Sequence heterogeneity in pneumonia virus of mice reveals G gene-dependent modulation of virulence

ABSTRACT Pneumonia virus of mice (PVM), the mouse homolog to respiratory syncytial virus (RSV), is increasingly used as a surrogate model to study pneumovirus pathogenesis in a more natural pathogen-host relationship. Two major strains of PVM, strain 15 and J3666, are currently used in laboratories, with preferences for either one or the other based on the well-documented isolation history of strain 15, or the suggested higher virulence of strain J3666. Using conventional and long-read sequencing, we found that the PVM strain J3666 represents two distinct virus populations, which are defined by the sequence and structure of the G and SH genes encoding the putative attachment and small hydrophobic proteins, in addition to further nucleotide polymorphisms. Specifically, a nucleotide polymorphism at position 65 in the G gene results in either an upstream open reading frame (uORF) preceding the main ORF in frame, or an extension of the major G ORF by 18 codons. The impact of the different forms of the J3666-G genes on PVM was examined by generating recombinant PVMs differing exclusively in the distinctive 5′ portion of the respective G gene. This revealed that the population with an extended main G ORF was more virulent than the population with a G gene containing an uORF or the parental virus. The presence of a uORF was associated with decreased expression levels of G, whereas the virus with the extended G ORF appeared to express slightly increased levels of G, which suggests that expression levels of G may modulate virulence. IMPORTANCE The pneumonia virus of mice strain J3666 is considered a more virulent and more suitable model for severe lower respiratory tract infections. The organization of the gene for the attachment protein G is reported to contain a small upstream open reading frame (uORF) preceding the main G ORF in frame. The translated G protein is predicted to comprise 396 amino acids. We report that this virus strain may be a mixture of two different populations, each with differing virulence. The more virulent population encodes a G protein of potentially 414 amino acids instead of a small uORF. The usage of the first start codon in this G gene organization remains to be determined. Importantly, this organization of the G gene is in line with that of several newly identified pneumoviruses, i.e., canine and swine pneumoviruses. These viruses may comprise a distinct group within the Pneumoviridae family.