Comparison of the intestinal mucosal microbiota in dogs diagnosed with idiopathic inflammatory bowel disease and dogs with food-responsive diarrhea before and after treatment

Comparison of the intestinal mucosal microbiota in dogs diagnosed with idiopathic inflammatory... Abstract We report the first study to evaluate the intestinal mucosal microbiota of dogs with inflammatory bowel disease (IBD) and dogs with food-responsive diarrhea (FRD) before and after treatment. It was hypothesized that differences in the microbial composition exist between both disease groups and within groups pre- vs. post-treatment. Duodenal and colonic biopsies were obtained endoscopically from 24 dogs (15 FRD, 9 IBD) before and after treatment. The intestinal microbiota was evaluated by Illumina sequencing of the bacterial 16S rRNA gene. The global bacterial composition did not differ between IBD and FRD dogs, nor between treatment status. However, several bacterial taxa showed a difference in abundance. Comparing disease groups, an unclassified genus of Neisseriaceae was abundant in the duodenum in the IBD group, whereas Bilophila occurred more frequently in the duodenum and Burkholderia in the colon of FRD dogs. Comparing the microbiota pre- and post-treatment revealed Enterococcus, Corynebacterium and Proteobacteria to be enriched in the duodenum of FRD dogs pre-treatment, while Bacteroides was abundant in the colon post-treatment. In dogs with IBD, Bacteroides also reached significant abundance in the colon post-treatment. In conclusion, some differences in individual bacterial taxa were identified between IBD and FRD dogs and between treatment status. mucosal microbiota, chronic enteropathies, canine, treatment, duodenum, colon INTRODUCTION Chronic enteropathies are a group of common disorders in dogs and are characterized by persistent or recurrent clinical signs of gastrointestinal disease, including diarrhea, vomiting, weight loss, inappetence or borborygm and flatulence (German, Hall and Day 2003; Hall and German 2010; Dandrieux 2016). Based on the response to treatment, chronic enteropathies are classified as food-responsive diarrhea (FRD), antibiotic-responsive diarrhea (ARD) or idiopathic inflammatory bowel disease (IBD) (Hall and German 2010; Dandrieux 2016). In dogs with FRD, clinical signs resolve after dietary modification to a novel source of protein and carbohydrate or to a hydrolyzed protein diet (Hall and German 2010; Mandigers et al.2010). Dogs with ARD respond to dietary management and antibiotic treatment, for example, with tylosin (German, Hall and Day 2003; Westermarck et al.2005; Kilpinen et al.2011). Idiopathic IBD is defined by the aforementioned chronic gastrointestinal signs and confirmation of intestinal inflammation by histology (German, Hall and Day 2003; Hall and German 2010; Dandrieux 2016). Until today, the pathogenesis of chronic enteropathies, and in particular idiopathic IBD, is not fully understood. The upregulation of Toll-like receptors in dogs with idiopathic IBD and the lack of significant changes after treatment in the face of obvious clinical improvement (Burgener et al.2008) suggest a genetic susceptibility as a contributing factor. This concept is further supported by the differential expression of Toll-like receptors 4 and 5 in German shepherd dogs with chronic enteropathies (Allenspach et al.2010). Additionally, the identification of non-synonymous single nucleotide polymorphisms in exon 3 of the NOD2 gene (Kathrani et al.2014) is in line with the findings of studies on the pathogenesis of IBD in human patients where several susceptibility genes could be identified (e.g. NOD2 gene) in patients with Crohn's disease (Xavier and Podolsky 2007). Furthermore, dietary and environmental factors are suspected to be main contributors in the development of idiopathic IBD (Hall and German 2010; Dandrieux 2016). While the exact mechanisms of host–microbe interactions remain elusive, evidence has grown to support that the intestinal microbiota plays a major role in the pathogenesis of idiopathic IBD (Xavier and Podolsky 2007; Suchodolski et al.2010, 2012a, 2012b; Minamoto et al.2015; Cassmann et al.2016; Vázquez-Baeza et al.2016). Moreover, advanced scientific techniques, such as next-generation sequencing, metagenomics and metabolomics, can facilitate research on the clinical relevance of the intestinal microbiota and their metabolites. Several studies have assessed the gastrointestinal microbiome in healthy dogs, dogs with idiopathic IBD and dogs with acute diarrhea. An interindividual diversity in the abundance of bacterial classes has been shown to exist even in healthy dogs (Handl et al.2011; Garcia-Mazcorro et al.2012; Guard and Suchodolski 2016), making inferences on the significance of changes in the intestinal microbiota somewhat difficult. Regardless of this overall variability in the microbial abundances, several studies have revealed an intestinal dysbiosis in dogs with idiopathic IBD. The intestinal dysbiosis was reflected mainly by an increase in Proteobacteria and a decrease in Faecalibacterium when compared to healthy dogs (Suchodolski et al.2010, 2012a, 2012b, Minamoto et al.2015; Vázquez-Baeza et al.2016). However, these recent studies have evaluated the fecal or duodenal mucosal microbiota of dogs with idiopathic IBD at a single timepoint only. To date, there is no study reported evaluating the mucosal microbiota in dogs with chronic enteropathies both before and after treatment. Furthermore, only little information is available on the differences in the intestinal microbiota of dogs diagnosed with IBD or FRD. Thus, the aims of this study were (i) to compare the duodenal and colonic mucosal microbiota between dogs with IBD and dogs with FRD, and (ii) to evaluate the effect of successful treatment on the microbial composition by comparing the mucosal microbiota of each dog before and after treatment. Our hypotheses were that (i) the mucosal microbial composition differs between the two disease classifications, and (ii) the mucosal microbiome also differs within each disease group depending upon the treatment status. MATERIALS AND METHODS Animals and study protocol Duodenal and colonic mucosal biopsies were retrieved from a former study on canine chronic enteropathies. The exact study protocol has been published elsewhere (Burgener et al.2008; Dumusc et al.2014) and is briefly summarized here. Dogs with chronic gastrointestinal signs were prospectively enrolled between 2006 and 2008. All of the dogs had diarrhea with or without vomiting for at least 6 weeks. Further inclusion criteria were the absence of an identifiable underlying disorder; histopathological evidence of intestinal inflammation; and no treatment with antibiotics, corticosteroids and/or antacids 2 weeks prior to enrollment into the study. Most dogs had already received dietary modifications prior to referral. Potential underlying disorders were ruled out by a CBC, biochemistry profile, measurement of serum trypsin-like-immunoreactivity (TLI), cobalamin and folate, ACTH stimulation test, urinalysis, parasitic and bacterial fecal examination, abdominal ultrasound and endoscopy of the gastrointestinal tract. Since the specific canine pancreatic lipase was not easily available between 2006 and 2008, the diagnosis of pancreatitis was ruled out considering amylase, lipase, TLI and sonographic findings. Also, all dogs received treatment with fenbendazole (50 mg/kg daily for 5 days) regardless of the fecal examination. All owners signed a letter of consent, and the study was reviewed and approved by the Cantonal Committee of Animal Experimentation, Bern, Switzerland. A clinical disease severity score (canine IBD activity index [CIBDAI]) (Jergens et al.2003) was assigned to each dog before and after treatment. The more detailed canine chronic enteropathy clinical activity index (CCECAI) (Allenspach et al.2007) was published during the course of this study and was assigned in some dogs. However, to obtain consistent results for all dogs, the CCECAI score was not further evaluated in the current study. In addition, every dog was categorized to have either mainly upper or lower gastrointestinal signs, or a combination of both. Both a gastroduodenoscopy and colonoscopy were performed in each dog enrolled in the study. Following the complete diagnostic evaluation, including gastrointestinal endoscopy, all dogs received a standardized elimination diet for 14 days. The elimination diet was a selected protein diet based on codfish and rice only, with codfish being a novel source of protein for all dogs enrolled in the study. That dry diet was specially produced for the current study (Biomill SA, Granges-Marnand, Switzerland). The diet was tested for contamination, and the adequacy of the nutritional composition was calculated by a veterinary nutritionist. Owners were thoroughly instructed on the principle of an elimination diet, including the importance of strictly feeding the prescribed diet. If clinical signs improved significantly or resolved within the first 14 days of feeding the diet, dogs were assigned to the FRD group. Although it is possible that few dogs with FRD had not yet responded, this length of the elimination trial was chosen according to previous publications that showed that most dogs with FRD usually respond within the first 2 weeks of a dietary trial (Marks, Laflamme and McAloose 2002; Allenspach et al.2007; Gaschen and Merchant 2011; Allenspach, Culverwell and Chan 2016). Dogs that did not respond to the elimination diet alone were assigned to the idiopathic IBD/steroid-responsive group and received additional prednisolone (1 mg/kg BID) for 14 days followed by a slow tapering of the dose. Cyclosporine (5 mg/kg SID) or other immunosuppressants (e.g. budesonide) were given to dogs that did not improve on prednisolone. Post-treatment assessment included the re-evaluation of the CIBDAI score and a repeat gastrointestinal endoscopy in all dogs. The FRD group of dogs was reassessed 4 weeks after starting the elimination diet, whereas the IBD group of dogs was re-evaluated at 10 weeks after starting treatment with prednisolone. Gastrointestinal endoscopy and histopathological evaluation Details on the endoscopic and the histopathological evaluation have been published elsewhere (Burgener et al.2008). Briefly, mucosal biopsy specimens were retrieved from the duodenum (∼10 cm below the caudal duodenal flexure) and the colon (the middle portion of the descending colon) or from areas with visible lesions. Samples were placed in 4% neutral-buffered formalin for 48 h before being embedded in paraffin and subsequently prepared for histopathological evaluation. In addition, three endoscopic biopsy samples were obtained from each intestinal section and were placed in RNA-later solution followed by storage at –70°C until DNA extraction. The endoscopic biopsies were examined histologically by a board-certified pathologist blinded to the number of endoscopy, diagnosis and treatment. The pathologist assigned a histologic lesion score reflecting the degree of inflammation and cellular infiltration (Jergens et al.1992). Updated histopathological guidelines were published by the World Small Animal Veterinary Association Gastrointestinal Standardization Group in 2008 (Day et al.2008). Similar to the CCECAI, these guidelines were not used in this study in order to apply the same histopathological standards to all dogs. Bacterial 16S rRNA gene quantitation and sequencing Genomic DNA was extracted from duodenal and colonic biopsies using a commercially available DNA extraction kit (PowerSoil®, Mo Bio, Carlsbad, CA, USA) according to the manufacturer's instructions. Amplification and sequencing of the V4 variable region (primers 515F/806R) of the 16S rRNA gene was performed on a MiSeq (Illumina) at the Molecular Research MR DNA laboratory (www.mrdnalab.com, Shallowater, TX, USA) as described previously (Bell et al.2014). The software Quantitative Insights Into Microbial Ecology (QIIME) v.1.8 (http://www.qiime.org) was used for processing and analysis of sequences (Caporaso et al.2010). The raw sequence data were de-multiplexed, and low-quality reads were filtered using default parameters. Chimeric sequences were detected using USEARCH (Edgar 2010) and were removed prior to further analysis. The remaining sequences were then assigned to operational taxonomic units (OTUs) using an open-reference OTU picking protocol in QIIME against the Greengenes (DeSantis et al.2006) database (v.13.8). Prior to the downstream steps, sequences that were assigned as chloroplast, mitochondria and low abundance OTUs were removed. The rarefaction depth was set at 15 170 sequences per sample for colon samples and 2530 sequences per sample for duodenal samples. The sequences were deposited in the Sequence Read Archive under the following accession number: SRP103535. Within-sample diversity was estimated with the alpha diversity indices Chao1, Shannon and Observed OTUs. Beta diversity, which refers to the similarity between samples and potential clustering patterns between sample groups, was visualized using principal coordinate analysis plot based on weighted and unweighted UniFrac distances. Statistical analysis Statistical analyses were performed using JMP® Pro v.12. A Shapiro–Wilk test was used to assess the data distribution for normality. Because the majority of the datasets did not meet the assumption of a normal distribution, comparisons of the alpha diversity and the bacterial taxa between FRD dogs and IBD dogs were performed using a Mann–Whitney U test. A Wilcoxon signed-rank test was used for comparison of paired samples (pre and post-treatment) within each disease group. A Benjamini–Hochberg false discovery rate was used to control for multiple testing. P- and q-values < 0.05 were considered statistically significant. The analysis of similarities (ANOSIM) function in the statistical software package PRIMER 6 (PRIMER-E Ltd, Luton, UK) was used on the weighted and unweighted UniFrac distance matrix to determine if any groups of samples contained significantly different bacterial communities. Linear discriminant analysis (LDA) effect size (LEfSe) (Segata et al.2011) was performed to identify bacterial groups that were significantly associated with disease classification and/or treatment status. LEfSe was used in the Galaxy workflow framework with the parameters set at α = 0.01, LDA score = 2.0. RESULTS Dogs Twenty-four dogs were included in the study: fifteen of these dogs responded to the dietary modification only (FRD group) and nine dogs needed additional immunosuppressant treatment (IBD group). Basic characteristics of all study dogs included in the study are summarized in Table 1. Two IBD dogs were panhypoproteinemic and were classified as having protein-losing enteropathy (PLE) as a result of severe lymphoplasmacytic inflammation due to idiopathic IBD. Both dogs responded to immunosuppressant therapy. One of these dogs, however, developed severe side effects while treated with prednisolone and was switched to budesonide (dosage 3 mg/m² SID), which was well tolerated. Table 1. Characteristics of dogs (n = 24) enrolled in the study. Disease  Breed  Age  Sex  Weight (kg)  BCS  CIBDAI  IBD  Shar Pei  4 y  f  12.4  2/9  9  IBD  Golden Retriever  6 y 10 mo  mn  36.5  6/9  7  IBD, PLE  Beauceron  4 y  fs  28.9  na  14  IBD, PLE  Bernese Mountain Dog  5 y  fs  35.5  na  12  IBD  Am. Cocker Spaniel  3 y 7 mo  mn  10.8  5/9  6  IBD  Mixed breed medium size  12 y 10 mo  mn  27.4  5/9  3  IBD  Cavalier King Charles Spaniel  4 y 6 mo  m  8.6  5/9  4  IBD  Malinois  