B. cepacia ATCC 25416 was sourced from the culture collection at the Food Microbiology Laboratory of The Ohio State University (Columbus, OH, USA). The bacterium is linked to the spoilage of onion bulbs and unpasteurized raw milk (Moore et al., 2001), and it is recognized for its robust ability to form biofilms (Tavares et al., 2020). The strain was streaked from frozen stock stored at –80 °C in 25% glycerol (Sigma-Aldrich, Burlington, MA, USA) onto a Tryptic Soy Agar (TSA) (Bacton, Dickson and Company, Franklin Lakes, NJ, USA) followed by aerobic incubation in static condition at 37 °C for 24 hours. Fresh frozen stock of the culture was used to initiate each experiment.

B. cepacia ATCC 25416 was grown using two media (Table 1); (i) tryptone yeast extract dextrose (TYD) broth, a nutrient rich medium, which was derived from Ashdown’s medium (Howard and Inglis, 2003) and B. cepacia selective agar medium (Thermo-Fisher Scientific, Waltham, MA, USA) after excluding the selective agents and modifying the composition, and (ii) yeast extract dextrose calcium carbonate broth (da Silva et al., 2021), a nutrient limited medium that was modified (mYDC broth) by excluding calcium carbonate, which tended to precipitate after autoclaving. Agar versions of these broths were made by including agar at a 2.0% level.

The growth curves of B. cepacia ATCC 25416 were determined in TYD and mYDC broth media as follows. To prepare the inoculum, the bacterium was grown in mYDC broth under aerobic conditions for 24 hours at 37 °C with shaking at 180 rpm. The overnight culture was diluted in TYD or mYDC broth to a final population of 10³–10⁴ CFU/mL. Aliquots (200 μL) of the diluted culture were dispensed in the well of a polystyrene 96-well plate and incubated statically at 37 °C. Samples (100 μL, each) were taken (from separate wells) after incubation for 0, 3, 6, 9, 12, 15, 20, 24, 36, and 48 hours. These samples were ten-fold serially diluted and plated on TYD and mYDC agar. The plates were incubated at 37 °C for 24 hours, and B. cepacia populations were counted. The population counts over time (growth curves), derived from three independent replicates, were fitted using the following Gompertz model (Ali et al., 2025):

where the dependant variable “Y” is B. cepacia population (log₁₀ CFU/mL), the independent variable “t” is time (hour), and model’s parameters were “a” is the lower asymptote (log₁₀ CFU/mL), “b” is the upper asymptote (log₁₀ CFU/mL), “c” is the growth rate (log₁₀ CFU/mL/hour), and “d” is the inflection point (hour). The model’s mathematical parameters (a, b, c, and d) were determined using statistical software (JMP Pro 17; SAS Institute Inc., Cary, NC).

Biofilm optimization in the two microbiological media (Table 1) was conducted using polystyrene 96-well microtiter plates (Corning Costar; Fisher Scientific). The colorimetric quantification method, using crystal violet, was used to assess biofilm formation (Rose et al., 2009). This experiment was designed as described previously (Reddersen et al., 2021) with modifications. The following four experimental variables were evaluated: inoculum size (10⁴–10⁷ CFU/mL), aerobic incubation conditions (static vs. shaking), incubation time (24 vs. 48 hours), and growth media (TYD vs. mYDC) for their impact on biofilm. The 24-h-old B. cepacia culture was adjusted to 0.1 OD_600nm (10⁷ CFU/mL as confirmed by plating on mYDC agar). Serial dilutions were performed to obtain inoculum sizes of 10⁶, 10⁵, and 10⁴ CFU/mL. A 200-μL aliquot of each diluted culture in each microbiological medium was added to the wells of polystyrene 96-well microtiter plates. Plates were incubated under aerobic static conditions at 37°C or under shaking conditions (140 rpm) at 37°C for 24 or 48 hours to allow biofilm formation. Sterile uninoculated media served as a negative control. Following the incubation, planktonic cells were carefully aspirated from each well, and wells were washed three times with sterile saline solution (0.85%) to remove non-adherent cells. Plates were then allowed to dry at 55°C for 20 min. Biofilms were stained by adding 250 µL of 0.1% crystal violet solution (Becton, Dickinson and Company, Franklin Lakes, NJ, USA) to each well, followed by incubation at room temperature (25 ± 2°C) for 25 min. The plate wells were washed with sterile water to remove any unbound or excess stain, and then the excess water was carefully removed. Subsequently, crystal violet in each well was solubilized in 300 µL of 70% ethanol (Decon Laboratories Inc., King of Prussia, PA, USA) for 45 min. Biofilm biomass was quantified by measuring absorbance at OD_590nm using a microtiter plate reader (SpectraMax 384 Plus; Molecular Devices, San Jose, CA, USA). Each condition was tested in four replicates.

