1st Page

Journal of Animal Science - 2011 and 2012 Early Careers Achievement Awards

2011 AND 2012 EARLY CAREERS ACHIEVEMENT AWARDS: Use of genomic biology to study companion animal intestinal microbiota1


This article in JAS

  1. Vol. 91 No. 6, p. 2504-2511
    Received: Dec 31, 2012
    Accepted: Feb 24, 2013
    Published: November 25, 2014

    2 Corresponding author(s):

  1. K. R. Kerr*†,
  2. A. N. Beloshapka and
  3. K. S. Swanson 2
  1. Division of Nutritional Sciences
    Department of Animal Sciences
    Department of Veterinary Clinical Medicine, University of Illinois, Urbana 61801


Although dogs and cats are quite different than many livestock species in that they have evolved by eating diets high in fat and protein and low in carbohydrates, the gastrointestinal microbiota still play a key role in the gut and overall host health of these species. Early experiments in this field used culture-based techniques to evaluate the effects of dietary ingredients, such as fibers and prebiotics, on microbiota and indices of gut health (e.g., fecal scores, pH, fermentative end products). Such studies, however, were limited in scope and lacked precision as it pertained to the microbiota. The DNA-based techniques that have become available over the past decade have greatly upgraded research capabilities and have provided a more encompassing view of the canine and feline gastrointestinal microbiomes. High-throughput sequencing techniques that are much cheaper and faster than Sanger sequencing have been a key development in this progress. Sequence data not only allow for the identification of all microbial taxa but also provide information regarding functional capacity when a shotgun sequencing approach is used. The few canine and feline studies that have used 454 pyrosequencing have identified the predominant microbial taxa and metabolic functions present in healthy populations, differences between healthy and diseased dog and cat populations, and the effects of diet (e.g., dietary fibers, prebiotics, protein to carbohydrate ratio) on gastrointestinal microbiota. Although these studies have provided a foundation from which to work, more research is needed to increase our general understanding of the gastro-intestinal microbiome, how it impacts host health, how its composition and activity may be altered by age, genetic, or environmental factors, and test whether specific pathogens or disease signatures can be identified and used in diagnosis and/or treatment of disease.


A complex and dense population of microorganisms, including bacteria, fungi, protozoa, and viruses, inhabit the feline and canine gastrointestinal tracts (GIT). Microbes are present throughout the entire GIT but are at greatest concentrations in the colon (109 to 1011 cfu/g digesta contents). Despite their simple GIT and natural selection for diets high in fat and protein and low in carbohydrates (Hewson-Hughes et al., 2011, 2013), recent studies have revealed that the feline and canine GIT microbiome is highly diverse, with several hundred phylotypes represented (Swanson et al., 2011; Barry et al., 2012).

A balanced gut microbiota is important for host health because of its role in numerous nutritional, developmental, immunological, and physiologic processes (Mackie et al., 1999; Hooper et al., 2001). In humans, alterations in the GIT microbiome have been associated with colon cancer (Flint et al., 2007), inflammatory bowel diseases (IBD; Friswell et al., 2010), and complex metabolic diseases such as obesity (Ley et al., 2006; Turnbaugh et al., 2008) and diabetes (Wen et al., 2008). Similar changes have been noted in dogs and cats with IBD (Inness et al., 2007; Janeczko et al., 2008; Suchodolski et al., 2010) and chronic diarrhea (Bell et al., 2008; Jia et al., 2010). Although those reports emphasize the importance of maintaining a balanced intestinal microbial ecosystem, cause–effect relationships are not so easily defined. In fact, many enteropathogens (e.g., Escherichia coli, Salmonella) are present in both diseased and healthy animals (Queen et al., 2012). Therefore, a better understanding of the microbiome, including its phylogenetic structure and functional capacity and how they relate to host health, is needed. The main objective of this review is to summarize recent studies that have used molecular methods to characterize the fecal microbiome of healthy and diseased cats and dogs and identify how diet may influence its composition and functional capacity.


