Inhibition of Staphylococcus epidermidis Biofilm Formation by a Synthetic Breitfussin Analogue

Abstract

Bacterial biofilms pose a major public health challenge by increasing the antimicrobial tolerance in pathogenic bacteria, thereby limiting the effect of medication-based treatment and promoting the development of antimicrobial resistance. Hence, there is a need to discover new molecules with the ability to prevent biofilm formation. We screened seven synthetic analogues of the breitfussin family of natural products for antimicrobial and antibiofilm activity using a broth microdilution and crystal violet method, respectively. Two compounds inhibited the growth of Gram-positive bacteria in their planktonic state at concentrations of 50 µM, of which one compound (2) demonstrated the ability to inhibit the biofilm formation of Staphylococcus epidermidis at sub-growth-inhibitory, low micromolar concentrations. Compound 2 did not inhibit biofilm growth in Staphylococcus aureus or Listeria monocytogenes, or the ability to eradicate pre-established biofilms. Initial Mode of Action (MoA) studies with compound 2 against S. epidermidis showed a modest impact on the cell surface hydrophobicity and early-stage adhesion to polystyrene. These findings highlight the breitfussin framework as a promising scaffold for the development of new antimicrobial and antibiofilm agents.

1. Introduction

Bacterial biofilms are aggregations of bacteria encapsulated in a self-produced extracellular matrix. They can form both on surfaces (adherent) and in suspension (non-adherent) and may consist of single or multiple bacterial species [1,2]. Biofilms are a major concern in clinical settings, because of their relevance to chronic infection and the colonization of medical devices [3,4]. Within the biofilm, bacteria can tolerate up to 1000-fold higher concentrations of antimicrobial agents compared to their planktonic counterparts. This tolerance poses a challenge for medication-based treatment and promotes the development of antimicrobial resistance [5]. Consequently, there is a pressing need for new compounds that not only inhibit bacterial growth but also specifically disrupt the biofilm lifestyle, thereby enhancing the effectiveness of both immune clearance and antibiotic treatments.

Potential drug targets may be found over the course of biofilm development, which can be described in three main stages: aggregation and attachment, growth and accumulation, and disaggregation and detachment [2]. Initially, bacteria adhere to surfaces or each other via non-specific (e.g., van der Waals) forces and specific adhesion (e.g., using fimbriae). Next, they produce an extracellular matrix of proteins, extracellular DNA (eDNA), and/or polysaccharide intercellular adhesins (PIAs). As the biofilm matures, nutrient and oxygen gradients create metabolic heterogenicity among cells. Eventually, parts of the biofilm disperse to colonize new sites, a process regulated by second messengers like cyclic diguanylate monophosphate (c-di-GMP) [6] and quorum-sensing molecules [7].

Nature has historically been an important source of new antimicrobial compounds, offering great structural and functional diversity [8]. Importantly, many of these natural products exhibit antibiofilm activity. For example, geraniol, a monoterpene from lemongrass, not only inhibits bacterial growth but also disrupts biofilm formation by interfering with membrane function and quorum sensing [9]. Similar activity profiles are found for natural products and natural product derivates currently in clinical use as antibiotics, including daptomycin [10] and azithromycin [11], respectively. Beyond individual compounds, complex extracts, often of plants, have demonstrated activity against biofilms in clinical studies [12,13,14]. While terrestrial ecosystems have been extensively explored for antimicrobial compounds, the marine environment represents an underexplored reservoir of natural products [15,16]. Nonetheless, molecules with biofilm inhibitory activity have been isolated from marine sources, e.g., the meridianin group of compounds isolated from the tunicate Aplidium meridianum [17] and a group of brominated indole alkaloid secondary metabolites from the bryozoan Flustra foliacea [18]. Natural products often possess novel molecular scaffolds with promising antimicrobial activities that can be further optimized by synthetic derivatization [19]. To develop the most effective possible molecule at the end of the optimization process, not only must the activity against a target be improved, but also the pharmacological properties concerning absorption, distribution, metabolism, and excretion (ADME). If these properties are sub-optimal, a molecule is unable to reach its target in sufficient doses to have a therapeutic effect [20].

The breitfussins are a group of natural products characterized by a halogenated indole–oxazole–pyrrole core, originally isolated from the Arctic hydrozoan Thuiaria breitfussi [21]. Hansen et al. [22] have reported their ability to act as kinase inhibitors, and Liu [23] has reported the ability of breitfussin B to inhibit bacterial growth, highlighting different bioactivity within this molecular scaffold. Based on this knowledge, we screened seven synthetic breitfussin analogues for their ability to inhibit bacterial and biofilm formation and investigated the possible modes of action for one molecule with promising biofilm inhibiting activity.

