THE ROLE OF HUMAN LEUKOCYTE ANTIGEN-G IN HEART TRANSPLANTATION … · 2010-02-08 · 1.2 HEART TRANSPLANTATION AND IMMUNOLOGY Heart Transplantation Heart failure, a condition characterized
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
THE ROLE OF HUMAN LEUKOCYTE ANTIGEN-G IN
HEART TRANSPLANTATION
by
Rohit Sheshgiri, BSc (Hons)
A thesis submitted in conformity with the requirements for the degree of Master of Science,
Graduate Department of the Institute of Medical Science University of Toronto
The Role of Human Leukocyte Antigen-G in Heart Transplantation MSc Thesis, 2008 Rohit Sheshgiri, BSc (Hons) The Institute of Medical Science, University of Toronto
Human leukocyte antigen-G (HLA-G), a protein expressed primarily by fetal trophoblasts,
plays an essential role in maintaining fetal immune tolerance and has previously been
detected following heart transplantation. We sought to establish the value of HLA-G in
identifying freedom from moderate or severe rejection post-heart transplant, and the
capability of its expression in vitro. After assessing myocardial HLA-G expression through
immunohistochemistry, we demonstrated that it was significantly more prevalent in non-
rejecting than rejecting heart transplant recipients. Utilizing vascular endothelial and smooth
muscle cell culture models, we also determined that while HLA-G expression remains tightly
regulated, its expression in vitro can be induced following progesterone treatment in a dose-
dependent manner. Hence, HLA-G may reliably identify patients with a low immunological
risk of developing subsequent clinically significant rejection post-heart transplant.
Furthermore, HLA-G expression can be induced in cultured endothelial and smooth muscle
cells, which might represent a strategy to protect against allograft rejection and vasculopathy.
iii
ACKNOWLEDGEMENT
I will forever remain grateful to my supervisors, colleagues, collaborators and friends who
have been instrumental to the completion of this thesis. Without their help, this body of work
would not have been possible.
I am especially appreciative of my supervisors, Dr. Vivek Rao and Dr. Diego
Delgado, for their mentorship, encouragement and guidance throughout my training program.
Both have been responsible for designing and overseeing this thesis, and have provided me
with the freedom to expand it into a variety of other studies. Most importantly, I was given
the opportunity to explore my own academic interests while undertaking this project. For
these reasons and more, I will never forget the impact they have had on my academic career.
I would like to express my gratitude to Dr. Heather Ross and Dr. Jagdish Butany for
providing me with invaluable advice and expertise during my time with the Heart Transplant
Program. I am also indebted to Laura Tumiati, Dr. Edgardo Carosella, Dr. Nathalie Rouas-
Freiss, Dr. Clifford Librach and Rong Xiao for offering much insight and technical help
along the way.
I am thankful to all my colleagues, especially Danny, Jessica, Mitesh and Elissa for
making my experience a pleasant and memorable one, and to my family for their constant
support and encouragement through the good times and bad. Finally, I thank the Heart and
Stroke/Richard Lewar Centre of Excellence in Cardiovascular Research, Astellas Pharma
Canada and the Institute of Medical Science for supporting our investigations.
iv
TABLE OF CONTENTS
Title Page …i
Abstract …ii
Acknowledgement …iii
Table of Contents …iv
List of Abbreviations …vii
List of Figures and Tables …ix
Chapter 1 KNOWLEDGE TO DATE …1
1.1 Introduction …2
1.2 Heart Transplantation and Immunology …21.2.1 Heart Transplantation …21.2.2 Innate and Adaptive Immunity …5 1.2.3 The Major Histocompatibility Complex …6
1.2.4. Allorecognition Pathways …8
1.3 Human Leukocyte Antigen-G …10 1.3.1 Structure …10 1.3.2 Receptors …11 1.3.3 Inhibition of Natural Killer Cell Function …121.3.4 Modulation of T Cell Function …141.3.5 Inhibition of Antigen-Presenting Cell Function …161.3.6 Pregnancy …181.3.7 Solid Organ Transplantation …191.3.8 Cancer …251.3.9 Inflammation …29
Chapter 2 PROPOSED INVESTIGATIONS …32
2.1 Rationale …33
2.2 Hypotheses …34
v
Chapter 3 MYOCARDIAL HLA-G RELIABLY INDICATES A LOW RISK OF ACUTE CELLULAR REJECTION FOLLOWING HEART TRANSPLANTATION …35
Chapter 4 PROGESTERONE INDUCES EXPRESSION OF HLA-G IN VASCULAR ENDOTHELIAL AND SMOOTH MUSCLE CELLS IN VITRO …44
4.1 Methods …454.1.1 Cell Cultures …454.1.2 Treatments and Interventions …464.1.3 Protein Extraction …474.1.4 Enzyme-Linked Immunosorbent Assays …484.1.5 Protein Determination …494.1.6 Flow Cytometry …504.1.7 Viability Assays …514.1.8 Statistical Analysis …52
4.2 Results …524.2.1 Induction Experiments …524.2.2 The Effect of Progesterone …534.2.3 Inhibition Experiments …534.2.4 Viability Studies …53
Chapter 5 FIGURES AND TABLES …55
Chapter 6 DISCUSSION …70
6.1 Myocardial HLA-G Reliably Indicates a Low Risk of Acute Cellular Rejection Following Heart Transplantation …71
vi
6.2 Progesterone Induces Expression of HLA-G in Vascular Endothelial and Smooth Muscle Cells in vitro …75
6.3 Future Perspectives …80
Chapter 7 REFERENCES …85
vii
LIST OF ABBREVIATIONS
AEC 3-amino-9-ethylcarbazole ANOVA Analysis of variance APC Antigen-presenting cell β2M Beta-2-microglobulin BEC Biliary epithelial cell BNP B-type natriuretic peptide BSA Bovine serum albumin CAV Cardiac allograft vasculopathy CD Cluster of differentiation CRP C-reactive protein CSF Cerebrospinal fluid DMEM Dulbecco's modified Eagle medium DMSO Dimethyl sulfoxide E- Exon EBM-2 Endothelial cell basal medium-2 EDTA Ethylene diamine tetra-acetic acid EGM-2 Endothelial cell growth medium-2 EGM-2 MV Microvascular endothelial cell growth medium-2 ELISA Enzyme-linked immunosorbent assay FasL Fas ligand FBS Fetal bovine serum H/R Hypoxia followed by reperfusion HAEC Human aortic endothelial cell HCAEC Human coronary artery endothelial cell HCASMC Human coronary artery smooth muscle cell HCl Hydrochloric acid HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HLA- Human leukocyte antigen HLA-G Human leukocyte antigen-G HLA-G- Human leukocyte antigen-G-negative HLA-G+ Human leukocyte antigen-G-positive HRP Horseradish peroxidase I- Intron IFN-γ Interferon-γIL Infiltrating leukocyte IL-10 Interleukin-10 ILT- Immunoglobulin-like transcript ISHLT International Society for Heart and Lung Transplantation IVF In vitro fertilization KIR2DL4 Killer cell immunoglobulin-like receptor 2DL4 mAb Monoclonal antibody MC Myocardial cell MHC Major histocompatibility complex
