Variable alterations of the microbiota, without metabolic or immunological change, following faecal microbiota transplantation in patients with chronic pouchitis

Faecal microbiota transplantation (FMT) is effective in the treatment of Clostridium difficile infection, where efficacy correlates with changes in microbiota diversity and composition. The effects of FMT on recipient microbiota in inflammatory bowel diseases (IBD) remain unclear. We assessed the effects of FMT on microbiota composition and function, mucosal immune response, and clinical outcome in patients with chronic pouchitis. Eight patients with chronic pouchitis (current PDAI ≥7) were treated with FMT via nasogastric administration. Clinical activity was assessed before and four weeks following FMT. Faecal coliform antibiotic sensitivities were analysed, and changes in pouch faecal and mucosal microbiota assessed by 16S rRNA gene pyrosequencing and 1H NMR spectroscopy. Lamina propria dendritic cell phenotype and cytokine profiles were assessed by flow cytometric analysis and multiplex assay. Following FMT, there were variable shifts in faecal and mucosal microbiota composition and, in some patients, changes in proportional abundance of species suggestive of a “healthier” pouch microbiota. However, there were no significant FMT-induced metabolic or immunological changes, or beneficial clinical response. Given the lack of clinical response following FMT via a single nasogastric administration our results suggest that FMT/bacteriotherapy for pouchitis patients requires further optimisation.


FMT Protocol
A single nasogastric infusion of donor faeces was given according to previously described protocols 2,3 . Stool donors were asked to provide a stool sample less than 6 hours prior to faecal transplantation. 30 g of stool was homogenised with a household blender in 50 ml of 0.9% saline until reaching a smooth consistency and filtered through sterile gauze to produce a faecal-saline solution. The night before and the morning of the procedure the recipient was treated with a proton pump inhibitor. A nasogastric tube was inserted and the position confirmed with chest radiograph and 30 ml of the faecal-saline solution was administered via the nasogastric tube which was then flushed with 50 ml of normal saline solution.

Assessment and sampling
Clinical assessment was made between zero and seven days prior to FMT. Pre-FMT assessment consisted of clinical (PDAI and Cleveland global quality of life scores) and endoscopic assessment and biopsies were taken for histological assessment. Stool was also collected for analysis of faecal coliform sensitivities as previously described 6 .
Clinical end points were defined as the number of patients in clinical remission (Clinical PDAI =0/ total PDAI ≤4) or clinical response (reduction in PDAI score ≥3 points) 7 four weeks after treatment with FMT and the number of patients demonstrating changes in pouch faecal bacterial sensitivities following FMT.
Biopsies (n=2) for microbiological assessment were immediately washed in phosphate buffered saline (PBS) and snap frozen in liquid nitrogen and stored at -80•C. 2 ml of stool for microbiological and metabolic assessment was mixed with 8 ml of PBS, vortexed and centrifuged at 250 g for 1 minute to exclude large particulate matter.
Aliquots (1:10) were stored in RNA later (Ambion) at -80•C. Biopsies were collected for isolation of dendritic cells (n=6) and for overnight culture (n=1) for assessment of cytokines in biopsy supernatants (see below). PCR and 16S rRNA gene sequencing and analysis 16S rRNA genes were amplified using Golay barcoded primers (Eurofins MWG Operon). Bacterial primers 454-338F (5'-CCTATCCCCTGTGTGCCTTGGCAGTCT CAGACTCCTACGGGAGGCAGCAG-3') and 454-926R (5'-CCATCTCATCCCTGCGTGTCTCCGACTCAG-barcode-CCGTCAATTCMTTTRAGT-3'), which span variable regions V3 to V5 of the 16S rRNA gene were used. Each sample in the study was amplified with a 926R primer that contained a unique barcode sequence. The Golay barcodes that were used are as follows:

Schedule of stool and mucosal biopsy analysis
Q5 TM Taq polymerase (New England Biolabs) was used for PCR reactions according to the product protocol with the following PCR cycling conditions: 98 •C for 2mins, followed by 25 cycles of 98 •C for 30 secs, 52 •C for 30 secs and 72 •C for 2 mins, followed by a final extension of 72 •C for 5 mins.
16S rRNA gene amplicons from each sample were then pooled in equimolar amounts into a mastermix for sequencing using the Lib-L kit on the 454 GS FLX Titanium platform. The resulting sequence data was processed using the mothur software package 8 as described previously 9 , except that sequences with less than a minimum length of 320 bp rather than 350 bp were discarded.
Any contaminant OTUs that were detected in the sequenced negative control samples were removed from the final dataset (see Supplementary Table 3 for full list of contaminants). Diversity comparisons (Chao, Shannon and inverse Simpson) were carried out after first sub-sampling the data down to 391 reads per sample to ensure equal sampling depth across all samples. Good's coverage (an estimate of completeness of species sampling) at 391 reads per sample was on average greater than 92% for all sample groups (overall median 95.9%, range 87.9 to 99.7%). Similarity indices were assessed using Bray-Curtis and Theta Yue & Clayton calculators using mothur. In order to identify differentially abundant taxa between study cohorts, the 50 most abundant taxa from the OTU, Genus and Family taxonomic levels, and all phyla, were assessed by Metastats 10 , as applied in mothur.
Significance thresholds were adjusted to account for false discovery rate when making multiple comparisons using the Benjamini-Hochberg approach 11 .

