GASTROENTEROLOGY

1992:102:598-804

Stimulatory and Inhibitory Effects of Bile Salts on Rat Pancreatic Secretion KYOKO MIYASAKA, AKIHIRO FUNAKOSHI, FUKUKO SHIKADO, and KENICHI KITANI Department of Clinical Physiology, Tokyo Metropolitan Institute of Gerontology, Tokyo, Japan; and National Ky&.hu Cancer Center, F;kuokay Japan

The effects of various species of bile salts (chenodeoxycholate, deoxycholate, ursodeoxycholate and cholate, and their taurine and glycine conjugates) on pancreatic exocrine secretion were studied in conscious rats with external bile and pancreatic fistulae. For examination of the stimulatory effects of bile salts, bile and pancreatic juice were collected for a basal period of 60 minutes and returned to the intestine, and then solutions of bile salts (60 mmol/ L) were infused intraduodenally at a rate of 1 mL/h for 2 hours. For examination of their inhibitory effects, pancreatic secretion was stimulated by exclusion of the bile and pancreatic juice; and then solutions of the bile salts were again infused intraduodenally. Chenodeoxycholate, glycochenodeoxycholate, ursodeoxycholate, deoxycholate, and its conjugates (glycodeoxycholate and taurodeoxycholate) significantly increased the fluid, bicarbonate and protein outputs, and bicarbonate concentration, with decrease in protein concentration. These increases were partially inhibited by infusion of either a cholecystokinin antagonist or secretin antibody. In contrast, cholate, taurocholate, tauroursodeoxycholate, glycoursodeoxycholate, and taurochenodeoxycholate inhibited pancreatic secretion and increase in the plasma cholecystokinin concentration produced by exclusion of bile and pancreatic juice. Thus, some bile salts, including taurocholate and taurochenodeoxycholate (major bile salts in rat bile) inhibited pancreatic secretion and cholecystokinin release, whereas some other bile salts increased pancreatic secretion via cholecystokinin release and secretin release. ile and pancreatic enzymes in the intestinal lumen are known to work together in the process of digestion. However, the effect of luminal bile (or bile salts) on the pancreas is controversial. Studies on rats have consistently shown that the exclusion of bile from the intestine increases pancreatic secretion and that infusion of bile or taurocholate (TC) prevents pancreatic hypersecretion.‘-4 However, Kon-

B

turek and Thor reported5 that in dogs intraduodenal infusion of bile or TC-stimulated basal pancreatic protein secretion, and Davies et al. reported” that the diversion of bile from the duodenum decreased the meal-stimulated plasma cholecystokinin (CCK) concentration and pancreatic secretion. On the other hand, Gomez et al. reported7 that triglyceride-induced release of CCK was significantly depressed by the presence of excess bile in the lumen. In humans, intraduodenal infusion of bovine bile or bile salt mixture has been found to stimulate pancreatic secretion via CCK8,’ or secretin release.‘“~” Recently, Bondesen et al.” examined the effects of individual bile salts and found that glycocholate (GC), cholate (CA), taurocholate (TC), and chenodeoxycholate (CDCA) increased the plasma secretin concentration, CDCA having the greatest stimulatory effect. They proposed that the effects of exogenous bile salts may differ from those of endogenous bile salts. A recent study by Koop13 suggested that CCK is a major candidate for this regulatory role under physiological conditions in humans. We found previously that ursodeoxycholate (UDCA) stimulated pancreatic exocrine secretion in conscious rats and in anesthetized rabbits but that no other bile salts tested had any stimulatory effect in rabbits.‘4,‘5 In the present study, we examined the effects of infusions of individual bile salts [CDCA, deoxycholate (DCA), UDCA, and CA and their conjugates] into the duodenum of conscious rats on pancreatic secretions. Materials and Methods Materials

and Chemicals

Synthetic CCK-octapeptide sulfate (CCK-8) was purchased from the Peptide Institute, Osaka, Japan. BOvine serum albumin, soybean trypsin inhibitor (type I-S), and chromatographically purified collagenase (type IV) were from Sigma Chemical Co., St. Louis, MO. Minimal Eagle’s medium amino acid supplement was obtained 0 1992 by the American Gastroenterological 0018-5085/92/$3.00

