Chimeric Antigen Receptors in Different Cell Types: New Vehicles Join the Race

Abstract

Adoptive cellular therapy has evolved into a powerful force in the battle against cancer, holding promise for curative responses in patients with advanced and refractory tumors. Autologous T cells, reprogrammed to target malignant cells via the expression of a chimeric antigen receptor (CAR) represent the frontrunner in this approach. Tremendous clinical regressions have been achieved using CAR-T cells against a variety of cancers both in numerous preclinical studies and in several clinical trials, most notably against acute lymphoblastic leukemia, and resulted in a very recent United States Food and Drug Administration approval of the first CAR-T-cell therapy. In most studies CARs are transferred to conventional αβT cells. Nevertheless, transferring a CAR into different cell types, such as γδT cells, natural killer cells, natural killer T cells, and myeloid cells has yet received relatively little attention, although these cell types possess unique features that may aid in surmounting some of the hurdles CAR-T-cell therapy currently faces. This review focuses on CAR therapy using effectors beyond conventional αβT cells and discusses those strategies against the backdrop of developing a safe, powerful, and durable cancer therapy.

Introduction

The ability to convert autologous or allogeneic immune cells into living drugs that specifically attack malignant cells has revolutionized cancer immunotherapy and forms the foundation on which modern adoptive cellular therapy (ACT) is based. Currently, engineered T cells reprogrammed to target tumor antigens are the frontrunners in the armamentarium of ACT. While using T cells transfected with an additional tumor-specific T-cell receptor (TCR) has yielded promising results in melanoma and synovial sarcoma,¹ the human leukocyte antigen (HLA) restriction of TCRs poses obstacles when targeting tumors with down-regulated major histocompatibility complex (MHC) expression. Thus, conferring HLA-independent tumor specificities on T cells via the transfer of chimeric antigen receptors (CARs) has gained attraction.

The basic construction plan of a CAR features an extracellular antigen-binding single-chain variable fragment (scFv) derived from an antibody fused to intracellular signaling domains. The initial CAR design, termed first-generation CAR, relied exclusively on the cluster of differentiation 3 zeta (CD3ζ) chain to induce activation, but other signaling domains (e.g.,CD28) were added to the CD3ζ chain giving rise to second-generation CARs, which exhibited improved proliferation and a superior overall performance.² In clinical trials, second-generation CAR-T cells induced drastic tumor regressions in patients suffering from hematological malignancies, with up to 90% of patients with relapsed B-cell acute lymphoblastic leukemia (ALL) responding to treatment with CD19-CAR-T cells.³ Efforts to further improve CAR signaling spawned third-generation CARs with three signaling domains linked in cis. Gearing up for the battle against solid tumors has led to the development of CAR-T cells redirected for universal cytokine killing (TRUCKS) that respond to receptor activation with the secretion of pro-inflammatory cytokines, such as interleukin-12 (IL-12).⁴ So far, however, the different CAR-T cell products evaluated against solid tumors did not possess the power to create satisfying response rates.⁵ Moreover, immune evasion of tumor cells by virtue of shutting down the target antigen remains another crucial chink in the armor of CAR-T cells, as evidenced by malignant cells relapsing without CD19 expression upon an initial successful response to CD19-CAR-T cells.⁶

In addition to improving antitumor activity and maintaining a tumor free survival, managing CAR-T cell related side effects poses a major challenge. Severe manifestations of excessive or aberrant CAR-T cell activity include cytokine release syndrome⁷ and on-target/off-tumor toxicities.⁸ Finally, the propensity of allogeneic αβT cells to evoke graft-versus-host disease restricts conventional CAR-T cell engineering to the autologous setting. Whereas the main focus to date has been on improving CAR-αβT cells, investigating the transfer of CARs into cell types other than conventional T cells has garnered growing interest. Capitalizing on the distinct functional traits inherent in effectors beyond αβT cells, such as intrinsic antitumor activity, the manifold interconnections to other immune cells, the absence of alloreactivity, and the reduced potential for autoimmunity could help circumvent some of the challenges CAR-T cell therapy currently faces. This review summarizes the latest preclinical and clinical data on CAR therapy using γδT cells, natural killer T (NKT) cells, NK cells, and myeloid cells and analyzes those strategies in the context of developing a safe, powerful, and durable cancer therapy. The few clinical trials involving CAR effectors beyond αβT cells, which are numerically eclipsed by those relying on CAR-αβT cells, are documented in Table 1.

Table 1.

Clinical trials using chimeric antigen receptors in different cell types

Cell type	Target antigen	Disease	Modification	Country	ClinicalTrials.gov identifier
γδ T cells	CD19	ALL, CLL, and B-NHL	NA	China	NCT02656147
NKT cells	GD2	Neuroblastoma	28ζ CAR + IL-15	United States	NCT03294954
NK cells	CD19	ALL	4-1BBζ CAR	United States	NCT00995137
		ALL	4-1BBζ CAR	Singapore	NCT01974479
		ALL, CLL, and B-NHL	28ζ CAR + IL-15 + iCasp9	United States	NCT03056339
		ALL, CLL, and B-NHL	28BBζ CAR	China	NCT02892695
	CD33	AML	28BBζ CAR	China	NCT02944162
	CD7	AML and T-NHL	28BBζ CAR	China	NCT02742727
	MUC1	HCC, NSCLC, pancreatic, gastric, breast, colorectal carcinoma, glioma	28BBζ CAR	China	NCT02839954

28BBζ CAR, chimeric antigen receptor with CD28 and 4-1BB co-stimulatory domains; 28ζ CAR, chimeric antigen receptor with CD28 co-stimulatory domain; 4-1BBζ CAR, chimeric antigen receptor with 4-1BB co-stimulatory domain; ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; B-NHL, B-Cell Non-Hodgkin lymphoma; CLL, chronic lymphocytic leukemia; HCC, hepatocellular carcinoma; iCasp9, inducible Caspase 9; IL-15, interleukin-15; NA, not available; NSCLC, non-small cell lung cancer; T-NHL, T-cell Non-Hodgkin lymphoma.

CAR-γδT cells

Armed with a VDJ-recombined γδT-cell receptor and several germline-encoded cytotoxicity receptors (e.g., natural killer group 2, member D NKG2D) γδT cells represent a crucial factor in the first-line defense against pathogens and malignant cells.⁹ After egress from the thymus, γδT cells expressing a Vδ1 chain localize to the subepithelial layers and those expressing a Vδ2 chain, usually paired with a Vγ9 chain, remain in the blood, making up around 2–10% of all circulating T cells.¹⁰ Contrary to αβTCR activation that requires MHC-bound peptides, γδTCRs are triggered in MHC-independent fashion (e.g., blood resident γ9δ2 T cells are activated by butyrophilin-bound aminobisphospohonates, such as the cholesterol precursor isopentylpyrophosphate).⁹ Artificial interference with cholesterol synthesis by applying the FDA-approved drug zoledronic acid (ZA) to peripheral blood mononuclear cells (PBMCs) results in a surface accumulation of isopentylpyrophosphate and a subsequent stimulation of γδT cells.¹¹ Based on this mechanism, ZA can be incorporated into an expansion protocol to generate large numbers of γδT cells from PBMCs.¹² One restriction of this approach may be the sole amplification of γ9δ2 T cells. Therefore several groups have devised protocols to expand more subtypes concomitantly. Reported strategies involve the expansion in the presence of artificial antigen-presenting cells¹³ and the use of an anti-CD2 monoclonal antibody.¹⁴ Collectively, these protocols enable the production of a large quantity of γδT cells needed for an effective adoptive T-cell therapy.

