Clinical Science (1979) 57,529-534
Release of dipeptide hydrolase activities from rat small intestine perfused in vitro and in vivo
M. L. G . G A R D N E R A N D J A N E A. P L U M B Department of Biochemistry. The University of Edinburgh Medical School, Edinburgh, Scotland, U.K.
(Received 14 March 1979; accepted 24 July 1979)
Summary 1. Hydrolase activities against three dipeptides were measured in mucosal cytoplasm in unperfused intestines and in mucosal cytoplasm, luminal effluents and serosal secretions after perfusion in vitro and in vivo for 1 h. Intestines in vitro were prepared both from anaesthetized rats and from freshly killed rats. 2. Only 0.6-1.9% of the initial cytoplasmic activity was recovered in the luminal effluent when intestines in vitro were prepared from anaesthetized rats. Recoveries in luminal effluents were similar (1.3-3.3%) during perfusion in vivo. 3. Losses of dipeptidases into the luminal effluent were four to eight times greater when intestines in vitro were prepared from freshly killed animals. 4. Similar losses of dipeptidases into the secretion on to the serosal surface were observed; they too were much greater when intestines were prepared from freshly killed animals. 5. Small losses of mucosal DNA during perfusion were also observed; however, losses of cytoplasmic peptidases were consistently slightly greater. 6. Enzyme loss therefore probably occurs both by sloughing of whole cells and by a more specific process which is greatly influenced by experimental procedure. Caution is necessary in the inter-
pretation of peptide transport experiments in vitro, although the possibility that intraluminal hydrolysis is of physiological significance must not be excluded. Key words: intestine, peptidase, peptide hydrolase. Introduction
Peptidase enzymes have been found in the contents of the small intestinal lumen in vivo in many species (including man, dog, rat and pig), but their physiological significance in digestion has not been resolved. Some intestinal preparations in vitro release considerable amounts of cytoplasmic peptidases into incubation media (Joseffson & Sjostrom, 1966; Lindberg, 1972; Lindberg, Norin & Sjostrom, 1975; Silk & Kim, 1976). It has been suggested that this process may be physiological (Joseffson & Sjostrom, 1966) and that it responds to various specific stimuli (e.g. Gotze, Adelson, Hadorn, Portmann & Troesch, 1972; Eloy, Vaultier, Raul, Mirhom, Clendinnen & Grenier, 1978); alternatively, it may be an artifact associated with a dying tissue since intestines in vivo apparently release much less of these enzymes (Lindberg, 1972; Silk & Kim, 1975). As Silk & Kim (1976) stressed, these observations raise questions as to the normal sites of peptide hydrolysis and the interpretation of studies in vitro into peptide absorption. If hydrolases are secreted into the intestinal lumen, then this may be an important site of terminal digestion of protein
Correspondence: Dr M. L. G. Gardner, Department of Biochemistry, The University of Edinburgh Medical School, Teviot Place, Edinburgh EH8 9AG, Scotland, U.K. 36
529
530
M . L . G . Gardner and J. A . Plumb
and it is important to characterize the process. If, however, peptidases leak abnormally into the lumen in experiments in vitro then it may not be valid to use these preparations to study the physiological relationship between peptide hydrolysis and transport. Furthermore if peptidases leak into the submucosal tissue fluid this might explain why most investigators have not observed transmural transport of intact peptides during experiments in vifro (see also Wiggans &Johnston, 1959). We have now measured dipeptide hydrolase activities released into a luminal perfusate and into the secretion on to the serosal surface of the isolated intestine during perfusion for an hour, and the peptide hydrolase activities remaining in the cytosol of the mucosal cells after perfusion. Since absorptive activity in vitro is greatly inferior if the animal is killed before intestinal perfusion is commenced (Gardner, 1978) we also studied the losses of peptide hydrolase activities from perfused intestines prepared rapidly from freshly killed rats. For comparison, losses from intestines perfused in vivo were also measured and the DNA contents of the mucosa after perfusion were determined. A preliminary abstract has reported part of this work (Plumb & Gardner, 1979).
anaesthesia. The segmented-flow perfusion technique and apparatus of Fisher & Gardner (1974) were used except that the small intestine was left in situ partly outside the abdominal cavity with the vasculature intact. The intestine was covered with moist cotton wool and a polyethylene sheet and was maintained at approximately 38OC by an overhead lamp. As with the experiments in vitro, the intraluminal distension pressure was approximately 30 cm water. Intestine from killed animals. In specific experiments in which the animals were not maintained under anaesthesia, ether-anaesthetized rats were killed rapidly by stunning and cervical dislocation. The intestine was then rinsed with oxygenated NaHCO,/NaCI solution at 38OC, connected to the perfusion apparatus, excised and placed in the organ chamber in the usual way as rapidly as possible. The time between killing of the rat and establishment of luminal perfusion was about 4 min. Perfusion medium. This was the modified Krebs-Henseleit bicarbonate-buffered solution (pH 7.4) equilibrated with CO, + 0, (5:95) at 38OC used by Fisher & Gardner (1974). It contained glucose (28 mmol/l) but not phenol red.
Experimental procedure Methods
Animals Female rats of a local Wistar strain and weighing 160-200 g were used. They had been kept in conditions of controlled day-length with free access to water and Oxoid Diet 86 (Oxoid Ltd, Basingstoke, Hants., U.K.) for at least a week before use.
