• Open Access

Upregulation of Toll-Like Receptors in Chronic Enteropathies in Dogs


  • The work was performed at the Small Animal Hospital of the University of Bern, Bern, Switzerland. Part of this work was presented at the 25th ACVIM Congress in Seattle, WA, USA, in June 2007.

Corresponding author: Iwan A. Burgener, med.vet., Dr.med.vet., DACVIM and DECVIM-CA, Department of Clinical Veterinary Medicine, Division of Small Animal Internal Medicine, Vetsuisse Faculty of the University of Bern, Länggassstrasse 128, 3001 Bern, Switzerland; e-mail: iwan.burgener@kkh.unibe.ch.


Background: Inflammatory bowel disease (IBD) is thought to result from a dysregulated interaction between the host immune system and commensal microflora. Toll-like receptors (TLRs) recognize microbe-associated molecular patterns (MAMPs), but their role in enteropathies in dogs is unknown.

Hypothesis: That there is a dysregulation of TLRs recognizing bacterial MAMPs in dogs with IBD.

Animals: Sixteen healthy beagles and 12 dogs with steroid-treated (ST) and 23 dogs with food-responsive (FR) diarrhea.

Methods: Prospective, observational study. mRNA expression of canine TLR2, 4, and 9 was evaluated by quantitative real-time RT-PCR in duodenal and colonic biopsies obtained before and after standard therapy. Samples from control dogs were taken at necropsy, with additional biopsies of stomach, jejunum, ileum, and mesenteric lymph node in 6 dogs.

Results: There were significant differences (P≤ .017) in expression of TLR2, 4, and 9 between the 6 sampled locations in healthy control dogs (lymph node > small intestine ≥ colon). Before therapy, ST expressed more mRNA than control dogs for all 3 receptors (P < .05). There were no significant differences between pretreatment and posttreatment values, even though 32/35 dogs improved clinically. No associations were found when comparing receptor mRNA expression with either histology or clinical activity scores.

Conclusions and Clinical Importance: Bacteria-responsive TLR2, 4, and 9 are upregulated in duodenal and colonic mucosa in IBD. This might lead to increased inflammation through interaction with the commensal flora. The absence of significant changes after therapy despite clinical improvement might point toward the existence of a genetic predisposition to IBD as described in human IBD.

Normal intestinal epithelium is not inflamed despite close contact with a high density of commensal organisms that would elicit inflammation when entering the body by other routes.1 Intestinal epithelial cells (IEC) serve as a barrier between the body and viruses, bacteria, and parasites present in the intestinal lumen.2 Rather than being a passive barrier, the intestinal epithelium is an active participant in the mucosal immune response through antigen processing and presentation, secretion of cytokines, and recruitment of inflammatory cells in response to pathogens and their products.3 The gastrointestinal-associated lymphoid tissue is the largest and most complex immunologic organ of the body and must be capable of mounting protective immune responses to pathogens while maintaining tolerance to harmless environmental antigens such as commensal bacteria and food. The breakdown of this tolerance is a key factor in the development of chronic intestinal inflammation.4

Cells of the vertebrate body have receptors by which they sense microbe-associated molecular patterns (MAMPs). These so-called pattern recognition receptors (PRR) convert the recognition process into a meaningful host defense reaction. A family of PRR are Toll-like receptors (TLRs), which are transmembrane receptors homologous to the Drosophila Toll protein.5 These receptors recognize MAMPs present on diverse microbes, including Gram-positive and Gram-negative bacteria, fungi, viruses, and parasites. The TLRs recognizing, among others, bacterial MAMPs, are TLR2 (lipopeptides, peptidoglycan, and lipoteichoic acid), TLR4 (lipopolysaccharide), TLR5 (flagellin), and TLR9 (bacterial and viral unmethylated CpG oligonucleotides). TLRs are expressed by cells belonging to the innate immune system, such as macrophages and dendritic cells, but also on IEC.6 Activation of TLRs leads to the production of proinflammatory cytokines, upregulation of costimulatory molecules, and synthesis of reactive oxygen and nitrogen intermediates.7

