Background

S. lugdunensis is a particularly virulent species of coagulase-negative staphylococci (CoNS) [1]. This microorganism is a skin commensal but is also an opportunistic pathogen able to cause potentially severe community-acquired infections as well as healthcare-associated infections [2, 3].

Several molecular typing methods, including multilocus sequence typing (MLST) and fbl-typing, revealed a clonal population structure, with no phylogenetic lineage associated with invasive infections [4, 5]. To date, seven clonal complexes (CCs) have been described (http://bigsdb.web.pasteur.fr/staphlugdunensis/). Strains isolated worldwide from hospitalized patients mainly belong to CC1 and CC3 [6,7,8]. Even if some methicillin resistant clones are circulating in Asia [8,9,10], this species currently remains highly sensitive to antibiotics [3]. Predominance of some CCs could then be explained by the low genetic diversity of the species [11] or by an increased ability of these CCs to survive in the hospital environment by biofilm formation [1] and/or resistance to antiseptics.

Two antiseptics, chlorhexidine digluconate (CHX) and benzalkonium chloride (BAC), are widely used in health care settings including skin and wound antisepsis [12, 13]. Antiseptic resistance of Staphylococcus aureus and CoNS is mediated by efflux pumps: NorA and QacA belonging to the major facilitator superfamily, and QacC, QacG, QacH and QacJ belonging to the multidrug resistant transporter family [14,15,16]. This resistance is phenotypically characterized by high minimum inhibitory concentrations (MICs) to antiseptics, even if there is no consensual definition [17, 18]. Antiseptic resistance can be associated with the presence of resistance genes, which can belong to the core genome such as the norA gene in S. aureus [19], or be encoded by a plasmid like qac genes, identified in S. aureus and various species of CoNS [14].

So far, little is known about S. lugdunensis resistance to antiseptics [20, 21]. Therefore, the aim of this study was to investigate the prevalence of resistance to the widely used antiseptics BAC and CHX by using both phenotypic and genotypic methods, in a collection of S. lugdunensis clinical isolates representing the seven CCs described to date.

Results

Antimicrobial susceptibility test results

The antibiotic susceptibility results of the 49 S. lugdunensis strains are shown in Table 1. Eleven strains (22.4%) were resistant to penicillin G. The rate of strains resistant to kanamycin, tobramycin, fusidic acid and tetracyclin was 4.1% (2/49 strains) and that of strains resistant to gentamicin, erythromycin and clindamycin was 2% (1/49 strains). One strain (2%) was resistant to methicillin (methicillin-resistant S. lugdunensis, MRSL). This strain was multidrug resistant and belonged to CC3.

Table 1 Antibiotic susceptibility testing for the 49 S. lugdunensis strains
Full size table

Determination of minimum inhibitory concentration and minimum bactericidal concentration

The MICs of CHX ranged from 0.5 to 2 mg/L and those of BAC from 0.25 to 8 mg/L (Table 2). The MIC50 and MIC90 were 1 mg/L and 2 mg/L respectively, for both CHX and BAC. Reduced susceptibility to CHX and BAC was observed for 12.2% (6/49) and 4.1% (2/49) of strains, respectively.

Table 2 Susceptibility testing of the 49 S. lugdunensis strains to CHX and BAC
Full size table

The minimum bactericidal concentrations (MBCs) ranged from 2 to 4 mg/L and 2 to 8 mg/L for CHX and BAC, respectively. The MBC50 and MBC90 of CHX were 4 mg/L. The MBC50 and MBC90 of BAC were 4 mg/L and 8 mg/L, respectively.

MIC and MBC distribution of CHX and BAC according to the CC is shown in Table 3. BAC MICs tended to vary according to the CC of the strains. The BAC MICs of the CC5 and singleton (belonging to any of the CC described) strains (mean of 0.55 mg/L and 0.81 mg/L, respectively) appeared to be lower than those of the other CCs. On the contrary, strains belonging to CC3 had the highest MIC values (mean of 2.18). No noticeable trend was highlighted regarding the MBC distributions of either CHX or BAC.

Table 3 MIC and MBC values of CHX and BAC according to the CC
Full size table

Detection ofnorAandqacgenes.

The norA gene was detected by PCR in all S. lugdunensis strains (Table 2). Two strains were positive for the qacA gene. Interestingly, these two strains had decreased susceptibility to both antiseptics (CHX and BAC) (Table 2) and belonged to CC3. The qacC, qacG, qacH and qacJ genes were not detected. To rule out false negative PCR in the highly variable qac genes, their absence was confirmed by blastn in the 11 strains whose genome sequences were available (SL_13, SL_29, SL_55, SL_117, SL_118, SL_122, 22FJ, 25AC, 27HJ, 33RM, 37BH).

Discussion

S. lugdunensis is recognized as a virulent species of CoNS responsible for severe infections. In hospital settings, predominant clones (CC1 and CC3) have been described [5, 6, 8], suggesting a potential adaptation to this environment. With the widespread use of antiseptics in healthcare facilities, there are concerns about antiseptic tolerance and resistance [17]. However, little is known about S. lugdunensis susceptibility to antiseptics.

