Antiseptic quaternary ammonium compound tolerance by gram-negative bacteria can be rapidly detected using an impermeant fluorescent dye-based assay

The study addresses the need for rapid, standardized methods to assess bacterial tolerance to quaternary ammonium compounds (QACs), widely used antiseptics/disinfectants implicated in antimicrobial resistance selection. Current testing relies on slow, labor-intensive AST without defined QAC breakpoints. The authors developed a rapid fluorescent dye-based membrane integrity assay (RFDMIA) using propidium iodide (PI) to detect QAC-induced changes in membrane permeability. Hypotheses: at identical PI concentration, increasing QAC levels will cause greater PI entry and fluorescence in QAC-susceptible cells than in QAC-tolerant cells; thus, the earliest significant rise in PI fluorescence over 30 minutes (ΔRFU30min) will occur at or below the MIC for susceptible isolates, while tolerant isolates will show increases at higher QAC concentrations due to greater membrane integrity. The assay’s ability to align with MIC and 30-minute MBC (30MBC) values and to generalize across Gram-negative species was tested.

QACs (e.g., benzalkonium chloride, BZK; cetrimide, CET) are cationic surfactants used extensively in healthcare, industry, and consumer products. They act by disrupting membrane lipids/proteins, altering membrane fluidity, and promoting cell leakage and death. Many proteobacterial pathogens (Enterobacterales, Acinetobacter spp., Pseudomonas spp.) show intrinsic tolerance and can adapt to sub-lethal QAC exposure, increasing QAC tolerance, cross-tolerance to other biocides, and cross-resistance to antibiotics. Despite widespread use (global QAC use exceeds antibiotic use), there are no EUCAST/CLSI standards or breakpoints for QAC susceptibility; thus, measurements are reported as tolerance. Existing methods (agar/broth microdilution) are slow (24–48 h). A rapid, sensitive assay is needed to quantify QAC susceptibility and study its role in resistance development.

Assay principle: RFDMIA monitors changes in membrane permeability via a membrane-impermeant fluorescent dye (primarily propidium iodide, PI; EX 544 nm/EM 620 nm). Increased dye entry and nucleic acid binding raise fluorescence. Dye emission is read every 5 min over 30 min in 96-well black, optical-bottom plates. Bacterial strains: Escherichia coli K-12 BW25113 parental (EC) and laboratory-adapted derivatives with increased tolerance to BZK (ECBZKT) or CET (ECCETT) via gradual subculture in sub-inhibitory QACs (~40 passages). Additional Gram-negative species: Acinetobacter baumannii DSM 6974 (AB), Pseudomonas aeruginosa PAO1 DSM 22644 (PA), Klebsiella pneumoniae DSM 6135 (KP) and its BZK-adapted variant (KPBZKT), Shigella sonnei DSM 5570 (SS) and its BZK-adapted variant (SSBZKT). Cultured in LB at 37 °C, 170 rpm. Adapted strains maintained with QAC selection (e.g., CET 50 µg/mL; BZK 10–40 µg/mL depending on strain) to preserve tolerance phenotype. Cell preparation: Three preparation states were compared: stationary phase (overnight; standardized to OD600=2.0, washed, diluted to OD600=0.2 in filtered PBS), mid-log phase (grown to OD600=0.5 then processed identically), and agar colony suspensions (from overnight plates; standardized as above). Heat-treated controls (121 °C, 30 min) provided maximal permeability references. RFDMIA conditions: Each well contained PI at 2 µg/mL (SYTOX Blue at 1 µM used in select experiments), bacteria or PBS blank, and a twofold QAC dilution series (0–600 µg/mL) of BZK or CET. Fluorescence measured (Polarstar Optima) at EX/EM appropriate for dye every 5 min for 30 min. Data processing: For each QAC concentration X, blank-subtracted sample RFUs were computed at T0 and T30 to obtain RFUΔ30min(X) = RFU30min − RFU0min. To control for dye uptake without QAC, ΔRFUΔ30min = RFUΔ30min(X) − RFUΔ30min(0). The first QAC concentration showing a statistically significant increase versus the lowest measured QAC level was identified (two-tailed Student’s t test; P<0.05 or P<0.01 as indicated). Inter-isolate comparisons at the same QAC also used t tests. Benchmarking: Broth microdilution AST determined MICs (lowest concentration with no growth after overnight incubation). A 30-min MBC (30MBC) was determined by spotting ~1–2 µL from RFDMIA wells onto LB agar (with selection for adapted strains where needed) immediately after the 30-min exposure; the lowest QAC concentration with no growth across replicate spots was recorded. Each measurement used 3 biological replicates with technical triplicates. SEM: Stationary-phase suspensions exposed to QACs for 30 min were fixed and imaged using a JEOL JCM-5700 SEM after gold sputtering (5000× magnification; five images per condition). Blinded analysis by two assessors quantified proportions of inflated/intermediate/deflated cells and measured lengths and widths (n=100 cells/condition). Statistical tests included Mann–Whitney U for morphology distributions.

