induced by phenylbutazone in the right dorsal colon of ponies. Ruth-Anne Richter, DVM, MS; David E. Freeman, MVB, PhD; Matthew Wallig, DVM, PhD;.
01-10-0294R.qxd
6/10/2002
1:15 PM
Page 934
In vitro anion transport alterations and apoptosis induced by phenylbutazone in the right dorsal colon of ponies Ruth-Anne Richter, DVM, MS; David E. Freeman, MVB, PhD; Matthew Wallig, DVM, PhD; Ted Whittem, BVSc, PhD; Gordon J. Baker, BVSc, PhD
Objectives—To study the functional and structural responses of the right dorsal colon (RDC) of ponies to phenylbutazone (PBZ) in vitro at a concentration that could be achieved in vivo. Animals—8 adult ponies. Procedure—Short circuit current and conductance were measured in mucosa from the RDC. Tissues incubated with and without HCO3– were exposed to PBZ, bumetanide, or indomethacin. Bidirectional Cl– fluxes were determined. After a baseline flux period, prostaglandin E2 (PGE2) was added to the serosal surfaces and a second flux period followed. Light and transmission electron microscopy were performed. Results—Baseline short circuit current was diminished significantly by PBZ and indomethacin, and increased significantly after addictions of PGE2. After PGE2 was added, Cl– secretion increased significantly in tissues in HCO3–-free solutions and solutions with anti-inflammatory drugs, compared with corresponding baseline measurements and with control tissues exposed to PGE2. Bumetanide did not affect baseline short circuit current and Cl– fluxes. The predominant histologic change was apoptosis of surface epithelial cells treated with PBZ and to a lesser extent in those treated with indomethacin. Conclusions and Clinical Relevance—Prostaglandin-induced Cl– secretion appeared to involve a transporter that might also secrete HCO3–. Both PBZ and indomethacin altered ion transport in RDC and caused apoptosis; PBZ can damage mucosa through a mechanism that could be important in vivo. The clinically harmful effect of PBZ on equine RDC in vivo could be mediated through its effects on Cl– and HCO3– secretion. (Am J Vet Res 2002;220:934–941)
P
henylbutazone (PBZ) is widely used for the treatment of lameness in horses, but concerns about its toxicity arose from studies conducted in 19791 and subsequently. Pathologic changes induced by PBZ in Received Oct 22, 2001. Accepted Mar 1, 2002. From the Departments of Veterinary Clinical Medicine (Richter, Freeman, Baker), Veterinary Pathobiology (Wallig), and Veterinary Biosciences (Whittem), College of Veterinary Medicine, University of Illinois, Urbana, IL 61802. Dr. Richter’s present address is Reid and Associates, 1630 F Rd, Loxahatchee, FL 33470. Dr. Whittem’s present address is 38 Austral St, Nelson Bay, NSW, Australia. Supported in part by the USDA Hatch Funds, Illinois Walking Horse Association Funds, and the Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois. This report represents a portion of a dissertation submitted by the senior author to the graduate faculty as partial fulfillment of the requirements for the MS degree. Address correspondence to Dr. Freeman. 934
horses are gastric ulceration, edema of the small intestine, edema, erosions and ulcers in the cecum and ventral and dorsal colons, rare venous thrombosis in the large colon, renal crest necrosis,2,3 right dorsal colitis,4 and, possibly, impaction of the large colon.5 Right dorsal colitis is an insidious disease attributed to PBZ toxicity and is characterized by diffuse and superficial epithelial necrosis with scattered erosions and ulcerations.4 Progression of this lesion to a chronic form can cause death from colonic stenosis, luminal impaction, and colonic rupture.4 Because the toxic effects of PBZ do not always correlate closely with serum concentrations,6 there might be considerable individual variation with regard to PBZ toxicity. Although slow release from binding sites in the right dorsal colon (RDC) could explain predisposition of this site to ulceration,7 direct luminal exposure is not required for gastrointestinal tract toxicosis caused by nonsteroidal anti-inflammatory drugs (NSAID).8 Toxicity of NSAID for the gastrointestinal tract has long been recognized in humans, with an estimated 16,500 NSAID-related deaths every year in the United States.9 However, despite importance of NSAID toxicity and widespread use of these drugs, the mechanism by which gastrointestinal tract damage is inflicted is not well understood. It is apparent that the topical irritant properties and ulcerogenic effects of these drugs are less important than their ability to inhibit cyclooxygenase.8 The NSAID-induced enteropathy in the human small intestine is possibly triggered by partial uncoupling of oxidative phosphorylation followed by mitochondrial swelling in enterocytes,10,11,a direct epithelial damage, collapse of the cytoskeleton, disruption of intercellular tight junctions, increased permeability, and neutrophil infiltration.8 The process is exacerbated by repeated exposure to the NSAID, achieved through its enterohepatic circulation.8 There is evidence that NSAID can cause apoptosis and inflammation in the colon of humans with inflammatory bowel disease.12 Although an inducible form of the cyclooxygenase enzyme, COX-2, is responsible for many manifestations of inflammation and is the target of NSAID treatment, PBZ and other drugs also inhibit the constitutive cyclooxygenase, COX-1 (prostaglandin endoperoxide synthase, EC 1.14.99.1), which is responsible for several normal cellular functions throughout the gastrointestinal tract.8 In the equine gastrointestinal tract in vitro, prostaglandins are involved in intestinal fluid and electrolyte transport in the right ventral colon (RVC),13 ion secretion induced by inflammatory mediAJVR, Vol 63, No. 7, July 2002
01-10-0294R.qxd
6/10/2002
1:15 PM
Page 935
