T cell receptor mimic antibodies for cancer immunotherapy

Abstract

Antibody-based immunotherapies show clinical effectiveness in various cancer types. However, the target repertoire is limited to surface or soluble antigens which are a relatively small percentage of the cancer proteome. Most proteins of the human proteome are intracellular. Short peptides from intracellular targets can be presented by major histocompatibility complex class I (MHC-I) molecules on cell surface, making them potential targets for cancer immunotherapy. Antibodies can be developed to target these peptide/MHC complexes, similar to the recognition of such complexes by the T cell receptor (TCR). These antibodies are referred to as T cell receptor mimic (TCRm) or TCR-like antibodies. Ongoing preclinical and clinical studies will help understand their mechanisms of action and selection of target epitopes for immunotherapy. This review will summarize and discuss the selection of intracellular antigens, production of the peptide/MHC complexes, isolation of TCRm antibodies for therapeutic applications, limitations of TCRm antibodies, and possible ways to advance TCRm antibody-based approaches in the clinic.

Introduction

A cancer proteome study analyzing the transcriptome of 17 major cancer types revealed more than 500 genes implicated in the transformation of normal cells to cancer cells, and nearly half of those genes encoded for intracellular proteins (1). Some proteins are on the cell surface, and they can be targeted by cancer immunotherapies including monoclonal antibody and chimeric antigen receptor (CAR) T-cell therapy. These extracellular proteins may include CD19, CD20, CD22, CD47, CD52, CD80, CD123, HER2, EGFR, mesothelin, GPC3, GPC2, GD2, B7-H3 and PD1/PDL1 (2–18). However, the intracellular cancer targets are not well accessible to those immunotherapies. Therefore, there is a need for the development of agents capable of targeting intracellular proteins. Some antibody-based approaches have been developed for intracellular antigens, including targeting externalized intracellular antigens and delivering the antibody into the cell cytoplasm by liposomal vehicles (19,20). However, there are limitations to these approaches. Only a few intracellular proteins are externalized by cancer cells, and it is challenging to retain the stability and binding capacity of the intracellular antibody to the target antigen within the cell. The protein encapsulation efficiency for liposomes or other materials like nanoparticles is still low and those materials might cause toxicities (21).

Another emerging approach is the development of antibodies to target the peptides derived from intracellular antigens. Intracellular proteins can be degraded by the proteasome into short peptides, commonly 8–10 amino acids long, which are presented on the cell surface by major histocompatibility complex class I (MHC-I), also known as human leukocyte antigen (HLA) system in humans (22). Some peptide/MHC complexes (pMHC) have been demonstrated to be implicated in various cancer and targeted by T cell receptor (TCR) therapy. These complexes can be utilized for antibody development. Antibodies targeting pMHC are commonly called TCR mimic (TCRm) or TCR-like antibodies referring to their ability to recognize the complex like TCR’s on T cells (23). The peptides originated from various intracellular tumor antigens such as viral oncogene products, transcription factors, oncofetal proteins, cancer-testis antigens, or neoantigens from mutated oncogenes. TCRm antibodies expand the range of therapeutic targets and have broad clinical potential. This review focuses primarily on the development of TCRm antibodies against pMHC complexes for cancer immunotherapy.

Identification of pMHC targets in cancer

When selecting a pMHC target for TCRm antibody development, we may consider the following factors: the abundance of the epitope, the specificity for cancer cells versus normal cells, the presentation of MHC molecules, and the heterogeneity of expression on tumor cells. The desired target for a TCRm antibody is a cancer-specific pMHC complex present at high density on the target cell surface while being absent on normal cells. The target antigen may have a functional or even crucial role in tumor growth to help avoid losing the antigen under subsequent therapeutic selection pressure. The peptide may also have a high affinity for the patient’s MHC and form a stable complex that persists on the cell surface, allowing for recognition by TCRm antibodies. A major challenge for TCRm antibody development is that tumor cells tend to downregulate MHC I expression. Consequently, the abundance of the pMHC complexes is generally low (24–26). However, a recent study showed that the TCRm antibody had potent anti-tumor activity against cancer cells with KRAS G12V MHC complex despite having a low number of the complex on cell surface (an average of three to ten copies per cell) (24).

