- Email: [email protected]
The inhibition of intestinal dipeptidylaminopeptidase-IV promotes the absorption of enterostatin and des-arginine-enterostatin across rat jejunum in vitro
Life Sciences, Vol. 59, No.6 25126, pp. 2147-2155.1996 Copydght Q 1996 Elsevier Science Inc. Printed in the USA. All rights rcscd 0024-32051% slS.00 t .oo
ELSEVIER
PII SOO24-3205(96)00571-l
THE INHIBITION OF INTESTINAL ABSORPTION OF ENTEROSTATIN
Mohammed Bouras,
DIPEPTIDYLAMINOPEPTIDASE-IV PROMOTES THE AND DES-ARGININE-ENTEROSTATIN ACROSS RAT JEJUNUM IN L’TRO
Jean-Frangois
Huneau’
and Daniel Tome
INRA, Unite de Nutrition Humaine et de Physiologie Intestinale, Institut National Agronomique Paris-Grignon, 16,rue Claude Bernard, 75005 Paris, France (Received in final form October 11,1996)
summary Transport of enterostatin (VPDPR) across rat jejunum was investigated usii GrassSweetana di&sion chambers. The rate of absorption of enterostatii and its metabolites were studied in absence and in presence of diisopropyltluorophosphate (DFP), a serine protease inhibitor. An extensive hydrolysis of enterostatin was observed during incubation with brush border membranes and when exposed to the mucosal side of the jejunal epithelium. No accumulation of enterostatin occured in serosal tissue. Addition of DFP delayed enterostatin disappearance and abolished desarg-enterostatii degradation. Under these conditions, a low amount of enterostatin was able to cross the epithelium intact. Moreover, a substantial amount of des-argenterostatin is absorbed across the jejunal epithelium, probably through passive diffision. Thus, a decreased metabolic activity increased the absorption of a tetrapeptide (VPDP). Dipeptidylaminopeptidase IV, remains a limiting step in transfer of intact enterostatin and its metabolite des-arg-enterostatin across intestinal wall. Key Words: enterostatin, diisopropylfluorophoaphate, dipeptidylpeptidase IV, jejunum, absorption The intestinal mucosa is a highly selective barrier for the absorption of bioactive peptides. The oral bioavailabii of peptides is hindered by both luminal proteolysis and a low permeability of intestinal epithelia related to their poor lipophilic character and to an enzymatic catabolism by brush border membrane peptidases. It is now well established that small di- and tri-peptides are transported through an energy-dependent carrier-mediated process whereas larger peptides are poorly absorbed, probably through an energy-independent trancellular or paracellular diffusion (1). Enterostatin is a pentapeptide released during tryptic cleavage of pancreatic procolipase in the intestinal lumen. In most mammals, its structure is Val-Pro-Asp-Pro-Arg. Previous works have shown that this peptide inhibit food intake in fasted rats, this effect being observed a&r either a central or a peripheral (i.e. intraperitoneal or intravenous) administration (2,3,4). Recently, an inhibition of high-fat food intake has also been observed in response to an intraduodenally int%sion of enterostatin in Osborne-Mendel rats (5). Intraduodenal infusion of enterostatin was also shown to reduce pancreatic exocrine secretion in pig (6). Taken together, these results suggest that the intestinal absorption of enterostatin is probably of importance regarding the expression of the different biological activities of this peptide. It has been ’ Corresponding author. Tel. (33) 144 08 18 19, Fax (33) 1 44 08 18 25
2148
Enterostatin Absorption
Vol. 59, No.s 25/X, 1996
shown in a previous experiment that the absorption of intact enterostatin across rabbit ileum was very poor and was probably hampered by a rapid enzymatic degradation (7). As shown in rat, the brushborder membrane proline-specific peptidases dipeptidylaminopeptidase IV (DDP IV) and carboxypeptidase P (CP) are both involved in the cleavage of enterostatin (8). The aim of our study was to investigate whether an inhibition of intestinal brush border membrane DPP IV would promote the absorption of intact enterostatin across rat jejunum. For that purpose, the role of the intestinal brush-border membranes as a barrier to absorption of a pentapeptide was assessed with both brush-border membrane vesicles and intestinal tissue mounted in vitro in Sweetana-Grass diI%.rsion chambers. We attempted to determine whether intact enterostatin could cross mucosal barrier in the presence of diisopropylfluorophosphate (DFP), an inhibitor of serine proteases.
Material and Methods
Reagents : Enterostatin (VPDPR) and diisopropylfluorophosphate (DFP) were purchased from Sigma Chemicals (La Verpilliere, France). [3H-Pro24] enterostatin, obtained by tritiation of Val-DehydroProAsp-DehydroPro-Arg, was a gifl from Pr Charlotte Erlanson-Albertsson and D-[14C]mannitol (57 mCimmo1’) was obtained from Amersham &es Ulis, France). Des-arginine-enterostatm (VPDP) was custom synthesized by Neosystem (Strasbourg, France). Valylproline (VP) was purchased from Bachem (Voisins-Le-Bretonneux, France). Other chemicals of analytical reagent grade were obtained from various commercial sources. Brush-border membrane vesicles experiments : Intestinal brush border membrane vesicles (BBMV) were prepared from everted rat intestine by calcium precipitation and differential centriI%gation as previously described (9). The membrane vesicles were suspended at a final concentration of 2-3 mg proteinml~’ with 300 mM D-mannitol, 10 mM N-2-hydroxyethylpiperazine-N-2-ethanesulphonic acid (HEPES)/Tris, pH 7.5. In the degradation study, 50 ul of diluted BBMV (30 pg proteins) were mixed with enterostatin (80 nmoles) at 37°C in a total volume of 150 pl. The reaction medium was buffered at pH 7.5 with 10 mM TritiCl, 50 mM NaCl. Inhibitors were preincubated with membranes for 30 min at room temperature. Sixty microliters of 85% trifluoroacetic acid (TFA) were added to terminate the reaction. After centritugation at 20,000 x g for 20 min, the supematant was analyzed by HPLC. Transport experiments : Rats were killed by cervical dislocation. An abdominal incision was performed and segments of jejunum were removed. The segments were immediately immersed in Krebs buffer solution consisting of (in n&I) 117 NaCl, 4.7 KCI, 2.5 CaCl2, 1.2 MgSO4, 24.8 NaHCO3, 1.2 KI-I&IPOd and 11.1 glucose (pH 7.4) gassed with carbogen (oz/COz 95:5). The tissue was cut along the mesenterial border and 4-cm-long segments mounted between two halves of Sweetana-Grass ditIi.rsion chambers (Precision Instruments design, Los Altos (CA), USA). Care was taken to avoid the Peyer’s patches. Each side of the tissue (exposed area 1.43 cm2) was bathed with 4 ml of Krebs solution. Oxygenation of the tissue was ensured by a gas litI of carbogen and the temperature was maintained at 37°C throughout the study. The spontaneous transepithelial electrical potential difference was short-circuited by an automatic voltage clamp (World Precision Instruments, Sarasota, FL, USA) throughout experiments and the short-circuit current corrected for fluid resistance. After stabiig the electrical parameters for 15 min, DFP was added to the mucosal and serosal reservoirs. Enterostatin (0.1 mM) was added to the mucosal compartment with simultaneous addition of 5 pCi of [3H-Proz4] enterostatin. In some experiments, the integrity of the intestinal preparation was checked by a
