Butyrate, which is produced by the human microbiome, is essential for a well-functioning colon. Bacteria that produce butyrate are phylogenetically diverse, which hinders their accurate detection based on conventional phylogenetic markers. As a result, reliable information on this important bacterial group is often lacking in microbiome research.
In this study we describe a gene-targeted approach for 454 pyrotag sequencing and quantitative polymerase chain reaction for the final genes in the two primary bacterial butyrate synthesis pathways, butyryl-CoA:acetate CoA-transferase (but) and butyrate kinase (buk). We monitored the establishment and early succession of butyrate-producing communities in four patients with ulcerative colitis who underwent a colectomy with ileal pouch anal anastomosis and compared it with three control samples from healthy colons. All patients established an abundant butyrate-producing community (approximately 5% to 26% of the total community) in the pouch within the 2-month study, but patterns were distinctive among individuals. Only one patient harbored a community profile similar to the healthy controls, in which there was a predominance of but genes that are similar to reference genes from Acidaminococcus sp., Eubacterium sp., Faecalibacterium prausnitzii and Roseburia sp., and an almost complete absence of buk genes. Two patients were greatly enriched in buk genes similar to those of Clostridium butyricum and C. perfringens, whereas a fourth patient displayed abundant communities containing both genes. Most butyrate producers identified in previous studies were detected and the general patterns of taxa found were supported by 16S rRNA gene pyrotag analysis, but the gene-targeted approach provided more detail about the potential butyrate-producing members of the community.
The presented approach provides quantitative and genotypic insights into butyrate-producing communities and facilitates a more specific functional characterization of the intestinal microbiome. Furthermore, our analysis refines but and buk reference annotations found in central databases.
The relationship between a healthy functioning gut microbiome and overall human well-being is firmly established. Recently, large-scale projects in this field, namely the Human Microbiome Project and the Metagenomics of the Human Intestinal Tract framework program, have been launched, with the goal of developing a holistic understanding of the composition and functional properties of intestinal bacteria and their effects on the human host. Numerous host-microbiome interactions have been reported and microbial-derived metabolites such as vitamins or short chain fatty acids have been of specific interest in many studies (see [1, 2]). Among these, butyrate is considered as one of the most important metabolites as it serves as the major energy source of colonocytes; has anti-inflammatory properties; and regulates gene expression, differentiation and apoptosis in host cells .
Much of the information on the diversity of butyrate-producing bacteria has depended on culture-independent methods, however recent cultivation efforts for some of these strict anaerobes have been successful . The existing isolates within this functional group are phylogenetically diverse, with the two most abundant groups related to Eubacterium spp. and Roseburia spp. (Clostridium cluster XIVa) and Faecalibacterium prausnitzii (Clostridium cluster IV) . However, both clusters include additional non-butyrate-producing species. As such, 16S rRNA gene-targeted analysis often cannot distinguish the butyrate-producing from the non-producing community in a sample . Furthermore, it is increasingly recognized that horizontal gene transfer, which uncouples bacterial function from phylogeny, plays an important role in shaping the human microbiome . The shortcomings of relying only on traditional 16S rRNA gene-based phylogenetic analysis for functional inferences are now recognized in many other fields of microbial ecology. To resolve this, functional gene-targeted sequencing has emerged as the method of choice to investigate microbial functionality independent of phylogeny. This method has been used in several studies examining the nitrogen cycle , degradation of xenobiotic compounds  and antibiotic resistance of gut bacteria . These studies have demonstrated the value of obtaining a detailed insight into specific microbial processes.
In the human gut, butyrate is produced through two main pathways, the butyryl-CoA:acetate CoA-transferase pathway (but) and the butyrate kinase (buk), and previous studies on colonic isolates of healthy individuals have illustrated that the but pathway predominates . Consequently, Louis and Flint  developed a semi-quantitative PCR protocol targeting a selection of but sequences and used the same primers to construct clone libraries from fecal samples that revealed high gene diversity, including several unknown operational taxonomic units (based on a 98% DNA similarity ).
In this study, we present a novel approach that targets a broad range of but and buk genes based on both 454 pyrotag sequencing in combination with the Ribosomal Database Project’s (RDP) functional gene pipeline  and on quantitative PCR targeting selected groups of butyrate producers. The presented methods were applied on luminal samples from patients with ulcerative colitis (UC) who underwent a colectomy followed by ileal pouch anal anastomosis (IPAA) as described in the accompanying paper by Young et al.. In this procedure, the entire colon is resected, the terminal ileum is fashioned into a pouch and connected to the anal canal, and intestinal flow is re-established. Previous data indicate that approximately half of patients will develop pouchitis within 1 year, an inflammatory condition similar to UC . Because of the clinical similarity between pouchitis and UC, it is thought that studying the development of pouchitis can be used to reveal the etiology of UC. Several studies reported dysbiosis of the intestinal microbiome in patients with UC [16, 17]. However, it is unclear whether the observed microbiome changes are the cause or the consequence of UC. These difficulties make pouchitis an ideal model system as it allows for the clinical observation of individuals from “time zero”, when fecal flow is initiated through the newly established, disease-free pouch. In this study, we specifically monitored the initial establishment (first 2 months) of butyrate-producing microbial communities in four patients after IPAA and compared the results with healthy controls.