Association

EFFECTS OF BILE SALTS ON RAT PANCREAS 599

February1992

Separate cannulae were introduced for draining bile and pancreatic juice (PJ). In addition, two duodenal cannulae and an extrajugular vein cannula were inserted. Details of the operative procedures and the maintenance of animals after surgery have been reported.‘8~‘g After the operation, the rats were placed in modified Bollman-type restraint cages and had free access to food and water in a room at 24°C with filtered air and light from 5 AM through 5 PM. Bile and PJ was continuously returned to the intestine via a duodenal cannula. Experiments were conducted on day 4 after the operation after withholding food for 6 hours.lg

from GIBCO Laboratories, Grand Island, NY. CR-1409 (sodium salt), a specific CCK antagonist, was a generous gift from Professors L. A. Rovati and I. Setniker, Rotta Research Laboratorium, Milan, Italy. Secretin antibody was a generous gift from Eisai Co., Ibaragi, Japan.16a17Rabbit antiserum was raised against synthetic porcine secretin and had a titer of l:lOO,OOO at a final dilution binding 50% of “?-labeled secretin in solution at a concentration of 10 pmol/L. 16,X7This antibody did not crossreact with glucagon or vasoactive intestinal peptide.16 CDCA, taurochenodeoxycholate (TCDC), glycochenodeoxycholate (GCDC), DCA, taurodeoxycholate (TDCA), glycodeoxycholate (GDCA), glycoursodeoxycholate (GU), CA and TC were obtained from Calbiochem, La Jolla, CA. Ursodeoxycholate and tauroursodeoxycholate (TU) were generous gifts from Tokyo Tanabe Co., Tokyo, Japan. Bile salts were dissolved in distilled water and the solutions were adjusted to pH 7.4-7.8, except that of UDCA, which was adjusted to pH 8.6. The cannulae used in this study were Silastic Medical Grade Tubing (Dow-Corning, Midland, MI.; 0.026-inch inner diameter, 0.037 inch outer diameter).

Experimental

Bile and PJ were collected separately for SO-minute periods and the volume of PJ was measured with a Hamilton syringe. Samples of 20 uL of PJ were used for determining protein and bicarbonate, and the rest was mixed with the bile and infused into the duodenum with a syringe pump (Compact Infusion Pump; Harvard Apparatus, South Natick, MA) over the next 30 minutes. The first SO-minute sample was not used for assay, and during the first 30 minutes of the experiment the previously pooled bile and PJ were infused. To examine the stimulatory effects of various bile salts on pancreatic secretion (during return of bile and PJ), each bile salt solution (60 mmol/L, 1 mL/h) was infused intraduodenally for z hours after collection of basal samples for 90 minutes. Some animals were killed 30 minutes after the beginning of CDCA infusion to measure the plasma CCK concentration

Animal Preparations Male Wistar rats (314-330 g) for experiments and female Sprague-Dawley rats (170-200 g) for CCK bioassay were obtained from Shizuoka Jikken Dobutsu (Shizuoka, Japan). Rats were given commercial rat chow (CRF-1, Oriental, Tokyo, Japan) before surgery and during recovery.

ProreIn mg/30mln Flow mlI30mln

HCO; pEq/30min

Design

CDCA 7

fi Flow

GCDC J

TCDC CCK

Figure 1. Effects of CDCA and its conjugates (GCDC and TCDC) on pancreatic secretion and plasma CCK concentration. Chenodeoxycholate and GCDC, hut not TCDC, significantly increased pancreatic fluid, bicarbonate, and protein outputs. n = 8 for CDCA, n = 5 for GCDC, n = 8 for TCDC. *Significantly different from the value before infusion of bile salts.

J

4

VI)

r

I

1

0

Time,

h

2

I

0

I

1 Time.

h

I

1

* * *:-hL * P I

I

0

1

1 Time,

h

I

2

600

MIYASAKA ET AL.

HCO, concentratior vEcv’ml

GASTROENTEROLOGY Vol. 102. No. 2

4

1

50

25

0 Protein concentration mdml

* * * * * * * r r UDCA infusion

CDCA infusion

I : I

1

I

DCA infusion

I

*

I

1

I

I

01

tl

i

2

Time, h

b

i

h

Li

i Time, h

Time, h

Infusion of CDCA resulted in marked increases in pancreatic secretions. The mechanism of its stimulatory action was examined by administration of either a specific CCK antagonist, CR-1409, or rabbit antiserum for porcine secretin. Intravenous infusion of CR-1409 at 0.3 mg. kg-’ - h, which has been reported to be sufficient to inhibit endogenously released CCK,” was started 30 minutes before CDCA infusion. In some cases, 10 pL of secretin antibody’6”7 was injected 30 minutes before CDCA infusion.

