Unraveling Enterococcus susceptibility to quaternary ammonium compounds: genes, phenotypes, and the impact of environmental conditions

ABSTRACT Quaternary ammonium compounds (QACs) have been extensively used in the community, healthcare facilities, and food chain, in concentrations between 20 and 30,000 mg/L. Enterococcus faecalis and Enterococcus faecium are ubiquitous in these settings and are recognized as nosocomial pathogens worldwide, but QACs’ activity against strains from diverse epidemiological and genomic backgrounds remained largely unexplored. We evaluated the role of Enterococcus isolates from different sources, years, and clonal lineages as hosts of QACs tolerance genes and their susceptibility to QACs in optimal, single-stress and cross-stress growth conditions. Only 1% of the Enterococcus isolates included in this study and 0.5% of publicly available Enterococcus genomes carried qacA/B, qacC, qacG, qacJ, qacZ, qrg, bcrABC or oqxAB genes, shared with >60 species of Bacillota, Pseudomonadota, Actinomycetota, or Spirochaetota. These genes were generally found within close proximity of antibiotics and/or metals resistance genes. The minimum inhibitory concentrations (MIC) and minimum bactericidal concentrations (MBC) of benzalkonium chloride (BC) and didecyldimethylammonium chloride ranged between 0.5 and 4 mg/L (microdilution: 37°C/20 h/pH = 7/aerobiosis) for 210 E. faecalis and E. faecium isolates (two isolates carrying qacZ). Modified growth conditions (e.g., 22°C/pH = 5) increased MICBC/MBCBC (maximum of eightfold and MBCBC = 16 mg/L) and changed bacterial growth kinetics under BC toward later stationary phases in both species, including in isolates without QACs tolerance genes. In conclusion, Enterococcus are susceptible to in-use QACs concentrations and rarely carry QACs tolerance genes. However, their potential gene exchange with different microbiota, the decreased susceptibility to QACs under specific environmental conditions, and the presence of subinhibitory QACs concentrations in various settings may contribute to the selection of particular strains and, thus, require a One Health strategy to maintain QACs effectiveness. IMPORTANCE Despite the increasing use of quaternary ammonium compounds (QACs), the susceptibility of pathogens to these antimicrobials remains largely unknown. Enterococcus faecium and Enterococcus faecalis are susceptible to in-use QACs concentrations and are not main hosts of QACs tolerance genes but participate in gene transfer pathways with diverse bacterial taxa exposed to these biocides. Moreover, QACs tolerance genes often share the same genetic contexts with antibiotics and/or metals resistance genes, raising concerns about potential co-selection events. E. faecium and E. faecalis showed increased tolerance to benzalkonium chloride under specific environmental conditions (22°C, pH = 5), suggesting that strains might be selected in settings where they occur along with subinhibitory QACs concentrations. Transcriptomic studies investigating the cellular mechanisms of Enterococcus adaptation to QACs tolerance, along with longitudinal metadata analysis of tolerant populations dynamics under the influence of diverse environmental factors, are essential and should be prioritized within a One Health strategy.

to QACs typically conduct tests under ideal standardized growth conditions.Thus, the impact of various physicochemical challenges present in diverse natural ecosystems, such as pH, temperature, and other pollutants, remains largely unknown (29).
In this study, we aimed to assess the occurrence of known transferable QACs tolerance genes in Enterococcus spp.isolates and public genomes and their flow between Enterococcus and other bacterial taxa.Also, we evaluated the activity of BC and DDAC in different growth conditions, among a collection of phylogenetically diverse E. faecalis and E. faecium from different sources, time spans, and antibiotic resistance profiles.

Enterococcus carry broad-host QACs tolerance genes widely spread across different bacterial taxa
The presence of QACs tolerance genes was investigated in Enterococcus from diverse epidemiological backgrounds, including 210 Enterococcus isolates from human and non-human sources (105 E. faecium and 105 E. faecalis) and 22,428 genomes publicly available at the NCBI database.

