Regulatory T cells: tolerance induction in solid organ transplantation

Summary

The concept of regulatory T cell (T_reg) therapy in transplantation is now a reality. Significant advances in science and technology have enabled us to isolate human T_regs, expand them to clinically relevant numbers and infuse them into human transplant recipients. With several Phase I/II trials under way investigating T_reg safety and efficacy it is now more crucial than ever to understand their complex biology. However, our journey is by no means complete; results from these trials will undoubtedly provoke both further knowledge and enquiry which, alongside evolving science, will continue to drive the optimization of T_reg therapy in the pursuit of transplantation tolerance. In this review we will summarize current knowledge of T_reg biology, explore novel technologies in the setting of T_reg immunotherapy and address key prerequisites surrounding the clinical application of T_regs in transplantation.

OTHER ARTICLES PUBLISHED IN THIS REVIEW SERIES

Immune tolerance in transplantation. Clinical and Experimental Immunology 2017, 189: 133–4.

Transplantation tolerance: the big picture. Where do we stand, where should we go? Clinical and Experimental Immunology 2017, 189: 135–7.

Operational tolerance in kidney transplantation and associated biomarkers. Clinical and Experimental Immunology 2017, 189: 138–57.

Immune monitoring as prerequisite for transplantation tolerance trials. Clinical and Experimental Immunology 2017, 189: 158–70.

Transplantation tolerance: don't forget about the B cells. Clinical and Experimental Immunology 2017, 189: 171–80.

Murine models of transplantation tolerance through mixed chimerism: advances and roadblocks. Clinical and Experimental Immunology 2017, 189: 181–9.

Chimerism‐based tolerance in organ transplantation: preclinical and clinical studies. Clinical and Experimental Immunology 2017, 189: 190–6.

Introduction

Innovations in surgery and therapeutics have led to remarkable advancements in solid organ transplantation. Despite this, allotransplantation is threatened by the prospect of graft rejection. While there have been improvements in averting acute rejection and supporting short‐term graft survival, there has been an accelerated rise in morbidity and mortality associated with the relative toxicity of immunosuppression regimens and chronic rejection 1, 2, 3, 4, 5. With a growing transplant waiting‐list, means by which we can ensure the prosperity of a transplanted organ are highly sought after.

A proposed panacea to this issue has been the induction of immunological tolerance: a concept centred on the manipulation of immune cells so as to design an immunological environment favouring regulation, supporting graft survival, without interfering with crucial immune surveillance mechanisms and, concurrently, abolishing immunosuppressant‐related toxicity. Indeed, there have been reports in the literature of a state of spontaneous operational tolerance in liver and renal transplant recipients 6, 7, 8. In such cases individuals forgo the need for therapeutic immunosuppressants, while synchronously maintaining normal graft function. However, this phenomenon is rare, and in the setting of liver transplantation has been shown to occur late after transplantation 9. Furthermore, ongoing immunosuppressant therapy has been thought to hinder the development of operational tolerance through its indiscriminate clearance of immune cells thought necessary for tolerance induction. Importantly, however, research has suggested that operational tolerance can be predicted before the cessation of immunosuppression through the detection of distinct immune signatures and clinical parameters, which will prove invaluable when devising future tolerance‐inducing protocols 6, 8, 10.

In the ongoing pursuit for the Holy Grail of transplant immunology, considerable effort has been directed on the search for regulatory immune cells. During the past few decades there have been countless discoveries of immunoregulatory cells spanning all lineages of immunity; arguably, however, one particular family that has received an abundance of attention has been regulatory T cells (T_regs).

T_regs are known for their importance in mediating immune tolerance to self‐antigens with their impairment linked to the development of organ‐specific autoimmune diseases, including type 1 diabetes (T1D) mellitus, autoimmune hepatitis and various haematological disorders 11, 12.

In the setting of transplantation, these cells have undoubtedly proved their worth. Animal models conducted from the point of their discovery have demonstrated their abilities to promote transplantation tolerance 11, 13, 14, 15. Furthermore, studies investigating the phenomenon of operational tolerance in liver transplantation have revealed that operationally tolerant individuals exhibited a significantly higher number of circulating and intrahepatic T_regs compared to non‐tolerant or healthy individuals 16. In addition, in both kidney and lung transplantation recent studies have shown how stable graft function is associated with increased circulating and intra‐organ T_regs 17, 18.

Having alluded to their merit in the field of tolerance induction, the biology of T_regs has come under intense investigation, with the aim of translating their apparent tolerogenic facets through to the clinic. In this review we will summarize current knowledge on these cells, prerequisites surrounding their clinical application in solid organ transplantation and recent advancements in the field.

