Antibiofilm assay for antimicrobial peptides combating the sulfate‐reducing bacteria Desulfovibrio vulgaris

Abstract In medical, environmental, and industrial processes, the accumulation of bacteria in biofilms can disrupt many processes. Antimicrobial peptides (AMPs) are receiving increasing attention in the development of new substances to avoid or reduce biofilm formation. There is a lack of parallel testing of the effect against biofilms in this area, as well as in the testing of other antibiofilm agents. In this paper, a high‐throughput screening was developed for the analysis of the antibiofilm activity of AMPs, differentiated into inhibition and removal of a biofilm. The sulfate‐reducing bacterium Desulfovibrio vulgaris was used as a model organism. D. vulgaris represents an undesirable bacterium, which is considered one of the major triggers of microbiologically influenced corrosion. The application of a 96‐well plate and steel rivets as a growth surface realizes real‐life conditions and at the same time establishes a flexible, simple, fast, and cost‐effective assay. All peptides tested in this study demonstrated antibiofilm activity, although these peptides should be individually selected depending on the addressed aim. For biofilm inhibition, the peptide DASamP1 is the most suitable, with a sustained effect for up to 21 days. The preferred peptides for biofilm removal are S6L3‐33, in regard to bacteria reduction, and Bactenecin, regarding total biomass reduction.


| INTRODUCTION
The majority of microorganisms live as aggregates forming a biofilm . Thereby, biofouling describes the phenomenon of undesirable biofilms (Flemming, 2002). Here, it can lead to numerous disruptions of processes, either in the medical field and human health due to the colonization of implants (usually titanium) or medical devices (usually stainless steel) through numerous multi-resistant bacteria like ESKAPE pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species) and their associated potential for infection (Vetrivel et al., 2021), or in industrial plants such as water circuits or the oil and gas industry (Flemming, 2002). In industrial processes, biocorrosion bacteria have an important role in attacking metals through various mechanisms (Chugh et al., 2020). In this context, sulfate-reducing bacteria (SRB) and sulfur-oxidizing bacteria (SOB) should be mentioned in particular, as they can cause immense corrosion damage by forming a sulfur cycle (Chugh et al., 2020).
The problem in combating biofilms is that in contrast to the planktonic mode, sessile cells are better protected through their and high cytotoxicity (Alhajjar et al., 2022;McLaughlin et al., 2021), the application of biocides is limited, and the development of new antimicrobial and antibiofilm candidates is crucial. In recent years, antimicrobial peptides (AMPs) have emerged as a new treatment strategy (Luo et al., 2023). Thereby, AMPs can lead to cell death via electrostatic interaction with the bacterial membrane and subsequent incorporation into the membrane and formation of pores (membranolytic mode of action) (Zhang et al., 2021). In addition, AMPs can inhibit intracellular metabolic pathways (Zhang et al., 2021). In addition to killing cells directly, AMPs can act on any state of biofilm formation, including regulation of the cell signaling pathway by the AMP LL-37 (Overhage et al., 2008) or reduction of EPS by targeting a major polysaccharide of EPS matrix by the AMP hepcidin 20 (Brancatisano et al., 2014). Although there are numerous modes of action of AMPs against biofilms, up to this time, only the effectiveness against the planktonic cells of biocorrosion bacteria has been shown (Stillger et al., 2023).
Analysis of the complete biofilm is not possible due to its complex composition. Nevertheless, various analytical methods (colorimetric, microscopic, etc.) and different assays (BioFilm Ring Test, flow chamber, etc.) are available to characterize the different biofilm aspects (Azeredo et al., 2017). A limitation of these methods is often insufficient throughput, which is important for the development of new compounds.
To overcome this, the Calgary Biofilm Device, a 96-well biofilm assay, has been developed. Ceri et al. (1999). The optimization of this assay, with the use of microcentrifuge tubes as growth surfaces in the respective wells, ensures flexibility to the respective conditions, as well as high throughput at the same time. Additionally, this method is more economical than the Calgary Biofilm Device and can be implemented with common laboratory equipment (Reiferth et al., 2022). It can be used to determine the minimum biofilm inhibitory concentration (MBIC), as well as the minimum biofilm eradication concentration (MBEC). Therefore, the biofilm is stained by crystal violet, and in the case of MBEC, viable microorganisms are measured through a recovery plate.
This study aims to establish this assay to determine the MBIC and MBEC on biocorrosive bacteria and quantify the activity of AMPs in a high-throughput method. For the first time, the application of AMPs against a representative of SRB, Desulfovibrio vulgaris, for inhibition and eradication of biofilm is presented. The utilization of steel rivets as growth surface ensures realistic material conditions.

