Acta Physiol Scund 1990, 139, 203-210
Villus and crypt electrolyte and fluid transport during intestinal secretion A. S J O Q V I S T and R. R E E U W K E S I11 Department of Pharmacology, Smith, Kline & French Laboratories, Philadelphia, PA, USA
SJOQVIST, A. & BEEUWKES 111, R, 1990. Villus and crypt electrolyte and fluid transport during intestinal secretion. Acta Ph,ysiol Scand 139, 203-210. Received 2 October 1989, accepted 18 December 1989. ISSN 0001-6772. Department of Pharmacology, Smith, Kline & French Laboratories, Philadelphia, PA, USA. The apical parts of jejunal villi of net-absorbing intestine have been shown to contain sodium chloride concentration gradients which are associated with water absorption (Sjoqvist & Beeuwkes 1989). T o determine whether these gradients are different in states of intestinal net secretion, jejunal segments of chloralose-anaesthetized cats were perfused with modified Krebs-Henseleit solutions while secretion was elicited by cholera toxin or vasoactive intestinal polypeptide (VIP). The segments were then rapidly frozen and freeze-dried, and sodium and chloride contents of the lamina propria of single villi were measured by X-ray microanalysis. The apical third of the villus was found to contain a concentration gradient of sodium and chloride when the lumen contained sodium, with no difference between secreting intestine and absorbing control intestine. When the intestine was perfused with hypotonic choline-mannitol solution, no sodium or chloride gradient was found. In this state, treatment of the intestines with secretagogues allowed development of an apical concentration gradient. .This demonstrated that the absorptive function of the villus tip was unimpaired during secretion and that secretion from the crypt could supply sufficient electrolyte to allow formation of an apical gradient. Key words: chloride, cholera toxin, electron microprobe, intestinal secretion, sodium, vasoactive intestinal polypeptide, X-ray microanalysis.
T h e intestine can transport fluid and electrolytes both from the lumen to tissue (absorption) and from the tissue to lumen (secretion). Net transport thus depends on the balance between absorptive and secretory rates. There is evidence that the absorptive and secretory processes are localized to the intestinal villi (Roggin et al. 1972, Hallback et al. 1979a, b) and crypts (Nasset & Ju 1973, Hallback et al. 1982, Welsh et al. 1982) respectively. T h e epithelial cells lining the mucosa originate from stem cells at the bottoms of the crypts. I n a process that takes less than a week the cells mature and migrate u p to the tip of the villus (Lipkin 1987). Since cells in crypt and tip differ in age by only a few days, one Correspondence : Anders Sjoqvist, Department of Physiology, University of Goteborg, Box 3303 I , S-400 33 Goteborg, Sweden.
might suppose that they have similar transport systems. However microvilli and glucocalyx develop as the cells move from crypt to villus, and certain transporting proteins and digestive enzymes are found only in villous cells. These differences probably explain the selective villus absorption of sodium, sugar and amino acids (King et al. 1981, Rowling & Sepulveda 1984, Madara & Trier 1987). Less is known about how secretagogues selectively induce secretion from crypt cells. T h e net fluid and electrolyte secretion characteristic of cholera diarrhoea is believed to be caused by increased secretion and not by decreased absorption (Hendrix & Paulk 1977). Indeed, the effectiveness of oral rehydration therapy indicates that absorption capacity is nearly unimpaired (Hirschhorn et al. 1968, Pierce et al. 1968). In the present study, we sought to
203
204
A . Sjogvist and R. Beeuwkes III
compare the mucosal distribution of sodium and chloride during net secretion with the pattern characteristic of absorption, in order to determine whether gradient-driven absorptive functions a t the villus tip could still be operative during net secretion. T w o different secretagogues, cholera toxin and vasoactive intestinal polypeptide (VIP), were used to determine whether the changes observed during secretion were of a general nature.
M A T E R I A L S A ND M E T H O D S Anaesthesia and operative procedure. Cats of both sexes were given free access to water but deprived of food for 16 h. Anaesthesia was induced by ketamine (10-20 mg kg-' im.) and maintained with achloralose (25-50 mg kg-' i.v.). The animals breathed spontaneously via a tracheostomy. The femoral vein and artery were cannulated to allow anaesthetic infusion and recording of arterial blood pressure. A glucose-bicarbonate buffer solution was infused (0.1 ml min-') during the experiments to maintain homeostasis (Haglund & Lundgren 1972). All animals were given atropine (0.5 mg kg-' i.v.) to suppress gut motility, and body temperature was kept at 38 "C with a heating pad. The abdomen was opened by midline incision and the spleen and greater omentum removed. The splanchnic nerves were divided preganglionically on both sides to minimize sympathetic nervous influence on fluid transport. A catheter was inserted into a proximal branch of the superior mesenteric artery to allow close intra-arterial administration of VIP to the intestine. The jejunum was divided into several I-15 cm long segments with intact vascular supply. The intestinal segments were perfused with modified Krebs-Henseleit electrolyte solutions. The perfusate was circulated through a circuit including lumen and a closed reservoir (0.7 1) and net fluid transport was measured with a volume transducer connected to the perfusion loop via a T-tube (Jodal et al. 1975). Net fluid secretion was induced by local intra-arterial infusion of VIP (30-60 pmol min-') or by preincubation of the intestinal mucosa with 20 pg cholera toxin. The VIP infusion rate used was chosen to induce a large secretion without significant effects on intestinal blood flow (Eklund et al. 1979). After a period of at least 30 min of steady-state fluid transport, a piece of the intestinal segment was rapidly extirpated (3-5 s) and frozen in vigorously stirred isopentane cooled by liquid nitrogen. Tissue preparation and electron microprobe analysis. The tissue was prepared for electron microprobe analysis by a method described in more detail elsewhere (Sjoqvist & Beeuwkes 1989). Briefly, the frozen intestine was cut into small pieces under dry ice
and freeze-dried at -40 "C in high vacuum. The tissue was then vacuum-embedded in electrolyte-free paraffin. The mucosa was cut into 5-pm-thick sections with an orientation along the villi and mounted on carbon discs, by an anhydrous technique. The contents of sodium, chloride and potassium in tissue compartments were analysed in a scanning electron microscope (Cameca, MBX). An electron beam (25 kV, zg nA) of size about 10x 50 pm was directed onto the specimen and characteristic X-rays generated by the elements were analysed with crystal spectrometers (Chandler 1977). The beam was moved in steps of 10 pm along the lamina propria from the apical part of the villus down to the mucosa between the crypts. A freeze-dried albumin standard with known electrolyte concentrations was studied under identical conditions to quantify the analysis. The spatial distribution of sodium was visualized by scanning the specimen with a highly focused electron beam. Characteristic X-rays generated by sodium during scanning triggered light flashes on the screen of a synchronously scanned cathode ray screen. Time photography of the screen provided X-ray fluorescence micrographs in which the image brightness was proportional to local elemental concentration. Solutions. The glucose-bicarbonate solution infused intravenously to maintain homeostasis contained IOO mM bicarbonate and 278 mM glucose. T w o different electrolyte solutions were used to perfuse the intestine. The first, an isotonic modified Krebs-Henseleit solution, contained 122 mM NaCI, 3.5 mM KC1, 1 . 2 mM KH,PO,, 25 mM NaHCO,, 1 . 2 mM MgCI,, 2.0 mM CaCI,, and 30 mM mannitol (Krebs-mannitol). In the second solution (choline-mannitol), NaHCO, was replaced by an equal concentration of choline-bicarbonate and sodium chloride was substituted with 72 mM choline chloride, making the solution hypotonic (about 200 mosmol) and sodium-free. Statzstics. The two-tailed Student t-test was used in the statistical evaluation of significance. Values are given as mean & SEM.