Regulatory T cells

Phenotype

T_regs exist as 5–10% of circulating CD4⁺ cells and are characterized by their high and stable expression of the interleukin (IL)‐2 receptor α chain, CD25, and the master transcription factor forkhead box protein 3 (FoxP3), crucial to their development and maintenance of function. In humans FoxP3 exists in two isoforms: the full‐length protein, consisting of 11 exons and a shorter isoform which lacks exon 2, both of which are expressed at equivalent levels 19, 20. Mutations in the FoxP3 gene have been linked with the development of immune dysregulation, polyendocrinopathy, enteropathy and X‐linked (IPEX) syndrome, a multi‐system autoimmune disease affecting the intestines, endocrine gland and the skin 21, 22. While FoxP3 expression in mouse T_regs is confined purely to the T_reg compartment, unfortunately in humans its expression is less faithful. It has been reported that CD4⁺CD25^– T effector (T_eff) cells can express the transcription factor transiently without displaying signs of regulation 23. Furthermore, its intracellular location makes the identification of T_regs through this marker unfeasible, requiring cell permeabilization to confirm its expression. As a result, until recently T_regs were grossly identified as CD4⁺CD25^hi cells.

More recent work has identified an inverse relationship between the surface antigen CD127, the α chain of the IL‐7 receptor, and FoxP3 expression 24, 25, with data demonstrating refined suppressor potency in CD4⁺CD25⁺CD127^low T_regs compared to T_regs classified by CD4⁺CD25⁺ expression alone 24, 26, 27. As such, demarcation of a bona fide population of human Tregs is now based on the markers CD4⁺CD25⁺CD127^low.

Miyara et al. compartmentalized T_regs further into three main subsets based on their expression of CD45RA and FoxP3. CD45RA is deemed to demarcate naive suppressive FoxP3⁺ cells (CD45RA⁺) present in the cord blood and in adult blood. T_regs were classified into: CD45RA⁺FoxP3^high resting T_regs (rT_regs), CD45RA^–FoxP3^high activated T_regs (aT_regs) and CD45RA^–FoxP3^low non‐suppressive T_regs. With regard to the latter population of T_regs described by Miyara et al., their initial ‘non‐suppressive’ identity has come under scrutiny. While data have corroborated their potential to differentiate into IL‐17‐producing cells 28 in association with CD161 expression, findings from our laboratory and others have suggested that this population do indeed possess suppressor capabilities in vitro 29, 30.

The phenotypic and functional diversity of human T_regs has been highlighted further by Duhen et al., who identified distinct populations of T_regs cells in human blood based on the expression of chemokine receptors. Additional analysis revealed that while each population was functionally suppressive they displayed differing expressions of lineage specific transcription factors and an array of cytokine production 31.

T_regs have also been classified in accordance with their origin: naive thymus‐derived T_regs (tT_regs), peripheral antigenically stimulated T_regs (pT_regs) and in‐vitro‐driven iT_regs. While in mice the expression of neuropilin‐1 on tT_regs can differentiate between pT_regs, in human counterparts this marker has not yet been identified 32, 33. More recently, the expression of the Ikaros transcription factor, Helios, on tT_regs has now potentially allowed for the phenotypical distinction between the two populations; however, its validity remains contentious 34, 35

To date, the only reliable way to distinguish between the subsets is through the analysis of epigenetic DNA methylation levels on the non‐coding conserved region of the FoxP3 gene termed the T_reg‐specific demethylation region (TSDR) 36. In conventional T_eff cells this region is fully methylated. tT_regs, however, exhibit a fully demethylated TSDR, whereas pT_regs only a partially demethylated region 36, 37, 38.

Indeed, these identified T_reg populations are by no means comprehensive 39. With the progression of technology we have discovered that T_regs are, in fact, far more heterogeneous than we imagined initially. The dawn of mass cytometry has brought with it the potential to characterize T_reg populations based on more than 40 different markers 40, 41. Here, by using cytometry time‐of‐flight (CyTOF) technology, where each cell is analysed through a multi‐dimensional deep phenotype, the complexity of the T_reg subsets has been laid bare. Our recent work has highlighted the intricacy of the T_reg compartment in healthy individuals, revealing 22 distinct subpopulations 42. Furthermore, recent work has proposed that results from this intense phenotypical interrogation can predict novel biomarkers for determining treatment success. Following CyTOF immune phenotyping of T_regs, Kordasti et al. were able to identify an immune signature in individuals with aplastic anaemia, which predicted the response to immunosuppressive therapy 43.

Mechanism of action

T_regs have been reported to exert their suppressive function through several different means (Fig. 1); indeed, the most reliable way of classifying T_regs would be through their function. Unfortunately, however, the expression of various phenotypical markers does not always correspond with the method of suppression utilized by a specific T_reg. With regard to the specific mode of suppression, T_regs have been described to possess several cell‐contact‐dependent and ‐independent mechanisms, ranging from cytokine release, receptor endocytosis and purinergic signalling to cell cytotoxic mechanisms (Fig. 1).

Figure 1.

