Introduction
Characterization of the bacterial composition within the intestinal tract is important to our understanding of the interplay between dysbiosis and disease. The most common method for determining the intestinal tract bacteriome in veterinary species is 16S rRNA gene sequencing by use of fecal specimens that have been naturally voided or collected during rectal examination.1 Although fecal specimens are easily obtained, they may not represent those bacteria adherent to the mucosal surface of the distal portion of the intestinal tract. Mucosal biopsy specimens may better represent this bacterial population; however, intestinal mucosal biopsy is an invasive procedure that requires that animals be anesthetized, and the procedure has inherent risks. On the other hand, rectal swab specimens may serve as a noninvasive and convenient alternative with potential to represent both bacteria adherent to the mucosa and those present in the intraluminal feces.
To the authors' knowledge, no comparison has been performed of the usefulness of various biological specimen types in the characterization of bacterial communities within the intestinal tract of dogs and cats by use of 16S rRNA gene sequencing. In people, the similarity of findings among fecal, rectal swab, and mucosal biopsy specimens is variable. For example, fecal and rectal swab bacterial profiles of hospital patients are reportedly similar, so these specimen types could be used interchangeably to assess bacterial composition by 16S rRNA gene sequencing.2 On the contrary, for people undergoing elective colonoscopy and patients with inflammatory bowel disease, bacterial profiles determined by 16S rRNA gene sequencing differ distinctly among fecal, rectal swab, and mucosal biopsy specimens in most cases.3 Similar results have been reported for people with a history of colorectal polyps for whom whole-genome shotgun sequencing was used to assess bacterial composition.4
Differences in findings for bacterial composition among fecal, rectal swab, and mucosal biopsy specimens could be expected to have important implications for the understanding and interpretation of changes in the intestinal tract bacteriome in diseased states or following treatment. From a practical standpoint, relying on owners to collect, preserve, and deliver fecal specimens for analysis may introduce various biases (eg, incorrect animal [as in multicat households], differences in specimen storage conditions, and lags from specimen collection to processing) with potential to affect the results. Therefore, the objective of the study reported here was to compare bacterial diversity and community composition determined by 16S rRNA gene sequencing among fecal, rectal swab, and CMB specimens from dogs and cats with a healthy or diseased (ie, CE) intestinal tract. We hypothesized that the bacterial composition would differ among these specimen types, particularly for dogs and cats with CE.
Materials and Methods
Animals and specimen collection
Fecal, rectal swab, and, when possible, mucosal biopsy specimens from the descending colon were collected from client-owned dogs and cats presented to the Colorado State University Veterinary Teaching Hospital between June and October 2019. Dogs and cats were each grouped as healthy or with CE. The study protocol was approved by the Institutional Animal Care and Use Committee (protocol No. 19-8701A) and the Clinical Review Board (protocol No. 2019-202) of Colorado State University. All animal owners provided written informed consent before participation.
Healthy dogs and cats
Dogs and cats presented for a routine wellness examination or prior to elective dental scaling (and anesthesia) were eligible for inclusion in the study. Health status of dogs was confirmed on the basis of medical history and unremarkable results of physical examination, a CBC, and serum biochemical analysis. Health status of cats was confirmed on the basis of these criteria as well as unremarkable results of urinalysis, fecal flotation, and serum total thyroxine concentration assay. Animals were excluded from the healthy group if they had a history of acute or chronic diarrhea or vomiting, known or suspected (eg, food-responsive) intestinal disease, uncontrolled systemic disease, or medications known to affect the intestinal bacteriome (ie, antimicrobials or antacids) received within 6 weeks before study enrollment.
Dogs and cats with CE
Dogs and cats presented for evaluation of CE (defined as weight loss, vomiting, diarrhea, inappetence, or hematochezia of at least 3 weeks' duration) and for which colonoscopy was planned as an appropriate part of the diagnostic evaluation were eligible for inclusion in the study. All dogs and cats with CE had negative results of fecal flotation, had undergone prophylactic deworming, or both prior to specimen collection. Animals were excluded if anesthesia or colonoscopy was contraindicated.
