Acta Physiol Scand 1990, 138, 75-84
Studies of cholera toxin-induced changes of alkaline secretion and transepithelial potential difference in the rat intestine in vivo M. H. T A N T I S I R A , L. F A N D R I K S , C. JONSSON, M. JODAL and 0. L U N D G R E N Department of Physiology, University of Gothenburg, Sweden TANTISIRA, M. H., FANDRIKS, L., JONSSON, C.,
JODAL, M. & LUNDGREN, 0. 1989. Studies of cholera toxin-induced changes of alkaline secretion and transepithelial potential difference in the rat intestine in vivo. Acta Physiol Scand 138, 75-84. Received 28 April 1989, accepted 26 June 1989. ISSN 000145772. Department of Physiology, University of Gothenburg, Sweden.
A pH-stat technique was used to investigate the effects of cholera toxin (CT) on alkaline secretion from denervated intestines (jejunum, ileum, colon) in anaesthetized rats. Transepithelial potential difference (PD) was also followed in some experiments. CT, given intraluminally, caused a marked increase in jejunal alkaline secretion, whereas only a small effect was observed in the ileum and no apparent effect was noted in the proximal colon. The pronounced increase in jejunal alkaline secretion was found to be inhibited by 10-25% by hexamethonium (10 mg kg-' body wt i.v.) and similarly by serosal application of lidocaine, whereas atropine (0.25 mg kg-' body wt i.v.) had no effect. Thus the cholera toxin-induced alkaline secretion in the jejunum is attributed mainly to a non-nervous mechanism. The small effect of C T on ileal alkaline secretion observed in this study contrasts with the high ileal bicarbonate concentration reported in cholera by authors who estimated the concentration from the total carbon dioxide/bicarbonate contents. This discrepancy may be explained by a CT-evoked increased transport of the coupled Na+/H+ and CI-/HCO; exchangers, which cannot be measured with the pHstat technique used in this study. Key words : cholera toxin, enteric nervous system, intestinal alkaline secretion.
Cholera toxin, produced by all pathogenic strains of Vibrio cholerae, causes secretory diarrhoea which is manifested clinically by a large-volume loss of a protein-free fluid. T h e cholera stool is an isotonic fluid with a bicarbonate concentration approximately twice and a potassium concentration 4-8 times that of normal plasma (Carpenter 1982). In humans, fluid and electrolyte transport has been studied in the jejunum, ileum and colon during acute cholera (Banwell et al. 1968, 1970, Speelman et al. 1986). A cholera toxin-induced bicarbonate secretion, measured as changes in
total CO,, was demonstrated in both the jejunum and the ileum. In these two sections of the tract, a reciprocal relationship was found between the concentrations of bicarbonate and chloride ions. T h e concentration of bicarbonate increased and that of chloride decreased as luminal fluid samples were taken lower in the small intestine (Banwell et al. 1968). In in-aitro studies, bicarbonate secretion, estimated as the residual flux in an Ussing chamber preparation, was not evoked from human ileal tissue exposed to cholera toxin. T h e major anion secreted in this preparation was chloride (Al-Awqati et al.
'973). Correspondence : Mayuree H. Tantisira, Department of Physiology, University of Goteborg, PO Box 33031, S-400 33 Goteborg, Sweden.
Furthermore, a discrepancy between the invivo and in-vitro effects of cholera toxin on
bicarbonate secretion has been demonstrated to
75
76
hl.H. Tantisira et al.
Anaesthesia was induced by an i.p. injection of exist in experimental animals. I n the in-vivo studies, bicarbonate secretion was measured in sodium pentobarbital (Mebumal Vet, 60 mg kg-' body wt). A trachael cannula was inserted to assure free terms of changes in the total carbon dioxide/ bicarbonate concentration in the intestinal fluid. airways and a femoral vein was cannulated for Measured in this way, the rabbit duodenum, administration of drugs as needed. Arterial blood pressure and heart rate were recorded by means of a jejunum and ileum (Ixitch & Burrows 1968, pressure transducer (Statham Pz3AC) via a catheter Love 1969, Norris et a[. 1969) as well as the in a femoral artery. This catheter was also used to canine jejunal and ileal loops (Carpenter et al. maintain anaesthesia with a continuous infusion of 1968,Moore et al. 1971)all exhibited bicarbonate chloralose (2-4 mg m1-l; 0.02 ml min-') provided in a secretion after being exposed to cholera toxin. solution containing I 10 mM glucose, 40 mM NaHCO, When the electrolyte secretion stimulated by and 50 mM NaCl to prevent dehydration and acidosis cholera toxin was studied in stripped rabbit during and after surgery. Body temperature was ileum under short-circuit current conditions, net maintained at 37 "C by a thermistor-controlled opfluxes of chloride into the intestinal lumen were eration table and lamp. In some animals, the acid-base status was deinduced, although no significant alteration was termined at the start and at the end of the experiment. observed in the residual flux (Field et al. 1972, Blood samples were collected in heparinized capillaries Powell et al. 1973, Sheerin & Field 1977). from the femoral artery for the determination of Several investigators have speculated on the plasma pH, and bicarbonate concentration by an difference betu-een the in-zico and in-wtro effects automatic acid-base analyser (ABL-jo, Radiometer, of cholera toxin on bicarbonate transport (Field Copenhagen, Denmark). 1974, Schultz et al. 1974, Powell 1987) but the Preparation of the intestine. The abdomen was question is still far from settled. This study was opened by a midline incision and the nerves surundertaken to investigate the in-cico effects of rounding the superior mesenteric artery were isolated chlolera toxin on bicarbonate secretion as studied and cut to eliminate the influence of extrinsic nerves. with a pH-stat technique which allows an Thereafter, an intestinal segment of 3-5 cm length uninterrupted measurement of luminal p H and a was isolated, cannulated in both ends and rinsed with continuous titration of the luminal alkalinization warm isotonic NaCl solution. The proximal ends of by HCI. Concomitantly, the transepithelial the jejunum, ileum and colon were defined as 5-8 cm membrane potential in the intestine was moni- distal to the ligament of Treitz, 8-10 cm proximal to the ileocaecal valve and 3-4 cm distal to the caecum tored during some experiments. respectively. We have previously proposed that intramural Measurement of alkaline secretion. The method used nervous reflexes are involved in cholera toxininduced secretion. T h i s hypothesis rests on the for recording intestinal alkaline secretion was a slight finding that cholera toxin-induced secretion in modification of the one described by Flemstrom et al. dcnervated intestinal segments of rats and cats is (1982). The intestinal segment was connected in a recirculating system to a titration chamber which niarkedlj- inhibited by various nerve blocking contained 10 ml isotonic NaCl solution, gassed with agents (Cassuto et a/. 1981, 1982, 1983). Thus, 100'; N, and maintained at 37 "C by a water jacket. the present study utilized denervated intestinal The perfusate was recirculated at a rate of 5 ml min-' segments from rats as the experimental model. by a roller pump (Ismatec, Switzerland) connected T h e work involves a study of the effects of nerve between the proximal cannula and the titration blocking agents in order to probe the possible chamber. The luminal pH was maintained at a role of the enteric nervous reflexes in cholera constant 7.4 by the addition of a NaCl solution containing either 0.05 M or 0.005 M HCI into the toxin-induced alkaline secretion. titration chamber. Titrant delivery was automatically monitored by a computer-controlled pump and a pH electrode (pH electrode GI( 2320 C, Radiometer, MATERIALS AND METHODS Copenhagen, Denmark) placed in the titration The experiments were performed on male Sprague chamber. Loss of fluid from the titration chamber due Dawley rats (Nab, Stockholm, Sweden) weighing to evaporation was compensated for by the continuous perfusion of an isotonic NaCl solution at a rate of 2.50- joo g. These animals were housed in the animal quarters under standardized environmental conditions approximately 0.09 ml min-'. The alkaline secretion, estimated from the rate of titration, was expressed as (22 "C, 55-60";, relative humidity, artificial lighting ~ surface area 06.00 to 18.00 h) for a t least 7 days prior to the pequiv. per min per ~ o o c r nserosal (pequiv. min-' I O O cm-"). euperiments.