Mechanism of regulatory T cell (T_reg) suppression. Contact‐independent: (a) anti‐inflammatory cytokine production. The secretion of anti‐inflammatory cytokines such as interleukin (IL)‐10, IL‐35 and transforming growth factor (TGF)‐ β has been linked with inhibition of T cell activation in vivo. (b) Exosomes: transfer of miRNA through extracellular vesicles of endosomal membranes can silence specific genes in T cells preventing cytokine production/proliferation. Contact‐dependent: modulation of antigen‐presenting cell (APC) maturation and function. The interaction of cytotoxic T lymphocyte antigen (CTLA)‐4 on T_regs with its ligand CD80/86 on APCs, delivers a negative signal for T cell activation. CTLA‐4's mechanism of action is varied, including: the capture of its APC expressed ligands and subsequent trans‐endocytosis and also the up‐regulation of indoleamine 2, 3‐dioxygenase (IDO) and the generation of kynurenines. Induction of apoptosis. T_regs have the capacity to induce apoptosis directly via granzyme A/B and perforin, inducible cyclic adenosine monophosphate (cAMP) early repressor (TRAIL), the Fas/Fas‐ligand pathway, the galectin‐9/T cell immunoglobulin and mucin domain‐3 (TIM‐3) pathway. Disruption of metabolic pathways. The ectoenzyme CD39, expressed on T_regs, result in the metabolism of adenosine triphosphate (ATP) to adenosine monophosphate (AMP), in turn producing the immunoregulatory purine, adenosine. T_regs have also been found to express high levels of intracellular cAMP. This is transferred to T effector cells through gap junctions, which leads to the up‐regulation of inducible cAMP early repressor (ICER) and in turn the inhibition nuclear factor of activated T cells (NFAT) and IL‐2 transcription leading to apoptosis by IL‐2 deprivation. High expression of IL‐2 receptor alpha chain, CD25, depletes surrounding IL‐2. DC = dendritic cells.

More recently, the role of exosomes in immunoregulation has been described. Exosomes are extracellular vesicles formed from endosomal membrane that carry various proteins, cytokines and microRNA 44, 45. Our laboratory first described their release from murine T_regs following activation 46. Furthermore, the ectonucleotidase CD73 was shown to play a significant role in mediating exosome‐related suppression through the purinergic generation of adenosine 46. Additionally, the transfer of miRNA in mechanisms of T_reg suppression has been described. Here, T_reg‐derived miRNA, in particular Let‐7d, transported in exosomes, have been shown to prevent T effector cell proliferation and cytokine secretion through specific gene silencing 47.

While no single predominant mechanism of suppression has been identified as of yet, it could be probable that T_reg suppression is, in fact, reliant upon several mechanisms operating at once.

T_reg clinical use: manufacture and considerations

T_reg isolation

The majority of clinical trials to date have looked to use adult peripheral blood as their source of T_regs. However, there have been studies investigating T_reg isolation from umbilical cord blood (UBC). It has been reported that approximately 5 − 7·5 ×10⁶ T_regs can be isolated from one unit of UBC 48. Furthermore, CD25^high T_regs are purified more readily from this source, and in addition are virtually devoid of CD25⁺ antigen‐experienced memory T cells 49, 50, 51.

More recently, T_reg isolation from discarded paediatric thymuses, resected routinely following cardiac surgery, has revealed a novel source for T_reg isolation 52. T_regs isolated from this source are in utter abundance. Here, 1 g of thymus yields T_regs in the order of 500 : 1 in comparison to 1 ml of peripheral blood. These isolated T_regs demonstrate stable properties and long telomeres, correlated with improved replicative capabilities, functional potency and in‐vivo survival. This source may prove invaluable in future trials of T_reg therapy, in the paediatric population in particular, given trials faced with isolated T_regs from blood/UBC 52.

In order for the translation of T_reg therapy through to the clinic, protocols outlining the manufacture of T_regs need to be in place that comply with good manufacturing practice (GMP). Because of the wealth of markers defining different populations of T_regs, much debate has been centred upon the chosen markers for T_reg isolation. Until only recently T_reg isolation for cell therapy has been limited to using the CliniMACs (Miltenyi Biotec, Bisley, UK) system, based on the selection of T_regs through a two‐step magnetic bead isolation. Methods have involved initial depletion of CD8⁺/CD19⁺ cells, followed subsequently by CD25 positive selection 53. However, this technique does not allow for T_reg selection based on multiple parameters, limiting its use for selection of T_regs with specific characteristics. Furthermore, this method is indiscriminate when it comes to selecting markers with broad expression patterns, and with the advent of the CyTOF system 42, 43 it may well be that disease‐specific optimal T_regs will be identified, with the potential for cell therapy application.

The concept of fluorescence‐activated cell sorting (FACS) has been acknowledged widely for many decades. However, it is only recently that this method of cell isolation has been deemed GMP compliant in the United Kingdom. FACS allows for cell sorting whereby each cell is interrogated on an individual level following fluorescent labelling. This method permits cell isolation based on several parameters. Because of its recent GMP accreditation it now opens up the possibility of T_reg isolation based on the highly researched markers of suppression, stability and specificity 54. While the concept of FACS isolation is shared, GMP‐certified machines used for this process of T_reg isolation differ around the world. Both the United States and Poland use the BD FACSAria™, Germany uses the BD Influx™ and the United Kingdom plans to use the MACSQuant^® Tyto, which is currently under validation.