In the past, knowledge regarding GIT microbes was primarily obtained using culture-based techniques. Because only a fraction of the organisms present in the GIT can be cultured and studied, progress in the field was greatly hindered until the recent availability of molecular assays. Multiple DNA-based, culture-independent methods for microbiome analysis have recently emerged and may be useful tools to effectively identify and quantify microbial populations. Several molecular tools based on the microbial 16S rRNA gene are available including quantitative PCR (qPCR), fluorescent in situ hybridization (FISH), gel-based techniques such as RFLP analysis, denaturing gradient gel electrophoresis (DGGE), and temperature gradient gel electrophoresis (TGGE), and sequencing techniques such as 454 pyrosequencing (Roche Applied Science, Indianapolis, IN), Illumina (Illumina Inc., San Diego, CA) sequencing, and Sanger sequencing. The next-generation sequencing techniques (i.e., 454 pyrosequencing, Illumina) are also commonly used for shotgun sequencing and functional assessment. These methods are briefly summarized in Table 1.

View Full Table | Close Full ViewTable 1.

Molecular methods used for microbiome analysis1

Method2 Description Characteristics Primary advantages Primary disadvantages
qPCR Amplifies and quantifies a targeted DNA molecule Dye or probe used to bind double-stranded DNA, which causes intensity of fluorescent emissions to increase Low cost; high sensitivity allows for detection of sequences at low concentrations Limited in scope
FISH Sensitive detection of specific nucleic acid sequences in metaphase or interphase cells Manual procedure of biological samples; fluorescence intensities measured using FLEX (a quantitative fluorescence microscope system) Allows for localization and study of spatial organization of cells as they occur in their natural habitat Costly; not easily scalable for disease screenings
RFLP High-throughput fingerprinting technique used to explore changes in structure and composition of microbial communities. DNA sample digested by restriction enzymes to characterize microbiota of specific regions. Fragments then separated according to length by gel electrophoresis. Provides a broad view of microbial systems Primers not specific
DGGE PCR-amplified 16S rRNA fragments separated on polyacrylamide gel containing gradient of denaturant (e.g., urea, formamide) Gel-based method of fingerprinting Provides a broad view of microbial systems Only semiquantitative and insensitive
TGGE PCR-amplified 16S rRNA fragments separated on polyacrylamide gel containing gradient of temperatures Gel-based method of fingerprinting Generate qualitative differences in microbial ecology Only semiquantitative and insensitive
454 sequencing Pyrosequencing light emission 400 to 600 base reads 16S coverage is good Cost limits shotgun coverage
Illumina sequencing Fluorescent, stepwise sequencing 100 to 150 base reads Very high coverage owing to high instrument output and very low cost Increased bioinformatics costs and time
Sanger sequencing Fluorescent, dideoxy terminator 750 base reads or greater High read length and accuracy Compared with next-generation sequencing, is costly and has low throughput
2qPCR = quantitative PCR; FISH = fluorescent in situ hybridization; DGGE = denaturing gradient gel electrophoresis; TGGE = temperature gradient gel electrophoresis.

The following citations provide a more in-depth review of the techniques for qPCR (Rigottier-Gois et al., 2003; Richards et al., 2005), FISH (Pinkel et al., 1986), RFLP (Liu et al., 1997; Schütte et al., 2008), DGGE and TGGE (Muyzer et al., 1993; Zoetendal et al., 1998), and sequencing (Petrosino et al., 2009; Weinstock, 2012). Over the past decade, these molecular techniques have dramatically changed the research landscape and have greatly enhanced our understanding of the composition, dynamics, and functionality of the host–microbiota ecosystem in dogs and cats (Ritchie et al., 2008; Desai et al., 2009; Suchodolski, 2011a,b; Swanson et al., 2011; Barry et al., 2012).


The major phyla present in the canine GIT are Firmicutes, Bacteroidetes, Proteobacteria, Fusobacteria, and Actinobacteria; however, proportions reported vary among studies. Xenoulis et al. (2008) constructed 16S rRNA gene clone libraries from duodenal contents of healthy dogs. Those researchers concluded that the following 6phyla were present in the healthy control dogs: Firmicutes (46.4% of clones), Proteobacteria (26.6% of clones), Bacteroidetes (11.2% of clones), Spirochaetes (10.3% of clones), Fusobacteria (3.6% of clones), and Actinobacteria (1.0% of clones). One-half of those sequences belonged to 3 orders: Clostridiales (19.6% of clones), Lactobacillales (14.1% of clones), and Campylobacteriales (13.9% of clones; Xenoulis et al., 2008). Suchodolski et al. (2008a), who also used gene clone libraries, reported increasing bacterial diversity from the duodenum to the colon. In that study, Clostridiales (approximately 40% of clones) and Enterobacteriales (approximately 30% of clones) were the predominant orders in the duodenum and jejunum whereas the ileum and colon were predominated by Clostridiales, Fusobacteriales, and Bacteroidiales (each approximately 25 to 30% of clones).