RESULTS Sodium-containing perfusate Cholera toxin produced a substantial net water secretion in jejunal segments perfused with isotonic Krebs-mannitol solution. T h r e e to four hours after exposing the intestine to the cholera toxin the secretion rate was 79 f32 pl min-' IOO cm-' serosal surface (Table I ) . The concentrations of sodium and chloride along the lamina propria of villi in secreting intestine are depicted in Figs I and 2 . T h e same figures include, for comparison, the villus tissue
Villus absorption and crypt secretion Table
I.
205
T h e effect of cholera toxin and VIP on intestinal fluid transport
Perfusate
Series
Blood pressure (mmHg)
Krebs-mannitol Krebs-mannitol
Control Cholera
148 k 1 0 104k3
Choline-mannitol Choline-mannitol Choline-mannitol
Control Cholera VIP
104k 10 100 k 3 106& 1 0
Fluid transport (,ul min-' IOO cm-')
Number of cats
5 6
212k52 -79k32" '35
*&
6 6
29
46* -166kz9" - 120
8
The intestine was perfused with Krebs-mannitol solution containing sodium and choline-mannitol solution in which choline replaced sodium. The choline-mannitol solution was hypotonic (zoo mosmol) while the Krebs-mannitol solution was isotonic (300 mosmol). The secretagogues elicited net fluid secretion in the small intestine (denoted by the negative fluid transport). Values are given as mean SEM. * Statistically significant secretion compared to conrrol ( P < 0.05).
*
1000{
700-
1
600-
--
500-
c
--$!2
400-
0
.-
300-
L
0 = 0
200100-
01
A. ,
0
,
,
20
,
I
40
,
,
60
,
I
80
I
I
100
!
,
120
Distance from villus tip (pet cent of villus length) Fig. I . Distribution of sodium along the small intestinal villus during nkt fluid secretion caused by and during control absorption cholera toxin (-) (----). The intestine was perfused with isotonic sodium-containing Krebs-mannitol solution in the lumen. The lamina propria of the villus was analysed from the tip down to the cryptal part of the mucosa in steps 1 0p m apart. The analysed area is schematically shown at the bottom of the figure. Specific X-rays from sodium (Ka line) were collected during 10-s intervals and plotted against the distance from the villus tip. The apical part of the villus contained significantly more sodium per unit volume of tissue than the basal part and there was no significant difference between secreting and absorbing intestine. The distance from the villus tip is given in per cent of villus length to account for varying length of the villi.
content of sodium and chloride in identically perfused absorbing intestine (redrawn from Sjoqvist & Beeuwkes 1989). In intestines exposed to cholera toxin, the concentration of sodium at villus tips was nearly twice that measured in the crypt region. T h e chloride concentration gradi-
0-
I
0
I
I
,
20
,
40
I
I
J
60
1
1
80
1
1
100
1
120
Distance from villus tip (pet cent of villus length) Fig. 2. Distribution of chloride along the small and intestinal villus during cholera secretion (-) control absorption (----). The lumen was perfused with isotonic Krebs-mannitol solution. There was a positive concentration gradient of chloride towards the tip of the villi in secreting and absorbing intestine. There was no significant difference in chloride concentration at the tip when comparing intestine during secretion and absorption.
ent along the villus was similar but the magnitude was not as large (Fig. 2 ) . In fact, there was no significant difference in sodium or chloride gradients along the villus between secreting and absorbing intestine (Table 2 ) T h e concentration of potassium in the apical region was higher in villi incubated with cholera toxin than in control villi.