HLA is a major part of the pMHC complex, making it very important to choose an HLA for cancer therapy. There are more than 20,000 HLA class I alleles in different human populations (27). An oncogenic peptide might be presented by different HLA alleles in different individuals. It should be a priority to develop TCRm antibody for peptides complexed with major HLA subtypes to broaden its clinical application. According to a study in Brazil (28), the most common HLA class I alleles found in Caucasians were HLA-A*02, 24, 01 and HLA-B*35, 44, 51; for Asians they were HLA-A*24, 02, 26 and HLA-B*40, 51, 52; for Afro-Brazilians they were HLA-A*02, 03, 30 and HLA-B*35, 15, 44. Apparently, HLA-A*02 is the most common with 27% among all the alleles at the HLA-A locus, making it a suitable HLA candidate for TCRm antibody development. Based on the origin of their short peptide component, the pMHC complexes can be classified broadly into the following two major groups: tumor associated antigens (TAAs) including overexpressed proteins, and tumor specific antigens (TSAs) (Table.1). TSAs include specific fusion proteins, oncogenic viral antigens, and mutated self-antigens. The antigens listed in the Table 1 have been validated to express pMHC complexes on cell surface making them validated targets for TCR cell therapy or for TCRm antibody development. The discovery and validation of the pMHC complex on cells normally uses sophisticated proteomics methods involving mass spectrometry (26). Consequently, there are not many antigens available at the moment. In addition, the vast majority of pMHC antigens are still unknown, which is a huge limitation in the development of TCRm antibodies.

Table 1.

Current targets for TCRm antibodies

Target class	Examples	Tumor type
Overexpressed proteins	WT1	Leukemia, ovarian cancer, colon cancer, mesothelioma
	AFP	Hepatocellular carcinoma
	PRAME	Leukemia, lymphoma, melanoma, breast cancer, colon cancer
	MAGE	Melanoma
	NY-ESO-1	Melanoma
Fusion proteins	BCR-ABL	Chronic myeloid leukemia
	PML-RARa	Acute promyelocytic leukemia
Oncogenic viral antigens	EBV proteins	EBV associated cancers: Burkitt’s lymphoma, Hodgkin’s lymphoma, and nasopharyngeal carcinoma
	HPV 16 E6/E7	HPV16+ associated cancers: cervical, oral and oropharyngeal cancers
	CMV proteins	Various cancers including breast, colon, prostate and glioblastoma
	HBV proteins	Hepatocellular carcinoma
Mutated antigens	Ras	Various cancers including pancreatic, lung, and colorectal cancers
	p53	Various cancers including breast, bladder, lung and brain tumors
	Myc	Various cancers including breast, colorectal, pancreatic and gastric cancers

Tumor associated antigens (TAAs)

TAAs are preferentially expressed by tumor cells and associated with a malignant cell phenotype, which is the basis of targeted cancer immunotherapy. Aberrantly overexpressed antigens are the largest group of TAAs. Studies have shown some success in targeting these TAAs by vaccines, adoptive cellular therapies and antibodies including oncofetal antigens like Wilms’ tumor 1 (WT1)(29), alpha-fetoprotein (AFP) (30) and GPC3 (31), and cancer-testis antigens such as the Preferentially Expressed Antigen in Melanoma (PRAME) (32) and the New York Esophageal Squamous Cell Carcinoma 1 (NY-ESO-1) (33), and those from the Melanoma Antigen Gene (MAGE) family (29,30,32–38). Oncofetal antigens WT1 and AFP are expressed during embryonic development and found in specific tumors but have substantially limited expression in healthy adult tissues. Cancer-testis antigens including those from the MAGE gene family and NY-ESO1 are highly expressed in several tumor types but their expression in healthy adult tissue is restricted to the testes. These TAAs have been well-documented and validated as targets for cytotoxic CTL, TCR T therapy, or TCRm antibodies (29,30,32,34–37).

Tumor specific antigens (TSAs)

TSAs are attractive targets for cancer immunotherapy since TSAs are expressed specifically by tumor cells, making therapeutic effects from immune approaches limited to tumors and mitigating off-target toxicities. Some fusion proteins are explicitly expressed by tumor cells, such as the breakpoint cluster region-proto-oncogene tyrosine-protein kinase (BCR-ABL) and the promyelocytic leukemia-retinoic acid receptor alpha fusion (PML-RARa). BCR-ABL fusion (Philadelphia chromosome) results from a translocation between chromosomes 9 and 22 t (9;22) (q34;q11). This translocation is critical for the pathogenesis of CML and can be detected in 95% of patients with CML while it is absent in normal cells (12,39). The PML-RARα fusion is caused a translocation of (15;17) and is a hallmark feature of acute promyelocytic leukemia (APL). The fusion protein created by this translocation functions as an abnormal retinoid receptor with aberrant transcriptional regulatory properties (40,41). Some studies using a BCR-ABL peptide vaccine have shown limited effects in clinical trials (42,43). However, immunotherapy approaches targeting those antigens still need further investigation.