2 y 8 mo  mn  32.6  4/9  15  IBD  Mixed breed medium size  2 y 11 mo  fs  20.3  5/9  4  FRD  Mixed breed medium size  3 y  mn  30.0  7/9  9  FRD  Yorkshire Terrier  8 y 6 mo  fs  2.9  6/9  6  FRD  French Bulldog  1 y 4 mo  m  14.6  5/9  8  FRD  Weimaraner  2 y  fs  23.0  4/9  4  FRD  Tervuren/Irish Wolfshound  9 mo  f  25.5  4/9  5  FRD  Samoyed/Border Collie/Swiss Mountain Dog  5 y 10 mo  mn  23.0  4/9  7  FRD  Cairn Terrier  3 y  m  9.4  na  4  FRD  Golden Retriever  1 y 2 mo  f  20.1  3/9  11  FRD  West Highland White Terrier  1 y  f  6.4  6/9  4  FRD  Labrador  2 y  m  46.0  6/9  9  FRD  Berger Blanc Suisse  2 y  fs  32.0  na  8  FRD  Pomeranian  10 mo  f  1.8  5/9  2  FRD  Labrador  11 y 2 mo  mn  32.5  6/9  4  FRD  Mixed breed large size  6 y 2 mo  fs  49.0  7/9  1  FRD  Newfoundland  6 y 9 mo  m  44.2  4/9  4  Disease  Breed  Age  Sex  Weight (kg)  BCS  CIBDAI  IBD  Shar Pei  4 y  f  12.4  2/9  9  IBD  Golden Retriever  6 y 10 mo  mn  36.5  6/9  7  IBD, PLE  Beauceron  4 y  fs  28.9  na  14  IBD, PLE  Bernese Mountain Dog  5 y  fs  35.5  na  12  IBD  Am. Cocker Spaniel  3 y 7 mo  mn  10.8  5/9  6  IBD  Mixed breed medium size  12 y 10 mo  mn  27.4  5/9  3  IBD  Cavalier King Charles Spaniel  4 y 6 mo  m  8.6  5/9  4  IBD  Malinois  2 y 8 mo  mn  32.6  4/9  15  IBD  Mixed breed medium size  2 y 11 mo  fs  20.3  5/9  4  FRD  Mixed breed medium size  3 y  mn  30.0  7/9  9  FRD  Yorkshire Terrier  8 y 6 mo  fs  2.9  6/9  6  FRD  French Bulldog  1 y 4 mo  m  14.6  5/9  8  FRD  Weimaraner  2 y  fs  23.0  4/9  4  FRD  Tervuren/Irish Wolfshound  9 mo  f  25.5  4/9  5  FRD  Samoyed/Border Collie/Swiss Mountain Dog  5 y 10 mo  mn  23.0  4/9  7  FRD  Cairn Terrier  3 y  m  9.4  na  4  FRD  Golden Retriever  1 y 2 mo  f  20.1  3/9  11  FRD  West Highland White Terrier  1 y  f  6.4  6/9  4  FRD  Labrador  2 y  m  46.0  6/9  9  FRD  Berger Blanc Suisse  2 y  fs  32.0  na  8  FRD  Pomeranian  10 mo  f  1.8  5/9  2  FRD  Labrador  11 y 2 mo  mn  32.5  6/9  4  FRD  Mixed breed large size  6 y 2 mo  fs  49.0  7/9  1  FRD  Newfoundland  6 y 9 mo  m  44.2  4/9  4  The canine inflammatory bowel disease activity index (CIBDAI) refers to the clinical activity score at the first visit. The body condition score (BCS) refers to the body condition score of the first visit. y, year; mo, months; f, female; m, male; n, neutered; s, spayed; na, not available. View Large In one dog diagnosed with FRD, duodenal biopsy samples were not sufficient for Illumina sequencing. Because duodenum and colon were analyzed separately, the colonic biopsies of this dog were still included in the analysis. In another dog with a diagnosis of FRD, the post-treatment sample did not meet the rarefaction depth that had been set for the analysis. Hence, this dog was excluded from the within-group evaluation of the effect of treatment on the mucosal microbial composition. Sequence analysis The sequence analysis yielded a total of 5436 076 quality sequences for all analyzed samples (n = 96, mean ± SD = 55 877 ± 39 144). The average of Good's coverage of all samples was 97.3 ± 0.4% (mean ± SD., ranging from 96.3% to 98.2%). Microbial communities in dogs with IBD or FRD Diversity analysis Alpha diversity, as described by species richness, Chao 1 and Shannon diversity index, was not significantly different between dogs with IBD and dogs with FRD in neither the duodenum nor colon (Fig. 1, Table 2). Also, within each disease group, significant differences were not seen before and after treatment (Fig. 2, Table 2). Figure 1. View largeDownload slide Rarefaction analysis of 16S rRNA gene sequences obtained from canine (A) duodenal and (B) colonic mucosa samples. Rarefaction depth was set at 15,170 (colon) and 2530 (duodenum) sequences per sample. The lines (red = dogs with FRD; blue = dogs with IBD) represent the average of each group. The error bars represent the standard deviation. Figure 1. View largeDownload slide Rarefaction analysis of 16S rRNA gene sequences obtained from canine (A) duodenal and (B) colonic mucosa samples. Rarefaction depth was set at 15,170 (colon) and 2530 (duodenum) sequences per sample. The lines (red = dogs with FRD; blue = dogs with IBD) represent the average of each group. The error bars represent the standard deviation. Figure 2. View largeDownload slide Rarefaction analysis of 16S rRNA gene sequences obtained from canine (A) duodenal and (B) colonic mucosa samples. The analysis was performed on a randomly selected subset of Table 2. Rarefaction depth was set at 15,170 (colon) and 2530 (duodenum) sequences per sample. The lines represent the average of each group. The error bars represent the standard deviation. Figure 2. View largeDownload slide Rarefaction analysis of 16S rRNA gene sequences obtained from canine (A) duodenal and (B) colonic mucosa samples. The analysis was performed on a randomly selected subset of Table 2. Rarefaction depth was set at 15,170 (colon) and 2530 (duodenum) sequences per sample. The lines represent the average of each group. The error bars represent the standard deviation. Table 2. Summary of alpha diversity indices comparing dogs with IBD and FRD in colonic and duodenal samples.   Median (min – max)    Duodenum  FRD pre  FRD post  IBD pre  IBD post  FRD vs IBD P-value*  FRD pre vs post P-value**  IBD pre vs post P-value**  Chao1  347(220–542)  352(170–773)  305(187–792)  273(248–434)  0.640  0.946  1  Observed OTU  153(103–250)  156(98–211)  168(127–230)  139(112–174)  0.480  0.893  0.129  Shannon  4.8(2.5–6.1)  5.1(2.6–6.1)  5.06(3.7–5.9)  4.62(4.03–5.43)  0.689  0.541  0.359  Colon  FRD pre  FRD post  IBD pre  IBD post  FRD vs IBD P-value  FRD pre vs post P-value  IBD pre vs post P-value  Chao1  1275(714–1747)  1397(911–1892)  1193(554–2118)  1409(898–1730)  0.975  0.833  0.468  Observed OTU  511(340–741)  573(388–761)  531(316–651)  539(380–728)  0.850  0.570  0.2969  Shannon  4.9(2.4–6.3)  5.4(4.4–6.3)  5.1(2.5–5.8)  5.5(3.5–6)  0.874  0.052  0.9375    Median (min – max)    Duodenum  FRD pre  FRD post  IBD pre  IBD post  FRD vs IBD P-value*  FRD pre vs post P-value**  IBD pre vs post P-value**  Chao1  347(220–542)  352(170–773)  305(187–792)  273(248–434)  0.640  0.946  1  Observed OTU  153(103–250)  156(98–211)  168(127–230)  139(112–174)  0.480  0.893  0.129  Shannon  4.8(2.5–6.1)  5.1(2.6–6.1)  5.06(3.7–5.9)  4.62(4.03–5.43)  0.689  0.541  0.359  Colon  FRD pre  FRD post  IBD pre  IBD post  FRD vs IBD P-value  FRD pre vs post P-value  IBD pre vs post P-value  Chao1  1275(714–1747)  1397(911–1892)  1193(554–2118)  1409(898–1730)  0.975  0.833  0.468  Observed OTU  511(340–741)  573(388–761)  531(316–651)  539(380–728)  0.850  0.570  0.2969  Shannon  4.9(2.4–6.3)  5.4(4.4–6.3)  5.1(2.5–5.8)  5.5(3.5–6)  0.874  0.052  0.9375  Rarefaction depth was set at 15,170 (colon) and 2530 (duodenum) sequences/sample. *P values obtained by Mann–Whitney U test. P values < 0.05 considered as significant. **P values obtained by Wilcoxon signed-rank test. P values < 0.05 considered as significant. View Large Additionally, principal coordinate analysis on unweighted (considering presence/absence OTU) and weighted (community membership and abundance of OTUs) UniFrac distance matrices did not reveal any significant difference in the microbial communities between dogs with a diagnosis of IBD and those dogs with FRD neither in the duodenum nor colon (Fig. 3). This was further confirmed with ANOSIM test (P > 0.05) as shown in Table 3. Similarly, ANOSIM did not reveal any significant differences pre- and post-treatment within each disease group (Table 3). No significant association was identified between microbial communities and the clinical disease severity (i.e. CIBDAI score) before treatment (ANOSIM P > 0.05). Figure 3. View largeDownload slide Three-dimensional principal coordinate analyses of unweighted uniFrac distances colored by disease in (A) duodenal and (B) colonic mucosal samples. Each dot represents the microbial composition of one dog. ANOSIM based on unweighted and weighted UniFrac distances did not show a significant difference between dogs diagnosed with FRD and dogs diagnosed with IBD in the duodenum and colon respectively (duodenum ANOSIM Punweighted = 0.694; colon ANOSIM Punweighted = 0.969; P values < 0.05 considered as significant). Figure 3. View largeDownload slide Three-dimensional principal coordinate analyses of unweighted uniFrac distances colored by disease in (A) duodenal and (B) colonic mucosal samples. Each dot represents the microbial composition of one dog. ANOSIM based on unweighted and weighted UniFrac distances did not show a significant difference between dogs diagnosed with FRD and dogs diagnosed with IBD in the duodenum and colon respectively (duodenum ANOSIM Punweighted = 0.694; colon ANOSIM Punweighted = 0.969; P values < 0.05 considered as significant). Table 3. ANOSIM test based on weighted and unweighted UniFrac distances.   Weighted  Unweighted    R value  P-value  R value  P-value  Duodenum          FRD vs IBD  –0.0090  0.474  –0.0365  0.649  FRD pre vs FRD post  –0.0103  0.554  0.0318  0.233  IBD pre vs IBD post  –0.1020  0.947  0.0082  0.414  Colon          FRD vs IBD  0.0399  0.264  –0.1192  0.969  FRD pre vs FRD post  0.0206  0.231  0.0338  0.181  IBD pre vs IBD post  0.0323  0.207  –0.0250  0.741    Weighted  Unweighted    R value  P-value  R value  P-value  Duodenum          FRD vs IBD  –0.0090  0.474  –0.0365  0.649  FRD pre vs FRD post  –0.0103  0.554  0.0318  0.233  IBD pre vs IBD post  –0.1020  0.947  0.0082  0.414  Colon          FRD vs IBD  0.0399  0.264  –0.1192  0.969  FRD pre vs FRD post  0.0206  0.231  0.0338  0.181  IBD pre vs IBD post  0.0323  0.207  –0.0250  0.741  The ANOSIM test was used to determine if any groups of samples contained significantly different bacterial communities. R values close to 1 indicate high separation between groups. R values close to 0 indicate similarity between groups. P values < 0.05 considered as significant. View Large To determine the differences in bacterial composition between the dogs with IBD and those dogs with FRD, an LEfSe was utilized. Several bacterial taxa were found to be enriched in the two disease groups. In the duodenum of dogs with IBD, Mycoplasmataceae, Microbacteriaceae and Unclassified_Rhizobiales were abundant at the family level (LDA scores: 3.659, 3.608 and 3.656, respectively), and one unclassified genus each of the families Neisseriaceae (LDA score: 4.156), Microbacteriaceae (LDA score: 3.659) and Rhizobiales (LDA score: 3.666) was abundant at the genus level. In dogs with FRD, the genus Bilophila (LDA score: 3.165) was abundant in the duodenal mucosa. Also, in dogs with FRD, the family Burkholderiaceae and the genera Carnobacterium, Burkholderia, Unclassified_Helicobacteraceae and Unclassified_Coriobacteriaceae (LDA scores: 3.176, 2.869, 2.978, 3.192 and 2.899, respectively) were found to be enriched within the colonic mucosa. An LEfSe was also used to assess the mucosal microbial composition before and after treatment. Several bacterial taxa were identified to differ in their abundance either before or after treatment (Tables 4–7). In dogs with FRD, for example, the genera Enterococcus (LDA score: 3.633), Corynebacterium (LDA score: 3.989) and Delftia (LDA score: 4.010) were abundant in the duodenum before treatment, whereas Comamonas (LDA score: 3.419) was significantly abundant in the duodenum after treatment. In the colon of dogs with FRD, the genera Carnobacterium (LDA score: 3.624) and Burkholderia (LDA score: 3.472) were significantly abundant before treatment, and the genera Bacteroides (LDA score: 4.548), Gemella (LDA score: 3.497) and Peptococcus (LDA score: 3.337) were abundant after treatment. Table 4. LDA scores for association between treatment status and bacterial taxa in duodenal mucosal specimens from dogs diagnosed with IBD (n = 9).   LDA score  Associated   Selected taxa  (log 10)  disease group  Phylum      Tenericutes  3.245  IBD pre  Class      Mollicutes  3.245  IBD pre  Order      No differentially abundant features found      Family      Micrococcaceae  3.407  IBD pre  Genus      Unclassified  4.085  IBD pre  Unclassified_Neisseriaceae  4.254  IBD pre  Unclassified_Bradyrhizobiaceae  4.219  IBD post    LDA score  Associated   Selected taxa  (log 10)  disease group  Phylum      Tenericutes  3.245  IBD pre  Class      Mollicutes  3.245  IBD pre  Order      No differentially abundant features found      Family      Micrococcaceae  3.407  IBD pre  Genus      Unclassified  4.085  IBD pre  Unclassified_Neisseriaceae  4.254  IBD pre  Unclassified_Bradyrhizobiaceae  4.219  IBD post  An LDA score >2.0 is considered significant. pre, status before treatment; post, status after treatment. View Large Table 5. LDA scores for association between treatment status and bacterial taxa in colonic mucosal specimens from dogs diagnosed with IBD (n = 9).   LDA score  Associated  Selected taxa  (log 10)  disease group  Phylum      Bacteroidetes  5.026  IBD post  Class      Bacteroidia  5.024  IBD post  Order      Bacteroidales  5.025  IBD post  Family      Planococcaceae  3.161  IBD pre  Oxalobacteraceae  3.200  IBD pre  Bacteroidaceae  4.542  IBD post  Genus      Unclassified_Oxalobacteraceae  3.868  IBD pre  Citrobacter  4.254  IBD pre  Burkholderia  3.884  IBD pre  Bacteroides  4.549  IBD post    LDA score  Associated  Selected taxa  (log 10)  disease group  Phylum      Bacteroidetes  5.026  IBD post  Class      Bacteroidia  5.024  IBD post  Order      Bacteroidales  5.025  IBD post  Family      Planococcaceae  3.161  IBD pre  Oxalobacteraceae  3.200  IBD pre  Bacteroidaceae  4.542  IBD post  Genus      Unclassified_Oxalobacteraceae  3.868  IBD pre  Citrobacter  4.254  IBD pre  Burkholderia  3.884  IBD pre  Bacteroides  4.549  IBD post  An LDA score >2.0 is considered significant. pre, status before treatment; post, status after treatment. View Large Table 6. LDA scores for association between treatment status and bacterial taxa in duodenal mucosal specimens from dogs diagnosed with FRD (n = 13).   LDA score   Associated  Selected taxa  (log 10)  disease group  Phylum      No differentially abundant features found      Class      No differentially abundant features found      Order      No differentially abundant features found      Family      Peptostreptococcaceae  3.694  FRD pre  Enterococcaceae  3.789  FRD pre  Oxalobacteraceae  3.766  FRD pre  Comamonadaceae  4.280  FRD pre  Bacillaceae  3.848  FRD pre  Corynebacteriaceae  3.952  FRD pre  Genus      Unclassified_Peptostreptococcaceae  3.664  FRD pre  Enterococcus  3.633  FRD pre  Unclassified_Coriobacteriaceae  3.855  FRD pre  Corynebacterium  3.989  FRD pre  Delftia  4.010  FRD pre  Comamonas  3.419  FRD post    LDA score   Associated  Selected taxa  (log 10)  disease group  Phylum      No differentially abundant features found      Class      No differentially abundant features found      Order      No differentially abundant features found      Family      Peptostreptococcaceae  3.694  FRD pre  Enterococcaceae  3.789  FRD pre  Oxalobacteraceae  3.766  FRD pre  Comamonadaceae  4.280  FRD pre  Bacillaceae  3.848  FRD pre  Corynebacteriaceae  3.952  FRD pre  Genus      Unclassified_Peptostreptococcaceae  3.664  FRD pre  Enterococcus  3.633  FRD pre  Unclassified_Coriobacteriaceae  3.855  FRD pre  Corynebacterium  3.989  FRD pre  Delftia  4.010  FRD pre  Comamonas  3.419  FRD post  An LDA score >2.0 is considered significant. pre, status before treatment; post, status after treatment. View Large Table 7. LDA scores for association between treatment status and bacterial taxa in colonic mucosal specimens from dogs diagnosed with FRD (n = 14).   