Genes & Genetic Systems
Online ISSN : 1880-5779
Print ISSN : 1341-7568
ISSN-L : 1341-7568
Full papers
Identification of A2059G 23S rRNA and G439A rplC gene mutations in Streptococcus criceti strain OMZ 61, a strain resistant to azithromycin, josamycin and clindamycin
Haruki Tamura Arisa YamadaHirohisa Kato
Author information
JOURNAL OPEN ACCESS FULL-TEXT HTML

2015 Volume 90 Issue 5 Pages 259-267

Details
ABSTRACT

Streptococcus criceti is a cariogenic organism that belongs to the mutans streptococci. Of the four S. criceti strains, strain OMZ 61 has been identified as being resistant to erythromycin. Antimicrobial susceptibility testing showed that strain OMZ 61 is also resistant to azithromycin, josamycin and clindamycin but susceptible to tetracycline and tiamulin. DNA hybridization analysis of the 23S rRNA genes revealed that the hybridization patterns in strain OMZ 61 differed from those in the other three strains. We further analyzed the nucleotide sequences of a ribosomal RNA operon, the rrnD operon, and the rpsJrpsQ region including rplC and rplD genes for ribosomal proteins L3 and L4, respectively, in the four strains studied. Nucleotide sequence analysis indicated that strain OMZ 61 contains an A-to-G substitution at nucleotide position 2059, equivalent to Escherichia coli numbering 2058, in a 23S rRNA gene (rrlD) and a G-to-A substitution at nucleotide position 439 in the rplC gene, suggesting an amino acid residue change at position 147 from valine to isoleucine, whereas no mutation in the rplD gene was found. DNA sequencing and polymerase chain reaction-restriction fragment length polymorphism analysis showed that most or all of the 23S rRNA genes in strain OMZ 61 contain the A2059G mutation. These findings suggest that the resistance to erythromycin, azithromycin, josamycin and clindamycin in strain OMZ 61 is conferred by alterations in 23S rRNA and/or ribosomal protein L3. This is the first description of mutations in the 23S rRNA and rplC genes in mutans streptococci.

INTRODUCTION

Streptococcus criceti is a cariogenic agent in experimental models and is of hamster origin, but has been isolated from humans on rare occasions (Loesche, 1986). This species is a member of the mutans streptococci, together with S. mutans, S. sobrinus, S. ratti, S. downei and S. macacae, on the basis of 16S rRNA and RNase P RNA gene sequences (Täpp et al., 2003). Recently, it has been proposed from genomic analysis that S. criceti and S. downei be grouped into a clade distinct from the mutans group (Richards et al., 2014). In a previous study using four S. criceti strains, we found that strains HS-6T, HS-1 and E49 were susceptible to erythromycin, but strain OMZ 61 was resistant to the antibiotic (Tamura et al., 2012). Strain OMZ 61 was isolated from a high-sucrose diet-fed rat, and was subsequently habituated to erythromycin by serial passage in broth containing increasing levels of erythromycin, up to 1,000 μg/ml (Guggenheim et al., 1965). Regarding strain OMZ 61 in our laboratory, its resistance to erythromycin seems to have been induced in the previous laboratory.

Macrolide resistance in S. pneumoniae, a mitis group member, is mediated by the mef(A) gene encoding a macrolide efflux pump or by the erm(B) gene encoding a ribosomal RNA methylase (Wierzbowski et al., 2007), which methylates the N-6 position of nucleotide A2058, in Escherichia coli numbering, in 23S rRNA domain V (Takaya et al., 2013). Furthermore, mutations in the 23S rRNA genes and the rplD gene encoding the ribosomal protein L4 have been found in clinical isolates and laboratory-generated mutants of macrolide-resistant S. pneumoniae (Tait-Kamradt et al., 2000; Wierzbowski et al., 2007). Of note is that mutation of A2058 in the 23S rRNA gene confers resistance to macrolide, lincosamide and streptogramin type B in S. pneumoniae as well as in E. coli, Helicobacter pylori, Propionibacteria and Streptomyces ambofaciens (for review, see Vester and Douthwaite, 2001).

