A robust deep learning workflow to predict CD8 + T-cell epitopes

BACKGROUND

T-cells play a crucial role in the adaptive immune system by triggering responses against cancer cells and pathogens, while maintaining tolerance against self-antigens, which has sparked interest in the development of various T-cell-focused immunotherapies. However, the identification of antigens recognised by T-cells is low-throughput and laborious. To overcome some of these limitations, computational methods for predicting CD8 + T-cell epitopes have emerged. Despite recent developments, most immunogenicity algorithms struggle to learn features of peptide immunogenicity from small datasets, suffer from HLA bias and are unable to reliably predict pathology-specific CD8 + T-cell epitopes.

METHODS

We developed TRAP (T-cell recognition potential of HLA-I presented peptides), a robust deep learning workflow for predicting CD8 + T-cell epitopes from MHC-I presented pathogenic and self-peptides. TRAP uses transfer learning, deep learning architecture and MHC binding information to make context-specific predictions of CD8 + T-cell epitopes. TRAP also detects low-confidence predictions for peptides that differ significantly from those in the training datasets to abstain from making incorrect predictions. To estimate the immunogenicity of pathogenic peptides with low-confidence predictions, we further developed a novel metric, RSAT (relative similarity to autoantigens and tumour-associated antigens), as a complementary to 'dissimilarity to self' from cancer studies.

RESULTS

TRAP was used to identify epitopes from glioblastoma patients as well as SARS-CoV-2 peptides, and it outperformed other algorithms in both cancer and pathogenic settings. TRAP was especially effective at extracting immunogenicity-associated properties from restricted data of emerging pathogens and translating them onto related species, as well as minimising the loss of likely epitopes in imbalanced datasets. We also demonstrated that the novel metric termed RSAT was able to estimate immunogenic of pathogenic peptides of various lengths and species. TRAP implementation is available at: https://github.com/ChloeHJ/TRAP .

CONCLUSIONS

This study presents a novel computational workflow for accurately predicting CD8 + T-cell epitopes to foster a better understanding of antigen-specific T-cell response and the development of effective clinical therapeutics.

Decitabine increases neoantigen and cancer testis antigen expression to enhance T-cell-mediated toxicity against glioblastoma

BACKGROUND

Glioblastoma (GBM) is the most common and malignant primary brain tumor in adults. Despite maximal treatment, median survival remains dismal at 14-24 months. Immunotherapies, such as checkpoint inhibition, have revolutionized management of some cancers but have little benefit for GBM patients. This is, in part, due to the low mutational and neoantigen burden in this immunogenically "cold" tumor.

METHODS

U87MG and patient-derived cell lines were treated with 5-aza-2'-deoxycytidine (DAC) and underwent whole-exome and transcriptome sequencing. Cell lines were then subjected to cellular assays with neoantigen and cancer testis antigen (CTA) specific T cells.

RESULTS

We demonstrate that DAC increases neoantigen and CTA mRNA expression through DNA hypomethylation. This results in increased neoantigen presentation by MHC class I in tumor cells, leading to increased neoantigen- and CTA-specific T-cell activation and killing of DAC-treated cancer cells. In addition, we show that patients have endogenous cancer-specific T cells in both tumor and blood, which show increased tumor-specific activation in the presence of DAC-treated cells.

CONCLUSIONS

Our work shows that DAC increases GBM immunogenicity and consequent susceptibility to T-cell responses in vitro. Our results support a potential use of DAC as a sensitizing agent for immunotherapy.

