproteinase (timp) in development of fibrosis in agnogenic myeloid metaplasia [Article], Leukemia & Lymphoma: Taylor & Francis Ltd.
Jiang, C.-G., Lv, L., Liu, F.-R., Wang, Z.-N., Na, D., Li, F., Li, J.-B., Sun, Z., & Xu, H.-M. (2012). Connective tissue growth factor is a positive regulator of epithelial–mesenchymal transition and promotes the adhesion with gastric cancer cells in human peritoneal mesothelial cells. Cytokine.
Jiang, N., Koh, G. S., Lim, J. Y., Kham, S. K., Ariffin, H., Chew, F. T., & Yeoh, A. E. (2011). Bim is a prognostic biomarker for early prednisolone response in pediatric acute lymphoblastic leukemia. Experimental Hematology, 39(3), 321-329.e323.
Jun, J.-I., & Lau, L. F. (2011). Taking aim at the extracellular matrix: Ccn proteins as emerging therapeutic targets. Nature reviews Drug discovery, 10(12), 945-963.
Kaatsch, P., Steliarova-Foucher, E., Crocetti, E., Magnani, C., Spix, C., & Zambon, P. (2006). Time trends of cancer incidence in european children (1978-1997): Report from the automated childhood cancer information system project. European Journal of Cancer, 42(13), 1961-1971.
204
Kamps, W. A., Boekkerink, J. P., Hakvoort-Cammel, F. G., Veerman, A. J., Weening, R. S., van Wering, E. R., van Weerden, J. F., Hermans, J., Slater, R., van den Berg, E., Kroes, W. G., & van der Does-van den Berg, A. (2002). Bfm-oriented treatment for children with acute lymphoblastic leukemia without cranial irradiation and treatment reduction for standard risk patients: Results of dclsg protocol all-8 (1991-1996). Leukemia, 16(6), 1099-1111.
Kang, H., Chen, I.-M., Wilson, C. S., Bedrick, E. J., Harvey, R. C., Atlas, S. R., Devidas, M., Mullighan, C. G., Wang, X., Murphy, M., Ar, K., Wharton, W., Borowitz, M. J., Bowman, W. P., Bhojwani, D., Carroll, W. L., Camitta, B. M., Reaman, G. H., Smith, M. A., Downing, J. R., Hunger, S. P., & Willman, C. L. (2010). Gene expression classifiers for relapse-free survival and minimal residual disease improve risk classification and outcome prediction in pediatric b-precursor acute lymphoblastic leukemia. Blood, 115(7), 1394-1405.
Kang, Y., Siegel, P. M., Shu, W., Drobnjak, M., Kakonen, S. M., Cordon-Cardo, C., Guise, T. A., & Massague, J. (2003). A multigenic program mediating breast cancer metastasis to bone. Cancer Cell, 3(6), 537-549.
Kasaragod, A. B., Lucia, M. S., Cabirac, G., Grotendorst, G. R., & Stenmark, K. R. (2001). Connective tissue growth factor expression in pediatric myofibroblastic tumors. Pediatric and Developmental Pathology, 4(1), 37-45.
Kees, U. R., Ford, J., Price, P. J., Meyer, B. F., & Herrmann, R. P. (1987). Per-117: A new human all cell line with an immature thymic phenotype. Leukemia Research, 11(5), 489-498.
Kees, U. R., Lukeis, R., Ford, J., Willoughby, M. L. N., & Garson, O. M. (1990). Establishment and characterization of a childhood pre-b acute lymphoblastic leukemia cell line, per-278, with chromosome translocations t (1; 19) and t (1; 9). Cancer Genetics and Cytogenetics, 46(2), 201-208.
Kees, U. R., Ford, J., Watson, M., Murch, A., Ringńer, M., Walker, R. L., & Meltzer, P. (2003). Gene expression profiles in a panel of childhood leukemia cell lines mirror critical features of the disease. Molecular Cancer Therapeutics, 2(7), 671-677.
Kern, W., & Estey, E. H. (2006). High-dose cytosine arabinoside in the treatment of acute myeloid leukemia. Cancer, 107(1), 116-124.
Kidd, M., Modlin, I., Shapiro, M., Camp, R., Mane, S., Usinger, W., & Murren, J. (2007). Ctgf, intestinal stellate cells and carcinoid fibrogenesis. World Journal of Gastroenterology, 13(39), 5208.
Kiel, M. J., Yilmaz, ñ. H., Iwashita, T., Yilmaz, O. H., Terhorst, C., & Morrison, S. J. (2005). Slam family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell, 121(7), 1109-1121.
Kiel, M. J., & Morrison, S. J. (2008). Uncertainty in the niches that maintain haematopoietic stem cells. Nat Rev Immunol, 8(4), 290-301.
205
Kikuchi, R., Tsuda, H., Kanai, Y., Kasamatsu, T., Sengoku, K., Hirohashi, S., Inazawa, J., & Imoto, I. (2007). Promoter hypermethylation contributes to frequent inactivation of a putative conditional tumor suppressor gene connective tissue growth factor in ovarian cancer. Cancer Research, 67(15), 7095-7105.
Klein, G. (1995). The extracellular matrix of the hematopoietic microenvironment. Experientia, 51(9-10), 914-926.
Knowles, J. P., Shi Wen, X., Haque, S. u., Bhalla, A., Dashwood, M. R., Yang, S., Taylor, I., Winslet, M. C., Abraham, D. J., & Loizidou, M. (2011). Endothelin 1 stimulates colon cancer adjacent fibroblasts. International Journal of Cancer, 130(6), 1264-1272.
Koliopanos, A., Friess, H., di Mola, F. F., Tang, W.-H., Kubulus, D., Brigstock, D., Zimmermann, A., & Büchler, M. W. (2002). Connective tissue growth factor gene expression alters tumor progression in esophageal cancer. World Journal of Surgery, 26(4), 420-427.
Kondo, S., Kubota, S., Shimo, T., Nishida, T., Yosimichi, G., Eguchi, T., Sugahara, T., & Takigawa, M. (2002). Connective tissue growth factor increased by hypoxia may initiate angiogenesis in collaboration with matrix metalloproteinases. Carcinogenesis, 23(5), 769-776.
Kostopoulou, F., Gkretsi, V., Malizos, K. N., Iliopoulos, D., Oikonomou, P., Poultsides, L., & Tsezou, A. (2012). Central role of srebp-2 in the pathogenesis of osteoarthritis. PloS one, 7(5), e35753.
Kouniavsky, G., Khaikin, M., Zvibel, I., Zippel, D., Brill, S., Halpern, Z., & Papa, M. (2002). Stromal extracellular matrix reduces chemotherapy-induced apoptosis in colon cancer cell lines. Clinical & Experimental Metastasis, 19(1), 55-60.
Kubo, H., Shimizu, M., Taya, Y., Kawamoto, T., Michida, M., & Kaneko, E. (2009). Identification of mesenchymal stem cell (msc)-transcription factors by microarray and knockdown analyses, and signature molecule-marked msc in bone marrow by immunohistochemistry. Genes to cells, 14(3), 407-424.
Kurosaka, D., LeBien, T. W., & Pribyl, J. A. R. (1999). Comparative studies of different stromal cell microenvironments in support of human b-cell development. Experimental Hematology, 27(8), 1271-1281.
Kuter, D. J., Bain, B., Mufti, G., Bagg, A., & Hasserjian, R. P. (2007). Bone marrow fibrosis: Pathophysiology and clinical significance of increased bone marrow stromal fibres. British Journal of Haematology, 139(3), 351-362.
Kwon, S., Munroe, X., Crawley, S., Lee, H., Spong, S., & Bradham, D. (2007). Expression of connective tissue growth factor in pancreatic cancer cell lines. International journal of oncology, 31(4), 693-703.
Lassailly, F., Foster, K., Lopez-Onieva, L., Currie, E., & Bonnet, D. (2013). Multimodal imaging reveals structural and functional heterogeneity in different bone marrow compartments: Functional implications on hematopoietic stem cells. Blood, 122(10), 1730-1740.
206
Lavall, D., Schuster, P., Selzer, C., Adam, O., Schaefers, H. J., Boehm, M., & Laufs, U. (2013). The mineralocorticoid receptor promotes pro-fibrotic remodeling in atrial fibrillation. European Heart Journal, 34(suppl 1).
Leask, A. (2002). Connective tissue growth factor: A new and important player in the pathogenesis of fibrosis. Current rheumatology reports, 4(2), 136-142.
Leask, A., & Abraham, D. J. (2006). All in the ccn family: Essential matricellular signaling modulators emerge from the bunker. J Cell Sci, 119(23), 4803-4810.
Lee, C. H., Shah, B., Moioli, E. K., & Mao, J. J. (2010). Ctgf directs fibroblast differentiation from human mesenchymal stem/stromal cells and defines connective tissue healing in a rodent injury model. The Journal of clinical investigation, 120(9), 3340.
Levental, K. R., Yu, H., Kass, L., Lakins, J. N., Egeblad, M., Erler, J. T., Fong, S. F. T., Csiszar, K., Giaccia, A., Weninger, W., Yamauchi, M., Gasser, D. L., & Weaver, V. M. (2009). Matrix crosslinking forces tumor progression by enhancing integrin signaling. Cell, 139(5), 891-906.
Li, L. Y., Li, E. M., Wu, Z. Y., Huang, X., Shen, J. H., Xu, X. E., Wu, J. Y., Huang, Q., & Xu, L. Y. (2010). Connective tissue growth factor expression in precancerous lesions of human esophageal epithelium and prognostic significance in esophageal squamous cell carcinoma. Diseases of the Esophagus, no-no.
Li, M.-H., Sanchez, T., Pappalardo, A., Lynch, K. R., Hla, T., & Ferrer, F. (2008). Induction of antiproliferative connective tissue growth factor expression in wilms' tumor cells by sphingosine-1-phosphate receptor 2. Molecular Cancer Research, 6(10), 1649-1656.