ators in the colon,b and repair of ischemic mucosa in the jejunum.14 The purpose of the study reported here was to examine the effects of PBZ on mucosal integrity and anion transport systems in the RDC in vitro. Our hypothesis was that mucosal integrity and anion transport systems in the RDC are controlled by prostaglandins, and can therefore be altered by NSAID with special emphasis on PBZ. Chloride and HCO3– were the anions of interest, because loss of important transport systems for these anions can induce clinical and histologic changes15,16 of possible relevance to colonic diseases in horses. Materials and Methods Horses and tissue preparation—The following procedures were approved by the Institutional Animal Care and Use Committee of the University of Illinois. Eight healthy adult ponies, fed hay twice daily and allowed ad libitum water, were euthanatized by administration of pentobarbital sodium at 88 mg/kg, IV, during the morning digestive period. A full-thickness segment was removed immediately from the lateral aspect of the RDC, immediately proximal to its attachment to the base of the cecum, and rinsed of intestinal contents. The tissue was divided into segments, and each segment was transported to the laboratory in a cold (4 C) solution of the same composition and with the same additions as used in the baseline experimental conditions. Each tissue segment was pinned on a rubber surface with the mucosal surface oriented upwards and submerged in a fresh batch of the same solution as for the experiments, at 20 to 22 C (room temperature). While the solution was constantly perfused with the appropriate gas for the bathing solution, sharp dissection was used to remove mucosal sheets that were mounted in Ussing chambers. These chambers had an aperture of 1.13 cm,2 and a solution volume of 7 ml bathed each tissue surface. The bicarbonate bathing solution was Krebs-Ringer-bicarbonate (KRB), which contained 112 mM NaCl, 25 mM NaHCO3, 10 mM glucose, 5 mM KCl, 3 mM Na acetate, 3 mM Na butyrate, 2.5 mM CaCl2, 1.2 mM MgSO4, 1.2 mM KH2PO4, and 0.01 mM mannitol. This solution was maintained at a pH of 7.4 by constant perfusion with 95% O2 and 5% CO2 at 37 C by circulating it with a gas lift through water-jacketed reservoirs. To determine the importance of HCO3– transport in the RDC under experimental conditions, HCO3– was replaced by 5 mM HEPES and 20 mM sodium isethionate. The solution was titrated to pH 7.4 with 1N HCl and perfused with 100% O2. Throughout the experimental period, the short circuit current (SCC) was recorded on voltage clamps through AgAgCl2 electrodes connected to 4% agar bridges of the same solution as in the relevant chamber. The ends of the bridges were placed to within 1 to 2 mm of the tissue surface. The potential difference (PD) between the calomel electrodes, when the tissue was not in place, was nullified by an offsetting voltage, and the transepithelial PD, with the tissue mounted, was recorded with a potentiometer.c The transepithelial PD was automatically and continuously nullified by the passage of an external current (ie, the SCC) across the tissue by use of an automatic voltage-current clamp amplifier,c except for 10-second periods at 15-minute intervals when spontaneous tissue PD was measured. The transmural electrical conductance was calculated as the inverse of resistance, which was determined by Ohm’s law as follows: PD = SCC X R
where R is resistance and spontaneous PD and SCC are used. The effect of the fluid resistance on the SCC was autoAJVR, Vol 63, No. 7, July 2002
matically corrected.c Conductance in millisiemens per centimeter2 was used as a measure of integrity of the colonic mucosa and permeability of the paracellular pathway. Experimental design and flux measurements—In addition to control tissues (KRB, no additions), tissues with and without HCO3– were selectively exposed to 3.24 X 10–6M (1µg/ml) PBZ,d 10-6M indomethacind or 10–4M bumetanide,d a potent inhibitor of the Na+-K+-2 Cl– (NKCC1) cotransporter. These treatments were applied to paired tissues to measure the electrical responses (joules [J])in duplicate and to determine bidirectional fluxes of Cl– in response to each treatment, such as flux from mucosa to opposite surface (mucosal to serosal flux [J m-s]) in 1 member of a pair and opposing flux (serosal to mucosal flux [J s-m]) in the other.17 Bidirectional Cl– fluxes were determined with chlorine 36e by use of tissues paired by matching conductances (< 25% difference). The tissues were allowed to stabilize at least 30 minutes before approximately 2 µCi of 36Cl was added to either the mucosal or serosal reservoir (ie, the hotside). The initial flux period was started at least 30 minutes after isotope addition. A single 250-µl aliquot of the hot-side reservoir was withdrawn immediately before the first flux period and placed in a liquid scintillation cocktail for counting. Two 45-minute flux periods were used, and duplicate 250-µl aliquots were withdrawn at the beginning and end of each period for scintillation counting. After 75 minutes, 10–6M prostaglandin E2 (PGE2)d was added to the serosal surfaces, and an interval of 30 minutes was allowed for equilibration of the 36Cl between the tissue spaces and bathing medium. A sample was then taken for liquid scintillation counting as for the first flux period, and the second flux period of 45 minutes followed. Measurements of SCC and PD were taken at 5-minute intervals for 15 minutes after the PGE2 addition because early responses to PGE2 were rapid and easily missed during longer intervals. Unidirectional fluxes of Cl– were calculated by the formula of Schultz and Zalusky,17 and net ion fluxes were calculated as follows: J net = J m-s – J s-m
Histologic examination—At the conclusion of the experiment, tissues were fixed in Karnovsky fixative for light and electron microscopy. This fixative was composed of 10% paraformaldehyde, 50% glutaraldehyde, calcium chloride, and cacodylate. Sections of tissues for light microscopy were 0.5-µm thick and were stained with toluidine blue. For electron microscopy, tissues were sectioned to a thickness of 70 nm. Histologic sections were evaluated by observers who were unaware of treatment groups. Statistical analyses—Changes in Cl– fluxes and SCC were compared for each incubation condition before (baseline) and after addition of PGE2 by use of a paired t test. Responses between different incubation conditions were compared under baseline conditions and after addition of PGE2 by use of ANOVA. The Student-Newman-Keuls test was used as the post hoc procedure to detect differences between means. The Kruskal-Wallis test was used for flux data with unequal variances. χ2 Analysis was used to compare the effects of treatments on number of apoptotic bodies per 100 crypt cells, based on counts obtained by light microscopic examination. A value of P < 0.05 was considered significant for all analyses.
