Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The isoprostanes are chemically stable novel prostaglandin (PG)-like compounds formed in vivo by free-radical-induced peroxidation of polyunsaturated fatty acids and are synthesized independent of cyclooxygenase (Morrow et al., 1990b, 1994). In contrast to their PG isomer, the isoprostanes predominantly have a cis orientation of their side chains in relation to the cyclopentane ring. The formation of the isoprostanes can occur from either free arachidonic acids, from free PGs (Morrow et al., 1990a), or in the esterified form such as phospholipids in cell membranes (Morrow et al., 1994). However, it has been suggested that the isoprostane 8-epi-PGF₂ is a product of a cyclooxygenase-dependent pathway (Pratico et al., 1995).

Reactive oxygen species (i.e., HO, HOO, ROO, superoxide anion, or nitric oxide) can abstract a hydrogen atom from the polyunsaturated fatty acid at carbons 7, 10, and 13, giving rise to four different classes of isoprostanes by a further reaction of two molecules of oxygen (Rokach et al., 1997). Each of the four regioisomers can be comprised of eight racemic diastereomers, thus a total of 64 different isoprostanes can be generated (Liu et al., 1999). Several of the isoprostanes possess potent biological activity and may thus participate as mediators of oxidant injury in vivo and in vitro (Morrow and Roberts, 1997). For instance, 8-iso-PGE₂ and 8-epi-PGF₂ have been observed to be renal vasoconstrictors (Takabashi et al., 1992; Fukunaga et al., 1993b).

The specific receptors involved in mediating the effects of isoprostanes have not been defined. Many observations suggest that biologically active isoprostanes exert their effects on thromboxane A₂ (TP) or TP-like receptors (Crankshaw, 1995; Audoly et al., 2000). However, other studies have indicated an activity of the isoprostanes on a "unique" receptor similar to, but distinct from, the TP receptor (Fukunaga et al., 1993a, 1997; Longmire et al., 1994). More recently, the possibility that prostanoid E (EP) receptors are involved was suggested by Elmhurst et al. (1997) in their studies on the canine proximal colonic epithelium. Similar observations have also been made on the rat fundus and the guinea pig ileum (Sametz et al., 2000).

Our objectives were to answer two questions: 1) Do the isoprostanes 8-iso-PGE₂ and 8-iso-PGF₂ have biological activities in vitro on the porcine small intestine? and 2) do these effects involve prostanoid receptors or a "unique" isoprostane receptor? The porcine small intestine is considered as an appropriate model for humans for the study of pathophysiological mechanisms in the gastrointestinal tract. Furthermore, the results may have direct interest for the pathophysiology of diarrhea. The Ussing chamber technique was used to record the responses of the tissues as changes in the bioelectrical parameters.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. Postweaned female pigs (Danish Landrace/Yorkshire crossbreed, 9-10 weeks old, weighing 18-20 kg) were fed a commercial standard diet. Before each experiment the pigs were fasted overnight but had free access to sterile drinking water containing 55 g/l D-glucose.

Preparation of Small Intestinal Tissue. Pigs were stunned with a bolt pistol, pinched, and immediately exsanguinated by cutting the aortic arch and superior vena cava. Each pig was opened along the abdominal midline, and a jejunal segment, approximately 25 cm long, was removed 30 cm distal to the duodenocolic ligament (ligament of Treitz). The intestine was immediately stripped of outer muscle layers, rinsed in bicarbonate Ringer's solution of the following composition (in mM): 145 Na⁺, 128 Cl, 0.32 PO₄³, 2.0 Ca²⁺, 1.0 Mg²⁺, 25 HCO₃, 1 SO₄², 6.3 K⁺, and 16 mM D-glucose; pH 7.4, opened along the mesenteric border, and rinsed again. The Ringer's solution also contained 1.4 µM indomethacin to suppress endogenous prostanoid synthesis.

