Batch culture fermentation

Faecal samples were obtained from three healthy human volunteers (two males, one female; age 30 to 38 years; BMI, body mass index of 18.5–25) who were free of known metabolic and gastrointestinal diseases (e.g., diabetes, ulcerative colitis, Crohn’s disease, irritable bowel syndrome, peptic ulcers and cancer). All healthy faecal donors had the experimental procedure explained to them and were given the opportunity to ask questions. All donors then provided written informed consent for the use of their faeces in the study, and a standard questionnaire to collect information regarding the health status, drug use, clinical anamnesis, and lifestyle was administered before the donor was asked to provide a faecal sample. For the coeliac donors (two females, one male; age 30 to 38 years; BMI of 18.5–25), written informed consent was also obtained, and the in vitro study was approved by the University of Reading Research Ethics Committee (UREC 15/20). All faecal samples from healthy and coeliac donors were collected on site, stored in an anaerobic cabinet (10% H₂, 10% CO₂ and 80% N₂) and used within a maximum of 15 min after collection. The samples were diluted 1/10 wt/vol in anaerobic phosphate-buffered saline (PBS, 0.1 M phosphate buffer solution, pH 7.4) and homogenised (Stomacher 400, Seward, West Sussex, UK) for 2 min (460 paddle-beats/min). To maintain the anaerobic conditions, the PBS was maintained in anaerobic cabins until the time of use. The resulting faecal slurries from each individual were used to inoculate the batch-culture systems.

Previously sterilized batch culture fermentation vessels (280 mL working volume) were filled with 45 mL of sterile complex colonic model growth medium. The composition of this medium included peptone water (5 g L^-1), yeast extract (4.5 g L^-1), starch (5 g L^-1), tryptone (5 g L^-1), NaCl (4.5 g L^-1), KCl (4.5 g L^-1), mucin (4 g L^-1), casein (3 g L^-1), pectin (2 g L^-1), xylan (2 g L^-1), arabinogalactan (2 g L^-1) and inulin (1 g L^-1) [22]. All media and chemicals were purchased from Oxoid and Sigma. Then, the vessels were connected to a circulating water bath at 37°C and sparged with O₂-free N₂ gas overnight to attain anaerobic conditions. The pH was adjusted to between 6.7 and 6.9 using pH meter controllers with NaOH or HCl (Electrolab260; Electrolab Ltd., Tewkesbury, UK), and 5 mL of faecal slurry was then inoculated in each vessel. In total, eighteen vessels were prepared and supplemented with 1 mL of CB or GFB digesta (3 vessels per type of donor, including a negative control, i.e., a sample containing faecal slurry but without bread, and samples with CB and GFB). The batch cultures were run for 48 h, and 5 mL of the samples were removed at 0, 6, 24 and 48 h for analysis of bacterial populations by fluorescence in situ hybridisation (FISH) and SCFA analysis by high-performance liquid chromatography (HPLC).

FISH was performed as described by Costabile et al. [23]. A 375 μL aliquot of the batch culture samples was fixed in three volumes of ice-cold 4% (w/v) paraformaldehyde for 4 h at 4°C, centrifuged at 13,000 g for 5 min and washed twice in 1 mL of sterile PBS. The cells were again pelleted by centrifugation and re-suspended in 150 μL of sterile PBS, to which 150 μL of ethanol was added. The samples were then mixed and stored at -20°C until used.

All probes were synthesised by Sigma-Aldrich. The following bacterial groups were identified using synthetic oligonucleotide probes that target specific regions of the 16S ribosomal RNA molecule, labelled with the fluorescent dye Cy3: Clostridium hystolyticum clusters I/II (Chis150, TTATGCGGTATTAATCTYCCTTT) [24], Lactobacillus/Enterococcus spp. (Lab158, GGTATTAGCAYCTGTTTCCA) [25], Clostridium clusters XIVa+b (Erec482, GCTTCTTAGTCARGTACCG) [24], Bacteroides/Prevotella group (Bac303, CCAATGTGGGGGACCTT) [26], Bifidobacterium spp. (Bif164, CATCCGGCATTACCACCC) [27] and Eub338 I-II-III (GCTGCCTCCCGTAGGAGT, GCAGCCACCCGTAGGTGT, GCTGCCACCCGTAGGTGT) [28].

Samples were taken from the batch culture vessels at each time point, and cell-free culture supernatants were obtained by centrifuging 1 mL at 13000 x g for 10 min, followed by filter sterilisation (0.2 μm Acrodisc^® syringe filters with a hydrophilic polyvinylidene fluoride (PVDF) membrane, 13 mm; Pall Corporation) to remove particulate matter. The SCFA content was quantified on an ion exclusion HPLC (LaChrom Merck Hitachi, Poole, Dorset, UK) instrument equipped with a pump (L-7100), RI detector (L-7490) and autosampler (L-7200). Samples (20 μL) were injected into the HPLC at a flow rate of 0.5 mL min^-1 with a prepacked Rezex ROA–Organic Acid H+ 80% (300 x 7.8 mm) column at 84°C and a detector wavelength of 210 nm. H₂SO₄ (2.5 mM) was used as the eluent, and the organic acids (lactic, acetic, propionic and butyric) were calibrated against standards at concentrations of 12.5, 25, 50, 75 and 100 mM. An internal standard of 2-ethylbutyric acid (20 mM) was included in the samples and external standards. All chemicals were provided from Sigma-Aldrich.

The results from the FISH and SCFA analyses were standardized and reported as increases/decreases relative to t₀ of the negative control (beginning of the experiment). The results were first analysed through one-way ANOVA and Tukey’s test using the homogeneous group approach [29]. The classical approach of one-way ANOVA and post-hoc offers an overview of the significant differences amongst different samples. However, it does not work well if the studied parameter does not show or possess well-defined statistical groups and is distributed in a continuous way and if each sample can be attributed to different statistical groups. When this type of distribution is observed in the results, the use of ANOVA by homogeneous groups is advisable: the parameters are organized in a table with different columns (homogeneous groups), and each column contains the samples belonging to the same homogeneous group. The novelty relies upon the fact that the same sample can be attributed to many groups.

Thereafter, the FISH and SCFA values were used as input data to run 3 different principal component analysis (PCA; for the results after 6, 24, and 48 h) experiments.

As a confirmatory experiment, a second standardization of the data was performed, i.e., for each sampling time, the results from healthy and coeliac donors supplemented with CB or GFB were reported as increases/decreases relative to the negative control at the same time, and the results were then analysed through one-way ANOVA. This type of modelling was used to exclude possible prebiotic activities of CB or GFB on both healthy and coeliac people.