One concern with isolating T_regs based on more stringent markers is the risk of obtaining poorer yields. Indeed, it has been hypothesized that sorting T_regs based on the high expression of CD25 will be too restrictive when considering the yield of cells required for ex‐vivo expansion. Putnam et al., however, devised a protocol for the FACS‐based isolation of CD4⁺CD25⁺CD127^low T_regs using the BD FACSAria II 55. Sorting based on these selected markers allowed for a greater overall yield through encompassing the spectrum of CD25‐expressing T_regs, and ensured the isolation of a pure T_reg population, > 95% FoxP3, through the inclusion of the marker CD127. Additionally, studies have highlighted the inclusion of isolating cells on CD45RA expression, demonstrating the superior purity and suppressive capabilities of CD45RA⁺ T_regs following isolation and culture 55, 56. However, one concern regarding this population is its waning with age 28. A recent study by our laboratory, however, demonstrated the feasibility of isolating and expanding this CD4⁺CD25⁺CD127^lowCD45RA⁺T_reg population in patients with inflammatory bowel disease, reiterating their superior function and stability 57. This has led subsequently to the planning of a clinical trial of T_reg immunotherapy in the setting of Crohn's disease at King's College London (TRIBUTE).

T_reg expansion

T_regs exist in relative paucity in the peripheral blood. The rationale behind tolerance induction stands to tip the balance in favour of regulation by the in‐vivo increase in T_reg numbers over T_effs. Extrapolated data from mouse models, where T_regs have been co‐infused with T_eff to determine efficacious ratios for tolerance, have suggested anywhere between 1 : 2–5 : 1, T_regs : T_eff 58, 59, 60. Therefore, where T_regs currently exist at 5–10% of circulating CD4⁺ T cells it has been suggested that this T_reg pool needs to be increased by a minimum of 33% to prevent transplant rejection 61. This requires the substantial expansion of the T_reg pool for clinical efficacy; as such, the feasibility of adoptive cell therapy is reliant upon protocols for the ex‐vivo expansion of T_regs to numbers needed for their clinical application.

T_regs can be expanded in vitro using polyclonal stimulation with bead‐bound or soluble anti‐CD3 and anti‐CD28 monoclonal antibodies concomitantly with high‐dose IL‐2 55, 62. To date, the GMP‐compatible protocols have been reliant upon the CliniMACS‐based isolation of the T_regs, the aforementioned of which can often be contaminated with T_eff cells. In these culture conditions T_effs will thrive in competition, leading to contamination of the final product. FACS‐sorting the starting product would circumvent this concern. However, there have been reports that even when starting with a highly pure population of T_regs repeated stimulation results in the loss of FoxP3 expression 63, 64, yet simply reducing the rounds of stimulation can often lead to insufficient overall T_reg yields 65. We and others have developed T_reg expansion protocols which ensure the purity of the final product, reaching clinically applicable numbers 62, 66, 67

Optimization of culture conditions has included the addition of the mammalian target of rapamycin (mTOR) inhibitor, rapamycin 68. This immunosuppressant has been shown to preferentially inhibit the proliferation and function of T_effs, as a result favouring T_reg expansion and stability, permitting the growth of T_regs even from a mixed starting population 62, 69, 70. In our recent publications we set to isolate and expand T_regs from patients with end‐stage renal and liver disease, with the aim of using these cells for adoptive immunotherapy post‐transplantation. T_regs were isolated by adopting the CliniMACS separation technique and the expansion of the cells for 36 days, in the presence of rapamycin, in the GMP Clinical Research Facility at Guy's Hospital. At final harvest we reported T_reg purities of more than 95%, reaching clinically applicable numbers for use in adoptive cell therapy and, importantly, showed the stability of the final T_reg product 62, 67. The clinical translation of this to date has seen 12 and five patients, respectively, having received the final product post‐renal/liver transplantation as part of the two‐dose escalation studies of T_reg adoptive transfer, ONE study (NCT02129881) and ThRIL (NCT02166177).

Stability

There is now growing evidence that T_regs are plastic with a potential to convert into proinflammatory cells 71. In agreement, the presence of a population of FoxP3⁺ T cells capable of secreting the proinflammatory cytokine, IL‐17, has been observed in human peripheral blood, thus questioning the stability of these cells. In this regard, the importance of assessing T_reg stability has been proposed, especially in view of the clinical application of these cells.

Despite this strict government of FoxP3 expression, emerging data suggest that T_regs can down‐regulate FoxP3 in the presence of inflammatory cytokines. In agreement, Yang et al. have shown that exposure of T_regs to IL‐6 and IL‐1 in vitro results in the expression of IL‐17 72. To determine the origin of the unstable T_regs, the extrinsic factors resulting in their instability and the fate of these ‘exT_regs’ is of particular importance.

In this regard, Hori et al. proposed a ‘heterogeneity model’, suggesting that FoxP3 expression does not necessarily determine T_reg lineage commitment. They showed that uncommitted FoxP3‐positive cells can lose FoxP3 expression, acquiring transient activation‐induced FoxP3 expression, converting to effector‐like ex‐FoxP3 cells under inflammatory conditions or committing to a T_reg fate (upon demethylation of TSDR) 73.