Lichtman, M. A. (1981). The ultrastructure of the hemopoietic environment of the marrow: A review. Experimental Hematology, 9(4), 391-410.
Liem, N. L. M., Papa, R. A., Milross, C. G., Schmid, M. A., Tajbakhsh, M., Choi, S., Ramirez, C. D., Rice, A. M., Haber, M., Norris, M. D., MacKenzie, K. L., & Lock, R. B. (2004). Characterization of childhood acute lymphoblastic leukemia xenograft models for the preclinical evaluation of new therapies. Blood, 103(10), 3905-3914.
Lin, B.-R., Chang, C.-C., Che, T.-F., Chen, S.-T., Chen, R. J.-C., Yang, C.-Y., Jeng, Y.-M., Liang, J.-T., Lee, P.-H., Chang, K.-J., Chau, Y.-P., & Kuo, M.-L. (2005). Connective tissue growth factor inhibits metastasis and acts as an independent prognostic marker in colorectal cancer. Gastroenterology, 128(1), 9-23.
Lin, B.-R., Chang, C.-C., Chen, R. J.-C., Jeng, Y.-M., Liang, J.-T., Lee, P.-H., Chang, K.-J., & Kuo, M.-L. (2011). Connective tissue growth factor acts as a therapeutic agent and predictor for peritoneal carcinomatosis of colorectal cancer. Clinical Cancer Research, 17(10), 3077-3088.
Lin, W.-W., & Karin, M. (2007). A cytokine-mediated link between innate immunity, inflammation, and cancer. The Journal of Clinical Investigation, 117(5), 1175-1183.
207
Lindner, U., Kramer, J., Rohwedel, J., & Schlenke, P. (2010). Mesenchymal stem or stromal cells: Toward a better understanding of their biology? Transfusion Medicine and Hemotherapy, 37(2), 75-83.
Liu, L., Han, Y., Wu, S., & Lv, Z. (2008). Expression of connective tissue growth factor in tumor tissues is an independent predictor of poor prognosis in patients with gastric cancer. World Journal of Gastroenterology, 14(13), 2110-2114.
Lock, R. B., Liem, N., Farnsworth, M. L., Milross, C. G., Xue, C., Tajbakhsh, M., Haber, M., Norris, M. D., Marshall, G. M., & Rice, A. M. (2002). The nonobese diabetic/severe combined immunodeficient (nod/scid) mouse model of childhood acute lymphoblastic leukemia reveals intrinsic differences in biologic characteristics at diagnosis and relapse. Blood, 99(11), 4100-4108.
Lu, H., Kojima, K., Battula, V., Korchin, B., Shi, Y., Chen, Y., Spong, S., Thomas, D., Kantarjian, H., Lock, R., Andreeff, M., & Konopleva, M. (2013). Targeting connective tissue growth factor (ctgf) in acute lymphoblastic leukemia preclinical models: Anti-ctgf monoclonal antibody attenuates leukemia growth. Annals of Hematology, 1-8.
Macanas-Pirard, P., Leisewitz, A., Broekhuizen, R., Cautivo, K., Barriga, F. M., Leisewitz, F., Gidi, V., Riquelme, E., Montecinos, V. P., & Swett, P. (2012). Bone marrow stromal cells modulate mouse ent1 activity and protect leukemia cells from cytarabine induced apoptosis. PloS one, 7(5), e37203.
Manabe, A., Coustansmith, E., Behm, F. G., Raimondi, S. C., & Campana, D. (1992). Bone marrow-derived stromal cells prevent apoptotic cell-death in b-lineage acute lymphoblastic-leukemia. Blood, 79(9), 2370-2377.
Manabe, A., Murti, K., Coustan-Smith, E., Kumagai, M., Behm, F., Raimondi, S., & Campana, D. (1994). Adhesion-dependent survival of normal and leukemic human b lymphoblasts on bone marrow stromal cells. Blood, 83(3), 758-766.
Mao, Z., Ma, X., Rong, Y., Cui, L., Wang, X., Wu, W., Zhang, J., & Jin, D. (2011). Connective tissue growth factor enhances the migration of gastric cancer through downregulation of e-cadherin via the nf-κb pathway. Cancer Science, 102(1), 104-110.
Martin, G. S. (2003). Cell signaling and cancer. Cancer Cell, 4(3), 167-174.
Mayani, H. (1992). Biology of the hemopoietic microenvironment. European journal of haematology, 49(5), 225.
Mazzocca, A., Fransvea, E., Dituri, F., Lupo, L., Antonaci, S., & Giannelli, G. (2010). Down-regulation of connective tissue growth factor by inhibition of transforming growth factor β blocks the tumor–stroma cross-talk and tumor progression in hepatocellular carcinoma. Hepatology, 51(2), 523-534.
McDoniels-Silvers, A. L., Nimri, C. F., Stoner, G. D., Lubet, R. A., & You, M. (2002). Differential gene expression in human lung adenocarcinomas and squamous cell carcinomas. Clinical cancer research, 8(4), 1127-1138.
208
Meads, M. B., Hazlehurst, L. A., & Dalton, W. S. (2008). The bone marrow microenvironment as a tumor sanctuary and contributor to drug resistance. Clinical Cancer Research, 14(9), 2519-2526.
Meyer, L. H., & Debatin, K.-M. (2011). Diversity of human leukemia xenograft mouse models: Implications for disease biology. Cancer Research, 71(23), 7141-7144.
Minn, A. J., Kang, Y., Serganova, I., Gupta, G. P., Giri, D. D., Doubrovin, M., Ponomarev, V., Gerald, W. L., Blasberg, R., & Massagué, J. (2005). Distinct organ-specific metastatic potential of individual breast cancer cells and primary tumors. J Clin Invest, 115(1), 44-55.
Mok, S. C., Chan, W. Y., Wong, K. K., Muto, M. G., & Berkowitz, R. S. (1996). Sparc, an extracellular matrix protein with tumor-suppressing activity in human ovarian epithelial cells. Oncogene, 12(9), 1895-1901.
Morrison, S. J., & Scadden, D. T. (2014). The bone marrow niche for haematopoietic stem cells. [Insight]. Nature, 505(7483), 327-334.
Mudry, R. E., Fortney, J. E., York, T., Hall, B. M., & Gibson, L. F. (2000). Stromal cells regulate survival of b-lineage leukemic cells during chemotherapy. Blood, 96(5), 1926-1932.
Mullighan, C. G., Phillips, L. A., Su, X., Ma, J., Miller, C. B., Shurtleff, S. A., & Downing, J. R. (2008). Genomic analysis of the clonal origins of relapsed acute lymphoblastic leukemia. Science, 322(5906), 1377-1380.
Mullighan, C. G. (2013). Genomic profiling of b-progenitor acute lymphoblastic leukemia. Best Practice & Research Clinical Haematology, 24(4), 489-503.
Munemasa, S., Sakai, A., Kuroda, Y., Okikawa, Y., Katayama, Y., Asaoku, H., Kubo, T., Miyakawa, Y., Serikawa, M., & Sasaki, T. (2007). Connective tissue growth factor is an indicator of bone involvement in multiple myeloma, but matrix metalloproteinase 9 is not. British journal of haematology, 139(1), 41-50.
Nagasawa, T. (2006). Microenvironmental niches in the bone marrow required for b-cell development. Nat Rev Immunol, 6(2), 107-116.
Nagasawa, T., Omatsu, Y., & Sugiyama, T. (2011). Control of hematopoietic stem cells by the bone marrow stromal niche: The role of reticular cells. Trends in Immunology, 32(7), 315-320.
Neesse, A., Frese, K. K., Bapiro, T. E., Nakagawa, T., Sternlicht, M. D., Seeley, T. W., Pilarsky, C., Jodrell, D. I., Spong, S. M., & Tuveson, D. A. (2013). Ctgf antagonism with mab fg-3019 enhances chemotherapy response without increasing drug delivery in murine ductal pancreas cancer. Proceedings of the National Academy of Sciences.
Nemeth, J. A., Nakada, M. T., Trikha, M., Lang, Z., Gordon, M. S., Jayson, G. C., Corringham, R., Prabhakar, U., Davis, H. M., & Beckman, R. A. (2007). Alpha-v integrins as therapeutic targets in oncology. Cancer Investigation, 25(7), 632-646.
209
Ness, K. K., Armenian, S. H., Kadan-Lottick, N., & Gurney, J. G. (2011). Adverse effects of treatment in childhood acute lymphoblastic leukemia: General overview and implications for long-term cardiac health. Expert Review of Hematology, 4(2), 185-197.
Noren-Nystrom, U., Roos, G., Bergh, A., Botling, J., Lonnerholm, G., Porwit, A., Heyman, M., & Forestier, E. (2007). Bone marrow fibrosis in childhood acute lymphoblastic leukemia correlates to biological factors, treatment response and outcome. Leukemia, 22(3), 504-510.
O'Brien, M. M., & Lacayo, N. J. (2008). Acute leukemia in children. Disease-a-Month, 54(4), 202-225.
O'Loughlin, A., & O'Brien, T. (2011). Topical stem and progenitor cell therapy for diabetic foot ulcers. Stem cells in clinic and research, Tech open acess publisher, Rijeka, 579-604.
Ono, M., Kubota, S., Fujisawa, T., Sonoyama, W., Kawaki, H., Akiyama, K., Oshima, M., Nishida, T., Yoshida, Y., Suzuki, K., Takigawa, M., & Kuboki, T. (2007). Promotion of attachment of human bone marrow stromal cells by ccn2. Biochemical and Biophysical Research Communications, 357(1), 20-25.
Pandey, D. P., Lappano, R., Albanito, L., Madeo, A., Maggiolini, M., & Picard, D. (2009). Estrogenic gpr30 signalling induces proliferation and migration of breast cancer cells through ctgf. EMBO J, 28(5), 523-532.