Processing of samples
In this study, four patients with a history of UC undergoing total abdominal colectomy with IPAA were identified from the outpatient and inpatient practices of gastroenterologists and colorectal surgeons at the University of Chicago Medical Center between 2010 and 2011. All four patients had a confirmed diagnosis of UC based on endoscopy and pathology findings, were scheduled for a total proctocolectomy with IPAA, and were willing and able to participate in the study. Exclusion criteria included pregnancy or inability to give informed consent. All patients gave written informed consent before screening. The Institutional Review Board of the University of Chicago Medical Center approved this study protocol. For each patient, one sample was collected prior to ileostomy takedown (except for patient 200) and an additional three samples were collected over a period of 2 months after connection of the pouch to the anal canal (Table 1). None of the patients received antibiotic treatment during the period of this study. All samples were obtained from stool aspirates. Sterile saline was injected to liquefy the stool and contents were sampled using the suction port of the colonoscope. Bulk DNA was extracted using the UltraClean Mega Soil DNA Isolation Kit (MO BIO Laboratories, Inc., Carlsbad, CA, USA) according to the manufacturer’s protocol. Healthy colon samples were obtained from the recto-sigmoid section of the colon without prior bowel preparation to ensure that the microbiota was not altered by this procedure. For additional details on sample collection and storage, see .
Samples analyzed in this study
Number of days from ileostomy takedown
Primers, amplicon generation and 454 pyrotag sequencing
Primers were designed based on the Fungene database for the butyryl-CoA:acetate CoA-transferase (but) and butyrate kinase (buk) genes (Table 2 - for more details see Additional file 1: Figures S1 and S2 and Tables S1 and S2). Three barcoded forward and three reverse primers with fused adaptors for the Lib-A system (454 Life Sciences, Branford, CT, USA) were designed for each gene. The aim was to obtain broad coverage without exceeding a degeneracy of 100. For PCR, each forward primer (0.4 μM final concentration) was used separately in triplicate samples and was mixed with all three corresponding reverse primers (0.16 μM final concentration each), except for but_1F, where each reverse primer was used in a separate reaction. Because of the low target concentrations in many samples, sufficient amplification was often difficult. Therefore, extracted DNA was subjected to whole genome amplification (WGA; illustra GenomiPhi V2 DNA Amplification Kit, GE Healthcare, Little Chalfont, UK) to increase template concentration. A total of 150 ng of WGA template was used for each PCR reaction using the GoTaq Flexi system (Promega, Madison, WI, USA; total volume of 12.5 μL). Because primers do not perfectly match all desired targets (Additional file 1: Tables S1 and S2), PCR stringency was low for both genes with an annealing temperature of 54°C and high MgCl2 concentrations of 3 mM. Furthermore, higher cycle numbers (35×) were used to increase yield. Thermocycling was done as follows: 2 min at 95°C; 45 s at 95°C, 45 s at 54°C, 45 s at 72°C (×35); 10 min at 72°C. PCR products were pooled for each forward primer (triplicate reactions), gel-extracted (QIAquick Gel Extraction Kit; Qiagen, Valencia, CA, USA) and purified (QIAquick Gel Purification Kit; Qiagen). Several bands were visible on gels (especially for but) and only the target bands located around 425 (but) and 500 (buk) were excised. Nonspecific binding of primers was reduced with increased target concentrations. A re-conditioning step of purified product was essential to avoid short reads during sequencing. Each sample was re-amplified (0.2 ng of generated amplicons as template, 60°C annealing temperature, 15 cycles, total volume of 50 μL) using the AccuPrime PCR system (Life Technologies, Grand Island, NY, USA) with primers (0.4 μM final concentration) targeting whole adaptor sequences. PCR products were gel-extracted and purified again. Sequencing was performed with a 454 Junior System according to the manufacturer (454 Life Sciences). For each run, eight samples (four from each gene) were mixed at equal concentrations. We are aware that the protocol used, including WGA followed by a high cycle number PCR and a final re-amplification step, may have introduced bias. However, comparing all obtained results derived from different methods suggested that the procedure did not alter the main trends (see main text).