viii
MHC-I Major histocompatibility complex class I MHC-I b Major histocompatibility complex class Ib MHC-II Major histocompatibility complex class II MICA Major histocompatibility complex class I-related chain A MLR Mixed lymphocyte reaction MS Multiple sclerosis N/A Not applicable NK Natural Killer PBMC Peripheral blood mononuclear cell PBS Phosphate buffered saline PBSG/BSA Phosphate buffered saline with glucose and bovine serum albumin PE Preeclampsia PIR-B Paired immunoglobulin-like inhibitory receptor-B PRA Panel reactive antibodies PRE Progesterone response element RCC Renal cell carcinoma RIPA Radio-immunoprecipitation assay sHLA-G Soluble human leukocyte antigen-G SMA Smooth muscle actin SmBM-2 Smooth muscle basal medium-2 SmGM-2 Smooth muscle growth medium-2 Treg Regulatory T VCAM-1 Vascular cell adhesion molecular-1 XTT Sodium 3’-[1-(phenylaminocarbonyl)-3,4-tetrazolium]-bis(4-methoxy-
6-nitro)-benzene sulfonic acid hydrate
ix
LIST OF FIGURES AND TABLES
Figure 1. Protein structures of the HLA-G isoforms …56
Figure 2. Structures of HLA-G mRNA transcripts …57
Figure 3. Immunohistochemical staining of MHC-I proteins in myocardial biopsies of heart transplant recipients …58
Figure 4. HLA-G expression profile in non-rejecting and rejecting patients …59
Figure 5. Characterization of cultured endothelial and smooth muscle cells through flow cytometric analysis …60
Figure 6. Assessment of HLA-G expression in endothelial and smooth muscle cell cultures following progesterone treatment …61
Figure 7. Flow cytometric analysis of HLA-G expression in cultured endothelial and smooth muscle cells following progesterone treatment …62
Figure 8. Time course experiments assessing progesterone treatment on HLA-G expression …63
Figure 9. The effect of mifepristone, a progesterone receptor antagonist, on progesterone-induced HLA-G expression in cultured endothelial and smooth muscle cells …64
Table 1. Baseline characteristics of non-rejecting and rejecting heart transplant patients …65
Table 2. Expression of HLA-G in non-rejecting and rejecting patients …66
Table 3. Expression of HLA-G in cultured endothelial and smooth muscle cells following exposure to interventions of interest …67
Table 4. Viability and proliferation ability of treated endothelial and smooth muscle cell cultures …68
Table 5. Viability and proliferation ability of treated endothelial and smooth muscle cell cultures following H/R stress …69
1
CHAPTER 1:
KNOWLEDGE TO DATE
2
1.1 INTRODUCTION
In 1986, Ellis et al. described a novel major histocompatibility complex (MHC) molecule
isolated from cytotrophoblast cell membranes (1), which was subsequently cloned the
following year as reported by Geraghty et al. (2). This protein, termed human leukocyte
antigen-G (HLA-G), was mapped to the short arm of chromosome 6 (3) and was initially
believed to be restricted to cytotrophoblast cells of the fetus during early gestation, based on
the results of pioneering studies (4-6). However, in recent years HLA-G has been shown to
be expressed during a number of physiological conditions, both pathological and non-
pathological. Considered a non-classical MHC molecule because of its role in immune
suppression rather than immune activation, its functions are now known to be much more
complex than initially believed and have been extended beyond the scope of pregnancy into a
variety of other milieus including transplantation, cancer and inflammation. This thesis
examines the role of HLA-G in the context of heart transplantation.
1.2 HEART TRANSPLANTATION AND IMMUNOLOGY
Heart Transplantation
Heart failure, a condition characterized by impaired cardiac function leading to insufficient
blood supply to the tissues, is increasing in prevalence worldwide and results in high rates of
morbidity and mortality while commanding tremendous human and economic resources (7-
9). Presently, heart transplantation remains the definitive treatment of choice for patients
suffering from end-stage congestive heart failure failing maximal medical therapy (10). Since
the first heart transplant operation performed by Christiaan Barnard in 1967 (11), over
3
76,000 procedures reported to the registry of the International Society for Heart and Lung
Transplantation (ISHLT) and tens of thousands of unreported cases have been performed
worldwide (12).
The vast majority of heart transplant operations have been undertaken due to existing
coronary artery disease or cardiomyopathy (12). While transplantation provides a reliable
therapeutic option for these conditions, advances in the cardiovascular field have introduced
a new set of problems. Improvements in patient outcomes have continued to expand the pool
of eligible recipients, resulting in longer waiting times and pre-operative mortalities (13) due
to the limited availability of suitable donor organs (14). To help combat this problem,
suboptimal or marginal hearts may be used, albeit at the expense of long-term cardiac
function and patient survival (15,16). Improvements in early post-transplant graft and patient
survival have also introduced a new set of late-onset complications (17).
Survival following heart transplantation is limited by several factors. While graft
failure and infection are primarily responsible for early death, malignancy and cardiac
allograft vasculopathy (CAV) are the leading causes of late morbidity and mortality (12). The
effects of lifelong immunosuppressive therapy in conjunction with the aging patient
population are widely regarded to be responsible for the increased risk of neoplastic
transformation post-cardiac transplant (18-21). It is also known that patients with a prior
history of cancer have an increased risk of developing post-transplant malignancies (20). Yet,
it is still unclear which components of an immunosuppressive regimen increase the risk of
developing the solid and hematologic malignancies seen following cardiac transplantation.
