|Home | About | Journals | Submit | Contact Us | Français|
Although there are over 90 serotypes of Streptococcus pneumoniae, antimicrobial resistance is predominantly found in a limited number of serotypes/serogroups, namely 6, 9, 14, 19 and 23. There is no compelling mechanism to account for this restriction. We aimed to determine whether serotypes commonly associated with drug resistance have higher transformation frequencies than those that are susceptible to antimicrobial agents. An in vitro investigation of the genetic transformation frequency of drug-resistant serotypes compared with that of susceptible serotypes under the influence of synthetic competence-stimulating peptides was performed. The transforming DNA was genomic DNA carrying a Tn916-like transposon containing the mefE gene that confers resistance to erythromycin. It was observed that serotypes 6, 9, 14, 19 and 23, which are highly associated with drug resistance, do not exhibit a higher degree of transformation efficiency than other serotypes. These findings suggest that the association of serotype with drug resistance is likely due to prolonged exposure to transforming DNA resulting from longer nasopharyngeal carriage and to a greater selective pressure from antimicrobials, particularly in children. This is the first study to compare the transformation frequencies of pneumococcal clinical isolates using genomic DNA that carries the composite Tn916-like element.
Streptococcus pneumoniae is a worldwide frequent cause of potentially life-threatening infections such as pneumonia, meningitis and bacteraemia, as well as otitis media and sinusitis. It causes excess morbidity and mortality in human immunodeficiency virus/acquired immune deficiency syndrome (HIV/AIDS) patients, young children, the elderly and debilitated patients [1,2]. Whilst the burden of pneumococcal disease is globally distributed, it is worse in developing countries where it is estimated to be responsible for greater than one million childhood deaths per year [3,4]. In sub-Saharan Africa, pneumococci are a significant cause of morbidity and mortality because of limited access to adequate health care and the high prevalence of HIV infection .
The number of reported antibiotic-resistant clinical isolates of S. pneumoniae has increased markedly over the last 30 years, and multidrug-resistant (MDR) strains have emerged [6–9]. The global emergence and spread of antimicrobial resistance in S. pneumoniae is a serious concern [9,10]. Many of the worldwide resistant clones have appeared with different capsular types [9,11]. Such variants of clones are thought to arise through natural transformation involving recombinational replacements, within and around the capsular biosynthesis (cps) locus . Between 3% and 80% of pneumococcal illness is due to drug-resistant S. pneumonia . Pneumococcal clones responsible for dispersion of resistance throughout the population locally and internationally have been identified in countries with a high prevalence of antibiotic-resistant strains .
Although over 90 pneumococcal serotypes exist and are assumed to be naturally competent for genetic transformation, antibiotic-resistant isolates, including those resistant to penicillin, usually belong to serotypes/serogroups 6, 9, 14, 19 or 23 and to a few defined clonal groups in most parts of the world [11,14–17]. Of these, the serotypes responsible for most pneumococcal disease and drug resistance are 6B, 14, 19F and 23F . This is particularly so in developing countries where the use of conjugate vaccine is still limited. The most successful clone in terms of geographical dispersion and prevalence is the MDR serotype 23F pandemic clone (Spain23F-1) , which originated in Spain and has been found in Mexico, South Africa, South Korea, Portugal, Croatia, France and the USA. Many penicillin-resistant strains are also resistant to other drugs such as chloramphenicol, erythromycin, tetracycline and trimethoprim/sulfamethoxazole . Even with the availability of antimicrobial agents, vaccines and intensive care, S. pneumoniae remains a frequent cause of life-threatening infections.
Hitherto, no compelling mechanism for the restriction of drug resistance to a few S. pneumoniae serotypes/serogroups has been proposed . Induction of competence for natural genetic transformation in S. pneumoniae depends on pheromone-mediated cell–cell communication and a signalling pathway initiated by a competence-stimulating peptide (CSP) . Two major CSP variants, CSP1 and CSP2, are found in members of this species, with individual strains responding to only one of these variants. Perhaps differences in the natural ability to develop competence for genetic transformation contribute to the development of drug resistance among various serotypes. Transformation with DNA from resistant strains of other species of streptococci is thought to be a major mechanism for the initial development of resistance in S. pneumoniae, since it is the major mechanism of horizontal gene transfer in this species. As such, we postulated that serotypes/serogroups commonly associated with drug resistance would have higher transformation frequencies than susceptible ones. The transformation frequencies of selected serotypes/serogroups of S. pneumoniae were compared and we report that serotypes/serogroups 6, 9, 14, 19 and 23, which are highly associated with drug resistance phenotypes, do not have a higher degree of transformation frequency than other serotypes/serogroups.