Inflammatory bowel disease (IBD) in humans is believed to result from predisposing genetic factors and environmental stimuli such as MAMPs being recognized by the host immune system and causing inflammation of the gastrointestinal mucosa.8 Recently, dysregulated expression of TLRs9 and TLR polymorphisms10 have been implicated in the pathogenesis of IBD. However, TLRs are involved not only in the host defense reaction and the etiology of IBD but also in the maintenance of intestinal homeostasis.1,11 In dogs with IBD, the breakdown of immunologic tolerance to luminal antigens (bacteria and dietary components) is thought to be critical, perhaps resulting from disruption of the mucosal barrier, dysregulation of the immune system, or disturbances in the intestinal microflora.12 Until recently, histiocytic ulcerative colitis was also considered a special form of IBD, but an adherent and invasive phenotype of E. coli was discovered as the etiologic agent of the disease13 explaining the success of enrofloxacin therapy.14 The expression of TLR2 and TLR4 mRNA has been determined in a primary canine colonic epithelial culture,15 but the role of TLRs in chronic enteropathies in dogs is unknown.

The purpose of this study was to determine the mRNA expression of canine TLR2, 4, and 9 in sections of healthy intestine and to test the hypothesis that there is a dysregulation of TLR2, 4, and 9 in the intestine of dogs with IBD.

Materials and Methods

Healthy Control Dogs (HCD)

Owing to limited access to HCD, control biopsies were taken at necropsy from 16 healthy Beagle dogs, which served as placebo controls in pharmacologic studies. This group included 8 males and 8 females (all intact), 10–16 months old (median 13 months), with body weights from 5.4 to 11.2 kg (median 8.1 kg). These dogs did not receive any drugs, were clinically healthy with no signs of diarrhea or vomiting, and revealed no abnormalities in CBC, serum biochemical profile, and urinalysis. Furthermore, parasitic and bacterial analysis of fecal samples was performed and the samples were free of Giardia sp., Salmonella sp., and Campylobacter sp. Within 5 minutes after euthanasia, 6 duodenal and colonic samples were taken from all 16 HCD with an endoscopic biopsy forceps and stored in RNAlater.a Additional biopsies from stomach, jejunum, ileum, and mesenteric lymph node were taken in 6 dogs and stored in RNAlater at −80 °C until analysis. As an inclusion criterion, all 16 HCD had to be free of histologic abnormalities.

Dogs with Chronic Enteropathies

Thirty-five dogs with signs of chronic gastrointestinal disease referred to the Small Animal Teaching Hospital between November 2003 and February 2006 were included in this prospective study. Some of the dogs described here were used in another study, and information about these dogs and the study protocol is available.16 Selection criteria included a history of chronic diarrhea with or without vomiting that lasted for at least 6 weeks, exclusion of identifiable underlying disorders, and histopathologic evidence of intestinal inflammatory cellular infiltrate. Owners of dogs signed a letter of consent in which they agreed to participate in initial and follow-up diagnostic evaluation, including endoscopic exams before and after therapy. None of the dogs had been treated with antibiotics, corticosteroids, or antacids in the 2 weeks before entering the study. All experimental procedures were approved by the Cantonal Committee for Animal Experimentation, Bern, Switzerland.

All dogs were given a clinical score using the canine IBD activity index (CIBDAI) as established by Jergens et al17 and were classified as clinically insignificant (score 0–3), mild (4–5), moderate (6–8), or severe (9–18) before and after therapy. Furthermore, the dogs were classified according to their predominant clinical signs as having upper or lower gastrointestinal disease or both. Duodenoscopy and colonoscopy were performed in all dogs except those with severe hypoalbuminemia (n = 3), where a 36-hour fasting period required for colonoscopy was considered to be detrimental.

After endoscopy, all dogs were treated initially with an elimination dietb for 10 days. Although recommendations usually state that dogs should be fed an appropriate formulation for at least 4–6 weeks,18 we limited the elimination diet trial to 10 days for owner-compliance reasons. Dogs that responded to the elimination diet in the first 10 days (clinical signs improved or resolved) were assigned to the food-responsive (FR) group. Although it is possible that these dogs still had IBD, we assigned them into the FR-group according to their prompt response to dietary treatment alone. The dogs that did not respond in the first 10 days of treatment (clinical signs persisted while on the elimination diet) were assigned to the steroid-treated (ST) group and were given oral prednisolone (1 mg/kg PO q12h) for 10 days followed by a tapering dosage. The FR group was reevaluated, including CIBDAI and endoscopy, 4 weeks after starting treatment with the elimination diet. The ST group was re-examined, including CIBDAI and endoscopy, at the end of the 10-week treatment period, which was 2 weeks after discontinuation of prednisolone treatment. All dogs were fed exclusively the elimination diet.


