Pre-encoded responsiveness to type I interferon in the peripheral immune system defines outcome of PD1 blockade therapy.

Type I interferons (IFN-Is) are central regulators of anti-tumor immunity and responses to immunotherapy, but they also drive the feedback inhibition underlying therapeutic resistance. In the present study, we developed a mass cytometry approach to quantify IFN-I-stimulated protein expression across immune cells and used multi-omics to uncover pre-therapy cellular states encoding responsiveness to inflammation. Analyzing peripheral blood cells from multiple cancer types revealed that differential responsiveness to IFN-Is before anti-programmed cell death protein 1 (PD1) treatment was highly predictive of long-term survival after therapy. Unexpectedly, IFN-I hyporesponsiveness efficiently predicted long-term survival, whereas high responsiveness to IFN-I was strongly associated with treatment failure and diminished survival time. Peripheral IFN-I responsive states were not associated with tumor inflammation, identifying a disconnect between systemic immune potential and 'cold' or 'hot' tumor states. Mechanistically, IFN-I responsiveness was epigenetically imprinted before therapy, poising cells for differential inflammatory responses and dysfunctional T cell effector programs. Thus, we identify physiological cell states with clinical importance that can predict success and long-term survival of PD1-blocking immunotherapy.

Characterization of Proteasome-Generated Spliced Peptides Detected by Mass Spectrometry.

CD8+ T cells recognize peptides displayed by HLA class I molecules and monitor intracellular peptide pools. It is known that the proteasome splices two short peptide fragments. Recent studies using mass spectrometry (MS) and bioinformatics analysis have suggested that proteasome-generated spliced peptides (PSPs) may account for a substantial proportion of HLA class I ligands. However, the authenticity of the PSPs identified using bioinformatics approaches remain ambiguous. In this study, we employed MS-based de novo sequencing to directly capture cryptic HLA ligands that were not templated in the genome. We identified two PSPs originating from the same protein in a human colorectal cancer line with microsatellite instability. Healthy donor-derived CD8+ T cells readily responded to the two PSPs, showing their natural HLA presentation and antigenicity. Experiments using minigene constructs demonstrated proteasome-dependent processing of two PSPs generated by standard and reverse cis splicing, respectively. Our results suggest a broader diversity of HLA class I Ag repertoires generated by proteasomal splicing, supporting the advantage of MS-based approaches for the comprehensive identification of PSPs.

Modification of the HLA-A*24:02 Peptide Binding Pocket Enhances Cognate Peptide-Binding Capacity and Antigen-Specific T Cell Activation.

The immunogenicity of a T cell Ag is correlated with the ability of its antigenic epitope to bind HLA and be stably presented to T cells. This presents a challenge for the development of effective cancer immunotherapies, as many self-derived tumor-associated epitopes elicit weak T cell responses, in part due to weak binding affinity to HLA. Traditional methods to increase peptide-HLA binding affinity involve modifying the peptide to reflect HLA allele binding preferences. Using a different approach, we sought to analyze whether the immunogenicity of wild-type peptides could be altered through modification of the HLA binding pocket. After analyzing HLA class I peptide binding pocket alignments, we identified an alanine 81 to leucine (A81L) modification within the F binding pocket of HLA-A*24:02 that was found to heighten the ability of artificial APCs to retain and present HLA-A*24:02-restricted peptides, resulting in increased T cell responses while retaining Ag specificity. This modification led to increased peptide exchange efficiencies for enhanced detection of low-avidity T cells and, when expressed on artificial APCs, resulted in greater expansion of Ag-specific T cells from melanoma-derived tumor-infiltrating lymphocytes. Our study provides an example of how modifications to the HLA binding pocket can enhance wild-type cognate peptide presentation to heighten T cell activation.

Identification of antigenic epitopes recognized by tumor infiltrating lymphocytes in high grade serous ovarian cancer by multi-omics profiling of the auto-antigen repertoire.