CAR-γδT cells – power

Gamma delta T cells can be rendered specific for tumor surface antigens by equipping them with a CAR. Similar to αβT cells, permanent DNA-based and transient RNA-based transfection methods have been explored. Using retroviral vectors, 73% of ZA-expanded γδT cells could be transduced with a CD19-specific CAR or a ganglioside GD2-specific CAR.¹⁵ Nonviral gene transfer of a CD19 CAR employing the sleeping beauty system transposon and transposase into PBMCs and subsequent activation with target cells could generate more than one billion CAR-γδT cells¹³. Electroporation of mRNA encoding a CSPG4-specific CAR into ZA-expanded γδT cells resulted in a high but transient CAR expression.¹⁶ Concerning functionality, CAR-γδT cells antigen-specifically generated tumor necrosis factor (TNF) and interferon gamma (IFNγ) in response to tumor cells. Notably, no IL-2 production could be observed.^13,16 Moreover, tumor lysis similar to conventional CAR-αβT cells was displayed by RNA-transfected CAR-γδT cells.¹⁶

The intrinsic feature of γδT cells to kill metabolically aberrant cells¹⁷ may synergize with the lymphodepleting chemotherapy given prior to CAR-T cell infusion. Hypothetically, those chemotherapy regimens could simultaneously disturb the metabolism in tumor cells and in the tumor-protective environment, to orchestrate a diversified attack combining CAR activity with intrinsic antitumor activity. The same effect might be achieved by co-administering CAR-γδT cells with ZA, which rendered tumor cells more susceptible to γδTCR mediated cytotoxicity.¹⁸ However, as this drug will act on the healthy tissue as well, it needs to be determined whether off-target toxicity will be increased. Unlike their αβ counterparts, γδT cells are able to target cells without MHC molecules on their surface, as evidenced by a strong lytic activity against the MHC-negative lymphoma cell line Daudi.¹⁹ While in vitro data on CAR-γδT cells are vastly outnumbered by reports on conventional CAR-T cells, studies evaluating the antitumor activity of CAR-γδT cells in in vivo models are even fewer. Deninger et al. examined genetically engineered CD19-CAR-γδT cells in the NALM-6 mouse model of ALL. Following infusion of ten million γδT cells, a significant tumor reduction could be registered, with the strongest response observed in the blood, followed by spleen and bone marrow.¹³

Besides striking with the introduced CAR and the endogenous receptor arsenal, γδT cells are also closely wired to other immune cells operating at the interface of adaptive and innate immunity. First, CAR-γδT cells can secrete macrophage inflammatory protein (MIP)-1α, MIP-1β, and RANTES to chemoattract macrophages.¹³ Second, it has been described that γδT cells can promote DC maturation and produce chemokine (C-X-C motif) ligand 13 to elicit antibody formation.⁹

CAR-γδT cells – durability

Attaining a successful tumor regression is difficult and maintaining it is even more challenging. Predicated on the trials with conventional CD19-CAR-T cells in ALL, it has become increasingly clear that tumor relapse after a successful CAR-T cell therapy can be rationalized with either loss of the T cell graft or antigen loss by the tumor.⁶

The former—loss of T cells—could be alleviated by recursive infusions of CAR-T cells. Contrary to conventional T cells, γδT cells display no alloreactivity and could thus be applied in an allogenic fashion after their isolation from healthy donors in large quantities.²⁰ The lymphodepleting chemotherapy that is commonly performed prior to ACT may possibly prevent host vs graft rejection of allogeneic CAR-γδT cells. Moreover, CAR-γδT cells obtained from healthy donors could boost T-cell functionality, as patient-derived T cells may exhibit functional impairments, for example a lower IFNγ secretion.²¹

The second crucial reason for tumor relapse uncovers the Achilles' heel of CAR-T cell therapy per se: the loss of the target antigen, which renders the CAR useless. CAR-γδT cells, however, could maintain antitumor activity via their endogenous TCR and other receptors, such as NKG2D, the ligands of which are widely expressed on malignant cells.²² Further possibilities how CAR-γδT cells might sustain the attack after antigen loss arise from evidence collected over the past two decades that γδT cells can act as antigen-presenting cells. Brandes et al. unveiled that activated γ9δ2T cells could process and display soluble antigens, as well as provide the necessary co-stimulation to activate naïve αβT cells. Induced proliferation in CD4⁺ and CD8⁺ T cells approximated the scale elicited by mature dendritic cells, highlighting an efficient MHC-II presentation and the ability to cross-present antigens.²³ Muto et al. showed that γδT cells co-cultured with MART-1-positive live tumors induced proliferation in MART-1-specific CD8⁺ T cells.²⁴ Data showing the activation of an αβT cell response by CAR-γδT cells are still missing and further research in this area is warranted. Nevertheless, the notion that CAR-γδT cells kill malignant cells with the CAR and present and cross-present antigens to CD4⁺ and CD8⁺ T cells brings three major advantages: (i) extending monotargeting via the CAR to polytargeting mediated by a variety of αβTCRs in CD4 and CD8 T cells, (ii) enlarging the target repertoire by intracellular antigens, which are recognized by TCRs, but not by CARs, and (iii) building up an endogenous antitumor memory. By these means, CAR-γδT cells possess the potential for creating a durable antitumor response, even against the backdrop of antigen loss, due to their capacity of mediating polyspecific αβT cell responses by acting as antigen-presenting cells to initiate epitope spreading.

CAR-γδT cells – safety

Minimizing side effects is a paradigm that applies to CAR-T cell therapy in general. In the clinical trials conducted so far, cross-reactivity toward nonmalignant host cells and excessive systemic cytokine production by infused T cells became apparent as the most severe side effects associated with CAR-T cell therapy.^7,8 Cross-reactivity either originates from CAR binding to cognate antigens on nonmalignant host cells or from CAR affinity to structurally similar antigens or isoforms. Additionally, it is conceivable that CAR signaling may stir up dormant autoreactive αβT cells previously silenced by peripheral tolerance, which may lead to off-target toxicities emanating from the endogenous αβTCR. This concern can be obviated by resorting to CAR-γδT cells, because the γδTCR does not recognize autoantigens and thus has no propensity to autoimmunity after activation. A mechanism to temporally limit CAR-related toxicity is represented by RNA-based transfection. In a recent study, CAR-γδT cells were electroporated with mRNA coding for a CAR, which resulted in a transient expression of the CAR.¹⁶ As for cytokine release syndrome, CAR-γδT cells may represent a safer substitute, as functional comparison to CAR-αβT cells regarding cytokine secretion and cytolytic capacity revealed a similar killing ability but a pronounced disparity in cytokine secretion with CAR-γδT cells producing no IL-2 and less IFNγ and TNF.¹⁶ This offers an alternative to αβT cells in cases with severe cytokine release syndromes or IL-2 mediated vascular leakage syndromes. Summarizing, with respect to safety, CAR-γδT cells may be utilized in some situations as less dangerous cell type than their αβ counterparts.

CAR-γδT cells – clinical trials

To date, the US National Institutes of Health ClinicalTrials.gov database includes only a single study using CAR-γδT cells. In a phase-I clinical trial from China, CD19-CAR transduced allogeneic γδT cells are administered to patients with ALL and CLL (chronic lymphocytic leukemia) after lymphodepleting chemotherapy (trial number NCT02656147). A trial examining the intrinsic tumor activity of γδT cells against renal cell carcinoma demonstrated a solid antitumor response, and no adverse events were reported.²⁵ In sum, CAR-γδT cells made their first step into the clinic, but far more clinical trials are needed to elucidate whether the promising theoretical advantages and preclinical findings regarding CAR-γδT cells really translate into a powerful tumor regression, long complete responses, and fewer adverse events.