Alkaline secretion in cholera Measurement of transepithelial electrical potential dzference (PD).A pair of 4% agar bridges in a balanced electrolyte solution (NaCI, 122mM; KCI, 4.7 RIM; NaHCO,, 25 mM; KH,PO,, 1.2 mM) were used, one agar bridge being inserted into the side arm of the titration chamber and the other being placed in the abdominal cavity. The potential difference (PD ; mV; lumen negative) was measured by a matched pair of calomel half-cells (calomel electrode K401, Radiometer, Copenhagen, Denmark) connected to the agar bridges through saturated KCI wells. All the measured variables (arterial blood pressure, heart rate, pH of the luminal perfusate and PD) were recorded continuously on a Grass polygraph. In addition, the computer (ABCSo, Luxor, Motala, Sweden) estimated the mean values of arterial pressure and heart rate during 5 min periods, as well as the amount of alkaline secreted by the intestinal segment. Histological procedures. At the end of the experiments, the animal was sacrificed by an intravenous bolus injection of a saturated KCI solution. The intestinal segment was removed and cut open so that its length and width could be measured. Intestinal tissue was taken from every third or fourth animal in each group for histological examinations. These tissues were fixed in 10% buffered formalin, embedded in paraffin, sectioned, stained with haematoxylin and eosin, and observed by light microscopy. Experimental protocol. Since alkaline secretion has been reported to be influenced by various factors, including handling and operative procedures (Flemstrom 1987), all preparations were done by the same investigator. After completion of the preparation, the intestinal segment was placed back in the abdomen, which was kept closed by metal clips. After the registration of basal alkaline secretion for 60 min, 60 pg of cholera toxin were added to the titration chamber. Atropine (0.25 mg kg-' body wt) and hexamethonium (10mg kg-' body wt) were administered via the femoral vein in some experiments, whereas in others a 1% lidocaine solution (5 mg 10cm-l intestine) was applied directly onto the serosa. In some experiments, concomitant changes in PD were also measured. Drugs and toxin. The following drugs were used: lidocaine (Astra AB, Sodertalje, Sweden), atropine sulphate and hexamethonium bromide (Sigma Chemicals, St Louis, MO, USA). Cholera toxin was purchased from Sigma Chemicals. The drugs and toxin were dissolved in isotonic NaCl solution.
Statistics. The differences in alkaline secretion between control and tested groups were evaluated for significance by an analysis of variance. Effects of drugs were analysed by the sign test. Differences resulting in probability values of 0.05 or less were considered significant. 5
77
RESULTS Basal alkaline secretion Basal alkaline secretion in the jejunum, the ileum and the proximal colon was continuously registered in three different groups of rats. Jejunum exhibited the lowest rate of basal alkaline secretion, whereas the highest rate was displayed by the ileum. During a 5-h observation period, basal alkaline secretion gradually increased in all segments (Figs. 1-3). Blood gas analyses were performed in several experiments as summarized in Table I. T h e blood samples were taken at the beginning and at the end of the experiments. A slight decrease in plasma bicarbonate concentration was noted in all experimental series. T h e rate of alkaline
h
2004 I
0
40
t
80
120
160
200
240
280
320
CT
Time ( m i d
Fig. I. Effects on alkaline secretion, heart rate and arterial pressure of intraluminal administration of 60 pg cholera toxin (CT) (0-0) to denervated jejunal segments. Control experiments (m-n) are illustrated as a comparison. Bars denote S S E . n = 7 in both groups. ACT 138
itl. H . Tantisira et nl.
27 N G
+I +I N r .
9:
- 5 N N
- N
secretion at the end of the experiments varied between the different groups. No consistent relationship between plasma bicarbonate concentration and alkaline secretion could be established.
-?;
Effects of cholera toxin on alkaline secretion
+I +I
As illustrated in Fig. I , the addition of 60 pg cholera toxin into the titration chamber caused a significant increase in alkaline secretion from the denervated jejunal segments ( n = 7). A time lag of 3-40 min after the addition of cholera toxin was observed for the alkaline secretion response, which thereafter progressively increased to reach a plateau within 3 h after the addition of cholera toxin. Washing the intestinal lumen after exposure to cholera toxin did not abolish the subsequent increase in alkaline secretion (data not shown). Cholera toxin caused a small but significant increase in ileal alkaline secretion in
lor?
-a
A$ 3
55 +I +I
o m ?p!
X G N N
m N
?-
- 0
+I +I m
Lc
$ ?
x
N
-
o m
zz 35
+I +I
m m x -
G' Ir. N N
sg
-r.
-r-3;
+I +I m o F, N N
:-1 E
5
m o
300
9 9
0 0
+I +I
2:
-4-G
m -
9 9 0 0 +I +I
xx 0 0
1
3
0
N
1
1
40
1
1
1
80
1
1
120
1
1
160
1
1
200
1
,
240
1
1
280
1
1
320
CT
Time ( m i d
on alkaline secretion, heart rate and arterial pressure of intraluminal administration of 60 pg cholera toxin (CT) (0-0) to denervated ileal are illussegments. Control experiments (0-0) trated as a comparison. Bars denote +SE. n = 7 in both groups. Fig.
0
E L
2. Effects
79
Alkaline secretion in cholera
-
--m
400
m
n
300
200
I
Q Y
o i 0
I
I
40
1
I
80
I
I
120
I
, 160
, ,
I
200
I
240
I
, 280
8
,
320
CT
Time (mid
Fig. 3. Effects on alkaline secretion, heart rate and arterial pressure of intraluminal administration of 60 pg cholera toxin (CT) (0-0) to denervated proximal colon segments. Control experiments (0-0) are illustrated as a comparison. Bars denote fSE. n = 5 in both groups. comparison to its corresponding control group (n = 7) (Fig. 2). However, cholera toxin failed to cause any significant change in alkaline secretion from the denervated proximal colon ( n = 5 ; Fig. 3).
Pharmacological analyses of the cholera toxininduced alkaline secretion in the denervated jejunum Three different types of nerve blocking agents, i.e. atropine (a muscarinic cholinergic receptor blocker), hexamethonium (a nicotinic cholinergic receptor blocker) and lidocaine (a local anaesthetic), were used to probe the possible involvement of the intramural nervous reflex in the alkaline response elicited in the jejunum by cholera toxin. The measurements made I 5-30 min after the administration of drug were used when evaluating the action of the drug.
0
40
1 CT
80
120
160 I 2 0 0
240
I
280
320
Atrocme Heramethenium
Time (mid
Fig. 4. Effects of intravenous administration of atropine and hexamethonium on alkaline secretion, heart rate and arterial pressure in denervated jejunal segments exposed to 60 ,ug cholera toxin (CT). Administrations of CT and drugs are indicated by arrows. Bars denote +SE. n = 7 in both ‘no drugs’ (0-0) and ‘drug’ (0-0) groups.
Atropine. As shown in Table 3 and Fig. 4, atropine had no significant effect on either cholera toxin-induced alkaline secretion or PD. In control experiments atropine significantly decreased the basal alkaline secretion and PD in both jejunum and ileum (Table 2). Hexamethonium. The increase in jejunal alkaline secretion caused by cholera toxin was partly inhibited by hexamethonium ( 1 0 mg kg-’ body wt i.v.) (Fig. 4). In experiments where changes of P D were concomitantly measured, hexamethonium decreased the increase in P D caused by the toxin (Table 3). I n control experiments hexamethonium significantly decreased basal alkaline secretion and PD in both jejunum and ileum (Table 2 ) . 5-2
M. H . Tantisiru et al.
80
T a b l e 2. Effects of atropine (0.25 mg kg-' body wrt) and hexamethonium alkaline secretion (..\S) and transepithelial potential difference (PD)
(10 mg
kg-' body wt) on basal
Ileumf
Jejunumt ~~
Ifter
Re fore
Before ~~
~
After
~~~
itropine
4s (pequi\ min
' I O O cm
I .06 f0 . 2 p
25.35 I:' 4 5 (5)
z1.56f 1.53"
')
4.00 I:0.44*
PD (mV)
5.25
fo.83 (5)
4.5 fo.85"
Hexamethonium
21.96f1.86'
AS (jtequit min-' PD (mV)
IOO cm-')
3.75f0.50"
Values are given as mean I:SE. Vumber of observations is given in parenthesis. * Denotes significant difference by sign test. t ..\tropine and hexamethonium were given intravenously at the end of the 3 h and 4 h respectively. 1 rZtropine and hesamethonium were given intravenously at the end of the z h and 3 h respectively.
Tablc 3 . Effects of drugs on cholera toxin-induced alkaline secretion (AS) and transepithelial potential difference (PD) on denervated jejunum Initial value
After
'3 6
0.9gf0.17 5.06 f0.08
5.37 f O . S Y 7.08 f0.40
'3 6
0.99f0.17 j.06f0.08
5.13f0.56 5.88 fo.61"
iS (jtequik. min
6
0.76 L0.13
PI) (m\ )
I
j.00
5.78+ r . r y * 5.50
I>rug\
n
4tropine ( 0 zj mg Lg IS (jtequn min I PI) (In\ )
I
bod\
100 cm
at)
')
I fexdmethonrum (10 mg kg ' bod! AS (pequn. min ' IOO cm ') PD ( n ~ T) 1,idocdine
(I
'lo
solution) ' ioo cm ')
\\
t)
I'alues represents mean value within a certain period of time f SE. Initial \slue = mean value of basal alkaline secretion and potential difference during the first hour. Before = 0-15 min before, and after = 15-30 min after the administration of drugs. n = number of observations. * denotes significant difference by sign test.