Vol. 59, Nos 25/26,19!36
Enterostatin Absorption
2149
concomitant addition of D-[‘4C]mannitol (final concentration 0.5 pCi.ml-1) to the mucosal compartment. Samples of 100 pl and 1 ml were withdrawn at diierent times from mucosal and serosal compartments, respectively, and replaced by an equivalent volume of fresh Krebs solution. Ten percent TFA was added to the remaking sample. A&r a 20,000 x g cemrifbgation, 100 pl of the supernantant were analyzed using reverse phase HPLC. Determination of 3H or 14C activities of each sample was done by liquid scintillation using a Beckman LS 9000 counter. HPLC analysis: HPLC analysis was performed on a Waters gradient system using a reverse-phase column (ClS-Bondapak, 5 x 250 mm) (Shandon, les Ulis, France). Peptides were separated at a constant flow rate (1 ml.&‘) using a linear gradient (5 to 40 %) of acetonitrile in 0.1% TFA. The different products were detected by measurement of 214 nm absorbance and on-line scintillation counting using a Flo-one l3 A500 (Packard, St Quentin Yvelines, France). Products were identified by co-elution with standard peptides and identification was confirmed by amino acid analysis. Calcuhtions and statistical analysis : In the study of enterostatin degradation by brush-border membrane peptidases, the amounts of enterostatin (VPDPR), des-arg-enterostatin (VPDP), valylproline (VP) recovered were quantiied by comparing the integrated peak areas of known amounts of standards. To quantii the amount of VPDP recovered in the transport experiments, the assumption was made that the specific activity of VPDP was similar to that of the VPDPR present in the mucosal compartment, i.e. 12.5 uCi.mmol-‘. Results are expressed as the percent of the initial amount of VPDPR added to the mucosal compartment absorbed per cm2 of jejunal mucosa. The data were analyzed with unpaired Student’s t test and a ~~0.05 was regarded as significant. Results Degrahtion of enterostatin in the presence of rat BBMV : In a first experiment 0.53 mmo1.1~’ enterostatin (VPDPR) was incubated with 2-3 g protein.1~’ purified BBMV in the absence or in the presence of either the serine protease inhibitor DFP (I mmol.K’), or the metalloenzyme inhibitor EDTA (1 mmol.1~‘)(Fig. I). In control conditions, the rate of degradation of VPDPR was 103 nmol.min~‘~mg“ protein and only 20 % of the initial amount of the peptide was recovered as intact enterostatin after 20 minutes of incubation. A transient appearance of VPDP was observed whereas the amount of VP recovered continuously increased. In the presence of the serine protease inhibitor DFP the rate of VPDPR degradation was reduced by SO%, a complete inhibition of des-arginine-enterostatin hydrolysis was achieved and the production of VP was completely inhibited. Addition of EDTA resulted in a 60% decrease of VPDPR and prevented the production of VPDP, whereas the formation of VP was unaffected. For DFP-treated membranes, the cumulative amount of the diierent valine-containing peptides recovered (VPDPR VPDP and VP) at the diierent times was never significantly different from the initial amount of enterostatin added to the incubation medium. However, in the absence of peptidase inhibitor or in the presence EDTA, a gradual loss of valine-containing material was observed, that reached 36 % and 59% of the initial amount of enterostatin at -0 minutes, for control and EDTA-treated membranes, respectively. A4ucosaI &gra&tion of enterostatin in rat jejunum mounted in Sweetana-Gr~ diJfusion chambers : In another set of experiments rat jejunum was mounted in Sweetana-Grass diision chambers in the absence or in the presence of the serine protease inhibitor DFP (1 mmol.1~‘). The viabiity of the tissue was continuously monitored through the recording of electrophysiological parameters. The epithelial conductance remained stable throughout the experiment (data not shown).
2150
Vol. 59, No.s 25/26, 1996
Enterostatin Absorption
0.6 ,
0
, 0.6,
10
20
30
Incubation time (min)
40
-,
0
10
20
30
Incubation time (min)
40
0.6,
0
10
Incubation
f0
30
40
time (min)
FIG. 1 Enterostatin hydrolysis (0) and VPDP (A) and VP (m) production by intestinal brush border membranes in control conditions (A), or in presence of 1mM DFP (B), or 1mM EDTA (C). Enterostatin (80 nmoles) was incubated with brush-border membranes (30 pg proteins) in 10 mM HEPES/Tris but&r pH 7.5 at 37°C and the reaction was stopped by adding TFA. Separation and quantitation of peptides were assesssed by HPLC. Data are the mean + S.E.M. of 3 experiments Following addition of 0.1 mmoles.1~’enterostatin trace-labeled with 5 uCi of [3H-Pro24] enterostatin to the mucosal compartment aliquots were collected at regular intervals in the mucosal reservoir. A rapid and almost linear disappearance of intact enterostatin was observed in the absence of DFP. VPDPR was rapidly hydrolyzed yielding a tetrapeptide (VPDP), a tripeptide (DPR) and two dipeptides (DP and VP) (Fig. 2). About 50% of the initial amount of radiolabeled enterostatin was hydrolyzed afler 30 minutes of incubation and the radioactivity was only recovered as fke proline and di- and tripeptides after 40 minutes. The presence of 1 mmol.1~’ DFP to the in vitro preparation reduced the rate of hydrolysis of enterostatin. Mucosal to scrod tmmport of enterostatin in rat jejmum in vitro : After the addition of 0.1 mmol.1” enterostatin trace-labeled with 5 uCi of [3H-Proz4] enterostatin to the mucosal compartment of the Sweetana-Grass difiksion chambers, the appearance of enterostatin and its degradation products were monitored in the serosal reservoir, Intact enterostatim was never detected in the serosal reservoir of untreated rat jejunum, even at time corresponding to a minimal degradation of the peptide in the mucosal compartment (i.e. 10 minutes) and the radiolabeled material was recovered as free proline and dipeptides (Fig. 3). In contrast, 10 minutes after the mucosal addition of enterostatin, a small amount of intact radiolabeled VPDPR was recovered in the serosal compartment of DFP-treated jejunum. Enterostatin was not recovered in the subsequent aliquots, suggesting that a degradation of this peptide occured in the serosal compartment of DFP-treated jejunum. At any time, VPDP was the main radiolabeled peptide recovered in the serosal compartment of DFP-treated jejunum and there was no fkther degradation of the VPDP produced during enterostatin hydrolysis. The amount of radiolabeled material recovered in the serosal reservoir linearly increased as a function of time and was significantly higher with control jejunum, compared to DFP-treated jejunum. (Fig. 4). In the later case, the serosal accumulation of VPDP increased gradually and reached about 0.30 % of the initial amount of radiolabeled peptide after 40 minutes. As shown on Fig 4(C), DFP did not affect the transepithelial absorption of D-[i4C]mannitol.
