Fish-meal diet enriched with omega-3 PUFA and treatment of canine chronic enteropathies

Authors

  • Edgar Corneille Ontsouka,

    Corresponding author
    1. Faculty of Medicine, Institute of Biochemistry and Molecular Medicine, University of Bern, Bern, Switzerland
    • Faculty of Medicine, Institute of Biochemistry and Molecular Medicine, University of Bern, Buehlstrasse 28, 3012 Bern, Switzerland Fax: +41-31-6313737.
    Search for more papers by this author
  • Iwan Anton Burgener,

    1. Vetsuisse Faculty, Department of Clinical Veterinary Medicine, University of Bern, Bern, Switzerland
    2. Current address: Veterinary Faculty, University of Leipzig, D-04103 Leipzig, Germany
    Search for more papers by this author
  • Nicole Luckschander-Zeller,

    1. Vetsuisse Faculty, Department of Clinical Veterinary Medicine, University of Bern, Bern, Switzerland
    2. Current address: Clinic for Internal Medicine, Department for Companion Animals and Horses, University of Veterinary Medicine, A-1210 Vienna, Austria
    Search for more papers by this author
  • Jürg Walter Blum,

    1. Vetsuisse Faculty, Veterinary Physiology, University of Bern, Bern, Switzerland
    Search for more papers by this author
  • Christiane Albrecht

    1. Faculty of Medicine, Institute of Biochemistry and Molecular Medicine, University of Bern, Bern, Switzerland
    Search for more papers by this author

Abstract

The effects of a fish-meal- and potato-protein-based diet enriched with omega-3 PUFA on intestinal inflammatory activity and expression of genes active in fatty acid (FA) uptake were tested in the duodenum of dogs with food responsive diarrhea (FRD; n = 14) and inflammatory bowel disease (IBD; n = 7). The clinical outcome was estimated by monitoring the canine IBD activity index (CIBDAI) before and after treatment. Dogs were treated with the diet alone (FRD) or the diet in combination with immunosuppressants (IBD). The duodenal mRNA levels of FA translocase, FA transport protein-1,-3,-4,-6, long chain acyl coenzyme synthetase-1,-4,-5,-6, liver- and intestinal-FA binding proteins, interleukin-1β (IL)-1β and -6, and tumor necrosis factor-α were measured by quantitative PCR. The CIBDAI significantly decreased after treatment in all dogs. The mRNA levels of target genes were associated both with disease phenotype and dietary treatment. Significantly different expression patterns were found for IL-1β, FA translocase, intestinal-FA binding protein, FA transport proteins-1,-3,-6, and long chain acyl coenzyme synthetase-5,-6. In conclusion, the mRNA levels of some of the genes involved in duodenal FA uptake may be altered by a fish-meal- and potato-protein-based diet enriched with omega-3 PUFA. This may be beneficial for the treatment of canine chronic enteropathies, particularly FRD.

Practical applications: In this study, feeding dogs on a fish-meal- and potato-protein-based diet enriched with omega-3 PUFA resulted in marked suppression of intestinal inflammatory activity, mainly in the duodenum of dogs with food responsive diarrhea with a concomitant alteration of some of the genes involved in FA uptake. In IBD, however, a combination of diet and immunosuppressive drugs was required. The present study provides preliminary insights into the importance of the herein tested dietary composition for the treatment of canine chronic enteropathies, revealing that an omega-3 PUFA-enriched diet can be beneficial to the animal's health and wellbeing. Furthermore, our results serve as the basis for future investigations using different food ingredients and FA compositions to identify the optimal dietary mixture that could be equally effective for dietary prophylactic or therapeutic treatment of canine chronic enteropathies, or both.

See commentary by Ferguson [p. 372–374]

Abbreviations:

Ascl, long chain acyl coenzyme synthetase; CE, chronic enteropathies; CIBDAI, canine IBD activity index; Ct, crossing point values; FA, fatty acid; FABP, fatty acid binding protein; FAT/CD36, fatty acid translocase; FATP, fatty acid transport protein; FRD, food responsive diarrhea; IBD, inflammatory bowel disease; IL, interleukin; TNF-α, tumor necrosis factor α

Introduction

Inflammatory bowel disease (IBD) and food responsive diarrhea (FRD) are the most commonly encountered forms of chronic enteropathies (CE) in dogs 1, 2, which can at present only be differentiated by their response to treatment involving an elimination diet 3, 4. The causes of these diseases are currently not well understood. An etiological role of exaggerated immune responses to normal gut microflora has been suggested for IBD 5, 6, whereas, FRD is mostly associated with allergic responses to food components 7. The canine IBD activity index (CIBDAI) is frequently used in veterinary practice to indirectly monitor the inflammatory activity during canine CE 8.

