1
UNIVERSITY OF PISA
Specialization School in Patologia Clinica
Director: Prof. Aldo Paolicchi
Thesis of Specialization
5-AZACYTIDINE AND RAPAMYCIN: DIFFERENT AND SYNERGISTIC
EFFECT ON EX VIVO EXPANSION OF NATURAL HUMAN T
REGULATORY CELLS
Candidate Supervisors
Dr. Giuseppina Conteduca Prof. Frédéric Baron
(University of Liège)
Prof. Cristina Bottino
(University of Genova)
2
INDEX
1. INTRODUTION………6
1.1 IMMUNE TOLERANCE...6
1.1.2 Natural Arising Regulatory T cell...7
1.1.3. Foxp3 controls the development and the function of Treg………7
1.1.4.How does Foxp3 orchestrate the cellular and molecular programs involved in Treg's operation...8
1.1.5. The antigen repertoire recognized by Treg...10
1.1.6. Interleukin 2 and Foxp3 + Treg cells………10
1.1.7. Generation of peripherals Treg cell from naive cells………12.
1.1.8. Treg cells localization…...13
1.1.9. Activation, proliferation and de-differentiation of Treg………14
1.1.10. Mechanisms of Treg immunosuppressive action………15
1.2. Graft versus host disease………..16
1.2.1. Treg and GVHD: Experimental studies………17
1.2.2. Treg and immunosuppressive agents………..18
1.2.3. Treg and xenogeneic GVHDa...18
1.2.4. Treg and GVHD: Translational studies………19
1.2.5. Treg Injection………21
1.2.6. Rapamycin...23
1.2.7. The mammalian Target of Rapamycin (mTOR)………23
1.2.8. Activation and role of mTOR……….24
1.2.9. Inhibition of mTOR by Rapamycin……….26
3
1.2.11. Rapamycin and B lymphocytes………27
1.2.12. Rapamycin and dendritic cells (DC)……….28
1.2.13.The effects of 5-azacytidine on the function and number of regulatory T cells and T-effectors………..28
2. Objectives and work plan………..30
3. Materials and Methods…………...31
3.1.Mononuclear cell separation………...31
3.2. Cell preparation...32
3.3. Antibodies, reagents, and flow cytometry………32
3.4. Isolation of regulatory T cells by FACS ARIA……….33
3.5. RNA isolation and quantitative real-time polymerase chain reaction………..33
3.6. Methylation assay...33
3.7. Functional analysis of regulatory T cells...34
3.8. Statistical analysis………...35
4. Results...36
4.1. Responses of Treg (CD4+CD25+Foxp3+) and conventional (CD4+CD25neg) T cells to RAPA and/or 5-Aza in the presence of CD3/CD28 Ab crosslinking and IL-2...36
4.2. Apoptosis of CD4+CD25neg T cells in the presence of RAPA and/or 5-Aza………37
4.3. Proliferation kinetics of Treg and conventional T cells + RAPA and/or 5-Aza………..39
4.4. Combination of azacytidin, rapamycin and IL-2 preserves nTreg cells suppressive function in vitro and in vivo………41
4.5. Methylation of upstream Foxp3 enhancer………...42
4.6. 5-Aza and Rapa differentially affects cytokine production by CD4+CD25+ cells in vitro…………44
4.7. 5-Aza effect on murine model of GVHD………46
5.Discussion………..47
4
Summary
BACKGROUND: Natural T regulatory cells (Treg) are challenging to expand ex vivo, and this has severely hindered in vivo evaluation of their therapeutic potential. 5-Azacytidine (5- azaC) and Rapamycin (RAPA) are immunosuppressive drugs that promote selectively the expansion of CD4+CD25highFoxp3+ regulatory T cells
OBJECTIVE: We investigated whether 5-azaC and RAPA could be used together to promote the ex vivo expansion of Tregs purified from adult human peripheral blood.
METHODS: CD4+CD25+ and CD4+CD25- T cells were isolated from PBMC of normal controls using Miltenyi beads and FACS ARIA sorting. These cells subsets were cultured in the presence of anti-CD3/CD28 antibodies and 200 IU/ml of IL2 for 2 weeks. RAPA (100nM) and 5-azaC (1μM) were added to half of the cultures. We monitored cell expansion and after harvest, the cell phenotype, gene expression, T suppression activity and Annexin V binding were determined and compared to values obtained with control cell culture and with freshly separated CD4+CD25+ and CD4+CD25- T cells. Also we tested the absolute number of Treg in the several culture conditions during the time of expansion.
RESULTS: We found that 5-azaC helped maintain FOXP3 expression during the expansion process probably by promoting the conversion of T conventional (Tconv) in Treg, instead Rapa induces selectively apoptosis in Tconv cells and expansion in Treg. The addition of 5- azaC to RAPA treated cultures improved gene expression of FOXP3, CD25, STAT5 and TGF-B resulted in enhanced Treg expansion and suppressive activity. Also Rapa and 5- AzaC combination sustain Bcl-2 protein expression in Treg conferring resistance to apoptosis process.
CONCLUSION: 5-azaC may have utility in ex vivo expansion of human Tregs, not as a single agent, but in combination with RAPA. These data may considerably accelerate the development of immunotherapeutic approaches for the treatment of autoimmune disease or posttransplant alloreactions by the adoptive transfer of nTreg cells
5
1. INTRODUCTION
The mammalian immune system protects the host from a wide spectrum of microorganisms, while avoiding, however, abnormal and excessive immune responses. Immune responses, both protective and harmful, are mainly mediated by T and B cells, which possess enormous diversity in antigen recognition, high antigenic specificity, powerful efficacious activity and a lasting immune memory. By virtue of these potentialities, aberrant immune system reactions such as autoimmunity and allergies can cause serious damage to the host. One of the major goals of Immunology and Medicine is to understand how to establish and maintain the state of non-responsiveness to self-antigens (Immune Tolerance).
1.1 Immune Tolerance
There are two types of "recessive" and "dominant" mechanisms for maintaining self-tolerance and immunostimulation homeostasis. In the recessive mechanism, the fate of self-reactive lymphocytes is determined intrinsically to the cell itself. For example, some lymphocytes are programmed to die for apoptosis when exposed to self-antigen during their development in central generative organs (thymus for T lymphocytes and bone marrow for lymphocytes B). Other T and B lymphocytes may replace Self-reactive receptors (TCR and BCR respectively), with other non-reactive ones. This mechanism is called editing. Cells that escape by cloning and editing may become anergic after meeting with the self-antigen, in addition to activating express inhibitors, or molecules involved in cell death signaling.
6
In the recessive mechanism, some T cell subtypes control the expansion of abnormal or exaggeratively reactive lymphocytes. For some years now it has been discovered that the normal immune system produces a population of T cells, called regulatory T cells (Treg). Disease in the development or function of Treg is the main cause of autoimmune and inflammatory diseases in humans and animals. All adaptive immune responses include lymphocytes T and B, but also Tregs, and the balance between the two populations is crucial to controlling the quality and response power and to establishing tolerance to self-antigens.
1.1.2 Natural Arising Regulatory T cells
Newborn neonatal timectomies of about three days of mice result in autoimmune damage in various organs (such as thyroid, ovary, testis, and intestines) and appear on tissue-specific autoantibodies in circulation. Timectomies in adult rats followed by radiation cycles produce autoimmune thyroiditis and type I diabetes. The inoculum of CD4+ T lymphocytes or CD4+CD8- thymocytes from untreated singular animals inhibits the autoimmunity development. These results indicate that normal animals produced not only self-reactive T cells but also T cells that can suppress self-immunity. CD4+ T cell total purified from normal mice or rats spleen and depleted by CD4Rblow populations transferred to syngeneic athymic mice induced in these last spontaneous autoimmune disorders, reconstituting mice with CD45Rblow subpopulation autoimmunity was inhibited (1; 2). Subsequently, further markers of these cells were identified such as CD25 (receptor chain of IL-2) (3). Treg depletion also produces an autoimmune type of colitis (Immune Bowel disease), due to an excessive response to intestinal bowel bacteria (4). Removal or reduction of CD25+ CD4+ Treg cells causes an increase in the response to both tumors and infections, leading to decreasing tumor mass and decaying the microbes. Instead, an enrichment of these cells leads to suppression of allergies, stabilization of organ transplant tolerance, prevention of graft versus host in bone marrow transplantation and promotion of mother-fetal tolerance.