Primers designed for this study are illustrated
Functional genes - pyro-sequencing
Functional genes - quantitative PCR
16S genes - quantitative PCR
Degenerate bases are shown in capital letters. The following sequences are targeted: buk_F/R - butyrate kinase (buk) genes; but_F/R - butyryl-CoA:acetate CoA-transferase (but) genes; G_buk_F/R - buk genes of Clostridium acetobutylicum, C. butyricum, C. perfringens; G_Acida - but genes of Acidaminococcus sp.; G_Fprsn - but genes of Faecalibacterium prausnitzii; G_RosEub, G_Ros_R, G_Eub_R - but genes of Eubacterium rectale and Roseburia sp; Cbuty_F/R -16S genes of C. butyricum; FPR-2 F, Fprau645R - 16S genes of F. prausnitzii; Rrec630F, RrecRi630F, Erec870R - 16S genes of E. rectale and Roseburia sp.; 1132 F, 1108R - universal for 16S. a Primers described in Ramirez-Farias et al.; b Primers described in Leigh et al.. For more details on targeted sequences see Additional file 1: Table S1 and S2.
Quantitative real time PCR
Primers designed for quantitative PCR (qPCR; Table 2) targeting the but/buk genes were based on the Fungene database and were specific to all desired target genes with at least two mismatches in one or both primers for other non-target but/buk genes. BLAST analysis illustrated no significant matches to other unrelated sequences. The 16S rRNA gene primers (Rrec2 and Fprau) targeting butyrate producers are described in Ramirez-Farias et al.. Total 16S rRNA gene community qPCR primers were based on Leigh et al.. Additionally, primers for the 16S rRNA genes of C. butyricum were designed based on the RDP database. Specific amplification of targets was verified for all primers using the following pure cultures (amplification efficiency per nanogram of pure culture is given in brackets): Bacillus licheniformis ATCC 14580, Bacteroidetes thetaiotaomicron E50, C. acetobutylicum ATCC 824 (2.65 × 105), C. difficile ATCC 630, C. perfringens ATCC 13124 (4.88 × 105), Eubacterium hallii DSM(Z) 3353, E. rectale DSM(Z) 17629 (4.06 × 105), Faecalibacterium prausnitzii DSM(Z) 17677 (5.53 × 105), Roseburia intestinalis DSM(Z) 14610 (1.80 × 105) and R. inulinivorans DSM(Z) 16841 (4.73 × 105). Strains were purchased either from ATCC or DSM(Z) (as indicated in name). B. licheniformis and B. thetaiotaomicron E50 were provided by Daniel Clemens. For the primers targeting Acidaminococcus (but gene) and C. butyricum, (16S rRNA gene), instead of a pure culture, a patient sample containing many target bacteria (based on all methods presented here) served as a positive control.
Amplification was performed with the SYBR Green Master Mix (Life Technologies) with 10 ng template DNA per reaction (total volume of 15 μL; no WGA except for the healthy control samples) in 384-well plates (ABI Prism 7900 HT, Life Technologies). Annealing temperatures and final primer concentrations were as follows: G_buk (64°C; 0.83 μM), Cbuty (66°C; 0.67 μM), FPR/Fprau (60°C; 0.83 μM), G_Acida (67°C; 0.83 μM), G_Fprsn (70°C; 0.83 μM), G_Ros/Eub (62°C; 0.83 μM; G_Ros_R and G_Eub_R were mixed together at equal final concentrations of 0.42 μM), Rrec/Erec (60°C; 0.83 μM; the two forward primers were mixed together at equal final concentrations of 0.42 μM) and total 16S (60°C; 0.67 μM). Thermocycling was done as follows; 2 min at 50°C; 10 min at 95°C; 45 s at 95°C; 45 s at individual annealing temperature; and 45 s at 72°C (for total 16S rRNA, elongation at 72°C was omitted) (×40). Analysis was performed in duplicate samples. Genomic DNA of R. inulinivorans, F. prausnitzii and C. perfringens (for functional gene qPCRs) and cloned amplified products (for 16S qPCRs and G_Acida; TOPO cloning kit, Life Technologies) at concentrations of 102 to 107 copies (10-fold dilutions) were used for standard curves to determine target concentrations. Genomic DNA of Desulfotomaculum acetoxidans DSM 771 with 10 16S rRNA gene copy numbers was used for the standard curve (103 to 108) for total 16S rRNA gene quantification. The detection limit was set as 102 target sequences for all primers and results are expressed as a percentage of the total bacterial community based on total 16S rRNA gene qPCR. For 16S rRNA gene copy number normalizations of specific 16S rRNA targets see below (comparing functional gene results to 16S pyrotag data). Because but/buk target sequences are present as a single copy per genome, qPCR results of functional genes were multiplied by five to account for multiple 16S rRNA gene copies (five on average) of the intestinal bacterial flora).