Intestinal perfusion Perfusion in vitro. Whole jejunum plus ileum from the ligament of Treitz to the ileo-caecal valve was rinsed with oxygenated NaHCO,/NaCI solution at 38OC and connected to the perfusion apparatus while the rat was maintained under tight ether anaesthesia. The intestine was removed and the animal killed only after complete luminal perfusion had been established. The apparatus and single-pass ‘segmented-flow’ perfusion technique were as detailed by Fisher & Gardner (1974). Perfusion in vivo. Animals were anaesthetized with urethane (1 ml of 0.8 mg/ml solution injected subcutaneously) while under temporary ether
Either after the intestine had been rinsed with the NaHCO,/NaCI solution (unperfused intestines) or after perfusion for 60 rnin the intestine was taken rapidly from the apparatus to a cold room. (In a few experiments intestines were removed from the perfusion apparatus at 20 and 40 rnin.) With the intestine on a dry glass sheet on ice the mucosa was scraped off. The mucosa was homogenized immediately with 7.5 ml of chilled NaCl solution (0-9 g/100 ml) and kept on ice. Within 20 min a portion of the homogenate (5.0 ml) was centrifuged at 20000 g for 60 rnin at 4OC and the supernatant taken for immediate assay of peptide hydrolase activities. (Estimations of sucrase activity suggested that less than 3% of the peptidase activity measured in this supernatant could have arisen from brush-border contamination.) Aliquots of the homogenate were also used within 3 h for DNA estimation, and a portion (1 ml) was dried at 37OC to constant weight for determination of water content. The luminal effluents and secretions were collected and kept on ice; the peptidase assays on them were started within 20 min of completion of the perfusion.
53 1
Intestinal peptidase release Peptide hydrolase assays Hydrolase activities against Gly-L-Met, L-L~uGly and L-Val-L-Leu were determined on 40 pl aliquots by the two-step amino acid oxidase method of Nicholson & Kim (1975), except that glycerol was omitted because of its inhibition of all three activities. Optimum substrate concentrations of Gly-L-Met, L-L~u-GIYand ~ - V a l - ~ - L e were u found to be 20, 15 and 15 mmol/l respectively. Samples were diluted on ice within the ranges 1 :320-1: 2560 for supernatants, 1 :4-1:30 for secretions and 0-1 : 2 for luminal effluents to give activities giving less than 1 0 0 nmol of amino acid released/20 min. All assays were carried out at least in triplicate. Peptide hydrolase activities were expressed in terms of units/ml or units/mg of DNA, where 1 unit is defined as the amount of enzyme catalysing the hydrolysis of 1 pmol of substrate/min under the assay condition of Nicholson & Kim (1975). The total activity in the supernatant was obtained from the activity/ml and the total water content of the homogenate: uniform distribution of enzyme throughout the water volume was assumed.
DNA estimation DNA in the mucosal homogenate of whole intestine was determined by the ethidium bromide fluorescence method of Le Pecq & Paoletti (1966) as modified by Karsten & Wollenberger (1972), except that the RNAase was used at 20 mg/ml, the incubation time at 37OC was extended to 1 h and all volumes were doubled. Samples were assayed in duplicate at each of three dilutions: 1 :80,1 : 160 and 1 :320. A standard curve was set up daily with freshly prepared solutions of calf thymus DNA (Sigma London Chemical Co.).
Results
Appearance of peptide hydrolase activities in Iuminal efluents and serosal secretions Dipeptidase activities were determined in the luminal effluents and the secretions on to the serosal surface of the intestine pooled over the perfusion period of 1 h. The results are shown in Fig. 1, which also shows in parentheses these activities expressed as a percentage of the activities in the mucosal cyto-
la1 Serosal secretion ( 1 1%1
20
0
Gly-Met
Leu-Gly
Val-Leu
FIG. 1. Peptide hydrolase activities recovered in (a) secretion on to the serosal surface of the intestine, and (b) luminal efRuent during perfusion for 1 h in uitro (10 intestines prepared from ether-anaesthetized rats) (open bars), in vitro (seven intestines prepared from freshly killed rats) (hatched bars) and in viuo (stippled bars, four rats). Values are means k SEM. Values in parentheses show mean values expressed as percentages of activities found in mucosal cytoplasm of unperfused intestines.
plasm of the unperfused intestines. It is clear that only small amounts of dipeptidase activities were found in the luminal effluents during perfusion in vivo or during perfusion in vitro of intestine prepared from anaesthetized rats. No significant differences were observed in these two types of perfusion. In contrast, when intestines for perfusion in vitro were prepared from freshly killed rats (hatched bars in Fig. l), the amounts of activity in the luminal effluent were dramatically greater by factors of from 3.8 (against Leu-Gly) to 7.8 (against Gly-Met). A corresponding increase (approximately threefold) was seen in the activities found in the serosal secretions from the intestines prepared from freshly killed animals. However, the total activity found in the secretion was invariably much less than that in the luminal effluent. When intestines from anaesthetized rats are perfused in vitro less than 1.9% of the initial cytoplasmic activity is recovered in the luminal effluent and less than 0.5% in the serosal secretion over 1 h.
532 1000
M . L. G . Gardner and J. A . Plumb
1
la1 Val-Leu T
800
-.?
600 400
a
(bl Leu-Gly
A
x 20
40
60
Duration of perfusion (min)
.-0
5 -ma V
(c) Gly-Met
2800
T
2400 5%)
2000
0 001
177 7%) P < 0 05
1600 1200 Prepn. in vitro from anaesthetired rats
Prepn. in vitro from freshly killed rats
Prepn. in viva
FIG. 2. Cytoplasmic peptide hydrolase activities in mucosa of small intestine before (open bars) and after (hatched bars) perfusion in uifro or in uiuo for 1 h. The intestines for perfusion in uifro were prepared both from ether-anaesthetized rats and freshly killed rats. Substrates were (a) L-Val-L-Leu, (b) L-Leu-Gly and (c) Gly-L-Met. Values are means ? SEM, with the number of animals shown within the bars. Values in parentheses show the activities in perfused intestines expressed as percentages of the activities in unperfused intestines, and the P values indicate the significance of the differences.