Results Short circuit current— Short circuit current in the PBZ- and indomethacin-treated tissues, with and without HCO3–, was significantly lower than that recorded for other treatments during the initial flux or baseline 935
01-10-0294R.qxd
6/10/2002
1:15 PM
Page 936
flux period. The SCC increased significantly over baseline current in response to PGE2 in control tissues and in tissues incubated with bumetanide, indomethacin, and PBZ (Fig 1), with and without HCO3– for indomethacin and PBZ (Table 1). The effect of the NSAID on baseline current was less pronounced in a HCO3–-free medium, compared with the response to either NSAID in the presence of HCO3–, but this difference was not significant. None of the treatments affected conductance. Chloride fluxes—Chloride fluxes from mucosa to serosa were similar under all experimental conditions
Figure 1 —Short circuit current response (Y axis; mean ± SEM) in mucosal tissues of the right dorsal colon of 8 ponies, incubated with phenylbutazone during baseline recording (Flux period 1 [Baseline]) and after prostaglandin E2 (PGE2) was added at 75 minutes (Flux period 2).
before (baseline) and after addition of PGE2 to the serosal solutions (Table 1). Chloride fluxes from serosa to mucosa were also unaffected by different incubation solutions before addition of PGE2. The serosa to mucosa flux was significantly increased by PGE2 in tissues incubated with all HCO3–-free solutions and in solutions containing indomethacin and PBZ. Serosal PGE2 also increased serosa to mucosa flux in all tissues, compared with control tissues, and in those incubated with indomethacin, compared with all other treatments. Bidirectional fluxes of Cl– resulted in a net absorption under all experimental conditions, except when PGE2 was added to tissues incubated with indomethacin, which caused a net secretion. The net flux after PGE2 was significantly different from baseline net flux in tissues treated with PBZ or indomethacin. Light microscopy—Compared with control tissues, tissues treated with PBZ had paler epithelial cells and a generalized decrease in cellularity throughout the lamina propria. Intercellular spaces on the mucosal surface and in crypts and the lamina propria were expanded by fluid accumulation. Apoptotic bodies and cells undergoing oncotic necrosis were evident at the surface extrusion zones, which were points approximately midway between crypts. Scattered on the luminal surface was a mucus layer containing bacteria and sloughed epithelial cells in various stages of degeneration. The nuclei of some of the sloughed cells had uniform nuclear chromatin condensation into crescentic masses along the intact nuclear membranes, consistent with apoptosis, whereas others had irregular chromatin condensation typical of oncotic necrosis. Cytoplasmic vac-
Table 1—Chloride fluxes and short circuit current (SCC) in mucosal tissues from the right dorsal colon of ponies Treatment Control Flux period 1 (baseline) Flux period 2 (+ PGE2) Bumetanide (Bumet) Flux period 1 (baseline) Flux period 2 (+ PGE2) Phenylbutazone (PBZ) Flux period 1 (baseline) Flux period 2 (+ PGE2) Bicarbonate-free Flux period 1 (baseline) Flux period 2 (+ PGE2) Bicarbonate-free + PBZ Flux period 1 (baseline) Flux period 2 (+ PGE2) Indomethacin (Indo) Flux period 1 (baseline) Flux period 2 (+ PGE2) Bicarbonate-free + Indo Flux period 1 (baseline) Flux period 2 (+ PGE2) Bicarbonate-free + Bumet Flux period 1 (baseline) Flux period 2 (+ PGE2)