Streptococcus pneumoniae strains used in this study were selected from the collection of one of the authors (MRJ). Strains of serotypes 1, 3, 6B, 9V, 14, 18A/B, 19F, 23F, 29, 33, 35B and R6, all susceptible to erythromycin, were used as recipients during transformation experiments. Strain S222 of serotype 19F, provided by Dr Krzystof Trzcinski (Department of Epidemiology, Harvard School of Public Health, Boston, MA), was used as a DNA donor in the study. S222 is an erythromycin-resistant strain that carries mefE and tetM genes on a composite Tn916-like element. All strains except R6 were clinical isolates that were previously characterised and whose antimicrobial susceptibility profiles were determined. Strains were grown in stationary media either in tryptic soy broth (TSB) (Liofilchem, Roseto degli Abruzzi, Italy) supplemented with 10% horse serum (Sigma, St Louis, MO) or on tryptic soy blood agar base No. 2 (TSBAB) (Difco, Becton Dickinson, Sparks, MD) supplemented with 5% defibrinated sheep blood. Erythromycin (Sigma) was used at a concentration of 1 µg/mL in TSB and TSBAB for selection of resistant transformants.
Polysaccharide capsular types of the strains were determined using the capsular swelling reaction with factor-specific typing sera (Statens Serum Institut, Copenhagen, Denmark).
The DNA donor strain S222 was incubated overnight at 37 °C on TSBAB/erythromycin from which five colonies were picked and inoculated in 30 mL of TSB/erythromycin supplemented with 10% horse serum. Cultures were grown without shaking at 37 °C for 6 h followed by centrifugation at 4000 × g for 15 min. Chromosomal DNA was extracted from the bacterial pellet using the Epicentre Master Pure DNA Purification Kit (Epicentre Biotechnologies, Madison, WI) following the manufacturer’s instructions. Extracted DNA was stored at −20 °C pending transformation experiments.
Serotypes of recipient strains were selected according to serotypes commonly associated with drug resistance (6B, 9V, 14, 19F and 23F), serotypes rarely associated with drug resistance (1, 3 and 18A/B) and serotypes that have recently emerged as drug-resistant (29, 33 and 35B). R6, the easily transformable laboratory pneumococcal strain, was used as a positive control for transformation efficiency. Recipient strains were incubated overnight at 37 °C on antibiotic-free TSBAB plates. Subsequently, two colonies of each strain were inoculated into 3 mL of antibiotic-free TSB medium [pH 6.8, adjusted with 1 M HCl and supplemented with 1 mM CaCl2 and 0.2 % bovine serum albumin (BSA) (Sigma-Aldrich, Steinheim, Germany)]. Cultures were then grown at 37 °C to an optical density at 600 nm (OD600) of 0.1–0.2, after which they were stored at −80 °C in TSB/10% glycerol.
Competent recipient strains were thawed rapidly at room temperature just until ice had melted. Thawed recipient cells were then centrifuged at 13 000 × g (Eppendorf, Hamburg, Germany) for 10 min at 4 °C and the cell pellets were re-suspended in TSB (pH 8.0, adjusted with 5 N NaOH and supplemented with 1 mM CaCl2, 0.2% BSA and 10% horse serum). Cells were then revived by incubating at 37 °C for 45 min without shaking. Synthetic CSP1 or CSP2 (Elim Biopharmaceuticals, Hayward, CA) was then added to a final concentration of 0.4 µg/mL, after which cells were incubated at 37 °C for 15 min. Then, chromosomal DNA extracted from the donor strain S222 was added to a final saturating concentration of 1 µg/mL. The mixture was further incubated at 37 °C for 150 min without shaking and 250 µL of the culture was plated onto TSBAB/erythromycin plates that were then incubated for 24–48 h at 37 °C. To determine total bacterial counts, 250 µL of the same culture was serially diluted (1 × 10−5, 1 × 10−6 and 1 × 10−7) and 250 µL was plated on antibiotic-free TSBAB plates and incubated for 24–48 h. Bacterial counts, expressed as colony-forming units per plate, were determined on plates and the transformation frequency (TF) was calculated as follows: TF = CFUE/(CFU × DF), where CFUE is the colony-forming units on the TSBAB/erythromycin plate, CFU is the colony-forming units on the antibiotic-free TSBAB plate and DF is the dilution factor.
Control transformations had water added in place of extracted chromosomal DNA. To rule out contamination, strains without added chromosomal DNA and CSP were plated on plain media and incubated under the same conditions.
Transformants obtained as described above were subcultured on TSBAB/erythromycin plates with optochin disks to verify that they were pneumococci and that they had acquired the erythromycin resistance phenotype of the S222 DNA donor strain. Acquisition of the mefE gene was determined by polymerase chain reaction (PCR) amplification of chromosomal DNA extracts from the recipient strains before and after transformation. Chromosomal DNA was extracted from transformants using the Epicentre Master Pure DNA Purification Kit.