Cytoplasmic peptide hydrolase activities before and after perfusion Fig. 2 shows the hydrolase activities against LVal-L-Leu, L-Lxu-GIY and Gly-L-Met in the mucosal cytoplasm from whole jejunum plus ileum before and after perfusion for 1 h. In all cases there was less cytoplasmic peptidase activity after perfusion than in the unperfused intestines, although this difference was not always significant (see Fig. 2). The peptidase activities in the unperfused, but saline-rinsed, intestines were slightly but not significantly ( P > 0.02) lower in the killed animals and those anaesthetized by urethane (i.e. the controls for those perfused in vivo) than in the normal ether-anaesthetized ones.
FIG. 3. Peptide hydrolase activities in mucosal supernatant of intestines after perfusion in uifro for 20, 40 or 60 min. Data have been expressed as a percentage of values found in unperfused intestines. Each point is a mean of observations on four intestines. Substrates were: o--o, L - V ~ I - L - L ~ U A---A, ; ~-Leu-Gly;L U Gly-L-Met.
In order to test whether the enzyme loss from the cytoplasm was a continual process during the hour of perfusion, some experiments on intestines perfused in vitro (from anaesthetized animals) were terminated at 20, 40 or 60 min. Fig. 3 shows that the loss of activity of each of the three peptide hydrolases was approximately linear with time. Mucosal DNA content before and after perfusion Table 1 shows the DNA content of the mucosa of unperfused intestines and of intestines after perfusion for 1 h. In all cases the mucosal DNA content was slightly less (by 6-14%) after perfusion than in the unperfused intestines, although this difference was never significant (P > 0.05). However, the difference between the DNA content of the perfused intestines prepared from dead animals and the unperfused ones prepared from anaesthetized rats was significant ( P < 0.05). Fig. 4 shows the cytoplasmic peptide hydrolase
activities expressed as units/mg of DNA in unperfused intestines and in intestines after perfusion for 1 h. Although in all cases there are losses of enzyme activity, relative to DNA content, during perfusion they are not significant (P > 0.1) except for Leu-Gly peptidase from the killed animals (P < 0.05).
Total recovery of peptide hydrolase activities after perfusion The total peptidase activities recovered in luminal effluent, serosal secretion and mucosal
Intestinal peptidase release
TABLE1. DNA content of whole small intestinal mucosa
,
533 T
Ic) Gly-Met
100
before and afler perfusion in vitro or in vivo for 1 h Intestines for perfusion in vitro were prepared both from ether-anaesthetized rats and freshly killed rats. Values are means ? SEM. Numbers of rats are shown in parentheses.
T,
75
50
Preparation
DNA content (rndintestine)
Unperfused
After perfusion for 1 h
T
z
25
. -E
.-e
I n virro. from
anaesthetized rats
26.5 f 1.2 (8)
22.7 f 0.8 (6)
In uitro, from freshly killed rats I n vivo
24.1 f 0.7 (8) 25.6 f 2.3 (6)
21.9 f 0.7 (7) 24.0 f 2.4 (4)
O Ib) Leu-Gly
.
cytoplasm after perfusion for 1 h amounted to 80.5% of the activities in the mucosal cytoplasm of unperfused intestines (mean of all experiments). ( a )Val-Leu T
r
Discussion
These results extend earlier studies (Silk & Kim, 1975; Lindberg, 1972) which demonstrated that intestines in uitro can release peptide hydrolases into perfusion and incubation media. However, we find that the amounts appearing in the luminal effluent are normally very small, amounting over 1 h to only 0.6-1.9% of the initial cytoplasmic content (Fig. 1). The enzyme activities recovered in the luminal effluent during perfusion in viuo and in vitro were closely similar provided that the preparations in vitro were set up from anaesthetized animals. However, when the segments for perfusion were excised from freshly killed animals, the enzyme activities found in the luminal effluent were four to eight times greater than when the animals were maintained under ether anaesthesia while the intestines were beiig connected to the perfusion apparatus. Similarly, the activities found in the secretion on to the serosal surface of the intestine were small (0.345% of the initial cytoplasmic activity) when the intestines were set up from anaesthetized rats. These values also increased substantially when the animals were killed before removal of the intestine. Fig. 3 indicates that the release of enzymes into the luminal effluent was a continual process throughout the hour of perfusion rather than one which commenced after a certain time. Thus these data indicate that the release of hydrolases from the perfused intestine is normally small although it can be increased experimentally. Although this suggests that the release of pepti-
l0 o
i
Prepn. in virro from anaesthetized rats
Prepn. in virro from freshly killed rnts
Prepn. in viw
FIG.4. Cytoplasmic peptide hydrolase activities per mg
of DNA in small intestine before (open bars) and aRer (hatched bars) perfusion in uitro or In vivo for 1 h. Intestines for perfusion in uitro were prepared both from ether-anaesthetized rats and freshly killed rats. Substrates were (a) ~-Val-~-Leu, (b) L-L~u-G~Y or (c) Gly-L-Met. Data in parentheses show the activities aRer perfusion for 1 h expressed as a percentage of activities in unperfused intestines. Values are means f SEM, with the number of intestines shown within the bars in (a).
dases is not a normal physiological process, it must be noted that the amount of peptidase activity within the mucosa is so large that the release of 1% cannot necessarily be dismissed as functionally insignificant, especially in view of the possibility that hormones may stimulate this process (Giitze et al., 1972; Eloy et al., 1978). The high rates of peptidase release in vitro, which Lmdberg (1972) and Silk & Kim (1976) reported, probably had been influenced by the method of preparation. Hence their remarks about the interpretation of studies in vitro into peptide absorption, while valid for many studies in uitro, are greatly dependent on the particular technique used.