Raw reads matching barcodes (106,708 for but and 84,222 for buk) were processed using the RDP pyro-sequencing pipeline , where 87% but and 94% buk sequences passed quality filtering. Subsequently, sequences were subjected to RDP FrameBot for frameshift corrections and closest match assignments. To develop a reference sequence set for FrameBot, we took the corresponding gene sequence sets from the Fungene database, developed through (Hidden Markov Model) HMM searches of the National Center for Biotechnology Information protein database, and removed partial sequences with less than 93% coverage (that is, last filled model position - first filled model position/model length) to the full gene length HMM model, giving 452 but and 422 buk reference sequences. For buk, 97% reads that passed the initial process passed FrameBot with minimum 30% identity to the closest match and 125 amino acids in length. On average, 1.6 frameshifts were corrected per sequence and 58% of the sequences contained at least one frameshift. For but, 59% reads that passed the initial process passed FrameBot with minimum 30% identity and 100 amino acids in length. The majority of non-passing sequences were identified as human origin. On average, 0.6 frameshifts were corrected per sequence, 30% of the sequences contained at least one frameshift. Sequences can be accessed at SRA062948.
Ordination and diversity analysis
For each gene, the frameshift-corrected protein sequences were aligned using HMMER3 and clustered using RDP mcClust with the complete-linkage algorithm. Only amplicons with an identity of ≥70% to the closest matches in the reference (97% of but and 93% of buk sequences) were used for additional phylogenetic tree and ordination analysis, as we were not confident that more distant matches were bona fide but or buk. Additional filtering was performed based on neighbor joining tree analysis of reference sequences (see Results and Additional file 1). The remaining sequences were binned according to closest match assignments with reference sequences showing less than 2% dissimilarity merged (based on Figures 1 and 2). Results of both genes were combined and the entire butyrate community of each sample was used for ordination analysis. The nonmetric multidimensional scaling based on Chao corrected Jaccard index distance was performed using the vegan package in the R environment . Both patients and time points were grouped for analysis. Diversity analysis (Shannon index) was calculated using the Biodiversity R package.
Comparing functional gene results to 16S pyrotag data
For library generation of 16S rRNA gene analysis and pyro-sequencing see Young et al.. For the first healthy control, no data on luminal aspirate were available and shown results are based on a colon biopsy sample of the same individual. Data were analyzed for known butyrate producers in the human colon at the genus level (based on  and obtained but/buk gene sequences) except for Clostridia, where species discrimination was applied. All results were normalized to five 16S rRNA gene copy numbers, which represented the average for Firmicutes and Bacteroidetes, the two most abundant phyla in the gut. Average copy number of each genus was derived from rrnDB  and the Integrated Microbial Genome database . A list of taxa searched as well as individual 16S rRNA gene copy numbers is presented in Additional file 1: Table S3.
Investigating but gene diversity
Several short chain fatty acid (SCFA) transferases have been characterized that exhibit broad substrate specificities and show remarkable sequence similarities . Consequently, existing annotations in public databases are often unreliable and misleading. In our established Fungene database, many known but sequences are wrongly annotated (due to GenBank’s annotation) and SCFA transferases similar to but such as 4-hydroxybutyrate CoA transferases (4hbt) are present. A neighbor joining tree of all sequences from Fungene’s butyryl-CoA:acetate CoA-transferase (but) database (>93% coverage to model; to ensure only full-length sequences were considered) was constructed where all functionally verified but genes cluster together and apart from 4hbt genes (Additional file 1: Figure S1). Primers were designed to specifically target those but sequences. However, it is still likely that SCFA transferases related to but are amplified as well. In order to quality filter our obtained but sequences (in addition to the processing pipeline presented in the Methods section), only sequences located within the cluster identified in Additional file 1: Figure S1 were regarded as likely real but, whereas the remaining amplicons (<1%) matching 16 references outside the cluster were excluded from further analysis. We detected a broad diversity of but genes in our samples and they were linked to almost all described but carrying species (Figure 1A). Four closest FrameBot matches were assigned to 75% of all obtained sequences, namely R. intestinalis L1-82, R. inulinivorans A2-194, Acidaminococcus sp. D21 and E. rectale ATCC 33656. To verify the closest match assignments all amplicons were mapped on a tree together with full-length reference sequences using Pplacer (; Additional file 1: Figure S3). We observed minimal deep branching; nearly all amplicons diverged in the terminal branches to the reference sequences, and the numbers assigned correlated well with the FrameBot closest match assignments. An exception was Clostridium sp. SS3/4 where many more amplicons than expected, that FrameBot had originally assigned to C. symbiosum and Clostridium sp. M62/1, mapped to that reference sequence. The discrepancies were most likely due to the different underlying assignment methods used by FrameBot and Pplacer. The former compares blossum62-corrected pairwise distances, whereas the latter is based on maximum likelihood criteria. Conservation analysis of but showed a remarkably similar pattern between the reference and amplicon sequences, and several well-conserved amino acid sites (>95% conservation in both groups) were identified (Figure 1B).