Here, we report the synthesis and biological evaluation of a series of breitfussin-inspired analogues (compounds 1–7, Figure 1). These analogues were screened for antibacterial, antibiofilm, and antifungal activity. Compounds 1 and 2 showed growth-inhibiting activity against Gram-positive bacteria. Upon further investigation, compound 2 was found to inhibit the formation of S. epidermidis biofilm. To explore its therapeutic potential, we also assessed the in vitro toxicity and selected pharmacological properties. To the best of our knowledge, this is the first study investigating anti-biofilm activity in an indole–oxazole–pyrrole molecular framework.

2. Materials and Methods

2.1. Synthesis of Breitfussin Analogues 1–7

Full experimental details and analytical data for the synthesis of compounds 1–7 are given in Appendix A, as well as electronic Supplementary Files. Structures were identified and confirmed using ¹H- and ¹³C-NMR.

2.2. Microorganism Strains and Growth Conditions

Enterococcus faecalis (ATCC 29212) and Streptococcus agalactiae (ATCC 12386) were grown and assayed in brain heart infusion broth (BHI, #53286, Merck KGaA, Darmstadt, Germany). Escherichia coli (ATCC 25922), Pseudomonas aeruginosa (ATCC 27853), and Staphylococcus aureus (ATCC 25923) were grown and assayed in Mueller Hinton broth (MH, #275730, BD, Franklin Lakes, NJ, USA). Staphylococcus epidermidis RP62A (ATCC 35984) and Staphylococcus haemolyticus (clinical isolate, provided by the Host–Microbe Interaction research group UiT) were grown and assayed in tryptic soy broth (TSB, #22092, Merck KGaA) and tryptic soy broth supplemented with 1% (w/v) glucose (TSBg). Candida albicans (ATCC 90028) was grown on potato dextrose agar (#70139, Merck KGaA) and assayed in RPMI-1640 medium (#R7755, Merck KGaA) supplemented with 3-(N-morpholino)propanesulfonic acid (#M3183, Merck KGaA) and L-glutamine (#X0551, Biowest, Nuaillé, France).

2.3. Bacterial Growth Inhibition

Bacterial growth inhibition was tested according to a previously described broth microdilution protocol [24]. In brief, bacterial suspensions in exponential phase were pipetted into a 96-well plate at 1500 CFU/well to 15,000 CFU/well, after which test compounds diluted in MilliQ^TM water (mqH₂O) were added at twice the assay concentration. If compound solubility did not permit preparation of 2× dilutions in mqH₂O, compounds were diluted directly in a 1:1 mixture of bacterial suspension and mqH₂O, of which 100 µL were added to each test well. After ~20 h incubation at 37 °C, the absorbance at 600 nm (Abs₆₀₀) was measured using a VICTOR3^TM 1420 Multilabel plate reader (PerkinElmer, Waltham, MA, USA). As a negative control, a 1:1 mixture of mqH₂O and bacterial suspension was used, and a 1:1 mixture of mqH₂O and sterile medium was used as blank. A gentamycin (#A2712, Biochrom GmbH, Berlin, Germany) dilution series was used as positive assay control for both Gram-positive and Gram-negative microbes. The screening was conducted with at least two biological replicates with two technical replicates each, except for compound 1, where limited availability allowed only one biological replicate. Against S. epidermidis, biological triplicates with technical triplicates were used. The minimum inhibitory concentration (MIC) was defined as the concentration at which no visible bacterial growth occurred.

2.4. Fungal Growth Inhibition

The antifungal activity of the test compounds was assayed using Candida albicans in a broth microdilution method as previously described [25]. Briefly, C. albicans freeze stock was spread onto a potato dextrose agar plate and grown overnight at 37 °C, after which a colony was transferred to 0.9% NaCl and cell density adjusted to 5 × 10⁵ CFU/mL using a Remel^TM 0.5 Mc Farland standard. This suspension was diluted 1:1000 in assay medium and 100 µL of it transferred to a 96-well plate where 100 µL sample at twice the assay concentration in mqH₂O had been previously added. Abs₆₀₀ was measured using a VICTOR3^TM 1420 Multilabel plate reader (PerkinElmer) before the incubation start, after 24 h of incubation, and after 48 h of incubation at 37 °C. The assay was conducted once with three technical replicates. Amphotericin B (#A2942, Merck KGaA) at a concentration of 8 µg/mL was used as positive control.

