Diversity of Multidrug Efflux Genes and Phenotypic Evaluation of the In vitro Resistance Dynamics of Clinical Staphylococcus Aureus Isolates Using Methicillin; a Model β-lactam

INTRODUCTION

MRSA is a Gram positive bacterium with a circular chromosome of 2.2-2.8 Mbp and varieties of extra-chromosomal elements that include plasmids, phages and other mobile elements [1]. The genome of S. aureus encodes chromosomal and extra-chromosomal virulence determinants, antibiotic resistance genes and multi-drug efflux genes (MDE) which play significant roles in the pathogenesis of S. aureus-associated infections and poor antibiotic therapeutic outcome [2]; thereby increasing patient suffering and sustained epidemics.

The MDE genes are bacterial survival mechanisms responsible for pumping out antibiotics and other biocides from the cytoplasm [3]. These efflux pumps can either be chromosomal or plasmid encoded and are able to extrude specific and/or different classes of antimicrobial compounds including biocides [2]. The MDEs are grouped into five families based on their structure and the kinetics of their activities: (i) the major facilitator superfamily (MFS), (ii) the small multi-drug resistance (SMR) family, (iii) the multi-drug and toxic compound extrusion (MATE) family, (iv) the resistance-nodulation-cell division (RND) superfamily and (v) the adenosine-triphosphate (ATP)-binding cassette (ABC) superfamily. Several MDEs have been described in S. aureus Table (S1) that contribute to the antimicrobial resistance mechanisms in this pathogen [2].

Previous studies have reported the prevalence and diversity of clinical MRSA isolate in Pretoria, South Africa [4, 5]. However, there are no reports on the distribution patterns of the multidrug efflux genes of the clinical isolates in South Africa. In addition, studies on antimicrobial profiling of S. aureus isolates often report the minimum inhibitor concentrations (MIC) of the tested antimicrobial and neglect the evaluation of the growth dynamics of the isolates; an assessment that could provide insight into the growth characteristics of the pathogen in response to antibiotics. This report reveals that all S. aureus isolates, be it of MRSA or MSSA genotype, do not contain the same set of MDEs despite being isolated from the same hospital setting and that there are at least four groups of specific growth-response pattern showing resistance to the model I lactam antibiotic, methicillin.

MATERIALS AND METHOD

RESULTS

Diversity and Distribution Pattern of MDEs in Clinical S. aureus Isolates

In other to evaluate the distribution pattern of the MDEs, all the isolates were reconfirmed as MRSA or MSSA and the PCR strategy used for the MDEs used proved to be robust and showed distinct and clear bands on the electrophoretic gel (Fig. S2). The distribution pattern of the detected MDEs in the clinical MRSA and MSSA isolates clustered into three major groups (Fig. S1 and Table S3). All the 11 MDE genes assessed in this study were detected in 35.05% (34/97) of the total isolates; out of which, 8.82% (3/34) were MSSA. The smr and qacA/B genes were not detected in 31.96% (31/97) of the all isolates, which, however; harboured nine other MDEs. The MSSA represented 16.1% (5/31) of this cluster. Moreover, MDEs including norC, sdrM and qacA/B were not detected in 12.37% (12/97) of the isolates, which hitherto had 8 other MDEs. This cluster included 50%(6/12) MSSA isolates. Other clusters/outliers included isolates numbers ranging from 1-4 and made up to 21.2% of the total S. aureus isolates. We did not observe any correlation between the genomic location (chromosomal or plasmid) of the MDEs and their distribution pattern in all the isolates.

A strong positive correlation (Pearson coefficient =1) was observed for the co-existence of norA, norB and tet38 efflux genes. This was followed by qacA/B and smr (Pearson coefficient = 0.89) which were present in 41.24% (40/97) and 44.33% (44/97) of the isolates, respectively (Table S4). Weak negative correlations were also observed for the co-existence of tet38 and other MDE genes including norA, norB and qacA/B (Table S4). Heatmap matrix Fig. (1) revealed the tendency of norC, sdrM, smr and qacA/B to be co-absent in some S. aureus strains especially in the MSSA (Table S5).

The observed individual prevalence of the other MDEs evaluated included norA (98.9%; 96/97); norB (98.9%; 96/97); mepA (97.9%; 95/97); tet38 (98.9%; 96/97); sepA (96.9%; 94/97); MdeA (95.9%; 93/97); lmrs (88.7%; 86/97); sdrM (85.6%; 83/97) and norC (79.4%; 77/97) (Table S5).