Paraguassu-Braga, F. H., & Paraguassu, B. (2003). Bone marrow stroma inhibits proliferation and apoptosis in leukemic cells through gap junction-mediated cell communication. Cell Death and Differentiation, 10(9), 1101.
Perez-Atayde, A. R., Sallan, S. E., Tedrow, U., Connors, S., Allred, E., & Folkman, J. (1997). Spectrum of tumor angiogenesis in the bone marrow of children with acute lymphoblastic leukemia. The American journal of pathology, 150(3), 815.
Pham, L. V., Tamayo, A. T., Li, C., Bueso-Ramos, C., & Ford, R. J. (2010). An epigenetic chromatin remodeling role for nfatc1 in transcriptional regulation of growth and survival genes in diffuse large b-cell lymphomas. Blood, 116(19), 3899-3906.
Potter, J. D. (2007). Morphogens, morphostats, microarchitecture and malignancy. Nat Rev Cancer, 7(6), 464-474.
Pritchard-Jones, K., Kaatsch, P., Steliarova-Foucher, E., Stiller, C. A., & Coebergh, J. W. W. (2006). Cancer in children and adolescents in europe: Developments over 20 years and future challenges. European Journal of Cancer, 42(13), 2183-2190.
Proudfoot, A. E. I. (2002). Chemokine receptors: Multifaceted therapeutic targets. Nat Rev Immunol, 2(2), 106-115.
Psaila, B., Lyden, D., & Roberts, I. (2012). Megakaryocytes, malignancy and bone marrow vascular niches. Journal of Thrombosis and Haemostasis, 10(2), 177-188.
210
Pui, C. H. (2006). Treatment of acute lymphoblastic leukemia. New England Journal of Medicine, The, 354(2), 166.
Pui, C. H. (2008). Acute lymphoblastic leukaemia. The Lancet, 371(9617), 1030.
Pui, C. H. (2009). Toward a total cure for acute lymphoblastic leukemia. Journal of Clinical Oncology, 27(31), 5121-5123.
Pui, C. H., Pei, D., Sandlund, J. T., Ribeiro, R. C., Rubnitz, J. E., Raimondi, S. C., Onciu, M., Campana, D., Kun, L. E., Jeha, S., Cheng, C., Howard, S. C., Metzger, M. L., Bhojwani, D., Downing, J. R., Evans, W. E., & Relling, M. V. (2010). Long-term results of st jude total therapy studies 11, 12, 13a, 13b, and 14 for childhood acute lymphoblastic leukemia. Leukemia, 24(2), 371-382.
Qi, J. H., Ebrahem, Q., Moore, N., Murphy, G., Claesson-Welsh, L., Bond, M., Baker, A., & Anand-Apte, B. (2003). A novel function for tissue inhibitor of metalloproteinases-3 (timp3): Inhibition of angiogenesis by blockage of vegf binding to vegf receptor-2. Nat Med, 9(4), 407-415.
Ravandi, F., Talpaz, M., & Estrov, Z. (2003). Modulation of cellular signaling pathways: Prospects for targeted therapy in hematological malignancies. Clinical Cancer Research, 9(2), 535-550.
Rebhan, M., Chalifa-Caspi, V., Prilusky, J., & Lancet, D. (1998). Genecards: A novel functional genomics compendium with automated data mining and query reformulation support. Bioinformatics, 14(8), 656-664.
Red-Horse, K., Crawford, Y., Shojaei, F., & Ferrara, N. (2007). Endothelium-microenvironment interactions in the developing embryo and in the adult. Developmental Cell, 12(2), 181-194.
Reya, T., & Rudolf, G. (1998). Transcriptional regulation of b-cell differentiation. Current opinion in immunology, 10(2), 158-165.
Rodriguez-Manzaneque, J. C., Lane, T. F., Ortega, M. A., Hynes, R. O., Lawler, J., & Iruela-Arispe, M. L. (2001). Thrombospondin-1 suppresses spontaneous tumor growth and inhibits activation of matrix metalloproteinase-9 and mobilization of vascular endothelial growth factor. Proceedings of the National Academy of Sciences, 98(22), 12485-12490.
Roecklein, B., & Torok-Storb, B. (1995a). Functionally distinct human marrow stromal cell lines immortalized by transduction with the human papilloma virus e6/e7 genes. Blood, 85(4), 997-1005.
Roecklein, B. A., & Torok-Storb, B. (1995b). Functionally distinct human marrow stromal cell lines immortalized by transduction with the human papilloma virus e6/e7 genes. Blood, 85(4), 997-1005.
Rosenwald, A., Wright, G., Chan, W. C., Connors, J. M., Campo, E., Fisher, R. I., Gascoyne, R. D., Muller-Hermelink, H. K., Smeland, E. B., & Giltnane, J. M. (2002). The use of molecular profiling to predict survival after chemotherapy for diffuse large-b-cell lymphoma. New England Journal of Medicine, 346(25), 1937-1947.
211
Safadi, F. F., Xu, J., Smock, S. L., Kanaan, R. A., Selim, A.-H., Odgren, P. R., Marks, S. C., Owen, T. A., & Popoff, S. N. (2003). Expression of connective tissue growth factor in bone: Its role in osteoblast proliferation and differentiation in vitro and bone formation in vivo. Journal of Cellular Physiology, 196(1), 51-62.
Said, N., Frierson, H. F., Sanchez-Carbayo, M., Brekken, R. A., & Theodorescu, D. (2013). Loss of sparc in bladder cancer enhances carcinogenesis and progression. The Journal of Clinical Investigation, 123(2), 751.
Sala-Torra, O., Gundacker, H. M., Stirewalt, D. L., Ladne, P. A., Pogosova-Agadjanyan, E. L., Slovak, M. L., Willman, C. L., Heimfeld, S., Boldt, D. H., & Radich, J. P. (2005). Ctgf (ccn2) predicts os and dfs in adult acute lymphoblastic leukemia. ASH Annual Meeting Abstracts, 106(11), 336-.
Sala-Torra, O., Gundacker, H. M., Stirewalt, D. L., Ladne, P. A., Pogosova-Agadjanyan, E. L., Slovak, M. L., Willman, C. L., Heimfeld, S., Boldt, D. H., & Radich, J. P. (2007). Connective tissue growth factor (ctgf) expression and outcome in adult patients with acute lymphoblastic leukemia. Blood, 109(7), 3080-3083.
Samarin, J., Cicha, I., & Goppelt-Struebe, M. (2009). Cell type-specific regulation of ccn2 protein expression by pi3k-akt-foxo signaling. Journal of Cell Communication and Signaling, 3(1), 79-84.
Schepers, K., Pietras, E. M., Reynaud, D., Flach, J., Binnewies, M., Garg, T., Wagers, A. J., Hsiao, E. C., & Passegue, E. (2013a). Myeloproliferative neoplasia remodels the endosteal bone marrow niche into a self-reinforcing leukemic niche. Cell Stem Cell, 13(3), 285-299.
Schepers, K., Pietras, E. M., Reynaud, D., Flach, J., Binnewies, M., Garg, T., Wagers, A. J., Hsiao, E. C., & PasseguÈ, E. (2013b). Myeloproliferative neoplasia remodels the endosteal bone marrow niche into a self-reinforcing leukemic niche. Cell stem cell, 13(3), 285-299.
Schofield, R. (1978). The relationship between the spleen colony-forming cell and the haemopoietic stem cell. Blood cells, 4(1-2), 7-25.
Schwartz, C. L. (1999). Long-term survivors of childhood cancer: The late effects of therapy. The oncologist, 4(1), 45.
Seftor, E. A., Meltzer, P. S., Kirschmann, D. A., Pe'er, J., Maniotis, A. J., Trent, J. M., Folberg, R., & Hendrix, M. J. (2002). Molecular determinants of human uveal melanoma invasion and metastasis. Clinical and Experimental Metastasis, 19(3), 233-246.
Seshi, B. (2000). Human bone marrow stromal cell: Coexpression of markers specific for multiple mesenchymal cell lineages. Blood cells, molecules & diseases, 26(3), 234.
Shain, K. H. (2000). The tumor microenvironment as a determinant of cancer cell survival: A possible mechanism for de novo drug resistance. Current opinion in oncology, 12(6), 557.
212
Shakunaga, T., Ozaki, T., Ohara, N., Asaumi, K., Nishida, K., Kawai, A., Nakanishi, T., Takigawa, M., & Inoue, H. (2000). Expression of connective tissue growth factor in cartilaginous tumors. Cancer, 89(7), 1466-1473.
Shen, Y., Winkler, I. G., Barbier, V., Sims, N. A., Hendy, J., & Levesque, J.-P. (2010). Tissue inhibitor of metalloproteinase-3 (timp-3) regulates hematopoiesis and bone formation in vivo. PloS one, 5(9), e13086.
Shi-Wen, X., Leask, A., & Abraham, D. (2008). Regulation and function of connective tissue growth factor/ccn2 in tissue repair, scarring and fibrosis. Cytokine & growth factor reviews, 19(2), 133-144.
Shimo, T., Nakanishi, T., Nishida, T., Asano, M., Kanyama, M., Kuboki, T., Tamatani, T., Tezuka, K., Takemura, M., & Matsumura, T. (1999). Connective tissue growth factor induces the proliferation, migration, and tube formation of vascular endothelial cells in vitro, and angiogenesis in vivo. Journal of Biochemistry, 126(1), 137-145.
Shimo, T., Nakanishi, T., Nishida, T., Asano, M., Sasaki, A., Kanyama, M., Kuboki, T., Matsumura, T., & Takigawa, M. (2001). Involvement of ctgf, a hypertrophic chondrocyte-specific gene product, in tumor angiogenesis. Oncology, 61(4), 315-322.