LDA score   Associated  Selected taxa  (log 10)  disease group  Phylum      No differentially abundant features found      Class      Bacteroidia  4.821  FRD post  Order      Gemellales  2.911  FRD post  Bacteroidales  4.491  FRD post  Family      Burkholderiaceae  3.132  FRD pre  Bacteroidaceae  4.512  FRD post  Peptococcaceae  3.342  FRD post  Gemellaceae  3.517  FRD post  Genus      Burkholderia  3.472  FRD pre  Carnobacterium  3.624  FRD pre  Gemella  3.497  FRD post  Peptococcus  3.337  FRD post  Bacteroides  4.548  FRD post    LDA score   Associated  Selected taxa  (log 10)  disease group  Phylum      No differentially abundant features found      Class      Bacteroidia  4.821  FRD post  Order      Gemellales  2.911  FRD post  Bacteroidales  4.491  FRD post  Family      Burkholderiaceae  3.132  FRD pre  Bacteroidaceae  4.512  FRD post  Peptococcaceae  3.342  FRD post  Gemellaceae  3.517  FRD post  Genus      Burkholderia  3.472  FRD pre  Carnobacterium  3.624  FRD pre  Gemella  3.497  FRD post  Peptococcus  3.337  FRD post  Bacteroides  4.548  FRD post  An LDA score >2.0 is considered significant. pre, status before treatment; post, status after treatment. View Large The family Micrococcaceae (LDA score: 3.407) and an unclassified genus of the family Neisseriaceae (LDA score: 4.254) were found to be enriched within the duodenum of dogs with IBD before treatment. Only an unclassified genus of the family Bradyrhizobiaceae (LDA score: 4.219) reached significant abundance in the duodenum after treatment. The families Planococcaceae (LDA score: 3.161) and Oxalobacteraceae (LDA score: 3.200), and the genera Citrobacter (LDA score: 4.254), Burkholderia (LDA score: 3.884) and Unclassified_Oxalobacteraceae (LDA score: 3.868) were significantly abundant in the colonic mucosa of dogs with IBD before treatment, whereas the genus Bacteroides (LDA score: 4.549) was abundant after treatment. DISCUSSION To our knowledge, this is the first study to evaluate the intestinal mucosal microbiota of dogs with IBD or FRD both before and after treatment. This study did not reveal any differences in the overall species richness in dogs diagnosed with IBD and dogs with FRD. This finding could be attributed to the fact that both conditions may represent a different spectrum of the same disease, reflected by similar histologic inflammatory lesions (Day et al.2008), and that can only be differentiated by their response to treatment (German, Hall and Day 2003; Simpson and Jergens 2011; Dandrieux 2016). It appears reasonable to assume that similar inflammatory histologic lesions might be associated with a similar effect on the mucosal microbiome. However, further studies are warranted to prove or disprove this hypothesis. Analysis of the specific bacterial taxa in dogs with FRD and dogs with idiopathic IBD showed a differential abundance of mainly bacteria belonging to the phylum of Proteobacteria (e.g. Bilophila in the duodenum, Burkholderia and Unclassified_Helicobacteraceae in the colon of FRD dogs; Unclassified_Neisseriaceae and Unclassified_Rhizobiales in the duodenum of IBD dogs). This finding agrees with recent studies revealing an increase of Proteobacteria in dogs with IBD (Suchodolski et al.2010, 2012a; Minamoto et al.2015). Proteobacteria have been shown to also belong to the most abundant phyla in the gastrointestinal tract of healthy dogs (Suchodolski, Camacho and Steiner 2008), where they are abundant especially in the small intestine and present at lower abundance in the colon (Suchodolski, Camacho and Steiner 2008; Schmitz and Suchodolski 2016). However, some members of this phylum have also been shown to have pathogenic characteristics. One variant of a novel family of Burkholderiales, for example, was associated with perianal Crohn's disease in humans (Sim et al.2010). However, the association of pathogenic Burkholderia and canine chronic enteropathies has not been investigated. Comparison of differentially abundant bacteria in dogs with FRD considering treatment status revealed the distinctive presence of bacteria in the duodenal mucosa before treatment. These bacteria belonged to the phyla Proteobacteria (Delftia), Actinobacteria (Corynebacterium) and Firmicutes (Enterococcus). The abundance of Proteobacteria before treatment is in line with the increase in Proteobacteria previously documented in dogs with IBD (Suchodolski et al.2012a; Minamoto et al.2015). The pathogenicity of particular strains of Corynebacterium has been known for many years (Dalal and Likhi 2008; Bernard 2012; Wagner et al.2012) and a recent study has also reported Corynebacterium to be associated with primary sclerosing cholangitis with concomitant colonic disease (Bajer et al.2017). Thus, the abundance of Corynebacterium before treatment in the current study might suggest involvement in the pathogenesis of chronic enteropathies. Certain Enterococcus strains are associated with beneficial effects and are used as probiotics (Kilpinen et al.2015; Schmitz and Suchodolski 2016). In our study, Enterococcus was significantly abundant before treatment in the duodenum of dogs diagnosed with FRD. This finding is in contrast to a previous study in dogs with FRD that showed no significant change in the abundance of Enterococcus spp. with treatment and with the addition of a probiotic mixture (Sauter et al.2006). This discrepancy might be explained by the different sampling methods, different methods to evaluate the microbiota or the difference in the study population or diet. One possible explanation for the differential abundance before treatment in the current study might be that Enterococcus with a higher abundance is more likely to modulate the inflammatory response and to keep the intestinal ecosystem in balance. Another consideration is the potential of Enterococcus to express virulence factors as has been shown in children with IBD (Golińska et al.2013). Such virulence factors could potentially contribute to the inflammatory state and could also be associated with the pathogenesis of the disease. However, in our study both Enterococcus and Corynebacterium were only identified at the genus level and the specific strain was not analyzed due to an inadequate amount of residual DNA. Therefore, no conclusions can be drawn from our findings as to the functional relationship between higher abundance of Enterococcus or Corynebacterium and chronic intestinal inflammation. Future studies analyzing the specific strains of significantly abundant bacteria via quantitative real-time PCR are warranted. In the colon of dogs with FRD, mainly bacteria of the phylum Firmicutes were differentially abundant both before and after treatment. However, one representative of the phylum Bacteroidetes—the genus Bacteroides—was also found to be enriched within the colonic mucosa after initiation of treatment. Since Bacteroidaceae have been detected to be more abundant in healthy dogs than in dogs with IBD (Suchodolski et al.2012a, Minamoto et al.2015), this finding might be associated with the effect of treatment and disease amelioration. This conclusion is also supported by the fact that certain Bacteroides strains have been recognized to be very beneficial to their host due to an important role in microbial degradation of carbohydrates (Flint et al.2012) and bile acid deconjugation (Pavlidis et al.2015). However, Bacteroides strains can also express virulence factors and promote abscess formation via their polysaccharide capsule, and could potentially have negative effects on the host (Wexler 2007; Reis et al.2014). Therefore, Bacteroides can be either protective or have virulent features, opening new avenues for further investigation of its role in the pathogenesis of chronic enteropathies in dogs, at best including the identification of specific strains. Evaluation for possible differences in the bacterial taxa in dogs with IBD depending on treatment status revealed an increased abundance of mainly members of the phylum Proteobacteria (unclassified genus of the family Neisseriaceae in the duodenum; unclassified genus of the family Oxalobacteraceae, and the genera Citrobacter and Burkholderia in the colon) before treatment. Further, one representative of the phylum Firmicutes (family Planococcaceae in the colon) was found at a higher abundance before treatment. To the authors’ knowledge, no specific pathogenic characteristics have been reported for this particular family found in our study. Therefore, the relevance of our finding remains unclear. Only an unclassified genus of the family Bradyrhizobiaceae was found to be enriched in the duodenum of dogs with IBD post-treatment. Again, to the authors’ knowledge, no pathogenic or constitutional trait has been reported for this bacterium, leaving the relevance of this finding unclear. An important finding, however, is that just as in dogs with FRD Bacteroides also reached significant abundance in the colon of dogs with IBD after treatment. This increase of Bacteroides in both disease groups after treatment possibly suggests disease amelioration, and therefore this bacterium could potentially be used as a respective marker of treatment response. The absence of a group of healthy control dogs poses a major limitation of this study. However, in order to follow a standardized study protocol, endoscopies involving general anesthesia would have been necessary before treatment and potentially also after a standardized treatment in these dogs. Because of this invasiveness and the fact that several studies evaluating the intestinal microbiota in healthy dogs have already been published (Suchodolski, Camacho and Steiner 2008; Handl et al.2011), performing repeated gastrointestinal endoscopies under general anesthesia in healthy dogs was considered unethical. Examination of the microbiota in fecal samples of healthy controls both before and after treatment with the study's diet could have been a less invasive alternative to endoscopy. However, a study by one of the authors (JS) has revealed bacterial diversity along the different sections of the canine intestine (Suchodolski, Camacho and Steiner 2008). Therefore, a direct comparison of the fecal microbiota of healthy dogs and, for example, the duodenal mucosal microbiota of diseased dogs would not have been meaningful. The mucosal samples of this study were purposefully retrieved as part of a larger study, and the intestinal microbiota was analyzed retrospectively for this specific study. Unfortunately, residual fecal samples of the dogs diagnosed with IBD and FRD were not available for an additional analysis of the fecal microbiota in the current study. Moreover, in the authors’ opinion, mucosal samples are superior to fecal samples in assessing the true intestinal microbiota. Genomic DNA was extracted using the Mo Bio PowerSoil® DNA Isolation Kit, which is used in the Human Microbiome Project and is currently recommended for the extraction of DNA. The use of this method yielded the currently most effective DNA extraction (Wagner Mackenzie, Waite and Taylor 2015) and kept the possible influence of sample preparation on this study's results to a minimum. The different time scales for the follow-up endoscopies in the two disease groups could have possibly influenced the results. However, as glucocorticoids were not instituted until there was no significant clinical improvement on the elimination diet only on day 14, those dogs had to be given more time to respond to treatment. In general, FRD dogs are considered to respond faster to appropriate treatment, whereas dogs requiring immunosuppressants usually take longer to improve clinically. The different time scales were adopted from previous studies (Allenspach et al.2007; Burgener et al.2008; Dumusc et al.2014) and are considered more representative of a similarly controlled disease status. As most dogs diagnosed with FRD show a significant clinical response to a dietary trial within 14 days (Marks, Laflamme and McAloose 2002; Allenspach et al.2007; Gaschen and Merchant 2011; Allenspach, Culverwell and Chan 2016), the elimination trial was limited to that period of time. Furthermore, this short period provided for good owner compliance. Dietary modification is also an inherent part of treatment in dogs with IBD. Thus, it is possible that some dogs with IBD could have initially improved on the diet alone and could have been erroneously grouped in the FRD group. Yet, in the authors’ experience, dogs diagnosed with IBD generally only show minor improvement on elimination diet and would have probably not been assessed as significant improvement, much less as clinical remission. Another limitation of our study is the small number of dogs included and the fact that endoscopic biopsies were only obtained from the duodenum and colon. A larger study cohort of dogs and inclusion of additional biopsy sites (particularly the ileum) might have yielded more obvious differences in the mucosal microbial composition. Another shortcoming of our study is that the possibility of a continuing effect of previous medications (e.g. antibiotic treatment) on the mucosal microbiota cannot be definitively excluded, although all efforts were made to minimize such an effect by including only dogs not given an antibiotic 2 weeks prior to enrollment in the study. According to studies in human medicine, a 4-week wash-out period (Dethlefsen et al.2008; Langdon, Crook and Dantas 2016) may be more appropriate than a 2-week period and may have resulted in finding a more restored and original intestinal microbial composition. A study in veterinary medicine by one of the authors (JS), however, showed that the intestinal microbiota of dogs that received metronidazole returned to its original composition within 14 days after cessation of therapy in the majority of dogs (Olson et al.2015). Until now, there are only limited studies examining the effect of glucocorticoids on the intestinal microbial composition. One study revealed no effect of prednisolone on the fecal microbiota of dogs (Igarashi et al.2014) and to the authors’ knowledge, there are no studies evaluating the influence of budesonide on the intestinal microbiota. The authors did not expect a major impact of budesonide on the microbial composition, and therefore, the one dog receiving budesonide was still included in the analysis. The possibility of a long-term effect of the previous diet and prebiotics or probiotics given prior to the study can also not be entirely ruled out. Studies in both human and veterinary medicine have shown that the diet, in particular macronutrients, influences the intestinal microbial composition (Hang et al.2012; Conlon and Bird 2014; Xu and Knight 2015; Herstad et al.2017; Li et al.2017). However, the diets used in those studies were diets with pronounced alterations in the composition of macronutrients, which do not conform to commonly used diets. Moreover, a recent study showed that alpha diversity did not correlate with fat or protein intake in a population of dogs with IBD and control dogs (Vázquez-Baeza et al.2016), concluding that the disease effect is stronger than the diet's effect on the microbial composition. It has been recognized that obese individuals harbor a different intestinal microbiota with especially a decreased abundance of Bacteroides and an increased abundance of Firmicutes, which leads to an increased capacity to harvest energy (Ley et al.2005; Turnbaugh et al.2006, 2009). Thus, the body condition score (BCS) of dogs could have affected the results of the current study. However, 16 dogs were assigned a BCS between 4 and 6 out of 9 at their initial visit, representing