Lzdocaine. Serosal application of lidocaine produced effects similar to those exerted by hexamethonium (Table 3). After application of lidocaine onto the intestinal serosa, the cholera toxin-induced alkaline secretion was partly inhibited in all animals, whereas no effect was seen a hen an isotonic NaCl solution was given in the 5ame manner (Fig. 5).
Histological obseraations Light microscopy revealed no evidence of histological damage caused by the perfusion per se o r by the cholera toxin. The integrity of the epithelial cells was well maintained in jejunum, ileum and proximal colon after 5 h perfusion with isotonic NaCl solution, with or without cholera toxin.
Alkaline secretion in cholera
80
60
2004 I
'0
1
T
CT
S
Lidocsine
Time (mid
Fig. 5. Effects of serosal administration of lidocaine and physiological saline solution(s) on alkaline se-
cretion, heart rate and arterial pressure in denervated jejunal segments exposed to 60 pg cholera toxin (CT). Administrations of CT, saline and drug are indicated by arrows. Bars denote fSE. n = 7 and 6 in the 'no drugs' (0-0) and 'drug' (0-0) groups re-
spectively.
DISCUSSION
Basal intestinal alkaline secretion A spontaneous alkalinization was observed in all three parts of the intestinal tract studied in the present report. This may reflect a transport of H+ from the lumen to surrounding tissues or a secretion of HCO, and/or OH- into the lumen. Titration to the end-point 7.4 of the intestinal contents precludes any large lumen-to-tissue gradient and, hence, any passive diffusion of protons. Furthermore, there are a number of
81
observations which suggest that bicarbonate is secreted in the ileum and the colon of several mammalian species including the rat (Parsons 1956, Wilson & Kazyak 1957, Edmond 1967, Hubel 1967, Phillips & Schmalz 1970, Podesta & Mettrick 1977a, Flemstrom et al. 1985). A carbon dioxide tension considerably higher than in plasma or tissue is found in the intestinal lumen (see, for example, Swallow & Code 1967). These observations suggest that the method used in this study indirectly records the net secretion of bicarbonate ions in the small intestine, although a secretion of OH- also may be present (Liedtke & Hopfer 1982). As the method used in the present study does not differentiate between HCO; and OH-, the term alkaline secretion is used. Several mechanisms of intestinal HCO, secretion have been proposed by Flemstrom (1987) : ( I ) electroneutral CI-/HCO; exchange; (2) electrogenic transcellular HCO; secretion stimulated by CAMP; and (3) passive paracellular diffusion. A Cl-/HCO; exchange has been demonstrated in the mammalian gut (Hubel 1967, 1969, Tai & Decker 1980) and found to be partly coupled to a Na+/H+ exchanger in the rat ileum and colon (Binder & Rawlins 1973, Podesta & Mettrick 1977, Lubcke et al. 1986). As one would expect if H+ and HCO; are secreted into the lumen simultaneously, no change in p H occurs. Therefore, such alkaline secretion cannot be detected by a pH-stat technique. Thus, this study provides information on the alkaline secretion which is not coupled to Hf secretion. As reported earlier, we found a higher rate of alkaline secretion in the ileum and the proximal colon (Parsons 1956, Hubel 1968). Whether the observed alkaline secretion is due to a normally operating mechanism of HCO, transport or merely a passive diffusion of HCO, from blood to lumen cannot be ascertained from the present study. However, changes of plasma bicarbonate concentration did not seem to influence alkaline secretion, suggesting that passive HCO; diffusion played a minor role (Table I). There is, however, experimental evidence also for an intestinal absorption of bicarbonate ions. The rat jejunum has been reported to absorb HCO, from the intestinal lumen in vivo by a mechanism that is partly Na+-dependent (Hubel 1973, Podesta & Mettrick 1977 b). The observed decrease of luminal p H could either reflect a
82
< H. % Tantiszra I, et al.
where the alkaline secretion arising from the coupling of Cl-/HCO, and Na+/H+ is not detected, cholera toxin produced only a small increase in ileal net alkaline secretion. This suggests that the high ileal COJbicarbonate content seen in cholera in animals and man zn r t r o resulted mainly from increased rates of transport in the coupled CI-/HCO; and Na+/H+ exchanges. It should be pointed out, however, that it is generally assumed from En-vztro studies that cholera toxin inhibits the coupled NaCl transport (Powell 1987). No attempt was made to probe the involvement of enteric nervous reflex in the observed ileal response for two reasons. Firstly, the increase in ileal alkaline secretion was of a small magnitude when compared either to its own basal secretion or to the jejunal alkaline secretion evoked by the enterotoxin. Secondly, atropine and hexamethonium were found to inhibit to a 6Jkl.1 ~ f r h o l e r atoxin on intestinal alkaline greater extent ileal basal alkaline secretion than s r u e t i o n trnd potentiul difference the ileal alkaline secretion evoked by cholera In the present study, intraluminal treatment toxin (Table 2, Fig. 2). with cholera toxin caused a marked increase in Jejunum. Jejunum is another site of alkaline jejunal alkaline secretion, whereas only a small loss in cholera diarrhoea. Studies in humans effect was observed in the ileum and apparently (Banwell et al. 1968, 1970) and experimental no effect was present in the proximal colon. animals (Carpenter et al. 1968) have demonSimilar results have been found in man, bi- strated that fluid as well as solute losses were carbonate concentration being significantly greater in the jejunum than in the ileum during higher in the jejunal and ileal contents during acute cholera diarrhoea. Furthermore, bicaracute cholera infection than during convalescence bonate was secreted into the jejunum, which (Ranu-ell et ill. 1968, 1970). I n the colon, on the normally absorbed this ion. T h e present results other hand, bicarbonate concentration was simi- are in line with those previous In-vzzw ohserlar during and after cholera (Speelman et al. \ ations in whichalkalinesecretion wasmeasuredas I ($36). Mou-ever, the profile of alkaline secretion total COJbicarbonate ions (Banwell et al. 1968, along the rat intestine in the present study was 1970, Carpenter et al. 1968, Love 1969) or as an different from the one observed in man. In increase in mucosal surface p H (McEwan et al. human cholera, the ileal bicarbonate concen- 1988). tration was significantly higher than the jejunal In the rat jejunum N a + / H + is unlikely to be one, the latter being approximately the same as coupled to a Cl-/OH- or a CI -/HCO, exchange that of plasma (Banwell et al. 1970). (Cassano et al. 1984). T h e Na+/H+ exchange is Ileum. In animal studies, in-riro cholera toxin partly responsible for sodium absorption, which causes a significant increase in ileal bicarbonate results in luminal acidification (Turnberg et al. secretion. In these studies bicarbonate transport 1970, Huhel 1973, Podesta & Mettrick ry77b). was measured as changes in the total Consequently, an inhibition of the Na+/H+ CO,/bicarbonate content in the lumen (Car- exchange may explain the increase in net alkaline penter et al. 1968, Leitch & Burrows 1968, secretion observed in the jejunum exposed to Norris el al. 1969, Moore r t al. 1971). However, cholera toxin. However, an increase in alkaline in z'itro, the absence of a significant effect of secretion into the lumen cannot be ruled out. In cholera toxin on the so-called residual flux has ritro rat jejunal crypts have been reported to been reported (Field et ul. 1972, Powell et al. hace an alkaline p H while jejunal villi exhibit an 1973), indicating that no electrogenic bicarbonate acid pH. T h e alkalinity of the crypt region was secretion occurred in z'itro. In the present study proposed to be due to an alkaline secretion by
N a + / H * exchange mechanism or a co-transport of sodium and hicarbonate. Since the decreased p H was accompanied by a rise in luminal PCO, (Turnberg et aI. 1970, Powell et a / . 1971,Hubel 1973, Podesta & Mettrick 1977b), it was concluded that a N a ' / H ' exchange was the major mechanism for luminal acidification. I n the present study, jejunum exhibited a net alkaline secretion when it was perfused by a IICOi-free saline solution. Similar results have been reported in previous studies using HCOifree luminal solutions (Powell et 01. 1971, Hubel 1973, Podesta 8; llettrick 1977 b). Since the pHstat method used in the present study measures net changes in pH, it is not possible to estimate to uhat extent a simultaneous proton secretion might be present in the response observed in the jejunum.