Following food intake, dietary fatty acids (FA) are absorbed in the small intestine. The mechanisms involved have not yet been fully elucidated, although both passive diffusion and active protein-mediated uptake have been suggested 9, 10. The active uptake of dietary FA involves a cascade of cellular events and interdependent mechanisms that mobilize integral membrane and membrane associated proteins such as FA translocase (FAT/CD36) 11, FA transport proteins (FATP) 12–14, long chain acyl coenzyme synthetase (Ascl) 13, and cytosolic FA binding proteins (FABP) 13, 15. Importantly, human patients with Crohn's disease and ulcerative colitis have an increased ileal and colonic mRNA expression of Ascl-1 and Ascl-4 in comparison with healthy controls 16, suggesting a possible role of these FA uptake genes in the pathogenesis of IBD.

In enterocytes, FA contribute considerably to both normal cell function and pathology by acting as activators of transcription as well as precursors of inflammatory mediators. For instance, omega-3 and omega-6 PUFA compete for prostaglandin and leukotriene synthesis at the cyclooxygenase and lipoxygenase level 17, 18. It has become evident that dietary omega-3 PUFA have anti-inflammatory effects in humans suffering from IBD 19, 20. Accordingly, it has been demonstrated that the alteration of dietary PUFA content in favor of omega-3 rather than omega-6 has the potential to modulate immune responses 17, 18. On the other hand, an increased dietary intake of fat and protein of animal origin was positively associated with a high risk for IBD in humans, while a negative association was observed between the prevalence of IBD and the intake of vegetable protein 26. In a study performed in dogs 4, those suffering from CE and fed exclusively a diet composed of fish-protein (salmon and trout), canola meal and rice, showed an improvement in clinical outcome. CIBDAI scores after treatment were still around 5, indicating the presence of mild disease according to the definition of CIBDAI 8.

To our knowledge, it has never been determined whether the mRNA abundance of genes involved in FA uptake is altered during canine CE, and if the mRNA abundance is affected by long-term exclusive feeding with a specific elimination diet containing increased amounts of omega-3 PUFA. Therefore, our objectives were to evaluate the impact of a fish-meal- and potato-protein-based diet enriched with omega-3 PUFA on both the CIBDAI and the mRNA abundance of selected genes of FA uptake in the duodenum of dogs with CE. The small intestine was chosen since it is the site in the digestive tract where dietary lipids are predominantly absorbed, and changes in the expression pattern of candidate genes are mainly expected to occur. We tested the hypotheses that a long-term and exclusive fish-meal- and potato-protein-based diet enriched with omega-3 PUFA will significantly (i) reduce the CIBDAI and (ii) alter the duodenal expression of selected genes of FA uptake and homeostasis in dogs suffering from CE. The results of the present study provide preliminary insights into the importance of FA transporters in the pathogenesis of CE in dogs. Furthermore, the data contribute to clarifying whether FA transporters, binding proteins and activators are functionally important in the treatment of canine IBD and which of these are involved, possibly through promoting the uptake of dietary FA species into enterocytes.

Materials and methods

Animals

Dogs suffering from spontaneous diarrhea lasting for at least 6 wk were referred by private veterinarians to the Small Animal Clinic of the Vetsuisse Faculty in Bern, Switzerland, for diagnostic gastro-, duodeno-, and colonoscopy. The animals were enrolled in the trial with the owners' written consent. Experimental protocols were approved by the Cantonal Committee for Animal Experimentation of the Canton of Bern, Switzerland (123/05).

Rationale for grouping

Procedures used for clinical examination and grouping of patients were as previously described 8, 22, 23. Briefly, the CIBDAI in all sick dogs was monitored by an in-house expert who was unaware of the study. This evaluation was used to estimate the severity of the disease, i.e., from clinically insignificant to clinically severe. The CIBDAI included scores for attitude and activity, appetite, vomiting, stool consistency, stool frequency, and weight loss. CIBDAI values indicated: 0–3 = healthy or clinically insignificant disease; 4–5 = mild disease; 6–8 = moderate disease; and ≥9 = severe disease. As previously described 23, after the first endoscopy all dogs initially received the elimination diet for 14 days. Animals that showed a CIBDAI ranging between 0 and 3 within 14 days of treatment with the elimination diet were grouped (retrospectively) as FRD (n = 14). They were re-evaluated with CIBDAI scoring and endoscopy analyses 4 wk after starting the elimination diet. Dogs with a CIBDAI ≥6 within 14 days were grouped as IBD and treated with prednisolone (2 mg/kg q12 h PO, cut in half every 2 wk over 8 wk) or cyclosporine (5 mg/kg q24 h PO) if the response to prednisolone was insufficient. These animals (n = 7) were re-examined with complete blood counts and chemistry profile, CIBDAI and endoscopy at the end of the 10 wk treatment period.

