SEARCH

SEARCH BY CITATION

Keywords:

  • horse;
  • strangulation colic;
  • myosin;
  • HSP20;
  • HSP27

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Manufacturers' addresses
  9. References

Reasons for performing study: Intestinal strangulation often leads to enterectomy after which ileus can develop. This has prompted research to look into possible pathophysiological processes triggering equine ileus. However, morphological changes of the small intestinal smooth muscle in relation to equine colic have not yet been studied.

Objectives: The presence of some smooth muscle proteins was morphologically assessed and quantified in control and colic horses. In addition, the up- or down-regulation of heat shock proteins (HSP20 and HSP27) influencing the contractility of smooth muscles was studied.

Methods: Cranial resection margins of 18 strangulated small intestinal samples were collected. Small intestinal control samples were collected from 11 horses subjected to euthanasia for other than gastrointestinal-related reasons. Formaldehyde-fixed tissue was paraffin-embedded and processed for conventional staining and immunohistochemistry. Snap-frozen full-thickness biopsies were collected for western blot analyses.

Results: Evaluating the muscle layer microscopically, colic samples showed significantly more signs of degradation than controls (P = 0.026) of which vacuolar degeneration was most prominent (P = 0.009). In colic samples, myosin protein levels were decreased (P = 0.022) whereas desmin (P = 0.049) and HSP20 protein levels (P = 0.005) were elevated.

Conclusions: In colic samples, microscopic lesions at the level of the muscle layer indicate a stress response. In addition, modified amounts of structural proteins such as myosin and desmin together with increased HSP20 levels could perhaps provide a basis for explaining the malfunctioning of the intestinal muscle layer.

Potential relevance: Post operative ileus, following small intestinal strangulation and resection, could be related in part to a dysfunctional muscle layer. In addition to microscopic signs of degeneration, myosin and HSP20 were affected. Pharmacological interventions might alter HSP20 expressions and thus serve a protective effect.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Manufacturers' addresses
  9. References

Small intestinal strangulation often leads to resection of the ischaemic intestine but with an increased risk of developing ileus post operatively (Holcombe et al. 2009; Torfs et al. 2009). The pathophysiology of ileus is partly explained by neuroimmune mechanisms. An increased sympathetic neural reflex, followed by a neutrophil influx, has been shown to inhibit smooth muscle function (Boeckxstaens and de Jonge 2009). Neutrophil influx into all small intestinal layers has been identified in both clinical and experimental cases of small intestinal strangulation (Little et al. 2005; Cook et al. 2008). Since anti-inflammatory drugs do not always prevent the development of ileus, this inflammatory response can only partly explain ileus (Holcombe et al. 2009). Normal gastrointestinal motility requires the coordinated function of the myenteric and submucosal plexi (Wood 1981), interstitial cells of Cajal (ICC) (Sanders et al. 2006) and smooth muscle cells (Gabella 1981). Strangulation can harm these gut wall components as a result of ischaemia, oedema and dilation. Malone and Kannan (2001) found no changes in the amount of neurons and in the production of neurotransmitters in jejunal samples obtained proximal and distal to an area made experimentally ischaemic. Nevertheless, in vitro contractility patterns of these samples were clearly aberrant. Although Fintl et al. (2004) identified a significant reduction in ICC density in horses with large colon disorders, no significant decrease was found in horses suffering from small intestinal strangulation. Histopathological studies of gastrointestinal motility disorders have mainly focused on enteric nerves and interstitial cells of Cajal, but rarely considered the enteric musculature. To our knowledge, studies seeking to document alterations in smooth muscle proteins were not yet performed in horse intestine. Smooth muscle cells in the gastrointestinal tract are found in several layers: the lamina muscularis mucosae, an inner circular layer, and an outer longitudinal layer. The smooth muscle cells are spindle-shaped and mononucleated and consist of different types of filaments each composed of typical proteins. The main contractile proteins are actin, myosin and tropomyosin. The main cytoskeletal proteins (actin, tubulin and desmin) function as protectors of muscle integrity and stability (Gabella 1981; Sharov et al. 2005). Both the function and the expression of these contractile and cytoskeletal proteins, have been reported to be disturbed in several pathological conditions such as myocardial ischaemia (Zhang and Riddick 1996), severe colorectal motility disorders (Wedel et al. 2006) and aganglionated bowel (Nemeth et al. 2002).

