Effects of Dietary Fiber on the Feline Gastrointestinal Metagenome

Oct 17, 2012 - It is possible that the host animal metagenome, consisting of genes produced by the host animal plus those of the microbial communities...
0 downloads 0 Views 4MB Size
Article pubs.acs.org/jpr

Effects of Dietary Fiber on the Feline Gastrointestinal Metagenome Kathleen A. Barry,†,‡ Ingmar S. Middelbos,†,§ Brittany M. Vester Boler,† Scot E. Dowd,∥ Jan S. Suchodolski,⊥ Bernard Henrissat,# Pedro M. Coutinho,# Bryan A. White,†,¶,○ George C. Fahey, Jr.,†,¶ and Kelly S. Swanson*,†,¶,▲ †

Department of Animal Sciences, ¶Division of Nutritional Sciences, and ▲Department of Veterinary Clinical Medicine, and ○The Institute for Genomic Biology, University of Illinois, Urbana, Illinois 61801, United States ∥ Research and Testing Laboratory and Medical Biofilm Research Institute, Lubbock, Texas 79407, United States ⊥ Gastrointestinal Laboratory, Texas A&M University, College Station, Texas 77843, United States # CNRS, Universités Aix-Marseille I & II, Marseille, France S Supporting Information *

ABSTRACT: Four healthy adult cats were used in a crossover design to determine phylogeny and metabolic functional capacity of the cat’s gastrointestinal microbiota using a metagenomic approach. Healthy adult cats (1.7 years old) were fed diets containing 4% cellulose, fructooligosaccharides (FOS), or pectin for 30 d, at which time fresh fecal samples were collected. Fecal DNA samples from each cat consuming each diet were subjected to 454 pyrosequencing. Dominant phyla determined using two independent databases (MG-RAST and IMG/M) included Firmicutes (mean = 36.3 and 49.8%, respectively), Bacteroidetes (mean = 36.1 and 24.1%, respectively), and Proteobacteria (mean = 12.4 and 11.1%, respectively). Primary functional categories as determined by KEGG were associated with carbohydrates, clustering-based subsystems, protein metabolism, and amino acids and derivatives. Primary functional categories as determined by COG were associated with amino acid metabolism and transport, general function prediction only, and carbohydrate transport and metabolism. Analysis of carbohydrate-active enzymes revealed modifications in several glycoside hydrolases, glycosyl transferases, and carbohydrate-binding molecules with FOS and pectin consumption. While the cat is an obligate carnivore, its gut microbiome is similar regarding microbial phylogeny and gene content to omnivores. KEYWORDS: feline gut, gastrointestinal bacteria, microbiome, metagenomics, pyrosequencing



INTRODUCTION The domestic cat is an obligate carnivore, yet most commercial cat foods include moderate to high amounts of carbohydrates for optimal kibble manufacture. Despite no dietary requirement, dietary fiber typically is added to commercial cat diets to promote a number of health outcomes in the cat, including laxation, weight management, prevention or modulation of diabetes, and promotion of gastrointestinal health.5,21,33 Few data have been reported regarding the impact of fiber on gut health outcomes in the cat. Short- and branched-chained fatty acid, phenol, indole, and biogenic amine concentrations, and microbial composition, are modified by the addition of supplemental fiber to the diet of the cat;3,11 however, little is known regarding the composition of this microbiome or the impact of supplemental dietary fiber on the microbiome. In the cat, several techniques have been used in an attempt to determine the microbiome of the large intestine. Plating, fluorescence in situ hybridization (FISH), and quantitative PCR all have provided some insight into the constituent members of © 2012 American Chemical Society

this complex community. From these analyses, it is known that the cat harbors a large proportion of clostridial species, as well as Lactobacillus spp., Escherichia spp., and, occasionally, bifidobacteria.14,26 These techniques, however, do not provide a broad enough scope with which to determine the full microbiome that is harbored by the cat. Next-generation pyrosequencing techniques allow a more in-depth view of this community.29 Analysis of 16S rRNA clones from cats demonstrated that the gastrointestinal tract of the cat is largely populated by Clostridiales.24 While elucidating the microbial composition of an environment is important to understanding host health response, it is not simply the presence of this community that is important. It is possible that the host animal metagenome, consisting of genes produced by the host animal plus those of the microbial communities harbored by these hosts, contains 100-fold more Received: July 23, 2012 Published: October 17, 2012 5924

dx.doi.org/10.1021/pr3006809 | J. Proteome Res. 2012, 11, 5924−5933

Journal of Proteome Research

Article

genes than believed to be produced by the host animal alone.2 Research into the metagenomes of mice and humans has demonstrated the vast array of microbial functions present in the intestinal microbiome, including those associated with energy harvest, macronutrient metabolism and transport, and the ability to repair and replicate microbial cells,2,15,23,31,32 and this research has just begun in the cat.29 Investigating the metagenome of this community helps researchers understand the functionality of the microbiota within the host animal. The objective of this research was to elucidate the fecal microbiome and metagenome of the cat as affected by diets varying in supplemental fiber sources, namely cellulose, fructooligosaccharides, or pectin, using 454 pyrosequencing techniques and compare these genomes to bacterial metagenomes of other species.