CAR-NKT cells

Comprising approximately 0.1–0.5% of T cells in the blood, NKT cells can be identified by an invariant αβTCR co-expressed with NK cell markers, such as CD161.²⁶ During thymic development NKT cell precursors pair a Vα24-Jα18 alpha chain with a Vβ11 chain to form an invariant TCR.²⁶ Contrary to conventional αβTCRs that recognize MHC-bound peptides, the invariant TCR of NKT cells is triggered by glycolipids presented via the monomorphic MHC class Ib molecule CD1d on antigen-presenting cells.²⁶ A well-studied CD1d-presented ligand is alpha-galactosylceramide (α-GalCer) inducing IL-4 and IFNγ production upon TCR ligation.²⁷ It has been demonstrated that cytokine production by NKT cells is dependent on the constitutive expression of the immunoglobuline superfamily member CD28, as CD28/CD80 interaction blockade interferes with cytokine production.²⁸ Similar to conventional αβT cells, NKT cells can be categorized on the basis of CD4 and CD8 expression into CD4⁺, CD8⁺, or double-negative NKT cells with distinct functionalities. For example, CD4⁺ NKT cells elaborate Th1 cytokines, as well as Th2 cytokines, and secrete perforin in response to pro-inflammatory stimuli, such as IL-12 and IL-2.²⁹ The CD4-negative NKT cells are confined to a Th1 cytokine profile and respond to PMA/ionomycin stimulation with perforin upregulation.²⁹

NKT cell expansion to numerically relevant quantities has been demonstrated by several investigators: Shimizu et al. amplified NKT cells from PBMCs with α-GalCer and IL-2 and performed fluorescence-activated cell sorting. The obtained NKT cells were restimulated with α-GalCer pulsed PBMCs and IL-2, IL-7, and IL-15 for one month.³⁰ Using α-GalCer–loaded monocyte derived dendritic cells, Takahashi et al. could show an expansion of CD4⁺ and double-negative NKT cells.³¹ Rogers et al. generated more than 10¹² NKT cells within 2 months of expansion using α-GalCer pulsed PBMCs and IL-2.³²

CAR-NKT cells – power

Opting for NKT cells as CAR carriers seems to be a promising approach as CAR activity may synergize with the intrinsic antitumor activity of NKT cells. In a variety of tumors, NKT cell infiltration bodes well for survival, for instance in colon carcinoma.³³ NKT cells unfold a strong cytotoxicity against malignant cells based on perforin and Fas ligand mediated killing.³⁴

In addition, cytokines—such as IL-2 and TNF—induce perforin upregulation in NKT cells, whereas triggering of the endogenous TCR in CD1d-dependent fashion results in IFNγ production, which is beneficial in the activation of NK and CD8⁺ T cells to promote greater tumor rejection.²⁹ A substantial difference that might help explain the dichotomy of CAR-T cell therapy achieving massive complete rejection rates in ALL versus moderate response rates in solid tumors consists of the protective and inflammatory tumor microenvironment (TME) surrounding solid tumors. From erecting fibrotic barriers to co-opting different cell types, such as regulatory T cells and tumor-associated macrophages (TAM), the TME poses a problem to T-cell therapy in general.³⁵ Using their endogenous receptor, NKT cells can kill CD1d⁺ TAM.³⁶ This trait may be beneficial in pioneering a way through the TME and create access to the bona fide malignant cells for CAR-targeting.

Heczey et al. retrovirally engineered NKT cells with a first-generation CAR and two second-generation constructs relying either on CD28 co-stimulation (28ζ) or on 4-1BB co-stimulation (BBζ) and a third-generation construct containing CD28, 4-1BB, and CD3ζ (28BBζ). All constructs were specific for GD2. Additionally, all these CAR-NKT cells exhibited a pronounced antigen-specific lysis of GD2⁺ neuroblastoma cells in vitro, while retaining their intrinsic capacity to kill the immunosuppressive CD1d⁺ M2 macrophages, supporting the aforementioned notion of pioneering and killing. Conspicuously, CARs possessing a 4-1BB domain polarized the NKT cells to a Th1 cytokine profile, and like those with the CD28 domain, enhanced the persistence in vivo. In a murine model of neuroblastoma, repeated injections of CAR-NKT cells bearing all those different CAR formats led to accumulation of those cells at the tumor site and significantly increased the survival of tumor bearing mice.³⁷ Apart from identifying the best CAR construct for NKT cells, elucidating the best composition of the NKT cell population before transduction is another crucial challenge. Some reports indicate that some NKT subpopulations can even be tumor protective by preventing effective immunosurveillance.³⁸

CAR-NKT cells – durability

Data from conventional T-cell therapy indicate that the differentiation state of a T cell after expansion and before infusion is a predictor for successful proliferation and persistence in vivo.³⁹ Seeking to establish parameters for NKT cell persistence and proliferation in vivo, Tian et al. identified CD62L expression as a predictor for robust proliferation. Antigenic stimulation led to a sole expansion of CD62L⁺ NKT cells, whereas the CD62L-negative fraction succumbed to early exhaustion. While exhibiting no difference in cytolytic activity, cytokine production could be exclusively attributed to CD62L-negative NKT cells. Upon transferal into immunodeficient NOD scid gamma (NSG) mice, CD62L⁺ NKT cells outproliferated their CD62L-negative counterparts and CD62L⁺ NKT cells could be detected five times longer. In addition, retroviral transfer of a CD19-specific CAR to CD62L⁺ NKT cells and subsequent infusion into mice with B-cell lymphoma resulted in tumor regression, while CAR-expressing CD62L-negative cells could not clear the malignant cells. Various modifications of target cell lines were assayed in order to determine the optimal yield of CD62L⁺ NKT cells, with HLA^nullCD1d^medCD86^high4-1BBL^medOX40L^high K562 cell–expanded CAR-NKT cells showcasing the best proliferation and the best antitumor activity against lymphoma and neuroblastoma cells.⁴⁰

In some instances, loss of CAR expression or failure to establish persistence by adoptively transferred effector cells necessitates repetitive injections. Heczey et al. employed allogeneic CAR-NKT cells in mice and no evidence of graft-versus-host disease (GVHD) was observed.³⁷ Hence, a register of healthy donors could be established serving as a NKT cell reservoir for repetitive injections and for patients where expansion of autologous NKT cells is difficult.

Finally, the ability of NKT cells to promote the maturation of dendritic cells via CD40/CD40L interactions⁴¹ creates avenues for eliciting an antigen spread and recruiting αβT cells specific for additional tumor antigens (and also intracellular ones).

CAR-NKT cells – safety

The safety of adoptively transferring NKT cells was corroborated by no severe toxicity occurring in patients with non-small-cell lung cancer following autologous NKT cell infusion.⁴² Equally to γδT cells another safety benefit originates from the invariant NKT cell TCR, which recognizes primarily glycolipid structures presented on CD1d that have not been associated with autoimmunity. Hence, stimulation through an introduced CAR does not entail the danger of activating an endogenous, self-reactive αβTCR inciting autoimmunity. To mitigate concerns on permanent CAR-related toxicities, evanescent CAR-transfection may represent an attractive alternative to genetically engineered CAR-NKT cells. So far, no RNA-based CAR expression has been reported, but the successful transfer of a functional αβ TCR via mRNA-electroporation into NKT cells has been shown.³⁰ In contrast to conventional T cells, NKT cells display no alloreactivity and donor NKT cells did even exert a protective effect against the emergence of GVHD after allogeneic bone marrow transplantation.⁴³ Up to now, no clinical data has been collected on administering donor CAR-NKT cells. Nevertheless, this strategy harbors some attractive and unique advantages: (i) a preparatory myeloablative chemotherapy may inflict a powerful first strike upon malignant cells, (ii) the ensuing bone marrow transplantation may sustain the initial strike by virtue of graft versus tumor effect, and (iii) administering donor CAR NKT cells might prevent the induction of GVHD while attacking residual malignant cells in an antigen-specific manner. Finally, NKT cell deficiency is associated with an enhanced susceptibility to autoimmunity. Reduced NKT cell numbers were reported in patients with type I diabetes,⁴⁴ as well as NKT cell dysfunctionalities in scleroderma.⁴⁵ Clinical trials are needed to find out whether the use of CAR-NKT cells can reconcile a strong antitumor activity and the suppression of autoimmunity.

CAR-NKT cells – clinical trials

Despite the advantages presented above, clinical studies with CAR-NKT cells are scarce. Motivated by promising results obtained in neuroblastoma patients using autologous CAR-T cells, investigators at Baylor have registered a clinical study employing CAR-NKT cells (NCT03294954), which is not yet open for recruitment. The major shortcomings of this anteceding T-cell trial arose from an insufficient CAR-T cell persistence.⁴⁶ In addition to the GD2-specific CAR, the NKT cells, infused after cyclophosphamide and fludarabine conditioning, are thus additionally engineered to secrete IL-15, speculating on an enhanced persistence and activity conferred on NKT cells by this cytokine. Moreover, several trials aiming to assay the clinical performance of untransfected NKT cells, especially against solid tumors, for example breast cancer (NCT01801852) and melanoma (NCT02619058), have commenced recruiting patients.