In pathological conditions, muscle contractility is influenced via small heat shock proteins (HSP). A typical ischaemic marker is HSP27. It is important in several cell functions such as cell survival, stress-induced microfilament organisation and smooth muscle contraction (Martin et al. 1997; Gerthoffer and Gunst 2001; Bitar 2002). In colonic smooth muscle, HSP27 when activated, is phosphorylated. Phosphorylated HSP27 induces prolonged phosphorylation of caldesmon, which is subsequently released from tropomyosin, sliding tropomyosin away from actin in order to expose the myosin binding sites (Bitar 2002; Somara and Bitar 2006).

The mechanisms upregulating HSP20 are not yet elucidated but stress situations, e.g. hypoxia or mechanical stress due to obstruction can trigger its expression (Batts et al. 2006). HSP20 has been reported to decrease the contractile activity in the distended smooth muscle of the bladder (Batts et al. 2006). In addition, HSP20 is found to exert a protective function after ischaemia/reperfusion by improving the recovery of the contractile mechanism in the heart (Fan et al. 2005).

It was our aim to study the expression of the most important smooth muscle contractile and cytoskeletal proteins in the tunica muscularis of the proximal border of resected small intestinal segments of colic cases suffering from small intestinal strangulation. In addition, a possible role of HSP27 and HSP20 was investigated.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Manufacturers' addresses
  9. References

Animals

The cranial border of resected small intestine of 18 colic horses diagnosed with small intestinal strangulation at the Department of Large Animal Internal Medicine of the Ghent University (Belgium) were collected during surgery. Samples were taken avoiding the surgical clamping site in order to prevent interference by iatrogenic damage. The horse's age ranged from <1–25 years (average age = 11.5 years). All horses (n = 18) were Warmblood horses with the exception of one Friesian horse, one Haflinger and one pony. The group consisted of 6 geldings, 11 mares and one stallion. The colic population consisted of samples taken from horses subjected to euthanasia before surgery (n = 6) and of horses submitted to surgery (n = 12). From the horses submitted to surgery (resection group), the outcome is divided in 3 groups: fully recovered (n = 3), ileus development (n = 2) and euthanasia during surgery or after surgery as a result of post operative complications other than ileus such as shock, laminitis, etc (n = 7).

Control small intestinal tissue mid-jejunum was collected at Ghent and Utrecht University Faculty of Veterinary Medicine from horses (n = 11) subjected to euthanasia for reasons other than colic. The average age of the control population was 13.5 years (range: 5–22 years).

Of each sample, four 8 mm full thickness punch biopsies, randomly taken, were rinsed and fixed in 10% buffered formalin overnight. After rinsing the biopsies were routinely processed for paraffin embedding. In addition, 6 full thickness biopsies were immediately frozen in liquid nitrogen and stored at -80°C.

Lesion score

Routinely, haematoxylin-eosin (HE)1,2 stained sections were used to microscopically score degeneration in the smooth muscle cells (tunica muscularis): eosinophilic cytoplasmic degeneration, influx of inflammatory cells, pyknosis, vacuolar degeneration, rupture of smooth muscle cells and wavy smooth muscle fibres were separately scored (0 = not present; 1 = sporadically present; 2 = average presence; 3 = abundantly present). Wavy fibres are smooth muscle cells bending in a wavy pattern often related with hypereosinophilia and sometimes containing pyknotic nuclei (Salinas-Madrigal et al. 1987).

The scores of all separate lesions were added and a general lesion score was created where 1 represents a total score from 0–5, 2 from 6–10 and 3 from 11–15.