Measurement of Bioelectrical Parameters. The bioelectrical parameters of the porcine small intestine were measured according to the short-circuit current technique described by Ussing and Zerahn (1951). The tissue was cut in sheets and mounted in modified perspex Ussing chambers (opening area 1.0 cm²) with a rubber O-ring on the mucosal side to minimize tissue edge damage. The tissues were bathed in bicarbonate Ringer's solution that was continuously oxygenated (95% O₂/5% CO₂) by a gas lift system, and the baths were kept at 39°C (porcine core temperature) by water jackets in connection to a water bath. The chambers and their reservoirs were coated with AquaSil (Pierce Chemical Co., Rockford, IL) to prevent drug adsorption to the surface.

Experimental Design. The bathing medium was Ringer's bicarbonate solution containing 16 mM D-glucose on the serosal side, which was substituted with 16 mM D-mannitol on the mucosal side to keep osmolality equal on both sides. Glucose was omitted in the mucosal Ringer's solution to avoid the electrical contribution of the Na⁺ absorption via the Na⁺/glucose cotransporter. Tissues were allowed to equilibrate for 30 min before starting the experiments to permit the epithelium and I_SC to stabilize. Unless otherwise stated, agonists were added to the serosal side and their responses were measured as the difference between the peak response and the baseline of the I_SC, i.e., I_SC. At the end, tissue viability and responsiveness were tested with 2.22 mM theophylline in bicarbonate Ringer's solution applied bilaterally.

Concentration-Response Experiments. Each tissue was used to evaluate the effects of the agonist at one concentration, i.e., a noncumulative concentration-response protocol was employed, and such that a range of different concentrations were added to intestinal preparations taken from the same pig. Concentration-response curves (log molar concentration of agonist versus effect) were constructed from the data by fitting the equation:
where E is the effect of the agonist, K is a power coefficient, C is the molar concentration of the agonist, and C₅₀ is the molar concentration of the agonist that produces a half-maximal response (EC₅₀). The value of log C₅₀ is equivalent to the pEC₅₀.

Desensitization Experiments. In these experiments, tissues were set up in pairs. The control tissue was exposed only to the agonist with no pretreatment. The corresponding pair was treated with the desensitizing agonist till no further responses were obtained. After the response to the desensitizing agonist had faded, the tissue was treated with the same concentration of the primary agonist. The peak responses after desensitizing were expressed as a percentage of the control response with the primary agonist.

Chemicals. PGE₂, PGF₂, 8-iso-PGE₂, 8-iso-PGF₂, U46619 (5-heptenoic acid, 7-[6-(3-hydroxy-1-octenyl)-2-oxabicyclo[2.2.1]hept-5-yl-[1R-[1,4,5(Z),6(1E,3S*)]]), and SQ29,548 (5-heptenoic, 7-[3-[[2-[(phenylamino)carbonyl]hydrazino]methyl]-7-oxabicyclo [2.2.1]hept-2-yl]-[1S-[1,2(Z),3,4]]), 17-phenyl-trinor-PGE₂, butaprost, misoprostol, and sulprostone were purchased from Cayman Chemical Co. (Ann Arbor, MI), and SC-51322 was from Biomol Research Laboratories, Inc. (Plymouth meeting, PA). These were all dissolved in 96% (v/v) ethanol at a stock concentration of 25 mM and stored at 20°C or 80°C until use (less than 6 months). Tetrodotoxin (TTX) was obtained from Calbiochem-Novabiochem GmbH (Band Soden, Germany). Indomethacin (Dumex, Denmark or Sigma, St. Louis, MO) was freshly prepared. All other chemicals were from Sigma. The vehicle concentration of ethanol showed no effects on the bioelectrical parameters.

Statistics. Data are presented as means ± S.E.M. (n = number of experiments). Either paired or unpaired Student's t tests were used to determine statistical significance, and values of p < 0.05 were considered to be significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Basal Bioelectrical Parameters. The following basal bioelectrical parameters were noted: for V_t, 0.36 ± 0.03 mV (n = 309); for I_SC, 11.6 ± 0.9 µA/cm²; and for R_t, 31.3 ± 0.6  × cm².

Effects of Isoprostanes and Prostaglandins on Bioelectrical Parameters. Serosal application of 8-iso-PGE₂, PGE₂, and PGF₂ induced a monophasic rise in I_SC with an initial rapid increase followed by a slower rise to a peak, which was achieved approximately 8 to 10 min after drug addition further followed by a slow decrease to a steady-state. Typical time courses of the effects on I_SC are given in Fig. 1. However, the PGE₂ and PGF₂ response showed occasionally a biphasic rise in I_SC (see Figs. 3B and 5E, respectively). No responses were seen with serosal addition of 50 µM 8-iso-PGF₂ or luminal addition of either PGE₂ or 8-iso- PGE₂ (data not shown).

