Introduction

Abstract Introduction Materials & Methods Results Discussion References

Clinically, acute pancreatitis exhibits a broad spectrum of pathologic changes of varying severity: mild edematous pancreatitis may resolve spontaneously or after conservative therapy; severe hemorrhagic pancreatitis has a high mortality due to multiple organ failure (Geokas et al., 1985; Pitchumoni et al., 1988). In spite of numerous studies, etiologic factors involved in the initiation and aggravation of acute pancreatitis have not been well defined, and the treatment of the disease is still to be established (Niederau and Schulz, 1993).

Although there is a general belief that autodigestion by activated pancreatic enzymes plays an important role in the initiation of pancreatitis (Becker, 1981; Geokas et al., 1972; Trapnell, 1981), impairment of the pancreatic blood flow has been stressed as an important event in its progression (Anderson and Schiller, 1968; Schiller and Anderson, 1975; Klar et al., 1990). Ligation of the pancreatic duct or injection of supramaximal doses of a pancreatic secretagogue, which caused edematous pancreatitis, combined with subsequent interruption of arterial supply (Popper et al., 1948) or production of venous thrombosis (Adams and Musselman, 1953; Anderson, 1963) consistently induced fatal necrotizing pancreatitis. It is thus interesting to note the observation by Prinz et al. (1984) that i.v. administration of pancreatic fluid caused platelet aggregation and resulted in the release of 5-HT. The released 5-HT then caused further platelet aggregation and 5-HT release (De Clerck et al., 1982; Hara et al., 1991), a positive feedback that no doubt led to a thrombus formation (Marcus, 1982). Also, 5-HT is considered a major mediator of the vasoconstriction induced by substances released from the platelets (McGoon and Vanhoutte, 1984). These results suggest that 5-HT may be an aggravating factor for acute pancreatitis initiated by activated pancreatic enzymes.

The present study was designed to explore the effect of 5-HT depletion and 5-HT receptor blockade on acute pancreatitis. We used a CDE diet to induce necrotizing acute pancreatitis in mice, the pathogenesis of which resembles that in the human (Lombardi et al., 1975; Niederau et al., 1994).

	Materials and Methods

Abstract Introduction Materials & Methods Results Discussion References

Drugs

The following drugs and chemicals were used in this study: p-CPA, pindolol, cyproheptadine hydrochloride (Sigma Chemical Co., St. Louis, MO), NAN-190 (1-(2-methoxy-phenyl)-4-[4-(2-phthalimido)butyl] piperazine hydrobromide), M-840 ([[3-(1-methyl-1H-indol-3-yl)-1,2,4-oxadiazol-5-yl]methyl] trimethylammonium iodide) (Cookson Chemicals Ltd., Southampton, England), ketanserin tartrate and ICS205-930 (3-tropanyl-indole-3-carboxylate hydrochloride) (Research Biochemicals Incorporated, Natick, MA). p-CPA was dissolved in saline, and the other drugs were suspended in 0.5% methylcellulose. All drug solutions were prepared freshly and administered p.o. in a volume of 10 ml/kg. The control mice were given the respective vehicles.

[³H]8-OH-DPAT (4621 GBq/mmol), [³H]ketanserin hydrochloride (2808 GBq/mmol) and [³H]GR65630 (2379 GBq/mmol) were obtained from DuPont/New England Nuclear (Boston, MA) and were used as 5-HT_1A, 5-HT₂ and 5-HT₃ receptor binding radioligands, respectively.

Animals

Female ICR mice that were 3 weeks old and weighed 12 to 17 g were purchased from Charles River Inc. (Atsugi, Japan). The animals were kept in our laboratory for 3 days under conditions of 22 ± 1°C and 12-hr light and dark cycles with lights on at 8:00 and then were used for the acute pancreatitis studies. Male Sprague-Dawley rats that were 7 weeks old and weighed 250 to 350 g (Japan Clea Inc., Tokyo, Japan) were used for membrane preparation in the 5-HT binding study. All animal procedures were carried out as approved by the Animal Care and Use Committee at Fujisawa Pharmaceutical Co. Ltd.

