Targeting outer membrane protein A (OmpA) – inhibitory effect of 2′-hydroxychalcone derivatives on Acinetobacter baumannii and Candida albicans dual-species biofilm formation

Abstract Biofilm production facilitates microbial colonization of wounds and catheters. Acinetobacter baumannii produces high levels of biofilm and causes difficult-to-treat nosocomial infections. Candida albicans is another strong biofilm producer which may facilitate A. baumannii adhesion by providing hyphae-mediated OmpA-binding sites. Here we tested the potential of 2′-hydroxychalcones to inhibit dual-species biofilm production of A. baumannii and Candida spp., and further predicted the mechanism of structure-related difference in activity. The results suggest that 2′-hydroxychalcones exhibit potent activity against Candida spp./A. baumannii dual-species biofilm production. Particularly active was trifluoromethyl-substituted derivative (p-CF3), which decreased C. albicans/A. baumannii biomass produced on vein-indwelling parts of the central venous catheterization set by up to 99%. Further, higher OmpA-binding affinity was also calculated for p-CF3, which together with demonstrated significant ompA-downregulating activity, suggests that superior antibiofilm activity of this chalcone against the tested dual-species community of A. baumannii is mediated through the OmpA. Graphical Abstract


Introduction
Acinetobacter baumannii is a nosocomial pathogen characterized by an extreme level of antimicrobial resistance, due to which infections caused by it are very hard to treat. Most commonly it causes chronic wound and device-related infections. These infections are induced by the high biofilm production, which facilitates colonization of wounds and medical devices such as central venous catheter (Alves et al. 2016;Roy et al. 2022). Further, the biofilm plays a crucial role in the wound pathogenesis and significantly hampers eradication of the pathogen. By equipping the embedded microbial cells with several additional mechanisms of tolerance and resistance to host immune responses and antimicrobial agents, it interferes with wound healing or causes defective delayed wound closure (Sen 2019). Also, the biofilm greatly contributes to the ability of A. baumannii to acquire new resistance mechanisms and to persist in healthcare environments and cause hospital outbreaks (Roy et al. 2022).
Further, wound infections are typically polymicrobial in nature, consisting predominantly of the members of skin microbiota and environmental contaminants (Kalan and Grice 2018;Klein et al. 2018). Therein, multiple species can grow together and interact with each other inside a common biofilm scaffold, often adding to the biofilm recalcitrance and occasionally resulting in synergistic morbidity of the infection (Stacy et al. 2016). Also, increasing evidence suggests that polymicrobial communities contribute to the rapid emergence of more resistant strains, due to the high rates of gene exchange enabled by the specific biofilm composition and high cell densities (Orazi and O'Toole 2019). Candida albicans is a prominent member of skin microbiota, frequently implicated in mixed fungal-bacterial biofilm-associated infections of wounds. Biofilm formation induces transition of C. albicans yeast cells to hyphae which contain multiple binding sites for various bacteria (Kalan and Grice 2018;Padmavathi et al. 2020). It was shown that, among others, A. baumannii can attach to C. albicans hyphae by exploiting outer membrane protein A (OmpA) as a binding ligand (Gaddy et al. 2009). In this way, C. albicans may promote colonization of wounds by A. baumannii. Further, C. albicans hyphae can actively penetrate through the epithelial barrier and therefore facilitate epithelial invasion of the attached bacteria in the form of hyphal passengers (Kalan and Grice 2018). In addition, C. albicans is also present as a widespread hospital environmental contaminant, frequently colonizing medical devices (Privett et al. 2010). Thus, it may facilitate A. baumannii concomitant colonization by providing OmpA-mediated hyphal adhesion sites, resulting in resistant polymicrobial devicerelated infections with worse outcomes (Giles et al. 2018). On the other hand, some studies have shown that when in direct contact, A. baumannii and C. albicans act antagonistically. A. baumannii can kill hyphae (but not yeast), following the attachment, and inhibit yeast-to-hyphae transition, whereas C. albicans quorum-sensing molecule farnesol can disrupt membrane integrity and impair virulence, biofilm production and motility of A. baumannii (Dhamgaye et al. 2016;Kostoulias et al. 2016). However, considering that biofilms are dynamic structures (Yang et al. 2011) it is likely that A. baumannii may utilize C. albicans hyphae for the promotion of its own virulence before killing it, thus preventing C. albicans to secrete the toxic molecules in self-defense.
Chalcones are compounds with well-described antibiofilm properties (Farhadi et al. 2019), which may be isolated as metabolic products from certain plants, or obtained by the process of total synthesis (Zhuang et al. 2017). In this work, two 2 0 -hydroxychalcones with previously described significant antibiofilm activity against A. baumannii (U sjak et al. 2019) were evaluated for the effects on mixed biofilm communities of A. baumannii with Candida species, and the mechanisms of their antibiofilm activity were further elucidated. In addition, docking analysis was performed to identify key structural entities contributing to the biofilm inhibiting activity of the 2 0 -hydroxychalcones.