2.5. Inhibition of Biofilm Formation

Inhibition of biofilm formation of S. epidermidis was tested using a crystal violet method described previously [24]. Briefly, overnight culture was diluted 1:100 in TSBg and pipetted into a 96-well plate (TC-treated, #734-2327, VWR Avantor, Radnor, PA, USA) where bacteria would be exposed to the test compounds. Assay concentrations of the test compounds were 10 µM and 50 µM for the initial screening and a range from 80 µM to 0.625 µM for the dose response testing of 2. After incubation at 37 °C for approximately 20 h, the liquid and the planktonic bacteria were removed and the remaining biofilms rinsed with tap water. Once the biofilms had been dried for 1 h at 65 °C, they were stained with 70 µL of a 0.1% (w/v) crystal violet (#V5265, Merck KGaA) solution for 5 min, before removing the crystal violet solution and rinsing the wells with tap water. Finally, the crystal violet that was left bound to the biofilm was re-dissolved with 70 µL of 70% (v/v) EtOH and the absorbance was measured at 570 nm using a Spark^® Multimode Microplate reader (TECAN, Männedorf, Switzerland). The initial screening was conducted using two biological replicates with three technical replicates each, while the dose–response testing was conducted using three biological replicates with three technical replicates each. As growth control, S. epidermidis that was not exposed to any test compound was used, S. haemolyticus a non-biofilm forming strain was used as a negative control, and medium without bacteria was used as blank control.

2.6. Eradication of Established Biofilms

Eradication of established S. epidermidis biofilm was tested using essentially the same approach as described above, with the only difference being that the biofilms were grown unexposed to test compounds in medium for 20 h, after which the planktonic bacteria were removed and the wells washed with sterile phosphate-buffered saline (PBS), and the medium with test compound was added. After incubation for 24 h, the biofilms were stained as described above, with the only difference being a change in the wavelength used to 490 nm, which was made because biofilm growth during the 2nd incubation period caused absorbance values to rise above the measurement limit of the plate reader. The eradication assay was conducted in three independent biological replicates with three technical replicates each.

2.7. Inhibition of Biofilm Formation in Additional Species

Biofilm assays against L. monocytogenes EGD-e, methicillin-resistant Staphylococcus aureus (MRSA) MB5393, and S. aureus ATCC29213 were conducted as follows: Each well of a sterile, flat-bottomed, 96-well polystyrene micro-well-plate was loaded with 10 µL antibiotics (vancomycin for both S. aureus strains, benzalkonium for L. monocytogenes), dimethyl sulfoxide (DMSO), or stock solution of the test compound using a Biomek i7 automated liquid handler (Beckman Coulter, Indianapolis, IN, USA). A volume of 90 µL medium (cation-adjusted Mueller Hinton II broth for S. aureus ATCC29213, BHI for MRSA, and TSB + 5% yeast extract for L. monocytogenes) containing the bacterial inoculum at 5 × 10⁷ CFU/well was added using Multidrop^TM Combi (Thermo Scientific, Waltham, MA, USA). The plates were incubated for 24 h at 37 °C in static conditions inside a plastic bag. The total biomass (planktonic and biofilm) was monitored by determination of the OD₆₁₂ before and after incubation using an EnVision^® multilabel plate reader (PerkinElmer). After incubation, plates were washed thrice with 200 µL sterile PBS 1×, after which 100 µL of resazurin with a final concentration of 40 µg/mL was added. The mixture was then incubated at 37 °C for 2 h to 3 h (S. aureus ATCC29213) or 4 h to 5 h (MRSA and L. monocytogenes). After incubation, fluorescence was measured with an excitation wavelength of 570 nm and an emission wavelength of 615 nm. Bacterial growth inhibition against L. monocytogenes EGD-e, MRSA MB5393, and S. aureus ATCC29213 was assessed based on the total biomass measurement. Two independent replicates of the assay were conducted.