Another group of TSAs is oncogenic viral antigens. Approximately 12% of all human cancers are caused by oncoviruses. There are seven oncoviruses including: Epstein–Barr virus (EBV), human papillomaviruses (HPVs), hepatitis B virus (HBV), hepatitis C virus (HCV), human T-lymphotropic virus 1 (HTLV-1), Kaposi sarcoma-associated herpesvirus (KSHV; also known as human herpesvirus 8 (HHV-8)) and Merkel cell polyomavirus (MCPyV) (44). Human viral oncogenesis is complicated and only a small percentage of infected individuals develop cancer, often many years or decades after the initial infection. The oncogenes from those viruses are constitutively expressed by cancer cells, making them attractive targets for cancer immunotherapy. Developing TCRm antibodies for EBV-associated cancers have shown to be feasible (45,46). EBV is associated with a number of human cancers such as Burkitt’s lymphoma, Hodgkin’s lymphoma, and nasopharyngeal carcinoma. The peptides from EBV proteins, including EBNA1_562–570, LMP1_125–133, and LMP2A_426–424 are presented by HLA-A*0201 on the cell surface and are potential targets for TCRm antibodies. Results showed that the xenograft growth of EBV transformed B lymphoblastoid cells in mice was inhibited and mice showed improved survival when treated with TCRm antibodies targeting those EBV peptides (45). Studies using TCR cell therapy to target HPV E6 or E7 peptide complex, exclusively expressed by HPV-associated cancers, demonstrated an excellent strategy to inhibit HPV-related epithelial cancers by targeting those oncoproteins (47,48). It would be interesting to investigate whether TCRm antibodies against the HPV pMHC complexes could inhibit HPV-related epithelial cancers.

Mutated self-antigens, or neoantigens, are excellent targets for cancer immunotherapy since they are specific for malignant cells. However, the majority of tumor mutations are not shared between patients and a small population of patients share the same neoantigen in most cases. It is difficult and challenging to develop TCRm antibodies for those diverse neoantigens since the clinical application is limited. Fortunately, research has revealed recurrent mutations (>1% of patients), which can be the recurrent neoantigens, such as p53, KRAS, and Myc (49–52). The tumor suppressor p53 is dysregulated in various cancers and a valuable target for immunotherapy. The p53 has two common mutations in colorectal cancer: R175H and R282W, with 9.5% frequency for each mutation (25,53). A recent study developed a TCRm antibody targeting p53 R175H mutation. The bispecific antibody derived from this TCRm antibody effectively activated T cells to lyse cancer cells with this mutation (25). Another study has isolated a TCRm antibody T1–116C targeting the p53 peptide complex and demonstrated its effectiveness in vitro and in vivo (52,54). KRAS mutations are enriched in various cancers, such as: pancreatic cancer (70%–90% of cases), colon cancer (35%–50%), small intestinal cancer (35%), biliary cancer (20%–30%), and lung cancer (20%–35%). Furthermore, specific KRAS mutations dominate in one type of tumor (55). KRAS G12D is the most common mutation in 32.4% of pancreatic cancer patients and G12V is the 2^nd most common mutations and has a percentage of 22% (55). A TCRm antibody was identified to target KRAS G12V mutant peptide complexed with HLA*A3. The bispecific antibody derived from this TCRm antibody effectively killed tumor cells (24).

Production of the pMHC complex

The pMHC complex is not a single protein; instead, it consists of MHC molecules, including HLA class I heavy chain, β2-microglobulin(β₂M), and the peptide from a tumor antigen. The peptide is loaded into the binding groove of the HLA molecule and the conserved hydrogen bonds between the sidechains of the MHC molecule and the backbone of the peptide creates the binding force (56). A pMHC complex in its native conformation is crucial to isolate a successful TCRm antibody. Producing the pMHC complexes as antigens for antibody screening using current technologies such as hybridoma or phage display is challenging, resultingly TCRm antibodies are not as commonly available as traditional antibodies. Initially, antigen-presenting cells harboring immunogenic peptides in the groove of their MHC molecules were used as immunogens with limited success in the generation of TCRm antibodies (57,58).

