Single-Cell and Spatial Transcriptomics Revealing HPV-Cervical Cancer Tumor Microenvironment Heterogeneity and Implications to Radiation and Immunotherapy

Abstract

<h3>Purpose/Objective(s)</h3> Cervical cancer (CC) is a common gynecological tumor and one of the leading causes of cancer death in women that is primarily cured by radiation-based therapy. However, 10%-40% of patients failed treatment. The high heterogeneity of tumors and their microenvironment (TME) is one of the main reasons for treatment failure which has been demonstrated recently through single-cell transcriptome (SCRNAseq) of the HPV+ CC. Since the tumor and TME of HPV- and HPV+ CC are significantly different, this study aimed to comprehensively understand the tumor and TME of HPV- CC at the single-cell level. <h3>Materials/Methods</h3> The primary method was single-cell and spatial transcriptome sequencing (STRNAseq) of tumor tissues. Newly diagnosed HPV- CC patients were eligible for testing of the tumor and tumor microenvironment (TME) heterogeneity. Fresh tumor tissue samples were obtained from HPV- CC patients by biopsy and divided into two parts: one was prepared as a single-cell suspension for SCRNAseq using the 10 × Genomics, and the other was frozen and embedded for STRNAseq. Seurat 4.0 was used to cluster and annotate cell clusters and map SCRNAseq data to the STRNAseq data. The CNV of epithelial cells (ECs) was estimated by InferCNV. CellChat and TCGA datasets were used to analyze cell interaction and survival, respectively. <h3>Results</h3> A total of 9800 cells were clustered 15 cell subsets in HPV- CC tissues, including 6 ECs subsets, 2 macrophage subsets, 2 fibroblasts subsets, 3 T cells subsets, plasma-B cell, and endothelial cell. Notably, ECs make up 90% of the entire tumor tissue. In addition, all ECs had high CNV compared with reference cells and high expression of TP63, the marker of malignancy, so all ECs were defined as tumor cells. The C1 and C4 tumor cells had 95% in G2M/S phase. According to pathological morphology and gene function enrichment of each cluster, tumor clusters were spatially mapped and divided into mature areas and naive areas with high proliferation and migration ability, and hypermetabolic tumor areas. A tight intercellular matrix was seen between tumor cells and fibroblasts through a variety of glycoprotein and collagen ligand-receptor pairs (LR). Moreover, we found that tumor cells recruited macrophages to migrate to the tumor periphery by secreting MIF, and macrophages interacted with tumor cells or fibroblasts through the SPP1/CD44 LR to form cell adhesion. The RNA expression of SPP1 in HPV- CC from the TCGA database showed that SPP1 high expression patients had poor survival (p = 0.08). <h3>Conclusion</h3> This study comprehensively analyzed the tumor and TME of HPV- CC at the single-cell transcriptional scale and the gene expression and functional characteristics of different tumor clusters, indicating the high heterogeneity and 'cold' tumor signature of HPV-CC TME. The high percentage of G2/M cells may indicate the sensitivity of this tumor to radiation. The ‘Cold' signature and the tight adhesion between cells in TME may explain the low infiltration of immune cells.