7
In the light of these evidences, it can be asserted that the immune system generates CD4+ CD25+ Cd45Rblow market cells which are engaged in self-responses, or self tumor antigens, and non-selfs (such as microbes, transplants). These cells are generated in thymus but persist in the periphery where they exert a constant control over self-reactive T lymphocytes. A decrease in Treg in the periphery generates autoimmune disorders and immunopathologies.
1.1.3. Foxp3 controls the development and the function of Treg.
Treg cells from thyme express specifically Foxp3 (forkhead box P3), a member of the transcription factor family of hairpin factors that bind the double helix of the DNA. Foxp3 is the largest regulator of Treg's development and function. The foxp3 gene is the first defective gene identified in the Scurfy mouse strain. Scurfy is a recessive X-linked mutant that is lethal for hemizygotic males around one month after birth, as they develop excessive activation of CD4+ T cells and overproduction of proinflammatory cytokines (5). Gene mutations for Foxp3 are the genetic cause of diseases such as IPEX (Immune Dysregulation, Polyendocrinopathy, Enteropathy, X-linked Syndrome), which is the human counterpart of Scurfy. The clinical and immunological symptoms of IPEX patients is similar to that developed by Tregs depleting mice and this has led to the idea that Foxp3 is crucial to the generation, development and function of Treg (6). Various studies have found that T cells and CD4+CD8-CD25+ thymocytes express Foxp3, while other T or timocyte populations are activated and resting do not express it. Retroviral gene transduction of foxp3 gene converts T CD4+ CD25- cells into CD4+ CD25+ cells that are capable of suppressing in vitro proliferation of other T cells and inhibiting the development of autoimmune pathologies in vivo. (7) Foxp3 gene transduction , also upregulates the expression of CD25 and other molecules associated with the market surface such as CTLA-4 (Cytotoxic T Cell associated antigen-4) and GITR (Glucocorticoid-induced TNF receptor familyrelated gene / protein), while repressing the production of IL-2, IFNγ, and IL-4. Both in mice knock out (ko) for foxp3 and in Scurfy mice, which lack DNA binding domain of Foxp3 factor T cell CD4+ CD25+ inoculates from normal mice prevent severe systemic inflammation(6). In chimeric mice whose bone marrow is made up of cells from wild type (wt) mice and foxp3 knockout (ko) mice, wt cells are able to generate mature Tregs that suppress the development of autoimmune disorders that
8
ko fail in the generation of Treg cells. Conversely, in mice that overexpress Foxp3 the number of CD4+ CD25+ cells is greater (8).
Thus, we can conclude that the Foxp3 transcription factor is important for positive cell α/β TCR differentiation in T CD4+ CD25+ T regulating cells, which high levels of foxp3 expression also confer abnormal activity on non-market cells, mutations In the foxp3 gene lead to autoimmune disorders both in the mouse and in humans and finally that Foxp3 is fundamental to the generation, development and function of the Tregs.
1.1.4. How does Foxp3 orchestrate the cellular and molecular programs
involved in Treg's operation
It has been shown that Foxp3 interacts with the factor of Transcription of Nuclear Factor of Activate T cells, AML1 (Acute Myeloid Leukemia 1), Runx1 (Run relatedtrnscription factor1), HAT/HDAC complex (Acetone transferase / histone deacetyl transferase) and NF-kB. NFAT activity is controlled by calcium and calcineurine calcium-dependent phosphatase. After NFAT activation is a complex with AP-1 and NF-kB, which promotes the expression of IL-2, IL-4, CTLA-4 and other genes involved in the activation of T lymphocytes and in the differentiation of efficacious lymphocytes. A recent study has shown that Foxp3 binds NFAT at the time both bind DNA (9). Amino acid substitutions in the Foxp3 hairpin domain prevent the formation of the Foxp3-NFAT complex and less Foxp3 functions as suppressing IL-2, activate CTLA-4 and CD25, confer abilities to suppress T cells. Ko mice for NFAT develop spontaneously lymphadenopathy, selective activation of Th2 cells accompanied by hyperproliferation of IgE, suggesting that lack of NFAT affects the Foxp3 function and consequently the Tregs. The AML1/Runx1 complex is crucial for normal hematopoiesis and for the development of T cells in thymus. Unlike NFAT which is dephosphorylated by calcineurin and translocated into the nucleus after activation of T cells, AML1/Runx1 binds the promoter II-2 upstream of the NFAT binding site of AP-1 NF-kB. After activation AML1/Runx1 facilitate the binding of NFAT to the promoter of Il-2. In Treg, however, the AML1/Runx complex binds to Foxp3 in the N-terminal region between the leucine zipper domination and the hairpin box. Ko mice for
9
AML1/Runx1 show autoimmune disorders similar to those that develop after the depletion of the Treg. Foxp3 signaling is also involved with HAT and HDAC. When Foxp3 is acetylated by HAT it increases its binding to the Il-2 promoter and increases the repression of this gene (10). Recent studies have shown that Foxp3 controls about 700 genes, of which 10% binds them directly (11). The genes controlled by Foxp3 coding for molecules involved in different signaling pathways (such as Zap70 and Ptpn22), transcription factors (such as Crem), cytokines (such as Il-2), and surface molecules (such as Il-2ra, CTLA-4, and Fas-L). All this suggests that Foxp3 activates a number of genes that lead to differentiation of T cells in Treg. To support this model, the observation that mice expressing amounts of Foxp3 lower than normal Tregs not only lose their suppressive capacity but spontaneously differentiate in T effectors producing IL-2, IFNγ, and IL-4 (12) lymphocytes. Foxp3 can also interact with RORγ, a key transcription factor for the differentiation of T naive lymphocytes in Th17, inhibiting development (13)
1.1.5. The antigen repertoire recognized by Treg.
A key feature of Foxp3 + Treg cells is that when they come out of thymus they are already mature (they are capable to suppress) since they already encounter antigens in thyme before in the periphery (14). The Treg possesses a high affinity TCR for self peptides MHC-bound on thymocytes unlike from other T cells. Transgenic T cells for a specific TCR, missing of the gene that activates the TCR recombination (Rag - / -) do not possess the Tregs. The development of the Treg cell line would be induced in response to the interaction of TCR with the MHC-peptide complex self a degree of greed that places halfway between the positive and the negative selection. Thus the avidity of interaction between stromal cells and thymocytes is critical to determining the development of the Treg. Recent studies suggest that Foxp3 is not required in this timing selection phase. Mice whose gene for Foxp3 is replaced by a reporter gene encoding for green fluorescence protein (GFP), non-expressing GFP + tiomocytes exhibit a phenotype typical of Treg (CD25 +) but are not capable of suppressing in vitro.
10
This means that the interaction between thymocytes and stromal cells involves a pathway of gene interaction other than in which Foxp3 is involved, but Foxp3's parallel activation stabilizes and supports the Treg phenotype and gives suppression activity.
