Gamma delta T lymphocytes (γδ T cells) are a unique subset of T lymphocytes that have gained traction in recent years as an immunotherapeutic platform. Exerting direct cytotoxicity against tumor targets, γδ T cells act as potent antitumor effectors in the context of several types of cancer [1]. Importantly, they also act as key modulatory cells, releasing activating cytokines, such as interferon-γ (IFNγ) and tumor necrosis factor-α (TNFα), to initiate signaling cascades that mount an additional immune response. γδ T cells have been found in a number of solid tumor infiltrates, and have been correlated positively with prognosis [2]. Further, the γδ T cell receptor (TCR) functions independently of the major histocompatibility complex (MHC), suggesting γδ T cells might be able to serve as an allogeneic cell therapy [3].

Different γδ T cell subsets, most notably Vδ1 and Vδ2, also display unique properties that are attractive for immunotherapeutic use. Specifically, Vδ2 cells constitute the predominant circulating γδ T cell population in humans, generally pairing with the Vγ9-chain. Vγ9Vδ2 cells respond ubiquitously in a TCR-dependent manner to non-peptidic pyrophosphates, or phosphoantigens (pAgs), that are intermediatesin the mammalian mevalonate and microbial metabolic pathways [4, 5]. pAgs mechanism of action entails conformational changes to butyrophilin molecules BTN3A1 and BTN3A2 which result in Vγ9Vδ2 engagement [6, 7]. As these pAgs are typically upregulated during either cellular transformation or microbial infection, Vγ9Vδ2 cells serve as robustly expanding effector cells primed for innate immune responses. They are also able to become professional antigen-presenting cells (pAPCs) following antigen stimulation, conferring multidimensional immune activation and indirect antitumor activity [8]. On the other hand, Vδ1 cells are more limited in circulating numbers, residing principally in adult peripheral tissues such as the gut [9] and skin [10]. Unlike the Vγ9Vδ2 subtype, Vδ1 cells do not pair preferentially with a particular γ-chain, yielding a TCR with responsivity to a diverse range of ligands that include stress-induced self-antigens and CD1c-presented glycolipids [11, 12]. They naturally home to a variety of tissue sites, leading to the notion that they carry intrinsic tissue-resident properties beneficial to intratumoral tracking and survival. Vδ1 cells are also resistant to activation-induced cell death, suggesting an enhanced ability to persist and function long-term in vivo [13]. In comparison, other human γδ T cell subsets are relatively poorly characterized in terms of phenotype and function. Despite the rising interest in γδ T cell therapy, much about their metabolic profiles, trafficking, signaling and co-stimulation requirements, as well as memory differentiation and exhaustion continues to be poorly understood for most subsets. Understanding such properties is essential for effective immunotherapeutic enhancement.

Chimeric antigen receptors (CARs) have dominated the recent field of cellular immunotherapy as a promising breakthrough capable of effective control of chemotherapy refractory cancers. The best documented success of CAR-based therapy involves genetically engineering αβ T cells with a CD19-specific single chain variable fragment (scFv) for leukemic B-cell targeting, either a CD28 or 4-1BB endodomain for co-stimulation, and a CD3ζ endodomain recapitulating TCR signaling. Several groups have investigated the antitumor efficacy of discrete αβ T cell-based CAR engineering strategies in γδ T cells in the context of both hematological and solid tissue malignancies. However, while many of the therapeutically relevant characteristics of distinct γδ T cell subtypes remain poorly defined, their distinctive and tissue-specific properties raise the hypothesis that γδ-CAR-T cells may have unique contributions to cancer therapy distinct from their αβ-CAR-T-counterparts. Hence, γδ T-CARs may have different requirements for sustained cell activation, long-term survival and maintenance, and ultimately, efficacious therapeutic impact. There are a relatively small number of γδ T cell-specific CAR strategies; each strategy has attempted to address activation requirements through cellular engineering tested both in vitro and in pre-clinical mouse models. Here, we discuss these strategies in various γδ T cell populations and tumor contexts and evaluate their collective findings to speculate on the next steps in tumor-specific γδ T-CAR development. We aim to highlight some of the cellular properties and unique findings from these and other authors to provide a glimpse of the full scope of γδ T-CAR therapy and the gaps that remain in the field.

Several key studies have examined the antitumor efficacy of γδ T cells engineered with second-generation CARs directed towards a number of tumor-specific antigens in the context of both hematological and solid tissue malignancies (Table 1). Importantly, these reports all demonstrated that γδ T cells are capable of stable CAR expression, regardless of the gene transfer method, cell expansion protocol, or antibody scFv design [14–19]. They also showed that γδ T-CARs mediated antigen-dependent antitumor activity against their respective cancer targets both in vitro and in vivo similar to equivalent CARs expressed in αβ T cells in terms of short-term cytokine production (TNFα and IFNγ), proliferation, and cytotoxicity [14–19].

