Sequence and in vitro expression of the M2 gene of turkey rhinotracheitis pneumovirus.

Negative-stranded virion RNA and oligonucleotide primers complementary to fusion (F) protein gene sequences were used to generate cDNA clones, revealing that the gene 5'-proximal to the F protein corresponded to the M2 (22K) gene, as in respiratory syncytial (RS) virus. The transcription start signal, GGGACAAGU, was identical to that of the F and matrix (M) proteins of turkey rhinotracheitis virus (TRTV). There were two sequences with the potential to function as transcription termination/poly(A) signals, located at nucleotides 751 to 762 and 777 to 787; 15 clones derived from mRNA indicated that the first of these sequences formed the major signal. Part of the next downstream (5') gene was sequenced; unlike mammalian pneumoviruses the TRTV M2 gene did not overlap the beginning of the 5'-proximal gene. Northern blotting indicated that infected Vero cells contained less M2 mRNA than F mRNA and that about half of the M2 mRNA was present as a F-M2 dicistronic mRNA. The M2 gene contained two overlapping open reading frames (ORFs 1 and 2), as with RS virus. ORF 1 comprised 558 nucleotides with the coding potential for a 186 amino acid polypeptide, M(r) 20959, eight or nine residues shorter than for human RS virus strains. The overall amino acid identity was 40%, the N-terminal one-third of the proteins sharing 62% of residues, the remainder 29%. A hydropathy plot of the TRTV M2 protein had close similarity to that of the M2 or RS virus. The protein was predicted to have a basic character with no N-terminal signal sequence or other major highly hydrophobic sequences. In vitro translation of a transcript comprising both ORFs 1 and 2 produced a single product of apparent M(r) 23000, corresponding to the M2 product of ORF 1. Site-directed mutagenesis confirmed that this product was derived from ORF 1 and that frameshifting was not involved. The second ORF was expressed only from a transcript which lacked the AUG codons of ORF 1 and, although occupying a similar position to that in the RS virus M2 gene, had virtually no amino acid identity in its 73 residue length and was approximately 25% shorter than the corresponding RS virus ORF 2. The hydropathy plot of the potential products of the second ORFs of TRTV and RS virus showed little resemblance. Taken together these results suggest that ORF 2 is unlikely to be expressed in vivo.(ABSTRACT TRUNCATED AT 400 WORDS)