Specimen collection
Owners were asked to collect a voided fecal specimen (≥ 1 g) at home. For dogs, the fecal specimen was collected immediately after defecation the day of or 1 day before colonoscopy. For cats, it was requested that the specimen be collected from the litter box within 12 hours after defecation and, for cats in the CE group, within 2 days before colonoscopy. Owners were asked to store the sample in a clean container or bag that was sealed and, if needed, stored on ice or in a 4°C refrigerator until transported on ice to the veterinary teaching hospital within 24 hours after defecation. Fecal specimens were frozen at–80°C immediately on receipt by the authors (SCS or AG). If fecal specimens could not be obtained at home, then a fecal specimen was manually collected from the rectum during anesthesia prior to colonoscopy and only from animals that had not undergone colonic preparation (ie, a warm water enema prior to colonoscopy). Rectal swab specimens were collected at the veterinary teaching hospital by the authors (SCS or AG) the day before or just prior to colonoscopy for animals that had not undergone colonic preparation. A dry sterile nylon swaba was inserted at least 2 cm into the anal canal and rotated, avoiding perianal skin or hair. Rectal swab specimens were frozen at–80°C within 1 hour after collection.
Colonoscopy was performed in healthy dogs that underwent anesthesia for dental prophylaxis and in dogs and cats with CE by use of a flexible video endoscope.b For some but not all animals with CE, 1 to 3 warm water enemas were performed in preparation for the procedure. Colonic mucosal biopsy was performed in the descending colon (≤ 20 cm into the rectum). Two biopsy specimens were obtained for gene sequencing by means of a sterile flexible biopsy instrument,c placed in sterile cryotubes, and frozen at–80°C within 1 hour after collection. In dogs and cats with CE, additional biopsy specimens of the stomach, duodenum, ileum, and colon were obtained for histologic evaluation. All specimens were stored at–80°C for < 6 months prior to analysis.
DNA extraction
Fecal and rectal swab specimens were thawed at room temperature (20°C) for approximately 20 minutes, a 250-mg portion of feces was obtained from each fecal specimen, and DNA was extracted from the 250-mg portion and swab specimens by use of commercial kitsd with the same lot number, in accordance with the manufacturer's instructions. For rectal swab specimens, the tip of the nylon swab was cut from the stem and placed directly into the tube for DNA extraction in the same manner as fecal specimens. The CMB specimens were thawed, and DNA was extracted with a different commercial kite in accordance with the manufacturer's instructions.
16S rRNA gene sequencing
The 16S rRNA V4 gene region in extracted DNA samples was amplified, and sequencing was performedf by use of primers 515F to 806R on a paired-end platform. Sequences were processed and analyzed with open-source bioinformatics softwareg as previously described.5 Sequences were filtered for chimeras, and sequence analysis was performed by use of all effective tags.6–8 Sequences with ≥ 97% similarity were assigned to the same OTUs. The representative sequence for each OTU was screened for further annotation. Open-source software for bioinformatics data processing9,h was used to generate taxonomic assignments for each representative sequence with a naïve Bayesian classifer10 and 16S rRNA gene database.11 Another software programi was used to determine the phylogenetic relationship of all OTU representative sequences as described elsewhere.12 Abundance information was normalized by use of a sequence number corresponding to the sample with the least sequences. Subsequent analyses of α diversity and β diversity were all performed on the basis of this normalized data output.
To evaluate bacterial community diversity within each specimen, within-specimen (α) diversity was measured by determination of the richness (ie, the number of observed OTUs), Chao1 estimate of total species richness, and Simpson diversity index for the specimen. The Chao1 estimate gives more weight to low-abundance species. The Simpson diversity index is a measure of diversity within communities and considers both species richness and evenness within a bacterial community. To assess between-specimen (β) diversity, the Bray-Curtis dissimilarity index was calculated. The Bray-Curtis dissimilarity index is a nonphylogenetic measure of β diversity that quantifies the species composition between specimens. An NMDS plot was generated by use of the Bray-Curtis dissimilarity index to visualize differences in bacterial communities among the various specimen types in the 4 subject groups. Sequences were deposited in the National Institute of Health Sequence Read Archive under accession No. PRJNA607571.
