Single-Cell and Spatial Analyses Characterize Distinct Subsets of Malignant T Cells in Angioimmunoblastic T Cell Lymphoma - Aix-Marseille Université Accéder directement au contenu
Article Dans Une Revue Blood Année : 2021

Single-Cell and Spatial Analyses Characterize Distinct Subsets of Malignant T Cells in Angioimmunoblastic T Cell Lymphoma

Résumé

Abstract Introduction Angioimmunoblastic T-cell lymphoma (AITL) is the most frequent of nodal peripheral T-cell lymphomas. AITL results from the transformation of T follicular helper (T FH) cells and is characterized by chemo-resistance and poor survival (5-year OS around 30%). Recent data from prospective clinical trials suggest that disease outcome may be impacted by factors other than genomic features, such as the tumor microenvironment (TME) and overall intra-tumoral heterogeneity. Our understanding of AITL intra-tumoral genetic, transcriptional and functional heterogeneity is limited because most molecular data generated so far have come from bulk analyses. Single-cell RNA sequencing (scRNA-seq) enables fine characterization of cell types and functional cell states. When focused on T or B cells, 5'-end scRNA-seq also yields the TCR or BCR sequences that allow tracking clonally related cells. Here we studied the intra-tumor heterogeneity of AITL tumors using integrative scRNA-seq. Methods We analyzed lymph node live cell suspensions from AITL patients (n=10) using droplet-based 10x Genomics 5'-end scRNA-seq. Malignant T cells from 4 AITL samples were also analyzed by FACS index sorting and plate-based 5'-end scRNA-seq to link cell surface phenotype and gene expression profile. We identified malignant T cell clones by intersecting the gene expression and TCR sequencing data, and performed separate focused analyses of TME subsets and malignant T cells. We compared subsets of malignant T cells from all patients using marker gene-based metaclustering to identify AITL T cell states conserved across patients. We explored the genetic heterogeneity of malignant T cells by mapping RHOA G17V mutations and inferring copy number variation (CNV) subclones from scRNA-seq data. In select cases, we performed in situ analysis by immunohistochemistry (IHC) or spatial transcriptomics to characterize the spatial distribution of malignant T cell subsets identified by scRNA-seq. Results Based on gene expression, malignant T cells grouped in patient-specific clusters, while non-malignant T, B and myeloid TME cells from all patients clustered by cell type or cell state. Among TME cells, we identified 7 subsets of B cells (including activated B cells, plasma cells, and one patient-specific monoclonal B cell proliferation), 6 subsets of myeloid cells (including macrophages, conventional and plasmacytoid dendritic cells), and 8 subsets of non-malignant T cells (including activated cytotoxic T lymphocytes (CTL) with clonal expansions). Patient-specific malignant T cells were heterogeneous and divided into several gene-expression based clusters. Metaclustering of malignant T cell subsets identified T central memory (T CM)-like and T FH-like states in 10/10 samples. We also identified in 3/10 samples clusters of CTL-like malignant T cells expressing characteristic marker genes (including NKG7, GNLY, GZMK, PRF1). We observed an intra-sample continuum of gene expression states from quiescent T CM-like to proliferating T FH-like states. T FH-like cells were larger in size and expressed higher levels of surface PD1 and ICOS than T CM-like and CTL-like subsets. We detected the RHOA G17V mutation in malignant T cells of 4/4 mutated cases, with no evidence of subclonal heterogeneity for that mutation. We detected clonal and subclonal CNV in most AITL malignant T cells. CTL-like states were enriched in specific CNV subclones, but the T CM-like to T FH-like continuum was observed in all CNV subclones, suggesting that functional plasticity and subclonal genetic evolution may occur independently. In situ staining of markers for T FH-like (PD1, ICOS, CD200) and CTL-like (GZMK, GZMA) cells showed that T FH-like and CTL-like cells occupied distinct tissue niches within the tumor. In spatial transcriptomics analysis, T FH-like cells mapped to follicular dendritic cell (FDC)-rich areas, while T CM-like cells were associated with T-zone reticular cells. Conclusions Our analyses recapitulate known characteristics of AITL TME, and uncover previously unrecognized heterogeneity among malignant T cells across multiple patients. The distinct gene expression programs, phenotypes, genetics, and locations of T FH-like, T CM-like and CTL-like states suggest that AITL malignant T cells undergo significant functional plasticity and genetic divergence, which could influence response to therapy and overall clinical course. Figure 1 Figure 1. Disclosures Lemonnier: Institut Roche: Research Funding; Gilead: Other: travel grant. Gaulard: Gilead: Consultancy; Innate Pharma: Research Funding; Sanofi: Research Funding; Alderaan: Research Funding; Takeda: Consultancy, Honoraria. Milpied: Institut Roche: Research Funding; Innate Pharma: Research Funding; Bristol Myers Squibb: Research Funding.

Dates et versions

hal-03659244 , version 1 (04-05-2022)

Identifiants

Citer

Francois Lemonnier, Chuang Dong, Bruno Tesson, Laurine Gil, Noudjoud Attaf, et al.. Single-Cell and Spatial Analyses Characterize Distinct Subsets of Malignant T Cells in Angioimmunoblastic T Cell Lymphoma. Blood, 2021, 138 (Supplement 1), pp.2393-2393. ⟨10.1182/blood-2021-147246⟩. ⟨hal-03659244⟩
25 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More