Microbial strains
A. baumannii type strain ATCC 19606 and 10 wound isolates obtained from Clinical Hospital Center Dr. Dragi sa Mi sovi c were used in this study. Species identification was performed by VITEK V R 2 (GN ID) and further by Fourier transform infrared (FTIR) spectroscopy (Sousa et al. 2014). All A. baumannii isolates were determined as extensively drug-resistant employing both VITEK V R 2 (GN 222) system and broth microdilution method (CLSI 2015). Also, type strain of C. albicans ATCC 10231 was used, as well as clinical isolates of C. albicans (sputum) and Candida krusei (urine), which were identified by VITEK V R 2 (YST ID).

Growth inhibition testing
Growth inhibition assays were conducted to determine concentration limits of chalcones that do not affect growth of the included strains. Minimum inhibitory concentrations (MICs) were determined by the broth-microdilution method (CLSI 2015). Mueller-Hinton broth and Sabouraud dextrose broth were used as growth media for bacteria and yeasts, respectively. Chalcones were tested over the concentration range of 25-400 mg/mL. Bacteria and yeasts were seeded at approximately 10 6 and 10 7 CFU/mL, respectively, and MICs were determined after 24 h of incubation at 37 C. A range of sub-MICs were then tested to identify the highest concentration that does not significantly affect the growth of any of the included strains. The strains were incubated in the tryptic soy broth supplemented with 1% (w/v) glucose (TSBG) (medium subsequently used for the biofilm cultivation), and the growth was monitored by measuring optical density (OD) at 600 nm.

Biofilm production assays
Assays were performed on polystyrene surface of the 96-well microtiter plates (Sarstedt, N€ umbrecht, Germany). Treatment was carried out by adding chalcones at 35 or 70 lg/mL to TSBG before incubation.

Mono-species biofilm
A. baumannii, C. albicans or C. krusei biofilm production was tested as described by Stepanovi c et al. (2007). Following the 24 h incubation at 37 C in TSBG, unbound cells were washed three times with phosphate-buffered saline (PBS) and the remaining mass was fixed by air-drying at 60 C for 1 h. The biomass was then stained with 0.5% (w/v) safranin for 15 min and the optical density of ethanol-extracted dye was measured at 490 nm using EZ Read 400 Microplate Reader (Biochrom, Cambridge, UK).

Dual-species biofilm
Dual-species biofilm production assay was performed as described (Raorane et al. 2019) with some modifications. TSBG was used as the growth medium and the 24 h-biofilms were prepared as described above for the spectrophotometric quantification at 490 nm. Combinations of reference (C. albicans ATCC 10231/A. baumannii ATCC 19606) and clinical strains (each A. baumannii wound isolate with C. albicans or C. krusei clinical isolates) were tested separately. During the inoculation, Candida spp. was seeded 1 h before the addition of A. baumannii to allow yeast-tohyphae transition.

Dual-species biofilm cell viability assay
The viability of biofilm cells was estimated by using 2,3,5-triphenyl-2H-tetrazolium chloride (TTC) (Sigma-Aldrich, St Louis, MO, USA) according to the method described by Sabaeifard et al. (2014). C. albicans ATCC 10231/A. baumannii ATCC 19606 dualspecies biofilm, obtained and treated as indicated above, was washed with PBS and further treated by adding 0.5% (w/v) TTC:TSBG, at ratio 1:5. Following the incubation in dark for 6 h at 37 C, ODS were measured at 405 nm.