The samples were also scored for mucosal damage. The mucosal damage score was adapted from the lesion score developed by Chiu et al. (1970). Briefly: in grade 0 no lesions were present; in grade 1, a small subepithelial space at the villus tip was visible; in grade 2 the lamina propria was exposed at the villus tips and epithelial cells were shedded from the underlying lamina propria; in grade 3 the epithelial cells were separating from the underlying lamina propria as far as halfway down the villus; in grade 4 the lamina propria was denuded as far as the base of the villus; and in grade 5 a total collapse of the lamina propria was found.

All scores were made by 2 blinded researchers, independently from each other.

Immunohistochemistry

Paraffin 4 µm sections were deparaffinised and rehydrated with successive alcohol baths. The sections were immunohistochemically stained using anti-myosin3, anti-panactin4, anti-desmin5, anti-HSP206 and anti-HSP276 antibodies. The detailed protocols are described in Table 1. Briefly, sections were pre-treated for antigen retrieval with Tris-EDTA buffer5 followed by a cool down period of 10 min. After thorough rinsing with Tris-phosphate buffered saline (TBS) (pH = 7.5), endogenous peroxidase was blocked by incubation in 3% hydrogen peroxide. Slides were incubated in 30% normal serum for 30 min, to prevent nonspecific binding of the secondary antibody. Subsequently primary and secondary antibodies were applied consecutively, both followed by rinsing. When necessary, sections were incubated with a tertiary reagent. A positive immunohistochemical reaction was visualised using 3,3′-diaminobenzidine (DAB)3 or 3-amino-9-ethyl-carbazole (AEC)5. All dilutions were made with antibody-diluent5. In the negative controls, antibody-diluent was used instead of primary antibody.

Table 1. Immunohistochemical protocols
TargetAntigen retrievalBlocking EP 30 minNormal serumPrimary antibodySecondary antibodyTertiary reagentDevelopment
  1. MW, microwave; PP, pressure cooker; H2O2, hydrogen peroxide; TBS, tris buffered saline; NGS, normal goat serum5; NRbS, normal rabbit serum5; ON, overnight; BS, 37°C incubator; GARb-hrp, goat anti-rabbit horseradish peroxide5, GARb-b, biotinylated goat anti-rabbit5; RbAG-b, biotinylated rabbit anti-goat5; Env Ms, envision anti-mouse5; S-HRP, streptavidin conjugated horseradish peroxidase5; DAB, 3,3′-diaminobenzidine5; AEC, 3-. amino-9-ethyl-carbazole5.

PanactinPP 10 min3%H2O2 in TBS30% NGS1:100 ON 4°CGARb-HRP 1 h 1:200DAB
MyosinPP 10 min3%H2O2 in TBS30% NGS1:100 ON 4°CGARb-HRP 1 h 1:200DAB
DesminMW 10 min3%H2O2 in methanol1:50 ON 4°CEnv Ms 30 minAEC
HSP27MW 10 min3%H2O2 in TBS30% NRbS1:75 BS 37°CRbAG-b 45 min 1:200S-HRP 1 h 1:200AEC
HSP20MW 10 min3%H2O2 in TBS30% NRbS1:100 ON 4°CRbAG-b 1 h 1:200S-HRP 1 h 1:200AEC

Western blot

The frozen tissue was crushed in liquid nitrogen and proteins were extracted using a lysis buffer (50 mmol/l Tris, 150 m mol/l NaCl, 1% Nonidet P40, 0.5% deoxycholate, half a tablet of protease inhibitor7. The samples were mixed, sonicated, cooled for 30 min and centrifuged for 2 min at 11,340 g. Protein concentrations in the supernatant were determined with a Bicinchoninic acid (BCA)-kit8. Samples were stored at -80°C till further use. Protein samples were prepared for loading by administering laurel-dodecyl-sulphate containing buffer9 and reducing agent9 followed by heating for 2 min in a boiling water bath. After loading 15 µg on a 4–12% Bis-Tris Gel9, samples were transferred to a nitrocellulose membrane9. Membranes were blocked with standard blocking buffer (SBB) (phosphate buffered saline + 1% bovine serum albumin1+ 0.2% Tween 20) during 1 h, after which primary antibodies were applied overnight at 4°C while shaking gently (Table 2). After rinsing with PBS + 0.2% Tween 20, secondary antibody was applied for 1 h at room temperature (RT). When necessary, a tertiary reagent was applied for 1 h at RT. For development enhanced chemiluminescense substrate8 was applied and the membrane was red in an OptiGo system with Isocam-285/20510. Results were analysed with TotalLab 1D Gel Analysis11. A normalised optical density (OD) was obtained by dividing the protein density by the density of the loading control β-tubulin3.