1.1.6. Interleukin 2 and Foxp3 + Treg cells.
Interleukin 2 is another fundamental molecule for Treg's function. CD25 is a high affinity component of the IL-2 receptor (IL-2R) and is functionally essential for the development of Treg. For a long time it was thought that IL-2 was the most important cytokine for proliferation and differentiation of T lymphocytes since in vitro is a potent growth factor. However, it has been shown that animals that do not express IL-2 have no serious problems in the differentiation and function of T cells but spontaneously develop inflammatory lymphoproliferative diseases with autoimmune components such as hemolytic anemia. Also ko-mice for CD25 succumb to similar autoimmune disorders. In humans defective in CD25 expression lead to severe autoimmunity and allergies with symptoms not very different from the IPEX. Various evidence suggests that this syndrome is due to a decrease or malfunction of the Tregs: first, in fact, IL-2 or CD25 ko mice exhibit a small number of Treg and that in these mice can be prevented by autoimmunity by inoculating CD25+ CD4+ T cells from mice wild-type (Wt). Second, Ko mice for STAT5 (nuclear factor involved in signaling activated by IL-2 receptor) do not develop Treg Foxp3+ (15) cells. Finally, inoculating an anti-IL2 monoclonal antibody in normal neonatal mice reduces the number of CD4+ CD25+ Foxp3+ cells for a limited period of time and autoimmune disorders similar to those produced by Treg depletion. Furthermore, it has been demonstrated in vitro that IL-2 enhances the function of Tregs to suppress T lymphocytes and is necessary to maintain the expression of Foxp3 and CD25 in these cells (16; 17). Although there are other cytokines whose receptors have in common the γ (γc) chain with the IL2 receptor (4, IL-7, IL-15), they can not fully satisfy the function of IL-2 when defects in expression or functionality of this cytokine are present. Conversely, if the IL-4, IL-7 or IL-15 expression is defective, the number of Foxp3+ cells remains unchanged, and the less autoimmunity occurs. It can be said that IL-2 is vital for the development and maintenance of Foxp3+ Treg cells. It is known that IL-2 has many target cells: T CD4+ and CD8+, B cells, natural killer (NK) lymphocytes, and thus, in the immune response, an
11
apparently contradictory pleiotropic function. IL-2 facilitates the differentiation of T CD4+ lymphocytes Th1 and Th2 and the expansion of CD8+ memory cells and NKs. On the other hand, IL-2 promotes apoptosis in activated T cells, maintains Foxp3+ Treg, with high doses induces expansion, facilitates TGFβ-dependent T cell cellular differentiation in Treg but inhibits TGFβ/ IL-6-dependent differentiation of naïve T cells in inflammatory cells Th17 (18). Assuming that the major source of IL-2 is activated T cells, there is a nagative feedback that controls the immune response: IL-2 is produced by activated cells, contributes to maintaining the expansion and activation of Foxp3+ Treg cells, which in turn suppress activated T cells. Interruption of this circuit causes inflammatory and autoimmune disorders.
1.1.7. Generation of peripherals Treg cell from naive cells.
Peripheral naive cells in particular experimental conditions such as in the presence of cultured TGFβ may acquire the expression Foxp3 and consequently the function of Treg (19). Treg TGFβ-dependent induction in vitro inhibits, this phenomenon of differentiation of naive T cells in Foxp3+ cells and induces them to differentiate at Th17 (20). IL-2 facilitates the differentiation of CD4+ T naive cells in Treg Foxp3+ but inhibits differentiation at Th17. Also retinoic acid secreted by a particular subset of dendritic cells (DCs) found in intestinal lymphatic organs inhibits the induction of Th17 IL-6-guided and in the presence of TGFb induces differentiation to Treg (21). Thus, by administering antigenic proteins, DCs associated with the intestine retinoic acid induce the Treg and this could be a possible mechanism underlying oral tolerance. It is still good to understand whether the Treg cells induced in the periphery are functionally stable in vivo, it has been observed in humans that peripheral Terg activate Foxp3 after stimulation of TCR more readily than the natural ones but the expression of this factor is lower and less stable. In addition to the Foxp3+ Treg cells, there are other types of regulatory T cells that are induced in peripherals such as IL-10 and TGFβ-producing T cells CD4+ called Tr1, in vitro induced by stimulating IL-10 (22) CD4+ T cells. Tr1 cells do not express Foxp3 but their in vitro properties are very similar to Treg Foxp3+ cells. Another type of regulatory T cells are Th3, they are antigen-specific cells that produce TGFβ and very often express Foxp3. Other T-cell subtypes with regulatory activity include CD8+, CD4-CD8-, and γ/δ but there is very little evidence that these
12
cells play a crucial role in self-tolerance. It is known that immunizing with strong adjuvants with not only do CD4+ Foxp3+ but also CD8+ Foxp3+ cells expand, these cells plausibly can correspond to CD8+ suppressors already described in the '70s (Figure1).
Figure1. Foxp3 Treg may mediate Treg inhibitory function through multiple soluble and
cell-surface factors. CTLA-4, TIGIT and PD_1 interact with costimulatory molecules on antigen
presenting cells (APC). CD25 binds the T cell growth factor IL-2. CD39 converts local ATP to adenosine. The cytokines IL-10, IL-35 and TGF-b have suppressive functions on nearby immune cells.
1.1.8. Treg cells localization.
Foxp3 + Treg cells specific to self-antigens tissues are located in their peripheral lymph nodes. Foxp3+ cells migrate into inflammatory tissues at the sites of infection and tumor-transformed tissues, in fact, in addition to the TCR specific antigen, Treg express a series of homing receptors as adhesion molecules and chemokine receptors. 80% of the Foxp3+ peripheral cells express high levels of CCR7, some express CD103 interacting with the E-caderine ligand from the epithelial cells, about 50% express high CD62L levels that interact with CD34 vascular addressing, GlyCAM-1 , And MAdCAM-1. These homologous receptors control the traffic and location of the Tregs, and the expression patterns of these receptors characterize different types and cellular features, for example, Treg Foxp3+ cells expressing CD103 are memorylike and have a 'Greater suppression activity than CD103- both in vitro and in vivo. In a model of murine colitis, Treg Foxp3+ CD103+ cells are able to control inflammation in the intestinal mucosa or Treg Foxp3+ CCR5+ cells migrate preferentially to
13
skin lesions due to Leishmania major infection while Treg Foxp3+ CCR6+ cells are recruited mainly in the " Rheumatoid arthritis. Ko cells for the homing receptor are no longer able to perform their function in infected tissue, transplantation, and in mother / fetal tolerance.
1.1.9. Activation, proliferation and de-differentiation of Treg.
Following the presence of antigen in regional lymph nodes, Treg Foxp3+ cells are activated and exert a lower antigen concentration than that in which T naive cells are activated. This suggests that the naturally occurring Treg can also be activated by immature dendritic cells expressing CD80/86 and the MHC/peptide complex too weak to activate T naive lymphocytes self-reactive so that the Treg exercised on these a stronger and dominant suppression. Contrary to the evidence obtained in vitro, Treg Foxp3+ cells in vivo proliferate continuously under the stimulation of self-antigens and commensal microbes. Several stimuli may induce Treg Foxp3+ cells to both clonal in vivo and in vitro expansion, such as stimulating GITR, expressively expressing Treg Foxp3+ cells in the presence of IL-2 or Toll like receptors (TLR-2, 4, 5 and 8). Despite intense proliferation, the percentage of Treg Foxp3+ cells is constant in normal mice (10-15% of CD4+ T cells), indicating that cell death holds homeostasis in fact antigen stimulation in Treg Foxp3+ down-regulates expression of Bcl-2 (anti-apoptotic protein) (23). In conclusion, Treg Foxp3+ cells migrate into lymphoid tissues and not, after being proliferated, they exert their suppressive activity and then die.
14 Figure 2: Regulatory T lymphocytes called natural or induced (details in the text) (Mills 2004).
Another subpopulation of T lymphocytes with regulatory activity also involved in maintenance of peripheral tolerance has also been discovered in CD8+ T lymphocytes. These CD8+ T lymphocytes with regulatory activity, expressing Foxp3, are generated from CD8+ CD25-T lymphocytes after antigenic stimulation. These CD8+ Foxp3+ have the ability to inhibit the proliferation of naïve CD4+ and effectors by direct contact with their target and/or by the secretion of inhibitory cytokines (IL-10 and TGF-β). Their immunoregulatory action is restricted to MHC class I and is specific for the antigen. More recently, another subpopulation of iTreg that does not express Foxp3 has been identified. Indeed, Collison et al. have shown that in the presence of IL-35, heterodimeric cytokine inhibitor of EBI3 for "EBV-induced gene 3" and p35 (IL12A), a member of the IL-12 family, activated naive CD4+ T cells exhibited immunoregulatory activity in vivo and in vitro via IL-35 secretion.