Catheter dual-species biofilm production
Vein-indwelling parts (guidewire, tissue dilator and central venous catheter) of the Double-Lumen Central Venous Catheterization Kit (Arrow International, Reading, PA, USA) were cut into multiple pieces, scrubbed with soapy water, rinsed three times with hot tap and distilled water and then autoclaved at 121 C/15 min. Sterilized pieces were then placed into tubes filled with 5 mL of TSBG containing C. albicans ATCC 10231/A. baumannii ATCC 19606, and incubated for 24 h at 37 C, with shaking at 180 rpm. The chalcones were added to the media at 35 or 70 lg/mL, prior to incubation. Washing, fixation and staining were performed as described above. The amount of produced biofilm was estimated using the modified biofilm index, based on the digital image analysis for color brightness quantification in ImageJ version 1.53t (NIH) (Yamamoto et al. 2008).
The following conditions were applied: 95 C/2 min and 40 cycles of 95 C/5 s and 60 C/32 s, and the normalization was done using the 2 -DDCt method (Livak and Schmittgen 2001).

Statistical analysis
The experiments were performed in triplicate and the results are presented as mean values ± standard deviations. Comparisons were performed using ANOVA with Tukey's post hoc test and p values less than .05 were considered as statistically significant. Analyses were performed using SPSS version 26.0 (IBM).

Antimicrobial activity of the chalcones
MICs of chalcones against A. baumannii strains were determined at 100-125 mg/mL. Candida spp. strains proved to be more resistant as their growth was inhibited only at 200-250 mg/mL (Supplementary Table 1). There was no significant difference in activity between two chalcones. All strains were then separately incubated in the TSBG treated with chalcones at series of concentrations below 100 mg/mL. Based on the measured growth at 24 h-interval (Supplementary  Tables 2 and 3), 70 mg/mL was identified as the single common highest concentration of both chalcones that does not significantly modify the growth of any of the included strains. This concentration was subsequently used for the biofilm inhibition assays, together with its two-fold dilution, 35 mg/mL, to test the dosedependency.

Inhibition of dual-species biofilm formation by the chalcones
Both chalcones significantly inhibited mono-species biofilm production by A. baumannii wound isolates at 70 mg/mL (Figure 1a). In average, p-CF 3 induced greater inhibition, but also a higher inter-strain variation. Candida spp. mono-species biofilm production, however, was not significantly affected by the chalcones (Figure 1b-d). When A. baumannii and C. albicans strains were cultivated together, the biofilm mass that was produced was by 33% higher in average compared to the sum of the amounts of their singlespecies biofilm counterparts (Figure 1e and f). This was not the case when C. krusei was used instead of C. albicans (Figure 1g). As with the A. baumannii single-species biofilm production, the addition of the chalcones resulted in significant inhibition of its dualspecies biofilm production with Candida spp. More specifically, p-CF 3 was notably more active against C. albicans/A. baumannii, whereas o-OCH 3 gave better results against C. krusei/A. baumannii biofilm production. Also, inter-strain variation between different A. baumannii wound isolates cultivated with C. albicans sputum isolate was much lower with p-CF 3 . On contrary, biofilm cell viability of C. albicans/A. baumannii was more affected by o-OCH 3 (Figure 1h).
The chalcones were even more active against the production of C. albicans/A. baumannii dual-species biofilm on different vein-indwelling parts of the central venous catheterization set ( Figure 2). Particularly, p-CF 3 decreased biofilm formation on the guidewire and dilator by up to 99%. The formation was also significantly reduced on the central venous catheter, however without noticeable difference between two chalcones. Interestingly, it was observed, based on the color change, that chalcones strongly adhered to the central venous catheter, in contrast to guidewire and dilator.

Interference between the chalcones and A. baumannii OmpA
The results obtained from quantitative real-time PCR suggest that tested chalcones significantly inhibit the expression of A. baumannii OmpA in mixed-culture with C. albicans (Figure 3). There was no observable difference between the effects of two chalcones at 35 mg/mL, but at 70 mg/mL o-OCH 3 was noticeably more active. The docking analysis further predicted that chalcones may bind with high affinity to one of the essential sites of OmpA, reportedly with Ala91 and Val137 pi-alkyl and Asp46 pi-anion as key interactions. Moreover, binding affinity of p-CF 3 was notably better (docking score ¼ À8.9 kcal/mol), due to the additional pi-alkyl and carbon-hydrogen interactions with Phe134 and Gly136, respectively, provided by the trifluoromethyl functional group (Figure 4).
Binding affinity of o-OCH 3 was calculated at À7.5 kcal/mol, which is even weaker compared to the non-substituted 2 0 -hydroxychalcone (docking score-¼ À7.7 kcal/mol), suggesting that methoxy group only added to steric hindrance at the particular site.