Genetic contexts of QACs tolerance genes are diverse and enriched in antimicrobial resistance genes
The co-occurrence of QACs tolerance genes with other antimicrobial resistance genes (e.g., antibiotics, metals) in the same genetic contexts and the similarity of these QACs tolerance genetic contexts between Enterococcus and other bacterial taxa were evalu ated.
Considering the few complete genomes included in the analysis for the different taxa, namely of Enterococcus spp., most QACs tolerance genes were found on plasmids of different sizes (n = 18/19; 2-129 kb) rather than in the chromosome (n = 1/19) (Fig. S1).For the most part, the genetic contexts of QACs tolerance genes studied presented high variability among the different bacterial taxa (Fig. S1-A-G), with exception of some isolates sharing few genes.This was the case for E. faecium C132 and Staph ylococcus aureus CC1-1 or Staphylococcus warneri Ani-LG-025, sharing the qacA/B and beta-lactam resistance genes, or E. faecalis C32 and Staphylococcus capitis 15-101 or Staphylococcus massiliensis P3, sharing the qacA/B and copper tolerance genes (Fig. S1-A), among others.Enterococcus spp.from the clinical and environmental settings shared qacZ genetic contexts carrying aac(6′)-Ie-aph(2′′)-Ia, coding for aminoglycoside resistance, insertion sequences, and recombinases (Fig. S1-D).Also, qrg genetic contexts were shared between E. faecalis and Streptococcus spp., mostly presenting insertion sequences, recombinases, and hypothetical proteins (Fig. S1-E).Finally, E. faecalis from different epidemiological backgrounds shared bcrABC genes, insertion sequences, genes coding for recombinases, hypothetical proteins, or bacteriocin associated (Fig. S1-F).
Regardless of the strains' source, geographical region, or date of isolation, we observed that different genetic determinants for antibiotic resistance and metal tolerance were located adjacent to QACs tolerance genes.Of note, several of the genetic contexts analyzed (n = 32) in diverse bacteria, including Enterococcus spp., harbored genes conferring resistance to aminoglycosides, beta-lactams, macrolides, lincosa mides, streptogramin B, mupirocin, bleomycin, tetracycline, chloramphenicol/florfeni col, trimethoprim, fosfomycin, and/or sulfonamide (Fig. S1-A,B,C,D,F,G).Metal tolerance genes, namely to copper, cadmium, zinc, or arsenic, were also detected within the vicinity of qacA/B, qacC, qacJ, qrg, and bcrABC genes (Fig. S1-A,B,C,E,F).Additionally, the genetic contexts analyzed were highly enriched in insertion sequences, recombinases, and replication associated proteins.

