Systematic analysis of REBASE identifies numerous Type I restriction-modification systems that contain duplicated, variablehsdSspecificity genes that randomly switch methyltransferase specificity by recombination

Abstract

that contain duplicated, variable hsdS specificity genes that randomly switch 2 methyltransferase specificity by recombination. 3 4 5 Abstract 29 N 6 -adenine DNA methyltransferases associated with some Type I and Type III restriction-30 modification (R-M) systems are able to randomly switch expression by variation in the length of 31 locus-encoded simple sequence repeats (SSRs). SSR tract-length variation causes ON/OFF 32 switching of methyltransferase expression, resulting in genome-wide methylation differences, and 33 global changes in gene expression. These epigenetic regulatory systems are called phasevarions, 34 phase-variable regulons, and are widespread in bacteria. A distinct switching system has also been 35 described in Type I R-M systems, based on recombination-driven changes in hsdS genes, which 36 dictate the DNA target site. In order to determine the prevalence of recombination-driven 37 phasevarions, we generated a program called RecombinationRepeatSearch to interrogate REBASE 38 and identify the presence and number of inverted repeats of hsdS downstream of Type I R-M loci. 39We report that 5.9% of Type I R-M systems have duplicated variable hsdS genes containing 40 inverted repeats capable of phase-variation. We report the presence of these systems in the major 41 pathogens Enterococcus faecalis and Listeria monocytogenes, which will have important 42 implications for pathogenesis and vaccine development. These data suggest that in addition to SSR-43 driven phasevarions, many bacteria have independently evolved phase-variable Type I R-M 44 systems via recombination between multiple, variable hsdS genes. 45 Importance 46Many bacterial species contain DNA methyltransferases that have random on/off switching of 47 expression. These systems called phasevarions (phase-variable regulons) control the expression of 48 multiple genes by global methylation changes. In every previously characterised phasevarion, genes 49 involved in pathobiology, antibiotic resistance, and potential vaccine candidates are randomly 50 varied in their expression, commensurate with methyltransferase switching. A systematic study to 51 determine the extent of phasevarions controlled by invertible Type I R-M systems has never before 52 been performed. Understanding how bacteria regulate genes is key to the study of physiology, 53 virulence, and vaccine development; therefore it is critical to identify and characterize phase-54 variable methyltransferases controlling phasevarions. 55 56

The streptococcal phase-variable Type I Restriction-Modification system SsuCC20p dictates the methylome ofStreptococcus suisand impacts virulence

“…In contrast to previously characterized phase-variable Type I RM systems, we demonstrate that additional hsdS alleles further downstream of hsdM are not always silent. To the best of our knowledge, in all phase-variable Type I RM systems described so far, only the hsdS allele directly downstream of the hsdM is transcribed and involved in genome methylation (10,13,26,27,30,31,34). Therefore, it was suggested that the other TRDs or hsdS within the locus are merely present to function as DNA templates for recombination (34).…”

“…S5A) (8), thus we speculate that SsuPORF1588P is not functional in methylation. SsuPORF1273P is phase-variable (26) and the methylation profile of its two dominant alleles (A and B) has been solved by SMRT sequencing of E. coli overexpressing the S. suis methyltransferases, giving CC m6 AN 8 CTT for allele A and CC m6 AN 6 DNH (D=A/G/T, H=A/C/T and N=A/C/G/T) for allele B (27). SsuPORF1273P expression when grown overnight on BHI plate has been demonstrated, but genome methylation in S. suis has not been shown yet (26).…”

“…In previously identified Type I RM systems, the hsdS allele that is not located directly downstream of the hsdM gene is silent, because it is encoded on the opposite strand, lacks a start codon, or lacks a promoter (27). In SsuCC20p, a complete and functional hsdS is predicted not only directly downstream of the hsdM gene in allele A and H, but also downstream of xerD in allele D and E (Fig.…”