The aim of this study was to deduce the mechanism of resistance to macrolide in strain OMZ 61. We examined susceptibility to antibiotics including macrolides, lincosamide, tetracycline and pleuromutilin in four S. criceti strains. We analyzed the 23S rRNA genes of the four strains by a DNA hybridization method, and genetically characterized the 23S rRNA gene rrlD and a region including the rplC and rplD genes in strain OMZ 61. We further performed DNA sequencing and polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis and identified an A2059G mutation in the 23S rRNA genes and a G439A mutation in the rplC gene in strain OMZ 61.

MATERIALS AND METHODS

Bacterial strains and growth conditions

S. criceti strains HS-6T (= ATCC 19642T; T superscript indicates type strain), E49, HS-1 and OMZ 61 were grown anaerobically in brain heart infusion broth (Difco Laboratories, Detroit, MI, USA) at 37 ℃. The type strain HS-6T and strain OMZ 61 were obtained from the American Type Culture Collection (ATCC, Rockville, MD, USA) and Dr. H. Mukasa (Department of Chemistry, National Defense Medical College, Saitama, Japan), respectively. E. coli strain Stellar was purchased from Clontech Laboratories (Mountain View, CA, USA) and grown on Luria–Bertani broth or agar (Difco Laboratories) at 37 ℃. To select for E. coli transformants, ampicillin (50 μg/ml) was added to the media as needed.

DNA preparation, manipulation and transformation

S. criceti genomic DNA was purified using the Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA). All primers used in this study were designed based on the sequence data of S. criceti strain HS-6T (Richards et al., 2014) (DDBJ/EMBL/GenBank accession number AEUV02000002) and are listed in Table 1. To amplify a 902-bp fragment of the 23S rRNA genes for cloning, direct sequencing and PCR-RFLP analysis, PCR was performed with TaKaRa Taq DNA polymerase (Takara Bio, Otsu, Japan) and 23S rRNA gene-specific primers 23S-F and 23S-R, which were designed to target all five copies of the 23S rRNA genes in strain HS-6T (Richards et al., 2014). The PCR conditions were as follows: 95 ℃ for 30 sec, and then 20 (for cloning) or 35 (for direct sequencing and PCR-RFLP analysis) cycles of 95 ℃ for 5 sec and 60 ℃ for 1 min. The leuSopuAA region containing the rrnD operon and the rpsJrpsQ region containing the rplD gene were amplified by PCR with TaKaRa LA Taq DNA polymerase (Takara Bio) and the primer pairs DF and DR and LF and LR, respectively. PCR conditions were as follows: 94 ℃ for 1 min; 35 cycles of 94 ℃ for 10 sec, 58 ℃ for 15 sec and 72 ℃ for 7 to 8 min; and finally 72 ℃ for 8 min. PCR products were purified using a GeneClean II kit (MP Biomedicals, Santa Ana, CA, USA). To construct clones for the preparation of DNA probes or for the examination of substitutions in clones, PCR fragments (902 bp) of the 23S rRNA genes were amplified from genomic DNA of strains HS-6T or OMZ 61, respectively, and cloned into a pGEM-T Easy vector (Promega), which possesses two EcoRI sites in the flanking regions of the A/T cloning site. The authenticity of the constructs was confirmed by sequencing on both strands. Competent cells of E. coli strain Stellar were transformed according to the supplier’s instructions (Clontech Laboratories).

Table 1. Primers used in this study
DesignationSequence (5′ to 3′)a
For amplification of the 23S rRNA genes
 23S-FGACGTGTCTGGCCAAGCAGT
 23S-RGACTTTCGTCCCTGCTCGAGT
For amplification of the leuSopuAA region containing the rrnD operon
 DFCCTAAAGACCTTCCGCGAGAA
 DRCCCCTCTCAATTCCAAACCATAC
For amplification of the rpsJrpsQ region containing the rplD gene
 LFGCCTTAGCTATGATGCAAGAGGT
 LRGACACGTCCAACAAGGGTTCTG
a  Primers were designed from the nucleotide sequence of S. criceti strain HS-6T (Richards et al., 2014); DDBJ/EMBL/GenBank accession number AEUV02000002.