J m-s µeq·cm–2·h–1) (µ
J s-m µeq·cm–2·h–1) (µ
J net µeq·cm–2·h–1) (µ
SCC µeq·cm–2·h–1) (µ
Conductance (msiemens/cm2)
6.11 ⫾ 0.52 6.96 ⫾ 0.65
4.50 ⫾ 0.48 4.38 ⫾ 0.30
1.61 ⫾ 0.76 2.58 ⫾ 0.53
1.06 ⫾ 0.14 1.43 ⫾ 0.21*
10.08 ⫾ 0.79 10.44 ⫾ 0.71
6.54 ⫾ 0.77 7.60 ⫾ 0.94
5.33 ⫾ 0.38 5.16 ⫾ 0.43
1.21 ⫾ 1.03 2.44 ⫾ 1.20
0.91 ⫾ 0.09 1.37 ⫾ 0.08*
10.34 ⫾ 0.77 10.36 ⫾ 0.70
6.75 ⫾ 0.55 6.34 ⫾ 0.78
4.04 ⫾ 0.46 5.55 ⫾ 0.39*†
2.72 ⫾ 0.37 0.79 ⫾ 0.73*
0.40 ⫾ 0.07† 1.23 ⫾ 0.12*
10.02 ⫾ 0.73 10.99 ⫾ 0.95
6.62 ⫾ 0.86 8.14 ⫾ 1.64
5.39 ⫾ 0.76 6.78 ⫾ 0.85*†
1.23 ⫾ 1.09 1.36 ⫾ 1.71
1.16 ⫾ 0.09 1.37 ⫾ 0.11
9.74 ⫾ 1.44 10.35 ⫾ 1.27
5.66 ⫾ 0.64 6.15 ⫾ 0.49
4.50 ⫾ 0.39 6.07 ⫾ 0.50*†
1.16 ⫾ 0.76 0.08 ⫾ 0.73
0.66 ⫾ 0.10† 1.22 ⫾ 0.12*
9.97 ⫾ 0.81 10.97 ⫾ 0.69
6.84 ⫾ 1.06 6.33 ⫾ 0.84
4.46 ⫾ 0.30 9.65 ⫾ 1.67*‡
2.38 ⫾ 0.99 –3.32 ⫾ 1.84*‡
0.47 ⫾ 0.09† 1.87 ⫾ 0.16*‡
9.66 ⫾ 1.00 9.31 ⫾ 0.85
6.06 ⫾ 0.90 8.03 ⫾ 0.91
4.35 ⫾ 1.22 6.56 ⫾ 1.37*†
1.71 ⫾ 0.56 1.46 ⫾ 0.86
0.71 ⫾ 0.10† 1.22 ⫾ 0.17*
8.76 ⫾ 0.60 9.99 ⫾ 0.70
5.70 ⫾ 1.03 8.46 ⫾ 1.35
4.03 ⫾ 0.43 5.66 ⫾ 0.51*†
1.67 ⫾ 0.65 2.30 ⫾ 0.43
1.09 ⫾ 0.09 1.07 ⫾ 0.06
8.96 ⫾ 0.63 11.17 ⫾ 1.36
Values are mean ⫾ SEM (6 to 8 ponies/group). J m-s = Flux from mucosa to serosa. J s-m = Flux from serosa to mucosa. J net = J m-s minus J s-m. PGE2 = Prostaglandin E2. *Significantly (P ⬍ 0.05) different from baseline value in the same treatment group. †Significantly (P ⬍ 0.05) different from others in the same flux period without this symbol. ‡Significantly (P < 0.05) different from all others in the same flux period.
936
AJVR, Vol 63, No. 7, July 2002
01-10-0294R.qxd
6/10/2002
1:15 PM
Page 937
uolation was apparent in the cells undergoing oncotic necrosis, whereas cytoplasmic condensation was more typical of the apoptotic cells. The indomethacin-treated colon had changes similar to those of the PBZ-treated tissues; however, histologic changes were milder in the indomethacin-treated tissues. Expressed as the number of apoptotic bodies per 100 crypts via light microscopy, PBZ (62 apoptotic bodies/100 crypts) and indomethacin (44 apoptotic bod-
ies/100 crypts) induced significantly more apoptosis than control conditions (29 apoptotic bodies/100 crypts), and PBZ produced significantly more than indomethacin. Transmission electron microscopy—Findings of TEM confirmed those described for light microscopy, with increased intercellular spaces (Fig 2) in all tissues after incubation and numerous apoptotic bodies at the extrusion zones in tissues treated with PBZ (Fig 3). The early histologic features of apoptosis were evident as rounding up and shrinkage of individual cells or portions of cells within a tissue, condensation of the cytoplasm with dilation of endoplasmic reticulum cisternae, crowding together of organelles, and formation of chromatin caps or crescents in the nucleus of cells.
Figure 3—Transmission electron photomicrograph of mucosal tissue from the right dorsal colon of a pony, after incubation in KRB with 1 µg ml/phenylbutazone and addition of PGE2. Notice numerous sloughed apoptotic bodies (AB) with intact nuclear membranes (white arrowheads), intact cell membranes (black arrows), and chromatin crescents (CC). Apoptotic fragments (AF) can be seen within a tissue macrophage (M).