Immunotherapeutic strategies aimed at enhancing tumor cell killing by tumor-specific T cells hold great potential for reducing tumor burden and prolonging survival of cancer patients. Although many potential tumor antigens have been described, identifying relevant targets when designing anti-cancer vaccines or targeted cell therapies remains a challenge. To identify novel, potentially immunogenic candidate tumor antigens, we performed integrated tumor transcriptomic, seromic, and proteomic analyses of high grade serous ovarian cancer (HGSC) patient tumor samples. We identified tumor neo-antigens and over-expressed antigens using whole exome and RNA sequencing and examined these in relation to patient-matched auto-antibody repertoires. Focusing on MHC class I epitopes recognized by CD8+ T cells, HLA-binding epitopes were identified or predicted from the highly expressed, mutated, or auto-antibody target antigen, or MHC-associated peptides (MAPs). Recognition of candidate antigenic peptides was assessed within the tumor-infiltrating T lymphocyte (TIL) population expanded from each patient. Known tumor-associated antigens (TAA) and cancer/testis antigens (CTA) were commonly found in the auto-antibody and MAP repertoires and CD8+ TILs recognizing epitopes from these antigens were detected, although neither expression level nor the presence of auto-antibodies correlated with TIL recognition. Auto-antibodies against tumor-mutated antigens were found in most patients, however, no TIL recognition of the highest predicted affinity neo-epitopes was detected. Using high expression level, auto-antibody recognition, and epitope prediction algorithms, we identified epitopes in 5 novel antigens (MOB1A, SOCS3, TUBB, PRKAR1A, CCDC6) recognized by HGSC patient TILs. Furthermore, selection of epitopes from the MAP repertoire identified 5 additional targets commonly recognized by multiple patient TILs. We find that the repertoire of TIL specificities includes recognition of highly expressed and immunogenic self-antigens that are processed and presented by tumors. These results indicate an ongoing autoimmune response against a range of self-antigens targeted by HGSC TILs.

[Development of universal CAR-T cells].

The efficacy of adoptive immunotherapy using CD19-targeting chimeric antigen receptor (CAR)-engineered T cells against B-cell malignancies has already been established in the clinic. However, high economic costs and heterogeneous quality of CAR-T cells derived from individual patients hinder further expansion of their applicability to various cancer types, including solid tumors. Mass CAR-T cell production from healthy donors is a promising approach to overcome these problems, given that allogeneic immunity elicited against donor CAR-T cells by the recipient's immune system is controlled. CAR-T cells genetically ablated with T-cell receptor and human leukocyte antigen molecules, referred to as universal CAR-T cells, may enable the use of allogeneic T cells for off-the-shelf adoptive cancer immunotherapy. However, several concerns, such as poor persistence of infused CAR-T cells and chromosomal abnormalities due to genome editing, remain to be addressed. Thus, recent clinical trials on universal CAR-T cells are summarized and future perspectives to overcome current challenges are discussed in this review.

Mechanisms of HLA-DP Antigen Processing and Presentation Revisited.

Polymorphisms in HLA-DP can modulate interactions with the invariant chain chaperone, contributing independently to differences in the peptide repertoire presented on DP. The resulting presentation of intracellular antigens directly to CD4+ T cells may partly explain genetic and clinical studies describing previously unexplained links between polymorphism in DP and disease.

A Subset of Human Autoreactive CD1c-Restricted T Cells Preferentially Expresses TRBV4-1+ TCRs.

In humans, a substantial portion of T cells recognize lipids presented by the monomorphic CD1 proteins. Recent studies have revealed the molecular basis of mycobacterial lipid recognition by CD1c-restricted T cells. Subsets of CD1c-restricted T cells recognize self-lipids in addition to foreign lipids, which may have implications in human diseases involving autoimmunity and malignancy. However, the molecular identity of these self-reactive T cells remains largely elusive. In this study, using a novel CD1c+ artificial APC (aAPC)-based system, we isolated human CD1c-restricted autoreactive T cells and characterized them at the molecular level. By using the human cell line K562, which is deficient in MHC class I/II and CD1 expression, we generated an aAPC expressing CD1c as the sole Ag-presenting molecule. When stimulated with this CD1c+ aAPC presenting endogenous lipids, a subpopulation of primary CD4+ T cells from multiple donors was consistently activated, as measured by CD154 upregulation and cytokine production in a CD1c-specific manner. These activated CD4+ T cells preferentially expressed TRBV4-1+ TCRs. Clonotypic analyses of the reconstituted TRBV4-1+ TCR genes confirmed CD1c-restricted autoreactivity of this repertoire, and the strength of CD1c reactivity was influenced by the diversity of CDR3ß sequences. Finally, alanine scanning of CDR1 and CDR2 sequences of TRBV4-1 revealed two unique residues, Arg30 and Tyr51, as critical in conferring CD1c-restricted autoreactivity, thus elucidating the molecular basis of the observed V gene bias. These data provide new insights into the molecular identity of human autoreactive CD1c-restricted T cells.

