Exploring the Role of EBV Infection in EBV-T/NKLPD With EBV Capture Technology Based on Single-Cell RNA Sequencing

Epstein-Barr virus (EBV)-associated T/NK cell lymphoproliferative diseases (EBV-T/NKLPD) is a highly heterogeneous group of disease, whose clinical manifestations range from an indolent course to aggressive disease, making it challenging to make the concise diagnosis and to decide the best time for allogenic hematopoietic stem-cell transplantation. EBV infection is crucial to the occurrence and development of EBV-T/NKLPD [1]. However, the underestimated morbidity, high heterogeneity and lack of specific cell line make it difficult to decipher the concrete function of EBV in this disease [2].

Single-cell RNA sequencing (scRNA-seq) detects gene expression in single cell, providing opportunities to dissect cell heterogeneity and the cell-to-cell interaction in tumor microenvironment, which is an excellent tool to analyze the mechanism of virus infection. 10X Genomics and SMART-seq. 2 are two most frequently-used scRNA-seq platforms. 10X Genomics can detect rare cell populations due to high cell throughout. Meanwhile, SMART-seq. 2 can detect more genes, especially low abundance transcripts. The two technologies are complementary and their combination become the dominant mode to study the virus-to-host interaction currently but the high cost is obviously unbearable to most people [3].

On the basis of 10X Genomics platform (Figure 1A), we added viral probes on the magnetic beads and used secondary enrichment to amplify the titer of EBV (Figure 1B). To evaluate the sensitivity and specificity of this technology, EBV+ Raji cells (n = 727) and EBV− A549 cells (n = 458) were mixed together, and were detected the level of EBV infection (Figure 1C,D). Using traditional 10X Genomics, only 303 (41.7%) Raji cells were tested EBV positive while the updated EBV capture technology detected 723 (99.4%) EBV+ Raji cells (Figure 1E). Obviously, the updated EBV capture technology improved the sensitivity and specificity of virus capturing.

This updated technology was then used to explore the role of EBV in EBV-T/NKLPD. We collected nasopharyngeal biopsy from one patient diagnosed with extranodal NK/T-cell lymphoma, nasal, as well as cutaneous biopsy from one patient diagnosed with systemic chronic active EBV disease before treatment (Figure 1F and Supporting Information: Figure 1A,B). Although EBV infection existed in all types of cells, the EBV-positive cell proportion (Figure 1G) and CNV score (Figure 1H and Supporting Information: Figure 1C) were the highest in T cells, indicating EBV tropism for T cells. In total cells, EBV+ cells had a higher CNV score compared with EBV− cells, which attributed mainly to the difference between EBV + T cells and EBV− T cells (Figure 1I and Supporting Information: Figure 1D).

Then we analyzed the gene expression of T cells (n = 1348) and categorized them into three clusters, including naive T cells, NK/T cells and proliferative NK/T cells, which exemplified the functional heterogeneity of the T cell population (Figure 2A). EBV infection existed in all three types of T cells. The proportion of EBV-positive cell, EBV high-expression cell and mean EBV burden were all the highest in proliferative NK/T cells (Figure 2B–E). In T cells, EBV burden was positively associated with CNV score, especially in proliferative NK/T cells, indicating the genome instability induced by EBV infection (Figure 2G and Supporting Information: Figure 1E). Comparing EBV+ cells with EBV− cells, gene set enrichment analysis (GSEA) revealed different pathways enriched in different cells, indicating varying susceptibilities to EBV. Proliferation-associated signaling pathways, including E2F targets, G2M checkpoint and MYC target signaling pathway, were enriched in EBV + T cells (Figure 2F), especially in proliferative NK/T cells with high EBV burden (Supporting Information: Figure 2A,B). Pseudotime trajectory analysis showed EBV infection persistsed during the evolution of all types of T cells (Supporting Information: Figure 3A). The expression of proliferation-associated signaling pathways and genes increase as the EBV burden increasing (Supporting Information: Figure 3B,C). The stemness score of naive T cells and NK/T cells increased after EBV infection while the result was opposite in proliferative NK/T cells (Supporting Information: Figure 3D). In other cell types, metabolism-associated pathways, including oxidative phosphorylation, adipogenesis, citrate cycle, and cysteine and methionine metabolism, were obviously enriched in EBV+ cells (Supporting Information: Figure 3E). Oxidation of fatty acids- and glycolysis-associated genes had an obvious increasing trend in proliferative NK/T cells, further implying the influence of EBV on metabolism (Supporting Information: Figure 3F). Intercellular communication analysis showed EBV + T cells had the most and strongest interaction with other cells (Figure 2H). In proliferative NK/T cells, the interaction between EBV− cells and monocytes was significantly stronger than EBV+ cells (Figure 2I). In naive T cells, the interaction between EBV+ cells and endothelial cells or B cells was significantly stronger than EBV− cells (Figure 2I).

Our results showed the updated EBV capture technology based on single-cell RNA sequencing was a practical technology to study the role of EBV. In T cells, EBV infection promoted cell proliferation, genomic instability, stemness increase, and interactions with other cells. This may explain the carcinogenic mechanism of EBV in EBV-T/NKLPD. However, limited by the small sample size, our results require validation by multicenter, large-scale cohorts.

Shunan Wang and Miao Zhong designed the study, and collected and analyzed the data. Shunan Wang provided a draft of the manuscript. Wenqi Zhu, Yeqin Sha and Ruize Chen participated in data correction and analysis. Yongle Li and Hanning Tang collected the clinical samples. Shunan Wang, Huayuan Zhu and Lei Fan provided funding. Huayuan Zhu and Lei Fan supervised the study. Lei Fan reviewed and revised the manuscript. All authors read and approved the final manuscript.

The Ethics Committee of Jiangsu Province Hospital of Nanjing Medical University granted its approval for this study (No. 2024-SRFA-693).

Informed consent was obtained from each patient before any samples were collected for experiments involving human tissue.

The authors declare no conflicts of interest.