Middelbos et al. (2010), the first to use 16S rRNA gene pyrosequencing to study fecal microbial communities of healthy dogs, reported a predominance of the phyla Firmicutes (15 to 28% of sequences), Bacteroidetes (32 to 34% of sequences), Fusobacteria (24 to 40% of sequences), Actinobacteria (0.8 to 1.4% of sequences), and Proteobacteria (5 to 6% of sequences). Subsequently, multiple studies have reported data for fecal microbial populations of healthy pet and laboratory dogs measured using 454 pyrosequencing (Garcia-Mazcorro et al., 2011; Handl et al., 2011, 2013; Suchodolski et al., 2012b; Beloshapka et al., 2013). Whereas the predominant phyla are the same, proportions vary greatly within (i.e., from dog to dog) and among studies [Firmicutes (27 to 97% of sequences), Fusobacteria (<1 to 50% of sequences), Bacteroidetes (<1 to 19% of sequences), Proteobacteria (<1 to 6% of sequences), and Actinobacteria (<1 to 3% of sequences)].

Several sources of variation may exist and contribute to the variability observed among studies, including animal, sample type, and methodological differences. Animal genetics, diet (e.g., macronutrient composition or form), age, and living conditions (e.g., exposure to pathogens, people, or other animals) may all be sources of variation. Differences in methodology, including sample collection, handling, and storage procedures, DNA extraction methodology, primer sequences, and sequencing method and depth, may all contribute. Finally, microbial variability may be attributed to sample type, as microbial populations of mucosal and digesta samples from the stomach, duodenum, jejunum, ileum, and colon would be expected to differ from each other and from fecal samples because of the differences in pH, substrate availability, host secretions, etc.

A metagenomics approach using shotgun sequencing to characterize the microbiome can be useful because this approach not only identifies all inhabitants of the gut but also provides information about the possible functions of the microbial genome. Swanson et al. (2011) used shotgun sequencing and reported the phylogeny and functional capacity of the healthy dog gastrointestinal (GI) microbiome, analyzing the same fecal samples from the study by Middelbos et al. (2010). As it pertains to functional capacity, Swanson et al. (2011) reported that the following functional categories were most prevalent in the canine GI microbiome: carbohydrates (12 to 13% of sequences), protein metabolism (8 to 9%), DNA metabolism (7%), cell wall and capsule (7 to 8%), AA and derivatives (7%), virulence (6 to 7%), and cofactors, vitamins, prosthetic groups, and pigments (6%).

The canine GI metagenome in Swanson et al. (2011) consisted of sequences from bacteria (98%), fungi (0.4%), archaea (1%), and viruses (0.3 to 0.4%). Of the 19 bacterial phyla, Bacteroidetes (36 to 37% of sequences), Firmicutes (31 to 35% of sequences), Proteobacteria (13 to 15% of sequences), Fusobacteria (7 to 9% of sequences), and Actinobacteria (1% of sequences) predominated. The proportions were different than those reported by Middelbos et al. (2010). Although it is difficult to identify the source of variation between these studies, it had to be methodology related because the same fecal samples were used in both studies. Biases involved with the generation of amplicons (e.g., primer bias) for the 16S rRNA gene-based method may have contributed to this discrepancy.