Table 2. Western blot protocols
TargetPrimary antibodySecondary antibodyTertiary reagentDevelopment
  1. RbAG-b, biotinylated rabbit anti-goat5; GARb-hrp, Goat anti-rabbit horseradish peroxide5; S-HRP, streptavidin conjugation horseradish peroxidase5; Env-ms, envision anti-mouse5; ECL, enhanced chemiluminescense substrate8.

Myosin1:500 ON 4°CGARb-HRP 1 h 1:200ECL
Panactin1:1000 ON 4°CGARb-HRP 1 h 1:200ECL
Desmin1:200 ON 4°CEnvision-Ms 30 minECL
HSP271:1000 ON 4°CRbAG-b 1 h 1:200S–HRP 1 h RTECL
HSP201:500 ON 4°CRbAG-b 1 h 1:200S–HRP 1 h RTECL
ß-tubulin1:500 ON 4°CGARb-HRP 1 h 1:500ECL

Statistical analysis

Results were expressed as mean ± s.e.m. For the statistical evaluations, SPSS 1812 was used. In case of categorical data (lesion score, degenerative lesions, sample collection, outcome and status) a Chi-squared test was used. Unpaired Student t test was used for all normally distributed continuous data (myosin-OD – panactin-OD – HSP27-OD) to compare control and colic samples and to compare between the lesion scores and sample collection. Mann-Witney U test was used when the criteria for using parameteric methods were not met (desmin-OD – HSP20-OD). ANOVA analyses (normally distributed continuous data) and Kruskal-Wallis tests (nonparametric datasets) were used in order to compare the parameters grouped according to the degenerative lesions and the outcome. Correlations between the normally distributed continuous data were calculated using a Pearson correlation test whereas not normally distributed continuous data were correlated using a Spearman test.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Manufacturers' addresses
  9. References

Mucosal lesion score

Haematoxylin and eosin staining showed a wide range of lesions. All control samples had grade 0. Twelve out of 18 colic samples (66.7%) showed grade 0 whereas 4 samples (22.2%) showed a discrete subepithelial space (grade 1). Grades 2 and 4 were found only once (5.6%). Grades 3 and 5 were not encountered.

In euthanasia samples, 5 showed mucosal damage whereas only one sample collected during surgery had mucosal damage (P = 0.001).

Muscle lesion score

All control and colic samples had a lesion score >1 (Fig 1). Colic samples showed significantly more degenerative changes than control tissue (P = 0.026). A general lesion score of 2 and 3 was seen in 22.2 and 77.8% of the colic samples, respectively, whereas in control tissue these scores were observed in 63.6 and 36.4% of the samples, respectively.

image

Figure 1. A detail of the normal muscle layer of the equine small intestinal tract (haematoxylin-eosin staining, Scale bar = 50 µm, magnification: 600×).

Download figure to PowerPoint

Eosinophilic cytoplasmic degeneration (P = 0.866, Fig 2a, Table 3), inflammatory cell influx (P = 0.378, Table 3), pyknosis (P = 0.453, Table 3), ruptured (P = 0.635, Table 3) and wavy myocytes (P = 0.463, Fig 2a, Table 3). Vacuolar degeneration occurred in all samples, but was more abundant in colic samples (P = 0.009). Within control samples, most samples showed only low degree vacuolar degeneration (54.5% of the samples), 27.3% showed more prominent vacuolar formation and 18.2% showed abundant vacuolar degeneration. This contrasts with colic samples where the majority showed abundant formation of vacuoles (55.5%, Fig 2b, Table 3). Only 5.6% of the colic samples showed few vacuoles, 38.9% had an average number of vacuoles.

image

Figure 2a. The muscle layer shows very prominent wavy fibres (arrows) and eosinophilic cytoplasmic degeneration (arrowheads) at the level of the circular muscle layer in cranially resected intestine after strangulation (haematoxylin-eosin staining, Scale bar = 50 µm, magnification: 600×).