Allograft vasculopathy, characterized by a progressive, diffuse and concentric
thickening of the arterial intima, is another major post-transplant complication that affects
4
long-term graft survival (22-24). CAV is a multifactorial process that is the consequence of
immunologic factors such as acute rejection episodes, HLA-mismatches and anti-HLA
antibodies, as well as non-immunologic risk factors including donor and recipient-related
USA) at an absorbance wavelength of 750 nm. Absorbance readings from blank wells were
subtracted from the values of the samples and standards. Total protein concentrations of
50
samples were calculated by plotting the corrected absorbances on the linear protein standard
curve.
Flow Cytometry
To assess cell surface protein expression, cells cultured in 6 cm dishes were washed with
PBS and detached by treating with 1 mL Accutase cell detachment medium (eBioscience,
Inc., San Diego, CA, USA) at 37 °C for 5 min. An equal number of cells from each 6 cm dish
were transferred into 2 polypropylene round-bottom tubes (BD Biosciences, San Jose, CA,
USA), to be assessed for protein expression or used as isotype controls. Non-specific binding
was controlled by decanting the supernatant, adding 50 µL of PBS with 20mMol glucose
(Fisher) and 5% BSA (PBSG/BSA) with 40% mouse serum (Sigma) and incubating at 4 °C
for 10 min. For determination of protein expression, 50 µL PBSG/BSA with a primary
antibody was added to the cell suspension and incubated for 45 minutes in the dark at 4 °C. If
necessary, cells were washed once with PBSG/BSA and incubated in the dark for 30 minutes
at 4 °C in 100 µL PBSG/BSA containing a secondary antibody. After two washing steps,
cells were resuspended in 200 µL PBSG/BSA and cell-surface HLA-G expression was
assessed using a Coulter Epics XL-MCL flow cytometer (Beckman Coulter). IsoFlow sheath
fluid (Beckman Coulter) was used to top up sample volumes in the round-bottom tubes. To
assess HLA-G expression we used a primary mouse anti-human HLA-G antibody, MEM-G/9
(AbD Serotec, Planegg, Germany), at 10 µg/mL. Cells serving as negative controls were
incubated with isotype-matched mouse IgG (AbD Serotec) at the same antibody
concentration in similar fashion. A goat anti-mouse IgG conjugated to Alexa Fluor
(Invitrogen) at 2.5 µg/mL was used as the secondary antibody. For phenotypic
characterization of HCAEC and HAEC cultures, cells were incubated with the primary
51
mouse anti-human CD31 antibody conjugated to phycoerythrin and compared to a
phycoerythrin-conjugated isotype matched mouse IgG (Beckman Coulter) at the same
concentration, without the need for a secondary antibody. For smooth muscle cell
characterization, cells were assessed for intracellular α-smooth muscle actin (SMA)
expression with a cyanine 3-primary conjugated mouse anti-SMA (Sigma) antibody. To
enable intracellular staining, cells were first simultaneously fixed and permeabilized by
resuspending in 100 µL Cytofix/Cytoperm solution (BD Biosciences) and incubating at 4 °C
for 20 min. Cells were washed and stained as previously described. For characterization of
smooth muscle cells, the Perm/Wash buffer (BD Biosciences) was used for all washing and
incubation steps to keep the HCASMC permeabilized.
Viability Assays
The XTT (sodium 3’-[1-(phenylaminocarbonyl)-3,4-tetrazolium]-bis(4-methoxy-6-nitro)-
benzene sulfonic acid hydrate) based Cell Proliferation Kit II (Roche Diagnostics
Corporation, Basel, Switzerland) was used for the quantification of cell proliferation and
viability. XTT labeling mixture was first prepared by mixing 2% electron coupling reagent in
XTT labeling reagent. Cells were then incubated for 4 h in 750 µL phenol red-free
Dulbecco's Modified Eagle Medium (DMEM, Invitrogen)supplemented with 5% FBS and
20% XTT labeling mixture per well. Cell viability was determined by the ability of
metabolically active cells to convert the yellow XTT salt to an orange formazan dye. Sample
absorbance was measured by an ELISA plate reader at 450 nm with a reference wavelength
of 650 nm. Cell viability was calculated as a percentage of the absorbance of control wells.
52
Statistical Analysis
ELISA, protein determination and viability experiments were all performed in duplicate. The
mean of the two readings was used as the absorbance of a sample or protein standard. Data
are represented as mean ± standard error. Comparisons between multiple groups were
performed by two-factor ANOVA. We used unpaired t tests for inhibition experiments
involving comparisons between two groups. We considered p < 0.05 to represent statistical
significance.
4.2 RESULTS
Induction Experiments
Phenotypic characterization of endothelial and smooth muscle cells was confirmed through
flow cytometric analysis. HCAEC and HAEC showed cell-surface expression of CD31,
while HCASMC cultures stained strongly for intracellular α-SMA (Figure 5). To determine
HLA-G expression in HCAEC, HAEC and HCASMC cultures, we subjected them to our
interventions of interest and assessed protein expression via ELISA and flow cytometry
(Table 3). Dose response and time course experiments were performed in all cell types for all
treatments. HCAEC, HAEC and HCASMC cultures did not express HLA-G at baseline.
Expression of HLA-G was not detected following treatment with cytokines (IFN-γ and IL-
10) and immunosuppressive agents (cyclosporine, sirolimus and tacrolimus), at all doses and
time points. We also did not detect HLA-G expression following exposure to different
periods of H/R. However, HLA-G expression was induced in progesterone-treated HCAEC,
HAEC and HCASMC cultures.
53
The Effect of Progesterone
Because HLA-G expression was only induced after incubation with progesterone, cells were
treated with varying progesterone doses for 24 h. Total HLA-G protein expression increased
in a dose-dependent manner (p < 0.001) as detected by ELISA (Figure 6), while cell surface
expression was confirmed by flow cytometry (Figure 7) in HCAEC, HAEC and HCASMC
cultures. Time course experiments showed maximal HLA-G expression within 12 h in all cell
cultures (p < 0.001) (Figure 8).
Inhibition Experiments
To determine if progesterone-induced HLA-G expression could be inhibited by the
progesterone receptor antagonist, mifepristone, cultured cells were incubated with
progesterone (10,000 ng/mL) in the presence or absence of mifepristone (1000 ng/mL) for 24
h. As measured by ELISA (Figure 9), HLA-G expression in HCAEC, HAEC and HCASMC
cultures was partially blocked by mifepristone (p < 0.05), indicating that the mechanism of
progesterone-induced HLA-G expression was through receptor binding.
