Susceptibility of Acinetobacter baumannii isolates from different sources to biocides: a matter of biofilm Authors: Leila Hamze, Estelle Saras, Sana Azaiez, Wejdene Mansour, Jean-Yves Madec, Thibault Destanque, Marisa Haenni, Agnese Lupo Published: 2025 ABSTRACT AbstractAcinetobacter baumannii is an opportunistic pathogen causing severe infections in patients residing in intensive care units. The ability of A. baumannii to acquire carbapenem-resistance makes these infections difficult to treat and poses a serious risk for global health. In fact, carbapenem-resistant A. baumannii isolates belonging to eleven international clones (IC1-11), especially IC2, have disseminated globally, generating outbreaks and epidemics. Persistence of A. baumannii in nosocomial settings is favored by its ability to produce biofilm and resist to biocides. However, investigations analyzing the efficacy of biocides by determining the minimum bactericidal concentration (MBC) on field A. baumannii isolates are rare. We determined the MBC of five commonly used biocides (hydrogen peroxide, ethanol, sodium hypochlorite, chlorhexidine digluconate, and benzalkonium chloride) adapting the EN1040 and EN1276 standard methods on 64 A. baumannii isolates collected from human hospital settings and diseased animals (multidrug resistant) and from food (multidrug susceptible). The 64 isolates belonged to diverse sequence types, some of them being part of ICs (IC2, IC7, IC9 and IC11). Efficacy of biocides was evaluated in presence of an interfering substance (bovine serum albumin, BSA) to simulate dirtiness, but also on in vitro formed biofilms. MBC50 and MBC90 values of the five biocides on bacteria in planktonic phase were lower than concentrations used in commercial formulations. Presence of BSA increased the MBC values of sodium hypochlorite and benzalkonium chloride. Isolates presenting the highest MBC values, mostly belonged to IC2 (multidrug resistant) and other expanding clonal complexes (multidrug susceptible). On biofilm matrix, none of the tested biocides could eradicate bacteria neither demonstrated bactericidal activity at the concentrations and exposure time tested. Based on a standardized methodology, our investigation provides data that could contribute to define epidemiological cut offs to interpret bacterial response to biocides. Investigations determining the efficacy of biocides on field isolates both in planktonic phase and in biofilms matrix are necessary to understand if hygienic measure to counteract environmental contamination are adapted and unveil bacterial adaptive strategies to these substances.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13756-025-01658-8. Background Acinetobacter baumannii is an opportunistic pathogen causing infections especially in patients residing in intensive care units (ICUs) [1]. The ability of A. baumannii to acquire genes conferring resistance to carbapenems poses serious difficulties to clinicians to treat infections sustained by these microorganisms [2]. Carbapenem-resistant A. baumannii (CRAB) isolates belong essentially to eleven international clones (IC1-11) that have disseminated globally [3]. CRAB isolates characterized in diseased animals were found to be genetically close to those commonly reported in humans, suggesting that the circulation of this opportunistic pathogen is not hindered by host barrier [4]. In veterinary settings, A. baumannii isolates are able to cause local outbreaks similarly to what is observed in human hospital settings [5]. Recently, we reported the occurrence of A. baumannii isolates from retailed raw meat samples. Some of these isolates, although susceptible to antibiotics used in human medicine, belonged to clonal complex (CC) or IC also found in humans. The repeated occurrence of isolates of the same sequence type (ST) in different meat processing sites suggested the persistence of this bacterium on the surfaces exposed to raw meat [6]. A. baumannii is able to produce biofilms, facilitating its persistence on abiotic surfaces [1]. Furthermore, this life-style provides protection against environmental challenges including biocides. In general, biofilm matrix physically impedes the diffusion of biocides and chemical components of the matrix may interact with biocides, thus reducing their efficacy [7]. In literature, A. baumannii was frequently reported as able to resist to biocides commonly present in cleaning products and disinfectants [8]. Molecular investigations on isogenic strains have demonstrated that efflux mediated by Resistance Nodulation Division (RND) systems AdeABC and members of Major Facilitator Systems (MFS) AmvA are major determinants underling A. baumannii tolerance to biocides [9–12]. However, investigations on the tolerance of A. baumannii isolates from the field are rare [13, 14]. Furthermore, these investigations have been mainly based on the determination of minimum inhibitory concentration (MIC) of biocides. While this method is widely implemented in microbiology laboratories, is easy to perform and can bring valuable information on the growth inhibition activity of an antibiotic, it introduces major drawbacks to assess the efficacy of biocides to exert their bactericidal activity under standard conditions. Biocides need to be bactericidal at a defined exposure time and therefore a growth inhibition assay is not suitable. To assess such efficacy, determination of biocides’ minimum bactericidal concentration (MBC) is the fit-for-purpose method [15]. To fill in this gap of knowledge, we determined the MBC of five among the most commonly used biocides (ethanol, hydrogen peroxide, sodium hypochlorite, chlorhexidine digluconate, benzalkonium chloride) [16] on A. baumannii isolates of different biological sources and belonging to several different CCs and ICs. We used an adaptation of the methodology proposed in the EN1040 standard according to a biocide-neutralizer test, exposing bacteria to determined concentrations of biocides for a determined period, at room temperature (20 ± 1 °C). Simulation of dirtiness was achieved using an interfering substance (bovine serum albumin, BSA) (EN1276 standard). Ability of isolates to produce biofilms was assessed and biocides MBC determination was attempted on in vitro biofilm matrix as well. The expression of the adeB gene, part of the adeABC operon, and of the amvA gene were evaluated with and without exposure to chlorhexidine. Methods Isolates collection A. baumannii isolates (n = 64) analyzed in this study were collected from different sources: food, in particular raw meat (n = 52) [6], animals (n = 3) [4], human patient (n = 1), hospital personnel (n = 3), and abiotic surfaces of clinical settings (n = 4) [17]. MBC of biocides was determined for the A. baumannii ATCC19606 strain that is a known biofilm producer [18]. Susceptibility of A. baumannii isolates to biocides in planktonic phase For the 64 A. baumannii isolates, susceptibility to five biocides (sodium hypochlorite, ethanol, hydrogen peroxide, chlorhexidine, benzalkonium chloride) was evaluated in absence or in presence of bovine serum albumin (BSA, Sigma-Aldrich, France) at 0.3 g/L and 3 g/L following the European standards NF EN 1040 and NF EN 1276. Different biocides concentrations were tested (Table S1). Biocides were diluted using hard water as reported in NF EN 1040, which was composed of solution A (19.84 g/L of magnesium chloride and 46.24 g/L of calcium chloride) and solution B (35.02 g/L of sodium bicarbonate (Sigma-Aldrich, France)). Both solutions were sterilized by membrane filtration and a mix of 6 mL of solution A and 8 mL of solution B into 1 L of water was freshly prepared for biocides dilutions the day of testing. From a 1.5 McFarland bacterial suspension, 10 μL were exposed to 80 μL of biocide of a given concentration for 30’’ in presence of 10 μL of BSA at 0.3 g/L or 3 g/L. Afterwards, 15 μL of this mixture and 15 μL of water was transferred into 120 μL of neutralizer containing lecithin (3 g/L), sodium thiosulfate (5 g/L), Tween 80 (7.5 g/L), and saponin (7.5 g/L) (Sigma-Aldrich, Germany). After 5’ incubation at room temperature, 10 μL of this mixture was spotted onto Mueller–Hinton agar plates (Biorad, Marnes-la-Coquette, France), that were incubated for 24 h at 37 °C. The biocides’ concentration at which no bacterial growth could be observed was defined as the minimum bactericidal concentration. The Pseudomonas aeruginosa ATCC15442 strain was used as control. Before biocide testing, survival of isolates to BSA was verified as follows: from a 1.5 McFarland bacterial suspension, 100 µL was added to 100 µL of BSA (at 0, 3 g/L and 3 g/L) after 2’ of incubation at a room temperature, 800 µl of hard water was added and 50 µl of a 1:103 dilution were plated onto three MH agar plates (EasySpiral Dilute ® Interscience, France). To assess neutralizer effect on A. baumannii isolates, from a 1,5 McFarland bacterial suspension, 100 µL was added to 800 µL of neutralizer and after 5’ incubation at room temperature, 50 µL of a 1:103 dilution were plated onto three MH agar plates. Colony forming unit (CFU/mL) were enumerated after 24 h incubation at 37 °C (Scan4000, Interscience, France). To validate the neutralizer’s action, 100 µL of BSA at 0.3 g/L or 3 g/L and 