Statistical analysis
Statistical analyses were conducted with the aid of statistical software.j,k The Friedman test with the Dunn multiple comparisons test or Wilcoxon matched-paired signed rank test was used to determine differences in α diversity indices among specimen types for each animal group. A software package designed for analyzing ecological communitiesl was used to conduct an analysis of similarities with the Bray-Curtis dissimilarity index to determine whether community structure (ie, β diversity) significantly differed between specimen types for each animal group. The Bray-Curtis dissimilarities were visualized by generation of NMDS plots. The R value of Bray-Curtis dissimilarities signifies the similarity in bacterial composition between groups of specimens and ranges between 0 and 1.0. An R value close to 0 indicates similar communities between specimen types, and an R value close to 1.0 indicates dissimilar communities between specimen types. The LDA effect size was used to highlight differentially abundant bacteria taxa that explained the differences among the 3 specimen types. The LDA effect size cutoff was set to 4. The larger the LDA effect size, the more the bacterial taxa explain the differentiating phenotypes among the 3 specimen types. For all analyses, a value of P < 0.05 was considered significant.
Results
Animals
Characteristics of dogs and cats that participated in the study, including any medications or nutritional supplements they were receiving at enrollment, were summarized (Supplementary Table S1, available at: avmajournals.avma.org/doi/suppl/10.2460/ajvr.82.6.494).
Healthy dogs
Fecal, rectal swab, and CMB specimens were collected from 9 healthy dogs with a median age of 5 years (range, 2 to 9 years). All healthy dogs were fed a commercial nonprescription adult dog food. At the time of enrollment, 7 of the 9 dogs were receiving monthly prophylactic antiparasitic medication (no fecal flotation performed), 1 dog had a negative fecal flotation result, and 1 dog had neither. None of the healthy dogs received a warm water enema prior to colonoscopy.
Healthy cats
Fecal and rectal swab specimens were collected from 8 healthy cats with a median age of 11.5 years (range, 10 to 13 years). Because no healthy cats underwent anesthesia for dental scaling, none underwent colonoscopy and no CMB specimens were collected. Cats were fed a commercial nonprescription adult or senior cat food. All healthy cats had negative fecal flotation results at the time of specimen collection.
Dogs with CE
Fecal, rectal swab, and CMB specimens were collected from 8 dogs with CE. Median age was 7.5 years (range, 1 to 12 years). Diets included a hypoallergenic (n = 3 dogs) or low-fat (1 dog) prescription dog food, home-cooked diet (2 dogs), and nonprescription adult dog food (2 dogs). Indications for colonoscopy included chronic diarrhea, vomiting, tenesmus, inappetence, or dysrexia for 6 of the 8 dogs. Two dogs had a protein-losing enteropathy characterized by hypoalbuminemia and chronic gastrointestinal signs. Of the 8 dogs with CE, 6 received a histologic diagnosis of inflammatory bowel disease characterized as lymphoplasmacytic or eosinophilic inflammation. One dog had neutrophilic inflammation with invasive Escherichia coli detected by fluorescence in situ hybridization and ulcerative colitis, and another dog had intestinal lymphangiectasia. Six dogs received warm water enemas prior to colonoscopy, whereas 2 did not.
Cats with CE
Fecal, rectal swab, and CMB specimens were collected from 9 cats with CE. Median age was 7 years (range, 4 to 14.5 years). Indications for colonoscopy included chronic diarrhea, vomiting, inappetence, hematochezia, or weight loss. Six of the 9 cats had small cell lymphoma as diagnosed by histologic and immunohistochemical evaluation. The remaining 3 cats had inflammatory bowel disease characterized as lymphoplasmacytic enteritis on histologic evaluation. One of the cats with CE received a warm water enema prior to colonoscopy.
Specimens
The DNA extraction was performed on a total of 94 specimens, including 27 specimens from healthy dogs, 16 from healthy cats, 24 from dogs with CE, and 27 from cats with CE. Quality genomic DNA was amplified from all specimens except the CMB specimens from 2 cats with CE; these 2 specimens were excluded from gene sequencing. Thus, a total of 92 specimens were represented in the 16S rRNA gene sequencing results.