Only two previous studies on antiseptic susceptibility have included a small number of S. lugdunensis strains: one in Sommer et al.’s work [22] and eight in the study of Addetia et al. [20]. Our study is the first to investigate the antiseptic susceptibility of such a large collection of S. lugdunensis strains from different clinical settings and geographical origins and representing the seven CCs described to date. Resistance was assessed by determining MIC and MBC by broth microdilution for the two commonly used antiseptics CHX and BAC. Standardized methods to determine MIC/MBC and consensus to define antiseptic resistance are missing [23]. Therefore, we applied protocols and MIC breakpoints widely used in literature for Staphylococcus spp. and defined reduced susceptibility to CHX and BAC when MIC > 1.5 mg/L and > 3 mg/L, respectively [18].

In our study, lower BAC MICs were observed (0.25 mg/L to 8 mg/L) compared to previous studies using broth microdilution. Indeed, they have shown that BAC MICs for clinical CoNS strains and S. aureus varied similarly between 0.25 mg/L and 64 mg/L [24,25,26,27,28]. The only two S. lugdunensis strains with reduced susceptibility to BAC in our study belonged to the CC3. In contrast, the CC5 and singleton ST strains tended to have lower MICs. For S. aureus, Kernberger-Fisher et al. have similarly showed that strains belonging to the major human lineages CC22 and CC5 had significantly higher CHX MIC values than the main animal lineage ST398 [29].

S. lugdunensis strains exhibited here low CHX MICs (MIC90 = 2 mg/L) with a narrow value distribution, similar to that obtained by Addetia et al. for eight S. lugdunensis strains (CHX MICs determined by broth microdilution ranging from 0.5 to 1 mg/L) [20]. In contrast, in literature, CHX MICs vary from 0.125 mg/L to 32 mg/L for other CoNS species [20, 24, 25, 28, 30], and from 0.25 mg/L to 8 mg/L for S. aureus [24, 25].

The CHX and BAC MBC determined here were close to MIC values, and remained much below the antiseptic concentration used in practice. Considering the epidemiological cut-off values (ECOFFs) defined for S. aureus by Morissey et al. (i.e. MIC = 8 mg/L and MBC > 64 mg/L for CHX, and MIC = 16 mg/L and MBC = 32 mg/L for BAC), all the 49 S. lugdunensis strains tested would be considered as wild-type [31].

In addition, resistance to CHX and BAC was assessed by a genotypic method. The norA gene was detected in all strains of our collection, suggesting that it belongs to the core genome of S. lugdunensis. Similarly, Costa et al. showed that norA is part of the core genome of S. aureus, but exists as multiple alleles [19]. Increased antiseptic resistance of S. aureus strains has been associated with NorA-mediated efflux via the overexpression of the norA gene [32, 33]. Thus, it would be interesting to compare norA expression levels of the strains in our collection.

On the contrary, the prevalence of the qacA gene was low (4.1%) in this study. This result contrasts with the variations of the qacA prevalence previously described for CoNS (42.4–62.4%) [24, 25, 34] and for S. aureus (from 10.5 to 21.8% for methicillin-susceptible strains and from 8,3 to 83.3% for methicillin-resistant strains) [18, 24, 25, 34,35,36]. Here, the qacC, qacG, qacH and qacJ genes were not detected by PCR. Due to lack of DNA conservation, some of the PCR primers described in the literature may fail to detect qac genes [14]. However, blast analysis confirmed the absence of qac gene sequences in the 11 whole genome sequences available. The prevalence of these genes in other CoNS species and in S. aureus varies greatly between studies (0–44.2% of isolates) [18, 24, 34,35,36]. The low prevalence of these plasmid-encoded qac genes observed for S. lugdunensis could be explained by the multiple barriers of this species genome that prevent horizontal gene transfer by mobile genetic elements [11, 37].

Interestingly, the two S. lugdunensis strains positive for the qacA resistance gene had reduced susceptibility for both CHX and BAC. This could suggest a relationship between qac genes and reduced susceptibility to antiseptics, as demonstrated in several studies for S. aureus and CoNs [20, 22,23,24] but to be confirmed on a larger number of strains. Whole genome studies of S. lugdunensis strains with reduced susceptibility to antiseptics could lead to identify mechanisms contributing to a reduced susceptibility to antiseptics in this species.

Cross-resistance to antibiotics and antiseptics remains controversial. Some studies have reported cross-resistance between CHX and antibiotics (e.g. cefoxitin, penicillin, ciprofloxacin, trimethoprim-sulfamethoxazole, clindamycin, tetracyclin), especially for S. epidermidis, S. warneri and S. aureus [25, 38, 39]. The only MRSL strain in our study showed reduced susceptibility to both CHX and BAC. MRSL, more prevalent in Asia [9, 10, 40], could represent a major health issue worldwide. Analysis of a larger number of MRSL strains would therefore be necessary to track such a putative link between methicillin resistance and reduced susceptibility to antiseptics.

