Fecal sampling and FMT preparation
Fresh fecal samples were collected from three healthy adult human volunteers (a 27 years-old woman and two men aged 35 and 50 years) with no history of antibiotic treatment 3 months prior to the beginning of the study. Fecal samples were kept under anaerobiosis for a maximum of 6 h before treatment. Fecal inocula for the in vitro colon model were prepared using 55 g of fresh fecal samples under strict anaerobic conditions in a vinyl anaerobic chamber (Coy, Grass Lake, MI, USA). Stools were mixed with 500 mL of a 30 mM anaerobic sodium phosphate buffer (pH 6.5) supplemented with 1.9 mM cysteine and the suspension was filtered through a double layer of gauze. The fecal suspension was divided into 100 mL aliquots that were rapidly transferred into each of the five bioreactors simultaneously inoculated.
Filtrates of the same fecal suspensions were used by MaaT Pharma (Lyon, France) to prepare AFMT enema and capsules, according to its manufacturing operating system. FMT enema is a fecal-microbiota suspension for rectal administration stored in a special bag at -80°C (volume of 128-150 mL for 30 g enema and 43-50 mL for 10 g enema). 30 g and 10 g enema contain about 30 x 1011 and 10 x 1011 bacteria, respectively, with a bacterial viability superior to 50 %. FMT capsule (0.45 g) is a caecum-release capsule containing the freeze-dried form of the enema formulation. A capsule contains about 0.35 x 1010 bacteria, with a bacterial viability superior to 50 %. For FMT capsules, only the active ingredient (enema in its freeze-dried form) was introduced into the ARCOL model.
In vitro artificial colon system ARCOL
Human colonic conditions were simulated in the ARCOL model using MiniBio 500 mL my-Control bundles and Lucullus® Lite software from Applikon (Applikon, Delft, The Netherlands). Fermentations were conducted under continuous conditions. The in vitro system reproduces, based on in vivo data, the main physicochemical and microbial conditions encountered in a healthy human adult colon [24,26]. Briefly, at the beginning of the experiment, fecal suspension (100 mL) was added in the bioreactor containing 200 mL of nutritive medium while flushing with O2-free N2 gas. Afterwards, anaerobic conditions were maintained exclusively through the activity of the resident microbiota and by ensuring the airtightness of the system. Overproduced gas were collected in a gas sampling bag connected to the condenser. The temperature of the fermentation was fixed at 37°C and maintained inside the bioreactor using an incorporated panel heater. Colonic pH and redox potential were constantly recorded (Applikon, Delft, The Netherlands) and pH was adjusted to a value of 6.3 with an automatic addition of 2 M NaOH. The amount of NaOH consumed was recorded daily. After one day of batch fermentation, the nutritive medium containing various sources of carbohydrates, proteins, lipids, minerals and vitamins [27] to closely mimic the composition of human ileal effluents, was continuously introduced into the bioreactor at a speed of 0.21 mL/min. The fermentation medium was stirred at a constant speed of 400 rpm. Its volume was monitored using a level sensor and maintained at a constant value of 300 mL by automatic withdrawal of the fermentation medium, ensuring a mean retention time of 24 h.
Experimental design of in vitro fermentations
Five bioreactors running in parallel were inoculated with the fecal suspension from one donor (Fig. 1) and applied as follows: the first bioreactor was used as a control with no antibiotic treatment (control) while the second bioreactor was treated with ciprofloxacin (17850-5G-F, Sigma-Aldrich, Darmstadt, Germany), with an initial addition of 150 mg on day 6 followed by a continuous supply of 500 µg/mL in the nutritive medium up to day 12 (ATB control). The three other bioreactors were treated with ciprofloxacin as described for ATB control and then received: 27 mL of enema preparation at day 14 and day 15 (30 g enema), 9 mL of enema at day 14 and day 15 (10 g enema) or the content of 3 capsules per day for 7 consecutive days from day 14 to day 20 (capsule). The total amount of bacteria administered in the ARCOL model with FMT treatment is 1.2 x 1012, 0.4 x 1012, 1.47 x 1010 for 30 g enema, 10 g enema and capsule, respectively. The experiments were performed in triplicate (biological replicates named Run1, Run2 and Run3) with the fecal samples from each of the three healthy donors. During fermentations, samples were collected daily from the fermentative medium and the atmospheric phase for further analyses.
Antibiotic dosage
Ciprofloxacin concentrations in the fermentative medium were determined using a TurboFlowTM technology (TLX) coupled to Liquid Chromatography-High Resolution Mass Spectrometry (LC-HRMS) at the pharmacological and toxicological analytic unit (CREPTA) of Clermont-Ferrand university hospital.
Gut microbiota activity
Gas. Analysis of O2, N2, CO2, CH4 and H2 found in the atmospheric phase of the bioreactor was performed using a 490 Micro-gas chromatograph (Agilent Technologies, USA) equipped with two columns, Molecular Sieve 5A and PoraPlot U, coupled with TCD detectors. Argon was used as the carrier gas. Gas composition was determined using calibration curves made from ambient air (78 % N2, 21 % O2, 0.04 % CO2) and two gas mixtures A (5 % CO2, 5 % H2, 90 % N2) and B (20 % CO2, 80 % H2). Results were expressed in relative percentages. The total volume of gases overproduced per day (in mL) was also measured by connecting a gas bag to the bioreactors.
