Background: Bacillus clausii is a gram-positive spore forming bacterium, used as a probiotic. Despite its wide use in acute diarrhea, little is known about its ability to modify the intestinal microbiota.

      Aims: Aim of this study is to investigate the effect of B.clausii on a mild DSS induced acute colitis in mice evaluating gut microbiota modulation.

      Methods: Mice were exposed to DSS and received B.clausii by gavage. Disease Activity Index (DAI) was calculated, stool samples were cultured for microbiological analysis at 3 time points, colonic inflammatory score was analyzed by histology. The direct impact of B.clausii on microbiota was evaluated in vitro using aerobic and anaerobic selective culture media.

      Results: B.clausii slightly ameliorated mild DSS colitis in mice, with reduction of the colonic inflammatory score. B.clausii treatment was associated with gut microbiota modifications in stool samples. Briefly lower counts of aerobic bacteria and higher anaerobes in fecal samples were recognized compared to healthy mice. A total reduction of Enterobacteriacee was also observed. In vitro experiments showed that B.clausii reduced the total count of P.Aeruginosa in aerobic conditions, when compared to control.

      Conclusions: B.clausii slightly improves mild acute colitis in mice, actively modulating gut microbiota.

      Bacillus clausii is a gram-positive spore forming bacterium, used as a probiotic [1]. Amongst probiotic microorganisms, the mix of four antibiotic resistant B.clausii strains (O/C, SIN, N/R, T) represents one of the most used probiotic in clinical practice. The application of preparations containing probiotics such as B.clausii spores in the treatment or prevention of gut barrier impairment has been largely supported during last years. Favorable effects of B.clausii have been linked to several properties: resistance to commonly used antibiotics, sporogenic activity in the intestine, adherence to intestinal wall, antimicrobial and immunomodulatory activity [2-4], regulation of cell growth and differentiation, cell-cell signaling, cell adhesion, transcription and signal transduction [4], vitamins production [5] and gut protection from genotoxic agents [6].

      Furthermore, B.clausii spores are extremely stable and capable to resist, at different terms, physical and chemical stress, such as low pH values (those present in the gastric barrier), biliary acids and at different temperatures [7-9].

      These properties are linked with its ability to reach the gut alive and to exercise its pharmacological effects. The ability to adhere to intestinal walls, reducing the permeability and potentiate the local immune response, especially through secretory Immunoglobulin (IgA) has been shown for probiotics and B.clausii [10,11].

      In human pharmacokinetic studies, B.clausii spores have shown to germinate in the human GI tract although the underlying mechanisms are still not completely understood [1-13].

      Little is known about the effect of B.clausii application on species or even genus level of the gut microbiota in dysbiosis, though in clinical practice B.clausii is effective in preventing from dysbiosis associated symptoms such as diarrhea and abdominal pain due to antibiotic treatment [14]. Besides preventing from GI symptoms, other investigations reported immunomodulatory properties of B.clausii, in vivo [3,15].

      The use of B.clausii has been associated to the induction of a Th1 response in nasal mucosa of patients with rhinitis, with good anti-inflammatory and immunomodulatory effect on mucosal cytokines and reduction of the pathogenic Th2 hyperactivation [15]. Starting from these observations, it could be argued that B.clausii, due to a potential strong Th1 response in the host, could have positive effects on experimental active murine colitis, a disease only partially resembling the human ulcerative colitis. Murine colitis, induced by DSS treatment, is in fact a good model of acute/chronic colitis, with clear signatures of Th2 colitis associated to an impairment of innate immunity mechanisms and gut microbiota alterations. Up to now, little is known about potential role of Bacillus clausii on clinical assessment of experimental colitis.

      In the present pre-clinical study we focused on the effect of B.clausii on mild acute colitis in mice induced by Dextran Sodium Sulfate (DSS) with the main attention to investigate its potential microbiota modulatory property.