Viability Studies
Cells were assessed for proliferation ability and viability following 24 h exposure to
progesterone with or without mifepristone to ensure that differences in HLA-G expression
were not the result of cellular injury. There were no significant differences in HCAEC,
HAEC or HCASMC viability following incubation with our interventions compared to
vehicle, indicating that these cultures were not injured by progesterone or mifepristone
(Table 4). In order to assess whether HLA-G expression might confer protection against H/R
injury, cells were subjected to 12 h hypoxia and 2 h reperfusion following 24 h progesterone
54
treatment. Viability studies demonstrated that HCAEC, HAEC and HCASMC cultures were
not protected against H/R injury following progesterone-induced HLA-G expression (Table
5). Relative to vehicle-treated cells under normoxia, there were no significant differences in
proliferation ability and viability following H/R stress in progesterone-treated cells compared
to vehicle-treated cells.
55
CHAPTER 5:
FIGURES AND TABLES
56
Figure 1. Protein structures of the HLA-G isoforms. To date, seven HLA-G variants have been reported. The transmembrane domain common to HLA-G1 to -G4 enable membrane anchoring. The HLA-G5 to -G7 structures are the soluble counterparts of HLA-G1 to -G3, respectively. HLA-G1 and -G5 possess all globular domains and non-covalently associate with β2M. Adapted from Carosella et al. (46).
57
Figure 2. Structures of HLA-G mRNA transcripts. HLA-G protein variants are the result of alternate splicing of the primary mRNA transcript. Exon 1 (E1) encodes the leader peptide, while exons 2, 3 and 4 encode the α1, α2 and α3 globular domains, respectively, of the heavy chain. Unlike the α1 domain which is common to all isoforms, α2 and α3 are absent in some HLA-G proteins. Exon 5 encodes the transmembrane region for the cell surface isoforms, HLA-G1 to -G4. Translation of these proteins terminates in exon 6 which encodes a shortened cytoplasmic domain. Intron 4 (I4) is retained in the mature mRNA transcripts of HLA-G5 and -G6, giving rise to a tail 21 amino acids in length. Similarly, the -G7 isoform has a short tail 2 amino acids long due to the presence of intron 2 in the mature transcript. The HLA-G5 to -G7 isoforms lack the transmembrane domain because of stop codons located within introns 2 and 4 and, thus, do not associate with the cell membrane. Adapted from Carosella et al. (46).
58
Figure 3. Immunohistochemical staining of MHC-I proteins in myocardial biopsies of heart transplant recipients. Expression of HLA-B and/or HLA-C in allograft myocardial cells (MC) and infiltrating leukocytes (IL) during Grade 0 rejection (a) or Grade ≥ 2R rejection (b) was revealed by the HC10 mAb to ensure tissue integrity. The MEM-G/2 mAb enabled detection of HLA-G+ (c) and HLA-G- (d) specimens, respectively, based on the presence or absence of red chromogen. HLA-G+ and HLA-G- biopsies were also identified by the staining (e) or absence of staining (f) with the 4H84 mAb, respectively.
59
Figure 4. HLA-G expression profile in non-rejecting and rejecting patients. Group A: patients without clinically significant rejection. Group B: patients with sustained moderate to severe rejection. The second biopsies of Group B patients were obtained during an episode of Grade ≥ 2R rejection. All other biopsy scores in both groups of patients were of rejection Grade ≤ 1R.
60
Figure 5. Characterization of cultured endothelial and smooth muscle cells through flow cytometric analysis. Cell surface protein expression of CD31 by HCAEC (a) and HAEC (b) confirmed endothelial phenotype. Intracellular α-smooth muscle actin (α-SMA) staining by HCASMC (c) confirmed smooth muscle lineage. More than 90% of cells stained positive for their respective markers.
61
Figure 6. Assessment of HLA-G expression in endothelial and smooth muscle cell cultures following progesterone treatment. HCAEC, HAEC and HCASMC were incubated for 24 h at 100-10,000 ng/ml. Treatment with vehicle alone failed to induce HLA-G expression in all cell types. HLA-G protein expression, determined by ELISA, increased with incremental treatment doses of progesterone in all cell types (p < 0.001). HLA-G concentrations were normalized for total cellular protein content.
62
Figure 7. Flow cytometric analysis of HLA-G expression in cultured endothelial and smooth muscle cells following progesterone treatment. HCAEC (a), HAEC (b) and HCASMC (c) were incubated with low (100 ng/mL) or high (10,000 ng/mL) doses for 24 h. Cell surface expression of HLA-G was not detected in vehicle-treated cells, but was present in all cultures following progesterone treatment.
63
Figure 8. Time course experiments assessing progesterone treatment on HLA-G expression. HCAEC (a), HAEC (b) and HCASMC cultures (c) showed maximal HLA-G expression within 6-12 hours of exposure following 10,000 ng/mL (p < 0.001). HLA-G levels were measured by ELISA and normalized for total cellular protein.
64
Figure 9. The effect of mifepristone, a progesterone receptor antagonist, on progesterone-induced HLA-G expression in cultured endothelial and smooth muscle cells. Compared to cells treated with 10,000 ng/mL progesterone alone (controls), HCAEC, HAEC and HCASMC incubated with 1000 ng/mL mifepristone in combination with 10,000 ng/mL progesterone had significantly reduced HLA-G expression. HLA-G levels were determined by ELISA and normalized for total protein content. All cultures were incubated for 24 h with treatment. *p < 0.05 compared to control.
65
Table 1. Baseline characteristics of non-rejecting and rejecting heart transplant patients.