534
M . L. G . Gardner and J. A . Plumb
Although it is not known that similar results to ours would be obtained from all other types of intestinal preparation in vitro, or whether temporary anoxia specifically causes the enzyme release, our results indicate the merit of using anaesthesia and the precautions of Fisher & Parsons (1949) and Fisher & Gardrnr (1974) for the preparation of isolated intestines for perfusion, especially during studies on peptide absorption and hydrolysis or on peptide hormones. Furthermore, absorptive activity of isolated perfused intestine prepared from anaesthetized rats is substantially greater than if the animals are killed before removal of the intestine (Gardner, 1978). The relatively large activity against Gly-L-Met found in the luminal efAuent and serosal secretion (Fig. 1) reflects the relatively large amount of GlyMet hydrolase activity in the cytoplasm (Fig. 2); this is probably not due to differences in the stability of the three enzyme activities studied since we find Gly-Met hydrolase activity to be the most labile (unpublished observations). We do not know whether the enzymes released into the lumen are of cytoplasmic or brush-border origin: however, the work of Silk & Kim (1976) and Silk, Nicholson & Kim (1976) suggests that they may be predominantly cytoplasmic with some contribution by ileal brush border. We find that DNA is also lost (6-14% in 1 h) from the perfused intestines (Table 1). The small DNA loss from intestines perfused in vitro accords with the good absorptive stability of the preparation reported by Fisher & Gardner (1974). However, in the experiments in vitro at least, the losses of enzyme activity from the cytoplasm during perfusion were proportionately slightly greater than the losses of DNA (Fig. 4). Although this difference is not quite significant it was consistently observed in all three groups of experiments; hence it is probable that enzymes are lost from the mucosa by a combination of sloughing of whole cells (leading to the slight DNA loss) and a more specific process. This is consistent with Lindberg’s (1972) observation that disaccharidase activities were not released in vitro along with the dipeptidase activities and with the substantially larger specific activity of intraluminal Gly-Leu hydrolase seen in vitro than in vivo (Lindberg et al., 1975, p. 228). The specific component of the release obviously requires further investigation.
Acknowledgments
We are indebted to the Medical Research Council and to Eaton Laboratories Ltd for financial support. We also thank Mrs Anne Pryde for technical assistance and Dr R. C. Heading for valuable discussions, and Miss Helen Scott for secretarial assistance.
References VAULTIER, J.P., RAUL, F., MIRHOM, R., CLENDMNEN, G. & GRENIER, J.F. (1978) Hormonal stimulation of intestinal brush border enzymes release. Research in Experimental Medicine, 172, 109- 12 I. FISHER,R.B. & GARDNER, M.L.G. (1974)A kinetic approach to the study of absorption of solutes by isolated perfused small intestine. Journal of Physiology (London), 241, 21 1ELOY, R.,
234. FISHER, R.B. & PARSONS,D.S. (1949) A preparation of surviving rat small intestine for the study of absorption. Journal of Physiology (London), 110,3646. GARDNER, M.L.G. (1978) The absorptive viability of isolated intestine prepared from dead animals. Quarterly Journal of Experimental Physiologv. 63,93-95. GOTZE,H., ADELSON,J.W., HADORN,H.B., PORTMANN, R. & TROESCH, V. (1972)Hormoneelicited enzyme release by the small intestinal wall. Gut, 13,471-476. JOSEFFSON, L. & SJOSTROM,H. (1966)Intestinal dipeptidases. IV. Studies on the release and subcellular distribution of intestinal dipeptidases of the mucosa cells of the pig. Acla Physiologica Scandinavica, 67.27-33. KARSTEN,U. & WOLLENBERCER, A. (1972)Determination of DNA and RNA in homogenised cells and tissues by surface Buorimetry. Analytical Biochemistry, 46, 135-148. LE PECQ, J.-B. & PAOLETTI,C. (1966) A new fluonmetric method for RNA and DNA determination. Analytical Biochemistry, 17,loQ-107. LINDBERG, T. (1972) Discussion. In: Peptide Transport in Bacteria and Mammalian Gut (CIBA Foundation Symposium), p. 91. Ed. Elliott, K. & O’Connor, M. Associated Scientific Publishers, Amsterdam. LMDBERC, T., NOREN,0. & SJOSTROM,H. (1975)Peptidases of the intestinal mucosa. In: Peptide Transport in Protein Nutrition, pp. 204-242. Ed. Matthews, D.M. & Payne, J.W. North-Holland, Amsterdam. NICHOLSON, J.A. & KIM,Y.S.(1975)A one-step L-amino acid oxidase assay for intestinal peptide hydrolase activity. AnalyticalBiochemistry, 63, 110-1 17. PLUMB,J.A. & GARDNER,M.L.G. (1979) The physiological viability of isolated intestine prepared from dead animals. Gastroenterologie Clinique el Biolagique, 3, 176. SILK, D.B.A. & KIM, Y.S. (1975) A study of intraluminal peptide hydrolase activity in the rat. Clinical Science and Molecular Medicine, 49, 523-526. SILK,D.B.A. & KIM,Y.S.(1976)Release of peptide hydrolases during incubation of intact intestinal segments in vitro. Journal of Physiology (London), 258489497. SILK,D.B.A., NICHOLSON, J.A. & KIM, Y.S. (1976)Hydrolysis of peptides within lumen of small intestine. American Journal of Physiology, 231, 1322-1329. WICCANS,D.S. & JOHNSTON,J.M. (1959)The absorption of peptides. Biochimica et Biophysica Acta. 32,69-73.