The availability of recombinant pMHC complexes has advanced the development of TCRm antibodies. To isolate antibodies using different strategies, a relatively high amount of purified MHC-peptide complexes are needed. At first, the MHC molecules including the heavy chain and β₂M are expressed separately in the bacteria and are purified from inclusion bodies. They are then refolded with the MHC-restricted peptide to generate correctly refolded monomers (Fig. 1B) (59,60). Structural and functional experiments are performed to verify the refolding of the monomers (61,62). Overall, the yield of those monomers is low and the production is tedious in contrast with other antigen preparation from bacterial or mammalian expression. To increase production efficiency, expression of a fusion protein consisting of the MHC heavy chain and β₂M was explored (63,64). The fusion protein has a flexible linker and can be loaded with the target peptide. Furthermore, another fusion protein has been engineered as a single chain trimer (SCT) which incorporates all three subunits of the MHC complex through two linkers (65,66). These chimeric complexes can be recognized by diverse TCRs, similar as native pMHC complex. A structural study revealed that there are some differences on the structure between SCT and the pMHC monomer, especially on the α1α2 domains. However, the differences on the antigen-binding platform might be small (67). The primary use of these fusion proteins is to generate tetramers or other multivalent staining reagents for visualizing T cells. A native conformation of the pMHC monomer may not be important in the tetramer application (66,67). Nevertheless, a native pMHC monomer may be the most common candidate for TCRm antibody screening. Development and optimization of pMHC antigen production would be the important area of research and technology improvement for TCRm antibody discovery.

Figure 1.

Flowchart of TCRm antibody development and preparation of pMHC complexes. A) To develop TCRm antibodies, it is important to identify the pMHC complexes that are associated with tumor. After production of the pMHC complexes, they can be used as antigens for antibody discovery by immunization and hybridoma methods, or phage display technologies, or even human single B cell cloning strategy. Those TCRm antibodies can be further developed into different formats like bispecific antibodies, ADCs, immunotoxins, or CAR T-cell therapy for cancer therapeutics. B) The antigen production of pMHC is important for TCRm antibody development. MHC class I molecules and β−2-microglobulin (β2m) are expressed and purified from bacteria. They are then assembled with a tumor associated or specific peptide to form a pMHC complex monomer that is used as an antigen for TCRm antibody screening.

Isolation and production of TCRm antibodies

TCRm antibodies have been traditionally difficult to generate. However, due to recent advances including the production of antigen as discussed above, there has been an increase in the development of TCRm antibodies targeting a growing repertoire of tumor and viral antigens including: WT1, AFP, PRAME, NY-ESO-1, MAGEA1, hTERT, TARP, Tyrosinase, hCGbeta, p53, p68 MIF, proteinase 3, MAGE3, and EBV proteins (Table 2) (33,37,38,52,68–80). Immunization and hybridoma strategy, and in vitro screening by phage display are the two main methods used to isolate TCRm antibodies (Fig. 1A). Alternatively, single B cell sorting and cloning strategy can be an alternative for the TCRm antibody isolation.

Table 2.