800 µL of the highest biocide concentration were mixed with 100 µL of hard water. From this mix, 100 µL were incubated with 800 µL of neutralizer and after 5’ incubation at room temperature, 100 µL of 1.5 McFarland bacterial suspension was added. After 30’ at room temperature a 1:103 dilution was plated on MH agar plates and incubated at 37 °C for 24 h. The bacterial growth after incubation indicated the effectiveness of the neutralizer against the biocide. Susceptibility of A. baumannii isolates to biocides in biofilm matrix Biofilm production was quantified for all 64 A. baumannii isolates by a crystal violet assay (CV, 0.1%) according to a previously described protocol [19] using 96-well flat-bottom, polystyrene microplates (ThermoScientific, France). The optical density (OD) of the crystal violet stained biofilm was measured at 600 nm (Magellan v7.5 STD 2PC, Tecan, France). A. baumannii ATCC19606 and ATCC17978, respectively biofilm- and non-biofilm producer, were used as controls. All isolates were tested in two independent experiments, each including two technical replicates. Data were interpreted according to the specific biofilm formation threshold (SBFT) set at three standard deviations above the SBF of ATCC17978 A. baumannii strain, thus isolates were classified as follows: non-biofilm former (SBF ≤ SBFT); weak biofilm former (SBFT < SBF ≤ 2 × SBFT); moderate biofilm former (2 × SBFT < SBF ≤ 4 × SBFT); strong biofilm former (4 × SBFT < SBF) [19]. The susceptibility of the five biocides was evaluated on biofilm matrix produced by isolates that were classified as strong biofilm former. In a 96-well microplate, 100 µL of bacterial suspension (OD600 = 0.05) was incubated at 37 °C in Luria–Bertani broth (Sigma-Aldrich, France). After 24 h, the cultivation broth was removed and the wells were washed with sterile 0.9% NaCl solution to remove planktonic cells. To each well, 150 µL of three biocides concentrations one corresponding to the MBC as determined in the planktonic phase for each isolate, and the other two corresponding to minimum and maximal concentrations of usage (Table S1) were added for 30’’, then the biocide solution was removed and 150 µL of neutralizing solution was added. After 5’ incubation, the neutralizing solution was removed and wells were washed with 150 µL of 0,9% NaCl solution. Then, wells were filled with 100 µL of sterile phosphate buffer and microplates underwent sonication for 30’’ with 3 times pulses (Advantage Lab, AL04_03). Finally, 50 µL from each well were diluted into 5 mL 0,9% NaCl solution and 50 µL, or an opportune dilution, were seeded on MH agar plates for CFUs enumeration after 24 h incubation at 37 °C. Decrease of colony forming units in biofilms exposed to different biocides concentrations was analyzed with a Wilcoxon test using R v. 4.2.2. and rstatix v.0.7.2 package [20]. Data were plotted using ggplot2 v.4.0.0 package [21]. Susceptibility of bacteria in biofilm matrix to biocides was evaluated based on the viability of bacterial cell according to a colorimetric assay (5% of Thiazolyl Blue Tetrazolium Bromide, MTT, Sigma-Aldrich, France) according to manufacturer’s instructions. A calibration curve (CFU/mL versus MTT absorbance at 600 nm) was carried out using a fresh culture of the A. baumannii ATCC19606 strain. CFU/mL in biofilms exposed to biocides and analyzed by the MTT assay were calculated according to the equation obtained from the calibration curve. For each concentration, isolates were tested in three technical replicates and three biological repetitions. Analysis of biocides resistance genes The genomes of the 64 A. baumannii isolates tested in this study were previously sequenced (Bio project: PRJNA1076118 and PRJNA999198). Presence of genes associated to biocides tolerance in A. baumannii were searched by tblastn (blast v.2.5.0) [22] using the BacMet v.2.0 database. The analysis focused on genes occurring in A. baumannii (n = 20) and experimentally confirmed to be involved in biocides tolerance genes (n = 20 for A. baumannii) [23]. Analysis of adeB and amvA genes expression by RTqPCR To investigate if efflux was involved in tolerance to biocides, selected isolates from different sources (#38,208, ST25 from urine infection in a dog; #58,425, ST2 from the colonized hands of hospital personnel; #58,475, ST85 from a soap dispenser in a hospital) together with the A. baumannii ATCC 19606 strain, were analyzed for the adeB (part of the AdeABC efflux system) and amvA gene expression. The isolates were cultivated under two conditions: brain heart infusion broth (BHI, Oxoid, United Kingdom) and in BHI supplemented with 0.5 mg/L