Sample preparation for NMR spectroscopic analysis
A total of 800 µl of faecal slurry was spun for 10 min 10,000 x g and 540 µl of supernatant was taken to thoroughly mix with 60 µl of 1.5 M potassium phosphate buffer. The mixture was subsequently centrifuged again at 10,000 x g for 10 min and 550 µl of supernatant was transferred into an NMR tube with an outer diameter of 5 mm. Ethanol signals were observed in the NMR spectra of fresh faecal samples, some of which contained extremely high concentrations, which affect the spectral quality and subsequent data analysis. To remove ethanol, all faecal water samples were dried out using a speed vacuum, re-suspended in 650 µl D 2 O and sonicated for 20 min. The resulting samples were spun for 5 min at 10,000 x g and 600 µl of supernatant was transferred into an NMR tube with an outer diameter of 5 mm. The water peak region [d 1 H 4.7-4.92] were removed to minimise the effect of the disordered baseline caused by water suppression. For the dataset of fresh faecal samples, ethanol peaks were also removed. Probabilistic quotient normalization was subsequently performed on the datasets in order to account for dilution of complex biological mixtures.

Isolation of lamina propria dendritic cells
The method used has been described and validated previously [12][13][14] . The epithelium was removed after a 60 minute treatment with 1 mM EDTA in calcium-and magnesium-free Hank's balanced salt solution at 37°C with gentle agitation. The remaining biopsy tissue was then digested in 1mg/mL collagenase D (Roche Diagnostics Ltd, Lewes, England) in RPMI 1640/HEPES (Sigma-Aldrich Co Ltd, Poole, England) containing 2% foetal calf serum and 20 μg/mL deoxyribonuclease I (Roche Diagnostics Ltd) agitating for 90-120 minutes at 37°C. After incubation, lamina propria mononuclear cells (LPMC) released from the tissue samples were passed through a cell strainer and washed in complete medium.

Cell Surface Labelling
Cells were labelled in FACS buffer (phosphate-buffered saline containing 1mmol/L EDTA and 0.02% sodium azide).
To prevent non-specific binding, unoccupied binding sites were blocked with foetal calf serum prior to antibodies being added at predetermined optimal concentrations. Following labelling, cells were washed twice in FACS buffer and resuspended in 300 μl 1% paraformaldehyde and stored at 4•C until acquisition within 24 hours.

Antibodies
Antibodies with the following specificities and conjugations were used: β7 integrin-PE (FIB504), CD3-PeCy5 Abcam. Appropriate isotype-matched control antibodies were purchased from the same manufacturers.

Cytokines in whole biopsy culture supernatants
Biopsy samples were blotted and weighed prior to being cultured overnight in complete medium (RPMI 1640 Dutch modification (Sigma Aldrich Co. Ltd, Irvine, UK) supplemented with 2 mM L-glutamine, 100 µg . ml -1 streptomycin, 100 units . ml -1 penicillin and gentamicin (50 µg . ml -1 ) with 10% foetal calf serum) at 37•C in a humidified 5% CO 2 atmosphere. Cell-free culture supernatants were analysed using a multiplex assay (BD Cytometric Bead Array) according to the manufacturer's instructions to determine levels of IL-6 and TNF in biopsy supernatants. Standard curves were plotted to calculate the detection limit of each cytokine. Values below the detection limit are reported as being equal to that level.  Percentage of sequences identified from the bacterial families of >1% total abundance in donor and patient samples at baseline. Percentage of total sequences from the predominant bacterial families in donor stool (n=8), patient stool pre FMT (n=7) and patient mucosal samples pre FMT (n=8). * denote where differences between donor and patient stool are significant after adjustment for FDR.   Table 3: Taxonomic list of contaminant OTUs that were removed from the final analysis after detecting their presence in negative "blank" control samples.