Of note, however, it is important to make the distinction between T_reg instability and functional specialization. Studies have also shown that while under inflammatory conditions T_regs can acquire the ability to produce effector cytokines while still maintaining high FoxP3 expression and suppressive activity. In support of this, it has been demonstrated that IL‐17 is produced by a subset of highly suppressive human T_regs that express CCR6, the chemokine receptor used by Th17 cells for their recruitment to sites of inflammation 74.

Moreover, our group and others have shown the expression of CD161, the killer cell lectin‐like receptor subfamily B, to be expressed on a subpopulation of human T_regs, that produce IL‐17 upon in‐vitro activation in the presence of IL‐1β, but not IL‐6. In addition, evidence has supported the suppressive capacity of these cells 30, 75.

These studies highlight that T_reg expression of effector cytokines alone cannot be viewed simply as a marker of plasticity or lack of stability, but may be a hallmark of their functional specialization. In support of this, Campbell et al. have shown that T_reg expression of transcription factors and cytokines specific for Th1, Th2 and Th17 enables them to control inflammation mediated by Th1, Th2 and Th17 cells by responding to the same environmental cues 76.

As the function of T_regs is highly dependent upon the constitutively high expression of FoxP3 77, many groups have sought to find ways to stabilize its expression. As discussed above, epigenetic modification of the FOXP3 locus has a major role in controlling FOXP3 transcription, with demethylation of key regions correlated with suppressive function and lineage stability 78. In this regard, in‐vitro treatment with demethylating agents such as azacytidine have shown to promote the stability of FoxP3 expression in T_regs, resulting in the potent ability of these treated cells to protect from graft‐versus‐host disease (GVHD) 79. In addition, a recent Phase I trial has shown that patients with acute myeloid leukaemia treated with azacytidine immediately after allogeneic stem cell transplantation had a higher proportion of T_regs compared to time‐matched controls 80.

FoxP3 levels are regulated not only through transcriptional control, but also through post‐translational modifications. In the context of transplantation, most work has focused upon acetylation of lysine residues, which is known to stabilize the FoxP3 protein 81, 82. It has been shown that inhibiting deacetylation with histone deacetylase (HDAC) inhibitors or genetically removing Sirtuin‐1, a histone and protein deacetylase, leads to an improvement in T_reg function and stability, leading ultimately to improved allograft survival 83. Thus, future directions of adoptive T_reg cell therapy will necessitate further understanding of factors that cause T_regs to lose FoxP3 expression and ways to stabilize its expression.

Dosing

It is important to note that despite protocols designed for the expansion of T_regs, there are limited data on the optimal T_reg dose for use in clinical trials of T_reg adoptive cell therapy. The majority of trials have focused initially upon the safety of T_reg cell therapy and are infusing cells at a much lower number than is likely to be clinically relevant.

Several predisposing factors must be taken into consideration when calculating the optimal dose. When looking to tip the balance in favour of regulation, the initial starting population of effector cells must also be considered, and therefore assess the current and predicted future of immunocompetency in the recipient. In the absence of lymphodepletion it is postulated that a dose of 49–79 × 10⁹ T_regs will be needed to increase the T_reg pool to clinically efficacious numbers 61. However, in concert with immunosuppressive agents such as anti‐thymocyte globulin (ATG), it is suggested that a single dose of 3–5 × 10⁹ of T_regs will serve to boost the T_reg percentage to > 33% 61. Furthermore, in association with the targeted effect of antigen‐specific cells, it has been suggested that only 1/100th–1/10th antigen‐specific T_regs (described in detail in the next section) are needed to achieve the same efficacy as polyclonal T_regs, further reducing the estimated clinically efficacious T_reg dose, although it is important to bear in mind that these extrapolations have been made from mice. At this point, the T_reg dose postulated to be efficacious in humans is unknown; as current clinical trials demonstrate the safety of T_reg therapy at increasing doses it is only a matter of time before we can begin to determine T_reg clinical efficacy accurately.

Antigen specificity

Although recent clinical trials have demonstrated the safety of T_reg therapy in a variety of contexts 51, 64, 65, 84, 85, 86, convincing data for the efficacy and therapeutic benefits of this treatment option remain to be obtained. With half of BMT recipients dying from GVHD, despite T_reg treatment 84, 85, and a limited efficacy being observed in T1D patients 86, it is clear that strategies to increase the potency of these cells are required. One approach is to use T_regs with a given antigen specificity 15, 87, 88, 89, 90, 91, 92; namely, in a transplant context, using T_regs with direct or indirect allospecificity.

In this context it is important to discuss briefly the pathways used by T_regs in the recognition of alloantigens, which will inevitably inform more directed T_reg therapy in this setting. Figure 2 depicts the three pathways of allorecognition; direct, the indirect and the semidirect.

Figure 2.

Direct, indirect and semidirect pathway of allorecognition. Direct pathway: in the direct pathway, intact major histocompatibility complex (MHC) on donor antigen‐presenting cells (APCs) is recognized directly by recipient CD4⁺ and CD8⁺ T cells. Indirect pathway: the indirect pathway is characterized by recipient APC uptake of allogeneic donor MHC that has been shed through apoptosis or necrosis. This is then processed, resulting in presentation of donor antigens in the context of recipient MHC class II to recipient CD4⁺ T cells. Semidirect pathway: semidirect allorecognition results from the transfer of cellular membrane components, including intact donor MHC, from donor APCs to recipient APCs. This process may occur through mechanisms such as cell–cell contact or through the transfer of donor exosomes that fuse with recipient APC cell membranes. Recipient APCs are then chimeric for MHC, and are able to stimulate both CD4⁺ and CD8⁺ recipient T cells.

