The complex existence of γδ T cells following transplantation: the good, the bad and the simply confusing

Abstract Gamma delta (γδ) T cells are a highly heterogeneous population of lymphocytes that exhibit innate and adaptive immune properties. Despite comprising the majority of residing lymphocytes in many organs, the role of γδ T cells in transplantation outcomes is under‐researched. γδ T cells can recognise a diverse array of ligands and exert disparate effector functions. As such, they may potentially contribute to both allograft acceptance and rejection, as well as impacting on infection and post‐transplant malignancy. Here, we review the current literature on the role and function of γδ T cells following solid organ and hematopoietic stem cell transplantation.


INTRODUCTION
Gamma delta (cd) T cells consist of~4% of the total T cell population in human peripheral blood; however, they typically comprise a higher proportion of T cells in skin and mucosal epithelium. 1 cd T cells are a highly heterogenous group of lymphocytes that display broad functional abilities, interacting with both innate and adaptive immune compartments. A body of evidence indicates that cd T cells are important in tissue homeostasis and repair, both in the skin and mucosa. 2 Mice deficient in cd T cells spontaneously develop inflammatory bowel disease 3 and succumb to dextran sodium sulphate-induced colitis (reviewed by Nanno et al. 4 ). In addition, through the production of TGFb, cd T cells limit damage to renal epithelial cells in a rat model of autoimmune-mediated glomerulonephritis 5 and protect pulmonary epithelial cells from damage after ozone exposure. 6 Furthermore, studies on antigenic tolerance in animal models have shown dependence on cd T cells. 7 Gamma delta T cells are mediators of both antiinflammatory and pro-inflammatory responses. cd T cells exert their effects largely through MHCindependent mechanisms and can be directly cytotoxic but can also be activated by other immune cells. Furthermore, cd T cells appear to be central in the control of post-transplant infection, particularly to cytomegalovirus (CMV). Their role in transplantation outcome remains unclear, with evidence suggesting they can be both effectors and suppressors of allogenic rejection, but nonetheless highlighting them as an important component of the post-transplant immune response.

cd T CELL RECEPTORS, LIGANDS AND EFFECTOR FUNCTIONS
Gamma delta T cells are controlled by a suite of cellsurface expressed molecules, including a T cell receptor (TCR) and several receptors more commonly associated with natural killer (NK) cells. The loci encoding the cd TCR genes are the T cell receptor gamma (TRG, encoding the gamma chain) and T cell receptor delta (TRD, encoding the delta chain). 8 Largely analogous to classical ab T cells, TCR rearrangement is dependent on the expression of recombination activating genes (RAG). However, compared to ab T cells, the repertoire of cd V and J gene segments is restricted, with the TRG locus containing only 12 Variable (V) segment genes, of which 6 are functional, and the TRD locus containing eight functional V region genes. This is in comparison with ab T cells, which have 52 Vb genes and 70 Va genes. 9 Furthermore, of the TRD genes, only four of these are frequently used: Vd1, Vd2, Vd3 and Vd5. However, cd TCR still has extreme sequence variation because of a high degree of junctional diversity as a result of D segment rearrangement. 10 Also, unlike ab T cells, the vast majority of cd T cells do not express either the CD4 or CD8 co-receptor. Important in the context of transplantation, cd T cells with different TCR localise to distinct regions. The vast majority of healthy adult peripheral blood cd T cells are Vc9Vd2, whereas cd T cells bearing Vd1, Vd3 or Vd5 TCR are located in the skin, intestine, lung and liver. 11,12 In addition to their TCR, cd T cells express many receptors in common with NK cells. The NK cell receptor NKG2D is expressed on a large proportion of cd T cells and recognises the stress-inducible ligands MHC class I chain-related proteins (MIC)-A and (MIC)-B and UL16 binding proteins (ULBPs), many of which may be upregulated following transplantation. [13][14][15] Vd1 cells reportedly recognise MIC-A via both TCR and NKG2D, although TCR interactions were not involved in their cytotoxic activity 16 (Table 1). Cytotoxic activity is also triggered by Vc9Vd2 cd T cells upon ligation of another NK cell receptor, DNAM-1 17 ( Table 1). Subsets of cd T cells also express other NK cell receptors, including NKp30, NKp44 and CD94-NKG2 receptors. 18 Another important receptor also shared with NK cells is CD16, a low-affinity receptor for the constant region of IgG. The expression of CD16 allows cd T cells to recognise IgG opsonised pathogens or target cells without a strict requirement for TCR engagement. 19 In contrast to ab T cells, cd T cells typically do not recognise ligands in the context of MHC molecules. Of the known ligands, Vc9Vd2 cd T cells are activated by phosphoantigens, which can be produced by microbes or as a result of malignant transformation, 20 whereas Vc4Vd5 TCRs bind to endothelial protein C receptor (EPCR) 21 (Table 1). The ligands for Vd1 cells have remained somewhat more elusive, but are reported to include MHC-like molecules, such as the CD1 family 22 and MIC-A/B 23 (Table 1). Another member of the CD1 family, CD1d, is recognised by subsets of Vd3 cd T cells, 24 whereas other subsets of Vd3 cells recognise annexin A2 25 ( Table 1). The ligands for TCR of other cd T cells are still largely undefined.