Shimo, T., Kubota, S., Yoshioka, N., Ibaragi, S., Isowa, S., Eguchi, T., Sasaki, A., & Takigawa, M. (2006). Pathogenic role of connective tissue growth factor (ctgf/ccn2) in osteolytic metastasis of breast cancer. Journal of Bone and Mineral Research, 21(7), 1045-1059.
Shultz, L. D., Schweitzer, P. A., Christianson, S. W., Gott, B., Schweitzer, I. B., Tennent, B., McKenna, S., Mobraaten, L., Rajan, T. V., & Greiner, D. L. (1995a). Multiple defects in innate and adaptive immunologic function in nod/ltsz-scid mice. The journal of immunology, 154(1), 180-191.
Shultz, L. D., Schweitzer, P. A., Christianson, S. W., Gott, B., Schweitzer, I. B., Tennent, B., McKenna, S., Mobraaten, L., Rajan, T. V., & Greiner, D. L. (1995b). Multiple defects in innate and adaptive immunologic function in nod/ltsz-scid mice. The journal of immunology, 154(1), 180-191.
Shur, I., Marom, R., Lokiec, F., Socher, R., & Benayahu, D. (2002). Identification of cultured progenitor cells from human marrow stroma. Journal of Cellular Biochemistry, 87(1), 51-57.
Smyth, G. K. (2004). Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol, 3(1), 3.
Smyth, G. K. (2005). Limma: Linear models for microarray data Bioinformatics and computational biology solutions using r and bioconductor (pp. 397-420): Springer.
Sonnylal, S., Shi-Wen, X., Leoni, P., Naff, K., Van Pelt, C. S., Nakamura, H., Leask, A., Abraham, D., Bou-Gharios, G., & de Crombrugghe, B. (2010). Selective expression of connective tissue growth factor in fibroblasts in vivo promotes systemic tissue fibrosis. Arthritis & Rheumatism, 62(5), 1523-1532.
213
Spriggs, D., Griffin, J., Wisch, J., & Kufe, D. (1985). Clinical pharmacology of low-dose cytosine arabinoside. Blood, 65(5), 1087-1089.
Stetler-Stevenson, W. G., & Seo, D.-W. (2005). Timp-2: An endogenous inhibitor of angiogenesis. Trends in Molecular Medicine, 11(3), 97-103.
Sugiyama, T., Kohara, H., Noda, M., & Nagasawa, T. (2006). Maintenance of the hematopoietic stem cell pool by cxcl12-cxcr4 chemokine signaling in bone marrow stromal cell niches. Immunity, 25(6), 977-988.
Suyani, E., Aki, S. Z., Yegin, Z. A., Ozkurt, Z. N., Altindal, N., Akyurek, N., Yagci, M., & Sucak, G. T. (2010). The impact of bone marrow fibrosis on the outcome of hematopoietic stem cell transplantation. Transplantation Proceedings, 42(7), 2713-2719.
Tabe, Y., Jin, L., Tsutsumi-Ishii, Y., Xu, Y., McQueen, T., Priebe, W., Mills, G. B., Ohsaka, A., Nagaoka, I., Andreeff, M., & Konopleva, M. (2007). Activation of integrin-linked kinase is a critical prosurvival pathway induced in leukemic cells by bone marrow-derived stromal cells. Cancer Research, 67(2), 684-694.
Tang, J., Hu, M., Lee, S., & Roblin, R. (2000). A polymerase chain reaction based method for detecting mycoplasma/acholeplasma contaminants in cell culture. Journal of Microbiological Methods, 39(2), 121-126.
Teitell, M. A., & Pandolfi, P. P. (2009). Molecular genetics of acute lymphoblastic leukemia. Annual Review of Pathology: Mechanisms of Disease, 4(1), 175-198.
Tesfai, Y., Ford, J., Carter, K. W., Firth, M. J., O‚'Leary, R. A., Gottardo, N. G., Cole, C., & Kees, U. R. (2012). Interactions between acute lymphoblastic leukemia and bone marrow stromal cells influence response to therapy. Leukemia Research, 36(3), 299-306.
Thomas, X., Le, Q., Danaila, C., Lheritier, V., & Ffrench, M. (2002). Bone marrow biopsy in adult acute lymphoblastic leukemia: Morphological characteristics and contribution to the study of prognostic factors. Leukemia Research, 26(10), 909-918.
Toh, B., Wang, X., Keeble, J., Sim, W. J., Khoo, K., Wong, W.-C., Kato, M., Prevost-Blondel, A., Thiery, J.-P., & Abastado, J.-P. (2011). Mesenchymal transition and dissemination of cancer cells is driven by myeloid-derived suppressor cells infiltrating the primary tumor. PLoS Biol, 9(9), e1001162.
Travlos, G. S. (2006). Normal structure, function, and histology of the bone marrow. Toxicologic Pathology, 34(5), 548-565.
Uhm, J. H., Dooley, N. P., Kyritsis, A. P., Rao, J. S., & Gladson, C. L. (1999). Vitronectin, a glioma-derived extracellular matrix protein, protects tumor cells from apoptotic death. Clinical Cancer Research, 5(6), 1587-1594.
Ushiki, T. (2002). Collagen fibers, reticular fibers and elastic fibers. A comprehensive understanding from a morphological viewpoint. Archives of histology and cytology, 65(2), 109-126.
214
van der Horst, G., van den Hoogen, C., Buijs, J. T., Cheung, H., Bloys, H., Pelger, R. C., Lorenzon, G., Heckmann, B., Feyen, J., Pujuguet, P., Blanque, R., Clement-Lacroix, P., & van der Pluijm, G. (2011). Targeting of alpha(v)-integrins in stem/progenitor cells and supportive microenvironment impairs bone metastasis in human prostate cancer. Neoplasia, 13(6), 516-525.
Veiga, J. P., Costa, L. F., Sallan, S. E., Nadler, L. M., & Cardoso, A. A. (2006). Leukemia-stimulated bone marrow endothelium promotes leukemia cell survival. Experimental Hematology, 34(5), 610-621.
Vora, A., Goulden, N., Wade, R., Mitchell, C., Hancock, J., Hough, R., Rowntree, C., & Richards, S. (2013). Treatment reduction for children and young adults with low-risk acute lymphoblastic leukaemia defined by minimal residual disease (ukall 2003): A randomised controlled trial. The Lancet Oncology, 14(3), 199-209.
Vorwerk, P., Wex, H., Hohmann, B., Oh, Y., Rosenfeld, R. G., & Mittler, U. (2000). Ctgf (igfbp-rp2) is specifically expressed in malignant lymphoblasts of patients with acute lymphoblastic leukaemia (all). British Journal of Cancer, 83(6), 756.
Wagner, W., Saffrich, R., & Ho, A. D. (2008). The stromal activity of mesenchymal stromal cells. Transfusion Medicine and Hemotherapy, 35(3), 185-193.
Wajapeyee, N., Serra, R. W., Zhu, X., Mahalingam, M., & Green, M. R. (2008). Oncogenic braf induces senescence and apoptosis through pathways mediated by the secreted protein igfbp7. Cell, 132(3), 363-374.
Wang, G.-B., Zhou, X.-Y., Yuan, T., Xie, J., Guo, L.-P., Gao, N., & Wang, X.-Q. (2010a). Significance of serum connective tissue growth factor in patients with hepatocellular carcinoma and relationship with angiogenesis. World Journal of Surgery, 34(10), 2411-2417.
Wang, J.-C., Lai, S., Guo, X., Zhang, X., de Crombrugghe, B., Sonnylal, S., Arnett, F., & Zhou, X. (2010b). Attenuation of fibrosis in vitro and in vivo with sparc sirna. Arthritis Research & Therapy, 12(2), R60.
Wang, J.-j., Ye, F., Cheng, L.-j., Shi, Y.-j., Bao, J., Sun, H.-q., Wang, W., Zhang, P., & Bu, H. (2009a). Osteogenic differentiation of mesenchymal stem cells promoted by overexpression of connective tissue growth factor. Journal of Zhejiang University SCIENCE B, 10(5), 355-367.
Wang, M.-Y., Chen, P.-S., Prakash, E., Hsu, H.-C., Huang, H.-Y., Lin, M.-T., Chang, K.-J., & Kuo, M.-L. (2009b). Connective tissue growth factor confers drug resistance in breast cancer through concomitant up-regulation of bcl-xl and ciap1. Cancer Research, 69(8), 3482-3491.
Welch, M. D., Greene, W. K., & Kees, U. R. (2013). Hypomethylation of the ctgf gene locus is a common feature of paediatric pre-b acute lymphoblastic leukaemia. British Journal of Haematology, 162(4), 537-541.
215
Wenger, C., Ellenrieder, V., Alber, B., Lacher, U., Menke, A., Hameister, H., Wilda, M., Iwamura, T., Beger, H. G., Adler, G., & Gress, T. M. (1999). Expression and differential regulation of connective tissue growth factor in pancreatic cancer cells. [Article]. Oncogene, 18(4), 1073-1080.
Willemze, R., Fibbe, W. E., & Zwaan, F. E. (1983). Experience with intermediate and high dose cytosine arabinoside in refractory acute leukaemia. Onkologie, 6(4), 200-204.
Wilson, A. (2006). Bone-marrow haematopoietic-stem-cell niches. Nature Reviews. Immunology, 6(2), 93.
Wong, Y.-F., Cheung, T.-H., Tsao, G. S. W., Lo, K. W. K., Yim, S.-F., Wang, V. W., Heung, M. M. S., Chan, S. C. S., Chan, L. K. Y., Ho, T. W. F., Wong, K. W. Y., Li, C., Guo, Y., Chung, T. K. H., & Smith, D. I. (2006). Genome-wide gene expression profiling of cervical cancer in hong kong women by oligonucleotide microarray. International Journal of Cancer, 118(10), 2461-2469.