DNA hybridization analysis

Genomic DNA was di-gested with EcoRI-HF, high-fidelity restriction endonuclease EcoRI (New England BioLabs, Beverly, MA, USA), or with EcoRI-HF and restriction endonuclease BbsI (New England BioLabs). The fragments were resolved in 0.7% agarose gels by gel electrophoresis and transferred to Hybond-N+ nylon membranes (GE Healthcare UK, Little Chalfont, Buckinghamshire, UK). A 0.9-kb EcoRI fragment of the construct was digoxigenin-labeled using the DIG High Prime DNA Labeling and Detection Starter Kit I (Roche Diagnostics, Mannheim, Germany). Hybridization was performed at 42 ℃ for 16 h, followed by washing according to the manufacturer’s instructions, and fragments that hybridized to the DNA probe were visualized with a chromogenic alkaline phosphatase substrate, NBT/BCIP, after a 16-h incubation.

PCR-RFLP analysis

PCR was performed to amplify 902-bp fragments of the 23S rRNA genes of four S. criceti strains with primers 23S-F and 23S-R, as described above. The authenticity of the amplified fragments was confirmed by direct sequencing on both strands. PCR products were incubated with or without 5 U of BbsI in a 40-μl reaction for 1 h at 37 ℃, and fragments were resolved by agarose gel electrophoresis and stained using ethidium bromide. The 902-bp PCR products containing the A2059G mutation were expected to be cut into two fragments of 586 and 316 bp after BbsI digestion.

Sequence analyses

Sequencing of purified DNA was performed on a 3130 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) using a BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems). Sequence analysis was carried out using the computational package GENETYX ver.11 (Genetyx, Tokyo, Japan). Alignment of DNA sequence traces was achieved using ATGC ver. 7 (Genetyx). Deduced amino acid sequences were aligned using Clustal W (Genetyx).

Antimicrobial susceptibility testing

The minimum inhibitory concentrations (MICs) of azithromycin, josamycin, clindamycin, tetracycline and tiamulin were determined by broth dilution after a 20-h incubation at 37 ℃, as described previously (Kitagawa et al., 2011).

RESULTS

Susceptibility of S. criceti strains to antibiotics

As we previously observed that only strain OMZ 61 showed resistance to erythromycin (14-membered-ring group) among the four S. criceti strains studied (Tamura et al., 2012), we further examined other antibiotics including the 15- and 16-membered-ring macrolides azithromycin and josamycin, respectively, and clindamycin (Table 2). There were marked differences in the MICs of azithromycin, josamycin and clindamycin for the four strains: OMZ 61 exhibited high resistance (MICs: 2000, > 80 and 7.36 μg/ml, respectively), but the others were susceptible (MICs: 0.40, 0.20 and 0.015−0.030 μg/ml, respectively). The MICs of tetracycline for all strains ranged from 0.80 to 1.60 μg/ml. This finding showed that all strains studied were susceptible to tetracycline. The MICs of tiamulin, a peptidyl transferase inhibitor (Bøsling et al., 2003), for all strains ranged from 0.20 to 0.80 μg/ml, indicating that they were susceptible to it. Thus, strain OMZ 61 exhibited high resistance to azithromycin, josamycin and clindamycin.

Table 2. Susceptibilities of S. criceti strains to antibiotics
StrainMIC (μg/ml)a
AzithromycinJosamycinClindamycinTetracyclineTiamulin
HS-6T0.400.200.030.800.20
HS-10.400.200.030.800.40
E490.400.200.0150.800.40
OMZ 612000>807.361.600.80
a  Azithromycin (a 15-membered-ring macrolide), josamycin (a 16-membered-ring macrolide), clindamycin (a lincosamide), tetracycline and tiamulin (a pleuromutilin) were examined, and MICs were determined in three independent experiments with similar results.