One of the first studies to investigate the therapeutic potential of allospecific T_reg therapy was performed by Taylor et al. in 2002 88. In this study, it was demonstrated that murine T_regs with direct allospecificity could be enriched from a polyclonal population of T_regs by stimulating the cells with allogeneic splenocytes and expanding these cells ex vivo. The enhanced protective capacity of these cells, compared to T_regs expanded using anti‐CD3, was demonstrated using a GVHD model. These findings were subsequently investigated further by our laboratory in various rodent solid‐tissue transplant models 89, 90. More recently, we highlighted the translational potential of this approach using a humanized mouse model in which human T_regs with direct allospecificity were generated by stimulating polyclonal cells with allogeneic DCs 92 or B cells 91. In both studies, T_regs with direct allospecificity were shown to protect human skin xenografts from alloimmune‐mediated injury more effectively than polyclonal T_regs. The success of these studies, combined with the recent development of GMP‐compatible protocols for expanding T_regs with direct allospecificity ex vivo 93, 94, 95, has facilitated the approval of clinical trials which aim to assess the safety and efficacy of these cells in kidney (NCT02711826 and NCT02244801) and liver (NCT01624077, NCT02188719 and NCT02474199) transplant recipients.

Enriching T_regs with direct allospecificity using a preferential expansion approach is due possibly to the disproportionately high precursor frequency with which these cells exist in individuals 96. Conversely, the preferential expansion of T_regs with indirect allospecificity has proved more challenging due to the low frequency of these cells naturally present in the periphery prior to a transplant procedure 97, 98. An alternative approach which may be adopted to obtain antigen‐specific T_regs is to confer specificity through genetic modification. Indeed, the concept of using α and β T cell receptor (TCR) chain gene‐modified antigen‐specific T_regs has been investigated on numerous occasions for the treatment of various autoimmune conditions 99, 100, 101, 102, 103, 104, as well as GVHD 88, 105 and solid‐tissue transplant rejection 58, 89, 90, 91, 92. In the case of the latter, we and others employed a TCR transduction approach to generate murine T_regs with dual direct and indirect allospecificity 15, 89, 90. Interestingly, our results demonstrated that in a fully mismatched heart transplant model (BALB/c → B6), multiple injections of T_regs with direct allospecificity alone were insufficient at inducing indefinite survival of an allograft. However, only T_regs with dual specificity protected the heart allografts from chronic vasculopathy, with maintenance of myocardial architecture and luminal obstruction inhibited 90. The fact that dual‐specific T_regs protected the integrity more effectively than any population of T_regs found naturally highlighted the clinical potential for gene‐modified cell therapy.

The studies summarized above describe the ability to generate antigen‐specific T_regs by delivering designated TCR α and β chain genes. However, a different approach to confer antigen specificity which has gained popularity during the past few years is to use chimeric antigen receptors (CAR) 106. CARs are clinically relevant synthetic fusion proteins which combine various protein elements to redirect the specificity of T cells towards desired antigens of interest. CARs have been developed principally as a means of generating tumour antigen‐specific T cells 107 and their clinical use 108 has seen resounding success, particularly in the treatment of various B cell malignancies 109, 110, 111, 112, 113, 114, 115, 116, 117, 118.

In addition to their use in various cancer settings, CAR technology has also been employed to generate antigen‐specific T_regs. Initial studies generated CAR T_regs to investigate their potential for treating various autoimmune conditions 119, 120, 121, 122. However, more recently, we and others have explored the potential of CAR T_regs for the treatment of xeno‐GVHD and allograft rejection 123, 124, 125. In these studies, CARs were developed to redirect human T_regs towards a designated human leucocyte antigen (HLA) class I molecule (HLA‐A2). A donor HLA class I molecule was selected as this alloantigen is expressed ubiquitously in allografts, unlike HLA class II. Together with the findings of Noyan et al., our results demonstrated that CAR T_regs inhibited alloimmune‐mediated injury of human skin xenografts more effectively than polyclonal T_regs 124, 125, thus suggesting a potential avenue for future clinical trials. This is particularly likely given the fact that the biotechnology company TxCell (Valbonne, France) recently announced the approval of a patent for the development of CAR T_regs for treating various autoimmune and inflammatory diseases (patent identification number: EP 2126054 A2).