Investigating buk gene diversity
A considerable diversity of buk sequences that included sequences similar to the majority of previously described butyrate producers were detected in our samples (Figure 2A). The Fungene database contains many sequences assigned to species not reported to produce butyrate, such as members of the phylum Bacteroidetes. Many of our amplicons closely matched sequences originating in Bacteroides and the established tree clusters them together with known butyrate producers and apart from acetate kinase, a closely related gene (Figure 2A; a neighbor joining tree of all Fungene sequences (93% cut-off) is shown in Additional file 1: Figure S2). Therefore, we included those sequences for analysis. Three quarters of all obtained buk amplicon sequences were assigned to four closest FrameBot matches; Bacteroides sp. D2, Bacteroides sp. 3_2_5, C. butyricum 5521 and C. perfringens. The resultant tree including the mapped amplicon sequences confirmed closest match assignments (Additional file 1: Figure S4). Sequence analysis revealed less similarity among buk genes than observed for but and fewer conserved amino acids could be detected (Figure 2B versus Figure 1B).
Ordination and diversity analysis of obtained data
Nonmetric multidimensional scaling analysis of the total butyrate-producing community (but and buk genes together) revealed a unique community pattern for individual patients, which all clustered distinct from the healthy control samples (Figure 3). However, the successional trend was different for all patients (Additional file 1: Figure S8). Diversity calculations also did not reveal a consistent successional pattern. Whereas Shannon diversity increased for patients 200 and 206, no change was detected for 207, and 210 demonstrated a decrease over time (Figure 4). At the fourth visit, all communities analyzed displayed a comparable diversity value, which was similar to that of the healthy control samples.
Quantitative analysis of but/buk genes
Functional gene pyro-sequencing only allows for relative abundance measurements in each sample. Hence, we developed a complementary qPCR approach. Clear patterns emerged from qPCR of the buk and but genes. Whereas no target genes could be detected in the luminal aspirate before ileostomy takedown (visit 1; no sample was available for patient 200), abundant butyrate-producing communities became established in all patients over time (Figure 5). The observed community profiles were distinct between patients. Patients 206 and 207 were enriched in buk genes (up to 19.9% of the total community), whereas but was almost absent. Patient 210 was unique in the development of a community similar to the healthy controls, harboring but genes most closely related to both F. prausnitzii and Roseburia sp./E. rectale. Additionally, this patient exhibited abundant but-carrying Acidaminococcus sp. communities, which were absent in the healthy control samples. At visits three and four, 15.5% and 26% of patient 210’s total microbial community exhibited but genes, which was within the broad range for the control samples with 4.4%, 2.6% and 74.1%, respectively. In patient 200, we initially detected only buk genes, but a considerable but-gene community linked to Roseburia sp./E. rectale was established over time as well.
Investigating the butyrate-producing community based on 16S rRNA gene analysis
We retrieved the major known butyrate-producing taxa from literature  and from the but and buk data and used this information to screen for those taxa in the total 16S 454 pyrotag analysis presented in the accompanying paper . Results are displayed in Figure 6A. Additionally, qPCR targeting specific butyrate producers was performed (Figure 6C). 16S rRNA gene analysis supported the functional gene results in that similar overall patterns were detected by the two different techniques. Communities linked to buk were dominated by sequences similar to those of C. butyricum and C. perfringens, whereas sequences similar to Acidaminococcus sp., F. prausnitzii and Roseburia sp. comprised the majority of but-associated bacteria in both methods (Figures 5 and 6, Additional file 1: Figures S3 and S4). Nevertheless, several differences between 16S rRNA gene and functional gene analysis were observed. Only a minute fraction from 16S rRNA gene pyrotag data was identified as Eubacterium sp., whereas many but sequences were assigned to strains of E. hallii and E. rectale. Other studies that utilized fluorescence in-situ hybridization and clone libraries reported high concentrations of those strains in the healthy colonic microbial flora [4, 25], which suggests that 16S rRNA gene-based analysis could not reliably discriminate them from other taxa. Furthermore, Subdoligranulum sp., which contain one butyrate-producing isolate, S. variabile ( has the gene buk), were not detected in the functional gene data. But if this genus is considered to be butyrate-producing, then the 16S rRNA gene analysis suggests a considerable abundance of buk genes in healthy control samples. Similarly, many more 16S rRNA gene sequences were assigned to Acidaminococcus sp., Anaerostipes sp., Coprococcus sp. and Peptoniphilus sp. in certain samples compared with the results obtained from the functional gene analysis. These findings support earlier reports that butyrate synthesis is often not a homogenous feature of all members of a genus [4, 5] and strengthens the application of higher taxonomic resolution techniques to adequately assess the butyrate-producing potential of bacterial communities. Species resolution is also crucial for the functionally diverse genus Clostridia. Several butyrate-producing members such as Clostridium sp. SS2/1, Clostridium sp. M62/1, C. acetobutylicum, C. carboxidivorans and C. symbiosum were matched to numerous functional gene sequences, but could not be detected in the 16S rRNA gene data.