Dogs were prepared for endoscopy by withholding food for 36 hours and administering a colonic lavage solutionc by gastric intubation (2 doses of 30 mL/kg of body weight 6–8 hours apart). Six mucosal biopsy specimens were obtained from the stomach, duodenum (∼10 cm below the caudal duodenal flexure), and middle portion of the descending colon, or from where lesions were visible. An endoscopic score19 was assigned based on mucosal appearance and on the severity of changes. Samples for subsequent histopathologic evaluation were placed in 4% neutral-buffered formalin for 48 hours before embedding in paraffin. Furthermore, 6 endoscopic biopsies from duodenum and colon were immediately put into RNAlatera and stored at −80 °C until RNA isolation.


A minimum of 5 biopsy specimens from each site were examined histopathologically. Blinded qualitative evaluation of the degree of inflammation and overall cellular infiltrate was performed by an ACVP board-certified pathologist, who assigned a grade (normal = 0, mild = 1, moderate = 2, and severe = 3) based on previously published guidelines.19 Mild lesions (=grade 1) were those with cellular infiltrates but without architectural distortion or mucosal epithelial immaturity. Moderate lesions (=grade 2) had cellular infiltrates accompanied by mucosal epithelial immaturity, solitary epithelial necrosis, or both. Severe lesions (=grade 3) consisted of cellular infiltrates accompanied by multifocal epithelial necrosis or extensive architectural distortion with epithelial immaturity. The pathologist was blinded in regard to number of endoscopy, clinical diagnosis, and treatments used and did the analysis in 1 sitting.

RNA Isolation

Three endoscopic biopsies (total tissue mass 20–30 mg) per spin column were mechanically disrupted and homogenized with a TissueLyserd as described by the manufacturer. Total RNA was isolated from the resulting lysate with an RNeasy Mini Kita as described by the manufacturer, including on-column RNase-free DNasea digestion during RNA purification. RNA concentration of the samples was measured at 260 nm (in ng/μL) with a NanoDrop ND-1000e spectrophotometer. The ratio of the sample at 260/280 nm was used to assess the purity of the RNA. All samples had a ratio >2.0 and were stored at −80 °C until further measurements.

Primer and Probe Design

Primers and probes were TaqMan gene expression products designed with Assays-by-design File Builder Software 2.0f using the dog-specific GenBank sequences (http://www.ncbi.nlm.nih.gov/Genbank/GenbankSearch.html) for TLR2 (AB189639),20 TLR4 (AB080363),21 TLR9 (AB104899),22β-actin (AF484115), and glyceraldehyde phosphate dehydrogenase (GAPDH; AB038240). All probes included a FAM fluorophore reporter (5′) and a nonfluorescent quencher bound to a minor grove binder (MGB; 3′). Probes with conjugated MGB groups form extremely stable duplexes with single-stranded DNA targets, allowing shorter probes to be used for hybridization-based assays.23 The gene expression products were delivered as premixed primers and TaqMan probe sets. The primer and probe sequences are summarized in Table 1.

Table 1.   Primer and TaqMan probe sequences used by real-time RT-PCR.
PrimerPrimer Sequence (5′–3′)Probe Sequence (5′–3′)
  1. All probes include a FAM reporter (5′) and a nonfluorescent quencher (3) bound to a minor grove binder. CAN, canine; F, forward; R, reverse.


Real-Time Polymerase Chain Reaction (qRT-PCR)

The RNA was thawed and diluted with nuclease-free diethyl pyrocarbonate (DEPC)-treated water to a concentration of 200 ng/μL used for PCR. Master mix and reverse transcriptase (RT) enzyme were used as suggested and provided in a commercially available TaqMan One-Step RT-PCR Master Mix Reagents kit.f The Multiscribe RT enzyme included in the kit is already premixed with RNAse inhibitor as a 40 × solution. The one-step master mix is a separate 2 × solution in the kit, which contains AmpliTaq Gold hot-start DNA polymerase, dNTPs with dUTP, ROX as a passive internal reference, and a PCR product carryover correction component. PCR strip tubes were purchased from Stratagene.g Each of the 25-μL one-step real-time RT-PCR reactions contained 12.5 μL one-step Master Mix, 0.625 μL Multiscribe RT enzyme, 1.25 μL mixture of forward primer (900 nM), reverse primer (900 nM), and fluorogenic probe (250 nM), 5.625 μL nuclease-free DEPC water, and 5 μL of RNA with a final amount of 1 μg.