2.8. Cell Surface Hydrophobicity

An alteration in cell surface hydrophobicity (CSH) may impact the formation of bacterial biofilm. To assess if this was the case in our study, CSH was measured based on an adaption of the bacterial adherence to hydrocarbons assay that was originally described by Rosenberg et al. [26,27]. Overnight cultures of S. epidermidis grown in TSB at 37 °C and 100 rpm were pelleted by centrifugation at 4300× g for 5 min. The pellets were then washed twice with phosphate urea magnesium sulfate (PUM) buffer [28] and resuspended in PUM to an absorbance between 0.4 and 0.6. From this bacterial suspension, 1 mL was transferred into an acid-washed test tube (Ø 13 mm). To this, 250 µL n-hexadecane (#A10322.AE, Thermo Scientific) was added, and (when testing the impacts of short exposure) 1 µL of test compound in a DMSO stock in the appropriate concentration (or the same volume of DMSO as vehicle control). After incubation of this mixture for 15 min at 37 °C, each test tube was vortexed for 2 min at max. speed and left to settle for 20 min to allow for phase separation between PUM buffer and n-hexadecane. Once the 20 min had passed, absorbance of the aqueous phase was measured. Absorbance measurements were conducted at 405 nm by transferring 100 µL of the aqueous phase into a microtiter plate which was read using a VICTOR3^TM 1420 multilabel plate reader (PerkinElmer). Based on the measurements of absorbance before (Abs₀) and after (Abs) addition of hexadecane and test compound, the change in surface hydrophobicity was calculated according to the formula %H = [(Abs₀ − Abs)/Abs₀] × 100%.

To test the impact of long exposure to compound 2 on CSH of S. epidermidis, the overnight culture was prepared in medium containing the appropriate concentration of test compound and vehicle control.

Both long and short exposure were tested in three biological replicates.

2.9. Bacterial Adhesion to Polystyrene

The impact of compound 2 on the early-stage adhesion of bacteria to polystyrene was tested as follows: Overnight cultures of S. epidermidis grown in TSB at 100 rpm and 37 °C were pelleted by centrifugation at 4300× g for 5 min, after which the cells were washed and resuspended to an Abs₆₀₀ of 0.275 in sterile PBS. Of this cell dilution, 50 µL were transferred to each well of a 96-well microtiter plate (TC-treated, #734-2327, VWR Avantor) containing 50 µL of the sample/control at 2× concentration in PBS. The test plate was incubated for 2 h at 37 °C before the bacterial adhesion was quantified with crystal violet as described above. Trypsin (#X0930, Biowest) (0.06%) was used as positive control, while DMSO was used as vehicle control. The assay was conducted in biological triplicates with six technical replicates each.

2.10. ADME Testing

The ADME properties of compound 2 were assessed as described previously [22]. The relevant method descriptions are reproduced here. The LC-MS analyses were conducted using a setup consisting of an Acquity I-class UPLC (Waters, Milford, MA, USA) coupled to a PDA detector and a Vion IMS QToF (Waters), using an Acquity BEH C18 UPLC column (1.7 µM, 2.1 mm × 100 mm) (Waters). The system was controlled and the data was processed using UNIFI 1.9.4 (Waters). Spectrophotometric measurements were conducted using a Spark^® Multimode Microplate reader (TECAN). ADME testing was conducted using three technical replicates.

2.10.1. Kinetic Solubility Assay

The aqueous solubility of compounds 2 and 3 was determined spectrophotometrically by measuring the kinetic solubility of a 500 μM compound solution in aqueous buffer (pH 7.4) after 90 min of vigorous shaking at room temperature, using a corresponding solution in acetonitrile as reference.

2.10.2. Parallel Artificial Membrane Permeability Assay

Parallel artificial membrane permeability (PAMPA) was assessed using an initial compound concentration of 500 μM in the donor compartment. After an incubation period of 20 h, the absorption of the receiver wells was measured by spectrophotometry, and permeation was calculated by normalization of the compound flux to that obtained across a blank filter.

2.10.3. Microsomal Stability Phase I Assay

Metabolic stability under oxidative conditions was measured in mouse liver microsomes (CD-1 male, pooled, #M1000, Xenotech (now Bio IVT, Westbury, NY, USA)) by LC-MS-based measuring of depletion of compound at a concentration of 3 μM over time up to 50 min at 37 °C. On the basis of compound half-life t_1/2, in vitro intrinsic clearance (Cl_int) was calculated: Cl_int = (V × 0.693)/(t_1/2 × mg).

2.10.4. Microsomal Stability Phase II Assay

Metabolic stability under conjugative conditions was measured in a glucuronidation assay by LC-MS-based determination of % remaining of selected compounds at a concentration of 5 μM in incubations with liver microsomes supplemented with UDP-glucuronic acid for 1 h at 37 °C.

2.10.5. Plasma Stability Assay

Plasma stability was measured by LC-MS-based determination of % remaining of selected compounds at a concentration of 5 μM after incubation in 100% mouse plasma for 1 h at 37 °C.