View larger version (6K): [in this window] [in a new window]   Fig. 1.   Time course of changes in short-circuit current (ISC) after serosal application of 8-iso-PGE2 (A), PGE2 (B), and PGF2 (C) to porcine small intestinal tissue. The traces should be read from left to right.

Concentration-Response Experiments. Concentration-response experiments were performed to further evaluate the effect of 8-iso-PGE₂, PGE₂, and PGF₂ on I_SC. A concentration-dependent increase in I_SC was observed (Fig. 2). The concentration-response curve for PGE₂ appeared bimodal, indicating that two EP receptors may be involved. Analysis showed that the data for PGE₂ were best fitted by a two-site model. PGF₂ induced a steep increase in I_SC at higher concentrations. The estimated pEC₅₀ and E_max values are shown in Table 1. A significant difference in the pEC₅₀ was observed between 8-iso-PGE₂ and the two pEC₅₀ values for PGE₂ (p < 0.01); and 8-iso-PGE₂ had a significant lower E_max value than the E_max for PGE₂ (2). A significant difference between the two pEC₅₀ and E_max values for PGE₂ was furthermore observed (p < 0.001). The E_max value for PGF₂ was lower than the PGE₂ (2) E_max value, and higher than the PGE₂ (1) E_max value (p < 0.05 and p < 0.001, respectively).

View larger version (10K): [in this window] [in a new window]   Fig. 2.   Concentration-response curves for 8-iso-PGE2 (n from 2 to 7) (A), PGE2 (n from 3 to 13) (B), and PGF2 (n from 2 to 5) (C) on the porcine small intestine. The concentration-response is measured as the change in the short-circuit current (ISC). The agonists were added as a single concentration to the serosal side of the preparations. The concentration-response experiments were performed when it was possible to compare the effect of different concentrations on different preparations taken from the pig. The data are means ± S.E.M.

View this table: [in this window] [in a new window]   TABLE 1 pEC50 and Emax values for 8-iso-PGE2, PGE2, and PGF2 induced peak changes in ISC (ISC; µA/cm2) in porcine small intestine Values are means ± S.E.M. Number of experiments: see Fig. 2.

View larger version (10K): [in this window] [in a new window]   Fig. 3.   Effect of TP receptor agonist U46619 on porcine intestinal epithelial ISC changes (ISC) before (A) and after (B) stimulation with PGE2.

Desensitization Experiments. We sought to determine whether 8-iso-PGE₂ acted via EP or prostanoid F (FP) receptors, because the TP receptor antagonist, SQ29,548, was unable to inhibit responses to either 8-iso-PGE₂ or PGE₂. Since 8-iso-PGE₂ is a cis isomer of PGE₂, it seemed likely that 8-iso-PGE₂ acted through an EP receptor. The investigations were performed by cross-desensitization protocols, as no selective EP and FP receptor antagonists are readily available. The results are presented in Table 2. The data show that both homologous and heterologous desensitization were observed with all three agonists, suggesting the involvement of EP and FP receptors in the responses elicited to the isoprostane. A representative trace of desensitization is shown in Fig. 4.

View larger version (7K): [in this window] [in a new window]   Fig. 4.   Representative trace of ISC showing the effect of 8-iso-PGE2 on porcine small intestinal tissue desensitized to PGE2.

Effects of TTX on Bioelectrical Parameters. To investigate the involvement of the enteric nervous plexus in the I_SC response to 8-iso-PGE₂, PGE₂, and PGF₂ the tissues were pretreated with TTX (1 µM in 30 min). The results are presented in Table 3 and illustrate no effects of TTX on the I_SC response to 8-iso-PGE₂, PGE₂, or PGF₂.

View this table: [in this window] [in a new window]   TABLE 3 Effects of 8-iso-PGE2, PGE2, and PGF2 after pretreatment with serosal TTX (1 µM in 30 min) on the porcine small intestine ISC responses Data are given as the changes () in ISC. Values are means ± S.E.M.; n, number of experiments.

