Cross-System Transcriptomics Suggests Enterovirus A71 Exploits Wound Healing Programs to Target Neural Progenitors in Human Spinal Cord Organoids.

INTRODUCTION

The mechanisms of EV-A71 immune evasion and neuroinvasion are poorly understood. Inactivated vaccines provide protective immunity without the dysregulated inflammation and viral persistence caused by active infection. In this bioinformatics-driven study, our objective was to utilize computational transcriptomics to investigate this dichotomy and determine the specific neural cell populations susceptible to enteroviral infection.

METHODS

We compared bulk RNA-seq data from THP-1 macrophages infected with active versus inactivated EV-A71 to identify viral hijacking programs. We then applied these systemic findings to the neural environment via an iterative Area Under the Curve (AUC)-based optimization strategy. This process refined the broad bulk-derived candidates into a 10-gene consensus signature, which was projected onto a human spinal cord organoid single-cell RNA-seq atlas to map viral susceptibility across distinct developmental lineages.

RESULTS

While active infection with EV-A71 is associated with a wound-healing program defined by the expression of PMEPA1, ALOX5, CCL1, and SPRED3, the administration of an inactivated virus elicits an anti-inflammatory response without activating a similar transcriptional program. In fact, decreased expression of inflammatory factors (ADORA2A, CCR1, CXCL1) marks this wound-healing stage and leads to the development of an immunosuppressive M2 phenotype. By projecting the highly robust 10-gene consensus (comprising ALDH2, SERPINE2, and L1CAM) against the organoid atlas, we have predicted that the cycling neural progenitors could be the most likely reservoir for EV-A71 replication. Using a Wilcoxon rank sum test with Benjamini-Hochberg correction, we observed the only statistically enriched signature in progenitor cells Discussion: Transcriptomic data indicate that EV-A71 might hijack TGF-β/PMEPA1 and ALOX5 signaling pathways to establish EMT-like behavior. The local acquisition of an M2-like immunosuppressive state might help promote early systemic dissemination, immunologic escape, followed by colonization of the proliferative neural niche prior to the development of neurological symptoms.

CONCLUSION

Computational identification of a macrophage gene signature provides robust discrimination of EV-A71 infection from vaccine-based protection in vivo. Mapping the signature to the context of human spinal cord organoids indicates that EV-A71 specifically targets proliferative and regenerative programs of neural stem cells.