| Synthesis of peptides
The peptides were synthesized with 9-fluorenylmethoxycarbonyl chemistry using the microwave-assisted peptide synthesizer LibertyBlue TM (CEM GmbH) in 0.1 mM scale. Peptides were checked for purity (reversed-phase high-performance liquid chromatography) and identity (liquid chromatography-mass spectrometry). For details regarding the above procedures see Stillger et al. (2023). The peptides with associated sequences used in this study are listed in Table 1.

| Stains and cultivation
D. vulgaris DSM 644 was cultivated in Postgate medium (DSMZ medium 63) with a modification of 0.004 g/L FeSO 4 × 7H 2 O and 0.3 g/L trisodium citrate. The medium was flushed with nitrogen gas.
Incubation was performed at 37°C without shaking.
The process was repeated a second time before sterilization using a UV lamp for 1 h. After 30 min, the rivets were rotated to ensure uniform exposure. The bacterial culture was subcultured to a starting OD600 of 0.007. A measure of 120 µL of bacterial culture per well was added as growth check (n = 14), 100 µL bacterial culture per well for AMP test (n = 3 per AMP concentration), and 120 µL medium per well as sterile control (n = 4) to a 96-microtiter plate (polystyrene).
Subsequently, each well was fitted with a rivet. Preincubation for 6 h was done. Afterward, 20 µL of AMP solution (sequential twofold dilution of 1 mg/mL to 1.95 µg/mL) was added to the corresponding well (in Figure A2, a pipetting scheme is shown). After a further incubation time of 66 h, biomass in the solution (OD600) and the biofilm formation were characterized by crystal violet staining. For crystal violet staining, the rivets were removed and washed with 0.9% NaCl. Subsequently, the biofilm was fixed using 99% methanol (incubation 5 min) and dried for 60 min. Subsequently, staining was performed using 0.5% crystal violet (incubation 30 min), washing 3× T A B L E 1 Collection of antimicrobial peptides used in this study with their corresponding sequence.

| MBEC assay
The preparation of the MBEC assay was similar to the MBIC. A measure of 120 µL of bacterial suspension was added per well as growth check (n = 8) or as AMP test (n = 6 per AMP concentration) and 120 µL of medium as sterile control (n = 8). The microtiter plate was incubated for 72 h. Afterward, the rivets were transferred to the challenge plate consisting of 120 µL of AMP solution (1 mg/mL to 1.95 µg/mL), 120 µL of medium for growth check (n = 8), 120 µL of medium (n = 4) or 120 µL of water (n = 2) or 120 µL of AMP with a concentration of 1 mg/mL (n = 2) for sterile control (in Figure A3, a pipetting scheme is shown). Incubation was performed for 24 h. The rivets were then washed with 0.9% NaCl and half of them were stained with crystal violet (see MBIC). The other half was transferred to the recovery plate containing 120 µL of medium per well and placed in the ultrasonic bath for 30 min.
The rivets were removed and the microtiter plate was incubated for 48 h. After the incubation period of 48 h, OD600 was measured. An overview of the individual steps of the MBIC/MBEC assay is shown in Figure 1.

| Adaptation of the growth surface for D. vulgaris
Based on Table 2, the OD570 of biofilms on PCR tubes indicated a low concentration of biomass with a high standard deviation of 50%.
When steel rivets were used, the OD570 increased significantly.
Additionally, OD570 increased with increasing pretreatment-OD570 of the polished rivets (3) higher than roughened rivets (2) higher than mechanically untreated rivets (1). The polished steel rivets showed the highest concentration of biofilm with an OD570 of 1.30 simultaneously with a low variation. For the polished rivets, a constant biofilm over the entire contact surface between the rivet and the bacterial suspension was observed. This was in distinct contrast to the untreated rivets, where the biofilm was mainly confined to the edge (see Figure A4).