QACs susceptibility assays of E. faecalis and E. faecium
The susceptibility of 105 E. faecalis and 105 E. faecium isolates (including two isolates carrying qacZ), with diverse epidemiological and clonal backgrounds (Table S2), to BC and DDAC was determined.
The MIC BC in standard conditions were similar for the E. faecalis and E. faecium studied (MIC 50 = 2 mg/L and MIC 90 = 2 mg/L for both) (Fig. 2).The highest MIC BC of 4 mg/L was observed in 1 E. faecalis and 9 E. faecium recovered from different sources, years, and clonal lineages.Of these, most were MDR (n = 8/10; resistant to three or more antibiotics from different families), and two of them harbored the gene qacZ (ST17 E. faecium-E241 and ST6 E. faecalis-V583).MBC BC distributions for both species also showed similar MBC 50 = 2 mg/L and MBC 90 = 4 mg/L (Fig. 2).Likewise, the 43% (n = 45/105) of E. faecalis and 19% (n = 20/105) of E. faecium isolates showing the highest MBC BC of 4 mg/L comprised, in both cases, isolates from different epidemiological and genetic backgrounds.On the contrary, E. faecium (n = 8) with the lowest MBC BC of 1 mg/L were mostly from the food chain (n = 7/8), belonging to different STs and years.The value of MIC BC and MBC BC of the control strain E. faecalis ATCC 29212 varied between 1-2 mg/L and 2-4 mg/L, respectively.
Both species were susceptible to lower concentrations of DDAC compared to BC (P ≤ 0.01) (Fig. 2).For DDAC, the MIC 50 was 1 mg/L for both species, and the MIC 90 was 1 mg/L or 2 mg/L for E. faecalis and E. faecium, respectively (Fig. 2).Strains with an MIC DDAC of 2 mg/L (n = 8 E. faecalis and n = 29 E. faecium) were diverse and comprised most of the Enterococcus with the highest MIC BC (n = 7/10, including the two isolates carrying qacZ).Finally, the MBC 50 was 2 mg/L for both species, and the MBC 90 was 4 mg/L and 2 mg/L for E. faecalis and E. faecium, respectively.Again, isolates showing the highest MBC DDAC of 4 mg/L or the lowest MBC DDAC of 1 mg/L (18%, n = 19/105, and 4%, n = 4/105, of the E. faecalis; 4%, n = 4/105, and 23%, n = 24/105, of the E. faecium populations tested, respectively) were associated with different sources, years, and STs.The MIC DDAC and MBC DDAC of the control strain E. faecalis ATCC 29212 varied between 0.5-1 mg/L and 1-2 mg/L, respectively.
Considering all sources, E. faecalis seem to be more tolerant to bactericidal concentra tions of the QACs tested (higher MBC) than E. faecium (P ≤ 0.01), although the latter presented a significantly higher MIC DDAC (P ≤ 0.01) (Fig. 2).The MIC and MBC distribu tions of the isolates tested were analyzed separately by source and time span (5-year intervals) (Fig. S2), with the following significant differences among them: the MIC DDAC and MIC BC were higher for E. faecium recovered from human infections compared to those from the food chain (P ≤ 0.01), but a significant increasing trend in the MIC BC over the years was detected in E. faecium isolates from the food chain (P ≤ 0.01) (Fig. S2).QACs susceptibility among MDR and non-MDR E. faecalis and E. faecium was similar (P > 0.01) (Fig. S3).BC and DDAC MIC and MBC values for vancomycin-or linezolid-resistant isolates varied within the ranges described for the whole population (Fig. S3). ; pH = 7.2; 37°C/20 h], followed by MBC determination, using the methodological approach proposed by the Clinical and Laboratory Standards Institute for antimicrobial susceptibility testing (33,34).The source and QACs genotype of each isolate are indicated with different colors and shapes, respectively.For more isolates' details see Table S2.Graphics were done in R, using the ggplot2 package v3.4.0 (35).