15
Tregs control the immune response by controlling the proliferation of naive T lymphocytes and their differentiation into effector T lymphocytes (TH1, TH2 and TH17). They can also inhibit the effector activity of differentiated CD4+ and CD8+ and the function of NK, NKT, lymphocytes B, macrophages, and DCs (24). It has been widely demonstrated that Tregs require activation of their TCR by the antigen they are specific for their immunosuppressive activity, which is not specific to the antigen. This is called "by-step suppression" (25). This suppressive activity is particularly associated with their ability to inhibit IL-2 production and promote the cessation of the cell cycle (anergy) (26). Although the Treg mechanism of action has not yet been fully elucidated, carried out in vitro and in vivo have revealed numerous mechanisms by which Treg can regulate the immune response. Among the mechanisms identified, Tregs could control the immune response by (Figure 3):
(1) Secretion of inhibitory cytokines
(2) Lysis of their target via the production of granzyme / perforin
(3) Deregulation of cellular metabolism
16 Figure 3: Mechanisms proposed to explain the mode of action of Treg. The Tregs could act in 4 ways: (1)
via secretion of inhibitory cytokines; (2) by cytolysis; (3) by deregulation of cellular metabolism; (4) by modulating maturation and function of DC. A2A: adenosine receptor; CAMP: cyclic adenosine monophosphate; IDO: "Indoleamine 2,3-dioxygenase"; TEff: effector T lymphocytes; CTLA-4: "cytotoxic T lymphocyte-associated antigen"; Pfr: perforin; GrzB / A: granzyme B / A; TGF- β: transforming growth factor; IL: interleukin (Collison and Vignali 2008).
1.2. Graft versus host disease
Regulatory T cells are an important mechanism for controlling immune reactions that affect post-transplant rejection. Hematopoietic stem cell allograft is the treatment of choice for some patients with malignant haemopathies. Its effectiveness depends largely on the destruction of tumour cells of the patient by the T lymphocytes present in the graft, also called the effect of graft-versus-leukemia (GVL). Unfortunately, same lymphocytes are also responsible for inducing graft versus host disease. Currently GVHD for "graft-versus-host disease", this disease is the consequence of the activation of donor (and particularly T-cell) immune cells against antigens specific to the recipient by the antigen presenting cells (APC).
17
The graft should contain immunocompetent cells
Histocompatibility between donor and recipient
Inability of the recipient to provide an effector response to eliminate or inactivat en grafted cells
GVHD has two main forms: an acute form (GVHDa) and a chronic form (GHVDc). Conventionally, these two forms are distinguished by their moment of appearance after transplantation (before or after day 100). However, it has been observed that in patients Non-myeloablative packaging, that is, does not destroy the bone marrow of the recipient (Mini-allografts), or of reduced intensity could develop GVHD of acute form beyond 100 days after transplantation. Therefore, the National Institute of Health (NIH) National Institutes of Health) proposed classifying the two forms of GVHD clinical symptoms observed rather than by time criteria.
1.2.1. Treg and GVHD: Experimental studies
Several studies using experimental mouse models of GVHDa have demonstrated the major role of CD4+ CD25+ Treg in the prevention of GVHDa. In 2002, Taylor et al. have observed in one model of allogeneic bone marrow transplantation, an acceleration of the appearance of GVHDa when the graft was depleted ex vivo to CD25+ or when the recipient was treated with antibodies anti-CD25+ before grafting. They also observed a significant increase in the survival of mice that received CD4+ CD25+ at the time of transplantation (27).
Later, the Negrin team showed, in a murine model of allogeneic transplant HLA-identical that the administration of high doses of Treg (ratio 1:1) induced a decrease significant mortality by GVHDa while preserving the GVL effect (28). Subsequently, the same team compared the effect of CD4+ CD25+ CD62L+ Treg and CD4+ CD25+ CD62L-Treg on the incidence of onset of GVHDa. They showed that only CD4+ CD25+ CD62L+ Treg protected GVHDa by their ability to migrate to lymph nodes and the spleen and thus to exert their suppressive activity on the CD4 + CD25- (29). Other studies have shown that Tregs are also capable of preventing the development of CD8+ T lymphocyte-mediated GVHDa in murine models of allogeneic graft with incompatibility for miHA
18
(30, 31). On the other hand, very few studies have studied the impact of Treg infusion in murine models of GVHDc. Among these, Zhao et al. have shown that CD4+ CD25+ CD103+ Treg activated in vivo were more effective in treating GVHDc than CD4+ CD25+ CD62L+ CD103-Treg freshly isolated or activated in vitro. CD103 is a heterodimeric integrin (αEβ7) which can be used as a marker in the identification of activated Tregs in vivo (CD4+ CD25+ CD62L- CCR5+ CD103+ Foxp3+) (32). This increase in therapeutic activity of CD103+ Treg compared to CD4+ CD25+ CD62L+ CD103- could be associated with they strongly express receptors to pro-inflammatory chemokines such as CCR5, which allows them to migrate to the inflammatory sites of GVHDa where they can induce apoptosis of activated CD4+ and T lymphocytes (33).
1.2.2. Treg and immunosuppressive agents
In 2005, Battaglia et al. showed that rapamycin favors the expansion of CD4+ CD25+ Foxp3+ Treg in vitro (34) and in vivo, in contrast to CsA and tacrolimus (35). Following these results, various studies have been carried out to study the influence of rapamycin as well as other immunosuppressants on suppressive activity and the expansion of Treg in different experimental mouse models of GVHDa.
The results of these different studies show a decrease in the incidence of GVHDa in mice that received rapamycin compared to mice that received MMF or CsA. This decreased proliferation of effector T lymphocytes as well as an increase in the percentage and absolute number of CD4+ CD25+ Foxp3+ Treg. These increases are due to an expansion of the CD4+ CD25+ Foxp3+ cells of the graft and to the conversion of CD4+ CD25+ CD4+ CD25+ T lymphocytes with immunosuppressive activity.
1.2.3. Treg and xenogeneic GVHDa
Other studies have evaluated the impact of amplified or non-ex vitro Treg infusion in xenogeneic murine models pathways of GVHDa. These different results demonstrate, as in the non-xenogeneic murine models of GVHDa, that the administration of human Treg freshly isolated or amplified ex vitro decreases the incidence of GVHDa in different murine xenogeneic models of GVHDa. However,
19
depending on the Treg amplification methods, their immunosuppressive activity is more or less influenced. Indeed, Hippen et al. have shown that Tregs derived from cord blood umbilical cells amplified ex vivo in the presence of CPA (KT32 = cell line of erythromyeloid leukemic cells) artificially expressing the co-stimulation signal OX40L or 4-1BB (two co-stimulation molecules that play an important role in the proliferation and survival of Tregs (36, 37) proliferated in 3 weeks, 1000 fold more than Tregs amplified in the presence of abCD3/CD28, without loss of their immunosuppressive activity. Moreover, they observed that Treg amplified in the presence of KT32/4.1BBL or KT32/OX40 expressed 2 fold the pro-apoptotic gene BH3 than Tregs amplified in the presence of abCD3/CD28, suggesting that this method of Treg amplification could be clinically adaptable (38). Some of these (not all) studies have also demonstrated that iTreg can prevent the development of GVHD. Among these, Hippen et al. have shown that generated iTreg from CD4+ CD25- and CD4+ CD25-CD45RA+ in the presence of rapamycin and TGF-β presented an immunosuppressive activity comparable to nTreg. They also observed that the activity of these iTreg was stable over time in contrast to iTreg generated in the presence of TGF-β, suggesting a synergistic effect of rapamycin and TGF-β (39). However, recently, Vermuren et al. have demonstrated for the first time that iTreg generated from CD4+ CD25- in the presence of IL-2 and anti-CD3 and CD28, in spite of their immunosuppressive activity in vitro, were unable to prevent the onset of GVHDa. This loss of activity was correlated with loss of expression of Foxp3+ compared to nTregs whose expression of Foxp3+ and immunosuppressive activity was maintained in vivo (40), suggesting that depending on the method of obtaining and amplifying iTregs, they may lose their immunosuppressive activity in vivo.