HLA-dependent variation in SARS-CoV-2 CD8 + T cell cross-reactivity with human coronaviruses

The conditions and extent of cross-protective immunity between severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and common-cold human coronaviruses (HCoVs) remain open despite several reports of pre-existing T cell immunity to SARS-CoV-2 in individuals without prior exposure. Using a pool of functionally evaluated SARS-CoV-2 peptides, we report a map of 126 immunogenic peptides with high similarity to 285 MHC-presented peptides from at least one HCoV. Employing this map of SARS-CoV-2-non-homologous and homologous immunogenic peptides, we observe several immunogenic peptides with high similarity to human proteins, some of which have been reported to have elevated expression in severe COVID-19 patients. After combining our map with SARS-CoV-2-specific TCR repertoire data from COVID-19 patients and healthy controls, we show that public repertoires for the majority of convalescent patients are dominated by TCRs cognate to non-homologous SARS-CoV-2 peptides. We find that for a subset of patients, >50% of their public SARS-CoV-2-specific repertoires consist of TCRs cognate to homologous SARS-CoV-2-HCoV peptides. Further analysis suggests that this skewed distribution of TCRs cognate to homologous or non-homologous peptides in COVID-19 patients is likely to be HLA-dependent. Finally, we provide 10 SARS-CoV-2 peptides with known cognate TCRs that are conserved across multiple coronaviruses and are predicted to be recognized by a high proportion of the global population. These findings may have important implications for COVID-19 heterogeneity, vaccine-induced immune responses, and robustness of immunity to SARS-CoV-2 and its variants.

A blood atlas of COVID-19 defines hallmarks of disease severity and specificity

Treatment of severe COVID-19 is currently limited by clinical heterogeneity and incomplete description of specific immune biomarkers. We present here a comprehensive multi-omic blood atlas for patients with varying COVID-19 severity in an integrated comparison with influenza and sepsis patients versus healthy volunteers. We identify immune signatures and correlates of host response. Hallmarks of disease severity involved cells, their inflammatory mediators and networks, including progenitor cells and specific myeloid and lymphocyte subsets, features of the immune repertoire, acute phase response, metabolism, and coagulation. Persisting immune activation involving AP-1/p38MAPK was a specific feature of COVID-19. The plasma proteome enabled sub-phenotyping into patient clusters, predictive of severity and outcome. Systems-based integrative analyses including tensor and matrix decomposition of all modalities revealed feature groupings linked with severity and specificity compared to influenza and sepsis. Our approach and blood atlas will support future drug development, clinical trial design, and personalized medicine approaches for COVID-19.

Potential CD8+ T Cell Cross-Reactivity Against SARS-CoV-2 Conferred by Other Coronavirus Strains

While individuals infected with coronavirus disease 2019 (COVID-19) manifested a broad range in susceptibility and severity to the disease, the pre-existing immune memory to related pathogens cross-reactive against SARS-CoV-2 can influence the disease outcome in COVID-19. Here, we investigated the potential extent of T cell cross-reactivity against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that can be conferred by other coronaviruses and influenza virus, and generated an in silico map of public and private CD8+ T cell epitopes between coronaviruses. We observed 794 predicted SARS-CoV-2 epitopes of which 52% were private and 48% were public. Ninety-nine percent of the public epitopes were shared with SARS-CoV and 5.4% were shared with either one of four common coronaviruses, 229E, HKU1, NL63, and OC43. Moreover, to assess the potential risk of self-reactivity and/or diminished T cell response for peptides identical or highly similar to the host, we identified predicted epitopes with high sequence similarity with human proteome. Lastly, we compared predicted epitopes from coronaviruses with epitopes from influenza virus deposited in IEDB, and found only a small number of peptides with limited potential for cross-reactivity between the two virus families. We believe our comprehensive in silico profile of private and public epitopes across coronaviruses would facilitate design of vaccines, and provide insights into the presence of pre-existing coronavirus-specific memory CD8+ T cells that may influence immune responses against SARS-CoV-2.