Phase II clinical trial of adoptive cell therapy for patients with metastatic melanoma with autologous tumor-infiltrating lymphocytes and low-dose interleukin-2.

Adoptive cell therapy using autologous tumor-infiltrating lymphocytes (TIL) has shown significant clinical benefit, but is limited by toxicities due to a requirement for post-infusion interleukin-2 (IL-2), for which high dose is standard. To assess a modified TIL protocol using lower dose IL-2, we performed a single institution phase II protocol in unresectable, metastatic melanoma. The primary endpoint was response rate. Secondary endpoints were safety and assessment of immune correlates following TIL infusion. Twelve metastatic melanoma patients were treated with non-myeloablative lymphodepleting chemotherapy, TIL, and low-dose subcutaneous IL-2 (125,000 IU/kg/day, maximum 9-10 doses over 2 weeks). All but one patient had previously progressed after treatment with immune checkpoint inhibitors. No unexpected adverse events were observed, and patients received an average of 6.8 doses of IL-2. By RECIST v1.1, two patients experienced a partial response, one patient had an unconfirmed partial response, and six had stable disease. Biomarker assessment confirmed an increase in IL-15 levels following lymphodepleting chemotherapy as expected and a lack of peripheral regulatory T-cell expansion following protocol treatment. Interrogation of the TIL infusion product and monitoring of the peripheral blood following infusion suggested engraftment of TIL. In one responding patient, a population of T cells expressing a T-cell receptor Vß chain that was dominant in the infusion product was present at a high percentage in peripheral blood more than 2 years after TIL infusion. This study shows that this protocol of low-dose IL-2 following adoptive cell transfer of TIL is feasible and clinically active. (ClinicalTrials.gov identifier NCT01883323.).

In vitro-generated MART-1-specific CD8 T cells display a broader T-cell receptor repertoire than ex vivo naïve and tumor-infiltrating lymphocytes.

The differentiation of human hematopoietic stem cells into CD8 T cells can be achieved in vitro with the OP9-DL4 system. We considered that in the absence of medullary thymic epithelial cells, which serve to restrict the breath of the T-cell receptor (TCR) repertoire by expressing tissue-restricted antigens, a distinct repertoire would be generated in vitro. To test this notion, we compared the TCR-Vα/Vß (TRAV/TRBV) gene usage of major histocompatibility complex-restricted antigen (MART-1)-specific T cells generated in vitro to that from ex vivo naïve T cells and tumor-infiltrating lymphocytes (TILs) using high-throughput DNA sequencing. In contrast to naïve T cells and TILs, which showed the expected narrow TRAV repertoire, in vitro-generated MART-1-specific T cells used almost all TRAV gene families and displayed unique CDR3 lengths. Our work demonstrates that the OP9-DL4 system supports the creation of a broad antigen-specific TCR repertoire, suggesting that T cells generated in vitro may undergo a different set of selection events that otherwise constrains the TCR repertoire of thymus-derived T cells.

Chaperones of the class I peptide-loading complex facilitate the constitutive presentation of endogenous antigens on HLA-DP84GGPM87.