CAR-NK cells

NK cells belong to the lymphoid branch of the immune system and account for up to 6% of circulating lymphocytes. After arising from the common lymphocyte progenitor, NK cell precursors do not undergo genetic rearrangement to create an antigen-specific receptor, but migrate to secondary lymphoid organs, where a distinct configuration of germline-encoded activating and inhibitory receptors emerges. NK cells residing in the lymphoid organs, predominantly CD56^bright/CD16^dim, are more proficient in cytokine secretion than their CD56^dim/CD16^bright counterparts in the blood, which exhibit a higher cytolytic capacity.⁴⁷ The internal circuitry determining the activation or inhibition of NK cells computes the various external signals received via the plethora of germline-encoded receptors into a single decision to act or to rest. Pro-activatory stimuli are transduced via receptors sensing disturbances in cell homeostasis indicating—for instance, an incipient malignant transformation. Those receptors include the NKG2D receptor recognizing the stress ligands MIC-A and MIC-B on tumor cells and the natural cytotoxicity receptors NKp30, 46, and p44, with the last only being present on activated NK cells.⁴⁸ Further stimulatory signals are transmitted by virtue of the receptor dimer CD94/NKG2C and DNAM-1, which empower NK cells to kill tumor cells expressing the corresponding ligands CD112 and CD155.⁴⁸ Via CD16 (FcγIII), NK cells can direct antibody-dependent cellular cytotoxicity to antibody-coated targets.⁴⁸ On the other hand, receptors pertaining to the killer cell immunoglobulin-like receptor (KIR) family bind to class I MHC molecules and inhibit NK cell activation. In sum, there are two major mechanisms to evoke NK cell effector functions: (i) missing self: absence of inhibitory ligands, for instance, as a result of down-regulated MHC presentation, and (ii) induced self: pro-activatory stimuli outweigh their inhibitory counterparts, for example upregulation of stress ligands or cells coated by antibodies. Upon activation, effector functions encompass elaborating cytokines, such as IFNγ and granulocyte-macrophage colony-stimulating factor, as well as cellular cytotoxicity mediated by perforin, granzyme, FAS/FAS ligand interactions and the TNF-related apoptosis-inducing ligand pathway.⁴⁹

To generate clinically relevant numbers of NK cells in close proximity to complete good manufacturing practice (GMP) compliance, Kloss et al. separated NK cells from PBMCs in fully automated vein using the Prodigy system of Miltenyi Biotec and propagated them for 14 days employing IL-2, IL-21, and autologous PBMCs as feeder cells.⁵⁰ Voskens et al. successfully expanded cytolytic NK cells from conventional PBMCs, as well as from a GMP-compliant lymphocyte enriched fraction of PBMCs with IL-2 and irradiated K562 cells engineered to express membrane-bound IL-15 and 4-1BBL.⁵¹ A feeder cell–free protocol involving an automated bioreactor for the clinical-grade generation of NK cells has also been reported.⁵² Unlimited access to NK-like cells is granted by harnessing the human NK-92 cell line, which does not require laborious isolation from blood and can be efficiently expanded requiring only IL-2 and no feeder cells.⁵³ Of note, NK-92 does not express CD16 and has a very small KIR repertoire.⁵⁴

CAR-NK cells – power

Given the important role bestowed on NK cells in tumor defense, malignant cells frequently deploy counter-mechanisms to evade NK cell tumor surveillance. Such mechanisms vary from shedding NKG2D ligands or inducing those on host cells leading to a general desensitization and internalization of NKG2D receptors.⁵⁵ The upregulation of NK cell inhibitory receptors to avoid NK cell mediated cytotoxicity has also been documented.⁵⁶ Retroviral gene transfer of a CD19-specific CAR containing either the CD3ζ chain alone or linked to the 4-1BB domain empowered primary NK cells to lyse CD19⁺ targets and autologous leukemic lines in vitro. Of note, 4-1BB co-stimulation led to a higher cytolytic capacity and cytokine production.⁵⁷ Apart from permanent transfection of CARs, where Suerth et al. could show superior transduction rates using alpharetroviral vectors compared with gammaretroviral and lentiviral vectors,⁵⁸ transient expression of CARs has also been investigated. Electroporation of mRNA coding for a CD19-specific CAR into NK cells resulted in an efficient transient transfection and functionality, reflected by considerable killing of CD19⁺ targets.^59,60 Additionally, NK cells nucleofected with mRNA encoding a CD20-specific CAR eliminated Rituximab-resistant cell lines and significantly prolonged the survival of tumor bearing mice.⁶¹ Noteworthy, the transfer of a functional CAR to NK cells can also be achieved by trogocytosis, as evidenced by the presence of CD19-specific CARs on NK cells after co-culture with a CAR-expressing donor cell line and the subsequent killing of ALL-blasts in vitro.⁶² Besides, CS1⁺ myeloma cells could be efficiently targeted with CAR-NK cells,⁶³ as well as CD123⁺ AML blasts.⁵⁰

Regarding solid tumors, which are often resistant to NK cell cytotoxicity, primary NK cells genetically reprogrammed to target HER-2 via a CD28 co-stimulated CAR could eliminate HER2⁺ carcinoma cells in vitro and in vivo.⁶⁴ In an effort to tailor CAR design to NK cells, Altvater et al. discovered that 2B4 co-stimulation led to an overall increase in functionality of CD19- and GD2-specific CARs and restored killing of autologous leukemic cells and neuroblastoma cells.⁶⁵ The functional comparison of CAR designs with either 4-1BB or 2B4 co-stimulation and a third-generation construct comprising both domains did not reveal a difference in antitumor activity, with all CAR-NK cells antigen-specifically killing GD2⁺ allogeneic Ewing sarcoma cells in vitro, but failing to exert substantial activity when locally administered to mice bearing Ewing sarcoma. Strikingly, the upregulation of HLA-G on the tumor cells was implicated in NK cell suppression, and highlights the importance of focusing research on NK cell specific checkpoint blockade.⁵⁶ A study examining the power of NK cells engineered to express a prostate stem cell antigen (PSCA)–specific CAR with a DAP-12 signaling motif found an augmented cytotoxicity compared to constructs including only the CD3ζ chain, and a complete tumor eradication of PSCA⁺ tumor cells in mice.⁶⁶ In order to facilitate tumor homing, lentiviral tandem transfer of a DAP-12 based CAR cognizant of endothelial growth factor receptor (EGFR)vIII and the C-X-C chemokine receptor type 4 was analyzed, resulting in an improved tumor chemotaxis and a significant extension of the survival in murine glioblastoma models.⁶⁷

Exploring the human NK-92 cell line as a vehicle for a CAR has gained increasing interest, owing to the unlimited availability and easy expansion. Upon irradiation, administering NK-92 cells to patients is well tolerated.⁵⁴ Equipping NK-92 cells with a CD20- or CD19-specific chimeric antigen receptor reinstated their capacity to kill primary lymphoma and leukemic cells.^68
–70 Further proof-of-principle data showing CAR functionality in NK-92 cells is derived from several studies transferring first-generation constructs eliciting NK-92 effector functions against CD138,⁷¹ GD2,⁷² and erbB2⁷³ expressing targets. As the evolution of CAR-technology brought up second-generation CARs, CD28, and 4-1BB co-stimulation was analyzed for promoting NK-92 effector functions in antigen-specific manner, with CD28 co-stimulated CAR-NK-92 cells displaying the highest cytotoxicity against erbB2⁺ targets⁷⁴ and CD19⁺ targets.⁵³ Further corroboration for the efficacy of CD28 co-stimulated CARs in NK-92 cells is provided by Liu et al. demonstrating that NK-92 cells genetically modified to express an erbB2-specific CAR could significantly prolong the survival of mice inoculated with erbB2⁺ breast cancer cells and induced the shrinkage of the primary tumor and lung metastases.⁷⁵ Unlike combating leukemic blasts, which primarily reside in the blood and bone marrow, thereby representing sitting ducks for adoptively transferred cells, solid tumors are often immersed in a protective tumor microenvironment (TME)³⁵ suppressing tumor clearance by CAR-NK cells. A key cytokine in the atmosphere of the TME is transforming growth factor beta (TGF-β), which paralyzes NK cells by interfering with granzyme and perforin secretion, as well as by downregulating NKGD2 receptors.⁷⁶ To reverse this effect to the opposite, Wang et al. fused the extracellular domain of the TGF-β receptor to the intracellular signaling machinery of the NKG2D receptor. NK-92 cells transfected with this CAR did not display TGF-β mediated suppression, maintained NKG2D expression, exhibited improvements in IFNγ secretion and lytic capacity, inhibited T-reg differentiation, showcased an enhanced chemoattraction to TGF-β-releasing cells, and proved to be effective in xenograft model of liver cancer.⁷⁷ Heterogeneous expression of target antigens on tumor cells poses another crucial issue, necessitating combination therapies. One group has combined NK-92 cells expressing a CAR dual-specific for EGFR and EGFRvIII with oncolytic herpes simplex virus (oHSV-1), speculating on the eradication of EGFR-negative targets by the virus and on enhanced infection rates due to epithelial disturbance incurred by the initial NK cell attack. In a murine model, the combination of oHSV-1 and CAR-NK92 could outperform the monotherapies with regard to tumor killing and mice survival.⁷⁸