Archaea from 2 phyla, Crenarchaeota and Euryarchaeota, have been identified in the intestine and feces of dogs; however, few data are available on relative proportions of each phylum (Suchodolski, 2011b; Swanson et al., 2011). As it pertains to fungal sequences, Suchodolski et al. (2008b), who used PCR, reported that all fungal phylotypes from the duodenal contents of healthy dogs were identified as belonging to 2 phyla, Ascomycota and Basidiomycota. More recently, the fungal community has been examined using 18S rRNA gene pyrosequencing (Handl et al., 2011; Suchodolski, 2011b). Suchodolski (2011b)reported sequences from canine and feline samples belonged predominantly to 2 fungal phyla [Ascomycota (>90% of fungal sequences) and Neocallimastigomycota (>5% of fungal sequences)]. Handl et al. (2011) also reported high proportions of Ascomycota (99.6% of fungal sequences). Saccharomyces (86% of fungal sequences) of the Ascomycota phylum were reported to be the most abundant fungal genera (Handl et al., 2011).

Similar to the dog, Firmicutes, Bacteroidetes, Proteobacteria, Fusobacteria, and Actinobacteria are the predominant phyla present in the feline GIT. Using traditional Sanger sequencing on constructed 16S rRNA gene clone libraries, the predominant bacterial phyla present in healthy cats were reported to be Firmicutes (68 to 87% of clones), Proteobacteria (8 to 14% of clones), Bacteroidetes (2 to 10% of clones), Actinobacteria (2 to 4% of clones), and Fusobacteria (<1 to 5% of clones; Ritchie et al., 2008, 2010). Handl et al. (2011) and Garcia-Mazcorro et al. (2011), who both used 454 pyrosequencing, reported that Firmicutes (92 to 95% of sequences) and Actinobacteria (4 to 7% of sequences) were the predominant bacterial phyla in cat feces, with Proteobacteria, Bacteroidetes, and Fusobacteria representing <1% of sequences.

A metagenomic evaluation of the healthy cat GIT demonstrated that the metagenome consisted of 97 to 99% bacteria, 0.02 to 0.4% fungi, 0.09 to 1% archaea, and 0.09 to 0.25% viruses (Barry et al., 2012; Tun et al., 2012). Bacteroidetes (36 to 68% of sequences), Firmicutes (13 to 36% of sequences), Proteobacteria (6 to 12% of sequences), and Actinobacteria (1 to 8% of sequences) were the predominant bacterial phyla present (Barry et al., 2012; Tun et al., 2012). Predominant archaeal divisions were Euryarchaeota (67 to 95% of archaeal sequences) and Crenarchaeota (7 to 33% of archaeal sequences; Barry et al., 2012; Tun et al., 2012). Additionally, the phyla Korarcheota, Nanoarchaeota, and Thaumarchaeota are present at lower amounts (<7% of archaea sequences; Barry et al., 2012). These data agreed with that of Suchodolski et al. (2011) who used archaea-specific 16S RNA primers and reported that the most common archaeal phyla in the feline GIT were Crenarchaeota, Euryarchaeota, and Korarchaeota. Among fungal sequences, Aspergillus (11% of fungal sequences) and Saccharomyces (58% of fungal sequences) of the Ascomycota phylum were reported to be the most abundant fungal genera in the feline GIT when measured using 18S rRNA pyrosequencing (Handl et al., 2011). The sources of microbiome variation among feline studies would be the same as those highlighted for canine studies discussed previously.

The functional metagenome of the feline fecal microbiota appears to be similar to that reported for dogs and other species. The most represented functional categories included carbohydrate (15%), clustering-based subsystems (14%), protein metabolism (8%), AA and derivatives (8%), cell wall and capsule (7%), DNA metabolism (7%), virulence (6%), and cofactors, vitamins, prosthetic groups and pigments (6%; Barry et al., 2012). Tun et al. (2012) reported similar findings, with carbohydrate (13%), protein metabolism (9%), DNA metabolism (8%), virulence (7%), and AA metabolism (6%) being the most predominant functional categories present. Despite the carnivorous nature of cats, preference for diets high in protein and fat (Hewson-Hughes et al., 2011), and limited ability to respond to high dietary starch or fiber, the microbiome that inhabits its gut appears to contain functional machinery similar to that of dogs, humans, and rodent models. Further research is necessary, however, to test the functional capacity and plasticity of the feline microbiome and how it compares with microbiomes of other host species in terms of substrate specificity, enzyme kinetics, production of primary and secondary metabolites, and virulence factors and how it may impact host health.