Group A (n = 29)
Group B (n = 38) P Value
Demographics
Gender (male/female) 24/5 31/7 0.90
Mean Age (range) 51±12 (25-68) 47±13 (18-69) 0.28
Reason For Transplantation
Idiopathic Cardiomyopathy 13 (45) 20 (53) 0.53
Ischemic Cardiomyopathy 11 (38) 12 (32) 0.59
Other 5 (17) 6 (16) 0.87
Immunosuppressive Therapy
Cyclosporine 23 (79) 29 (76) 0.77
Tacrolimus 8 (28) 14 (37) 0.42
Mycophenolate Mofetil 25 (86) 36 (95) 0.23
Sirolimus 3 (10) 15 (39) 0.01
Prednisone 28 (97) 34 (89) 0.27
Donor Characteristics
Mean Age (range) 38±16 (15-67) 35±12 (15-56) 0.35
Gender (male/female) 20/9 26/12 0.96
Gender Mismatches 10 (34) 11 (29) 0.63
Pre-transplant PRA
0% 20 (69) 30 (79) 0.35
1-19% 6 (21) 4 (11) 0.25
20-100% 3 (10) 4 (11) 0.98
Unless specified, values in parentheses denote percentages. Non-rejecting and rejecting patients were separated into Groups A and B, respectively. PRA: Panel Reactive Antibodies.
66
Table 2. Expression of HLA-G in non-rejecting and rejecting patients.
Group A Group B P Value
HLA-G+ 25 (86) 4 (11) Patients
HLA-G- 4 (14) 34 (89) < 0.001
HLA-G+ 46 (53) 4 (4) Biopsies
HLA-G- 41 (47) 110 (96) < 0.001
Values in parentheses denote percentages. Non-rejecting and rejecting patients were separated into Groups A and B, respectively.
67
Table 3. Expression of HLA-G in cultured endothelial and smooth muscle cells following exposure to interventions of interest.
Intervention Dose Time HLA-G
IFN-γ 0.1-100 ng/mL 2-24 hr -
IL-10 0.1-100 ng/mL 2-24 hr -
H/R N/A 2-12 hr/2 hr -
Cyclosporine 100-1000 ng/mL 2-24 h -
Sirolimus 0.1-100 ng/mL 2-24 h -
Tacrolimus 0.1-100 ng/mL 2-24 h -
Progesterone 100-10,000 ng/mL 2-24 hr +
Plus signs indicate the presence of HLA-G following treatment in HCAEC, HAEC and HCASMC cultures. Minus signs indicate undetectable HLA-G expression that fell below the lower detection limit of the ELISA. N/A: not applicable.
68
Table 4. Viability and proliferation ability of treated endothelial and smooth muscle cell cultures.
Progesterone/Mifepristone Dose (ng/mL)
0/0 100/0 1000/0 10,000/0 10,000/1000
HCAEC 100 ± 5 102 ± 6 94 ± 9 98 ± 4 93 ± 8
HAEC 100 ± 6 100 ± 10 98 ± 7 96 ± 10 93 ± 9
HCASMC 100 ± 6 97 ± 5 104 ± 5 93 ± 5 96 ± 5
Data are represented as percentages of vehicle-treated cells.
69
Table 5. Viability and proliferation ability of treated endothelial and smooth muscle cell cultures following H/R stress.
Progesterone/Mifepristone Dose (ng/mL)
0/0 100/0 1000/0 10,000/0 10,000/1000
HCAEC 70 ± 3 73 ± 7 74 ± 7 74 ± 4 74 ± 5
HAEC 65 ± 7 70 ± 5 67 ± 6 67 ± 5 68 ± 5
HCASMC 82 ± 5 82 ± 4 84 ± 6 84 ± 4 87 ± 4
Data are represented as percentages of vehicle-treated cells under normoxic conditions.
70
CHAPTER 6:
DISCUSSION
71
6.1 MYOCARDIAL HLA-G RELIABLY INDICATES A LOW RISK OF ACUTE CELLULAR REJECTION FOLLOWING HEART TRANSPLANTATION
Currently, allograft rejection post-heart transplant is best monitored through myocardial
biopsies (170). However, assessment of rejection by this procedure is expensive, invasive
and subjective (171). For these reasons, it is important to develop non-invasive techniques to
monitor allograft status. Although molecules such as BNP, CRP and cardiac troponins have
been described as potential biological markers post-transplant, they indicate cardiac damage
rather than rejection. Therefore, our retrospective investigation was undertaken to assess the
value of HLA-G expression in determining freedom from clinically important rejection in
heart transplant patients. This study demonstrates that myocardial HLA-G expression
following heart transplantation is more prevalent in patients who exhibit lower rejection
scores. Both, the frequency of HLA-G expression and the proportion of HLA-G+ patients
were significantly greater in Group A (non-rejecting) versus Group B (rejecting), indicating
that HLA-G expression portrays a low risk of cardiac allograft rejection. Thus, myocardial
HLA-G expression post-transplant might reliably aid in identifying patients with a low
immunological risk of developing subsequent clinically relevant moderate or severe acute
cellular rejection episodes.
Our results are comparable to observations from other studies which demonstrate that
HLA-G expression is associated with fewer acute rejection episodes in the first post-
operative year, and absence of chronic rejection as assessed by coronary angiography in heart
transplant patients (104,105). However, using a case-control design we have illustrated for
the first time differential expression of HLA-G in two distinct populations of heart transplant
recipients, whereby non-rejecting patients were found to have significantly increased
72
myocardial HLA-G expression compared to rejecting patients post-transplant. Additionally,
we show for the first time that HLA-G expression is inversely associated with moderate and
severe rejection episodes after the first post-operative year, as Grade ≥ 2R rejection was
undetected in Group A patients up to 9 years post-transplant. Interestingly, unlike previous
reports of unvarying HLA-G status (104,105), HLA-G appears to have a dynamic pattern of
expression in our study population, a finding which is consistent with fluctuations in post-
transplant soluble HLA-G levels observed in other studies (107,108). While there was no
direct association between HLA-G status and rejection score at the time of a myocardial
biopsy, detection of an HLA-G+ biopsy appears to reliably indicate low risk of developing
moderate or severe graft rejection post-transplant, since 86% of recipients with no Grade ≥
2R rejection were HLA-G+. Nonetheless, prospective investigations involving larger patient
cohorts are required to investigate the utility and variability of HLA-G expression post-heart
transplant.
In addition to determining HLA-G status of heart transplant patients through
immunohistochemical staining of biopsies obtained from allografts, measuring sHLA-G in
these recipients might represent another potential strategy to monitor the degree of allograft
rejection. High sHLA-G concentrations in heart transplant patients appear to be associated
with reduced incidences of acute and chronic rejection (106-108). Similar phenomena have
been reported across the solid organ transplantation milieu. Soluble HLA-G levels in liver
and/or kidney transplant patients have a negative relationship with rejection and graft failure
(110,117,119). Moreover, post-transplant sHLA-G is inversely correlated with anti-HLA IgG
antibody production (110), possibly suggesting protection from humoral rejection. Since
HLA antibodies are linked to poor patient outcomes and rejection episodes post-cardiac
73
transplant (172), sHLA-G might possess further clinical relevance. Hence, quantifying
sHLA-G levels in conjunction with myocardial HLA-G staining might improve stratification
of patients according to immunological risk.