Interestingly, cd T cell effector function depends on their niche. For example, intestinal epitheliumresident cd T cells produce keratinocyte growth factor, contributing to the intestinal barrier health and homeostasis. 26 Firmly placed at the interface of innate and adaptive immunity, following recognition of ligands by the TCR and/ or activating NK cell receptors, cd T cells are potent producers of pro-inflammatory cytokines (IFN-c, TNF-a, IL-17) and can directly lyse infected or transformed cells via perforin-and granzymedependent mechanisms. Following activation, cd T cells can also induce several cell types into antigen-presenting cells, thereby promoting dendritic cell maturation, CD4 + and CD8 + T cell priming, as well as antibody production. 27 cd T cells can also produce inflammatory and chemotactic chemokines such as RANTES, CXCL10 and lymphotactin. They are also capable of crosspresenting antigens, thereby inducing CD8 + T cell responses. 28 In addition, cd T cells do not require TCR engagement for cytokine production. Instead, they can be activated to produce IL-17 by cytokines such as IL-1b and IL-23. 29 In summary, given the complexity of receptors expressed, ligands bound and responses exerted by cd T cells, it is not surprising that they have been implicated as playing diverse roles in transplantation outcome.

EVIDENCE FOR cd T CELLS IN ADVERSE OUTCOMES FOLLOWING TRANSPLANTATION
A large proportion of the research implicating cd T cells in adverse outcomes following transplantation comes from small animal models ( Figure 1). Although cd T cell phenotypes and function in mice and humans are broadly consistent, there are also distinct differences between species, most notably the types of TCR ligands that have been identified ( Table 2).
In small animal models, cd T cells have been implicated in playing a role in ischaemiareperfusion injury (IRI). This has been demonstrated by reduced IRI in TCR cd-deficient mice in a model of kidney transplantation 30 and the observation that IL-17A, produced by cd T cells, is elevated in a mouse model of cardiac transplantation. 31 However, the proposed mechanisms differ between the studies, with cd T cells either inducing the recruitment ab T cells into the allograft, 30,32 or alternatively by inducing neutrophil recruitment through the production of IL-17. 31 The production of IL-17 from cd T cells also is reported to contribute to acute and chronic allograft dysfunction in small animal models of skin, 33 heart 34-36 and lung 37 transplantation. However, in the mouse model of lung transplantation, despite being potent producers of intragraft IL-17, there was no effect of cd T cell depletion on the development of acute rejection or fibrosis. 37 In addition, the literature is void of a link between IL-17 producing cd T cells and rejection following solid organ transplantation in humans.