TCRm antibodies in cancer immunotherapy

Target	tumor types	Peptide sequence	MHC haplotype	Isotype/format	isolation method	Institute/Company	Status	Reference
WT1	Leukemia, ovarian, colon, mesothelioma	RMFPNAPYL	HLA-A*0201	hIgG1	Phage	MSKCC/Eureka	Preclinical	(29)
WT1	Leukemia	RMFPNAPYL	HLA-A*0201	Fab	Phage	Technion-Israel Institute of Technology	Preclinical	(76)
WT1	Leukemia	RMFPNAPYL	HLA-A*0201	scFv	Phage	MSKCC	Preclinical	(38)
AFP	Hepatocyte carcinoma	FMNKFIYEI	HLA-A*0201	scFv-CAR	Phage	Eureka	Clinical Phase I/II NCT03998033	(30)
PRAME	Leukemia, lymphoma, melanoma, breast, colon	ALYVDSLFFL	HLA-A*0201	hIgG1	Phage	MSKCC	Preclinical	(32)
NY-ESO-1	Melanoma	SLIMWITQC	HLA-A*0201	Fab	Phage	Saarland University Medical School	Preclinical	(33)
MAGEA1	Melanoma	EADPTGHSY	HLA-A*0101	Fab	Phage	Leiden University	Preclinical	(68)
MAGEA1	Melanoma	EADPTGHSY	HLA-A*0101	Fab	Phage	Maastricht University	Preclinical	(69)
hTERT	Melanoma, prostate	ILAKFLHWL	HLA-A*0201	Fab	Phage	Technion-Israel Institute of Technology	Preclinical	(72)
hTERT	Melanoma, prostate	RLVDDFLLV	HLA-A*0201	Fab	Phage	Technion-Israel Institute of Technology	Preclinical	(72)
TARP	Breast, prostate	FLRNFSLML	HLA-A*0201	Fab-PE38	Phage	Technion-Israel Institute of Technology	Preclinical	(70)
Tyrosinase	Melanoma	YMDGTMSQV	HLA-A*0201	Fab	Phage	Technion-Israel Institute of Technology	Preclinical	(74)
Ras	Lung cancer	VVVGAVGVGK	HLA-A*03:01	scFv	Phage	Johns Hopkins University	Preclinical	(24)
P53	myeloma	HMTEVVRHC	HLA-A*0201	scFv	Phage	Johns Hopkins University	Preclinical	(25)
P53	Breast	RMPEAAPPV	HLA-A*0201	IgG1	Hybridoma	University of Oxford	Preclinical	(52)
P53	Breast	RMPEAAPPV	HLA-A*0201	IgG1, IgG2b	Hybridoma	University of Oxford	Preclinical	(73)
P53	Breast	GLAPPQHLIRV	HLA-A*0201	IgG1, IgG2a	Hybridoma	University of Oxford	Preclinical	(52)
hCGbeta	Ovarian, colon, breast	TMTRVLQGV	HLA-A*0201	IgG1	Hybridoma	Texas Tech University Health Sciences Center	Preclinical	(75)
p68	Breast	YLLPAIVHI	HLA-A*0201	mIgG2a	Hybridoma	Texas Tech University Health Sciences Center	Preclinical	(79)
MIF	Breast	FLSELTQQL	HLA-A*0201	IgG2a	Hybridoma	Texas Tech University Health Sciences Center	Preclinical	(71)
Proteinase 3	AML	VLQELNVTV	HLA-A*0201	IgG2a	Hybridoma	M. D. Anderson Cancer Center	Preclinical	(77)
MAGE3	Melanoma	FLWGPRALV	HLA-A*0201		Hybridoma	Universite Nantes	Preclinical	(37)
EBV	EBV associated cancer	FMVFLQTHI	HLA-A*0201	IgG1	Hybridoma	National University of Singapore	Preclinical	(45)
EBV	EBV associated cancer	CLGGLLTMV	HLA-A*0201	IgG1	Hybridoma	National University of Singapore	Preclinical	(45)
EBV	EBV associated cancer	YLLEMLWRL	HLA-A*0201	IgG1	Hybridoma	National University of Singapore	Preclinical	(45)

The immunization methods with hybridoma technology were first adopted to produce TCRm antibodies (57). Early on, the antigens used were APC cells presenting the pMHC complexes (57,58). Few TCRm antibodies were generated using this method and many efforts proved to be unsuccessful. Several groups started to use the recombinant pMHC complexes in the immunization methods (Table 2) (37,71,75,77,79–81). For each screening, there might be thousands of clones and high-throughput screening was performed to identify the specific TCRm antibodies. The major advantage of the hybridoma technology is the high probability of isolating antibodies with high affinity binding to the pMHC complexes. If the animal is immunized with the antigen multiple times before sacrificing for the cell clones, there is a process of in vivo affinity maturation of antibodies. However, this approach might only work better with stable MHC complexes binding to the peptide with high affinity, which is necessary for the induction of effective immunization and in vivo IgG maturation. In addition, humanization will be an extra step if the antibody needs to be advanced into the clinic.

