# HPV-driven transcriptome and splicing rewiring under SRPK1 inhibition in cervical cancer

## TL;DR

This study explores how inhibiting SRPK1 affects gene expression and splicing in HPV-positive and HPV-negative cervical cancer cells, revealing distinct responses linked to cancer signaling and metabolism.

## Contribution

The study reveals how SRPK1 inhibition differentially rewires the transcriptome and splicing in HPV+ versus HPV- cervical cancer cells.

## Key findings

- SRPK1 inhibition in HPV+ SiHa cells suppressed oncogenic signaling pathways like Hippo, Wnt, and PI3K-AKT.
- HPV- C33A cells showed increased ribosomal and metabolic gene expression after SRPK1 inhibition.
- Splicing changes in SiHa cells had larger effect sizes, while C33A cells exhibited numerous exon skipping and intron retention events.

## Abstract

Serine/arginine protein kinase 1 phosphorylates serine-arginine-rich (SR) proteins to regulate splice-site selection during alternative splicing. While its role in general RNA regulation is established, its contribution to the HPV-dependent transcriptome and splicing stratification in cervical cancer remains unclear. Therefore, we sought to determine how SRPK1 inhibition differentially remodels gene expression and alternative splicing in HPV+ versus HPV- cervical cancer cells.

HPV16+ SiHa and HPV- C33A cervical cancer cells were treated with the SRPK1 inhibitor, SPHINX31. RNA profiling was performed, and differentially expressed genes were defined as |log2FC| ≥ 1.5. AS events were classified by SUPPA as exon skipping (SE), intron retention (RI), mutually exclusive exons (MXE), alternative 3′ splice site (A3SS), and alternative 5′ splice site (A5SS). Pathway enrichment was assessed using Gene Ontology/KEGG, STRING protein-protein interaction (PPI) networks, and Molecular Complex Detection (MCODE) was used to identify protein hubs. To determine computational prediction of docking, SPHINX31 was docked into SRPK1 (PDB 5MY8) using SP/XP docking and MM-GBSA rescoring.

SRPK1 inhibition was associated with distinct responses that were HPV-related. In C33A cells, upregulated genes were enriched for translation, RNA processing, and glycosylation, with KEGG highlighting ribosome and metabolic modules. Ribosomal hubs dominated the PPI/MCODE, suggesting possible translational and metabolic adjustments. In contrast, SiHa cells exhibited transcriptomic changes consistent with reduced expression of genes linked to Hippo, Wnt, PI3K-AKT, ERK1/2 signaling, migration, angiogenesis, and growth factor cytokine networks. Targets of YAP/TAZ (e.g., CCND1, BIRC5, SNAI2, SERPINE1) and their regulators (RASSF1, CSNK1E) were suppressed. At the splicing level, SiHa cells displayed fewer total AS events but with larger effect sizes, particularly in A3SS/A5SS. C33A cells showed abundant SE (59,234 events; small median ΔPSI) and RI (1,770 events, often binary), including complete shifts in HLA-DRB1/PLIN2 (+1.00) and KLF4 (-1.00). Notable A5SS switches included LEF1 (+1.00) and CDK6 (-1.00) in C33A, and DLX1/MRPL14/THAP5 (-1.00) in SiHa. Docking computationally predicted the low-energy poses of SPHINX31 in the SRPK1 ATP pocket. While not definitive, this evidence may potentially support the transcriptomic and splicing findings.

SRPK1 inhibition may remodel the cervical cancer transcriptome in an HPV-linked manner, with SiHa cells exhibiting changes consistent with suppression of oncogenic signaling, whereas C33A cells adapt through translational and metabolic reprogramming.