Download figure to PowerPoint

Table 3. Determination of the lesions and the histopathological score for the control and the colic population with their P values
Lesion and scoreControlColicP value
General lesion score 00%0%P = 0.026
General lesion score 10%0%
General lesion score 263.6%22.2%
General lesion score 336.4%77.8%
Eosinophilic cytoplasmic degeneration 018.2%11.1%P= 0.866
Eosinophilic cytoplasmic degeneration 136.3%27.8%
Eosinophilic cytoplasmic degeneration 227.3%38.9%
Eosinophilic cytoplasmic degeneration 318.2%22.2%
Inflammation 027.3%16.7%P= 0.378
Inflammation 145.4%27.7%
Inflammation 227.3%38.9%
Inflammation 30%16.7%
Pyknosis 09.1%5.6%P = 0.453
Pyknosis 154.5%27.8%
Pyknosis 227.3%44.4%
Pyknosis 39.1%22.2%
Ruptured myocytes 09.1%16.7%P = 0.635
Ruptured myocytes 154.5%33.3%
Ruptured myocytes 227.3%27.8%
Ruptured myocytes 39.1%22.2%
Wavy fibres 00%11.2%P = 0.463
Wavy fibres 118.2%22.2%
Wavy fibres 236.3%44.4%
Wavy fibres 345.5%22.2%
Vacuolar degeneration 154.5%5.6%P = 0.009
Vacuolar degeneration 227.3%38.9%
Vacuolar degeneration 318.2%55.5%
image

Figure 2b. The muscle layer shows moderate vacuolar degeneration (arrows), which extends into the formation of large vacuoles (arrowhead) at the level of the longitudinal muscle layer in cranially resected intestine after strangulation (haematoxylin-eosin staining, Scale bar = 50 µm, magnification: 600×).

Download figure to PowerPoint

There was no correlation between the muscle lesion score and the mucosal lesion score (P = 0.228).

The muscle lesion score of the samples was not different if samples were collected during surgery or after euthanasia (P = 0.688). Of the samples resected during surgery with muscle lesion score 3, 33.3% (n = 3) recovered, 22.2% (n = 2) developed ileus and 44.4% (n = 4) were subjected to euthanasia for other reasons (P = 0.240).

Myosin

Myosin-IR (immunoreactivity) was found in the cytoplasm of smooth muscle cells throughout the small intestine including vascular smooth muscle cells. In colic samples, the myosin protein content was significantly lower when compared to control tissue (control: 199.00 ± 35.34 vs. colic: 100.22 ± 15.181; P = 0.022) (Fig 3). The myosin-OD did not differ when taking into account the muscle lesion score (Score 2: 165.67 ± 36.93 vs. Score 3: 116.06 ± 17.08; P = 0.242). Myosin-OD was not related to the presence of eosinophilic cytoplasmic degeneration (P = 0.153), inflammatory cells (P = 0.800), pyknotic cells (P = 0.207), ruptured myocytes (P = 0.605), wavy fibres (P = 0.873) or vacuolar degeneration (P = 0.353).

image

Figure 3. Optical densities (OD) values for myosin, panactin, desmin, HSP27 and HSP20 obtained by western blot analysis for control and colic samples. Protein estimation was made. Myosin content significantly changed in colic samples (a vs. b) (P=0.022), panactin-OD and HSP27OD did not differ between control and colic samples (P=0.257 and P=0.115), desmin-OD is augmented in colic samples (c vs. d) (P=0.049) and HSP20 is significantly upregulated in colic samples (e vs. f) (P=0.005). Results were expressed as means ± s.e.m.