Induction of experimental murine chronic colitis by DSS

      All animal used for this study were studied and treated in accordance with our institution’s guidelines and current animal research laws, furtherly specified in the approved protocol of the present study. DSS colitis was induced on female mice C57BL/6 exposed to 1cycle of 5 days of 2.5% dextran sodium sulfate (DSS, molecular weight 36-44 kDa, MP Biomedicals, Aurora, OH, USA) in drinking water as indicated by several publications [16-18]. Animals were checked every day by measuring body weight, stool consistency and presence of fecal occult blood, as indicated [17,18]. Disease Activity Index (DAI) was calculated as described previously [18]. Animals were divided into 3 groups: 1 receiving DSS only, and 2 receiving 2 different doses of B.clausii by gavage daily for 5 days during the DSS cycle (dose 1 consisting in 80 mln of B.clausii spores and dose 2 consisting in 20 mln of B.clausii spores). Three parallel groups of mice not receiving DSS were used as controls. In particular the first group received only water (placebo), the second group only dose 1 consisting in 80 mln of B.clausii spores and the third group only dose 2 consisting in 20 mln of B.clausii spores). Of each group, 50% of animals were sacrificed at day 5 while all the others were sacrified at day 14th. Of each animal, plasma, faecal samples, colon, last loop of the small intestine and abdominal lymph nodes were taken for histology and molecular biology analysis. Same experimental settings are utilized on healthy mice receiving drinking water instead of DSS.

Bacillus clausii doses ad source

      The probiotic preparations (Enterogermina, Sanofi-Aventis SpA, Milan, Italy) consisted of a mixture of spores of four antibiotic-resistant B.clausii strains named O/C, N/R, SIN and T. Each probiotic vial was stated to contain 2x109 CFU of B.clausii spores in 5 ml of oral suspension commercially available. Two doses of B.clausii were administered: a higher dose (dose 1) of 80 million spores and a lower dose equal to 20 million spores (dose 2), both provided in water for a total of 200 microliters by oral gavage, on day 1-5.

      Dose 1: containing 80 million spores, corresponded to 200 microliters of solution available in commerce.

      Dose 2: containing 20 million spores, corresponded to 50 microliters of solution available in commerce, diluted in 150 microliters of water, thus reaching a final volume of 200 microliters of solution.

Gut Microbiota assessment from feces of treated mice

      Mice stool samples from all experimental groups were collected at time 0, after 5 days of treatment (time 1) and at 14º day, the end of the experiment (time 2). The specimens were then weighed and homogenized in Phosphate-Buffered Saline (PBS) to normalize for dry weight of faeces. The homogenized samples were plated in 10-fold-diluted onto TSA Agar (Becton Dickinson) and onto CNA Agar/Schaedler Agar (BioMerieux) for aerobes and anaerobes assessment. All plates were incubated at 37oC for 24-48 h. The colonies were subsequently counted and the results are given as Colony-Forming Units (CFU)/g dry matter. Single colonies were selected for characterization of bacteria by MALDI-TOF mass spectrometry (Bruker).

      Gut Microbiota assessment for in vivo evaluation of Bacillus clausii effect on fecal microbiota in experimental colitis. Mice stool samples from all experimental groups were collected at time 0 (base line), after 5 days (time 1) and at the end of the experiment (14 days) (time 2). The specimens were weighed and collected into FECALSWABTM with 2ml of Cary-Blair medium (COPAN) and then conducted at the microbiology laboratory to proceed to the analysis of the microbial flora.

Cultures

      Culture was the first tool for the exploration of the gut microbiota and bacterial ecosystem [19-21].

      We used Columbia CNA AGAR +5% sheep blood (BIOMERIEUX), a selective medium that includes colistin and nalidixic acid to inhibit gram negative growth, in order to enables important Gram-positive cocci to be recognised more readily and isolated easily from the mixed bacterial populations. Chocolate PVX AGAR (BIOMERIEUX), a variant of the blood agar plate, containing red blood cells that have been lysed by slowly heating to 80 °C, was used for the isolation of more aggressive bacteria. SCHAEDLER AGAR with Vitamin K1 and 5% Sheep Blood (BENKTON-DICKINSON), a non-selective but highly nutritive media was used for the isolation of obligate anaerobes. Each stool sample was seeded on agar media using a volumetric loop calibrated to hold 0.01 ml of sample. We used a semi-quantitative streak method. Each sample plate with its specific culture media was then incubated in Aerobic condition with 5% CO2 for PVX agar medium and anaerobic condition for CNA and SCHAEDLER agar using an anaerobic hood. Incubation time ranged from 48h to 72 h at 37°C to identify slow growing bacterial species. After incubation, each agar plate was observed and all different colonies were identified by mass spectrometry. The microbial loads were expressed in CFU/g.

