Spatial Multiomics Profiling of Angioimmunoblastic T-Cell Lymphoma

Introduction

Angioimmunoblastic T-cell lymphoma (AITL) is a T-cell lymphoma characterized by T follicular helper cell (TFH) phenotype. It comes with a strong inflammatory response and highly complex tumor microenvironment, where tumor cells are supported by a scaffold of highly proliferative CD21+ follicular dendritic cells (FDCs) in proximity to high endothelial venules (HEVs). The structured organization of the lymph node, with its distinct B-cell and T-cell zones, is disrupted in lymph node samples from patients with AITL. Based on prior literature in this space, background “benign” inflammatory bystanders and abnormal non-neoplastic vasculature and stromal cells appear to enable tumor cells to escape immune surveillance and unchecked proliferation. We hypothesize that genomic spatial characterization of the tumor microenvironment can lead to improved targeting of these tumor enablers, and for the first time, AITL has been profiled by single cell spatial ATAC-seq, paired with multiplexed immunofluorescence and DBiT-seq for this purpose.

Methods

The sample cohort consists of 23 de-identified, archived formalin-fixed paraffin-embedded (FFPE) human lymphoma tissue blocks and 7 OCT-frozen blocks, 6 with paired FFPE. The quality of RNA in each tissue block was assessed via an RNA Integrity Number (RIN) test and blocks with DV200 > 50% (minimal RNA degradation levels) were the ones included in our cohort. For each sample, we performed highly multiplexed immunofluorescence imaging (CODEX) using a custom immune cell panel of 40 markers on the PhenoCycler-Fusion system (Akoya Biosciences). The panel includes cell-type specific markers such as Podoplanin (lymphatic vasculature), CD20, CD3e, as well as all TFH markers. Cell segmentation was performed using a StarDist-based model in QuPath and downstream analysis done using the Seurat package. Serial section was then used for spatial transcriptomics using DBiT-seq to characterize the whole transcriptional landscape of the lymphoma. For the OCT-frozen sections, spatial ATAC-seq was performed on one slide to explore the epigenetic landscape of AITL.

Results

Cell type annotation of the CODEX data allowed us to identify the cell types within the tumor microenvironment and quantify their relative proportions. We observed a background of EBV positive B cells marked by the expression of LMP1. Interestingly, we also discovered a subset of non-neoplastic T cells that were EBV-positive. We observed a strong correlation between the proteomic data from CODEX and the spatial transcriptomic data for key cell type populations. Ligand-receptor interaction analysis of the transcriptomic data also showed communication between CXCL13 and its receptor CXCR5, heightened in association with FDCs. We also identified a cluster of IGHM positive cells that highly express KMT2C, one of the histone modification proteins frequently mutated in T-cell lymphomas. Preliminary epigenetic data suggests high chromatin accessibility score in the T-cell rich regions nearby HEVs. BCL6 seems to show high gene activity score in this region, despite its relatively lower transcript and protein expression level. We are currently in the process of fully integrating the RNA profiling and protein expression as well as epigenetic datasets using novel MaxFuse pipeline to provide a comprehensive understanding of the biology underlying AITL.

Conclusion

Combining multiplexed immunofluorescence with DBIT-seq and spatial ATAC-seq is providing a comprehensive picture of the mechanisms underlying epigenetic, transcriptional and proteomic landscape of AITL. Data in process of curation reveal novel molecular architecture that may explain underlying cell-cell interactions between TFH tumor cells, reactive B cells and the supporting stromal and vascular apparatus. Specifically, that there may be immune-suppressed EBV+ T cells in the background in addition to spatially segregated activation of tumor cells by surrounding non-neoplastic cells. Broadly, our approach can further the understanding of AITL’s complex microenvironment, leading to improved targeting of tumor-enabling signals that facilitate tumor evasion of immune surveillance and drug resistance.