Page 106 - Read Online

Page 14 of 19 Ambros et al. Microbiome Res Rep 2023;2:34 https://dx.doi.org/10.20517/mrr.2023.18

Yoshikawa et al. demonstrated for Xanthomonas citri jumbo phage XacN1, that codon usage varies between
phage XacN1 and its host and that the phage-encoded tRNAs target codons used at a higher frequency by
[52]
the phage than its host .

The intact predicted L. curvatus (pro)phages of this study harboured predominantly isoleucine tRNA genes,
in which 7/10 use the CAU anticodon. Tomikawa et al. have shown that, for Lactiplantibacillus plantarum,
Ile2
the isoleucine tRNA using the CAU anticodon [tRNA (CAU)] is post-transcriptionally modified by the
enzyme tRNA -lysidine synthetase (TilS) to produce tRNA (LAU), which decodes the minor AUA
Ile2
Ile
[53]
isoleucine codon . Therefore, the high frequency of tRNA genes using the CAU anticodon in L. curvatus
prophages could be explained by the higher frequency of the minor AUA isoleucine codon used in the
coding regions of those phages in contrast to the coding regions of their hosts. Furthermore, all L. curvatus
strains with isoleucine tRNA gene harbouring intact prophages also carry the aforementioned tilS gene. The
hypothesis of L. curvatus phages bringing tRNA genes to offset the codon usage of their host is tentatively
supported by these results for this specific isoleucine tRNA. In contrast, L. curvatus (pro)phages harbouring
phenylalanine tRNA genes that exclusively use GAA as anticodon used the corresponding codons at a
similar rate as their hosts. It should be noted that the number of samples analysed in these comparisons was
relatively small [Phage/host genomes with respective tRNA-harbouring phages: n(tRNA ) = 8;
Ile
n(tRNA ) = 7] due to the limited number of accessible genomes for L. curvatus. As more genomes become
Phe
available for this species, more robust results can be obtained.

If the differences in codon usage between phage and host increase the amount of corresponding tRNA
genes in phage genomes, this phenomenon could theoretically also depend on the number of corresponding
tRNA genes within the host genome and the particular tRNA gene itself. In a previous study in Escherichia
coli, McDonald et al. analysed the retention of tRNA genes as a consequence of rapid influxes of
maladapted, foreign genes through lateral gene transfer, for example, via temperate phage infection/
[54]
prophage integration . The authors concluded that tRNA genes might selectively be kept by the host to
accommodate shifting demands in codon usage . Therefore, the transduction of tRNA genes by intact
[54]
predicted prophages in our study could partially support the observed differences in codon usage, such as
for the isoleucine codon AUA, although corresponding tRNA genes could also originate from other sources.

Interestingly, methyltransferase (MTases) genes were also annotated in some of the intact predicted
L. curvatus prophages. These MTases generally can occur alone, so-called “orphan” MTases, or in
combination with restriction endonucleases (REases), forming a restriction-modification system (R-M
system) . While R-M systems can be found in most bacterial and archaeal genomes as a protection
[55]
mechanism against foreign DNA (e.g., viral genomes, or transposons) , MTases are sometimes used by
[56]
phages to protect their genomes against R-M systems after host infection . In our analyses, we mainly
[57]
found “orphan” MTases within the prophages of L. curvatus, although three prophages (phage FAM25164
P1, phage MRS6 P1, and phage TMW 1.591 P1) harboured genes annotated as “integrase (tyrosine
recombinase)”, or “HNH homing endonuclease” directly downstream of those MTases.

For Escherichia coli phage P1, DNA methylation has been demonstrated to have a positive effect on the
efficiency of DNA packaging into the phage capsid . The question arises whether the supportive function
[58]
of DNA methylation is a general requirement for phage DNA packaging during the lytic cycle, and whether
this also applies to phages infecting L. curvatus. If this were the case, MTases should be expected in all intact
predicted prophages in this study. Three potential explanations could address this observation, which
requires future investigation. DNA methylation might (i) not be a ubiquitous requirement during phage
packaging; (ii) methylation might be performed by MTases provided by the host; and (iii) MTases are