Env-less endogenous retroviruses are genomic superspreaders

Abstract

Endogenous retroviruses (ERVs) differ from typical retroviruses in being inherited through the host germline and therefore are a unique combination of pathogen and selfish genetic element. Some ERV lineages proliferate by infecting germline cells, as do typical retroviruses, whereas others lack the env gene required for virions to enter cells and thus behave like retrotransposons. We wished to know what factors determined the relative abundance of different ERV lineages, so we analyzed ERV loci recovered from 38 mammal genomes by in silico screening. By modeling the relationship between proliferation and replication mechanism in detail within one group, the intracisternal A-type particles (IAPs), and performing simple correlations across all ERV lineages, we show that when ERVs lose the env gene their proliferation within that genome is boosted by a factor of ∼30. We also show that ERV abundance follows the Pareto principle or 20/80 rule, with ∼20% of lineages containing 80% of the loci. This rule is observed in many biological systems, including infectious disease epidemics, where commonly ∼20% of the infected individuals are responsible for 80% of onward infection. We thus borrow simple epidemiological and ecological models and show that retrotransposition and loss of env is the trait that leads endogenous retroviruses to becoming genomic superspreaders that take over a significant proportion of their host's genome.

Fig. 1.

Phylogeny of mammals (57) with ERV megafamilies (see text) shown as colored circles (area is proportional to the percentage of the ERV loci in the genome represented by that family). The placing of megafamilies on the tree shows relative age but not origin (which may be considerably earlier). Scale bar shows approximate dates in host phylogeny. Asterisked taxa are treated as duplicates and excluded from our analysis of all ERV families. Name color shows how many IAP loci were found in each species (Table S1). A typical megafamily in one genome (Spermophilus) is shown colored red.

Fig. 2.

Phylogenetic tree of IAP loci. Expansions in host species that have had multiple invasions are colored. Integrity of env gene is shown by color of terminal branch: orange indicates the longest ORF (at least 75% of the full length); red indicates an ORF between 25 and 75% of the full length; blue indicates an ORF <25% of the full length. Black shows loci for which we could not extract sequences >13 kb. Solid and open circles show Shimodaira–Hasegawa (SH) support values > 0.90 and >0.75, respectively. The two blue triangles show switches of env. The published IAPE and IAP sequences are indicated.

Fig. 3.

Phylogenetic tree of the IAPs with the inferred ancestral states of their host species. Expansions are collapsed into single taxa (white triangles), and cross-species transmission events are indicated by yellow pentagons. Colored lines show ancestral states that, according to the available host sampling, can be attributed to a single host. Dashed lines show ancestral states that could not be resolved.

Fig. 4.

Histograms showing (A) how common are ERV families of different size (Inset shows right-hand tail expanded for clarity) and (B) how many loci in total are in these families. Lines are generated assuming a lognormal (solid black) or generalized Pareto (dashed red) distribution. (C) env integrity (relative to gag) for megafamilies and randomly selected smaller families. The horizontal axes have been scaled using the logarithm to base 2.