The antimicrobials used were tetracycline hydrochloride (Sigma-Aldrich, Inc.), chloramphenicol (IBI Scientific, Peosta, IA), kanamycin sulfate (Sigma-Aldrich, Inc.), ceftazidime (European pharmacopoeia reference standard), erythromycin (Sigma-Aldrich, Inc.), ampicillin (Sigma-Aldrich, Inc.), ciprofloxacin hydrochloride (MP biochemicals, USA), and meropenem trihydrate (ThermoFisher Scientific). Each antimicrobial was prepared by dissolving in the appropriate solvent or water to obtain stock concentrations of 2000 μg/mL (Jorgensen, 2012). Resazurin solution was prepared by dissolving 0.05 g resazurin powder (Sigma Aldrich, Inc.) in 50 mL sterile distilled water. The solution was thoroughly mixed and then diluted 1:10 to obtain a final working concentration of 0.01% for further experiments. To protect the dye degradation from light exposure, the resazurin solution was stored in a sterile tube wrapped with aluminum foil.

The resazurin-based microdilution assay was used to evaluate the inhibitory effects of the antimicrobials against B. cepacia planktonic cells using the broth microdilution method (Elshikh et al., 2016). The B. cepacia strain was grown in the mYDC broth for 24 hours at 37°C with shaking at 180 rpm. Following incubation, the inoculum was prepared by diluting the overnight culture 1:1000 in the mYDC to achieve a cell density of ~10⁴ CFU/mL. The selected antimicrobials were serially diluted two-fold in mYDC broth to generate final concentrations ranging from 0.24–1000 μg/mL. In the polystyrene 96-well plates, 100 µL of the appropriate antimicrobial dilution and 10 µL of the diluted culture were added. The plates were briefly shaken to ensure uniform mixing before incubation at 37°C for 24 hours. After incubation, turbidity readings were taken at OD₆₀₀ nm and then 10 μL of resazurin dye (0.01%) were added to each well and incubated at room temperature for 30 min. Active bacterial cells will reduce the non-fluorescent resazurin dye (blue) to the fluorescent resorufin (pink) (O’Brien et al., 2000). MIC values were interpreted using the standard CLSI M100 Performance Standards for Antimicrobial Susceptibility Testing, 33^rd edition, 2023 (Lewis et al., 2023). The lowest concentration before turbidity and resazurin color change was considered as the MIC. The color change of the dye was assessed visually, with a shift from blue to pink, indicating cell growth and no change indicating absence of growth (Chakansin et al., 2022). Microdilution was performed in triplicate for each antimicrobial.

The enzymes α-amylase from Bacillus spp. (Sigma-Aldrich, Inc), proteinase K (Invitrogen, Carlsbad, CA), and DNase I (Thermo Fisher Scientific) were included in this test. The enzymes were solubilized according to the manufacturer’s instructions, as follows. The α-amylase (10 mg/mL) was prepared by solubilizing the powder in a buffer solution consisting of 23 mM potassium phosphate and 6.6 mM NaCl, and the pH was adjusted to 6.9. To prepare DNase-I, 100 μL of the commercial enzyme preparation was diluted in 900 μL of DNase reaction buffer (Thermo Fisher Scientific), and further diluted 1:10 to achieve a final concentration of 0.01 U/μL. Proteinase K was prepared by dissolving the commercial preparation in 50 mM Tris-HCl (pH 8.0), containing 2 mM calcium acetate (Mallinckrodt chemicals, St. Louis, MO, USA), to prepare a solution containing 2 mg/mL. All enzyme solutions were stored on ice while performing the experiments.