Gut microbiota characterization and bacterial strain Identification

      Colony identification was performed by mass spectrometry (MALDI-TOF), permitting a rapid and effective identification of bacterial strain as reported [22]. Briefly, the MALDI-TOF uses a soft ionization mechanism, which is obtained by using a matrix added to the bacterial colonies on metal plates. Ionization is obtained by using a UV laser beam. The measurement of the time of flight into the tube to reach a detector generated spectra. Spectral comparison with data from the defined database available is automatically performed and allows identification. This is a rapid and low-cost identification method [23].

Detection of B.clausii in fecal samples overtime

      Stool samples from healthy and DSS-colic mice were thermally inactivated by exposing the stool suspension to 80ºC for 15 minutes prior to plating (10 μl per plate) on Mueller-Hinton Agar (OXOID) and TSA AGAR (BENKTON-DICKINSON). The plates were incubated at 37ºC for 24-48 hours. The detection limit of this method was estimated to be 10 bacteria for 50 mg of stool sample. To confirm the presence of B.clausii strains we used a Gram staining of selected colonies and then a mass spectrometry approach with MALDI–TOF.

Histological assessment

      Colon samples from scarified mice were fixed on 4% formalin and embedded in paraffin. After hydration, all sections are de-paraffined and Haematoxylin/Eosin (H&E) staining was performed. Rachmilewitz score was used for histology assessment, as shown [24].

Statistical analysis

      Mean, standard deviations, standard errors were used for continuous variables. Comparisons were performed upon statistician consultant. Non-parametric tests (Willcoxon, etc) or parametric tests (T test, Chi square, Anova) were used properly.

Bacillus clausii slightly improves mild DSS murine colitis

      The highest dose of B.clausii (dose 1) slightly ameliorates mild DSS colitis compared to placebo, as shown by the assessment of disease activity index (DAI) at days 2-5 (p<005) (Figure 1A).

      At sacrifice, body weight loss and colon length did not differ significantly among different groups (Figure 1B and Figure 1C). Histologic analysis of colonic samples confirmed the mild extent of colitis, with no fibrosis. However, the group that obtained the best inflammatory score was the one treated with lower dose (Figure 1D).

Bacillus clausii modulates gut microbiota composition in mild DSS colitis in stool samples of mice with colitis

      B.clausii was associated to a moderate increase in total count of aerobic bacteria compared to basal time T0 (p<0.05), particularly at higher dose (dose 1) (Figure 2A). Similar effect was also found in healthy mice (Figure 2C). No major changes were observed in anaerobic bacteria, treated or not with B.clausii (Figure 2 B and D).

      B.clausii was isolated from the fecal samples of mice at 5 days (data not shown) but also following its withdrawal (time 2), following the recovery procedure after heat inactivation (Figure 2 E). B.clausii strains produce identical white and grayish colonies on MHA (Muller Hinton Agar) as previously reported [12].

      Finally, when moving to single germ analysis (Figure 3), the treatment with B.clausii was associated to a measurable effect on gut microbiota composition, resulting, in particular, in a significant (p<0.05) reduction of count of Staphilococcus aureus, Clostridium innocuum and total count of Enterobacteriacae like Klebsiella oxytoca, Enterococcus casselliflavus and Enterococcus gallinarum.

      Other variations, which did not reach the statistical significance, included an increase in Mycobacterium, Bacteroides vulgatus, Bacteroides uniformis, Lactobacillus gasseri, Lactobacillus johnsonii and the increase in Bacteroides fragilis.

Direct modulatory effect of Bacillus clausii on fecal microbiota

      The bactericidal effect of B.clausii in relation to the other bacterial components was then evaluated in vitro.

      A co-culture of B.clausii was set up under aerobic conditions in physiological solution.

      When B.clausii was incubated in co-culture with P.aeruginosa, a significant decrease of the total count of P.aeruginosa itself was observed, as shown by a reduced hemolysis in culture plate (Figure 4).

      To our knowledge, the present paper assesses for the first time that the probiotic B.clausii (strains O/C, SIN, N/R, T) modulates the gut microbiota while acting as a mild modulator of subacute murine colitis. On one hand, these observations are consistent with the well described immunomodulatory properties of B.clausii, described in literature. On the other side, it opens new possibilities to study further usage in conditions that could benefit from its these pharmacological effects.