Release of dipeptide hydrolase activities from rat small intestine perfused in vitro and in vivo
M. L. G . G A R D N E R A N D J A N E A. P L U M B Department of Biochemistry. The University of Edinburgh Medical School, Edinburgh, Scotland, U.K.
(Received 14 March 1979; accepted 24 July 1979)
Summary 1. Hydrolase activities against three dipeptides were measured in mucosal cytoplasm in unperfused intestines and in mucosal cytoplasm, luminal effluents and serosal secretions after perfusion in vitro and in vivo for 1 h. Intestines in vitro were prepared both from anaesthetized rats and from freshly killed rats. 2. Only 0.6-1.9% of the initial cytoplasmic activity was recovered in the luminal effluent when intestines in vitro were prepared from anaesthetized rats. Recoveries in luminal effluents were similar (1.3-3.3%) during perfusion in vivo. 3. Losses of dipeptidases into the luminal effluent were four to eight times greater when intestines in vitro were prepared from freshly killed animals. 4. Similar losses of dipeptidases into the secretion on to the serosal surface were observed; they too were much greater when intestines were prepared from freshly killed animals. 5. Small losses of mucosal DNA during perfusion were also observed; however, losses of cytoplasmic peptidases were consistently slightly greater. 6. Enzyme loss therefore probably occurs both by sloughing of whole cells and by a more specific process which is greatly influenced by experimental procedure. Caution is necessary in the inter-
pretation of peptide transport experiments in vitro, although the possibility that intraluminal hydrolysis is of physiological significance must not be excluded. Key words: intestine, peptidase, peptide hydrolase. Introduction
Peptidase enzymes have been found in the contents of the small intestinal lumen in vivo in many species (including man, dog, rat and pig), but their physiological significance in digestion has not been resolved. Some intestinal preparations in vitro release considerable amounts of cytoplasmic peptidases into incubation media (Joseffson & Sjostrom, 1966; Lindberg, 1972; Lindberg, Norin & Sjostrom, 1975; Silk & Kim, 1976). It has been suggested that this process may be physiological (Joseffson & Sjostrom, 1966) and that it responds to various specific stimuli (e.g. Gotze, Adelson, Hadorn, Portmann & Troesch, 1972; Eloy, Vaultier, Raul, Mirhom, Clendinnen & Grenier, 1978); alternatively, it may be an artifact associated with a dying tissue since intestines in vivo apparently release much less of these enzymes (Lindberg, 1972; Silk & Kim, 1975). As Silk & Kim (1976) stressed, these observations raise questions as to the normal sites of peptide hydrolysis and the interpretation of studies in vitro into peptide absorption. If hydrolases are secreted into the intestinal lumen, then this may be an important site of terminal digestion of protein
Correspondence: Dr M. L. G. Gardner, Department of Biochemistry, The University of Edinburgh Medical School, Teviot Place, Edinburgh EH8 9AG, Scotland, U.K. 36
529
530
M . L . G . Gardner and J. A . Plumb
and it is important to characterize the process. If, however, peptidases leak abnormally into the lumen in experiments in vitro then it may not be valid to use these preparations to study the physiological relationship between peptide hydrolysis and transport. Furthermore if peptidases leak into the submucosal tissue fluid this might explain why most investigators have not observed transmural transport of intact peptides during experiments in vifro (see also Wiggans &Johnston, 1959). We have now measured dipeptide hydrolase activities released into a luminal perfusate and into the secretion on to the serosal surface of the isolated intestine during perfusion for an hour, and the peptide hydrolase activities remaining in the cytosol of the mucosal cells after perfusion. Since absorptive activity in vitro is greatly inferior if the animal is killed before intestinal perfusion is commenced (Gardner, 1978) we also studied the losses of peptide hydrolase activities from perfused intestines prepared rapidly from freshly killed rats. For comparison, losses from intestines perfused in vivo were also measured and the DNA contents of the mucosa after perfusion were determined. A preliminary abstract has reported part of this work (Plumb & Gardner, 1979).
anaesthesia. The segmented-flow perfusion technique and apparatus of Fisher & Gardner (1974) were used except that the small intestine was left in situ partly outside the abdominal cavity with the vasculature intact. The intestine was covered with moist cotton wool and a polyethylene sheet and was maintained at approximately 38OC by an overhead lamp. As with the experiments in vitro, the intraluminal distension pressure was approximately 30 cm water. Intestine from killed animals. In specific experiments in which the animals were not maintained under anaesthesia, ether-anaesthetized rats were killed rapidly by stunning and cervical dislocation. The intestine was then rinsed with oxygenated NaHCO,/NaCI solution at 38OC, connected to the perfusion apparatus, excised and placed in the organ chamber in the usual way as rapidly as possible. The time between killing of the rat and establishment of luminal perfusion was about 4 min. Perfusion medium. This was the modified Krebs-Henseleit bicarbonate-buffered solution (pH 7.4) equilibrated with CO, + 0, (5:95) at 38OC used by Fisher & Gardner (1974). It contained glucose (28 mmol/l) but not phenol red.
Experimental procedure Methods
Animals Female rats of a local Wistar strain and weighing 160-200 g were used. They had been kept in conditions of controlled day-length with free access to water and Oxoid Diet 86 (Oxoid Ltd, Basingstoke, Hants., U.K.) for at least a week before use.