CAR-NK cells – durability

CAR-NK cell persistence is a vital prerequisite to maintaining tumor control. Capitalizing on the absence of the T-cell marker CD5 on NK cells, a recent study reported on potent and specific activity exerted by NK-92 cells equipped with a third-generation CD5-specific CAR against T-cell leukemia and primary tumor cells in vitro. When challenged in xenograft murine models of T-ALL, over 50% of the tumor mass could be eliminated by day 11. Nevertheless, after discontinuation of CAR-NK infusions, tumor cells remerged as a result of insufficient CAR-NK cell persistence exemplifying the short-lived nature of these cells in vivo.⁷⁹ Hence, repetitive dosing will be necessary to maintain tumor regression. Sufficient numbers can be provided by employing the clonally expanding NK-92 cell line or creating donor banks with suitable healthy NK cells donors for the use in the allogeneic setting, which additionally harbors the advantage that NK cells are less inhibited by KIR signaling triggered by self-MHC molecules. Deriving NK cells from hematopoietic stem cells has also been successfully explored, offering another source for supplying a sufficient quantity of NK cells.⁸⁰ Persistence could be further boosted by co-engineering CAR-NK cells to constitutively secrete IL-15, which can act in autocrine fashion to sustain proliferation in the absence of exogenous cytokines without impairing CAR functionality as shown for an EpCAM-specific CAR.⁸¹ Investigators from Houston generated NK cells from cord blood simultaneously expressing a CD19-CAR and IL-15. Engineered cord blood NK cells outperformed equally engineered autologous NK cells when cytotoxicity against CLL cells was assayed, which could be largely traced back to a NKG2A mediated inhibition of autologous NK cells. In xenograft mouse models of Raji lymphoma, IL-15 co-transduced cord blood CAR-NK achieved complete tumor eradication, while mono-transduced CAR-NK cells did not establish tumor control. The in vivo presence of IL-15 co-transduced cells could be detected for 68 days, supporting the beneficial effects of IL-15 for the persistence of CAR-NK cells.⁸² Undoubtedly, one of the most efficient mechanisms to undermine a durable tumor response is disposing of the target antigen. CAR-NK cells bispecific for EGFR and EGFRvIII were capable of dually hitting glioblastoma cells and extend the symptom-free survival of mice in an orthotopic xenograft model of glioblastoma.⁸³ Hence, multispecific CAR-NK cells offer a strategy to reduce the dependence on a single tumor antigen. Chang et al. constructed a CAR by assembling the extracellular part of the NKG2D receptor with the intracellular signaling domain DAP-10 linked in cis with the CD3ζ domain. After retroviral transfer into primary NK cells, ligation of the NKG2D-CAR resulted in a pronounced cytotoxicity and cytokine secretion against leukemia cells and solid tumors, while sparing nontransformed blood cells and mesenchymal cells. Intraperitoneally injected into a xenograft murine model of osteosarcoma, those CAR-NK cells markedly outperformed their untransduced counterparts by significantly reducing the overall tumor burden.⁸⁴ As the NKG2D receptor recognizes several different ligands, which are frequently upregulated during cellular stress,^48,52 NKG2D-CAR-NK cells, are also multispecific and may be less prone to antigen-loss-mediated blindfolding of CAR-effector cells.

CAR-NK cells – safety

Administering CAR-transfected T cells to cancer patients led to massive tumor regressions but has also raised awareness of the severe and potentially lethal side effects this therapy could entail. As for CAR-NK cells, no comparable adverse events were reported. Although being an obstacle for maintaining a durable response, the presumed limited persistence of mature CAR-NK-cells adds to the safety of this approach, rendering permanent side effects rather unlikely. NK cells with greater longevity, such as those derived from cord blood or hematopoietic stem cells, however, harbor a greater risk for long-term toxicities. Therefore, incorporating suicide systems into CAR-NK cells, which could prompt a rapid elimination of the transfected cells, is utilized to obviate such concerns. In a recent study the swift action of the inducible caspase 9 (iCAS-9) suicide systems was showcased, causing apoptosis in CD19-CAR, IL-15 and iCAS-9 triple transfected NK cells within 4 hours upon addition of the corresponding small molecule dimerizer.⁸² To temporally restrict CAR activity in NK cells, resorting to transient transfection methods (e.g., mRNA-electroporation) seems to be a conceivable option and several studies have reported on the successful transient receptor transfer.^59
–61,68 Autoreactivity from the endogenous NK cell receptor repertoire has not been deemed a crucial issue, as autologous nonmalignant cells are protected by engaging inhibitory receptors on NK cells and endogenous activity against autologous malignant cells may aid in tumor elimination. To minimize the occurrence of GVHD, allogeneic NK cells are obtained from HLA-matched or haploidentical donors. Nevertheless, HLA-matched NK cells infused to patients after nonmyeloablative stem cell transplantation induced grade 4 GVHD in several subjects.⁸⁵ On the other hand, NK cells can also ameliorate GVHD by killing host dendritic cells.⁸⁶ Nonetheless, the only way to find out about the true potential for autoreactivity and alloreactivity emanating from allogeneic CAR-NK cells are clinical trials.

CAR-NK cells – clinical trials

Various clinical trials evaluating safety and efficacy of CAR-NK cells against one solid tumor, and several hematological malignancies have been registered. To battle CD19⁺ malignant cells in 14 patients with refractory ALL, haploidentical NK cells were expanded with K562 cells expressing 4-1BBL and membrane-bound IL-15 and subsequently transduced with a CD19-specific CAR (NCT00995137). The trial has been completed and is now awaiting data presentation. A second trial (NCT01974479) led by Singaporean investigators involving CD19-CAR-NK cells targeting ALL has currently suspended participant recruiting. This year, a study predicated on cord blood derived NK cells engineered to express a CD19-specific CAR, IL-15 and inducible caspase 9, has been opened at Anderson cancer center, Texas, enrolling patients with relapsed or refractory B-cell lymphoma or leukemia. Modified NK cells will be administered after fludarabine and cyclophosphamide lymphodepletion (NCT03056339). Allogeneic NK-92 cells transfected with a CD19-specific third-generation CAR containing CD28 and 4-1BB domains, are currently being trialed in China for eradicating CD19⁺ lymphoma or leukemic cells prior to stem cell transplantation (NCT02892695). Moreover, NK-92 cells redirected with a CD28 and 4-1BB co-stimulated CAR to CD33 against AML (NCT02944162) and to CD7 against T-cell malignancies (NCT02742727) are tested. With respect to solid tumors, only one study has been opened so far by investigators from China recruiting patients with MUC1-positive relapsed or refractory solid tumors. In this study NK-92 expressing a MUC1-specific third-generation CAR will be administered (NCT02839954). With several trials set to gather clinical data over the next couple of years, CAR-NK cells will emerge as the best-characterized non-αβT cell CAR carrier for clinical application.