Diet-Induced Acute Pancreatitis

Acute pancreatitis was induced by feeding the mice a choline-deficient, 0.5% D,L-ethionine supplemented diet (Japan Clea Inc., Tokyo, Japan) ad libitum for 48 hr, starting at 10:00. Before and after the CDE diet, the mice were fed a normal diet. The control mice were fed a normal diet throughout the experiment.

Time course study. Five mice each were used for the CDE diet groups and the corresponding controls. The animals were killed by heart puncture under ether anesthesia before (day 0) or 1, 2, 3 or 4 days after the beginning of the experiment (10:00-12:00). Blood samples were collected for determination of serum amylase and lipase levels. In another set of experiments, blood samples were collected using a syringe containing 60 µl of EDTA (27 mmol/l in saline), and the plasma was separated by centrifugation and determined for 5-HT and 5-HIAA levels. The blood samples that clotted during the procedure were not used for 5-HT and 5-HIAA measurement. Pancreas samples were taken and used for histologic examinations. The experiments were performed twice to confirm the reproducibility of the results. Some of the CDE diet mice died during the experiment and were excluded from the data. The results were combined and expressed as the mean ± S.E. (n = 4  10).

Drug effect. Mice were randomly assigned to different treatment groups (n = 10) and fed the CDE diet as described above. p-CPA was administered p.o. 72 hr and 48 hr before the beginning of the CDE diet to deplete endogenous 5-HT. 5-HT antagonists were given p.o. twice a day (10:00 and 17:00) for 4 days, starting with the introduction of the CDE diet. All the mice were followed up to for 7 days, and mortality in each group was determined. The results of two or three series of experiments were combined and expressed as the mean of 20 or 30 animals.

Histologic Examination

The pancreata were fixed with 10% buffered formaldehyde, embedded in paraffin and sectioned to 3-µm thickness. The sections were stained with hematoxylin and eosin and examined by light microscopy.

To examine the capillary blood distribution, mice were given an i.v. injection of carbon particle solution (drawing ink, Rötringwerke, Hamburg, FRG, average particle size about 160 µm) to make a 30% suspension in saline containing 1% gelatin and were killed 5 min later by CO₂ inhalation. The pancreata were fixed with 10% buffered formaldehyde, dehydrated by ethanol and clarified with methyl salicylate. The capillaries were observed stereomicroscopically.

Biochemical Analysis of Blood

Serum amylase and lipase levels were determined by the modified (blocked) p-nitrophenylmaltoheptaoside method (Denka Seiken Co., Ltd., Tokyo, Japan) and the monoglyceridelipase method (Nippon Shoji Co., Ltd., Osaka, Japan), respectively. It is possible that -amylase isoenzymes of salivary type contributed to the amylase assay (Dupuy et al., 1987). The lipase assay is specific for the pancreatic enzymes. Both enzyme assays were performed using an auto analyzer (model TBA-20R, Toshiba, Tokyo, Japan).

Plasma levels of 5-HT and 5-HIAA, a metabolite of 5-HT, were measured using a HPLC-ECD method according to Kumar et al. (1990) with slight modification. The HPLC-ECD system consisted of a pump (EP-10, Eicom, Kyoto, Japan), a reverse-phase chromatographic column (Eicompak MA-5ODS, 4.6 × 150 mm; Eicom) and an electrochemical detector (ECD-100; Eicom) equipped with a WE-3G glassy carbon electrode. The applied potential at the working electrode was +750 mV vs. an Ag/AgCl reference. The mobile phase consisted of 0.1 M citric acid-sodium acetate buffer (pH 3.5) containing 5 mg/l EDTA(2Na), 100 mg/l sodium octyl sulfate and 15% methanol that had been filtered and degassed before use. The flow rate was 1 ml/min. The plasma samples (250 µl) were combined with 50 µl of isoproterenol (10 ng/ml) as an internal standard, deproteinized by vortexing with 7 µl of 70% perchloric acid and centrifuged at 20,000 × g for 15 min at 4°C. The supernatant was aliquoted, adjusted to pH 3 using 1 M sodium acetate and injected into the HPLC-ECD within 24 hr after it was prepared. The extracts were found to be stable at 4°C for 30 hr, after which there was a slow degradation.