Sodium-free perfusate Net fluid absorption occurred in jejunal segments perfused with hypotonic sodium-free
206
A . Sjogvist and R. Beeuwkes III
Table 2. The effect of cholera toxin and VIP on electrolyte concentration in the lamina propria of the intestinal mucosa
Control Isotonic Krebs-mannitol solutzon Sodium concentration (mmol I-') Villus tip Villus base Chloride concentration (mmol 1 I ) Villus tip Villus base Potassium concentration (mmol I-') Villus tip Villus base Number of villi Rypotonic sodium-fee solution Sodium concentration (mmol I-') Villus tip Villus base Chloride concentration (mmol I-') Villus tip Villus base Potassium concentration (mmol IF') Villus tip Villus base Number of villi
Cholera
I 23 f I 8***
VIP
-______
__
93 Ik 36**
52i6
53f6
108 & 1 3 ~ " " 61 f 6
82 f29** 56f6
Izt 1ozf9
84f9 93fIO
107
27
52
42f6 45 f 5 5of9 52f5 77 f9** 91 f 9 47
78 f7" 88f I 1 48
The concentration of sodium and chloride along the lamina propria of the villus was unaffected by the cholera secretion. When the perfusate was devoid of sodium, the secretagogues increased the concentration of sodium and chloride in the apical part of the villus in spite of a net secretion. The tip concentration is the mean concentration in the apical 2.5o/b of the villus. The crypt value is from the lamina propria at a distance of I O ~ I Z O : / ~of the distance from the villus tip. The concentration of sodium, chloride, and potassium was calculated using an albumin standard with known concentration of the elements and is given as concentration per volume tissue. Valyes are given as mean fSEM. t Significant difference from corresponding control (P < 0.001). * P < 0.05,** P < 0.01,*** P < 0.001,statistically significant difference between tip and base. choline-mannitol solution. T h i s absorption was reversed to secretion by both V I P and cholera toxin (Table I ) . VIP infused intra-arterially (30-60 pmol min-l) elicited net fluid secretion within a few minutes, in contrast to choleratoxin induced secretion, which took several hours to develop. There was no sodium gradient in villi from segments that showed a net fluid absorption due to the hypotonicity of the sodium-free perfusate. However, gradients were observed in villi from segments secreting water and electrolytes after either V I P or cholera toxin treatment. I n this state the concentration of sodium at the tip was more than twice that found in the crypt region of the mucosa (Fig. 3). T h e distribution of sodium in the villi can be visualized in an X-ray fluorescence micrograph
from an intestine in a secretory state due to VIPinfusion (Fig. 5 ) . T h e distribution of chloride in the mucosa was similar to that of sodium (Fig. 4). Chloride concentration was significantly higher at the tip than the base during fluid and electrolyte secretion caused by either V I P or cholera toxin (Table 2). There was a small, but statistically significant, negative gradient of potassium towards the tip of the villus in absorbing intestine and in intestine exposed to cholera toxin but not when VIP was used to elicit the secretion. DISCUSSION T h e present study clearly demonstrates that the 'absorptive gradient' of the villus remains
207
Villus absorption and crypt secretion 800
--
600
v)
8
3
400 0)
2
r 0
I
0
I
I
I
20
I
40
I
I
60
I
I
80
0 1
/
100
1
1
200
120
Distance from villus tip cent of villus length)
(per
Fig. 3. Distribution of sodium along the small intestinal villus during control absorption ( . . . . )and and during secretion caused by cholera toxin (-) VIP (---). The lumen contained hypotonic sodiumfree choline-mannitol solution. Sodium concentration was constant in villi exposed to a hypotonic sodiumfree solution since no sodium was present to be absorbed by the epithelium. The fluid absorption seen was due to the hypotonicity of the luminal solution. Note that secreting intestines had increased concentration of sodium apically in the villus.
0
1
0
1
1
20
1
1
40
1
1
60
1
1
80
1
1
1
100
1
120
Distance from villus tip (per cent of villus length) Fig. 4. Distribution of chloride along the small intestinal villus during control absorption ( . . . . ) and and during secretion caused by cholera toxin (-) VIP (---). The lumen contained hypotonic sodiumfree choline-mannitol solution. Intestines had an increased concentration of chloride apically in the villus during secretion but not during control absorption.
Fig. 5 . X-ray ‘fluorescence’ mappings of the relative distribution of villus sodium in intestine during secretion caused by VIP (left) and during absorption in intestine not treated with any secretagogue (right). The lumen was perfused with sodium-free choline-mannitol solution. The intestines were oriented with the villi pointing upwards. The density of the white dots corresponds to the sodium concentration of the tissue. The morphology of the mucosa can be readily visualized in the left map since there is some space between the villi. Note the increased sodium content in the apical region of the villi from the intestine in which sodium could be reabsorbed from the secretion caused by VIP, while the non-secreting intestine had more uniform distribution of sodium along its villi. The sodium-rich material outside the tissue, located in the vicinity of the villus tip, was probably freeze-dried mucus.
208
A . Sjoqvist and R. Beeuwkes 111 absorption
secretion
Fig. 6. Schematic illustration of the location of absorption and secretion in the small intestinal mucosa.
The villus epithelium absorbed sodium and chloride into the villus provided sodium was present in the lumen. The sodium was derived either from the luminal perfusate (absorption) or from secretion coming from the crypts (reabsorption). Secretagogues such as cholera toxin and VIP increased the crypt secretion and consequently also villus reabsorption. This was readily detected when lumen did not contain any other source of sodium since the sodium content of the apical villus increased as a response to the increased reabsorption of sodium. basically unchanged during net secretion caused by cholera toxin or VIP. The sodium gradient in the lamina propria of the villus was not statistically different in the intestine during net fluid secretion or n i t fluid absorption provided sodium was present in the luminal fluid. This accumulation of sodium in the apical third of the villus apparently reflects continuing active absorption of sodium in the outer part of the villus despite net intestinal loss of fluid and electrolytes. The constant concentration of sodium and chloride found along villi in segments perfused with hypotonic sodium-free solution indicated the lack of significant electrolyte absorption in the villus and showed that sodium is needed to absorb chloride. When two secretagogues were used to elicit net fluid and electrolyte secretion in the intestines perfused with the sodium-free hypotonic solution, electrolyte gradients reappeared. Since the hypotonicity of the perfusate prevented water secretion from the villus, the increased concentration of sodium at the villus tip must have been due to sodium absorption. Since the perfusate was sodium-free, the absorbed sodium must have originated from