Cluster-Dependent Heterogenic Growth Response to Methicillin and Co-Clustering of Clinical MRSA and MSSA Isolates in Relation to Their Growth Dynamics in the Presence of Methicillin

Hierarchical clustering based on the spectrophotometric measurement of the growth of clinical S. aureus isolates in the presence of different concentration of methicillin, revealed two major clusters made up of four groups which showed heterogeneity in their response to methicillin Fig. (2A). Moreover, MRSA and MSSA isolates co-clustered (Fig. 2B); Groups 3 and 4) in relation to their growth trend. Assessment of the individual MRSA and MSSA isolates within the co-clusters showed that they had similar growth-response phenotype (Fig. 3). However, MRSA isolates exclusively made up other clusters (Fig. 2B); Groups 1 and 2). There was no correlation between the specimen type or the hospital unit where the specimen were collected (Table S6).

The phenotypic characteristics of the four groups are as follows: Group 1: Those that were resistant at all tested concentrations of methicillin and therefore showed no dose-dependent difference in the level of growth. Group 2: Those that showed typical dose-dependent growth reduction in the presence of methicillin. Group 3: Those that appeared to display a growth trend that was directly proportional to the methicillin dose increment from 1-16 µg/ml. The S. aureus MU50 ATCC 700699 was clustered in this group. Group 4: Isolates that showed similar growth trend were observed in Group 3, however, with a lower average growth.

Correlation analysis of quantitative spectrophotometric growth measurements and the distribution of MDEs in the 24 S. aureus isolates revealed no specific correlation between the distribution pattern of MDEs and the concentrations of methicillin tested in this study (Data not shown).

DISCUSSION

Staphylococcus aureus-associated diseases remain a threat to human and animal health. Methicillin-resistant Staphylococcus aureus (MRSA) across the world often leave clinicians with little or no choice of treatment options [9, 10]. As the genetic diverity of this pathogen increases [11, 12] and fewer new antibiotics are entering into clinical use, efforts are being renewed to boost the activity of old antimicrobials to which pathogens have developed resistance. These efforts include the implementation of antibiotic stewardship which includes strategic use of antibiotics [13] and the development of antibiotic adjuvants such as efflux pumps inhibitors to counter resistance and/or enhance the activity of antibiotics in clinical use [14].

To contribute to this effort, this study revealed three major clusters of MDE distribution patterns in the clinical S. aureus isolates obtained from a hospital in South Africa suggesting that all members of the same bacterial species may not contain the same MDEs as previously suggested [15].

The prevalence of qacA/B gene in this study (35.08%) is higher compared to that reported in a Japanese hospital (23.6%) (16). Moreover, smr gene which has a lower activity compared to qacA/B gene (2) was present in 43.3% of the isolates and higher in prevalence compared to that of the Japanese study (4.2%) [16].

It has been shown that most efflux pumps inhibitors have been focused on norA [14]. The observations in this study indicate the importance of expanding this effort to cover other MDEs. Moreover, the detection of norA, norB norC, mepA and sdrM in all the isolates corroborates with the focus on these important MDE efflux pumps and suggest the potential for resistance to flouroquinolones incuding norfloxacin, ciprofloxacin and sparfloxacin [2, 14]. Our laboratory is looking for the possible correlation between the presence of these genes and resistance to flouroquinolones.

The high prevalence of mdeA, sepA, and mepA is particularly worrisome and suggests a high risk of resistance to a wide range of commonly used antiseptics, biocides and antibiotics of different chemical class [2]. Similarly, tet38 which is involved in the extrusion of tetracycline [2] was absent only in one MRSA isolate indicating a high prevalence of this gene in the S. aureus isolated from patients attending the health care facility.

Interestingly, we found an isolate (TA162) which harboured only qacA/B and smr (Table S3). Repeated PCR assays indicated a consistent result for this isolate. It is not known whether this was due to mutations in the genes that prevented primer recognition of target sites in this isolate or this isolate simply evolved differently from the other isolates tested in this study. Further assessment of this isolate is the next step to be followed in our laboratory. Moreover, this study showed the co-existence and co-absence of some MDEs in the S. aureus strains suggesting a possible co-evolution of these genes in the pathogens. It is not certain if the co-absence of norC, sdrM, smr and qacA/B in the MSSA strains plays a role in susceptibility to antibiotics compared to the MRSA.