1.2.4. Treg and GVHD: Translational studies
Following the results obtained in vivo, various clinical studies have been induced to study the correlation between the number of Tregs present in the graft and the incidence of GVHD after transplantation allogeneic. Numerous studies have shown that the incidence of GVHD was inversely proportional to the number of Tregs present in the graft (41-44). Recently, Matsuoka et al. have studied the immune reconstitution of conventional T lymphocytes and Tregs in a heterogeneous
20
population of 45 patients who received an allogeneic CSH transplant and observed that Treg number declined 9 months after the transplant and that this decrease was correlated with an increase in the incidence of extensive GVHDc. These authors have shown that the decline in time of Treg was due to CD4+ persistent lymphopenia that selectively influenced proliferation and survival of Treg (45). Kawano et al. have analyzed telomere length, and Treg telomerase activity in 61 patients who survived 2 years after receiving an allogeneic CSH transplant. They observed that in patients with GVHDc or without GVHDc, Treg telomerase activity was associated with a increased expression of the anti-apoptotic protein Bcl-2, an increase in the number of Treg and, consequently, the establishment of a peripheral tolerance. On the other hand, patients with severe GVHDc showed an increase in Treg apoptosis resulting in a decrease of Treg number and the loss of the establishment of a peripheral tolerance (46).
However, other clinical studies did not observe a correlation between the number of Treg and the incidence of GVHD. For example, Ratajczak et al. did not observe a decrease in the percentage of Treg in 96 patients with severe (acute and chronic) GVHD digestive tract or skin compared to patients without GVHD. They also observed increase in TH17 in the target tissues of GVHD. However, they observed that patients with GVHD had a TH17/Treg <1 ratio suggesting that in vitro in situ TH17/Treg ratio could be a specific marker in the diagnosis of GVHD (47). Recently, in a retrospective study, Rosenzwajg et al. have allogeneic bone marrow transplant without prior manipulation, there was no correlation between the absolute number or frequency of Treg present in the graft and the incidence of GVHDa (48) contrary to what was observed in peripheral blood stem cells transplantation (49-51). These results suggest that the Treg level differs depending on the source of stem cells (from peripheral blood vs. bone marrow), joining what had been previously described by Blache et al. (52) and which therefore influences the incidence of GVHDa. Lord et al. observed no correlation between the percentage of Treg found in biopsies done at the upper gastrointestinal tract and the incidence of GVHDa in a cohort of 60 patients having received an allogeneic CSH transplant, suggesting that the frequency of Treg is not a good marker in the diagnosis of GVHDa in the upper digestive tract (53).
21
1.2.5. Treg Injection
Not amplified
Following the observations obtained in vivo and in the various clinical studies, one of the strategies to prevent the development of GVHD is therefore to enrich the graft in Treg. Recently, Di Ianni et al. were the first to show that infusion of freshly isolated Treg 4 days prior to haploid identical CSH and conventional T lymphocytes (Tconv ratio:Treg=1:2) could prevent the onset of GVHDa and promote immunity reconstitution in the absence of immunosuppressive treatment after transplantation (54).
Amplification ex vivo
The major difficulty in ex vivo Treg amplification is to maintain their suppressive activity and their migration capacity while avoiding the amplification of the contaminating effecting T lymphocytes. Preliminary studies have shown that Tregs derived from umbilical cord blood more rich in Treg than adult peripheral blood and could be easily amplified ex-vivo in presence of supra-physiological concentrations of IL-2 and stimulation of TCR (microbeads anti-CD3/CD28) without loss of either their suppressive activity or expression of their homing "(55, 56). In 2011, Brunstein et al. published the first phase I clinical trial using Treg derivatives of umbilical cord blood amplified and activated ex vivo in patients who received a double graft of CSH of umbilical cord blood with non-myeloablative packaging in order to prevent GVHDa (57). One of the major limitations of the use of umbilical cord is indeed the often too small amount of stem cells for their use therapeutic in adolescents or adults. A double umbilical cord blood therefore more easily reach the therapeutic dose (3.0 × 107 cells / kg) required in (58, 59). However, unlike the simple umbilical cord blood graft, the double graft increases the risks of grade II GVHDa (60). In their study, Brunstein et al. demonstrated the efficacy of infusion of one or two doses at 2 weeks interval of amplified Treg in vivo from the same donor as cord blood stem cells on the prevention of GVHDa after double umbilical cord blood graft. They also compared the effect of combination of MMF/rapamycin with the combination of MMF/CsA on Treg expansion and observed an (not significant) increase in the absolute number of Tregs in the MMF/rapamycin as compared to the MMF/CsA group (61).
22
Amplification in vivo
In 2009, Zorn et al. have examined the impact of IL-2 low-dose injection in vivo of CD4+ CD25+ Foxp3+ Treg. They observed that the injection of IL-2 into patients who had received an allogeneic CSH graft with infusion of CD4+ lymphocytes from the donor (CD4+ DLI) increased the percentage of CD4+ CD25+ Foxp3+ Treg in peripheral blood compared to patients who received only IL-2 or CD4+ DLI (62). More recently, Koreth et al. evaluated the impact of daily subcutaneous injection for 8 weeks of IL-2 low doses on Treg amplification in patients with GVHDc refractory to glucocorticoids. In this study, they did not receive any injection of IL-2. Eight weeks after the onset of IL-2 injections, a partial response was observed in half of the patients, with an improvement of the sclerodermic manifestations and fibrotic factors correlated with an increase in the number of Treg (CD4+ CD25+ CD127-) and the ratio Treg:Tconv (CD4+ and CD8+) in peripheral blood (63). Unlike the work on the impact of Treg infusion in the prevention of GVHD, there have been very few studies on the effect of Treg administration in the treatment of GVHD. One of the major limitations of this type of study is that a large number of Treg with a very high purity so as not to aggravate the GVHD by infusion of Treg contaminated with conventional T lymphocytes. The dose and the time of administration also remain poorly known. In addition, the amplification of Treg requires 2 or 3 weeks of culture, which proves to be too long when the patient presents GVHDa with rapid and severe progression. Recently, Edinger et al. have conducted a study in a small number of patients with GVHDa refractory to corticosteroids. In this study, they observed an improvement in the signs
of GVHDa at the level of the digestive tract after the CD4+ CD25+ Treg (95% Foxp3 +) transcription amplified ex vitro (64). Although these results are encouraging, further studies should be carried out in order to confirm these preliminary data.
1.2.6. Rapamycin
Rapamycin (Sirolimus®) is a cyclic macrolide with antifungal and antibiotic properties, initially isolated in the 1970s on Easter Island (also known as Rapa Nui) from the fungus Streptomyces hygroscopicus (65). In 1977, Martel et al. have discovered that rapamycin has been
23
immunosuppressive agent capable of inhibiting the proliferation of a large number of cell types (66). Later, the study of the mechanism of action of rapamycin led to the discovery of a protein called TOR for "Target of Rapamycin" initially discovered in yeasts, this protein has its homologous in mammals which is mTOR for "Mammalian target of rapamycin" (67).
1.2.7. The mammalian Target of Rapamycin (mTOR)
mTOR is a serine-threonine protein kinase belonging to the PIKK family ("phosphoinositide Kinase-related kinase ") that plays a major role in cell growth and proliferation, and also in the control of homeostasis (68). mTOR is part of two distinct multiprotein complexes, mTORC1 and mTORC2 (Figure 4).
Figure 4: TORC1 and TORC2 multiprotein complexes. mTOR: "mammalian Target of Rapamycin ";
RAPTOR: "regulatory-associated protein of mTOR"; mLST8: "mammalian lethal with Sec 13 protein 8"; RICTOR: "rapamycin-insensitive companion of mTOR"; MAPKAP1: "mitogen-activated protein kinase-associated Protein 1 "; PIKK: "phosphoinositide kinase-related kinase"; HEAT: "Huntington, Elongation Factor 3, PR65 / A, TOR" (Thomson, Turnquist et al., 2009).
mTORC1 is composed of RAPTOR for "regulatory-associated protein of mTOR" and mLST8
24
mTORC2 also contains mLST8 but also RICTOR for "rapamycin-insensitive companion of
mTOR " and MAPKAP1 for " mitogen-activated protein kinase-associated protein 1 " (69).
1.2.8. Activation and role of mTOR
Although several molecular mechanisms leading to the activation of mTORC1 have been described, information on the activation of mTORC2 are currently available, due to the lack of a specific inhibitor of mTORC2 (70). Activation of mTORC1 can be controlled by various stimuli:
• Growth factors and cytokines
Certain cytokines such as IL-1, IL-2, IL-3, IL-4, and IL-6 and growth factors that the IGF for insulin-like growth factor, EGF for epidermal growth factor, PDGF for platelet-derived growth factor activate the PI3K ("phosphoinositid-3-kinase") -AKT pathway leading to inactivation of the TSC complex for tuberculous sclerosis complex consisting of TSC1 and TSC2, by phosphorylation of TSC2. TSC is an Rheb inhibitor complex for "RAS homolog-enriched in brain "which is a powerful regulator of the mTORC1 activity. When TSC is inactive, Rheb sets itself to mTORC1 and active.