There is growing evidence that the proportions of GI microbes are altered in some disease states, including cancer, GI diseases, and metabolic diseases (Ley et al., 2006; Flint et al., 2007; Turnbaugh et al., 2008; Wen et al., 2008; Friswell et al., 2010). No clear differences in microbial populations were observed between lean and obese dogs in a recent study, however (Handl et al., 2013). Gastrointestinal diseases (e.g., IBD, small intestinal dysbiosis) are relatively common in dogs and cats, yet only a few studies using molecular techniques have examined the relationship between GIT microbial populations and disease in companion animals. Jia et al. (2010), using FISH analysis, aimed to investigate the fecal microbial communities of dogs with chronic diarrhea compared with healthy control dogs. Firmicutes was the most abundant phyla in both chronic diarrhea (21% of total bacteria) and control (31% of total bacteria) dogs. However, LactobacillusEnterococcus was greater (by 8%) in control dogs versus the dogs with chronic diarrhea. Chaban et al. (2012) used pyrosequencing of the cpn60 target gene and reported a decrease in the proportion of Bacteroidetes in dogs with unspecified diarrhea compared with healthy controls (8 vs. 50% of sequences). Broad groupings of disease states, however, may not be adequate to fully profile gut dysbiosis associated with disease. Suchodolski et al. (2012b) reported differences in microbial proportions measured using 454 pyrosequencing for dogs with acute hemorrhagic and nonhemorrhagic diarrhea. For example, compared with healthy dogs, Blautia was decreased (dogs with acute diarrhea regardless of type (0.2 vs. 10% of sequences); however, Clostridium were increased in dogs with nonhemorrhagic diarrhea (82 vs. 34% of sequences) whereas those with hemorrhagic diarrhea were intermediate (44% of sequences).

Differences in microbial populations between healthy dogs and dogs with IBD have also been reported. Xenoulis et al. (2008), using 16S rRNA gene clone libraries, reported small intestinal microbial communities of dogs with spontaneous IBD containing Firmicutes (60.6% of clones), Proteobacteria (30.3% of clones), Bacteroidetes (2.7% of clones), Fusobacteria (3.2% of clones), and Actinobacteria (1.2% of clones). Compared with healthy dogs, those with spontaneous IBD had fewer sequences belonging to the phylum Bacteroidetes (11.2 vs. 2.7% of clones) and more sequences belonging to the Clostridiaceae family (13.9 vs. 34.2% of clones) and Enterobacteriaceae family (0.7 vs. 20.9% of clones). Suchodolski et al. (2012a) investigated intestinal mucosal biopsies from 14 dogs with moderate (n = 7) or severe (n = 7) IBD and control dogs (n = 6) using 454 pyrosequencing. The proportion of Fusobacteria were decreased in dogs with IBD (<1 vs. 15% of sequences). Proteobacteria tended to be greater (P = 0.07) in dogs with IBD (73 vs. 32% of sequences) whereas the proportion of Firmicutes and Bacteroidetes tended to be less (4 vs. 15% of sequences and 7 vs. 29% of sequences, respectively; P = 0.06 to P = 0.10). Samples from IBD dogs also had decreased proportions of Clostridia (<1% vs. 7% of sequences). Suchodolski et al. (2010) reported similar data for duodenal content. Suchodolski et al. (2012b) also reported decreased concentrations of Fusobacteria and Fecalibacterium in the feces of dogs with active IBD compared with control dogs (6.4 vs. 7.3 log DNA and 4.2 vs. 5.8 log DNA, respectively) when fecal samples were analyzed using qPCR. Fusobacteria and Fecalibacterium concentrations in control and active IBD groups were not different from dogs with therapeutically controlled IBD (7.1 and 5.5 log DNA, respectively). However, few differences between healthy dogs and dogs with IBD were noted when 454 pyrosequencing was used to compare fecal microbes (Suchodolski et al., 2012b).

Less data are available characterizing the microbial populations of felines in the diseased state. Using FISH, Janeczko et al. (2008) reported increased mucosal Enterobacteriaceae in duodenal biopsies of cats with IBD compared with healthy cats and a positive relationship between total bacterial number and histological inflammation, including density of T lymphocytes and macrophages. Inness et al. (2007), who also used FISH to examine cats with IBD compared with healthy cats, reported that cats with IBD had slightly less total bacteria, Bacteroides spp., and Bifidobacterium spp. but greater Desulfovibrio spp. Abecia et al. (2010) used FISH to examine fecal microbial populations and did not detect differences in bacterial populations when IBD and healthy cats were fed the same diet.

Although IBD cases are similar in the sense that intestinal inflammation is present, the term is used to encompass a group of diseases that are quite variable and difficult to diagnose and treat. Therefore, in addition to the potential variability attributed to animal genetics, environment, diet, sampling and handling procedures, and laboratory assays in healthy animals, microbiome populations measured in intestinal disease studies may also be quite variable due to the lack of disease conformity, stage of disease, and medications administered either at the time of active disease or those administered in past bouts. Also, it is unclear whether the perturbations in microbial populations in the diseased state develop as a result of an abnormal interaction between bacteria and the host or if the differences are secondary to the disease.


Commercial dog and cat foods vary widely in macronutrient distribution, with extruded diets being rich in carbohydrate (30 to 60% of diet on DM basis) whereas carbohydrate concentrations of canned diets are more similar to wild-type and raw meat diets (<10% on DM basis) and much greater in protein and fat content (Hill et al., 2009; Plantinga et al., 2011; Kerr, 2012; Kerr et al., 2012, 2013). Given their carnivorous nature, it has been hypothesized that the protein to carbohydrate ratio of feline diets is important for feline health (i.e., obesity, feline diabetes, and gut microbiota). Therefore, several research projects examining the impact of macronutrient composition and dietary format in cats have been performed in the past few years (Lubbs et al., 2009; Vester et al., 2009; Jia et al., 2011; Hooda et al., 2013; Kerr et al., 2012). Not surprisingly, the variability in dietary nutrient concentration and source may alter fecal microbial populations (Kerr et al., 2012; Bermingham et al., 2013). Bermingham et al. (2013) reported decreased diversity in cats fed a commercial dry diet (CP = 33%, DM basis; crude fat = 11%, DM basis) compared with those maintained on a commercial wet diet (CP = 42%, DM basis; crude fat = 42%, DM basis). Additionally, the proportion of Proteobacteria and Fusobacteria were decreased in cats fed dry (0.04 vs. 1.1% and 0.3 vs. 23.1% of sequences) whereas Bacteroidetes tended to decrease (P = 0.06; 9 vs. 16% of sequences) and Actinobacteria tended to increase (P = 0.08; 17 vs. 0.1% of sequences). These differences were likely driven by differences in macronutrient composition and dietary ingredients.

Hooda et al. (2013) reported large fecal microbial shifts due to dietary treatments, including those at the phylum, family, and genus levels. In that study, fecal samples from kittens (8 to 16 wk old) fed a moderate-protein, moderate-carbohydrate diet (CP = 34%, DM basis; fat = 20%, DM basis) were compared with those fed a high-protein, low-carbohydrate diet (CP = 52%, DM basis; fat = 23%, DM basis) using 454 pyrosequencing. The proportion of Firmicutes (12- and 16-wk-old kittens only), Fusobacteria, and Proteobacteria were greater in kittens fed high-protein, low-carbohydrate compared with those fed moderate-protein, moderate-carbohydrate diet (79% versus 72%; 13 versus <1%; 4% versus 1% of sequences) whereas Actinobacteria were less (4 to 8% versus 18 to 29% of sequences; Hooda et al., 2013). Hang et al. (2012) also reported increased Fusobacteriales in fecal samples of dogs fed a high-protein diet (CP = 61%, DM basis; fat = 15%, DM basis) compared with those fed a high-carbohydrate diet (CP = 19%, DM basis; fat = 13%, DM basis); however, sequences were isolated primarily from the high guanine–cytosine peaks that were separated by density gradient centrifugation, so the data may not have been representative of the complete microbial population.