2151
Enterostatin Absorption
Vol. 59, Nos 25/26,1996
B: DF’P treated tissue
A: Untreated tissue
ti
d8
tzdynl
L 1
0
I
I
I
I
I
5
10
15
20
25
Elution time (min)
I
1
30 0
I
I
I
I
I
I
5
10
15
20
25
30
Elution time (min)
FIG. 2 HPLC analysis of the radiolabeled material recovered in the mucosal compartment of control (A) and DFP-treated (B) jejunum mounted in Sweetana-Grass diliksion chamber at different times following the addition of 0.1 mM enterostatii trace-labeled with 5 pCi [3HPro’“] enterostatin. Results are those of one typical experiment.
The pentapeptide enterostatin is released in the intestinal lumen during the N-terminal tryptic cleavage of procolipase. This peptide has been shown to reduce the intake of a high-fat food and to decrease the exocrine pancreatic output r&r a duodenal infusion (4, 6). However, a previous
2152
Enterostatin Absorption
Vol. 59, No.s 25/26, 1996
experiment suggests that the absorption of enterostatin across the rabbit ileum was only minimal (7). The present study reported that the inhibition of the post-proline dipeptidylaminopeptidase IV promotes the absorption of the intact pe@ide and may thus increase the biological response to this peptide a&x luminal ir&sion. Bioactive peptides present in the intestinal lumen are attacked by a barrage of metabolizing enzymes, including pancreatic enzymes - namely trypsin, chymotrypsin, elastase and carboxypeptidases - and brush-border membrane peptidases. A: Untreated jejunum
B: DFP treated jejunum
I
’
’
’
’
’
0
5
10
15
20
25
Elution time (min)
c 30
I
’
’
’
’
’
0
5
10
15
20
25
c 30
Elution time (min)
FIG. 3 HPLC analysis of the serosal compartment of the Sweetana-Grass ditision chambers atIer the addition of 0.1mM of enterostatin radiolabeled with S uCi of [3H-Pro”]enterostatin to the mucosal compartment in absence (A) and presence (B) of 1mM DFP inhibitor. Results are those of one typical experiment. The latters have been shown to be responsible for the metabolism of prohne-containing peptides such as enterostatm and casomorphins which are resistant to the action of pancreatic endo- and exopeptidases (8,lO). In a previous study, we have shown that enterostatin hydrolysis involved two P and namely carboxypeptidase brush-border membrane post-proline peptidases, dipeptidylaminopeptidase IV (8).
Enterostatin Absorption
Vol. 59, No.s 2.5/26, 1996
,o
20
30
40
50
0
10
20
30
40
so
2153
0
IV
zu
30
4u
50
Time (min)
FIG. 4 Effect of DFP on the in vitro intestinal absorption of enterostatin. DFP 0 (A) or 0.1 rnmol. 1’ (0) was added to the mucosal and serosal compartment of rat jejunum mounted in Sweetana-Grass diffusion chambers. At t=O, enterostatin (0.4 pmol, 5 l&i) was added to the mucosal compartment and the accumulation of total radioactivity (A) and radiolabeled VPDP (B) was measured sequentially in the serosal compartment. The effect of DFP on the paracellular absorption of solute was assessed by measuring the transport of D[‘4C]mannito1 in the same conditions (C). Results are means f SD of n=6 (A and B) or n=4 (C) experiments and are expressed as percent of the mucosal amount of radiolabeled peptide (A and B) or mannitol (C) absorbed per cm’ of mucosa. * significantly diierent from control, pcO.05. In that study, diprotm A, a DPP IV specific inhibitor, was responsible for a 40% reduction in enterostatin degradation rate. The results reported here confirm the involvement of this two enzymes. However, the replacement of diprotin A by the DPP IV inhibitor DFP results in a reevaluation of the role of DPP IV in the degradation of enterostatin. Addition of 1 mmol.1~’ DFP reduces the rate of degradation of enterostatin by 80 % with BBh4V. This discrepancy may arise from a partial degradation of diprotin A (Ile-Pro-Ile) by DPP IV (II). The inhibition by DFP of a post-proline cleaving enzyme (PPCE) differing from DPP IV may represent an alternative explanation. DFP is a serine protease inhibitor and has been shown to inhibit DPP IV as well as PPCE and some post-proline carboxypeptidases (12). The latter enzymes are mainly present in brain and are involved in the inactivation of neuropeptides (12) and the presence of PPCE was also reported in the intestine and kidney membranes (13). Therefore, a limited contribution of PPCE to the intestinal degradation of enterostatin may be responsible for the difference in the inhibition potencies of DFP and diprotin A The loss of valine-contakg peptides observed upon incubation of enterostatin with control and EDTA-treated brush-border membranes also suggest that the dipeptide VP, but not VPDP, may be further processed to Valine and Proline, which can not be detected on the chromatograms. The strict dipeptidase prolidase is likely to be responsible for this degradation. This enzyme is not inhibited by EDTA and may be recovered with membrane fraction, although it is mainly cytosolic (14,15). Our results also indicate that DFP does not completely prevent the degradation of enterostatin in the mucosal and serosal compartment of the chamber. This observation is consistent with the results obtained with puritied brush-border membranes indicating the involvement of a membrane-bound carboxypeptidase P in the degradation of this peptide. It seems conceivable that such an activity may also be responsible for the conversion of enterostatin to its des-arginine metabolite in the serosal
2154
Enterostatin Absorption
Vol. 59, No.s 25/24, 1996