• Amino acids
Some amino acids (mainly leucine (71) influence the activity of mTOR when leucine is present in the cellular environment, it is transported into the cell by α glutamine-dependent mechanism and, once in the cell, it promotes the binding of Rag protein (a GTPase) to RAPTOR, which leads to the interaction of TORC1 with Rheb.
• The energy status
The energy status of the cell, determined by the intracellular ATP/ADP ratio, may also influence the activity of mTOR. Indeed, a decrease in the ATP/ADP ratio leads to the activation of AMP-kinase (AMP-activated protein kinase) which inhibits mTORC1 by phosphorylating TSC2.
25
Hypoxia can inhibit mTORC1 by promoting the expression of REDD 1 ("DNA damage response 1 "). REDD 1 activates mTORC1 by inactivating TSC2 (70).
Activation of mTORC1 can be measured by the phosphorylation rate of the p70S6K1 protein for "p70 ribosomal S6 kinase 1" or 4EBP-1 for "translational inhibitor eukaryotic initiation factor 4E binding protein 1". Depending on the stimulus responsible for the activation of mTORC1, various consequences, such as the inhibition of autophagy, the activation of lipid metabolism, protein synthesis, or mitochondrial biogenesis, but the main consequence of the activation of mTORC1 is the control of growth and cellular differentiation.
Activation of mTORC2 leads to the inactivation of Foxo1 transcription factors for "Transcription factor Forkhead box O1" and Foxo3a. The members of the family of factors of Foxo transcription activate the transcription of genes involved in apoptosis and cell cycle. mTORC2 is also involved in the reorganization of the actin filaments of cytoskeleton by activation of PKC-α for "protein kinase C alpha" or RHO for "small GTPase RAS homolog ". Activation of mTORC2 can be measured by the level of phosphorylation from AKT to Ser473 (70).
1.2.9. Inhibition of mTOR by Rapamycin
Rapamycin inhibits mTORC1 by forming a complex with FKBP12 for "immunophilin FK506-binding protein 1A, 12 kDa". The rapamycin-FKBP12 complex binds to a kinase domain (FRB for "FKBP12-rapamycin-binding domain") of mTOR, preventing RAPTOR binding to this and hence inhibits the activity of mTORC1 (72). Contrary to mTORC1, mTORC2 appears to be resistant to direct inhibition by rapamycin (73). Inhibition of mTORC1 by rapamycin may have different effects on differentiation, proliferation, and function of various cell types (74).
1.2.10. Rapamycin and T lymphocytes
Inhibition of proliferation of T lymphocytes by stopping their progression in the cellular cycle (75).
26
Reduction in the expression of transcription factors necessary for the differentiation of T lymphocytes naive to effector T lymphocytes (TH1, TH2, or TH17) (76).
Differentiation of CD8 + T-lymphocytes with memory effectors in a viral infection (77).
Decreased expression of homing molecules such as CD62L and CCR7 that prevents
migration of effector T lymphocytes to target tissues (78).
Increased frequency of CD4+ CD25+ Foxp3+. Unlike T lymphocytes Rapamycin does not
affect the proliferation or immunosuppressive activity of Rapamycin.
Treg but on the contrary favors their relative expansion in vitro (79) and in vivo (70). Valmori et al. have suggested that the increase in the proportion of CD4+ CD25+ Foxp3+ in the presence of Rapamycin was due to the acquisition of a regulatory phenotype by peripheral CD4+ T lymphocytes (iTreg) than by an expansion of Treg (80). Currently, the mechanisms of this conversion are still poorly understood. This difference in sensitivity to the Rapamycin between Treg and effector T-cells suggests that they not have the same signaling path after stimulation of their TCR. Indeed, the Treg do not use the signaling pathway PI3K-Akt-mTOR but the signaling pathway IL-2R-STAT5 Pim2-dependent. Pim 2 is a serine / threonine protein kinase constitutively expressed by Treg but not by effector T lymphocytes, of which the expression is induced by the Foxp3 transcription factor.
In 2008, Basu et al. have shown that the resistance of Treg to Rapamycin intimately related to the expression of Pim 2 in the presence of a sufficient concentration of IL-2
(81).
More recently, Procaccini et al. have shown that Tregs express leptin and receiver. They observed that upon neutralization of this hormone or the deletion of gene encoding its receptor, the activity of mTOR in the Tregs was decreased, suggesting a close link between autocrine leptin secretion and activation of mTOR in Treg. They also demonstrated that activation of mTOR by leptin retained anergy of Treg. Indeed, it has been observed that the neutralization of leptin or mTOR inhibition induced proliferation of Treg in the presence of TCR (82).
27
1.2.11. Rapamycin and B lymphocytes
The proliferation of the B lymphocytes passes through the PI3K-mTOR pathway after stimulation of their BCR. Rapamycin inhibits the proliferation of B lymphocytes, their production of cytokines and their differentiation into plasma cells (83, 84).
1.2.12. Rapamycin and dendritic cells (DC)
Rapamycin has different effects on differentiation, maturation, and function of DC. The drug interferes with capture and presentation of antigen as well as with the production of cytokines and the expression of chemokines (85). The effect of mTOR inhibition by rapamycin largely depends on the context in which the DC are (In the case of a viral infection or an allograft) (86). This ability to inhibit the proliferation of DC and effector T lymphocytes and to promote Treg expansion makes rapamycin in combination or not with other immunosuppressants one of the therapeutic strategies of choice in the induction of immunological tolerance during allografts (87, 88) and prevention of GVHD (89-92) and in the prevention of autoimmune diseases (70; 93, 94).
1.2.13.The effects of 5-azacytidine on the function and number of
regulatory T cells and T-effectors
DNA methyltransferase inhibitors, particularly 5-azacytidine (5-Aza), have been shown to increase the survival of patients with GVHD. It is not entirely clear whether this improvement in patients’
28
survival is related to the effects of DNA methyltransferase inhibitors on the immune system and/or the direct effect of these drugs on the dysplastic clone. Goodyear and colleagues show that 5-Aza after allogeneic stem cell transplantation (allo-SCT) can increase the number of regulatory T cells (Tregs) while inducing a cytotoxic CD8+ T cell response, suggesting a potential mechanism for augmenting the graft-versus-leukemia (GVL) effect without increase in graft-versus host disease (GVHD) (95). 5-Aza, is an analog of cytidine, functions as a DNA methyltransferase inhibitor and has shown substantial potency in reactivating epigenetically silenced tumor suppressor genes (96). The antineoplastic activity of 5-Aza is thought to be mainly due to incorporation into RNA with disruption of RNA metabolism, and inhibition of DNA methylation. However, the precise mechanism by which this agent exerts an antitumor effect remains unknown, and it appears that clinical responses are influenced both by epigenetic alterations and by apoptosis induction. In addition, recent evidence suggested that 5-Aza can significantly impact some important immune functions via epigenetic modifications, making it an attractive candidate for pharmacologic manipulation of the immune system. (97). Moreover, 5-Aza administration induced a CD8+ T cell response to 1 or more tumor-specific peptides in more than half of the patients who had received 3 or more cycles of azacytidine 1. In patients who could receive at least 6 cycles of 5-Aza, a CD8+ CTL response to a range of tumor antigens including MAGE-A1, MAGE-A2, MAGE-A3, BAGE-1, RAGE-1, and WT1 could be documented both in the peripheral blood and bone marrow. From the functional standpoint, the detectable tumorspecific responses were also active in response to peptide (98).This is, to our knowledge, the first demonstration that azacitidine has the capacity to increase both circulating Tregs and CD8+ tumor-specific T cell responses in humans in the context of allo-SCT, without increasing the incidence of acute and chronic GVHD compared to both control patients and previous studies using a similar conditioning regimen. The strengths of this study are its prospective nature, inclusion of a fair number of homogenously treated allo-SCT AML patients, and use of state-of-the-art, modern immune monitoring tools. The combined beneficial effect of azacitidine on both Tregs and CD8+ T cells represents an ideal scenario after allo-SCT toward induction of tolerance and improving disease control. Obviously, the role and impact of Tregs in GVHD remains controversial. However, recent convincing evidence showed that sustained Treg expansion in vivo correlated with the amelioration of
29
the manifestations of active chronic GVHD (99). Thus, administration of 5-Aza after transplantation is an appealing and potentially interesting alternative strategy for Treg expansion. On the other hand, the true clinical relevance of CD8+ T cell responses to tumor antigens after allo-SCT is yet to be established. However, it is an attractive mechanism by which the GVL effect might be augmented after allo-SCT. Furthermore, one could anticipate that the use of 5-Aza in combination with DLI may further augment the safety and antitumor effect of DLI. Such trials are currently ongoing. The exact mechanisms underlying the effects of azacytidine after allo-SCT are still to be deciphered. It is possible that 5-Aza can regulate several genes and pathways that are involved in regulation of T-cell differentiation. For instance, epigenetic regulation of some cytokine genes is a key event in the initiation of immune responses (100). In conclusion, these far-reaching findings suggest that 5-Aza administration after allo-SCT is feasible and can allow for Treg expansion, as well as an impressive induction of CD8+ T-cell responses to tumor antigens without GVHD. It raises the possibility of therapeutically separating GVHD and GVL. While the optimal dosing, schedule timing, and duration of azacytidine administration after allo-SCT are still to be established (101), and further refined understanding of molecular events associated with the immunomodulatory effects of novel epigenetic therapies is needed, this agent could eventually allow selective targeting of the GVHD and GVL pathways.