Different growth conditions occurring in the environment affect Enterococcus spp. susceptibility to QACs
To assess the activity of QACs on Enterococcus growth or survival under diverse environmental conditions that mimic real contexts, MIC BC and MBC BC were determined in anaerobic conditions (e.g., occurring in sewage) (37,38), room temperature (22°C; representing abiotic surfaces in healthcare or agri-food sectors) (39), and mildly acidic pH (pH = 5; resembling human skin and fresh fruits/vegetables contaminated with QACs residues) (40)(41)(42).BC was the QAC chosen for the susceptibility assays with modified conditions, as both species demonstrated higher tolerance and phenotype variability (wider MIC and MBC ranges) to this compound compared to DDAC under standard conditions.MIC BC and MBC BC distributions for the modified conditions studied are shown in Fig. 3.
The analysis revealed similar results between aerobic (standard) and anaerobic conditions (P > 0.01), but increased tolerance to BC at room temperature (22°C) and/or mildly acidic pH (pH = 5) conditions for E. faecalis and E. faecium strains (P ≤ 0.01) (Fig. 3).In anaerobic conditions, MIC BC or MBC BC increased by no more than twofold in a few strains (six E. faecalis and six E. faecium) belonging to diverse sources, dates, and STs.The MBC BC of qacZ + ST17 E. faecium E241 increased from 4 to 8 mg/L, outside the range obtained in standard conditions.Also, the MIC BC and MBC BC increased from 4 to 8 mg/L in a human E. faecium strain (ST412), recovered from a clinical infection in 2011, that did not contain any of the known QACs tolerance genes.
Contrastingly, four to eightfold MIC BC and MBC BC increases were observed at 22°C and/or pH = 5 conditions in diverse strains of both species.When isolates were tested at pH = 5, the MIC BC increased significantly for E. faecalis (P ≤ 0.01) but not for E. faecium (P > 0.01), although MBC BC were higher for both (P ≤ 0.01) (Fig. 3).The MIC BC of one E. faecalis increased fourfold (from 1 to 4 mg/L; urban wastewater treatment plant from Tunisia, ST23, 2014) and the MIC BC of the qacZ + E. faecalis V583 increased from 4 to 8 mg/L.The highest MBC BC obtained among the investigated E. faecalis strains was also 8 mg/L, observed in nine strains exhibiting a twofold (n = 8; from several sources and STs) or fourfold rise (n = 1; poultry carcass, ST843, 2018).In the case of E. faecium isolates, the highest MBC BC at pH = 5 was 16 mg/L, corresponding to qacZ + E. faecium E241 and one strain isolated from the faeces of a long-term care facility patient in 2016 (ST262).It is noteworthy that these strains exhibited an MBC BC of 4 mg/L under standard conditions.
An increased tolerance to BC was more pronounced at 22°C than at pH = 5 for E. faecium, reflected both in MIC BC (n = 3 isolates with fourfold and n = 2 isolates with eightfold increases at 22°C, compared to standard conditions) and MBC BC (n = 6 isolates with fourfold and n = 2 isolates with eightfold increases at 22°C) values.The highest MBC BC of 16 mg/L was detected in two food chain isolates (ready-to-eat salad and poultry carcass, ST12 and ST352, 2010-2019), while the remaining strains had an MBC BC of 8 mg/L (n = 18).Among E. faecalis, in all but one of the isolates, the MBC BC increased to 8 mg/L [following twofold (n = 17) or fourfold (n = 4) MBC BC increases] at 22°C, whereas the MIC BC did not change (P > 0.01) for most strains (n = 19).
The effect of the combination of the test conditions (growth at 22°C and pH = 5) was also studied, excluding anaerobiosis since it did not significantly influence QACs susceptibility.Indeed, higher MIC BC or MBC BC were observed for both species under 22°C and pH = 5 stress than when each condition was tested separately (P ≤ 0.01).MBC BC increased from 1 to 4 mg/L (standard conditions) to 8 mg/L at 22°C (93%; n = 21/22 E. faecalis and n = 18/20 E. faecium) and 16 mg/L (100%; n = 7/7 E. faecalis and n = 7/7 E. faecium) for the combination of 22°C and pH = 5 for most isolates, regardless of the species, epidemiological or genetic background.MIC BC also reached the highest values of 16 mg/L in four E. faecium strains from human infection and colonization, hospital sewage (E.faecium ST17 carrying qacZ), and food chain (2002-2020), and of 8 mg/L in four E. faecalis from human infection (n = 2, including an ST6 strain carrying qacZ), food chain, and environmental origins .
Two E. faecalis (V583 with qacZ; S37-25 without qacZ) and two E. faecium (E241 with qacZ; F1651 without qacZ) were included in kinetic assays, showing changes in their growth dynamics under the modified conditions studied compared with the standard ones (Fig. S4 and S5).Such modifications in the growth curves were strain specific, with E. faecalis isolates showing more similarities among each other compared to the E. faecium isolates.This was observed in the kinetic assays conducted both with subinhibitory concentrations of BC (0.5 mg/L for E. faecium F1651 and 1 mg/L for E. faecium E241, E. faecalis V583, and S37-25) as well as without BC (Fig. 4; Fig. S4 and S5).Comparing the growth kinetics of each strain under BC plus modified conditions with the standard ones (Fig. 4), different bacterial adaptations were observed, including extended lag phases and, in several cases, slower exponential growth, resulting in a delayed entry into the stationary phase.Also, for the E. faecium F1651 and the two E. faecalis, their growth with BC plus 22°C or 22°C + pH = 5 surpassed that occurring in standard conditions during the time of the assay, which may suggest that these modified conditions are better for bacterial multiplication under BC stress.
Specifically, at pH = 5 or 22°C, we described earlier that both E. faecium and E. faecalis increased their MBC BC (two to eightfold) compared to standard conditions, which may be explained by the longer exponential growth phase in such modified conditions (Fig. 4).Additionally, the highest MBC BC increase observed for E. faecium (eightfold) may be related to their exponential growth for even longer periods compared to E. faecalis, which enter much earlier into the stationary phase (<20 hours) (Fig. 4).Of note, for the combination of 22°C + pH = 5 conditions, growth kinetics for all E. faecium and E. faecalis Enterococcus faecium isolates and a combination of 22°C and pH = 5 for 7 strains of each species.For more isolates' details see Table S3.Graphics were built using the ggridges R package v0.5.4 (36).

