Resistance Toward Chlorhexidine in Oral Bacteria – Is There Cause for Concern?

1The 2016 Review on Antimicrobial Resistance has predicted an alarming scenario that the number of annual deaths attributable to antimicrobial resistance will increase globally from the current 700,000 to 10 million in the year 2050. In this light, resistance toward antibiotics has attracted strong interest in the scientific as well as the medical community during the last few decades, and provided a major stimulus for establishing stewardship programs and in the search for alternative antimicrobial measures such as antimicrobial peptides, natural compounds, cold atmospheric plasma, or light-based approaches. Furthermore, the oral cavity has recently been highlighted as a potential reservoir of antibiotic resistance genes that can be transferred via horizontal gene transfer among the bacteria present in oral biofilms.

2Conversely, much less attention has been paid to the directly related problem of resistance toward the different classes of antiseptics and biocides. This is somewhat surprising as bacterial resistance toward commonly used antiseptics such as benzalkonium chloride (BAC; a quaternary ammonium compound, QAC), triclosan (TCS; a polychloro phenoxy phenol) and chlorhexidine (CHX), as well as induction of cross-resistances between these agents and a range of clinically important antibiotics, have been known for many years.

3While TCS has already been banned from its use in household wash products by the FDA in 2016 due to its risk of triggering cross-resistances and resistance toward QACs like BAC has also been highlighted in numerous reports (e.g.,; ), CHX has come into the spotlight only recently. For example, it was reported that exposure to CHX may be associated with emergence of stable resistance to the last-resort antibiotic colistin, which may be associated with mutations in the two-component regulator phoPQ. The lack of focus on CHX resistance is somewhat surprising since it was found as early as 1980 that the extensive and liberal use of CHX in the management of patients with long-term bladder catheters resulted in urinary tract infections from CHX-resistant Gram-negative bacteria (i.e., strains of Proteus mirabilis, Providencia stuartii, and Pseudomonas aeruginosa), which were also found to be multidrug-resistant. More recently, a high prevalence of reduced susceptibility toward CHX (69%) was found in microorganisms that caused central line-associated bloodstream infections in a large academic medical center with significantly higher proportion in patients who received daily CHX bathing (86% vs. 64%; p = 0.028).

4Interestingly, the use of CHX in the field of dentistry for oral biofilm control has not been considered so far in this context as a possible source for the development of resistance against CHX itself or cross-resistances against other antiseptics or antibiotics. This is in spite of the widespread use of CHX by dental clinicians as gold-standard antiseptic, e.g., for plaque control and managing gingivitis or for treatment of periodontitis. In addition, CHX is included in a wide range of oral care consumer products.

5Therefore, the objectives of this work are to review the evidence for resistance in oral bacteria toward CHX and to summarize our understanding of potential resistance mechanisms toward CHX. This review further aims to raise awareness among the dental community about the potential risk of resistance toward CHX as well as concomitant cross-resistances toward antibiotics and to propose future research directions related to CHX resistance in oral bacteria.

6Chlorhexidine is a symmetric bis-biguanide molecule comprising two chloroguanide chains that are connected by a central hexamethylene chain and carries two positive charges at physiological pH (see Figure 1A for structural formula). It acts as a strong base and is practically insoluble in water but reacts with acids to form salts that exhibit varying water solubility characteristics. Today, mostly the digluconate or diacetate salts are used, whereby the water solubility of CHX digluconate is considerably higher than that of CHX diacetate (50% w/v vs. 2% w/v). For the sake of clarity, in the present paper the acronym CHX will be used for all of its salts.

7Chlorhexidine was first synthesized in Great Britain in the early 1950s within the scope of a broad screening exercise for active agents against malaria by Imperial Chemical Industries (Manchester, United Kingdom). It was described by under the name “Hibitane®” as “a new antibacterial agent of high potency” using its diacetate and dihydrochloride salts to ensure water solubility. Since then CHX has extensively been used in various medical fields (e.g., urology, gynecology, ophthalmology, otolaryngology) as well as for disinfection of surgical environments before it was introduced to dentistry in the late 1960s.