Although it is not known what factor drives the different distribution of efflux genes in the S. aureus isolates in this study, it is thought that antibiotic resistance is a natural ecological phenomenon resulting from years of evolutionary process that may include a combination of different mechanisms [17].

It has been suggested that MDEs may also play a role in virulence of bacteria as these factors assist in colonization of the host tissue by removing host bile salts and antimicrobial peptides that may prevent colonization process [18-20]. The finding of different distribution patterns in individual isolates begs the question of ‘what roles do the different distribution patterns have in these isolates? Pearsons correlation analysis suggested that none of the combinations of MDEs evaluated in this study had an obvious effect on the β-lactam, methicilin. However, the effect of individual MDE on the resistance pattern could not be assessed using this experimental design. No correlation was observed between the MDEs distribution pattern in an isolate and SCCmec type (Data not shown).

By evaluating the visual MIC, SC and the growth dynamics in the presence of methicilin, the resistance phenotype of the clinical MRSA and MSSA isolates could be determined. Importantly, some MRSA and MSSA isolates demonstrated similar susceptibility even at 0.5 µg/ml methicilin, or resistance at concentration as high as 16 µg/ml (Fig. 2). This suggest that despite the lack of mecA genes in MSSA isolates, phenotypic resistance characteristics should not be overlooked in clinical settings.

Similarly, phenotypic assessment of the resistance mechanism in the clinical S. aureus isolates using spectrophotometric measurements of the growth in TSB media, showed two major clusters consisting of four groups of specific growth-response pattern showing the resistance to the model β lactam antibiotic, methicillin. The clustering of the well characterized S. aureus MU50 ATCC 799600 to Group 2 which showed a dose-dependent response in contrast to other groups (Fig. 2A) clearly showed that there exist phenotypic differences in the resistance pattern of clinial MRSA to methicilin and perhaps other β lactams with similar physicochemical properties. This quantitative analysis proved valuable in being able to reveal the growth dynamics of the isolates evaluated and showed that a group of isolates (Fig. 2B); Groups 3 and 4) actually had higher growth as the methicillin concentration increased. Another instance of the advantage provided by spectrophotometric bacterial growth measurement can be seen in Group 3 (Fig. 2A) in which isolates SA11, SA28, TA2, SA12, SA19 and TA41 showed increasing optical densities in response to higher concentrations (4-16 µg/ml) of methicillin, despite the fact that the MSSA isolates SA11, SA28 and SA12 were classified as susceptible at this concentrations (Table 2) based on the visual MIC assessment criteria for describing susceptibility as indicated in the method section. Interestingly, these isolates co-clustered with MRSA isolates TA41 and TA12 which displayed resistance at these concentrations thus indicating that the co-clustering was based mainly on the growth dynamics and independent of the genotype. These observations are contrary to reported hormesis effect where low doses of antibiotics are associated with increased bacterial growth [21, 22].

Subsistence of environmental bacteria on antibiotics has been previously reported [23]. However, to our knowledge, this event has not been reported in clinical S. aureus isolates. Although it is not known if this observation was a result of direct utilization of methicillin as a carbon source or a trigger of other mechanisms that promote the bacterial growth, this observation represents a potential S. aureus resistance mechanism to methicilin and perhaps other β lactam antibiotics with similar structural and physicochemical properties. Theoretically, the clinical importance of these findings is that patients infected with these set of MSSA isolates may likely be prescribed/administered with antibiotic regimen that may not be able to kill the pathogen.

In conclusion, this study describes the diversity of MDEs in a hospital in South Africa and shows that there were at least three major distribution patterns of MDEs in the clinical isolates evaluated. This information is important for the development of antibiotic adjuvants that target the inhibition of MDE in S. aureus. In addition, we report that S. aureus isolates, be it of MRSA or MSSA genotype, do not contain the same set of MDEs despite being recovered from the same hospital setting. The high prevalence of MDEs suggestive of resistance to a broad spectrum of biocides and fluoroquinolones is particularly disturbing. A limitation of this study is the lack of CFU/mL data which would have shown the actual log increase or decrease in bacterial growth. However, a previous report demonstrated a strong correlation between optical density (absorption) measurement of S. aureus growth and CFU/mL [24-30]. Finally, at least four specific growth-response patterns of S. aureus in the presence of a model β lactam antibiotic (methicillin) were observed and included the enhancement of S. aureus growth in the presence of the antimicrobial agent.

REFERENCES