of chlorhexidine digluconate. The rpoB gene was used as normalizer [24] and values obtained from cultures not exposed to chlorhexidine were used as calibrators for determining folds expression difference according to the method proposed by Pfaffl, taking in account the efficiency of amplification [25]. Further details, oligonucleotides used for the assay and amplification conditions are reported in supplemental material 1. Results Determination of minimum bactericidal concentration of biocides in planktonic phase The 64 tested A. baumannii isolates were of different biological sources including diseased animals, raw meat, human patients, hospital environment and were assigned to several STs (n = 52), in turn belonging to several clonal complex (CCs) (n = 17) or ICs (IC2, IC7, IC9, IC11) (Table S2). Isolates from animals, human patients and hospital environment were carbapenem-resistant and carried the blaOXA-23 or blaNDM-1 gene (Table S2). MBC50 and MBC90 were reported for all five biocides (Table 1). Table 1Minimal bactericidal concentration (MBC50 and MBC90) values of five biocides on 64 Acinetobacter baumannii isolates in planktonic phase in presence of bovine serum albumin (BSA) at 0.3 g/L and 3 g/LMBCBSASodium hypochlorite (mg/L)Ethanol (%)Hydrogen peroxide (%)Chlorhexidine digluconate (mg/L)Benzalkonium chloride (ppm)MBC500.3 g/L < 100400.53280150 ppm3 g/L800400.53280 > 200 ppmMBC900.3 g/L < 1004016560200 ppm3 g/L8004016560 > 200 ppm Minimal bactericidal concentration (MBC50 and MBC90) values of five biocides on 64 Acinetobacter baumannii isolates in planktonic phase in presence of bovine serum albumin (BSA) at 0.3 g/L and 3 g/L For ethanol MBC50 and MBC90 values were centred on 40% irrespective of the amount of BSA present (Fig. 1). Only one isolate, #58511 (ST2, IC2), isolated from the hands of hospital personnel, presented an ethanol MBC at 50% in presence of BSA at 3 g/L. Plot of colony forming unit enumeration (CFU/mL) for 64 A. baumannii isolates in planktonic phase and biocides concentrations. Green and red bars represent minimum and maximal biocides concentrations found in commercial formulations On the contrary, MBC50 and MBC90 values of sodium hypochlorite changed according to the concentration of BSA present, with MBC50/90 values inferior to 100 mg/L in presence of 0.3 g/L of BSA and with values centred on 800 mg/L in presence of 3 g/L of BSA (Table 1). Two isolates from raw meat (#53403, ST357, CC331 and #54684, ST106, CC1303) necessitated a higher concentration than MBC90 of sodium hypochlorite to be killed in presence of 0.3 g/L of BSA (Fig. 1, Table S2), whereas in presence of 3 g/L of BSA sodium hypochlorite MBC exceeded MBC90 for seven isolates (Fig. 1, Table S2). Isolates (n = 9) for which sodium hypochlorite MBC was the highest mostly belonged to CC331, CC1303, CC462 and CC1017 or ICs (IC2, IC7, and IC11). MBC50/90 values of hydrogen peroxide were distributed on a large concentration range (< 0.25–2%) and were not affected by the amount of BSA present (Table 1). The 17 isolates presenting the highest hydrogen peroxide MBC (1%) were collected from raw meat and hospital settings and 13 out of 17 belonged to several CCs, with four belonging to IC2 (Table S2). Similarly, MBC50/90 values of chlorhexidine spanned from < 800 mg/L to 6560 mg/L (Fig. 1). The amount of BSA did not affect the MBC50 and MBC90 values (Table 1). For eight isolates, chlorhexidine MBC (6560 mg/L) exceeded the minimum concentration of chlorhexidine used in commercial formulations. These eight isolates were assigned to several CCs, and to IC2 (n = 4), whereas two were sporadic (Table S2). MBC of benzalkonium chloride spanned from 100 ppm to values > 200 ppm (Fig. 1). Presence of BSA affected the MBC50 and MBC90 increasing with the increase of the BSA concentration (Table 1). For several isolates (n = 37) benzalkonium chloride MBC was higher than minimum concentration used in commercial formulations (Fig. 1). Susceptibility to biocides of A. baumannii isolates in biofilm matrix Isolates were screened for their ability to produce biofilm on abiotic surfaces at a solid surface. Among the 64 A. baumannii isolates, ten did not produce biofilm whereas four, 23 and 27 were weak, medium and strong biofilm former, respectively. Among the 27 isolates that were strong biofilm former, one isolate representing a CC or a ST was chosen (n = 21, including A. baumannii ATCC 19606 strain) to evaluate the susceptibility to biocides in biofilm matrices after 24 h production. In average biofilm matrix not exposed to biocides contained 4.3 × 107 CFU/mL. Exposure of biofilms to different biocides concentrations