a similar BCS for the majority of dogs. Furthermore, the study by Vázquez-Baeza et al. also revealed no correlation of the BCS and alpha diversity in the enrolled dogs. Therefore, the authors consider the influence of the BCS on this study's results negligible. An influence of lifestyle (e.g. smoking, exercising, stress) on the intestinal microbiota has been shown in humans (Conlon and Bird 2014), with, for example, a higher diversity of the intestinal microbes in extreme athletes (Clarke et al.2014). Although similar effects can be assumed in athletic or working dogs and dogs living in a smoking household, these effects are again only expected with extreme lifestyles, not reflecting the average companion dog. One study reported both age and breed, respectively size, to have an impact on the intestinal microbial composition in dogs (Simpson et al.2002), whereas another study did not find a correlation between age and the alpha diversity in dogs (Vázquez-Baeza et al.2016). More studies are needed to evaluate the true impact of these factors on the intestinal microbiota. All of the aforementioned possible influences could not be fully excluded in this study, as the dogs’ characteristics and environment could not be standardized in the case of a clinical trial. However, in the authors’ opinion, the diversity of dogs enrolled in this study represents a more realistic and therefore meaningful patient population. In conclusion, this study is the first to assess the intestinal mucosal microbiota in dogs with IBD or FRD before and after treatment. Some differences in individual bacterial taxa were identified both between disease groups and in relation to the treatment status. The relevance of these bacterial groups in the pathogenesis of IBD and FRD requires further research. However, our results suggest that Bacteroides might be a possible marker of successful response to treatment. Larger scale studies are warranted to confirm our findings and to shed more light on the functional consequences of the changes in the intestinal microbial composition shown in this study. Evaluation of their role in the pathophysiology of IBD and FRD and of their potential therapeutic benefit warrants further study. Acknowledgements The authors gratefully acknowledge the assistance of Deepti Gupta (Gastrointestinal Laboratory, Department of Small Animal Clinical Science, College of Veterinary Medicine and Biomedical Sciences, Texas A&M University, College Station, TX, USA) with the Illumina sequencing and statistical analysis. Conflict of interest. None declared. REFERENCES Allenspach K, Culverwell C, Chan D. Long-term outcome in dogs with chronic enteropathies: 203 cases. Vet Rec  2016; 178: 368. Google Scholar CrossRef Search ADS PubMed  Allenspach K, House A, Smith K et al.   Evaluation of mucosal bacteria and histopathology, clinical disease activity and expression of Toll-like receptors in German shepherd dogs with chronic enteropathies. Vet Microbiol  2010; 146: 326– 35. Google Scholar CrossRef Search ADS PubMed  Allenspach K, Wieland B, Gröne A et al.   Chronic enteropathies in dogs: evaluation of risk factors for negative outcome. J Vet Intern Med  2007; 21: 700– 8. Google Scholar CrossRef Search ADS PubMed  Bajer L, Kverka M, Kostovcik M et al.   Distinct gut microbiota profiles in patients with primary sclerosing cholangitis and ulcerative colitis. World J Gastroenterol  2017; 23: 4548– 58. Google Scholar CrossRef Search ADS PubMed  Bell ET, Suchodolski JS, Isaiah A et al.   Faecal microbiota of cats with insulin-treated diabetes mellitus. PLoS One  2014; 9: e108729. Google Scholar CrossRef Search ADS PubMed  Bernard K. The genus Corynebacterium and other medically relevant coryneform-like bacteria. J Clin Microbiol  2012; 50: 3152– 8. Google Scholar CrossRef Search ADS PubMed  Burgener IA, König A, Allenspach K et al.   Upregulation of Toll-like receptors in chronic enteropathies in dogs. J Vet Intern Med  2008; 22: 553– 60. Google Scholar CrossRef Search ADS PubMed  Caporaso JG, Kuczynski J, Stombaugh J et al.   QIIME allows analysis of high-throughput community sequencing data. Nat Methods  2010; 7: 335– 6. Google Scholar CrossRef Search ADS PubMed  Cassmann E, White R, Atherly T et al.   Alterations of the ileal and colonic mucosal microbiota in canine chronic enteropathies. PLoS One  2016; 11: e0147321. Google Scholar CrossRef Search ADS PubMed  Clarke SF, Murphy EF, Sullivan O et al.   Exercise and associated dietary extremes impact on gut microbial diversity. Gut  2014; 63: 1913. Google Scholar CrossRef Search ADS PubMed  Conlon MA, Bird AR. The impact of diet and lifestyle on gut microbiota and human health. Nutrients  2014; 7: 17– 44. Google Scholar CrossRef Search ADS PubMed  Dalal A, Likhi R. Corynebacterium minutissimum bacteremia and meningitis: a case report and review of literature. J Infect  2008; 56: 77– 79. Google Scholar CrossRef Search ADS PubMed  Dandrieux JR. Inflammatory bowel disease versus chronic enteropathy in dogs: are they one and the same? J Small Anim Pract  2016; 57: 589– 99. Google Scholar CrossRef Search ADS PubMed  Day MJ, Bilzer T, Mansell J et al.   Histopathological standards for the diagnosis of gastrointestinal inflammation in endoscopic biopsy samples from the dog and cat: a report from the World Small Animal Veterinary Association Gastrointestinal Standardization Group. J Comp Pathol  2008; 138( Suppl 1): S1– 43. Google Scholar CrossRef Search ADS PubMed  DeSantis TZ, Hugenholtz P, Larsen N et al.   Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl Environ Microb  2006; 72: 5069– 72. Google Scholar CrossRef Search ADS   Dethlefsen L, Huse S, Sogin ML et al.   The pervasive effects of an antibiotic on the human gut microbiota, as revealed by deep 16S rRNA sequencing. PLoS Biol  2008; 6: e280. Google Scholar CrossRef Search ADS PubMed  Dumusc SD, Ontsouka EC, Schnyder M et al.   Cyclooxygenase-2 and 5-lipoxygenase in dogs with chronic enteropathies. J Vet Intern Med  2014; 28: 1684– 91. Google Scholar CrossRef Search ADS PubMed  Edgar RC. Search and clustering orders of magnitude faster than BLAST. Bioinformatics  2010; 26: 2460– 1. Google Scholar CrossRef Search ADS PubMed  Flint HJ, Scott KP, Duncan SH et al.   Microbial degradation of complex carbohydrates in the gut. Gut Microbes  2012; 3: 289– 306. Google Scholar CrossRef Search ADS PubMed  Garcia-Mazcorro JF, Dowd SE, Poulsen J et al.   Abundance and short-term temporal variability of fecal microbiota in healthy dogs. MicrobiologyOpen  2012; 1: 340– 7. Google Scholar CrossRef Search ADS PubMed  Gaschen FP, Merchant SR. Adverse food reactions in dogs and cats. Vet Clin North Am-Small  2011; 41: 361– 79. Google Scholar CrossRef Search ADS   German AJ, Hall EJ, Day MJ. Chronic intestinal inflammation and intestinal disease in dogs. J Vet Intern Med  2003; 17: 8– 20. Google Scholar CrossRef Search ADS PubMed  Golińska E, Tomusiak A, Gosiewski T et al.   Virulence factors of Enterococcus strains isolated from patients with inflammatory bowel disease. World J Gastroenterol  2013; 19: 3562– 72. Google Scholar CrossRef Search ADS PubMed  Guard BC, Suchodolski JS. HORSE SPECIES SYMPOSIUM: Canine intestinal microbiology and metagenomics: From phylogeny to function. J Anim Sci  2016; 94: 2247– 61. Google Scholar CrossRef Search ADS PubMed  Hall EJ, German AJ. Diseases of the small intestine. In: Ettinger SJ, Feldman EC (eds). Textbook of Veterinary Internal Medicine , Vol. 2. Missouri, USA: Saunders Elsevier, St. Louis, 2010, 1560– 6. Handl S, Dowd SE, Garcia-Mazcorro JF et al.   Massive parallel 16S rRNA gene pyrosequencing reveals highly diverse fecal bacterial and fungal communities in healthy dogs and cats. FEMS Microbiol Ecol  2011; 76: 301– 10. Google Scholar CrossRef Search ADS PubMed  Hang I, Rinttila T, Zentek J et al.   Effect of high contents of dietary animal-derived protein or carbohydrates on canine faecal microbiota. BMC Vet Res  2012; 8: 90. Google Scholar CrossRef Search ADS PubMed  Herstad KMV, Gajardo K, Bakke AM et al.   A diet change from dry food to beef induces reversible changes on the faecal microbiota in healthy, adult client-owned dogs. BMC Vet Res  2017; 13: 147. Google Scholar CrossRef Search ADS PubMed  Igarashi H, Maeda S, Ohno K et al.   Effect of oral administration of metronidazole or prednisolone on fecal microbiota in dogs. PLoS One  2014; 9: e107909. Google Scholar CrossRef Search ADS PubMed  Jergens A, Moore F, Haynes J et al.   Idiopathic inflammatory bowel disease in dogs and cats: 84 cases (1987-1990). J Am Vet Med Assoc  1992; 201: 1603– 8. Google Scholar PubMed  Jergens AE, Schreiner CA, Frank DE et al.   A scoring index for disease activity in canine inflammatory bowel disease. J Vet Intern Med  2003; 17: 291– 7. Google Scholar CrossRef Search ADS PubMed  Kathrani A, Lee H, White C et al.   Association between nucleotide oligomerisation domain two (Nod2) gene polymorphisms and canine inflammatory bowel disease. Vet Immunol Immunop  2014; 161: 32– 41. Google Scholar CrossRef Search ADS   Kilpinen S, Rantala M, Spillmann T et al.   Oral tylosin administration is associated with an increase of faecal enterococci and lactic acid bacteria in dogs with tylosin-responsive diarrhoea. Vet J  2015; 205: 369– 74. Google Scholar CrossRef Search ADS PubMed  Kilpinen S, Spillmann T, Syrja P et al.   Effect of tylosin on dogs with suspected tylosin-responsive diarrhea: a placebo-controlled, randomized, double-blinded, prospective clinical trial. Acta Vet Scand  2011; 53: 26. Google Scholar CrossRef Search ADS PubMed  Langdon A, Crook N, Dantas G. The effects of antibiotics on the microbiome throughout development and alternative approaches for therapeutic modulation. Genome Med  2016; 8: 39. Google Scholar CrossRef Search ADS PubMed  Ley RE, Backhed F, Turnbaugh P et al.   Obesity alters gut microbial ecology. P Natl Acad Sci USA  2005; 102: 11070– 5. Google Scholar CrossRef Search ADS   Li Q, Lauber CL, Czarnecki-Maulden G et al.   Effects of the dietary protein and carbohydrate ratio on gut microbiomes in dogs of different body conditions. mBio  2017; 8: e01703– 01716. Google Scholar PubMed  Mandigers PJ, Biourge V, van den Ingh TS et al.   A randomized, open-label, positively-controlled field trial of a hydrolyzed protein diet in dogs with chronic small bowel enteropathy. J Vet Intern Med  2010; 24: 1350– 7. Google Scholar CrossRef Search ADS PubMed  Marks S, Laflamme DP, McAloose D. Dietary trial using a commercial hypoallergenic diet containing hydrolyzed protein for dogs with inflammatory bowel disease. Vet Ther  2002; 3: 109– 18. Google Scholar PubMed  Minamoto Y, Otoni CC, Steelman SM et al.   Alteration of the fecal microbiota and serum metabolite profiles in dogs with idiopathic inflammatory bowel disease. Gut Microbes  2015; 6: 33– 47. Google Scholar CrossRef Search ADS PubMed  Olson E, Honneffer JB, Waddle M et al.   Evaluation of the effects of a 2 week treatment with metronidazole on the fecal microbiome of healthy dogs (abstr.). J Vet Intern Med  2015; 29: 1184. Pavlidis P, Powell N, Vincent RP et al.   Systematic review: bile acids and intestinal inflammation-luminal aggressors or regulators of mucosal defence? Aliment Pharm Ther  2015; 42: 802– 17. Google Scholar CrossRef Search ADS   Reis AC, Silva JO, Laranjeira BJ et al.   Virulence factors and biofilm production by isolates of Bacteroides fragilis recovered from dog intestinal tracts. Braz J Microbiol  2014; 45: 647– 50. Google Scholar CrossRef Search ADS PubMed  Sauter SN, Benyacoub J, Allenspach K et al.   Effects of probiotic bacteria in dogs with food responsive diarrhoea treated with an elimination diet. J Anim Physiol An N  2006; 90: 269– 77. Google Scholar CrossRef Search ADS   Schmitz S, Suchodolski J. Understanding the canine intestinal microbiota and its modification by pro-, pre- and synbiotics - what is the evidence? Vet Med Sci  2016; 2: 71– 94. Google Scholar CrossRef Search ADS PubMed  Segata N, Izard J, Waldron L et al.   Metagenomic biomarker discovery and explanation. Genome Biol  2011; 12: R60. Google Scholar CrossRef Search ADS PubMed  Sim WH, Wagner J, Cameron DJ et al.   Novel Burkholderiales 23S rRNA genes identified in ileal biopsy samples from children: preliminary evidence that a subtype is associated with perianal Crohn's disease. J Clin Microbiol  2010; 48: 1939– 42. Google Scholar CrossRef Search ADS PubMed  Simpson JM, Martineau B, Jones WE et al.   Characterization of fecal bacterial populations in canines: effects of age, breed and dietary fiber. Microb Ecol  2002; 44: 186– 97. Google Scholar CrossRef Search ADS PubMed  Simpson KW, Jergens AE. Pitfalls and progress in the diagnosis and management of canine inflammatory bowel disease. Vet Clin N Am Small  2011; 41: 381– 98. Google Scholar CrossRef Search ADS   Suchodolski JS, Camacho J, Steiner JM. Analysis of bacterial diversity in the canine duodenum, jejunum, ileum, and colon by comparative 16S rRNA gene analysis. FEMS Microbiol Ecol  2008; 66: 567– 78. Google Scholar CrossRef Search ADS PubMed  Suchodolski JS, Dowd SE, Wilke V et al.   16S rRNA gene pyrosequencing reveals bacterial dysbiosis in the duodenum of dogs with idiopathic inflammatory bowel disease. PLoS One  2012a; 7: e39333. Google Scholar CrossRef Search ADS   Suchodolski JS, Markel ME, Garcia-Mazcorro JF et al.   The fecal microbiome in dogs with acute diarrhea and idiopathic inflammatory bowel disease. PLoS One  2012b; 7: e51907. Google Scholar CrossRef Search ADS   Suchodolski JS, Xenoulis PG, Paddock CG et al.   Molecular analysis of the bacterial microbiota in duodenal biopsies from dogs with idiopathic inflammatory bowel disease. Vet Microbiol  2010; 142: 394– 400. Google Scholar CrossRef Search ADS PubMed  Turnbaugh PJ, Ley RE, Mahowald MA et al.   An obesity-associated gut microbiome with increased capacity for energy harvest. Nature  2006; 444: 1027– 131. Google Scholar CrossRef Search ADS PubMed  Turnbaugh PJ, Hamady M, Yatsunenko T et al.   A core gut microbiome in obese and lean twins. Nature  2009; 457: 480– 4. Google Scholar CrossRef Search ADS PubMed  Vázquez-Baeza Y, Hyde ER, Suchodolski JS et al.   Dog and human inflammatory bowel disease rely on overlapping yet distinct dysbiosis networks. Nat Microbiol  2016; 1: 16177. Google Scholar CrossRef Search ADS PubMed  Wagner KS, White JM, Lucenko I et al.   Diphtheria in the postepidemic period, Europe, 2000–2009. Emerg Infect Dis  2012; 18: 217– 25. Google Scholar CrossRef Search ADS PubMed  Wagner Mackenzie B, Waite DW, Taylor MW. Evaluating variation in human gut microbiota profiles due to DNA extraction method and inter-subject differences. Front Microbiol  2015; 6: 130. Google Scholar CrossRef Search ADS PubMed  Westermarck E, Skrzypczak T, Harmoinen J et al.   Tylosin-responsive chronic diarrhea in dogs. J Vet Intern Med  2005; 19: 177– 86. Google Scholar CrossRef Search ADS PubMed  Wexler HM. Bacteroides: the good, the bad, and the nitty-gritty. Clin Microbiol Rev  2007; 20: 593– 621. Google Scholar CrossRef Search ADS PubMed  Xavier RJ, Podolsky DK. Unravelling the pathogenesis of inflammatory bowel disease. Nature  2007; 448: 427– 34. Google Scholar CrossRef Search ADS PubMed  Xu Z, Knight R. Dietary effects on human gut microbiome diversity. Br J Nutr  2015; 113( Suppl): S1– 5. Google Scholar CrossRef Search ADS PubMed  © FEMS 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png FEMS Microbiology Ecology Oxford University Press

Comparison of the intestinal mucosal microbiota in dogs diagnosed with idiopathic inflammatory bowel disease and dogs with food-responsive diarrhea before and after treatment

Loading next page...