DNA Extraction

Genomic DNA was extracted and isolated from fecal samples using a modification of the method of Yu and Morrison.37 Briefly, sterile glass beads (0.2 g of 0.5 mm and 0.07 g of 0.1 mm; Biospec Products, Inc., Bartlesville, OK) were added to each sample (∼200 mg) to facilitate the disruption of feces by vortexing. The aggressive bead-beater steps of the assay were not performed to reduce DNA shearing that may affect the nebulization of genomic DNA for random pyrosequencing. After extraction, DNA was quantified using a ND-1000 spectrophotometer (Nanodrop Technologies, Wilmington, DE). Finally, all four samples from each treatment were sequenced individually. Pyrosequencing and Bioinformatics

Samples were subjected to pyrosequencing using a 454 Genome Sequencer using FLX titanium reagents (Roche Applied Science, Indianapolis, IN). Both unassembled reads and assembled contigs were analyzed separately. Sequences derived from pyrosequencing were quality trimmed based upon N50 values. For unassembled analyses, the data set was depleted of sequences 10; remaining 12% is composed of glucose and fructose) (FOS; SynergyC, BENEO Group, Tienen, Belgium), or high-methoxy pectin (HM Pectin, TIC Gums, White Marsh, MD; viscosity: ∼1150 cP). All diets were similar in crude protein (mean = 34.2%), fat (mean = 20.7%), and ash (mean = 8.8%) composition. Diets contained 4% intrinsic fiber, which brought the total dietary fiber concentration to approximately 8%. Complete dietary ingredient and chemical composition are presented in Barry et al.3

p

djk =

∑i = 1 |δijk| P

, where δijk = zij − zjk

CAZy Analysis

A total of 3 238 600 of unassembled, filtered sequences were compared against the carbohydrate-active enzyme database (CAZy).6 Sequences were attributed to protein families in a two round approach: (i) an initial BLASTX analysis (PMID: 9254694) was performed using the BLOSUM62 substitution matrix against a library composed of all prokaryote and unclassified CAZy protein sequences with all results having an E-value above 1 × 10−6 being filtered out; (ii) the remaining sequences were subject to a FASTX analysis (PMID: 9403055) against a library exclusively composed of catalytic and carbohydrate-binding modules derived from CAZy, all the hits satisfying a criteria E-value of 1 × 10−6 being retained. This approach ensures that only the hits against the core functional elements of CAZy proteins are retained, and is an adaptation of the analysis regularly performed by our system to analyze GenBank updates or perform the analysis of new genomes.6 Given the high volume of sequences to be analyzed, the computation was performed on the grid hosted by the Computing Department of the Centre de Physique des Particules de Marseille (CPPM), Marseille, France. In order to compare the results obtained from the different samples, the number of CAZy family hits was adjusted to that expected by the analysis of a hypothetical set of 300K prior to a

Experimental Procedures

A replicated 3 × 3 Latin square design with three 30-d periods was used. Cats were randomly assigned to one of three diets in the first period and received the other diets in the second and third periods. Cats were fed twice daily to meet their individual metabolic energy requirements based on the equation of Edstadtler-Pietsch (2003) following NRC20 recommendations. At each feeding, refusals from the previous feeding were collected and weighed. After a 26-d diet adaptation phase, a fresh fecal sample was collected during a 4-d collection phase. Fresh feces were immediately flash-frozen in liquid nitrogen and stored at −80 °C until DNA extraction. 5925

dx.doi.org/10.1021/pr3006809 | J. Proteome Res. 2012, 11, 5924−5933

Journal of Proteome Research

Article

Figure 1. Fecal microbial communities of cats fed diets containing 4% cellulose (a), fructooligosaccharides (FOS; b), or pectin (c) as determined by metagenomics rapid analysis server technique (MG-RAST) conducted by 454 pyrosequencing using Titanium reagents.

Ward method34 was performed following distance analysis. All these were performed using GINKGO7 (http://pfgrc.jcvi.org/ index.php/bioinformatics/ginkgo.html). The trees correspond-

distance and clustering analysis. Distances between the different sets of results were obtained using the Bray−Curtis and ChiSquare distance analysis. Agglomerative clustering with the 5926

dx.doi.org/10.1021/pr3006809 | J. Proteome Res. 2012, 11, 5924−5933

Journal of Proteome Research

Article

RAST.18 There were 32 750 hits against the nonredundant protein database. For the FOS data set, 71.8% of the sequences matched the SEED protein nonredundant database, with 30 274 nonredundant hits. For the pectin data set, 73.3% of the sequences matched the SEED protein nonredundant database, with 31 035 nonredundant hits. All three metagenomes had a similar microbial profile when viewed at the major taxonomic levels when using functional genes for identification. Bacteroidetes/Chlorobi group and Firmicutes were the predominant phyla in the feline fecal samples, which represented 36.1 and 36.3% of all sequences, respectively. Proteobacteria (12.4%) and Actinobacteria (7.7%) were the other predominant phyla present in these samples. Statistics comparing phylum proportions among samples demonstrated that cats consuming pectin-supplemented diets had a greater percentage of Firmicutes (P = 0.05) than cellulose- or FOS-supplemented diets (Figure 1; Supporting Information Table 3). Microbiota of cats consuming FOS had a greater percentage of Actinobacteria (P < 0.05) than cats consuming diets supplemented with cellulose or pectin. Using the database associated with the integrated microbial genomes system for microbial genomes (IMG/M), a metagenome gene count (Supporting Information Table 4) was generated. Metagenome gene counts reflected a similar pattern in proportions of phyla present in the cat fecal samples. These data seem to agree well with the dominant phyla determined by MG-RAST. Considering the metagenome gene counts, several phyla within the cat’s fecal microbiome were affected by diet. Actinobacteria increased (P < 0.05) as a result of microbial adaptation to FOS compared to adaptation to the other diets (Supporting Information 3). With pectin supplementation, Chlorobi (P < 0.05), Elusimicrobia (P < 0.05), Proteobacteria (P < 0.05), and total bacteria (P < 0.05) increased compared to cellulose or FOS supplementation. In both analyses of the fecal microbiome of the cat, Actinobacteria increased as a percentage of gene counts. Fructooligosaccharides are one of three true prebiotic compounds25 and have been demonstrated to increase Bifidobacterium spp. in many animal species.9,12,19,35,36 Because Bif idobacterium spp. are members of Actinobacteria, the increase in this phylum was anticipated. The predominant phyla discussed by Tun and collegues,29 the Bacteroidetes/Chlorobi group, were reported at nearly double the concentration analyzed in the current study; however, the dominant phyla are consistent between the two studies. In other studies, Firmicutes were noted as the dominant phyla of the gastrointestinal tract of healthy conventionally raised cats24 and pet cats10 using 16S rRNA gene analysis to determine the feline microbiome. In addition to genetic differences between the cat populations studied here vs those by Tun et al.,29 the diets fed (research diets vs Royal Canin diet), living environment (laboratory in US vs homedwelling cats in Hong Kong), and fecal collection and DNA extraction methodologies used between laboratories may have influenced the microbial taxa measured. Archaea constituted a minor part of the feline metagenome, representing approximately 1% of all sequencing reads using MG-RAST, similar to studies with cats29 and dogs.28 Using MG-RAST, no significant diet effects on archaea were observed. Using IMG/M, however, pectin-supplemented cats had decreased (P < 0.05) total archaea compared to cats fed the other diets. A consideration that must be made with regard to the observation of microbiota and archaea in this study are the