Figure 2—Transmission electron photomicrographs of mucosal tissue of the right dorsal colon of a pony. A—After incubation in Krebs-Ringer’s-bicarbonate solution (KRB) with PGE2. B—After incubation in KRB with 1 µg of phenylbutazone/ml and addition of PGE2. Notice increased intercellular spaces (arrowheads). Mv = Microvilli. Mu = Mucin. Mt = Mitochondria. Nu = Nucleus. TJ = Tight junctions. Eos = Eosinophil. AJVR, Vol 63, No. 7, July 2002
Figure 4—Transmission electron photomicrograph of mucosal tissue from the right dorsal colon of a pony after incubation in KRB with 1 µg of phenylbutazone/ml and addition of PGE2, illustrating early and late stages of apoptosis. Early stages are evident in 2 cells with large protuberances or blebs being pinched off from the surface (arrowheads). Late stages are evident as apoptotic bodies (AB) and chromatin crescents (CC). 937
01-10-0294R.qxd
6/10/2002
1:15 PM
Page 938
Figure 5—Transmission electron photomicrograph of mucosal tissue from the right dorsal colon of a pony, after incubation in KRB with 1 µg of phenylbutazone/ml and addition of PGE2. Notice diffusely swollen cells in the early stages of necrosis, with pale cytoplasm, irregular chromatin clumping, intracytoplasmic vacuolation (V), and swollen mitochondria (arrowheads).
Figure 6—Transmission electron photomicrograph of mucosal tissue from the right dorsal colon of a pony, after incubation in KRB with 1 µg of phenylbutazone/ml and addition of PGE2. Notice 2 cells in advanced stages of ballooning degeneration and necrosis (N) in the same field as an apoptotic body (AB).
Other changes characteristic of apoptosis were also evident, including development of numerous large surface protuberances or blebs that would eventually pinch off (zeiosis), trapping the intact organelles or the nucleus within them (Fig 4). All the round bodies thus formed, both cytoplasmic and nuclear, were surrounded by intact membranes with no visible loss of membrane integrity. In later stages, the apoptotic bodies were ingested by adjacent tissue cells or by resident tissue macrophages to be degraded. Some cells in the early stages of necrosis were also evident, with pale cytoplasm, irregular chromatin clumping, intracytoplasmic vacuolation, and swollen but intact mitochondria (Fig 5). Cells in more advanced stages of ballooning degeneration and necrosis were also evident, often in the same field as apoptotic bodies (Fig 6). 938
Figure 7—Schematic diagram of transporters that could be involved in ion transport in the right dorsal colon of equids. In this model of a crypt cell, the H+ produced by dissociation of carbonic acid could be transported back to blood via a basolateral H+-ATPase, Na+/H+ exchanger,23 or both or could add a proton to an absorbed volatile fatty acid. NBC = Sodium bicarbonate cotransporter. CFTR = Cystic fibrosis transmembrane conductance regulator. AE = Anion exchanger. NKCC1 = Sodium potassium chloride cotransporter. Na+K+ATPase = Basolateral sodium potassium pump. cAMP = Cyclic adenosine monophosphate. PKA = Protein kinase A. NSAID = Nonsteroidal anti-inflammatory drug. The letter n preceding HCO3– on the NBC cotransporter indicates that the number of HCO3– ions is unknown but probably is > 1.23 A positive sign indicates activation and a negative sign indicates inhibition.
Discussion Findings of this study confirmed our hypothesis, because mucosal integrity and anion transport systems in the RDC were altered by PBZ in vitro at a concentration similar to that achieved in plasma after IV administration,18,19 corrected for 96 to 98% proteinbinding.19,20 Luminal and serosal addition of PBZ was intended to achieve immediate concentrations around surface epithelial cells, as would be expected in vivo by diffusion from capillaries. Results of this study were consistent with the presence of a PG-controlled transporter in mucosa from the RDC that involved some secretion of Cl– and, possibly, HCO3–. The role of HCO3– was underscored by the finding that removal of HCO3–reduced intensity of the shortcircuit response to PGE2 in indomethacin-treated tissues. Also, control tissues and tissues treated with bumetanide in KRB solution had an increased SCC in response to PGE2 but had no change in Cl– secretion. In all tissues treated with an NSAID, the significant decrease in SCC could not be explained by a change in Cl– flux. The effect of an NSAID on baseline current was less pronounced in a HCO3–-free medium than in the presence of HCO3–, but this difference was not significant. However, the role of Na+ was not investigated by use of radioisotope or amiloride in this study, and important transport systems for this cation could explain some of our findings.13,21 All attempts were made to reduce HCO3– concentration in the immediate vicinity of transporting cells; however, the final concentration was not measured in the bathing solutions, and it is possible that sufficient HCO3– remained or was produced by the cells to allow some residual interaction with transporters and mitigation of HCO3–-free responses. AJVR, Vol 63, No. 7, July 2002
01-10-0294R.qxd
6/10/2002
1:15 PM
Page 939