Intestinal perfusion Perfusion in vitro. Whole jejunum plus ileum from the ligament of Treitz to the ileo-caecal valve was rinsed with oxygenated NaHCO,/NaCI solution at 38OC and connected to the perfusion apparatus while the rat was maintained under tight ether anaesthesia. The intestine was removed and the animal killed only after complete luminal perfusion had been established. The apparatus and single-pass ‘segmented-flow’ perfusion technique were as detailed by Fisher & Gardner (1974). Perfusion in vivo. Animals were anaesthetized with urethane (1 ml of 0.8 mg/ml solution injected subcutaneously) while under temporary ether
Either after the intestine had been rinsed with the NaHCO,/NaCI solution (unperfused intestines) or after perfusion for 60 rnin the intestine was taken rapidly from the apparatus to a cold room. (In a few experiments intestines were removed from the perfusion apparatus at 20 and 40 rnin.) With the intestine on a dry glass sheet on ice the mucosa was scraped off. The mucosa was homogenized immediately with 7.5 ml of chilled NaCl solution (0-9 g/100 ml) and kept on ice. Within 20 min a portion of the homogenate (5.0 ml) was centrifuged at 20000 g for 60 rnin at 4OC and the supernatant taken for immediate assay of peptide hydrolase activities. (Estimations of sucrase activity suggested that less than 3% of the peptidase activity measured in this supernatant could have arisen from brush-border contamination.) Aliquots of the homogenate were also used within 3 h for DNA estimation, and a portion (1 ml) was dried at 37OC to constant weight for determination of water content. The luminal effluents and secretions were collected and kept on ice; the peptidase assays on them were started within 20 min of completion of the perfusion.
53 1
Intestinal peptidase release Peptide hydrolase assays Hydrolase activities against Gly-L-Met, L-L~uGly and L-Val-L-Leu were determined on 40 pl aliquots by the two-step amino acid oxidase method of Nicholson & Kim (1975), except that glycerol was omitted because of its inhibition of all three activities. Optimum substrate concentrations of Gly-L-Met, L-L~u-GIYand ~ - V a l - ~ - L e were u found to be 20, 15 and 15 mmol/l respectively. Samples were diluted on ice within the ranges 1 :320-1: 2560 for supernatants, 1 :4-1:30 for secretions and 0-1 : 2 for luminal effluents to give activities giving less than 1 0 0 nmol of amino acid released/20 min. All assays were carried out at least in triplicate. Peptide hydrolase activities were expressed in terms of units/ml or units/mg of DNA, where 1 unit is defined as the amount of enzyme catalysing the hydrolysis of 1 pmol of substrate/min under the assay condition of Nicholson & Kim (1975). The total activity in the supernatant was obtained from the activity/ml and the total water content of the homogenate: uniform distribution of enzyme throughout the water volume was assumed.
DNA estimation DNA in the mucosal homogenate of whole intestine was determined by the ethidium bromide fluorescence method of Le Pecq & Paoletti (1966) as modified by Karsten & Wollenberger (1972), except that the RNAase was used at 20 mg/ml, the incubation time at 37OC was extended to 1 h and all volumes were doubled. Samples were assayed in duplicate at each of three dilutions: 1 :80,1 : 160 and 1 :320. A standard curve was set up daily with freshly prepared solutions of calf thymus DNA (Sigma London Chemical Co.).
Results
Appearance of peptide hydrolase activities in Iuminal efluents and serosal secretions Dipeptidase activities were determined in the luminal effluents and the secretions on to the serosal surface of the intestine pooled over the perfusion period of 1 h. The results are shown in Fig. 1, which also shows in parentheses these activities expressed as a percentage of the activities in the mucosal cyto-
la1 Serosal secretion ( 1 1%1
20
0
Gly-Met
Leu-Gly
Val-Leu
FIG. 1. Peptide hydrolase activities recovered in (a) secretion on to the serosal surface of the intestine, and (b) luminal efRuent during perfusion for 1 h in uitro (10 intestines prepared from ether-anaesthetized rats) (open bars), in vitro (seven intestines prepared from freshly killed rats) (hatched bars) and in viuo (stippled bars, four rats). Values are means k SEM. Values in parentheses show mean values expressed as percentages of activities found in mucosal cytoplasm of unperfused intestines.
plasm of the unperfused intestines. It is clear that only small amounts of dipeptidase activities were found in the luminal effluents during perfusion in vivo or during perfusion in vitro of intestine prepared from anaesthetized rats. No significant differences were observed in these two types of perfusion. In contrast, when intestines for perfusion in vitro were prepared from freshly killed rats (hatched bars in Fig. l), the amounts of activity in the luminal effluent were dramatically greater by factors of from 3.8 (against Leu-Gly) to 7.8 (against Gly-Met). A corresponding increase (approximately threefold) was seen in the activities found in the serosal secretions from the intestines prepared from freshly killed animals. However, the total activity found in the secretion was invariably much less than that in the luminal effluent. When intestines from anaesthetized rats are perfused in vitro less than 1.9% of the initial cytoplasmic activity is recovered in the luminal effluent and less than 0.5% in the serosal secretion over 1 h.
532 1000
M . L. G . Gardner and J. A . Plumb
1
la1 Val-Leu T
800
-.?
600 400
a
(bl Leu-Gly
A
x 20
40
60
Duration of perfusion (min)
.-0
5 -ma V
(c) Gly-Met
2800
T
2400 5%)
2000
0 001
177 7%) P < 0 05
1600 1200 Prepn. in vitro from anaesthetired rats
Prepn. in vitro from freshly killed rats
Prepn. in viva
FIG. 2. Cytoplasmic peptide hydrolase activities in mucosa of small intestine before (open bars) and after (hatched bars) perfusion in uifro or in uiuo for 1 h. The intestines for perfusion in uifro were prepared both from ether-anaesthetized rats and freshly killed rats. Substrates were (a) L-Val-L-Leu, (b) L-Leu-Gly and (c) Gly-L-Met. Values are means ? SEM, with the number of animals shown within the bars. Values in parentheses show the activities in perfused intestines expressed as percentages of the activities in unperfused intestines, and the P values indicate the significance of the differences.