Wu, S., Korte, A., Kebelmann-Betzing, C., Gessner, R., Henze, G., & Seeger, K. (2005). Interaction of bone marrow stromal cells with lymphoblasts and effects of predinsolone on cytokine expression. Leukemia Research, 29(1), 63-72.
Xiang, Z.-L., Zeng, Z.-C., Fan, J., Tang, Z.-Y., & Zeng, H.-Y. (2012). Expression of connective tissue growth factor and interleukin-11 in intratumoral tissue is associated with poor survival after curative resection of hepatocellular carcinoma. Molecular biology reports, 1-6.
Xie, D., Nakachi, K., Wang, H., Elashoff, R., & Koeffler, H. P. (2001). Elevated levels of connective tissue growth factor, wisp-1, and cyr61 in primary breast cancers associated with more advanced features. Cancer Research, 61(24), 8917-8923.
Xie, D., Yin, D., Wang, H.-J., Liu, G.-T., Elashoff, R., Black, K., & Koeffler, H. P. (2004). Levels of expression of cyr61 and ctgf are prognostic for tumor progression and survival of individuals with gliomas. Clinical Cancer Research, 10(6), 2072-2081.
Xie, H., Zhao, Y., Caramuta, S., Larsson, C., & Lui, W.-O. (2012). Mir-205 expression promotes cell proliferation and migration of human cervical cancer cells. PloS one, 7(10), e46990.
Xiu, M., Liu, Y.-H., Brigstock, D. R., He, F.-H., Zhang, R.-J., & Gao, R.-P. (2012). Connective tissue growth factor is overexpressed in human hepatocellular carcinoma and promotes cell invasion and growth. World Journal of Gastroenterology: WJG, 18(47), 7070.
Yamada, E., Tobe, T., Yamada, H., Okamoto, N., Zack, D. J., Werb, Z., Soloway, P. D., & Campochiaro, P. A. (2001). Timp-1 promotes vegf-induced neovascularization in the retina.
Yang, F. (2005). Stromal expression of connective tissue growth factor promotes angiogenesis and prostate cancer tumorigenesis. Cancer Research, 65(19), 8887.
216
Yang, M., Lin, B., Chang, C., Chen, S., Lin, S., Kuo, M. Y., Jeng, Y., Kuo, M., & Chang, C. (2011). Connective tissue growth factor modulates oral squamous cell carcinoma invasion by activating a mir-504/foxp1 signalling. Oncogene, 31(19), 2401-2411.
Yang, X., Zhang, Z., Qiu, M., Hu, J., Fan, X., Wang, J., Xu, L., & Yin, R. (2013). Glypican-5 is a novel metastasis suppressor gene in non-small cell lung cancer. Cancer Letters, 341(2), 265-273.
Yin, D., Chen, W., O'Kelly, J., Lu, D., Ham, M., Doan, N. B., Xie, D., Wang, C., Vadgama, J., Said, J. W., Black, K. L., & Koeffler, H. P. (2010). Connective tissue growth factor associated with oncogenic activities and drug resistance in glioblastoma multiforme. International Journal of Cancer, 127(10), 2257-2267.
Zeisberg, M., & Kalluri, R. (2013). Cellular mechanisms of tissue fibrosis. 1. Common and organ-specific mechanisms associated with tissue fibrosis. American Journal of Physiology - Cell Physiology, 304(3), C216-C225.
Zhou, Y., Capuco, A. V., & Jiang, H. (2008). Involvement of connective tissue growth factor (ctgf) in insulin-like growth factor-i (igf1) stimulation of proliferation of a bovine mammary epithelial cell line. Domestic animal endocrinology, 35(2), 180-189.
Zhou, Z.-Q., Cao, W.-H., Xie, J.-J., Lin, J., Shen, Z.-Y., Zhang, Q.-Y., Shen, J.-H., Xu, L.-Y., & Li, E.-M. (2009). Expression and prognostic significance of thbs1, cyr61 and ctgf in esophageal squamous cell carcinoma. BMC Cancer, 9(1), 291.
Zirn, B., Samans, B., Spangenberg, C., Graf, N., Eilers, M., & Gessler, M. (2005). All-trans retinoic acid treatment of wilms tumor cells reverses expression of genes associated with high risk and relapse in vivo. Oncogene, 24(33), 5246-5251.
Zirn, B., Hartmann, O., Samans, B., Krause, M., Wittmann, S., Mertens, F., Graf, N., Eilers, M., & Gessler, M. (2006). Expression profiling of wilms tumors reveals new candidate genes for different clinical parameters. International Journal of Cancer, 118(8), 1954-1962.
Appendix: Supplementary Figures
Tabl
e S1
. Ing
enui
ty p
athw
ay a
naly
sis
(IPA)
for d
iffer
entia
lly re
gula
ted
gene
s be
twee
n st
rom
al c
ells
(HS5
) tre
ated
with
rhCT
GF
(100
ng/m
l) or
con
trol b
uffe
r for
16
hour
s . T
he IP
A ge
ne lis
t com
prise
s ge
nes
iden
tifie
d sig
nific
antly
diff
eren
t by
LIM
MA
and
with
an
expr
essio
n di
ffere
nce
grea
ter t
han
two-
fold
com
pare
d to
unt
reat
ed c
ontro
l cel
ls (B
enja
min
i and
Hoc
hber
g-ad
just
ed p
-val
ue <
0.05
, fol
d ch
ange
>2
and
<2).
Num
ber o
f upr
egul
ated
(·) and
dow
nreg
ulat
ed (‚) g
enes
indi
cate
d we
re u
sed
to g
ener
ate
IPA
signi
fican
t net
work
, ca
noni
cal p
athw
ay a
nd k
nown
bio
logi
cal f
unct
ion
asso
ciatio
ns.
Cellu
lar A
ssem
bly
and
Org
aniza
tion
Cellu
lar F
unct
ion
and
Mai
nten
ance
Amin
oacy
l-tRN
A Bi
osyn
thes
is
Glyc
ine,
Ser
ine
and
Thre
onin
e M
etab
olism
1.Re
nal a
nd U
rolo
gica
l Dise
ase,
Gen
etic
Diso
rder
, Sk
elet
al a
nd M
uscu
lar D
isord
ers
2. N
ervo
us S
yste
m D
evel
opm
ent a
nd F
unct
ion,
Ti
ssue
Mor
phol
ogy,
Amin
o Ac
id M
etab
olism
·136 ‚23
Cellu
lar M
ovem
ent
Amyo
troph
ic La
tera
l Scle
rosis
Si
gnal
ing
3.
Infe
ctio
n M
echa
nism
, Inf
ectio
us D
iseas
e, C
ell
Mor
phol
ogy
Cell M
orph
olog
y M
ethi
onin
e M
etab
olism
Amin
o Ac
id M
etab
olism
Mito
chon
dria
l Dys
func
tion
IPA
Netw
ork
# G
enes
Top
Biol
ogic
al F
unct
ions
IPA
Cano
nica
l
Tabl
e S2
. Ing
enui
ty p
athw
ay a
naly
sis
(IPA)
for d
iffer
entia
lly re
gula
ted
gene
s be
twee
n st
rom
al c
ells
(HS5
) tre
ated
with
rhCT
GF
(100
ng/m
l) or
con
trol b
uffe
r for
48
hour
s . T
he IP
A ge
ne lis
t com
prise
s ge
nes
iden
tifie
d sig
nific
antly
diff
eren
t by
LIM
MA
and
with
an
expr
essio
n di
ffere
nce
grea
ter t
han
two-
fold
com
pare
d to
unt
reat
ed c
ontro
l cel
ls (B
enja
min
i and
Hoc
hber
g-ad
just
ed p
-val
ue <
0.05
, fol
d ch
ange
>2
and
<2).
Num
ber o
f upr
egul
ated
(·) and
dow
nreg
ulat
ed (‚) g
enes
indi
cate
d we
re u
sed
to g
ener
ate
IPA
signi
fican
t net
work
, ca
noni
cal p
athw
ay a
nd k
nown
bio
logi
cal f
unct
ion
asso
ciatio
ns.
Cellu
lar A
ssem
bly
and
Org
aniza
tion
Cellu
lar F
unct
ion
and
Mai
nten
ance
Hepa
tic F
ibro
sis /
Hepa
tic
Stel
late
Cel
l Act
ivatio
n
Coag
ulat
ion
Syst
em
1. L
ipid
Met
abol
ism, S
mal
l Mol
ecul
e Bi
oche
mist
ry,
Cellu
lar D
evel
opm
ent
2. G
enet
ic Di
sord
er, H
emat
olog
ical D
iseas
e,
Ce
ll Dea
th
·89 ‚180Ce
llula
r Mov
emen
t Pr
opan
oate
Met
abol
ism3.