30
Graft versus host disease is one of the major complications after allogeneic CSH. Many therapeutic strategies are and have been considered to prevent and / or treat this disease. Baron,F. et al have developed a model of xenogeneic GVHD (xGVHD), by infusing human peripheral blood mononuclear cells (PBMNC) into NOD-scid IL-2Rγnull (NSG) mice (102). In that model he severity of GVHD was closely correlated with the dose of PBMC infused. More recently, they investigated the ability of clinical-grade enriched human Treg to attenuate experimental xGVHD induced by PBMNCs (autologous to Treg) infusion in NSG mice, as well as verified their inability to induce xGVHD when infused alone (103). We investigated whether 5-azaC and RAPA could be used together to promote the ex vivo expansion of Tregs purified from adult human peripheral blood to better understanding Treg homeostasis and expansion in vitro and assessing the impact of different drugs combination in Treg subsets for GVHD (xGVHD) prevention/ treatment.
3. Materials and Methods
3.1.Mononuclear cell separation
Mononuclear cells were separated from peripheral blood by density gradient sedimentation (LSM 1077 Lymphocyte; PAA Laboratories GmbH, Austria). Also peripheral blood mononuclear cells (PBMC) were frozen in serum-free cryopreservation medium (CryoMaxx I, PAA Laboratories GmbH, Austria) and stored in liquid nitrogen.
31
3.2. Cell preparation
Human peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll gradient (Sigma, St. Louis, MO). After purification CD4+ CD25+ and CD4+ CD25- T cells received Rapa and/or 5-Aza treatment for 14 days.
CD4+CD25+Treg cells populations were isolated from PBMC using a CD4+CD25+Treg Cells Isolation Kit with according to the manufacturer's protocols (Miltenyi, Bergisch Gladbach, Germany). Briefly, PBMC were incubated with a Biotin-antibody cocktail against: CD8α (Ly2), CD11b (Mac-1),CD45R (B220),CD49B (DX5) and Ter-119, for 20min at 4°°C. The cell suspension was loaded on a LD column, which is placed in magnetic field of a MACS Separator. The remaining fraction in the column is the enriched CD4+ cells. For the isolation of CD4+CD25+ cells, the CD25+ cells in the enriched CD4+ cells fraction were magnetically labeled with anti-CD25 MicroBeads. The magnetically labeled CD4+CD25+ cells were enriched from the CD4+ cells fraction by MACS Separator. The purities of the sorted CD4+CD25+ populations were always 95% as confirmed by flow cytometry (FCM).
3.3. Antibodies, reagents, and flow cytometry
At least 1×106 T cells were initially labeled for dead cells with the 7-Aminoactinomycin D (7 AAD)
(Becton-Dickinson, UK) and antibodies against surface antigens, anti-CD3 V500, anti-CD4 V450 or PerCP Cy5.5, anti-CD25 APC, anti-CD45RA V510, anti-CD8 PE, anti-HLA DR APC Cy7, were purchased from BD Bioscience (USA). Anti-Bcl2 PE, IFNγ PECy7, Annexine V FITC, anti-CD127 V510, andti-IL-10 PE anti-Ki67 APC, anti-CD45 murine PE, anti-IL-17 APC, anti-TNFa APC Cy7, anti-IL-4 PE were purchased from eBioscience (USA).
For Treg staining we used the anti-human FOXP3 FITC conjugate after fixation and permeabilization according to manufacturer’s instructions (eBioscience, San Diego, CA, USA). For cytokines staining cells were stimulated for 4 h with phorbol myristate acetate andionomycin in the presence of brefeldin A. The stimulated cells were fixed and permeabilized prior to intracellular staining according to the manufacturer’s instruction. Flow cytometry was performed on a FACSCantoII using FACSDiva
32
software (Becton Dickinson). Data were analyzed on FlowJoVX software (Tree Star,Ashland, OR, USA).
3.4. Isolation of regulatory T cells by FACS ARIA
To obtain CD4+CD25highCD27+ Treg, PBMC were first enriched for CD4+ T cells using a negative isolation kit (Miltenyi-Biotec Ltd., Surrey, UK) and stained with anti-human CD4, CD25 and CD27. Purified Treg, defined as CD3+CD4+CD25highCD27+, were sorted on a FACSAria (Becton-Dickinson)
3.5. RNA isolation and quantitative real-time polymerase chain reaction
Total RNA was isolated from Treg and total CD4+ T-cells obtained from healthy donors, using a combination of TriZOL (Invitrogen) to lyse the cells and RNAeasy Mini prep(Qiagen) to purify the RNA. The quantity and integrity of RNA were evaluated by measuring the absorbance at 260 nm with a Nanodrop ND8000 (Labtech) and bioanalyzer (Agilent). cDNA was synthesized from 1 μg of RNA using the RT cDNA kit (Invitrogen) according to the manufacturer’s protocol. Real time PCR were performed with specific primers and Taqman probe for IL-17A, CCR7, CTLA-4,TGFβ, IFNγ, STAT5 and CD25 genes acquired by Applied Biosystems. Probes with the following Applied
Biosystems assay identification numbers were used: CTLA4,Hs00175480_m1;
CCR7,Hs01013469_m1 ; STAT5 Hs00560026_m1;CD25, Hs00907777_m1; IL17, Hs00907777_m1; IFNg Hs00989291_m1; and TGFb. The relative gene expression was analyzed in relation to TBP, Hs00427620_m1 housekeeping gene levels.
3.6. Methylation assay
A methyl-sensitive PCR was used to analyze the methylation status of the CpG island of the Foxp3 enhancer using MS-qPCR as previously described (104). Genomic DNA was prepared using PureLinkTM Genomic DNA Mini Kit (Invitrogen, California, USA) and quantified by NanoDrop 1000 V3.7.1 (Thermo Fisher Scientific, Massachusetts, USA). 150ng of genomic DNA isolated from
33
murine splenocytes was digested with 10 U of HpaII or MspI enzymes overnight at 37°C. The Foxp3 CpG island and B-act were PCR amplified using the following specific primers: Foxp3 enhancer CpG
island, 5'_ATCCTCGCCATCGTCTTCCTCAT_3' (forward) and
5'_CCTGTTCTGGCTTTCTCATTGGCT_3' (reverse); IL2A promoter CpG island,
5'_GTAACTGCTGTGTGACTTGT_3'(forward) and 5'_TTTCACTTCCCTGGTGAGG_3' (reverse);
IFNg promoter CpG 5'_ GGCAGAAGACACGCGAATAG_3'(forward) and
5'_TGCTCATGCTGTTTCTTTGG_3' (reverse); IL17 promoter CpG island, 5'_
ACTTATATGATGGGAACTTGAG_3' (forward) and 5'_GCTATGCTATGGGTCAATATCA_3' (reverse); B-act (GenBank accession no. U89400), 5_TAGCACCATGAAGATCAAG-3_(forward) and 5_CCTGCTTGCTGATCCACAT-3_(reverse) Quantitative real-time PCR was performed using the LightCycler 480 thermocycler and the KAPA SYBR FAST qPCR kit, following manufacturer's instructions (Kapa Biosystems, Wilmington, USA,). Data were normalized using the b-Act Endogenous Control and B-actin specific primer for uncut region by HpaII and MspI.