Determination of the Ionic Component Involved in the Effects of 8-Iso-PGE₂, PGE₂, and PGF₂. The Na⁺/K⁺/2Cl cotransporter located in the basolateral membrane is the principal mechanism for intracellular accumulation of Cl to a level above its electrochemical equilibrium (Matthews et al., 1994; D'Andrea et al., 1996). The effects of 8-iso-PGE₂, PGE₂, and PGF₂, after pretreatment of the tissues with the Na⁺/K⁺/2Cl-cotransport inhibitor, bumetanide (0.1 mM in 50 min), on the I_SC and R_t, are shown in Fig. 5. Bumetanide significantly reduced, but did not abolish, the increases in I_SC and R_t. The results indicate that 8-iso-PGE₂, PGE₂, and PGF₂ induced a secretion of Cl from the serosal to the mucosal side. Furthermore, in five of the nine tissues pretreated with bumetanide, PGE₂ elicited sharp increases in I_SC, in contrast to a more sustained response seen in the controls (Fig. 6).

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Effects of 8-iso-PGE2 (A and B), PGE2 (C and D), and PGF2 (E and F) on the short-circuit current (ISC) and the transepithelial electrical resistance (Rt) of porcine jejunal tissue in nontreated or pretreated tissues with bumetanide (0.1 mM in 50 min). The values are given as the changes () compared with the value of ISC or Rt measured at time = 0 when the agonist was applied. Control: 8-iso-PGE2 (n = 5), PGE2 (n = 11), and PGF2 (n = 5). Bumetanide: 8-iso-PGE2 (n = 6), PGE2 (n = 9), and PGF2 (n = 6).

View larger version (7K): [in this window] [in a new window]   Fig. 6.   Representative trace of ISC showing that PGE2 elicited an increase of short duration in the presence of 0.1 mM bumetanide.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The objectives were to answer two questions: 1) Do the isoprostanes 8-iso-PGE₂ and 8-iso-PGF₂ have biological activities in vitro on the epithelium of the porcine small intestine? and 2) Are these effects exerted on prostanoid (i.e., TP, EP, or FP) receptors or on a unique isoprostane receptor?

The I_SC is a measure of net charge movement caused by all ion transport processes across the jejunal epithelium. Previous studies on porcine jejunum have shown that the I_SC, in the presence of indomethacin, is nearly equal to the net absorption of Na⁺ (coupled to D-glucose) with a net flux for Cl near zero, i.e., a flux ratio for Cl of unity (Holtug and Skadhauge, 1991). Glucose was omitted from the mucosal solution, and the recorded bioelectrical parameters are in agreement with other results on the porcine jejunum (Holtug and Skadhauge, 1991; Erlwanger et al., 1999).

8-Iso-PGE₂ produced concentration-dependent increases in I_SC, whereas no effect of 8-iso-PGF₂ was observed. The same isoprostane has previously been reported to induce elevations in I_SC in the canine proximal colon (Elmhurst et al., 1997). Pretreatment with PGE₂ induced a concentration-dependent reduction in the stimulant responses to 8-iso-PGE₂ and PGF₂, and vice versa. These desensitization experiments suggested that the stimulant responses elicited by 8-iso-PGE₂ appeared to involve both EP and FP receptors in the porcine small intestine. The concentration-response curve of PGE₂ suggested that at least two receptors were involved. In a study by Bunce and Spraggs (1990), the potency of different prostanoid agonists and antagonists was compared, demonstrating that different prostanoid receptors are involved in the prostanoid agonist stimulated I_SC in guinea pig ileum and gastric mucosa.