crypt secretions, recycled in the villus as depicted in Fig. 6. A negative potassium gradient was seen in some experiments. It might be due to high volume absorption and accumulation of extracellular fluid at the villus tip, resulting in decreased potassium concentration per volume tissue as a consequence of a decreased fraction of intracellular volume. However, the finding was not consistent for reasons unknown to us. The observation that the absorptive capacity of the small intestine is largely unaffected during cholera secretion i s not new. However, the use of an advanced X-ray microanalytic technique enabled us to localize accurately an absorptive sodium chloride gradient in the villus tip. This observation thus materially extends the finding of increased osmolality in the apical part of the villus during net fluid and electrolyte absorption (Jodal et al. 1978, Hallback et al. 1979b) and also during net fluid secretion (Hallback et a/. 1979a, Hallback et al. 1982). It is interesting to notice that although both crypt and villous cells were exposed to VIP and cholera toxin, only the crypt cells responded with net secretion. The results from the present study also reemphasize that the villus and crypt parts of the mucosa have complementary transport functions. As suggested by Hendrix & Bayless (1970) these mechanisms seem to operate more or less independently of each other. During ‘normal’ function villus absorption exceeds crypt secretion, resulting in net absorption of fluid and electrolytes. However, the mucosa is exposed to stimulators of secretion during normal digestion (Sjovall et al. 1983, Miazza et al. 1985) and if villus (re)absorption is exceeded net fluid and electrolyte secretion will result. The physiological significance of this secretion might be to supply sodium to the villus, thus improving the absorption of water and of substances cotransported with sodium. Indeed, it has been calculated that the absorption of carbohydrates and protein in man demands more sodium than is delivered via the diet and secretion from other parts of the gastrointestinal tract (Halperin et al. 1986). Intestinal secretion is also turned on by bacterial toxins and harmful substances such as bile salt. This is probably part of a defence system washing the noxious substances away from the mucosa and thus controlling the microclimate. Rapidly multiplying cells, such as
Villus absorption and crypt secretion
209
HENDRIX, T.R. & BAYLESS, T.M. 1970. Digestion: intestinal secretion. A n n Rev Physiol32, 139-164. HENDRIX,T.R. & PAULK,H.T. 1977. Intestinal secretion. In: R. Crane (ed.) International Review of Physiology; vol. 1 2 Gastrointestinal Physiology I I , pp. 257-284. University Park Press, Baltimore. HIRSCHHORN, N., KINZIE, J.L., SACHAR,D.B., NORTHRUP, R.S., TAYLOR,J.O., AHMAD,Z. & PHILLIPS R.A. 1968. Decrease in net stool output in cholera during intestinal perfusion with glucosecontaining solutions. N E n g l f Med 279, 1766181. JODAL, M., HALLBACK, D.-A. & LUNDGREN, 0. 1978. Tissue osmolality in intestinal villi during luminal perfusion with isotonic electrolyte solutions. Acta Physiol Scand 102, 94-107. JODAL, M., HALLBACK, D.-A,, SVANVIK, J. & LUNDGREN, 0. 1975. A method for the continuous study of net water transport in the feline small bowel. Acta Physiol Scand 95, 4 1 - 4 4 7 . KING,I S . , SEPULVEDA, F.V. & SMITH,M.W. 1981. Cellular distribution of neutral and basic amino Expert photographic assistance was provided by acid transport systems in rabbit ileal mucosa. f Elwood Stack. This work was performed during Dr Physiol319, 355-368. Sjoqvist’s tenure as a post-doctoral fellow at Smith, LIPKIN,M. 1987. Proliferation and differentiation of Kline & French laboratories. normal and diseased gastrointestinal cells. In : L.R. Johnson. (ed.) Physiology of the Gastrointestinal Tract, 2nd edn, pp. 255-284. Raven Press, New York. REFERENCES LUNDGREN, O., SVANVIK, J. & JIVEGKRD, L. 1989. Enteric nervous system. I. Physiology and CHANDLER, J.A. 1977. X-ray Microanalysis in the pathophysiology of the intestinal tract. Dig Dis Sci Electron Microscope. North-Holland, Amsterdam. 34, 264-283. EKLUND, S., JODAL, M., LUNDGREN, 0. & SJOQVIST, A. J. & TRIER,J. 1987. Functional morphology 1979. Effects of vasoactive intestinal polypeptide on MADARA, of the mucosa of the small intestine. In: L.R. blood flow, motility and fluid transport in the Johnson (ed.) Physiology of the Gastrointestinal gastrointestinal tract of the cat. Acta Physiol Scand Tract, 2nd edn, pp. 1209-1250. Raven Press, New 105, 461-468. York. HAGLUND, U. & LUNDGREN, 0. 1972. Reactions B., PALMA, R., LACHANCE, J.R., CHAYVIALLE, within consecutive vascular sections of small MIAZZA, J.A., JONARD, P.P. & MODIGLIANI, R. 1985.Jejunal intestine of the cat during prolonged hypotension. secretory effect of intraduodenal food in humans. Acta Physiol Scand 84, 151-163. Gastroenteroloxy 88, 1 2 1 5-1222. HALLBACK, D.A., JODAL, M. & LUNDGREN, 0. 1979a. E.S. & Ju, J.S. 1973. Micropipet collection of Effects of cholera toxin on villous tissue osmolality NASSET, succus entericus at crypt ostia of guinea-pig and fluid and electrolyte transport in the small jejunum. Digestion 9, 205-21 I . intestine of the cat. Acta Physiol Scand 107, PIERCE, N.F., BANWELL,J.G., MITRA, R.C., 239-249. CARANASOS, G.J., KEIMOWITZ, R.I., MONDAL, A. & HALLBACK,D.A., JODAL, M., SJOQVIST,A. & MANJI, P.M. 1968. Effect of intragastric LUNDGREN, 0. 1979b. Villous tissue osmolality and glucose-lectrolyte infusion upon water and elecintestinal transport of water and electrolytes. Acta trolyte balance in Asiatic cholera. Gastroenterology Physiol Scand 107, 115-126. HALLBACK,D.A., JODAL, M., SJOQVIST,A. & 55, 333-343. LUNDGREN, 0. 1982. Evidence for cholera secretion ROGGIN,G.M., BANWELL, J.G., YARDLEY, J.H. & HENDRIX, T.R. 1972.Unimpaired response of rabbit emanating from the crypts. Gastroenterology 83, 1051~1056. jejunum to cholera toxin after selective damage to HALPERIN, M.L., WOLMAN, A.L. & GREENBERG, G.R. villus epithelium. Gastroentrology 63, 981-989. 1986. Paracellular recirculation of sodium is esROWLING,P. & SEPULVEDA, F.W. 1984. The dissential to support nutrient absorption in the tribution of (Na++ K+)-ATPase along the villus gastrointestinal tract: an hypothesis. Clin Invest crypt-axis in the rabbit small intestine. Biochim Biophys Acta 771, 35-41. Med 9 , 209-21 I
the stem cells in the bottom of the crypts, are more sensitive to disturbances in their environment. It is therefore very appropriate that the intestinal secretion originates in the crypts, making it difficult for dangerous substances to enter the crypt lumen against the secretion. Furthermore, the mucosa might have ‘an early warning system ’, since there is evidence that many potentially harmful secretagogues elicit secretion by activation of enteric nerve reflexes (Lundgren et al. 1989).T h i s enables detection of the substances at the villus level turning on the crypt secretion before they have reached down into the crypts. Secretory diarrhoea may be the result of an inappropriate stimulation of these physiological mechanisms if the reabsorptive capacity of the villi is exceeded.
210
A . Sjogvist and R. Beeuwkes 111
SJOQVIST, A. & BEEUWKES, R. 1989. Villous sodium gradient associated with volume absorption in the feline intestine : an electron-microprobe study on freeze-dried tissue. Acta Physiol Scand 136, 27 1-279. SJOVALL, H., REDFORS, S., JODAL, M. & LUNDGREN, 0. 1983. On the mode of action of the sympathetic
fibres on intestinal fluid transport: evidence for the existence of a glucose-stimulated secretory nervous pathway in the intestinal wall. Acta Physiol Scand 119, 39-48.
WELSH,M.J., SMITH, P., FROMM, M. & FRIZZELL, R. 1982. Crypts are the site of intestinal fluid and electrolyte secretion. Nature 218, 1219-1221.