8In general, it is often difficult to distinguish the primary mechanism of action of a given antiseptic or biocide from secondary effects that are merely a consequence of the action. CHX is usually described to act as a bacteriostatic agent at low concentrations and to be bactericidal at higher concentrations. Its antibacterial mechanism of action is described as damage of bacterial membranes and subsequent leakage of cytoplasmic components (see Figure 1B for scanning electron microscopic images of CHX-mediated damage to bacterial cell surfaces). First-line targets of CHX (at lower concentrations) are cytoplasmic membrane integrity as well as function of membrane-bound enzymes (see Figure 1C–F for a scheme describing the detailed interaction of CHX with bacterial cytoplasmic membranes), while secondary effects (at higher concentrations) are cytoplasmic leakage and, ultimately, coagulation and precipitation of intracellular constituents such as proteins and nucleic acids. As cytoplasmic membranes are the main action sites of CHX, the outer membrane in Gram-negative bacteria may act as a permeability barrier for CHX and limit its antibacterial efficacy. For instance, cationic CHX molecules may be “locked up” in the outer membrane due to interactions with negatively charged lipopolysaccharide-moieties and thus may not be able to even reach the cytoplasmic membrane. Furthermore, outer membrane vesicles released by Gram-negative bacteria like Porphyromonas gingivalis may bind CHX, thus allowing protection of bacteria. As there is not much data on the actual role or influence of the outer membrane in Gram-negative bacteria with regard to the action of CHX, this point merits closer attention.

9Studies of the effects of CHX at sub-inhibitory, bacteriostatic, concentrations are very scarce. Analogous to QACs or antibiotics at sub-inhibitory concentrations, the effects of CHX at such concentrations may be far more complicated including multiple processes such as loss of osmoregulation around the cytoplasmic membrane, disturbance of processes related to transport and respiratory activity, dissipation of proton motive force and oxidative stress, which triggers SOS responses and in turn induces error-prone DNA replication leading to mutations and gene transfers.

10Intrinsic resistance is defined as natural property of an organism, while acquired resistance is resulting from genetic changes and arising either by mutation or by the acquisition of the genetic material, e.g., via plasmids. Furthermore, one has to clearly distinguish between phenotypic adaptation, which is reversible when the exposure to the agent ends, and acquired resistance, which is genetically defined and therefore stable. Resistance must also be differentiated from tolerance, whereby resistance means the inherited ability of microorganisms to survive high concentrations of an antimicrobial drug (quantified by the minimum inhibitory concentration, MIC), while tolerance describes the ability to withstand a transient exposure to high concentrations of an antimicrobial (which otherwise would be lethal) without a change in MIC mostly due to the deceleration of metabolic processes. In contrast, persistence describes the presence of a sub-population that is able to survive a treatment with concentrations higher than the MIC of a given antimicrobial despite the population being clonal. These terms have often been used synonymously in the literature (especially in older studies), which hampers differentiation.

11Furthermore, the definition of “resistance” toward antiseptics is not entirely clear. There are internationally recognized and standardized methods for susceptibility testing of antibiotics with defined breakpoints by means of which given isolates can be classified as resistant, intermediate or susceptible. This practically means that given isolates whose MIC values exceed the breakpoint concentrations are deemed resistant to a given antibiotic. In contrast, similar frameworks are missing for susceptibility testing of biocides like CHX and there is a lack of well-defined MIC cut-off values indicating resistance. While in vitro methods originally developed for systemic antibiotics (e.g., MICs) are still used for antiseptics, their results must be interpreted with caution. In this context, measurable MIC increases by a given factor (often 4- to 16-fold) have been considered relevant and may be seen as a parallel definition of “resistance”, while on the other hand “in-use” concentrations of these biocides may be much higher than the measured MICs. Despite these limitations, investigation of MICs can still be valuable for antiseptics for studying potential resistance mechanisms in vitro.

