Achieving immunosuppression-free immune tolerance to an allograft is one of the central goals of transplantation. In this article, we review recent developments in the fields of T cell–based therapies and T cell engineering using chimeric Ag receptors and their potential for effective and targeted immune modulation of T and B cell activity in an effort to eliminate pre-existing alloantibodies (desensitization) and achieve long-term tolerance. Approaches that span preclinical to early clinical studies in transplantation will be reviewed, with specific emphasis on advances in T cell immunotherapy that have shown promise. Lastly, we conclude with a forward-looking discussion of how T cell–based therapies in other fields of medicine can be potentially applied to solid organ transplantation.
Solid organ transplantation is the standard of care and definitive treatment option for end-stage diseases of various organs, including kidney, liver, heart, and lung. Current armamentarium of pharmacologic and biologic immunosuppression has achieved remarkable patient and allograft survival rates in the first few years posttransplantation (1–5). However, a number of critical challenges have limited successful long-term outcomes in solid organ transplant, namely, toxicities associated with chronic and nonspecific immunosuppression (6–9) and graft loss caused by Ab-mediated rejection despite immunosuppression (10–12). Furthermore, a distinct but related immune barrier to transplant is the presence of preformed anti-HLA alloantibodies that preclude sensitized recipients from the lifesaving option of transplantation but also impact graft survival posttransplant (13–15). In this regard, current strategies for desensitization targeting B cells (anti-CD20) (16–18), plasma cells (protease inhibitors) (19, 20), and Abs (IVIG, plasmapheresis) (21, 22) have been largely ineffective. Thus, given the limited availability of precious organs, there is a clear need for creative approaches to achieve safe and effective immunomodulation that will realize the goal of “one [organ] for life” (23).
Over the last decade, T cell therapies using both endogenous and gene-edited T cells have yielded remarkable breakthroughs in cancer biology, spurring the development of several T cell platforms for application in other disease arenas, including autoimmunity and alloimmunity. T cell therapies in immuno-oncology have reached clinical-phase development, ushering an exciting era in the field of alloimmune modulation (24, 25). In the context of organ transplantation, both regulatory and engineered chimeric Ag receptor (CAR) T cells may afford therapeutic benefit. In this article, we discuss the potential of polyclonal and Ag-specific T cells (regulatory T cells [Tregs]) in modulating alloimmune responses, as well as engineered CAR T cells that could be harnessed for tolerance induction or desensitization.
Induction of immunologic tolerance using Treg therapy
The therapeutic potential of Tregs has been long recognized since the phenotypic identification of CD4+CD25+ T cells with immune-suppressive properties by Sakaguchi et al. (26). Subsequent work identified FOXP3 as a critical transcription factor for Tregs and highlighted the pleiotropic suppressive effects of Tregs by virtue of their impact on APC, T cell, and B cell functions. The demonstration that adoptively transferred Tregs can prolong the survival of skin allografts had clear therapeutic implications and has prompted clinical development of Treg therapies for the induction of immunologic tolerance in solid organ transplantation (27, 28). Indeed, correlations between the level or frequency of circulating endogenous Tregs and rejection-free graft survival have been demonstrated (29–33).
Clinical evaluation of Treg therapies for organ transplantation began with infusion of polyclonal Tregs (34). Once this was determined to be safe in the setting of kidney and liver transplant, several trials were undertaken to evaluate the efficacy of polyclonal Treg therapy in transplant and whether this approach could help minimize the immunosuppressive regimen, recently reviewed in “Regulatory T Cells: Tolerance Induction in Solid Organ Transplantation” (35). The ONE Study conducted by the multicenter consortium separately evaluated seven cell-based therapies using Tregs, tolerogenic dendritic cells, or macrophages as suppressive agents in living-donor kidney transplant recipients (36). Although there was no statistically significant difference in the rate of biopsy-proven rejection compared with a control cohort, there was importantly no increased incidence of infection supporting the safety of polyclonal Treg therapy. Another phase 1 study conducted in recipients of cadaveric liver transplants who received autologous polyclonal Tregs 2–16 mo after transplant also demonstrated the safety of the approach (37). Although there were no cases of rejection during the 6-mo follow-up, there was no clear evidence of efficacy attributable to the infused cells. These early studies that support feasibility and safety have primarily used nonspecific, polyclonal Tregs. Collectively, their results have inspired studies of Ag-specific or donor-reactive Tregs in both kidney and liver transplant recipients (36, 37). Compared with nonspecific polyclonal Tregs, a product enriched for alloantigen-specific Tregs has multiple theoretical advantages. For instance, specificity may limit global immunodeficiency, which could be a precursor to infection and malignancy. Its improved efficacy has been demonstrated in preclinical studies (38, 39).