CAR-transfected myeloid cells

The myeloid compartment encompasses more than half of all blood leukocytes, predominantly monocytes, and granulocytes. Autologous blood monocyte-derived macrophages, administered to patients with pleural mesothelioma, proved to be safe but exhibited little antitumor activity.⁸⁷ Clinical trials directly involving receptor-engineered myeloid cells have not been conducted yet but some few preclinical studies exploring CAR-transfected myeloid cells were carried out.

Mature neutrophils are short-lived in vitro rendering direct receptor transfer difficult. Therefore, precursors were transduced and then differentiated to neutrophils. In an early study, neutrophils expressing a CD4-CD3ζ chimeric immunoreceptor were generated either in vivo from retrovirally transduced murine hematopoietic stem cells or in vitro from retrovirally transduced human hematopoietic stem cells.⁸⁸ In cytotoxicity assays, those transfected neutrophils mediated specific cytotoxicity against CD4⁺ targets.⁸⁸

Another group lentivirally engineered cord blood derived human CD34⁺ hematopoietic stem cells with either a first or a second-generation CD19-specific CAR. Following cytokine-driven differentiation into different myeloid cell types, the CAR was detected on monocytes, macrophages, and granulocytes. Receptor expression did not impair cell morphology and canonical myeloid functions. Additionally, antigen-specific killing of CD19⁺ targets was observed, with no difference between first- and second-generation CAR discernible.⁸⁹ After transplantation of CD19-CAR-engineered hematopoietic stem cells, CAR-expressing neutrophils are expected to appear after 2 weeks, providing early antigen-specific activity against remaining malignant cell.

Seeking to dissect the functionality of erbB2-specific-CAR-modified leukocytes in general, Yong et al. created transgenic mice harboring the CAR gene under the control of the vav promotor to restrict expression to hematopoietic cells. ErbB2-redirected macrophages engulfed erbB2-expressing tumor cells and antigen-specifically secreted pro-inflammatory cytokines, such as TNF and IL-6.⁹⁰

Another study specifically redirecting blood monocytes to tumor cells is based on the adenoviral integration of a CAR composed of a CEA-specific scFv and the intracellular domain of CD64. CAR activity induces antigen-specific TNF secretion and reduces tumor progression in vitro and in an in vivo mouse model.⁹¹

In summary, it is feasible to reprogram subsets of myeloid cells with CARs and evoke antigen-specific tumor activity in those cells. Contrary to lymphocytes, mature myeloid cells are devoid of proliferative capacity. Whereas clinically relevant numbers of macrophages can be obtained, in vitro culture of the short-lived granulocytes is not practicable. Regarding safety, a potential caveat arises from the cytokine profile exhibited by CAR macrophages, with a preponderance of TNF and IL-6, which are the primary drivers in cytokine release syndrome.⁷ Therefore, further studies on CAR-transfected myeloid cells are warranted to make more educated assessments about their role in future CAR therapy.

Conclusion

The advent of CAR-T cells has marked one of the biggest breakthroughs in cellular cancer immunotherapy, with unprecedented responses attained in patients suffering from relapsed and refractory hematological malignancies. However, these exciting results, which led to the U.S. Food and Drug Administration approving the use of CD19-CAR-T cells in the treatment of ALL this year, have not been equaled in the battle against solid tumors. Furthermore, the occurrence of severe side effects has stirred concerns about the safety of CAR-T-cell therapy. Hence, exploring immune cells beyond conventional αβT cells as suitable CAR drivers and exploiting their distinct advantages for CAR therapy has gained attention. CAR-engineered γδT cells may induce an antigen spread via presenting tumor antigens to conventional T cells and thereby prolong tumor regression. CAR-engineered NKT cells can also target the tumor microenvironments with the endogenous receptors and destroy TAMs. CAR-engineered mature NK cells are relatively short-lived in vivo obviating concerns on permanent target toxicities. Nevertheless, few clinical data have been published yet and the majority of the clinical trials involving non-αβCAR-T cells are in early stages. But the available preclinical data regarding CAR-engineered γδT cells, NKT cells, and NK cells clearly indicate the potential for those new vehicles to join the race and may even take over CAR-transfected αβT cells in certain situations.

Footnotes

Author Disclosure

No competing financial interests exist.

References

Robbins

, Morgan

, Feldman

, et al. Tumor regression in patients with metastatic synovial cell sarcoma and melanoma using genetically engineered lymphocytes reactive with NY-ESO-1. J Clin Oncol, 2011; 29:917–924.

Maher

, Brentjens

, Gunset

, et al. Human T-lymphocyte cytotoxicity and proliferation directed by a single chimeric TCRzeta /CD28 receptor. Nat Biotechnol, 2002; 20:70–75.

Turtle

, Hanafi

, Berger

, et al. CD19 CAR-T cells of defined CD4+:CD8+ composition in adult B cell ALL patients. J Clin Invest, 2016; 126:2123–2138.

Chmielewski

, Hombach

, Abken

. Of CARs and TRUCKs: chimeric antigen receptor (CAR) T cells engineered with an inducible cytokine to modulate the tumor stroma. Immunol Rev, 2014; 257:83–90.

Abken

Driving CARs on the Highway to Solid Cancer: Some Considerations on the Adoptive Therapy with CAR T Cells. Hum Gene Ther, 2017; 28:1047–1060.

Gardner

, Wu

, Cherian

, et al. Acquisition of a CD19-negative myeloid phenotype allows immune escape of MLL-rearranged B-ALL from CD19 CAR-T-cell therapy. Blood, 2016; 127:2406–2410.

Perales

, Kebriaei

, Kean

, et al. Building a Safer and Faster CAR: Seatbelts, Airbags and Crispr. Biol Blood Marrow Transplant, 2017; 24:27–31.

Lamers

, Sleijfer

, van Steenbergen

, et al. Treatment of metastatic renal cell carcinoma with CAIX CAR-engineered T cells: clinical evaluation and management of on-target toxicity. Mol Ther, 2013; 21:904–912.

Vantourout

, Hayday

. Six-of-the-best: unique contributions of gammadelta T cells to immunology. Nat Rev Immunol, 2013; 13:88–100.

10.

Mirzaei

, Mirzaei

, Lee

, et al. Prospects for chimeric antigen receptor (CAR) gammadelta T cells: A potential game changer for adoptive T cell cancer immunotherapy. Cancer Lett, 2016; 380:413–423.

11.

Nussbaumer

, Gruenbacher

, Gander

, et al. Essential requirements of zoledronate-induced cytokine and gammadelta T cell proliferative responses. J Immunol, 2013; 191:1346–1355.

12.

Kondo

, Izumi

, Fujieda

, et al. Expansion of human peripheral blood gammadelta T cells using zoledronate. J Vis Exp, 2011;(55):3182.

13.

Deniger

, Switzer

, Mi

, et al. Bispecific T-cells expressing polyclonal repertoire of endogenous gammadelta T-cell receptors and introduced CD19-specific chimeric antigen receptor. Mol Ther, 2013; 21:638–647.

14.

Lopez

, Xu

, Guo

, et al. CD2-mediated IL-12-dependent signals render human gamma delta-T cells resistant to mitogen-induced apoptosis, permitting the large-scale ex vivo expansion of functionally distinct lymphocytes: implications for the development of adoptive immunotherapy strategies. Blood, 2000; 96:3827–3837.

15.

Rischer

, Pscherer

, Duwe

, et al. Human gammadelta T cells as mediators of chimaeric-receptor redirected anti-tumour immunity. Br J Haematol, 2004; 126:583–592.

16.

Harrer

, Simon

, Fujii

, et al. RNA-transfection of gamma/delta T cells with a chimeric antigen receptor or an alpha/beta T-cell receptor: a safer alternative to genetically engineered alpha/beta T cells for the immunotherapy of melanoma. BMC Cancer, 2017; 17:551.

17.

Gober

, Kistowska

, Angman

, et al. Human T cell receptor gammadelta cells recognize endogenous mevalonate metabolites in tumor cells. J Exp Med, 2003; 197:163–168.

18.

D'Asaro

, La

, Di

, et al. V gamma 9V delta 2 T lymphocytes efficiently recognize and kill zoledronate-sensitized, imatinib-sensitive, and imatinib-resistant chronic myelogenous leukemia cells. J Immunol, 2010; 184:3260–3268.