Survival in vivo

An additional concern post‐T_reg infusion is their survival in vivo. Initially, research suggested that T_reg survival post‐adoptive administration is short‐lived, at approximately 14 days after monitoring T_reg donor‐specific HLA markers post‐infusion 51. However, Bluestone et al. in his most recent trial of T_reg cell therapy in type 1 diabetes described an elegant method for the tracking of T_regs post‐infusion 86. In this trial, during expansion T_regs were labelled with deuterium (²H₂), which was incorporated into replicating T_reg DNA. Following this, T_regs were analysed at various time‐points post‐administration for deuterium enrichment. Results from this trial showed that the number of T_regs reached a maximum at 7–14 days post‐transplantation, with a subsequent decline. Despite this initial fall‐off, after 14 days the percentage of T_regs soon stabilized and were detectable up to a year after infusion. As such, it was postulated that T_reg survival is centred on a two‐phase decline curve, t_1/2 of the first decay at 19·2 days and t_1/2 of the second decay at a year or more 86. Further work in this area is warranted not only to define T_reg survival in vivo, but also to investigate T_reg localization post‐infusion.

Cryopreservation

Based on the suggested doses for T_reg therapy, it is uncertain whether a single dose will be sufficient or whether multiple doses are required for tolerance induction. The feasibility of multiple injections and, along this line, the practicality of cryopreservation has been questioned. Cryopreservation in T_reg manufacturing allows for the long‐term storage of isolated and expanded T_regs until they are needed, giving more flexibility in terms of ease of transport to patients’ location, follow‐up injections and allows for investigation into whether the final product fits the set release criteria 126; however, it can pose various technical challenges. Our recent publication concerning the manufacture of GMP‐compatible T_regs for cell therapy investigated the feasibility of cryopreservation. Here cells were suspended in 10% dimethylsulphoxide (DMSO) and placed into a controlled‐rate freezer before being transferred to liquid nitrogen (vapour phase) for long‐term storage. After 3 months, these cells were thawed and assessed. We demonstrated that T_reg viability post‐defrosting was > 85% with concurrent maintenance of function and phenotype 62.

T_reg therapy in solid organ transplantation to date

There has been a wealth of clinical trials investigating T_reg adoptive cell therapy in bone marrow transplantation and T1D 51, 64, 65, 84, 86. The main readouts from the trials were safety and tolerability, which were demonstrated. However, although not tested explicitly, there were hints of efficacy, evidenced in trials of BMT with the amelioration of GVHD.

Definitive trials demonstrating the relative safety and efficacy of adoptive T_reg therapy in the context of solid organ transplantation are scarce. Only recently have a Japanese group published the results of T_reg cell therapy in the setting of liver transplantation 127, where iT_regs were generated through the culture of recipient lymphocytes with irradiated donor cells in the presence of anti‐CD80/CD86 monoclonal antibodies. These iT_regs were then infused 13 days after living donor liver transplantation in 10 recipients. Immunosuppression weaning was commenced 6 months after transplantation, with complete cessation established at 18 months. Of the 10 LT recipients, seven allowed for complete cessation of immunosuppression while maintaining tolerance for more than 12 months, with four now demonstrating tolerant graft function between 29 and 33 months post‐IS withdrawal 127. The proportion of T_regs dosed amounted to an average of 3·4 × 10⁶/kg, a value much lower than that postulated by Tang et al. to be necessary for the induction of tolerance. However, in this study the iT_regs were deemed to be donor‐antigen specific, recipients were treated with cyclophosphamide prior to infusion and subjects were splenectomized, all of which may have contributed to the apparent efficacy of T_reg therapy at this reduced dose.

Currently there are several ongoing/planned trials in the setting of liver and kidney transplantation (Table 1). The clinical trials ThRIL and the ONE study are the leading trials of autologous T_reg immunotherapy worldwide in the setting of liver and kidney transplantation, respectively.

Table 1.

Table listing details of all ongoing trials of T_reg adoptive cell therapy in solid organ transplantation recorded on clinicaltrials.gov