Frequency determination of breast tumor-reactive CD4 and CD8 T cells in humans: unveiling the antitumor immune response

As cancer immunotherapy gains importance, the determination of a patient's ability to react to his/her tumor is unquestionably relevant. Though the presence of T cells that recognize specific tumor antigens is well established, the total frequency of tumor-reactive T cells in humans is difficult to assess, especially due to the lack of broad analysis techniques. Here, we describe a strategy that allows this determination, in both CD4 and CD8 compartments, using T cell proliferation induced by tumor cell-lysate pulsed dendritic cells as the readout. All 12 healthy donor tested had circulating CD4 and CD8 tumor cell-reactive T cells. The detection of these T cells, not only in the naïve but also in the memory compartment, can be seen as an evidence of tumor immunosurveillance in humans. As expected, breast cancer patients had higher frequencies of blood tumor-reactive T cells, but with differences among breast cancer subtypes. Interestingly, the frequency of blood tumor-reactive T cells in patients did not correlate to the frequency of infiltrating tumor-reactive T cells, highlighting the danger of implying a local tumor response from blood obtained data. In conclusion, these data add T cell evidence to immunosurveillance in humans, confirm that immune parameters in blood may be misleading and describe a tool to follow the tumor-specific immune response in patients and, thus, to design better immunotherapeutic approaches.

Tolerogenic IDO(+) Dendritic Cells Are Induced by PD-1-Expressing Mast Cells

Mast cells (MCs) are tissue resident cells, rich in inflammatory mediators, involved in allergic reactions, and with an increasingly recognized role in immunomodulation. Dendritic cells (DCs), on the other hand, are central to the determination of immune response patterns, being highly efficient antigen-presenting cells that respond promptly to changes in their microenvironment. Here, we show that direct cell contact between immature monocyte-derived DCs (iDCs) and MC bends DCs toward tolerance induction. DCs that had direct contact with MC (MC-iDC) decreased HLA-DR but increased PD-L1 expression and stimulated regulatory T lymphocytes, which expresses FoxP3⁺, secrete TGF-β and IL-10, and suppress the proliferation of mitogen-stimulated naïve T lymphocytes. Furthermore, MC-iDC expressed higher levels of indoleamine-2,3-deoxigenase (IDO), a phenomenon that was blocked by treatment of MC with anti-PD-1 or by the treatment of DCs with anti-PD-L1 or anti-PD-L2, but not by blocking of H1 and H2 histamine receptors on DCs. Contact with MC also increased phosphorylated STAT-3 levels in iDCs. When a STAT-3 inhibitor, JSI-124, was added to the DCs before contact with MC, the MC-iDC recovered their ability to induce allogeneic T cell proliferation and did not increase their IDO expression.

Dendritic Cells Stimulated by Cationic Liposomes

Immunotherapy of cancer aims to harness the immune system to detect and destroy cancer cells. To induce an immune response against cancer, activated dendritic cells (DCs) must present tumor antigens to T lymphocytes of patients. However, cancer patients' DCs are frequently defective, therefore, they are prone to induce rather tolerance than immune responses. In this context, loading tumor antigens into DCs and, at the same time, activating these cells, is a tempting goal within the field. Thus, we investigated the effects of cationic liposomes on the DCs differentiation/maturation, evaluating their surface phenotype and ability to stimulate T lymphocytes proliferation in vitro. The cationic liposomes composed by egg phosphatidylcholine, 1,2-dioleoyl-3-trimethylammonium propane and 1,2-dioleoylphosphatidylethanolamine (50/25/25% molar) were prepared by the thin film method followed by extrusion (65 nm, polydispersity of 0.13) and by the dehydration-rehydration method (95% of the population 107 nm, polydispersity of 0.52). The phenotypic analysis of dendritic cells and the analysis of T lymphocyte proliferation were performed by flow cytometry and showed that both cationic liposomes were incorporated and activated dendritic cells. Extruded liposomes were better incorporated and induced higher CD86 expression for dendritic cells than dehydrated-rehydrated vesicles. Furthermore, dendritic cells which internalized extruded liposomes also provided stronger T lymphocyte stimulation. Thus, cationic liposomes with a smaller size and polydispersity seem to be better incorporated by dendritic cells. Hence, these cationic liposomes could be used as a potential tool in further cancer immunotherapy strategies and contribute to new strategies in immunotherapy.