- RFDMIA tracked QAC-induced permeability changes that aligned near MICs for susceptible and adapted E. coli. For BZK: EC showed the first significant PI ΔRFU30min increase at 18.8 µg/mL (its MIC), while ECBZKT showed it at 75 µg/mL, coinciding with its higher tolerance. For CET: EC showed the first significant increase at 37.5 µg/mL (its MIC), whereas ECCETT showed it at 150 µg/mL, twofold below its MIC (300 µg/mL), indicating higher baseline permeability or membrane changes in adapted cells. - Across broad and narrow concentration ranges, RFDMIA distinguished tolerant from susceptible phenotypes between 10–40 µg/mL for both BZK and CET in E. coli, enabling discrimination at sub-MIC concentrations. - The maximum PI ΔRFU30min rarely predicted 30MBC; for EC with BZK, 30MBC occurred at 75 µg/mL while the RFDMIA maximum was at 150 µg/mL, indicating maxima are not reliable for bactericidal thresholds. - Growth state impacted performance: stationary-phase suspensions provided the most reproducible, lowest-error discrimination concordant with MICs. Mid-log cells yielded higher signals but greater variability and less reliable MIC alignment. Colony suspensions poorly differentiated tolerant/susceptible phenotypes (similar PI responses across isolates), though EC MICs were still detected. - High QAC concentrations reduced dye signal (quenching/convergence of live and heat-treated signals) at ≥150 µg/mL BZK and ≥300 µg/mL CET, limiting detection accuracy. This coincided with reported critical micelle concentrations (CMC), implicating micellization and membrane–detergent interactions in fluorescence loss. - Species robustness: RFDMIA correctly predicted MIC-aligned first significant increases for A. baumannii with both BZK (18.8 µg/mL) and CET (37.5 µg/mL). P. aeruginosa produced high background fluorescence at all tested concentrations using PI and also SYTOX Blue, preventing accurate susceptibility discrimination due to intrinsic pigment fluorescence. - Enterobacterales beyond E. coli: BZK-adapted K. pneumoniae (KPBZKT) and S. sonnei (SSBZKT) were distinguished from their parental strains (KP, SS) at or below MICs; the first significant increase in KPBZKT coincided with its MIC (75 µg/mL BZK), and SSBZKT at 37.5 µg/mL BZK. - SEM revealed morphological alterations in adapted E. coli: ECBZKT and ECCETT cells were longer and thinner than EC. ECCETT exhibited a chain-like morphology and a CET-dependent phenotype—cells were deflated without CET but regained inflated bacilliform morphology at sub-MIC CET; at MIC, deflation increased again. These morphological shifts align with RFDMIA signals and suggest altered membrane integrity in adapted isolates. - Overall, RFDMIA rapidly (30 min) discriminated QAC susceptibility near MICs for several Gram-negative species, particularly Enterobacterales, using low QAC concentrations (10–40 µg/mL).

The findings support that monitoring impermeant dye fluorescence over 30 minutes provides a sensitive proxy for QAC-induced membrane perturbation and thus tolerance phenotype. The first significant ΔRFU30min rise typically occurred at or near MICs for susceptible isolates and at higher concentrations for adapted/tolerant isolates, enabling rapid discrimination without 24–48 h incubation. Stationary-phase preparations offered the most reliable performance, likely due to consistent membrane composition compared with mid-log or colony-derived cells, where lipid content, capsule/exopolymers, and aggregation can alter dye access. The observed high-concentration fluorescence loss underscores an inherent assay ceiling related to surfactant micellization (near CMC) and potential counter-ion effects on dye emission, necessitating operation below these limits. RFDMIA generalized to A. baumannii and Enterobacterales but failed for P. aeruginosa due to intrinsic fluorescence, highlighting dye/spectral constraints. SEM corroborated RFDMIA by revealing membrane-related morphological changes and a CET-dependent integrity phenotype in adapted E. coli, suggesting adaptation-driven envelope remodeling that increases baseline permeability and can underpredict MIC relative to AST. Collectively, RFDMIA addresses the need for a rapid, biocide-focused susceptibility screen and can complement MIC testing for QACs where standardized breakpoints are lacking.

This study introduces RFDMIA, a rapid (30-min), sensitive fluorescent microplate assay that differentiates QAC-susceptible from QAC-tolerant Gram-negative bacteria by quantifying impermeant dye uptake as a surrogate for membrane integrity. RFDMIA aligned near MICs for multiple species and adapted isolates, functioned best with stationary-phase suspensions, and detected phenotypic membrane changes corroborated by SEM, including a novel CET-dependent morphology in adapted E. coli. The assay is practical for screening at low QAC concentrations (10–40 µg/mL) but should be used below CMC and avoided for intrinsically fluorescent species like P. aeruginosa. Future work should: explore alternative dyes/spectral windows to overcome intrinsic fluorescence; refine calibration to better estimate bactericidal thresholds; investigate longer exposure times or kinetic parameters for improved MIC prediction; and expand to broader taxa and biocides to develop standardized rapid susceptibility testing for antiseptics.

- Intrinsic fluorescence in Pseudomonas aeruginosa (and likely other fluorescent species) prevents accurate discrimination using PI or SYTOX Blue due to high background emission. - Signal quenching and convergence at high QAC concentrations (near or above CMC; ≥150 µg/mL BZK, ≥300 µg/mL CET) limit the dynamic range and preclude accurate susceptibility estimation at these levels. - RFDMIA often underpredicted MICs for adapted isolates, reflecting methodological and physiological differences from broth microdilution (short exposure in PBS vs 18–24 h growth in media) and possible QAC-dependent membrane integrity in adapted strains. - Colony-derived and mid-log preparations introduced higher variability or reduced discrimination compared to stationary phase. - Reliance on laboratory-adapted isolates (lack of standardized QAC-resistant reference strains) may limit generalizability; adaptation-induced envelope remodeling could alter dye permeability independent of true QAC tolerance mechanisms. - No established QAC breakpoints; results indicate tolerance rather than standardized resistance classifications.