Alkaline secretion in cholera the crypt cells (Daniel et al. 1985). T h e present results d o not provide direct evidence in favour of or against any of the mentioned mechanisms. Cholera toxin-induced fluid secretion has been proposed to be mediated partly by an intramural nervous reflex (Cassuto et al. 1981, 1982, 1983). I n the present study the effects of various nerve blocking agents were tested on the cholera toxininduced jejunal alkaline secretion and, in some experiments, also on the toxin effect on PD. Atropine, fornierly reported to have no effect on cholera toxin-induced fluid secretion, showed no significant effect on either stimulated jejunal alkaline secretion or PD. T h u s cholinergic muscarinic mechanisms d o not appear to be involved in the cholera toxin responses observed. Hexamethonium and serosally applied lidocaine exhibited a small but significant inhibitory effect on the cholera toxin-induced alkaline secretion in the jejunum (Table 3, Figs. 4 and 5). T h e effect of hexamethonium on PD seemed to be comparatively greater than the effect on alkaline secretion, although hexamethonium also significantly decreased the jejunal basal alkaline secretion and PD (Table 2 ) . T h e latter observation makes it difficult to estimate the true importance of intramural nervous reflex in cholera toxin-induced alkaline secretion. I t can be roughly estimated that the maximal inhibition of toxin-induced jejunal alkaline secretion by hexamethonium and lidocaine is 1-25 yo,while their effect on fluid transport is at least 60% (Cassuto et al. 1982). Thus, the present study suggests that a non-neural mechanism mainly underlies the cholera toxin-induced jejunal alkaline secretion. T h e increase in PD produced by cholera toxin may be explained by an active electrogenic ion transport and/or by a fluid transport producing a streaming potential. The discrepancy between the effects of nerve blocking agents on toxininduced changes in PD and fluid transport on the one hand and alkaline secretion on the other indicates that the influence of intramural nervous reflexes on different secretory mechanisms may vary. T h e results of this study suggest that the enteric nerves mainly control an electrogenic non-alkaline secretion. This research was supported by grants from the Swedish Medical Research Council (2855), The Medical Council of Swedish Life Insurance Companies and the Faculty of Medicine, University of
83
Gothenburg. M. H. Tantisira was supported by a scholarship from the Chulalongkorn University, Bangkok, Thailand.
REFERENCES AL-AWQATI, Q., CAMERON, J.L. & GREENOUGH 111, W.B. 1973. Electrolyte transport in human ileum: effect of purified cholera exotoxin. A m 3 Physiol 224, 818-823. BANWELL, J.G., PIERCE, N.F., MITRA,R., CARANASOS, G.J., KEIMOWITZ, R.I., MONDAL,A. & MANJI, P.M. 1968. Preliminary results of a study of small intestinal water and solute movement in acute and convalescent human cholera. Indian 3Me d Res 56, 633-639. BANWELL, J.G., PIERCE, N.F., MITRA,R.C., BRIGHAM, G.J., KEIMOWITZ, R.I., FEDSON, K.L., CARANASOS, D.S., THOMAS, J., GORBACH, S.L., SACK,R.B. & MONDAL,A. 1970. Intestinal fluid and electrolyte transport in human cholera. 3 Clin Invest 49, 183-195. BINDER,H.J. & RAWLINS, C.L. 1973. Electrolyte transport across isolated large intestine mucosa. Am 3 Physiol225, 1232-1239. CARPENTER, C.C.J. 1982. The pathophysiology of secretory diarrhoeas. Med Clin 66, 597-610. CARPENTER, C.C.J., SACK,R.B., FEELEY, J.C. & STEENBERG, R.W. 1968. Site and characteristics of electrolyte loss and effect of intraluminal glucose in experimental canine cholera. 3 Clin Invest 47, 12IcFI220
CASSANO, G., STIEGER, B. & MURER,H. 1984. Na/H and CI/OH-exchange in rat jejunal and rat proximal tubular brush border membrane vesicles. Studies with acridine orange. Pfugers Arch 400, 39-317. CASSUTO, J., JODAL, M., TUTTLE,R. & LUNDGREN, 0. 1981. On the role of intramural nerves in the pathogenesis of cholera toxin-induced intestinal secretion. Scand 3 Gastroenterol 16, 377-384. CASSUTO, J., JODAL, M. & LUNDGREN, 0. 1982. The effect of nicotinic and muscarinic receptor blockade on cholera toxin induced intestinal secretion in rats and cats. Acta Physiol Scand 114, 573-599. CASSUTO, J., SIEWERT, A., JODAL, M. & LUNDGREN, 0. 1983. The involvement of intramural nerves in cholera toxin-induced intestinal secretion. Acta Physiol Scand 117, 195-202. DANIEL, H., NEUGEBAUER, B., KRATZ,A. & REHNER, G. 1985. Localization of acid microclimate along intestinal villi of rat jejunum. A m 3 Physiol 248, G293-Gz98. EDMOND,C.J. 1967. Transport of sodium and secretion of potassium and bicarbonate by the colon of normal and sodium-depleted rats. 3Physiol 193, 589402. FIELD, M. 1974. Intestinal secretion. Gastroenterology 66, 1063-1084.
84
M. H . Tantisira et al.
FIELD,M., FROMM,D., AL-AWQATI,Q & GREENOUGHT 111, W.B. 1972.Effect of cholera enterotoxin on ion transport across isolated ileal mucosa. 3 Clin Invest 51, 796-804. FLEMSTROM, G. 1987.Gastric and duodenal mucosal bicarbonate secretion. In : L.R. Johnson (ed.) Physiologv of the Gastrointestinal Tract, 2nd edn, pp. I O I 1-1029.Raven Press, New York. FLEMSTROM, G., GARNER, A., NYLANDER, O., HURST, B.C. & HEYLINGS, J.R. 1982. Surface epithelial HCO; transport by mammalian duodenum in vivo. Am 3'Physiol 243, G348-G358. FLEMSTROM, G., JEDSTEDT, G. & KIVILAAKSO, E. 1985. Bicarbonate secretion by colon and duodenum in the rat in vivo. .4cta Ph,ysiol Scand 124 (Suppl. 542),
360.
Modification of intestinal secretion in experimentalcholera. 3 Infect Dis 119, 117-125. PARSONS, D.S. 1956. The absorption of bicarbonatesaline solutions by the small intestine and colon of the white rat. Q3 Exp Physiol 41, 410-420. PHILLIPS,S.F.& SCHMALZ, P.F. 1979.Bicarbonate secretion by the rat colon: effect of intraluminal chloride and acetazolamide. Proc Soc Exp Biol Med 135, 116-122. PODESTA, R.B. & METTRICK, D.F. 1977a.HCO; and H' secretion in rat ileum in vivo. Am 3 Physiol232,
E574-E579. PODESTA,R.B. & METTRICK,D.F. 1977b. HCO, transport in rat jejunum : relationship to NaCl and H,O transport in vivo. Am 3Physiol232, E62-E68. POWELL,D.W.1987.Intestinal water and electrolyte transport. I n : L.R. Johnson (ed.) Physiology ojthe Gastrointestinal Tract, 2nd edn, pp. 1267-1305, Raven Press, New York. POWELL, D.W., BINDER, H.J. & CURRAN, P.F. 1973. Active electrolyte secretion stimulated by choleragen in rabbit ileum in vitro. Am 3 Physiol 225,
HCBEL,K..k 1967. Bicarbonate secretion in the rat ileum and its dependence on intraluminal chloride. Am 3 Physiol 213, 1409-1413. IiUBEL, K.A. 1968.The ins and outs of bicarbonate in the alimentary tract. Gastroenterology 54, 64745I . HUBEL,K.A. 1969.Effect of luminal chloride con781-787. centration on bicarbonate secretion in rat ileum. POWELL,D.W., SOLBERG,L.I., PLOTKIN,G.R., .4m 3 Physiol 217,40-45. S.B. CATLIN,D.H., MAENZA, R.M. & FORMAL, HUBEL,K.A. 1973.Effect of luminal sodium con1971. Experimental diarrhea. 111. Bicarbonate centration on bicarbonate absorption in rat jejunum. transport in rat salmonella enterocolitis. Gastro7 Clin Inrest 52, 3172-3179. enterology 60, 10761086. L ~ I T C HG.J. , & BURROWS, W. 1968.Experimental R.A. & NELLANS,H.N. cholera in the rabbit ligated intestine: ion and water SCHULTZ,S.G.,FRIZZELL, 1974.Ion transport by mammalian small intestine. accumulation in the duodenum, ileum and colon. 3 Ann Rev Physiol36, 51-91. Infect Dis 118, 349-359. H.E. & FIELD, M. 1977.Ileal mucosal cyclic LIEDTKE, C.M.& HOPFER,U. 1982.Mechanism of SHEERIN, AMP and CI secretion : serosal vs. mucosal addition C 1 ~translocation across small intestinal brushof cholera toxin. Am 3 Physiol 232, E210-Ez15. border membrane. 11. Demonstration of P., BUTLER,T., KABIR,I., ALI, A. & C1 -OH-exchange and CI- conductance. Am 3 SPEELMAN, BANWELL, J. 1986. Colonic dysfunction during Ph,ysiol 242, G272-Gz80. cholera infection. Gastroenterology 91, I 164-1 170. LOVE,A.H.G. 1969.Water and sodium absorption by SWALLOW, J.H. & CODE, C.F. 1967. Intestinal the intestine in cholera. Gut 10. 6347. transmucosal fluxes of bicarbonate. Am 3 Physiol LCBCKE,R., HAAG,K., BERGER, E., KNAUF,H. & 212, 717-723. GEROK,W. 1986. Ion transport in rat proximal TAI,Y. & DECKER,R.A. 1980. Mechanisms of colon in vivo. Am 3 Physiol 251, G 1 3 2 4 1 3 9 . MCEWAN,G.T.A., DANIEL, H., FETT,C., BURGESS, electrolyte transport in rat ileum. Am 3Physiol238, Gzo8-Gz I 2. M.N. & LVCAS, M.L. 1988.The effect of Escherichia L.A., FORDTRAN, J.S., CARTER, N.W. & roli STa entrotoxin and other secretagogues on TURNBERG, RECTOR JR, F.C. 1970.Mechanism of bicarbonate mucosal surface pH of rat small intestine in vivo. absorption and its relationship to sodium transport Proc R SOCLand 234, 219-237. in human jejunum. 3 Clin Invest 49, 548-556. MOOREJR, W.L., BIEBERDORF, F.A., MORAWSKI, S.G., L. 1957.Acid-base changes FINKELSTEIN, R.A. & FORDSTRAN, J.S. 1971.Ion WILSON,T.H. & KAZYAK, across the wall of hamster and rat intestine. Biochim transport during cholera-induced ileal secretion Biophys Acta 24, 124-132. in the dog. 3 Clin Invest 50, 312-318. NORRIS,H.T., CURRAN, P.F. & SCHULTZ, S.G. 1969.
