Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Irinotecan, a semisynthetic derivative of camptothecin, is a DNA topoisomerase I inhibitor and has been clinically used as an antitumor drug in the United States, Europe, and Japan. One of major dose-limiting toxicities of this drug is severe diarrhea. To resolve this problem, concomitant application of irinotecan with high doses of loperamide, an antidiarrhea drug, has been used in Europe and the United States (Abigerges et al., 1994; Armand et al., 1996; Rothenberg, 1998; Saliba et al., 1998).

Diarrhea is generally accompanied with excess secretion of electrolytes, especially Cl (Field and Semrad, 1993). We found recently that irinotecan indirectly stimulated Cl secretion in isolated rat colonic mucosa via the release of thromboxane A₂ (TXA₂) from subepithelial layer (Sakai et al., 1995, 1997). Irinotecan-released TXA₂ binds to a TXA₂ receptor in epithelial crypt cells and increases the secretion of Cl (Sakai et al., 1997). These findings are important because endogenous TXA₂ has been established for the first time as a novel secretagogue in isolated animal colon. Interestingly, TXA₂ has been suggested to be associated with human ulcerative colitis (Rampton and Collins, 1993; Casellas et al., 1995).

Because TXA₂ is unstable and quickly transforms into TXB₂ in aqueous solutions (Hamberg et al., 1975), several stable analogs of TXA₂ such as U46619 (Phillips and Hoult, 1988; Smith et al., 1988), carbocyclic thromboxane A₂ (Diener and Rummel, 1991), and 9,11-epithio-11,12-methano-thromboxane A₂ (STA₂) (Katsura et al., 1983; Sakai et al., 1997) have been used to mimic the effect of endogenous TXA₂ on ion transport in the intestine. U46619 caused the Cl secretion in the rat ileum (Smith et al., 1988) and rat colon (Phillips and Hoult, 1988). However, the effect of U46619 was mediated by the release of cyclooxygenase metabolites (Smith et al., 1988). Carbocyclic thromboxane A₂ did not cause Cl secretion but inhibited Cl absorption in the rat colon (Diener and Rummel, 1991). In contrast to U46619 and carbocyclic thromboxane A₂, the action of STA₂ has been found to be similar to that of irinotecan-released endogenous TXA₂; that is, STA₂ directly caused the Cl secretion via TXA₂ receptor located in epithelial crypt cells (Sakai et al., 1997). At present, STA₂ is only an analog, which can mimic the effect of endogenous TXA₂, at least, in the animal colonic mucosa.

Loperamide is a synthetic opiate derivative, and it shows inhibitory effects against electric field- and secretagogue(s)-stimulated ion secretion in the colon (Diener et al., 1988b; Burleigh, 1991; Kromer, 1995). The mechanisms of antisecretory action of loperamide have been discussed with reference to 1) opiate agonism, 2) block of calcium channels, and 3) inhibition of calmodulin (Ooms et al., 1984; Diener et al., 1988b; Awouters et al., 1993).

So far, the effect of loperamide on the TXA₂-induced secretion has not been reported. In the present study with isolated rat colonic mucosa, we tested the effects of loperamide on the Cl secretion induced by irinotecan and STA₂. We also investigated the mechanism of action of loperamide on the STA₂-induced secretion in the colon.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. 7-Ethyl-10-[4-(1-piperidino)-1-piperidino]carbonyloxycamptothecin (irinotecan; Daiichi Pharmaceutical Co., Tokyo, Japan, and Yakult Honsha Co., Tokyo, Japan); STA₂ (ONO Pharmaceutical Co., Osaka, Japan), and prostaglandin E₂ (PGE₂; Toray Industries, Tokyo, Japan) were generous gifts. Loperamide hydrochloride was obtained from Research Biochemicals International (Natick, MA). STA₂, PGE₂, and loperamide were dissolved in ethanol, and irinotecan was dissolved in dimethyl sulfoxide. Ethanol and dimethyl sulfoxide concentrations in the final solutions never exceeded 0.5%, at which concentration the vehicle per se did not affect the short-circuit current (Isc), the potential difference across the mucosa (Pd), and the tissue conductance (Gt). Atropine monohydrate (Wako Pure Chemical Industries, Osaka, Japan), N-(6-aminohexyl)-5-chloro-1-naphthalenesulfonamide hydrochloride (W-7; Seikagaku Co., Tokyo, Japan), and N-(6-aminohexyl)-1-naphthalenesulfonamide hydrochloride (W-5; Seikagaku Co.) were dissolved in the Parsons solution described below just before use.