| Biofilm inhibition through AMPs
The peptides L5K5W-W4I6 and Bactenecin showed no biofilm inhibitory effects. The peptide S6L3-33 had a MBIC95% of 1000 µg/mL. The peptide DASamP1 had the lowest minimum biofilm inhibitory concentration with a MBIC95% of 250 µg/mL ( Figure 2). These results were consistent for both OD600 and OD570.

| Long-term effectiveness of biofilm inhibition through AMPs
DASamP1 was selected for long-term study based on its best activity in the previous MBIC assay. For both the uninhibited samples (black circles) and the growth check without rivet (black triangles), an increase in

| Biofilm eradication through AMPs
S6L3-33 was detected as the most effective peptide in regard to OD600 with a MBEC95% of 125 µg/mL. L5K5W-W4I6 and DASamP1 reached each a MBEC95% of 250 µg/mL. Bactenecin exhibited a 95% reduction of OD600 up to 500 µg/mL and a partial STILLGER ET AL.
| 3 of 16 effect could be detected at two subsequent concentration levels (250 and 125 µg/mL). For the removal of the total biofilm mass (OD570), only the peptide Bactenecin was able to contribute. The concentration curve is comparable to the measurement at OD600. Up to a peptide concentration of 500 µg/mL, complete removal of the biofilm was observed. For the concentrations 250 and 125 µg/mL, a partial removal could be detected ( Figure 5).