In this study we show that functional gene-targeted analysis of the intestinal bacterial butyrate-producing community can overcome limitations imposed by relying solely on 16S rRNA gene targeted investigations. A combination of 454 pyrotag sequencing with qPCR analysis was essential to resolve the full differences among samples. Pyro-sequencing provided specific community profiles at great depth, whereas qPCR enabled the absolute quantification of genes. Ordination analysis based on pyrotag data revealed individual community patterns for each patient distinct from those of the healthy controls (Figure 3); however, only qPCR could demonstrate that overall gene concentrations differed over several orders of magnitude (Figure 5). Notably, the presented protocol for amplicon generation enabled amplification of genes for all samples, although actual abundance of individual targets was often below qPCR thresholds.
An abundant butyrate-producing community is essential for a well-functioning colon . Butyrate is also the preferred energy generating substrate for the pouch epithelium and it is believed that supply deficiencies could initiate or promote development of pouchitis . The pouch was aerobic before ileostomy takedown and only became anoxic after it was connected to the anal canal, which limits oxygen influx and promotes the establishment of anaerobes. Anoxic/oxic ratios of cultivars steadily increased over time after ileostomy takedown in all investigated patients . In this study, we could demonstrate that these environmental changes were accompanied by the development of butyrate-producing communities at abundances similar to healthy participants of other studies  and to the healthy controls of this study. However, only patient 210 displayed a community pattern comparable to healthy control samples, which was also the case in the companion global 16S rRNA community analysis . Patients 206 and 207 exhibited abnormal communities with buk genes predominating and only very few detectable but genes. Patient 200 displayed an ‘in-between’ community harboring both genes. Currently, the buk pathway is not considered to be important for butyrate production in healthy individuals , a finding further supported by this study. Whether the highly abundant buk-containing communities in patients 206 and 207 can compensate for low concentration of but is unclear. Unfortunately, no SCFA data are available to address this question. Enzyme assays on 17 butyrate-producing isolates demonstrated considerably higher activities for but than for buk, suggesting that the but pathway yields more butyrate in comparison to synthesis via buk. Interestingly, patient 210 is the only individual who did not show onset of inflammation 25 months post ileostomy takedown, whereas patients 200 (8 months), 206 (16 months) and 207 (17 months) all developed pouchitis. Although the patient number is low in this study, it does suggest that the initial establishment of a ‘healthy type’ butyrate-producing community is important to maintain a well-functioning pouch and to prevent the development of disease. The specific question of how butyrate production affects the development of disease will be addressed in a follow-up study where community profiles of patients undergoing IPAA will be monitored until the onset of inflammation and compared with those derived from asymptomatic individuals.
Our approach directly targets the genes coding for butyrate-synthesizing enzymes. We did observe some discrepancies between phylogeny and predicted function, which was especially true for the obtained buk gene sequences assigned to members of the genus Bacteroides. Bacteroides are currently not considered butyrate producers and several culture-based investigations point out their inability to synthesize butyrate (for examples, see [29, 30]). This also applies to many other sequences presented in Additional file 1: Figure S2. Interestingly, some early studies from the 1980s indicated butyrate production by closely related bacteria, namely certain Porphyromonas (former Bacteroides) strains [31, 32]. However, additional studies specifically investigating butyrate synthesis including more Bacteroides strains (and other candidates) under several different physiological conditions are needed to address this issue. Furthermore, even for known butyrate-synthesizing bacteria, gene detection does not automatically imply production of butyrate. Gene expression and a functioning pathway are determined by environmental conditions, with oxygen concentration as likely the most important factor . Most butyrate producers are considered to be strict anaerobes with their growth and function strongly coupled. However, it has been recently shown that certain butyrate producers, namely F. prausnitzii, can also grow under microaerophilic conditions using extracellular oxygen as the final electron acceptor . Butyrate production by this bacterium was still detected under these conditions but at a reduced rate.