Real-time RT-PCR and subsequent data analysis were performed using the Mx4000 Multiplex Quantitative PCR Systemg equipped with Version 4.20 software. The thermocycling conditions for fluorogenic one-step RT-PCR were 30 minutes at 50 °C, 10 minutes at 95 °C, and 50 cycles of 30 seconds at 95 °C and 1 minute at 60 °C during which the fluorescence data were collected. All primer and probe sets used for targets and endogenous controls yielded approximately 100% efficient reactions. Controls were performed without RT and by the addition of nuclease-free water instead of RNA. None of the RNA samples showed evidence of amplifiable genomic DNA with this assay. Duplicate RT reactions were performed for each sample, and a mean threshold cycle (Ct) value was calculated. Canine GAPDH and β-actin were used as housekeeping genes for normalization. Threshold cycles for the housekeeping genes were calculated, and ΔCt (Ct probe—Ct housekeeping gene) was used for analysis. A lower ΔCt value indicates higher mRNA expression. Therefore, to simplify the understanding of the figures, the results were expressed as 1/ΔCt.

Statistical Analysis

All statistical analyses were performed with NCSS 2004h (http://www.ncss.com). ΔCt was used as the basis for comparisons. The results were assessed for normality by the Shapiro-Wilk normality test and subsequently described and plotted using means and standard deviations. Measurements were compared between disease groups and over regions (within dogs) by a repeated-measures ANOVA with Tukey-Kramer multiple-comparison tests. To account for possible inhomogeneity between group variances, the Geisser-Greenhouse corrected P-values of the repeated-measures ANOVA procedure are presented. A paired T-test was used to compare measurements within groups and regions before and after treatment. Within the colon and duodenum, the association between measured values and CIBDAI or histology scores was tested by a one-way ANOVA. CIBDAI or histology scores were entered as factor variables with 4 or fewer classes, whereas TLR expression was considered as the continuously measured outcome. Separate analyses were run for each region (duodenum, colon) and before and after therapy. The overall α level of statistical significance was set at 0.05.


Dogs with Chronic Enteropathies

The group of dogs classified as FR (n = 23) included 15 males and 8 females, of which 2 and 3, respectively, were neutered. The dogs were 7 to 77 months old (median 24 months), and 15 mostly large breeds were represented with 1 or 2 individuals (5 mixed breed dogs). Body weights in the FR group ranged from 7.5 to 61 kg (median 27 kg). Two dogs were presented with signs of upper and 4 dogs with lower gastrointestinal tract problems only, whereas the other 17 showed mixed signs.

The group of dogs classified as ST (n = 12) included 7 males and 5 females, of which 3 and 5 were neutered, respectively. The dogs were 9 to 117 months old (median 87 months). Breeds in this group included Dachshund (2), Yorkshire Terrier (2), mixed breed (2), and 1 of each belonging to Bull Mastiff, Boxer, Shar Pei, Rottweiler, German Shepherd Dog, and West Highland White Terrier breed. Body weights in the ST group ranged from 2.9 to 71.5 kg (median 17.55 kg). Two dogs were presented with signs of upper gastrointestinal tract problems only, whereas the other 10 showed mixed clinical signs.

Canine IBD Activity Index

In the FR group before treatment (n = 23), 2 dogs were classified as insignificant, 4 as mild, 15 as moderate, and 2 as severe. After therapy, 22 dogs were classified as insignificant and 1 as mild. In the ST group before treatment (n = 12), 1 was classified as insignificant, 5 as moderate, and 6 as severe. After therapy, 4 were classified as insignificant, 3 as mild, 3 as moderate, and 2 as severe. Compared with FR, in which all 23 dogs got clinically better, 1 ST dog got worse despite therapy (from moderate to severe) and 2 dogs were stable under therapy (1 moderate, 1 severe). These 3 dogs were weaned from the prednisolone treatment and received cyclosporinei at 5 mg/kg PO q24h for 8 weeks, whereupon all 3 improved (2 classified as insignificant at the end of treatment, 1 as moderate).

Endoscopy Score

There was no difference in endoscopy score between the FR and ST groups nor within the groups before and after treatment (results not shown).