I

I

L***

*

%

25-

3

1

Figure 2. Changes

in hicarbonate and protein concentrations in PJ during infusions of CDCA, UDCA, and DCA. The bicarbonate concentration increased, but the protein concentration decreased. n = 7 for UDCA, n = 5 for DCA. *Significantly different from the value before infusion of bile salts.

To examine the inhibitory effect of bile salts, we excluded bile and PJ from the intestinal lumen and stimulated pancreatic secretion. The bile and PJ were replaced by 0.05 mol/L NaHCO, (1 mL/h) for 3 hours. Then, after collection of PJ for 90 minutes, each bile salt solution was infused into the duodenum at a rate of 1 mL/h for 2 hours. Before and at the end of bile salt infusion, blood was drawn and the proximal quarter of the small intestine was removed, washed with 10 mL of distilled water, and immediately lyophilized. Assays

Table 1. Increments Outputs

of Fluid, Bicarbonate,

and Protein CDCA +

Fluid (mL/.Z h) Bicarbonate kE@ h) Protein (mg/2 h)

Secretin

CDCA (n = 8)

CDCA + CR-1409 (n = 6)

antibody [n = 4)

0.411 + 0.038

0.321 f 0.054

0.164 f 0.057”

23.09 + 1.98

19.10 + 1.63

12.57 f 2.76b

20.77 f 2.36

6.88 f 3.53b

9.70 + 4.41

NOTE. Increments of pancreatic secretion were calculated as values in Z-hour increments during CDCA infusion minus those in 30 minute increments before the infusion X 4. F(2,15) = 7.42 (P < 0.01) for fluid output, 5.67 (P < 0.05) for bicarbonate output, and 6.06 (P < 0.05) for protein output. “Significantly lower than the other two values for fluid output. bSignificantly lower than the value with CDCA only by the multiple comparison test.

Protein in PJ was determined by measuring the optical density at 280 nmzl of samples diluted 200-fold with 0.04 mol/L Tris buffer, pH 7.8. The bicarbonate concentration in 10 pL samples immediately after their collection was measured with a Natelson microgasometer. Plasma CCK concentrations were measured by bioassay using dispersed acini as described by Liddle et al.” Isolated rat pancreatic acini were prepared by collagenase digestion of pancreases from fasted, ovariectomized female Sprague-Dawley rats as described previously.” A 2-mL sample of plasma was obtained from each rat, and CCK was extracted from the plasma by adsorption onto Sep-Pak cartridges. The cartridges were washed with 20 mL of water, and CCK was then eluted with 1 mL of acetonitrile/ water (l:l, vol/vol), collected in a polyethylene scintillation vial, and dried with a flow of nitrogen at 45’Cz3 Aliquots of 1 mL of acini suspension were added to vials containing the plasma extracts or various concentrations of CCK-8 and incubated for 30 minutes at 37°C. Amylase re-

February

EFFECTS OF BILE SALTS ON RAT PANCREAS

1992

UDCA I-low Figure 3. Comparison of stimulatory effects of unconjugated and conjugated bile salts on pancreatic exocrine secretion. Values are shown as percentages of those with the unconjugated bile salts. n = 4-8 for each experiment.

41.6(P < 0.01)for flow, 68.3(P i 0.01) for bicarbonfor ate, and 17.1(P < 0.01)

HCO,

601

CDCA votein

Flow

HC6;

Protein

Flow

HCO;

Protein

% 100

F =

proteinonUDCAadministration; F = 42.2 (P < 0.01)for flow, 67.9 (P < 0.01)for bicarbonate, and 21.3 (P < 0.01) for protein on CDCA administration: and F = 16.0 (P < 0.01) for flow, 7.17 (P < 0.01) for bicarbonate, and 2.04 (P > 0.05) for protein on

DCA administration. *Significantly higher than the values with the respective conjugates.

50

0 - 50

u

free

bile

salt

I

taurine

conjugates

glycine

conjugates

- 100

lease into the medium and total acinar amylase contents were measured. Values were calculated as CCK-8 equivalents by comparison with a standard curve for CCK-8. Trypsin activity in the proximal intestine was measured spectrophotometrically, with tosyl-I-arginine methyl ester as a substrate.24 Lyophilized intestinal contents were homogenized with 1 mL of distilled water at O”C, and samples of 50-200 PL were used for assay.‘* Statistics Values are expressed as the means + SE. Significances of differences were determined by one-way analysis of variance or Student’s t test. Post hoc analysis was performed using Newman Keul’s multiple-comparison test. A P value of