Recent work has delineated key differences in the antigen processing and presentation mechanisms underlying HLA-DP alleles encoding glycine at position 84 of the DPß chain (DP84GGPM87). These DPs are unable to associate with the class II-associated Ii peptide (CLIP) region of the invariant chain (Ii) chaperone early in the endocytic pathway, leading to continuous presentation of endogenous antigens. However, little is known about the chaperone support involved in the loading of these endogenous antigens onto DP molecules. Here, we demonstrate the proteasome and TAP dependency of this pathway and reveal the ability of HLA class I to compete with DP84GGPM87 for the presentation of endogenous antigens, suggesting that shared subcellular machinery may exist between the two classes of HLA. We identify physical interactions of prototypical class I-associated chaperones with numerous DP alleles, including TAP2, tapasin, ERp57, calnexin, and calreticulin, using a conventional immunoprecipitation and immunoblot approach and confirm the existence of these interactions in vivo through the use of the BioID2 proximal biotinylation system in human cells. Based on immunological assays, we then demonstrate the ability of each of these chaperones to facilitate the presentation of endogenously derived, but not exogenously derived, antigens on DP molecules. Considering previous genetic and clinical studies linking DP84GGPM87 to disease frequency and severity in autoimmune disease, viral infections, and cancer, we suggest that the above chaperones may form the molecular basis of these observable clinical differences through facilitating the presentation of endogenously derived antigens to CD4+ T cells.

Arginine methylation of FOXP3 is crucial for the suppressive function of regulatory T cells.

Forkhead box transcription factor 3 (FOXP3) plays a pivotal role in the suppressive function of regulatory T cells. In addition to mRNA levels, FOXP3 activity can also be controlled by posttranslational mechanisms, which have not been studied in a comprehensive manner. Through extensive screening using selective inhibitors, we demonstrate that the inhibition of type I protein arginine methytransferases (PRMTs) attenuates the suppressive functions of regulatory T cells. FOXP3 undergoes methylation on arginine residues at positions 48 and 51 by interacting with protein arginine methyltransferase 1 (PRMT1). The inhibition of arginine methylation confers gene expression profiles representing type I helper T cells to FOXP3+ T cells, which results in attenuated suppressive activity. A methylation-defective mutant of FOXP3 displays less potent activity to suppress xenogeneic graft-versus-host disease in vivo. These results elucidate an important role of arginine methylation to enhance FOXP3 functions and are potentially applicable to modulate regulatory T cell functions.

Key Residues at Third CDR3ß Position Impact Structure and Antigen Recognition of Human Invariant NK TCRs.

The human invariant NK (iNK) TCR is largely composed of the invariant TCR Vα24-Jα18 chain and semivariant TCR Vß11 chains with variable CDR3ß sequences. The direct role of CDR3ß in Ag recognition has been studied extensively. Although it was noted that CDR3ß can interact with CDR3α, how this interaction might indirectly influence Ag recognition is not fully elucidated. We observed that the third position of Vß11 CDR3 can encode an Arg or Ser residue as a result of somatic rearrangement. Clonotypic analysis of the two iNK TCR types with a single amino acid substitution revealed that the staining intensity by anti-Vα24 Abs depends on whether Ser or Arg is encoded. When stained with an anti-Vα24-Jα18 Ab, human primary invariant NKT cells could be divided into Vα24 low- and high-intensity subsets, and Arg-encoding TCR Vß11 chains were more frequently isolated from the Vα24 low-intensity subpopulation compared with the Vα24 high-intensity subpopulation. The Arg/Ser substitution also influenced Ag recognition as determined by CD1d multimer staining and CD1d-restricted functional responses. Importantly, in silico modeling validated that this Ser-to-Arg mutation could alter the structure of the CDR3ß loop, as well as the CDR3α loop. Collectively, these results indicate that the Arg/Ser encoded at the third CDR3ß residue can effectively modulate the overall structure of, and Ag recognition by, human iNK TCRs.

Human cell-based artificial antigen-presenting cells for cancer immunotherapy.

Adoptive T-cell therapy, where anti-tumor T cells are first prepared in vitro, is attractive since it facilitates the delivery of essential signals to selected subsets of anti-tumor T cells without unfavorable immunoregulatory issues that exist in tumor-bearing hosts. Recent clinical trials have demonstrated that anti-tumor adoptive T-cell therapy, i.e. infusion of tumor-specific T cells, can induce clinically relevant and sustained responses in patients with advanced cancer. The goal of adoptive cell therapy is to establish anti-tumor immunologic memory, which can result in life-long rejection of tumor cells in patients. To achieve this goal, during the process of in vitro expansion, T-cell grafts used in adoptive T-cell therapy must be appropriately educated and equipped with the capacity to accomplish multiple, essential tasks. Adoptively transferred T cells must be endowed, prior to infusion, with the ability to efficiently engraft, expand, persist, and traffic to tumor in vivo. As a strategy to consistently generate T-cell grafts with these capabilities, artificial antigen-presenting cells have been developed to deliver the proper signals necessary to T cells to enable optimal adoptive cell therapy.