Research Article Microbiology Spectrum
September/October 2023 Volume 11 Issue 5 10.1128/spectrum.02324-238 strains included in the analysis have shown that they were still in the exponential multiplication phase after 24 hours.This may explain the higher MBC BC observed for both species under the combination of 22°C and pH = 5 stress than for any other growth condition tested, with all isolates reaching the highest MBC BC of 16 mg/L.No clear  S3.Growth curves were determined using a Biotek Synergy HT plate reader (Marshall Scientific), with absorbance at 600 nm (A600) recorded every 20 minutes for 24 hours.Graphics were built using the Prism software v9.0 (GraphPad Software; www.graphpad.com).

Pre-exposure to BC does not increase E. faecalis and E. faecium tolerance to the biocide
In order to assess if a pre-exposure to QACs would increase QACs-tolerant phenotypes, we exposed seven E. faecalis and seven E. faecium strains to subinhibitory concentrations of BC (0.5× MIC BC ) at different contact times (3, 5, and 10 minutes), but no change in their susceptibility to BC was observed (P > 0.01).This was also the case for the E. faecium and E. faecalis isolates harboring qacZ.

DISCUSSION
This study highlights the increased tolerance of Enterococcus spp. to QACs under different environmental stresses, although QACs typical in-use concentrations remain effective in all conditions tested.Furthermore, it shows that Enterococcus spp.share known QACs tolerance genes with diverse bacterial taxa within the same ecosystems, despite the low content of QACs tolerance genes among enterococci.These findings show the importance of understanding the complex dynamics of QACs tolerance in microbial communities and provide new opportunities for further exploration in this field.
Despite the increased use of QACs in recent years (4, 5), E. faecalis and E. faecium from various epidemiological backgrounds remain susceptible to QACs concentrations lower than those recommended for disinfection or preservation, with MIC and MBC distributions for BC and DDAC being consistent with previous reports (10,20,22,27,43).The observed similar QACs phenotypes in isolates with and without qacZ gene, along with supporting published data (20,22), suggest that QACs tolerance genes may not have a significant impact on the survival of Enterococcus spp.under QACs exposure.It is also possible that the standard susceptibility determination methods did not accurately replicate the necessary conditions for the expression of QACs tolerance genotypes.The apparent lack of selective advantage conferred by QACs tolerance genes in Enterococcus seems to be also corroborated by their low prevalence among Entero coccus isolates, which could be explained by rare horizontal transfer events, lack of stability of the acquired genes or genetic contexts due to, for example, high fitness costs (11).Accordingly, literature mostly supports that Enterococcus from diverse sources do not seem to be relevant reservoirs of known QACs tolerance genes (20)(21)(22)43), with few exceptions (24,25).In addition, we found that pre-exposure to BC did not induce a higher tolerance to this biocide in isolates carrying qacZ, which may indicate that the BC concentration and exposure times tested were not able to induce the gene expression of the efflux pump encoded by qacZ, that increased expression levels of qacZ did not lead to a detectable increase in Enterococcus tolerance to BC, or that other gene expression regulatory systems or environmental inductors are needed.In a few studies, bacterial adaptation and higher tolerance to QACs have been observed following repeated exposure to sublethal/increasing concentrations of QACs (44,45), but QACs tolerance genes were either not considered or detected, making it difficult to draw conclusions about their role on the increased phenotypes.
The main source and bacterial species harboring identical QACs proteins that were found in Enterococcus spp.might reflect the respective microbial communities and settings in which specific QACs tolerance genes circulate.For example, the genes qac and qrg, identified mostly in human Enterococcus were, respectively, predominant in Staphylococcus and Streptococcus isolates mainly associated with human colonization and infection (46)(47)(48), whereas bcrABC and oqxAB, mostly found in Enterococcus spp.recovered from the food chain, were largely detected in bacterial species often occurring as food chain pathogens or hygiene indicators, such as Listeria spp.and E. coli (49)(50)(51).The genetic contexts of QACs tolerance genes studied were very diverse in Enterococ cus and other taxa, probably as a result of a high number of recombination events, as suggested by the abundance of insertion sequences and recombinases detected.They were often co-located with genes coding for resistance or tolerance to other antimicrobial agents, such as antibiotics or metals, coexisting as selective agents in many ecosystems (52)(53)(54), and highlighting the possibility of co-and cross-selection events.