DNA hybridization analysis of the 23S rRNA genes in S. criceti strains

As mutations of the 23S rRNA genes and the rplD gene of S. pneumoniae are associated with resistance to macrolides (Tait-Kamradt et al., 2000; Wierzbowski et al., 2007), we first examined the 23S rRNA genes (rrl genes) of the four S. criceti strains by DNA hybridization. Genome sequence analysis of strain HS-6T indicated that the genome contains five rrl genes and that five fragments, 12.8, 9.4, 9.2, 7.9 and 7.1 kb in size, should be detected after EcoRI digestion, whereas five fragments, 7.4, 6.0, 5.3, 3.8 and 3.3 kb in size, should be detected after EcoRI and BbsI double digestion (Richards et al., 2014). After EcoRI digestion, similar patterns of fragments hybridized to the probe were observed in all four strains: bands ranging from 18 to 6.3 kb were detected in each lane (lanes 1−4 in Fig. 1). However, after EcoRI and BbsI double digestion, a distinct pattern was found in strain OMZ 61 compared with those in strains HS-6T, HS-1 and E49: bands ranging from 7.4 to 5.3 kb and from 3.8 to 3.3 kb were observed in strains HS-6T, HS-1 and E49 (lanes 5−7), while four bands at 5.7, 4.3, 3.6 and 1.7 kb were observed in strain OMZ 61 (lane 8). This finding showed that the positions of the BbsI sites of the 23S rRNA genes (rrl genes) in strain OMZ 61 are distinct from those in the other strains.

Fig. 1.

DNA hybridization of four S. criceti strains for detection of 23S rRNA genes. Genomic DNA was digested with EcoRI (lanes 1–4) or with EcoRI and BbsI (lanes 5–8). Fragments hybridizing to the 0.9-kb 23S rRNA gene-specific probe were visualized. Lanes 1 and 5, strain HS-6T; lanes 2 and 6, strain HS-1; lanes 3 and 7, strain E49; lanes 4 and 8, strain OMZ 61; lane M, size marker. Positions of DNA markers are indicated on the left.

Identification of the leuSopuAA region containing the rrnD operon in S. criceti strains

To confirm the 23S rRNA gene mutation in strain OMZ 61, we selected the rrnD operon, which is similar to the rrnA operon of S. criceti strain HS-6T with respect to gene arrangement ‘16S-tRNAAla-23S-5S-tRNAAsn’ and is situated between the leuS and opuAA genes for leucyl-tRNA synthetase and a protein homologous to the ABC transporter, respectively, in strain HS-6T (Richards et al., 2014). We isolated and sequenced a DNA fragment of 6,557 bp in each strain (E49, HS-1 and OMZ 61; DDBJ/EMBL/GenBank accession numbers AB981371−AB981373, respectively), and found that the rrnD operon of all four strains contained the rrsD, rrlD and rrfD genes for 16S rRNA, 23S rRNA and 5S rRNA, respectively, and the tRNAAla (TGC) and tRNAAsn (GTT) genes (Fig. 2). The nucleotide sequences of strains HS-1 and E49 were identical to that of strain HS-6T and were divergent from that of strain OMZ 61 with respect to a substitution in the rrlD gene. The A-to-G substitution at nucleotide position 2059, which is equivalent to 2058 in the E. coli 23S rRNA gene, in the rrlD gene of strain OMZ 61 generated a site for BbsI, which recognizes the sequence 5′-GAAGACNN-3′ (N, any nucleotide), at nucleotide position 4758.

Fig. 2.