Unlike conventional CARs, CCRs incorporate an antibody-binding (scFv) and costimulatory (CD28, 4-1BB, DAP10, NKG2D, etc.). domain without an ITAM-containing signaling domain, such as CD3ζ. In this way, CCRs confer enhanced tumor-specific costimulation to γδ T cells, providing enhanced tumor antigen-dependent activation, whilst preserving native TCR function and specificity. Because they depend on endogenous γδ TCR activation, CCRs offer the same tumor-targeting advantages as do γδ T cells. Specifically, γδ TCRs function independently of MHC-binding, enabling them to recognize a number of tumors utilizing immune evasive mechanisms. The γδ TCR also innately recognizes cellular stress ligands that are not expressed on healthy tissue, allowing them to provide natural on-target, on-tumor cytotoxicity and natural avoidance of reactivity to non-malignant cells (Figure 1).

Fisher et al. [19] described a neuroblastoma-targeting CCR that incorporates natural killer receptor (NKR) costimulation in γδ T cells (i.e., GD2-DAP10). Comparing a GD2-28ζ CAR and a GD2-DAP10 CCR expressed in Vδ2+ cells, respectively, they demonstrated that CCR-expressing cells exerted a cytolytic response and released inflammatory cytokines similarly to CAR-expressing cells. Importantly, they found that full cytotoxicity and cytokine release occurred only when both the endogenous TCR (signal 1) and the CAR (signal 2) were engaged, but not when either is engaged alone. They also asserted that Vδ2-CCRs expressed lower frequencies of the exhaustion markers PD-1 and TIM-3 than did Vδ2-CARs, arguing for the long-term survival advantages of CCRs in γδ T cells.

In later studies, Fisher et al. [31] analysed CCRs comprised of a range of tumor-specific antigens (e.g., GD2, CD33, and CD19) and incorporate either TCR-mediated (i.e., CD28) or NKR-mediated (i.e., DAP10/NKG2D) costimulation. Evaluating cytotoxicity and cytokine release, all of their tested CCRs were capable of mediating cytolysis and producing TNFα and IFNγ in response to antigen-expressing tumors, but not healthy cells. However, when exploring downstream signaling strength, DAP10 yielded the most robust response. Not only did the GD2-DAP10 CCR incite the most potent TNFα production, but it also supported increased cell network plasticity wherein cells were able to flux between different signaling pathways, suggesting its utility in allowing sustained cell responsivity over time. Although more work is needed on the costimulatory needs for persistent function and maintenance in γδ T cells, the conclusion from this work is that CCRs pose a promising format of gene engineering in γδ T cells.

While the strategies discussed above yield γδ T cells with enhanced cytotoxic capacity against cancer, other approaches have aimed to direct γδ T cells towards tumor targets without providing costimulation to prevent tonic CAR signaling. First-generation CARs provide the tumor-targeting scFv antibody-binding domain and CD3ζ signaling domain of a second-generation CAR without any T cell costimulatory domain. Importantly, Ang et al. [34] demonstrated that Vγ9Vδ2 cells electroporated with an NKG2D-specific first-generation RNA CAR exhibited cytotoxicity against multiple human solid tumor cell lines. Further, these cells were able to delay disease progression in tumor-bearing mice. While these results are encouraging for the use of a first-generation CAR in γδ T cells, there are several points to be addressed in its therapeutic development. The authors postulated that the transient nature of electroporated RNA expression would help prevent on-target/off-tumor toxicity. Although it was not examined in their work, such a design could potentially abate the concern for T cell exhaustion; however, it would require multiple injections for sufficient disease treatment, as demonstrated by their xenograft data.

As another means of addressing tonic signaling associated with traditional CARs, non-signaling CARs remove all activation endodomains from a chimeric receptor and anchor a tumor-targeting scFv to a transmembrane domain (Figure 1). Like CCRs, non-signaling CARs (NSCARs) utilize the innate function of the γδ TCR to engage with stress ligands expressed on tumor cells and provide an additional tumor-homing signal on the cell surface. Fleischer et al. examined the cytotoxicity of CD5– and CD19-directed NSCARs in γδ T cells [35]. Testing them against CD5+ Jurkat and Molt-4 cell lines, as well as CD19+ 697 cells, they showed that NSCARs could confer enhanced cytotoxicity to antigen-expressing targets in vitro.

In a similar vein, one group has mounted γδ T cells with a tumor-targeting scFv bound to a CD3ɛ-binding antibody domain and a CD4 hinge, transmembrane, and cytosolic tail. These T-cell antigen couplers (TACs, Figure 1) redirect T cells towards tumor antigens, but rely on native TCR-ligand binding for full activation and costimulation [36]. This group first showed TAC function and efficacy targeting both human epidermal growth factor receptor 2 (HER2) and CD19 in αβ T cells. Comparing TACs to their corresponding second-generation CD28ζ and 4-1BBζ CARs, respectively, they showed that TACs did not upregulate exhaustion markers or promote T cell terminal memory differentiation to the same extent as CARs. The TACs were also able to mediate potent anticancer efficacy against solid and liquid tumor models. Most notably, HER2-TAC-T cells showed enhanced antitumor activity and tumor penetration against a solid tumor model as compared to the second-generation HER2-CAR. The same group have gone on to examine TAC function and antitumor efficacy when expressed in γδ T cells. In a brief report, they asserted that TACs induced tumor cytotoxicity in vitro and in vivo in γδ T cells, arguing for further development as a solid cancer treatment [37].