Cytotoxicity of the chalcones
The chalcones manifested cytotoxic activity toward epithelial and epidermal cells ( Supplementary Figures  1 and 2). Lower cytotoxicity was demonstrated for p-CF 3 since over 40% keratinocytes survived the exposure to 35 mg/mL. In all other instances, about 20% keratinocytes/epithelial cells survived.

Discussion
The production of biofilm is a key feature that enables A. baumannii persistence in hospital environment (Roy et al. 2022). A. baumannii adheres to various surfaces, including non-living material like hospital furniture, equipment and medical devices, and also human tissues (Peleg et al. 2008). Prior colonization of surfaces by C. albicans may even enhance A. baumannii adherence ability, by providing additional hyphae-mediated binding sites, and result in a dual-species biofilm structure that is even more resilient (Gaddy et al. 2009;Giles et al. 2018). Previously, it was shown that 2 0 -hydroxychalcones significantly inhibit biofilm production of A. baumannii and that the inhibition may be mediated through OmpA interference (U sjak et al. 2021). OmpA is a porin-forming outer membrane protein involved in multiple virulence determinants of A. baumannii, including the biofilm formation (Nie et al. 2020). It was shown that the ompA-deficient A. baumannii produces significantly less biofilm on polystyrene surface. Most probably, OmpA stimulates biofilm formation by affecting the initial attachment, since it promotes the adhesion to both non-living, and living surfaces such as human epithelial cells and extracellular matrix proteins (Gaddy et al. 2009;Smani et al. 2012). Notably, it also promotes the adhesion to C. albicans hyphae, by binding to hyphal Hyr1 protein (Hyr1p) (Uppuluri et al. 2018). Therefore, we assumed that the chalcones could significantly affect C. albicans/A. baumannii dual-species biofilm formation. Besides, Messier and Grenier (2011) showed that licochalcone A, which is a 4 0 -hydroxychalcone derivative, exhibit strong inhibitory effect on yeast-tohyphae transition of C. albicans. This way, by reducing the number of hyphal forms in the biofilm, hydroxychalcones might additionally decrease the number of bacterial adhesion sites.
In this work, the effects of two synthetic 2 0 -hydroxychalcone derivatives with previously demonstrated antibiofilm activity against A. baumannii (U sjak et al. 2019) were tested. The latter activity was confirmed against 10 newly isolated, wound-originating A. baumannii strains. Variable strain-specific responsiveness was however evident at 70 mg/mL, especially with  p-CF 3 in case of which the biofilm production of four isolates was much less affected compared to the other six. Strain-specific effects of antimicrobial compounds against A. baumannii biofilm production were documented before (Navidifar et al. 2019) and are probably caused by high genomic plasticity of this bacterial species (Colquhoun and Rather 2020). In contrast to A. baumannii, the biofilm production of either of the included Candida species was not affected, suggesting the existence of selectivity in antibiofilm activity of the chalcones. Yet, the dual-species biofilm production of Candida species with A. baumannii strains was significantly reduced by both tested chalcones. Particularly strong activity, interestingly with relatively low inter-strain variation, was displayed by p-CF 3 against the biofilm production of putatively synergistic mixed-species community of A. baumannii and C. albicans. On the other hand, p-CF 3 had lower activity against the dual-species biofilm production of A. baumannii and C. krusei, the community which based on the total produced biomass indicated no signs of synergy. The lack of synergy is not surprising, since C. krusei does not produce regular hyphae (Jamiu et al. 2021) and to date there is no described A. baumannii binding protein expressed by this fungal pathogen. Interestingly, the viability of mixed A. baumannii and C. albicans biofilm cells was also less affected by p-CF 3 . The reduction of viability is significant since the biofilm is populated by large number of viable but non-culturable cells, which being more resistant to stresses of various origin, together with the dormant cells, may repopulate the biofilm upon damage, and prolong its lifespan (Wang et al. 2021). Altogether, the p-CF 3 proved to be superior in reducing the level of C. albicans/A. baumannii total produced dual-species biomass, and this could be due to the specific activity against the synergydriven, hyphal-binding protein OmpA. To test this possibility, gene expression analysis was performed, which showed that the expression of the ompA is indeed significantly downregulated in chalcone-treated A. baumannii cultivated