Breeds and body weight

Dogs with FRD had a mean (±SEM) age of 45 ± 11 months (range 9–134 months) and a body weight (BW) of 25.1 ± 4 kg (range 2–49 kg). They were of the following breeds: 1 Yorkshire Terrier, 1 French Bulldog, 1 Weimaranian, 1 West Highland White Terrier, 2 Labrador Retrievers, 1 Berger Blanc Suisse, 1 Pomeranian, 1 Newfoundlander, 1 Golden Retriever, and 4 Mongrels. Dogs with IBD had the mean (±SEM) age of 76 ± 17 months (range 36–154 months) and a BW of 21.5 ± 5 kg (range 3–37 kg). They were of the following breeds: 1 Shar Pei, 1 Papillon, 1 Golden Retriever, 1 Bauceron, 1 American Cocker Spaniel, and 2 Mongrels. Healthy dogs used as controls (n = 14) were Beagles that were free of any gastrointestinal tract disorders. They had on average an age of 108 months (range 78–154 months).

Diets

Diseased dogs were fed an elimination diet (produced by Biomill SA, Granges-Marnand; Switzerland) that contained fish-meal (not salmon-derived), potato-protein, rice, whole egg meal, beet pulp, wheat middlings, minerals, trace elements, vitamins, (Tables 1 and 2), with saturated FA (3.8%), monosaturated FA (5%), omega-6 PUFA (3.7%), and omega-3 (0.8%) PUFA. All dogs were fully grown and were, therefore, fed according to maintenance requirements.

Table 1. Ingredients of the diets
Ingredients (g/kg)Elimination diet fed to dogs with FRD and IBDStandard diet fed to healthy control dogs
  1. Values are expressed in g/kg of the diet and values in brackets represent the contribution of individual components to the total protein amounts (100%) of the diets. The ingredients making up the composition of the standard diet are mostly present in commercially available feeds fed to healthy adult dogs.

  2. NP, not present.

Chicken mealNP168.4 (64.8%)
Corn glutenNP20.1 (7.72%)
WheatNP18.0 (6.92%)
Soy beansNP17.2 (6.61%)
CornNP6.8 (2.63%)
LinseedNP6.5 (2.48%)
LiquidNP3.5 (1.35%)
Fish-meal145 (55.7%)7.0 (2.69%)
Potato-protein73.1 (28.1%)NP
Rice22.9 (8.80%)6.2 (2.36%)
Whole egg meal9.4 (3.64%)NP
Beet pulp6.8 (2.61%)2.6 (1.00%)
Wheat middlings3.0 (1.15%)3.8 (1.44%)
Table 2. Average diet composition
Variablea)Elimination diet fed to dogs with FRD and IBDStandard diet fed to healthy control dogs
  • a)

    The elimination and standard diets additionally contained per kg: vitamin A (12 000 and 13 400 UI, resp.), vitamin D3 (1300 and 1500 UI, resp.), vitamin E (0.1 g), vitamin B1 (0.01 g), vitamin B2 (0.02 g), vitamin B6 (0.08 and 0.01 g, resp.), vitamin B12 (0.1 mg), biotine (0.8 and 0.76 mg, resp.), choline (2.38 and 2.78 g, resp.), calcium (11.5 and 12.8 g, resp.), inorganic phosphorus (8 and 9.2 g, resp.), potassium (6 and 5.8 g, resp.), sodium (3.5 and 4.3 g, resp.), magnesium (1.0 and 1.1 g, resp.), iron (0.1 g), copper (15 and 10 mg, resp.), manganese (20 mg), zinc (200 mg), selenium (0.20 mg), cobalt (0.1 mg), and carnitine (0.1 g).

Dry matter (DM) (g)950910
Crude protein (g/kg) DM280285
Crude fat (g/kg) DM174115
Nitrogen-free extract (g/kg) DM395445
Crude fiber (g/kg) DM7127
Ash (g/kg) DM8177

The healthy control dogs were fed commercially available feeds for healthy adult dogs, termed standard diet in this study (for details see Tables 1 and 2). The overall composition in vitamins, minerals, and oligo-elements were similar in both diets.