Schematic representation of the leuSopuAA region containing the rrnD operon of four S. criceti strains. Arrows represent genes for proteins or RNA, and double slanted lines indicate the lack of an apparent start codon in leuS and stop codon in opuAA. The open arrow represents the rrlD gene, which harbors an A-to-G nucleotide substitution at position 2059 of the gene in strain OMZ 61. Numbers inside and between arrows indicate the sizes of genes and intergenic regions, respectively, in base pairs. A PCR amplicon generated by the 23S rRNA gene-specific primers 23S-F and 23S-R from the rrlD gene is indicated below the open arrow. Physical maps of EcoRI and BbsI restriction enzymes are shown at the bottom for strains HS-6T, HS-1 and E49 (upper) and strain OMZ 61 (lower). The numbers mark positions of the indicated restriction enzyme cleavage sites (in parentheses) and lengths in base pairs of the fragments (in italics). The nucleotide sequence from 4751 to 4758 of strain OMZ 61 is shown, and the A-to-G substitution at position 4751, corresponding to position 2059 of the rrlD gene, in strain OMZ 61 is in bold. The leuSopuAA region of strain HS-6T originated from nucleotide sequence data (DDBJ/EMBL/GenBank accession number AEUV02000002) (Richards et al., 2014).

Identification of the rpsJrpsQ region including the rplD gene for ribosomal protein L4 in S. criceti strains

Next, we isolated and sequenced the rpsJrpsQ region (5,238 bp) in each strain (E49, HS-1 and OMZ 61; DDBJ/EMBL/GenBank accession numbers LC009004− LC009006, respectively). In this region of each strain, genes encoding 30S ribosomal proteins, S3 (rpsC gene), S10 (rpsJ), S17 (rpsQ) and S19 (rpsS), and 50S ribosomal proteins, L2 (rplB gene), L3 (rplC), L4 (rplD), L16 (rplP), L22 (rplV), L23 (rplW) and L29 (rpmC), have been identified. As an example, the region of strain OMZ 61 is schematically shown in Fig. 3A. Comparison of the nucleotide sequences among the four strains showed no mutation in any gene, including the rplD gene for protein L4, other than a G-to-A substitution at nucleotide position 439 of the rplC gene for protein L3 in strain OMZ 61 (G439A). This change in the rplC gene was predicted to produce an amino acid substitution from valine (GTT) to isoleucine (ATT) in codon 147 of the protein L3 (Val147Ile) (Fig. 3B). Interestingly, a mutation at position 149 in E. coli L3 protein and mutations at positions 148 and 149 of Brachyspira spp. L3 proteins have been shown to be associated with reduced susceptibility to the peptidyl transferase inhibitor tiamulin (Bøsling et al., 2003; Pringle et al., 2004). By analysis of the deduced amino acid sequences, L3 protein of S. criceti strain OMZ 61 showed 93.8, 49.1 and 43.4% identity to L3 proteins of S. mutans (DDBJ/EMBL/GenBank accession number NP_722322), E. coli (CAA26460) and B. pilosicoli (AAG27264), respectively. As strain OMZ 61 showed susceptibility to tiamulin, as did strains HS-6T, E49 and HS-1 (Table 2), the rplC gene mutation in strain OMZ 61 is unlikely to be associated with susceptibility to tiamulin.

Fig. 3.

The rpsJrpsQ region including the rplC and rplD genes for ribosomal proteins L3 and L4, respectively, of S. criceti strain OMZ 61 and the deduced amino acid sequence alignment of L3 proteins. (A) Schematic representation of the rpsJrpsQ region of strain OMZ 61. Arrows with and without double slanted lines indicate open reading frames (ORFs) lacking an apparent stop codon and ORFs, respectively, and numbers inside and between arrows indicate the sizes of ORFs and intergenic regions, respectively, in base pairs. The open arrow represents the rplC gene harboring a nucleotide substitution, G439A, which is predicted to lead to a missense mutation at codon 147 (GTT→ATT) of the protein L3 (Val147Ile), in strain OMZ 61. (B) Deduced amino acid sequence alignment of L3 proteins with the L3 protein of S. criceti strain OMZ 61. Dashes represent gaps introduced for alignment. Asterisks and dots denote completely conserved and strongly similar residues, respectively, in the four sequences. Abbreviations: Sc, S. criceti strain OMZ 61; Sm, S. mutans (DDBJ/EMBL/GenBank accession number NP_722322); Ec, E. coli (CAA26460); Bp, Brachyspira pilosicoli (AAG27264). The 147th codon isoleucine (I) in strain OMZ 61 is in bold. The 147th codon valine in strains HS-6T, HS-1 and E49 is shown as 147V above the alignment. The positions of L3 protein amino acid residues in E. coli and Brachyspira spp. isolates that are associated with reduced susceptibility to a peptidyl transferase inhibitor, tiamulin, are underlined.