Solutions. The Parsons solution for tissue preparation and Ussing chamber experiments consisted of 107 mM NaCl, 4.5 mM KCl, 25 mM NaHCO₃, 1.8 mM Na₂HPO₄, 0.2 mM NaH₂PO₄, 1.25 mM CaCl₂, 1 mM MgSO₄, and 12 mM glucose. The solution was gassed with carbogen (5% CO₂ in 95% O₂) at a pH of 7.4. The Ca²⁺-free EDTA solution for the isolation of crypts from distal colon contained 127 mM NaCl, 5 mM KCl, 1 mM MgCl₂, 5 mM EDTA, 10 mM HEPES, 5 mM glucose, and 5 mM sodium pyruvate, with 10 mg/ml BSA. The pH of the solution was adjusted with NaOH to 7.4. The high K⁺ Tyrode's solution for the storage of the crypts contained 100 mM potassium gluconate, 30 mM KCl, 20 mM NaCl, 1.25 mM CaCl₂, 1 mM MgCl₂, 10 mM HEPES, 12 mM glucose, and 5 mM sodium pyruvate, with 1 mg/ml BSA. The pH was adjusted with KOH to 7.4.

Tissue Preparation. The following procedures were performed in accordance with the guidelines presented by the Animal Care and Use Committee of Toyama Medical and Pharmaceutical University. The mucosa-submucosa preparation (hereafter, simply described as the mucosa) was obtained from female Wistar rats (Japan SLC, Shizuoka, Japan) with a weight of 140 to 200 g. The animals had free access to water and food until the day of the experiment. Animals were sacrificed rapidly by stunning and cervical dislocation. The serosa and muscularis propria were stripped away by hand to obtain the mucosa preparation of the distal part of the colon descendens.

Ussing Chamber Experiments. The tissue was fixed in a modified Ussing chamber and bathed with 4 ml of the Parsons solution incubated at 37°C on each side of the mucosa. The exposed surface of the tissue was 0.3 cm². Isc was continuously measured at zero voltage difference with an amplifier (CEZ-9100; Nihon Kohden Co., Tokyo, Japan). The fluid resistance was compensated. The direction of Isc from the mucosal-to-serosal side was expressed as positive; that is, an increase in Cl movement from the serosal-to-mucosal side (Cl secretion) corresponded to an increase in Isc. The transepithelial Pd under open-circuit conditions was measured in the current clamp mode of the amplifier, and the reference was taken on the serosal side. Gt was determined from the deviation of Isc in response to the command voltage pulse of 0.5 mV (its duration was 100 ms).

Preparation of Isolated Colonic Crypts. Crypts were isolated from the distal colon as previously described (Ecke et al., 1996). Briefly, the distal colon was resected and turned inside out. The inverted sac was filled with 3 to 5 ml of the Ca²⁺-free EDTA solution, and it was incubated in the Ca²⁺-free solution for 10 min at 35°C. Isolated crypts were collected and resuspended in the high K⁺ Tyrode's solution.

Measurements of [Ca²⁺]_i of Colonic Crypt Cells. The isolated crypts were loaded with indo-1 AM (10 µM) in the BSA-free high K⁺ Tyrode's solution containing the detergent Pluronic F127 (0.025%, w/v) for 60 min at 23°C. The indo-1-loaded crypts were washed with the Parsons solution and placed in a glass chamber, the bottom of which was coated with poly(L-lysine). The indo-1 fluorescence of the single cells located at the middle of isolated crypts was monitored at emission wavelengths of 405 and 485 nm by using the ACAS 570 interactive confocal laser cytometer (Meridian, Okemos, MI) as described elsewhere (Ikari et al., 1999). After correction for background fluorescence, the intensity ratio (405/485 nm) and [Ca²⁺]_i were calculated as previously described (Grynkiewicz et al., 1985).