Tissue sampling

Dogs were fasted for 48 h prior to endoscopy. Biopsies of ∼3 mg each were taken at two locations in the duodenum and colon 23. Tissues from gastro-intestinally healthy dogs were collected after euthanasia with pentobarbiturate and stored in stabilization buffer (RNAlater, Ambion Inc, Austin, Tex) according to the manufacturer's instructions until analysis. As lipid absorption occurs primarily at the proximal rather than the distal intestine 11, 24, colonic tissues were used as an internal control and served to calibrate the expression in the duodenum.

Feed analyses

Dry matter, crude protein, crude fat, crude fiber, crude ash, and nitrogen-free extract were analyzed at Biomill SA by the Weende methods 25. Crude protein was determined after the procedure of Kjehldal (factor: 6.25) and crude fat was determined after extraction with petroleum ether. The profile of FA was determined by GC of the methyl esters 26 by UFAG Laboratories, Sursee, Switzerland.

Real-time RT-PCR analysis

Materials and procedures for RNA extraction, quantification, and cDNA synthesis were as previously described 27. The PCR amplification of targets was performed in duplicate on a real-time PCR machine (Rotor Gene, Corbett Life Science, Mortlake, Australia Corbet) in a final volume of 10 mL, the PCR mixture containing 10 ng cDNA, 5 mmol/L of gene-specific forward and reverse primers, and the PCR Master mix (Sensimix NoRef PCR MasterMix, Quantace, London, UK) as previously described 32. Primers used for the amplification of canine FAT/CD36, FATP-1, -3, -4, -6, L-FABP, I-FABP, Ascl-1, -4, -5, -6, interleukin (IL)-1β, and of cyclophilin B were designed using specific canine sequences at http://www.ensembl.org/ (see Table 3). Primers for the determination of ubiquitin and GAPDH mRNA levels were previously published 28.

Table 3. Details on forward (for) and reverse (rev) primer sequences used for real-time PCR amplification of fatty acid uptake genes
Genea)Sequence, 5′–3′Position/sizeExon spanningGene ID
  • a)

    FAT/CD36, fatty acid translocase; FATP-1, 2, 3, 4, 6, fatty acid transport protein-1, 2, 3, 4, 6; I-FABP, intestinal fatty acid binding protein (fabp2); L-FABP, liver fatty acid binding protein (fabp1); Ascl, long chain acyl coenzyme synthetase; IL-1β, interleukin-1beta; nt, nucleotide; bp, base pairs. The genes' IDs are available at http://www.ensembl.org.

FAT/CD36
 forcatattggtcaagccagcaant1170–137310–12ENSCAFG00000006401
 revgcaacaaacatcaccacacc204 bp  
FATP-1
 fortgactgcctacccctgtaccnt858–10205–7ENSCAFT00000024320
 revgatctccccgatgtactgga163 bp  
FATP-2
 foragccaattttgggatgactgnt883–10554–5ENSCAFT00000023903
 revctcctgatgaattccctcca173 bp  
FATP-3
 forgggaaagctgctgaagaant1329–152110–11ENSCAFT00000027642
 revcacctcctggagaaaatcca193 bp  
FATP-4
 forcggttctgggacgactgtatnt955–11356–7ENSCAFT00000031925
 revggatgtggaaacggctagaa181 bp  
FATP-6
 fortggcagttggaaatggtgtant1001–12385–6ENSCAFT00000001116
 revcaccagccctgttcatttct238 bp  
I-FABP
 foraccgtcaaggaatcaagcant145–3002–3ENSCAFG00000012462
 revtccattgtctacccgtttg156 bp  
L-FABP
 foraaatcgtgcagaatgggaagnt166–3562–3ENSCAFG00000007413
 revttcggtcacagacttgatgc191 bp  
Ascl-1
 forgtggtcgttcccctctatgant517–6935–6ENSCAFT00000012314
 revgccataggagtccatgagga177 bp  
Ascl-4
 foragccgaatggaaaggtttttnt359–5391–2ENSCAFT00000028710
 revgtctgagctgcaatcatcca181 bp  
Ascl-5
 fortgtatgacaccttgggagcant542–7057–8ENSCAFT00000012348
 revctcaaagggatccatgagga164 bp  
Ascl-6
 forgaggatgcgtgaggatgattnt1352–155315–16ENSCAFT00000001273
 revatatgattgcagggcagagg202 bp  
IL-1β
 forccctggaaatgtgaagtgctnt96–3362–5ENSCAFG00000007249
 revtatccgcatctgttttgcag241 bp  
CyclophilinB
 forggtcatcggtctctttggaant168–3422–4ENSCAFT00000027006
 revgatgctctttcctccagtgc175 bp  

