Inflammatory bowel disease affects density of nitrergic nerve fibers in the mucosal layer of the canine gastrointestinal tract

Abstract

The objective of this study was to determine the effect of inflammatory bowel disease (IBD) on the density of nitrergic nerve fibers in the mucosal layer of different sections of the gastrointestinal tract of dogs. Twenty-eight German shepherd hybrid dogs of both sexes, weighing from 15 to 25 kg and aged 6 to 10 y, were studied. The dogs were divided into 4 groups with 7 animals in each group: healthy animals, as well as dogs suffering from mild, moderate, and severe IBD. Immunoreactivity to neuronal isoform of nitric oxide synthase, which is a marker of nitrergic neurons, in samples of the mucosal layer in the duodenum, jejunum, and descending colon was studied using the single immunofluorescence method and the number of nerve fibers was evaluated in each observation field. The obtained results showed that IBD causes an increase in the number of nitrergic nerve fibers in all intestinal segments studied and these changes are directly proportional to the intensity of the disease process. These observations may be useful in diagnostic evaluation of the stage of canine inflammatory bowel disease in veterinary practice. The pathological mechanisms of these observed changes and the specific reasons for them are still not completely explained, however, and further study is required.

Introduction

The gastrointestinal (GI) tract is innervated by the enteric nervous system (ENS) located in the wall of the esophagus, stomach, and intestine (1,2). There is also extrinsic innervation, in which bodies of neuronal cells supplying the GI tract are situated in parasympathetic, sympathetic, and sensory ganglia (3,4). The ENS and extrinsic innervation system regulate all functions of the GI tract, such as intestinal motility, blood flow, and excretive activity (2). It is well-known that neurons within the ENS and extrinsic neuronal cells supplying the gut may undergo the process of neuronal plasticity, which is reflected by changes within the neurons under various stimuli. These stimuli can be grouped into 2 categories: physiological conditions, such as ontogenesis of the GI tract or diet (5,6), and pathological factors, such as intestinal and extra-intestinal diseases, mycotoxin poisoning, and nerve injury (7,8).

It should be noted that neurons innervating the GI tract show the presence of a wide range of active substances, which may act as neuromediators and/or neuromodulators (2). Changes in the expression of these substances are the main symptom of neuronal plasticity (8). Nitric oxide (NO) is one of many active substances that occur in the intestinal neuronal structures (9). It is a gaseous neurotransmitter, which is synthesized from the guanidine group of L-arginine during reaction, catalyzed by a family of enzymes called nitric oxide synthases (NOSs) (10). Three isoenzymes are recognized here: neuronal (nNOS or NOS-1); cytokine-inducible (iNOS or NOS-2); and endothelial (eNOS or NOS-3) (10), among which nNOS is the marker of nitrergic neurons.

Until now, nitric oxide has been described in the intestinal neuronal structures of various parts of the GI tract in a wide range of mammal species including humans (11–15). Most nitrergic neurons have been noted within the myenteric plexus, which is part of the ENS located between the longitudinal and circular muscle layers of the GI tract (12). Most nitrergic nerve fibers have been observed within the muscular layers of the intestine, although mucosal nerves that are immunoreactive to nNOS have also been described (13,14).

Nitric oxide in the GI tract has an inhibitory function, namely it suppresses intestinal motility, both by directly influencing muscular cells and by inhibiting the release of stimulating neuromediators (16). Nitric oxide (NO) takes part in the reduction of excretive action in the GI tract, regulation of blood flow in the intestinal wall (by dilation of blood vessels), as well as neuroprotective and/or adaptive processes in the enteric nervous system (11,17,18). Finally, NO is probably involved in changes of neuronal nitric oxide synthase-like immunoreactivity (nNOS-LI) in intestinal nervous structures, which have been noted in the presence of various pathological factors, such as Crohn’s disease, inflammatory processes, or mycotoxin poisoning (19,20).

Idiopathic inflammatory bowel disease (IBD) in dogs, however, is a group of chronic and recurrent pathological states within the GI tract, which is an important problem in recent veterinary practice. Both the diagnosis of IBD, which consists of ruling out other reasons for intestinal symptoms characteristic of this disease, such as loss of appetite, vomiting, and diarrhea (21–23), and its treatment are difficult. Many pathological mechanisms accompanying IBD, such as the influence of IBD on intestinal nervous structures and changes in expression of neurotransmitters and/or neuromodulators, are not fully explained and knowledge about them is limited (24–26).