 
/lp/ou_press/comparison-of-the-intestinal-mucosal-microbiota-in-dogs-diagnosed-with-SbdwYAsmM7
Publisher
Blackwell
Copyright
© FEMS 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0168-6496
eISSN
1574-6941
D.O.I.
10.1093/femsec/fix173
Publisher site
See Article on Publisher Site

Abstract

Abstract We report the first study to evaluate the intestinal mucosal microbiota of dogs with inflammatory bowel disease (IBD) and dogs with food-responsive diarrhea (FRD) before and after treatment. It was hypothesized that differences in the microbial composition exist between both disease groups and within groups pre- vs. post-treatment. Duodenal and colonic biopsies were obtained endoscopically from 24 dogs (15 FRD, 9 IBD) before and after treatment. The intestinal microbiota was evaluated by Illumina sequencing of the bacterial 16S rRNA gene. The global bacterial composition did not differ between IBD and FRD dogs, nor between treatment status. However, several bacterial taxa showed a difference in abundance. Comparing disease groups, an unclassified genus of Neisseriaceae was abundant in the duodenum in the IBD group, whereas Bilophila occurred more frequently in the duodenum and Burkholderia in the colon of FRD dogs. Comparing the microbiota pre- and post-treatment revealed Enterococcus, Corynebacterium and Proteobacteria to be enriched in the duodenum of FRD dogs pre-treatment, while Bacteroides was abundant in the colon post-treatment. In dogs with IBD, Bacteroides also reached significant abundance in the colon post-treatment. In conclusion, some differences in individual bacterial taxa were identified between IBD and FRD dogs and between treatment status. mucosal microbiota, chronic enteropathies, canine, treatment, duodenum, colon INTRODUCTION Chronic enteropathies are a group of common disorders in dogs and are characterized by persistent or recurrent clinical signs of gastrointestinal disease, including diarrhea, vomiting, weight loss, inappetence or borborygm and flatulence (German, Hall and Day 2003; Hall and German 2010; Dandrieux 2016). Based on the response to treatment, chronic enteropathies are classified as food-responsive diarrhea (FRD), antibiotic-responsive diarrhea (ARD) or idiopathic inflammatory bowel disease (IBD) (Hall and German 2010; Dandrieux 2016). In dogs with FRD, clinical signs resolve after dietary modification to a novel source of protein and carbohydrate or to a hydrolyzed protein diet (Hall and German 2010; Mandigers et al.2010). Dogs with ARD respond to dietary management and antibiotic treatment, for example, with tylosin (German, Hall and Day 2003; Westermarck et al.2005; Kilpinen et al.2011). Idiopathic IBD is defined by the aforementioned chronic gastrointestinal signs and confirmation of intestinal inflammation by histology (German, Hall and Day 2003; Hall and German 2010; Dandrieux 2016). Until today, the pathogenesis of chronic enteropathies, and in particular idiopathic IBD, is not fully understood. The upregulation of Toll-like receptors in dogs with idiopathic IBD and the lack of significant changes after treatment in the face of obvious clinical improvement (Burgener et al.2008) suggest a genetic susceptibility as a contributing factor. This concept is further supported by the differential expression of Toll-like receptors 4 and 5 in German shepherd dogs with chronic enteropathies (Allenspach et al.2010). Additionally, the identification of non-synonymous single nucleotide polymorphisms in exon 3 of the NOD2 gene (Kathrani et al.2014) is in line with the findings of studies on the pathogenesis of IBD in human patients where several susceptibility genes could be identified (e.g. NOD2 gene) in patients with Crohn's disease (Xavier and Podolsky 2007). Furthermore, dietary and environmental factors are suspected to be main contributors in the development of idiopathic IBD (Hall and German 2010; Dandrieux 2016). While the exact mechanisms of host–microbe interactions remain elusive, evidence has grown to support that the intestinal microbiota plays a major role in the pathogenesis of idiopathic IBD (Xavier and Podolsky 2007; Suchodolski et al.2010, 2012a, 2012b; Minamoto et al.2015; Cassmann et al.2016; Vázquez-Baeza et al.2016). Moreover, advanced scientific techniques, such as next-generation sequencing, metagenomics and metabolomics, can facilitate research on the clinical relevance of the intestinal microbiota and their metabolites. Several studies have assessed the gastrointestinal microbiome in healthy dogs, dogs with idiopathic IBD and dogs with acute diarrhea. An interindividual diversity in the abundance of bacterial classes has been shown to exist even in healthy dogs (Handl et al.2011; Garcia-Mazcorro et al.2012; Guard and Suchodolski 2016), making inferences on the significance of changes in the intestinal microbiota somewhat difficult. Regardless of this overall variability in the microbial abundances, several studies have revealed an intestinal dysbiosis in dogs with idiopathic IBD. The intestinal dysbiosis was reflected mainly by an increase in Proteobacteria and a decrease in Faecalibacterium when compared to healthy dogs (Suchodolski et al.2010, 2012a, 2012b, Minamoto et al.2015; Vázquez-Baeza et al.2016). However, these recent studies have evaluated the fecal or duodenal mucosal microbiota of dogs with idiopathic IBD at a single timepoint only. To date, there is no study reported evaluating the mucosal microbiota in dogs with chronic enteropathies both before and after treatment. Furthermore, only little information is available on the differences in the intestinal microbiota of dogs diagnosed with IBD or FRD. Thus, the aims of this study were (i) to compare the duodenal and colonic mucosal microbiota between dogs with IBD and dogs with FRD, and (ii) to evaluate the effect of successful treatment on the microbial composition by comparing the mucosal microbiota of each dog before and after treatment. Our hypotheses were that (i) the mucosal microbial composition differs between the two disease classifications, and (ii) the mucosal microbiome also differs within each disease group depending upon the treatment status. MATERIALS AND METHODS Animals and study protocol Duodenal and colonic mucosal biopsies were retrieved from a former study on canine chronic enteropathies. The exact study protocol has been published elsewhere (Burgener et al.2008; Dumusc et al.2014) and is briefly summarized here. Dogs with chronic gastrointestinal signs were prospectively enrolled between 2006 and 2008. All of the dogs had diarrhea with or without vomiting for at least 6 weeks. Further inclusion criteria were the absence of an identifiable underlying disorder; histopathological evidence of intestinal inflammation; and no treatment with antibiotics, corticosteroids and/or antacids 2 weeks prior to enrollment into the study. Most dogs had already received dietary modifications prior to referral. Potential underlying disorders were ruled out by a CBC, biochemistry profile, measurement of serum trypsin-like-immunoreactivity (TLI), cobalamin and folate, ACTH stimulation test, urinalysis, parasitic and bacterial fecal examination, abdominal ultrasound and endoscopy of the gastrointestinal tract. Since the specific canine pancreatic lipase was not easily available between 2006 and 2008, the diagnosis of pancreatitis was ruled out considering amylase, lipase, TLI and sonographic findings. Also, all dogs received treatment with fenbendazole (50 mg/kg daily for 5 days) regardless of the fecal examination. All owners signed a letter of consent, and the study was reviewed and approved by the Cantonal Committee of Animal Experimentation, Bern, Switzerland. A clinical disease severity score (canine IBD activity index [CIBDAI]) (Jergens et al.2003) was assigned to each dog before and after treatment. The more detailed canine chronic enteropathy clinical activity index (CCECAI) (Allenspach et al.2007) was published during the course of this study and was assigned in some dogs. However, to obtain consistent results for all dogs, the CCECAI score was not further evaluated in the current study. In addition, every dog was categorized to have either mainly upper or lower gastrointestinal signs, or a combination of both. Both a gastroduodenoscopy and colonoscopy were performed in each dog enrolled in the study. Following the complete diagnostic evaluation, including gastrointestinal endoscopy, all dogs received a standardized elimination diet for 14 days. The elimination diet was a selected protein diet based on codfish and rice only, with codfish being a novel source of protein for all dogs enrolled in the study. That dry diet was specially produced for the current study (Biomill SA, Granges-Marnand, Switzerland). The diet was tested for contamination, and the adequacy of the nutritional composition was calculated by a veterinary nutritionist. Owners were thoroughly instructed on the principle of an elimination diet, including the importance of strictly feeding the prescribed diet. If clinical signs improved significantly or resolved within the first 14 days of feeding the diet, dogs were assigned to the FRD group. Although it is possible that few dogs with FRD had not yet responded, this length of the elimination trial was chosen according to previous publications that showed that most dogs with FRD usually respond within the first 2 weeks of a dietary trial (Marks, Laflamme and McAloose 2002; Allenspach et al.2007; Gaschen and Merchant 2011; Allenspach, Culverwell and Chan 2016). Dogs that did not respond to the elimination diet alone were assigned to the idiopathic IBD/steroid-responsive group and received additional prednisolone (1 mg/kg BID) for 14 days followed by a slow tapering of the dose. Cyclosporine (5 mg/kg SID) or other immunosuppressants (e.g. budesonide) were given to dogs that did not improve on prednisolone. Post-treatment assessment included the re-evaluation of the CIBDAI score and a repeat gastrointestinal endoscopy in all dogs. The FRD group of dogs was reassessed 4 weeks after starting the elimination diet, whereas the IBD group of dogs was re-evaluated at 10 weeks after starting treatment with prednisolone. Gastrointestinal endoscopy and histopathological evaluation Details on the endoscopic and the histopathological evaluation have been published elsewhere (Burgener et al.2008). Briefly, mucosal biopsy specimens were retrieved from the duodenum (∼10 cm below the caudal duodenal flexure) and the colon (the middle portion of the descending colon) or from areas with visible lesions. Samples were placed in 4% neutral-buffered formalin for 48 h before being embedded in paraffin and subsequently prepared for histopathological evaluation. In addition, three endoscopic biopsy samples were obtained from each intestinal section and were placed in RNA-later solution followed by storage at –70°C until DNA extraction. The endoscopic biopsies were examined histologically by a board-certified pathologist blinded to the number of endoscopy, diagnosis and treatment. The pathologist assigned a histologic lesion score reflecting the degree of inflammation and cellular infiltration (Jergens et al.1992). Updated histopathological guidelines were published by the World Small Animal Veterinary Association Gastrointestinal Standardization Group in 2008 (Day et al.2008). Similar to the CCECAI, these guidelines were not used in this study in order to apply the same histopathological standards to all dogs. Bacterial 16S rRNA gene quantitation and sequencing Genomic DNA was extracted from duodenal and colonic biopsies using a commercially available DNA extraction kit (PowerSoil®, Mo Bio, Carlsbad, CA, USA) according to the manufacturer's instructions. Amplification and sequencing of the V4 variable region (primers 515F/806R) of the 16S rRNA gene was performed on a MiSeq (Illumina) at the Molecular Research MR DNA laboratory (www.mrdnalab.com, Shallowater, TX, USA) as described previously (Bell et al.2014). The software Quantitative Insights Into Microbial Ecology (QIIME) v.1.8 (http://www.qiime.org) was used for processing and analysis of sequences (Caporaso et al.2010). The raw sequence data were de-multiplexed, and low-quality reads were filtered using default parameters. Chimeric sequences were detected using USEARCH (Edgar 2010) and were removed prior to further analysis. The remaining sequences were then assigned to operational taxonomic units (OTUs) using an open-reference OTU picking protocol in QIIME against the Greengenes (DeSantis et al.2006) database (v.13.8). Prior to the downstream steps, sequences that were assigned as chloroplast, mitochondria and low abundance OTUs were removed. The rarefaction depth was set at 15 170 sequences per sample for colon samples and 2530 sequences per sample for duodenal samples. The sequences were deposited in the Sequence Read Archive under the following accession number: SRP103535. Within-sample diversity was estimated with the alpha diversity indices Chao1, Shannon and Observed OTUs. Beta diversity, which refers to the similarity between samples and potential clustering patterns between sample groups, was visualized using principal coordinate analysis plot based on weighted and unweighted UniFrac distances. Statistical analysis Statistical analyses were performed using JMP® Pro v.12. A Shapiro–Wilk test was used to assess the data distribution for normality. Because the majority of the datasets did not meet the assumption of a normal distribution, comparisons of the alpha diversity and the bacterial taxa between FRD dogs and IBD dogs were performed using a Mann–Whitney U test. A Wilcoxon signed-rank test was used for comparison of paired samples (pre and post-treatment) within each disease group. A Benjamini–Hochberg false discovery rate was used to control for multiple testing. P- and q-values < 0.05 were considered statistically significant. The analysis of similarities (ANOSIM) function in the statistical software package PRIMER 6 (PRIMER-E Ltd, Luton, UK) was used on the weighted and unweighted UniFrac distance matrix to determine if any groups of samples contained significantly different bacterial communities. Linear discriminant analysis (LDA) effect size (LEfSe) (Segata et al.2011) was performed to identify bacterial groups that were significantly associated with disease classification and/or treatment status. LEfSe was used in the Galaxy workflow framework with the parameters set at α = 0.01, LDA score = 2.0. RESULTS Dogs Twenty-four dogs were included in the study: fifteen of these dogs responded to the dietary modification only (FRD group) and nine dogs needed additional immunosuppressant treatment (IBD group). Basic characteristics of all study dogs included in the study are summarized in Table 1. Two IBD dogs were panhypoproteinemic and were classified as having protein-losing enteropathy (PLE) as a result of severe lymphoplasmacytic inflammation due to idiopathic IBD. Both dogs responded to immunosuppressant therapy. One of these dogs, however, developed severe side effects while treated with prednisolone and was switched to budesonide (dosage 3 mg/m² SID), which was well tolerated. Table 1. Characteristics of dogs (n = 24) enrolled in the study. Disease  Breed  Age  Sex  Weight (kg)  BCS  CIBDAI  IBD  Shar Pei  4 y  f  12.4  2/9  9  IBD  Golden Retriever  6 y 10 mo  mn  36.5  6/9  7  IBD, PLE  Beauceron  4 y  fs  28.9  na  14  IBD, PLE  Bernese Mountain Dog  5 y  fs  35.5  na  12  IBD  Am. Cocker Spaniel  3 y 7 mo  mn  10.8  5/9  6  IBD  Mixed breed medium size  12 y 10 mo  mn  27.4  5/9  3  IBD  Cavalier