Relative quantification of gene expression

Dietary lipids are predominantly absorbed in the small intestine rather than in the large intestine 11, 24. In this study, the mRNA expression of target genes was additionally measured in the colon. This serves both as an internal control and as a calibrator for the determination of gene expression in the duodenum. Figure 1 compares the duodenal and colonic mRNA levels of two randomly chosen genes analyzed within our study. The results indicate a greater mRNA abundance (p < 0.001) in the duodenum than in the colon for both genes. This distribution pattern was representative for all selected genes of FA uptake measured (data not shown).

Figure 1.

Comparison of mRNA abundance of FAT/CD36 and Ascl-5 measured by real-time quantitative PCR in the duodenum (left panel) and the colon (right panel) of dogs with FRD and IBD. These exemplary data represent the general trend in the expression pattern of selected genes of FA uptake in the two intestinal locations. The delta Ct (where Ct is the amplification cycle number at which the measured signal becomes higher than background) value was calculated as the mean Ct of target minus the mean Ct of housekeepers and transformed to a relative expression as 2−ΔCt27. The box represents 50% of values and the lines within boxes indicate the median.

The evaluation procedure for the gene expression data using a calibrator was adapted from that previously described 29. First, the delta crossing point values (Ct; where Ct is the amplification cycle number at which the measured signal becomes higher than background) in the duodenum and colon was obtained as Ct of the target minus Ct of the mean of housekeeping genes (GAPDH, ubiquitin and cyclophilin B). Second, the delta delta Ct for the duodenum was calculated as the delta Ct for the duodenum minus the delta Ct for the colon. Finally, as previously described 30, the relative mRNA abundance of targets was expressed as 2−ΔΔCt. Intra-assay and inter-assay coefficients of variation of replicates were determined and values varied between 1.9 and 6.8%, respectively.

Statistics

The mRNA data were statistically evaluated with the GraphPad Prism program (GraphPad Software Inc., La Jolla, CA, USA). Since the relative gene expression of target genes was not normally distributed according to the Kolmogorov Smirnov distribution, Kruskal–Wallis one way ANOVA on Ranks was used to identify the differences between dogs within one disease group (before and after treatment) and healthy controls. The Dunn's multiple comparison test served to adjust for multiple testing. The Mann–Whitney U-test was used to determine the difference between IBD and FRD. The Spearman rank order correlation test was performed to evaluate the correlation among target genes and their correlation to CIBDAI. The level of significance was set at p < 0.05.

Results and discussion

Gene expression calculation

The relative mRNA abundance of targets was expressed according to the delta delta Ct 30, by using the mRNA levels of ubiquitin, GAPDH, and cyclophilin B, whose mRNA levels were unaltered by treatment. The mean ± SEM of housekeeping genes in dogs with FRD before treatment, FRD after treatment, IBD before treatment, IBD after treatment, and in controls were similar (p = 0.11) for the duodenum (17.03 ± 0.13, 17.24 ± 0.09, 16.54 ± 0.16, 17.24 ± 0.14, and 17.87 ± 0.16 (p = 0.224) and for the colon (16.89 ± 0.11, 17.10 ± 0.11, 16.43 ± 0.18, 16.62 ± 0.07, and 17.03 ± 0.18).

Animals and intestinal inflammatory activity

All dogs included in the study were evaluated for their CIBDAI to indirectly monitor the consequences of the intestinal inflammatory activity before and after treatment. As shown in Table 4, the CIBDAI scores significantly decreased after treatment both in dogs with FRD (p < 0.0001) and IBD (p < 0.01). As expected, the CIBDAI values indicated that clinical manifestations of the disease were more pronounced in IBD than in FRD, in accordance with previous studies 8. The decrease of the CIBDAI after treatment to levels indicative of insignificant disease, i.e., CIBDAI ≤3, suggests that the fish-meal- and potato-protein-based diet was sufficient for the treatment of FRD. On the other hand, a combination of diet and anti-inflammatory drugs was required to reduce the CIBDAI values to levels similar to those seen in FRD after dietary treatment alone. Unlike previous results reporting CIBDAI situated around 5 after treatment 4, in the present study the value of CIBDAI dropped to ∼2. The discrepancy between the two studies may suggest a positive effect of the nutrient composition of the elimination diets. Indeed, unlike the dogs in the present study, those in the previous study 4 were treated with a diet based on salmon and trout in addition to canola meal and rice (Purina Canine LA® Limited Antigen Diet, St. Louis, MO). In the present study, anti-inflammatory effects exerted by dietary omega-3 PUFA possibly contributed to the modulation of the inflammatory activity in the duodenum of sick dogs, in accordance with 31, 32. However, interpretations should be made with caution, since other vegetable ingredients might also have had similar beneficial health effects 21.