In 17 of 23 dogs with FR as well as 6 of 9 dogs with ST (3 PLE duodenum only), the histopathologic changes were more severe in the duodenum compared with the colon. In the group of dogs classified as FR (n = 23), 21/23 had histopathologic abnormalities in the duodenum (91%) and 21/23 in the colon (91%). Although all dogs improved clinically after therapy, the histopathologic scoring did not change in 11 dogs; 8 dogs improved and 4 dogs worsened. In the group of dogs classified as ST (n = 12), all had histopathologic abnormalities in the duodenum (12/12; 100%) and 8/9 in the colon (89%). Although all but 3 dogs were improved clinically after therapy, the histopathologic scoring did not change in 7 dogs, whereas 3 dogs improved and 2 got worse. Three dogs with hypoalbuminemia (both Yorkshire Terriers and the West Highland White Terrier) and clinical signs of protein-losing enteropathy were included in the ST group because of moderate to severe lymphoplasmacytic duodenitis and lymphangiectasia revealed by the histopathologic examination. The histopathologic results are summarized in Table 2.

Table 2.   Summary of histopathologic scores from 23 dogs with FR and 12 dogs with ST before and after therapy.
Duod. FR
FR After
Colon FR
Colon FR
Duod. ST
ST After
Colon ST
ST After
  1. The numbers in each cell represent the number of dogs. Duodenal biopsies were taken and scored for all FR (n = 23) and ST dogs (n = 12). Colonic biopsies were taken from all dogs except 3 ST dogs with severe protein-losing enteropathy.

  2. FR, food-responsive diarrhea; ST, steroid-treated diarrhea; Duod., duodenum; LP, lymphoplasmacellular infiltrate; Eos, eosinophilic infiltrate.

0No infiltrates2223  1 
1LP43913  33
LP + Eos1211 121
Fibrosis  2     
LP + Eos751145  
Eos31 11  1
Fibrosis  1     
3LP  1     
LP + Eos 2  1112
Eos    1111
Total 23232323121299

Comparison of TLR Expression between Different Regions in HCD

Canine β-actin had significantly lower expression in the stomach when compared with small and large intestine and mesenteric lymph node, which disqualified β-actin as an endogenous control. Besides the stomach, all results for β-actin were almost identical to GAPDH in regard to significant results. GAPDH was used as endogenous control for further calculations.

For all 3 receptors tested, there were significant differences between the 6 regions in the 6 HCD (TLR2: P= .017; TLR4: P= .001; TLR9: P= .008). The mesenteric lymph node expressed more mRNA than all other regions for all 3 receptors measured. Transcription of TLR2 and 4 was significantly lower in the colon when compared with lymph node, stomach, duodenum, and ileum. The ileum expressed significantly more mRNA for TLR9 than did stomach, jejunum, and colon, whereas there were no differences between the other regions (Fig 1). These results were confirmed when comparing the mRNA expression between duodenum and colon within all 16 HCD, in which TLR2 and 4 were significantly more expressed in the duodenum (P < .001 for both), whereas there was no significant difference in regard to TLR9 (P= .113).

Figure 1.

 Toll-like receptor (TLR) mRNA expression in different regions in 6 healthy control dogs. The results are expressed in means of 1 /ΔCt± standard deviation. Ct, threshold cycle; ΔCt=Ct gene of interestCt GAPDH; S, stomach; D, duodenum; J, jejunum; I, ileum; C, colon; L, mesenteric lymph node.

Comparison of TLR Expression between Groups

When comparing the 3 groups before therapy, ST had the highest and HCD the lowest amount of mRNA for all 3 receptors in duodenum and colon. When comparing the different groups, all 3 TLRs had significantly higher expression in ST than in HCD (TLR2: P= .046; TLR4: P= .006; TLR9: P= .033), but there were no significant difference between ST and FR or FR and controls. When comparing regions, the differences between duodenum and colon were significant for TLR4 (P= .003), but not TLR2 (P= .060) and TLR9 (P= .990). Furthermore, there were significant interactions between group and region for TLR2 and 4 (P= .007 and P < .001), but not for TLR9 (P= .576) (Fig 2).

Figure 2.