Tissues incubated in HCO3–-free solutions had a significantly greater Cl– secretion than tissues in KRB, probably because the lack of HCO3– eliminated competition for a shared anion secretory mechanism.22,23 The weak and unexpected Cl– secretion induced by PGE2 in tissues incubated in a HCO3–-free solution with bumetanide was opposed by a Cl– absorption of similar magnitude, so there was no change in net Cl– flux or SCC. This finding could be explained by increased intracellular delivery of Cl– on an anion exchanger that normally would use HCO3–. Then, in the absence of HCO3–, exchange of a radiolabeled Cl– on the mucosal side for a nonlabeled Cl– on the serosal side could be enhanced and thereby increase movement of the label from mucosa to serosa. The contribution of HCO3– to anion transport in equine RDC could involve a number of different transporters (Fig 7).23 Under all conditions, the net flux of Cl– favored absorption before and after PGE2, with the exception of tissues treated with indomethacin. However, there was no evidence that Cl– absorption was stimulated by an NSAID or blocked by PGE2. These findings suggest that the RDC differs from the RVC in horses by lacking electroneutral Na+ and Cl– absorption, which may involve parallel ion exchangers (Na+-H+ and Cl–HCO3–) blocked by PG.13 In another study,21 electroneutral Na+ and Cl– absorption was evident in equine RDC but not in the equine small colon. In that study,21 which reported flux data for Cl– similar to our study, indomethacin had no effect on ion transport in the RDC, and the SCC was considerably lower than our recordings. These differences could be explained by the possibility of a more distal site of tissue sampling in our study so that our tissue was closer in function to the small colon. We did not attempt to control plasma aldosterone concentrations, which can influence ion transport in equine colon, but we minimized aldosterone fluctuations as much as possible by harvesting tissues from ponies in equivalent postprandial states.21 Intestinal secretion of Cl– followed by the osmotic flow of water is important in maintaining intestinal contents in a liquid form during digestion.23,24 In general, Cl– enters the secreting epithelial cell at the basolateral membrane through the electroneutral NKCC1 cotransporter (Fig 7) and is secreted though an apical channel, the cystic fibrosis transmembrane conductance regulator (CFTR).24 Expression of the CFTR gene is greatest at the base of the colonic crypts in humans25 and mice,26 and the CFTR is probably the main mechanism by which water and salts are secreted across epithelia.15,16,23,27 The intestinal lesions of cystic fibrosis in humans and mice are caused by mutational inactivation of the cAMP-sensitive CFTR and include accumulations of amorphous mucus plugs in the crypts, crypt dilatation, goblet cell hyperplasia, erosions of the epithelium, and inflammatory infiltrates within the subepithelial layers.15,27 These histologic changes can be seen in horses with PBZ-induced right dorsal colitis.4 Also, because Cl– secretion through the CFTR is an important mechanism for luminal hydration, impairment of this secretion can dehydrate lumen contents, causing intestinal impaction and obstruction.15,27 Ibuprofen can inhibit Cl– secretion, partly AJVR, Vol 63, No. 7, July 2002
through the blockade of the CFTR, and inhibition of Cl– secretion by NSAID may exacerbate the clinical symptoms of cystic fibrosis.28 This may, in part, be a mechanism by which NSAID may cause the large colon and cecum in horses to be predisposed to impactions. In the accepted model for Cl– secretion in the gastrointestinal tract, an agonist that increases intracellular cAMP activates CFTR, and the resulting Cl– secretion decreases intracellular Cl– concentration and depolarizes apical and basolateral membranes.29 These events upregulate the expression of basolateral NKCC1,29 which restores intracellular Cl– and increases the electrochemical equilibrium for Cl– secretion by the CFTR.29,30 Specificity of pharmacologic inhibition by bumetanide is an important criterion for identification of this cotransporter.30 Bumetanide had no effect on Cl– fluxes in our study, possibly because it was used under conditions that did not alter Cl– fluxes or did not induce maximal chloride secretion.23,31-33 The concentration of bumetanide we used was similar to that reported for intestine from other species,23,31 although dose-response curves would be needed to determine the effectiveness of the concentration used.30 However, there are examples of intestinal absence of the NKCC1 transporter23,31 or insensitivity of Cl– secretion to serosal bumetanide.32,33 In contrast to the apparent insensitivity of equine RDC to bumetanide in our study, furosemide can reduce baseline SCC in control RVC and block the SCC response to PGE2 in indomethacin-treated equine RVC.13 Under all conditions, epithelium of the RDC developed some increased intercellular spaces and fluid-filled lamina propria, which is not surprising in a water-absorbing tissue that cannot remove absorbed material via vascular drainage (Fig 2). Epithelial injury could contribute to this change, although absence of structural damage and measured electrical parameters suggest that epithelial cells were viable. An important histologic change in the RDC in our study was an increase in apoptotic bodies in tissues incubated with an NSAID, compared with control tissues. This is a recognized intestinal response to an NSAID12 and could explain the chemopreventive action of these drugs against colon cancer.34-36 Apoptosis is a genetically programmed mechanism of ordered cell death37,38 that