We sought to define the receptor involved in the responses to 8-iso-PGE₂. Previous studies had raised the possibility that either a unique isoprostane receptor or a variant of the TP receptor was involved (Fukunaga 1993a, 1997). Limited evidence exists for the former possibility from a number of studies. A variety of test systems from platelets, vascular and nonvascular smooth muscles, colonic epithelia, and iris-ciliary body preparations have suggested the involvement of TP receptors (Audoly et al., 2000; Awe et al., 2000; Cayatte et al., 2000; Oliveira et al., 2000; Sametz et al., 2000). In a recent study, Audoly et al. (2000) reported the effects of isoprostanes on vascular responses and platelet aggregation in transgenic mice that either lacked the TP gene or overexpressed the TP-beta isoform of the TP receptor. Pressor responses and platelet function were abolished in mice lacking the TP gene. These mice responded normally to another vasoconstrictor, angiotensin II. By contrast, the pressor responses to the isoprostanes were exaggerated in the mice that overexpressed the TP receptor, and these responses were predictably inhibited by the TP receptor antagonist, SQ29,548. Oliveira and coworkers (2000) found that human umbilical arteries in vitro responded to a number of isoprostanes as well as to U46619. A TP antagonist, GR 32191, shifted the contractile responses to the right. In the present study, the TP agonist, U46619, elicited transient responses and the TP receptor antagonist, SQ29,548, significantly inhibited these. However the antagonist did not alter responses to the isoprostane, making it unlikely that TP receptors were involved. By contrast, the desensitization experiments showed that responses to the isoprostane were significantly reduced by pretreatment with PGE₂ and PGF₂. Thus both EP and FP receptors could play a significant role. Desensitization has been reported for the EP receptor (Bastepe and Ashby, 1997). This is not surprising, because 8-iso-PGE₂ is a stereoisomer of PGE₂ and because several naturally occurring PGs exerted actions at several PG receptors (Coleman et al., 1994). It is thus not particularly surprising that the effects of 8-iso-PGE₂ involve EP receptors.

In contrast to the information available on the role of TP receptors in isoprostane effects, the roles of EP receptors remain relatively unexplored. The involvement of EP receptors in the responses to isoprostanes was initially demonstrated in the canine proximal colonic epithelium (Elmhurst et al., 1997). In that tissue, the EP responses were stimulant in contrast to the TP effects, which were inhibitory. In the porcine jejunum, no inhibitory effects were seen that might represent a species variation. More recently, similar observations have been made in the rat fundus and guinea pig ileum (Sametz et al., 2000). It is worth noting that the involvement of EP receptors have been noted so far with tissues from the gastrointestinal tract, whereas effects on vascular tissues involve TP receptors.

The particular subtype of the EP receptor involved in the responses to isoprostane requires definition. The bimodal shape of the concentration-responses seen with PGE₂ suggests that such subtypes are relevant. The pEC₅₀ values were significantly different. The expression of EP receptor subtypes in the mouse gut has been explored (Narumiya et al., 1999). There appear to be differential localization in different regions. With particular reference to the intestinal lining, EP4 mRNA was present in intestinal epithelial cells. The definition of the particular subtypes present in the pig jejunum requires far more detailed study and was beyond the scope of the present study. Thus it is difficult to state which particular subtype of the EP receptor was involved in eliciting the responses to 8-iso-PGE₂. The lack of selective antagonists makes the task more difficult. We have, however, some preliminary data on this issue. The EP₁ antagonist, SC-51322 (5 µM) had no significant effect on the responses to either PGE₂ or 8-iso-PGE₂, suggesting that EP₁ receptors are not involved (Abramovitz et al., 2000). We tested the effects of single concentrations of different agonists in the micromolar range in a few experiments. The average responses to each of the agonists tested are indicated alongside each. Thus, slight responses were obtained with the EP₁ agonist, 17-phenyl-trinor-PGE₂ (11.2 µA/cm²), as well as butaprost (3.83 µA/cm²), although better responses were seen with misoprostol (44.2 µA/cm²). No responses were seen with sulprostone. Given that only misoprostol gave reasonable responses, we are tempted to suggest that EP4 receptors are involved. Such an inference must be made very cautiously, as detailed concentration-response curves were not constructed. This is clearly an area that requires far more detailed analysis.