decreased significantly the amount of viable bacteria compared to non-exposed biofilms (Wilcoxon test, p < 0.01 or p < 0.001), with the exception of the smallest concentration of hydrogen peroxide (Fig. 2). Plot of colony forming unit enumeration (CFU/mL) from 21 A. baumannii isolates after exposure of biofilm matrix to different biocides concentrations. The first concentration tested corresponded to the minimum bactericidal concentration of the biocides as determined on planktonic phase (MBCp, ethanol: 40%; sodium hypochlorite: 100 mg/L; hydrogen peroxide: 0.25–2%; chlorhexidine digluconate: 820–6560 mg/L; benzalkonium chloride: 100–200 ppm). NT: not treated For ethanol the smallest concentration tested was 40%, which caused a drop of viable bacteria from an average of 3.2 × 107 to 5.1 × 106 CFU/mL. Exposure to an increased ethanol concentration (60%) caused a further decrease in viable bacteria (7.9 × 105 CFU/mL). However, exposure to an increased ethanol concentration (90%) did not change significantly bacteria killing, with an average bacterial enumeration equal to 3.1 × 105 CFU/mL (Fig. 2). Sodium hypochlorite at 100 mg/L, corresponding to the sodium hypochlorite MBC of A. baumannii isolates in planktonic phase, caused a significant reduction of viable bacteria in biofilm. Exposure of biofilms to higher concentrations of sodium hypochlorite caused further decrease of viable bacteria. At 60,000 mg/L, sodium hypochlorite caused a 3 log10 reduction of viable bacteria compared to unexposed biofilms (from 6.4 × 107 to 7.5 × 104 CFU/mL) (Fig. 2). The smallest concentration of hydrogen peroxide tested varied between 0.25 and 1%, corresponding to the hydrogen peroxide MBC of A. baumannii isolates in planktonic phase. Overall, these concentrations did not decrease the amount of viable bacteria in biofilms. On the contrary, 3% hydrogen peroxide caused a significant decrease of viable bacteria, likewise 6% hydrogen peroxide that caused a 2 log10 decrease compared to unexposed biofilms (from 3.3 × 107 to 7.4 × 105 CFU/mL in average) (Fig. 2). MBC of chlorhexidine digluconate in presence of 0.3 g/L BSA varied between 820 and 6560 mg/L for isolates in planktonic phase. For 11 isolates chlorhexidine digluconate MBC was higher than the minimum chlorhexidine digluconate concentration used in commercial formulations (5000 mg/L), thus the amount of viable bacteria did not change in biofilm exposed to MBC chlorhexidine digluconate concentration or chlorhexidine digluconate at 5000 mg/L. On the contrary, chlorhexidine digluconate at 40,000 mg/L significantly decreased the amount of viable bacteria in biofilm compared to unexposed biofilms causing a 3 log10 decrease (from 3.7 × 107 to 1.7 × 104 CFU/mL, in average) (Fig. 2). For benzalkonium chloride, all tested concentrations caused a significant decrease of bacterial load compared to unexposed biofilms. No differences between the tested concentrations was observed (Fig. 2). Thiazolyl blue tetrazolium bromide calibration assay An assay calibrating CFU/mL enumeration of a bacterial suspension versus the optical density resulting by the MTT reduction of the same bacterial suspension was carried out. Bacterial enumeration between 1.1 and 2.4 × 108 CFU/mL resulted in an optical density (600 nm) ranging between 0.36 and 0.48, whereas bacterial enumeration between 3.5 and 4.3 × 107 CFU/ml resulted in an optical density ranging between 0.142 and 0.145. Enumeration varying between 1.7 and 2.3 × 106 CFU/ml resulted in an optical density between 0.101 and 0.104 (Figure S1), similar to the absorbance of LB medium without bacterial inoculum (0.093–0.101). Overall, calculated CFU/ml from the MTT assay over-estimated bacterial amount compared to enumerated CFU/mL from biofilms exposed to biocides. Furthermore, no differences in viable bacteria could be observed between biofilms not exposed to biocides and those exposed to the MBC defined in the planktonic phase. Similarly, no differences in viable bacteria could be observed between the minimum and maximal concentrations used in commercial formulations (Table S3). Presence of genes associated to biocides non susceptibility All isolates were positive for the presence of the abeM, adeG, adeJ, adeL, adeT2, and abuO genes (with 90% amino acid identity and 60% coverage with sequences present in the BacMet database v.2.). Genes adeK, adeI, adeH, adeF, adeC, abeS, were present in all isolates but ATCC 19606 A. baumannii strain, whereas other genes were present in a variable number of isolates (adeA, n = 52; adeB, n = 58; adeN, n = 59; adeT1, n = 47; nreB, n = 63). Genes qacΔE1, adeD, and adeE were not