Table 4. Canine intestinal bowel disease index activity (CIBDAI)a)
Treatmentsb)Dogs with FRDDogs with IBD
  • Data are given as median (and range) for dogs with FRD (n = 14) and dogs with IBD (n = 7).

  • (*) and (**) indicate significant differences (*p < 0.01; **p < 0.001) before and after treatment within each treatment group.

  • a)

    The CIBDAI was monitored to indirectly assess the severity of disease (see Section 2 for additional details). The CIBDAI included scores for attitude and activity, appetite, vomiting, stool consistency, stool frequency, and weight loss. CIBDAI values indicated: 0–3 = healthy or clinically insignificant disease; 4–5 = mild disease; 6–8 = moderate disease; ≥9 = severe disease.

  • b)

    The dogs were fed a fish-meal- and potato-protein-based diet enriched in omega-3 PUFA exclusively (FRD) or in combination with anti-inflammatory drugs (IBD; see Section 2 for additional details).

Before5.5** (1–11)9* (3–14)
After1.07 (0–4)1.5 (0–5)

In this study, dog ages did not significantly differ between FRD and IBD groups (p = 0.062). The median and range values were 24 months (9–134 months) for dogs with FRD and 48 months (36–154 months) for those with IBD. However, although the difference did not reach a statistical significance, dogs with FRD were numerically younger than those with IBD.

Control dogs were clinically healthy and had no signs of GIT disease (CIBDAI scores <1).

Cytokine mRNA expression in the duodenum

The mRNA abundance of IL-1β, IL-6, and tumor necrosis factor α (TNF-α) was measured to assess the intestinal inflammatory status. The data demonstrated a decrease of IL-1β mRNA levels in the duodenum of dogs with FRD (p < 0.05) and IBD (p < 0.05) after treatment (Fig. 2A). Furthermore, the mRNA abundance of IL-1β in dogs with FRD and IBD before treatment was higher (p < 0.05 and 0.001, respectively), than in controls. After treatment, the mRNA abundance of IL-1β in both dogs with FRD and IBD was similar (p > 0.05) to controls. There were no differences in mRNA levels between dogs with FRD and IBD both before and after treatment (p > 0.05 in both cases; Fig. 2A). Unlike IL-1β, the mRNA abundance of IL-6 (Fig. 2B) and of TNF-α (Fig. 2C) was unaltered by the treatment both in dogs with FRD (p > 0.05, respectively), and with IBD (p > 0.05, respectively). The mRNA levels of IL-6 and TNF-α before and after treatment were comparable to controls (p > 0.05, in both cases; Fig. 2B and C). The results indicate that the duodenal mRNA abundance of IL-6 and TNF-α was unaltered by treatment in all dogs. This is not surprising since conflicting findings have been previously reported regarding the gene expression of cytokines in intestine of dogs with CE 4, 33–35. On the other hand, the decrease of the mRNA abundance of IL-1β after treatment in both FRD and IBD suggests, at least in our study, that IL-1β is an inflammatory marker in the duodenum. Furthermore, the data indicate that omega-3 PUFA, which were present in increased amounts in the elimination diet, may suppress the IL-1β production by intestinal immune cells 36, 37.

Figure 2.

Box-plots showing mRNA levels of IL-1β (panel A), IL-6 (panel B), and TNF-α (panel C), as measured by real-time quantitative PCR in biopsies from the duodenum of dogs with FRD and IBD, before and after dietary treatment. Tissues from healthy dogs served as controls. The gene expression in the colon was used as an internal control and served for the calibration of the gene expression in the duodenum. Samples were determined in duplicate and the relative gene expression was normalized to three housekeeping genes as 2−ΔΔCt according to 34 (see Section 2 for details). The boxes represent 50% of values and the solid line within boxes represents the median value. a,bValues with different superscripts differ significantly (p < 0.05) within FRD and compared with healthy controls. x,yValues with different superscripts differ significantly (p < 0.05) within IBD and compared with healthy controls. Values with different capital letters differ significantly (p < 0.05) between FRD and IBD before treatment (A and B) and after treatment (X and Y).

mRNA expression of genes involved in fatty acid uptake at the duodenal cell surface

The mRNA levels of selected genes involved in FA uptake at the cell surface 11–15 were measured to evaluate their potential involvement and the beneficial potential of omega-3 PUFA for the treatment of canine CE. Of the measured genes, FATP-2, and FATP-4 mRNA levels were unaltered by treatment in both FRD and IBD (data not shown).