The present study reports for the first time on how IBD affects nitrergic nervous structures within the mucosal layer of the canine digestive tract. Results obtained may not only add to knowledge about the pathological mechanisms of IBD, but could also contribute to improving diagnostic methods for this disease.

Materials and methods

Twenty-eight German shepherd hybrid dogs of both sexes, weighing from 15 to 25 kg and aged 6 to 10 y, participated in this study. The control group consisted of 7 healthy dogs approved for the experiment during screening tests for IBD conducted at a dog shelter in Olsztyn, Poland. The experimental groups were selected from patients of the Veterinary Clinic at the University of Warmia and Mazury in Olsztyn, Poland. All experimental animals were approved for the experiment based on results of clinical, laboratory, and endoscopic examinations, as well as histopathology of duodenal, jejunal, and colonic mucosa.

Dogs with IBD were subjected to biochemical, radiological, parasitological, bacteriological, and mycological stool tests and provocation trials in order to exclude other diseases associated with chronic diarrhea. Based on their Clinical Inflammatory Bowel Disease Activity Index (CIBDAI) scores (22), animals suffering from IBD were divided into the following groups: Group I — mild IBD, CIBDAI score — 4 to 5 points, histopathological score “+”; Group II — moderate IBD, CIBDAI score — 6 to 8 points, histopathological score “++”; and Group III — severe IBD, CIBDAI score — 9 to 16 points, histopathological score “+++”. Each group consisted of 7 dogs. The clinical CIBDAI scores for all groups as well as for individual dogs are listed in Table I. All procedures were carried out according to the instructions of the Local Ethical Committee in Olsztyn, Poland (decision number 47/2009/DTN).

Table I.

Clinical CIBDAI scores for groups as well as individual dogs

Group	CIBDAI score	Patient number	Individual CIBDAI score	X CIBDAI	SD CIBDAI
Control group	0 to 3	1	0	0	0
		2	0
		3	0
		4	0
		5	0
		6	0
		7	0
Group I — mild IBD	4 to 5	1	4	4.43	0.57
		2	5
		3	4
		4	4
		5	5
		6	4
		7	5
Group II — moderate IBD	6 to 8	1	7	7.29	0.71
		2	8
		3	8
		4	7
		5	7
		6	7
		7	7
Group III — severe IBD	< 9	1	9	9.57	1.43
		2	9
		3	9
		4	9
		5	11
		6	9
		7	11

CIBDAI — Clinical Inflammatory Bowel Disease Activity Index.

X — mean average.

SD — standard deviation.

Tissue samples were obtained from all dogs, both experimental and control, during gastroscopic or colonoscopic examinations with the use of FB-24U-1 biopsy forceps with a diameter of 2.5 mm and FB-50U-1 biopsy forceps with a diameter of 3.7 mm (Olympus Medical System Corp., Tokyo, Japan). Three biopsy specimens from every section of the gastrointestinal tract from each dog were fixed by immersion in 4% buffered paraformaldehyde solution for 15 min and rinsed in phosphate buffer (pH 7.4) for 3 d. Tissue samples were then transferred to 18% saccharose solution in phosphate buffer and stored at 4°C for at least a week and then frozen (at –22°C), sectioned into slices of 10 μm thickness in the Microm Cryostat HM 525 (Thermo Scientific, Walldorf, Germany), and placed on gelatin-coated slides.

Slices were subjected to routine single immunofluorescence method as described in a previous study (24). In brief, the sections were dried for 45 min at room temperature, rinsed in buffered PBS solution (PBS, 0.1 mol, pH 7.4, 3 × 15 min), and incubated for 1 h with blocking solution containing 10% goat serum, 0.1% bovine serum albumin (BSA), 0.01% sodium azide (NaN₃), Triton X-100, and thimerosal in PBS. After rinsing in NaCl solution, tissue samples were incubated overnight in a humidity chamber at room temperature with primary antibodies directed against the neuronal form of NOS (Rabbit; Cappel, Aurora, Ohio, USA), at a working dilution of 1:4000. The next day, slices were again rinsed in PBS (3 × 15 min) and incubated with specific secondary antibody conjugated to Alexa Fluor 594 at a working dilution of 1:1000 (Donkey; Invitrogen, Carlsbad, California, USA) for 1 h at room temperature. After repeated rinsing in NaCl solution, the specimens were mounted on slides with glycerol solution and PBS (1:2; pH 7.4) and covered with coverslips.