Download figure to PowerPoint

Myosin-OD of resected samples did not differ according to the outcome of the surgery (recovery: 62.41 ± 27.59, ileus 94.06 ± 14.62 and euthanasia for other reasons 102.21 ± 20.64, P = 0.541).

Actin

Panactin-IR was found in the cytoplasm of smooth muscle cells throughout the various layers of the small intestinal wall. Western blot analysis found no significant difference between the control and colic group (control: 126.16 ± 36.43 vs. colic: 80.53 ± 21.34; P = 0.257) (Fig 3). The panactin-OD did not differ when data were grouped according to the muscle lesion score (Score 2: 87.41 ± 28.66 vs. Score 3: 102.92 ± 25.74; P = 0.698).

Panactin-OD was not related to eosinophilic cytoplasmic degeneration (P = 0.896), inflammatory cells (P = 0.795), pyknotic cells (P = 0.386), rupture of the myocytes (P = 0.259), wavy myocytes (P = 0.758) and vacuolar degeneration (P = 0.479).

Panactin-OD was not related to the outcome of the surgery (recovery: 58.34 ± 14.60, ileus 67.47 ± 39.93 and euthanasia for other reasons 66.48 ± 8.50 [P = 0.910]).

Desmin

Desmin-IR showed the same staining pattern as myosin and panactin. Desmin-OD was higher in the colic samples compared to the control samples (colic: 131.93 ± 31.49 vs. control: 56.79 ± 21.07 P = 0.049) (Fig 3). It was, however, not related to the muscle lesion score (Score 2: 71.19 ± 23.53 vs. Score 3: 127.03 ± 33.10 P = 0.196).

Desmin was not related to the presence of eosinophilic cytoplasmic degeneration (P = 0.228), inflammatory cells (P = 0.806), pyknotic cells (P = 0.415), ruptured myocytes (P = 0.709), wavy smooth muscle cells (P = 0.410) or vacuolar degeneration (P = 0.128).

Desmin was also unrelated to outcome of the surgery (recovery: 129.63 ± 48.06, ileus 62.29 ± 7.99 and euthanasia for other reasons 115.43 ± 43.61, P = 0.566).

The desmin-OD was correlated to that of HSP20 (P = 0.008; r = 0.493).

HSP27

Expression of HSP27-IR showed a patchy distribution in the cytoplasm of the smooth muscle cells throughout the intestinal wall (Fig 4). It was often found close to the nucleus and a more intense staining pattern could be observed in conjunction with presence of histopathological signs such as rupture of the myocytes.

image

Figure 4. An overview of the immunohistochemical anti-HSP27 staining in a biopsy from the equine small intestine. The immunohistochemical localisation of HSP27 is found in all smooth muscle cells. It is found in the tunica muscularis (arrow), the lamina muscularis mucosae (black arrowhead) with extensions towards the mucosa and in the blood vessel wall (thick arrow) (haematoxylin counterstaining, Scale bar = 50 µm, magnification: 600×).

Download figure to PowerPoint

The OD values were not related to the occurrence of strangulation colic (control: 188.32 ± 30.58 vs. colic: 132.85 ± 18.56 P = 0.115) (Fig 3) and to the muscle lesion score (Score 2: 151.81 ± 16.95 vs. Score 3: 150.96 ± 25.79 P = 0.980).

Values for HSP27 were not related to the presence of eosinophilic cytoplasmic degeneration (P = 0.341), inflammatory cells (P = 0.162), pyknotic nuclei (P = 0.171), ruptured smooth muscle cells (P = 0.378), wavy myocytes (P = 0.418) and vacuolar degeneration (P = 0.350).

The HSP27-OD was not related to the outcome of the intestinal resection (recovery: 144.42 ± 23.88, ileus 93.64 ± 34.94 and euthanasia for other reasons 147.14 ± 21.85, P = 0.476). HSP27-OD was negatively correlated with HSP20-OD (P = 0.026; r = -0.428) and positively correlated with myosin-OD (P = 0.029; r = 0.421).