Conclusions

This study conducted on a large collection of strains shows that, unlike other CoNS, S. lugdunensis is highly susceptible to CHX and BAC. However, the first description of a reduced susceptibility to these antiseptics in two CC3 strains, highlights a potential risk for infection control in healthcare settings.

Materials and methods

Bacterial strains

Forty-nine clinical strains of S. lugdunensis isolated from 49 patients were included in this study [see Additional file 1]. They were recovered from carriage (n = 12) or infections (n = 37) and collected in seven French regions (Strasbourg, Rouen, Tours, Nancy, Montpellier, Nantes, Bordeaux) and Sweden (one strain from Kronoberg) between 2013 and 2016. All 49 strains were identified by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry (Bruker Daltonik GmbH, Bremen, Germany). Strains were previously characterized by MLST and/or fbl-typing [5]. Forty-five strains belonged to the seven CCs described [(CC1, n = 11), (CC2, n = 5), (CC3, n = 11), (CC4, n = 4), (CC5, n = 5), (CC6, n = 5) and (CC7, n = 4)], three were singleton STs [(ST13, n = 1) and (ST28, n = 2)] and for one strain data were not obtained. S. aureus ATCC 25923 was used as a quality control strain for all the susceptibility tests [20, 41].

Antibiotic susceptibility testing

Antimicrobial susceptibility was assessed by the disk diffusion method on Mueller-Hinton agar (Bio-Rad, Marnes-la-Coquette, France) according to the European Committee on Antimicrobial Susceptibility Testing recommendations (https://www.eucast.org/) [41]. The following antibiotic disks were tested: penicillin G (1 UI), kanamycin (30 µg), tobramycin (10 µg), gentamicin (10 µg), erythromycin (15 µg), clindamycin (2 µg), quinupristin/dalfopristin (15 µg), cefoxitin (30 µg), norfloxacin (10 µg), trimethoprim/sulphamethoxazole (25 µg), fusidic acid (10 µg), rifampicin (5 µg), levofloxacin (5 µg), tetracyclin (30 µg), minocyclin (30 µg), and linezolid (10 µg) (I2A, Montpellier, France).

Minimum inhibitory concentration and minimal bactericidal concentration determination

The MIC of CHX (Sigma-Aldrich, Saint Louis, USA) and BAC (Sigma, Saint Louis, USA) was determined by broth microdilution method according to the Clinical and Laboratory Standards Institute guidelines [42]. The concentration range tested for CHX and BAC varied from 0.25 mg/L to 128 mg/L. Strains with MIC below or equal to 1 mg/L were categorized susceptible to CHX ; reduced susceptibility was defined for strains with MIC between 1.5 mg/L and 3 mg/L. Strains with MIC below or equal to 3 mg/L were considered susceptible to BAC and strains with MIC values higher than 3 mg/L exhibited reduced susceptibility [18].

For MBC determination, 10 µL of the suspension in MIC wells without visible microbial growth were subcultured onto Mueller-Hinton agar and incubated at 37 °C for 24 h. The MBC was noted as the lowest antiseptic concentration for which no growth was observed [33].

All MIC and MBC experiments were carried out in triplicate. MIC and MBC obtained twice were recorded as the final values.

Molecular detection of antiseptic resistance genes

DNA extraction was performed using the InstaGene™ Matrix kit (Bio-Rad), according to the manufacturer’s recommendations. Each strain’s DNA was screened by PCR for the presence of norA, qacA, qacC, qacG, qacH, and qacJ genes using primers presented in Table 4 [43]. PCR reaction was composed of 0.25 µL (0.50 µM) of each primer, 12.5 µL of Go Taq® G2 Green Master Mix, 5 µL of DNA and 7 µL of sterile water, for a final volume of 25 µL. PCRs were performed, using a Veriti Thermal Cycler (Applied Biosystems, Foster City, CA, USA), as follows: initial denaturation step at 94 °C for 3 min, 30 cycles of 94 °C for 1 min, 50 °C for 1 min, 72 °C for 1 min, and final extension step at 72 °C for 5 min. PCR products were visualized under UV after migration for 45 min at 110 V on 1.5% gel containing 0.005 mg/L ethidium bromide.

Table 4 List of PCR primers
Full size table

Research ofqacgenes in whole genomes.

We collected the 11 whole genomes of strains used in this study available on NCBI (accession numbers: GCA_008728755.1, GCA_008728775.1, GCA_008728795.1, GCA_008728815.1, GCA_008728715.1, GCA_008728735.1, GCA_002097035.1, GCA_002096135.1, GCA_002096155.1, GCA_002104555.1, GCA_002096075.1) and the reference sequences of the qacA, qacC, qacG, qacH and qacJ genes (NCBI accession numbers: NC_007931.1, GQ900464.1, NG_051904.1, NC_019081.1, NG_048046.1) cited in the publication by Worthing KA et al., 2018 [44]. Each qac gene reference sequence was searched for in the whole genomes of the strains using blastn.