Sequence analyses of the 23S rRNA genes in S. criceti strains

As an A2059G substitution in the rrlD gene was identified in strain OMZ 61, fragments of 902 bp were amplified from the 23S rRNA genes in each strain and sequenced directly. No mixed trace was found in the sequenced regions of any of the four strains. Strains HS-6T, HS-1 and E49 showed identical sequences and an A at position 2059, which is identical to A2058 in E. coli 23S rRNA and is conserved in bacteria (Xie et al., 2012). Strain OMZ 61 showed an A-to-G change at position 2059 (Fig. 4), suggesting that the A2059G alteration was predominant in the 23S rRNA genes of strain OMZ 61. To examine this possibility, we separately cloned the 902-bp fragments amplified from the OMZ 61 chromosome into a plasmid and sequenced the inserts. Comparing 20 cloned inserts excluding the primer sequences with an 861-bp region of the 23S rRNA gene sequence (nucleotides 1501 to 2361 of rrlA of strain HS-6T), a total of 13 substitution types, namely, a single substitution, 10 double substitutions and 2 triple substitutions, were detected (Table 3). Among these substitutions, the A2059G substitution was observed in all 20 clones examined. This finding supported the suggestion that the A2059G mutation was predominant in the 23S rRNA genes of strain OMZ 61. We assume that the majority of mutations other than A2059G identified are due to PCR errors, although we cannot exclude the possibility that they might result from differences among the alleles of the 23S rRNA genes.

Fig. 4.

Sequencing traces of PCR products of the 23S rRNA genes of four S. criceti strains. Sequencing traces were aligned with the nucleotide sequence of the rrlA gene, nucleotides 2040 to 2071, of strain HS-6T (Richards et al., 2014) (DDBJ/EMBL/GenBank accession number AEUV02000002) as a reference. Numbering of the reference sequence is shown, with the E. coli numbering given in parentheses. The arrow represents an A-to-G substitution at nucleotide position 2059, E. coli numbering 2058, of the 23S rRNA genes in strain OMZ 61. Bars represent half the peak height generated automatically by the alignment software ATGC ver. 7 (Genetyx).

Table 3. Substitutions in 20 clones containing part of the 23S rRNA genes of strain OMZ 61
Substitution(s)aNumber of clones
A2059G7
A1502T/A2059G1
T1629C/A2059G1
G1660A/A2059G1
A1758T/A2059G1
A1761G/A2059G1
T1804A/A2059G1
A2059G/T2076C1
A2059G/T2206C1
A2059G/T2230C2
A2059G/T2317C1
T1618A/A1928G/A2059G1
A1673G/T2017A/A2059G1
a  Nucleotide substitutions were determined by comparison with the nucleotide sequence of rrlA (nucleotides 1501 to 2361) of S. criceti strain HS-6T (Richards et al., 2014); DDBJ/EMBL/GenBank accession number AEUV02000002.

PCR-RFLP analysis

PCR for 23S rRNA genes was performed and the products were then incubated with or without BbsI. After mock treatment, a single 0.9-kb fragment was found in the four strains (lanes 1−4 in Fig. 5). After BbsI digestion, a single 0.9-kb fragment was found in strains HS-6T, HS-1 and E49 (lanes 5−7), whereas two 0.6- and 0.3-kb fragments were observed in strain OMZ 61 (lane 8), indicating that the 0.9-kb fragment contained a site for BbsI.

Fig. 5.