The presented protocols provide a new approach to more specifically resolve the butyrate-producing community. We could clearly demonstrate that butyrate producers were established at high abundance (approximately 5% to 26% of total bacterial community) in the pouch of all patients undergoing an IPAA within the first 2 months after ileostomy takedown. Community profiles were distinctive among patients. Most important, one individual harbored a community profile similar to the healthy controls with but genes predominating and an almost complete absence of buk genes, whereas the other three patients had other variants. Only the former patient remained healthy 25 months later. 16S rRNA gene analysis showed similar overall patterns as the functional gene-targeted approach, but only the latter could reveal specific details on butyrate-producing taxa that were essential to assess the entire butyrogenic potential of the microbial communities analyzed. Furthermore, our analysis refines but and buk reference annotations found in central databases. In the near future, these methods will be complemented by metagenomic tools that will provide full-length gene sequences without prior amplification and will facilitate the investigation of not only individual genes of interest but also complete synthesis pathways.
Basic Local Alignment Search Tool
Hidden Markov Model
Ileal pouch anal anastomosis
Polymerase chain reaction
Quantitative polymerase chain reaction
Ribosomal Database Project
Short chain fatty acid
Whole genome amplification
Financial support was provided by NIH Human Microbiome Project Demonstration Project (UH3 DK083993) and the University of Chicago Digestive Disease Research Core Center (P30DK42086). Special thanks to Charles Falkiewicz, Jiarong Gao, Stephanie LaHaye, Craig McMullen, Derek St. Louis and Andrew Worden for technical assistance, and to Christopher Radek for his support during sequencing. The authors acknowledge Daniel Clemens for providing several strains.
Center for Microbial Ecology, Michigan State University
Department of Internal Medicine, University of Michigan
Argonne National Laboratory
Marine Biological Laboratory
Knapp Center for Biomedical Discovery, University of Chicago
Department of Medicine, University of Chicago
Mortensen P, Clausen M: Short-chain fatty acids in the human colon: relation to gastrointestinal health and disease.Scand J Gastroenterol Suppl 1996, 216:132–143.PubMedView Article
Arumugam M, Raes J, Pelletier E, Le Paslier D, Yamada T, Mende DR, Fernandes GR, Tap J, Bruls T, Batto JM, Bertalan M, Borruel N, Casellas F, Fernandez L, Gautier L, Hansen T, Hattori M, Hayashi T, Kleerebezem M, Kurokawa K, Leclerc M, Levenez F, Manichanh C, Nielsen HB, Nielsen T, Pons N, Poulain J, Qin J, Sicheritz-Ponten T, Tims S: Enterotypes of the human gut microbiome.Nature 2011, 473:174–180.PubMedView Article
Hamer HM, Jonkers D, Venema K, Vanhoutvin S, Troost FJ, Brummer RJ: Review article: the role of butyrate on colonic function.Aliment Pharmacol Ther 2008, 27:104–119.PubMedView Article
Louis P, Young P, Holtrop G, Flint HJ: Diversity of human colonic butyrate-producing bacteria revealed by analysis of the butyryl-CoA:acetate CoA-transferase gene.Environ Microbiol 2010, 12:304–314.PubMedView Article
Louis P, Flint HJ: Diversity, metabolism and microbial ecology of butyrate-producing bacteria from the human large intestine.FEMS Microbiol Lett 2009, 294:1–8.PubMedView Article
Pryde SE, Duncan SH, Hold GL, Stewart CS, Flint HJ: The microbiology of butyrate formation in the human colon.FEMS Microbiol Lett 2002, 217:133–139.PubMedView Article
Ventura M, O’Flaherty S, Claesson MJ, Turroni F, Klaenhammer TR, van Sinderen D, O’Toole PW: Genome-scale analyses of health-promoting bacteria: probiogenomics.Nat Rev Microbiol 2009, 7:61–71.PubMedView Article
Braker G, Fesefeldt A: Development of PCR primer systems for amplification of nitrite reductase genes (nirK and nirS) to detect denitrifying bacteria in environmental samples.Appl Environ Microbiol 1998, 64:3769–3775.PubMed
Iwai S, Chai B, Sul WJ, Cole JR, Hashsham SA, Tiedje JM: Gene-targeted-metagenomics reveals extensive diversity of aromatic dioxygenase genes in the environment.ISME J 2009, 4:279–285.PubMedView Article