reported. Expression of adeB and amvA genes Expression levels of adeB and amvA genes differed after chlorhexidine exposure. In particular the amvA gene displayed 2.4 folds increased expression in ATCC 19606 cultured in presence of 0.5 mg/L of chlorhexidine. However, exposure to chlorhexidine did not affect the amvA gene expression level in the other tested isolates. On the contrary, the adeB gene was over-expressed in all isolates exposed to chlorhexidine with the lowest fold difference observed in the ATCC 19606 strain (1.9), followed by the #58425 (2.8), #58475 (3.7), and #38208 (7.5) isolates. Discussion Biocides efficacy is crucial to preserve health, in particular in the hospital environment where hospital-acquired infections sustained by antibiotic resistant bacteria might occur in fragile patients [26, 27]. This study demonstrated that five biocides that are commonly used in disinfectants and antiseptics were able to kill a diverse population of A. baumannii, including isolates from hospital settings, diseased animals and from food sources, after 30’’ exposure. Observed MBCs proved that biocides were effective in eradicating bacteria in the planktonic phase at concentrations lower than those that are used in commercial formulations [16]. For instance, 40% ethanol could kill all bacteria in planktonic phase in a condition mimicking cleanliness (0.3 g/L BSA), whereas only one isolate needed 50% ethanol to be killed in dirty condition (3 g/L BSA) after 30’’ exposure. These results suggest that common products containing 70% ethanol should be effective to disinfect contaminated hands [28]. Ethanol at 40–100% concentrations has previously demonstrated bactericidal activity also against other species like Pseudomonas aeruginosa, and several species of the Enterobacterales family [29]. Similarly, sodium hypochlorite killed A. baumannii isolates at concentrations (MBC < 100 mg/L) 5000-fold lower than that contained in household bleach (52500–61500 mg/L) [30], although sodium hypochlorite killing was reduced in presence of BSA 3 g/L (MBC 800–1600 mg/L). Hydrogen peroxide MBC values displayed a unimodal distribution for the 64 A. baumannii isolates, with most of isolates killed after exposure to 0,5% solution for 30’’. The highest MBC value was 2% for one isolate. This concentration is still lower than the minimum hydrogen peroxide concentration (3–6%) used to disinfect inanimate surfaces [31]. Vaporization of high hydrogen peroxide concentration (30%) is efficacious in reducing multidrug resistant bacteria contaminating ICU rooms [32] and in counteracting outbreaks caused by multidrug resistant A. baumannii isolates [33]. Also MBC values of chlorhexidine displayed a unimodal distribution, with most of isolates killed with 6560 mg/L (MBC90). None of the 64 A. baumannii isolates presented a MBC higher than the highest concentrations used in commercial disinfectants, but several displayed a chlorhexidine MBC higher than minimum usage concentration. As expected, chlorhexidine MIC values are lower than MBC as a recent report showed on 96 A. baumannii isolates of environmental origin (MIC90: 64 mg/L) [14]. MBC values of benzalkonium chloride were more sparse and, in presence of interfering substance at 3 g/L, MBC could not be determined for several isolates (> 200 ppm). Overall, BSA amount affected the benzalkonium chloride efficacy, shifting MBC towards higher values exceeding the minimum concentrations used and approaching maximal concentrations of usage (800 ppm). Benzalkonium chloride efficacy are known to be affected by factors like high water hardness and materials such as cotton that can favour insoluble precipitates and decrease benzalkonium chloride bactericidal activity [16]. Most isolates that showed the highest MBC values for each biocide (hydrogen peroxide: 1–2%, n = 15; sodium hypochlorite: 1600 mg/L, n = 7; ethanol: 50%, n = 1; chlorhexidine: 6560 mg/L, n = 22) belonged to CCs or ICs. For instance 11 out of the 15 isolates displaying hydrogen peroxide MBC 1–2% belonged to IC2 (n = 2), to the expanding CC33 and CC32 (n = 2 and 3, respectively), CC288, CC331, CC1017 and CC1303 each including one isolate [6]. The only isolate, #58511, with ethanol MBC 50% belonged to IC2 and was collected from the hands of hospital personnel where circulation of carbapenem-resistant A. baumannii was endemic [17]. Furthermore, two IC2 isolates, collected in the same hospital of #58511 isolate, were among the seven isolates displaying the highest sodium hypochlorite MBC (1600 mg/L). Of the remaining isolates with high sodium hypochlorite MBC (1600 mg/L) only one was assigned to ST2143 that was a