Tissue samples were observed under the Olympus BX51 fluorescence microscope equipped with appropriate filters and the density of nNOS-like immunoreactive (nNOS-LI) mucosal nerve fibers was defined by a semi-quantitative evaluation based on counting nNOS-positive nerves in the field of view (0.1 mm²). Such fibers were counted in 4 fields of view in 3 sections of every biopsy specimen from the duodenum, jejunum, and colon. A total of 36 fields of view in every intestinal part of each dog was evaluated. Observed fields of view were located at least 100 μm apart, which prevented the repeated counting of the same nerves. Obtained data were pooled and presented as a mean ± standard deviation (SD).

Standard controls of specificity of the method were carried out. Antibody against nNOS was tested by pre-absorption of antiserum with appropriate antigen and omission and replacement of primary antiserum by non-immune serum. The pre-absorption test was done as described in a previous study (13) and included incubation of the intestinal slices with “working” dilutions of antibody directed towards nNOS, which had already been pre-absorbed for 18 h at 37°C with 1.0 μM solution of nitric oxide synthase (Sigma Aldrich, St. Louis, Missouri, USA). These controls completely eliminated specific stainings.

The specificity of the primary antibody was verified by Western blot analysis on canine ileum, which was obtained from another dog euthanized due to cancer at the Veterinary Clinic of the University of Warmia and Mazury in Olsztyn, Poland. Frozen tissue (200 mg) was homogenized in 1800 μL of sodium dodecyl sulfate (SDS) sample buffer (27) using an Ultra-Turrax homogenizer (Janke & Kunkel, IKA Works GmbH & Co. KG, Staufen, Germany). Samples were heated to 95°C for 4 min, cooled on ice, and centrifuged for 5 min at 2000 × g. Supernatant (30 μL) was loaded to 10% SDS-polyacrylmide gel electrophoresis (PAGE) gel (27), together with 5 μL of molecular weight standard (PageRuler Broad Range Protein Ladder, ThermoFisher, USA). Electrophoresis was done for 90 min at 100 V in a Mini-Protean 3 Cell Electrophoresis System (BioRad, Richmond, California, USA). After electrophoresis, the proteins were blotted to polyvinylidene difluoride (PVDF) membrane (BioRad) in Mini Trans-Blot Electrophoretic Transfer Cell (BioRad) using transfer buffer (25 mM TRIS, 192 mM glycine, pH 8.3) at 100 V for 80 min.

After transfer, proteins in the blot were stained with 0.05% Amido Black in 10% acetic acid, following destaining in 10% acetic acid-20% methanol. Positions of lanes and molecular weight marker bands were marked on the membrane with a soft pencil and the membrane was rinsed in PBS. The membrane was then blocked with a blocking solution (3% BSA in TBS-0.05% Tween 20) for 1 h at 25°C. The blocking solution was replaced with a primary anti-nNOS antibody, working dilution 1:4000, in a blocking solution (Rabbit; Cappel) and incubated overnight at 25°C. After incubation, the membrane was washed 3 × 10 min with PBS and then incubated for 90 min in secondary biotinylated antibody solution (VectaStain; Vector Laboratories, Burlingame, California, USA) in the blocking solution. After incubation, the membrane was again washed 3 × 10 min in TBS-0.05% Tween and incubated for 40 min with a streptavidin-conjugated alkaline phosphatase (R&D Systems, Minneapolis, Minnesota, USA), diluted 1:500 in TBS-0.05% Tween. The membrane was then washed 3 × 10 min in TBS-0.05% Tween and incubated for 5 min with alkaline phosphatase-activation buffer [100 mM NaCl, 5 mM magnesium chloride (MgCl₂), 0.1% Tween 20, 100 mM TRIS, pH 9.5]. The color reaction was developed in the alkaline phosphatase-activation buffer containing 6 μL of BCIP [5-Bromo-4-chloro-3-indolyl phosphate, 20 mg/mL, (Sigma Aldrich)] and 8 μL of Nitroblue tetrazolium (NBT) (50 mg/mL, Sigma Aldrich) per 10 mL. The color reaction was stopped by rinsing in tap water, followed by ultrapure water. The membrane was dried and scanned with a desktop scanner. The image was digitally processed and the locations of selected molecular weight standard bands were placed in the image (Figure 1). The “<” mark in the figure shows the location of a 150 kDa band.

Figure 1.

Western blot test result.

The significance of differences between groups was determined by the Kruskal-Wallis test at P ≤ 0.05 (significant) and P ≤ 0.01 (highly significant). The results were processed using Statistica 9.1 Software (StatSoft, Tulsa, Oklahoma, USA).