Specific roles of each TCR hemichain in generating functional chain-centric TCR.

TCRα- and ß-chains cooperatively recognize peptide-MHC complexes. It has been shown that a "chain-centric" TCR hemichain can, by itself, dictate MHC-restricted Ag specificity without requiring major contributions from the paired TCR counterchain. Little is known, however, regarding the relative contributions and roles of chain-centric and its counter, non-chain-centric, hemichains in determining T cell avidity. We comprehensively analyzed a thymically unselected T cell repertoire generated by transducing the α-chain-centric HLA-A*02:01(A2)/MART127-35 TCRα, clone SIG35α, into A2-matched and unmatched postthymic T cells. Regardless of their HLA-A2 positivity, a substantial subset of peripheral T cells transduced with SIG35α gained reactivity for A2/MART127-35. Although the generated A2/MART127-35-specific T cells used various TRBV genes, TRBV27 predominated with >10(2) highly diverse and unique clonotypic CDR3ß sequences. T cells individually reconstituted with various A2/MART127-35 TRBV27 TCRß genes along with SIG35α possessed a wide range (>2 log orders) of avidity. Approximately half possessed avidity higher than T cells expressing clone DMF5, a naturally occurring A2/MART127-35 TCR with one of the highest affinities. Importantly, similar findings were recapitulated with other self-Ags. Our results indicate that, although a chain-centric TCR hemichain determines Ag specificity, the paired counterchain can regulate avidity over a broad range (>2 log orders) without compromising Ag specificity. TCR chain centricity can be exploited to generate a thymically unselected Ag-specific T cell repertoire, which can be used to isolate high-avidity antitumor T cells and their uniquely encoded TCRs rarely found in the periphery because of tolerance.

CDR3ß sequence motifs regulate autoreactivity of human invariant NKT cell receptors.

Invariant natural killer T (iNKT) cells are a subset of T lymphocytes that recognize lipid ligands presented by monomorphic CD1d. Human iNKT T cell receptor (TCR) is largely composed of invariant Vα24 (Vα24i) TCRα chain and semi-variant Vß11 TCRß chain, where complementarity-determining region (CDR)3ß is the sole variable region. One of the characteristic features of iNKT cells is that they retain autoreactivity even after the thymic selection. However, the molecular features of human iNKT TCR CDR3ß sequences that regulate autoreactivity remain unknown. Since the numbers of iNKT cells with detectable autoreactivity in peripheral blood is limited, we introduced the Vα24i gene into peripheral T cells and generated a de novo human iNKT TCR repertoire. By stimulating the transfected T cells with artificial antigen presenting cells (aAPCs) presenting self-ligands, we enriched strongly autoreactive iNKT TCRs and isolated a large panel of human iNKT TCRs with a broad range autoreactivity. From this panel of unique iNKT TCRs, we deciphered three CDR3ß sequence motifs frequently encoded by strongly-autoreactive iNKT TCRs: a VD region with 2 or more acidic amino acids, usage of the Jß2-5 allele, and a CDR3ß region of 13 amino acids in length. iNKT TCRs encoding 2 or 3 sequence motifs also exhibit higher autoreactivity than those encoding 0 or 1 motifs. These data facilitate our understanding of the molecular basis for human iNKT cell autoreactivity involved in immune responses associated with human disease.

IL-21 can supplement suboptimal Lck-independent MAPK activation in a STAT-3-dependent manner in human CD8(+) T cells.