      In the present experiments we aimed to rule out potential toxic effects of B.clausii. Therefore we used relatively low doses of B.clausii, that allowed us to detect, at the same time, a slight improvement of colitis and a significant microbiota modulatory effect in a dose dependent manner.

      The combined immunological and microbiological effect of this probiotic is supported by the in vitro findings. In facts B.clausii itself, when not given orally but incubated on fecal samples of mice not previously exposed to it, only exerts a minor effect, including a decrease of the total count of P. Aeruginosa over E. Coli. On the other hand, when given orally during an acute phase of colitis, it is associated to an overall increase in the total count of aerobic bacteria, modulating the microbiota of colic mice toward the microbiota of the healthy phenotype.

      Furthermore, the use of B.clausii is associated to a significant reduction of potentially harmful commensals like Staphilococcus aureus, Clostridium innocuum and total count of Enterobacteriacae (Klebsiella oxytoca, Enterococcus casselliflavus and Enterococcus gallinarum).

      Lat but not least, the presence of B.clausii in stool samples of colitic mice in form of spore, as demonstrated by the growth capacity following heat inactivation, suggests that B.clausii remains within the colon up to 9 days from the beginning of therapy. This data suggest that in the intestine B.clausii acts not only as vegetative bacteria but also as a spore form.

      The last aspect that needs to be addressed is that B.clausii seems to ameliorate the overall histology of colitic mice when given at lower dose. However, this effect was observed during the recovery phase of the colitis, when the clinical symptoms were almost silent in both control and treated mice. We believe that these findings have to be discussed in the light of the mucosal healing phase following the acute phase of colitis. Further investigations are needed to fuel these discussions.

      This paper has several limitations. First of all we did not perform more mechanistic analysis about the cytokine modulation by B.clausii at the intestinal level that is associated to gut microbiota modulation. We believe that this topic is worth to be investigated in dedicated trials. The experimental setup, to answer these questions, would need blood and biopsy samples at the time of animal sacrifice, in order to perform the analysis of the inflammatory cytokine expression on both serum and tissue samples.

      Furthermore, we performed the analysis of the microbiota through “culturomics” instead of metagenomics. We believe that this is not a real limitation, as also “culturomics”, have shown to deliver valid results in the investigation of the modulation of the gut microbiome in health and disease models [25]. The culturomics choice over the genomic is more sensitive for “quantitative” data on the overall effect on B.clausii; furthermore we wanted to assess also the “growth” of B.clausii itself over its vegetative form in order to induce a microbial change.

      Surprisingly, although we clearly detected B.clausii in treated mice, the amount of the bacterium recovered in the feces was low. That suggests that B.clausii acts not just as a gut microbe covering a niche in the intestine, but potentially induces immunological changes that are able to reinforce its “ecological” effects.

      The present paper strongly encourages further studies to investigate its effect on the gut microbiome in health and disease models, in order to define new therapeutic areas for this probiotic besides the treatment of dysbiosis associated symptom like diarrhea.

Institutional review board statement

      The study was reviewed and approved by the Catholic University Institutional Review Board.

Institutional animal care and use committee statement

      All experiments were approved by the Local Ethics Committee for Animal Research Studies at the UNIVERSITA’ CATTOLICA DEL SACRO CUORE of Rome (Protocol Number 15699/15 - 08/07/2015 and the Superior Service Ethics Commission Animal Health and the Ministry of Health 1075/2015-PR09/10/2015). All animal used for this study received “humane care” and all experiments complied with our institution’s guidelines and current animal research laws, as well specified in the approved protocol of the present study.

Conflict-of-interest statement

      Antonio Gasbarrini, Franco Scaldaferri and Loris Riccardo Lopetuso have been lecturer for SANOFI. This project has been conducted with an unrestricted grant provided by SANOFI to UNIVERSITÀ CATTOLICA DEL SACRO CUORE of Rome, under the supervision of Franco Scaldaferri

Data sharing statement

      Technical appendix, statistical code, and dataset available from the corresponding author at francoscaldaferri@gmail.com.

      The study was supported by Sanofi Aventis Grant.