Statistics. Results are presented as the mean ± S.E. Differences between groups were analyzed by one-way ANOVA, and correction for multiple comparisons was made by using Dunnett's multiple comparison test. If necessary, Tukey's multiple comparison test was used. Comparison between the two groups was made with Student's t test. Statistically significant differences were assumed at P < .05. The IC₅₀ values of data were calculated by using the KaleidaGraph program, version 3.08 (Synergy Software, Reading, PA).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Cl Secretion Induced by Irinotecan, STA₂, and PGE₂ in Isolated Rat Colon. We found recently that exogenous addition of STA₂ (Katsura et al., 1983), a stable TXA₂ analog, mimics the action of endogenous TXA₂, released by antitumor drug irinotecan, in the rat colon; that is, both of them caused the Cl secretion via a TXA₂ receptor located in epithelial crypt cells (Sakai et al., 1997). These effects were not affected by tetrodotoxin, a neuronal inhibitor (Sakai et al., 1995, 1997).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Inhibitory effect of loperamide on the irinotecan-induced increase in Isc. A, typical trace of Isc is shown where atropine (5 µM at the serosal side) was added before stimulation with irinotecan (500 µM at both the serosal and mucosal sides). When the plateau phase of Isc was observed after the addition of irinotecan, loperamide (30 µM at the serosal side) was added. B, loperamide was added cumulatively at the serosal side when the plateau phase of Isc was observed after the addition of 500 µM irinotecan in the presence of 5 µM atropine. The Isc values were measured when the effect of loperamide had become steady, and data are expressed as net increases from the Isc just before addition of irinotecan (Isc). Data are mean ± S.E. from five experiments. **, significantly different from the value in the absence of loperamide (P < .01).

Inhibition of Irinotecan-Induced Cl Secretion by Loperamide. Beubler and Badhri (1990) reported that loperamide does not act from the intestinal luminal side, but acts after delivery from the blood in vivo. Loperamide was therefore added to the serosal side after stimulation with irinotecan. Figure 1B shows that the irinotecan-induced Cl current was inhibited in a concentration-dependent manner by loperamide, and its IC₅₀ value was 0.7 µM. Increases of Pd and Gt elicited by irinotecan were significantly suppressed by 30 µM loperamide (Fig. 2). Loperamide (30 µM) slightly decreased the basal Isc in the absence of irinotecan, but the effect was not significant: the values before and after the addition of loperamide were 20.0 ± 2.0 and 15.7 ± 2.3 µA/cm², respectively (n = 5; P > .05).

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Effects of loperamide on the irinotecan-induced increases in Pd and Gt. The Pd (A) and Gt (B) values () were measured just before addition of irinotecan (500 µM at both the serosal and mucosal sides), when values of Isc had established their steady states after addition of irinotecan (finely hatched columns), and after the additional application of loperamide (roughly hatched columns) (30 µM at the serosal side). The data were obtained from five experiments. *P < .05 and **P < .01.

Inhibition of STA₂- and PGE₂-Induced Cl Secretion by Loperamide. Both STA₂- and PGE₂-induced Cl currents were inhibited in a concentration-dependent manner by loperamide (Fig. 3). But the inhibitory effect of loperamide on the STA₂-induced current (IC₅₀ = 1.2 µM; Fig. 3A) was greater than that on the PGE₂-induced current (IC₅₀ = 23 µM; Fig. 3B). The efficacy of loperamide for the STA₂-induced current (Fig. 3A) is comparable with that for the irinotecan-induced current (Fig. 1B). Increases of Pd and Gt elicited by STA₂ and PGE₂ were significantly suppressed by 30 µM loperamide (Fig. 4).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Inhibitory effects of loperamide on the STA2- and PGE2-induced increases in Isc. Loperamide was added cumulatively at the serosal side when the plateau phase of Isc was observed after the addition of STA2 (1 µM at the serosal side; A) or PGE2 (0.5 µM at the serosal side; B) in the presence (, ) or absence () of atropine (5 µM at the serosal side). The Isc values were measured when the effect of loperamide had become steady, and data are expressed as net increases from the Isc just before addition of STA2 or PGE2 (Isc). Data are mean ± S.E. from five to seven experiments. *P < .05 and **P < .01, significantly different from the value in the absence of loperamide. Inset, typical traces showing the inhibition of Isc by 30 µM loperamide.

View larger version (41K): [in this window] [in a new window]   Fig. 4.   Effects of loperamide on the STA2- and PGE2-induced increases in Pd and Gt. The Pd (A and B) and Gt (C and D) values () were measured just before addition of STA2 (1 µM at the serosal side; A and C) or PGE2 (0.5 µM at the serosal side; B and D), when values of Isc had established their steady states after addition of the secretagogue (finely hatched columns), and after the additional application of loperamide (roughly hatched columns) (30 µM at the serosal side). The data were obtained from five experiments. *P < .05 and **P < .01.