Receptor Binding Assay

The receptor binding assay was performed using rat cerebral cortex membrane. Membrane preparation and the binding assay were performed as described previously (Nomura et al., 1994).

Frozen (80°C) cerebral cortex samples obtained by decapitation of male Sprague-Dawley rats were homogenized in ice-cold 0.32 M sucrose (1:10, w/v), centrifuged at 900 × g for 10 min at 4°C and the supernatant centrifuged at 70,000 × g for 15 min at 4°C. The pellet was resuspended in 10 volumes of 50 mM Tris-HCl, pH 7.5, incubated at 37°C for 15 min and then centrifuged at 70,000 × g for 15 min. The final pellet was resuspended in 2.5 volumes of 50 mM Tris-HCl, pH 7.7, containing 4 mM CaCl₂ and 0.1% ascorbic acid and stored at 80°C until it was used. In the 5-HT₃ receptor binding assay, 50 mM HEPES was used instead of Tris-HCl. The final tissue concentrations for the 5-HT_1A, 5-HT₂ and 5-HT₃ receptor binding assays were prepared in 20, 80 and 15 volumes, respectively, by using the buffers described.

Receptor binding assays for 5-HT_1A and 5-HT₂ receptor were performed in duplicate in a final volume of 1 ml. The assay buffers were identical to the tissue buffers described. Competition analyses were performed using 1.5 and 2 nM of [³H]8-OH-DPAT and [³H]ketanserin, respectively. Assay tubes were incubated for 30 min at 37°C, filtered through Whatman (Clifton, NJ) GF/B filters presoaked in 0.5% polyethyleneimine and washed with 20 ml of cold buffer. The filters were counted by liquid scintillation spectrometry (TRI/CARB 4530, Packard, Downers Grove, IL) in 8 ml of aqueous counting scintillant (Aquazol, DuPont/New England Nuclear) after overnight equilibration. Nonspecific binding of 5-HT_1A and 5-HT₂ binding was determined in the presence of 10 µM of 5-HT and 1 µM of mianserin, respectively. The 5-HT₃ binding assay was performed in a final volume of 0.8 ml in the aforementioned buffer containing 10 µM pargyline and 0.1% ascorbic acid. Competition analysis was performed using 0.2 nM of [³H]GR65630, and the nonspecific binding was determined in the presence of 30 µM of metoclopramide. Incubation, filtration and counting were performed as described above. Each IC₅₀ value was obtained from an inhibition curve by a computerized program, and the K_i value was calculated.

Statistical Analysis

Statistical significance of time course changes in serum and plasma parameters were evaluated by two-way analysis of variance (ANOVA), followed by Student's t test for determination of the differences between CDE diet mice and the control at each time-point. For evaluation of the drug effect on the serum pancreatic enzymes, we used one-way ANOVA followed by Dunnett's t test. Mortality of the treatment groups was compared with the vehicle administered, CDE diet mice, using a Fisher's exact probability test. Significance was assumed for P < .05.

	Results

Abstract Introduction Materials & Methods Results Discussion References

Morphologic and biochemical evidence in CDE diet mice. Figure 1A shows the stereomicroscopy of the pancreas of a control mouse (day 0), in which the pancreatic vascular tree was made clearly visible by injection of carbon particles. The capillaries, arterioles and/or arteries were evenly filled with the carbon particles. Compared with the stereomicroscopy of the control animal, there was no difference in the distribution of the carbon particles in the CDE diet mice on day 1, but the particles were concentrated in the dilated large vessels on day 2 (figures not shown). The change was more marked on day 3 (fig. 1C), when hardly any filling of the capillaries and marked dilation of the prestenotic parts were observed.