Looft T, Johnson T, Allen HK, Bayles DO, Alt DP, Stedtfeld RD, Sul WJ, Stedtfeld TM, Chai B, Cole JR, Hashsham S, Tiedje JM, Stanton TB: In-feed antibiotic effects on the swine intestinal microbiome.Proc Natl Acad Sci U S A 2012, 109:1691–1696.PubMedView Article
Louis P, Duncan SH, McCrae SI, Millar J, Jackson MS, Flint HJ: Restricted distribution of the butyrate kinase pathway among butyrate-producing bacteria from the human colon.J Bacteriol 2004, 186:2099–2106.PubMedView Article
Louis P, Flint HJ: Development of a semiquantitative degenerate real-time PCR-based assay for estimation of numbers of butyryl-coenzyme A (CoA) CoA transferase genes in complex bacterial samples.Appl Environ Microbiol 2007, 73:2009–2012.PubMedView Article
Young VB, Raffals LH, Huse S, Schloss PD, Vital M, Dai D, Bruic JM, Antonopoulos DA, Arrieta R, Kwon J, Reddy G, Hubert N, Dalal S, Morrison H, Eren MA, Schmidt TM, Tiedje JM, Sogin M, Chang E: Multiphasic analysis of the temporal development of the distal gut microbiota in patients following ileal pouch anal anastomosis.BMC Microbiome
Salemans JM, Nagengast FM, Lubbers EJ, Kuijpers JH: Postoperative and long-term results of ileal pouch-anal anastomosis for ulcerative colitis and familial polyposis coli.Dig Dis Sci 1992, 37:1882–1889.PubMedView Article
Frank DN, St Amand AL, Feldman R, Boedeker EC, Harpaz N, Pace NR: Molecular-phylogenetic characterization of microbial community imbalances in human inflammatory bowel diseases.Proc Natl Acad Sci U S A 2007, 104:13780–13785.PubMedView Article
Peterson D, Frank DN, Pace NR, Gordon JI: Metagenomic approaches for defining the pathogenesis of inflammatory bowel diseases.Cell Host Microbe 2008, 3:417–427.PubMedView Article
Ramirez-Farias C, Slezak K, Fuller Z, Duncan A, Holtrop G, Louis P: Effect of inulin on the human gut microbiota: stimulation ofBifidobacterium adolescentisandFaecalibacterium prausnitzii.Br J Nutr 2009, 101:541–550.PubMedView Article
Leigh MB, Pellizari VH, Uhlík O, Sutka R, Rodrigues J, Ostrom NE, Zhou J, Tiedje JM: Biphenyl-utilizing bacteria and their functional genes in a pine root zone contaminated with polychlorinated biphenyls (PCBs).ISME J 2007, 1:134–148.PubMedView Article
Cole JR, Wang Q, Cardenas E, Fish J, Chai B, Farris RJ, Kulam-Syed-Mohideen AS, McGarrell DM, Marsh T, Garrity GM, Tiedje JM: The Ribosomal Database Project: improved alignments and new tools for rRNA analysis.Nucleic Acids Res 2009, 37:D141-D145.PubMedView Article
Matsen F, Kodner RB, Armbrust EV: pplacer: linear time maximum-likelihood and Bayesian phylogenetic placement of sequences onto a fixed reference tree.BMC Bioinformatics 2010, 11:538.PubMedView Article
Hold GL, Schwiertz A, Aminov RI, Blaut M, Flint HJ: Oligonucleotide probes that detect quantitatively significant groups of butyrate-producing bacteria in human feces.Appl Environ Microbiol 2003, 69:4320.PubMedView Article
Holmstrøm K, Collins MD, Møller T, Falsen E, Lawson PA: Subdoligranulum variabilegen. nov., sp. nov. from human feces.Anaerobe 2004, 10:197–203.PubMedView Article
Van Immerseel F, Ducatelle R, de Vos M, Boon N, van De Wiele T, Verbeke K, Rutgeerts P, Sas B, Louis P, Flint HJ: Butyric acid-producing anaerobic bacteria as a novel probiotic treatment approach for inflammatory bowel disease.J Med Microbiol 2010, 59:141–143.PubMedView Article
Chapman M, Hutton M, Grahn MF, Williams NS: Metabolic adaptation of terminal ileal mucosa after construction of an ileoanal pouch.Br J Surg 1997, 84:71–73.PubMedView Article
Scott K, Duncan S, Flint HJ: Dietary fibre and the gut microbiota.Nutr Bull 2008, 33:201–211.View Article
Touw JJ, van Steenbergen TJ, De Graaff J: Butyrate: a cytotoxin for Vero cells produced byBacteroides gingivalisandBacteroides asaccharolyticus.Antonie Van Leeuwenhoek 1982, 48:315–325.PubMedView Article
Grenier D, Mayrand D: Cytotoxic effects of culture supernatants of oral bacteria and various organic acids on Vero cells.Can J Microbiol 1985, 31:302–304.PubMedView Article
Khan MT, Duncan SH, Stams AJM, van Dijl JM, Flint HJ, Harmsen HJM: The gut anaerobeFaecalibacterium prausnitziiuses an extracellular electron shuttle to grow at oxic-anoxic interphases.ISME J 2012, 6:1–8.View Article
Guindon S, Dufayard JF, Lefort V, Anisimova M, Hordijk W, Gascuel O: New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0.Syst Biol 2010, 59:307–321.PubMedView Article
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.