HSP20

Heat shock protein 20-IR was equally distributed in the smooth muscle cells throughout the small intestinal wall (Fig 5). The HSP20-OD was significantly upregulated in colic tissue compared to control samples (colic: 83.81 ± 30.79 vs. control: 139.66 ± 27.45; P = 0.005) (Fig 3). HSP20-OD did not depend on the general lesion score (Score 2: 107.97 ± 26.88 vs. Score 3: 127.31 ± 30.56; P = 0.906).

image

Figure 5. An overview of the immunohistochemical anti-HSP20 staining in a biopsy from the equine small intestine. The immunohistochemical localisation of HSP20 is found in all smooth muscle cells. It is found in the tunica muscularis (arrow), the lamina muscularis mucosae (black arrowhead) with extensions towards the mucosa (thick arrow) and the blood vessel wall (haematoxylin counterstaining, Scale bar = 50 µm, magnification: 600×).

Download figure to PowerPoint

Values for HSP20 were not related to the amount of eosinophilic cytoplasmic degeneration (P = 0.290), inflammatory cells (P = 0.381), pyknotic nuclei (P = 0.667), ruptured myocytes (P = 0.368), wavy myocytes (P = 0.621) and vacuolar degeneration (P = 0.202). HSP20-OD was not related to the outcome the intestinal resection (recovery: 113.14 ± 19.78, ileus 85.67 ± 10.81 and euthanasia for other reasons 116.78 ± 17.49, P = 0.690).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Manufacturers' addresses
  9. References

The gastrointestinal tract is one of the most susceptible organ systems to ischaemia.

Not only mucosal injury but also alterations of intestinal motility have been reported in response to ischaemia (Malone and Kannan 2001). However, there are few reports focusing on changes occurring in the enteric musculature. To our knowledge, studies documenting smooth muscle proteins have not previously been performed in ischaemic horse intestine.

In our study, all control and colic samples showed histological lesions in the muscle layer. The presence of lesions in control samples is intriguing. These samples were taken under the most ideal conditions, without important ischaemic challenge and following a protocol that has been applied in other equine studies (Oikawa et al. 2007). Nevertheless, colic samples showed significantly more degenerative changes at the level of the muscle layer than control tissue. This finding entails that despite the fact that the cranial borders of resected small intestinal segments are generally considered to be vital and macroscopically normal, there are microscopical signs of degeneration. Of all degenerative lesion types, abundant formation of vacuoles (vacuolar degeneration), was more prominent in colic samples (55.5%) in comparison to controls (18.2%). Small intestinal smooth muscle vacuolar degeneration has been described in ischaemia-reperfusion challenged small intestine (Ming and McNiff 1976), in conjunction with small intestinal dilatation (Cheng et al. 2001) and in experimentally-induced endotoxaemia in horses (Oikawa et al. 2007). In addition, 22.2% of the colic samples showed profound rupture of the smooth muscle fibres, which is the final event leading to cell death (Ganote and Armstrong 1993). Early degenerative changes of smooth muscle cells such as contraction resulting in wavy fibres occur in relation to ischaemia (Ming and McNiff 1976) and rigor mortis. The latter explains why these early degenerative changes were seen equally in colic and control tissue.

The myosin protein content was significantly lower in colic samples when compared to control tissue. However, no significant relationship could be demonstrated with lesion severity score, nor was there a correlation with any type of degeneration found in the intestinal smooth muscle. Since myosin is an important contractile filament, this finding opens interesting considerations on what effects its reduced expression could have on intestinal smooth muscle contractility. In human colorectal motility disorders, such as Hirschsprung's disease, idiopathic megacolon and slow-transit constipation, myosin-IR was reported to be significantly reduced, suggesting an involvement in the pathogenesis of these gastrointestinal motility disorders (Wedel et al. 2006). Hein et al. (1995) used myosin-IR as a marker for contractile vitality of cardiac muscle cells and reported a decreased myosin-IR expression in cases of human cardiac ischaemia (Hein et al. 1995). It has been suggested that ischaemia causes myosin and actin filaments to depolymerise, a process that is stimulated by proteolytic enzymes, which are known to be even more active in an acidic environment in cases of human cardiac ischaemia (Hein et al. 1995). Myosin heavy chain mRNA is also proven to be significantly decreased in experimentally induced ileum hypertrophy in guinea-pigs (Lofgren et al. 2002). These reported findings could support the suggestion that in our study both intestinal ischaemia and distention have contributed to an altered expression or a breakdown of myosin protein.