View larger version (152K): [in this window] [in a new window]   Fig. 1.   Morphologic changes in the pancreatitis mice. A) Stereomicroscopy of the pancreas from the control mice shows the carbon particles evenly distributed among the intralobular capillaries (LC) and intralobular arterioles (LA). Original magnification ×6. B) Light microscopy of the pancreas from the control mice shows normal architecture. Hematoxylin and eosin staining (H&E); original magnification ×100. C) Stereomicroscopy of the pancreas from the CDE diet mice on day 3 (surviving mice). Note lack of filling of the intralobular capillaries and markedly dilated intralobular arterioles. Original magnification ×6. D) Light microscopy of the pancreas from the CDE diet mice on day 3. Note interstitial edema characterized by expansion of interlobular and interacinar spaces, multiple cytoplasmic vacuolization in acinar cells (short arrows), inflammatory cells, neutrophils predominant in the interstitial space (arrow heads) and massive necrosis of the acinar cells (pale areas, long arrows). H&E; original magnification ×100.

Effect of drugs on mortality of the CDE diet mice. After the morphologic and biochemical changes mentioned, death occurred among the CDE diet mice, mostly between days 2 and 4 (fig. 2). Usually none of the animals died later than day 4. The average mortalities at days 3, 4 and 7 were 41.8, 74.5 and 75.4%, respectively. No deaths occurred among the mice fed the control diet.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Cumulative mortality in the pancreatitis mice. Data were collected from the control groups of the experiments on drug effect. Each value represents the mean ± S.E. of 11 separate experiments (n = 10 for each experiment).

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Effect of p-CPA treatment on mortality (bars) and food consumption (lines) of the pancreatitis mice. ** P < .01, *** P < .001; statistically significant compared with the control by Fisher's exact probability test.

View larger version (34K): [in this window] [in a new window]   Fig. 4.   Effects of serotonin receptor antagonists on mortality (bars) and food consumption (lines) of the pancreatitis mice. * P < .05, ** P < .01, *** P < .001; statistically significant compared with the control by Fisher's exact probability test.

Binding affinities for 5-HT receptor subtypes of 5-HT antagonists. Binding affinities of drugs for 5-HT_1A, 5-HT₂ and 5-HT₃ receptors in rat cerebral cortex are presented as K_i values in table 3. NAN-190 showed the highest affinity to 5-HT_1A receptor, followed by pindolol > cyproheptadine > ketanserin, whereas ICS205-930 and M-840 had little affinity to this receptor. In the 5-HT₂ binding study, the highest affinity was observed with ketanserin and cyproheptadine, followed by NAN-190 > pindolol > ICS205-930 > M-840. M-840 had the highest affinity to 5-HT₃ receptor, followed by ICS205-930, but the other compounds had little affinity to it.

Relationship between in vivo and in vitro potencies of 5-HT antagonists. Figure 5 shows the relationship between the antipancreatitis activities and the 5-HT receptor binding affinities of the 5-HT antagonists; pK_i values of the binding affinities to 5-HT_1A (fig. 5A), 5-HT₂ (fig. 5B) and 5-HT₃ (fig. 5C) receptors were plotted against the pED₃₀ values of the antipancreatitis activities. The pED₃₀ values were significantly (r = 0.914, P < .001) correlated with pK_i values for the 5-HT₂ receptors, but not with those for the 5-HT_1A or 5-HT₃ receptors.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Relationship between in vivo and in vitro potency of serotonin receptor antagonists.

Effects of ketanserin and cyproheptadine on CDE diet-induced changes in serum enzyme levels and pancreatic morphology. The smaller doses of ketanserin (1.0 and 3.2 mg/kg) hardly affected the increased serum amylase and lipase levels in the CDE diet mice on day 3, which, however, were significantly attenuated by the largest dose of the drug (table 4). On the other hand, cyproheptadine hardly affected the serum amylase and lipase levels in the mice at any dose.