However, phenotypes of reduced susceptibility to QACs have not only been attributed to the acquisition of particular genes.It has been suggested that environ mental conditions may contribute to bacterial cellular adaptations protective against the action of biocides (29).Here, when testing bacterial susceptibility to QACs under conditions such as lower temperatures than ideal for growth (e.g., room temperature, as found on many abiotic surfaces) (39) and/or mildly acidic pH (e.g., on human skin or on fruits and vegetables) (40)(41)(42), Enterococcus tolerance to BC significantly increased and growth kinetics changed, regardless of whether they contained or not known QACs tolerance genes.Environmental stresses can cause cellular modifications in both Gram-positive and Gram-negative bacteria (e.g., changes in membrane composition and fluidity, cellular metabolism), with implications on the activity of cationic compounds, such as QACs (29,(55)(56)(57).Although the effects of these stresses on Enterococcus spp.susceptibility to QACs are still rather unexplored, in other Bacillota, such as Listeria spp. or S. aureus, studies have shown that adaptation to acid pH, lower temperature, and anaerobiosis resulted in strains more tolerant to different QACs (58)(59)(60)(61).Also, similar findings have been reported for Gram-negative bacteria (55,62).
While the more tolerant phenotypes detected in this study at pH = 5 and/or 22°C were found to be below the in-use QACs concentrations, the potential implications of environmental stress on biocide efficacy emphasize the importance of conducting susceptibility tests under conditions mimicking those occurring in settings where Enterococcus are exposed to QACs.Also, QACs concentrations within or below the MIC and MBC ranges obtained in this study have been detected in wastewater and surface waters (less than 0.001 to 6 mg/L) as well as in various types of food, including fruits and nuts, meat, or dairy products (up to 14.4 mg/kg) (6,7,42), and might promote the selection and persistence of particular strains, possibly increased by environmental factors.Moreover, exposure to sublethal concentrations of disinfectants, including QACs, has been associated with an increased horizontal gene transfer via conjugation by upregulating the SOS response, enhancing the membrane permeability and production of reactive oxygen species (63)(64)(65).
In conclusion, this study provides a comprehensive phenotypic and genomic analysis of the E. faecalis and E. faecium tolerance to QACs.Although Enterococcus do not seem to be significant hosts of QACs tolerance genes, they can adapt to modified growth conditions occurring in the environment toward more tolerant phenotypes.Further studies, including whole transcriptome analyses, are needed to understand the expression of known QACs tolerance genes and the cellular mechanisms influencing Enterococcus adaptation to QACs.Continuous surveillance, namely through longitudinal metadata studies considering the influence of environmental interventions and local physicochemical factors on the dynamics of the microbiota composition in different settings, will also be crucial to assess the selection of tolerant Enterococcus populations exposed to varying concentrations of QACs.Such research can offer valuable insights for One Health intervention strategies aimed at preventing future biocide inefficacy, with significant implications for Public Health.
Additionally, the frequency and distribution of Enterococcus strains with QACs tolerance genes were also assessed among Enterococcus genomes (n = 22,428, until 28 April 2022; https://www.ncbi.nlm.nih.gov/datasets/genome/?taxon=1350) and proteins from other collections available at the NCBI database.Amino acid sequences with ≥90% of query-cover and ≥97% identity were considered.The source of isolates carrying QACs tolerance genes was retrieved from the NCBI and BV-BRC (Bacterial and Viral Bioinformat ics Resource Center) databases and their clonality determined at the PubMLST website (72)(73)(74).
Also, QACs tolerance proteins variants identified in Enterococcus were searched in other NCBI published bacterial taxa genomes (n = 1,162,672, until 28 April 2022) and proteins, in order to infer about the genetic flow of QACs tolerance genes.This was performed using the blastp tool (81), by looking at the 100% identical amino acid sequences (coverage and identity) of each of the proteins found in Enterococcus strains.The source of the isolates of all taxa carrying such QACs tolerance proteins variants was retrieved from the NCBI Biosample database (http://www.ncbi.nlm.nih.gov/biosample/), when available.