12Intrinsic resistance to CHX is known from bacterial spores and mycobacteria and is due to their outer cell layers which form an impermeable barrier to the ingress of CHX molecules. Acquired genetically defined mechanisms conferring resistance toward CHX include multidrug efflux pumps and cell membrane changes. The latter have been suspected to be responsible for CHX resistance in Pseudomonas stutzeri, where changes in the outer membrane protein and lipopolysaccharide profiles were found in CHX-resistant strains as compared to CHX-sensitive ones. Accordingly, electron-microscopic investigations confirmed that treatment with CHX led to considerably greater morphological changes in CHX-sensitive cells, while CHX-resistant cells showed no structural damage. Furthermore, it was confirmed by energy-dispersive spectroscopy that there was less uptake of CHX in resistant cells. The CHX-resistant strains also showed decreased susceptibility toward different classes of antibiotics which was suggested to be due to a non-specific decrease in cell permeability limiting uptake of chemically unrelated molecules into the resistant cells.

13Multidrug efflux pumps are membrane proteins that contain multiple transmembrane domains forming channels to remove toxic substances from the cytoplasm and the cytoplasmic membrane (see Figure 2 for schematic depiction of two exemplary efflux systems conferring CHX resistance in Gram-positive or Gram-negative bacteria, respectively). For instance, the plasmid-borne qac (standing for “quaternary ammonium compound”) gene family (e.g., qacA, qacB, qacC; qacA, and qacB usually recorded as qacA/B and qacC synonymously termed smr) is well known in Gram-positive bacteria (mainly in staphylococci). These qac genes encode for Qac efflux proteins that belong to the “Major Facilitator Superfamily” (MFS, i.e., QacA/B) or to the “Small Multidrug Resistance” (SMR) family (i.e., Smr) and have cationic biocides like CHX as substrates. Accordingly, introduction of a CHX-based surface antiseptic protocol in an intensive care unit with methicillin-resistant Staphylococcus aureus (MRSA) prevalence of about 20% was found to lead to an immediate and sustained reduction in transmission of susceptible MRSA strains while strains carrying qacA/B genes were not affected after introduction of this protocol. In contrast, a recent study found no correlation in S. aureus isolates between susceptibility toward CHX and carriage of qacA/B genes. Interestingly, when comparing two pairs of isolates with fourfold differences in susceptibility to CHX (as shown by minimum bactericidal concentrations, MBCs) but very close relation according to phylogeny data, single-nucleotide polymorphisms within chromosomal efflux systems encoded by norA or norB were identified indicating that norA/norB may functionally complement qacA/qacB in these strains.

14In Gram-negative bacteria, efflux proteins from the SMR family (e.g., QacE, QacEΔ1, QacF, QacG) and from the “Resistance-Nodulation-Division” (RND) superfamily (like the Mex efflux systems in P. aeruginosa) have been described. For instance, the activity of the MexCD-OprJ multidrug efflux system was found to be upregulated in P. aeruginosa upon exposure to sub-inhibitory concentrations of CHX or BAC. Expression of this efflux gene was found in various mutants, suggesting that MexCD-OprJ is a determinant of CHX resistance. Furthermore, the four RND superfamily efflux pumps MexAB-OprM, MexCD-OprJ, MexEF-OprN, and MexXY-OprM are also well recognized to confer resistance to fluoroquinolones. Likewise, the “Proteobacterial Antimicrobial Compound Efflux” (PACE) family is wide-spread in Gram-negative bacteria conferring resistance to a wide range of biocides whose first member, the active CHX efflux protein AceI, was identified by investigating the transcriptomic response in Acinetobacter baumannii toward CHX. In the same study, it was also found that expression of the RND efflux protein AdeAB was even more up-regulated upon CHX exposure.