Results

During the present study, nitrergic nerve fibers were noted in all intestinal segments studied, both in control animals and in dogs suffering from IBD, but the population of such fibers was low (Figures 2, 3). In healthy dogs, the average number of nerve processes immunoreactive to nNOS in all studied regions of the intestine fluctuated at around 0.40 per observation field and was even in all intestinal fragments studied (Table II). In individual observation fields, usually 1 or none of the nitrergic nerve was noted (Figures 2A, 3AI, 3AII). Observation fields with 2 such nerve processes were only observed occasionally.

Figure 2.

Distribution pattern of neuronal isoform of nitric oxide synthase-like immunoreactive nerve fibers (arrows) in the mucosal layer of the canine duodenum. A — physiological conditions; B — mild inflammatory bowel disease (IBD); C — moderate IBD; D — severe IBD. Scale bar = 50 μm.

Figure 3.

Distribution pattern of neuronal isoform of nitric oxide synthase-like immunoreactive nerve fibers (arrows) in the mucosal layer of the canine jejunum (I) and descending colon (II). A — physiological conditions; B — moderate inflammatory bowel disease (IBD); C — severe IBD.

Table II.

The number of NOS-immunoreactive nerve fibers per observation field in the duodenum, jejunum, and descending colon in control dogs (Group C) and in dogs suffering from mild (Group I), moderate (Group II), and severe (Group III) IBD

	Group C	Group I	Group II	Group III
Duodenum	0.48 ± 0.14c	0.55 ± 0.21c	0.77 ± 0.22	1.26 ± 0.24a,b
Jejunum	0.42 ± 0.18c	0.56 ± 0.23	0.54 ± 0.07	0.84 ± 0.15a
Colon	0.48 ± 0.28c	0.54 ± 0.30c	0.61 ± 0.06	0.97 ± 0.15a,b

^aSignificantly different from control group.

^bSignificantly different from group I.

^cSignificantly different from group III.

Kruskal-Wallis test — P < 0.05 — lowercase letters; P < 0.01 — uppercase letters.

Obtained results show that the density of the mucosal nerve fibers increased in proportion to the intensity of the IBD process (Figures 2, 3, 4). In the duodenum (Figure 2), a gradual increase in the number of nitrergic nerves proportional to the intensity of the disease process was noticed. The average number of nerve processes immunoreactive to nNOS per observation field amounted to 0.48 ± 0.14 in control animals and 0.55 ± 0.21, 0.77 ± 0.22, and 1.26 ± 0.24 in groups I, II, and III, respectively (Table II, Figure 4). In the duodenum of dogs with severe IBD (group III), the population of nitrergic nerves was the most numerous among all intestinal fragments studied. Some observation fields with 3 nNOS-positive nerves were noted only in this group (Figure 2D).

Figure 4.

Immunoreactivity of NOS fibers in the digestive tract in healthy dogs and in animals with IBD of various severity.

In the jejunum of healthy dogs, the number of nNOS-like immunoreactive nerves was slightly lower than within the duodenum and came to 0.42 ± 0.18 (Table II, Figure 3AI). Mild and moderate IBD (groups I and II) caused a slight increase of such nerves (0.56 ± 0.23 and 0.54 ± 0.07, respectively), whereas in dogs suffering from severe IBD (group III), the increase in the number of described nerve processes was greater and amounted to 0.84 ± 0.15 per observation field (Table II, Figure 4).

Changes observed in the mucosal layer of the descending colon (Figure 3B) were similar to those observed in the duodenum and jejunum. Namely, in control animals, the average number of nNOS-positive nerves amounted to 0.48 ± 0.28 per observation field. This value progressively increased as the intensity of IBD pathological processes rose and amounted to 0.54 ± 0.30 in group I, 0.61 ± 0.06 in group II, and 0.97 ± 0.15 in group III (Table II, Figure 4).

Statistical analysis showed essential differences, especially between the control group and group I and dogs with severe IBD (group III). Highly statistically significant differences were observed in the duodenum and descending colon of healthy dogs (group C) and dogs with mild IBD (group I) compared to those in dogs with severe IBD (group III). In the case of the jejunum, highly statistically significant differences were observed only between control dogs and those with severe IBD.

In regard to the localization of nerves studied, most nNOS-positive fibers, both in physiological conditions as well as during IBD, were observed in the superficial layer of the mucosa and probably supplied the enterocytes and especially enteroendocrine cells. Considerably fewer nNOS-LI nerves were noted in the deep layer of mucosa, not far from blood vessels.