accounts for normal cell senescence and maintains the tissue at a given size, thereby protecting the organ from overgrowth under neoplastic conditions.39,40 Apoptosis is also a mechanism by which an injured cell self-destructs in an orderly, sequential fashion, with minimal disruption to surrounding tissues and the organism itself, generally without scarring, and with little or no gross evidence of an inflammatory response.37-40 By comparison, oncotic necrosis involves cell swelling with disrupted membranes and organelles in a large number of cells, followed by the inflammatory response that frequently accompanies an injury.35,40 The large number of apoptotic bodies in our control tissues (29/100 crypts), compared with human large colon in vivo (< 1/100 crypts),12 could be explained by low concentration of butyrate in the bathing medium,41 the length of incubation, and absence of growth factors and other factors that reduce apoptosis in vivo.35 Also, activation of the mucosal immune system or tissue dam939
01-10-0294R.qxd
6/10/2002
1:15 PM
Page 940
age from hypoxia, free radicals, and proinflammatory cytokines during the isolation and incubation procedures could have triggered some apoptosis in our control and treated tissues.35,42-44 Apoptosis occurs rapidly and well within the timeframe of our experiments.41,42 Possible causes for increased apoptosis in the presence of an NSAID include reactive oxygen metabolites and proteases, in addition to inhibition of PG synthesis.8 In our study, the higher conductance of RDC than that reported in other studies for equine RDC21 and RVC13,45 was attributed to increased passive ion conductivity that resulted from greater edge damage, which was induced by the high edge-to-surface ratio in our 1.13-cm2 chambers.46 In our study, the lack of an effect on conductance despite histologic injury could be explained by the nature of the injury, combined with immediate restitution.44 Cells neighboring those undergoing apoptosis extend processes beneath the apoptotic cell and form a tight junction.47 This point of union migrates toward the apices of the cells like a zipper, pushing the apoptotic cell toward the lumen while preserving the epithelial barrier and conductance.39,47 Even after extensive cell loss through apoptosis, barrier function is preserved by rearrangement of tight junctions and desmosomes and by flattening and spreading neighboring cells over the defect.44 However, the inflammatory response associated with NSAIDinduced apoptosis in the human colon suggests that apoptosis can achieve sufficient magnitude to disrupt the epithelial barrier and cause inflammation.12 On the basis of our results, we believe that PBZ can affect an important secretory pathway for anions and water in equine RDC and concurrently cause cell injury. Because the affected transporter could directly or indirectly involve HCO3– secretion, it could play a role in alkalinization of lumen contents in a segment subjected to intense acidification from microbial digestion. In addition, loss of this anion secretory process in other species can be associated with mucosal lesions and intestinal impaction. These findings could explain predisposition to right dorsal colitis in horses. a
Hallyar J, Somasundaram S, Sarathchandra P, et al. Early cellular events in the pathogenesis of NSAID enteropathy in the rat (abstr). Am J Gastroenterol 1991;100:216. b Inoue OJ, Freeman DE, Clarkson RB. Effects of reactive oxygen metabolites on equine colonic mucosa in vitro (abstr). Vet Surg 1996;25:430. c World Precision Instruments, Sarasota, Fla. d Sigma Chemical Co, St Louis, Mo. e New England Nuclear, Boston, Mass.
References 1. Snow DH, Bogan JA, Douglas TA, et al. Phenylbutazone toxicity in ponies. Vet Rec 1979;105:26–30. 2. MacAllister CG, Morgan SJ, Borne AT, et al. Comparison of adverse effects of phenylbutazone, flunixin meglumine, and ketoprofen in horses. J Am Vet Med Assoc 1993;202:71–77. 3. Meschter CL, Gilbert M, Krook L, et al. The effects of phenylbutazone on the intestinal mucosa of the horse: a morphological, ultrastructural and biochemical study. Equine Vet J 1990;22: 255–263. 4. Karcher LF, Dill SG, Anderson WI, et al. Right dorsal colitis. J Vet Intern Med 1990;5:247–253. 5. Dabareiner RM, White NA. Large colon impaction in horses: 147 cases (1985–1991). J Am Vet Med Assoc 1995;206:679–685. 940
6. Rohde C, Anderson DE, Bertone AL, et al. Effects of phenylbutazone on bone activity and formation in horses. Am J Vet Res 2000;61:537–543. 7. Lees P, Taylor JBO, Higgins AJ, et al. In vitro binding of phenylbutazone and related drugs to equine feeds and digesta. Res Vet Sci 1988;44:50–56. 8. Wallace JL. Nonsteroidal anti-inflammatory drugs and gastroenteropathy: the second hundred years. Gastroenterology 1997; 112:1000–1016. 9. Wolfe MM, Lichtenstein DR, Singh G. Medical progress: gastrointestinal toxicity of nonsteroidal antiinflammatory drugs. N Engl J Med 1999;340:1888–1899. 10. Lichtenstein DR, Syngal S, Wolfe MM. Review: nonsteroidal antiinflammatory drugs and the gastrointestinal tract: the doubleedged sword. Arthritis Rheum 1995;38:5–18. 11. Somasundaram S, Rafi S, Hayllar J, et al. Mitochondrial damage: a possible mechanism of the “topical” phase of NSAID induced injury to the rat intestine. Gut 1997;41:344–353. 12. Lee FD. Importance of apoptosis in the histopathology of drug related lesions in the large intestine. J Clin Pathol 1993;46: 118–122. 