The Effect of Selected Bee Products on Adhesion and Biofilm of Clostridioides difficile Strains Belonging to Different Ribotypes

There is an ongoing search for alternative treatments for Clostridioides difficile infections. The aim of the study was to investigate the antibacterial and antibiotic activity of bee products against C. difficile strains with different polymerase chain reaction ribotypes (RTs). The minimum inhibitory concentration (MICs) of Manuka honey 550+, goldenrod honey, pine honey, and bee bread were determined by the broth dilution method. C. difficile adhesion to HT-29, HT-29 MTX, and CCD 841 CoN cell lines was assessed. Biofilm was cultured in titration plates and visualized by confocal microscopy. The MICs of Manuka honey for C. difficile 630 and ATCC 9689 strains and control strain, M 120, were 6.25%, 6.25%, and 1.56% (v/v), respectively; of goldenrod honey, 50%, 50%, and 12.5%, respectively; of pine honey, 25%, 25%, and 25%, respectively; and of bee bread, 100 mg/L, 50 mg/L, and 100 mg/L, respectively. Manuka honey (1%) increased adhesion of C. difficile RT176 strains, and one strain of RT023, to the CCD 841 cell line. Pine honey (1%) increased RT027 adhesion to the HT-29 cell line. Manuka honey, pine honey, and bee bread at subinhibitory concentrations increased the adhesion of C. difficile. Our research proved that bee products are active against the tested strains of C. difficile.


Introduction
Clostridioides difficile (previously Clostridium difficile; C. difficile) is an anaerobic grampositive rod that produces highly resistant spores capable of spreading in the environment. Over the past four decades, C. difficile has become the leading cause of antibiotic-associated diarrhea worldwide, imposing a heavy burden on the healthcare system. The histopathological hallmark of C. difficile infection (CDI) is damage to the colonic epithelial cells with the accompanying acute inflammatory response and the formation of pseudo-membranes. C. difficile is a multi-drug resistant microorganism. Until recently, standard therapy for CDI included vancomycin and metronidazole. However, while active against vegetative organisms, these antibiotics do not effectively inhibit spore formation. Currently, fidaxomicin and vancomycin are preferred as standard of care [1]. Antibiotics such as ampicillin, amoxicillin, cephalosporins, clindamycin, and fluoroquinolones disrupt the healthy microflora in the digestive tract, which leads to proliferation of C. difficile [2].
Bacterial adhesion to the host's biological membranes is the first step in colonization and biofilm formation. It determines the long-term presence of pathogens in the body and their impact on human health. The ability of C. difficile to form biofilm has significantly contributed to the increasing antibiotic resistance [3,4]. In fact, according to Semenyuk et. al. [5], C. difficile biofilms cause a 100-fold increase in resistance to metronidazole [5]. Although little information has been published so far on the ability of C. difficile to form biofilm in vivo, it is well known that multicellular structures can potentially protect bacteria against cellular immune response and antibiotic activity [3,4]. Moreover, CDI relapse can be associated with the persistence of biofilm [4]. the MBC was 6.25%, 6.25%, and 3.125% (v/v), respectively. The MIC of goldenrod honey was 50%, 50%, and 12.5%, respectively; of pine honey, 25%, 25%, and 25%, respectively; and of bee bread, 100 mg/L, 50 mg/L, and 100 mg/L, respectively. Table 1. Minimum inhibitory concentrations (MIC) and minimum bactericidal concentrations (MBC) of Manuka honey 550+, goldenrod honey, pine honey, and bee bread on C. difficile strains. 6 The reference strains did not show MBC resistance values under the influence of goldenrod honey, pine honey, or bee bread. The MIC value of clinical strains with given RTs was different for the bee products tested and ranged from 1.56% to 6.25% for Manuka honey, from 1.56% to 50% for goldenrod honey, from 1.56% to 25% for pine honey, and from 25 mg/L to 100 mg/L for bee bread. The MBC values of clinical C. difficile strains for Manuka honey ranged from 12.5% to 3.125%. The MBC values for RT176 strains and one RT027 strain under the influence of goldenrod honey ranged from 50% to 25%, and under the influence of pine honey, from 50% to 12.5%. On the other hand, MBC values for RT023 strains and one RT027 strain were above 100% under the influence of goldenrod honey and pine honey. All clinical strains were resistant to bee bread, and the MBC value was above 200 mg/L (Table 1).