Comparative analysis of the genetic contexts of QACs tolerance genes
The synteny and nucleotide identity of the genetic contexts of QACs tolerance genes among Enterococcus and other bacterial taxa genomes were analyzed using the BLASTN option of Easyfig v2.2.2 (82) and based on NCBI genome annotations.To assess the genetic context of each QACs tolerance gene, Enterococcus strains and one representa tive species from other bacterial taxa from several sources and time spans were selected for comparison.The criteria used to select the boundaries of the analyzed sequences were the limit of the contigs in which the QACs tolerance genes were located or the presence of genetic elements of interest (e.g., antibiotic resistance, biocides or virulence genes, mobile genetic elements).

QACs susceptibility assays in standard conditions
The MIC BC (BC: CAS 68391-01-5, VWR) and MIC DDAC (DDAC: CAS 7173-51-5, Sigma Aldrich) of the 210 Enterococcus isolates were established by broth microdilution, using the methodological approach proposed by the Clinical and Laboratory Standards Institute (CLSI) for antimicrobial susceptibility testing [Mueller-Hinton broth (MHB); pH = 7.2; 37°C/20 hours] (33).A 96-well microtiter plate containing serial twofold dilutions of the disinfectant (concentration range of 0.125 to 128 mg/L) was used to assess the susceptibility of bacterial suspensions in log-phase growth (adjusted to reach a final inoculum of 5 × 10 5 CFU/mL in each well) at 37°C for 20 hours.Microdilution panels were freshly prepared before each assay.The first concentration of QAC without visible growth was considered the MIC (33).
To determine the MBC BC and MBC DDAC , 10 µL of each well without visible growth from the 96-well MIC plate was incubated onto brain heart infusion (BHI) agar plates at 37°C for 24 hours, as defined by the CLSI (34).The MBC was defined as the lowest QAC concentration for which the number of colonies was equal or less than the rejection value defined by CLSI guidelines, based on the final inoculum of each well confirmed by actual count (34).Each experiment was repeated three to six times, and the MIC/MBC values corresponded to the mean of the determinations.The E. faecalis ATCC 29212 strain (without any known QACs tolerance genes) was included as control to guarantee the reproducibility of all assays.The MIC 50 and the MIC 90 (minimum concentration of an antimicrobial agent that inhibits the growth of 50% or 90% of the bacterial population, respectively) as well as the MBC 50 and the MBC 90 (lowest concentration of an antimicro bial agent that is required to kill 50% or 90% of the bacterial population, respectively) were determined.

QACs susceptibility in modified environmental conditions
QACs assays performed in modified environmental conditions (anaerobiosis, pH = 5, 22°C, or pH = 5 + 22°C) were performed for 20 E. faecium and 22 E. faecalis (Table S3).The bacterial strains selected were representative of the MIC and MBC distribution ranges in standard conditions, different geographical regions, sources, time spans, antimicro bial resistance profiles, and genomic backgrounds of the initial collections.Modified conditions were initially tested separately (anaerobiosis, pH = 5, 22°C), and subsequently, MIC BC and MBC BC were determined with cross-environmental conditions (pH = 5 + 22°C) for seven E. faecium and seven E. faecalis (Table S3), based on the results obtained from individual assays.From these 14 isolates, two E. faecium isolates (E241 with qacZ from hospital sewage and F1651 without qacZ from poultry meat, belonging to ST17 and ST12, respectively) and two E. faecalis isolates (V583 with qacZ from clinical origin and S37-25 without qacZ from a ready-to-eat salad, identified as ST6 and ST309) were selected for kinetic assays.These assays, which provide more quantitative information compared to the endpoint measurements used in MIC and MBC determinations, were performed under different growth conditions including standard conditions (37°C; pH = 7.2), pH = 5, 22°C, and pH = 5 + 22°C.BC concentrations tested corresponded to the 0.5× MIC BC of the lowest MIC BC obtained in the three replicates in standard conditions for each strain.Growth curves were determined using a Biotek Synergy HT plate reader (Marshall Scientific), with absorbance at 600 nm (A600) recorded every 20 minutes for 24 hours.Graphics were built using the Prism software v9.0 (GraphPad Software; www.graphpad.com).