 Comparison of Toll-like receptor (TLR) mRNA expression between healthy control dogs (HCD, n = 16) and dogs with food-responsive diarrhea (FR, n = 23) or steroid-treated diarrhea (ST, n = 12) in duodenum (D) and colon (C) before therapy. The results are expressed in 1/ΔCt. P-values from a repeated measures ANOVA with Geisser-Greenhouse correction (for inhomogeneity between group variances) and Tukey-Kramer multiple comparison test. Group is defined as HCD, FR, or ST, whereas region is defined as duodenum or colon. Open circles represent mild outliers, whereas full circles represent severe outliers. Ct, threshold cycle; ΔCt=Ct gene of interestCt GAPDH.

Comparison of TLR Expression before and after Therapy within Groups

There were no significant differences in mRNA expression for all 3 receptors before and after therapy in duodenum and colon of ST and FR.

Comparison of TLR Expression with CIBDAI and Histology

There were no significant differences in ST or FR when comparing expression of the 3 different TLRs in duodenum or colon before or after therapy with the CIBDAI classes (insignificant, mild, moderate, severe). The same was true when comparing it with the histopathologic grading (normal or changes grade 1, 2, or 3).


In this study, TLRs responsive to bacterial MAMPs were upregulated at the level of mRNA in duodenal and colonic mucosa of dogs with ST. If this upregulation exists also at the protein level, the interaction of the commensal flora with TLRs could lead to increased inflammation as in human IBD, where TLR4 (but not TLR2) is strongly upregulated in both active ulcerative colitis and Crohn's disease in IECs.9

Previously, canine TLR2, 4, and 9 were sequenced and moderate expression of TLR2 mRNA detected in small and large intestine20 and TLR4 in small intestine,21 but no TLR922 with conventional PCR. In our study, mRNA coding for all 3 TLRs tested was detected in duodenum and colon, which can be explained by the increased accuracy and sensitivity of real-time RT-PCR compared with conventional PCR.24 TLR2 and 4 expression was higher in the duodenum than in the colon, whereas for TLR9, there was no significant difference. Consistent with the previous studies, TLR9 expression appeared lower than TLR2 and 4 expression (Ct values mostly 1–3 cycles higher, ie, ∼2–8 times less mRNA). Little is known about TLRs in canine intestine with the exception that TLR2 and TLR4 are expressed at low levels in nonstimulated canine primary colonic epithelial cells and can be upregulated in response to challenge with their respective agonists peptidoglycan or lipopolysaccharide.15

The decreased expression of TLR2 and 4 mRNA in the colon of healthy dogs compared with the duodenum is meaningful given the fact that the bacterial density rises from proximal to distal in the bowel, with the greatest increase across the ileocecal valve. The diversity of the flora is highest in the colon,25 where a single layer of epithelial cells is lining the lumen with the impressive amount of 1012 bacteria per gram of intestinal content.26 Interestingly, this difference between duodenum and colon almost disappears in diseased dogs, which might be attributable to the high percentage of colonic inflammation (∼90% in FR and ST) as well as the higher amount of bacteria that can lead to an upregulation in the colon. When discussing the missing difference in expression for TLR9 between duodenum and colon, one has to consider that TLR2 and 4 appear to be expressed on the apical epithelial surface in direct contact with the commensal flora,2 whereas TLR9 generally seems to be intracellular without direct contact to bacteria.27 When upregulated, TLR9 can also be found on the surface, where it can lead to inflammation or tolerance in mice.11

Tolerance in healthy individuals mostly means maintaining hyporesponsiveness to harmless luminal commensals.28,29 When TLRs on the surface of IEC are upregulated, frontline recognition is increased and leads to inflammation. Whether this is a primary event rather than a sequel to inflammation caused by other factors is difficult to distinguish. Because of the extensive work-up of these cases, most well-known causes of inflammation are already excluded. Furthermore, the TLRs tested herein are mostly responsive to bacterial (TLR2, 4, and 9) or viral (TLR9) products and are, to the best of our knowledge, not nonspecifically upregulated. The fact that the expression of the different TLRs did not change significantly after 4 (FR) to 10 (ST) weeks of treatment despite clinical improvement also points toward a primary event rather than a sequel to (unspecific) inflammation.

The upregulation of all 3 receptors in ST compared with HCD could explain the sustained inflammation in canine IBD. The absence of significant changes after therapy despite clinical improvement on the other hand might point toward genetic predisposition as known from humans, where IBD is believed to result from predisposing genetic factors and environmental triggers acting on the immunoregulatory system.8 Nevertheless, another possible explanation for the missing difference before and after therapy could be the low case number in our study. This could cause low power with only mild changes, even though a very small difference before and after therapy despite obvious clinical improvement would be of questionable clinical impact.