In contrast, the mRNA levels of FATP-1 and FATP-3 increased (p < 0.01 and 0.001, respectively), after the treatment in the duodenum of dogs with IBD (Fig. 3A and B), but not in dogs with FRD. The mRNA levels after treatment were higher than in controls (p < 0.001 and 0.05, respectively; Fig. 3A and B). Both FATP-1 and -3 thus seem to be more active in processes associated with treatment in dogs with IBD. The results also suggest that different mechanisms involving cell surface FA transporters exist in the two forms of CE. Moreover, the mRNA abundance of FATP-1 in the duodenum of dogs with FRD, before and after treatment, was lower (p < 0.001, in both cases) than in dogs with IBD (Fig. 3A), indicating that FATP-1 might be less important in FRD.

Figure 3.

Box-plots showing mRNA levels of FAFATP-1 (panel A), FATP-3 (panel B), FATP-6 (panel C), and FA translocase (panel D) as measured by real-time quantitative PCR in biopsies from the duodenum of dogs with FRD and IBD, before and after dietary treatment. Tissues from healthy dogs served as controls. The gene expression in the colon was used as an internal control and served for the calibration of the gene expression in the duodenum. See Fig. 2 for details.

Unlike the above-mentioned FATPs, the mRNA abundance of FATP-6 and of FAT/CD36 in the duodenum of dogs with FRD increased (p < 0.05) after treatment, while remaining unchanged (p > 0.05) in dogs with IBD (Fig. 3C and D). This suggests that FATP-6 and FAT/CD36 are more important for processes involving FA uptake in FRD than in IBD. Furthermore, the mRNA levels of FAT/CD36 in the duodenum of dogs with FRD before and after treatment were higher (p < 0.05 and 0.001, respectively), than in controls. In contrast, mRNA levels in the duodenum of dogs with IBD were unaltered (p > 0.05) by treatment, and levels before and after treatment were similar (p > 0.05, in both cases) to controls (Fig. 3D). The mRNA abundance of FAT/CD36 in the duodenum of dogs with FRD before and after treatment was higher (p < 0.05 and 0.001, respectively), than in those with IBD (Fig. 3D). The results support the assumption that FAT/CD36 is less important in IBD than in FRD. In this study, the clinical manifestations of CE were more pronounced in dogs with IBD. Because the mRNA abundance of FATP-6 and FAT/CD36 in these dogs was similar to controls, this also suggests that they might not be very important for processes associated with the improvement of the health status of dogs with IBD.

mRNA expression of fatty acid binding proteins in the duodenum

As expected, the mRNA transcripts of I-FABP and L-FABP were expressed in the duodenum of all tested dogs. Indeed, the small intestine is the initial site of dietary FA uptake and it has been reported that intestinal enterocytes co-express both I-FABP and L-FABP 38. In contrast to earlier findings 38 reporting a greater abundance of L-FABP than I-FABP in humans and comparable mRNA levels between L-FABP and I-FABP in rodents, our data in the duodenum demonstrated a greater mRNA abundance (>10-fold) of I-FABP than L-FABP. Taken together, this may indicate a species-dependent role of these cytosolic binding proteins. To date, the precise role of FABPs has not been clarified, although they have been clearly associated with intra-cellular lipid transport and metabolism 13, 15, 39, 40. In the present study, the mRNA abundance of I-FABP in the duodenum of dogs with FRD and IBD increased (p < 0.01 and 0.05, respectively), after treatment (Fig. 4A), suggesting a response to the dietary treatment, i.e., supposedly to the increased PUFA content. In both FRD and IBD, the mRNA levels of I-FABP before treatment were comparable to controls (p > 0.05, in both cases), whereas, the mRNA levels of I-FABP after treatment were higher than in controls (p < 0.001 and 0.05, respectively; Fig. 4A). The mRNA levels of I-FABP in dogs with FRD before and after treatment were similar (p > 0.05, in both cases) to that in dogs with IBD (Fig. 4A). Our results indicate that FA were absorbed during dietary treatment and thus affected FA binding proteins.

Figure 4.