Sequence analysis yielded 10,385,788 quality sequences for all 92 analyzed specimens (median, 117,518 sequences; range, 49,152 to 174,962 sequences). The relative abundances of 16S rRNA gene sequences (V4 region), classified to the order level, were graphically displayed for each specimen type for each animal group (Figure 1). A Venn diagram was created to compare the shared OTUs in the bacterial profiles of the 3 specimen types within each animal group (Supplementary Figure S1, available at: avmajournals.avma.org/doi/suppl/10.2460/ajvr.82.6.494).
Bacterial community composition (relative abundance) for fecal specimens (F), rectal swab specimens (S), and, when available, CMB specimens (B) in healthy cats (FH; n = 8) and dogs (CH; 9) and diseased cats (FD; 9) and dogs (CD; 8) with CE. The relative abundances of 16S rRNA gene sequences (V4 region) are shown, classified by bacterial order. Data for CMB specimens from cats with CE represent only the 7 cats for which they were available.
Citation: American Journal of Veterinary Research 82, 6; 10.2460/ajvr.82.6.494
Healthy dogs
Among healthy dogs, bacterial richness, as described by the number of observed OTUs, was significantly (P = 0.03) higher for CMB specimens than for rectal swab specimens (Table 1). No significant difference in richness was found between fecal and rectal swab specimens (P > 0.99) or between fecal and CMB specimens (P = 0.06). Chao1 values were significantly (P = 0.03) higher for CMB specimens than for fecal specimens. No significant difference in Chao1 values was found between rectal swab and fecal specimens (P > 0.99) or between rectal swab and CMB specimens (P = 0.06). No significant (P = 0.15) difference in Simpson diversity index values was found among the 3 specimen types.
Median (range) α diversity indices (richness [No. of OTUs], Chao1, and Simpson) for fecal, rectal swab, and CMB specimens collected from healthy dogs (n = 9), healthy cats (8), dogs with CE (8), and cats with CE (7*).
| Specimen type | Richness | Chao1 | Simpson |
|---|---|---|---|
| Healthy dogs | |||
| Fecal | 925 (538–2,594) | 1,263 (761–3,827)a | 0.947 (0.836–0.967) |
| Rectal swab | 912 (544–2,281)a | 1,538 (886–3,395) | 0.946 (0.908–0.979) |
| CMB | 1,871 (1,278–3,332)b | 2,914 (2,066–4,879)b | 0.969 (0.901–0.979) |
| Healthy cats | |||
| Fecal | 666 (440–10,93) | 1,104 (554–1,677) | 0.942 (0.810–0.970) |
| Rectal swab | 931 (685–1,352) | 1,422 (1,135–1,933) | 0.928 (0.736–0.960) |
| Dogs with CE | |||
| Fecal | 743 (510–1,285)a | 1,102 (608–2,053)a | 0.912 (0.805–0.961) |
| Rectal swab | 431 (346–1,725)a | 612 (466–2,133)a | 0.927 (0.600–0.971) |
| CMB | 2,317 (1,035–3,298)b | 3,571 (1,713–4,236)b | 0.916 (0.811–0.985) |
| Cats with CE | |||
| Fecal | 595 (450–834)a | 816 (581–1,284)a | 0.945 (0.579–0.972) |
| Rectal swab | 905 (670–1525) | 1,425 (1,041–2,623) | 0.940 (0.708–0.956) |
| CMB | 1,187 (902–1,922)b | 1,692 (1,453–2,858)b | 0.964 (0.662–0.972) |
For paired analysis, only cats that had all 3 specimen types (7/9 cats with CE) were included in the analysis.
Within a column and animal group, values with different superscript letters differ significantly (P < 0.05).
When individual bacterial taxa were analyzed on the basis of the LDA effect size, the relative abundance of the genus Collinsella (LDA score, 4.1) was higher for fecal specimens versus rectal swab and CMB specimens. The relative abundance of the phylum Proteobacteria (LDA score, 4.6) was higher for CMB specimens versus fecal and rectal swab specimens.