The use of ISHLT rejection scores represents a major limitation of this study. As all
biopsies were graded by one cardiac pathologist, rejection scores are subjective and may not
have been graded similarly by other pathologists. Furthermore, because the biopsies were not
graded concurrently, rejection scores are susceptible to intra-observer variability. Patient
clinical information may have also influenced the rejection grading. Another limitation of
this investigation is the use of retrospective biopsy samples. Incorporating such a procedure
into clinical practice might be challenging due to tissue degradation and different biopsy
storage approaches across heart transplant centers. Several patients were excluded from our
study as a result of poor tissue quality. Furthermore, tissue specimens show diminished
immunoreactivity over time. Even with the reassurance that the biopsies stained positively
for HLA-B and/or HLA-C, HLA-G might degrade differently and harbour false negatives.
The retrospective nature of this study also did not allow for quantification of soluble HLA-G;
however, previous studies have shown a significant correlation between myocardial and
soluble HLA-G expression (104,105). Nevertheless, to effectively utilize the HLA-G status
of patients in the clinical setting as a prognostic indicator of immunological risk, tissue
samples used for HLA-G measurement should be obtained prospectively during routine
myocardial biopsy procedures and employed with soluble HLA-G levels to help predict the
rejection profile of heart transplant recipients. It is important to note that HLA-G testing
should be performed repeatedly due to the risk of false negative staining.
74
Although our investigation in conjunction with other studies have demonstrated the
beneficial nature of HLA-G expression as it applies to protection from allograft rejection, it
remains unknown whether myocardial HLA-G expression following heart transplantation is
donor- or recipient-specific, why this expression is only presented by certain patients and
how it is induced. Nevertheless, both genetic and non-genetic factors are believed to be
responsible for this phenomenon. Distinct HLA-G alleles have been associated with
variations in HLA-G splicing patterns and mRNA levels (173). Additionally, different HLA-
G alleles have been shown to differentially influence soluble HLA-G concentrations (109),
which raises the possibility for differential HLA-G expression post-transplant based on
genotype. Immunosuppressive therapy has also been suggested to increase sHLA-G levels in
some heart transplant patients soon after administration (107,108), which might partially
explain the presence or absence of HLA-G expression in certain patients at different stages
post-transplant. With the exception of sirolimus, however, we did not detect significant
differences in immunosuppressive regimens between the two patient cohorts. It is unclear
whether sirolimus influenced our results because of the limited sample size. Given that HLA-
G has not been detected in healthy cardiac tissue and that sHLA-G levels are generally higher
in heart transplant patients, stressors in the peri- and/or post-operative period may represent
another likely explanation for HLA-G expression. This situation is not unexpected
considering mechanical and pathological stress may induce re-expression of cardiac fetal
genes (174-180). Ultimately, HLA-G expression in certain heart transplant patients is likely
the result of numerous non-genetic peri- and post-transplant factors affecting those who are
genetically-susceptible.
75
Is summary, our retrospective study illustrates an applicable inverse relationship
between myocardial HLA-G expression and clinically relevant cardiac allograft rejection,
suggesting that myocardial HLA-G staining indicates an improved tolerance profile post-
transplant. Yet, while HLA-G status may prove clinically useful as a prognostic indicator of
immunological risk, the ultimate goal of research in the field of organ transplantation is to
inhibit the initiation and progression of allograft rejection in transplant patients, without the
administration of lifelong immunosuppression to thereby uphold viable immune responses
against other pathogens. Therefore, it is important to scrutinize the potential strategies that
can hinder the rejection process, such as elucidating the mechanisms that contribute to HLA-
G expression following organ recovery, storage or transplantation. In heart transplant
patients, CAV (22-24) due in part to immune-mediated factors, and malignancy (18-21) due
to constant immunosuppression, represent the most prevalent long-term complications.
Induction of HLA-G expression in transplant recipients could represent a possible therapeutic
strategy to limit allograft rejection and reduce immunosuppressive medications along with
their associated negative side effects, and thus potentially improve long-term outcomes by
protecting against CAV and malignancy.
6.2 PROGESTERONE INDUCES EXPRESSION OF HLA-G IN VASCULAR ENDOTHELIAL AND SMOOTH MUSCLE CELLS IN VITRO
Following the results of our retrospective investigation which suggested that HLA-G
expression might reliably indicate freedom from clinically relevant rejection post-heart
transplant, we assessed whether vascular endothelial and smooth muscle cultures were
capable of in vitro HLA-G expression, as these could represent potential targets for future
76
therapies aimed at protecting against allograft vasculopathy. While HLA-G has been detected
in endothelial cells of chorionic fetal vessels during embryonic development, this expression
is lost in endothelial cells lining mature vessels (181,182). Our experiments, for the first time,
illustrate in vitro HLA-G expression in adult HCAEC, HAEC and HCASMC cultures.
Interestingly, increasing HLA-G levels were detected following treatment with incremental
progesterone doses in all cell types, without any changes in cellular proliferation ability or
viability; yet no expression was found after exposure to our other interventions including
cytokines, H/R stress or immunosuppressive agents, which may be present in the peri- or
post-transplant milieu, and have been shown experimentally to induce or upregulate HLA-G
mRNA and/or protein expression.
Treatment of normal blood monocytes, macrophage cell lines and constitutive HLA-
G-expressing cell lines (183-186) with IFN-γ has been shown to raise HLA-G mRNA levels,
leading to increased intracellular and cell-surface protein expression in vitro. This effect is
not unexpected considering IFN-γ and IFN-γ receptors are synthesized in first trimester
human trophoblasts (187,188), which coincide with HLA-G expression. Additionally, IL-10
has been shown to induce HLA-G mRNA and protein expression in peripheral blood
monocytes, trophoblasts and cells lines which constitutively express HLA-G (189,190).