Chromosomal DNA from the donor strain S222 was used as a positive control in the PCR amplifications. The PCR reaction included 20 pmol of forward and reverse primers (mefE-fwd, 5’-AGTATCATTATTCACTAGTGC-3’ and mefE-rev, 5’-TCCTCCTGGTACTAAAAGTGG-3’ that amplify a 346-bp fragment of the mefE gene), 1.5 U of Taq polymerase, Custom PCR Master Mix (Thermo Scientific, Epsom, UK), template DNA and nuclease-free water in a total volume of 10 µL. The reaction was carried out in a Peltier thermocycler (MJ Research, Watertown, MA) at the following conditions: initial denaturation at 96 °C for 3 min; 30 cycles each consisting of 95 °C for 1 min, 52 °C for 1 min and 72 °C for 1 min; and a final extension at 72 °C for 5 min. Amplicons were separated by electrophoresis on a 2% agarose gel in TBE [Tris base–boric acid–ethylene diamine tetra-acetic acid (EDTA)]. For each strain, transformation and PCR were repeated at least three times for confirmation.
Transformation frequency raw data were converted into logarithms of transformation frequency. Differences in transformation frequencies among various serotypes were analysed by statistical analysis of variance (ANOVA) using SPSS software (SPSS Inc., Chicago, IL).
Of the 12 strains studied, only 8 (R6, 6B, 14, 19F, 3, 18A/B, 29 and 23F) resulted in growth of pneumococci on TSBAB/erythromycin following transformation. Of these eight strains recovered, seven had acquired phenotypic erythromycin and tetracycline resistance by disk testing and were shown by PCR to have acquired the erythromycin resistance-conferring gene mefE by showing an amplicon of 346 bp (Fig. 1). The remaining strain (serotype 23F) that acquired phenotypic erythromycin resistance without tetracycline resistance did not show any mefE band upon PCR and was presumed to have developed mutational erythromycin resistance (Fig. 1). The seven pneumococcal strains transformed were R6, 6B, 14, 19F, 3, 18A/B and 29, with transformation stimulated by CSP1 for five strains (R6, 19F, 3, 18A/B and 29) and CSP2 for two strains (6B and 14). Transformation frequency results are shown in Table 1. Serotype 3 had the highest transformation efficiency (2.7 × 10−5) and serotype 18A/B had the lowest (7.9 × 10−8). Mean log transformation frequencies of the transformant strains were similar (P = 0.679) (Table 1). None of the strains was transformed in parallel cultures (negative controls) without the addition of chromosomal DNA. The five strains that did not show transformation with CSP1 or CSP2 were 1, 9V, 23F, 33 and 35B.
Natural competence for genetic transformation in S. pneumoniae is mediated by a quorum-sensing-regulated mechanism in which CSP functions as an autoinducer [21,22]. The ability of synthetic CSP to induce competence in clinical encapsulated isolates  and the possibility that transformation can constitute a means for entry for transposons, which provide flanking homologous sequences that mediate homology-dependent recombination events , rendered transformation successful in this experimental system. Few studies have been performed so far to compare the competence and transformation frequencies of various clinical isolates of S. pneumoniae. To the best of our knowledge, this is the first study to compare the transformation frequencies of various clinical isolates of S. pneumoniae using whole genomic DNA that carries the composite Tn916-like element.
We investigated whether serotypes commonly associated with drug resistance have higher transformation frequencies than susceptible ones. All successful transformants were resistant to erythromycin as well as tetracycline, indicating that the entire Tn916-like transposon carrying mefE and tetM genes had been incorporated into the recipient genome. Six of eleven clinical serotypes were transformed, in which serotype 3 had the highest transformation frequency of 2.7 × 10−5. A higher transformation frequency of 4 × 10−3 was observed in a previous study for this serotype, which was transformed with genomic DNA carrying an erythromycin resistance gene . Differences in reported transformation frequencies could be due to the use of different competence media for transformation. In contrast, Hsieh et al.  assessed the competency of serotypes using plasmid DNA in which a very low transformation frequency of 7.8 × 10−9 was reported for serotype 3. The use of plasmid DNA as the transforming DNA could account for this difference. Furthermore, serotype 3 has a relatively rich amount of capsular polysaccharide, which apparently does not block uptake of exogenous DNA, making this serotype easily transformable at high frequency. These findings contradict earlier reports that large amounts of capsular polysaccharide have an inhibitory effect on transformation in S. pneumoniae .