Villus and crypt electrolyte and fluid transport during intestinal secretion A. S J O Q V I S T and R. R E E U W K E S I11 Department of Pharmacology, Smith, Kline & French Laboratories, Philadelphia, PA, USA
SJOQVIST, A. & BEEUWKES 111, R, 1990. Villus and crypt electrolyte and fluid transport during intestinal secretion. Acta Ph,ysiol Scand 139, 203-210. Received 2 October 1989, accepted 18 December 1989. ISSN 0001-6772. Department of Pharmacology, Smith, Kline & French Laboratories, Philadelphia, PA, USA. The apical parts of jejunal villi of net-absorbing intestine have been shown to contain sodium chloride concentration gradients which are associated with water absorption (Sjoqvist & Beeuwkes 1989). T o determine whether these gradients are different in states of intestinal net secretion, jejunal segments of chloralose-anaesthetized cats were perfused with modified Krebs-Henseleit solutions while secretion was elicited by cholera toxin or vasoactive intestinal polypeptide (VIP). The segments were then rapidly frozen and freeze-dried, and sodium and chloride contents of the lamina propria of single villi were measured by X-ray microanalysis. The apical third of the villus was found to contain a concentration gradient of sodium and chloride when the lumen contained sodium, with no difference between secreting intestine and absorbing control intestine. When the intestine was perfused with hypotonic choline-mannitol solution, no sodium or chloride gradient was found. In this state, treatment of the intestines with secretagogues allowed development of an apical concentration gradient. .This demonstrated that the absorptive function of the villus tip was unimpaired during secretion and that secretion from the crypt could supply sufficient electrolyte to allow formation of an apical gradient. Key words: chloride, cholera toxin, electron microprobe, intestinal secretion, sodium, vasoactive intestinal polypeptide, X-ray microanalysis.
T h e intestine can transport fluid and electrolytes both from the lumen to tissue (absorption) and from the tissue to lumen (secretion). Net transport thus depends on the balance between absorptive and secretory rates. There is evidence that the absorptive and secretory processes are localized to the intestinal villi (Roggin et al. 1972, Hallback et al. 1979a, b) and crypts (Nasset & Ju 1973, Hallback et al. 1982, Welsh et al. 1982) respectively. T h e epithelial cells lining the mucosa originate from stem cells at the bottoms of the crypts. I n a process that takes less than a week the cells mature and migrate u p to the tip of the villus (Lipkin 1987). Since cells in crypt and tip differ in age by only a few days, one Correspondence : Anders Sjoqvist, Department of Physiology, University of Goteborg, Box 3303 I , S-400 33 Goteborg, Sweden.
might suppose that they have similar transport systems. However microvilli and glucocalyx develop as the cells move from crypt to villus, and certain transporting proteins and digestive enzymes are found only in villous cells. These differences probably explain the selective villus absorption of sodium, sugar and amino acids (King et al. 1981, Rowling & Sepulveda 1984, Madara & Trier 1987). Less is known about how secretagogues selectively induce secretion from crypt cells. T h e net fluid and electrolyte secretion characteristic of cholera diarrhoea is believed to be caused by increased secretion and not by decreased absorption (Hendrix & Paulk 1977). Indeed, the effectiveness of oral rehydration therapy indicates that absorption capacity is nearly unimpaired (Hirschhorn et al. 1968, Pierce et al. 1968). In the present study, we sought to
203
204
A . Sjogvist and R. Beeuwkes III
compare the mucosal distribution of sodium and chloride during net secretion with the pattern characteristic of absorption, in order to determine whether gradient-driven absorptive functions a t the villus tip could still be operative during net secretion. T w o different secretagogues, cholera toxin and vasoactive intestinal polypeptide (VIP), were used to determine whether the changes observed during secretion were of a general nature.
M A T E R I A L S A ND M E T H O D S Anaesthesia and operative procedure. Cats of both sexes were given free access to water but deprived of food for 16 h. Anaesthesia was induced by ketamine (10-20 mg kg-' im.) and maintained with achloralose (25-50 mg kg-' i.v.). The animals breathed spontaneously via a tracheostomy. The femoral vein and artery were cannulated to allow anaesthetic infusion and recording of arterial blood pressure. A glucose-bicarbonate buffer solution was infused (0.1 ml min-') during the experiments to maintain homeostasis (Haglund & Lundgren 1972). All animals were given atropine (0.5 mg kg-' i.v.) to suppress gut motility, and body temperature was kept at 38 "C with a heating pad. The abdomen was opened by midline incision and the spleen and greater omentum removed. The splanchnic nerves were divided preganglionically on both sides to minimize sympathetic nervous influence on fluid transport. A catheter was inserted into a proximal branch of the superior mesenteric artery to allow close intra-arterial administration of VIP to the intestine. The jejunum was divided into several I-15 cm long segments with intact vascular supply. The intestinal segments were perfused with modified Krebs-Henseleit electrolyte solutions. The perfusate was circulated through a circuit including lumen and a closed reservoir (0.7 1) and net fluid transport was measured with a volume transducer connected to the perfusion loop via a T-tube (Jodal et al. 1975). Net fluid secretion was induced by local intra-arterial infusion of VIP (30-60 pmol min-') or by preincubation of the intestinal mucosa with 20 pg cholera toxin. The VIP infusion rate used was chosen to induce a large secretion without significant effects on intestinal blood flow (Eklund et al. 1979). After a period of at least 30 min of steady-state fluid transport, a piece of the intestinal segment was rapidly extirpated (3-5 s) and frozen in vigorously stirred isopentane cooled by liquid nitrogen. Tissue preparation and electron microprobe analysis. The tissue was prepared for electron microprobe analysis by a method described in more detail elsewhere (Sjoqvist & Beeuwkes 1989). Briefly, the frozen intestine was cut into small pieces under dry ice
and freeze-dried at -40 "C in high vacuum. The tissue was then vacuum-embedded in electrolyte-free paraffin. The mucosa was cut into 5-pm-thick sections with an orientation along the villi and mounted on carbon discs, by an anhydrous technique. The contents of sodium, chloride and potassium in tissue compartments were analysed in a scanning electron microscope (Cameca, MBX). An electron beam (25 kV, zg nA) of size about 10x 50 pm was directed onto the specimen and characteristic X-rays generated by the elements were analysed with crystal spectrometers (Chandler 1977). The beam was moved in steps of 10 pm along the lamina propria from the apical part of the villus down to the mucosa between the crypts. A freeze-dried albumin standard with known electrolyte concentrations was studied under identical conditions to quantify the analysis. The spatial distribution of sodium was visualized by scanning the specimen with a highly focused electron beam. Characteristic X-rays generated by sodium during scanning triggered light flashes on the screen of a synchronously scanned cathode ray screen. Time photography of the screen provided X-ray fluorescence micrographs in which the image brightness was proportional to local elemental concentration. Solutions. The glucose-bicarbonate solution infused intravenously to maintain homeostasis contained IOO mM bicarbonate and 278 mM glucose. T w o different electrolyte solutions were used to perfuse the intestine. The first, an isotonic modified Krebs-Henseleit solution, contained 122 mM NaCI, 3.5 mM KC1, 1 . 2 mM KH,PO,, 25 mM NaHCO,, 1 . 2 mM MgCI,, 2.0 mM CaCI,, and 30 mM mannitol (Krebs-mannitol). In the second solution (choline-mannitol), NaHCO, was replaced by an equal concentration of choline-bicarbonate and sodium chloride was substituted with 72 mM choline chloride, making the solution hypotonic (about 200 mosmol) and sodium-free. Statzstics. The two-tailed Student t-test was used in the statistical evaluation of significance. Values are given as mean & SEM.