There is also a disconnect between animal studies and human transplantation with respect to the role of cd T cells in graft-versus-host disease (GvHD) following hematopoietic stem cell transplantation (HSCT). Early animal studies linked cd T cells to the progression of GvHD. For example, Blazar and others 38 created a transgenic mouse model where a large proportion of T cells expressed the cd TCR. These transgenic cells proliferated and killed mismatched cells in vitro. Moreover, when the transgenic cells were infused into mismatched mice following bone-marrow transplantation, they infiltrated GvHD target tissues, indicating their capacity to cause pathology. 38 Another early study in mice revealed that depletion of cd T cells resulted in reduced GvHD. 39 However, the evidence for cd T cells contributing to GvHD following HSCT in humans is varied. While some studies showed that higher numbers of cd T cells were correlated with increased incidence of acute GvHD, 40,41 other studies have either found no correlation between numbers of cd T cells and GvHD 42 or that lower numbers were associated with increased incidence of GvHD. 43 However, it is also possible that only specific subsets of cd T cells adversely contribute to GvHD, notably Vd2 cd T cells which were implicated in the study by Viale et al. 40 Interestingly, these same Vd2 cd T cells may also be associated with poorer outcomes following solid organ transplantation. Yu et al. 44 showed higher proportions of Vd2 cells in liver transplant patients with acute allograft rejection. Similarly, lower proportions of Vd2 cd T cells were observed in operationally liver transplant recipients, having not received immunosuppression for at least 12 months. 45 However, these findings need to be interpreted with caution as an expansion of Vd1 cd T cells (thereby decreasing the proportion of Vd2 cd T cells) was observed following liver and kidney transplantation, regardless of immunosuppression treatment. It is possible that Vd1 T clonotypes expand in the blood as a result of post-transplant infections, such as CMV, as reported in healthy individuals 46 and following transplantation.
While cd T cells may contribute to the control of post-transplant infection to enhance clinical outcomes, the co-expression of CD16 may allow them to participate in antibody-mediated rejection. One study found the expansion of CD16 + cd T cells in kidney transplant patients with donor-specific antibodies was associated with renal dysfunction. 47 However, in patients without donor-specific antibodies, such cd T cells seem to be correlated with positive outcomes following transplantation, because of their ability to control CMV.  8 However, CMV is the most common infectious complication following transplantation and cd T cells are emerging as a significant player in the immunity to CMV. Following murine CMV (MCMV) infection, cd T cells prevented an increase in viral load in all organs and were as effective as ab T cells at controlling viral load in the lungs. 48 The same authors also showed that transfer of MCMVinduced cd T cells into mice lacking innate and adaptive lymphocytes rescued the animals from MCMV-induced death, indicating that cd T cells were important in the response to MCMV. 48 Another study confirmed that cd T cells can effectively control MCMV in the absence of CD4 + T cells, CD8 + T cells and B cells. 49 Figure 1. Gamma delta T cells in transplantation: the good, the bad and the simply confusing. Adverse and favorable associations between cd T cells and outcomes following transplantation have been reported. Left: 'Bad' outcomes in animal studies include ischaemia-reperfusion injury (in heart and kidney), and acute rejection of heart and lung allografts via IL-17-mediated mechanisms. Other adverse outcomes in humans include the presence of cd T cell infiltration in kidney and heart allografts; the presence of Vd2 cd T cells in patients with liver allograft rejection; antibody-mediated rejection mediated by Vd2-negative (Vd2 À ) cd T cells via recognition of donor-specific antibodies (DSA) through CD16, and an increased incidence of graft-versus-host disease (GvHD). Right: 'Good' outcomes include decreased GvHD incidence; increase in Vd1-positive (Vd1 + ) infiltration in tolerant liver recipients; secretion of IL-4 and IL-10 leading to allograft protection (observed in skin, kidney and liver); control of cytomegalovirus (CMV) infection by Vd2 À cells via IFNc and the killing of infected cells through their T cell receptor (TCR) or CD16 engagement; and control of post-transplant malignancies by Vd2 À cells which recognise tumor cells through CD16, TCR or other receptor engagements.