View larger version (145K): [in this window] [in a new window]   Fig. 6.   Effects of ketanserin (left) and cyproheptadine (right) on the pancreatic morphology in the pancreatitis mice. Drugs (3.2 mg/kg) were given p.o. twice a day from the beginning of the CDE diet, and pancreas samples were taken from the surviving mice on day 3. A) Stereomicroscopy of the pancreas from the pancreatitis mice treated with ketanserin shows carbon particles distributed evenly over the intralobular capillaries (LC) and shows intralobular arterioles (LA) of almost the same diameter as those in the normal control in figure 1A. Original magnification ×6. B) Light microscopy of the pancreas from the pancreatitis mice treated with ketanserin shows marked diminution of interstitial edema (decreased interlobular spaces) and acinar cell necrosis (pale areas, long arrows), as well as lack of inflammatory cellular infiltration, but multiple cytoplasmic vacuoles in acinar cells (short arrows) are still evident, compared with the untreated pancreatitis mice in figure 1D. H&E; original magnification ×100. C) Stereomicroscopy of the pancreas from the pancreatitis mice treated with cyproheptadine. Note the attenuated vascular changes. Original magnification ×6. D) Light microscopy of the pancreas from the pancreatitis mice treated with cyproheptadine shows moderate interlobular edema and acinar cell necrosis (pale areas, long arrows). There is no evidence of inflammatory cellular infiltration, but acinar cell vacuolization (short arrows) is present. H&E; original magnification ×100.

	Discussion

Abstract Introduction Materials & Methods Results Discussion References

The present study confirmed the work of others who reported that feeding a CDE diet caused fatal acute pancreatitis associated with increased circulating pancreatic enzymes in mice (Lombardi et al., 1975; Niederau et al., 1994). In addition, the present stereomicroscopy indicated circulatory disturbance in the pancreas of the CDE diet mice: i.v. injected carbon particles were distributed evenly over the pancreas in the control mice but in the CDE diet mice were concentrated mainly in the dilated arterioles and were scarcely seen in the microcirculation. It has been speculated that a CDE diet inhibits the biosynthesis of lecithins, a major membrane constituent, to cause pancreatic enzyme leakage into the blood (Lombardi et al., 1975). Although activated pancreatic enzymes, when they gain access to the parenchyma, cause local autodigestion that initiates pancreatitis (Becker, 1981; Geokas et al., 1972; Trapnell, 1981), pancreatic blood flow is also considered an important factor in the progression of pancreatitis (Anderson, 1963; Anderson and Schiller, 1968; Schiller and Anderson, 1975). Pfeffer et al. (1962) showed that various degrees of pancreatitis could be produced by an intra-arterial injection of polyethylene microspheres of various sizes in dogs. Further, edematous pancreatitis can be turned into fatal pancreatic necrosis by venous ligation (Anderson, 1963) and hemorrhagic shock (Kyogoku et al., 1992). Taken together, these results suggest that the local autodigestion by activated enzymes, which appears to be minimal in the normal pancreas because of the absorption of the enzymes into circulation, could be exaggerated by circulatory disturbance in the pancreas of the CDE diet mice.

One of the most important findings in the present study is that the plasma levels of 5-HIAA were significantly increased by the treatment. Plasma 5-HT levels also tended to increase in the CDE diet mice compared with the control, though the difference between the two was not statistically significant. It is well known that the half-life of the released 5-HT is short, which may explain the different changes in plasma 5-HIAA and 5-HT levels. On the other hand, plasma 5-HIAA is rather stable in the circulation but biologically inactive. We thus regarded 5-HIAA as an index of 5-HT release and speculated that 5-HT release may be enhanced in the pancreatitis. We also found that pretreatment with p-CPA, a 5-HT depletor, significantly attenuated the mortality of the CDE diet mice and that p.o. dosing with 5-HT antagonists such as pindolol, NAN-190, ketanserin and cyproheptadine likewise had a significantly attenuating effect. Their activities significantly correlated with their binding affinities for 5-HT₂ receptors but not for 5-HT_1A or 5-HT₃ receptors. Furthermore, ketanserin and cyproheptadine attenuated to a large extent the interstitial edema, necrosis of the acinar cells and neutrophilic infiltration in the CDE diet mice. Although we have not identified the pancreatitis-associated protein that indicates the severity of pancreatitis (Iovanna et al., 1994), Gukovskaya et al. (1996) reported that the severity of pancreatitis is correlated with necrosis and neutrophillic infiltration. These results, taken together with ours, suggest that 5-HT₂ receptor activation through endogenous 5-HT release plays an important role in the development of acute pancreatitis. However, there still remains the possibility that some of the actions of the drugs in preventing the pathologic changes were unrelated to 5-HT-mediated mechanisms, because both of the 5-HT2 antagonists we used in this study are known to have actions unrelated to its serotonergic effects in addition to blockade of 5-HT2 receptors. Further study may necessary to clarify this point.

