HLA-DR Presentation of the Tumor Antigen MSLN Associates with Clinical Outcome of Ovarian Cancer Patients

Comprehensive analysis of the interaction of antigen presentation during anti-tumour immunity and establishment of AIDPS systems in ovarian cancer

There are hundreds of prognostic models for ovarian cancer. These genes are based on different gene classes, and there are many ways to construct the models. Therefore, this paper aims to build the most stable prognostic evaluation system known to date through 101 machine learning strategies. We combined 101 algorithm combinations with 10 machine learning algorithms to create antigen presentation-associated genetic markers (AIDPS) with outstanding precision and steady performance. The inclusive set of algorithms comprises the elastic network (Enet), Ridge, stepwise Cox, Lasso, generalized enhanced regression model (GBM), random survival forest (RSF), supervised principal component (SuperPC), Cox partial least squares regression (plsRcox), survival support vector machine (Survival-SVM). Then, in the train cohort, the prediction model was fitted using a leave-one cross-validation (LOOCV) technique, which involved 101 different possible combinations of prognostic genes. Seven validation data sets (GSE26193, GSE26712, GSE30161, GSE63885, GSE9891, GSE140082 and ICGC_OV_AU) were compared and analysed, and the C-index was calculated. Finally, we collected 32 published ovarian cancer prognostic models (including mRNA and lncRNA). All data sets and prognostic models were subjected to a univariate Cox regression analysis, and the C-index was calculated to demonstrate that the antigen presentation process should be the core criterion for evaluating ovarian cancer prognosis. In a univariate Cox regression analysis, 22 prognostic genes were identified based on the expression profiles of 283 genes involved in antigen presentation and the intersection of genes (p < 0.05). AIDPS were developed by our machine learning-based integration method, which was applied to these 22 genes. One hundred and one prediction models are fitted using the LOOCV framework, and the C-index is calculated for each model across all validation sets. Interestingly, RSF + Lasso was the best model overall since it had the greatest average C-index and the highest C-index of any combination of models tested on the validated data sets. In comparing external cohorts, we found that the C-index correlated AIDPS method using the RSF + Lasso method in 101 prediction models was in contrast to other features. Notably, AIDPS outperformed the vast majority of models across all data sets. Antigen-presenting anti-tumour immune pathways can be used as a representative gene set of ovarian cancer to track the prognosis of patients with cancer. The antigen-presenting model obtained by the RSF + Lasso method has the best C-INDEX, which plays a key role in developing antigen-presenting targeted drugs in ovarian cancer and improving the treatment outcome of patients.

Claudin-4 modulates autophagy via SLC1A5/LAT1 as a tolerance mechanism for genomic instability in ovarian cancer

Genome instability is key for tumor heterogeneity and derives from defects in cell division and DNA damage repair. Tumors show tolerance for this characteristic, but its accumulation is regulated somehow to avoid catastrophic chromosomal alterations and cell death. Claudin-4 is upregulated and closely associated with genome instability and worse patient outcome in ovarian cancer. This protein is commonly described as a junctional protein participating in processes such as cell proliferation and DNA repair. However, its biological association with genomic instability is still poorly-understood. Here, we used CRISPRi and a claudin mimic peptide (CMP) to modulate the cladudin-4 expression and its function, respectively in in-vitro (high-grade serous carcinoma cells) and in-vivo (patient-derived xenograft in a humanized-mice model) systems. We found that claudin-4 promotes a protective cellular-mechanism that links cell-cell junctions to genome integrity. Disruption of this axis leads to irregular cellular connections and cell cycle that results in chromosomal alterations, a phenomenon associated with a novel functional link between claudin-4 and SLC1A5/LAT1 in regulating autophagy. Consequently, claudin-4's disruption increased autophagy and associated with engulfment of cytoplasm-localized DNA. Furthermore, the claudin-4/SLC1A5/LAT1 biological axis correlates with decrease ovarian cancer patient survival and targeting claudin-4 in-vivo with CMP resulted in increased niraparib (PARPi) efficacy, correlating with increased tumoral infiltration of T CD8+ lymphocytes. Our results show that the upregulation of claudin-4 enables a mechanism that promotes tolerance to genomic instability and immune evasion in ovarian cancer; thus, suggesting the potential of claudin-4 as a translational target for enhancing ovarian cancer treatment.