Following kidney transplantation in humans, reactivation of CMV drives a persistent expansion of cd T cells expressing predominantly Vd1 and Vd3 TCR, collectively referred to as Vd2-negative cd T cells 50 (Figure 1). Their expansion parallels that of CMV-specific CD8 + T cells, 51 often resulting in an increase from 1% of circulating T cells to more than 10% of the total lymphocyte count. 52 The expanded CMV-specific Vd2-negative cd T cells persisted for more than 1 year in kidney transplant recipients 53 and their presence correlated with the resolution of viraemia, whereas their absence was associated with recurrent CMV disease. 54 Similar to CMV-specific CD8 + T cells, CMV-specific cd T cells possess an effector memory phenotype, in contrast to CMVnegative patients, where they exhibited a na€ ıve phenotype. 55 Both effector memory Vd2-negative cd T cells and CMV-specific CD8 + ab T cells of CMV-infected renal transplant patients produced high levels of perforin, granzyme B, and expressed the activating NK cell receptor NKG2D. They appeared to be fully differentiated effector cells with a lower surface expression of CD28 compared to na€ ıve T cells. 52 Not only do effector memory Vd2-negative cd T cells have the same differentiated effector phenotype as CD8 + ab T cells, but they expand more rapidly in patients with CMV reactivation as opposed to primary CMV infection, which suggests that they may have an adaptive memory function. 55 The persistent expansion of Vd2-negative cd T cells following CMV infection, coupled with their differentiation into an effector/memory phenotype with expression of cytotoxic agents, implies that cd T cells respond to CMV in an adaptive manner similar to cytotoxic CD8 + T cells. Like CD8 + T cells, recognition of CMV-infected targets by Vd2 negative cd T cells is TCR-dependent, although this occurs independent of MHC. 56 The nature of the ligand(s) for Vd2-negative cd T cells remains unknown but may include EPCR. 21 However, EPCR expression is not upregulated by CMV infection and recognition of target cells by EPCR-reactive clones requires costimulatory ligands. 21 Unlike CD8 + ab T cells, Vd2-negative cd T cells may have the capacity to contribute to CMV immune control via antibody-dependent cellmediated cytotoxicity (ADCC). CD16 is expressed by the majority of CMV-induced cd T cells, whereas it is expressed only by a small amount of Vd2-negative cd T cells in renal transplant patients without CMV, suggesting that CD16 on Vd2- HLA-E (inhibitory) [99][100][101] Role in anti-CMV response cd T cells provide protection from CMV-induced death negative cd T cells is upregulated in the response to CMV. 19 However, the presence of CD16 + Vd2negative cd T cells may be problematic in transplant recipients with donor-specific antibodies because of their ability to lyse antibody-coated target cells. 51 The expansion of CMV-specific Vd2-negative cd T cells was first observed in kidney transplant recipients but has subsequently been shown to occur in heart and lung transplant recipients 57 and following HSCT. 58 Longitudinal monitoring of cd TCR repertoires in HSCT patients using nextgeneration sequencing revealed that the CMVinduced Vd2-negative cd T cells were clonal in nature. 59 Reactivation of CMV following HSCT induced significant changes in both the TRG (TCRc) and TRD (TCRd) repertoires. There were no public or shared sequences specific to CMV, as individual patients had distinct clonal cd TCR responses to CMV, although there was some homology. 59 Another study also showed that the TRD repertoire had reduced diversity in patients with CMV, further demonstrating the remarkable impact CMV can exert on cd T cells. 60

cd T cells in the control of post-transplant malignancies
Interestingly, in addition to their antiviral function, CMV-induced Vd2-negative cd T cells have been associated with reduced occurrence of skin and solid cancers in kidney transplant patients. 61 Patients who had not experienced CMV infection either prior to or following transplantation, and therefore lacked CMVinduced cd T cells, experienced a higher rate of malignancies. The expansion of CMV-specific Vd2negative cd T cells was associated with reduced cancer occurrence, and these CMV-specific Vd2negative cd T cells were shown to be able to kill tumor cells as efficiently as CMV-infected cells in vitro. 51 Akin to recognition of CMV-infected cells, the killing of tumor targets by Vd2-negative cd T cells was dependent on TCR engagement. 56 This implies that CMV infection and transformation causes the upregulation of a common antigen that is recognised by the TCR of Vd2-negative cd T cells. This phenomenon is not restricted to kidney transplant patients, as CMV-associated Vd2negative cd T cells show anti-leukaemic effects following HSCT. 62,63 However, the anti-leukaemic effector functions of Vd1-positive cd T cells were only partially dependent on TCR and strongly dependent on the expression of B7-H6, a ligand for the NK cell receptor NKp30. 64 Vd2-positive cd T cells, in particular the Vc9Vd2 subset, have also been found to exert anti-tumor effects. Vc9Vd2 cells isolated from the blood of patients following HSCT can be expanded in vitro and efficiently lyse lymphoid and myeloid targets. 63 This subset is selectively expanded in vitro by phosphoantigen stimulation following exposure of cells to zoledronic acid. 18 The in vivo activity of the Vc9Vd2 subset can be further boosted by direct infusion of zoledronic acid to the patient. These features have seen clinical trials of Vc9Vd2 cd T cells in cell therapy for the treatment of solid tumors and haematological malignancies. 18 Additionally, CD16 + Vc9Vd2 cd T cells have been shown to lyse lymphoma, chronic lymphocytic leukaemia and breast cancer cells coated with antibodies via ADCC. 65 Moreover, cd T cells were shown to have a beneficial role against refractory leukaemia by specifically targeting the recipient's cancer cells without GvHD. 66 Taken together, the data suggest that cd T cells are efficient in controlling post-transplant malignancies by multiple mechanisms including direct recognition of tumor antigens, ADCC and through the recognition of stress-associated antigens.