Healthy cats
Among healthy cats, no significant (P = 0.06, P = 0.15, and P = 0.20, respectively) difference in bacterial richness, Chao1, and Simpson diversity index values was found between fecal and rectal swab specimens (Table 1). The LDA effect size indicated no differentially abundant bacterial taxa between fecal and rectal swab specimens.
Dogs with CE
Among dogs with CE, bacterial richness and Chao1 values for CMB specimens were significantly higher than for fecal specimens (P = 0.04 for both comparisons) and rectal swab specimens (P = 0.001 for both comparisons; Table 1). No significant (P = 0.95) difference in richness or Chao1 values was found between fecal and rectal swab specimens. Simpson diversity index values did not differ among the 3 specimen types (P > 0.99).
The LDA effect size indicated several distinguishing bacterial taxa for fecal, rectal swab, and CMB specimens. Fecal specimens had a higher relative abundance of the family Enterococcaceae (LDA score, 4.1) and the genera Blautia (LDA score, 4.6) and Enterococcus (LDA score, 4.1) than did rectal swab and CMB specimens. Results for CMB specimens indicated a higher relative abundance of the class Alpha-proteobacteria (LDA score, 4.1) than did fecal and rectal swab specimens.
Cats with CE
Among cats with CE, bacterial richness and Chao1 values were significantly (P = 0.02 and P = 0.049, respectively) higher for CMB specimens than for fecal specimens (Table 1). No significant difference in richness and Chao1 values was found between rectal swab and fecal specimens (P = 0.10 and P = 0.33, respectively) or between rectal swab and CMB specimens (P > 0.99 for both comparisons). No significant (P = 0.19) difference in Simpson diversity index values was found among the 3 specimen types. Several distinguishing bacterial taxa were identified by LDA effect size for the 3 specimen types (Figure 2).
Histogram of LDA effect size (logarithmic scale) showing differentially abundant bacterial taxa from CMB (red), fecal (green), and rectal swab (blue) specimens for 9 cats with CE. Bacterial species classification is reported at the phylum (p), class (c), order (o), family (f), and genus (g) level.
Citation: American Journal of Veterinary Research 82, 6; 10.2460/ajvr.82.6.494
β Diversity
A significant difference in β diversity between fecal and rectal swab specimens was found for healthy cats and cats with CE and between fecal and CMB specimens for dogs and cats with CE (Table 2; Figures 3 and 4). The relative abundances of the top 10 bacterial taxa (classified to order level) captured for each specimen type for each individual cat or dog were summarized (Supplementary Figure S2, available at: avmajournals.avma.org/doi/suppl/10.2460/ajvr.82.6.494).
Bray-Curtis dissimilarity index (R) values for bacterial communities in fecal, rectal swab, and CMB specimens collected from the animals of Table 1.
| Specimen type | R value | P value |
|---|---|---|
| Healthy dogs | ||
| Fecal vs rectal swab | –0.012 | 0.54 |
| CMB vs fecal | 0.079 | 0.08 |
| CMB vs rectal swab | 0.10 | 0.06 |
| Healthy cats | ||
| Fecal vs rectal swab | 0.815 | 0.001 |
| Dogs with CE | ||
| Fecal vs rectal swab | –0.017 | 0.49 |
| CMB vs fecal | 0.251 | 0.02 |
| CMB vs rectal swab | 0.106 | 0.09 |
| Cats with CE | ||
| Fecal vs rectal swab | 0.272 | 0.001 |
| CMB vs fecal | 0.192 | 0.02 |
| CMB vs rectal swab | –0.007 | 0.49 |
Results of NMDS analysis based on Bray-Curtis dissimilarity of 16S sequencing of rRNA genes from fecal (circles), rectal swab (triangles), and CMB specimens (squares) for the healthy dogs (A), healthy cats (B), dogs with CE (C), and cats with CE (D) of Figure 1. Each data point represents a single specimen. The distance between data points reflects the extent of variation. Values for stress reflect how well the ordination summarizes the observed distances among the specimens. A plot with a stress value < 0.2 provides a good representation. The MDS1 and MDS2 are axes showing NMDS coordinates.