QACs susceptibility after bacteria pre-exposure to BC
The seven E. faecalis and seven E. faecium strains previously selected for cross-stress testing (including qacZ + ST17 E. faecium E241 and ST6 E. faecalis V583) (Table S3) were pre-exposed to BC.The protocol was designed based on the European standards EN 1276 (83) and EN 13727 (84) and a methodology previously described by Skive et al. (85).Briefly, MHB bacterial suspensions in log-phase of growth were split into two glass tubes and exposed to either 0.5× MIC BC or to only MHB as a control.The cultures were left in the presence of BC for 3, 5, or 10 minutes at room temperature to determine the effect of usually recommended disinfectant contact times (17,86,87).Following exposure, 1 mL of the suspension was added to 9 mL of neutralizer solution prepared according to EN 13727 (84) containing Tween 80 (30 g/L; CAS: 9005-65-6, Sigma Aldrich), sodium dodecyl sulfate (4 g/L; CAS: 151-21-3, VWR), and lecithin (3 g/L; MP Biomedicals).
Cells were centrifuged at 2,500 × g for 10 minutes.Supernatant was removed and cells were resuspended in 3 mL of PBS (phosphatebuffered saline).Cells were centrifuged at 2,500 × g for 10 minutes, and supernatant was removed from the pelleted cells.The washed cells were resuspended in 2 mL of NaCl 0.9% and volume adjusted to reach a final inoculum of 5 × 10 5 CFU/mL in each well.MIC BC and MBC BC were then determined as described above, in standard conditions, following the CLSI guidelines (33,34).
The effectiveness and toxicity of the neutralizer were evaluated according to the protocol described by Vali et al. (88).First, neutralizer toxicity was tested by adding 1 mL of neutralizer to a bacterial suspension and left in contact for 5 minutes.Cells were serially diluted to reach a final inoculum of 5 × 10 5 CFU/mL and incubated in Mueller-Hinton agar plates at 37°C for 24 hours.The number of CFU/mL was compared to a control with NaCl 0.9% replacing the neutralizer.To confirm BC was being quenched effectively by the neutralizer, a bacterial suspension was split into three glass tubes and, in each, was added a mixture containing 1 mL of biocide (at bactericidal concentration) and 9 mL of neutralizer, a mixture of 1 mL biocide (at bactericidal concentration) and 9 mL NaCl 0.9%, or 10 mL NaCl 0.9%.After 5 minutes of contact time, cells were diluted and incubated as previously, and the number of CFU/mL was compared.The results for the effectiveness and toxicity test of the neutralizer are shown in Table S4.

Statistical analysis
The statistical significance of the differences between MIC and MBC distributions of isolates from the diverse sources, time spans, and with disparate antibiotic resistance profiles in standard conditions, as well as differences in MIC BC and MBC BC between standard and modified conditions, were assessed using the Wilcoxon Signed Rank and Mann-Whitney Wilcoxon tests (R Statistical Software, v4.1.2) (89).The statistical analysis of potential associations between QACs tolerance genes found in Enterococcus and particular bacterial taxa or sources was performed by Fisher's exact test (Prism software, v9.0, GraphPad Software; www.graphpad.com).P values ≤0.01 were considered significant.

FIG 3
FIG 3 Ridgeline chart depicting the distribution of benzalkonium chloride minimum inhibitory concentrations (MIC) and minimum bactericidal concentrations (MBC) determined in standard or modified conditions.The non-standard conditions tested were anaerobiosis, room temperature (22°C), or mildly acidic pH (pH = 5) for 22 Enterococcus faecalis and 20

FIG 4
FIG 4 Growth kinetics of two Enterococcus faecium and two Enterococcus faecalis in the presence of subinhibitory concentrations of benzalkonium chloride [1 mg/L for E. faecium E241 with qacZ (A), E. faecalis V583 with qacZ (C), and E. faecalis S37-25 without qacZ (D) and 0.5 mg/L for E. faecium F1651 without qacZ (B)] in standard (light-yellow circles) or modified conditions.The non-standard conditions tested were 22°C (light pink squares), mildly acidic pH = 5 (brown triangles) and a combination of 22°C and pH = 5 (inverted green triangles).For more isolates' details see TableS3.Growth curves were determined using a Biotek