Studies of cholera toxin-induced changes of alkaline secretion and transepithelial potential difference in the rat intestine in vivo M. H. T A N T I S I R A , L. F A N D R I K S , C. JONSSON, M. JODAL and 0. L U N D G R E N Department of Physiology, University of Gothenburg, Sweden TANTISIRA, M. H., FANDRIKS, L., JONSSON, C.,
JODAL, M. & LUNDGREN, 0. 1989. Studies of cholera toxin-induced changes of alkaline secretion and transepithelial potential difference in the rat intestine in vivo. Acta Physiol Scand 138, 75-84. Received 28 April 1989, accepted 26 June 1989. ISSN 000145772. Department of Physiology, University of Gothenburg, Sweden.
A pH-stat technique was used to investigate the effects of cholera toxin (CT) on alkaline secretion from denervated intestines (jejunum, ileum, colon) in anaesthetized rats. Transepithelial potential difference (PD) was also followed in some experiments. CT, given intraluminally, caused a marked increase in jejunal alkaline secretion, whereas only a small effect was observed in the ileum and no apparent effect was noted in the proximal colon. The pronounced increase in jejunal alkaline secretion was found to be inhibited by 10-25% by hexamethonium (10 mg kg-' body wt i.v.) and similarly by serosal application of lidocaine, whereas atropine (0.25 mg kg-' body wt i.v.) had no effect. Thus the cholera toxin-induced alkaline secretion in the jejunum is attributed mainly to a non-nervous mechanism. The small effect of C T on ileal alkaline secretion observed in this study contrasts with the high ileal bicarbonate concentration reported in cholera by authors who estimated the concentration from the total carbon dioxide/bicarbonate contents. This discrepancy may be explained by a CT-evoked increased transport of the coupled Na+/H+ and CI-/HCO; exchangers, which cannot be measured with the pHstat technique used in this study. Key words : cholera toxin, enteric nervous system, intestinal alkaline secretion.
Cholera toxin, produced by all pathogenic strains of Vibrio cholerae, causes secretory diarrhoea which is manifested clinically by a large-volume loss of a protein-free fluid. T h e cholera stool is an isotonic fluid with a bicarbonate concentration approximately twice and a potassium concentration 4-8 times that of normal plasma (Carpenter 1982). In humans, fluid and electrolyte transport has been studied in the jejunum, ileum and colon during acute cholera (Banwell et al. 1968, 1970, Speelman et al. 1986). A cholera toxin-induced bicarbonate secretion, measured as changes in
total CO,, was demonstrated in both the jejunum and the ileum. In these two sections of the tract, a reciprocal relationship was found between the concentrations of bicarbonate and chloride ions. T h e concentration of bicarbonate increased and that of chloride decreased as luminal fluid samples were taken lower in the small intestine (Banwell et al. 1968). In in-aitro studies, bicarbonate secretion, estimated as the residual flux in an Ussing chamber preparation, was not evoked from human ileal tissue exposed to cholera toxin. T h e major anion secreted in this preparation was chloride (Al-Awqati et al.
'973). Correspondence : Mayuree H. Tantisira, Department of Physiology, University of Goteborg, PO Box 33031, S-400 33 Goteborg, Sweden.
Furthermore, a discrepancy between the invivo and in-vitro effects of cholera toxin on
bicarbonate secretion has been demonstrated to
75
76
hl.H. Tantisira et al.
Anaesthesia was induced by an i.p. injection of exist in experimental animals. I n the in-vivo studies, bicarbonate secretion was measured in sodium pentobarbital (Mebumal Vet, 60 mg kg-' body wt). A trachael cannula was inserted to assure free terms of changes in the total carbon dioxide/ bicarbonate concentration in the intestinal fluid. airways and a femoral vein was cannulated for Measured in this way, the rabbit duodenum, administration of drugs as needed. Arterial blood pressure and heart rate were recorded by means of a jejunum and ileum (Ixitch & Burrows 1968, pressure transducer (Statham Pz3AC) via a catheter Love 1969, Norris et a[. 1969) as well as the in a femoral artery. This catheter was also used to canine jejunal and ileal loops (Carpenter et al. maintain anaesthesia with a continuous infusion of 1968,Moore et al. 1971)all exhibited bicarbonate chloralose (2-4 mg m1-l; 0.02 ml min-') provided in a secretion after being exposed to cholera toxin. solution containing I 10 mM glucose, 40 mM NaHCO, When the electrolyte secretion stimulated by and 50 mM NaCl to prevent dehydration and acidosis cholera toxin was studied in stripped rabbit during and after surgery. Body temperature was ileum under short-circuit current conditions, net maintained at 37 "C by a thermistor-controlled opfluxes of chloride into the intestinal lumen were eration table and lamp. In some animals, the acid-base status was deinduced, although no significant alteration was termined at the start and at the end of the experiment. observed in the residual flux (Field et al. 1972, Blood samples were collected in heparinized capillaries Powell et al. 1973, Sheerin & Field 1977). from the femoral artery for the determination of Several investigators have speculated on the plasma pH, and bicarbonate concentration by an difference betu-een the in-zico and in-wtro effects automatic acid-base analyser (ABL-jo, Radiometer, of cholera toxin on bicarbonate transport (Field Copenhagen, Denmark). 1974, Schultz et al. 1974, Powell 1987) but the Preparation of the intestine. The abdomen was question is still far from settled. This study was opened by a midline incision and the nerves surundertaken to investigate the in-cico effects of rounding the superior mesenteric artery were isolated chlolera toxin on bicarbonate secretion as studied and cut to eliminate the influence of extrinsic nerves. with a pH-stat technique which allows an Thereafter, an intestinal segment of 3-5 cm length uninterrupted measurement of luminal p H and a was isolated, cannulated in both ends and rinsed with continuous titration of the luminal alkalinization warm isotonic NaCl solution. The proximal ends of by HCI. Concomitantly, the transepithelial the jejunum, ileum and colon were defined as 5-8 cm membrane potential in the intestine was moni- distal to the ligament of Treitz, 8-10 cm proximal to the ileocaecal valve and 3-4 cm distal to the caecum tored during some experiments. respectively. We have previously proposed that intramural Measurement of alkaline secretion. The method used nervous reflexes are involved in cholera toxininduced secretion. T h i s hypothesis rests on the for recording intestinal alkaline secretion was a slight finding that cholera toxin-induced secretion in modification of the one described by Flemstrom et al. dcnervated intestinal segments of rats and cats is (1982). The intestinal segment was connected in a recirculating system to a titration chamber which niarkedlj- inhibited by various nerve blocking contained 10 ml isotonic NaCl solution, gassed with agents (Cassuto et a/. 1981, 1982, 1983). Thus, 100'; N, and maintained at 37 "C by a water jacket. the present study utilized denervated intestinal The perfusate was recirculated at a rate of 5 ml min-' segments from rats as the experimental model. by a roller pump (Ismatec, Switzerland) connected T h e work involves a study of the effects of nerve between the proximal cannula and the titration blocking agents in order to probe the possible chamber. The luminal pH was maintained at a role of the enteric nervous reflexes in cholera constant 7.4 by the addition of a NaCl solution containing either 0.05 M or 0.005 M HCI into the toxin-induced alkaline secretion. titration chamber. Titrant delivery was automatically monitored by a computer-controlled pump and a pH electrode (pH electrode GI( 2320 C, Radiometer, MATERIALS AND METHODS Copenhagen, Denmark) placed in the titration The experiments were performed on male Sprague chamber. Loss of fluid from the titration chamber due Dawley rats (Nab, Stockholm, Sweden) weighing to evaporation was compensated for by the continuous perfusion of an isotonic NaCl solution at a rate of 2.50- joo g. These animals were housed in the animal quarters under standardized environmental conditions approximately 0.09 ml min-'. The alkaline secretion, estimated from the rate of titration, was expressed as (22 "C, 55-60";, relative humidity, artificial lighting ~ surface area 06.00 to 18.00 h) for a t least 7 days prior to the pequiv. per min per ~ o o c r nserosal (pequiv. min-' I O O cm-"). euperiments.
