A rapid, sensitive and specific quantitative PCR assay has been developed for myxoma virus which complements previously described qPCR assays for orthopoxviruses (e.g. vaccinia virus, ) and parapoxviruses (e.g. Orf virus, ). This assay has been shown to function on tissue culture grown viruses and also directly on material extracted from clinical samples. It removes the requirement for specific antisera (which can be variable) or an electron microscope.
Using this assay for myxoma virus, we tested samples from Danish rabbits with clinically determined myxomatosis and we confirmed the diagnosis of the unusually large number of cases of myxomatosis in Denmark during 2007 (see Table 1). The cases were predominantly located in the eastern part of Denmark around Copenhagen, however a small number of cases were also observed elsewhere (as in most previous years). Unexpectedly, a specific frame-shift mutation was present within the coding sequence of the M135R gene in most of the current Danish samples. The insertion of a single additional G, into a run of 5 G's, was present in 10 out of 13 samples analysed. This was the only sequence difference in the M135R gene between each of these samples and the myxoma virus vaccine (which is also identical in this gene to the pathogenic Lausanne strain, ). The same nucleotide insertion was also detected in subsequent, independent, PCR reactions which used a different primer set. The generation of these amplicons enabled the fusion of the various M135R gene sequences to the coding sequence of the CAT gene (lacking a termination codon). Expression of these fusion gene plasmids within mammalian cells, in transient expression assays, provided clear evidence that the frame-shift mutation, as detected by the DNA sequencing, indeed resulted in the synthesis of a greatly truncated protein from the mutant M135R sequence (Figure 3).
The M135R gene has recently been described as a virulence factor in myxoma virus  since targeted deletion (removing 86% of the coding region) of the gene resulted in severe attenuation of the virus within European rabbits. Experimental inoculation of rabbits with the modified virus gave rise to only moderate symptoms and all animals survived whereas all animals receiving viruses with the intact M135R gene developed severe disease and had to be euthanized. Hence it was surprising that viral DNA obtained from rabbits with clinical disease encoded a greatly truncated version of this gene product comprising only 19 (out of the total 178) amino acids (plus 21 residues translated in a different reading frame). Clearly, the presence of this frame-shift mutation does not severely attenuate the virus within rabbits, since the samples analysed came from animals which had died or were showing significant clinical disease which had lead veterinarians to euthanize the animals. This result is not without precedent. Initial studies, using gene deletion, aimed at determining whether particular poxvirus genes were essential for growth indicated that the vaccinia virus F11L gene was essential for growth , however this gene is split in the modified vaccinia Ankara strain indicating that the intact gene is not essential . Subsequent studies using an improved approach demonstrated that this gene is indeed non-essential and showed that misleading results can be obtained from the use of insertional mutagenesis to test for viability . It is, therefore, possible that gene deletion can give misleading data on the requirement of specific genes for virulence. Thus, we conclude from our observations that expression of the full length M135R gene product is not essential for virulence in European rabbits. It would be interesting to assess by transcription profiling (e.g., as described for vaccinia virus [20, 21]) whether the targeted deletion of the myxoma virus M135R gene has an influence on the expression of neighbouring genes.
Very recently, Morales et al.  reported the complete genome sequence of a nonpathogenic myxoma virus strain (6918) (see ) and found that it encoded thirty-two amino acid substitutions plus six insertions or deletions leading to four frame-shift mutations (plus two intergenic region changes) compared to the parental Lausanne strain of the virus (which was 99.95% identical in sequence). One of these frame-shift mutations, within strain 6918, is within the M135R gene and is identical to the change observed within most of the recent Danish samples analysed. It is interesting to note that analysis of recent Portuguese isolates of myxoma virus found evidence of another frame-shift mutation within a virulent strain of this virus which also occurred at a homopolymeric region .
To analyse the similarity between the Danish samples and the avirulent 6918 strain, five other regions of the myxoma virus genomes from Danish samples with or without the M135R insertion were amplified by PCR and sequenced. These analyses showed that none of the other frame-shift mutations were present in the Danish samples although the deletion of a single T from within the M138L/M139R intergenic region was a common feature (Table 3). Morales et al.  suggested that the frame-shift mutation within the M135R gene may be an important determinant of the attenuation of the 6918 strain since, as described above, deletion of the gene had lead to attenuation . However, the properties of the Danish samples would suggest that some other sequence changes within the 6918 strain are responsible for this phenotype. The critical change(s) could either be amongst the insertions/deletions that are not present in the Danish samples or within the changes leading to the thirty-two different amino acid substitutions.