Effect of Bee Products on Adhesive Properties of C. difficile
The tested reference strains 630 and M 120 and clinical strains 1468 (RT027), 25694 (RT023), and 2628 (RT176) showed adhesion to the CCD 841, HT-29, and HT-29 MTX cell lines, and with addition of 1% of tested bee products. Manuka honey increased the adhesion of strain RT176 and strain RT023 to the CCD 841 cell line and that of strain RT023 to the HT-29 MTX cell line (p < 0.05). Pine honey increased the adhesion of strain RT027 to the HT-29 cell line (p < 0.05). Finally, bee bread increased the adhesion of strain M 120 to the HT-29 cell line (p < 0.05) (Figure 1).

Confocal Laser Scanning Microscopy
Reference strain 630 produced an irregular, heterogeneous biofilm with sparse and thin three-dimensional (3D) architecture. Under the subinhibitory concentration of Manuka honey 550+ (3.125%), the biofilm was regular and sparse, with a high 3D architecture containing microaggregates and significantly elongated plankton cells. Under the influence of a subinhibitory dose of goldenrod honey (25%), the biofilm was irregular and sparse, with high 3D architecture containing microaggregates and short plankton cells. Under the influence of a subinhibitory dose of pine honey (12.5%), the biofilm was irregular and heterogeneous, with a thick 3D architecture with very short, twisted plankton cells and tightly packed microaggregates in places. Under the influence of a subinhibitory concentration of bee bread (50%), the biofilm was regular and sparse, with a mediumhigh 3D architecture containing a small number of microaggregates and very elongated plankton cells.
The reference strain M 120 produced a homogeneous, regular biofilm with a dense architecture with microaggregates, composed of cells tightly arranged next to each other along the entire width. Under the subinhibitory concentration of Manuka 550+ honey (0.78%), the biofilm was homogeneous, dense, carpet-like, and composed of tightly packed and contiguously arranged microaggregates. Under the influence of a subinhibitory concentration of goldenrod honey (6.25%), the biofilm was very thick, regular, carpet-like, and consisted of tightly packed and contiguously arranged microaggregates. Under the subinhibitory concentration of pine herb honey (12.5%), the biofilm was very thick, regular, carpet-like, and consisted of tightly packed and contiguously arranged microaggregates with high 3D architecture, irregularly shaped, with significantly elongated plankton cells and microaggregates.
The biofilm of the clinical strain with RT027 (No 4. 1468/12) showed irregular, heterogeneous, thin, and sparse architecture of both microaggregates and cells. Under the influence of a subinhibitory concentration of Manuka 550+ honey (3.125%), the biofilm was thin, irregular, heterogeneous, and dense with the architecture of microaggregates. Under the influence of a subinhibitory concentration of goldenrod honey (6.25%), the biofilm was homogeneous, dense, carpet-like, but thin with tightly packed microaggregates. Under the influence of a subinhibitory dose of pine herb honey (1.56%), the biofilm was regular, dense, with a low 3D architecture containing tightly packed microaggregates. Under the influence of a subinhibitory concentration of bee bread (25%), the biofilm was homogeneous and thin, with large amounts of plankton cells.
The biofilm of the clinical strain RT023 (No 7. 25694/12) was irregular, heterogeneous, and sparse in both microaggregate and cell architecture. Under the influence of a subinhibitory dose of Manuka 550+ honey (3.125%), the biofilm was regular, with low 3D architecture containing microaggregates. Under a subinhibitory dose of goldenrod honey (25%), the biofilm was homogeneous, dense, and carpet-like, with tightly packed microaggregates and cells, but with a low 3D architecture. Under the influence of a subinhibitory dose of bee bread (12.5%), the biofilm was regular with microaggregates and very short plankton cells, with a thin 3D architecture.
The biofilm of the clinical strain RT176 (No 9. 2628/12) was irregular, heterogeneous, and sparse in the architecture of both microaggregates and cells. Under the influence of a subinhibitory dose of Manuka 550+ honey (1.5625%), the biofilm had a very thin 3D architecture containing few microaggregates and cells. Under the influence of a subinhibitory dose of goldenrod honey (1.5625%), the biofilm was homogeneous, dense, and carpet-like with lots of regular plankton cells, with low 3D architecture. Under the influence of a subinhibitory dose of pine herb honey (0.78%), the biofilm was very sparse, irregular, with a small number of microaggregates and plankton cells, with a very thin 3D architecture. Under the influence of a subinhibitory dose of bee bread (50%), the biofilm was regular, sparse, with many short plankton cells and with low 3D architecture ( Figure 2).