Citation: American Journal of Veterinary Research 82, 6; 10.2460/ajvr.82.6.494
Results of NMDS analysis based on Bray-Curtis dissimilarity of 16S sequencing of rRNA genes by individual animal for the animals (each represented by a different color) and specimens (each represented by a different symbol) of Figure 1. See Figure 3 for remainder of key.
Citation: American Journal of Veterinary Research 82, 6; 10.2460/ajvr.82.6.494
Discussion
The purpose of the present study was to determine similarities and differences in colonic bacterial communities among fecal, rectal swab, and CMB specimens from dogs and cats. Although similarities were found among these 3 specimen types, there were also distinct differences, particularly between fecal and CMB specimens. The bacterial richness, with the exception of healthy dogs, and the estimated richness based on Chao1 values were higher for CMB specimens than for fecal specimens. However, no difference in Simpson diversity index values signifying richness and evenness within the bacterial communities was identified among the specimen types for any animal group. Comparison of the bacterial community profiles among specimen types by means of a nonphylogenetic β diversity index (Bray-Curtis dissimilarity) showed that rectal swab specimens yielded results similar to those for fecal specimens in dogs and to results for CMB specimens in dogs and in cats with CE. Whereas bacterial communities shared some similarities, as a whole, the 3 specimen types appeared to capture distinct bacterial communities, and the bacterial profiles among the 3 specimen types differed within individual animals.
Although fecal specimens are convenient to collect, especially for dog owners, specimen collection by pet owners may introduce bias by contamination during collection, delay in specimen freezing, and risk of freeze-thaw during transport. Rectal swab specimens may provide a more convenient and reliable specimen for researchers. In addition, rectal swab specimens can be collected in the hospital, allowing for more standardized specimen collection and processing. Studies involving the use of 16S rRNA gene sequencing to compare bacterial communities between rectal swab and fecal specimens in both healthy and diseased people have yielded variable results. In healthy people, fecal and rectal swab specimens from the same individual had similar bacterial profiles.2,13 In people with inflammatory bowel disease and oncological patients, bacterial profiles for fecal specimens resembled those of rectal swab specimens, especially when specimens were collected at the same time.3,13 In a group of critically ill humans, bacterial diversity and composition differed between rectal swab and fecal specimens.14 In the present study, rectal swab and fecal specimens from both dogs and cats had similar bacterial diversity as indicated by α diversity indices and those from dogs had a similar bacterial community structure; the bacterial community structure differed between these 2 specimen types in both groups of cats. Although the bacterial profiles for rectal swab and fecal specimens were not interchangeable in our study, rectal swab collection appeared to be a valid sampling method for bacteriome characterization in situations when fresh fecal specimen collection is challenging, such as in multicat households.
In the study reported here, bacterial diversity was higher in CMB specimens than in fecal specimens and bacterial composition differed between these 2 specimen types for dogs and cats with CE. The CMB specimens had a higher abundance of bacteria belonging to the phylum Proteobacteria for both groups of dogs. For cats with CE, the abundance of bacteria belonging to the families Campylobacteraceae and Pasteurellaceae was greater in CMB specimens than in fecal and rectal swab specimens. Colonic mucosal biopsy specimens will capture adherent bacteria that reside in the mucosal layer of the colon, but potentially miss bacterial taxa within the intestinal lumen. On the other hand, fecal specimens contain transient luminal bacteria and may not represent the mucosa-associated bacterial community.15,16 Therefore, CMB specimens might be a more appropriate specimen type to use when evaluating the interface between the microbiota and the host immune system, microbiota-related intestinal abnormalities, or the impact of treatments aimed at altering the bacteriome. It appears that a rectal swab specimen may capture some of the adherent mucosal bacteria. The bacterial community captured by rectal swab specimens in our study was not significantly different from that captured by CMB specimens (excluding the results from healthy cats) as indicated by Bray-Curtis dissimilarity and Simpson diversity index values. Thus, rectal swab specimens may be an important sampling method for capturing adherent mucosal bacteria that does not require animals to undergo anesthesia and colonoscopy.