Desmin, an intermediate filament protein found in muscle cells only, plays an essential role in the maintenance of muscle cytoarchitecture by anchoring neighbouring Z discs. In our study, the level of desmin was inversely related to the level of myosin. Its expression was significantly more pronounced in colic samples when compared to control samples. A similar increased desmin expression was reported in experimentally induced ileum hypertrophy. Results of functional contractile experiments performed in this study suggest that structural protein alterations affected contractility leading to a slower and more economical function of the small intestine (Lofgren et al. 2002).

Both HSP27 and HSP20 are known to influence smooth muscle contractility (Batts et al. 2006; Somara and Bitar 2006). HSP27 upregulation after, e.g. ischaemia/reperfusion seems to be triggered via reactive oxygen species (ROS) (Ferns et al. 2006). In this study, however, the protein level of HSP27 was not significantly affected. A possible reason could be that ROS only massively appear after reperfusion, a process that has not taken place yet in our samples. Therefore, we postulate that strangulation triggers a molecular pathway leading to HSP20 upregulation and not affecting HSP27. HSP20 is constitutively expressed in smooth muscle cells (Gusev et al. 2005), which accords with our immunohistochemical results. The HSP20-OD was significantly upregulated in colic tissue compared to control samples but was not related to lesion score. Severe distension of bladder smooth muscle has been reported to upregulate HSP20 causing relaxation of the muscle layer (Batts et al. 2006). The exact mechanism is not yet known but it is suggested that a general relaxation is caused by force suppression in consequence of HSP20 binding to thin filaments. As a result, myosin filaments can no longer connect to actin filaments, leading to reduced contraction (Batts et al. 2006). The pathogenesis of post operative ileus is not yet precisely known. An immediate neurogenic cause in combination with leucocyte influx are put forward as the main, but not only, reasons. During strangulation, distention, ischaemia and congestion are stress factors occurring in the oral resection site. These can trigger upregulation of HSP20. The inhibitory effect of HSP20 combined with increased smooth muscle damage might provide a molecular approach to the development of ileus. However, both myosin-OD and HSP20-OD were not significantly different in horses developing ileus vs. those that recovered. However, group sizes (ileus vs. nonileus) were too small and unequally large to be able to draw unambiguous conclusions.

In conclusion, this study suggests a morphological basis for reduced contractility in the smooth muscle cells of horses suffering from small intestinal strangulation. The colic group is more prone to develop post operative ileus. Since HSP20 was already related to hypocontractility, ileus might be improved by pharmacological interventions aiming at inhibiting HSP20.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Manufacturers' addresses
  9. References

We would like to thank Katty Huybrechts and Gunther Vrolix for their help with the laboratory experiments. We would like to thank the medical biochemistry laboratory for their technical assistance. We would like to thank the Department of Internal Medicine and Clinical Biology of Large Animals of Ghent and Utrecht University for the aid in collecting the samples.

Manufacturers' addresses

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Manufacturers' addresses
  9. References

1 Sigma Aldrich, Saint Louis, Missouri, USA.

2 Merck KGaA, Darmstadt, Germany.

3 Abcam, Cambridge, UK.

4 Cell Signaling, Danvers/Boston, USA.

5 Dakocytomation, Glostrup, Denmark.

6 Santa Cruz Biotechnology Inc., Santa Cruz, California, USA.

7 Roche, Basel, Switzerland.

8 Pierce, Rockford, Illinois, USA.

9 Invitrogen, Carlsbad, California, USA.

10 Isogen Life Science, De Meern, The Netherlands.

11 Nonlinear Dynamics, Newcastle-upon-Tyne, UK.

12 SPSS Inc, Chicago, Illinois, USA.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Manufacturers' addresses
  9. References