2.11. Antiproliferative Activity

The antiproliferative effects were determined against the human cell lines MRC5 (nonmalignant human lung fibroblasts, ATCC CCL-171, LGC Standards, Teddington, UK) and HepG2 (human hepatoblastoma, ATCC HB-8065, LGC Standards) using the MTS assay. HepG2 cells were cultured and assayed in MEM Earle’s (#F0325, Biochrom GmbH) and MRC-5 cells in MEM Eagle (#M7278, Merck KGaA) medium, both supplemented with 10% Fetal Bovine Serum (#S1810, Biowest), 10 mg/L gentamicin (#A2712, Biowest), 2 mM glutamine stable (#X0551, Biowest), non-essential amino acids (#K0293, Biochrom GmbH), and 1 mM sodium pyruvate (#L0473, Biochrom GmbH). All cells were incubated in 5% CO₂ at 37 °C. When assayed, cells were seeded in 96-well microtiter plates at 2000 cells/well (MRC-5) or 20,000 cells/well (HepG2). After 24 h, both cell lines were treated with various concentrations of compound 2 and incubated for 72 h (MRC-5) or 24 h (HepG2). DMSO at the highest concentration used in the treatment wells was used as vehicle control. As positive control, 10% DMSO was used. After the incubation period, 10 μL of MTS solution (Cell Titer 96^® AQueous One Solution Reagent, #G358B, Promega, Madison, WI, USA) was added to each well, and the cells were incubated for 1 h at 37 °C. The absorbance was measured at 490 nm using a Spark^® Multimode Microplate reader (TECAN). Cell viability was calculated as follows: cell survival (%) = [(absorbance treated wells − absorbance positive control)/(absorbance negative control − absorbance positive control)] × 100. From this, the half-maximal inhibitory concentration (IC₅₀) was calculated. The assays were conducted in three biological replicates.

2.12. Statistical Analysis

All statistical analyses were performed with GraphPad Prism version 11.0.2 for Windows (GraphPad Software, Boston, MA, USA, www.graphpad.com). The concentration at which 50% of planktonic growth was inhibited (MIC₅₀), the concentration at which 50% of biofilm growth was inhibited (MBIC₅₀), and the IC₅₀ values were calculated using the [Inhibitor] vs. normalized response—Variable slope method provided by GraphPad Prism.

3. Results

3.1. Synthesis of Breitfussin Analogues

In this work, two synthesis strategies were employed toward the synthesis of breitfussin analogues 1–7. Compounds 1–5, which contain substituted indoles, were prepared using the same strategy that Hansen et al. (2019) and Pandey et al. (2015) reported for the synthesis of breitfussin A–D (Scheme 1) [22,29].

For compounds 6 and 7 (Scheme 2), we used the method reported by Xiang et al. [30], which allowed for the one-pot preparation of compound 7 and intermediate 8 from 3-acetylindole by a reaction with either phenyl glycine or glycine, respectively.

Intermediate 8 was TIPS-protected at the indole and the resulting 9 was submitted to the previously described conditions for the iodination of the oxazole using 3 equivalents of LiHMDS and 1.2 equivalents of iodine. Interestingly, performing the iodination at −78 °C gave a 1:3 mixture of 10 and the 4-iodo isomer with a 53% yield alongside 14% of the 2,4-diiodinated analogue. By treating 9 with LiHMDS at −78 °C followed by the addition of a room-tempered solution of iodine and allowing the temperature to rise from −78 °C to −45 °C during the iodination step, 10 could be obtained in 73% isolated yield alongside 4% of the 2,4-diiodinated analogue. By increasing the temperature to −42 °C before adding iodine and allowing the reaction mixture to attain room temperature over 1.5 h, the 2-iodinated indolyl-oxazole is generated exclusively with an 87% yield. We had initially planned to prepare 6 through selective coupling on the 2,4-diiodinated oxazole derivative followed by acid-mediated de-iodination as reported previously [22,29], but found that the coupling of 10 with N-Boc-pyrrol-2-boronic acid gave 11 with a satisfactory yield of 60% after 72 h at room temperature. Finally, the removal of the Boc- and TIPS- protecting groups using TFA and TBAF, respectively, gave compound 6 with an excellent yield.

3.2. Antimicrobial Screening

The test compounds were screened for antimicrobial activity by evaluating their ability to inhibit the growth of four Gram-positive and two Gram-negative bacterial strains, as well as the fungal pathogen C. albicans. In this initial screening, two compounds inhibited the planktonic growth of at least one Gram-positive strain at 50 µM or below. Activity was observed against S. aureus ATCC 25923, S. agalactiae, E. faecalis, and S. epidermidis. No activity was observed against Gram-negative bacteria (E. coli and P. aeruginosa) or C. albicans. The results of the initial screening as well as subsequent investigations are gathered in

Table 1. Overview of the MIC values of the compounds that showed activity in at least one strain and test concentration (n = 3 for compound 2, with exception of strains marked with asterisk where n = 2 and C. albicans where n = 1; n = 1 for compound 1, due to compound availability limitations; positive controls were gentamycin for bacterial growth inhibition and amphotericin B for fungal growth inhibition).