Serotype 18 had the lowest transformation frequency of 7.9 × 10−8, similar to that observed in a previous study . Similarly, serotype 6B had a higher transformation frequency than serotypes 14, 19F and 18, and the profile of transformation frequencies for serotypes 6B, 14 and 19F were similar to those reported by Hsieh et al. . These similarities are intriguing since the mechanisms involved in the uptake and establishment of plasmid DNA is different compared with the uptake and homologous recombination events involved in processing genomic DNA . Although transforming DNA was extracted from serotype 19F, recipient strains of this serotype did not exhibit a higher transformation frequency than the majority of serotypes. This implies that capsule homology does not favour transformation frequency of the Tn916-like element, as previously observed .
Serotypes/serogroups 6, 9, 14, 19 and 23, which are commonly associated with drug resistance phenotypes [14–17], did not show higher transformation frequencies than serotypes 1, 3 and 18 that are commonly susceptible to antimicrobials. The reason why serotypes/serogroups 6, 9, 14, 19 and 23 are commonly associated with drug resistance phenotypes is still elusive. The reason could be that these serotypes are carried for longer periods in the nasopharynx of children who are exposed to antimicrobials more frequently than adults . In humans, the nasopharynx is one of the environments most colonised by microorganisms; it offers the availability of exogenous transforming DNA from bacteria co-colonising with pneumococci. While in the nasopharynx S. pneumoniae serotypes come into contact with other commensal bacteria, such as viridans group streptococci, Gemella spp., Streptococcus pyogenes and other closely related oral streptococci. These commensal bacteria may none the less carry a number of drug resistance genes and ultimately exchange genetic material with pneumococci sharing their habitat owing to the fact that pneumococci are naturally competent for genetic transformation. The acquired genes will subsequently increase the pneumococcal overall fitness. Many genes that confer drug resistance, virulence and colonisation have been acquired this manner [4,30,31]. Recent studies have shown that S. pneumoniae can also be transformed in vivo with whole genomic DNA from different species of viridans group streptococci and from Gemella spp. containing the mefE gene at a transformation frequency as high as 10−6 . Transfer of the mefE gene by transformation clearly indicates that gene transfer can occur between commensal bacteria and S. pneumoniae, as has already occurred for genes encoding penicillin-binding proteins [30,32]. This further supports the fact that S. pneumoniae in the nasopharynx can take up exogenous DNA that may contain drug resistance genes from commensal microorganisms.
Five serotypes were not transformed, probably due to the low homology that could have blocked homologous recombination processes. Integration of genetic markers into the chromosome in naturally competent cells requires DNA processing during entry. Lefrancois et al.  reported a number of possible reasons that can result in failure to obtain chromosomal transformants, which can probably be extrapolated to the inability to obtain transformants for these serotypes. The inability to detect transformants if they are very rare, the absence of induced competence or the absence of requirements of the complex process at the entry stages during natural competence to integrate the recipient chromosome were reported as possible reasons for failed natural transformation in pneumococci . Alternatively, the homologous recombination machinery of the non-transformed cells could have failed to process the large size of the composite transposon . Failure to transform encapsulated pneumococci with whole genomic DNA is common, especially with larger markers . Among the strains that did not transform was serotype 23F, where a presumed resistant mutant at a mutation frequency of 8.7 × 10−8 was detected. This strain was also tested with primers that amplify the ermB gene, but this was negative (results not shown). We therefore speculated that this strain probably has a mutation in 23S ribosomal RNA and/or ribosomal protein L4 and/or L22; a few examples of clinical isolates with mutational resistance have been described [35–37].
The shortcoming of this study is that the Tn916-like transposon element used as a marker may not be representative enough for generalisation of transformation frequencies among various S. pneumoniae serotypes. Therefore, to establish the role of natural genetic transformation in relation to drug resistance among S. pneumoniae serotypes/serogroups, we recommend further studies involving a large number of serotypes/serogroups using several genetic markers in the transformation procedures. Furthermore, since this was an in vitro study, an in vivo study in mice aimed at assessing transformation frequency of the various S. pneumoniae serotypes is also necessary.
In addition, more studies on the role of CSP in genetic competence and transformation of S. pneumoniae are necessary since CSP is known to stimulate a variety of genes that drive virulence, pathogenesis, recombination and colonisation. Knowledge from such studies will be invaluable for drug and vaccine discovery since the genes are involved in important cellular processes of the pneumococcus. Currently, information from sequenced genomes of several pneumococcal serotypes is a boon to these studies.
The authors thank Dr Krzystof Trzcinski (Department of Epidemiology, Harvard School of Public Health, Boston, MA) for technical advice and for providing genomic DNA containing a ‘Janus Cassette’, which was used in optimising transformation experiments.
This work was supported in part by the Tuberculosis Research Unit (TBRU), Case Western Reserve University (Cleveland, OH) and the National Institutes of Health grant to MLJ (#SRO1A1075637-4). DPK is supported by the Fogarty International Center (award number U2RTW006879).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.