| DISCUSSION
For the establishment of the Calgary assay modified by Reiferth et al. (2022) for the SRB D. vulgaris, all growth-critical steps were performed in an anaerobic chamber, and only the crystal violet staining was performed under aerobic conditions. All required fluids were flushed with N 2 . The selection of growth medium and F I G U R E 1 Overview of the individual steps of MBIC and MBEC assay for Desulfovibrio vulgaris.
temperature was based on the optimal conditions for the planktonic growth of D. vulgaris. In naturally occurring biocorrosion, such as in water cycles, the nutrient supply is very different from the medium used here, and the temperature could reach higher or lower values.
The analysis under realistic conditions may be considered in further experiments. In addition to transferring the assay to anaerobic conditions, the incubation periods of D. vulgaris were adjusted. This adjustment was oriented to the growth curve of Zhang et al. (2016), where the mean exponential phase of sessile growth was reached after an incubation period of 72 h. A modification of the material's adhesion surface is necessary due to very low and heterogeneous biofilm formation on previously used polypropylene tubes (see Table 2 This is not necessary in this case, since the significance of the OD600 is not affected. The peptide DASamP1 is the most effective against biofilm inhibition (MBIC95 of 250 µg/mL) and is only slightly lower compared to its planktonic effectiveness against D. vulgaris. In further studies, this peptide was also able to inhibit an early biofilm stage of multiresistant Staphylococcus aureus (Menousek et al., 2012) and consequently has great potential as a short and "simple" peptide for inhibiting biofilms. The peptide structure of DASamP1 suggests a membranolytic mode of action (Menousek et al., 2012). However, the exact mode of action, in particular the effective inhibition of a biofilm, is not yet known. A key factor in preventing biofilm formation by corrosive bacteria is to maintain the effect of the peptides for as long as possible. Due to the good effectiveness of DASamP1, this peptide was selected for long-term observation. Both MBIC (250 µg/mL) and double MBIC (500 µg/mL) were used as peptide concentrations.
From Day 10 onwards, bacterial growth and biofilm formation could be detected using the MBIC. This is due to the decreasing stability of the peptide in the bacterial supernatant (see Figure 4) (500 µg/mL) was effective over the entire period of 21 days. Halving the peptide concentration over time resulted in a concentration of around 250 µg/mL, which is equivalent to MBIC and thus still effective. To achieve the longest possible biofilm inhibition, it is necessary to either apply a higher concentration than the real MBIC to buffer peptide degradation or to avoid peptide degradation by using D-amino acids, nonnatural amino acids, or cyclization while maintaining peptide activity (Gentilucci et al., 2010;Molhoek et al., 2011). However, a longer observation was not further possible with this setup due to starting evaporations. Instead, the long-term study should be performed in a closed system. Hereby, it was possible to show the principal biofilm inhibition of different peptides after a certain period (72 h) as well as their long-term efficacy, while applying the same materials. Therefore, different aims can be targeted with this assay material. The other three tested peptides showed a lower biofilm inhibitory activity. In the case of S6L3-33, its amphiphilic structure suggests a membranolytic mode of action (He et al., 2007). The same is true for L5K5W, so it is assumed that the modified version L5K5W-W4I6 acts identically to the original peptide (Kang et al., 2009;Kim et al., 2013). Bactenecin, which is linear due to reducing conditions of the medium, also has a membranolytic effect, and barrel stave is assumed in this context (Wu & Hancock, 1999). All tested peptides showed a significant loss of activity regarding biofilm inhibition compared to their effectiveness against planktonic cells. This can be explained by the time lag of peptide addition in the MBIC assay compared to MIC determination and the associated challenges of biofilm structure. In particular, the EPS matrix acts as a diffusion barrier, and therefore higher concentrations are often necessary to combat biofilms (Dunsing et al., 2019;Saleh et al., 2021). Why the peptide DASamP1 is best able to overcome these challenges is currently unexplored and more detailed analyses of biofilm mechanism are needed.
In the MBEC assay, all four peptides showed killing activity against the bacteria, based on the OD600 measurement of the recovery plate. However, as this is only a turbidity measurement, where bacterial fragments can also be considered, this method is flawed. A more accurate method is the determination of cell number, which is a more complex method and therefore not considered for high throughput. For the reduction of the total biomass, only the peptide Bactenecin was successful. The remarkably strong antibiofilm activity of Bactenecin, specifically in eradicating total biofilm mass, may be related to its structure. Bactenecin has a cyclic form with a hydrophobic ring and a hydrophilic tail in an oxidative milieu due to the presence of two cysteines (Yari et al., 2019). This characteristic structure enables surfactant-like interactions. Thereby, surface-active peptides promote the degradation and detachment of biofilms by reducing surface tension (Banaschewski et al., 2015;Mandal et al., 2013). Especially, the MBEC assay indicates that different substances have different points of interaction. For efficient removal of the biofilm, it is necessary to develop a mixture of different peptides, each with different biofilm target points.
In this study, the successful application of AMPs, both to inhibit a biofilm and to remove an existing biofilm, was shown against D.
vulgaris, a representative of SRB, which is considered a major agent against biocorrosion. Therefore, a modified Calgary assay was implemented. This assay allows realistic conditions through the use of metal rivets and is flexible depending on the required conditions, allowing different bacteria (also anaerobic) and different growth materials to be used. Thus, this assay can be used to screen new (a) (b) F I G U R E 3 Long-term analysis for the peptide DASamP1 against Desulfovibrio vulgaris: OD600 (a) and OD570 (b) over time (days) for uninhibited bacteria culture (black circle), growth check without rivets (black triangles), and inhibited bacteria culture with a peptide concentration of 0.25 mg/mL (light gray) and 0.5 mg/mL (dark gray); mean and standard deviation were determined by the propagation of error.
F I G U R E 4 Long-term stability for the peptide DASamP1 in Postgate C medium (black) and in the supernatant of a 7-day-old culture of Desulfovibrio vulgaris (gray), analyzed with reversed-phase high-performance liquid chromatography; 100% peptide content corresponds with the peptide content at 0 h. Values are shown as means with standard deviation, n = 3.
F I G U R E 5 Minimum biofilm eradication concentration (MBEC) assay for Desulfovibrio vulgaris: OD600 of the recovery plate (48 h incubation) and OD570 of the challenge plate (72 h incubation) for different concentrations of four antimicrobial peptides: L5K5W-W4I6, S6L3-33, Bactenecin, DASamP1 (black striped), and uninhibited sample (gray); mean and standard deviation were determined by error propagation; asterisk indicates significant reduction (p = 0.005).
biofilm-active substances fast, simply, and cost-efficiently. In addition to the determination of the MBIC and the MBEC, this assay can be used for further analyses such as long-term studies, and thus enables multiple possibilities with a single setup.

CONFLICT OF INTEREST STATEMENT
None declared.

DATA AVAILABILITY STATEMENT
The data generated and analyzed during this study are included in this published article.

ETHICS STATEMENT
None required.