ing to the distance and clustering analysis were calculated with FastME8 and represented using Dendroscope.13 Statistical Analysis

Data (sequence numbers and percentages obtained from the various raw data analyses) were analyzed by the MIXED procedure (SAS Inst. Inc., Cary, NC). The statistical model included the random effects of animal and period and the fixed effect of treatment. Differences among treatments were determined using least significant differences. Differences among treatment level least-squares means with a probability of P ≤ 0.05 were accepted as statistically significant, and mean differences with 0.05 < P ≤ 0.10 were accepted as trends.



RESULTS AND DISCUSSION Cats fed commercial diets regularly consume low dietary fiber concentrations and do not rely heavily on fermentation to meet their energy requirements.20 However, maintaining a stable intestinal microbiota is crucial for both gastrointestinal and host health. Several attempts have been made to describe the colonic microbiome of the cat previously using plating,26 FISH,14 quantitative PCR,3 16S rRNA gene analysis,24 and also using 454 junior pyrosequencing.29 The latter study29 did not investigate the impact of diet on the microbiome, but was the first to describe the feline gut metagenome. The current data set has direct implications on feline health, but also on human health because cats can serve as surrogates for human pathogens. While a baseline assessment of the microbiota present in the cat gastrointestinal tract was not performed, the cellulose diet was selected to serve as the “control” diet. A diet devoid of supplemental dietary fiber may cause intestinal irregularities in the cat and would provide insufficient fecal matter to be collected for subsequent analysis. Cellulose is not readily fermented by the microbiota that reside in the cat’s gastrointestinal tract,8,27 which permits the diet to serve as an effective control diet that contains a similar amount of fiber as the other dietary treatments. Cats fed FOS and pectin did have altered fecal scores at times during the adaptation phase (score = 4 on 5 point scale on some days), but were considered normal (2.7−2.8 on 5 point scale) by the time samples for analyses were collected. Pyrosequencing generated a total of 4 192 192 sequences with an average read length of 725 bp. Assembly of the 1 282 677 feline cellulose metagenomic sequences resulted in 1 022 173 sequences being assembled into 181 758 contigs with 5591 contigs >2 kb. Average sequence length was 705 bp, with an average of 5.3 sequences per contig. Of the 1 249 104 feline FOS sequences, 1 003 806 sequences were assembled into 168 330 contigs with 5853 contigs >2 kb. Average sequence length was 688 bp, with an average of 5.5 sequences per contig. Assembly of the 1 660 411 feline pectin metagenomic sequences resulted in 1 427 289 sequences being assembled into 169 319 contigs with 7152 contigs >2 kb. Average sequence length was 782 bp, with an average of 8 sequences per contig. While this does not cover the entire metagenome of the cat gastrointestinal tract, it is the greatest coverage to date for the cat. Phylogenetic Analysis of Bacteria, Archaea, Eukarya, Fungi, and Viruses

For the cellulose data set, 72.5% of the sequences evaluated were matched to the SEED protein nonredundant database (using an E-value of 1 × 10−5)22 within the program MG5927

dx.doi.org/10.1021/pr3006809 | J. Proteome Res. 2012, 11, 5924−5933

Journal of Proteome Research

Article

Figure 2. Double dendrogram of phylogenetic analysis of fecal microbial communities of individual cats fed diets containing 4% cellulose (Cel), fructooligosaccharides (FOS), or pectin (Pec) as determined by 454 pyrosequencing. Cat samples named with the following system: Name-Perioddiet, where name and diet are abbreviated. Names: A−D. Diets: Cel, cellulose; FOS, fructooligosaccharides; Pec, pectin.

concentrations of viral sequences were similar to those detected in the present study. Low virus counts were reported in the previous feline metagenomics study.29 Because the gastrointestinal microbiome is complex and diverse, it would be unreasonable to expect large changes in composition (i.e., loss of dominant phyla) based on a relatively small changes in dietary habit. However, it might be reasonable to anticipate clustering of microbiomes by diet when all cats consumed the same diet. Feline fecal samples from the current study were compared to each other in a double-dendrogram format (Figure 2). Likewise, Figure 3 presents clustering using the IMG/M results. In both cases, cats do not appear to cluster according to individual or diet. Despite using cats that were genetically similar (half-siblings) and maintaining a very similar living environment, it appears that 4% supplemental fiber did not result in a dramatic shift in the gastrointestinal microbiota. This stability is thought to be a benefit to the host and other species as well,16,30 because it allows the host to consume a variety of diets without great stress on the body. It is possible, however, that changes in amount of “rare” species or the activity of the microbes present occurred, but was not measured.

methods of extraction and detection used. Because environmental gene tags and not 16S RNA sequences were used for annotation, it is possible that genes related to these species have been transferred to other microbes in the gastrointestinal tract through lateral or horizontal gene transfer. Because two independent databases returned hits against archaeal species, however, the cat most likely harbors these species. It does not appear that these archaea are very active in healthy adult cats because no production of methane was observed in a recent in vitro analysis using the same four cats in the present study.4 The possibility exists, however, that the archaea sequenced in this study were not methanogenic. Thus, methane production cannot be the only method to determine archaea presence or activity. Eukaryotic sequences comprised a small percentage of the overall fecal microbiome of the cats sequenced. Using MGRAST, 0.35% of sequences were identified as eukaryotes, with the largest proportion (0.31%) determined to be Fungi/ Metazoa. The dominant phyla were Apicomplexa (0.21%), Ascomycota (0.25%), and Chordata (0.17%). Ascomycota have been observed previously in the feces of cats.10 Eukaryota were reported to represent approximately four times as many sequences on a percentage basis by Tun et al.;29 however, in both studies, the relative percentages of Eukaryotic sequences observed in the gastrointestinal tract of the cat were low. Plasmids and viruses were observed in very low concentrations in the cat fecal microbiome. Combined with the broadrange host plasmids, total plasmids comprised 0.01% of total sequences in data generated using MG-RAST. Viruses were identified as 0.24% of sequences from MG-RAST. While plasmids were not detected in canine fecal samples,28 the

Metagenomics-based Metabolic Profiles

Approximately half (49.1% for cellulose; 48.7% for FOS; and 50.6% for pectin) of all sequences in this data set were classified metabolically and are summarized in Figure 4 (KEGG from MG-RAST; Supporting Information Table 5) and Figure 5 (COG from IMG/M; Supporting Information Table 6). Not surprisingly, carbohydrates (14.6%); protein metabolism (8.2%); amino acids and derivatives (8.0%); cell wall and 5928

dx.doi.org/10.1021/pr3006809 | J. Proteome Res. 2012, 11, 5924−5933

Journal of Proteome Research

Article

transport and metabolism (10.5%); carbohydrate transport and metabolism (9.1%); replication, recombination, and repair (8.8%); translation, ribosomal structure, and biogenesis (8.5%); cell wall/membrane/envelope biogenesis (6.6%); transcription (6.2%); energy production and conversion (5.7%); and inorganic ion transport and metabolism (4.9%) (Table 1). Not surprisingly, predominant functional categories were related to microbial metabolism. Pathways pertaining to amino acid metabolism and their derivatives were highly prevalent. The prevalence of amino acid metabolism-related genes in the present study was similar to that observed in the human gut metagenome,23,30,32 but slightly higher than what has been observed in the dog gastrointestinal metagenome.28 These pathways likely would increase with an increase in microbial mass (amino acid requirements for replicating cells). The composition and quality of diet may also impact amino acid metabolism, including pathways contributing to their synthesis and degradation. Because the cat is an obligate carnivore, it consumes a diet high in crude protein and amino acids. Although the majority of dietary amino acids and proteins are readily digestible, some enter the colon due to processing (e.g., maillard browning during kibble manufacture) or presence of connective tissue (e.g., high amounts of proline and hydroxyproline) that decrease digestibility. The abundance of genes associated with carbohydrate metabolism in the present study was similar to that observed in the human and dog gut

Figure 3. Clustering analysis of cats by diet based on sequences gathered from IMG/M. Cat samples named with the following system: Name-Period-diet, where name and diet are abbreviated. Names: A− D. Diets: Cel, cellulose; FOS, fructooligosaccharides; Pec, pectin.

capsule (7.2%); DNA metabolism (6.8%); virulence (6.0%); and cofactors, vitamins, prosthetic groups and pigments (5.5%) were the most prominent KEGG pathways represented and are similar to those reported previously for the dog28 and cat.29 The major COG functions represented were amino acid

Figure 4. Metabolic function of fecal microbial communities of cats fed diets containing 4% cellulose, fructooligosaccharides (FOS), or pectin as determined by KEGG using MG-RAST. (a,b) Different superscript letters in the same functional category denote differences (P < 0.05) among treatments for percentage of sequences. (c,d) Different superscript letters in the same functional category denote differences (P < 0.10) among treatments for percentage of sequences. 5929

dx.doi.org/10.1021/pr3006809 | J. Proteome Res. 2012, 11, 5924−5933

Journal of Proteome Research

Article

Figure 5. Metabolic function of fecal microbial communities of cats fed diets containing 4% cellulose, fructooligosaccharides (FOS), or pectin as determined by COG analysis of assembled sequences using IMG/M. (a,b) Different superscript letters in the same functional category denote differences (P < 0.05) among treatments for percentage of sequences. (c,d) Different superscript letters in the same functional category denote differences (P < 0.10) among treatments for percentage of sequences.

contained sources of carbohydrate (e.g., corn; brewer’s rice) and at least 4% dietary fibers of varying fermentability and solubility. Data from CAZy analysis (discussed below) supports the KEGG and COG data. The inclusion of 4% pectin or FOS did not greatly alter gene sequence number of any KEGG or COG categories, but some minor changes were present. Similar to what was observed for the microbial phylogeny, a double dendrogram comparing the functional gene data failed to cluster the cats based on diet or individual (Supporting Information Figure 1). Using MGRAST (KEGG), the metagenome of FOS-fed cats had increased (P < 0.05) amino acid metabolism compared to pectin-fed cats. In contrast, nitrogen metabolism was increased (P < 0.05) in the metagenome of pectin-fed vs FOS-fed or cellulose-fed cats. Within COG data, the number of hits for chromatin structure and dynamics was increased (P < 0.05) and number for coenzyme transport and metabolism decreased (P < 0.05) in FOS-fed cats compared to those fed pectin or cellulose. Increased fecal polyamine concentrations, including tryptamine, putrescine, cadaverine, and spermidine, were observed in cats consuming the FOS and pectin diets.3 Because these compounds are derived directly from amino acids (tryptamine, putrescine, cadaverine, and spermidine are derived from tryptophan, arginine, lysine, and arginine, respectively), it would suggest either an increase in dietary amino acids reaching the colon or increased bacterial amino acid metabolism in general. The same could be said for the general increase in

Table 1. COG Functional Categories C − Energy production and conversion D − Cell cycle control, cell division, and chromosome partitioning E − Amino acid transport and metabolism F − Nucleotide transport and metabolism G − Carbohydrate transport and metabolism H − Coenzyme transport and metabolism I − Lipid transport and metabolism J − Translation, ribosomal structure, and biogenesis K − Transcription L − Replication, recombination, and repair M − Cell wall/membrane/envelope biogenesis N − Cell motility O − Posttranslational modification, protein turnover, and chaperones P − Inorganic ion transport and metabolism Q − Secondary metabolites biosynthesis, transport and catabolism R − General function prediction only S − Function unknown T − Signal transduction mechanisms U − Intracellular trafficking, secretion, and vesicular transport V − Defense mechanisms

metagenomes.23,28,30,32 In addition to microbial enzymes used to target long polysaccharides, pathways related to monosaccharide, disaccharide, and oligosaccharide metabolism were all highly prevalent. Even though cats are carnivores, the abundance of such genes was not surprising because all diets 5930

dx.doi.org/10.1021/pr3006809 | J. Proteome Res. 2012, 11, 5924−5933

Journal of Proteome Research

Article

Figure 6. Double dendrogram of the glycoside hydrolase (GH), polysaccharide lyase (PL), and carbohydrate esterase (CE) families of carbohydrate degrading enzymes found in the metagenome (determined by 454 pyrosequencing) of cats consuming diets containing 4% cellulose, fructooligosaccharides, or pectin using the carbohydrate-active enzyme (CAZy) database. Cat samples named with the following system: NamePeriod-diet, where name and diet are abbreviated. Names: A-D. Diets: Cel, cellulose; FOS, fructooligosaccharides; Pec, pectin.

fructose and glucose for the microbial cells, hence increasing the activity of CBM48 to decrease the release of glycogen. While up-regulation of mannanases is not entirely obvious or intuitive, a small amount of brewer’s yeast (3% of diet) was added as a palatant and, thus, may have acted synergistically with the microbiota associated with the FOS treatment. Pectinadapted microbes had increased (P < 0.05) GH8, GH53, GT9, GT51, GT83, CBM6, and CBM20 genes compared to microbes adapted to cellulose or FOS. Glycoside hydrolase 8 is a chitosanase and cellulase, GH53 is an endogalactanase, GT52 is a murein polymerase, CBM 6 has been shown to bind cellulose and β-1,3- and β-1,4-glucans, CBM20 binds starches, and GT9 is an N-acetylglucosaminyltransferase. Because pectin is a highly viscous molecule, it may have bound starch from the dietary matrix (brewer’s rice and corn were primary starch sources) as the intestinal chyme traveled the length of the small intestine, thus bringing undigested starch to the colon. Why a higher prevalence of enzymes related to cellulose and chitin metabolism occurred in this treatment is unknown. All cats exhibited a high prevalence of GH2 and GH3. Glycoside hydrolase 2 is a β-galactosidase, glucuronidase, and mannosidase, while GH3 is a β-glucosidase and xylosidase. All diets contained brewer’s rice as a carbohydrate source, which may have been modified during the extrusion process and escaped upper gastrointestinal digestion and absorption due to these modifications, as well as brewer’s yeast, which is high in mannans. This data set is not without limitations. First, the number of animals used for sequencing was low. Second, the diets compared were quite similar, with only 4% of the diet being different. The effects of greater dietary differences, including those pertaining to macronutrient shifts (proteins vs carbohydrates vs fat), the inclusion of functional ingredients (e.g., phytochemicals), etc. should be documented in the future. Third, the sequencing depth was not sufficient to determine the effects of diet on the complete phylogeny or functional capacity of the feline gastrointestinal tract, but should provide large changes influenced by dietary fiber type. Lastly, methodology and analytical methods have continued to change rapidly and

nitrogen metabolism genes observed in cats fed pectin. Although FOS does not increase digesta viscosity, pectin is a viscous fiber. Thus, pectin may have decreased protein digestibility in the small intestine, thereby increasing protein or amino acid content reaching the colon, serving as a substrate for microbes. Because pectin and FOS are both rapidly and highly fermentable in the colon of cats, an increase in microbial activity vs cellulose would be expected. While this change in activity can be measured using in vitro fermentation assays by measuring metabolite concentrations, these changes may not be brought out in terms of DNA content, as measured in this study. Thus, mRNA, protein, or metabolite profiles that provide a better picture of microbial metabolic activity may be more appropriate in future studies. Carbohydrate Enzyme-based Metabolic Profiles

The CAZy database is used to determine the genes present in an environment capable of catabolizing carbohydrates.6 To the best of the authors’ knowledge, this is the first data set to report CAZy analysis for the cat gut metagenome (Figure 6). Results of these analyses are presented in Supporting Information Tables 7 (glycoside hydrolase; GH), 8 (glycosyl transferase; GT), 9 (polysaccharide lyase and carbohydrate esterase; PL and CE, respectively), and 10 [carbohydrate binding molecule (CBM) and dockerin]. When cats were adapted to a diet containing FOS or pectin, the GH112 and CBM48 genes were increased (P < 0.05) compared to the cellulose-containing diets. Glycoside hydrolase 112 is a phosphorylase, and CBM48 often is appended to GH13 molecules and can be found on the glycogen-binding beta subunit of AMP-activated protein kinases. The glycoside hydrolase 38 and CBM23 genes were increased (P < 0.05) when fecal microbes were adapted to FOS compared to adaptation to cellulose. The GT66 gene was increased (P < 0.05) when fecal microbes were adapted to FOS compared to adaptation to pectin. Glycoside hydrolase 38 is a mannanase, GT66 is a β-oligotransferase, and CBM23 has been shown to bind mannans. The oligotransferase function is most likely up-regulated because of the dietary FOS in this treatment group. This action may provide enough energy in the form of 5931

dx.doi.org/10.1021/pr3006809 | J. Proteome Res. 2012, 11, 5924−5933

Journal of Proteome Research

Article

(5) Bueno, A. R.; Cappel, T. G.; Sunvold, G. D.; Moxley, R. A.; Reinhart, G. A.; Clemens, E. T. Feline colonic microbes and fatty acid transport: Effects of feeding cellulose, beet pulp and pectin/gum arabic fibers. Nutr. Res. (N.Y.) 2000, 20, 1319−1328. (6) Cantarel, B. L.; Coutinho, P. M.; Rancurel, C.; Bernard, T.; Lombard, V.; Henrissat, B. The Carbohydrate-Active EnZymes database (CAZy): An expert resource for Glycogenomics. Nucleic Acids Res. 2009, 37, D233−D238. (7) De Caceres, M.; Oliva, F.; Font, X.; Vives, S. Ginkgo, a program for non-standard multivariate fuzzy analysis. Adv. Fuzzy Sets Syst. 2007, 2, 41−56. (8) Desper, R.; Gascuel, O. Fast and accurate phylogeny reconstruction algorithms based on the minimum-evolution principle. J. Comput. Biol. 2002, 9, 687−705. (9) Estrada, A.; Drew, M. D.; Van Kessel, A. Effect of the dietary supplementation of fructooligosaccharides and Bif idobacterium longum to early-weaned pigs on performance and fecal bacterial populations. Can. J. Anim. Sci. 2001, 81, 141−148. (10) Handl, S.; Dowd, S. E.; Garcia-Mazcorro, J. F.; Steiner, J. M.; Suchodolski, J. S. Massive parallel 16S rRNA gene FLX-Titanium amplicon pyrosequencing reveals species-specific differences in fecal bacterial and fungal communities in healthy dogs and cats. Proceedings of the 14th Congress of the European Society of Veterinary and Comparative Nutrition, Studentendruckerei, University of Zurich: Zurich, Switzerland, 2010. (11) Hesta, M.; Janssens, G. P. J.; Debraekeleer, J.; De Wilde, R. The effect of oligofructose and inulin on faecal characteristics and nutrient digestibility in healthy cats. J. Anim. Physiol. Anim. Nutr. 2001, 85, 135−141. (12) Howard, M. D.; Gordon, D. T.; Pace, L. W.; Garleb, K. A.; Kerley, M. S. Effects of dietary supplementation with fructooligosaccharides on colonic microbiota populations and epithelial cell proliferation in neonatal pigs. J. Pediatr. Gastroenterol. Nutr. 1995, 21, 297−303. (13) Huson, D. H.; Richter, D. C.; Rausch, C.; Dezulian, T.; Franz, M.; Rupp, R. Dendroscope: An interactive viewer for large phylogenetic trees. BMC Bioinform. 2007, 8, 460. (14) Inness, V. L.; McCartney, A. L.; Khoo, C.; Gross, K. L.; Gibson, G. R. Molecular characterization of the gut microflora of healthy and inflammatory bowel disease cats using fluorescence in situ hybridization with special reference to Desulfovibrio spp. J. Anim. Physiol. Anim. Nutr. 2007, 91, 48−53. (15) Kurokawa, K.; Itoh, T.; Kuwahara, T.; Oshima, K.; Toh, H.; Toyoda, A.; Takami, H.; Morita, H.; Sharma, V. K.; Srivastava, T. P.; Tayor, T. D.; Noguchi, H.; Mori, H.; Ogura, Y.; Ehrlich, D. S.; Itoh, K.; Takagi, T.; Sakaki, Y.; Hayashi, T.; Hattori, M. Comparative metagenomics revealed commonly enriched gene sets in human gut microbiomes. DNA Res. 2007, 14, 169−181. (16) Ley, R. E.; Hamady, M.; Lozupone, C.; Turnbaugh, P. J.; Ramey, R. R.; Bircher, J. S.; Schlegel, M. L.; Tucker, T. A.; Schrenzel, M. D.; Knight, R.; Gordon, J. I. Evolution of mammals and their gut microbes. Science 2008, 320, 1647−1651. (17) Markowitz, V. M.; Ivanova, N. N.; Szeto, E.; Palaniappan, K.; Chu, K.; Dalevi, D.; Chen, I.-M. A.; Grechkin, Y.; Dubchak, I.; Anderson, I.; Lykidis, A.; Mavromatis, K.; Hugenholtz, P.; Kyrpides, N. C. IMG/M: A data management and analysis system for metagenomes. Nucleic Acid Res. 2008, 36, D534−D538. (18) Meyer, F.; Paarmann, D.; D’Souza, M.; Olson, R.; Glass, E. M.; Kubal, M.; Paczian, T.; Rodriguez, A.; Stevens, R.; Wilke, A.; Wilkening, J.; Edwards, R. A. The Metagenomics RAST server - a public resource for the automatic phylogenetic and functional analysis of metagenomes. BMC Bioinform. 2008, 9, 386−393. (19) Middelbos, I. S.; Godoy, M. R.; Fastinger, N. D.; Fahey, G. C., Jr. A dose-response evaluation of spray-dried yeast cell wall supplementation of diets fed to adult dogs: Effects on nutrient digestibility, immune indices, and fecal microbial populations. J. Anim. Sci. 2007, 85, 3022−3032. (20) National Research Council. Nutrient Requirements of Dogs and Cats; National Academies Press: Washington, D.C., 2006.

databases have expanded since these analyses were performed. As time progresses, scientists will be able to gain a deeper understanding of the feline gastrointestinal tract metagenome. At this time, however, this is the most comprehensive data set of its type, and provides microbiologists and nutritionists with more knowledge of the feline gastrointestinal microbiome and how it may be impacted by dietary intervention.



CONCLUSION This data set is one of the first to evaluate the effects of dietary fiber on the feline gastrointestinal metagenome using MGRAST, IMG/M, and CAZy profiles. While the percentage of several microbial and functional gene sequences were changed with respect to diet, the overall gene counts and, thus, the microbiome itself, was not heavily modified by 4% of the dietary fiber sources tested. Given the lack of change in the feline metagenome, it appears to be largely unaffected by different dietary fibers, indicating that microbial function is highly conserved in the gastrointestinal tract. Cat to cat variation appeared to greatly impact the results of this study, however, justifying the need to study larger cat populations in the future. This data set provides much information regarding the microbiome of the healthy feline laboratory cat, and may be a useful tool to investigate changes to the microbiome and metagenome of clinically diseased cats. As such, this data set provides baseline microbiome and metagenome values with which to base future investigations into the gastrointestinal microbiome of the cat.



ASSOCIATED CONTENT

S Supporting Information *

Supplementary tables and figure. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*Tel: +1 217 333 4189. Fax: +1 217 333 7861. E-mail: [email protected]. Present Addresses ‡ Mars Petcare, Inc., 315 Cool Springs Blvd., Franklin, TN 37067. § Novus International, Inc., 20 Research Park Dr., St. Charles, MO 63304.

Notes

The authors declare no competing financial interest.



REFERENCES

(1) AAFCO Official Publication, 100th ed.; Association of American Feed Control Officials: Oxford, IN, 2010. (2) Backhed, F.; Ding, H.; Wang, T.; Hooper, L. V.; Koh, G. Y.; Nagy, A.; Semenkovich, C. F.; Gordon, J. I. The gut microbiota as an environmental factor that regulates fat storage. Proc. Natl. Acad. Sci. 2004, 101, 15718−15723. (3) Barry, K. A.; Wojcicki, B. J.; Middelbos, I. S.; Vester, B. M.; Swanson, K. S.; Fahey, G. C., Jr. Dietary cellulose, fructooligosaccharides, and pectin modify fecal protein catabolites and microbial populations in adult cats. J. Anim. Sci. 2010, 88, 2978−2987. (4) Barry, K. A.; Wojcicki, B. J.; Bauer, L. L.; Middelbos, I. S.; Vester Boler, B. M.; Swanson, K. S.; Fahey, G. C., Jr. Adaptation of healthy adult cats to select dietary fibers in vivo affects gas and short-chain fatty acid production from fiber fermentation in vitro. J. Anim. Sci. 2011, 89, 3163−3169. 5932

dx.doi.org/10.1021/pr3006809 | J. Proteome Res. 2012, 11, 5924−5933

Journal of Proteome Research

Article

(21) Nelson, R. W.; Scott-Moncrieff, J. C.; Feldman, E. C.; DeVriesConcannon, S. E.; Kass, P. H.; Davenport, D. J.; Kiernan, C. T.; Neal, L. A. Effect of dietary insoluble fiber on control of glycemia in cats with naturally acquired diabetes mellitus. J. Amer. Vet. Med. Assoc. 2000, 216, 1082−1088. (22) Overbeek, R.; Begley, T.; Butler, R. M.; Choudhuri, J. V.; Chuang, H. Y.; Cohoon, M.; de Crecy-Lagard, V.; Diaz, N.; Disz, T.; Edwards, R.; Fonstein, M.; Frank, E. D.; Gerdes, S.; Glass, E. M.; Goesmann, A.; Hanson, A.; Iwata-Reuyl, D.; Jensen, R.; Jamshidi, N.; Krause, L.; Kubal, M.; Larsen, N.; Linke, B.; McHardy, A. C.; Meyer, F.; Neuweger, H.; Olsen, G.; Olson, R.; Osterman, A.; Portnoy, V.; Pusch, G. D.; Rodionov, D. A.; Ruckert, C.; Steiner, J.; Stevens, R.; Thiele, I.; Vassieva, O.; Ye, Y.; Zagnitko, O.; Vonstein, V. The subsystems approach to genome annotation and its use in the project to annotate 1000 genomes. Nucleic Acids Res. 2005, 33, 5691−5702. (23) Qin, J.; Li, R.; Raes, J.; Arumugam, M.; Burgdorf, K. S.; Manichanh, C.; Nielson, T.; Pons, N.; Levenez, F.; Yamada, T.; Mende, D. R.; Li, J.; Xu, J.; Li, S.; Li, D.; Cao, J.; Wang, B.; Liang, H.; Zheng, H.; Xie, Y.; Tap, J.; Lepage, P.; Bertelan, M.; Batto, J.-M.; Hansen, T.; Le Paslier, D.; Linneburg, A.; Bjorn Nielsen, H.; Pelletier, E.; Renault, P.; Sicheritz-Ponten, T.; Turner, K.; Zhu, H.; Yu, C.; Li, S.; Jian, M.; Zhou, Y.; Li, Y.; Zhang, X.; Li, S.; Qin, N.; Yang, H.; Wang, J.; Brunak, S.; Dore, J.; Guarner, F.; Kristiansen, K; Pedersen, O.; Parkhill, J.; Weissenbach, J.; Bork, P.; Dusko Erlich, S.; Wang, J. A human gut microbial gene catalogue established by metagenomic sequencing. Nature 2010, 464, 59−65. (24) Ritchie, L. E.; Steiner, J. M.; Suchodolski, J. S. Assessment of microbial diversity along the feline intestinal tract using 16SrRNA gene analysis. FEMS Microbiol. Ecol. 2008, 66, 590−598. (25) Roberfroid, M. Prebiotics: Concept, definition, criteria, methodologies, and products. In Handbook of Prebiotics; Gibson, G. R., Roberfroid, M., Eds.; Taylor and Francis Group, LLC: Boca Raton, FL, 2008; pp 39−68. (26) Sparkes, A. H.; Papasouliotis, K.; Sunvold, G.; Werrett, G.; Gruffydd-Jones, E. A.; Egan, K.; Gruffydd-Jones, T. J.; Reinhart, G. Effect of dietary supplementation with fructooligosaccharides on fecal flora of healthy cats. Am. J. Vet. Res. 1998, 59, 436−440. (27) Sunvold, G. D.; Fahey, G. C.; Merchen, N. R.; Bourquin, L. D.; Titgemeyer, E. C.; Bauer, L. L.; Reinhart, G. A. Dietary fiber for cats: In vitro fermentation of selected fiber sources by cat fecal inoculum and in vivo utilization of diets containing selected fiber sources and their blends. J. Anim. Sci. 1995, 73, 2329−2339. (28) Swanson, K. S.; Dowd, S. E.; Suchodolski, J. S.; Middelbos, I. S.; Vester, B. M.; Barry, K. A.; Nelson, K. E.; Torralba, M.; Henrissat, B.; Coutinho, P. M.; Cann, I. K. O.; White, B. A.; Fahey, G. C., Jr. Phylogenetic and gene-centric metagenomics of the canine gastrointestinal microbiome reveals similarities with human and mouse gut metagenomes. ISME J. 2011, 5, 639−649. (29) Tun, H. M.; Brar, M. S.; Khin, N.; Jun, L.; Hui, R.K.-H.; Dowd, S. E.; Leung, F.C.-C. Gene-centric metagenomics analysis of feline intestinal microbiome using 454 junior pyrosequencing. J. Microbiol. Methods 2012, 88, 369−376. (30) Turnbaugh, P. J.; Hamady, M.; Yatsunenko, T.; Cantarel, B. L.; Duncan, A.; Ley, R. E.; Sogin, M. L.; Jones, W. J.; Roe, B. A.; Affourtit, J. P.; Egholm, M.; Henrissat, B.; Heath, A. C.; Knight, R.; Gordon, J. I. A core gut microbiome in obese and lean twins. Nature 2009, 457, 480−484. (31) Turnbaugh, P. J.; Ley, R. E.; Mahowald, M. A.; Magrini, V.; Mardis, E. R.; Gordon, J. I. An obesity-associated gut microbiome with increased capacity for energy harvest. Nature 2006, 444, 1027−1031. (32) Vaishampayan, P. A.; Kuehl, J. V.; Froula, J. L.; Morgan, J. L.; Ochman, H.; Francino, M. P. Comparative metagenomics and population dynamics of the gut microbiota in mother and infant. Genome Biol. Evol. 2012, 2, 53−66. (33) Verbrugghe, A.; Hesta, M.; Gommeren, K.; Daminet, S.; Wuyts, B.; Buyse, J.; Janssens, G. P. J. Oligofructose and inulin modulate glucose and amino acid metabolism through propionate production in normal-weight and obese cats. Br. J. Nutr. 2006, 102, 694−702.

(34) Ward, J. H. Hierarchical grouping to optimize an objective function. J. Am. Stat. Assoc. 1963, 58, 236−244. (35) Xu, Z. R.; Hu, C. H.; Xia, M. S.; Zhan, X. A.; Wang, M. Q. Effects of dietary fructooligosaccharide on digestive enzyme activities, intestinal microflora and morphology of male broilers. Poult. Sci. 2003, 82, 1030−1036. (36) Xu, Z. R.; Zuo, X. T.; Hu, C. H.; Xia, M. S.; Zhan, X. A.; Wang, M. Q. Effects of dietary fructooligosaccharide on digestive enzyme activities, intestinal microflora and morphology of growing pigs. AsianAust. J. Anim. Sci. 2002, 15, 1784−1789. (37) Yu, Z.; Morrison, M. Improved extraction of PCR-quality communityDNA from digesta and fecal samples. Biotechniques 2004, 36, 808−812.

5933

dx.doi.org/10.1021/pr3006809 | J. Proteome Res. 2012, 11, 5924−5933