compartment of the Sweetana-Grass ditision chambers. With brush-border membrane preparation, an almost complete inhibition of the production of VPDP is achieved with 1 mM EDTA, thus indicating that the carboxypeptidase P is sensitive to chelating agents. Although many reports now suggest that small but biologically significant amounts of intact oligopeptides are absorbed across the epithelium of the small intestine (16-20) the mechanisms underlying this absorption remain unclear. The size of the peptides appears to be a critical determinant for the interaction with the I-I+-peptide cotransporter (21) and their is no evidence at present for the existence of specific transport mechanisms for peptides above 4 residus. It is usually assumed that these molecules are absorbed by passive transcellular or paracellular dif%.rsion and that the brush-border peptidases are the main limiting step of this process. The absorption of peptidase-resistant analogues of vasopressin across everted segment of rat jejunum has be shown to be more efficient than that of the natural peptide (19). In the same way, a promotion of the absorption of intact bioactive peptides has been reported for vasopressin and morphiceptin analogues atler treatment of intestinal mucosal sheets mounted in ussing chambers with peptidase inhibitors (19, 20). A reevaluation of the role of dipeptidylaminopeptidase IV in the degradation of enterostatin by intestinal mucosal peptidases suggests that an increase in the intestinal absorption of this peptide can be achieved through the inhibition of this enzyme. Our results indicate that inhibition of mucosal serine proteases by DFP promotes the absorption of enterostatin across rat jejunum in vitro. Addition of DPP resulted in a transient recovery of intact enterostatin and an accumulation of des-arginineenterostatin in the serosal compartment of the chambers. A promotting effect of DF’P through a non specific increase in the paracellular absorption of solute can be ruled out since the addition of DPP to the incubation buffer neither increased the absorption of the inert paracellular marker D-[‘4C]mannitol, nor affected the electrical conductance of the rat jejunum. Rather, it seems clear that the absorption of enterostatin and des-arginine-enterostatin across DFP-treated rat jejunum is a consequence of the inhibition of serine peptidases. No attempt was done to investigate the consequence of carboxypeptidase P inhibition on the in vitro transport of enterostatin across rat jejunum since BDTA has been shown to a&ct tight junctions through calcium chelation, and thus to increase the paracellular absorption of solute across the intestinal epithelium (22). Therefore, we would not have been able to discriminate between effects resulting from enzymatic inhibition and paracellular shunting. In conclusion, our results provide evidence that the absorption of intact enterostatin is hampered by brush-border membrane peptidases and that the inhibition of DPP IV stabilize peptide, hence improving its intestinal absorption, Some reports also suggest that absorption of enterostatin may also occurs in viva. In human a biphasic increase in the concentration of enterostatin in serum has been reported to occur in response to a meal (23) and plasma enterostatin level was shown to increase after a high fat, a high fat/sucrose or a low-fat diet in rats (J. Mei, unpublished results). The production of enterostatin in the intestinal lumen as a consequence of the tryptic activation of pancreatic secretion occurs conwmitantly to the release of digestion products, including glucose, peptides, amino acids and free f&y acids. Whether some of these products may exert an enhancing potency effect on enterostatin absorption through a modulation of brush-border membrane peptidase remains to be determined.
References 1. D.M. MATTHEWS, Protein Absorption: Developement and Present State of the Subject, Wiley-Liss, New York. (1991). 2. C. ERLANSON-ALBERTSSON and A. LARSSON, Regul. Pept. 22 325-331 (1988). 3. N.S. SHARGILL, S. TSUJII, G.A. BRAY, C. ERLANSON-ALBERTSSON, Brain Res. 544 137-140 (1991).
Vol. 59, Nos 25/X, 1996
4. J. MEI and C. ERLANSON-ALBERTSSON,
Enterostatin Absorption
215.5
Regul. Pept. 23 209-218 (1992). 5. J. ME1 and C. ERLANSON-ALBERTSSON, Obesity Research 41161-165 (1996). 6. C. ERLANSON-ALBERTSSON, B. WESTRGM, S. PIERZYNOWSKI, S. KARLSSON and B. AI-&N, Pancreas 6 619-624 (1991). 7. J.F. HUNEAU, C. ERLANSON-ALBERTSSON, C. BEAUVALLET and D. TOME, Regul. Pept. 54 495-503 (1994). 8. M. BOURAS, J.F. HUNEAU, C. LUENGO, C. ERLANSON-ALBERTSSON and D. TOME, Peptides 16 399-405 (1995). 9. E. BROT-LAROCHE and F. ALVARADO, Biochim. Biophys. Actam 175-181 (1981). 10. Y. MIYAMOTO, V. GANAPATHY, A. BARLOS, K. NEUBERT, A. BARTH and F. H. LEIBACH, Am. J. Physiol. 252 F670-F677 1987. 11. J. RAHFELD, M. SCHIERHORN, B. HARTRODT, K. NEUBERT and J. HEINS, Biochim. Biophys. Acta 1076 314-316 (1991). 12. S. KALWANT and A.G. PORTER Biochem. J. 276 237-244 (1991). 13. T. YOSHIMOTO, K. OGITA, R. WALTER, M. KOIDA and D. TSURU. Biochim. Biophys. Acta 569 184- 192 (1979). 14. R. MENTLEIN. FEBS lett. 234 251-256 (1988). 15. R. WALTER, W.H. SIMMONS and T. YOSHIMOTO, Mol. Cell. Biochem., 30, 11 l-125 (1980). 16. K. ATISOOK and J.L. MADARA, Gastroenterology 100 719-724. (1991). 17. D. TOME, A.M. DUMONTIER, M. HAUTEFEUILLE and J.F. DESJEUX, Am. J. Physiol. 253 G737-G744 (1987). 18. S. LUNDIN, H.G. FOLKESSON, S.G. PIERZYNOVSKI and H. BENGTSSON, Peptides fi 809-814 (1994). 19. A. UNGELL, A. ANDREASSON, K. LUNDIN and L. UTTER J. Pharm. Sci. 81640-645 (1992). 20. S. MAHE, D. TOME, A.M. DUMONTIER and J.F. DESJEUX. Peptides B 45-52 (1989). 21. Y.J. FEI, Y. KANAI., S. NUSSBEREGER, V. GANAPATHY, F.H. LEIBACH, M.F. ROMERO, S.K.SINGH, W.F. BORON and M.A. HEDIGER. Nature 368 563-566 (1994). 22. B.J. AUNGST, J.A. BLAKE and M.A. HUSSAIN, J. Pharmacol. Exp. Ther. 259 139-145 (1991). 23. R.C. BOWYER, W.M. ROWSTON, A.M.T. JEHANLI and J.H. HERMON-TAYLOR, Gut 34 1520-1525 (1993).