Alkaline secretion in cholera Measurement of transepithelial electrical potential dzference (PD).A pair of 4% agar bridges in a balanced electrolyte solution (NaCI, 122mM; KCI, 4.7 RIM; NaHCO,, 25 mM; KH,PO,, 1.2 mM) were used, one agar bridge being inserted into the side arm of the titration chamber and the other being placed in the abdominal cavity. The potential difference (PD ; mV; lumen negative) was measured by a matched pair of calomel half-cells (calomel electrode K401, Radiometer, Copenhagen, Denmark) connected to the agar bridges through saturated KCI wells. All the measured variables (arterial blood pressure, heart rate, pH of the luminal perfusate and PD) were recorded continuously on a Grass polygraph. In addition, the computer (ABCSo, Luxor, Motala, Sweden) estimated the mean values of arterial pressure and heart rate during 5 min periods, as well as the amount of alkaline secreted by the intestinal segment. Histological procedures. At the end of the experiments, the animal was sacrificed by an intravenous bolus injection of a saturated KCI solution. The intestinal segment was removed and cut open so that its length and width could be measured. Intestinal tissue was taken from every third or fourth animal in each group for histological examinations. These tissues were fixed in 10% buffered formalin, embedded in paraffin, sectioned, stained with haematoxylin and eosin, and observed by light microscopy. Experimental protocol. Since alkaline secretion has been reported to be influenced by various factors, including handling and operative procedures (Flemstrom 1987), all preparations were done by the same investigator. After completion of the preparation, the intestinal segment was placed back in the abdomen, which was kept closed by metal clips. After the registration of basal alkaline secretion for 60 min, 60 pg of cholera toxin were added to the titration chamber. Atropine (0.25 mg kg-' body wt) and hexamethonium (10mg kg-' body wt) were administered via the femoral vein in some experiments, whereas in others a 1% lidocaine solution (5 mg 10cm-l intestine) was applied directly onto the serosa. In some experiments, concomitant changes in PD were also measured. Drugs and toxin. The following drugs were used: lidocaine (Astra AB, Sodertalje, Sweden), atropine sulphate and hexamethonium bromide (Sigma Chemicals, St Louis, MO, USA). Cholera toxin was purchased from Sigma Chemicals. The drugs and toxin were dissolved in isotonic NaCl solution.
Statistics. The differences in alkaline secretion between control and tested groups were evaluated for significance by an analysis of variance. Effects of drugs were analysed by the sign test. Differences resulting in probability values of 0.05 or less were considered significant. 5
77
RESULTS Basal alkaline secretion Basal alkaline secretion in the jejunum, the ileum and the proximal colon was continuously registered in three different groups of rats. Jejunum exhibited the lowest rate of basal alkaline secretion, whereas the highest rate was displayed by the ileum. During a 5-h observation period, basal alkaline secretion gradually increased in all segments (Figs. 1-3). Blood gas analyses were performed in several experiments as summarized in Table I. T h e blood samples were taken at the beginning and at the end of the experiments. A slight decrease in plasma bicarbonate concentration was noted in all experimental series. T h e rate of alkaline
h
2004 I
0
40
t
80
120
160
200
240
280
320
CT
Time ( m i d
Fig. I. Effects on alkaline secretion, heart rate and arterial pressure of intraluminal administration of 60 pg cholera toxin (CT) (0-0) to denervated jejunal segments. Control experiments (m-n) are illustrated as a comparison. Bars denote S S E . n = 7 in both groups. ACT 138
itl. H . Tantisira et nl.
27 N G
+I +I N r .
9:
- 5 N N
- N
secretion at the end of the experiments varied between the different groups. No consistent relationship between plasma bicarbonate concentration and alkaline secretion could be established.
-?;
Effects of cholera toxin on alkaline secretion
+I +I
As illustrated in Fig. I , the addition of 60 pg cholera toxin into the titration chamber caused a significant increase in alkaline secretion from the denervated jejunal segments ( n = 7). A time lag of 3-40 min after the addition of cholera toxin was observed for the alkaline secretion response, which thereafter progressively increased to reach a plateau within 3 h after the addition of cholera toxin. Washing the intestinal lumen after exposure to cholera toxin did not abolish the subsequent increase in alkaline secretion (data not shown). Cholera toxin caused a small but significant increase in ileal alkaline secretion in
lor?
-a
A$ 3
55 +I +I
o m ?p!
X G N N
m N
?-
- 0
+I +I m
Lc
$ ?
x
N
-
o m
zz 35
+I +I
m m x -
G' Ir. N N
sg
-r.
-r-3;
+I +I m o F, N N
:-1 E
5
m o
300
9 9
0 0
+I +I
2:
-4-G
m -
9 9 0 0 +I +I
xx 0 0
1
3
0
N
1
1
40
1
1
1
80
1
1
120
1
1
160
1
1
200
1
,
240
1
1
280
1
1
320
CT
Time ( m i d
on alkaline secretion, heart rate and arterial pressure of intraluminal administration of 60 pg cholera toxin (CT) (0-0) to denervated ileal are illussegments. Control experiments (0-0) trated as a comparison. Bars denote +SE. n = 7 in both groups. Fig.
0
E L
2. Effects
79
Alkaline secretion in cholera
-
--m
400
m
n
300
200
I
Q Y
o i 0
I
I
40
1
I
80
I
I
120
I
, 160
, ,
I
200
I
240
I
, 280
8
,
320
CT
Time (mid
Fig. 3. Effects on alkaline secretion, heart rate and arterial pressure of intraluminal administration of 60 pg cholera toxin (CT) (0-0) to denervated proximal colon segments. Control experiments (0-0) are illustrated as a comparison. Bars denote fSE. n = 5 in both groups. comparison to its corresponding control group (n = 7) (Fig. 2). However, cholera toxin failed to cause any significant change in alkaline secretion from the denervated proximal colon ( n = 5 ; Fig. 3).
Pharmacological analyses of the cholera toxininduced alkaline secretion in the denervated jejunum Three different types of nerve blocking agents, i.e. atropine (a muscarinic cholinergic receptor blocker), hexamethonium (a nicotinic cholinergic receptor blocker) and lidocaine (a local anaesthetic), were used to probe the possible involvement of the intramural nervous reflex in the alkaline response elicited in the jejunum by cholera toxin. The measurements made I 5-30 min after the administration of drug were used when evaluating the action of the drug.
0
40
1 CT
80
120
160 I 2 0 0
240
I
280
320
Atrocme Heramethenium
Time (mid
Fig. 4. Effects of intravenous administration of atropine and hexamethonium on alkaline secretion, heart rate and arterial pressure in denervated jejunal segments exposed to 60 ,ug cholera toxin (CT). Administrations of CT and drugs are indicated by arrows. Bars denote +SE. n = 7 in both ‘no drugs’ (0-0) and ‘drug’ (0-0) groups.
