TEXT

Top Abstract Text References

Restriction-modification (R-M) systems are distributed in a variety of microorganisms (29). The simplest systems are type II R-M systems, which usually consist of two separate enzymes, a restriction endonuclease and a methyltransferase (4, 40). Genes for R-M systems have sometimes been found to be located on transferable elements, such as plasmids and bacteriophages, and in some cases, genes encoding proteins involved in DNA mobility, such as transposases, integrases, and invertases, are found in the vicinity of the genes for R-M systems (3, 5, 18, 21, 30, 35, 38-40). These genetic structures may facilitate the transfer of R-M systems and may contribute to the wide distribution of the R-M systems (17). The above findings, together with recent studies of the complete sequences of bacterial genomes, have led to a proposal that R-M systems are likely to be mobile genetic elements (2, 8, 17, 19, 25, 26, 36). However, a few examples in which genes for R-M systems are not colocalized with a gene for DNA transfer are known, indicating that other mechanisms are involved in the spread of R-M systems (9, 32, 41).

Streptococcus suis is a gram-positive, indifferent anaerobic coccus that has been implicated as the cause of a wide range of clinical diseases in swine and other domestic animals (10). Recently, we showed that some S. suis strains of serotypes 1 and 2 possessed a type II R-M system, whereas the type strain, NCTC10234, as well as some other strains, lacked a type II R-M system (32). This R-M system is an isoschizomer of Moraxella bovis MboI (12), which recognizes 5'-GATC-3', and the one found in the strain DAT1 was designated SsuDAT1I. SsuDAT1I is unique among all type II R-M systems described to date in that it contains two methyltransferase genes and two isoschizomeric restriction endonuclease genes. Nucleotide sequence analysis revealed that the R-M gene region comprising 3,503 bp has moved from distant sources and inserted in the intergenic sequence between purH and purD. Since the genetic organization was conserved among the strains tested, we suggested that the R-M genes have been horizontally transferred among S. suis strains. However, because the strains studied comprised a limited number of serotypes, we could not rule out the possibility that the strains having the R-M system were clonal. S. suis strains are presently classified into 35 capsular serotypes (13, 14, 16, 28), and the genetic heterogeneity and the phylogenetic diversity of the serotype reference strains have been described previously (7, 15, 27). In this communication, using the reference strains of serotypes 1 through 28 and a field isolate of serotype 1/2, we present further evidence that the R-M system was horizontally transferred among S. suis strains of different serotypes.

The bacterial strains used in this study are listed in Table 1. These strains, except DAT1 and NIAH11318, are the reference strains of each serotype. Strain NIAH11318 was isolated from a pig with meningitis in Japan. Strains NCTC10234 and NCTC10237 were purchased from the National Collection of Type Cultures, Central Public Health Laboratory, London, England. Strains of serotypes 3 through 8 and 9 through 28 were supplied by J. Henrichsen and M. Gottschalk, respectively, via Y. Kataoka. The strains described hereafter, except DAT1 and NCTC10234, are referred to by their serotypes. Bacteria were grown in Todd-Hewitt broth or agar medium (Difco Laboratories, Detroit, Mich.) at 37°C with 5% CO₂ for 18 h. Genomic DNA was isolated from the S. suis strains as described previously (32). Methylation of the adenine residue in the 5'-GATC-3' sequence was determined by testing the susceptibility of the DNA to restriction endonucleases DpnI and MboI, which are specific for methylated and unmethylated DNA, respectively. Restriction digests were analyzed by agarose gel electrophoresis using standard procedures (31). The genomic DNAs isolated from serotypes 1/2, 3, 7, 23, 24, and 26 could be digested to small fragments by DpnI but not by MboI, as was seen in strain DAT1 (32), while the converse was seen in the other serotypes (data not shown). This indicated that the genomic DNAs of the six serotypes above were methylated, whereas those of the others were unmethylated. Crude cellular extracts of the six serotypes were then prepared, and their DNA cleavage activities were examined using unmethylated pUC19 DNA substrate as described previously (32, 42). As shown in Fig. 1, the crude extracts prepared from serotypes 1/2, 3, 7, 23, and 26 showed restriction endonuclease activity indistinguishable from that of MboI. Thus, these five strains possessed an R-M system isoschizomeric to SsuDAT1I. The DNA cleavage activity of the extract from serotype 24 was weak, and nonspecific degradation of the substrate DNA occurred (Fig. 1); therefore, we could not determine the specificity of the enzyme.