15The close association frequently found between antibiotic and antiseptic resistance may be explained by the fact that genetic determinants of resistance are commonly linked to each other; accordingly, the qac genes are frequently located on plasmids with various other resistance genes. For example, a clinical strain of S. aureus with high-level resistance to vancomycin isolated in June 2002 harbored a 57.9 kb multi-resistance conjugative plasmid, pLW1043, comprising elements that encode for resistance toward vancomycin (vanA), trimethoprim (dfrA), β-lactams (blaZ), aminoglycosides (aacA-aphD) as well as antiseptics (qacC). Likewise, the plasmid pSAJ1 from a methicillin- and gentamicin-resistant strain of S. aureus isolated in the 1980s conferred resistance toward CHX (most likely due to qacA) along with resistance toward kanamycin, gentamicin, tobramycin, amikacin, BAC, acriflavine, and ethidium bromide.

16Recently, it was found that adaptation of clinical Klebsiella pneumoniae isolates to CHX in vitro can lead not only to stable resistance toward CHX but also cross-resistance toward the last line antibiotic colistin. Whole genome sequencing revealed mutations in phoPQ and/or smvR, whereby the smvR mutation in turn may lead to upregulation of smvA which encodes for expression of SmvA, an efflux pump belonging to the MFS, which was suggested to play a major role in CHX resistance of K. pneumoniae strains. The cross-resistance toward colistin upon adaptation to CHX was likely due to the mutation in PhoPQ affecting the regulatory targets pmrD and pmrK. The operon pmrK functions to alter lipopolysaccharides by replacement of phosphate groups by 4-amino-4-deoxy-L-arabinose, which in turn reduces the net negative charge of lipid A and results in a reduction of its binding affinity to colistin.

17Chlorhexidine has been extensively used in dental practice since 1970, when described total inhibition of plaque formation and gingivitis development in patients applying a 0.2% CHX mouth rinse twice daily despite stopping all other oral hygiene measures (e.g., tooth brushing) (see ). Already in 1972 two studies were published reporting clinical isolates of Streptococcus sanguinis that showed slightly reduced susceptibility toward CHX after long-term use of CHX-containing mouth rinses, which was, however, considered to be “relatively inconspicuous”. Isolates of S. sanguinis with reduced susceptibility to CHX were also found in a later clinical study upon daily tooth brushing with 0.5% CHX-containing gel, whereby the authors concluded that it was unclear whether this finding was a result of an adaptation toward CHX or due to a selection for less sensitive mutants within the original oral microbiota.

18described adaptation to CHX in three CHX-sensitive strains of S. sanguinis when grown in vitro as continuous cultures in a fermenter containing medium with increasing CHX concentrations. This resistance persisted after continuous growth in CHX-free medium and, interestingly, extracted DNA from the resistant mutants transformed competent sensitive S. sanguinis strains to increased CHX resistance. Therefore, they concluded that appearance of less susceptible isolates in the oral cavity after long-term use of CHX may be explained by genetic changes and not solely by selection for naturally occurring less susceptible strains. Likewise, when screening 315 isolates from subgingival plaque for their susceptibility to a mouth rinse containing 0.2% CHX, evidence for what the authors called “relative resistance” in different strains of Streptococcus mitis, S. sanguinis, and Capnocytophaga spp., was found. reported significantly higher MICs for CHX in bacteria recovered from patients that had been using a 1% CHX toothpaste for 6 months as compared to those using a toothpaste without CHX. However, this finding was not thought to be of clinical significance.