Opiate Receptor-Independent Action of Loperamide. Loperamide is known to behave like an opiate agonist (Ooms et al., 1984; Awouters et al., 1993). We therefore checked the effects of naloxone, a competitive opiate antagonist, on the STA₂- and PGE₂-induced Cl secretion in isolated colonic mucosa. Naloxone (10 µM at the serosal side) per se did not affect the basal Isc; that is, the values before and after the addition of naloxone were 33.3 ± 1.5 and 31.9 ± 1.7 µA/cm², respectively (n = 6; P > .05). Figure 5 shows that naloxone (10 µM) does not significantly attenuate the inhibitory effects of loperamide (30 µM) on the STA₂- and PGE₂-induced Cl secretion, indicating that these loperamide effects appear not to be mediated via opiate receptors.

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Effects of naloxone on the loperamide-induced inhibition of Isc. Typical traces of Isc are shown where naloxone (10 µM at the serosal side) and atropine (5 µM at the serosal side) were added before stimulation with STA2 (1 µM at the serosal side; A) or PGE2 (0.5 µM at the serosal side; B). When the plateau phase of Isc was observed after the addition of STA2 or PGE2, loperamide (30 µM at the serosal side) was added. C and D, inhibitory effects of 30 µM loperamide in the presence () and absence () of 10 µM naloxone are expressed as percentage of inhibition, where the data were measured as described in the legend to Fig. 3. One hundred percent corresponds to complete inhibition of the net STA2- (C) or PGE2-induced (D) increase of Isc. Note that 30 µM loperamide inhibits the STA2-induced effect (C) more efficiently than the PGE2-induced effect (D). NS, not significant.

Inhibitory Effects of a Calmodulin Antagonist on the STA₂- and PGE₂-Induced Cl Secretion. Herein, we examined whether W-7, a specific calmodulin antagonist (Hidaka et al., 1981b), mimics the antisecretory effect of loperamide. Figure 6 shows that W-7 inhibits both the STA₂- and PGE₂-induced secretion in a concentration-dependent manner. The inhibitory effect of the drug on the STA₂-induced secretion (IC₅₀ = 5 µM; Fig. 6A) was greater than the PGE₂-induced secretion (IC₅₀ = 36 µM; Fig. 6B), similar to the case found with loperamide (Fig. 3). We also checked the effect of W-5 on the STA₂-induced secretion. Because W-5 is a dechlorinated derivative of W-7 and has about 7 times lower affinity than W-7 for calmodulin (Hidaka et al., 1981a), it has been used to confirm the specificity of W-7 on calmodulin. In the present preparation, the inhibition of the STA₂-induced secretion by 30 µM W-5 (34.9 ± 8.3% inhibition; n = 4) was much less than that by 30 µM W-7 (80.7 ± 8.8% inhibition; n = 7).

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Inhibitory effects of W-7 on the STA2- and PGE2- induced increases in Isc. W-7 was added cumulatively at the serosal side when the plateau phase of Isc was observed after the addition of STA2 (1 µM at the serosal side; A) or PGE2 (0.5 µM at the serosal side; B) in the presence of atropine (5 µM at the serosal side). The values of Isc were measured when the effect of W-7 had become steady, and data are expressed as net increases from the Isc just before addition of STA2 or PGE2 (Isc). Data are mean ± S.E. from four experiments. *P < .05 and **P < .01, significantly different from the value in the absence of W-7. Inset, typical traces showing the inhibition of Isc by 30 µM W-7.

Effect of Loperamide on STA₂-Induced Increase in [Ca²⁺]_i in Colonic Crypt Cells. We have recently shown that STA₂ increased [Ca²⁺]_i with a transient peak phase followed by a subsequent plateau phase in single cells of isolated colonic crypts, whereas PGE₂ did not increase the [Ca²⁺]_i (Ikari et al., 1999). As shown in Fig. 7, loperamide (10 µM) did not affect on the STA₂ (0.1 µM)-induced increase in [Ca²⁺]_i, indicating that loperamide does not inhibit Ca²⁺ channels in colonic crypt cells. It is noted that colonic crypts were preincubated with loperamide for 10 min to see its effect on intracellular Ca²⁺ stores (Fig. 7A).

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Effect of loperamide on the STA2-induced increase in [Ca2+]i of single cells in isolated colonic crypts. The STA2 (0.1 µM)-induced increase in [Ca2+]i was measured in the presence () and absence () of 10 µM loperamide. Loperamide was added 10 min before (A) and 70 s after (B) the addition of STA2. Data are mean ± S.E. from 10 to 20 experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Diarrhea is a serious side effect and a dose-limiting toxicity of irinotecan, which has a strong anticancer activity against many types of tumor. Loperamide has been used to allow administration of higher doses of irinotecan in Europe and the United States. In fact, loperamide was reported to be effective in controlling diarrhea in the patients receiving irinotecan (Abigerges et al., 1994; Armand et al., 1996; Rothenberg, 1998; Saliba et al., 1998). In contrast to the case for irinotecan, diarrhea is not a dose-limiting toxicity of other topoisomerase I inhibitors such as topotecan (9-dimethylaminomethyl-10-hydroxycamptothecin) and rubitecan (9-nitrocamptothecin), although they infrequently cause diarrhea (Gerrits et al., 1997).