Atropine. As shown in Table 3 and Fig. 4, atropine had no significant effect on either cholera toxin-induced alkaline secretion or PD. In control experiments atropine significantly decreased the basal alkaline secretion and PD in both jejunum and ileum (Table 2). Hexamethonium. The increase in jejunal alkaline secretion caused by cholera toxin was partly inhibited by hexamethonium ( 1 0 mg kg-’ body wt i.v.) (Fig. 4). In experiments where changes of P D were concomitantly measured, hexamethonium decreased the increase in P D caused by the toxin (Table 3). I n control experiments hexamethonium significantly decreased basal alkaline secretion and PD in both jejunum and ileum (Table 2 ) . 5-2
M. H . Tantisiru et al.
80
T a b l e 2. Effects of atropine (0.25 mg kg-' body wrt) and hexamethonium alkaline secretion (..\S) and transepithelial potential difference (PD)
(10 mg
kg-' body wt) on basal
Ileumf
Jejunumt ~~
Ifter
Re fore
Before ~~
~
After
~~~
itropine
4s (pequi\ min
' I O O cm
I .06 f0 . 2 p
25.35 I:' 4 5 (5)
z1.56f 1.53"
')
4.00 I:0.44*
PD (mV)
5.25
fo.83 (5)
4.5 fo.85"
Hexamethonium
21.96f1.86'
AS (jtequit min-' PD (mV)
IOO cm-')
3.75f0.50"
Values are given as mean I:SE. Vumber of observations is given in parenthesis. * Denotes significant difference by sign test. t ..\tropine and hexamethonium were given intravenously at the end of the 3 h and 4 h respectively. 1 rZtropine and hesamethonium were given intravenously at the end of the z h and 3 h respectively.
Tablc 3 . Effects of drugs on cholera toxin-induced alkaline secretion (AS) and transepithelial potential difference (PD) on denervated jejunum Initial value
After
'3 6
0.9gf0.17 5.06 f0.08
5.37 f O . S Y 7.08 f0.40
'3 6
0.99f0.17 j.06f0.08
5.13f0.56 5.88 fo.61"
iS (jtequik. min
6
0.76 L0.13
PI) (m\ )
I
j.00
5.78+ r . r y * 5.50
I>rug\
n
4tropine ( 0 zj mg Lg IS (jtequn min I PI) (In\ )
I
bod\
100 cm
at)
')
I fexdmethonrum (10 mg kg ' bod! AS (pequn. min ' IOO cm ') PD ( n ~ T) 1,idocdine
(I
'lo
solution) ' ioo cm ')
\\
t)
I'alues represents mean value within a certain period of time f SE. Initial \slue = mean value of basal alkaline secretion and potential difference during the first hour. Before = 0-15 min before, and after = 15-30 min after the administration of drugs. n = number of observations. * denotes significant difference by sign test.
Lzdocaine. Serosal application of lidocaine produced effects similar to those exerted by hexamethonium (Table 3). After application of lidocaine onto the intestinal serosa, the cholera toxin-induced alkaline secretion was partly inhibited in all animals, whereas no effect was seen a hen an isotonic NaCl solution was given in the 5ame manner (Fig. 5).
Histological obseraations Light microscopy revealed no evidence of histological damage caused by the perfusion per se o r by the cholera toxin. The integrity of the epithelial cells was well maintained in jejunum, ileum and proximal colon after 5 h perfusion with isotonic NaCl solution, with or without cholera toxin.
Alkaline secretion in cholera
80
60
2004 I
'0
1
T
CT
S
Lidocsine
Time (mid
Fig. 5. Effects of serosal administration of lidocaine and physiological saline solution(s) on alkaline se-
cretion, heart rate and arterial pressure in denervated jejunal segments exposed to 60 pg cholera toxin (CT). Administrations of CT, saline and drug are indicated by arrows. Bars denote fSE. n = 7 and 6 in the 'no drugs' (0-0) and 'drug' (0-0) groups re-
spectively.
DISCUSSION
Basal intestinal alkaline secretion A spontaneous alkalinization was observed in all three parts of the intestinal tract studied in the present report. This may reflect a transport of H+ from the lumen to surrounding tissues or a secretion of HCO, and/or OH- into the lumen. Titration to the end-point 7.4 of the intestinal contents precludes any large lumen-to-tissue gradient and, hence, any passive diffusion of protons. Furthermore, there are a number of
81
observations which suggest that bicarbonate is secreted in the ileum and the colon of several mammalian species including the rat (Parsons 1956, Wilson & Kazyak 1957, Edmond 1967, Hubel 1967, Phillips & Schmalz 1970, Podesta & Mettrick 1977a, Flemstrom et al. 1985). A carbon dioxide tension considerably higher than in plasma or tissue is found in the intestinal lumen (see, for example, Swallow & Code 1967). These observations suggest that the method used in this study indirectly records the net secretion of bicarbonate ions in the small intestine, although a secretion of OH- also may be present (Liedtke & Hopfer 1982). As the method used in the present study does not differentiate between HCO; and OH-, the term alkaline secretion is used. Several mechanisms of intestinal HCO, secretion have been proposed by Flemstrom (1987) : ( I ) electroneutral CI-/HCO; exchange; (2) electrogenic transcellular HCO; secretion stimulated by CAMP; and (3) passive paracellular diffusion. A Cl-/HCO; exchange has been demonstrated in the mammalian gut (Hubel 1967, 1969, Tai & Decker 1980) and found to be partly coupled to a Na+/H+ exchanger in the rat ileum and colon (Binder & Rawlins 1973, Podesta & Mettrick 1977, Lubcke et al. 1986). As one would expect if H+ and HCO; are secreted into the lumen simultaneously, no change in p H occurs. Therefore, such alkaline secretion cannot be detected by a pH-stat technique. Thus, this study provides information on the alkaline secretion which is not coupled to Hf secretion. As reported earlier, we found a higher rate of alkaline secretion in the ileum and the proximal colon (Parsons 1956, Hubel 1968). Whether the observed alkaline secretion is due to a normally operating mechanism of HCO, transport or merely a passive diffusion of HCO, from blood to lumen cannot be ascertained from the present study. However, changes of plasma bicarbonate concentration did not seem to influence alkaline secretion, suggesting that passive HCO; diffusion played a minor role (Table I). There is, however, experimental evidence also for an intestinal absorption of bicarbonate ions. The rat jejunum has been reported to absorb HCO, from the intestinal lumen in vivo by a mechanism that is partly Na+-dependent (Hubel 1973, Podesta & Mettrick 1977 b). The observed decrease of luminal p H could either reflect a
82
< H. % Tantiszra I, et al.
where the alkaline secretion arising from the coupling of Cl-/HCO, and Na+/H+ is not detected, cholera toxin produced only a small increase in ileal net alkaline secretion. This suggests that the high ileal COJbicarbonate content seen in cholera in animals and man zn r t r o resulted mainly from increased rates of transport in the coupled CI-/HCO; and Na+/H+ exchanges. It should be pointed out, however, that it is generally assumed from En-vztro studies that cholera toxin inhibits the coupled NaCl transport (Powell 1987). No attempt was made to probe the involvement of enteric nervous reflex in the observed ileal response for two reasons. Firstly, the increase in ileal alkaline secretion was of a small magnitude when compared either to its own basal secretion or to the jejunal alkaline secretion evoked by the enterotoxin. Secondly, atropine and hexamethonium were found to inhibit to a 6Jkl.1 ~ f r h o l e r atoxin on intestinal alkaline greater extent ileal basal alkaline secretion than s r u e t i o n trnd potentiul difference the ileal alkaline secretion evoked by cholera In the present study, intraluminal treatment toxin (Table 2, Fig. 2). with cholera toxin caused a marked increase in Jejunum. Jejunum is another site of alkaline jejunal alkaline secretion, whereas only a small loss in cholera diarrhoea. Studies in humans effect was observed in the ileum and apparently (Banwell et al. 1968, 1970) and experimental no effect was present in the proximal colon. animals (Carpenter et al. 1968) have demonSimilar results have been found in man, bi- strated that fluid as well as solute losses were carbonate concentration being significantly greater in the jejunum than in the ileum during higher in the jejunal and ileal contents during acute cholera diarrhoea. Furthermore, bicaracute cholera infection than during convalescence bonate was secreted into the jejunum, which (Ranu-ell et ill. 1968, 1970). I n the colon, on the normally absorbed this ion. T h e present results other hand, bicarbonate concentration was simi- are in line with those previous In-vzzw ohserlar during and after cholera (Speelman et al. \ ations in whichalkalinesecretion wasmeasuredas I ($36). Mou-ever, the profile of alkaline secretion total COJbicarbonate ions (Banwell et al. 1968, along the rat intestine in the present study was 1970, Carpenter et al. 1968, Love 1969) or as an different from the one observed in man. In increase in mucosal surface p H (McEwan et al. human cholera, the ileal bicarbonate concen- 1988). tration was significantly higher than the jejunal In the rat jejunum N a + / H + is unlikely to be one, the latter being approximately the same as coupled to a Cl-/OH- or a CI -/HCO, exchange that of plasma (Banwell et al. 1970). (Cassano et al. 1984). T h e Na+/H+ exchange is Ileum. In animal studies, in-riro cholera toxin partly responsible for sodium absorption, which causes a significant increase in ileal bicarbonate results in luminal acidification (Turnberg et al. secretion. In these studies bicarbonate transport 1970, Huhel 1973, Podesta & Mettrick ry77b). was measured as changes in the total Consequently, an inhibition of the Na+/H+ CO,/bicarbonate content in the lumen (Car- exchange may explain the increase in net alkaline penter et al. 1968, Leitch & Burrows 1968, secretion observed in the jejunum exposed to Norris el al. 1969, Moore r t al. 1971). However, cholera toxin. However, an increase in alkaline in z'itro, the absence of a significant effect of secretion into the lumen cannot be ruled out. In cholera toxin on the so-called residual flux has ritro rat jejunal crypts have been reported to been reported (Field et ul. 1972, Powell et al. hace an alkaline p H while jejunal villi exhibit an 1973), indicating that no electrogenic bicarbonate acid pH. T h e alkalinity of the crypt region was secretion occurred in z'itro. In the present study proposed to be due to an alkaline secretion by
N a + / H * exchange mechanism or a co-transport of sodium and hicarbonate. Since the decreased p H was accompanied by a rise in luminal PCO, (Turnberg et aI. 1970, Powell et a / . 1971,Hubel 1973, Podesta & Mettrick 1977b), it was concluded that a N a ' / H ' exchange was the major mechanism for luminal acidification. I n the present study, jejunum exhibited a net alkaline secretion when it was perfused by a IICOi-free saline solution. Similar results have been reported in previous studies using HCOifree luminal solutions (Powell et 01. 1971, Hubel 1973, Podesta 8; llettrick 1977 b). Since the pHstat method used in the present study measures net changes in pH, it is not possible to estimate to uhat extent a simultaneous proton secretion might be present in the response observed in the jejunum.