Box-plots showing mRNA levels of intestinal (I)-FA binding protein (FABP; panel A) and liver (L)-FABP (panel B), as measured by real-time quantitative PCR in biopsies from the duodenum of dogs with FRD and IBD, before and after dietary treatment. Tissues from healthy dogs served as controls. The gene expression in the colon was used as an internal control and served for the calibration of the gene expression in the duodenum. See Fig. 2 for details.

Contrary to I-FABP, the mRNA levels of L-FABP were unaltered by the dietary treatment. This could be explained by the fact that FABPs elicit differences in their selectivity, affinity, and binding mechanisms relative to long chain PUFA and various acyl metabolites 39–41. The mRNA abundance of L-FABP in the duodenum of dogs with FRD and IBD was unaltered by treatment (p > 0.05, in both cases), although the trend to an increase was observed in IBD after treatment. The mRNA abundance in dogs with FRD before and after treatment was higher (p < 0.01 and 0.001, respectively), than in controls (Fig. 4B). The mRNA abundance of L-FABP in dogs with FRD before and after treatment was comparable to that in dogs with IBD (p > 0.05, in both cases; Fig. 4B). It is worthy noting that L-FABP is essential for the uptake of PUFA 42 and that the elimination diet in this study also contained high amounts of FA species other than PUFA.

mRNA expression of fatty acid activators in the duodenum

The Acsl are membrane-associated proteins that promote the acylation of long-chain FA into intra-cellular FA-acyl CoA 43. It has been reported that Acsl-1 and -5 similarly activate most unsaturated FA 44, 45, while Ascl-3, Ascl-4, and Ascl-6 preferentially activate PUFA 44, 46. In the present study, the mRNA abundance of Ascl-1 and Ascl-4 did not vary with the treatment (data not shown). In addition, their mRNA levels were much lower than those of Ascl-5 and Ascl-6, suggesting that the latter could be of major importance in biological processes occurring in the duodenum of patients 47. The mRNA abundance of Ascl-5 increased (p < 0.05) after treatment in dogs with FRD, but remained stable (p > 0.05) in dogs with IBD (Fig. 5A). In contrast, the mRNA abundance of Ascl-6 did not change (p > 0.05) after treatment in the duodenum of dogs with FRD, but showed a dramatic increase (p < 0.001) in dogs with IBD (Fig. 5B). Because it has been demonstrated in vitro that over-expression of Ascl-6 increases PUFA uptake 48, our results suggest a variable importance of these two most abundant FA activators in the two forms of canine CE.

Figure 5.

Box-plots showing mRNA levels of Ascl-5 (panel A), and Ascl-6 (panel B), as measured by real-time quantitative PCR in biopsies from the duodenum of dogs with FRD and IBD, before and after dietary treatment. Tissues from healthy dogs served as controls. The gene expression in the colon was used as an internal control and served for the calibration of the gene expression in the duodenum. See Fig. 2 for details.

In summary, the present study describes a beneficial role of the fish-meal- and potato-protein-based diet enriched with omega-3 PUFA in the treatment of canine CE, and describes its effects on the mRNA expression of genes involved in FA uptake. The obvious lack of previously published data on the expression pattern of genes of FA uptake in the context of intestinal disorders has limited further critical interpretation of data reported here. In the present study, dogs with FRD were successfully treated with an elimination diet alone, while a combination of an elimination diet with anti-inflammatory drugs was necessary in the case of IBD. Whether the immunosuppressive drugs can per se affect the duodenal mRNA expression of investigated genes was not specifically tested. However, a study performed in rats revealed that utilization of immunosuppressive drugs in therapeutic concentrations did not have a direct and immediate influence on gastrointestinal function 49. Based on data presented here, dietary FA may contribute to biological processes in the duodenum of canine patients, but different pathways seem to be involved in the uptake of dietary FA.

Conclusions

The present study revealed that feeding a fish-meal- and potato-protein-based diet enriched with omega-3 FA can alter the mRNA expression of genes involved in FA uptake and that this seems to be beneficial in dogs with CE, specifically those suffering from FRD. Our findings imply a potential effect of FA intake on genes that are candidate targets in the treatment of CE. However, additional studies using different diets and FA compositions are warranted.

Acknowledgements

This study was supported by grants from the SwissLife Jubiläumsstiftung, the Novartis Foundation, and by the Department of Clinical Veterinary Medicine of the Vetsuisse Faculty. We thank Maria Feher at the Institute of Biochemistry and Molecular Medicine for mRNA isolation. We also thank Biomill SA (Granges-Marnand, Switzerland) for preparing the diets and particularly Dr. Dorothee Isler for preparing the elimination diet.

The authors have declared no conflict of interest.

Ancillary