together with C. albicans. The level of the ompA downregulation was even greater compared to the previously documented downregulation achieved by the chalcones against A. baumannii in the form of single-species culture (U sjak et al. 2021). This difference may be due to the interference with putative ompA gene expressioninducing effect of C. albicans Hyr1p. The extent of downregulation was however greater with o-OCH 3 . On the other hand, computational analysis predicted that tested chalcones could bind to the extracellular site of OmpA, thus blocking the subsequent interaction with Hyr1p, and greater binding affinity was calculated for p-CF 3 . Theoretically, the ompA downregulation may have been triggered by the negative feedback response to binding of chalcones to OmpA, but at this point, it is not possible to predict why o-OCH 3 induced greater downregulation. In overall, these findings suggest that the inhibitory activity of p-CF 3 against C. albicans/A. baumannii dual-species biofilm production may be more potent due to the overwhelming OmpA-Hyr1p interaction blocking effect. This could also explain why C. krusei/A. baumannii dual-species biofilm, where no interactions between OmpA and fungal proteins are known, was less affected by p-CF 3 . Comparable advantage of p-CF 3 was also demonstrated through lower cytotoxic activity. Keratinocytes proved to be more resistant toward p-CF 3 in comparison to epithelial cells. Nonetheless, the results of cytotoxic assay suggest avoiding the eventual application to dermis-penetrating deeper wounds, associated with a heavier damage in the epithelium.
Notably, it was detected that tested 2 0 -hydroxychalcones significantly reduced C. albicans/A. baumannii dual-species biofilm on vein-indwelling parts of the central venous catheterization set: guidewire, dilator and catheter. Tested guidewire and dilator are made of nickel-titanium alloy (nitinol) and polyether-polyamide block copolymer (PEBAX), respectively, which are less prone to biofilm formation (Louie et al. 2006;Song et al. 2016). PEBAX was even shown to be an effective antibiofouling coating for reverse osmosis membrane, by creating a smoother hydrophilic surface (Louie et al. 2006). However, the results of the current work proved that C. albicans in conjunction with A. baumannii can still form significant amounts of biofilm on these materials, as well. The chalcones though strongly stimulated their antibiofouling properties. The biomass that was formed in the presence of p-CF 3 was barely visible and calculated as almost 99% reduction. This is important since the guidewire and dilator are scarcely investigated for biofilm production, although it is obvious that the pathogens can be introduced through these parts as well, and simple chalcone-based coating could practically prevent the colonization. Catheter that was investigated, on the other hand, is made of polyurethane, a material commonly used to build catheters due to the favorable mechanical properties and stability (Wildgruber et al. 2016), but at the same time highly prone to biofilm formation (Lim et al. 2022). Although the chalcones exhibited potent antibiofilm activity, C. albicans/A. baumannii mixed culture was still able to produce significant amounts of biofilm on the catheter. Interestingly, however, it could be observed based on the post-incubation color change that the chalcones strongly adhered to the polyurethane surface of the catheter, which indicates that they could be easily utilized, with some improvements, to design coating materials.
Altogether, in this work it was demonstrated that 2 0 -hydroxychalcones possess potent activity against C. albicans/A. baumannii dual-species biofilm production. Greater activity was exhibited by the trifluoromethyl-substituted derivative (p-CF 3 ) probably due to more avid binding to OmpA, enabled by additional pi-alkyl and carbon-hydrogen interactions. Moreover, this binding putatively triggered the ompA gene downregulation, which in addition to direct blocking diminished biofilm-associated function of the OmpA. The viability of biofilm cells was also significantly inhibited by the chalcones. Finally, the chalcones significantly reduced C. albicans/A. baumannii biofilm production on vein-indwelling parts of the central venous catheterization set, especially on the guidewire and dilator, strongly stimulating antibiofouling properties of their surfaces. The biofilm production was also significantly reduced on polyurethane surface of the central venous catheter, on which it was observed that in addition to the microbes the chalcones also strongly bound, indicating the potential for coating design.

Disclosure statement
No potential conflict of interest was reported by the authors.