Cel
l-To-
Cell S
igna
ling
and
Inte
ract
ion,
Hep
atic
Syst
em D
iseas
e, In
flam
mat
ory
Dise
ase
Post
-Tra
nsla
tiona
l Mod
ificat
ion
Intri
nsic
Prot
hrom
bin
Activ
atio
n Pa
thwa
y
Cellu
lar D
evel
opm
ent
p53
Sign
alin
g
IPA
Netw
ork
# G
enes
Top
Biol
ogic
al F
unct
ions
IPA
Cano
nica
l
secreted genes upregulated at 8 hours
Gene
Symbol Aliases & Descriptions
Fold
change
COL3A1 collagen, type 3, alpha 1 15.07LOX lysyl oxidase 7.08PLCB4 phospholipase C, beta 4 6.99PAPPA pregnacy-associated plasma protein A 6.26EFEMP1 EGF-containing fibulin-like extracellular matrix protein 1 5.42CYP51A1 cytochrome P450, family 51, subfamily A, polypeptide 1 5.35LUM lumican 4.81DPP4 dipeptidyl-peptidase 4 4.44MATN2 matrilin 2 4.30COL1A2 collagen, type 1, alpha 2 4.23FN1 fibronectin 1 4.00ROBO1 roundabout homolog 1 3.88NUCB2 nucleobindin 2 3.83NPTX2 neuronal pentraxin 2 3.68LAMB1 laminin subunit beta 1 3.66CCL2 chemokine (C-C motif) ligand 2 3.60TGFBR3 transforming growth factor beta receptor type 3 3.55NID2 nidogen 2 3.36COL5A1 collagen, type 5, alpha 1 3.33GALC Galactosylceramidase 3.26FBN2 fibrillin 2 3.26SEMA3C semaphorin-3C 3.21CALU calumenin 3.10PTX3 pentraxin 3 3.05SEC24A SEC24 related gene family, member A 3.05TNFAIP6 tumor necrosis factor alpha-inducible protein 6 3.01DCN decorin 2.95C5orf13 chromosome 5 open reading frame 13 2.94
TMEFF1transmembrane protein with EGF-like and two follistatin-like domains 1 2.87
PSG1 pregnancy specific beta-1-glycoprotein 1 2.82CCL7 chemokine (C-C motif) ligand 7 2.77TGFB2 transforming growth factor beta 2 2.75COLEC12 collectin sub-family member 12 2.75IL1RAP interleukin 1 receptor accessory protein 2.67ASAH1 N-acylsphingosine amidohydrolase (acid ceramidase) 1 2.67TUFT1 tuftelin 1 2.67ZNF23 zinc finger protein 32 2.67RANBP9 ran-binding protein 9 2.60SYNJ1 synaptojanin 1 2.60GNS glucosamine (N-acetyl)-6-sulfatase 2.59
TFPItissue factor pathway inhibitor (lipoprotein-associated coagulation inhibitor) 2.59
AGA aspartylglucosaminidase 2.59ADAM12 a disintegrin and metalloproteinase domain 12 (meltrin alpha) 2.55MATN3 matrilin 3 2.53CCL8 chemokine (C-C motif) ligand 8 2.50
ENPP2ectonucleotide pyrophosphatase/phosphodiesterase family member 2 2.49
secreted genes upregulated at 8 hours
Gene
Symbol Aliases & Descriptions
Fold
change
WNT5A wingless-type MMTV integration site family, member 5A 2.49TGOLN2 trans-golgi network protein 2 2.48CLIC4 chloride intracellular channel protein 4 2.45
ALG5asparagine-linked glycosylation 5, dolichyl-phosphate beta-glucosyltransferase homolog (S. cerevisiae) 2.45
TTC3 tetratricopeptide repeat domain 3 2.43THOC2 THO complex 2 2.41INHBA inhibin beta A chain 2.39MYLK myosin light chain kinase 2.38BDNF brain-derived neurotrophic factor 2.36ENPP1 ectonucleotide pyrophosphatase/phosphodiesterase 1 2.36FBN1 fibrillin 1 2.35SHOX2 short stature homeobox protein 2 2.33
SERPINE2 serpin peptidase inhibitor, clade E (nexin, plasminogen activator inhibitor type 1), member 2 2.32
ITGBL1 integrin beta-like protein 1 2.31CHSY1 chondroitin synthase 1 2.27RBM3 RNA binding motif protein 3 2.21
B4GALT4beta-N-acetylglucosaminyl-glycolipid beta-1,4-galactosyltransferase 4 2.19
STAM2signal transducing adaptor molecule (SH3 domain and ITAM motif) 2 2.17
GALNT7 UDP-GalNAc:polypeptide N-acetylgalactosaminyltransferase 7 2.15CRIM1 cysteine-rich motor neuron 1 2.14FADS1 fatty acid desaturase 1 2.14C1S complement component 1, s subcomponent 2.12GRSF1 G-rich RNA sequence binding factor 1 2.11RDH11 retinol dehydrogenase 11 (all-trans/9-cis/11-cis) 2.10FAM3C family with sequence similarity 3, member C 2.08IL1R1 interleukin 1 receptor, type 1 2.07CYLD ubiquitin carboxyl-terminal hydrolase CYLD 2.05TFPI2 tissue factor pathway inhibitor 2 2.05TFRC transferrin receptor (p90, CD71) 2.05PDGFC platelet derived growth factor C 2.05
C1GALT1core 1 synthase, glycoprotein-N-acetylgalactosamine 3-beta-galactosyltransferase, 1 2.04
C1orf9 chromosome 1 open reading frame 9 2.03
Table S3(HS5) in the presence of CTGF. HS5 cells were incubated for 8 hours in the presence of rhCTGF (100ng/ml) or control buffer. RNA was used for microarray analysis. Gene list comprises secreted genes identified significantly upregulated by LIMMA (FDR-adjusted P-value <0.05, fold change >2) and present on the Secreted Protein Discovery Initiative (SPDI, Clarke et al., 2003).
. Significantly upregulated genes secreted from stromal cells
secreted genes downregulated at 8 hours
Gene
Symbol Aliases & Descriptions
Fold
Change
CXCL5 neutrophil-activating peptide ENA-78 13.35
CXCL1chemokine (C-X-C motif) ligand 1 (melanoma growth stimulating activity, alpha) 8.85
COL6A2 collagen, type 6, alpha 2 8.09CALR calregulin 5.97
CXCL6chemokine (C-X-C motif) ligand 6 (granulocyte chemotactic protein 2) 5.83
PLAU plasminogen activator, urokinase 5.72GAL galanin peptides 5.28MIF macrophage migration inhibitory factor 4.91
CSF2 colony stimulating factor 2 (granulocyte-macrophage) 4.75COL6A1 collagen, type 6, alpha 1 4.51BMP2 bone morphogenetic protein 2 4.16CST3 cystatin 3 3.88
PLAUR urokinase plasminogen activator surface receptor 3.83STC1 stanniocalcin 1 3.78
PTPRF receptor-type tyrosine-protein phosphatase F 3.62
MCL1 myeloid cell leukemia sequence 1 (BCL2-related) 3.55TRIM28 transcription intermediary factor 1-beta 3.49SSR4 signal sequence receptor subunit delta 3.41SPRY4 sprouty homolog 4 (Drosophila) 3.40TGFB1 transforming growth factor, beta 1 3.39PRKCSH protein kinase C substrate, 80 Kda protein 3.28NMU neuromedin U 3.27FGF5 fibroblast growth factor 5 3.27
NDUFS8NADH dehydrogenase [ubiquinone] iron-sulfur protein 8, mitochondrial 3.16
MGAT4Balpha-1,3-mannosyl-glycoprotein beta-1,4-N-acetylglucosaminyltransferase 3.03
CD44 hematopoietic cell E- and L-selectin ligand 2.88
TIMP1tissue inhibitor of metalloproteinase 1 (erythroid potentiating activity, collagenase inhibitor) 2.77
CPNE1 copine 1 2.75COL1A1 collagen, type 1, alpha 1 2.73ADRM1 adhesion regulating molecule 1 2.71EMILIN1 elastin microfibril interfacer 1 2.58ANGPTL4 Angiopoietin-like protein 4 2.53AEBP1 adipocyte enhancer-binding protein 1 2.53CCL20 chemokine (C-C motif) ligand 20 2.53EREG proepiregulin 2.50LTBP3 latent-transforming growth factor beta-binding protein 3 2.50PGLS 6-phosphogluconolactonase 2.48GPX3 glutathione peroxidase 3 2.45GNAS guanine nucleotide regulatory protein 2.44ZNF185 zinc finger protein 185 (LIM domain) 2.41RCN3 reticulocalbin 3 2.41ESM1 endothelial cell-specific molecule 1 2.41
Table S4. Significantly downregulated genes secreted from stromal cells (HS5) in the presence of CTGF. HS5 cells were incubated for 8 hours in the presence of rhCTGF (100ng/ml) or control buffer. RNA was used for microarray analysis. Gene list comprises secreted genes identified significantly downregulated by LIMMA (FDR-adjusted P-value <0.05, fold change >2) and present on the Secreted Protein Discovery Initiative (SPDI, Clarke et al., 2003).
secreted genes downregulated at 8 hours
Gene
Symbol Aliases & Descriptions
Fold
Change
C12orf10 chromosome 12 open reading frame 10 2.37G0S2 G0/G1 switch regulatory protein 2 2.34TIMP2 TIMP metallopeptidase inhibitor 2 2.34
CREB3 cyclic AMP response element (CRE)-binding protein 3 (luman) 2.33C12orf11 chromosome 12 open reading frame 11 2.32G0S3 G0/G1 switch regulatory protein 3 2.31TIMP3 TIMP metallopeptidase inhibitor 3 2.30
CREB4 cyclic AMP response element (CRE)-binding protein 3 (luman) 2.28MFAP2 microfibrillar-associated protein 2 2.21EPHA2 ephrin type-A receptor 2 2.20SFRP1 secreted apoptosis-related protein 2 2.16CS citrate synthase, mitochondrial 2.16MMP12 matrix metallopeptidase 12 (macrophage elastase) 2.14IGFBP7 insulin-like growth factor-binding protein 7 2.12VNN1 vascular non-inflammatory molecule 1 2.11MAZ myc-associated zinc finger protein 2.10IGFBP3 insulin-like growth factor-binding protein 3 2.09LAMC3 laminin subunit gamma-3 2.07ECM1 extracellular matrix protein 1 2.07
SERPINH1serine (or cysteine) proteinase inhibitor, clade H (heat shock protein 47), member 2, (collagen-binding protein 2) 2.07
LGALS3BP lectin, galactoside-binding, soluble, 3 binding protein 2.03CTSD cathepsin D (lysosomal aspartyl protease) 2.03TGFA protransforming growth factor alpha 2.01
PSMD2proteasome (prosome, macropain) 26S subunit, non-ATPase, 2 2.00
Figure S1.NOD/SCID xenografts treated with high dose ARAC. Mice were injected
5 6via the tail-vein with 2x10 PER-278 High and Low cells or 1x10 PER-371 High and Low cells. At 1% engraftment in peripheral blood ARAC treatment was started at 100mg/kg/day according to treatment protocol a. PER-278 Low, and b. PER-278 High injected mice were administered saline or ARAC. c. PER-371 Low, and d. PER-371 High xenografts were administered saline or ARAC. Data are expressed as the percent survival of mice against time. (n=5-7). (Log-Rank test: **, p<0.01, ***, p<0.001). Arrow indicates first and last ARAC injection, drug toxicity is responsible for treated mice sacrificed before last ARAC injection.