13. Clarke LL, Argenzio RA. NaCl transport across equine proximal colon and the effect of endogenous prostanoids. Am J Physiol 1990;259:62–69. 14. Campbell NB, Blikslager AT. The role of cyclooxygenase inhibitors in repair of ischaemic-injured jejunal mucosa in the horse. Equine Vet J 2000;32(suppl):59–64. 15. Clarke LL, Gawenis LR, Franklin CL, et al. Increased survival of CFTR knockout mice with an oral osmotic laxative. Lab Anim Sci 1996;46:612–618. 16. Grubb BR, Gabriel SE. Intestinal physiology and pathology in gene-targeted mouse models of cystic fibrosis. Am J Physiol 1997; 273:258–266. 17. Schultz SG, Zalusky R. Ion transport in isolated rabbit ileum. I. Short-circuit current and Na fluxes. J Gen Physiol 1964;47: 567–584. 18. Lees P, Taylor JBO, Higgins AJ, et al. Phenylbutazone and oxyphenbutazone distribution into tissue fluids in the horse. J Vet Pharmacol Ther 1986;9:204–212. 19. Maitho TE, Lees P, Taylor JB. Absorption and pharmacokinetics of phenylbutazone in Welsh mountain ponies. J Vet Pharmacol Ther 1986;9:26–39. 20. Tobin T, Chay S, Kamerling S, et al. Phenylbutazone in the horse: a review. J Vet Pharmacol Ther 1986;9:1–25. 21. Clarke LL, Roberts MC, Grubb BR et al. Short-term effect of aldosterone on Na-Cl transport across equine colon. Am J Physiol 1992;262:939–946. 22. Grubb BR. Ion transport across the normal and CF neonatal murine intestine. Am J Physiol 1999;277:167–174. 23. Grubb BR, Lee E, Pace AJ, et al. Intestinal ion transport in NKCC1-deficient mice. Am J Physiol 2000;279:707–718. 24. Akabas MH. Cystic fibrosis transmembrane conductance regulator— structure and function of an epithelial chloride channel. J Biol Chem 2000;275:3729–3732. 25. Eggermont E. Gastrointestinal manifestations in cystic fibrosis. Eur J Gastroenterol Hepatol 1996;8:731–738. 26. Illek B, Fischer H, Machen TE. Genetic disorders of membrane transport. II. Regulation of CFTR by small molecules including HCO3–. Am J Physiol 1998;275:1221–1226. 27. Wyllie R. Gastrointestinal manifestations of cystic fibrosis. Clin Pediatr 1999;38:735–738. 28. Devor DC, Schultz BD. Ibuprofen inhibits cystic fibrosis transmembrane conductance regulator-mediated Cl– secretion. J Clin Invest 1998;102:679–687. 29. Shumaker H, Soleimani M. CFTR upregulates the expression of the basolateral Na+-K+-2Cl– cotransporter in cultured pancreatic duct cells. Am J Physiol 1999;277:1100–1110. 30. O’Grady SM, Palfrey HC, Field M. Characteristics and functions of Na-K-Cl cotransport in epithelial tissues. Am J Physiol 1987; 253:177–192. 31. Flagella M, Clarke LL, Miller ML, et al. Mice lacking the basolateral Na-K-2Cl cotransporter have impaired epithelial chloride secretion and are profoundly deaf. J Biol Chem 1999;274:26946–26955. 32. Charney AN, Gianella RA, Egnor RW. Effect of short-chain AJVR, Vol 63, No. 7, July 2002
01-10-0294R.qxd
6/10/2002
1:15 PM
Page 941
fatty acids on cyclic 3',5'-guanosine monophosphate-mediated colonic secretion. Comp Biochem Physiol A Physiol 1999;124:169–178. 33. Schulteiss G, Horger S, Diener M. The bumetanide-resistant part of forskolin-induced anion secretion in rat colon. Acta Physiol Scand 1998;164:219–228. 34. Shiff SJ, Rigs B. Nonsteroidal anti-inflammatory drugs and colorectal cancer: evolving concepts of their chemopreventive actions. Gastroenterology 1997;113:1992–1998. 35. Tarnawski AS, Szabo I. Apoptosis—programmed cell death and its relevance to gastrointestinal epithelium: survival signal from the matrix. Gastroenterology 2001;120:294–298. 36. Chan TA, Morin PJ, Vogelstein B, et al. Mechanisms underlying nonsteroidal anti-inflammatory drug-mediated apoptosis. Proc Natl Acad Sci U S A 1998;95:681–686. 37. Jones BA, Gores GJ. Physiology and pathophysiology of apoptosis in epithelial cells of the liver, pancreas, and intestine. Am J Physiol 1997;273:1174–1188. 38. Thompson CB. Apoptosis in the pathogenesis and treatment of disease. Science 1995;267:1456–1462. 39. Watson AJM, Pritchard DM. Lessons from genetically engineered animal models. VII. Apoptosis in intestinal epithelium: lessons from transgenic and knockout mice. Am J Physiol 2000;278:1–5.
AJVR, Vol 63, No. 7, July 2002
40. Watson AJM. Necrosis and apoptosis in the gastrointestinal tract. Gut 1995;37:165–167. 41. Hass R, Busche R, Luciano L, et al. Lack of butyrate is associated with induction of Bax and subsequent apoptosis in the proximal colon of guinea pig. Gastroenterology 1997;112:875–881. 42. Noda T, Iwakiri R, Fujimoto K, et al. Programmed cell death induced by ischemia-reperfusion in rat intestinal mucosa. Am J Physiol 1998;274:270–276. 43. Coopersmith CM, O’ Donnell D, Gordon JI. BCl–2 inhibits ischemia-reperfusion-induced apoptosis in the intestinal epithelium of transgenic mice. Am J Physiol 1999;276:677–686. 44. Abreu MT, Palladino AA, Arnold ET, et al. Modulation of barrier function during Fas-mediated apoptosis in human intestinal epithelial cells. Gastroenterology 2000;119:1524–1536. 45. Freeman DE, Inoue O, Eurell T. Effects of flunixin meglumine on short circuit in equine colonic mucosa in vitro. Am J Vet Res 1997;58:915–919. 46. Dobson JG, Kidder GW. Edge damage effect in in vitro frog skin preparations. Am J Physiol 1968;214:719–724. 47. Iwanaga T, Han H, Adachi K, et al. A novel mechanism for disposing of effete epithelial cells in the small intestine of guinea pigs. Gastroenterology 1993;105:1089–1097.
941