We found previously that irinotecan induces Cl secretion in the rat colon, which may account for one of mechanisms of the diarrhea (Sakai et al., 1995). Recently, we found that the irinotecan-induced Cl secretion in the colon is mainly mediated by release of TXA₂, and that a stable TXA₂ analog (STA₂) mimics the effect of irinotecan (Sakai et al., 1997). STA₂ acts on the TXA₂ receptor in epithelial crypt cells (Sakai et al., 1997; Ikari et al., 1999). TXA₂ has been suggested as a mediator of inflammatory bowel diseases in an animal model (Taniguchi et al., 1997) and in human (Rampton and Collins, 1993; Casellas et al., 1995). In fact, oral administration of a thromboxane synthase inhibitor to patients with the disease showed clinical and colonoscopic improvements, and the inhibitor significantly and selectively reduced the release of TXB₂ (Casellas et al., 1995). These results suggest that TXA₂ is a novel pathophysiological mediator in the colon.

In the present study, we have shown that loperamide inhibits the irinotecan- and STA₂-induced Cl secretion more effectively than the PGE₂-induced secretion in isolated rat colonic mucosa (Figs. 1 and 3), indicating that loperamide is highly involved in the TXA₂-elicited pathway. TXA₂ receptor was cloned and found to link with both the Ca²⁺- and cAMP-signaling pathways (Hirata et al., 1996). However, the colonic PGE₂ receptor is an EP2 subtype and coupled to only a cAMP-signaling pathway (Homaidan et al., 1995). In fact, we found that PGE₂ did not increase [Ca²⁺]_i in colonic crypt cells but STA₂ did (Ikari et al., 1999). Diener et al. (1988b) reported that ~10 times higher concentration of loperamide was necessary to block the secretion caused by forskolin (mediated via cAMP pathway) than to block the secretion caused by carbachol (mediated via the Ca²⁺ pathway) in isolated rat colon.

The inhibitory effects of loperamide on the STA₂- and PGE₂-induced Cl secretion were not mediated by opiate receptors (Fig. 5). In isolated colonic mucosa, naloxone, an opiate antagonist, did not reverse inhibitory effects of loperamide on secretion stimulated by electric field (Diener et al., 1988b; Burleigh, 1991), carbachol, forskolin (Diener et al., 1988b), or PGE₁ plus theophylline (Kromer, 1995). In the rat colon, loperamide did not affect the choleratoxin-increased cAMP levels, whereas it inhibited the choleratoxin-induced fluid secretion (Farack et al., 1981).

We examined whether the Ca²⁺-dependent pathway is a target of loperamide (Figs. 6 and 7). Besides opiate receptors, Ca²⁺ channels (Reynolds et al., 1984; Chang et al., 1986) and calmodulin (Zavecz et al., 1982; Diener et al., 1988b) have been considered to be target molecules for antisecretory action of loperamide. In the present study, a specific calmodulin antagonist, W-7, mimicked the antisecretory effects of loperamide; that is, W-7 inhibited the STA₂-induced Cl secretion more effectively than the PGE₂-induced Cl secretion (Fig. 6). Diener et al. (1988b) reported that trifluoperazine, a calmodulin antagonist, mimicked the inhibitory effects of loperamide against the carbachol- and forskolin-induced secretion in the rat colon. However, loperamide did not affect the STA₂-induced increase in [Ca²⁺]_i of colonic crypt cells (Fig. 7). Taken together, we suggest that loperamide blocks the TXA₂-induced Cl secretion by inhibiting the calmodulin system. Ca²⁺ channels are not the sites of action of loperamide, at least, in isolated rat colonic mucosa. Higher concentration of loperamide may interfere with cross talk between the PGE₂-elicited cAMP-dependent system and the endogenous calmodulin system.

Loperamide is a safe and effective antidiarrhea drug (Ericsson and Johnson, 1990). We found in the present study that the irinotecan- and STA₂-induced Cl secretion is highly sensitive to loperamide in contrast to the PGE₂-induced Cl secretion. Our results may explain, at least in part, why loperamide is clinically efficient for arresting the irinotecan-induced diarrhea.