Kaplan Meier survival plots for PER-278 and PER-371
b.
c.
a.
d.
Human L1000 Array, Membrane (AAH-BLM-1000)Detects 1000 Human Proteins.
11b-HSD1, 2B4, 4-1BB, 6Ckine, A1BG, A2M, ABL1, ACE, ACE-2, ACK1, ACPP, ACTH, Activin A, Activin B, Activin C, Activin RIA / ALK-2, Activin RIB / ALK-4, Activin RII A/B, Activin RIIA, ADAM-9, ADAMTS-1, ADAMTS-10, ADAMTS-13, ADAMTS-15, ADAMTS-17, ADAMTS-18, ADAMTS-19, ADAMTS-4 , ADAMTS-5, ADAMTS-L2, Adiponectin / Acrp30, Adipsin, Afamin, AFP, AgRP, ALBUMIN, ALCAM, Aldolase A, Aldolase B, Aldolase C, ALK, Alpha 1 AG, Alpha 1 Microglobulin, Alpha Lactalbumin, ALPP, AMICA, AMPKa1, Amylin, Angiogenin, Angiopoietin-1, Angiopoietin-2, Angiopoietin-4, Angiopoietin-like 1, Angiopoietin-like 2, Angiopoietin-like Factor, Angiostatin, ANGPTL3, ANGPTL4, Annexin A7, APC, APCS, Apelin, Apex1, APJ, APN, ApoA1, ApoA2, ApoA4, ApoB, ApoB100, ApoC1, ApoC2, ApoC3, ApoD, ApoE, ApoE3, ApoH, ApoM, APP, APRIL, AR (Amphiregulin), Artemin, ASPH, Attractin, Axl, B3GNT1, B7-1 /CD80, BAF57, BAFF, BAFF R / TNFRSF13C, BAI-1, BCAM, BCMA / TNFRSF17, BD-1, BDNF, Beta 2M, Beta Defensin 4, Beta IG-H3, beta-Catenin, beta-Defensin 2, beta-NGF, Biglycan, BIK, BLAME, BLC / BCA-1 / CXCL13, BMP-15, BMP-2, BMP-3, BMP-3b / GDF-10, BMP-4, BMP-5, BMP-6, BMP-7, BMP-8, BMP-9, BMPR-IA / ALK-3, BMPR-IB / ALK-6, BMPR-II, BMX, BNIP2, BNP, BTC, Btk, C2, C3a, C5/C5a, C7, C8B, C9, CA125, CA15-3, CA19-9, CA9, Cadherin-13, Calbindin, Calbindin D, Calcitonin, Calreticulin, Calsyntenin-1, Cardiotrophin-1 / CT-1, CART, Caspase-3, Caspase-8, Cathepsin B, Cathepsin D, Cathepsin L, Cathepsin S, CBP, CCK, CCL14 / HCC-1 / HCC-3, CCL28 / VIC, CCR1, CCR2, CCR3, CCR4, CCR5, CCR6, CCR7, CCR8, CCR9, CD 163, CD14, CD200, CD23, CD24, CD27 / TNFRSF7, CD30 / TNFRSF8, CD30 Ligand / TNFSF8, CD36, CD38, CD40 / TNFRSF5, CD40 Ligand / TNFSF5 /CD154, CD44, CD45, CD46, CD47, CD55, CD59, CD61, CD71, CD74, CD79 alpha, CD90, CD97, CEA, CEACAM-1, Cerberus 1, Ceruloplasmin, CFHR2, Chem R23, Chemerin, CHI3L1, Chordin-Like 1, Chordin-Like 2, Chromogranin A, Chymase, cIAP-2, Ck beta 8-1, CK-MB, Claudin-3, Claudin-4, CLC, CLEC3B, Clusterin, CNDP1, CNTF R alpha, CNTF, Coagulation Factor III / Tissue Factor, COCO, Complement factor H, Contactin-1, Contactin-2, Corticosteroid-binding globulin, COX-2, C-peptide, CPN2, Creatinine, CRIM 1, Cripto-1, CRP, CRTAM, CRTH-2, Cryptic, CSH1, Csk, CTACK / CCL27, CTGF / CCN2, CTLA-4 /CD152, cTnT / Troponin T, CutA, CV-2 / Crossveinless-2, CXCL14 / BRAK, CXCL16, CXCR1 / IL-8 RA, CXCR2 / IL-8 RB, CXCR3, CXCR4 (fusin), CXCR5 /BLR-1, CXCR6, Cyclin D1, Cystatin A, Cystatin B, Cystatin C, Cytochrome C, Cytokeratin 8, Cytokeratin18, Cytokeratin19, D6, DAN, DANCE, DBI, DCBLD2, DcR3 / TNFRSF6B, D-Dimer, Decorin, DEFA1/3, Defensin, Desmin, Dkk-1, Dkk-3, Dkk-4, DLL1, DLL4, DMP-1, DPPIV, DR3 / TNFRSF25, DR6 / TNFRSF21, Dtk, E-Cadherin, EDA-A2, EDAR, EDG-1, EGF, EGF R / ErbB1, EG-VEGF / PK1, EMAP-II, ENA-78, Endocan, Endoglin / CD105, Endorphin Beta, Endostatin, Endothelin, Endothelin Receptor A, Enolase 2, ENPP2, EN-RAGE, Eotaxin / CCL11, Eotaxin-2 / MPIF-2, Eotaxin-3 / CCL26, EpCAM, EphA1, EphA2, EphA3, EphA4, EphA5, EphA6, EphA7, EphA8, EphB1, EphB2, EphB3, EphB4, EphB6, Epiregulin, ErbB2, ErbB3, ErbB4, ERRa, Erythropoietin R, Erythropoietin, ESAM, E-Selectin, EV15L, EXTL2, FABP1, FABP2, FABP3, FABP4, Factor XIII A, Factor XIII B, FADD, FAK, FAM3B, FAP, Fas / TNFRSF6, Fas Ligand, Fc RIIB/C, Fen 1, FER, Ferritin, Fetuin A, Fetuin B, FGF Basic, FGF R3, FGF R4, FGF R5, FGF-10 / KGF-2, FGF-11, FGF-12, FGF-13 1B, FGF-16, FGF-17, FGF-18, FGF-19, FGF-20, FGF-21, FGF-23, FGF-4, FGF-5, FGF-6, FGF-7 / KGF, FGF-8, FGF-9, FGF-BP, FGFR1, FGFR1 alpha, FGFR2, Fibrinogen, Fibrinopeptide A, Fibronectin, Ficolin-3, FIH, FLRG, Flt-3 Ligand, Follistatin, Follistatin-like 1, FOLR1, FOXN3, FoxO1, FoxP3, Fractalkine, Frizzled-1, Frizzled-3, Frizzled-4, Frizzled-5, Frizzled-6, Frizzled-7, FRK, FSH, Furin, Fyn, GADD45A, Galanin, Galectin-1, Galectin-3, Galectin-3BP, Galectin-7, gamma-Thrombin, Gas1, GASP-1 / WFIKKNRP, GASP-2 / WFIKKN, Gastrin, GATA-3, GATA-4, GCP-2 / CXCL6, GCSF, G-CSF R / CD 114, GDF1, GDF11, GDF-15, GDF3, GDF5, GDF8, GDF9, GDNF, Gelsolin, GFR alpha-1, GFR alpha-2, GFR alpha-3, GFR alpha-4, Ghrelin, GITR / TNFRF18, GITR Ligand / TNFSF18, GLP-1, Glucagon, Glut1, Glut2, Glut3, Glut5, Glypican 3, Glypican 5, GM-CSF, GM-CSF R alpha, GMNN, GPBB, GPI, GPR-39, GPX1, GPX3, Granzyme A, GREMLIN, GRO, GRO-a, Growth Hormone (GH), Growth Hormone R (GHR), GRP, GRP75, GRP78, GSR, GST, HADHA, HAI-1, HAI-2, Haptoglobin, HB-EGF, HCC-4 / CCL16, hCG alpha, hCGb, Hck, HCR / CRAM-A/B, HE4, Hemopexin, Hepassocin, Hepcidin, Heregulin / NDF / GGF / Neuregulin, HGF, HGFR, HOXA10, HRG-alpha, HRG-beta 1, HSP10, HSP20, HSP27, HSP32, HSP40, HSP60, HSP70, HSP90, HSPA8, HTRA2, HVEM / TNFRSF14, I-309, IBSP, ICAM-1, ICAM-2, ICAM-3 (CD50), ICAM-5, IFN-alpha / beta R1, IFN-alpha / beta R2, IFN-beta, IFN-gamma, IFN-gamma R1, IGF2BP1, IGFBP-1, IGFBP-2, IGFBP-3, IGFBP-4, IGFBP-5, IGFBP-6, IGFBP-rp1 / IGFBP-7, IGF-I, IGF-I SR, IGF-II, IGF-II R, IL-1 alpha, IL-1 beta, IL-1 F10 / IL-1HY2, IL-1 F5 / FIL1delta, IL-1 F6 / FIL1 epsilon, IL-1 F7 / FIL1 zeta, IL-1 F8 / FIL1 eta, IL-1 F9 / IL-1 H1, IL-1 R3 / IL-1 R AcP, IL-1 R4 /ST2, IL-1 R6 / IL-1 Rrp2, IL-1 R8, IL-1 R9, IL-1 ra, IL-1 sRI, IL-1 sRII, IL-10, IL-10 R alpha, IL-10 R beta, IL-11, IL-12 p40, IL-12 p70, IL-12 R beta 1, IL-12 R beta 2, IL-13, IL-13 R alpha 1, IL-13 R alpha 2, IL15, IL-15 R alpha, IL-16, IL-
Table S5membrane detecting 1000 proteins (Human Array L-1000, Ray Biotech). A combination of L-507 and L-493 arrays. For the simultaneous detection of the relative expression of 1000 human proteins in cell culture supernatants. Cat# AAH-BLM-1000-4.
