Endogenous gammaretrovirus acquisition in Mus musculus subspecies carrying functional variants of the XPR1 virus receptor

Abstract

The xenotropic and polytropic mouse leukemia viruses (X-MLVs and P-MLVs, respectively) have different host ranges but use the same functionally polymorphic receptor, XPR1, for entry. Endogenous retroviruses (ERVs) of these 2 gammaretrovirus subtypes are largely segregated in different house mouse subspecies, but both MLV types are found in the classical strains of laboratory mice, which are genetic mosaics of 3 wild mouse subspecies. To describe the subspecies origins of laboratory mouse XP-MLV ERVs and their coevolutionary trajectory with their XPR1 receptor, we screened the house mouse subspecies for known and novel Xpr1 variants and for the individual full-length XP-MLV ERVs found in the sequenced C57BL mouse genome. The 12 X-MLV ERVs predate the origins of laboratory mice; they were all traced to Japanese wild mice and are embedded in the 5% of the laboratory mouse genome derived from the Asian Mus musculus musculus and, in one case, in the <1% derived from M. m. castaneus. While all 31 P-MLV ERVs map to the 95% of the laboratory mouse genome derived from P-MLV-infected M. m. domesticus, no C57BL P-MLV ERVs were found in wild M. m. domesticus. All M. m. domesticus mice carry the fully permissive XPR1 receptor allele, but all of the various restrictive XPR1 receptors, including the X-MLV-restricting laboratory mouse Xpr1(n) and a novel M. m. castaneus allele, originated in X-MLV-infected Asian mice. Thus, P-MLV ERVs show more insertional polymorphism than X-MLVs, and these differences in ERV acquisition and fixation are linked to subspecies-specific and functionally distinct XPR1 receptor variants.

Fig 1

Detection of a representative intact provirus, Mpmv11, and a deleted provirus, Pmv16, in common inbred strains. (A) At the top is the Mpmv11 provirus structure with primers designed from the LTR and from flanking sequences, followed by a diagram of the empty preintegration locus. At the bottom right are PCR tests of 7 inbred strains that had been typed for this ERV by Southern blotting (8). (B) The Pmv16 genome, the empty locus, and the solo LTR are diagrammed with PCR primers and expected products, and PCR test results are provided for 8 strains.

Fig 2

Distribution of XP-MLV ERVs and Xpr1 receptor alleles in Mus taxa and in M. musculus-derived laboratory strains. A phylogenetic tree shows the house mouse (M. musculus) radiation. The inbred strain genomes are largely M. m. domesticus derived, but some of these strains acquired the restrictive Xpr1ⁿ receptor and X-MLV ERVs from Japanese fancy mice (M. m. molossinus).

Fig 3

Subspecific origins of C57BL XP-MLV ERVs. Horizontal tracks for each of 7 inbred strains represent 7- or 17-Mb genomic segments surrounding the ERV insertion sites from the Mouse Phylogeny Viewer (19). ERV sites are indicated by yellow lines, and the 3 track colors represent blocks that originated from the 3 subspecies that contributed to the laboratory mouse genome. To the right of each panel are PCR typing data for each ERV, color coded for all Xmvs except Xmv15, to match track colors at the ERV site.

Fig 4

Subspecific origins of Xpr1 alleles and XP-MLV ERVs and haplotype diversity at insertion sites. (A) Regions of haplotype similarity at a representative P-MLV ERV, Mpmv6. The black line marks the Chr 1 insertion site, and the different track colors represent different shared haplotypes. The Mpmv6 insertion lies in a region of haplotype similarity (tan color) shared by the 12 positive strains as determined by Mouse Phylogeny Viewer (19). (B and C) The unexpected detection of Xmv12 in M. m. domesticus strains RBB and SOD1 and Mpmv6 in M. m. molossinus strains MOLD and MOLG is because their insertion sites in these strains are M. m. musculus and M. m. domesticus derived, respectively, due to genomic contamination. (D) Subspecific origins of the two laboratory mouse Xpr1 alleles and two linked X-MLVs. In panels B to D, ERV sites are indicated by yellow lines, and track colors represent 3 Mus musculus subspecies. To the right are PCR typing data for each marker.

Fig 5

Identification of a novel Xpr1 variant in wild-caught M. m. castaneus. (A) Sequence variation in the receptor determining ECLs of the Xpr1 of 10 individual M. m. castaneus animals. At the top are sequences for the permissive allele, Xpr1^sxv, the common laboratory mouse allele, Xpr1ⁿ, and the Xpr1^c variant identified in inbred CAST/EiJ animals. Below are the two alleles identified in sequences of 10 wild-caught M. m. castaneus animals, Xpr1^c and Xpr1^c2. (B) Infectivity of XP-MLV lacZ pseudoviruses in E36 hamster cells expressing XPR1 variants. Light bars represent Xpr1^sxv with and without mutation F584I; the dark bars include the K500E mutation, which disables the ECL3 receptor determinant. Cells were infected with pseudoviruses expressing Env proteins of 4 XP-MLV host range variants. Infected cells were stained to detect the reporter gene 24 h after infection. Titers represent blue cells in 50 μl of virus stock ± standard errors of the means, and significant P values were determined from 3 independent tests. E36 cells are inefficiently infected by CAST-X. The Western blot shows expression of V5-tagged XPR1s and tubulin in transfected E36 cells.