19Passaging 24 periodontitis-associated bacteria (i.e., 10 oral streptococci, 8 P. gingivalis strains, 4 Aggregatibacter actinomycetemcomitans strains, and 2 enterobacteria) on agar plates containing sub-inhibitory concentrations of CHX resulted in a transitory moderate increase in the tolerance to CHX in five of the tested isolates (i.e., 1 oral streptococcus strain, 3 P. gingivalis strains, and 1 A. actinomycetemcomitans strain) after 25 passages, which vanished again after 50 passages in spite of the agar still containing sub-inhibitory concentrations of CHX. reported two- to fourfold increased MICs in two out of five P. gingivalis strains after culturing for 20–30 passages in sub-inhibitory concentrations of CHX, whereas no increase in MIC was observed for two Streptococcus mutans and two Streptococcus sobrinus strains.

20Recently, found that repeated exposure of Enterococcus faecalis to CHX by serial passaging (10 cycles of exposure to CHX for MIC testing followed by re-growth in CHX-free medium) resulted in continuous increase in its MIC. The adapted cells showed increased surface hydrophobicity. Furthermore, SDS-PAGE of the adapted mutant strain revealed a novel protein of about 19-kDa that had also been found in vancomycin-resistant enterococci. The authors searched the proteome of E. faecalis 62 in the NCBI database for a candidate protein and reported that the most relevant protein in the 19-kDa range with regard to resistance was a 19.93-kDa MFS efflux pump protein. This protein is unusually small for an MFS transporter and it is not clear whether it directly exports vancomycin. However, it is noteworthy that the candidate protein is a partial match to a larger 45-kDa MFS protein of E. faecalis V583, encoded by ef2068.

21evaluated changes in MIC in eight common oral bacterial species over 10 passages of CHX-challenge and re-growth in CHX-free medium and reported adaptation in Streptococcus gordonii, E. faecalis, Fusobacterium nucleatum, and P. gingivalis. Further analysis of the adapted S. gordonii strain showed a delayed growth with prolonged log-phase and decelerated growth rate as compared to its parental strain indicating reduced metabolic activity that may be responsible for the reduced susceptibility toward CHX.

22In another recent study, bacteria were isolated from dental plaque from five healthy individuals to screen for strains resistant toward CHX. CHX-resistant isolates also displayed variable resistance to a range of antibiotics including ampicillin, kanamycin, gentamicin and tetracycline. Exposure of the most resistant isolate, a strain identified as Chryseobacterium indologenes, to CHX (16 μg/mL) resulted in a 19-fold up-regulation of the gene CIN01S_RS05745 encoding expression of the HlyD-like periplasmic adaptor protein of a tripartite efflux pump. However, it was not investigated whether the other components of this efflux pump were upregulated as well. The authors highlighted the requirement for increased vigilance for the presence of multidrug resistant bacteria within dental plaque and raised awareness of the potential risk of long-term use of oral care products containing antimicrobial agents such as CHX for the control of oral biofilms.

23Interestingly, treatment with CHX has also been shown to induce the formation of antifungal-tolerant persister cells in Candida albicans biofilms in vitro. In another study, 150 isolates of C. albicans and Candida glabrata obtained from cancer patients, who were at high risk for the development of oral candidiasis and who had been treated with topical CHX once a day, were investigated in terms of the development of persisters. The authors found that persister cells are clinically relevant and that antimicrobial therapy with CHX and long-term carriage of Candida select for high-persister strains in vivo. There are also reports of the induction of S. mutans persisters by CHX in single-species biofilms.

24The increased clinical use of CHX makes it important to be alert to the possibility of the emergence of new clones with reduced susceptibility. It seems reasonable that microorganisms will be exposed to sub-inhibitory concentrations of antiseptics in a clinical environment. Such sub-inhibitory concentrations can be reached in the oral cavity upon treatment with CHX, as the antimicrobial effect of CHX seems to be diminished in the presence of organic substance such as saliva or serum due to inactivation by salivary or serum proteins. Also inside oral biofilms, treatment with CHX results in a concentration gradient from the biofilm surface toward its lower strata which may lead to biofilm layers with sub-inhibitory concentrations of CHX. Accordingly, the limited antimicrobial effects of CHX in the inner layers of oral biofilms have been shown using confocal laser scanning microscopy in combination with LIVE/DEAD staining (see Figure 3). Furthermore, starvation may play a major role in these biofilms, which may increase adaptation of the bacteria within the biofilm. It was shown that S. mutans exhibited significantly increased MBCs toward CHX under starvation conditions in vitro than in nutrient-rich medium.