Complete detectable protein list for human protein array
IL-15, IL-15 R alpha, IL-16, IL-17, IL-17B, IL-17B R, IL-17C, IL-17D, IL-17E, IL-17F, IL-17R, IL-17RC, IL-17RD, IL-18 BPa, IL-18 R alpha /IL-1 R5, IL-18 R beta /AcPL, IL-19, IL-2, IL-2 R alpha, IL-2 R beta /CD122, IL-2 R gamma, IL-20, IL-20 R alpha, IL-20 R beta, IL-21, IL-21 R, IL-22, IL-22 BP, IL-22 R, IL-23, IL-23 R, IL-23p19, IL-24, IL-26, IL-27, IL-28A, IL-29, IL-3, IL-3 R alpha, IL-31, IL-31 RA, IL-33, IL-34, IL36RN, IL-4, IL-4 R, IL-5, IL-5 R alpha, IL-6, IL-6 R, IL-7, IL-7 R alpha, IL-8, IL-9 , Inhibin A, Inhibin B, INSL3, INSRR, Insulin, Insulin R, Insulysin / IDE, Integrin alpha V, IP-10, I-TAC / CXCL11, Itk, ITM2B, Kallikrein 10, Kallikrein 11, Kallikrein 14, Kallikrein 2, Kallikrein 5, Kallikrein 6, Kallikrein 7, Kallikrein 8, KCC3, KCTD10, KIF3B, Kininostatin / kininogen, KLF4, Kremen-1, Kremen-2, LAG-3, Latent TGF-beta bp1, Layilin, LBP, Lck, LDL R, LECT2, Lefty - A, Legumain, Leptin (OB), Leptin R, LFA-1 alpha, LH, LIF R alpha, LIF, LIGHT / TNFSF14, LIMPII, LIN41, Lipocalin-1, Livin, LOX-1, LPS, LRG1, LRP-1, LRP-6, L-Selectin (CD62L), LTF, LTK, Luciferase, Lumican, Lymphotactin / XCL1, Lymphotoxin beta / TNFSF3, Lymphotoxin beta R / TNFRSF3, Lyn, LYRIC, LYVE-1, LZTS1, MAC-1, Mammaglobin A, Marapsin, MATK, MBL, MBL-2, MCP-1, MCP-2, MCP-3, MCP-4 / CCL13, M-CSF, M-CSF R, MDC, Mer, Mesothelin, MFG-E8, MFRP, MICB, Midkine, MIF, MIG, MINA, MIP 2, MIP-1a, MIP-1b, MIP-1d, MIP-3 alpha, MIP-3 beta, MMP-1, MMP-10, MMP-11 /Stromelysin-3, MMP-12, MMP-13, MMP-14, MMP-15, MMP-16 / MT3-MMP, MMP-19, MMP-2, MMP-20, MMP-24 / MT5-MMP, MMP-25 / MT6-MMP, MMP-3, MMP-7, MMP-8, MMP-9, MSHa, MSP alpha Chain, MSP beta-chain, MTUS1, Musk, Myoglobin, NAIP, Nanog, NAP-2, NCAM-1 / CD56, NELL2, NEP, Nesfatin, Nestin, NET1, Netrin G2, Netrin-4, Neuritin, NeuroD1, Neurokinin-A, Neuropeptide Y, Neuropilin-2, Neurturin, NF1, NGF R, NM23-H1/H2, Notch-1, NOV / CCN3, NPTX1, NPTXR, NR3C3, NRG1 Isoform GGF2, NRG1-alpha / HRG1-alpha, NRG1-beta1 / HRG1-beta1, NRG2, NRG3, NT-3, NT-4, Ntn1, OCT3/4, Omentin, Orexin A, Orexin B, OSM, Osteoactivin / GPNMB, Osteocalcin, Osteocrin, Osteopontin, Osteoprotegerin / TNFRSF11B, OX40, OX40 Ligand / TNFSF4, p21, p27, p53, PAI-1, PAK7, Pancreastatin, Pancreatic Polypeptide, Pappalysin-1, PARC / CCL18, PARK7, P-Cadherin, PCAF, PD-1, PD-ECGF, PDGF R alpha, PDGF R beta, PDGF-AA, PDGF-AB, PDGF-BB, PDGF-C, PDGF-D, PDX-1, PECAM-1 /CD31, PEDF , Pentraxin3 / TSG-14, PEPSINOGEN I, PEPSINOGEN II, Persephin, PF4 / CXCL4, PGRP-S, PI 16, PI 3Kinase p85 beta, PIM2, PKM2, Plasminogen, PlGF, PLUNC, Podocalyxin, POMC, PON1, PON2, PPARg2, PPP2R5C, Pref-1, Presenilin 1, Presenilin 2, Pro-BDNF, Procalcitonin, Pro-Cathepsin B, Progesterone, pro-Glucagon, Progranulin, Prohibitin, Prolactin, Pro-MMP-13, Pro-MMP-7, Pro-MMP-9, ProSAAS, Prostasin, Protein p65, PSA-Free, PSA-total, P-selectin, PSP, PTH, PTHLP, PTN, PTPRD, PYK2, PYY, RAGE, RANK / TNFRSF11A, RANTES, Ras, RBP4, RECK, RELM alpha, RELM beta, RELT / TNFRSF19L, Resistin, RET, RIP1, ROBO4, ROCK1, ROCK2, ROR1, ROR2, ROS, RYK, S100 A8/A9, S100A10, S100A4, S100A6, S100A8, S-100b, SAA, SART1, SART3, SCF, SCF R /CD117, SCG3, SDF-1 / CXCL12, Selenoprotein P, SEMA3A, Serotonin, Serpin 1, Serpin A1, Serpin A12, Serpin A3, Serpin A4, Serpin A5, Serpin A8, Serpin A9, Serpin B5, Serpin D1, Serpin I1, SERTAD2, sFRP-1, sFRP-3, sFRP-4, sgp130, SHBG, SIGIRR, Siglec-5/CD170, Siglec-9, SLPI, SMAC, Smad 1, Smad 4, Smad 5, Smad 7, Smad 8, SMDF / NRG1Isoform, SNCG, Soggy-1, Somatotropin, Sonic Hedgehog (Shh N-terminal), SOST, SOX17, SOX2, SPARC, SPARCL1, Spinesin, SPINK1, SRMS, SSEA-1, SSEA-4, SSTR2, SSTR5, Survivin, SYK, Syndecan-1, Syndecan-3, TACE, TACI / TNFRSF13B, TAF4, Tarc, TCCR / WSX-1, Tec, TECK / CCL25, TFF1, TFF3, TFPI, TGF-alpha, TGF-beta 1, TGF-beta 2, TGF-beta 3, TGF-beta 5, TGF-beta RI / ALK-5, TGF-beta RII, TGF-beta RIIb, TGF-beta RIII, Thrombin, Thrombomodulin, Thrombopoietin (TPO), Thrombospondin (TSP), Thrombospondin-1, Thrombospondin-2, Thrombospondin-4, Thymidine Kinase-1, Thymopoietin, Thyroglobulin, Tie-1, Tie-2, TIM-1, TIMP-1, TIMP-2, TIMP-3, TIMP-4, TL1A / TNFSF15, TLR1, TLR2, TLR3, TLR4, TMEFF1 / Tomoregulin-1, TMEFF2, TNF RI / TNFRSF1A, TNF RII / TNFRSF1B, TNF-alpha, TNF-beta, TNK1, TOPORS, TPA, TPM1, TRA-1-60, TRA-1-81, TRADD, TRAIL R1 / DR4 / TNFRSF10A, TRAIL R2 / DR5 / TNFRSF10B, TRAIL R3 / TNFRSF10C, TRAIL R4 / TNFRSF10D, TRAIL / TNFSF10, TRANCE, Transferrin, Trappin-2, TREM-1, TRKB, Troponin C, Troponin I, TROY / TNFRSF19, TRPC1, TRPC6, TRPM7, Trypsin 1, TSG-6, TSH, TSLP, TSLP, TWEAK / TNFSF12, TWEAK R / TNFRSF12, TXK, Tyk2, TYRO10, Ubiquitin+1, uPA, uPAR, Uromodulin, Vasopressin, Vasorin, VCAM-1 (CD106), VDUP-1, VE-Cadherin, VEGF, VEGF R1, VEGF R2 (KDR), VEGF R3, VEGF-B, VEGF-C, VEGF-D, VEGI / TNFSF15, VGF, VIP Receptor 2, Visfatin, Vitamin D Receptor, Vitamin D-BP, Vitamin K-dependent protein S, Vitronectin, VWF, WIF-1, Wilms Tumor 1, WISP-1 / CCN4, XEDAR, XIAP, ZAG, ZAP70
c.
a.Membrane 1 (L-507) Membrane 2 (L-493)
b.
Figure S2proteins (Human Array L-1000, Ray Biotech). Protein expression was determined for secreted media from a. HS5 in low serum with red boxes indicating positive controls. b. HS5 co-incubated with PER-371 Low in low serum and c. HS5 co-incubated with PER-371 High in low serum. The left hand column depicts membrane 1 (L-507) and the right hand column depicts membrane 2 (L-493).
. Raw data for human protein array membrane detecting 1000
Top Related