Cytoplasmic peptide hydrolase activities before and after perfusion Fig. 2 shows the hydrolase activities against LVal-L-Leu, L-Lxu-GIY and Gly-L-Met in the mucosal cytoplasm from whole jejunum plus ileum before and after perfusion for 1 h. In all cases there was less cytoplasmic peptidase activity after perfusion than in the unperfused intestines, although this difference was not always significant (see Fig. 2). The peptidase activities in the unperfused, but saline-rinsed, intestines were slightly but not significantly ( P > 0.02) lower in the killed animals and those anaesthetized by urethane (i.e. the controls for those perfused in vivo) than in the normal ether-anaesthetized ones.
FIG. 3. Peptide hydrolase activities in mucosal supernatant of intestines after perfusion in uifro for 20, 40 or 60 min. Data have been expressed as a percentage of values found in unperfused intestines. Each point is a mean of observations on four intestines. Substrates were: o--o, L - V ~ I - L - L ~ U A---A, ; ~-Leu-Gly;L U Gly-L-Met.
In order to test whether the enzyme loss from the cytoplasm was a continual process during the hour of perfusion, some experiments on intestines perfused in vitro (from anaesthetized animals) were terminated at 20, 40 or 60 min. Fig. 3 shows that the loss of activity of each of the three peptide hydrolases was approximately linear with time. Mucosal DNA content before and after perfusion Table 1 shows the DNA content of the mucosa of unperfused intestines and of intestines after perfusion for 1 h. In all cases the mucosal DNA content was slightly less (by 6-14%) after perfusion than in the unperfused intestines, although this difference was never significant (P > 0.05). However, the difference between the DNA content of the perfused intestines prepared from dead animals and the unperfused ones prepared from anaesthetized rats was significant ( P < 0.05). Fig. 4 shows the cytoplasmic peptide hydrolase
activities expressed as units/mg of DNA in unperfused intestines and in intestines after perfusion for 1 h. Although in all cases there are losses of enzyme activity, relative to DNA content, during perfusion they are not significant (P > 0.1) except for Leu-Gly peptidase from the killed animals (P < 0.05).
Total recovery of peptide hydrolase activities after perfusion The total peptidase activities recovered in luminal effluent, serosal secretion and mucosal
Intestinal peptidase release
TABLE1. DNA content of whole small intestinal mucosa
,
533 T
Ic) Gly-Met
100
before and afler perfusion in vitro or in vivo for 1 h Intestines for perfusion in vitro were prepared both from ether-anaesthetized rats and freshly killed rats. Values are means ? SEM. Numbers of rats are shown in parentheses.
T,
75
50
Preparation
DNA content (rndintestine)
Unperfused
After perfusion for 1 h
T
z
25
. -E
.-e
I n virro. from
anaesthetized rats
26.5 f 1.2 (8)
22.7 f 0.8 (6)
In uitro, from freshly killed rats I n vivo
24.1 f 0.7 (8) 25.6 f 2.3 (6)
21.9 f 0.7 (7) 24.0 f 2.4 (4)
O Ib) Leu-Gly
.
cytoplasm after perfusion for 1 h amounted to 80.5% of the activities in the mucosal cytoplasm of unperfused intestines (mean of all experiments). ( a )Val-Leu T
r
Discussion
These results extend earlier studies (Silk & Kim, 1975; Lindberg, 1972) which demonstrated that intestines in uitro can release peptide hydrolases into perfusion and incubation media. However, we find that the amounts appearing in the luminal effluent are normally very small, amounting over 1 h to only 0.6-1.9% of the initial cytoplasmic content (Fig. 1). The enzyme activities recovered in the luminal effluent during perfusion in viuo and in vitro were closely similar provided that the preparations in vitro were set up from anaesthetized animals. However, when the segments for perfusion were excised from freshly killed animals, the enzyme activities found in the luminal effluent were four to eight times greater than when the animals were maintained under ether anaesthesia while the intestines were beiig connected to the perfusion apparatus. Similarly, the activities found in the secretion on to the serosal surface of the intestine were small (0.345% of the initial cytoplasmic activity) when the intestines were set up from anaesthetized rats. These values also increased substantially when the animals were killed before removal of the intestine. Fig. 3 indicates that the release of enzymes into the luminal effluent was a continual process throughout the hour of perfusion rather than one which commenced after a certain time. Thus these data indicate that the release of hydrolases from the perfused intestine is normally small although it can be increased experimentally. Although this suggests that the release of pepti-
l0 o
i
Prepn. in virro from anaesthetized rats
Prepn. in virro from freshly killed rnts
Prepn. in viw
FIG.4. Cytoplasmic peptide hydrolase activities per mg
of DNA in small intestine before (open bars) and aRer (hatched bars) perfusion in uitro or In vivo for 1 h. Intestines for perfusion in uitro were prepared both from ether-anaesthetized rats and freshly killed rats. Substrates were (a) ~-Val-~-Leu, (b) L-L~u-G~Y or (c) Gly-L-Met. Data in parentheses show the activities aRer perfusion for 1 h expressed as a percentage of activities in unperfused intestines. Values are means f SEM, with the number of intestines shown within the bars in (a).
dases is not a normal physiological process, it must be noted that the amount of peptidase activity within the mucosa is so large that the release of 1% cannot necessarily be dismissed as functionally insignificant, especially in view of the possibility that hormones may stimulate this process (Giitze et al., 1972; Eloy et al., 1978). The high rates of peptidase release in vitro, which Lmdberg (1972) and Silk & Kim (1976) reported, probably had been influenced by the method of preparation. Hence their remarks about the interpretation of studies in vitro into peptide absorption, while valid for many studies in uitro, are greatly dependent on the particular technique used.