25The question, whether there already is reason for concern with regard to enhanced tolerance or even resistance toward CHX in oral bacteria, cannot be answered that easily and must be treated in a more differentiated way: On the one hand, CHX is still very effective in its various fields of clinical application, e.g., intensive care, hand soaps and as an oral antiseptic. On the other hand, there unquestionably are reports about emergence of isolates exhibiting enhanced tolerance toward CHX. True resistance toward CHX will have horrendous consequences for infection control. For example, recently reported the spread of a unique USA300 clone of community-acquired MRSA in an Orthodox Jewish Community in Brooklyn. Genetic analyses showed that a metabolic change (mutation in pyrA) and acquisition of a clone-specific prophage (Φ11) probably have primed the clonal variant for success by promoting colonization and abscess formation. However, it was also found that emergence of a dominant clone coincided with acquisition and evolution of a plasmid (pBSRC1) with genes conferring resistance to CHX (qacA/B) and mupirocin (mupA), strongly suggesting that resistance toward antimicrobials used for decolonization therapy like CHX and mupirocin were key elements in the spread of this clone. Moreover, low level exposure to CHX (as potentially occurring in deeper layers of oral biofilms) may result in development of cross-resistances toward antibiotics as it has already been described in biocide-sensitive strains from organic foods. As there currently is little awareness about the potential risks accompanying the widespread use of CHX in dentistry, this research subject should be particularly highlighted as a hot topic in dental research. From the authors’ point of view, there are two key questions relating to CHX resistance that urgently need to be addressed:

26Does the widespread use of CHX lead to an enrichment of resistant strains in oral biofilms and does it further encourage the development of cross-resistances in oral biofilms?

27What are the molecular mechanisms conferring CHX resistance in oral bacteria?

28The emergence of resistant bacteria as a result of the widespread use of CHX in dental practice has not been studied systematically so far. Culturing biofilms in oral in situ biofilm models and treating them with CHX sequentially would give insights into ecological changes of the microbiota of the oral biofilm after long-term treatment with this antimicrobial agent. In this light, it will be important to analyze the impact of CHX by microbiome analyses by high-throughput sequencing methods as well as by culture techniques that can distinguish between viable and non-viable cells. In addition, subsequent testing of antibiotic resistance in representative oral isolates may facilitate the study of potential effects of CHX on cross-resistance toward antibiotics and emergence of multidrug resistant bacteria. What remains unclear is the clinical relevance of persister cells in oral biofilms which have frequently been treated with CHX. The research on persister development in oral bacteria under the pressure of CHX treatment is still in its early stage.

29Furthermore, novel techniques like transposon sequencing (Tn-seq) are now available that can be used to identify key genetic mechanisms of CHX resistance in oral bacteria by determining gene disruptions differentially represented in mutant populations on a genome-wide scale. For this purpose, resistant strains must either be isolated from patient samples or induced in vitro by multiple passaging of bacteria under CHX-treatment (see Figure 4), whereby the latter has the advantage that the mutant strain exhibiting phenotypic adaptation to CHX can be compared to the wildtype strain. Finally, it is important to consider the respective fitness cost of a given mutation related to CHX resistance as the magnitude of this cost is the main biological parameter determining the rate of development of resistance.

30All datasets generated for this study are included in the manuscript.

31FC and AA-A took the lead in writing the manuscript. NJ, WB, TM, and EH provided the critical feedback and reviewed the manuscript for high intellectual content.

32The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.