19.

Rothenfusser

, Buchwald

, Kock

, et al. Missing HLA class I expression on Daudi cells unveils cytotoxic and proliferative responses of human gammadelta T lymphocytes. Cell Immunol, 2002; 215:32–44.

20.

Oevermann

, Lang

, Feuchtinger

, et al. Immune reconstitution and strategies for rebuilding the immune system after haploidentical stem cell transplantation. Ann N Y Acad Sci, 2012; 1266:161–170.

21.

Critchley-Thorne

, Simons

, Yan

, et al. Impaired interferon signaling is a common immune defect in human cancer. Proc Natl Acad Sci U S A, 2009; 106:9010–9015.

22.

Lehner

, Gotz

, Proff

, et al. Redirecting T cells to Ewing's sarcoma family of tumors by a chimeric NKG2D receptor expressed by lentiviral transduction or mRNA transfection. PLoS One, 2012; 7:e31210.

23.

Brandes

, Willimann

, Moser

. Professional antigen-presentation function by human gammadelta T Cells. Science, 2005; 309:264–268.

24.

Muto

, Baghdadi

, Maekawa

, et al. Myeloid molecular characteristics of human gammadelta T cells support their acquisition of tumor antigen-presenting capacity. Cancer Immunol Immunother, 2015; 64:941–949.

25.

Kobayashi

, Tanaka

, Shimmura

, et al. Complete remission of lung metastasis following adoptive immunotherapy using activated autologous gammadelta T-cells in a patient with renal cell carcinoma. Anticancer Res, 2010; 30:575–579.

26.

Godfrey

, Stankovic

, Baxter

. Raising the NKT cell family. Nat Immunol, 2010; 11:197–206.

27.

Kawano

, Cui

, Koezuka

, et al. CD1d-restricted and TCR-mediated activation of valpha14 NKT cells by glycosylceramides. Science, 1997; 278:1626–1629.

28.

Hayakawa

, Takeda

, Yagita

, et al. Differential regulation of Th1 and Th2 functions of NKT cells by CD28 and CD40 costimulatory pathways. J Immunol, 2001; 166:6012–6018.

29.

Gumperz

, Miyake

, Yamamura

, et al. Functionally distinct subsets of CD1d-restricted natural killer T cells revealed by CD1d tetramer staining. J Exp Med, 2002; 195:625–636.

30.

Shimizu

, Shinga

, Yamasaki

, et al. Transfer of mRNA encoding invariant NKT cell receptors imparts glycolipid specific responses to T cells and gammadeltaT cells. PLoS One, 2015; 10:e0131477.

31.

Takahashi

, Nieda

, Koezuka

, et al. Analysis of human V alpha 24+ CD4+ NKT cells activated by alpha-glycosylceramide-pulsed monocyte-derived dendritic cells. J Immunol, 2000; 164:4458–4464.

32.

Rogers

, Matsumoto

, Naidenko

, et al. Expansion of human Valpha24+ NKT cells by repeated stimulation with KRN7000. J Immunol Methods, 2004; 285:197–214.

33.

Tachibana

, Onodera

, Tsuruyama

, et al. Increased intratumor Valpha24-positive natural killer T cells: a prognostic factor for primary colorectal carcinomas. Clin Cancer Res, 2005; 11:7322–7327.

34.

Kawano

, Nakayama

, Kamada

, et al. Antitumor cytotoxicity mediated by ligand-activated human V alpha24 NKT cells. Cancer Res, 1999; 59:5102–5105.

35.

Quail

, Joyce

. Microenvironmental regulation of tumor progression and metastasis. Nat Med, 2013; 19:1423–1437.

36.

Song

, Asgharzadeh

, Salo

, et al. Valpha24-invariant NKT cells mediate antitumor activity via killing of tumor-associated macrophages. J Clin Invest, 2009; 119:1524–1536.

37.

Heczey

, Liu

, Tian

, et al. Invariant NKT cells with chimeric antigen receptor provide a novel platform for safe and effective cancer immunotherapy. Blood, 2014; 124:2824–2833.

38.

Terabe

, Matsui

, Noben-Trauth

, et al. NKT cell-mediated repression of tumor immunosurveillance by IL-13 and the IL-4R-STAT6 pathway. Nat Immunol, 2000; 1:515–520.

39.

Gattinoni

, Klebanoff

, Restifo

. Paths to stemness: building the ultimate antitumour T cell. Nat Rev Cancer, 2012; 12:671–684.

40.

Tian

, Courtney

, Jena

, et al. CD62L+ NKT cells have prolonged persistence and antitumor activity in vivo . J Clin Invest, 2016; 126:2341–2355.

41.

Gottschalk

, Mettke

, Kurts

. The role of invariant natural killer t cells in dendritic cell licensing, cross-priming, and memory CD8(+) T cell generation. Front Immunol, 2015; 6:379.

42.

Motohashi

, Ishikawa

, et al. A phase I study of in vitro expanded natural killer T cells in patients with advanced and recurrent non-small-cell lung cancer. Clin Cancer Res, 2006; 12:6079–6086.

43.

Pillai

, George

, Dutt

, et al. Host NKT cells can prevent graft-versus-host disease and permit graft antitumor activity after bone marrow transplantation. J Immunol, 2007; 178:6242–6251.

44.

Beristain-Covarrubias

, Canche-Pool

, Gomez-Diaz

, et al. Reduced iNKT cells numbers in type 1 diabetes patients and their first-degree relatives. Immun Inflamm Dis, 2015; 3:411–419.

45.

van der Vliet

, von Blomberg

, Nishi

, et al. Circulating V(alpha24+) Vbeta11+ NKT cell numbers are decreased in a wide variety of diseases that are characterized by autoreactive tissue damage. Clin Immunol, 2001; 100:144–148.

46.

Louis

, Savoldo

, Dotti

, et al. Antitumor activity and long-term fate of chimeric antigen receptor-positive T cells in patients with neuroblastoma. Blood, 2011; 118:6050–6056.

47.

Mandal

, Viswanathan

. Natural killer cells: In health and disease. Hematol Oncol Stem Cell Ther, 2015; 8:47–55.

48.

Morvan

, Lanier

. NK cells and cancer: You can teach innate cells new tricks. Nat Rev Cancer, 2016; 16:7–19.

49.

Bollino

, Webb

. Chimeric antigen receptor-engineered natural killer and natural killer T cells for cancer immunotherapy. Transl Res, 2017; 187:32–43.

50.

Kloss

, Oberschmidt

, Morgan

, et al. Optimization of human NK cell manufacturing: fully automated separation, improved ex vivo expansion using IL-21 with autologous feeder cells, and generation of anti-CD123-CAR-expressing effector cells. Hum Gene Ther, 2017; 28:897–913.

51.

Voskens

, Watanabe

, Rollins

, et al. Ex-vivo expanded human NK cells express activating receptors that mediate cytotoxicity of allogeneic and autologous cancer cell lines by direct recognition and antibody directed cellular cytotoxicity. J Exp Clin Cancer Res, 2010; 29:134.

52.

Sutlu

, Stellan

, Gilljam

, et al. Clinical-grade, large-scale, feeder-free expansion of highly active human natural killer cells for adoptive immunotherapy using an automated bioreactor. Cytotherapy, 2010; 12:1044–1055.

53.

Oelsner

, Friede

, Zhang

, et al. Continuously expanding CAR NK-92 cells display selective cytotoxicity against B-cell leukemia and lymphoma. Cytotherapy, 2017; 19:235–249.

54.

Klingemann

, Boissel

, Toneguzzo

. Natural killer cells for immunotherapy - advantages of the NK-92 cell line over blood NK cells. Front Immunol, 2016; 7:91.

55.

Lanier

LL.

NKG2D receptor and its ligands in host defense. Cancer Immunol Res, 2015; 3:575–582.

56.

Kailayangiri

, Altvater

, Spurny

, et al. Targeting Ewing sarcoma with activated and GD2-specific chimeric antigen receptor-engineered human NK cells induces upregulation of immune-inhibitory HLA-G. Oncoimmunology, 2017; 6:e1250050.