Owing to limited access to HCD, biopsies from healthy Beagles serving as placebo controls in pharmacological studies were taken at necropsy. These dogs were living in different conditions than the dogs with diarrhea, and the bacterial flora might be different compared with the cases suffering from chronic enteropathies. Moreover, these control dogs probably did not have contact with as many stimuli than the clinical cases. This may be 1 reason why the difference between healthy controls and dogs with chronic enteropathies is so obvious. Nevertheless, Beagles are known to have a higher percentage of small intestinal bacterial overgrowth and increased intestinal permeability together with no or only minimal histologic changes.30,31 Furthermore, increased interferon γ and interleukin 10 expression were demonstrated by in situ hybridization when comparing Beagles with Irish Setters.32 These factors may all influence the expression of TLRs, but an increase of (commensal) bacteria and especially increased permeability would rather lead to an upregulation of TLRs in Beagles to avoid bacterial translocation. Therefore, the significant changes found between Beagles and ST could even be more pronounced between healthy non-Beagles and ST.

Another shortcoming of the study is the fact that the histology of the control dogs was not read by the same pathologist as the clinical cases. The control dogs were placebo controls in pharmacologic studies at Novartis and therefore interpreted by a Novartis pathologist (J.B.). Nevertheless, the most important comparisons were between the clinical cases and within every case before and after therapy. These analyses have all been performed by the same pathologist (A. Gröne) in a blinded fashion.

Last but not least, no significant differences were found in ST or in FR when comparing the expression of the 3 different TLRs in duodenum or colon before or after therapy with the CIBDAI classes or with the histopathologic grading. Up to now and with the scoring systems used, histology has rarely proven to be a very helpful tool to relate it to the clinical status in canine IBD. Substantial intraobserver variation in histopathologic evaluations33 can be excluded in this study, where histology of all clinical cases was interpreted by the same pathologist. Nevertheless, several reports revealed the unreliability of histopathologic lesions in regard to clinical status and response to treatment.34,j Another possible explanation is the relatively low case number in our study. Especially when splitting the cases into 4 subgroups for CIBDAI (insignificant, mild, moderate, severe) and histology (grades 0–3), there were subgroups with very low numbers, which made adequate statistics impossible. Taking these problems into considerations, we tried to regroup CIBDAI and histology subgroups in <4 groups to avoid very small numbers, but also in this case, there were no significant differences.

In conclusion, expression of bacteria-responsive TLRs is increased in duodenal and colonic mucosa in ST, putatively leading to increased inflammation through interaction with the commensal flora. It is currently not clear whether this is caused by an upregulation of TLRs in IEC or recruitment of inflammatory cells expressing TLRs, which warrants further investigations. The absence of significant changes after therapy despite clinical improvement might point toward genetic predisposition, as known from human IBD.


aQiagen AG, Basel, Switzerland

bPurina Veterinary Diets LA, Société des Produits Nestlé SA, Vevey, Switzerland

cEach liter contains 60 g polyethylene glycol (PEG) 4000, 1.46 g sodium chloride, 0.745 g potassium chloride, 1.68 g sodium bicarbonate, and 5.68 g sodium sulfate

dTissueLyser, Retsch GmbH & Co KG, Haan, Germany

eNanoDrop Technologies Inc, Wilmington, DE

fApplied Biosystems, Applera International Inc, Nieuwerkerk aan den Ijssel, the Netherlands

gStratagene Europe, Amsterdam Zuidoost, the Netherlands

hNumber Cruncher Statistical Systems (NCSS), Kaysville, UT

iAtopica, Novartis Animal Health, Basel, Switzerland

jSchreiner N et al. No changes in histological scoring, total number of infiltrating cells and number of T cells after treatment in dogs with chronic enteropathies. 15th ECVIM-CA Congress, Glasgow, Scotland, 2005 (abstract)


The authors would like to thank Kay-Sara Sauter for her technical support with qRT-PCR and Andrea Gröne for reading the histology. This work was supported by Research Grants from the Canine Health Foundation of the American Kennel Club (Grant No. 803) and from the Research Foundation of the Vetsuisse Faculty of the University of Bern.