The doses of 8-iso-PGE₂ and PGs in the present study were 0.1 to 100 µM, which are relatively high, i.e., pharmacological doses, to evoke a response in I_SC. To elicit a maximal effect of PGE₂, similar results have been observed in the rat gastrointestinal tract (Racusen and Binder, 1980; Flemstrom and Kivilaakso, 1983; Stewart and Turnberg, 1989). Conversely, PGE₂ and PGF₂ were effective in low, i.e., physiological, doses (1-10 pM) in the human jejunum (Bukhave and Rask-Madsen, 1980). The reasons for the variations in sensitivity to PGs between species may reflect differences in tissue preparations. With specific reference to the isoprostanes, they are likely to be produced in large amounts under pathophysiological states. Thus the effects seen in this study under relatively high concentrations may in fact have relevance under those conditions.

PGs are known to induce electrogenic Cl secretion (Bunce and Spraggs, 1990; Deachapunya and O'Grady, 1998), a process that underlies for instance secretory diarrhea (Hecht et al., 1999). The accumulation of Cl into the cells to sustain high rates of secretion in Cl-secreting epithelia occurs primarily via a bumetanide-sensitive Na⁺/K⁺/2Cl-cotransport (O'Grady et al., 1987). Bumetanide clearly reduced the 8-iso-PGE₂, and PG induced increases in I_SC and R_t. This indicates that the Na⁺/K⁺/2Cl cotransporter contributes significantly to the basolateral uptake of Cl under secretory conditions and suggests that the 8-iso-PGE₂ and the PGs induce a Cl secretion to the luminal side of the small intestine. An activation of the Na⁺/K⁺/2Cl cotransporter by cAMP may indicate that the main response of 8-iso-PGE₂ and PGE₂ is mediated via EP₂ and/or EP₄ receptors (Coleman et al., 1994). It is thus not surprising that the isoprostane activates a Cl channel, because it presumably activates identical receptors as the PGs.

TTX had no effect on 8-iso-PGE₂ and PGE₂. The induced increase in I_SC indicates that the effect of 8-iso-PGE₂ and the PGs is not mediated through TTX-sensitive neurons. It is likely that the effects of 8-iso-PGE₂ and the PGs occur directly on the enterocyte cell, which may express PG receptors (Eberhart and DuBois, 1995). It must be emphasized that the concentration of TTX used in these studies was 10 time higher than that shown to significantly reduce responses to SP and neurokinin A (Thorbell et al., 1998). Thus, had there be an indirect effect of the eicosanoids, it would have been revealed in these experiments.

The eicosanoids have been implicated in inflammatory bowel disease (Gaginella, 1990; Davies and Rampton, 1997). The role of the isoprostanes in gastrointestinal pathophysiology is undefined. Nevertheless, as the isoprostanes are formed subsequent to oxidative stress, the inflammatory state might enhance the production of isoprostanes, which could affect the intestinal epithelium. The concentration of isoprostanes in human biological fluids has been shown to be of several orders of magnitude greater than the products of the cyclooxygenase (Morrow et al., 1990a).

Since an important route of transepithelial passage of solutes is via the paracellular space, it has been suggested that a physiological control over tight junctions might be of major physiological importance (Anderson and Van Itallie, 1995). The tight junctions are crucial to baseline intestinal barrier function and form a barrier that limits the paracellular diffusion of hydrophilic solutes. It has been shown that cAMP modulates tight junctional permeability and increases R_t (Duffey et al., 1981). The present study demonstrates that 8-iso-PGE₂, and the PGs, induced an increase in R_t. A restoration of the intestinal barrier, measured as an increase of R_t, in the ischemia-injured porcine ileum has recently been suggested to be accounted for by PGs involving a cytoskeleton-mediated tight junction closure via intracellular cAMP and Ca²⁺ (Blikslager et al., 1997). The restoration is induced by secretion of Cl and inhibition of Na⁺ absorption (Blikslager et al., 1999).

Thus, the results in the present study might indicate that isoprostanes and PGs affect cellular as well as paracellular transport pathways in the porcine small intestine. The effect on the paracellular shunt might indicate a tight junction closure. Since the intestinal paracellular space is a major route of transepithelial solute passage, it is therefore likely that the isoprostanes and PGs exert a cytoprotective function. Cytoprotection designates the property of a compound to protect the mucosa of the intestine from becoming inflamed and necrotic, when the intestine is exposed to enteropathogenic bacteria (Robert, 1979). The precise mechanisms of isoprostanes involved in eliciting these effects need further definition.