View larger version (125K): [in this window] [in a new window]   FIG. 1.   Restriction cleavage activities of the crude extracts using unmethylated pUC19 as the substrate. Lane numbers represent the serotypes of the strains from which the crude extracts were prepared. M, pUC19 digested with MboI; S, 100-bp ladder size standards (GIBCO/BRL).

To examine the presence of genes homologous to SsuDAT1I genes, we performed genomic Southern hybridization using the same procedures as described previously (32). The 3,503-bp fragment of the SsuDAT1I region was amplified by PCR from the genomic DNA of strain DAT1 using primers 5'-AAGTATAGCACCCCAGCTGGAGAAG-3' and 5'-CTTGATTATCTAAACAAATCATGC-3', which were complementary to the 5' and 3' ends, respectively, of the 3,503-bp SsuDAT1I region. The conditions of the PCR were essentially the same as described previously (32). The amplified product was labeled with digoxigenin to probe S. suis genomic DNAs that had been digested with PvuII, for which no cutting site is present in the 3,503-bp SsuDAT1I region. A DNA fragment that strongly hybridized with the probe was seen in the digested DNAs of serotypes 1/2, 3, 7, 23, and 26 (Fig. 2). The sizes of the fragments in serotypes 1/2, 3, 7, and 26 were the same (4.5 kb) as that which appeared in the strain DAT1, while that which appeared in serotype 23 was larger (5.5 kb) than the others. These results indicated that the five strains possessed R-M genes which were highly homologous to the SsuDAT1I genes. However, no hybridizing fragment appeared in the digested DNAs of other serotypes, including serotype 24, which showed a methylated phenotype and a low level of restriction cleavage activity. This indicated that the strains which showed an unmethylated phenotype do not possess R-M genes homologous to the SsuDAT1I genes and that the R-M enzymes in serotype 24 may be encoded by nonhomologous genes which do not hybridize the SsuDAT1I probe under the experimental conditions.

View larger version (16K): [in this window] [in a new window]   FIG. 2.   Genomic Southern hybridization of S. suis strains probed with 3,503-bp fragment of the SsuDAT1I gene region. DNA was digested with PvuII and separated by agarose gel electrophoresis. Lane numbers represent the serotypes of the strains used. N, NCTC10234; D, DAT1. Molecular sizes of the signals that appeared are indicated on the left in kilobases.

The location and genetic organization of the R-M system described above were analyzed by PCR using primers 5'-GCGCTAGCTATTTTGACCAATA-3' and 5'-GCAAAGTCTTTTGACCACTCTA-3', which were designed to correspond to purH and purD sequences, respectively (32). The genomic DNAs of serotypes 1/2, 3, 7, 23, 24, and 26 were used as templates for test samples, and those of DAT1 and NCTC10234 were used for controls. A 4.3-kb fragment was amplified from serotypes 1/2, 3, 7, 23, and 26 as well as strain DAT1 (data not shown). Thus, these five strains were considered to possess the same genetic organization in the corresponding DNA region. On the other hand, when the genomic DNA of serotype 24 was used, a 1.0-kb fragment was amplified, as was also seen in strain NCTC10234 (data not shown), and the R-M genes were thus completely missing in this region. Therefore, the R-M genes, if they exist in serotype 24, may be present in a separate location.

The nucleotide sequences of the R-M genes in serotypes 1/2, 3, 7, 23, and 26 were directly determined using the 4.3-kb fragments described above. The R-M gene regions of all these serotypes consisted of 3,503 bp. The sequence alignment of the junction regions where the R-M genes were inserted showed that several features previously found in the junction regions (32) were also observed in these serotypes, i.e., (i) the insertion target site, 5'-GA(-/G)(T/A)  TTTG-3', as indicated by the arrow, (ii) a direct repeat of 3 bp, AAG, at both ends of the R-M region, and (iii) a lack of transposable element and long repeated sequences. Therefore, the unique structure of the SsuDAT1I genes, two methyltransferase genes and two restriction endonuclease genes without long repeated sequences or mobility genes, was substantially conserved among the six serotypes, indicating that these R-M systems shared the same origin.