3.7. Functional analysis of regulatory T cells
Expanded cells were assessed for suppressive function between 10 and 14 days of culture. 1 × 104
CFSE (1μM Invitrogen)-labeled responder CD25- CD45RA+CD4+ T cells were cocultured with 1 × 104 unlabeled cells assessed for their suppressive capacity. Cells were stimulated with 1μg/mL anti-CD3 (OKT3 mAb) and 1μg/mL anti-CD8 (15E8) in 96-well round-bottom plate in RPMI medium supplemented with 10% fetal bovin serum (Bio West), 2mM L-glutamin, 1mM sodium pyruvate,1% non essential amino acid MEM, 100U/mL penicillin, 100 μg/ml streptomycin and amphotericin B (all from Gibco). Proliferation of CFSE-labeled cells was assessed by flow cytometry after 84-90 hr of culture. Percent suppression was calculated as follows: [1 - (number of proliferating CFSE diluting responder cells in the presence of suppressor cells at a 1 to 1 ratio / number of proliferating responder cells when cultured alone)] x 100.
34
3.8. Statistical analysis
A paired samples t-test was used to compare paired samples treated in-vitro with 5-azaC and/or with Rapa. P values <0.05 were considered statistically significant. The search for statistically significant differences between means of different variables was performed by paired Student’s t-test. Calculations were performed using the GraphPad Prism version 5.0 software. The analysis of
quantitative Real Time PCR data was based on 2-CT values.
4. RESULTS
4.1.
Responses
of
Treg
(CD4+CD25+Foxp3+)
and
conventional
(CD4+CD25neg) T cells to RAPA and/or 5-Aza in the presence of
CD3/CD28 Ab crosslinking and IL-2.
We analysed absolute cell number of CD4+CD25+and CD4+CD25neg T cells expanded in vitro in the presence of RAPA and/or 5-Aza for 2–3 weeks. Our initial experiments indicated that RAPA and 5-Aza can selectively promotes expansion of CD4+CD25highFoxp3+human Treg isolated from the CD25high T cell subset used in our experiments was previously shown to consist largely of Foxp3+cells (105-107). in the absence of RAPA, and 5-Aza, (control culture) we observed a minor frequency of CD4+CD25+ cells. In contrast, the parallel RAPA and 5-Aza cultures , were enriched in CD4+ CD25high T cells (Figure 5B). In the Rapa and 5-Aza cultures, CD4+CD25+and CD4+CD25high T cells were the expanding cell populations. In contrast, control cultures largely contained proliferating CD4+ CD25neg T cells (Figure 5A). These results suggested that RAPA inhibited expansion of CD4+ CD25neg cells while promoting proliferation of CD4+CD25+T cells. Also we observed CD4+CD25neg T cells before and after their culture in the presence of RAPA and/or 5-Aza for 2–3 weeks, and we found that in control culture there was a higher number of CD4+CD25neg T cell than in RAPA and/or 5-Aza culture (Figure 5A,5B).
35
A
B
Figure 5 Growth curve of CD4+ T cells.A. In-vitro effects of 5-AzaC and/or Rapa on CD4+CD25+ T cells.
The proliferative capacity of CD4+CD25+ T cells was assessed following 14 days of stimulation in the presence of AzaC only (1 μM), in presence of Rapa only (100nM), and in presence of both Rapa and 5-AzaC.The date are mean fold expansion ± SD from 3 indipendent experiments. Note the highest fold proliferation of CD4+CD25+ T cells in the presence of 5-azaC in both culture with Rapa or without Rapa (P<0.001), and in culture with Rapa only compared to control culture (p<0.05). B . In-vitro effects of 5-AzaC and/or Rapa on CD4+CD25- T cells. Expansion into 2 weeks of CD4+CD25- Tcells was observed in presence of 5-AzaC only (1 μM), in presence of Rapa only (100nM), and in presence of both Rapa and 5-AzaC.The date are mean fold expansion ± SD from 3 independent experiments. The expansion of CD4+CD25- T cells decrease significantly in the presence of 5-AzaC compared to the control culture, Rapa culture and culture with Rapa and 5-AzaC (P<0.05)
4.2. Apoptosis of CD4+CD25neg T cells in the presence of RAPA and/or
5-Aza.
To analyse whether RAPA and/or 5-Aza inhibit selectively inducing apoptosis of CD4+CD25neg T cells, we examined ANX V binding to CD4+CD25+and CD4+CD25neg T cells before and after their culture in the presence of RAPA and/or 5-Aza for 2 weeks. Fresh MACS-separated CD4+ CD25neg cells failed to bind ANX V, while 7% of autologous CD4+ CD25+T cells were ANX V+ (Figure 6B, 6A). This result suggests that these CD4+CD25+T cells were undergoing activation induced cell death (AICD). In contrast, ANX V binding to CD4+ CD25neg cells cultured in the presence of RAPA
36
(+R) for 2 weeks was very high (44%), while significantly lower ANX V binding to CD4+ CD25neg cells (p<0.0001) was evident in proliferating R0 cultures (Figure 6A,). As also shown in Figure 2B for a representative culture, significantly higher percentages of ANXV+CD4+CD25neg T cells were observed in 2 week Rapa and 5-Aza cultures than of CD4+ CD25high T cells (79% vs. 5%). Analogous data were obtained when these cell subsets were cultured with 5-Aza, or with RAPA for 2 weeks (Figure 6A, 6B). In addition to ANX V binding, anti-apoptotic Bcl-2 protein production was evaluated by flow cytometry staining in CD4+ CD25high and CD4+ CD25neg cell subsets +RAPA and/ or 5-Aza (Figure 7). High level of Bcl-2 production was observed only in CD4+ CD25neg cells +RAPA and 5-Aza. These results suggest that apoptosis-sensitive CD4+ CD25high T cells become resistant to apoptosis following TCR-mediated+IL-2 activation and subsequent culture in the presence of RAPA or RAPA and 5-Aza. In contrast, CD4+ CD25neg T cells become highly sensitive to apoptosis upon TCR and IL-2-mediated activation in the presence of RAPA and 5-Aza and die.
37
A B
Figure 6. The phenotype and sensitivity to apoptosis of CD4+CD25- and CD4+CD25+T cells expanding with Rapa, 5-AzaC, or both Rapa and 5-AzaC. A. Note that after 2 weeks of culture with RAPA, or together
with Rapa and 5-AzaC the greatest part of cells CD4+CD25-binding Annexin V protein compared fresh cells. B. Percentages of ANX V binding to fresh, uncultured MACS and FACS ARIA-isolated CD4+CD25+, and of ANX V binding to expanded CD4+CD25+ T cells after 2 weeks in culture with RAPA, or/and 5-AzaC or in medium culture only. Note that only few CD4+CD25+ cells expanding in cultures with RAPA or/and with 5-AzaC bind ANX V, compared to CD4+CD25+ in control culture with no Rapa and no 5-5-AzaC. Data are representative of 3 indipendent experiments.
38
Figure7. The phenotype analysis of Bcl-2+ cell population on CD4+CD25+ or CD4+CD25. Expression
levels the anti-apoptotic Bcl-2 proteins in CD4+CD25- (A) or CD4+CD25+T cells (B) measured before and after 2 weeks of culture in the presence of RAPA, or 5-AzaC or both Rapa and 5-AzaC. T cell subsets were separated by MACS and by FACS ARIA from PBMC of 3 different buff coat and cultured as described above. These are representative figures of 3 indipendent experiments