	Compound
	1	2
Gram-positive bacteria
S. aureus ATCC 25923	50 µM	10 µM
S. aureus ATCC 29213 1		>50 µM
S. agalactiae	10 µM	10 µM
E. faecalis	50 µM	>50 µM
S. epidermidis		20 µM
MRSA MB5393 1		>50 µM
Gram-negative bacteria
P. aeruginosa	>50 µM	>50 µM
E. coli	>50 µM	>50 µM
Fungal pathogen
C. albicans	>50 µM	>50 µM
Biofilm
S. epidermidis		10 µM
S. aureus ATCC29213 1		>50 µM
MRSA MB5393 1		>50 µM
L. monocytogenes EGD-e 1		>50 µM

¹ These strains were tested using a different method by co-authors, as described in Section 2.7. Benzalkonium and vancomycin were used as controls.

for the active compounds and in Table S1 for all tested compounds.

Based on the results of this screening and compound availability considerations, compound 2 was chosen for further investigation.

3.3. Antibiofilm Testing

To narrow down the possible modes of action, compound 2 was evaluated in more detail for its ability to either inhibit biofilm formation or to disrupt the already established biofilm in S. epidermidis. As shown in Figure 2, treatment with compound 2 resulted in a reduction in biofilm formation, with an MBIC₅₀ of 2.5 µM compared to an MIC₅₀ of 13.9 µM. At 10 µM, the biofilm growth was not significantly (p < 0.05; one-way ANOVA with Dunnett’s multiple comparisons test) different from those observed for a non-biofilm-forming S. haemolyticus strain, while the planktonic growth remained at 80% of that observed in untreated controls (Figure 2). In contrast, the effect on pre-established biofilms was limited. Even at the highest concentration tested (80 µM), the disruption was modest and unlikely to be of practical relevance (see Figure S1). The ability of compound 2 to inhibit biofilm formation at sub-MIC concentrations, but not to disrupt established biofilms, suggests that its mechanism of action is associated with the early stages of biofilm development, prior to the formation of a robust extracellular matrix.

In addition to the assays that were conducted using S. epidermidis, the potential of compound 2 to inhibit biofilm growth and eradicate established biofilms was tested in a metabolic assay against L. monocytogenes, MRSA, and S. aureus ATCC29213. In this screening, compound 2 did not show any activity in inhibiting biofilm formation or disrupting the established biofilm at concentrations up to 50 µM, suggesting a species-specific activity. Likewise, no inhibition of the planktonic growth of L. monocytogenes, MRSA, and S. aureus ATCC29213 was observed either. Due to the different measurement principles used for these strains as compared to S. epidermidis, caution must be taken when comparing the results. However, it can still be concluded that there is no inhibition of biofilm formation at the test concentration in the additional strains.

3.4. Cell Surface Hydrophobicity

The attachment and adhesion of bacterial cells, which are the initial steps of biofilm formation, increased with a higher cell surface hydrophobicity. Therefore, an alteration in the bacterial cell surface hydrophobicity might be a reason for decreased biofilm formation [31]. In Gram-positive bacteria, major contributors to surface hydrophobicity are autolysins such as AtlE or teichoic acids [32,33]. For example, lipoteichoic acid in S. aureus affects both cell surface hydrophobicity and biofilm formation on polystyrene [32]. In the present study, the cell surface hydrophobicity of S. epidermidis was marginally increased after the exposure to compound 2 at 20 µM for 15 min and 18 h (see Figure 3A).

3.5. Adhesion

The initial attachment of staphylococci to polystyrene is mediated by a combination of non-specific interactions (e.g., hydrophobic interactions) [34] and surface proteins such as AtlE [35], SdrF [36], and Aap [37]. Similar to the results observed when testing the impact on cell surface hydrophobicity, exposure to compound 2 led to a slight but non-significant (p > 0.05) increase in the early-stage adhesion of S. epidermidis to the surface of polystyrene tissue plates (Figure 3B). In comparison to this, exposure to trypsin, which digests adhesion-related surface proteins [38], led to a strong decrease in early-stage adhesion.

