The piRNA pathway response to retroviral invasion in wild koalas
Date:
Endogenous retroviruses (ERVs) are significant contributors to mammalian genomes, composing ~8% of the human genome. Although ERVs drive genome evolution, their activities are closely related to genomic instability and diseases. Most ERVs integrated into mammalian genomes millions of years ago, making it challenging to study the earliest stages of endogenization. Koala retrovirus subtype A (KoRV-A) is one of the youngest ERVs in mammalian genomes, with its first endogenization events occurring < 50,000 years ago. KoRV-A provides a unique opportunity to directly analyze the endogenization of an infectious retrovirus and the host response to genome invasion.
PIWI-interacting RNAs (piRNAs) are 23- to 32-nt small RNAs expressed in the germline. piRNAs have a conserved role in repressing established transposable elements, including ERVs. Antisense piRNAs transcriptionally silence target elements by binding nascent transcripts and directing repressive DNA and chromatin modifications, while post-transcriptionally silencing targets through sequence-specific transcript cleavage. Sense piRNAs cleave complementary transcripts, generating precursors for antisense piRNAs and driving processing through ping-pong amplification and phased biogenesis. However, the initial response of the germline to retroviral invasion is not well understood.
We investigated the genome, transcriptome, DNA methylome, and piRNA profiles of eight wild koalas from two populations surrounding the Brisbane River. We found that in koalas south of the Brisbane River, KoRV-A is expressed in the testis, but its unspliced transcripts are processed into sense strand-biased piRNAs, resulting in pattern-specific post-transcriptional silencing. In striking contrast, KoRV-A expression is 10–fold lower in koalas north of the Brisbane River, sense and antisense piRNAs are of roughly equal abundance, and KoRV-A promoter DNA is methylated, consistent with sequence-specific transcriptional silencing. Our further analysis linked the reduction of KoRV-A expression to an antisense provirus in the 3´ UTR of the MAP4K4 gene, where antisense piRNAs are produced. Analysis of hundreds of additional koala genomes suggested that the MAP4K4 KoRV-A provirus is sweeping through the northern population and possibly under positive selection.
Our results indicate a two-phase genome immune response to retroviral invasion mediated by the piRNA pathway. Initially, unspliced retroviral transcripts are recognized as molecular patterns and processed into sense piRNAs, providing an immediate but inefficient barrier to retroviral activity. As the retrovirus spreads, its sequence is captured within the 3´ UTRs of host genes, generating antisense piRNAs, methylating retroviral promoters, and establishing efficient transcriptional silencing.