534
M . L. G . Gardner and J. A . Plumb
Although it is not known that similar results to ours would be obtained from all other types of intestinal preparation in vitro, or whether temporary anoxia specifically causes the enzyme release, our results indicate the merit of using anaesthesia and the precautions of Fisher & Parsons (1949) and Fisher & Gardrnr (1974) for the preparation of isolated intestines for perfusion, especially during studies on peptide absorption and hydrolysis or on peptide hormones. Furthermore, absorptive activity of isolated perfused intestine prepared from anaesthetized rats is substantially greater than if the animals are killed before removal of the intestine (Gardner, 1978). The relatively large activity against Gly-L-Met found in the luminal efAuent and serosal secretion (Fig. 1) reflects the relatively large amount of GlyMet hydrolase activity in the cytoplasm (Fig. 2); this is probably not due to differences in the stability of the three enzyme activities studied since we find Gly-Met hydrolase activity to be the most labile (unpublished observations). We do not know whether the enzymes released into the lumen are of cytoplasmic or brush-border origin: however, the work of Silk & Kim (1976) and Silk, Nicholson & Kim (1976) suggests that they may be predominantly cytoplasmic with some contribution by ileal brush border. We find that DNA is also lost (6-14% in 1 h) from the perfused intestines (Table 1). The small DNA loss from intestines perfused in vitro accords with the good absorptive stability of the preparation reported by Fisher & Gardner (1974). However, in the experiments in vitro at least, the losses of enzyme activity from the cytoplasm during perfusion were proportionately slightly greater than the losses of DNA (Fig. 4). Although this difference is not quite significant it was consistently observed in all three groups of experiments; hence it is probable that enzymes are lost from the mucosa by a combination of sloughing of whole cells (leading to the slight DNA loss) and a more specific process. This is consistent with Lindberg’s (1972) observation that disaccharidase activities were not released in vitro along with the dipeptidase activities and with the substantially larger specific activity of intraluminal Gly-Leu hydrolase seen in vitro than in vivo (Lindberg et al., 1975, p. 228). The specific component of the release obviously requires further investigation.
Acknowledgments
We are indebted to the Medical Research Council and to Eaton Laboratories Ltd for financial support. We also thank Mrs Anne Pryde for technical assistance and Dr R. C. Heading for valuable discussions, and Miss Helen Scott for secretarial assistance.
References VAULTIER, J.P., RAUL, F., MIRHOM, R., CLENDMNEN, G. & GRENIER, J.F. (1978) Hormonal stimulation of intestinal brush border enzymes release. Research in Experimental Medicine, 172, 109- 12 I. FISHER,R.B. & GARDNER, M.L.G. (1974)A kinetic approach to the study of absorption of solutes by isolated perfused small intestine. Journal of Physiology (London), 241, 21 1ELOY, R.,
234. FISHER, R.B. & PARSONS,D.S. (1949) A preparation of surviving rat small intestine for the study of absorption. Journal of Physiology (London), 110,3646. GARDNER, M.L.G. (1978) The absorptive viability of isolated intestine prepared from dead animals. Quarterly Journal of Experimental Physiologv. 63,93-95. GOTZE,H., ADELSON,J.W., HADORN,H.B., PORTMANN, R. & TROESCH, V. (1972)Hormoneelicited enzyme release by the small intestinal wall. Gut, 13,471-476. JOSEFFSON, L. & SJOSTROM,H. (1966)Intestinal dipeptidases. IV. Studies on the release and subcellular distribution of intestinal dipeptidases of the mucosa cells of the pig. Acla Physiologica Scandinavica, 67.27-33. KARSTEN,U. & WOLLENBERCER, A. (1972)Determination of DNA and RNA in homogenised cells and tissues by surface Buorimetry. Analytical Biochemistry, 46, 135-148. LE PECQ, J.-B. & PAOLETTI,C. (1966) A new fluonmetric method for RNA and DNA determination. Analytical Biochemistry, 17,loQ-107. LINDBERG, T. (1972) Discussion. In: Peptide Transport in Bacteria and Mammalian Gut (CIBA Foundation Symposium), p. 91. Ed. Elliott, K. & O’Connor, M. Associated Scientific Publishers, Amsterdam. LMDBERC, T., NOREN,0. & SJOSTROM,H. (1975)Peptidases of the intestinal mucosa. In: Peptide Transport in Protein Nutrition, pp. 204-242. Ed. Matthews, D.M. & Payne, J.W. North-Holland, Amsterdam. NICHOLSON, J.A. & KIM,Y.S.(1975)A one-step L-amino acid oxidase assay for intestinal peptide hydrolase activity. AnalyticalBiochemistry, 63, 110-1 17. PLUMB,J.A. & GARDNER,M.L.G. (1979) The physiological viability of isolated intestine prepared from dead animals. Gastroenterologie Clinique el Biolagique, 3, 176. SILK, D.B.A. & KIM, Y.S. (1975) A study of intraluminal peptide hydrolase activity in the rat. Clinical Science and Molecular Medicine, 49, 523-526. SILK,D.B.A. & KIM,Y.S.(1976)Release of peptide hydrolases during incubation of intact intestinal segments in vitro. Journal of Physiology (London), 258489497. SILK,D.B.A., NICHOLSON, J.A. & KIM, Y.S. (1976)Hydrolysis of peptides within lumen of small intestine. American Journal of Physiology, 231, 1322-1329. WICCANS,D.S. & JOHNSTON,J.M. (1959)The absorption of peptides. Biochimica et Biophysica Acta. 32,69-73.