ELSEVIER
PII SOO24-3205(96)00571-l
THE INHIBITION OF INTESTINAL ABSORPTION OF ENTEROSTATIN
Mohammed Bouras,
DIPEPTIDYLAMINOPEPTIDASE-IV PROMOTES THE AND DES-ARGININE-ENTEROSTATIN ACROSS RAT JEJUNUM IN L’TRO
Jean-Frangois
Huneau’
and Daniel Tome
INRA, Unite de Nutrition Humaine et de Physiologie Intestinale, Institut National Agronomique Paris-Grignon, 16,rue Claude Bernard, 75005 Paris, France (Received in final form October 11,1996)
summary Transport of enterostatin (VPDPR) across rat jejunum was investigated usii GrassSweetana di&sion chambers. The rate of absorption of enterostatii and its metabolites were studied in absence and in presence of diisopropyltluorophosphate (DFP), a serine protease inhibitor. An extensive hydrolysis of enterostatin was observed during incubation with brush border membranes and when exposed to the mucosal side of the jejunal epithelium. No accumulation of enterostatin occured in serosal tissue. Addition of DFP delayed enterostatin disappearance and abolished desarg-enterostatii degradation. Under these conditions, a low amount of enterostatin was able to cross the epithelium intact. Moreover, a substantial amount of des-argenterostatin is absorbed across the jejunal epithelium, probably through passive diffision. Thus, a decreased metabolic activity increased the absorption of a tetrapeptide (VPDP). Dipeptidylaminopeptidase IV, remains a limiting step in transfer of intact enterostatin and its metabolite des-arg-enterostatin across intestinal wall. Key Words: enterostatin, diisopropylfluorophoaphate, dipeptidylpeptidase IV, jejunum, absorption The intestinal mucosa is a highly selective barrier for the absorption of bioactive peptides. The oral bioavailabii of peptides is hindered by both luminal proteolysis and a low permeability of intestinal epithelia related to their poor lipophilic character and to an enzymatic catabolism by brush border membrane peptidases. It is now well established that small di- and tri-peptides are transported through an energy-dependent carrier-mediated process whereas larger peptides are poorly absorbed, probably through an energy-independent trancellular or paracellular diffusion (1). Enterostatin is a pentapeptide released during tryptic cleavage of pancreatic procolipase in the intestinal lumen. In most mammals, its structure is Val-Pro-Asp-Pro-Arg. Previous works have shown that this peptide inhibit food intake in fasted rats, this effect being observed a&r either a central or a peripheral (i.e. intraperitoneal or intravenous) administration (2,3,4). Recently, an inhibition of high-fat food intake has also been observed in response to an intraduodenally int%sion of enterostatin in Osborne-Mendel rats (5). Intraduodenal infusion of enterostatin was also shown to reduce pancreatic exocrine secretion in pig (6). Taken together, these results suggest that the intestinal absorption of enterostatin is probably of importance regarding the expression of the different biological activities of this peptide. It has been ’ Corresponding author. Tel. (33) 144 08 18 19, Fax (33) 1 44 08 18 25
2148
Enterostatin Absorption
Vol. 59, No.s 25/X, 1996
shown in a previous experiment that the absorption of intact enterostatin across rabbit ileum was very poor and was probably hampered by a rapid enzymatic degradation (7). As shown in rat, the brushborder membrane proline-specific peptidases dipeptidylaminopeptidase IV (DDP IV) and carboxypeptidase P (CP) are both involved in the cleavage of enterostatin (8). The aim of our study was to investigate whether an inhibition of intestinal brush border membrane DPP IV would promote the absorption of intact enterostatin across rat jejunum. For that purpose, the role of the intestinal brush-border membranes as a barrier to absorption of a pentapeptide was assessed with both brush-border membrane vesicles and intestinal tissue mounted in vitro in Sweetana-Grass diI%.rsion chambers. We attempted to determine whether intact enterostatin could cross mucosal barrier in the presence of diisopropylfluorophosphate (DFP), an inhibitor of serine proteases.
Material and Methods
Reagents : Enterostatin (VPDPR) and diisopropylfluorophosphate (DFP) were purchased from Sigma Chemicals (La Verpilliere, France). [3H-Pro24] enterostatin, obtained by tritiation of Val-DehydroProAsp-DehydroPro-Arg, was a gifl from Pr Charlotte Erlanson-Albertsson and D-[14C]mannitol (57 mCimmo1’) was obtained from Amersham &es Ulis, France). Des-arginine-enterostatm (VPDP) was custom synthesized by Neosystem (Strasbourg, France). Valylproline (VP) was purchased from Bachem (Voisins-Le-Bretonneux, France). Other chemicals of analytical reagent grade were obtained from various commercial sources. Brush-border membrane vesicles experiments : Intestinal brush border membrane vesicles (BBMV) were prepared from everted rat intestine by calcium precipitation and differential centriI%gation as previously described (9). The membrane vesicles were suspended at a final concentration of 2-3 mg proteinml~’ with 300 mM D-mannitol, 10 mM N-2-hydroxyethylpiperazine-N-2-ethanesulphonic acid (HEPES)/Tris, pH 7.5. In the degradation study, 50 ul of diluted BBMV (30 pg proteins) were mixed with enterostatin (80 nmoles) at 37°C in a total volume of 150 pl. The reaction medium was buffered at pH 7.5 with 10 mM TritiCl, 50 mM NaCl. Inhibitors were preincubated with membranes for 30 min at room temperature. Sixty microliters of 85% trifluoroacetic acid (TFA) were added to terminate the reaction. After centritugation at 20,000 x g for 20 min, the supematant was analyzed by HPLC. Transport experiments : Rats were killed by cervical dislocation. An abdominal incision was performed and segments of jejunum were removed. The segments were immediately immersed in Krebs buffer solution consisting of (in n&I) 117 NaCl, 4.7 KCI, 2.5 CaCl2, 1.2 MgSO4, 24.8 NaHCO3, 1.2 KI-I&IPOd and 11.1 glucose (pH 7.4) gassed with carbogen (oz/COz 95:5). The tissue was cut along the mesenterial border and 4-cm-long segments mounted between two halves of Sweetana-Grass ditIi.rsion chambers (Precision Instruments design, Los Altos (CA), USA). Care was taken to avoid the Peyer’s patches. Each side of the tissue (exposed area 1.43 cm2) was bathed with 4 ml of Krebs solution. Oxygenation of the tissue was ensured by a gas litI of carbogen and the temperature was maintained at 37°C throughout the study. The spontaneous transepithelial electrical potential difference was short-circuited by an automatic voltage clamp (World Precision Instruments, Sarasota, FL, USA) throughout experiments and the short-circuit current corrected for fluid resistance. After stabiig the electrical parameters for 15 min, DFP was added to the mucosal and serosal reservoirs. Enterostatin (0.1 mM) was added to the mucosal compartment with simultaneous addition of 5 pCi of [3H-Proz4] enterostatin. In some experiments, the integrity of the intestinal preparation was checked by a