King Charles Spaniel  4 y 6 mo  m  8.6  5/9  4  IBD  Malinois  2 y 8 mo  mn  32.6  4/9  15  IBD  Mixed breed medium size  2 y 11 mo  fs  20.3  5/9  4  FRD  Mixed breed medium size  3 y  mn  30.0  7/9  9  FRD  Yorkshire Terrier  8 y 6 mo  fs  2.9  6/9  6  FRD  French Bulldog  1 y 4 mo  m  14.6  5/9  8  FRD  Weimaraner  2 y  fs  23.0  4/9  4  FRD  Tervuren/Irish Wolfshound  9 mo  f  25.5  4/9  5  FRD  Samoyed/Border Collie/Swiss Mountain Dog  5 y 10 mo  mn  23.0  4/9  7  FRD  Cairn Terrier  3 y  m  9.4  na  4  FRD  Golden Retriever  1 y 2 mo  f  20.1  3/9  11  FRD  West Highland White Terrier  1 y  f  6.4  6/9  4  FRD  Labrador  2 y  m  46.0  6/9  9  FRD  Berger Blanc Suisse  2 y  fs  32.0  na  8  FRD  Pomeranian  10 mo  f  1.8  5/9  2  FRD  Labrador  11 y 2 mo  mn  32.5  6/9  4  FRD  Mixed breed large size  6 y 2 mo  fs  49.0  7/9  1  FRD  Newfoundland  6 y 9 mo  m  44.2  4/9  4  Disease  Breed  Age  Sex  Weight (kg)  BCS  CIBDAI  IBD  Shar Pei  4 y  f  12.4  2/9  9  IBD  Golden Retriever  6 y 10 mo  mn  36.5  6/9  7  IBD, PLE  Beauceron  4 y  fs  28.9  na  14  IBD, PLE  Bernese Mountain Dog  5 y  fs  35.5  na  12  IBD  Am. Cocker Spaniel  3 y 7 mo  mn  10.8  5/9  6  IBD  Mixed breed medium size  12 y 10 mo  mn  27.4  5/9  3  IBD  Cavalier King Charles Spaniel  4 y 6 mo  m  8.6  5/9  4  IBD  Malinois  2 y 8 mo  mn  32.6  4/9  15  IBD  Mixed breed medium size  2 y 11 mo  fs  20.3  5/9  4  FRD  Mixed breed medium size  3 y  mn  30.0  7/9  9  FRD  Yorkshire Terrier  8 y 6 mo  fs  2.9  6/9  6  FRD  French Bulldog  1 y 4 mo  m  14.6  5/9  8  FRD  Weimaraner  2 y  fs  23.0  4/9  4  FRD  Tervuren/Irish Wolfshound  9 mo  f  25.5  4/9  5  FRD  Samoyed/Border Collie/Swiss Mountain Dog  5 y 10 mo  mn  23.0  4/9  7  FRD  Cairn Terrier  3 y  m  9.4  na  4  FRD  Golden Retriever  1 y 2 mo  f  20.1  3/9  11  FRD  West Highland White Terrier  1 y  f  6.4  6/9  4  FRD  Labrador  2 y  m  46.0  6/9  9  FRD  Berger Blanc Suisse  2 y  fs  32.0  na  8  FRD  Pomeranian  10 mo  f  1.8  5/9  2  FRD  Labrador  11 y 2 mo  mn  32.5  6/9  4  FRD  Mixed breed large size  6 y 2 mo  fs  49.0  7/9  1  FRD  Newfoundland  6 y 9 mo  m  44.2  4/9  4  The canine inflammatory bowel disease activity index (CIBDAI) refers to the clinical activity score at the first visit. The body condition score (BCS) refers to the body condition score of the first visit. y, year; mo, months; f, female; m, male; n, neutered; s, spayed; na, not available. View Large In one dog diagnosed with FRD, duodenal biopsy samples were not sufficient for Illumina sequencing. Because duodenum and colon were analyzed separately, the colonic biopsies of this dog were still included in the analysis. In another dog with a diagnosis of FRD, the post-treatment sample did not meet the rarefaction depth that had been set for the analysis. Hence, this dog was excluded from the within-group evaluation of the effect of treatment on the mucosal microbial composition. Sequence analysis The sequence analysis yielded a total of 5436 076 quality sequences for all analyzed samples (n = 96, mean ± SD = 55 877 ± 39 144). The average of Good's coverage of all samples was 97.3 ± 0.4% (mean ± SD., ranging from 96.3% to 98.2%). Microbial communities in dogs with IBD or FRD Diversity analysis Alpha diversity, as described by species richness, Chao 1 and Shannon diversity index, was not significantly different between dogs with IBD and dogs with FRD in neither the duodenum nor colon (Fig. 1, Table 2). Also, within each disease group, significant differences were not seen before and after treatment (Fig. 2, Table 2). Figure 1. View largeDownload slide Rarefaction analysis of 16S rRNA gene sequences obtained from canine (A) duodenal and (B) colonic mucosa samples. Rarefaction depth was set at 15,170 (colon) and 2530 (duodenum) sequences per sample. The lines (red = dogs with FRD; blue = dogs with IBD) represent the average of each group. The error bars represent the standard deviation. Figure 1. View largeDownload slide Rarefaction analysis of 16S rRNA gene sequences obtained from canine (A) duodenal and (B) colonic mucosa samples. Rarefaction depth was set at 15,170 (colon) and 2530 (duodenum) sequences per sample. The lines (red = dogs with FRD; blue = dogs with IBD) represent the average of each group. The error bars represent the standard deviation. Figure 2. View largeDownload slide Rarefaction analysis of 16S rRNA gene sequences obtained from canine (A) duodenal and (B) colonic mucosa samples. The analysis was performed on a randomly selected subset of Table 2. Rarefaction depth was set at 15,170 (colon) and 2530 (duodenum) sequences per sample. The lines represent the average of each group. The error bars represent the standard deviation. Figure 2. View largeDownload slide Rarefaction analysis of 16S rRNA gene sequences obtained from canine (A) duodenal and (B) colonic mucosa samples. The analysis was performed on a randomly selected subset of Table 2. Rarefaction depth was set at 15,170 (colon) and 2530 (duodenum) sequences per sample. The lines represent the average of each group. The error bars represent the standard deviation. Table 2. Summary of alpha diversity indices comparing dogs with IBD and FRD in colonic and duodenal samples.   Median (min – max)    Duodenum  FRD pre  FRD post  IBD pre  IBD post  FRD vs IBD P-value*  FRD pre vs post P-value**  IBD pre vs post P-value**  Chao1  347(220–542)  352(170–773)  305(187–792)  273(248–434)  0.640  0.946  1  Observed OTU  153(103–250)  156(98–211)  168(127–230)  139(112–174)  0.480  0.893  0.129  Shannon  4.8(2.5–6.1)  5.1(2.6–6.1)  5.06(3.7–5.9)  4.62(4.03–5.43)  0.689  0.541  0.359  Colon  FRD pre  FRD post  IBD pre  IBD post  FRD vs IBD P-value  FRD pre vs post P-value  IBD pre vs post P-value  Chao1  1275(714–1747)  1397(911–1892)  1193(554–2118)  1409(898–1730)  0.975  0.833  0.468  Observed OTU  511(340–741)  573(388–761)  531(316–651)  539(380–728)  0.850  0.570  0.2969  Shannon  4.9(2.4–6.3)  5.4(4.4–6.3)  5.1(2.5–5.8)  5.5(3.5–6)  0.874  0.052  0.9375    Median (min – max)    Duodenum  FRD pre  FRD post  IBD pre  IBD post  FRD vs IBD P-value*  FRD pre vs post P-value**  IBD pre vs post P-value**  Chao1  347(220–542)  352(170–773)  305(187–792)  273(248–434)  0.640  0.946  1  Observed OTU  153(103–250)  156(98–211)  168(127–230)  139(112–174)  0.480  0.893  0.129  Shannon  4.8(2.5–6.1)  5.1(2.6–6.1)  5.06(3.7–5.9)  4.62(4.03–5.43)  0.689  0.541  0.359  Colon  FRD pre  FRD post  IBD pre  IBD post  FRD vs IBD P-value  FRD pre vs post P-value  IBD pre vs post P-value  Chao1  1275(714–1747)  1397(911–1892)  1193(554–2118)  1409(898–1730)  0.975  0.833  0.468  Observed OTU  511(340–741)  573(388–761)  531(316–651)  539(380–728)  0.850  0.570  0.2969  Shannon  4.9(2.4–6.3)  5.4(4.4–6.3)  5.1(2.5–5.8)  5.5(3.5–6)  0.874  0.052  0.9375  Rarefaction depth was set at 15,170 (colon) and 2530 (duodenum) sequences/sample. *P values obtained by Mann–Whitney U test. P values < 0.05 considered as significant. **P values obtained by Wilcoxon signed-rank test. P values < 0.05 considered as significant. View Large Additionally, principal coordinate analysis on unweighted (considering presence/absence OTU) and weighted (community membership and abundance of OTUs) UniFrac distance matrices did not reveal any significant difference in the microbial communities between dogs with a diagnosis of IBD and those dogs with FRD neither in the duodenum nor colon (Fig. 3). This was further confirmed with ANOSIM test (P > 0.05) as shown in Table 3. Similarly, ANOSIM did not reveal any significant differences pre- and post-treatment within each disease group (Table 3). No significant association was identified between microbial communities and the clinical disease severity (i.e. CIBDAI score) before treatment (ANOSIM P > 0.05). Figure 3. View largeDownload slide Three-dimensional principal coordinate analyses of unweighted uniFrac distances colored by disease in (A) duodenal and (B) colonic mucosal samples. Each dot represents the microbial composition of one dog. ANOSIM based on unweighted and weighted UniFrac distances did not show a significant difference between dogs diagnosed with FRD and dogs diagnosed with IBD in the duodenum and colon respectively (duodenum ANOSIM Punweighted = 0.694; colon ANOSIM Punweighted = 0.969; P values < 0.05 considered as significant). Figure 3. View largeDownload slide Three-dimensional principal coordinate analyses of unweighted uniFrac distances colored by disease in (A) duodenal and (B) colonic mucosal samples. Each dot represents the microbial composition of one dog. ANOSIM based on unweighted and weighted UniFrac distances did not show a significant difference between dogs diagnosed with FRD and dogs diagnosed with IBD in the duodenum and colon respectively (duodenum ANOSIM Punweighted = 0.694; colon ANOSIM Punweighted = 0.969; P values < 0.05 considered as significant). Table 3. ANOSIM test based on weighted and unweighted UniFrac distances.   Weighted  Unweighted    R value  P-value  R value  P-value  Duodenum          FRD vs IBD  –0.0090  0.474  –0.0365  0.649  FRD pre vs FRD post  –0.0103  0.554  0.0318  0.233  IBD pre vs IBD post  –0.1020  0.947  0.0082  0.414  Colon          FRD vs IBD  0.0399  0.264  –0.1192  0.969  FRD pre vs FRD post  0.0206  0.231  0.0338  0.181  IBD pre vs IBD post  0.0323  0.207  –0.0250  0.741    Weighted  Unweighted    R value  P-value  R value  P-value  Duodenum          FRD vs IBD  –0.0090  0.474  –0.0365  0.649  FRD pre vs FRD post  –0.0103  0.554  0.0318  0.233  IBD pre vs IBD post  –0.1020  0.947  0.0082  0.414  Colon          FRD vs IBD  0.0399  0.264  –0.1192  0.969  FRD pre vs FRD post  0.0206  0.231  0.0338  0.181  IBD pre vs IBD post  0.0323  0.207  –0.0250  0.741  The ANOSIM test was used to determine if any groups of samples contained significantly different bacterial communities. R values close to 1 indicate high separation between groups. R values close to 0 indicate similarity between groups. P values < 0.05 considered as significant. View Large To determine the differences in bacterial composition between the dogs with IBD and those dogs with FRD, an LEfSe was utilized. Several bacterial taxa were found to be enriched in the two disease groups. In the duodenum of dogs with IBD, Mycoplasmataceae, Microbacteriaceae and Unclassified_Rhizobiales were abundant at the family level (LDA scores: 3.659, 3.608 and 3.656, respectively), and one unclassified genus each of the families Neisseriaceae (LDA score: 4.156), Microbacteriaceae (LDA score: 3.659) and Rhizobiales (LDA score: 3.666) was abundant at the genus level. In dogs with FRD, the genus Bilophila (LDA score: 3.165) was abundant in the duodenal mucosa. Also, in dogs with FRD, the family Burkholderiaceae and the genera Carnobacterium, Burkholderia, Unclassified_Helicobacteraceae and Unclassified_Coriobacteriaceae (LDA scores: 3.176, 2.869, 2.978, 3.192 and 2.899, respectively) were found to be enriched within the colonic mucosa. An LEfSe was also used to assess the mucosal microbial composition before and after treatment. Several bacterial taxa were identified to differ in their abundance either before or after treatment (Tables 4–7). In dogs with FRD, for example, the genera Enterococcus (LDA score: 3.633), Corynebacterium (LDA score: 3.989) and Delftia (LDA score: 4.010) were abundant in the duodenum before treatment, whereas Comamonas (LDA score: 3.419) was significantly abundant in the duodenum after treatment. In the colon of dogs with FRD, the genera Carnobacterium (LDA score: 3.624) and Burkholderia (LDA score: 3.472) were significantly abundant before treatment, and the genera Bacteroides (LDA score: 4.548), Gemella (LDA score: 3.497) and Peptococcus (LDA score: 3.337) were abundant after treatment. Table 4. LDA scores for association between treatment status and bacterial taxa in duodenal mucosal specimens from dogs diagnosed with IBD (n = 9).   LDA score  Associated   Selected taxa  (log 10)  disease group  Phylum      Tenericutes  3.245  IBD pre  Class      Mollicutes  3.245  IBD pre  Order      No differentially abundant features found      Family      Micrococcaceae  3.407  IBD pre  Genus      Unclassified  4.085  IBD pre  Unclassified_Neisseriaceae  4.254  IBD pre  Unclassified_Bradyrhizobiaceae  4.219  IBD post    LDA score  Associated   Selected taxa  (log 10)  disease group  Phylum      Tenericutes  3.245  IBD pre  Class      Mollicutes  3.245  IBD pre  Order      No differentially abundant features found      Family      Micrococcaceae  3.407  IBD pre  Genus      Unclassified  4.085  IBD pre  Unclassified_Neisseriaceae  4.254  IBD pre  Unclassified_Bradyrhizobiaceae  4.219  IBD post  An LDA score >2.0 is considered significant. pre, status before treatment; post, status after treatment. View Large Table 5. LDA scores for association between treatment status and bacterial taxa in colonic mucosal specimens from dogs diagnosed with IBD (n = 9).   LDA score  Associated  Selected taxa  (log 10)  disease group  Phylum      Bacteroidetes  5.026  IBD post  Class      Bacteroidia  5.024  IBD post  Order      Bacteroidales  5.025  IBD post  Family      Planococcaceae  3.161  IBD pre  Oxalobacteraceae  3.200  IBD pre  Bacteroidaceae  4.542  IBD post  Genus      Unclassified_Oxalobacteraceae  3.868  IBD pre  Citrobacter  4.254  IBD pre  Burkholderia  3.884  IBD pre  Bacteroides  4.549  IBD post    LDA score  Associated  Selected taxa  (log 10)  disease group  Phylum      Bacteroidetes  5.026  IBD post  Class      Bacteroidia  5.024  IBD post  Order      Bacteroidales  5.025  IBD post  Family      Planococcaceae  3.161  IBD pre  Oxalobacteraceae  3.200  IBD pre  Bacteroidaceae  4.542  IBD post  Genus      Unclassified_Oxalobacteraceae  3.868  IBD pre  Citrobacter  4.254  IBD pre  Burkholderia  3.884  IBD pre  Bacteroides  4.549  IBD post  An LDA score >2.0 is considered significant. pre, status before treatment; post, status after treatment. View Large Table 6. LDA scores for association between treatment status and bacterial taxa in duodenal mucosal specimens from dogs diagnosed with FRD (n = 13).   LDA score   Associated  Selected taxa  (log 10)  disease group  Phylum      No differentially abundant features found      Class      No differentially abundant features found      Order      No differentially abundant features found      Family      Peptostreptococcaceae  3.694  FRD pre  Enterococcaceae  3.789  FRD pre  Oxalobacteraceae  3.766  FRD pre  Comamonadaceae  4.280  FRD pre  Bacillaceae  3.848  FRD pre  Corynebacteriaceae  3.952  FRD pre  Genus      Unclassified_Peptostreptococcaceae  3.664  FRD pre  Enterococcus  3.633  FRD pre  Unclassified_Coriobacteriaceae  3.855  FRD pre  Corynebacterium  3.989  FRD pre  Delftia  4.010  FRD pre  Comamonas  3.419  FRD post    LDA score   Associated  Selected taxa  (log 10)  disease group  Phylum      No differentially abundant features found      Class      No differentially abundant features found      Order      No differentially abundant features found      Family      Peptostreptococcaceae  3.694  FRD pre  Enterococcaceae  3.789  FRD pre  Oxalobacteraceae  3.766  FRD pre  Comamonadaceae  4.280  FRD pre  Bacillaceae  3.848  FRD pre  Corynebacteriaceae  3.952  FRD pre  Genus      Unclassified_Peptostreptococcaceae  3.664  FRD pre  Enterococcus  3.633  FRD pre  Unclassified_Coriobacteriaceae  3.855  FRD pre  Corynebacterium  3.989  FRD pre  Delftia  4.010  FRD pre  Comamonas  3.419  FRD post  An LDA score >2.0 is considered significant. pre, status before treatment; post, status after treatment. View Large Table 7. LDA scores for association between treatment status and bacterial taxa in colonic mucosal specimens from dogs diagnosed with FRD (n = 14).   