More recently, antibody phage display provided a breakthrough and showed success in the isolation of TCRm antibodies (29,30,32). Specificity is important for the success of TCRm antibodies; therefore, phage display can combine negative and positive selections to isolate specific antibodies for a target pMHC complex while hybridoma normally cannot do that. Another vital part of the phage display is the size of an antibody library, which normally possesses a large diversity. The antibody format in the library is usually an antigen-binding fragment (Fab) or single-chain variable fragment (scFv) fragment, displayed on each phage particle. Through several rounds of phage panning, phage particles carrying specific antibodies are selected and enriched against the target. One of the first reports with the phage display approach is the isolation of an antibody against influenza virus hemagglutinin peptide HA_255–262 (FESTGNLI) presented by a murine MHC class I (61). With the advancement of technology, several cancer-related pMHC complexes have been targeted for TCRm antibody development using phage display technology (24,25,30,82). The major advantage of this approach is the power of extensive selection toward the target which is achieved within a relatively short procedure. Another advantage is the generation of fully humanized antibodies, which does not require the antibody humanization step that is normally needed in the hybridoma strategy. However, TCRm antibodies from phage libraries usually possess a lower binding affinity compared to the hybridoma strategy. It might be necessary to perform the affinity maturation in vitro to improve the affinity and anti-tumor activity.

Structural insights of TCRm antibodies revealed from their complexes with MHC

Structural information on the binding of TCRm antibodies to the pMHC complex is valuable for the development of therapy for genetically diverse patient populations. TCRs have moderate affinity to the complexes ranging from 1 to 100 μM while TCRm antibodies can have much higher affinity. For example, ESK1, a TCRm antibody against WT-1, has a subnanomolar affinity, 1000-fold higher than that of natural TCRs (29). Comparing the binding to the pMHC complex with TCRs and other TCRm Fabs, ESK1 has a different mode (83). The variable domains of ESK1 bind the regions of HLA that TCRs typically do not reach. In addition, ESK1 binds part of the peptide, which accounts for only 15% of the total pMHC contact surface. The percentage of binding on the peptide is different among various TCRm antibodies. A recent study examined the binding structure of TCRm antibody H2 binding to p53 R175H peptide complexed with HLA-A*02:01 and found H2-Fab binds to both the peptide and HLA-A*02, around 50% of the contact surface respectively (25). In addition, the crystal structural study revealed that ESK1 binds largely to HLA receptor along with the peptide, with 85% of the total complex contact region on the HLA. Interestingly, this region is conserved among HLA-A*02 subtypes. It was predicted that ESK1 could bind to other subtypes of the HLA-A*02 family and confirmed by affinity assays (83). This potentially broadens the target patient populations with different HLA-A*02 subtypes. One major limitation of TCR-based drugs besides soluble TCR-based constructs and TCR engineered cells is their restriction to a specific HLA subtype. The structural finding with ESK1 implies that TCRm antibody can possibly overcome the limits of HLA-subtype restriction. TCRm therapy might have broader therapeutic applications than TCR-based drugs do.

Therapeutic applications of TCRm antibodies

The identification of TCRm antibodies provides new opportunities for cancer treatment with antibody-based immunotherapeutic approaches (Table. 2). Weidanz et al. isolated a specific mouse antibody to human chorionic gonadotropin (hCG)-β presented by HLA-A2 and demonstrated this antibody slowed down breast tumor growth in a mouse xenograft model (80). Another TCRm antibody was produced specifically for the PR1 peptide presented by HLA-A*0201 and inhibited AML progenitor cell growth (77). Human TCRm antibodies can be produced from human antigen-naïve PBMC phage libraries (25,30). The development of a human TCRm antibody against the pMHC complex from WT1 oncoprotein is a typical example of cancer immunotherapy. WT1 is an intracellular transcriptional factor and overexpressed in a wide range of cancers. The 9-mer peptide, WT1_126–134 (RMFPNAPYL), is processed and presented by HLA-A*0201 on the cancer cell surface. The TCRm antibody ESK1 specifically binds to tumor cells restricted to the WT1 and HLA-A2 and not to normal control cells. It showed potent antitumor effect in a mouse xenograft model (29).

For antibody-based approaches, the mechanism of action of TCRm antibodies is similar to conventional mAbs targeting specific tumor antigens through antibody‐dependent cellular cytotoxicity (ADCC), antibody-dependent cellular phagocytosis (ADCP), or complement-dependent cytotoxicity (CDC). To enhance the potency of a TCRm antibody, different methods can be utilized including enhancing ADCC, ADCP, engineering bi-specific mAbs, immunotoxins, antibody drug conjugates (ADCs), and combination with other treatment modalities. A modified version of ESK1 with an enhanced Fcγ receptor (FcγR) binding, ESKM, showed enhanced ADCC through increased affinity for activating FcγRs and decreased affinity for the inhibitory FcγRIIb receptor (84). ADCP was also achieved in ESKM through increased FcγR binding on macrophages. In addition, combination therapy of ESKM and a variety of tyrosine kinase inhibitors could improve therapeutic efficacy for the treatment of chronic myeloid leukemia (82).

