Nucleotide sequence comparisons between several strains and isolates of human cytomegalovirus reveal alternate start...

Mutations abound in all viral populations, which are thus rendered adaptable to changes in environmental conditions. Human cytomegalovirus (HCMV) is an important human pathogen for investigating nucleotide sequence variations because they can affect its potential to cause disease. We have determined part of the nucleotide sequence of the Toledo strain and compared it to the published sequences of the strains AD169, Toledo, and Towne and of three clinical isolates. Overall nucleotide sequence divergence between strains AD169 and Toledo amounts to roughly 2%, with considerable variations across the viral genome. In aligning the Toledo nucleotide sequences with those of the other strains and clinical isolates, numerous amino-terminal extensions of the known open reading frames (ORFs) have been noted. These extensions carry additional AUG or non-canonical CUG or GUG translational initiation codons. CUG and GUG have previously been shown to serve as translational start codons in prokaryotic and eukaryotic systems. Six of the more closely inspected extensions start with an AUG, 26 with a CUG, and 26 with a GUG. Some of these extended sequences might bestow altered biological properties upon HCMV proteins. These ORF extensions are common to the sequenced genomes of most of the HCMV strains or isolates. Supporting evidence for their functionality comes from studies on HCMV mRNAs that were isolated from HCMV-infected human cells. Several of these viral mRNA sequences carry the identified ORF extensions. Moreover, in the amino-terminal ORF extensions, codon usage in general resembles that in the main parts of several of the HCMV genes analyzed for this property.