Checkerboard microdilution assays were conducted to determine the MIC of individual agents (antimicrobial or enzyme) and their combinations against preformed biofilms in polystyrene 96-well microtiter plates. To distinguish the MIC measured against preformed biofilms from the one previously determined against planktonic cells (MIC-Plank), the former will be designated as MIC-Bio. MIC-Bio was quantified using the crystal violet assay, and absorbance was measured at OD_590nm. Briefly, a 24 - hour old B. cepacia culture was adjusted to approximately 10⁴ CFU/mL in mYDC broth, and 200 µL of this suspension was added to each well of a polystyrene 96-well microtiter plate in a 6 × 7 matrix format. Plates were incubated at 37 °C for 48 hours to allow biofilm formation. After the incubation, the planktonic cells in each well were gently removed and washed twice with sterile saline solution (0.85%) to yield pre-formed biofilms for assessing the efficacy of enzyme-antimicrobial combinations. In a separate sterile 96-well plate, selected antimicrobials and enzymes were prepared at different concentrations, resulting in 6 combinations (Table 2). The prepared combinations were then added to the appropriate wells containing the pre-formed biofilms, and plates were incubated at 37°C for an additional 24 hours. After the incubation, the planktonic cells in each well were gently removed, and the wells were washed three times with sterile saline solution (0.85%). The excess water was allowed to dry by holding the plates at 55°C for 20 minutes. Biofilm formation was quantified by crystal violet assay using 250 µL of 0.1% crystal violet (Becton, Dickinson and Company) as described previously. The fractional inhibitory concentration index (FICI) for enzyme-antimicrobial combinations was calculated, as described previously (Liu et al., 2021), using the following equation:

The interaction was defined as follows: synergy, FICI ≤ 0.5; additivity, FICI > 0.5 to 1.0; and antagonism, FICI ≥ 2 (Fratini et al., 2017).

Data were analyzed using a commercial statistical analysis software (GraphPad Prism 9.0.0; GraphPad software, San Diego, CA, USA). The analysis of variance (ANOVA) was paired with Tukey’s test to rank pairs for multiple comparisons. Tests with p < 0.05 were considered significant.

The growth curves of B. cepacia in TYD and mYDC broths, incubated under static conditions at 37 °C, are illustrated in Figure 1. A typical growth curve was observed in the nutrient-rich, TYD broth, where the planktonic population steadily increased from an initial value of 3.2 ± 0.58 log₁₀ CFU/mL at 0 hour to a maximum of 8.7 ± 0.08 log₁₀ CFU/mL by the 20^th hour of incubation and remained steady throughout the remainder of the incubation period. The predicted growth, fitted using the Gompertz model, was governed by the following equation:

where “Y” is the B. cepacia population (log₁₀ CFU/mL) and “t” is the incubation time (hour). Based on the model’s parameters, the estimated maximum growth rate was 0.24 log₁₀ CFU/mL/hour, the predicted maximum growth was 8.6 log₁₀ CFU/mL, and the correlation coefficient (R²) was 0.99, indicating a good fit of data by the model.

The growth behavior of B. cepacia in the mYDC broth was different than that observed in the TYD broth (Figure 1). Despite its poor nutritional composition, mYDC broth supported typical growth behavior during the first 15 hours of incubation, which was evident when this phase of growth was fitted using the Gompertz model (0.99 R2):

Growth parameters predicted by the model revealed a higher maximum growth rate (0.36 log₁₀ CFU/mL/hour) and a lower maximum growth (8.2 log₁₀ CFU/mL) in mYDC broth when compared with the growth patterns exhibited in TYD broth. After the growth reached its peak at 15^th hour (8.2 log₁₀ CFU/mL, measured as well as model-predicted), the measured population decreased during the subsequent 21 hour to 5.9 log₁₀ CFU/mL (2.3 log decrease), then increased again during the following 12 hours to 6.3 log₁₀ CFU/mL (0.4 log increase). This cyclic growth pattern in the mYDC medium suggests a phenotypic switch from the planktonic state to biofilm-associated cells’ adherence to the surfaces of the polystyrene microtiter plate, which was followed by biofilm maturation and release of cells from the biofilm matrix. Expression of this cyclic behavior in mYDC broth but not in TYD broth is probably induced by the nutrient-limited conditions of the former medium.