Clinical trial	Sponsor	Setting	Phase	Status	Product	Treg doses	Timing of infusion
ONETreg1 NCT02129881	Guy's and St Thomas’ NHS Foundation Trust	Kidney transplantation	1/2	Recruiting	Polyclonal, autologous CD4+CD25+ Tregs	Single infusion of 1–10 × 106/kg	5 days post‐renal transplantation
DART NCT02244801	University of California	Kidney transplantation	1	Recruiting	Donor‐alloantigen‐reactive autologous regulatory T cells	Single infusion of 300 × 106 and 900 × 106	Timing not specified
ONEnTreg13 NCT02371434	Charite University, Berlin, Germany	Kidney transplantation	1/2	Recruiting: invitation only	Polyclonal, autologous CD4+CD25+ Tregs	Single infusion of 0.5 × 106, 1 ×  106 and 3 × 106 cells/kg	Timing not specified
NCT02091232	Massachusetts General Hospital	Kidney transplantation	1	Ongoing: no longer recruiting	Belatacept anergized donor‐alloantigen‐reactive Tregs	Infusion dose not specified	3 days post‐renal transplantation
TRACT NCT02145325	Northwestern University	Kidney transplantation	1	Ongoing: no longer recruiting	Polyclonal, autologous CD4+CD25+ Tregs	Infusion dose not specified	2 months post‐renal transplantation
TASK NCT02088931	University of California	Kidney transplantation	1	Ongoing: no longer recruiting	Polyclonal, autologous CD4+CD25+CD127– Tregs	Single infusion of 320 × 106	On their 6‐month surveillance biopsy post‐renal transplantation
TASK NCT02711826	National Institute of Allergy and Infectious Diseases	Subclinical inflammation in kidney transplantation	1/2	Recruiting	Polyclonal and donor‐alloantigen‐reactive autologous regulatory T cells	Single infusion 400 × 106	On their 6‐month surveillance biopsy post‐renal transplantation
NCT01446484	Pirogov Russian National Research Medical University	Kidney transplantation	1/2	Unknown	Polyclonal, autologous CD4+CD25+ CD127lowFoxP3+ Tregs	Two injections at 200 × 106	30 days and 6 months post‐renal transplantation
ThRIL NCT02166177	Guy's and St Thomas’ NHS Foundation Trust	Liver transplantation	1	Ongoing: no longer recruiting	Polyclonal autologous CD4+CD25+ Tregs	Single infusion of 1 × 106 cells/kg and 4.5 × 106 cells/kg	12 weeks post‐liver transplantation
NCT01624077	Nanjing Medical University	Liver transplantation	1	Ongoing: no longer recruiting	Polyclonal and donor‐alloantigen‐reactive regulatory autologous T cells CD4+CD25+CD127– Tregs	Multiple doses of 1 ×  106 cells/kg at several intervals	Timing not specified
deLTa NCT02188719	National Institute of Allergy and Infectious Diseases	Liver transplantation	1	Recruiting	Donor‐alloantigen‐reactive autologous Tregs	Single infusion 50 × 106, 200 × 106 and 800 × 106	Timing not specified
ARTEMIS NCT02474199	National Institute of Allergy and Infectious Diseases	Liver transplantation	1/2	Recruiting	Donor‐alloantigen‐reactive, autologous Tregs	Single infusion of 400 × 106	Timing not specified

The ONE Study (NCT02129881) is a multi‐centre Phase I/II study funded by the European Union FP7 programme. This trial investigates the safety of and potential efficacy of infusing ex‐vivo‐expanded T_regs, among other regulatory cells, in the setting of renal transplantation.

ThRIL (NCT02166177) is a Phase I clinical trial of T_reg immunotherapy the setting of liver transplantation, and has been initiated at King's College London, in which the safety, tolerability and efficacy of polyclonally expanded T_regs in combination with depletion therapy and short‐term immunosuppression will be assessed. At present, three patients have been dosed, with encouraging safety data to date.

Future

The understanding and use of T_reg therapy is an important and emerging area of science/medicine. Following completion of the ongoing clinical trials, we will have safety and tolerability data as well as initial efficacy data in the transplantation population. This will serve as proof‐of‐principle for use of these cells in solid organ transplantation and provide the basis for progression to larger Phase II/III studies. Concurrent immune monitoring integrated as part of the T_reg therapy trials will help to gain further mechanistic insight on T_reg function in patients, further informing future trials. However, as highlighted in this review, with the advent of new technology such as the CyTOF we now appreciate the phenotypical complexity of T_reg compartment and can characterize comprehensively rare and complex disease‐specific populations. With the GMP certification of the FACS cell sorter, as discussed before, and the information gleaned from CyTOF analysis, we will be able to isolate the optimal cell populations for expansion and cell therapeutic application.

Parallel to this work, we will also see improvements in the clinical protocol and the use of adjunct therapy to ensure T_reg survival in vivo. Currently, ongoing trials have looked at optimizing T_reg therapy through modulating immunosuppressive regimens which favour T_reg expansion and stability (NTC02166177) and pretreating T_regs with anergistics such as belatacept (NCT02091232). As T_regs depend upon exogenous IL‐2 for survival, recent strategies to ensure the longevity of the injected cells have explored the use of low‐dose IL‐2, which lacks the toxicity and immunostimulatory effects of the higher IL‐2 doses used to treat cancer patients 128, 129, 130.

This approach has been shown recently to increase the number of endogenous T_regs in patients with chronic GVHD 131, providing supporting evidence that this therapeutic strategy may be an ideal adjuvant to adoptive T_reg cell therapy by promoting T_reg expansion in an otherwise inflammatory setting. However, further studies on the isolation and expansion of patient‐derived T_regs are warranted, in view of publications highlighting the differential ability of ‘healthy’ and patient‐derived T_regs to respond to IL‐2. In line with these reports, the ability of T_regs to expand and respond to treatments such as low‐dose IL‐2 might be compromised in certain disease settings due to defective T_reg responses to IL‐2 and IL‐2 signalling 132, 133.

It is also pertinent to note that at present the cost of manufacturing a single ‘personalized’ injection of T_regs in our GMP facility is £25–30 000. We anticipate that the future optimization of the manufacturing process for larger‐scale trials and commercialization would reduce the cost, enabling the broader application of this therapy 134.

In conclusion, these advances have highlighted further the relative depth and individuality of T_reg biology in disease, and as such efforts will be focused upon tailoring T_reg therapy with the goal of developing an optimized, personalized treatment in the striving for tolerance in solid organ transplantation.

Disclosure

The authors declare no financial or commercial conflicts of interest.