Abstract B069: Systemic alterations in T cell subpopulations of breast cancer patients

Breast cancer patients have immune system functional deviations, that affect immune cells that infiltrate the tumor, but also cells from the blood compartment, even in the absence of a metastatic disease. Blood monocytes of these patients, for example, have impaired ability to differentiate into dendritic cells. Little is known, however, about systemic alterations in T cells from these patients. To address this issue, we evaluated T cell populations in peripheral blood of cancer patients, comparing it to those of healthy donors. For this, peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Paque centrifugation, stained with antibodies to different T cell markers, and evaluated by flow cytometry. Patients had lower CD4:CD8 ratio then healthy donors (Healthy: 2.91±0.31; Cancer: 2.06±0.32; n=10; p=0.04). There was a positive correlation between the histological grade of the tumor and the CD4:CD8 ratio, where those with the lower grade had the lower CD4:CD8 ratio (r2=0.5679; p=0.03). Regarding CD4+ T cells, cancer patients had the same percentage of effector (CD45RA+CCR7-) and central memory (CD45RA-CCR7+) CD4+ T cells, but significantly lower naïve (CD45RA+CCR7+) CD4+ T cells (Healthy: 30.82±5.94, n=8; Cancer: 14.21±3.46, n=10; p=0.02) and a tendency to have higher effector memory (CD45RA-CCR7-) CD4+ T cells (Healthy: 35.13±5.54, n=8; Cancer: 48.64±4.52, n=10; p=0.08). When comparing CD8+ T cells, patients also had significantly lower naïve CD8+ T cells (Healthy: 42.74±6.97, n=8; Cancer: 19.16±3.42, n=10; p=0.005) and higher effector memory CD8+ T cells (Healthy: 29.69±5.35, n=8; Cancer: 45.37±5.15, n=10; p=0.05). Interestingly, cancer patients seemed to have higher percentage of effector CD8+ T cells than healthy patients (Healthy: 19.54±5.30, n=8; Cancer: 30.30±5.49, n=10; p=0.18). No difference in the expression of the IL-2 high affinity receptor (CD25) was observed between individuals with or without breast cancer. In conclusion, breast cancer patients had significant alterations in the percentage of the different peripheral blood T cell populations, suggesting that these may affect their ability to respond to the tumor, but also to other antigens.

Financial Support: FAPESP (2014/25988-1).

Citation Format: Mariana Pereira Pinho, José Alexandre Marzagão Barbuto. Systemic alterations in T cell subpopulations of breast cancer patients. [abstract]. In: Proceedings of the CRI-CIMT-EATI-AACR Inaugural International Cancer Immunotherapy Conference: Translating Science into Survival; September 16-19, 2015; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2016;4(1 Suppl):Abstract nr B069.

Dendritic cell membrane CD83 enhances immune responses by boosting intracellular calcium release in T lymphocytes

CD83 is a marker of mDCs directly related to their lymphostimulatory ability. Some data suggest that it has a central role in the immune system regulation, but how this function is performed remains to be determined. This work aimed to analyze the influence of CD83, present in mDCs, in the modulation of calcium signaling in T lymphocytes. Mo were differentiated into iDCs and activated with TNF-α. iDCs were treated, 4 h before activation, with siRNACD83, to reduce CD83 expression. Purified allogeneic T lymphocytes were labeled with the calcium indicator Fluo-4-AM, and calcium mobilization in the presence of mDCs was analyzed. CD83 knockdown mDCs induced lower calcium signal amplitude in T lymphocytes (29.0±10.0) compared with siRNAscr-treated mDCs (45.5±5.3). In another set of experiments, surface mDC CD83 was blocked with a specific mAb, and again, decreased calcium signaling in T lymphocytes was detected by flow cytometry and microscopy (fluorescence and confocal). In the presence of antibody, the percentage of responding T cells was reduced from 58.14% to 34.29%. As expected, anti-CD83 antibodies also reduced the proliferation of T lymphocytes (as assessed by CFSE dilution). Finally, in the absence of extracellular calcium, CD83 antibodies abrogated T cell signaling induced by allogeneic mDCs, suggesting that the presence of CD83 in mDC membranes enhances T lymphocyte proliferation by boosting calcium release from intracellular stores in these cells.