Elusive preferred hosts or nucleic acid level selection? A commentary on: Evolutionary interpretations of mycobacteriophage biodiversity and host-range through the analysis of codon usage bias (Esposito et al. 2016)

While confirming the long held view that “viruses do not closely imitate the use of the [host’s] ... codon catalogue” (Grantham et al., 1986), it is nevertheless considered a “surprising finding” that “despite having the ability to infect the same host, many mycobacteriophages share little or no genetic similarity” (i.e. similarity in their “GC contents and codon utilization patterns;” Esposito et al., 2016). Arguing correctly that “efficient translation of a phage’s proteins within a host is optimized by the phage’s ability to match the codon usage patterns of their hosts,” the authors conclude that “the preferred host of many mycobacteriophages is not M. smegmatis, despite their having been isolated on M. smegmatis.” Thus, a virus and its elusive preferred hosts would have had similar GC% and codon usages, but the same virus could still infect a less-preferred host (M. smegmatis), where the virus-host similarity would be less evident.