LDA score   Associated  Selected taxa  (log 10)  disease group  Phylum      No differentially abundant features found      Class      Bacteroidia  4.821  FRD post  Order      Gemellales  2.911  FRD post  Bacteroidales  4.491  FRD post  Family      Burkholderiaceae  3.132  FRD pre  Bacteroidaceae  4.512  FRD post  Peptococcaceae  3.342  FRD post  Gemellaceae  3.517  FRD post  Genus      Burkholderia  3.472  FRD pre  Carnobacterium  3.624  FRD pre  Gemella  3.497  FRD post  Peptococcus  3.337  FRD post  Bacteroides  4.548  FRD post    LDA score   Associated  Selected taxa  (log 10)  disease group  Phylum      No differentially abundant features found      Class      Bacteroidia  4.821  FRD post  Order      Gemellales  2.911  FRD post  Bacteroidales  4.491  FRD post  Family      Burkholderiaceae  3.132  FRD pre  Bacteroidaceae  4.512  FRD post  Peptococcaceae  3.342  FRD post  Gemellaceae  3.517  FRD post  Genus      Burkholderia  3.472  FRD pre  Carnobacterium  3.624  FRD pre  Gemella  3.497  FRD post  Peptococcus  3.337  FRD post  Bacteroides  4.548  FRD post  An LDA score >2.0 is considered significant. pre, status before treatment; post, status after treatment. View Large The family Micrococcaceae (LDA score: 3.407) and an unclassified genus of the family Neisseriaceae (LDA score: 4.254) were found to be enriched within the duodenum of dogs with IBD before treatment. Only an unclassified genus of the family Bradyrhizobiaceae (LDA score: 4.219) reached significant abundance in the duodenum after treatment. The families Planococcaceae (LDA score: 3.161) and Oxalobacteraceae (LDA score: 3.200), and the genera Citrobacter (LDA score: 4.254), Burkholderia (LDA score: 3.884) and Unclassified_Oxalobacteraceae (LDA score: 3.868) were significantly abundant in the colonic mucosa of dogs with IBD before treatment, whereas the genus Bacteroides (LDA score: 4.549) was abundant after treatment. DISCUSSION To our knowledge, this is the first study to evaluate the intestinal mucosal microbiota of dogs with IBD or FRD both before and after treatment. This study did not reveal any differences in the overall species richness in dogs diagnosed with IBD and dogs with FRD. This finding could be attributed to the fact that both conditions may represent a different spectrum of the same disease, reflected by similar histologic inflammatory lesions (Day et al.2008), and that can only be differentiated by their response to treatment (German, Hall and Day 2003; Simpson and Jergens 2011; Dandrieux 2016). It appears reasonable to assume that similar inflammatory histologic lesions might be associated with a similar effect on the mucosal microbiome. However, further studies are warranted to prove or disprove this hypothesis. Analysis of the specific bacterial taxa in dogs with FRD and dogs with idiopathic IBD showed a differential abundance of mainly bacteria belonging to the phylum of Proteobacteria (e.g. Bilophila in the duodenum, Burkholderia and Unclassified_Helicobacteraceae in the colon of FRD dogs; Unclassified_Neisseriaceae and Unclassified_Rhizobiales in the duodenum of IBD dogs). This finding agrees with recent studies revealing an increase of Proteobacteria in dogs with IBD (Suchodolski et al.2010, 2012a; Minamoto et al.2015). Proteobacteria have been shown to also belong to the most abundant phyla in the gastrointestinal tract of healthy dogs (Suchodolski, Camacho and Steiner 2008), where they are abundant especially in the small intestine and present at lower abundance in the colon (Suchodolski, Camacho and Steiner 2008; Schmitz and Suchodolski 2016). However, some members of this phylum have also been shown to have pathogenic characteristics. One variant of a novel family of Burkholderiales, for example, was associated with perianal Crohn's disease in humans (Sim et al.2010). However, the association of pathogenic Burkholderia and canine chronic enteropathies has not been investigated. Comparison of differentially abundant bacteria in dogs with FRD considering treatment status revealed the distinctive presence of bacteria in the duodenal mucosa before treatment. These bacteria belonged to the phyla Proteobacteria (Delftia), Actinobacteria (Corynebacterium) and Firmicutes (Enterococcus). The abundance of Proteobacteria before treatment is in line with the increase in Proteobacteria previously documented in dogs with IBD (Suchodolski et al.2012a; Minamoto et al.2015). The pathogenicity of particular strains of Corynebacterium has been known for many years (Dalal and Likhi 2008; Bernard 2012; Wagner et al.2012) and a recent study has also reported Corynebacterium to be associated with primary sclerosing cholangitis with concomitant colonic disease (Bajer et al.2017). Thus, the abundance of Corynebacterium before treatment in the current study might suggest involvement in the pathogenesis of chronic enteropathies. Certain Enterococcus strains are associated with beneficial effects and are used as probiotics (Kilpinen et al.2015; Schmitz and Suchodolski 2016). In our study, Enterococcus was significantly abundant before treatment in the duodenum of dogs diagnosed with FRD. This finding is in contrast to a previous study in dogs with FRD that showed no significant change in the abundance of Enterococcus spp. with treatment and with the addition of a probiotic mixture (Sauter et al.2006). This discrepancy might be explained by the different sampling methods, different methods to evaluate the microbiota or the difference in the study population or diet. One possible explanation for the differential abundance before treatment in the current study might be that Enterococcus with a higher abundance is more likely to modulate the inflammatory response and to keep the intestinal ecosystem in balance. Another consideration is the potential of Enterococcus to express virulence factors as has been shown in children with IBD (Golińska et al.2013). Such virulence factors could potentially contribute to the inflammatory state and could also be associated with the pathogenesis of the disease. However, in our study both Enterococcus and Corynebacterium were only identified at the genus level and the specific strain was not analyzed due to an inadequate amount of residual DNA. Therefore, no conclusions can be drawn from our findings as to the functional relationship between higher abundance of Enterococcus or Corynebacterium and chronic intestinal inflammation. Future studies analyzing the specific strains of significantly abundant bacteria via quantitative real-time PCR are warranted. In the colon of dogs with FRD, mainly bacteria of the phylum Firmicutes were differentially abundant both before and after treatment. However, one representative of the phylum Bacteroidetes—the genus Bacteroides—was also found to be enriched within the colonic mucosa after initiation of treatment. Since Bacteroidaceae have been detected to be more abundant in healthy dogs than in dogs with IBD (Suchodolski et al.2012a, Minamoto et al.2015), this finding might be associated with the effect of treatment and disease amelioration. This conclusion is also supported by the fact that certain Bacteroides strains have been recognized to be very beneficial to their host due to an important role in microbial degradation of carbohydrates (Flint et al.2012) and bile acid deconjugation (Pavlidis et al.2015). However, Bacteroides strains can also express virulence factors and promote abscess formation via their polysaccharide capsule, and could potentially have negative effects on the host (Wexler 2007; Reis et al.2014). Therefore, Bacteroides can be either protective or have virulent features, opening new avenues for further investigation of its role in the pathogenesis of chronic enteropathies in dogs, at best including the identification of specific strains. Evaluation for possible differences in the bacterial taxa in dogs with IBD depending on treatment status revealed an increased abundance of mainly members of the phylum Proteobacteria (unclassified genus of the family Neisseriaceae in the duodenum; unclassified genus of the family Oxalobacteraceae, and the genera Citrobacter and Burkholderia in the colon) before treatment. Further, one representative of the phylum Firmicutes (family Planococcaceae in the colon) was found at a higher abundance before treatment. To the authors’ knowledge, no specific pathogenic characteristics have been reported for this particular family found in our study. Therefore, the relevance of our finding remains unclear. Only an unclassified genus of the family Bradyrhizobiaceae was found to be enriched in the duodenum of dogs with IBD post-treatment. Again, to the authors’ knowledge, no pathogenic or constitutional trait has been reported for this bacterium, leaving the relevance of this finding unclear. An important finding, however, is that just as in dogs with FRD Bacteroides also reached significant abundance in the colon of dogs with IBD after treatment. This increase of Bacteroides in both disease groups after treatment possibly suggests disease amelioration, and therefore this bacterium could potentially be used as a respective marker of treatment response. The absence of a group of healthy control dogs poses a major limitation of this study. However, in order to follow a standardized study protocol, endoscopies involving general anesthesia would have been necessary before treatment and potentially also after a standardized treatment in these dogs. Because of this invasiveness and the fact that several studies evaluating the intestinal microbiota in healthy dogs have already been published (Suchodolski, Camacho and Steiner 2008; Handl et al.2011), performing repeated gastrointestinal endoscopies under general anesthesia in healthy dogs was considered unethical. Examination of the microbiota in fecal samples of healthy controls both before and after treatment with the study's diet could have been a less invasive alternative to endoscopy. However, a study by one of the authors (JS) has revealed bacterial diversity along the different sections of the canine intestine (Suchodolski, Camacho and Steiner 2008). Therefore, a direct comparison of the fecal microbiota of healthy dogs and, for example, the duodenal mucosal microbiota of diseased dogs would not have been meaningful. The mucosal samples of this study were purposefully retrieved as part of a larger study, and the intestinal microbiota was analyzed retrospectively for this specific study. Unfortunately, residual fecal samples of the dogs diagnosed with IBD and FRD were not available for an additional analysis of the fecal microbiota in the current study. Moreover, in the authors’ opinion, mucosal samples are superior to fecal samples in assessing the true intestinal microbiota. Genomic DNA was extracted using the Mo Bio PowerSoil® DNA Isolation Kit, which is used in the Human Microbiome Project and is currently recommended for the extraction of DNA. The use of this method yielded the currently most effective DNA extraction (Wagner Mackenzie, Waite and Taylor 2015) and kept the possible influence of sample preparation on this study's results to a minimum. The different time scales for the follow-up endoscopies in the two disease groups could have possibly influenced the results. However, as glucocorticoids were not instituted until there was no significant clinical improvement on the elimination diet only on day 14, those dogs had to be given more time to respond to treatment. In general, FRD dogs are considered to respond faster to appropriate treatment, whereas dogs requiring immunosuppressants usually take longer to improve clinically. The different time scales were adopted from previous studies (Allenspach et al.2007; Burgener et al.2008; Dumusc et al.2014) and are considered more representative of a similarly controlled disease status. As most dogs diagnosed with FRD show a significant clinical response to a dietary trial within 14 days (Marks, Laflamme and McAloose 2002; Allenspach et al.2007; Gaschen and Merchant 2011; Allenspach, Culverwell and Chan 2016), the elimination trial was limited to that period of time. Furthermore, this short period provided for good owner compliance. Dietary modification is also an inherent part of treatment in dogs with IBD. Thus, it is possible that some dogs with IBD could have initially improved on the diet alone and could have been erroneously grouped in the FRD group. Yet, in the authors’ experience, dogs diagnosed with IBD generally only show minor improvement on elimination diet and would have probably not been assessed as significant improvement, much less as clinical remission. Another limitation of our study is the small number of dogs included and the fact that endoscopic biopsies were only obtained from the duodenum and colon. A larger study cohort of dogs and inclusion of additional biopsy sites (particularly the ileum) might have yielded more obvious differences in the mucosal microbial composition. Another shortcoming of our study is that the possibility of a continuing effect of previous medications (e.g. antibiotic treatment) on the mucosal microbiota cannot be definitively excluded, although all efforts were made to minimize such an effect by including only dogs not given an antibiotic 2 weeks prior to enrollment in the study. According to studies in human medicine, a 4-week wash-out period (Dethlefsen et al.2008; Langdon, Crook and Dantas 2016) may be more appropriate than a 2-week period and may have resulted in finding a more restored and original intestinal microbial composition. A study in veterinary medicine by one of the authors (JS), however, showed that the intestinal microbiota of dogs that received metronidazole returned to its original composition within 14 days after cessation of therapy in the majority of dogs (Olson et al.2015). Until now, there are only limited studies examining the effect of glucocorticoids on the intestinal microbial composition. One study revealed no effect of prednisolone on the fecal microbiota of dogs (Igarashi et al.2014) and to the authors’ knowledge, there are no studies evaluating the influence of budesonide on the intestinal microbiota. The authors did not expect a major impact of budesonide on the microbial composition, and therefore, the one dog receiving budesonide was still included in the analysis. The possibility of a long-term effect of the previous diet and prebiotics or probiotics given prior to the study can also not be entirely ruled out. Studies in both human and veterinary medicine have shown that the diet, in particular macronutrients, influences the intestinal microbial composition (Hang et al.2012; Conlon and Bird 2014; Xu and Knight 2015; Herstad et al.2017; Li et al.2017). However, the diets used in those studies were diets with pronounced alterations in the composition of macronutrients, which do not conform to commonly used diets. Moreover, a recent study showed that alpha diversity did not correlate with fat or protein intake in a population of dogs with IBD and control dogs (Vázquez-Baeza et al.2016), concluding that the disease effect is stronger than the diet's effect on the microbial composition. It has been recognized that obese individuals harbor a different intestinal microbiota with especially a decreased abundance of Bacteroides and an increased abundance of Firmicutes, which leads to an increased capacity to harvest energy (Ley et al.2005; Turnbaugh et al.2006, 2009). Thus, the body condition score (BCS) of dogs could have affected