RESULTS Sodium-containing perfusate Cholera toxin produced a substantial net water secretion in jejunal segments perfused with isotonic Krebs-mannitol solution. T h r e e to four hours after exposing the intestine to the cholera toxin the secretion rate was 79 f32 pl min-' IOO cm-' serosal surface (Table I ) . The concentrations of sodium and chloride along the lamina propria of villi in secreting intestine are depicted in Figs I and 2 . T h e same figures include, for comparison, the villus tissue
Villus absorption and crypt secretion Table
I.
205
T h e effect of cholera toxin and VIP on intestinal fluid transport
Perfusate
Series
Blood pressure (mmHg)
Krebs-mannitol Krebs-mannitol
Control Cholera
148 k 1 0 104k3
Choline-mannitol Choline-mannitol Choline-mannitol
Control Cholera VIP
104k 10 100 k 3 106& 1 0
Fluid transport (,ul min-' IOO cm-')
Number of cats
5 6
212k52 -79k32" '35
*&
6 6
29
46* -166kz9" - 120
8
The intestine was perfused with Krebs-mannitol solution containing sodium and choline-mannitol solution in which choline replaced sodium. The choline-mannitol solution was hypotonic (zoo mosmol) while the Krebs-mannitol solution was isotonic (300 mosmol). The secretagogues elicited net fluid secretion in the small intestine (denoted by the negative fluid transport). Values are given as mean SEM. * Statistically significant secretion compared to conrrol ( P < 0.05).
*
1000{
700-
1
600-
--
500-
c
--$!2
400-
0
.-
300-
L
0 = 0
200100-
01
A. ,
0
,
,
20
,
I
40
,
,
60
,
I
80
I
I
100
!
,
120
Distance from villus tip (pet cent of villus length) Fig. I . Distribution of sodium along the small intestinal villus during nkt fluid secretion caused by and during control absorption cholera toxin (-) (----). The intestine was perfused with isotonic sodium-containing Krebs-mannitol solution in the lumen. The lamina propria of the villus was analysed from the tip down to the cryptal part of the mucosa in steps 1 0p m apart. The analysed area is schematically shown at the bottom of the figure. Specific X-rays from sodium (Ka line) were collected during 10-s intervals and plotted against the distance from the villus tip. The apical part of the villus contained significantly more sodium per unit volume of tissue than the basal part and there was no significant difference between secreting and absorbing intestine. The distance from the villus tip is given in per cent of villus length to account for varying length of the villi.
content of sodium and chloride in identically perfused absorbing intestine (redrawn from Sjoqvist & Beeuwkes 1989). In intestines exposed to cholera toxin, the concentration of sodium at villus tips was nearly twice that measured in the crypt region. T h e chloride concentration gradi-
0-
I
0
I
I
,
20
,
40
I
I
J
60
1
1
80
1
1
100
1
120
Distance from villus tip (pet cent of villus length) Fig. 2. Distribution of chloride along the small and intestinal villus during cholera secretion (-) control absorption (----). The lumen was perfused with isotonic Krebs-mannitol solution. There was a positive concentration gradient of chloride towards the tip of the villi in secreting and absorbing intestine. There was no significant difference in chloride concentration at the tip when comparing intestine during secretion and absorption.
ent along the villus was similar but the magnitude was not as large (Fig. 2 ) . In fact, there was no significant difference in sodium or chloride gradients along the villus between secreting and absorbing intestine (Table 2 ) T h e concentration of potassium in the apical region was higher in villi incubated with cholera toxin than in control villi.