57.

Imai

, Iwamoto

, Campana

. Genetic modification of primary natural killer cells overcomes inhibitory signals and induces specific killing of leukemic cells. Blood, 2005; 106:376–383.

58.

Suerth

, Morgan

, Kloess

, et al. Efficient generation of gene-modified human natural killer cells via alpharetroviral vectors. J Mol Med (Berl), 2016; 94:83–93.

59.

, Liu

, Feller

, et al. Expression of chimeric antigen receptors in natural killer cells with a regulatory-compliant non-viral method. Cancer Gene Ther, 2010; 17:147–154.

60.

Shimasaki

, Fujisaki

, Cho

, et al. A clinically adaptable method to enhance the cytotoxicity of natural killer cells against B-cell malignancies. Cytotherapy, 2012; 14:830–840.

61.

Chu

, Hochberg

, Yahr

, et al. Targeting CD20+ aggressive B-cell non-Hodgkin lymphoma by anti-CD20 CAR mRNA-modified expanded natural killer cells in vitro and in NSG mice. Cancer Immunol Res, 2015; 3:333–344.

62.

Cho

, Chang

, Shu

, et al. Enhanced cytotoxicity of natural killer cells following the acquisition of chimeric antigen receptors through trogocytosis. PLoS One, 2014; 9:e109352.

63.

Chu

, Deng

, Benson

, et al. CS1-specific chimeric antigen receptor (CAR)-engineered natural killer cells enhance in vitro and in vivo antitumor activity against human multiple myeloma. Leukemia, 2014; 28:917–927.

64.

Kruschinski

, Moosmann

, Poschke

, et al. Engineering antigen-specific primary human NK cells against HER-2 positive carcinomas. Proc Natl Acad Sci U S A, 2008; 105:17481–17486.

65.

Altvater

, Landmeier

, Pscherer

, et al. 2B4 (CD244) signaling by recombinant antigen-specific chimeric receptors costimulates natural killer cell activation to leukemia and neuroblastoma cells. Clin Cancer Res, 2009; 15:4857–4866.

66.

Topfer

, Cartellieri

, Michen

, et al. DAP12-based activating chimeric antigen receptor for NK cell tumor immunotherapy. J Immunol, 2015; 194:3201–3212.

67.

Muller

, Michen

, Tietze

, et al. Engineering NK cells modified with an EGFRvIII-specific chimeric antigen receptor to overexpress CXCR4 improves immunotherapy of CXCL12/SDF-1alpha-secreting glioblastoma. J Immunother, 2015; 38:197–210.

68.

Boissel

, Betancur

, Lu

, et al. Comparison of mRNA and lentiviral based transfection of natural killer cells with chimeric antigen receptors recognizing lymphoid antigens. Leuk Lymphoma, 2012; 53:958–965.

69.

Boissel

, Betancur-Boissel

, Lu

, et al. Retargeting NK-92 cells by means of CD19- and CD20-specific chimeric antigen receptors compares favorably with antibody-dependent cellular cytotoxicity. Oncoimmunology, 2013; 2:e26527.

70.

Muller

, Uherek

, Maki

, et al. Expression of a CD20-specific chimeric antigen receptor enhances cytotoxic activity of NK cells and overcomes NK-resistance of lymphoma and leukemia cells. Cancer Immunol Immunother, 2008; 57:411–423.

71.

Jiang

, Zhang

, Shang

, et al. Transfection of chimeric anti-CD138 gene enhances natural killer cell activation and killing of multiple myeloma cells. Mol Oncol, 2014; 8:297–310.

72.

Esser

, Muller

, Stefes

, et al. NK cells engineered to express a GD2 -specific antigen receptor display built-in ADCC-like activity against tumour cells of neuroectodermal origin. J Cell Mol Med, 2012; 16:569–581.

73.

Uherek

, Tonn

, Uherek

, et al. Retargeting of natural killer-cell cytolytic activity to ErbB2-expressing cancer cells results in efficient and selective tumor cell destruction. Blood, 2002; 100:1265–1273.

74.

Schonfeld

, Sahm

, Zhang

, et al. Selective inhibition of tumor growth by clonal NK cells expressing an ErbB2/HER2-specific chimeric antigen receptor. Mol Ther, 2015; 23:330–338.

75.

Liu

, Yang

, Sun

, et al. Specific growth inhibition of ErbB2expressing human breast cancer cells by genetically modified NK92 cells. Oncol Rep, 2015; 33:95–102.

76.

Crane

, Han

, Barry

, et al. TGF-beta downregulates the activating receptor NKG2D on NK cells and CD8+ T cells in glioma patients. Neuro Oncol, 2010; 12:7–13.

77.

Wang

, Guo

, Song

, et al. Augmented anti-tumor activity of NK-92 cells expressing chimeric receptors of TGF-betaR II and NKG2D. Cancer Immunol Immunother, 2017; 66:537–548.

78.

Chen

, Han

, Chu

, et al. A combinational therapy of EGFR-CAR NK cells and oncolytic herpes simplex virus 1 for breast cancer brain metastases. Oncotarget, 2016; 7:27764–27777.

79.

Chen

, Wada

, Pinz

, et al. Preclinical targeting of aggressive T-cell malignancies using anti-CD5 chimeric antigen receptor. Leukemia, 2017; 31:2151–2160.

80.

Lowe

, Truscott

, De Oliveira

. In vitro generation of human NK cells expressing chimeric antigen receptor through differentiation of gene-modified hematopoietic stem cells. Methods Mol Biol, 2016; 1441:241–251.

81.

Sahm

, Schonfeld

, Wels

. Expression of IL-15 in NK cells results in rapid enrichment and selective cytotoxicity of gene-modified effectors that carry a tumor-specific antigen receptor. Cancer Immunol Immunother, 2012; 61:1451–1461.

82.

Liu

, Tong

, Dotti

, et al. Cord blood NK cells engineered to express IL-15 and a CD19-targeted CAR show long-term persistence and potent antitumor activity. Leukemia, 2017 Jul 20. doi: 10.1038/leu.2017.226. [Epub ahead of print].

83.

Genssler

, Burger

, Zhang

, et al. Dual targeting of glioblastoma with chimeric antigen receptor-engineered natural killer cells overcomes heterogeneity of target antigen expression and enhances antitumor activity and survival. Oncoimmunology, 2016; 5:e1119354.

84.

Chang

, Connolly

, Shimasaki

, et al. A chimeric receptor with NKG2D specificity enhances natural killer cell activation and killing of tumor cells. Cancer Res, 2013; 73:1777–1786.

85.

Shah

, Baird

, Delbrook

, et al. Acute GVHD in patients receiving IL-15/4-1BBL activated NK cells following T-cell-depleted stem cell transplantation. Blood, 2015; 125:784–792.

86.

Ruggeri

, Mancusi

, Burchielli

, et al. NK cell alloreactivity and allogeneic hematopoietic stem cell transplantation. Blood Cells Mol Dis, 2008; 40:84–90.

87.

Monnet

, Breau

, Moro

, et al. Intrapleural infusion of activated macrophages and gamma-interferon in malignant pleural mesothelioma: a phase II study. Chest, 2002; 121:1921–1927.

88.

Roberts

, Cooke

, Tran

, et al. Antigen-specific cytolysis by neutrophils and NK cells expressing chimeric immune receptors bearing zeta or gamma signaling domains. J Immunol, 1998; 161:375–384.

89.

De Oliveira

, Ryan

, Giannoni

, et al. Modification of hematopoietic stem/progenitor cells with CD19-specific chimeric antigen receptors as a novel approach for cancer immunotherapy. Hum Gene Ther, 2013; 24:824–839.

90.

Yong

, John

, Devaud

, et al. A role for multiple chimeric antigen receptor-expressing leukocytes in antigen-specific responses to cancer. Oncotarget, 2016; 7:34582–34598.

91.

Biglari

, Southgate

, Fairbairn

, et al. Human monocytes expressing a CEA-specific chimeric CD64 receptor specifically target CEA-expressing tumour cells in vitro and in vivo . Gene Ther, 2006; 13:602–610.