When biofilm development was quantified using the colorimetric crystal violet assay (OD_590nm) (Figure 2), the largest biofilm density was observed at an inoculum size of 10⁴ CFU/mL under shaking conditions (140 rpm) in mYDC broth, with optical densities (OD_590nm) of 3.0 ± 0.62 after 24 hours and 3.8 ± 0.25 after 48 hours of incubation. The significant difference in biofilm density (p < 0.0001) was observed with the 10⁴ CFU/mL inoculum size compared to the other inocula (Figure 2). Compared to shaking conditions, biofilm density under aerobic static incubation was consistently smaller across all inoculum sizes (p < 0.0001), with the largest OD_590nm being 1.1 ± 0.41 for 10⁴ CFU/mL in mYDC broth. Compared to mYDC broth, biofilm formation in TYD broth was comparatively smaller across all incubation conditions and inoculum sizes, with OD_590nm values below ~1.0. These findings indicate that TYD broth, being nutrient-rich, likely promoted the planktonic growth rather than biofilm formation, which is often induced under nutrient-limiting conditions, as in the case of mYDC broth.

The MIC values of the antimicrobials tested against planktonic cells of B. cepacia (MIC-Plank) were interpreted using the standard CLSI M100 Performance Standards for Antimicrobial Susceptibility Testing, 33^rd Edition, 2023 (Lewis et al., 2023). The MIC values ranged from 1000 μg/mL to 4 μg/mL for the tested antimicrobials (Table 3). The highest MIC (1000 µg/mL) was observed for erythromycin, whereas the lowest MICs, 4.0 and 8.0 μg/mL, were observed against ciprofloxacin and meropenem, respectively; for these two antibiotics, the absence of resazurin color change corresponded to the absence of turbidity at OD₆₀₀ nm. According to the CLSI MIC breakpoints for the B. cepacia complex, the test bacterium was resistant to all tested antimicrobials except ciprofloxacin and meropenem, to which intermediate susceptibility was exhibited (Table 3).

The MICs of selected antimicrobials were tested, alone or in combination with potentially synergistic enzymes, against preformed B. cepacia biofilms (MIC-Bio), and the results are shown in Figure 3 and analyzed in Table 4. When applied individually, ciprofloxacin at four times its MIC-Plank was required to eliminate B. cepacia preformed biofilm, whereas meropenem at twice its MIC-Plank eliminated the preformed biofilm (Tables 3, 4). However, the two antimicrobials eliminated the biofilm at levels lower than their individual MIC-Bio when combined with the matrix-degrading enzymes. Out of the six combinations (Table 4), α-amylase and ciprofloxacin possessed the highest synergistic effect (FICI, 0.50), whereas proteinase K and ciprofloxacin (FICI, 0.625), and α-amylase and meropenem (FICI, 0.750) displayed an additive effect (Table 4). In contrast, DNase I in combination with either ciprofloxacin or meropenem exhibited an antagonistic effect (FICI, ≥ 2), showing no impact on biofilm eradication. Additionally, the combination of meropenem with proteinase K (FICI≥ 2) displayed an antagonistic interaction.

Stainless-steel is a widely used food contact material in processing equipment due to its corrosion resistance, ease of cleaning, and exceptional mechanical strength (Ciolacu et al., 2022). To mimic an industrial environment, biofilms were developed on stainless-steel coupons, and conditions that affect the adhesion of B. cepacia to their surfaces were investigated. Considering the suitability of mYDC medium and 10⁴ CFU/mL inoculum for robust biofilm development in the wells of microtiter plates, these conditions were applied in stainless-steel experiments. During the first 24 hours of incubation (Figure 4), the biofilm populations reached 7.5 ± 0.1 and 7.0 ± 0.26 log₁₀ CFU/coupon, under aerobic static and shaking conditions, respectively, but the difference between these two populations was not significant (p > 0.05). During the subsequent 24 hours of incubation, the biofilm populations did not increase considerably, but the maximum growth of 7.8 ± 0.10 and 6.9 ± 0.32 log₁₀ CFU/coupon was reached for biofilms formed under aerobic static and shaking incubation, respectively, and these two populations were significantly different (p < 0.01). During the last 24 hours of the 72-hours incubation, biofilm log₁₀ CFU/coupon decreased slightly, possibly due to nutrient depletion or biofilm detachment.