Collectively, although Tregs have been shown to be safe in the preclinical and clinical studies performed thus far, Ag receptor–engineered Tregs with defined specificities have the potential to provide a more effective and safer therapy to modulate alloimmune responses.
CAR T cell engineering
Autologous T cells, engineered to express a CAR, have provided breakthroughs in oncology (40). CARs are synthetic receptors that provided defined-target specificity in an MHC-independent manner. This is achieved by linking variable domains of an Ab via a transmembrane domain to intracellular T cell activation domains. First-generation CARs used the ζ-chain of the CD3 complex on the intracellular side to induce T cell activation. Optimized CARs include a costimulatory domain promoting complete T cell activation. These so-called second-generation CARs have proved highly efficacious against hematological malignancies. Several CD19-directed CAR T cells have now been US Food and Drug Administration approved for treatment of B cell lineage malignancies, and additional CAR therapies (e.g., anti–B cell maturation Ag [BCMA]) are expected to be approved soon. These successes lay the foundation for expansion of CAR-based approaches beyond oncology, including the field of organ transplantation.
Induction of immunologic tolerance.
The efficacy of CAR Tregs to dampen alloimmune response was demonstrated by CAR Tregs recognizing HLA-A2 in mouse models, which proved effective in preventing graft-versus-host disease (41). Similarly, CAR Tregs were demonstrated to be more potent than polyclonal Tregs at minimizing graft injury in immunodeficient mice engrafted with human HLA-A2+ skin (42–44). More recently, donor-specific CAR Tregs were evaluated for their ability to prevent skin allograft rejection in naive and in previously sensitized mice (40). Six weeks posttransplant, despite the absence of any detectable CAR T cells in circulation, the recipients had a significantly lower concentration of donor specific Abs (DSAs), suggesting that Ag-specific CAR T cells can potentially prevent DSA formation by inhibiting the generation of donor-reactive memory B and T cells (45). However, this approach is likely limited to prevention because the CAR Tregs were ineffective at eliminating established DSAs in sensitized mice.
Although preclinical murine models have demonstrated the efficacy of CAR Tregs in the context of solid organ transplantation, most of these models have targeted HLA-A2–reactive immune cells (46); expanding this approach to other HLA epitopes is warranted. For instance, an increasing body evidence supports the finding that Abs against HLA-C, which were previously thought to be clinically insignificant, may contribute to the evolution of Ab-mediated rejection (47, 48). In addition, studies have found that Abs commonly arise against HLA-A24, B27, and B7 alleles at frequencies comparable with those against HLA-A2 (49, 50). Thus, to create therapies that have a wide applicability, a method that can target DSAs that arise to HLA epitopes beyond HLA-A2 will be necessary.
CAR T cell therapy to achieve desensitization.
CAR T cells targeting CD19 have resulted in durable remission of B cell lineage malignancies. Although global B cell depletion is an undesirable off-tumor on-target side effect, it can be leveraged as a tool to deplete alloreactive or autoreactive pathogenic B cells. Indeed, Schett and colleagues (24) used CD19-targeted CAR T therapy solely for the purpose of eliminating autoimmune B cells in a patient with severe systemic lupus erythematosus refractory to numerous chemoimmunotherapies, including the B cell–depleting agents rituximab and belimumab. Administration of a single dose of CAR T cells resulted in B cell aplasia, which was sustained for at least 44 d and which was accompanied by decline of anti-dsDNA Abs to an undetectable level, restoration of complement component levels, and normalization of proteinuria, all within 44 d of CAR T infusion. These were also accompanied by a remarkable improvement in SLEDAI score of disease activity to the point of complete remission by day 44. Certainly, evaluation of more patients with longer post-CAR T follow-up is needed. Importantly, the safety profile appears favorable for further evaluation of CAR T cells in these and other noncancer patient cohorts.