Several possible drawbacks exist regarding the use of rectal swab specimens to sample the intestinal bacteriome. One is the potential for contamination of the swab with skin bacteria during specimen collection.13 We evaluated the relative abundance of skin commensals and found that Streptococcus, Staphylococcus, and Corynebacterium spp represented < 1% of all bacteria in each group; therefore, we believe skin contamination is unlikely to affect analysis, at least for specimens with a similarly high biomass. Another concern with rectal swab specimens is the yield of genomic DNA for sequencing. In the present study, all rectal swab specimens yielded adequate DNA (≥ 150 ng) and specimen processing and gene sequencing yielded a number of reads per specimen (median, 108,721 reads; range, 43,618 to 156,929 reads) similar to that for fecal specimens (median, 106,957 reads; range, 46,810 to 163,793) and CMB specimens (median, 104,846 reads; range, 38,937 to 145,720 reads). Although our study was not designed to assess safety, adverse events related to rectal swab collection were uncommon in the study dogs and cats. One rectal swab specimen collected from a healthy dog had gross evidence of blood contamination; however, results of rectal examination immediately after specimen collection were unremarkable and the owner reported no concerns about the dog once at home.
The present study had some limitations. No negative control samples were included during DNA extraction, and therefore the possibility of sample contamination during DNA extraction could not be excluded. The kitome was not characterized. The DNA extraction kit used for rectal swab and fecal specimens was different from the kit used for CMB specimens, which could have affected comparisons of results between the specimen types. Because only 1 specimen was collected at a single time point for each animal, we were unable to determine the repeatability of these findings over time. In addition, the animal groups were heterogeneous in composition. For example, 3 healthy dogs were receiving medication for treatment of hypothyroidism or atopy, and some of the dogs and cats with CE were receiving a probiotic or antacid for treatment of their intestinal disease at the time of specimen collection. Although these medications may have altered the bacteriome of the descending colon to a degree, the bacteriome captured by each specimen type should have been equally affected and therefore the group heterogeneity was unlikely to have affected our results. Although most (7/9) healthy dogs were receiving a monthly prophylactic antiparasitic medication, no fecal flotation was performed for most (8/9) of them. Therefore, a subclinical intestinal parasite infection could not be excluded in those healthy dogs. We did not evaluate the differences in functional gene potential among the specimen types, and findings related only to descriptive compositional analysis of the intestinal bacteriome. Lastly, rectal swab specimens were collected from dogs and cats with CE prior to warm water enemas, which may have increased the chance of specimen contamination with luminal contents. This may explain the similarity between fecal and rectal swab specimens in the dogs of the present study because fecal contamination of the swab tip subjectively appeared to occur more commonly for dogs than for cats. Similarly, 6 of 8 dogs with CE and 1 of 9 cats with CE had at least 1 warm water enema prior to collection of CMB specimens, which might have altered the bacteriome captured by the specimens, albeit to an unknown degree, and thus impacted the results for these animal groups.
Although specimen types are not interchangeable for the purpose of microbiota gene sequencing, all 3 specimen types (fecal, rectal swab, and CMB) in the study reported here were successfully used to characterize the intestinal bacteriome in dogs and cats by 16S rRNA gene sequencing. The CMB specimens had a distinct bacterial community with high diversity in dogs and cats with CE, compared with feces, and might be a better representation of the bacterial biofilm of the colonic mucosa. Findings suggested that rectal swab specimens are practical alternatives for use in dogs and cats to study the intestinal bacterial composition and may be an alternate option to CMB and fecal specimen collection. This important possibility requires further research.
Acknowledgments
Funded by the Comparative Gastroenterology Society/IdexxVeterinary Student Summer Scholar Grant.
The authors declare that there were no conflicts of interest.
Presented as a poster at the American College of Veterinary Internal Medicine Forum, June 2020.
The authors thank Rae Isdale (Colorado State University) for helping with patient recruitment and specimen collection.