Another popular area of companion animal research has been the impact of dietary fiber, prebiotics, and probiotics on GI microbiota. A specific source of dietary fiber is often added to today’s commercial pet foods and prebiotic and probiotic dietary supplements are readily available. Recent experiments have incorporated the use of 454 pyrosequencing and metagenomic analyses to provide a more in-depth view of the GI microbiota and how it may be impacted by these fermentable substrates. Middelbos et al. (2010) evaluated the effects of 7.5% beet pulp fiber diet [total dietary fiber (TDF) = 4.5%, DM basis] compared with a 0% supplemental fiber diet (TDF = 1.4%, DM basis) on fecal microbiota. The proportions of Actinobacteria (0.8 vs. 1.4% of sequences) and Fusobacteria (24 vs. 40% of sequences) were less compared with control dogs whereas Firmicutes were greater (28 vs. 15% of sequences). Within Firmicutes, the proportion of Clostridia was greater in dogs fed beet pulp compared with controls (90 vs. 83% of sequences). Similarly in cats, when using metagenomic analysis, the dietary inclusion of 4% fructooligosaccharides increased the fecal proportions of Actinobacteria compared with 4% cellulose or pectin treatments (11 vs. 5 to 7% of sequences, respectively) whereas 4% pectin treatment increased the proportions of Firmicutes (41 vs. 34% of sequences) compared with the other treatments (Barry et al., 2012). Beloshapka et al. (2013) used 454 pyrosequencing to evaluate fecal microbial communities of healthy adult dogs fed raw meat-based diets with or without 1.4% inulin or yeast cell wall extracts (CP = 25 to 30%, DM basis; fat = 45 to 50%, DM basis). No changes were noted at the phylum level, but dogs fed the inulin-supplemented diets tended (P < 0.10) to have decreased fecal Fusobacterium (P < 0.10). These data indicate that the inclusion of some fermentable fibers may, in part, counteract the shifts in microbial proportions reported previously in cats (e.g., increased concentrations of Fusobacteria for high-protein diets; Hooda et al., 2013).

The effects of probiotic and synbiotic supplements on canine and feline GI microbiota are also of interest. Marshall-Jones et al. (2006) used FISH to evaluate the effects of Lactobacillus acidophilus DSM13241 (3.2 to 4.1 × 109 cfu/kg) on the fecal microbiota of healthy adult cats. Compared with baseline, cats fed the probiotic had greater fecal Lactobacillus spp. and L. acidophilus and lower Enterococcus fecalis, demonstrating that probiotics may not only affect the species being fed but other bacterial taxa as well. Garcia-Mazcorro et al. (2011) evaluated treatment with synbiotic (Proviable-DC; Nutramax Laboratories, Edgewood, MD), which included a probiotic mixture (total 5 × 109 cfu; Enterococcus faecium, Streptococcus salivarus thermophiles, Bifidobacterium longum, L. acidophilus, Lactobacillus casei rhamnosus, Lactobacillus plantarum, and Lactobacillus delbrueckii bulgaricus) and prebiotic substances (i.e., fructooligosaccharides and arabinogalactans). Although qPCR measures indicated increased levels of some of the probiotic species, according to 454 pyrosequencing, phyla, class, order, family, or genus level proportions were not altered. These data indicate that targeted sequence analyses may be necessary when evaluating probiotics.


The availability of DNA-based methods that allow for the identification and characterization of GI microbiota has dramatically changed the research landscape. The field is still in its infancy, but these techniques have begun to identify the bacterial composition and functional capacity. Although the primary taxa and metabolic functions of microbiota present in healthy and diseased dog and cat populations have been identified at a general level, more in-depth research is needed to fully characterize these populations. Not only will this research increase our understanding of the role of the GI microbiome in healthy populations, but it may identify specific pathogens or disease signatures that will be useful in disease diagnosis and/or treatment. Because most research has been performed in research colonies, more research is also needed in the pet population. Such studies will be key in identifying how age, genetics, living environment, and diet impact GI microbiota and what relevance that has in regards to disease.

The data reported above demonstrate that diet can alter the composition of the feline and canine GI microbiome, but much more research is needed in this area as well. Because most of the past research was performed at a time when traditional plating techniques or qPCR were used to quantify a limited number of bacteria (e.g., Lactobacillus, Bifidobacterium, Clostridium perfringens, E. coli), the outcomes were of limited scope. A more holistic view is needed in the future to evaluate the effects of diet on gut microbiota, including studies focused on testing the prebiotic potential of novel fibers. Moreover, because most studies testing dietary intervention were performed in healthy adult research populations, more clinical studies are needed to test the efficacy of dietary intervention in ill or susceptible (e.g., weanling, geriatrics) animals.




Be the first to comment.

Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.