PCR-RFLP patterns of four S. criceti strains. PCR products were incubated with or without BbsI. DNA fragments were separated through a 2% agarose gel and visualized by staining with ethidium bromide. Lane M, size marker; lanes 1 and 5, HS-6T; lanes 2 and 6, HS-1; lanes 3 and 7, E49; lanes 4 and 8, OMZ 61. The sizes of representative DNA standards are indicated on the right.

DISCUSSION

Antimicrobial susceptibility testing showed that the MICs of azithromycin, josamycin and clindamycin for strain OMZ 61 were the highest among those of the four S. criceti strains studied. In conjunction with our previous study showing that strain OMZ 61 is resistant to erythromycin (Tamura et al., 2012), these findings suggested that this strain possesses determinants of resistance to 14-, 15- and 16-membered-ring macrolides and lincosamide. Regarding macrolide resistance determinants, little is known about them in mutans group streptococci other than S. mutans strain V1, which was isolated from the heart valve of an infected endocarditis patient and showed resistance to erythromycin by carrying a gene homologous to erm(B) of S. pneumoniae (Nemoto et al., 2008). The erm(B) gene was shown to confer resistance to lincosamides and 14-, 15- and 16-membered-ring macrolides (Syrogiannopoulos et al., 2003); it was carried by either Tn917 or omega elements and each element was integrated into the Tn916 transposon, which carries the tetracycline resistance gene tet(M) in S. pneumoniae (Croucher et al., 2011). Of note is that the Tn916-related transposons Tn6002, Tn6003 and Tn3872 containing an unexpressed ‘silent’ tet(M) were found in erm(B)-positive tetracycline-susceptible isolates of S. pneumoniae, S. mitis and S. oralis (Cochetti et al., 2007; Brenciani et al., 2014). As erythromycin-resistant strain OMZ 61 was shown to be susceptible to tetracycline, we analyzed this strain in advance by PCR with primers for erm(B) (Amezaga et al., 2002) and tet(M) genes [primers tetmSb and tetmAb (Zeng et al., 2006)] and found it to be PCR-negative for both erm(B) and tet(M). Furthermore, we analyzed this strain by PCR with primers mefSb and mefAb (Zeng et al., 2006) for mef(A or E) genes encoding an efflux pump, and found it to be PCR-negative. These results suggest that other mechanisms are involved in the resistance to macrolides and lincosamide.

Genome sequence analysis indicated that S. criceti HS-6T contains five rrl genes (rrlArrlE), 2,896 to 2,906 nucleotides in length, and that the rrlA and rrlD genes are identical (2,901 nucleotides) and all five genes have A at positions corresponding to position 2059 of the rrlA gene (Richards et al., 2014) (accession number AEUV02000002). Moreover, sequence analysis of strain HS-6T indicated that a 12.8-kb EcoRI fragment after single digestion and a 3.3-kb BbsI fragment after double digestion should contain the rrlD gene (Fig. 2; Richards et al., 2014). As expected, the 3.3-kb band of strain HS-6T after double digestion was detected by the DNA hybridization procedure; however, the 12.8-kb band after single digestion was masked by other rrl gene-containing bands. DNA hybridization analysis of the rrl genes showed that the positions of BbsI sites in strain OMZ 61 were distinct from those in strains HS-6T, HS-1 and E49. Indeed, nucleotide sequence analysis revealed the A2059G mutation in the rrlD gene of strain OMZ 61, resulting in the generation of a site for BbsI. Sequence analysis of the rrlD gene of strain OMZ 61 predicted that 1.7-kb and 1.5-kb BbsI fragments should be detected after double digestion; however, DNA hybridization analysis showed only the 1.7-kb band. It is presumed that, under the wash conditions, the probe hybridizing to the 0.3-kb portion of the 1.5-kb fragment might have washed out. Similarly, the weak signals of the upper bands of strain OMZ 61 (Fig. 1, lane 8) may be related to the complementary segment length of the targets interacting with the probe; however, the possibility of partial digestion cannot be excluded.