the results of the current study. However, 16 dogs were assigned a BCS between 4 and 6 out of 9 at their initial visit, representing a similar BCS for the majority of dogs. Furthermore, the study by Vázquez-Baeza et al. also revealed no correlation of the BCS and alpha diversity in the enrolled dogs. Therefore, the authors consider the influence of the BCS on this study's results negligible. An influence of lifestyle (e.g. smoking, exercising, stress) on the intestinal microbiota has been shown in humans (Conlon and Bird 2014), with, for example, a higher diversity of the intestinal microbes in extreme athletes (Clarke et al.2014). Although similar effects can be assumed in athletic or working dogs and dogs living in a smoking household, these effects are again only expected with extreme lifestyles, not reflecting the average companion dog. One study reported both age and breed, respectively size, to have an impact on the intestinal microbial composition in dogs (Simpson et al.2002), whereas another study did not find a correlation between age and the alpha diversity in dogs (Vázquez-Baeza et al.2016). More studies are needed to evaluate the true impact of these factors on the intestinal microbiota. All of the aforementioned possible influences could not be fully excluded in this study, as the dogs’ characteristics and environment could not be standardized in the case of a clinical trial. However, in the authors’ opinion, the diversity of dogs enrolled in this study represents a more realistic and therefore meaningful patient population. In conclusion, this study is the first to assess the intestinal mucosal microbiota in dogs with IBD or FRD before and after treatment. Some differences in individual bacterial taxa were identified both between disease groups and in relation to the treatment status. The relevance of these bacterial groups in the pathogenesis of IBD and FRD requires further research. However, our results suggest that Bacteroides might be a possible marker of successful response to treatment. Larger scale studies are warranted to confirm our findings and to shed more light on the functional consequences of the changes in the intestinal microbial composition shown in this study. Evaluation of their role in the pathophysiology of IBD and FRD and of their potential therapeutic benefit warrants further study. Acknowledgements The authors gratefully acknowledge the assistance of Deepti Gupta (Gastrointestinal Laboratory, Department of Small Animal Clinical Science, College of Veterinary Medicine and Biomedical Sciences, Texas A&M University, College Station, TX, USA) with the Illumina sequencing and statistical analysis. Conflict of interest. None declared. REFERENCES Allenspach K, Culverwell C, Chan D. Long-term outcome in dogs with chronic enteropathies: 203 cases. Vet Rec  2016; 178: 368. Google Scholar CrossRef Search ADS PubMed  Allenspach K, House A, Smith K et al.   Evaluation of mucosal bacteria and histopathology, clinical disease activity and expression of Toll-like receptors in German shepherd dogs with chronic enteropathies. Vet Microbiol  2010; 146: 326– 35. Google Scholar CrossRef Search ADS PubMed  Allenspach K, Wieland B, Gröne A et al.   Chronic enteropathies in dogs: evaluation of risk factors for negative outcome. J Vet Intern Med  2007; 21: 700– 8. Google Scholar CrossRef Search ADS PubMed  Bajer L, Kverka M, Kostovcik M et al.   Distinct gut microbiota profiles in patients with primary sclerosing cholangitis and ulcerative colitis. World J Gastroenterol  2017; 23: 4548– 58. Google Scholar CrossRef Search ADS PubMed  Bell ET, Suchodolski JS, Isaiah A et al.   Faecal microbiota of cats with insulin-treated diabetes mellitus. PLoS One  2014; 9: e108729. Google Scholar CrossRef Search ADS PubMed  Bernard K. The genus Corynebacterium and other medically relevant coryneform-like bacteria. J Clin Microbiol  2012; 50: 3152– 8. Google Scholar CrossRef Search ADS PubMed  Burgener IA, König A, Allenspach K et al.   Upregulation of Toll-like receptors in chronic enteropathies in dogs. J Vet Intern Med  2008; 22: 553– 60. Google Scholar CrossRef Search ADS PubMed  Caporaso JG, Kuczynski J, Stombaugh J et al.   QIIME allows analysis of high-throughput community sequencing data. Nat Methods  2010; 7: 335– 6. Google Scholar CrossRef Search ADS PubMed  Cassmann E, White R, Atherly T et al.   Alterations of the ileal and colonic mucosal microbiota in canine chronic enteropathies. PLoS One  2016; 11: e0147321. Google Scholar CrossRef Search ADS PubMed  Clarke SF, Murphy EF, Sullivan O et al.   Exercise and associated dietary extremes impact on gut microbial diversity. Gut  2014; 63: 1913. Google Scholar CrossRef Search ADS PubMed  Conlon MA, Bird AR. The impact of diet and lifestyle on gut microbiota and human health. Nutrients  2014; 7: 17– 44. Google Scholar CrossRef Search ADS PubMed  Dalal A, Likhi R. Corynebacterium minutissimum bacteremia and meningitis: a case report and review of literature. J Infect  2008; 56: 77– 79. Google Scholar CrossRef Search ADS PubMed  Dandrieux JR. Inflammatory bowel disease versus chronic enteropathy in dogs: are they one and the same? J Small Anim Pract  2016; 57: 589– 99. Google Scholar CrossRef Search ADS PubMed  Day MJ, Bilzer T, Mansell J et al.   Histopathological standards for the diagnosis of gastrointestinal inflammation in endoscopic biopsy samples from the dog and cat: a report from the World Small Animal Veterinary Association Gastrointestinal Standardization Group. J Comp Pathol  2008; 138( Suppl 1): S1– 43. Google Scholar CrossRef Search ADS PubMed  DeSantis TZ, Hugenholtz P, Larsen N et al.   Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl Environ Microb  2006; 72: 5069– 72. Google Scholar CrossRef Search ADS   Dethlefsen L, Huse S, Sogin ML et al.   The pervasive effects of an antibiotic on the human gut microbiota, as revealed by deep 16S rRNA sequencing. PLoS Biol  2008; 6: e280. Google Scholar CrossRef Search ADS PubMed  Dumusc SD, Ontsouka EC, Schnyder M et al.   Cyclooxygenase-2 and 5-lipoxygenase in dogs with chronic enteropathies. J Vet Intern Med  2014; 28: 1684– 91. Google Scholar CrossRef Search ADS PubMed  Edgar RC. Search and clustering orders of magnitude faster than BLAST. Bioinformatics  2010; 26: 2460– 1. Google Scholar CrossRef Search ADS PubMed  Flint HJ, Scott KP, Duncan SH et al.   Microbial degradation of complex carbohydrates in the gut. Gut Microbes  2012; 3: 289– 306. Google Scholar CrossRef Search ADS PubMed  Garcia-Mazcorro JF, Dowd SE, Poulsen J et al.   Abundance and short-term temporal variability of fecal microbiota in healthy dogs. MicrobiologyOpen  2012; 1: 340– 7. Google Scholar CrossRef Search ADS PubMed  Gaschen FP, Merchant SR. Adverse food reactions in dogs and cats. Vet Clin North Am-Small  2011; 41: 361– 79. Google Scholar CrossRef Search ADS   German AJ, Hall EJ, Day MJ. Chronic intestinal inflammation and intestinal disease in dogs. J Vet Intern Med  2003; 17: 8– 20. Google Scholar CrossRef Search ADS PubMed  Golińska E, Tomusiak A, Gosiewski T et al.   Virulence factors of Enterococcus strains isolated from patients with inflammatory bowel disease. World J Gastroenterol  2013; 19: 3562– 72. Google Scholar CrossRef Search ADS PubMed  Guard BC, Suchodolski JS. HORSE SPECIES SYMPOSIUM: Canine intestinal microbiology and metagenomics: From phylogeny to function. J Anim Sci  2016; 94: 2247– 61. Google Scholar CrossRef Search ADS PubMed  Hall EJ, German AJ. Diseases of the small intestine. In: Ettinger SJ, Feldman EC (eds). Textbook of Veterinary Internal Medicine , Vol. 2. Missouri, USA: Saunders Elsevier, St. Louis, 2010, 1560– 6. Handl S, Dowd SE, Garcia-Mazcorro JF et al.   Massive parallel 16S rRNA gene pyrosequencing reveals highly diverse fecal bacterial and fungal communities in healthy dogs and cats. FEMS Microbiol Ecol  2011; 76: 301– 10. Google Scholar CrossRef Search ADS PubMed  Hang I, Rinttila T, Zentek J et al.   Effect of high contents of dietary animal-derived protein or carbohydrates on canine faecal microbiota. BMC Vet Res  2012; 8: 90. Google Scholar CrossRef Search ADS PubMed  Herstad KMV, Gajardo K, Bakke AM et al.   A diet change from dry food to beef induces reversible changes on the faecal microbiota in healthy, adult client-owned dogs. BMC Vet Res  2017; 13: 147. Google Scholar CrossRef Search ADS PubMed  Igarashi H, Maeda S, Ohno K et al.   Effect of oral administration of metronidazole or prednisolone on fecal microbiota in dogs. PLoS One  2014; 9: e107909. Google Scholar CrossRef Search ADS PubMed  Jergens A, Moore F, Haynes J et al.   Idiopathic inflammatory bowel disease in dogs and cats: 84 cases (1987-1990). J Am Vet Med Assoc  1992; 201: 1603– 8. Google Scholar PubMed  Jergens AE, Schreiner CA, Frank DE et al.   A scoring index for disease activity in canine inflammatory bowel disease. J Vet Intern Med  2003; 17: 291– 7. Google Scholar CrossRef Search ADS PubMed  Kathrani A, Lee H, White C et al.   Association between nucleotide oligomerisation domain two (Nod2) gene polymorphisms and canine inflammatory bowel disease. Vet Immunol Immunop  2014; 161: 32– 41. Google Scholar CrossRef Search ADS   Kilpinen S, Rantala M, Spillmann T et al.   Oral tylosin administration is associated with an increase of faecal enterococci and lactic acid bacteria in dogs with tylosin-responsive diarrhoea. Vet J  2015; 205: 369– 74. Google Scholar CrossRef Search ADS PubMed  Kilpinen S, Spillmann T, Syrja P et al.   Effect of tylosin on dogs with suspected tylosin-responsive diarrhea: a placebo-controlled, randomized, double-blinded, prospective clinical trial. Acta Vet Scand  2011; 53: 26. Google Scholar CrossRef Search ADS PubMed  Langdon A, Crook N, Dantas G. The effects of antibiotics on the microbiome throughout development and alternative approaches for therapeutic modulation. Genome Med  2016; 8: 39. Google Scholar CrossRef Search ADS PubMed  Ley RE, Backhed F, Turnbaugh P et al.   Obesity alters gut microbial ecology. P Natl Acad Sci USA  2005; 102: 11070– 5. Google Scholar CrossRef Search ADS   Li Q, Lauber CL, Czarnecki-Maulden G et al.   Effects of the dietary protein and carbohydrate ratio on gut microbiomes in dogs of different body conditions. mBio  2017; 8: e01703– 01716. Google Scholar PubMed  Mandigers PJ, Biourge V, van den Ingh TS et al.   A randomized, open-label, positively-controlled field trial of a hydrolyzed protein diet in dogs with chronic small bowel enteropathy. J Vet Intern Med  2010; 24: 1350– 7. Google Scholar CrossRef Search ADS PubMed  Marks S, Laflamme DP, McAloose D. Dietary trial using a commercial hypoallergenic diet containing hydrolyzed protein for dogs with inflammatory bowel disease. Vet Ther  2002; 3: 109– 18. Google Scholar PubMed  Minamoto Y, Otoni CC, Steelman SM et al.   Alteration of the fecal microbiota and serum metabolite profiles in dogs with idiopathic inflammatory bowel disease. Gut Microbes  2015; 6: 33– 47. Google Scholar CrossRef Search ADS PubMed  Olson E, Honneffer JB, Waddle M et al.   Evaluation of the effects of a 2 week treatment with metronidazole on the fecal microbiome of healthy dogs (abstr.). J Vet Intern Med  2015; 29: 1184. Pavlidis P, Powell N, Vincent RP et al.   Systematic review: bile acids and intestinal inflammation-luminal aggressors or regulators of mucosal defence? Aliment Pharm Ther  2015; 42: 802– 17. Google Scholar CrossRef Search ADS   Reis AC, Silva JO, Laranjeira BJ et al.   Virulence factors and biofilm production by isolates of Bacteroides fragilis recovered from dog intestinal tracts. Braz J Microbiol  2014; 45: 647– 50. Google Scholar CrossRef Search ADS PubMed  Sauter SN, Benyacoub J, Allenspach K et al.   Effects of probiotic bacteria in dogs with food responsive diarrhoea treated with an elimination diet. J Anim Physiol An N  2006; 90: 269– 77. Google Scholar CrossRef Search ADS   Schmitz S, Suchodolski J. Understanding the canine intestinal microbiota and its modification by pro-, pre- and synbiotics - what is the evidence? Vet Med Sci  2016; 2: 71– 94. Google Scholar CrossRef Search ADS PubMed  Segata N, Izard J, Waldron L et al.   Metagenomic biomarker discovery and explanation. Genome Biol  2011; 12: R60. Google Scholar CrossRef Search ADS PubMed  Sim WH, Wagner J, Cameron DJ et al.   Novel Burkholderiales 23S rRNA genes identified in ileal biopsy samples from children: preliminary evidence that a subtype is associated with perianal Crohn's disease. J Clin Microbiol  2010; 48: 1939– 42. Google Scholar CrossRef Search ADS PubMed  Simpson JM, Martineau B, Jones WE et al.   Characterization of fecal bacterial populations in canines: effects of age, breed and dietary fiber. Microb Ecol  2002; 44: 186– 97. Google Scholar CrossRef Search ADS PubMed  Simpson KW, Jergens AE. Pitfalls and progress in the diagnosis and management of canine inflammatory bowel disease. Vet Clin N Am Small  2011; 41: 381– 98. Google Scholar CrossRef Search ADS   Suchodolski JS, Camacho J, Steiner JM. Analysis of bacterial diversity in the canine duodenum, jejunum, ileum, and colon by comparative 16S rRNA gene analysis. FEMS Microbiol Ecol  2008; 66: 567– 78. Google Scholar CrossRef Search ADS PubMed  Suchodolski JS, Dowd SE, Wilke V et al.   16S rRNA gene pyrosequencing reveals bacterial dysbiosis in the duodenum of dogs with idiopathic inflammatory bowel disease. PLoS One  2012a; 7: e39333. Google Scholar CrossRef Search ADS   Suchodolski JS, Markel ME, Garcia-Mazcorro JF et al.   The fecal microbiome in dogs with acute diarrhea and idiopathic inflammatory bowel disease. PLoS One  2012b; 7: e51907. Google Scholar CrossRef Search ADS   Suchodolski JS, Xenoulis PG, Paddock CG et al.   Molecular analysis of the bacterial microbiota in duodenal biopsies from dogs with idiopathic inflammatory bowel disease. Vet Microbiol  2010; 142: 394– 400. Google Scholar CrossRef Search ADS PubMed  Turnbaugh PJ, Ley RE, Mahowald MA et al.   An obesity-associated gut microbiome with increased capacity for energy harvest. Nature  2006; 444: 1027– 131. Google Scholar CrossRef Search ADS PubMed  Turnbaugh PJ, Hamady M, Yatsunenko T et al.   A core gut microbiome in obese and lean twins. Nature  2009; 457: 480– 4. Google Scholar CrossRef Search ADS PubMed  Vázquez-Baeza Y, Hyde ER, Suchodolski JS et al.   Dog and human inflammatory bowel disease rely on overlapping yet distinct dysbiosis networks. Nat Microbiol  2016; 1: 16177. Google Scholar CrossRef Search ADS PubMed  Wagner KS, White JM, Lucenko I et al.   Diphtheria in the postepidemic period, Europe, 2000–2009. Emerg Infect Dis  2012; 18: 217– 25. Google Scholar CrossRef Search ADS PubMed  Wagner Mackenzie B, Waite DW, Taylor MW. Evaluating variation in human gut microbiota profiles due to DNA extraction method and inter-subject differences. Front Microbiol  2015; 6: 130. Google Scholar CrossRef Search ADS PubMed  Westermarck E, Skrzypczak T, Harmoinen J et al.   Tylosin-responsive chronic diarrhea in dogs. J Vet Intern Med  2005; 19: 177– 86. Google Scholar CrossRef Search ADS PubMed  Wexler HM. Bacteroides: the good, the bad, and the nitty-gritty. Clin Microbiol Rev  2007; 20: 593– 621. Google Scholar CrossRef Search ADS PubMed  Xavier RJ, Podolsky DK. Unravelling the pathogenesis of inflammatory bowel disease. Nature  2007; 448: 427– 34. Google Scholar CrossRef Search ADS PubMed  Xu Z, Knight R. Dietary effects on human gut microbiome diversity. Br J Nutr  2015; 113( Suppl): S1– 5. Google Scholar CrossRef Search ADS PubMed  © FEMS 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

FEMS Microbiology EcologyOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial