Eosinophilic granulocytes are physiologic resident cells of the gastrointestinal lamina propria, but the precise function of these cells in resting or inflammatory conditions is not currently understood.1,2 Traditionally, functions of eosinophils have been associated primarily with immune-mediated or parasitic diseases. However, eosinophils are proinflammatory leukocytes with a wide range of possible functions and have the potential to sustain or augment the immune response, inflammatory reaction, and tissue repair.3 Activated eosinophils can elicit potent cytotoxic effects by generation of unstable oxygen radicals and by direct degranulation of mast cells and basophils.1,4 They can also generate large amounts of cytokines (including leukotriene C4, tumor necrosis factor, transforming growth factor, granulocyte-macrophage colony-stimulating factor, IL-4, IL-5, IL-8, IL-13, and platelet-activating factor), which indicates a wide range of possible effects on the inflammatory reaction and tissue repair. Finally, eosinophils can also initiate antigen-specific immune responses by acting as antigen-presenting cells.1 In diseased bowel of human patients with eosinophil-associated gastrointestinal diseases (including gastroesophageal reflux, eosinophilic gastroenteritis, and inflammatory bowel disease), eosinophils can be found in all layers of the tissue,5 and there is a correlation between degree of eosinophil accumulation and disease severity.6–9 In some diseases (eg, gastrointestinal hypersensitivity disorders), eosinophils are believed to be the principal effector cell, and they initiate and sustain tissue injury and disease pathogenesis in at least a subset of these disorders.9 However, when eosinophilic disorders are associated with other primary gastrointestinal disorders (eg, gastroesophageal reflux), eosinophils do not appear to degranulate, which suggests that they also may have a regulatory function on the inflammatory process or tissue repair, rather than directly promoting tissue injury.9
Eotaxin-1 is a chemokine considered to be responsible for eosinophil migration into the gastrointestinal tract, and its mRNA is strongly expressed in equine jejunum and colon.10 This could be the reason that physiologically normal equine gastrointestinal mucosa has large numbers of resident eosinophils. Histamine and substance P have induced activation, adherence, and migration of equine eosinophils during in vitro studies.11,12 Protein kinase C can be involved in regulating adherence and superoxide production of equine eosinophils.13 Histochemical evidence of superoxide generation in jejunal eosinophils has been reported for horses with experimentally induced ischemia and reperfusion.14 Furthermore, accumulation of eosinophils has been described for horses with experimentally induced acute colitis15 and experimentally induced ischemia-reperfusion injury.16 Focal and diffuse eosinophilic gastroenteritis have been reported as clinical conditions in horses; clinical signs include weight loss, diarrhea, signs of depression, acute and chronic colic, and hypoalbuminemia.17–22 The cause of eosinophil accumulation is unknown in these affected horses, but eosinophil accumulation has been attributed to immune-mediated processes, such as a food allergy or parasite infection.17–22
The mucosal accumulation and distribution of eosinophilic granulocytes throughout the gastrointestinal tract of horses free of gastrointestinal diseases have been described in another study23 conducted by our research group. The objective of the study reported here was to evaluate eosinophilic granulocytes in the mucosa of horses with experimentally induced intestinal ischemia and reperfusion and those with naturally occurring intestinal strangulation. We hypothesized that eosinophils would participate in the inflammatory response to these types of injury, and this participation would be evident by a change in the location of eosinophils within the mucosa and by an accumulation of eosinophils in the mucosal tissue layer.
Materials and Methods
Sample
Intestinal mucosal samples were collected from adult horses of both sexes and various breeds and body sizes that were admitted with signs of colic and underwent exploratory laparotomy at the University of Illinois (n = 35 horses) or North Carolina State University (6). Samples also were obtained from horses with experimentally induced ischemia and reperfusion; these horses were the subject of studies24–26 performed at the University of Florida (n = 33 horses) or North Carolina State University (6). The animal care and use committee at the respective university approved those studies.24–26 Control samples were harvested from horses with experimentally induced ischemia and reperfusion and from 13 horses of both sexes and various breeds and body sizes immediately after those horses were euthanized at the University of Illinois for reasons other than gastrointestinal tract disease. All samples were full-thickness samples of the antimesenteric side of the intestinal wall (except for samples from horses with naturally occurring large colon volvulus, from which only mucosal samples were obtained), and all samples were fixed in 4% formalin immediately after harvesting. For horses with naturally occurring disease, the number of samples equaled the number of horses, whereas for horses with experimentally induced conditions, 1 horse could provide 1 sample each for control, ischemia, and reperfusion groups.
Naturally occurring lesions included jejunal strangulation (n = 24 samples), jejunal distention (6), and large colon volvulus (11). Jejunal strangulation samples were obtained from 24 horses subjected to jejunal resection and anastomosis performed after a small intestinal strangulating lesion was diagnosed; tissues from the strangulated intestine were harvested immediately after resection was completed. Jejunal distension samples were obtained from 6 horses with naturally occurring strangulating lesions that necessitated jejunal resection. Tissues were harvested from the proximal resection margin representing distended but not strangulated bowel immediately after resection was completed. Large colon volvulus samples were obtained from 11 horses in which an enterotomy of the pelvic flexure was deemed necessary. A mucosal sample (approx 1 × 1 cm) was harvested at the enterotomy site immediately after the enterotomy was established.
Experimentally induced lesions were evaluated in tissues that included jejunum after 1 (n = 6 samples) or 2 (4) hours of ischemia and large colon from the pelvic flexure after 1 hour of ischemia (12), 1 hour of ischemia followed by 30 minutes of reperfusion (6), 2 hours of ischemia (15), or 2 hours of ischemia followed by 30 minutes of reperfusion (8). Jejunal samples were obtained from 6 horses24,26: each horse provided 1 control sample and 1 sample after 1 hour of ischemia, and 4 horses provided 1 sample after 2 hours of ischemia (2 samples were missing for that experiment). Samples of the large colon from the area of the pelvic flexure were obtained from 33 horses subjected to experimental ischemia and reperfusion,25 but several samples were missing for these experiments, which resulted in the following sample sizes: 12 for 1 hour of ischemia, 15 for 1 hour of ischemia and 30 minutes of reperfusion, 15 for 2 hours of ischemia, and 8 for 2 hours of ischemia and 30 minutes of reperfusion.
For control tissues, mucosal samples (19 samples of jejunum and 29 samples of the large colon from the pelvic flexure) were harvested from horses (6 horses for jejunum and 33 horses for colon) included in the aforementioned studies24–26 and from horses euthanized for reasons other than gastrointestinal tract disease (13 horses that provided 1 sample each for the control jejunum and control colon groups). All 52 horses had a history of regularly scheduled anthelmintic administration and no history of recent medications.
Histologic examination
Histologic examination was performed as previously described.23 Mucosal samples were fixed in 4% formalin immediately after harvesting, embedded in paraffin, and cut into 5-µm-thick sections. Slides were stained with Luna eosinophil stain27 and examined to determine eosinophil accumulation and distribution. For Luna eosinophil staining, slides were desiccated in xylene, stained with Biebrich scarlet–hematoxylin, differentiated in 1% acid alcohol, and subsequently stained with lithium carbonate. Eosinophil granules and erythrocytes were reddish-orange, Charcot-Leyden crystals were red, and all nuclear elements were blue. One investigator (AKR), who was unaware of the source of the tissues, performed all histologic examinations.
A computer-based programa was used for histomorphometric analysis of images obtained by use of light microscopy. The mucosa was divided into 5 regions as follows: mean distance from the muscularis mucosa to the luminal surface was divided into successive quarters for regions 1 (quartile closest to the muscularis mucosa) through 4 (quartile closest to the luminal surface), and the luminal surface of epithelial cells was region 5. Absolute numbers of eosinophils were counted in 3 hpfs,16,28 and the mean number of eosinophils per region was calculated. Each hpf was arbitrarily chosen for 1 tissue section (1 section/slide) on the basis of the presence of intact mucosa with a straight transection through villi or crypts (or both) in the entire hpf (ie, each hpf was obtained on different sections spaced several cuts apart within the tissue block, and the hpf considered most acceptable on the basis of the aforementioned criteria was selected). These mean numbers were then used to calculate the percentage of eosinophils in each of the 5 regions, relative to the total eosinophil count in all regions, for each horse. The area occupied by regions 1 to 4 was then measured, and the number of eosinophils per mm2 of mucosa was calculated as a measure of the degree of eosinophil density.
Statistical analysis
Data were expressed as least squares geometric mean ± SEM. Data that were not normally distributed were logarithmically transformed or ranked before repeated-measures ANOVA was performed by use of a compound symmetry covariance matrix. Whenever indicated by a significant F test for group, least squares mean values for each group were compared with least squares mean values for the control group by use of a Fisher protected least significant difference test. A statistical software programb was used for analysis. Values were considered significant at P < 0.05.
Results
Jejunal strangulation
Number of eosinophils per mm2 of intestinal mucosa was significantly (P = 0.007) lower in jejunum with strangulating lesions (24 samples) than in the control jejunum (19 samples; Table 1). Migration of eosinophils toward the luminal surface was evident in strangulating lesions of the small intestine. There was a significantly (P = 0.002) lower percentage of total eosinophils in region 1 of the mucosa of jejunum with strangulating lesions than in control jejunum. There was a significantly (P < 0.001) higher percentage of total eosinophils on the luminal surface in region 5 of jejunum with strangulating lesions than in control jejunum.
Eosinophil distribution (expressed as the percentage for each region, relative to the total eosinophil count for all regions) and accumulation in samples of equine jejunal mucosa.
| Region* | |||||||
|---|---|---|---|---|---|---|---|
| Group | No. of samples | 1 | 2 | 3 | 4 | 5 | Eosinophils/mm2 of jejunal mucosa |
| Control | 19 | 65.0 | 20.3 | 1.1 | 0 | 0 | 168 |
| (33.3–100) | (0–66.7) | (0–25.0) | (0–3.0) | (0–0) | (97–290) | ||
| 1 h of ischemia | 6 | 59.5 | 30.3 | 1.4 | 1.0 | 1.0 | 177 |
| (31.8–100) | (13.0–70.3) | (0.8–2.5) | (0.7–1.5) | (0.6–1.7) | (67–467) | ||
| 2 h of ischemia | 4 | 69.1 | 20.0 | 0.2 | 0.1 | 0 | 152 |
| (41.0–100) | (0–51.3) | (0–5.1) | (0–2.6) | (0–0) | (46–498) | ||
| Strangulation | 24 | 34.2† | 20.7 | 0.5 | 0.3 | 1.3† | 61† |
| (0–100) | (0–100) | (0–40.0) | (0–81.0) | (0–73.3) | (37–98) | ||
| Distention | 6 | 85.5 | 1.9† | 0 | 0 | 0 | 7† |
| (50.0–100) | (0–50.0) | (0–0) | (0–0) | (0–0) | (2–17) | ||
Data reported are least squares geometric mean (95% confidence interval). Within a row, values for distribution of each group do not sum to 100% because geometric means were calculated.
The mucosa was categorized into 5 regions as follows: mean distance from the muscularis mucosa to the luminal surface was divided into successive quarters for regions 1 (quartile closest to the muscularis mucosa) through 4 (quartile closest to the luminal surface), and the luminal surface of epithelial cells was region 5.
Within a column, value differs significantly (P < 0.05) from the value for the control group.
Jejunal distention
Number of eosinophils per mm2 of intestinal mucosa was significantly (P < 0.001) lower in jejunum with jejunal distention (6 samples) than in control jejunum (19 samples; Table 1). Mucosal distribution of eosinophils differed in distended jejunum, compared with the distribution in control jejunum, with a significantly (P < 0.001) lower eosinophil percentage in region 2 of distended jejunum.
Experimentally induced jejunal ischemia
Intestinal ischemia for 1 hour (6 samples) or 2 hours (4 samples) did not change the eosinophil count in the jejunal mucosa (Table 1). There was no difference in the eosinophil distribution in jejunal mucosa between jejunum with experimentally induced ischemia and control jejunum (19 samples).
Colon from horses with naturally developing large colon volvulus
Total mucosal eosinophil count for colon samples of large colon volvulus (11 samples) was similar to that of control colon (29 samples; Table 2). Significantly more eosinophils were seen on the luminal surface (region 5; P = 0.19) and in region 4 (P = 0.011) of colon samples after large colon volvulus than in control colon samples. In addition, significantly (P = 0.012) fewer eosinophils remained near the muscularis mucosae (region 1) in colon samples of large colon volvulus than in control colon samples.
Eosinophil distribution (expressed as the percentage for each region, relative to the total eosinophil count for all regions) and accumulation in samples of equine colonic (area of the pelvic flexure) mucosa.
| Region* | |||||||
|---|---|---|---|---|---|---|---|
| Group | No. of samples | 1 | 2 | 3 | 4 | 5 | Eosinophils/mm2 of colonic mucosa |
| Control | 29 | 50.3 (0–87.6) | 29.6 (8.3–64.7) | 8.8 (0–38.5) | 1.5 (0–19.2) | 0 (0–6.1) | 428 (330–550) |
| 1 h of ischemia | 12 | 85.6 (38.8–93.3) | 26.9 (5.0–50.0) | 6.1 (0–22.1) | 2.0 (0–7.7) | 0.2 (0–2.2) | 368† (279–486) |
| 1 h of ischemia followed by 30 min of reperfusion | 6 | 49.3 (33.7–62.6) | 33.1 (24.7–45.9) | 10.5 (2.3–18.5) | 3.0† (0–12.1) | 0.7† (0–3.2) | 456 (337–615) |
| 2 h of ischemia | 15 | 48.3 (16.7–88.4) | 29.8 (7.0–55.8) | 9.2 (0–30.0) | 3.9† (0–27.3) | 0.5† (0–4.7) | 362† (275–476) |
| 2 h of ischemia followed by 30 min of reperfusion | 8 | 43.5 (19.2–63.5) | 28.7 (14.7–51.9) | 10.6 (2.2–26.5) | 4.6† (0–38.4) | 2.2† (0–19.2) | 378 (282–505) |
| Large colon volvulus | 11 | 30.7* (0–85.7) | 25.3 (0–50) | 11.7 (0–47.1) | 3.7† (0–50) | 0.5† (0–100) | 348 (229–531) |
See Table 1 for key.
Colon from horses with experimentally induced ischemia and reperfusion
Experimentally induced ischemia for 1 hour (12 samples) significantly (P = 0.031) decreased the total eosinophil count, compared with results for the control samples, but did not alter eosinophil distribution (Table 2). Experimentally induced ischemia for 2 hours (15 samples) also significantly (P = 0.011) decreased the total eosinophil count, compared with results for the control samples, and more eosinophils were present near the luminal surface (region 4) and on the luminal surface (region 5). In contrast, reperfusion did not change the absolute numbers of eosinophils but did cause a migration of eosinophils toward and onto the luminal side of the mucosa. Compared with eosinophil counts for regions in the control samples, significantly more eosinophils were seen in region 4 and on the luminal surface (region 5) in samples after 1 hour of ischemia followed by 30 minutes of reperfusion (6 samples; P = 0.009 and P = 0.001, respectively) and after 2 hours of ischemia (8 samples; P < 0.001 for both regions).
Discussion
The study reported here was focused on eosinophil accumulation and distribution in jejunal and colonic tissues. Other histologic evidence of mucosal damage was not evaluated in this study. For the samples from the experimentally induced ischemia and control groups and for horses with jejunal distention, evaluation of histologic damage was performed as described in the respective studies24–26 and therefore was not repeated for the present study. In the study25 of horses with experimentally induced colonic ischemia and reperfusion, an increase of the interstitium-to-crypt ratio and hemorrhage score after ischemia and after ischemia and reperfusion was used as an indication of edema formation and bleeding into the lamina propria. For horses with naturally occurring large and small intestinal strangulation, mucosal damage appeared similar to that described in the literature18 (ie, mucosal hemorrhage and edema, epithelial elevation, and sloughing) but was not specifically evaluated in the present study.
Experimentally induced ischemia had no effects on eosinophil accumulation or distribution in the jejunal mucosa in the present study, even though samples from the same ischemic-injured jejunal mucosa had severe histologic and electrophysiologic alterations when evaluated in other studies.24–26 In contrast, experimentally induced ischemia decreased the number of eosinophils per mm2 of mucosa in colonic mucosa, and 2 hours of ischemia resulted in a larger percentage of eosinophils near the luminal surface (region 4) and on the luminal surface (region 5). The latter finding is consistent with that of another study25 in which investigators identified activation of eosinophils in equine large colon during experimentally induced ischemia, which suggests that eosinophils could be involved in tissue changes during ischemia of the colon.
Reperfusion of ischemic large colon (pelvic flexure) resulted in migration of resident mucosal eosinophils toward and onto the luminal surface but did not change the total eosinophil count. Eosinophilic migration appeared more pronounced after 2 hours of ischemia followed by reperfusion, compared with results after 1 hour of ischemia followed by reperfusion, although this was not compared statistically. This corresponds to findings for a model of low-flow ischemia and reperfusion injury of the equine large colon,16 whereby accumulation of eosinophils was only significantly different after prolonged (3.25 and 4 hours) low-flow ischemia and after reperfusion. Similarly, in horses with naturally occurring strangulating lesions, migration of eosinophils was evident in both the jejunum and colon. It is possible that some degree of reperfusion amplified a mucosal eosinophilic migratory response. Clinical strangulating lesions usually have a certain amount of blood flow that remains, and strangulation usually is relieved before mucosal samples can be collected, which allows for a period of reperfusion. This would correspond with the similar eosinophilic response seen in the horses after experimentally induced ischemia and reperfusion of the colon and might explain the reason for an eosinophilic response in clinically affected horses.
The most prominent change during reperfusion is in the availability of oxygen to previously ischemic tissues and subsequently the production of oxygen-derived free radicals. Intracellular reactive oxygen species are involved in the signaling pathway leading to cAMP production in human eosinophils.29 The exact mechanisms by which oxygen-derived free radicals might induce a mucosal eosinophilic response remain unclear, but the hypothesis that they activate eosinophils to migrate is supported by findings from another study28 in which in vitro exposure to hypochlorous acid induced eosinophil migration toward and onto the luminal surface of equine colonic mucosa. Migration of eosinophils in that in vitro study28 was clearly evident; in particular, a higher number of eosinophils was detected on the luminal surface (region 5). Eosinophils and other cells on the surface layer could have been removed from the luminal surface by the passage of intestinal contents or might have been inadvertently dislodged by cleaning ingesta from the mucosa before fixation in formalin.
Migration of eosinophils to the mucosal surface could represent a mechanism to regulate eosinophil activity during inflammation.25 In experimentally induced ischemia and reperfusion of the large colon, increased production of nitrotyrosine has been detected in resident mucosal eosinophils, and nitrotyrosine is regarded as a marker for the generation of reactive nitrogen species that are indicative of oxidative stress.25 Nitrotyrosine can form after exposure of tyrosine residues to peroxynitrite.30 During reperfusion, oxygen-derived free radicals such as nitric oxide and superoxide are formed by many cells, including neutrophils and macrophages. Superoxide and nitric oxide can react to produce peroxynitrite,31 and peroxynitrite modulates eosinophil migration through effects of eotaxin in human eosinophils in vitro.30 The authors of that study30 suggested that oxidants may play an important role in regulation of eotaxin-induced eosinophil chemotaxis.
Eosinophils involved in the eosinophilic response noticed in the present study appeared to be resident mucosal eosinophils, with little or no migration of submucosal eosinophils into the mucosa based on the fact that the absolute number of eosinophils in the lamina propria of the mucosa was minimally changed in tissues after experimentally induced ischemia and reperfusion or naturally occurring strangulation in colonic mucosa. A change in absolute eosinophil count in the present study was attributable to a reduction of eosinophils in jejunum after naturally occurring distention and strangulation. Similarly, numbers of eosinophils were lower in colon (pelvic flexure) samples of horses with large colon volvulus, even though this change was not significant. Reduction of the eosinophil count per mm2 of lamina propria could be explained by the prominent edema and hemorrhage in these tissues, although variables to quantify edema (eg, interstitium-to-crypt ratio) were not evaluated. Another possible explanation is the loss of eosinophils from the lamina propria into the lumen in severely damaged tissues. The submucosa contains large numbers of resident eosinophils, and these were not quantitatively evaluated in the study reported here. Submucosal eosinophils could replace mucosal eosinophils that are lost into the lumen of the bowel. The reduction in the total number of eosinophils in the mucosal samples from distended jejunum was most likely a result of edema formation because migration of eosinophils was not observed in this tissue group.
We did not evaluate histologic changes, except for eosinophil migration or accumulation, in the mucosal samples of the present study. However, tissues did have histologic changes, including hemorrhage, severe epithelial denudation, and edema formation, that have previously been described14 for intestinal mucosa after ischemia and reperfusion. Because the investigator was not aware of the source of the tissues during histologic evaluation in the present study, no statement can be made concerning the severity of these changes in each of the groups. Samples from clinically affected horses probably had variations in the severity of histologic changes.
It is currently unclear whether eosinophilic granulocytes are detrimental or protective in tissues damaged by ischemia and reperfusion. Although not systematically or quantitatively evaluated, eosinophils appeared to degranulate in tissues after reperfusion and after strangulation, and degranulation of eosinophils has been associated with detrimental effects to surrounding tissues.9 Also, histochemical evidence of superoxide generation in submucosal jejunal eosinophils has been detected after 60 and 120 minutes of reperfusion in experiments with in vivo–induced jejunal ischemia and reperfusion.14 These findings would suggest that eosinophils could have a detrimental effect on equine intestinal mucosa after ischemia and reperfusion.
In the study reported here, eosinophils responded to mucosal damage evoked by ischemia and reperfusion by migration toward and onto the luminal surface during reperfusion. Functions of the potent inflammatory eosinophilic granulocytes in this disease complex should be further evaluated.
Acknowledgments
The authors thank Dr. Anthony Blikslager for providing tissue samples.
ABBREVIATIONS
| IL | Interleukin |
Footnotes
Image Pro Express 4.5, Media Cybernetics Inc, Rockville, Md.
PROC MIXED, SAS, version 9.3, SAS Institute Inc, Cary, NC.
References
1. Rothenberg ME, Mishra A, Brandt EB, et al. Gastrointestinal eosinophils in health and disease. Adv Immunol 2001; 78: 291–328.
2. Strauman A, Simon H-U. The physiological and pathophysiological roles of eosinophils in the gastrointestinal tract. Allergy 2004; 59: 15–25.
3. Kita H. The eosinophil: a cytokine-producing cell? J Allergy Clin Immunol 1996; 97: 889–892.
4. Rothenberg ME. Eosinophilia. N Engl J Med 1998; 338: 1592–1600.
5. Kelly KJ. Eosinophilic gastroenteritis. J Pediatr Gastroenterol Nutr 2000; 30 (suppl): S28–S35.
6. Desreumaux P, Bloget F, Seguy D, et al. Interleukin 3, granulocyte-macrophage colony-stimulating factor, and interleukin 5 in eosinophilic gastroenteritis. Gastroenterology 1996; 110: 768–774.
7. Talley NJ, Shorter RG, Phillips SF, et al. Eosinophilic gastroenteritis: a clinicopathological study of patients with disease of the mucosa, muscle layer, and subserosal tissues. Gut 1990; 31: 54–58.
8. Torpier G, Colombel JF, Mathieu-Chandelier C, et al. Eosinophilic gastroenteritis: ultrastructural evidence for a selective release of eosinophil major basic protein. Clin Exp Immunol 1988; 74: 404–408.
9. Hogan SP, Rothenberg ME. Review article: the eosinophil as a therapeutic target in gastrointestinal disease. Aliment Pharmacol Ther 2004; 20: 1231–1240.
10. Benarafa C, Cunningham FM, Hamblin AS, et al. Cloning of equine chemokines eotaxin, monocyte chemoattractant protein (MCP)-1, MCP-2 and MCP-4, mRNA expression in tissues and induction by IL-4 in dermal fibroblasts. Vet Immunol Immunopathol 2000; 76: 283–298.
11. Foster AP, Cunningham FM. Histamine-induced adherence and migration of equine eosinophils. Am J Vet Res 1998; 59: 1153–1159.
12. Foster AP, Cunningham FM. Substance P induces activation, adherence and migration of equine eosinophils. J Vet Pharmacol Ther 2003; 26: 131–138.
13. Sepulveda MF, Greenaway EC, Avella M, et al. The role of protein kinase C in regulating equine eosinophil adherence and superoxide production. Inflamm Res 2005; 54: 97–105.
14. Kooreman K, Babbs C, Fessler J. Effect of ischemia and reperfusion on oxidative processes in the large colon and jejunum of horses. Am J Vet Res 1998; 59: 340–346.
15. McConnico RS, Weinstock D, Poston ME, et al. Myeloperoxidase activity of the large intestine in an equine model of acute colitis. Am J Vet Res 1999; 60: 807–813.
16. Moore RM, Bertone AL, Bailey MQ, et al. Neutrophil accumulation in the large colon of horses during low-flow ischemia and reperfusion. Am J Vet Res 1994; 55: 1454–1463.
17. Stanar LS, Little D, Redding WR, et al. Idiopathic eosinophilic enteritis in a 10-week-old colt. Compend Contin Educ Pract Vet 2002; 24: 342–347.
18. Edwards GB, Kelly DF, Proudman CJ. Segmental eosinophilic colitis: a review of 22 cases. Equine Vet J Suppl 2000;(32):86–93.
19. Southwood LL, Kawcak CE, Trotter GW, et al. Idiopathic focal eosinophilic enteritis associated with small intestinal obstruction in 6 horses. Vet Surg 2000; 29: 415–419.
20. French RA, Meier WA, Zachary JF. Eosinophilic colitis and hepatitis in a horse with colonic intramucosal ciliated protozoa. Vet Pathol 1996; 33: 235–238.
21. van der Kolk JH, Grinwis GCM. Een paard met een chronische eosinofiele enteritis. Tijdschr Diergeneeskd 1995; 120: 680–682.
22. Morton LD, Morton DG, Baker GJ, et al. Chronic eosinophilic enteritis attributed to Phytium sp. in a horse. Vet Pathol 1991; 28: 542–544.
23. Roetting AK, Freeman DE, Constable PD, et al. Mucosal distribution of eosinophilic granulocytes within the gastrointestinal tract of horses. Am J Vet Res 2008; 69: 874–879.
24. Tomlinson JE, Wilder BO, Young KM, et al. Effects of flunixin meglumine or etodolac treatment on mucosal recovery of equine jejunum after ischemia. Am J Vet Res 2004; 65: 761–769.
25. Grosche A, Freeman DE, Morton AJ, et al. Effects of ischemia and reperfusion on production or nitrotyrosine, activation of eosinophils, and apoptosis in the large colonic mucosa of horses. Am J Vet Res 2012; 73: 53–61.
26. Little D, Tomlinson JE, Blikslager AT. Post operative neutrophilic inflammation in equine small intestine after manipulation and ischemia. Equine Vet J 2005; 37: 329–335.
27. Luna LG. In: Histopathologic methods and color atlas of special stains and tissue artifacts. Gaithersburg, Md: American Histolabs Inc Publications Division, 1992 255–256.
28. Roetting AK, Freeman DE, Eurell JAC, et al. Effects of acetylcysteine and migration of resident eosinophils in an in vitro model of mucosal injury and restitution in equine right dorsal colon. Am J Vet Res 2003; 64: 1205–1212.
29. Ezeamuzie CI, Taslim N. Reactive oxygen species mediate phorbol ester-stimulated cAMP response in human eosinophils. Eur J Pharmacol 2006; 543: 174–180.
30. Sato E, Simpson KL, Grisham MB, et al. Effects of reactive oxygen and nitrogen metabolites on eotaxin-induced eosinophil chemotactic activity in vitro. Am J Respir Cell Mol Biol 2000; 22: 61–67.
31. Huie RE, Padmaja S. The reaction of NO with superoxide. Free Radic Res Commun 1993; 18: 195–199.
