Discovery of a novel murine type C retrovirus by data mining

Abstract

Analysis of genomic and expression data allows both identification and characterization of novel retroviruses. We describe a recombinant type C murine retrovirus, similar to the Mus dunni endogenous retrovirus, with VL30-like long terminal repeats and murine leukemia virus-like coding sequences. This virus is present in multiple copies in the mouse genome and expressed in a range of mouse tissues.

FIG. 1

Maximum-likelihood phylogeny of an alignment of conserved regions of coding sequences from available whole genome sequences of mammalian type C leukemia retroviruses. Porcine endogenous retrovirus (PERV, accession AF038600), M. dunni endogenous retrovirus MDEV (AF053745), koala endogenous retrovirus (7) (KoRV, AF151794), Gibbon ape leukemia virus (GALV, M26927), gibbon ape leukemia virus strain X (GALVX, GLU60065), feline leukemia virus (FELV, M18247 and M19392), murine leukemia virus (MLV, MLU13766), Friend murine leukemia virus (FMLV, M93134 and M81185), rat leukemia virus (RLV, MLVGAG), Moloney murine leukemia virus (MMLV, REMLM). Sequences were aligned by eye; positions too variable to be confidently aligned were excluded from the analysis, leaving a final alignment of 4,779 bp. We assumed an HKY+Γ model, with transition-transversion ratio (ti/tv), (ti/tv) base composition, and gamma shape parameter (α) estimated from the data) (ti/tv = 3.1, α = 0.4). Scale in substitutions per site.

FIG. 2

Nucleotide sequence of the proviral genome of a novel type C retrovirus (MmERV) from positions 112341 to 121005 of M. musculus PAC clone 657p21 (accession no. AC005743) has a total length of 8,665 bp. The 9-bp imperfect inverted repeats that define the LTRs have been underlined. Note that 2 bp are deleted from the terminal copies of these repeats upon insertion. The CAT box is not clearly identifiable (13), though a candidate sequence appears upstream of the TATA box on the opposite strand. The repeat region spanning U3 and R is marked by bold horizontal bars, and the following features are identified: flanking 9-bp repeats (double underline), pol-like motif (bold type), 12-bp inverse palindrome (boxed), and polyadenylation signal (boxed). This region containts a putative ORF (213 bp, positions 378 to 590: MSFDPRALVYVLSCCCFIKSCLQHFEFGLSVFLGPRLSRGLSEGLPSGVFHLGARPGSVRPPRDPRPTLR). Upstream of the gag ATG start codon, a CTG start codon defines the start of 99 bp of glycosylated gag. The untranslated region between U5 and the start of glycosylated gag contains a variable number (two to six) of perfect 35-bp repeats.

FIG. 2

Nucleotide sequence of the proviral genome of a novel type C retrovirus (MmERV) from positions 112341 to 121005 of M. musculus PAC clone 657p21 (accession no. AC005743) has a total length of 8,665 bp. The 9-bp imperfect inverted repeats that define the LTRs have been underlined. Note that 2 bp are deleted from the terminal copies of these repeats upon insertion. The CAT box is not clearly identifiable (13), though a candidate sequence appears upstream of the TATA box on the opposite strand. The repeat region spanning U3 and R is marked by bold horizontal bars, and the following features are identified: flanking 9-bp repeats (double underline), pol-like motif (bold type), 12-bp inverse palindrome (boxed), and polyadenylation signal (boxed). This region containts a putative ORF (213 bp, positions 378 to 590: MSFDPRALVYVLSCCCFIKSCLQHFEFGLSVFLGPRLSRGLSEGLPSGVFHLGARPGSVRPPRDPRPTLR). Upstream of the gag ATG start codon, a CTG start codon defines the start of 99 bp of glycosylated gag. The untranslated region between U5 and the start of glycosylated gag contains a variable number (two to six) of perfect 35-bp repeats.

FIG. 2

Nucleotide sequence of the proviral genome of a novel type C retrovirus (MmERV) from positions 112341 to 121005 of M. musculus PAC clone 657p21 (accession no. AC005743) has a total length of 8,665 bp. The 9-bp imperfect inverted repeats that define the LTRs have been underlined. Note that 2 bp are deleted from the terminal copies of these repeats upon insertion. The CAT box is not clearly identifiable (13), though a candidate sequence appears upstream of the TATA box on the opposite strand. The repeat region spanning U3 and R is marked by bold horizontal bars, and the following features are identified: flanking 9-bp repeats (double underline), pol-like motif (bold type), 12-bp inverse palindrome (boxed), and polyadenylation signal (boxed). This region containts a putative ORF (213 bp, positions 378 to 590: MSFDPRALVYVLSCCCFIKSCLQHFEFGLSVFLGPRLSRGLSEGLPSGVFHLGARPGSVRPPRDPRPTLR). Upstream of the gag ATG start codon, a CTG start codon defines the start of 99 bp of glycosylated gag. The untranslated region between U5 and the start of glycosylated gag contains a variable number (two to six) of perfect 35-bp repeats.