While both NSCARs and TACs show promise as a successful means of tumor tethering, much remains to be determined about their therapeutic efficacy, especially their longevity without the addition of other TCR signals and their tumor trafficking and anticancer activity in vivo when expressed in γδ T cells. Additionally, much like traditional first and second-generation CARs, both NSCARs and TACs rely on the presence of tumor-specific antigens for targeting. A novel tethering strategy recently published involves γδ TCR anti-CD3 bispecific molecules (GABs) that combine the tumor targeting capacity of an extracellular Vγ9Vδ2 TCR domain with the activating pan-CD3 scFv OKT-3 [38]. Using this method, van Deist et al. found that GABs were able to redirect αβ T cells towards hematologic and solid tumor cell lines, as well as primary patient-derived tumor cells. Moreover, treatment with these GABs in myeloma xenograft models yielded a significant reduction in tumor growth. Thus, GABs represent another promising route towards an effective γδ-mediated tumor metabolite immunotherapy against a broad range of cancers.

As an alternative approach for enhancing γδ T cell tumor targeting and activation, transducing γδ T cells with a tumor-specific αβ TCR has been explored. In one key study, van der Veken et al. [39] transduced γδ T cells with a leukemia-specific αβ TCR to implement tumor specificity to an effector population capable of robust killing and cytokine production without mixed TCR dimerization. They showed that transduction with an HA-2-specific αβ TCR conferred γδ T cells with potent anti-leukemic activity [39]. Similarly, Hiasa et al. [40] showed that γδ T cells transduced with an MAGE-A4-specific CD8+ αβ TCR acquired cytotoxicity against antigen-expressing tumor cells and produced cytokines in both αβ- and γδ-TCR-dependent manners. More recently, Harrer et al. [15] showed successful RNA transfection of γδ T cells with a melanoma-specific αβ TCR, whereby receptor-transfected cells exhibited melanoma-specific lysis while also retaining intrinsic γδ cytotoxic activity. Such a strategy is appealing, as it allows for patient-derived tumor targeting, but ongoing work is required to elucidate its safety and efficacy.

Another appealing approach to provide γδ-mediated therapeutic benefit targeting tumor targets involves transducing canonical αβ T cells with a γδ TCR. In this way, highly proliferative effector lymphocytes, readily attainable in high numbers from a single donor, acquire both the broad tumor reactivity and healthy self-protection of a γδ TCR without the need for a particular tumor-specific antigen. As demonstrated by Kuball and researchers, αβ T cells can efficiently express a Vγ9Vδ2 TCR that redirects them against cancer with tumor-specific proliferation and effector function [41]. One caveat to note in their study is the need for bisphosphonate application for maximal in vivo efficacy of transduced cells. However, Strijker et al. [42] showed that Vγ9Vδ2-expressing αβT cells were competent at killing neuroblastoma organoids independent of MHC-I expression. Further, these γδ-engineered αβ T cells demonstrated superior effector function compared to donor-matched untransduced αβ T and γδ T cells, respectively, suggesting preliminarily therapeutic promise for this method.

Altogether, we assert that γδ T cells serve as an up-and-coming therapeutic product. Whether as a platform for CAR-T cell therapy or for any of the alternative CAR developments, γδ T cells provide a robust effector cell with the ability to home to tumor sites, infiltrate, and exert tumor-specific cytotoxicity with concomitant cytokine production. However, as we have highlighted here, much remains to be evaluated in these cells, including exhaustion, memory phenotype and differentiation, as well as costimulation requirements. Understanding these properties both in natural γδ T cells and in transduced γδ T cells will allow for further advancements in therapeutic designs for optimal tumor targeting. In addition to this incomplete knowledge on therapeutically relevant properties within γδ T cells, γδ T-CAR development has also been confined by the limited application to human patients. The studies described here, while valuable to the field, have only addressed antitumor function in contrived cell co-culture and in pre-clinical mouse models. However, there have been a number of γδ-based immunotherapy clinical trials that supported the use of allogeneic γδ T adoptive cell transfer in cancer treatment. Several completed trials have demonstrated safe and efficient administration of γδ T cells against breast (NCT03183206), liver (NCT03183219), lung (NCT03183232), and pancreatic (NCT03180437) cancers, respectively. Further, ex vivo expanded γδ T cells have been approved in ongoing phase 1 trials against hepatocellular carcinoma (NCT04518774) and acute myeloid leukemia (NCT05015426, NCT05001451), while CAR-engineered γδ T cells targeting NKG2D ligand (NKG2DL) are being examined against relapsed or refractory solid tumors (NCT04107142). Thus, γδ T cells provide a promising platform for the future of cell-based cancer immunotherapy.