Discussion
Honey, which also stimulates inflammatory cytokines [47], has been identified as a potent superoxide anion scavenger and a highly effective reactive oxygen species inhibitor that is released from human neutrophils [38]. Evidence from published studies shows that honey destroys Pseudomonas aeruginosa cell walls [48] and disrupts cell division in methicillin-resistant S. aureus [49]. Honey has been shown to have a broad spectrum of antibacterial activity against many pathogens, including antibiotic-resistant organisms and their biofilms [50,51].
Both raw and sterilized (pasteurized or irradiated) honey showed the desired broad spectrum of activity against gram-positive and gram-negative bacteria, including clinically important pathogens such as Burkholderia cepacia [52], P. aeruginosa [31,53], Salmonella enterica, Serratia marcescens, E. coli, Klebsiella pneumoniae [54], S. aureus, Salmonella typhimurium, Shigella sonnei [55], and Streptococcus pyogenes [56]. In addition, several multidrug-resistant bacteria, including methicillin-resistant S. aureus and vancomycin-resistant Enterococci, were susceptible to honey [57]. Research on activity-related biomarkers has shown that hydrogen peroxide and methylglyoxal contribute significantly to the antibacterial activity of honey [58]. However, neither the action of hydrogen peroxide nor methylglyoxal can account for the total antibacterial activity of the honey since their removal did not completely abolish its cytotoxicity. It is an undeniable fact that almost all types of honey exhibit at least bacteriostatic activity, regardless of their botanical or geographical origin, which makes the antibacterial activity an inherent property of honey and a desirable source for drug development [59].
In the presented study, we did not observe antiadhesive action by the selected bee products. Instead, we observed an increase in adhesion under 1% concentration of bee products. It is hypothesized that this may be due to the low concentrations of tested substances. According to Mizzi et al., high sugar concentrations inhibit bacterial growth, while low concentrations stimulate bacterial growth. The same may apply to bacterial adhesion [60].
In our previous study, RT027 strains showed the highest potential for biofilm formation compared with other PCR-RTs [28]. C. difficile strains belonging to RT017, RT023, RT027, and RT046 were completely inhibited by Manuka honey MGO 400+ at a concentration of 6.25%. Both the MIC and MBC values were found to be 6.25%. The study demonstrated that Manuka honey MGO 400+ could inhibit biofilm formation in vitro in clinically significant strains of C. difficile, especially those belonging to RT027 [28]. Manuka honey, derived from the Manuka tree (Leptospermum scoparium), is known to exhibit antimicrobial properties that are associated with its significant content of methylglyoxal, a natural antibiotic. In our study, the MIC and MBC values for Manuka honey 550+ were 6.25% and were comparable with the values observed for strains with RT027 and RT023 in the previous study. For RT176, these values ranged from 1.56% to 12.5%. Under the influence of high concentrations (6.25-100%), Manuka Honey 550+ inhibited the formation of biofilm in strains belonging to RT027, RT023, and some RT176. Goldenrod honey, under the influence of 50% and 100% concentrations, inhibited the biofilm formation of clinical strains. Pine honeydew biofilm formation was significantly inhibited at all concentrations in the tested clinical strains, except for one strain RT176, RT023 and RT027. Pine herb honey also changed the appearance of plankton cells. Bee bread did not enhance or inhibit biofilm formation. Our findings confirm the previous literature [61][62][63].
The limitation of the presented study is a lack of information on the phytochemical composition of tested products.
The clinical strains of C. difficile were isolated from Polish hospitals. All C. difficile strains were obtained from the collection of the Anaerobic Laboratory of the Chair and Department of Medical Microbiology, Medical University of Warsaw, Poland. The strains were stored in the Microbank™ bacterial storage system (ProLab Diagnostics, Bromborough, Wirral, UK) at −70 • C. C. difficile strains 630 and M 120 were kindly provided as a gift by Professor Brendan Wren from the Department of Pathogen Molecular Biology at the London School of Hygiene and Tropical Medicine, London, UK. Strain ATCC 9689 was purchased from bioMérieux (Marcy l'Etoile, France).
The tested strains were thawed, plated on Columbia agar, and incubated at 37 • C for 48 h under anaerobic conditions. All isolates were cultivated in brain-heart infusion medium (BHI; Difco, Detroit, MI, USA) at 37 • C for 48 h under anaerobic conditions.