Vol. 59, Nos 25/26,19!36
Enterostatin Absorption
2149
concomitant addition of D-[‘4C]mannitol (final concentration 0.5 pCi.ml-1) to the mucosal compartment. Samples of 100 pl and 1 ml were withdrawn at diierent times from mucosal and serosal compartments, respectively, and replaced by an equivalent volume of fresh Krebs solution. Ten percent TFA was added to the remaking sample. A&r a 20,000 x g cemrifbgation, 100 pl of the supernantant were analyzed using reverse phase HPLC. Determination of 3H or 14C activities of each sample was done by liquid scintillation using a Beckman LS 9000 counter. HPLC analysis: HPLC analysis was performed on a Waters gradient system using a reverse-phase column (ClS-Bondapak, 5 x 250 mm) (Shandon, les Ulis, France). Peptides were separated at a constant flow rate (1 ml.&‘) using a linear gradient (5 to 40 %) of acetonitrile in 0.1% TFA. The different products were detected by measurement of 214 nm absorbance and on-line scintillation counting using a Flo-one l3 A500 (Packard, St Quentin Yvelines, France). Products were identified by co-elution with standard peptides and identification was confirmed by amino acid analysis. Calcuhtions and statistical analysis : In the study of enterostatin degradation by brush-border membrane peptidases, the amounts of enterostatin (VPDPR), des-arg-enterostatin (VPDP), valylproline (VP) recovered were quantiied by comparing the integrated peak areas of known amounts of standards. To quantii the amount of VPDP recovered in the transport experiments, the assumption was made that the specific activity of VPDP was similar to that of the VPDPR present in the mucosal compartment, i.e. 12.5 uCi.mmol-‘. Results are expressed as the percent of the initial amount of VPDPR added to the mucosal compartment absorbed per cm2 of jejunal mucosa. The data were analyzed with unpaired Student’s t test and a ~~0.05 was regarded as significant. Results Degrahtion of enterostatin in the presence of rat BBMV : In a first experiment 0.53 mmo1.1~’ enterostatin (VPDPR) was incubated with 2-3 g protein.1~’ purified BBMV in the absence or in the presence of either the serine protease inhibitor DFP (I mmol.K’), or the metalloenzyme inhibitor EDTA (1 mmol.1~‘)(Fig. I). In control conditions, the rate of degradation of VPDPR was 103 nmol.min~‘~mg“ protein and only 20 % of the initial amount of the peptide was recovered as intact enterostatin after 20 minutes of incubation. A transient appearance of VPDP was observed whereas the amount of VP recovered continuously increased. In the presence of the serine protease inhibitor DFP the rate of VPDPR degradation was reduced by SO%, a complete inhibition of des-arginine-enterostatin hydrolysis was achieved and the production of VP was completely inhibited. Addition of EDTA resulted in a 60% decrease of VPDPR and prevented the production of VPDP, whereas the formation of VP was unaffected. For DFP-treated membranes, the cumulative amount of the diierent valine-containing peptides recovered (VPDPR VPDP and VP) at the diierent times was never significantly different from the initial amount of enterostatin added to the incubation medium. However, in the absence of peptidase inhibitor or in the presence EDTA, a gradual loss of valine-containing material was observed, that reached 36 % and 59% of the initial amount of enterostatin at -0 minutes, for control and EDTA-treated membranes, respectively. A4ucosaI &gra&tion of enterostatin in rat jejunum mounted in Sweetana-Gr~ diJfusion chambers : In another set of experiments rat jejunum was mounted in Sweetana-Grass diision chambers in the absence or in the presence of the serine protease inhibitor DFP (1 mmol.1~‘). The viabiity of the tissue was continuously monitored through the recording of electrophysiological parameters. The epithelial conductance remained stable throughout the experiment (data not shown).
2150
Vol. 59, No.s 25/26, 1996
Enterostatin Absorption
0.6 ,
0
, 0.6,
10
20
30
Incubation time (min)
40
-,
0
10
20
30
Incubation time (min)
40
0.6,
0
10
Incubation
f0
30
40
time (min)
FIG. 1 Enterostatin hydrolysis (0) and VPDP (A) and VP (m) production by intestinal brush border membranes in control conditions (A), or in presence of 1mM DFP (B), or 1mM EDTA (C). Enterostatin (80 nmoles) was incubated with brush-border membranes (30 pg proteins) in 10 mM HEPES/Tris but&r pH 7.5 at 37°C and the reaction was stopped by adding TFA. Separation and quantitation of peptides were assesssed by HPLC. Data are the mean + S.E.M. of 3 experiments Following addition of 0.1 mmoles.1~’enterostatin trace-labeled with 5 uCi of [3H-Pro24] enterostatin to the mucosal compartment aliquots were collected at regular intervals in the mucosal reservoir. A rapid and almost linear disappearance of intact enterostatin was observed in the absence of DFP. VPDPR was rapidly hydrolyzed yielding a tetrapeptide (VPDP), a tripeptide (DPR) and two dipeptides (DP and VP) (Fig. 2). About 50% of the initial amount of radiolabeled enterostatin was hydrolyzed afler 30 minutes of incubation and the radioactivity was only recovered as fke proline and di- and tripeptides after 40 minutes. The presence of 1 mmol.1~’ DFP to the in vitro preparation reduced the rate of hydrolysis of enterostatin. Mucosal to scrod tmmport of enterostatin in rat jejmum in vitro : After the addition of 0.1 mmol.1” enterostatin trace-labeled with 5 uCi of [3H-Proz4] enterostatin to the mucosal compartment of the Sweetana-Grass difiksion chambers, the appearance of enterostatin and its degradation products were monitored in the serosal reservoir, Intact enterostatim was never detected in the serosal reservoir of untreated rat jejunum, even at time corresponding to a minimal degradation of the peptide in the mucosal compartment (i.e. 10 minutes) and the radiolabeled material was recovered as free proline and dipeptides (Fig. 3). In contrast, 10 minutes after the mucosal addition of enterostatin, a small amount of intact radiolabeled VPDPR was recovered in the serosal compartment of DFP-treated jejunum. Enterostatin was not recovered in the subsequent aliquots, suggesting that a degradation of this peptide occured in the serosal compartment of DFP-treated jejunum. At any time, VPDP was the main radiolabeled peptide recovered in the serosal compartment of DFP-treated jejunum and there was no fkther degradation of the VPDP produced during enterostatin hydrolysis. The amount of radiolabeled material recovered in the serosal reservoir linearly increased as a function of time and was significantly higher with control jejunum, compared to DFP-treated jejunum. (Fig. 4). In the later case, the serosal accumulation of VPDP increased gradually and reached about 0.30 % of the initial amount of radiolabeled peptide after 40 minutes. As shown on Fig 4(C), DFP did not affect the transepithelial absorption of D-[i4C]mannitol.