Abbreviations
| CE | Chronic enteropathy |
| CMB | Colonic mucosal biopsy |
| LDA | Linear discriminant analysis |
| NMDS | Nonmetric multidimensional scaling |
| OTU | Operational taxonomic unit |
Footnotes
FLOQSwabs, COPAN Diagnostics Inc, Murrieta, Calif.
Model No. 60714 NK, Karl Storz, Tuttlingen, Germany.
Standard 2.8-mm alligator biopsy forceps, Micro-Tech Endoscopy, Ann Arbor, Mich.
QIAamp PowerFecal DNA Kit, Qiagen, Germantown, Md.
DNeasy Blood and Tissue Kit, Qiagen, Germantown, Md.
Novaseq 6000 (PE250), Illumina Inc, San Diego, Calif.
Mothur. Available at: mothur.org/wiki. Accessed Apr 14, 2020.
MUSCLE, version 3.8.31. Available at: www.drive5.com/muscle. Accessed Apr 14, 2020.
Prism, version 8.3.0, Graph Pad Software Inc, La Jolla, Calif.
R Studio, version 2.15.3, R Foundation for Statistical Computing, R Core Team, Vienna, Austria.
Vegan package, version 2.1-43, R Foundation for Statistical Computing, R Core Team, Vienna, Austria.
References
- 1. ↑
Suchodolski JS. Diagnosis and interpretation of intestinal dysbiosis in dogs and cats. Vet J 2016;215:30–37.
- 2. ↑
Bassis CM, Moore NM, Lolans K, et al. Comparison of stool versus rectal swab samples and storage conditions on bacterial community profiles. BMC Microbiol 2017;17:78.
- 3. ↑
Budding AE, Grasman ME, Eck A, et al. Rectal swabs for analysis of the intestinal microbiota. PLoS One 2014;9:e101344.
- 4. ↑
Jones RB, Zhu X, Moan E, et al. Inter-niche and inter-individual variation in gut bacterial community assessment using stool, rectal swab, and mucosal samples. Sci Rep 2018;8:4139.
- 5. ↑
Caporaso JG, Kuczynski J, Stombaugh J, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods 2010;7:335–336.
- 6. ↑
Edgar RC, Haas BJ, Clemente JC, et al. UCHIME improves sensitivity and speed of chimera detection. Bioinformatics 2011;27:2194–2200.
- 7.
Haas BJ, Gevers D, Earl AM, et al. Chimeric 16S rRNA sequence formation and detection in Sanger and 454-pyrosequenced PCR amplicons. Genome Res 2011;21:494–504.
- 8. ↑
Edgar RC. UPARSE: highly accurate OTU sequences from bacterial amplicon reads. Nat Methods 2013;10:996–998.
- 9. ↑
Schloss PD, Westcott SL, Ryabin T, et al. Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl Environ Microbiol 2009;75:7537–7541.
- 10. ↑
Wang Q, Garrity GM, Tiedje JM, et al. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol 2007;73:5261–5267.
- 11. ↑
Quast C, Pruesse E, Yilmaz P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res 2013;41:D590–D596.
- 12. ↑
Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 2004;32:1792–1797.
- 13. ↑
Biehl LM, Garzetti D, Farowski F, et al. Usability of rectal swabs for bacteriome sampling in a cohort study of hematological and oncological patients. PLoS One 2019;14:e0215428.
- 14. ↑
Fair K, Dunlap DG, Fitch A, et al. Rectal swabs from critically ill patients provide discordant representations of the gut bacteriome compared to stool samples. MSphere 2019;4:e00358-e19.
- 15. ↑
Durbán A, Abellan JJ, Jimenez-Hernandez N, et al. Assessing gut bacterial diversity from feces and rectal mucosa. Microb Ecol 2011;61:123–133.
- 16. ↑
Zoetendal EG, von Wright A, Vilpponen-Salmela T, et al. Mucosa-associated bacteria in the human gastrointestinal tract are uniformly distributed along the colon and differ from the community recovered from feces. Appl Environ Microbiol 2002;68:3401–3407.