Bee Products
MGO™ Manuka honey 550+, certified to contain at least 550 mg/kg of methylglyoxal, was obtained from Manuka Health, New Zealand. Goldenrod honey (Solidago virgaurea), pine herb honey (Pinus sylvestris L.), and bee bread (a combination of flower pollen, honey, royal jelly, and bee enzymes) were obtained from an apiary in southern Poland.
All bee products were stored at an appropriate temperature and were used for research immediately after harvesting. Stock solutions were prepared by dissolving in BHI medium and then filtered using 0.2-µm syringe filters (Corning, Corning, NY, USA).

Determination of the Minimum Inhibitory and Minimum Bactericidal Concentrations of Bee Products for C. difficile Strains Belonging to Different PCR-RTs
The minimum inhibitory concentration (MIC) for the growth of bacteria and the minimum bactericidal concentration (MBC) for the tested C. difficile strains were assessed according to the methodology described previously [62]. A 96-well titration plate (Nunc, Roskilde, Denmark) was used for the study. The tested products were dissolved in the BHI medium at the concentrations of 100%, 50%, 25%, 12.5%, 6.25%, 3.125%, and 1.56%, while bee bread was at the concentrations of 200 mg/L, 100 mg/L, 50 mg/L, 25 mg/L, and 12.5 mg/L. Next, 20 mL of a 3.0 McFarland suspension of tested strain with 0.85% NaCl was added to the wells of the titration plate containing 180 µL of the tested concentrations of bee products and incubated at 37 • C for 48 h under anaerobic conditions. The positive control was the BHI medium with 20 µL of suspension of C. difficile adjusted to 3.0 McFarland, and the negative control was the BHI medium. All strains were tested in triplicate. Following incubation, optical density at 600 nm was measured using a microplate reader (Bio-Rad, Hercules, CA, USA). The MBC was determined by plating the cell suspensions in 96-well plates used for MIC tests onto Columbia agar containing 5% sheep blood (Becton Dickinson, Heidelberg, Germany) and incubated at 37 • C for 48 h under anaerobic conditions. The bacterial growth was then visually observed. To ensure the continuity of the cell culture and to obtain the appropriate number of cells for the study, the cell-line cultures were passaged when reaching a confluence of 90%. The medium above the growth surface was harvested and washed with phosphate buffer saline (PBS; Lonza, Basel, Switzerland), then the cells were trypsinized with EDTA (0.25%) (Sigma-Aldrich, Saint Louis, MI, USA) at 37 • C for 10 min. The trypsin effect was inactivated by adding 10 mL of fresh medium with 10% FBS. Then, the contents of the bottle were centrifuged at 1500× g for 5 min, after which the supernatant was removed, and 1 mL of medium was added to the remaining pellet. The cells were then counted with a Thoma chamber and diluted to the desired amount in a new flask with a fresh medium and 10% FBS. The excess cells were transferred to a new culture vessel. Cells intended for the experiment of bacterial adhesion to colon cells were cultured by means of the above method using 24-well plates in an incubator with a flow of 5% CO 2 and a temperature of 37 • C. On the day before the experiment, the culture fluid was replaced with a new one, without the addition of antibiotics. Experiments were performed on mature cells, which were 15-days-after-seeding HT-29 and CCD 841 CoN cells and 21-days-after-seeding HT-29 MXT cells, all passaged 15-25 times.

The Effect of Bee Products on the Adhesion of the Tested C. difficile Strains to Cell Lines
The 24-well plates of the 15-day culture of HT-29, HT-29 MTX, and CCD 841 CoN cell lines were washed twice with PBS buffer. After rinsing, the wells were supplemented with 400 µL of DMEM culture fluid with 1% content of individual tested bee products and incubated for 4 h at 37 • C with a flow of 5% CO 2 . Then, 100 µL of the suspension of the tested bacteria was added to the wells and incubated for another hour. The inoculum consisted of bacterial suspensions with a density of 3.0 on the McFarland scale prepared from 24 h bacterial cultures on a solid medium. The blank was the inoculum cell culture without the addition of bee products. After 1 h, the medium was harvested and washed twice with PBS buffer, then the cells were trypsinized for 10 min at 37 • C. The trypsin effect was stopped by adding 500 µL of medium with 10% FBS. The content of each well in the amount of 100 µL was transferred to sterile Eppendorf tubes containing 900 µL of sterile water, and 20 µL was spread on a solid medium with 5% sheep blood and incubated for 48 h at 37 • C under anaerobic conditions. The colonies were counted and averaged [27].