TCRm antibody can also be developed for CAR T-cell therapy which is an emerging approach for cancer immunotherapy. CAR constructs are composed of the extracellular antibody parts specific for target antigen, hinge region, transmembrane region, and cytoplasmic signaling region including CD3 zeta chain and several costimulatory domains. TCRm can be engineered in the format of CAR T cells which target intracellular proteins directly by cellular immunotherapy. Several CARs derived from TCRm antibodies have been reported. Some groups reported CAR T cells based on TCRm antibodies including ESK1 against WT1 pMHC complex showed efficacy in vitro and in vivo (85,86). Zhang et al. generated CAR T cells against melanoma with a TCRm antibody targeting a gp100 pMHC complex(87). CAR T cells derived from this TCRm showed specific cytotoxicity against tumor cells and suppressed melanoma progression in a xenograft model (87). Liu et al isolated a human TCRm antibody against AFP_158–166 peptide complexed with HLA-A*0201 and developed it into CAR T-cell therapy against liver cancer (30). The AFP CAR T cells showed potent antitumor efficacy in liver cancer lines and xenograft models. Currently, the AFP CAR T-cell therapy is in clinical trial (NCT03998033) and the clinical effectiveness is being evaluated.

Concluding remarks

TCRm antibodies are mostly being evaluated in preclinical status. There are several important issues still facing the development of TCRm antibodies. The first major challenge and rate limiting step is the production of the pMHC complexes. It is both time consuming and costly to produce enough high-quality pMHC monomers. The second major concern is the off-target effects impairing the therapeutic application of TCRm antibodies. Most pMHC complexes are from proteins that might not be expressed exclusively in tumor cells and may still have low expression levels in normal tissues. In addition, cross reactivity of TCRm with other nonspecific peptides is a concern since the peptides loaded into the MHC binding grove are short. It is crucial that TCRm antibodies do not recognize MHC-I alone since it is expressed on most cells and must be specific for the pMHC complex. One possible approach is to develop TCRm from single domain antibody libraries such as camelid V_HHs. The rationale is that small heavy chain single domain antibodies might access further into the protein cavities, such as the MHC binding groove, compared to conventional antibodies, and recognize the conformation epitopes of the antigen (88–91). Another advantage of single domain antibodies is that it may be well fit for applications such as bi-paratopic or even multi-paratopic binding (92). Although camelid single domain antibodies are close to human by sequence, immunogenicity remains largely unknown. Humanization of camelid single domain antibodies may minimize the immunogenicity. The third challenge is the low epitope density of pMHC complexes on the cell surface and the downregulation of antigen presentation, the mechanism for therapy resistance (93). There are possible methods to address this challenge, including increasing pMHC-I expression in tumors, making TCRm antibodies more sensitive to low-density epitopes, combination therapy with other agents, and engineering bispecific format of TCRm antibodies. There are ways to upregulate MHC-I expression on tumors cells, such as treatment with cytokines (IFNγ and TNFα) or increase of Fhit gene expression (93,94). Alternatively, some tumors losing expression of MHC-I might retain expression of MHC-II, making tumor associated MHC class II complex a potential target (95,96). An additional limitation is the MHC-restricted manner of TCRm antibodies. Although TCRm antibody is less restricted by a specific HLA subtype than TCRs, it is generally challenging to isolate a TCRm antibody binding specifically to the peptide and the conserved region of the HLA molecules at the same time. If there is such a TCRm antibody, it might be able to cover a larger population of patients with various HLA types. The binding structure of TCRm antibodies to MHC molecules is still limited. More insights of the pMHC complex structures will facilitate the development of “universal” TCRm antibody specific for a tumor associated peptide target but not restricted to a HLA type. TCRm antibodies can expand the antigen repertoire to intracellular proteins and potentially make many “undruggable” targets accessible to immunotherapy. Although many efforts need to be made to isolate specific and successful TCRm antibodies, they have promising therapeutic potential for targeting intracellular proteins. Isolation of more TCRm antibodies to existing and novel antigens, improvement of pMHC antigen production, TCRm antibody discovery including the use of single domain antibodies, antibody delivery, structure investigations, and affinity maturation and deimmunization of TCRm will advance the therapy against the intracellular targets and hopefully bring effective antibodies into the clinic pipelines in the future.