It was important to determine whether a similar approach would also be effective in elimination of preformed anti-HLA alloantibodies. This was studied in patients who achieved remission after CART-19 treatment with ongoing B cell aplasia. We and others demonstrated persistence of numerous pre-existing protective vaccine/pathogen-specific Abs and anti-HLA alloantibodies in patients who achieved durable B cell aplasia after CD19-directed CAR T cell therapy (51–53). These findings highlight the importance of CD19-negative long-lived plasma cells (LLPCs) in maintaining such immune responses. Thus, elimination of anti-HLA alloantibodies likely requires depletion of both B cells and LLPCs (51, 54, 55). In this regard, several CAR platforms currently in development for the treatment of multiple myeloma, a malignancy of PCs, might be effective. For instance, a CAR T cell targeting the BCMA has already demonstrated efficacy in patients with multiple myeloma, resulting in depletion of malignant plasma cells in the bone marrow (56), and possibly even physiological PCs (57, 58), although more evidence is needed to conclude the latter definitively. Combination of a CAR T cell targeting BCMA with a CAR T cell targeting CD19 has resulted in depletion of anti-HLA alloantibodies in human subjects (59), and this strategy can be readily adapted to desensitize candidates on the wait list for solid organ transplantation with high alloantibody burden. For example, for kidney transplant alone, 3000 candidates are very highly allo-sensitized with anti-HLA alloantibodies against 99.9–100% of the potential donor population. For this subset of patients who find it extremely challenging to find a compatible organ for transplant, CAR T cell therapy for depletion of donor-specific B cells and PCs may result in a clinically meaningful reduction of alloantibodies. This strategy is expected to be tested in an upcoming clinical trial (Clinical Trials in Organ Transplant, National Institute of Allergy and Infectious Diseases, National Institutes of Health). It is also worth mentioning that bispecific Abs that bind CD3 and BCMA, thus bridging cytolytic T cells with PCs, are also being evaluated as a desensitization strategy in transplant candidates (ClinicalTrials.gov: NCT05137054).
CAR T cell therapies that target lineage-specific Ags (e.g., CD19) are unfortunately relatively nonspecific. A novel CAR platform that targets Ag-specific B cells is a more attractive option, to avoid depletion of protective B cells and PCs. Proof-of-concept studies have used target autoantibodies or alloantigens, such as desmoglein 3 or factor VIII, as the extracellular CAR domain to redirect engineered T cells to B cells expressing only cognate Ag-specific BCRs (60, 61). Could such a strategy be used to target the highly polyclonal pool of allo-HLA–specific B cells and/or T cells? In this setting, with HLA being the immune target, the platform would call for converting HLA molecules into T cell–activating receptors. For example, β2-microglobulin (β2M) complexes with the H chain of HLA class I and thus could be engineered to deliver activation signals. β2M linked to an intracellular T cell–activating domain (β2M-CAR) could theoretically associate with endogenous HLA class I H chain, endowing the engineered T cells to selectively target HLA class I alloreactive B and/or T cells. A similar strategy may be used to convert HLA class II molecules into CARs but would require allele-specific constructs given the absence of a monomorphic binding partner such as β2M.
Indeed, this strategy of leveraging MHC molecules as CARs has been tested in the context of autoimmunity to target diabetogenic T cells. Gross and colleagues (62, 63) linked β2M-CD3ζ to peptides such as ones derived from B cell autoantigens islet-specific glucose-6-phosphatase catalytic subunit-related protein and insulin. T cells, from NOD mice, expressing this CAR construct were able to target autoreactive T cells and reduce insulitis and hyperglycemia (62).
Application of this and similar strategies to target alloreactive T cells in the context of transplantation must contend with additional important challenges such as the vast and undefined peptide-MHC specificity of the alloreactive T cell pool. In a peptide-agnostic manner, Quach et al. (64) used the β2M-CAR without peptide linkage, relying on its complex with endogenously loaded MHC class I (MHC-I) H chains. In vitro studies demonstrated the ability of CAR T cells to suppress activation and proliferation of alloreactive T cells in mixed lymphocyte reactions. Interestingly, they found suppression of both alloreactive CD8 and CD4 T cells despite using only an HLA class I–based CAR, which may point to direct and indirect methods of activity. Harnessing HLA complexes as activating receptors has also been explored for targeting MHC-II–directed pathogenic autoimmunity (65). Transgenic mice in which hCD2 promoter-driven expression of class II chains (IASα and IASβ), each linked to CD3ζ, along with linkage to major basic protein peptide (to IASβ), resulted in CAR-expressing CD4 and CD8 T cells. These T cells exhibited peptide-MHC-specific in vitro cytotoxicity, and in vivo they were able to suppress expansion of major basic protein-specific T cells and suppress experimental autoimmune encephalitis in both prevention and treatment models.
Although engineered T cell platforms to target alloreactive T cells are still in a very early phase of development, it is important to consider potential application challenges that are unique to transplant. One, for instance, is the choice between use of engineered recipient (autologous T cell product) versus donor T cells (allogeneic T cell product). While only the latter expresses allo-HLA that could be CAR modified, additional modification, such as TCR deletion, may be needed to eliminate the risk for graft-versus-host disease. The alternative use of recipient T cells presents a distinct challenge: multiple allo-HLA molecules must be codelivered. However, advances in CAR manufacturing (see Future Directions), particularly in the delivery of CAR genes/mRNA, may make this a feasible approach. Regardless of the T cell source, another consideration is the timing of CAR treatment relative to transplantation. Strategies focused on pretransplant desensitization are clinically feasible, while attempting to eliminate alloreactive cells after transplant to treat rejection must address concomitant T cell immunosuppression. Certainly, posttransplant immunosuppression may also impact the persistence and long-term efficacy of pretransplant-administered T cell therapies.