Another interesting finding of the present study is that the aforementioned changes in 5-HIAA level in the CDE diet mice preceded the increase in the serum levels of pancreatic enzymes such as amylase and lipase; the enzyme levels were either unchanged or reduced on the first day, whereas the 5-HIAA level was increased by about 2 to 3-fold compared with the control. In the present study, we found that ketanserin dose-dependently attenuated the increase in the levels of enzymes in the CDE diet mice. Similar results were obtained by Oguchi et al. (1992), who reported that ketanserin reduced the increase in serum amylase concentration in cerulein-induced pancreatitis rats. However, we also found that cyproheptadine hardly attenuated the increase in the serum enzyme levels. It is thus reasonable to assume that 5-HT₂ receptor activation does not play a role in the release of the pancreatic enzymes.

Ketanserin and cyproheptadine markedly attenuated not only the histologic changes but also the vascular changes of the pancreas at doses that hardly affected the serum enzymes levels. Attenuation of the pancreatitis-induced alteration of pancreatic circulation, rather than enzyme release, appears to be the main mechanism of action of these drugs, because the increase in the 5-HT and 5-HIAA levels in the CDE diet mice preceded the vascular change. An increase in circulating 5-HT in the CDE diet mice would cause platelet aggregation by synergistically potentiating the effect of the other endogenous substances (De Clerck et al., 1982; Hara et al., 1991), which would result in formation of a thrombus. On the other hand, 5-HT is considered a major mediator of the vasoconstriction among the substances released from platelets (McGoon and Vanhoutte, 1984). These two actions of 5-HT, which are attributed to 5-HT₂ receptor activation, no doubt lead to circulatory disturbance and may have contributed to the aggravation of pancreatitis in the CDE diet mice. In this respect, it is interesting to note that edematous pancreatitis can be turned into fatal pancreatic necrosis by venous thrombosis (Adams and Musselman, 1953) or by dosing with phenylephrine, a vasoconstrictor (Klar et al., 1991), whereas fibrinolysin (Wright and Goodhead, 1970), heparin (Wright and Goodhead, 1970) and sympathetic blockade (Goodhead and Wright, 1969) prevent the development of hemorrhagic pancreatitis.

As for the source of 5-HT, one possibility is platelets; Prinz et al. (1984) reported that i.v. administration of pancreatic fluid caused platelet aggregation and resulted in the release of 5-HT. If this is the case, however, increased serum enzyme levels should precede or at least coincide with the 5-HT release. As we have noted, the contrary was observed in the present study. Thus only the later (not the initial) increase in 5-HT release could be explained by platelet aggregation. Enterochromaffin cells distributed throughout the GI tract are another possibility. Celinski et al. (1995) have recently shown that there was a significant fall in 5-HT level accompanied by a relevant increase in 5-HIAA level in the stomach of rats with cerulein-induced pancreatitis. The inhibitory effects of a CDE diet on lecithin formation, as mentioned above, may derange the membrane permeability of enterochromaffin cells and/or platelets, and this should be clarified by further experiments.

In conclusion, our results suggest that activation of 5-HT₂ receptors by an increase in circulating 5-HT plays an important role in the pathogenesis of CDE diet-induced acute pancreatitis in mice, possibly through the induction of pancreatic circulatory disturbances.