Testing the Formation of Biofilm of C. difficile Strains under the Influence of Bee Products
In the strains grown on Columbia agar, one colony was excised and transferred to BHI liquid medium. All C. difficile strains were incubated overnight in a BHI medium at 37 • C under anaerobic conditions. The study was performed on a 96-well titer plate. The suspension of bee products in BHI at a volume of 180 µL was added to the wells at concentrations of 100%, 50%, 25%, 12.5%, 6.25%, 3.125%, 1.56%, and 20 µL of overnight cultures of the test strains. Wells with BHI broth without the inoculum were used as controls. The plates were incubated at 37 • C for 48 h under anaerobic conditions for biofilm formation. After 48 h, the liquid phase was aspired using a sterile pipette, washed twice with PBS, and air dried at 37 • C for 15 min. Each well was then stained with crystal violet (Analab, Warsaw, Poland) for 10 min. The crystal violet was removed, and the wells were washed eight times with PBS. After air drying for 15 min at 37 • C, the crystal violet within the biofilms was dissolved in ethanol, and the absorbance was measured at 620 nm (A 620 ) using a Bio-Rad 550 Microplate Reader (Bio-Rad, Hercules, CA, USA). All strains were tested six times. The average values for each C. difficile strain were calculated.

Confocal Laser Scanning Microscopy
The effects of bee products at different concentrations were visualized in culture with five different C. difficile strains, namely, 630 and M 120, and three randomly selected strains with representative RTs: 1468/12 (RT027), 25694/12 (RT023), and 2628/12 (RT176). Biofilms were grown on sterile 10 mm diameter glass-bottom dishes (Nunc, Roskilde, Denmark). Biofilm culture was prepared by dilution of overnight cultures of C. difficile in fresh BHI and BHI medium with subinhibitory doses of bee products: for strain 630, BHI medium supplemented with 25% of goldenrod honey, 3.125% of Manuka honey 550+, 12.5% of pine honey, and 50 mg/L of bee bread; for strain M 120, BHI medium supplemented with 6.25% of goldenrod honey, 0.78% of Manuka honey 550+, 12.5% of pine honey, and 50 mg/L of bee bread; for strain 1468/12, BHI medium supplemented with 6.25% of goldenrod honey, 3.125% of Manuka honey 550+, 1.56% of pine honey, and 25 mg/L of bee bread; for strain 25694/12, BHI medium supplemented with 25% of goldenrod honey, 3.125% of Manuka honey 550+, 12.5% of pine honey, and 12.5 mg/L of bee bread; and for strain 2628/12, BHI medium supplemented with 1.56% of goldenrod honey and Manuka honey 550+, 0.78% of pine honey, and 50 mg/L of bee bread. As a result, 2500 µL of BHI-supplemented medium and 500 µL of overnight cultures of C. difficile strains were obtained. The control was the biofilm of the tested strain without bee products. The biofilms were allowed to grow for 48 h at 37 • C under anaerobic conditions. The mature biofilms were then washed twice using 10 mM MgSO 4 before staining with acridine orange (10 µg/mL) for 30 min in the dark. The dishes were washed twice with 10 mM MgSO 4 . Imaging was performed using a Nikon A1R MP microscope with a Nikon Ti Eclipse series (Nikon, Tokyo, Japan), using a ×60 objective lens and immersion oil. Images were acquired at 2040 × 2048 pixels using a Z-step of 0.1 µm. Acridine orange was detected using an excitation wavelength of 488 nm and an emission wavelength of 500-550 nm. Images were processed and analyzed using NIS-Elements AR v. 4.10 software.

Statistical Analysis
Statistical analysis was performed using Statistica software (version 13, StatSoft, TIBCO Software, Warsaw, Poland). The effect of bee products on C. difficile adhesion and biofilm formation was assessed using the Kruskal-Wallis one-way analysis of variance (ANOVA) followed by the Dunn's test for comparison. Differences in biofilm formation were assessed using ANOVA followed by the Tukey's post-hoc test. Differences were considered as statistically significant for p values < 0.05.

Conclusions
The RT023 and RT027 strains deserve attention. The RT023 strains showed a high MIC sensitivity to all bee products and strongly adhered to the healthy CCD 841 CoN line and the HT-29 MTX tumor line. Similarly, the strains with RT027 were also sensitive to all bee preparations and adhered strongly to the HT-29 line. In both cases, strains with these RTs produced a thin biofilm.
Our research proved that bee products had differential activity against the tested strains of C. difficile belonging to different PCR-RTs.

Data Availability Statement:
The data underlying this article will be shared on reasonable request to the corresponding author.