3.6. ADMET

To gain insight into the drug likeness of the compounds showing antimicrobial activity, their ADMET properties were evaluated using a panel of assays assessing the kinetic solubility, permeability through artificial membranes, metabolic stability, plasma stability, and antiproliferative activity against MRC5 and HepG2 cells. The results of these tests are summarized in

Table 2. ADMET properties of compound 2 (MS = microsomal stability, n = 3 for cell-based results).

	Cpd. 2	Target Value	Unit
Kinetic Solubility 1	85	>100	µM
PAMPA	17	>25	% flux
Mouse MS phase I Clint	5	<30	µL/min/mg
Mouse MS phase II	31	>80	% remaining after 1 h
Mouse plasma stability	69	>85	% remaining after 1 h
MRC5 IC50	15	>25	µM
HepG2 IC50	30	>25	µM

¹ The solubility of compound 3 was determined as 35 µM.

. Limited aqueous solubility can limit drug delivery options and complicate in vitro testing. Compound 2 showed a kinetic solubility of 84.8 µM, indicating a moderate solubility below the desired target value. Most drugs are absorbed through passive membrane permeation, something that was modelled in this study by measuring the flux of compound 2 from a donor- to an acceptor-compartment through an artificial membrane. With a flux of 16.6% over the course of 20 h, compound 2 was able to pass membranes but at a slower pace than desired. Looking at the metabolic stability, compound 2 shows a high stability against phase I metabolizing enzymes while the stability against phase II metabolism, specifically glucuronidation, is relatively low (31% remaining after 1 h). To attain sufficient plasma concentrations, it is important that a compound is stable over time in blood plasma. In our study, 31% of compound 2 was degraded in mouse plasma over 1 h. Lastly, the toxicity of compound 2 against HepG2 (human hepatoblastoma) and MRC5 (human lung fibroblast) was tested. When compared to the MBIC₅₀, there is a 12× window in selectivity against the hepatoblastoma cell line and a 6× window in activity against the lung cell line.

4. Discussion

The majority of hospital-acquired infections are associated with bacterial biofilms, and biofilm formation is especially relevant in infections related to implanted medical devices. Among the biofilm formers infecting medical devices, S. epidermidis is the most common pathogen [39,40]. In addition, S. epidermidis is the most common cause of (usually hospital-acquired) neonatal sepsis [39,41]. To effectively combat these biofilm-related infections, new antimicrobial strategies targeting not only bacterial survival but also disrupting the biofilm lifestyle are needed [42,43,44]. To avoid (or delay) resistance formation, it is preferable to not kill pathogens directly but instead target virulence factors [42,43]. To this end, natural products have often provided a good starting point for the development of new antimicrobial compounds [8]. In this study, the starting point is the indole–oxazole–pyrrole framework of the breitfussin group of compounds, of which seven synthetic derivatives were tested. Of these seven compounds, two were observed to inhibit the growth of at least one of the tested pathogens at 50 µM or lower, with compound 2 being the most potent of them. This is an improvement over the previously reported activity of breitfussin B against S. aureus, where the growth inhibition was reported at 1 mg/mL to 2 mg/mL (2.3 mM to 4.6 mM) [23]. Additionally, the biofilm formation of S. epidermidis was inhibited by compound 2 at sub-MIC concentrations. The higher activity observed in compounds 1 and 2 compared to the other compounds might be due to them both containing iodine. Compound 3, the only other iodinated compound, showed comparatively low solubility, which is likely to have hindered its activity. The increased biofilm activity observed in compound 2 might be related to the fused benzene ring in the indole moiety, as proposed by Yap et al. (2023) for structurally similar compounds inhibiting the biofilm formation of S. aureus [45].

The lifecycle of a biofilm is a tightly regulated process that relies on a variety of environmental factors, which provides a wide range of molecular pathways as possible targets for small molecules with antibiofilm activity. Because compound 2 did not disrupt the established biofilms, it is likely that the mechanism by which it inhibits biofilm formation is to be found in the early stages of biofilm formation.

In the early attachment stage, bacteria often rely on nonspecific interactions to attach to a surface, resulting in the interaction with surfaces showing properties resembling those of the bacteria themselves [31]. In the case of this study, where the attachment surface is hydrophobic polystyrene, a reduction in the cell surface hydrophobicity (which corresponds to an increase in hydrophilicity) might correspond to a reduction in biofilm formation [46]. However, we observed a marginal increase in the surface hydrophobicity of S. epidermidis cells, suggesting that the exposure to compound 2 might have a minor impact on surface-associated components involved in cell surface hydrophobicity. However, the magnitude of this change makes it questionable whether the impact is high enough to be biologically relevant. This is supported by the results of the initial adhesion assay which shows no significant change resulting from the exposure to compound 2 in the bacterial attachment to the surface of polystyrene culture plates after 2 h of incubation at 37 °C. Together, these results suggest that the reason for the decrease in biofilm formation is not related to the adhesion and attachment phase.