Suppression of post-transplant immune responses by cd T cells cd T cells may also contribute to favorable outcomes through suppression of immune responses. Lower proportions of CD8 + regulatory cd T cells were found in the blood of renal transplant recipients with acute or chronic rejection. 67 Similarly, higher numbers of CD8 + regulatory cd T cells in renal allografts were associated with prolonged survival in a rat model of renal transplantation. 68 The proposed mechanism is through the production of IL-4 and IL-10 from CD8 + regulatory cd T cells, which acts to effectively dampen Th1 responses. Supporting this notion, improved graft survival was associated with expansions of cd T cells and the increased production of IL-4 and IL-10 in an animal model of skin transplantation. 69 IL-4 in turn has a profound effect on the cd T cell population and favors the survival of IL-10-producing Vd1 cells. 70 Improved survival in this model was lost following the administration of an antibody to cd TCR. Interestingly, the production of IL-10 from Vd1 cd T cells has been hypothesised to induce operational tolerance following paediatric liver transplantation. 71 Likewise, higher proportions of regulatory Vd1 cd T cells that co-expressed CD4 and CD25 were found in the blood of tolerant adult liver transplant recipients. 45 Therefore, both animal models and human studies indicate regulatory cd T cells can positively contribute to engraftment following transplantation, possibly by the production of IL-4 and/or IL-10.
An increase in regulatory cd T cells also reportedly reduces the occurrence of GvHD following HSCT. Novel subsets of regulatory cd T cell that express Foxp3 were associated with lower GvHD in HSCT patients. 72 Interestingly, the Foxp3positive subsets utilised both Vd1 and Vd2 TCR segments, and a follow-up study narrowed the effective subset to be CD27 + Vd1 + . 73 However, in direct contrast, grafts containing higher proportion of CD8 + cd T cells were associated with increased incidence of GvHD. 74 Therefore, as reported in the above section, the role of cd T cells in the prevention or promotion of GvHD following HSCT is far from clear.

CONCLUSIONS AND FUTURE DIRECTIONS
cd T cells represent an under-researched population of immune cells with the propensity to significantly contribute to adverse and positive outcomes following transplantation, via both innate and adaptive pathways ( Figure 1). However, as the underlying cause of transplantation and the infectious insults following transplantation vary widely between recipients, the role of cd T cells needs to be carefully evaluated in the specific context.
Adverse functions of cd T cells appear to be largely linked to the production of IL-17. On the one hand, CD16 + , CMV-specific cells may exert ADCC on transplanted cells coated in donorspecific antigens, thereby contributing to antibody-mediated rejection. On the other hand, these same CMV-specific cd T cells effectively control viral replication and post-transplant malignancies. Furthermore, other cd T cell subsets can efficiently suppress adaptive immune responses and aid in immune tolerance following transplantation. The role of cd T cells in preventing or promoting GvHD following HSCT is highly controversial and may be dependent on different subsets exerting opposite effects.
Although the role of particular subsets of cd T cells is dependent on the individual context, it is clear these cells are an active and dynamic component of the transplant environment. An identification of the ligands for cd T cells will significantly aid in harnessing their therapeutic potential following transplantation. Indeed, more research is required to unveil specific subsets of cd T cells with a view to develop novel therapies that can meaningfully contribute to positive outcomes following transplantation.