To date, no study has reported the use of HLA-based CAR T strategies to target allo-specific B cells (Fig. 1). A B cell–targeting strategy could be used as a desensitization approach before transplantation and may even provide a clinically meaningful effect with a single or limited set of HLA alleles to create “windows” in a patient’s broad alloimmune repertoire. Targeting humoral immune responses in an Ag-specific manner is confounded by IgG-secreting LLPCs, which may not display their specificity to engineered T cells as surface BCR (66). This potential challenge will need to be tested empirically; if they are found to be resistant to elimination by Ag-specific CAR T cells, nonspecific PC-targeted approaches (e.g., BCMA CAR T cells) may have to be combined with the Ag-specific B cell–targeted methods.
Developments in basic science and translational T cell immunology and bioengineering have enormous potential of providing more effective precision therapies for modulation of alloimmunity. Among the most exciting advances in the field are those related to manufacturing engineered T cells. For example, recent studies that have shortened CAR T cell manufacturing from 10–14 d to a 1- to 2-d process provide an opportunity for ease of delivery and also improve the biologic potency of the product (i.e., less ex vivo T cell expansion–induced exhaustion) (67–69). All three platforms demonstrate robust in vivo proliferation of infused CAR T cells, including in first-in-human clinical trials conducted by two of the groups (68, 69). Preclinical studies reported by Ghassemi et al. (67) and Yang et al. (68) also demonstrated superiority of their shortened manufacturing process compared with conventionally manufactured CAR T cells in preclinical models. Conventional CAR T manufacturing involves ex vivo T cell activation typically followed by CAR gene introduction using lentiviral or retroviral vectors. In contrast, Ghassemi et al. (67) described a process in which T cells may be transduced, using a modified formulation of the culture medium and the surface area-to-volume ratio of the culture vessel, to yield efficient lentiviral transduction without the need for T cell activation. One potential advantage of their method is the avoidance of T cell activation ex vivo, which may further mitigate adverse effects of activation using artificial anti-CD3/anti-CD28 reagents.
Other exciting developments in the CAR T field are approaches for in vivo CAR T cell “manufacturing” that completely avoid all issues related to ex vivo T cell expansion (70, 71). Recently, Rurik et al. (71) combined mRNA therapeutics, targeted lipid nanoparticle delivery, and CAR technology to treat pathogenic cardiac fibrosis after myocardial injury. Delivery of CD5-targeted lipid nanoparticle encapsulating mRNA encoding a CAR-targeting fibroblast activation protein created CAR-targeting fibroblast activation protein T cells in vivo, which were effective at reducing fibrosis that had already been established in a murine model of hypertension-induced cardiac injury (71). The inherently transient nature of these mRNA CAR T cells is also an attractive feature for noncancer applications such as solid organ transplant.
Given the vast complexities and challenges of alloimmunity, it is heartening to acknowledge that CAR T approaches are one among several other technologies that are currently in development for desensitization and tolerance induction. As in the field of oncology, we may need a combination of approaches to achieve successful long-term graft survival. Combining CAR T therapy with standard immunosuppression in the context of organ transplant presents unique challenges that will require careful consideration and evaluation. For instance, although desensitization strategies in the pretransplant setting may be more straightforward, attempting to eliminate alloreactive cells after transplant to treat rejection must address concomitant T cell immunosuppression, which may impact the persistence and long-term efficacy of CAR T cells (72). Nonetheless, the recent developments in basic biology and clinical applications of CAR T cell therapy make this an attractive tool in the armamentarium of immune intervention strategies. As these enter the clinical arena, the immunology community will be witness to the elimination of two major barriers in the field of organ transplantation: desensitization and tolerance induction.
This work was supported by the Institute for Translational Medicine and Therapeutics at Penn (K.M.), Gift of Life Transplant Foundation (A.N. and V.G.B.), Burroughs Wellcome Fund (Grant 100000861 to V.G.B.), and The Colton Center for Autoimmunity at Penn (V.G.B.). V.G.B. received research funding from Cabaletta Bio.
V.G.B. is a coinventor on a patent, titled “Compositions and methods of chimeric autoantibody receptor t cells,” United States patent 20170051035A1, licensed to Cabaletta Bio. The other authors have no financial conflicts of interest.