Sodium-free perfusate Net fluid absorption occurred in jejunal segments perfused with hypotonic sodium-free
206
A . Sjogvist and R. Beeuwkes III
Table 2. The effect of cholera toxin and VIP on electrolyte concentration in the lamina propria of the intestinal mucosa
Control Isotonic Krebs-mannitol solutzon Sodium concentration (mmol I-') Villus tip Villus base Chloride concentration (mmol 1 I ) Villus tip Villus base Potassium concentration (mmol I-') Villus tip Villus base Number of villi Rypotonic sodium-fee solution Sodium concentration (mmol I-') Villus tip Villus base Chloride concentration (mmol I-') Villus tip Villus base Potassium concentration (mmol IF') Villus tip Villus base Number of villi
Cholera
I 23 f I 8***
VIP
-______
__
93 Ik 36**
52i6
53f6
108 & 1 3 ~ " " 61 f 6
82 f29** 56f6
Izt 1ozf9
84f9 93fIO
107
27
52
42f6 45 f 5 5of9 52f5 77 f9** 91 f 9 47
78 f7" 88f I 1 48
The concentration of sodium and chloride along the lamina propria of the villus was unaffected by the cholera secretion. When the perfusate was devoid of sodium, the secretagogues increased the concentration of sodium and chloride in the apical part of the villus in spite of a net secretion. The tip concentration is the mean concentration in the apical 2.5o/b of the villus. The crypt value is from the lamina propria at a distance of I O ~ I Z O : / ~of the distance from the villus tip. The concentration of sodium, chloride, and potassium was calculated using an albumin standard with known concentration of the elements and is given as concentration per volume tissue. Valyes are given as mean fSEM. t Significant difference from corresponding control (P < 0.001). * P < 0.05,** P < 0.01,*** P < 0.001,statistically significant difference between tip and base. choline-mannitol solution. T h i s absorption was reversed to secretion by both V I P and cholera toxin (Table I ) . VIP infused intra-arterially (30-60 pmol min-l) elicited net fluid secretion within a few minutes, in contrast to choleratoxin induced secretion, which took several hours to develop. There was no sodium gradient in villi from segments that showed a net fluid absorption due to the hypotonicity of the sodium-free perfusate. However, gradients were observed in villi from segments secreting water and electrolytes after either V I P or cholera toxin treatment. I n this state the concentration of sodium at the tip was more than twice that found in the crypt region of the mucosa (Fig. 3). T h e distribution of sodium in the villi can be visualized in an X-ray fluorescence micrograph
from an intestine in a secretory state due to VIPinfusion (Fig. 5 ) . T h e distribution of chloride in the mucosa was similar to that of sodium (Fig. 4). Chloride concentration was significantly higher at the tip than the base during fluid and electrolyte secretion caused by either V I P or cholera toxin (Table 2). There was a small, but statistically significant, negative gradient of potassium towards the tip of the villus in absorbing intestine and in intestine exposed to cholera toxin but not when VIP was used to elicit the secretion. DISCUSSION T h e present study clearly demonstrates that the 'absorptive gradient' of the villus remains
207
Villus absorption and crypt secretion 800
--
600
v)
8
3
400 0)
2
r 0
I
0
I
I
I
20
I
40
I
I
60
I
I
80
0 1
/
100
1
1
200
120
Distance from villus tip cent of villus length)
(per
Fig. 3. Distribution of sodium along the small intestinal villus during control absorption ( . . . . )and and during secretion caused by cholera toxin (-) VIP (---). The lumen contained hypotonic sodiumfree choline-mannitol solution. Sodium concentration was constant in villi exposed to a hypotonic sodiumfree solution since no sodium was present to be absorbed by the epithelium. The fluid absorption seen was due to the hypotonicity of the luminal solution. Note that secreting intestines had increased concentration of sodium apically in the villus.
0
1
0
1
1
20
1
1
40
1
1
60
1
1
80
1
1
1
100
1
120
Distance from villus tip (per cent of villus length) Fig. 4. Distribution of chloride along the small intestinal villus during control absorption ( . . . . ) and and during secretion caused by cholera toxin (-) VIP (---). The lumen contained hypotonic sodiumfree choline-mannitol solution. Intestines had an increased concentration of chloride apically in the villus during secretion but not during control absorption.
Fig. 5 . X-ray ‘fluorescence’ mappings of the relative distribution of villus sodium in intestine during secretion caused by VIP (left) and during absorption in intestine not treated with any secretagogue (right). The lumen was perfused with sodium-free choline-mannitol solution. The intestines were oriented with the villi pointing upwards. The density of the white dots corresponds to the sodium concentration of the tissue. The morphology of the mucosa can be readily visualized in the left map since there is some space between the villi. Note the increased sodium content in the apical region of the villi from the intestine in which sodium could be reabsorbed from the secretion caused by VIP, while the non-secreting intestine had more uniform distribution of sodium along its villi. The sodium-rich material outside the tissue, located in the vicinity of the villus tip, was probably freeze-dried mucus.
208
A . Sjoqvist and R. Beeuwkes 111 absorption
secretion
Fig. 6. Schematic illustration of the location of absorption and secretion in the small intestinal mucosa.
The villus epithelium absorbed sodium and chloride into the villus provided sodium was present in the lumen. The sodium was derived either from the luminal perfusate (absorption) or from secretion coming from the crypts (reabsorption). Secretagogues such as cholera toxin and VIP increased the crypt secretion and consequently also villus reabsorption. This was readily detected when lumen did not contain any other source of sodium since the sodium content of the apical villus increased as a response to the increased reabsorption of sodium. basically unchanged during net secretion caused by cholera toxin or VIP. The sodium gradient in the lamina propria of the villus was not statistically different in the intestine during net fluid secretion or n i t fluid absorption provided sodium was present in the luminal fluid. This accumulation of sodium in the apical third of the villus apparently reflects continuing active absorption of sodium in the outer part of the villus despite net intestinal loss of fluid and electrolytes. The constant concentration of sodium and chloride found along villi in segments perfused with hypotonic sodium-free solution indicated the lack of significant electrolyte absorption in the villus and showed that sodium is needed to absorb chloride. When two secretagogues were used to elicit net fluid and electrolyte secretion in the intestines perfused with the sodium-free hypotonic solution, electrolyte gradients reappeared. Since the hypotonicity of the perfusate prevented water secretion from the villus, the increased concentration of sodium at the villus tip must have been due to sodium absorption. Since the perfusate was sodium-free, the absorbed sodium must have originated from