Recent evidence also suggests that HLA-G expression in PBMC may be the result of an IL-
10 autocrine feedback loop (191), consistent with the fact that human cytotrophoblasts
produce and secrete IL-10 (192). However, treatment with IFN-γ or IL-10 did not induce
HLA-G expression in our studies, indicating that HLA-G expression appears to be tightly
regulated and specific for cell type.
77
We also examined the effect of H/R on HLA-G expression in our cell culture models,
because allografts are subjected to ischemia-reperfusion injury in the peri-operative period. It
has been shown that soluble and membrane-bound HLA-G mRNA expression is inversely
related to oxygen concentrations in primary cultures of extravillous cytotrophoblasts (193).
Furthermore, hypoxia has been demonstrated to differentially affect HLA-G expression in
various tumor cell lines by inducing HLA-G gene transcription in some HLA-G- lines and
decreasing constitutive expression in certain HLA-G+ lines, indicating hypoxic injury is
capable of modulating HLA-G expression in different cell types (194). Interestingly, the
HLA-G promoter contains a heat shock element which binds to heat shock factor 1 (195), a
transcriptional factor activated during conditions of environmental stress (196), thus
providing more evidence that HLA-G expression might be stress-inducible. For this reason,
we mimicked the ischemic and reperfusion times commonly observed in clinical
transplantation by exposing our vascular and smooth muscle cell cultures to H/R injury, but
were unable to detect any expression following this intervention.
Differential patterns of HLA-G expression following heart transplantation have been
noted in patients from our retrospective study and other investigations (104,105). Yet it
remains unknown why HLA-G is detected in only some patients post-transplant. While
genetic predisposition of the donor allograft and/or recipient represents a likely explanation
(109), immunosuppressive therapy might also play a role since different regimes are tailored
for different patients. We therefore assessed HLA-G expression after treatment with
cyclosporine, sirolimus or tacrolimus, as recent studies have implicated immunosuppressive
agents with increased serum HLA-G concentrations following heart transplantation
(107,108). After treating cell cultures with clinically relevant doses of these interventions, we
78
were unable to detect HLA-G protein expression, leading us to believe that HLA-G
expression in response to immunosuppressive therapy might be specific for cell type.
Finally, we determined the effects of progesterone and its receptor antagonist,
mifepristone, on HLA-G expression. Progesterone has been recently demonstrated to
enhance HLA-G expression in the JEG-3 choriocarcinoma cell line and isolated first
trimester cytotrophoblasts (197) through receptor activation followed by binding to a
progesterone response element (PRE), which shares 60% homology with the wild-type
mouse mammary tumor virus PRE sequence (198). Furthermore, progesterone receptors are
present in human heart as well as vascular endothelial and smooth muscle cells (199,200),
indicating that these tissues are possible targets for progesterone-induced HLA-G expression.
Intriguingly, our results show induction of this expression following treatment with
progesterone, and partial inhibition after co-incubation with mifepristone in HCAEC, HAEC
and HCASMC cultures, suggesting that the effect of progesterone was the result of
progesterone receptor activation. Complete inhibition was not noticed, however, since
mifepristone is a competitive inhibitor of progesterone. Following this finding, we assessed
the effect of HLA-G expression on protection from H/R injury. However, while HLA-G
upregulation might represent a protective response against hypoxic injury in tumor cells
(194), we found that progesterone-induced HLA-G expression did not protect against H/R
injury in our cell cultures.
Induction of HLA-G expression in endothelial cells is of particular interest as they are
primary targets of circulating T cells post-transplant, due to expression of classical MHC-I
and II antigens (161). After binding to classical MHC proteins, T cells secrete a host of
cytokines, leading to recruitment of inflammatory cells and proliferation of smooth muscle
79
cells, which can eventually progress into CAV (201,202). Endothelial HLA-G expression
following progesterone treatment could represent a strategy to inhibit adjacent T cells and
prevent the progression of CAV. Intriguingly, progesterone appears to have numerous other
beneficial effects on the vasculature such as suppressing endothelial VCAM-1 mRNA and
protein expression (203), which has been implicated in leukocyte adhesion (204), and
inhibiting proliferation of endothelial (205) and smooth muscle cells (206), all of which
contribute to atherosclerosis (206,207) and CAV (201,202).
Vasculopathy remains a primary complication post-heart transplant and represents a
major cause of late morbidity and mortality (22-24). Although it is the result of immune and
non-immune factors (25), the fact that transplant patients require lifelong
immunosuppression indicates that the immune system constantly directs responses to the
allograft likely through the indirect (40) and semi-direct (43) allorecognition pathways and,
over time, contributes heavily to CAV development. Hence, induction of HLA-G expression
in these patients might prove clinically relevant due to its inhibitory effects on NK cells (70-
75,77,79-81), T cells (65,66,82-89) and APC (92-96), and its ability to protect against CAV
(104-106). Expression of HLA-G post-transplant might also reduce the requirement for
immunosuppressive agents and thus lessen their associated negative side effects such as
malignancy, the leading complication post-heart transplant (18-21).
The beneficial effects of post-transplant soluble and membrane-bound HLA-G
expression are evident throughout the solid organ transplant milieu, including heart (104-
108), kidney (110,113), liver (88,119) and combined kidney-liver (87,88,114,117).
Investigations involving murine models also demonstrate longer skin allograft survival in the
presence of transgenic or recombinant HLA-G (92-96). Yet, the mechanisms of post-
80
transplant HLA-G expression remain vague. Peri- and post-transplant interventions are likely
to be responsible for this phenomenon because HLA-G is not detected in healthy organ
tissue. Therefore, our investigations were designed to assess whether factors such as
cytokines, H/R stress or immunosuppressive agents resulted in HLA-G expression in vitro.
While we did not detect HLA-G expression in response to these interventions, HLA-G was
detected after progesterone treatment. This result is consistent with the fact that progesterone
levels increase during pregnancy, a physiological situation resulting in fetal HLA-G
expression. Nevertheless, it still remains important to elucidate the precise mechanisms that
contribute to HLA-G expression following heart transplantation. Our experiments
demonstrate that although HLA-G is tightly regulated in human tissue, vascular endothelial
and smooth muscle cells are capable of HLA-G expression in vitro. Induction of this
expression in vivo might represent a novel therapeutic strategy to protect against acute
rejection episodes and CAV.
6.3 FUTURE PERSPECTIVES
Transplantation remains the definitive option for patients with advanced heart disease (10).