singleton, the others belonged to IC7 (n = 1), IC11 (n = 1) and expanding CCs (CC1017, CC462). Similarly, among the 22 isolates for which chlorhexidine MBC was 6560 mg/L only two were assigned to sporadic STs, five belonged to IC2 and were collected from hospital settings and other belonged to expanding CCs (CC33, n = 2; CC32, n = 3; CC1303, n = 2). Two isolates, #54720 from raw meat and assigned to ST2143 that is currently a singleton and #58424 belonging to IC2 collected from the bed of an ICU room, were among isolates displaying the highest MBC of sodium hypochlorite, hydrogen peroxide, and chlorhexidine. Previously, Wispinghoff et al. investigating the biocides susceptibility of sporadic A. baumannii isolates and those from outbreaks with a methodology similar to that used in our study, highlighted that bacterial cell reduction after exposure to diluted chlorhexidine was lower for epidemic strains compared to the sporadic ones, although such differences did not reach statistical significance [34]. Among isolates presenting the highest biocides MBC, no peculiar pattern of genes involved in biocides tolerance was identified, but isolates belonging to IC2 had a complete adeABC operon encoding for a RND efflux pump, which could mediate cross-resistance to biocides and antibiotics [9]. In selected isolates of this study, adeB over-expression occurred after exposure to chlorhexidine, while such over-expression was not observed in chlorhexidine-free medium. This result is in line with previous observation from Singkham-in et al. who also highlighted that chlorhexidine over-expression of adeABC is mediated trough alternative regulators to adeRS [35]. All the investigated isolates (ATCC 19606, #58425, #58475, and #38208) presented high chlorhexidine digluconate and benzalkonium chloride MBC ranging from 3280–6560 mg/L and 150–200 ppm, respectively. Other efflux mechanisms could contribute to biocides resistance, like the AmvA pump. However, the gene encoding for this major facilitator system was not over-expressed even after chlorhexidine exposure. Other genes encoding for other efflux systems were detected in the A. baumannii isolates of this study, like the AbeS that has been demonstrated to mediate chlorhexidine and benzalkonium efflux [36]. Also the AdeIJK system with low-level constitutional expression contributes to chlorhexidine and belzalkonium tolerance [37] and the genes encoding for this operon were present in all isolates of this study. Similarly, the TolC homolog AbuO present in all isolates could contribute to chlorhexidine and benzalkonium tolerance and to oxidative stress survival [38]. This study mainly focused on determining the phenotypic base-line of A. baumannii susceptibility to biocides and a dedicated study on the contribution to biocides tolerance of other pumps than adeABC or amvA could be pertinent. If in planktonic phase, concentrations needed to kill A. baumannii isolates were inferior to those commonly present in commercial products, susceptibility of isolates embedded in biofilm matrix changed dramatically. In fact, exposure of biofilms also to the highest used concentrations of biocides failed to eradicate bacteria and even no bactericidal action was observed after 30’’ exposure (≥ 5 log10 reduction). This does not mean that biocides did not conserve bactericidal action at those concentrations, but exposure time for the concentrations tested was not sufficient. Previous investigations have reported that 70% ethanol was not able to eradicate biofilms even after 15’ exposure and that 0.1% (1000 mg/L) of sodium hypochlorite was efficient in eliminating biofilms after 15’ exposure [39]. Also hydrogen peroxide has proved to be efficacious in significantly reducing bacterial load in biofilm matrix causing > 6 log10 bacteria reduction in 50’ [40]. Considering the impact of biofilm formation on biocides susceptibility [7], evaluating biocides susceptibility in biofilm matrix of field isolates is crucial to understand the efficacy of currently used biocides. Indeed, clinical A. baumannii isolates frequently produce biofilm [41], similarly to what we observed in the current population where 50 out 64 isolates were biofilm former and 27 strong biofilm former. More information should be gathered also on the amount of bacteria in biofilms occurring on fields and the contribution of multispecies biofilms. However, the study of biofilms using CFU enumeration is time consuming and requires a large amount of laboratory consumables, in particular cultivation plates. We thus conducted a parallel assay to the CFU counting based on the deduction of bacterial viability