Although both MHC class II/CD8α double-knockout and CD8ß null mice show a defect in the development of MHC class I-restricted CD8(+) T cells in the thymus, they possess low numbers of high-avidity peripheral CTL with limited clonality and are able to contain acute and chronic infections. These in vivo data suggest that the CD8 coreceptor is not absolutely necessary for the generation of Ag-specific CTL. Lack of CD8 association causes partial TCR signaling because of the absence of CD8/Lck recruitment to the proximity of the MHC/TCR complex, resulting in suboptimal MAPK activation. Therefore, there should exist a signaling mechanism that can supplement partial TCR activation caused by the lack of CD8 association. In this human study, we have shown that CD8-independent stimulation of Ag-specific CTL previously primed in the presence of CD8 coligation, either in vivo or in vitro, induced severely impaired in vitro proliferation. When naive CD8(+) T cells were primed in the absence of CD8 binding and subsequently restimulated in the presence of CD8 coligation, the proliferation of Ag-specific CTL was also severely hampered. However, when CD8-independent T cell priming and restimulation were supplemented with IL-21, Ag-specific CD8(+) CTL expanded in two of six individuals tested. We found that IL-21 rescued partial MAPK activation in a STAT3- but not STAT1-dependent manner. These results suggest that CD8 coligation is critical for the expansion of postthymic peripheral Ag-specific CTL in humans. However, STAT3-mediated IL-21 signaling can supplement partial TCR signaling caused by the lack of CD8 association.

Ex vivo activation of the GCN2 pathway metabolically reprograms T cells, leading to enhanced adoptive cell therapy.

The manipulation of T cell metabolism to enhance anti-tumor activity is an area of active investigation. Here, we report that activating the amino acid starvation response in effector CD8+ T cells ex vivo using the general control non-depressible 2 (GCN2) agonist halofuginone (halo) enhances oxidative metabolism and effector function. Mechanistically, we identified autophagy coupled with the CD98-mTOR axis as key downstream mediators of the phenotype induced by halo treatment. The adoptive transfer of halo-treated CD8+ T cells into tumor-bearing mice led to robust tumor control and curative responses. Halo-treated T cells synergized in vivo with a 4-1BB agonistic antibody to control tumor growth in a mouse model resistant to immunotherapy. Importantly, treatment of human CD8+ T cells with halo resulted in similar metabolic and functional reprogramming. These findings demonstrate that activating the amino acid starvation response with the GCN2 agonist halo can enhance T cell metabolism and anti-tumor activity.

Tumor reactive γδ T cells contribute to a complete response to PD-1 blockade in a Merkel cell carcinoma patient.

Immunotherapies targeting PD-1/PD-L1 are now widely used in the clinic to treat a variety of malignancies. While most of the research on T cell exhaustion and PD-1 blockade has been focused on conventional αß T cells, the contribution of innate-like T cells such as γδ T cells to anti-PD-1/PD-L1 mediated therapy is limited. Here we show that tumor reactive γδ T cells respond to PD-1 blockade in a Merkel cell carcinoma (MCC) patient experiencing a complete response to therapy. We find clonally expanded γδ T cells in the blood and tumor after pembrolizumab treatment, and this Vγ2Vδ1 clonotype recognizes Merkel cancer cells in a TCR-dependent manner. Notably, the intra-tumoral γδ T cells in the MCC patient are characterized by higher expression of PD-1 and TIGIT, relative to conventional CD4 and CD8 T cells. Our results demonstrate that innate-like T cells could also contribute to an anti-tumor response after PD-1 blockade.

Molecular, metabolic, and functional CD4 T cell paralysis in the lymph node impedes tumor control.

CD4 T cells are central effectors of anti-cancer immunity and immunotherapy, yet the regulation of CD4 tumor-specific T (TTS) cells is unclear. We demonstrate that CD4 TTS cells are quickly primed and begin to divide following tumor initiation. However, unlike CD8 TTS cells or exhaustion programming, CD4 TTS cell proliferation is rapidly frozen in place by a functional interplay of regulatory T cells and CTLA4. Together these mechanisms paralyze CD4 TTS cell differentiation, redirecting metabolic circuits, and reducing their accumulation in the tumor. The paralyzed state is actively maintained throughout cancer progression and CD4 TTS cells rapidly resume proliferation and functional differentiation when the suppressive constraints are alleviated. Overcoming their paralysis established long-term tumor control, demonstrating the importance of rapidly crippling CD4 TTS cells for tumor progression and their potential restoration as therapeutic targets.