crypt secretions, recycled in the villus as depicted in Fig. 6. A negative potassium gradient was seen in some experiments. It might be due to high volume absorption and accumulation of extracellular fluid at the villus tip, resulting in decreased potassium concentration per volume tissue as a consequence of a decreased fraction of intracellular volume. However, the finding was not consistent for reasons unknown to us. The observation that the absorptive capacity of the small intestine is largely unaffected during cholera secretion i s not new. However, the use of an advanced X-ray microanalytic technique enabled us to localize accurately an absorptive sodium chloride gradient in the villus tip. This observation thus materially extends the finding of increased osmolality in the apical part of the villus during net fluid and electrolyte absorption (Jodal et al. 1978, Hallback et al. 1979b) and also during net fluid secretion (Hallback et a/. 1979a, Hallback et al. 1982). It is interesting to notice that although both crypt and villous cells were exposed to VIP and cholera toxin, only the crypt cells responded with net secretion. The results from the present study also reemphasize that the villus and crypt parts of the mucosa have complementary transport functions. As suggested by Hendrix & Bayless (1970) these mechanisms seem to operate more or less independently of each other. During ‘normal’ function villus absorption exceeds crypt secretion, resulting in net absorption of fluid and electrolytes. However, the mucosa is exposed to stimulators of secretion during normal digestion (Sjovall et al. 1983, Miazza et al. 1985) and if villus (re)absorption is exceeded net fluid and electrolyte secretion will result. The physiological significance of this secretion might be to supply sodium to the villus, thus improving the absorption of water and of substances cotransported with sodium. Indeed, it has been calculated that the absorption of carbohydrates and protein in man demands more sodium than is delivered via the diet and secretion from other parts of the gastrointestinal tract (Halperin et al. 1986). Intestinal secretion is also turned on by bacterial toxins and harmful substances such as bile salt. This is probably part of a defence system washing the noxious substances away from the mucosa and thus controlling the microclimate. Rapidly multiplying cells, such as
Villus absorption and crypt secretion
209
HENDRIX, T.R. & BAYLESS, T.M. 1970. Digestion: intestinal secretion. A n n Rev Physiol32, 139-164. HENDRIX,T.R. & PAULK,H.T. 1977. Intestinal secretion. In: R. Crane (ed.) International Review of Physiology; vol. 1 2 Gastrointestinal Physiology I I , pp. 257-284. University Park Press, Baltimore. HIRSCHHORN, N., KINZIE, J.L., SACHAR,D.B., NORTHRUP, R.S., TAYLOR,J.O., AHMAD,Z. & PHILLIPS R.A. 1968. Decrease in net stool output in cholera during intestinal perfusion with glucosecontaining solutions. N E n g l f Med 279, 1766181. JODAL, M., HALLBACK, D.-A. & LUNDGREN, 0. 1978. Tissue osmolality in intestinal villi during luminal perfusion with isotonic electrolyte solutions. Acta Physiol Scand 102, 94-107. JODAL, M., HALLBACK, D.-A,, SVANVIK, J. & LUNDGREN, 0. 1975. A method for the continuous study of net water transport in the feline small bowel. Acta Physiol Scand 95, 4 1 - 4 4 7 . KING,I S . , SEPULVEDA, F.V. & SMITH,M.W. 1981. Cellular distribution of neutral and basic amino Expert photographic assistance was provided by acid transport systems in rabbit ileal mucosa. f Elwood Stack. This work was performed during Dr Physiol319, 355-368. Sjoqvist’s tenure as a post-doctoral fellow at Smith, LIPKIN,M. 1987. Proliferation and differentiation of Kline & French laboratories. normal and diseased gastrointestinal cells. In : L.R. Johnson. (ed.) Physiology of the Gastrointestinal Tract, 2nd edn, pp. 255-284. Raven Press, New York. REFERENCES LUNDGREN, O., SVANVIK, J. & JIVEGKRD, L. 1989. Enteric nervous system. I. Physiology and CHANDLER, J.A. 1977. X-ray Microanalysis in the pathophysiology of the intestinal tract. Dig Dis Sci Electron Microscope. North-Holland, Amsterdam. 34, 264-283. EKLUND, S., JODAL, M., LUNDGREN, 0. & SJOQVIST, A. J. & TRIER,J. 1987. Functional morphology 1979. Effects of vasoactive intestinal polypeptide on MADARA, of the mucosa of the small intestine. In: L.R. blood flow, motility and fluid transport in the Johnson (ed.) Physiology of the Gastrointestinal gastrointestinal tract of the cat. Acta Physiol Scand Tract, 2nd edn, pp. 1209-1250. Raven Press, New 105, 461-468. York. HAGLUND, U. & LUNDGREN, 0. 1972. Reactions B., PALMA, R., LACHANCE, J.R., CHAYVIALLE, within consecutive vascular sections of small MIAZZA, J.A., JONARD, P.P. & MODIGLIANI, R. 1985.Jejunal intestine of the cat during prolonged hypotension. secretory effect of intraduodenal food in humans. Acta Physiol Scand 84, 151-163. Gastroenteroloxy 88, 1 2 1 5-1222. HALLBACK, D.A., JODAL, M. & LUNDGREN, 0. 1979a. E.S. & Ju, J.S. 1973. Micropipet collection of Effects of cholera toxin on villous tissue osmolality NASSET, succus entericus at crypt ostia of guinea-pig and fluid and electrolyte transport in the small jejunum. Digestion 9, 205-21 I . intestine of the cat. Acta Physiol Scand 107, PIERCE, N.F., BANWELL,J.G., MITRA, R.C., 239-249. CARANASOS, G.J., KEIMOWITZ, R.I., MONDAL, A. & HALLBACK,D.A., JODAL, M., SJOQVIST,A. & MANJI, P.M. 1968. Effect of intragastric LUNDGREN, 0. 1979b. Villous tissue osmolality and glucose-lectrolyte infusion upon water and elecintestinal transport of water and electrolytes. Acta trolyte balance in Asiatic cholera. Gastroenterology Physiol Scand 107, 115-126. HALLBACK,D.A., JODAL, M., SJOQVIST,A. & 55, 333-343. LUNDGREN, 0. 1982. Evidence for cholera secretion ROGGIN,G.M., BANWELL, J.G., YARDLEY, J.H. & HENDRIX, T.R. 1972.Unimpaired response of rabbit emanating from the crypts. Gastroenterology 83, 1051~1056. jejunum to cholera toxin after selective damage to HALPERIN, M.L., WOLMAN, A.L. & GREENBERG, G.R. villus epithelium. Gastroentrology 63, 981-989. 1986. Paracellular recirculation of sodium is esROWLING,P. & SEPULVEDA, F.W. 1984. The dissential to support nutrient absorption in the tribution of (Na++ K+)-ATPase along the villus gastrointestinal tract: an hypothesis. Clin Invest crypt-axis in the rabbit small intestine. Biochim Biophys Acta 771, 35-41. Med 9 , 209-21 I
the stem cells in the bottom of the crypts, are more sensitive to disturbances in their environment. It is therefore very appropriate that the intestinal secretion originates in the crypts, making it difficult for dangerous substances to enter the crypt lumen against the secretion. Furthermore, the mucosa might have ‘an early warning system ’, since there is evidence that many potentially harmful secretagogues elicit secretion by activation of enteric nerve reflexes (Lundgren et al. 1989).T h i s enables detection of the substances at the villus level turning on the crypt secretion before they have reached down into the crypts. Secretory diarrhoea may be the result of an inappropriate stimulation of these physiological mechanisms if the reabsorptive capacity of the villi is exceeded.
210
A . Sjogvist and R. Beeuwkes 111
SJOQVIST, A. & BEEUWKES, R. 1989. Villous sodium gradient associated with volume absorption in the feline intestine : an electron-microprobe study on freeze-dried tissue. Acta Physiol Scand 136, 27 1-279. SJOVALL, H., REDFORS, S., JODAL, M. & LUNDGREN, 0. 1983. On the mode of action of the sympathetic
fibres on intestinal fluid transport: evidence for the existence of a glucose-stimulated secretory nervous pathway in the intestinal wall. Acta Physiol Scand 119, 39-48.
WELSH,M.J., SMITH, P., FROMM, M. & FRIZZELL, R. 1982. Crypts are the site of intestinal fluid and electrolyte secretion. Nature 218, 1219-1221.