Comparison of the 3,503-bp R-M gene sequences among serotypes 1/2, 3, 7, and 23 and DAT1 showed striking similarity (99.86 to 99.97% identity). However, relatively low similarity between serotype 26 and the other serotypes (94.38 to 94.43% identity) was noted. The deduced amino acid sequences of the R-M enzymes in serotypes 1/2, 3, 7 and 23 were completely identical to those of SsuDAT1I. On the other hand, some amino acid differences were found between the R-M enzymes of serotype 26 and those of SsuDAT1I. Therefore, the R-M system found in serotype 26 (strain 89-4109-1) was designated Ssu4109I. The amino acid sequences of four enzymes in Ssu4109I, designated M.Ssu4109IA, M.Ssu4109IB, R.Ssu4109IA, and R.Ssu4109IB, were aligned with those of SsuDAT1I and other closely related enzymes, DpnII and LlaDCHI (20, 23, 32) (Fig. 3). Among the amino acids that differed between Ssu4109I and SsuDAT1I, several amino acids, Asn₁₂₀, Asn₁₆₀, and Leu₁₉₂ (in M.Ssu4109IA), Lys₁₀₅ and Val₂₁₈ (in R.Ssu4109IA), and Glu₆₇, Ser₁₀₂, and Val₂₅₀ (in R.Ssu4109IB), are the same as the corresponding amino acids in DpnII and/or LlaDCHI.

View larger version (126K): [in this window] [in a new window]   FIG. 3.   Sequence alignment of M.SsuDAT1IA (MA.DAT1), M.Ssu4109IA (MA.4109), M.DpnII, and M.LlaA (A); M.SsuDAT1IB (MB.DAT1), M.Ssu4109IB (MB.4109), M.DpnA, and M.LlaB (B); and R.SsuDAT1IA (RA.DAT1), R.Ssu4109IA (RA.4109), R.DpnII, R.SsuDAT1IB (RB.DAT1), R.Ssu4109IB (RB.4109), and R.LlaDCHI (R.LlaC) (C). Black background, amino acids identical among all the sequences aligned; dashes, gaps in the aligned sequences; dots, positions where the amino acids of SsuDAT1I are different from those of Ssu4109I. DDBJ/EMBL/GenBank accession numbers of the sequences of the SsuDAT1I, DpnII, and LlaDCHI systems are AB045609 (32), M14339 (20), and U16027 (23), respectively.

Highly virulent strains of S. suis, especially strains of serotype 2, which caused meningitis and/or septicemia, have been shown to be genetically homogenous, whereas other strains isolated from pigs with pneumonia or healthy pigs were diverse (1, 6, 33, 34). Since the S. suis field isolates examined in our previous study were mostly isolated from diseased pigs with meningitis (32), it was possible that the strains we had used were clonal or closely related. In contrast, the S. suis strains used in this study had different origins in terms of species, health status, tissue origin, and geographical location (13, 14, 28). Phylogenetic analysis using 16S rRNA gene sequences showed that the S. suis reference strains were closely related and that the strains of serotypes 1/2 and 1 through 31 constituted a major group. However, they could be divided into three clusters (7). According to the cluster analysis (7), the reference strains of serotypes 1, 2, 3, and 23 were classified into cluster I, whereas those of serotype 7 and 26 were classified into clusters II and III, respectively. Furthermore, restriction fragment polymorphism analysis of rRNA genes showed that most of the serotype reference strains, especially serotypes 1/2, 3, 7, 23, and 26, were not closely related (15, 27). Therefore, the reference strains which were found to possess the R-M system appeared to belong to different clonal lineages.

The R-M system isoschizomeric to SsuDAT1I was found in five of the 29 strains of different serotype that were examined, and in each case the genes for the system were inserted in the same place in the genome, between purH and purD. No long repeated DNA sequences or transposable elements were found in the vicinity of the R-M systems. These findings support the hypothesis that the SsuDAT1I system was originally inserted in S. suis from a foreign source by illegitimate recombination and was subsequently transferred among S. suis strains by homologous recombination via conserved flanking housekeeping genes rather than by insertion of mobile genetic elements. The R-M gene sequences in serotypes 1/2, 3, 7, and 23 were very similar to those of SsuDAT1I. This may be due to the fact that the S. suis serotype 2 is predominant among isolates from pigs. Since many healthy pigs harbor S. suis as an early colonizer (11, 22, 24), the bacteria may encounter each other in vivo, and frequent colonization by these bacteria in pigs may facilitate an exchange of DNA. On the other hand, there was some deviation in sequences between the genes for Ssu4109I and the others, indicating their evolutionary divergence. This finding suggests an early original insertion in S. suis with transmission along two lines and recent lateral transfers from one line into the other strains, all of which were apparently mediated by homologous recombination.

Nucleotide sequence accession number. The nucleotide sequences for the R-M region in serotypes 1/2, 3, 7, 23, and 26 determined in this study have been deposited in the DDBJ/EMBL/GenBank database under accession no. AB058941, AB058942, AB058943, AB058944, and AB058945, respectively.