A2059 of the rrlD gene in S. criceti corresponds to A2058 of the 23S rRNA gene in E. coli, with the A at position 2058 being a bacteria-specific base, in contrast to G, which is shared between archaea and eukaryotes (for instance, G2400 in the 25S rRNA sequence of Saccharomyces cerevisiae) (Xie et al., 2012). Furthermore, nucleotides A2058, A2059 and A2062 of the 23S rRNA in E. coli have been shown to be essential for macrolide binding (Tenson et al., 2003). The A2058G mutation of the 23S rRNA gene in E. coli was shown to confer erythromycin resistance (Vester and Garrett, 1987), and the affinity of erythromycin was reduced 104-fold in the A2058G mutant (Douthwaite and Aagaard, 1993). A2058G mutations in S. pneumoniae confer resistance to 14-, 15- and 16-membered-ring macrolides and lincosamide (Tait-Kamradt et al., 2000); thus, it is suggested that the A2059G mutation in the rrlD gene of S. criceti strain OMZ 61 confers the resistance to macrolides and lincosamide.

Nucleotide sequence analysis revealed the G439A mutation of rplC in strain OMZ 61, indicating a Val147Ile alteration in ribosomal protein L3. As strain OMZ 61 was shown to be susceptible to a peptidyl transferase inhibitor, tiamulin, like the other strains, this change is unlikely to affect such susceptibility. In E. coli protein L3, an Asn149Asp mutation situated close to the peptidyl transferase center was found to confer resistance to tiamulin (Bøsling et al., 2003). Similarly, three mutations, Asn148Ser, Asn148Lys and Ser149Ile, in L3 protein of Brachyspira spp. isolates were associated with reduced susceptibility to tiamulin (Pringle et al., 2004). To the best of our knowledge, no reports suggesting that ribosomal protein L3 is responsible for resistance to macrolides and lincosamide have been published, although we cannot rule out the possibility that the mutation in rplC is associated with the antimicrobial resistance in strain OMZ 61. As we failed to transform the S. criceti strains studied, we could not confirm whether the two mutations are responsible for resistance to macrolides and lincosamide in strain OMZ 61. Furthermore, no mutants resistant to erythromycin at a concentration of 10 μg/ml were obtained from strain HS-6T by nitrosoguanidine treatment and serial passage in broth containing increasing concentrations of the antibiotic; thus, we could not show an involvement of mutations in the 23S rRNA and rplC genes in erythromycin resistance.

It is noteworthy that, in mutans streptococci, mutations in the 23S rRNA and rplC genes were first identified in S. criceti strain OMZ 61. In S. pneumoniae, a single A2058G mutation was sufficient to cause elevation of erythromycin MICs, but macrolide resistance was conferred only when two or more alleles harbored an A2058G mutation (Tait-Kamradt et al., 2000). Direct sequencing analysis and sequence analysis of clones indicated that the A2059G mutation (A2058G in E. coli numbering) is predominant in alleles of strain OMZ 61. To confirm this, domain V of each 23S rRNA gene should be sequenced.

In accordance with the results of sequence analyses, the PCR-RFLP assay successfully identified the A2059G mutation in the 23S rRNA genes, suggesting its utility for detecting this mutation. Although we used BbsI in the assay, isoschizomers such as BbvII and BpuAI could be substituted (Pogge von Strandmann et al., 1992). Since the acquisition of macrolide resistance and the prevalence of the resistance determinants in pathogenic agents and commensal bacteria are serious issues, our goal is to develop a simple detection system for 23S rRNA mutations conferring macrolide resistance.

In conclusion, we observed that S. criceti strain OMZ 61 showed resistance to azithromycin, josamycin and clindamycin and it possessed two gene mutations, A2059G in the 23S rRNA genes and G439A in the rplC gene. The mutations in 23S rRNA and L3 protein are proposed to account for the resistance characteristics of strain OMZ 61.

ACKNOWLEDGMENTS

We are deeply indebted to Dr. H. Mukasa at the National Defense Medical College for providing S. criceti strain OMZ 61.

REFERENCES
 
© 2015 by The Genetics Society of Japan
feedback
Top