2151
Enterostatin Absorption
Vol. 59, Nos 25/26,1996
B: DF’P treated tissue
A: Untreated tissue
ti
d8
tzdynl
L 1
0
I
I
I
I
I
5
10
15
20
25
Elution time (min)
I
1
30 0
I
I
I
I
I
I
5
10
15
20
25
30
Elution time (min)
FIG. 2 HPLC analysis of the radiolabeled material recovered in the mucosal compartment of control (A) and DFP-treated (B) jejunum mounted in Sweetana-Grass diliksion chamber at different times following the addition of 0.1 mM enterostatii trace-labeled with 5 pCi [3HPro’“] enterostatin. Results are those of one typical experiment.
The pentapeptide enterostatin is released in the intestinal lumen during the N-terminal tryptic cleavage of procolipase. This peptide has been shown to reduce the intake of a high-fat food and to decrease the exocrine pancreatic output r&r a duodenal infusion (4, 6). However, a previous
2152
Enterostatin Absorption
Vol. 59, No.s 25/26, 1996
experiment suggests that the absorption of enterostatin across the rabbit ileum was only minimal (7). The present study reported that the inhibition of the post-proline dipeptidylaminopeptidase IV promotes the absorption of the intact pe@ide and may thus increase the biological response to this peptide a&x luminal ir&sion. Bioactive peptides present in the intestinal lumen are attacked by a barrage of metabolizing enzymes, including pancreatic enzymes - namely trypsin, chymotrypsin, elastase and carboxypeptidases - and brush-border membrane peptidases. A: Untreated jejunum
B: DFP treated jejunum
I
’
’
’
’
’
0
5
10
15
20
25
Elution time (min)
c 30
I
’
’
’
’
’
0
5
10
15
20
25
c 30
Elution time (min)
FIG. 3 HPLC analysis of the serosal compartment of the Sweetana-Grass ditision chambers atIer the addition of 0.1mM of enterostatin radiolabeled with S uCi of [3H-Pro”]enterostatin to the mucosal compartment in absence (A) and presence (B) of 1mM DFP inhibitor. Results are those of one typical experiment. The latters have been shown to be responsible for the metabolism of prohne-containing peptides such as enterostatm and casomorphins which are resistant to the action of pancreatic endo- and exopeptidases (8,lO). In a previous study, we have shown that enterostatin hydrolysis involved two P and namely carboxypeptidase brush-border membrane post-proline peptidases, dipeptidylaminopeptidase IV (8).
Enterostatin Absorption
Vol. 59, No.s 2.5/26, 1996
,o
20
30
40
50
0
10
20
30
40
so
2153
0
IV
zu
30
4u
50
Time (min)
FIG. 4 Effect of DFP on the in vitro intestinal absorption of enterostatin. DFP 0 (A) or 0.1 rnmol. 1’ (0) was added to the mucosal and serosal compartment of rat jejunum mounted in Sweetana-Grass diffusion chambers. At t=O, enterostatin (0.4 pmol, 5 l&i) was added to the mucosal compartment and the accumulation of total radioactivity (A) and radiolabeled VPDP (B) was measured sequentially in the serosal compartment. The effect of DFP on the paracellular absorption of solute was assessed by measuring the transport of D[‘4C]mannito1 in the same conditions (C). Results are means f SD of n=6 (A and B) or n=4 (C) experiments and are expressed as percent of the mucosal amount of radiolabeled peptide (A and B) or mannitol (C) absorbed per cm’ of mucosa. * significantly diierent from control, pcO.05. In that study, diprotm A, a DPP IV specific inhibitor, was responsible for a 40% reduction in enterostatin degradation rate. The results reported here confirm the involvement of this two enzymes. However, the replacement of diprotin A by the DPP IV inhibitor DFP results in a reevaluation of the role of DPP IV in the degradation of enterostatin. Addition of 1 mmol.1~’ DFP reduces the rate of degradation of enterostatin by 80 % with BBh4V. This discrepancy may arise from a partial degradation of diprotin A (Ile-Pro-Ile) by DPP IV (II). The inhibition by DFP of a post-proline cleaving enzyme (PPCE) differing from DPP IV may represent an alternative explanation. DFP is a serine protease inhibitor and has been shown to inhibit DPP IV as well as PPCE and some post-proline carboxypeptidases (12). The latter enzymes are mainly present in brain and are involved in the inactivation of neuropeptides (12) and the presence of PPCE was also reported in the intestine and kidney membranes (13). Therefore, a limited contribution of PPCE to the intestinal degradation of enterostatin may be responsible for the difference in the inhibition potencies of DFP and diprotin A The loss of valine-contakg peptides observed upon incubation of enterostatin with control and EDTA-treated brush-border membranes also suggest that the dipeptide VP, but not VPDP, may be further processed to Valine and Proline, which can not be detected on the chromatograms. The strict dipeptidase prolidase is likely to be responsible for this degradation. This enzyme is not inhibited by EDTA and may be recovered with membrane fraction, although it is mainly cytosolic (14,15). Our results also indicate that DFP does not completely prevent the degradation of enterostatin in the mucosal and serosal compartment of the chamber. This observation is consistent with the results obtained with puritied brush-border membranes indicating the involvement of a membrane-bound carboxypeptidase P in the degradation of this peptide. It seems conceivable that such an activity may also be responsible for the conversion of enterostatin to its des-arginine metabolite in the serosal
2154
Enterostatin Absorption
Vol. 59, No.s 25/24, 1996