after reduction of MTT that generates optical absorbance proportional to the amount of viable bacterial cells. This assay proved to be adapted for the quantification of bacterial suspension containing 107–108 CFU/mL load but it failed to highlight bacterial load differences between the different biocides concentrations tested on bacteria embedded in biofilms. One major limitation of the current study is the lack of comprehensive information on the level of expression of genes encoding for mechanisms involved in biocides tolerance and the regulatory response to biocides exposure. Here we tested only three isolates for adeB and amvA expression and we evaluated only response after chlorhexidine exposure. A dedicated study on the expression of efflux systems present in these isolates should be performed to nourish the understanding of efflux in tolerance to biocides and the interplay with antibiotic resistance. One other limitation is the lack of guidelines to estimate the exposure periods to biocides in order to evaluate isolates susceptibility in biofilm matrix. Here, we tested the same exposure time as the one used for planktonic bacteria (30’’), but longer exposures might be tested. However, this study presents strengths. It is unique in investigating a diverse population of A. baumannii isolates conjugating information on clonality, and biocides susceptibility characterization, demonstrating that isolates presenting the highest MBC values were those assigned to the expanding CC32, CC33, CC1017, and CC462 or IC2 and IC7, suggesting that these isolates have the potential to disseminate. Those belonging to IC2 were collected in an endemic context and were multidrug resistant. The fact that these isolates presented higher MBC values compared to other isolates could be linked to selective pressure exerted in hospital settings. However, among isolates presenting higher MBC values, those collected from raw meat demonstrated a capacity to colonize several meat processing sites. In this context, selective pressure by intensive cleaning practices is conceivable as well. Besides, among the 30 isolates with the highest biocides MBC, 22 were medium or strong biofilm former. In this lifestyle, repeated exposure to sub-eradicating concentrations of biocides could have selected isolates with higher biocides tolerance [42, 43]. Comparison of our results to other studies was possible for few that used a similar methodology [34, 39]. The majority of studies focus on the evaluation of biocides MIC values and the absolute concentrations of tested biocides are sometimes not clearly explicated [44, 45]. Conclusions The methodology we used to determine biocides action is robust taking into consideration parameters that are essential to evaluate biocides action like the temperature of incubation, the exposure period and the presence of interfering substance. MBC values determined in this study for certain biocides like ethanol, hydrogen peroxide, chlorhexidine and sodium hypochlorite, which present a unimodal distribution, could provide the first data set for the elaboration of epidemiological cut offs that could definitely facilitate comparison of results from studies using different methodologies and in different epidemiological context. Future efforts should focus on harmonized methodologies and understanding if changes in biocides MIC values of tested isolates correlate with a change of efficacy of biocides in proper conditions without neglecting testing on biofilm matrix. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Supplementary Material 4. Supplementary Material 5. Author contributions Conceptualization: AL, MH, JYM; Data curation: LH, ES, TD; Formal analysis: AL, VP; Funding acquisition: MH, JYM; Investigation: LH, SA; Methodology: ES, LH, AL; Project administration: MH; Supervision: AL, MH, WM; Validation: AL, MH, JYM, WM; Visualization: LH, TD; Statistical test: TD; Roles/Writing—original draft: LH, AL; and Writing—review & editing: all authors. Funding Internal funds from Agence Nationale de Sécurité Sanitaire de l’Alimentation,de l’Environnement et du Travail. Data availability and materials Genome sequences of isolates are available from two bio projects (PRJNA1076118 and PRJNA999198). Isolates are available at request. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests No competing interest to declare for this study. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Supplementary Material 4. Supplementary Material 5. Data Availability Statement Genome sequences of isolates are available from two bio projects (PRJNA1076118 and PRJNA999198). Isolates are available at request.