After the initial attachment to a surface, biofilm-forming bacteria begin production of an extracellular matrix that provides stability and protection from threats to the biofilm (antimicrobial treatment, immune response, etc.). Interrupting the formation of this matrix can significantly hamper the formation of a strong biofilm. Among the biofilm-forming species that were tested in this study, there are differences in the type of macromolecules found in the extracellular matrix. The S. epidermidis strain RP62A, whose biofilm formation was inhibited by compound 2, produces a matrix consisting almost entirely of polysaccharides [47]. In contrast, the matrices of the other tested strains are less polysaccharide-dependent with that of L. monocytogenes consisting up to roughly 80% of proteins [48] and the S. aureus matrix showing a 3:2 ratio between the protein and polysaccharide mass [49]. These differences in the extracellular matrix could be among the reasons for the differences in activity, as a compound with a mode of action targeting PIA production would likely have a stronger impact on the biofilm formation of the strongly PIA-producing S. epidermidis strain compared to the other two pathogens.

Altogether, the lack of activity against pathogens with less PIA-dependent matrices, as well as the low impact on CSH and early-stage adhesion, suggest that the mode of action of compound 2 involves the disruption of PIA production in S. epidermidis. This disruption of PIA production could also provide a possible explanation for the results observed in the biofilm disruption assay, as the existing matrix remains intact, but additional PIA production is hindered at higher concentrations of compound 2. In both S. aureus and S. epidermidis, PIA production is regulated by the icaADBC operon, which is mediated by a variety of regulatory factors, both direct or indirect [50,51]. Many of these regulatory factors are shared between both species, which suggests that, if one of the shared regulatory factors were the target, both strains would be affected, not only S. epidermidis. However, there are differences—for example, in the function of SarX, which promotes biofilm formation in an ica-dependent manner in S. epidermidis, while the same is not the case in S. aureus [50]. Therefore, if interference with PIA production is the mode of action of compound 2, it is likely due to one of the ica-regulating factors that do not share functionality between Staphylococcus species. This conclusion may be strengthened by future studies including assays specifically focusing on PIA production, and a wider range of bacterial species and S. epidermidis isolates.

We cannot fully rule out other mechanisms of antibiofilm activity, for example, interference with quorum sensing or c-di-GMP signaling. Furthermore, it is often the case that an increased tolerance of antimicrobials can be seen in mature biofilms compared to planktonic bacteria, which leads to the concentrations at which biofilm formation and planktonic growth are inhibited being far lower than the ones at which established biofilms are eradicated [52,53]. As a result of the maximal solubility being only approximately 4× higher than the MIC against the test strain S. epidermidis, we cannot rule out that a theoretical biofilm disrupting activity exists at concentrations above the solubility ceiling.

In addition to the bioactivity, we also investigated the drug-likeness of the most active compound to provide a starting point for possible future structure optimization. One well-established method to predict if a chemical will show favorable oral bioavailability is the “rule of 5” proposed by Lipinski in 1997 [54]. It states that compounds which exceed 5 H-bond donors, 10 H-bond acceptors, a molecular weight of 500, and a clogP of 5 are likely to exhibit poor oral absorption. None of these parameters are exceeded by compound 2 but more insight is gained through the tested ADMET properties where compound 2 shows promise, but also potential for improvement. The kinetic solubility and permeation through the artificial membrane of the PAMPA are lower than desired. Both could negatively affect the absorption of the compound in vivo [20]. Looking at predictors of metabolic stability, the intrinsic clearance rate of compound 2 by mouse liver microsomes is low, predicting an acceptable stability against phase I metabolism. However, the stability against phase II metabolism and degradation in plasma is not sufficiently high.

5. Conclusions

This study shows that compounds based on the indole–oxazole–pyrrole molecular framework have potential use as inhibitors of bacterial growth and biofilm formation. The synthetic analogues described in this study improve upon the reported activity of the natural products they are based on. Additionally, compound 2 shows strong biofilm-inhibiting activity at sub-MIC concentrations, possibly based on the interference with PIA production. These activities are favorable and the compound shows promising drug-like characteristics, but there is potential for the improvement of ADMET properties. Therefore, additional studies focusing on the improvement of both the efficacy and ADMET properties are warranted for the development of biofilm-inhibiting drugs or coatings based on this molecular framework.