Short chain fatty acids (SCFAs). Samples collected from the fermentative medium were centrifuged at 18,000 g for 15 min at 4°C and supernatants were filtered (0.45 µm). Concentrations of the three main SCFAs (acetate, propionate and butyrate) were determined using high performance liquid chromatography (HPLC) (Elite LaChrom, Merck HITACHI, USA) coupled with a diode-array detector. The HPLC column (ICSep ICE-COREGEL 87H3 9 µm 150 x 7.8 mm, Concise Separations, USA) and its guard column were maintained at 50°C. Sulfuric acid 0.008 N was used as mobile phase and SCFAs were separated at a flow rate of 0.6 mL/min. Data was analyzed by the EZChrom Elite software at 205 nm. SCFAs concentrations (expressed in mM or relative percentages) were calculated from calibration curves established with known concentrations of acetate, propionate and butyrate (0, 10, 25 and 40 mM).
Gut microbiota composition
Flow cytometry analysis. Concentrations of viable bacteria in the fermentation medium were determined by a live/dead analysis with flow cytometry. Samples were 10-fold diluted in sterile physiological water to reach 10-4 dilution factor. Bacteria were double-stained with the green-fluorescent DNA SYTO 9 dye labelling all bacteria and the red-fluorescent Propidium Iodide (PI) dye only penetrating and staining cells with damaged membranes (LIVE/DEAD™ BacLight™ Bacterial Viability and Counting Kit, for flow cytometry, Molecular probes, Whaltham, MA). Bacterial suspensions were thus incubated for 15 min at room temperature in the dark with 3.3 mM SYTO 9 and 0.375 mM PI and transferred into BD Trucount™ Tubes (BD Biosciences, San Jose, CA). Flow cytometry analysis was performed on a BD™ LSR II cytometer (BD Biosciences, San Jose, CA) and data were collected with BD FACSDivaTM software. Gating on forward-angle light scatter/side-angle light scatter was used in order to differentiate bacteria from the background. Then, the combined red and green fluorescence dot-plots were used to distinguish among the various subpopulations. Results were expressed as viable cells per mL of fermentative medium.
qPCR analysis. Total DNA was extracted from fermentative medium using the SmartExtract-DNA Extraction Kit (Eurogentec, Seraing, Belgium), according to the manufacturer’s instructions. DNA quantity was evaluated with a NanoDropTM 2000 (Thermo Scientific, Wilmington, Delaware USA). Samples were stored at -20°C prior to analysis. Total bacteria concentration was quantified by qPCR analysis performed on a Stratagene Mx3005P apparatus (Agilent, Waldbronn, Germany), using the Takyon Low ROX SYBR 2X MasterMix blue dTTP kit (Eurogentec, Seraing, Belgium). Each reaction was run in duplicate in a final volume of 10 µL with 5 µL of Master Mix, 0.45 µL of each primer (10 µM), 1 µL of DNA sample (10 ng/µL), and 3.1 µL of ultra-pure water. The following primers targeting 16S rRNA gene were used: BAC338R, 5’- ACTCCTACGGGAGGCAG-3’ and BAC516F, 5’-GTATTACCGCGGCTGCTG-3’, as described by Yu and colleagues [28]. Amplifications conditions consisted of 1 cycle at 95°C for 5 min followed by 40 cycles of 95°C for 30 s, 58°C for 30 s, and 72°C for 30 s. A final cycle of 5 min at 95°C was included. Standard curves were generated from genomic DNA extracted from ARCOL samples, standardized to 10 ng/μL and serially diluted from 1010 to 100 gene copy/mL. Final results were expressed as copy/mL.
16S rRNA gene sequencing and bioinformatics analysis. Genomic DNA was extracted using the NucleoSpin Soil kit (Machery Nagel) and samples stored at -20°C before analysis. 16S rRNA gene sequencing was performed by Eurofins Genomics (Ebersberg, Germany). A sequencing library targeting the V3-V4 region of the 16S rRNA gene was constructed for each sample using the MyTaq HS-Mix 2X, Bioline, according to the manufacturer’s instructions. Libraries were then pooled in an equimolar mixture and sequenced in paired-end (2 x 300 bp) MiSeq V3 runs, Illumina. After amplicon merging using FLASH [29] and quality filtering using Trimmomatic [30], host sequence decontamination was performed with Bowtie2 [31]. Operational Taxonomic Unit (OTU) sequence clustering was performed with an identity threshold of 97 % using VSEARCH [32] and taxonomic profiling was then performed with the Silva SSU database Release 128 [33]. Taxonomic and diversity analyses were performed with R Statistical Software (R Core Team 2015, version 3.4.4) [34] using vegan and phyloseq packages. For fair comparison, the sequence number of each sample was randomly normalized to the same sequencing depth i.e. 50,000 amplicons per sample and normalized by total bacteria count based on qPCR results. Diversity measures correspond to the median value of 20 subsamplings per sample.
Dysbiosis criteria
Criteria selected to determine microbial dysbiosis periods in the ARCOL system were based on modifications of both gut microbiota activity and gut microbiota composition compared to stabilized conditions. For microbiota activity, the following parameters were selected: redox potential values, NaOH consumption, total gas production, CO2 concentration, and SCFA concentrations. Regarding microbiota composition, the selected parameters were the following: total viable bacteria as determined by flow cytometry, total bacterial populations measured by qPCR, richness, Shannon and Bray Curtis indexes. In order to establish a dysbiotic period, each day of fermentation from day 6 for a treated bioreactor was compared to day 6 (corresponding to the end of stabilization phase) of the same bioreactor for all 16S rRNA gene analysis-related criteria (abundance and diversity indexes). For all the other criteria, each day of fermentation from day 6 for a treated bioreactor was compared to the same day of the control bioreactor. All the selected criteria and threshold values are summarized in Table 1.
Statistical analysis
Data were analyzed using a one-way repeated measure analysis of variance (ANOVA) followed by a Newman-Keuls multiple comparisons test. The statistical analysis was performed using GraphPad Prism software 8.0 (GraphPad Software, Inc., San Diego, CA). Results were expressed as means ± SEM (n=3). Differences were considered statistically significant when P < 0.05.