compartment of the Sweetana-Grass ditision chambers. With brush-border membrane preparation, an almost complete inhibition of the production of VPDP is achieved with 1 mM EDTA, thus indicating that the carboxypeptidase P is sensitive to chelating agents. Although many reports now suggest that small but biologically significant amounts of intact oligopeptides are absorbed across the epithelium of the small intestine (16-20) the mechanisms underlying this absorption remain unclear. The size of the peptides appears to be a critical determinant for the interaction with the I-I+-peptide cotransporter (21) and their is no evidence at present for the existence of specific transport mechanisms for peptides above 4 residus. It is usually assumed that these molecules are absorbed by passive transcellular or paracellular dif%.rsion and that the brush-border peptidases are the main limiting step of this process. The absorption of peptidase-resistant analogues of vasopressin across everted segment of rat jejunum has be shown to be more efficient than that of the natural peptide (19). In the same way, a promotion of the absorption of intact bioactive peptides has been reported for vasopressin and morphiceptin analogues atler treatment of intestinal mucosal sheets mounted in ussing chambers with peptidase inhibitors (19, 20). A reevaluation of the role of dipeptidylaminopeptidase IV in the degradation of enterostatin by intestinal mucosal peptidases suggests that an increase in the intestinal absorption of this peptide can be achieved through the inhibition of this enzyme. Our results indicate that inhibition of mucosal serine proteases by DFP promotes the absorption of enterostatin across rat jejunum in vitro. Addition of DPP resulted in a transient recovery of intact enterostatin and an accumulation of des-arginineenterostatin in the serosal compartment of the chambers. A promotting effect of DF’P through a non specific increase in the paracellular absorption of solute can be ruled out since the addition of DPP to the incubation buffer neither increased the absorption of the inert paracellular marker D-[‘4C]mannitol, nor affected the electrical conductance of the rat jejunum. Rather, it seems clear that the absorption of enterostatin and des-arginine-enterostatin across DFP-treated rat jejunum is a consequence of the inhibition of serine peptidases. No attempt was done to investigate the consequence of carboxypeptidase P inhibition on the in vitro transport of enterostatin across rat jejunum since BDTA has been shown to a&ct tight junctions through calcium chelation, and thus to increase the paracellular absorption of solute across the intestinal epithelium (22). Therefore, we would not have been able to discriminate between effects resulting from enzymatic inhibition and paracellular shunting. In conclusion, our results provide evidence that the absorption of intact enterostatin is hampered by brush-border membrane peptidases and that the inhibition of DPP IV stabilize peptide, hence improving its intestinal absorption, Some reports also suggest that absorption of enterostatin may also occurs in viva. In human a biphasic increase in the concentration of enterostatin in serum has been reported to occur in response to a meal (23) and plasma enterostatin level was shown to increase after a high fat, a high fat/sucrose or a low-fat diet in rats (J. Mei, unpublished results). The production of enterostatin in the intestinal lumen as a consequence of the tryptic activation of pancreatic secretion occurs conwmitantly to the release of digestion products, including glucose, peptides, amino acids and free f&y acids. Whether some of these products may exert an enhancing potency effect on enterostatin absorption through a modulation of brush-border membrane peptidase remains to be determined.
References 1. D.M. MATTHEWS, Protein Absorption: Developement and Present State of the Subject, Wiley-Liss, New York. (1991). 2. C. ERLANSON-ALBERTSSON and A. LARSSON, Regul. Pept. 22 325-331 (1988). 3. N.S. SHARGILL, S. TSUJII, G.A. BRAY, C. ERLANSON-ALBERTSSON, Brain Res. 544 137-140 (1991).
Vol. 59, Nos 25/X, 1996
4. J. MEI and C. ERLANSON-ALBERTSSON,
Enterostatin Absorption
215.5
Regul. Pept. 23 209-218 (1992). 5. J. ME1 and C. ERLANSON-ALBERTSSON, Obesity Research 41161-165 (1996). 6. C. ERLANSON-ALBERTSSON, B. WESTRGM, S. PIERZYNOWSKI, S. KARLSSON and B. AI-&N, Pancreas 6 619-624 (1991). 7. J.F. HUNEAU, C. ERLANSON-ALBERTSSON, C. BEAUVALLET and D. TOME, Regul. Pept. 54 495-503 (1994). 8. M. BOURAS, J.F. HUNEAU, C. LUENGO, C. ERLANSON-ALBERTSSON and D. TOME, Peptides 16 399-405 (1995). 9. E. BROT-LAROCHE and F. ALVARADO, Biochim. Biophys. Actam 175-181 (1981). 10. Y. MIYAMOTO, V. GANAPATHY, A. BARLOS, K. NEUBERT, A. BARTH and F. H. LEIBACH, Am. J. Physiol. 252 F670-F677 1987. 11. J. RAHFELD, M. SCHIERHORN, B. HARTRODT, K. NEUBERT and J. HEINS, Biochim. Biophys. Acta 1076 314-316 (1991). 12. S. KALWANT and A.G. PORTER Biochem. J. 276 237-244 (1991). 13. T. YOSHIMOTO, K. OGITA, R. WALTER, M. KOIDA and D. TSURU. Biochim. Biophys. Acta 569 184- 192 (1979). 14. R. MENTLEIN. FEBS lett. 234 251-256 (1988). 15. R. WALTER, W.H. SIMMONS and T. YOSHIMOTO, Mol. Cell. Biochem., 30, 11 l-125 (1980). 16. K. ATISOOK and J.L. MADARA, Gastroenterology 100 719-724. (1991). 17. D. TOME, A.M. DUMONTIER, M. HAUTEFEUILLE and J.F. DESJEUX, Am. J. Physiol. 253 G737-G744 (1987). 18. S. LUNDIN, H.G. FOLKESSON, S.G. PIERZYNOVSKI and H. BENGTSSON, Peptides fi 809-814 (1994). 19. A. UNGELL, A. ANDREASSON, K. LUNDIN and L. UTTER J. Pharm. Sci. 81640-645 (1992). 20. S. MAHE, D. TOME, A.M. DUMONTIER and J.F. DESJEUX. Peptides B 45-52 (1989). 21. Y.J. FEI, Y. KANAI., S. NUSSBEREGER, V. GANAPATHY, F.H. LEIBACH, M.F. ROMERO, S.K.SINGH, W.F. BORON and M.A. HEDIGER. Nature 368 563-566 (1994). 22. B.J. AUNGST, J.A. BLAKE and M.A. HUSSAIN, J. Pharmacol. Exp. Ther. 259 139-145 (1991). 23. R.C. BOWYER, W.M. ROWSTON, A.M.T. JEHANLI and J.H. HERMON-TAYLOR, Gut 34 1520-1525 (1993).