# Evolutionary divergence of induced versus constitutive antiviral gene expression levels between primates and rodents

## TL;DR

This study compares how primates and rodents respond to viral infections at the gene expression level, finding that some genes are more active in one group than the other, which may explain differences in antiviral immunity.

## Contribution

The study introduces a method to identify genes with divergent expression in immune responses between primates and rodents.

## Key findings

- 578 genes show divergent expression in immune responses between humans and mice.
- Transcriptional divergence is linked to specific pathways and chromatin accessibility differences.
- Some genes are constitutively expressed in one species, potentially aiding rapid antiviral defense.

## Abstract

Hundreds of genes are upregulated in response to pathogen infection. These genes’ sequences often diverge across mammals, to counteract rapid pathogen evolution. However, the transcriptional divergence of these genes, their relative levels before and after infection in different host species, remains poorly understood. We studied this divergence by comparing gene expression before and after viral stimulation in cells from primates and rodents. We developed a method to identify orthologs strongly upregulated in one species that are unchanged in response to stimulus in another species. Using human and mouse data, we detected 578 transcriptionally divergent orthologous genes. For example, genes related to the NFκB complex are only upregulated in mouse. While most divergent genes do not belong to the same cellular process, several pathways and protein complexes are enriched in this set, suggesting that divergence in immune responses between closely related mammals is limited to specific modules rather than involving entire pathways. Transcriptional divergence between human and mouse orthologs was also observed when ortholog expression from different primates and rodents were compared, when responses were studied in several other cell types, and was recapitulated at the chromatin level, using histone mark patterns that denote active promoter regions. Surprisingly, these transcriptional changes were associated with evolutionary changes in coding sequences only when the genes are lowly expressed. In summary, we found genes whose orthologs diverge between primates and rodents in response to immune stimulation. Some of these genes are constitutively expressed in one species even before infection, potentially facilitating rapid antiviral activity that may be linked to clade-specific adaptation to confer greater resistance against pathogens. Further studies are required to test which of these transcriptional changes are adaptive, and what are their functional consequences. Moreover, comparative studies on diverse infections can point to additional species-specific responses and how they enable different species to overcome infection.

To limit viral replication, infected cells mount an innate immune response where many genes encoding for proteins with antiviral activities are upregulated. Some of these genes are continuously expressed in cells even without infection, to provide a rapid defense against incoming invaders. The identity of these genes and their level of expression before and after infection can vary between host species, in a manner that can impact the ability of different species to inhibit infection but is poorly understood. In this work, we develop an approach to detect these divergent genes, and characterize the genes that differ in expression between human and mouse cells. We show that these genes also vary in transcription between other primates and rodents and that transcriptional changes are in agreement with differences in chromatin accessibility. The divergent genes belong to specific pathways, suggesting that divergence in these pathways may contribute to primate- or rodent-specific antiviral resistance. Our work thus highlights potential differences between human, primates and rodents in their antiviral immunity, and can point to limitations of animal models in infection studies.