Discussion

In spite of the numerous important functions of nitric oxide (NO) in the gastrointestinal tract, the number of nitrergic mucosal nerve fibers observed during the present study seems to be very slight. Values noted herein are lower than those observed in the gastrointestinal tract of other species of mammals (11), which may be a result of interspecies differences. The present study, however, seems to be partially compatible with previous studies, where most nitrergic neuronal structures have been noted first within the intestinal muscular layer and the myenteric plexus of the ENS (11,13,14).

The present results show that canine IBD may change the number of nNOS-positive mucosal nerve fibers in various segments of the digestive tract. Moreover, a directly proportional relationship was noted between the number of these nerves and the intensity of pathological processes connected with inflammatory infiltration of the lamina propria of the mucosal layer.

It should be pointed out that knowledge about IBD-induced changes in the expression of active substances within nerve structures in the canine digestive tract is limited (24,26,28). Both the pathological mechanisms of these observed changes and the specific reasons for them remain unknown. The increase in the number of nNOS-like immunoreactive mucosal nervous structures observed during the present study could be due to factors responsible for triggering canine IBD, which include bacterial, allergic, and environmental agents (21–23). These noted changes could also be due to secondary processes, such as pain or digestive dysfunction, that occur during IBD. The increase in the density of fibers immunoreactive to nNOS may reflect changes in synthase of this enzyme at the transcriptional, translational, or metabolic level, as well as modifications in transport of its molecules from cell bodies to nervous processes.

Observations during the present study may also confirm the findings of previous studies, which suggest that nitric oxide is a neuroprotective neuronal factor that can function in adaptive processes of the nervous system in response to various pathological agents (25,29). This function of NO within the digestive tract is arguable. It is commonly accepted that expression of neuronal neuroprotective factors increases in the presence of other pathological factors. Such changes have been observed both in the present study and in previous investigations of Crohn’s disease or mycotoxin poisoning (19,20). Other pathological processes, such as diabetes mellitus (30), however, caused a decrease in the number of nNOS-immunoreactive intestinal nervous structures. These discrepancies strongly suggest that nitric oxide plays different roles in various pathological processes.

Another reason for changes observed during the present study may be related to inflammatory processes. Until now, it has been well-established that nitric oxide (NO) can have both pro-inflammatory properties, e.g., during arthritis, and anti-inflammatory properties, e.g., during lung inflammatory disorders, depending on the type of inflamed tissues (10). Within the gastrointestinal tract, NO is primarily a pro-inflammatory mediator. It is produced in mucosal neutrophils during acute inflammatory processes, as well as in monocytes and lymphocytes during chronic disease (31). Nitric oxide (NO) also causes an increase in TNF-α concentration, which contributes to infiltration of the mucosal layer by activated neutrophils (32).

Previously mentioned functions of nitric oxide are connected with activity of one of the cytokine-inducible isoenzyme of NOS (iNOS) (10), which among other things, is activated during intestinal inflammatory processes and contributes to damage of the mucosal macrophages (33). This type of NOS, induced by tumor necrosis factor, starts tumor protein p53-dependent processes that result in apoptosis of intestinal epithelium during chronic colitis ulcerosa (34). It is known that inhibitors of iNOS activity significantly reduce the intensity of inflammatory processes and also reduce damage to the mucosal layer of the inflamed intestine (35).

Knowledge about the participation of neuronal isoenzymes of NOS (nNOS) is limited. Previous studies (36) and relatively well known interactions between the enteric nervous system and immunological structures within the gastrointestinal tract, especially during intestinal diseases (19), suggest that results obtained in the present study may be due to the previously described roles of NO in inflammatory processes. Observed changes, however, may be due to the disturbances of intestinal motility that are often observed during IBD. It is possible that the increase in density of nitrergic nerves, which play a role in muscular relaxation (11,16), could be the enteric nervous system’s adaptive response to intestinal irritation and diarrhea.

In conclusion, results obtained in the present study show that IBD causes an increase in the number of mucosal nitrergic nerve fibers in the canine digestive tract. These changes are directly proportional to the intensity of the disease process, which despite the difficulties connected with the small number of mucosal nNOS-LI nerves and relatively high costs and labor intensity of the immunofluorescence technique, could be useful in diagnostic evaluation of the stage of canine inflammatory bowel disease in veterinary practice. The mechanisms of these observed changes and the specific reasons for them are still not completely explained, however, and further study is required.