TOFIMS mass spectrometry-based immunopeptidomics refines tumor antigen identification

T cell recognition of human leukocyte antigen (HLA)-presented tumor-associated peptides is central for cancer immune surveillance. Mass spectrometry (MS)-based immunopeptidomics represents the only unbiased method for the direct identification and characterization of naturally presented tumor-associated peptides, a key prerequisite for the development of T cell-based immunotherapies. This study reports on the implementation of ion mobility separation-based time-of-flight (TOFIMS) MS for next-generation immunopeptidomics, enabling high-speed and sensitive detection of HLA-presented peptides. Applying TOFIMS-based immunopeptidomics, a novel extensive benignTOFIMS dataset was generated from 94 primary benign samples of solid tissue and hematological origin, which enabled the expansion of benign reference immunopeptidome databases with > 150,000 HLA-presented peptides, the refinement of previously described tumor antigens, as well as the identification of frequently presented self antigens and not yet described tumor antigens comprising low abundant mutation-derived neoepitopes that might serve as targets for future cancer immunotherapy development.

Immune Surveillance of Acute Myeloid Leukemia Is Mediated by HLA-Presented Antigens on Leukemia Progenitor Cells

Therapy-resistant leukemia stem and progenitor cells (LSC) are a main cause of acute myeloid leukemia (AML) relapse. LSC-targeting therapies may thus improve outcome of patients with AML. Here we demonstrate that LSCs present HLA-restricted antigens that induce T-cell responses allowing for immune surveillance of AML. Using a mass spectrometry-based immunopeptidomics approach, we characterized the antigenic landscape of patient LSCs and identified AML- and AML/LSC-associated HLA-presented antigens absent from normal tissues comprising nonmutated peptides, cryptic neoepitopes, and neoepitopes of common AML driver mutations of NPM1 and IDH2. Functional relevance of shared AML/LSC antigens is illustrated by presence of their cognizant memory T cells in patients. Antigen-specific T-cell recognition and HLA class II immunopeptidome diversity correlated with clinical outcome. Together, these antigens shared among AML and LSCs represent prime targets for T cell-based therapies with potential of eliminating residual LSCs in patients with AML.

SIGNIFICANCE

The elimination of therapy-resistant leukemia stem and progenitor cells (LSC) remains a major challenge in the treatment of AML. This study identifies and functionally validates LSC-associated HLA class I and HLA class II-presented antigens, paving the way to the development of LSC-directed T cell-based immunotherapeutic approaches for patients with AML. See related commentary by Ritz, p. 430 . This article is featured in Selected Articles from This Issue, p. 419.

Immunopeptidome Diversity in Chronic Lymphocytic Leukemia Identifies Patients with Favorable Disease Outcome

Simple Summary

Immunosurveillance of cancer is mediated by T cell-based recognition of tumor-associated antigens, i.e., short peptides that are presented on the surface of cells on human leukocyte antigen (HLA) molecules. This encourages the analysis of the entirety of HLA-presented peptides, the so-called immunopeptidome, of malignant and benign cells, in order to identify novel therapeutic targets presented exclusively on malignant cells. In the present study, we aim to investigate the role of previously described immunopeptidome-defined antigen presentation in chronic lymphocytic leukemia (CLL) patients for clinical characteristics and disease outcome. We observed that higher yields of presented total and CLL-exclusive peptides were associated with a more favorable disease course, suggesting efficient immunosurveillance in a subgroup of patients and the possibility of further investigating T cell-based therapeutic approaches for CLL.

Abstract

Chronic lymphocytic leukemia (CLL) is characterized by recurrent relapses and resistance to treatment, even with novel therapeutic approaches. Despite being considered as a disease with low mutational burden and thus poor immunogenic, CLL seems to retain the ability of eliciting specific T cell activation. Accordingly, we recently found non-mutated tumor-associated antigens to play a central role in CLL immunosurveillance. Here, we investigated the association of total and CLL-exclusive HLA class I and HLA class II peptide presentation in the mass spectrometry-defined immunopeptidome of leukemic cells with clinical features and disease outcome of 57 CLL patients. Patients whose CLL cells present a more diverse immunopeptidome experienced fewer relapses. During the follow-up phase of up to 10 years, patients with an HLA class I-restricted presentation of high numbers of total and CLL-exclusive peptides on their malignant cells showed a more favorable disease course with a prolonged progression-free survival (PFS). Overall, our results suggest the existence of an efficient T cell-based immunosurveillance mediated by CLL-associated tumor antigens, supporting ongoing efforts in developing T cell-based immunotherapeutic strategies for CLL.