The most promising enzyme-antimicrobial combinations, based on the checkerboard assay in the microtiter plates (Table 4), were tested against B. cepacia biofilm that was pre-formed on the stainless-steel coupons. Overall, significant biofilm reduction was observed in all treatment groups compared to the untreated control. One of these combinations was α-amylase and ciprofloxacin at 625 μg/mL and at 4 μg/mL, respectively (Figure 5A). The B. cepacia biofilm on stainless-steel coupons significantly decreased (p < 0.0001) from 8.4 ± 0.2 log₁₀ CFU/coupon to 6.03 ± 0.2, 5.3 ± 0.3, and 4.5 ± 0.4 log₁₀ CFU/coupon with α-amylase, ciprofloxacin, and a combination thereof, respectively (Figure 5A). In the combination of α-amylase (1250 μg/mL) and meropenem (4 μg/mL), B. cepacia biofilm on stainless-steel coupons significantly decreased (p < 0.0001) from 7.5 ± 0.5 log₁₀ CFU/coupon to 5.8 ± 0.2, 5.6 ± 0.2, and 3.8 ± 1.0 log₁₀ CFU/coupon with α-amylase, meropenem, and a combination of both agents, respectively (Figure 5B). Hence, the last combination significantly (p < 0.0001) degraded the pre-existing biofilm as compared to the untreated control. When proteinase K (62.5 μg/mL) and ciprofloxacin (8 μg/mL) were tested, the count of the preformed biofilm decreased significantly (p < 0.0001), from 7.7 ± 0.2 log₁₀ CFU/coupon to 6.0 ± 0.3, 6.3 ± 0.3, and 5.3 ± 0.6 log₁₀ CFU/coupon with ciprofloxacin, proteinase K, and their combination, respectively (Figure 5C). Overall, α-amylase in combination with either meropenem or ciprofloxacin demonstrated greater efficacy in reducing preformed B. cepacia biofilms, compared to combinations with proteinase K.

The scanning electron microscope was used to assess the morphological changes of B. cepacia biofilms formed on stainless-steel surfaces, in response to promising treatments deduced from the checkerboard assay (Table 4). Uninoculated stainless-steel coupons (Figure 6A) had a generally smooth surface with minor cracks and irregularities, showing the baseline topography before bacterial colonization. In untreated control samples (Figure 6B), B. cepacia established dense, well-developed biofilms. Cells were embedded in a thick EPS matrix that firmly adhered to the substrate. Morphologically, cells within the EPS appeared as short rods, whereas those cells outside the matrix were longer rods. Bacterial morphology changed significantly after treatment with ciprofloxacin alone (Figure 6C). The cells were elongated and filamentous, indicating impaired septation and potential interference with DNA replication and cell division (Wickens et al., 2000). Biofilm samples treated with α-amylase alone (625 μg/mL; Figure 6D) showed a noticeable reduction in biofilm biomass and EPS matrix. The cells appeared dispersed and loosely attached to the surface, lacking the cohesive structure typical of mature biofilms. The combination of α-amylase and ciprofloxacin (Figure 6E) produced a synergistic effect consistent with the previous results (Figure 3). Even though individual cells were present, cell density was significantly reduced, and there was no EPS or structured biofilm. The synergism between the enzymatic degradation of EPS and increased antimicrobial penetration probably contributed to the improved antimicrobial efficacy. The second combination to which biofilm was exposed was α-amylase (1250 μg/mL) and meropenem (4 μg/mL) (Table 4; Figure 6F). Meropenem treatment caused morphological changes in cells, such as V-shaped cell arrangements. Exposure to the α-amylase alone had the same effect as previously mentioned; there was no EPS production or thick biofilm mass (Figure 6G). However, samples treated with α-amylase and meropenem together showed visible morphological changes (Figure 6H). The bacterial population seemed to have clearly decreased, and the remaining cells showed deformed shapes, including shrinking and spheroplast-like appearances.