Successful clinical and basic science research has translated into improved short-term
outcomes following heart transplantation. Additionally, medical advances have resulted in
higher risk patients being eligible for transplantation, and the use of donor hearts that
previously may have not been suitable. For these reasons, improvements in patient survival
may actually be more meaningful than what they appear (12). However, despite such
advancements in early patient survival, the mortality rate in patients surviving past the first
post-operative year has remained unchanged (12). Therefore it is worthwhile to aim current
81
and future strategies towards solving the long-term problems which arise post-transplant. In
the context of heart transplantation, HLA-G is negatively associated with long-term rejection
as demonstrated by this investigation and those conducted by others (104-106). It is therefore
an attractive protein in this context because of its potential clinical implications.
In order to minimize inter-assay variability, standardized HLA-G assays need to be
developed if sHLA-G concentrations are to be used for future investigation or clinical
assessment. This will ensure that levels are measured accurately thus enabling results to be
reproducible and applicable across different transplant centres. HLA-G detection protocols
across centres must also remain consistent to maintain reproducibility. All subsequent assays
must be conducted with plasma rather than serum samples, since sHLA-G levels in plasma
are often greater than in serum (208). Therefore, serum HLA-G measurements, as reported in
several earlier clinical investigations, may not accurately true sHLA-G levels.
If HLA-G is to be utilized therapeutically in heart transplant recipients, either through
administration of recombinant protein or by induction of its expression via pharmacologic
mechanisms or gene therapy, the side-effects of such strategies need to be assessed. An
important consideration is the risk of malignancy which is a major complication post-heart
transplant (12). The relationship between HLA-G expression and cancers has been well
documented, suggesting that this expression might enable tumor survival (125,128,129).
However, it is important to emphasize that HLA-G has not been shown to induce malignant
transformation. Its association with cancer is likely the consequence of HLA-G upregulation
by malignant cells as a survival strategy. There is no evidence suggesting that pre-existing
HLA-G leads to the generation of new tumours. Furthermore, the onset of carcinogenesis
following heart transplantation is likely caused by immunosuppressive agents affecting the
82
cancer immunosurveillance capability of the recipient (18-21). If HLA-G decreases or
ultimately abolishes the need for lifelong immunosuppression, the risk of post-transplant
malignancies may decrease.
There remain many unanswered questions with respect to HLA-G in the realm of
heart transplantation. There is a poor understanding of the relationship between HLA-G on
antibody-mediated rejection, which is clinically relevant in heart transplant patients (172).
Given that HLA-G inhibits CD4+ cells, it is likely to modulate the humoral response to some
degree. It should also be assessed whether pre-transplant sHLA-G levels have any clinical
value. These levels might be associated with freedom from rejection and vasculopathy post-
transplant. Moreover, it remains unknown why HLA-G is detected in certain patients only.
Genetic factors likely play a role since as different HLA-G alleles have been shown to result
in differential protein expression (109). Yet, the fact that HLA-G has not been detected in
healthy cardiac tissue but in biopsies of post-transplant allografts indicates that there are
factors present in the peri- or post-transplant stages that induce expression of this gene. Our
experiments revealed that stressors such as cytokines, H/R or immunosuppressive agents
were unable to induce this expression in vitro. Nonetheless, genotype might predispose
certain individuals to express HLA-G post-transplant. Future investigations might assess the
HLA-G genotypes of organ donors and recipients to determine whether this has an affect on
post-transplant expression. Such issues may be better clarified through prospective
investigations.
Our in vitro investigations illustrate induction of HLA-G expression in vascular
endothelial and smooth muscle cells following progesterone treatment. Yet, age and gender
did not appear to have a relationship with HLA-G expression among patients in our clinical
83
investigation. Future experiments may examine the role of progesterone on HLA-G
expression and transplant outcomes, to determine if it is indeed protective in vivo. These
investigations might also be expanded to co-culture models to assess whether progesterone
treated cells are protected from T cell and NK cell responses. Additionally, it may be
worthwhile to examine the role of circulating progesterone levels on HLA-G expression in
vivo. For instance, fluctuations in progesterone levels during the menstrual cycle may
influence sHLA-G levels.
Transplanted organs, undoubtedly, do not survive as long or perform as well as
healthy organs due to a multitude of factors. Research in the realm of transplantation is
directed towards advancements in several areas including surgical technique, donor organ
preservation, immunosuppressive therapy, tolerance induction and xenotransplantation.
Improvements in these areas can improve patient survival, reduce negative side effects and
potentially solve the organ shortage problem. The generation of tolerance induction,
however, is considered the ultimate goal of transplantation research. Since the fetus can be
considered a semi-allograft, it might be worthwhile to study the mechanisms of fetal
protection during pregnancy and to determine whether these mechanisms can be applied to
the realm of transplantation. Expression of HLA-G is one such strategy by which an
immunosuppressive protein expressed primarily during pregnancy is applicable to the post-
transplant setting. Other immunosuppressive molecules that participate in fetal tolerance,
such as galectin-1 (209), might represent useful candidate proteins for gene therapy
following transplantation to combat rejection and vasculopathy. Applying concepts from
pregnancy to transplantation might ultimately represent an effective strategy to induce
allograft tolerance.
84
The negative relationship between allograft and/or soluble HLA-G expression and
graft dysfunction secondary to rejection has been well documented following solid organ
transplantation. The clinical relevance of HLA-G in the transplant milieu can be attested to
the fact that such findings are based on studies encompassing over one thousand heart, liver
and kidney transplant patients (46). Allograft HLA-G expression inhibits NK cells, T cells
and APC in the local milieu and, unlike medical therapy, does not systemically
immunosuppress the host, making it particularly attractive. Our studies have shown that
HLA-G expression reliably indicates freedom from clinically significant rejection and that
this expression can be induced in vitro. Therefore, there is considerable potential for HLA-G
expression as a prognostic indicator or as a therapeutic endpoint to protect against allograft
rejection and vasculopathy in heart transplant recipients.
85
CHAPTER 7:
REFERENCES
86
1. Ellis SA, Sargent IL, Redman CW, McMichael AJ. Evidence for a novel HLA
antigen found on human extravillous trophoblast and a choriocarcinoma cell line.
Immunology. 1986; 59: 595-601.
2. Geraghty DE, Koller BH, Orr HT. A human major histocompatibility complex
class I gene that encodes a protein with a shortened cytoplasmic segment. Proc Natl Acad Sci