Alkaline secretion in cholera the crypt cells (Daniel et al. 1985). T h e present results d o not provide direct evidence in favour of or against any of the mentioned mechanisms. Cholera toxin-induced fluid secretion has been proposed to be mediated partly by an intramural nervous reflex (Cassuto et al. 1981, 1982, 1983). I n the present study the effects of various nerve blocking agents were tested on the cholera toxininduced jejunal alkaline secretion and, in some experiments, also on the toxin effect on PD. Atropine, fornierly reported to have no effect on cholera toxin-induced fluid secretion, showed no significant effect on either stimulated jejunal alkaline secretion or PD. T h u s cholinergic muscarinic mechanisms d o not appear to be involved in the cholera toxin responses observed. Hexamethonium and serosally applied lidocaine exhibited a small but significant inhibitory effect on the cholera toxin-induced alkaline secretion in the jejunum (Table 3, Figs. 4 and 5). T h e effect of hexamethonium on PD seemed to be comparatively greater than the effect on alkaline secretion, although hexamethonium also significantly decreased the jejunal basal alkaline secretion and PD (Table 2 ) . T h e latter observation makes it difficult to estimate the true importance of intramural nervous reflex in cholera toxin-induced alkaline secretion. I t can be roughly estimated that the maximal inhibition of toxin-induced jejunal alkaline secretion by hexamethonium and lidocaine is 1-25 yo,while their effect on fluid transport is at least 60% (Cassuto et al. 1982). Thus, the present study suggests that a non-neural mechanism mainly underlies the cholera toxin-induced jejunal alkaline secretion. T h e increase in PD produced by cholera toxin may be explained by an active electrogenic ion transport and/or by a fluid transport producing a streaming potential. The discrepancy between the effects of nerve blocking agents on toxininduced changes in PD and fluid transport on the one hand and alkaline secretion on the other indicates that the influence of intramural nervous reflexes on different secretory mechanisms may vary. T h e results of this study suggest that the enteric nerves mainly control an electrogenic non-alkaline secretion. This research was supported by grants from the Swedish Medical Research Council (2855), The Medical Council of Swedish Life Insurance Companies and the Faculty of Medicine, University of
83
Gothenburg. M. H. Tantisira was supported by a scholarship from the Chulalongkorn University, Bangkok, Thailand.
REFERENCES AL-AWQATI, Q., CAMERON, J.L. & GREENOUGH 111, W.B. 1973. Electrolyte transport in human ileum: effect of purified cholera exotoxin. A m 3 Physiol 224, 818-823. BANWELL, J.G., PIERCE, N.F., MITRA,R., CARANASOS, G.J., KEIMOWITZ, R.I., MONDAL,A. & MANJI, P.M. 1968. Preliminary results of a study of small intestinal water and solute movement in acute and convalescent human cholera. Indian 3Me d Res 56, 633-639. BANWELL, J.G., PIERCE, N.F., MITRA,R.C., BRIGHAM, G.J., KEIMOWITZ, R.I., FEDSON, K.L., CARANASOS, D.S., THOMAS, J., GORBACH, S.L., SACK,R.B. & MONDAL,A. 1970. Intestinal fluid and electrolyte transport in human cholera. 3 Clin Invest 49, 183-195. BINDER,H.J. & RAWLINS, C.L. 1973. Electrolyte transport across isolated large intestine mucosa. Am 3 Physiol225, 1232-1239. CARPENTER, C.C.J. 1982. The pathophysiology of secretory diarrhoeas. Med Clin 66, 597-610. CARPENTER, C.C.J., SACK,R.B., FEELEY, J.C. & STEENBERG, R.W. 1968. Site and characteristics of electrolyte loss and effect of intraluminal glucose in experimental canine cholera. 3 Clin Invest 47, 12IcFI220
CASSANO, G., STIEGER, B. & MURER,H. 1984. Na/H and CI/OH-exchange in rat jejunal and rat proximal tubular brush border membrane vesicles. Studies with acridine orange. Pfugers Arch 400, 39-317. CASSUTO, J., JODAL, M., TUTTLE,R. & LUNDGREN, 0. 1981. On the role of intramural nerves in the pathogenesis of cholera toxin-induced intestinal secretion. Scand 3 Gastroenterol 16, 377-384. CASSUTO, J., JODAL, M. & LUNDGREN, 0. 1982. The effect of nicotinic and muscarinic receptor blockade on cholera toxin induced intestinal secretion in rats and cats. Acta Physiol Scand 114, 573-599. CASSUTO, J., SIEWERT, A., JODAL, M. & LUNDGREN, 0. 1983. The involvement of intramural nerves in cholera toxin-induced intestinal secretion. Acta Physiol Scand 117, 195-202. DANIEL, H., NEUGEBAUER, B., KRATZ,A. & REHNER, G. 1985. Localization of acid microclimate along intestinal villi of rat jejunum. A m 3 Physiol 248, G293-Gz98. EDMOND,C.J. 1967. Transport of sodium and secretion of potassium and bicarbonate by the colon of normal and sodium-depleted rats. 3Physiol 193, 589402. FIELD, M. 1974. Intestinal secretion. Gastroenterology 66, 1063-1084.
84
M. H . Tantisira et al.
FIELD,M., FROMM,D., AL-AWQATI,Q & GREENOUGHT 111, W.B. 1972.Effect of cholera enterotoxin on ion transport across isolated ileal mucosa. 3 Clin Invest 51, 796-804. FLEMSTROM, G. 1987.Gastric and duodenal mucosal bicarbonate secretion. In : L.R. Johnson (ed.) Physiologv of the Gastrointestinal Tract, 2nd edn, pp. I O I 1-1029.Raven Press, New York. FLEMSTROM, G., GARNER, A., NYLANDER, O., HURST, B.C. & HEYLINGS, J.R. 1982. Surface epithelial HCO; transport by mammalian duodenum in vivo. Am 3'Physiol 243, G348-G358. FLEMSTROM, G., JEDSTEDT, G. & KIVILAAKSO, E. 1985. Bicarbonate secretion by colon and duodenum in the rat in vivo. .4cta Ph,ysiol Scand 124 (Suppl. 542),
360.
Modification of intestinal secretion in experimentalcholera. 3 Infect Dis 119, 117-125. PARSONS, D.S. 1956. The absorption of bicarbonatesaline solutions by the small intestine and colon of the white rat. Q3 Exp Physiol 41, 410-420. PHILLIPS,S.F.& SCHMALZ, P.F. 1979.Bicarbonate secretion by the rat colon: effect of intraluminal chloride and acetazolamide. Proc Soc Exp Biol Med 135, 116-122. PODESTA, R.B. & METTRICK, D.F. 1977a.HCO; and H' secretion in rat ileum in vivo. Am 3 Physiol232,
E574-E579. PODESTA,R.B. & METTRICK,D.F. 1977b. HCO, transport in rat jejunum : relationship to NaCl and H,O transport in vivo. Am 3Physiol232, E62-E68. POWELL,D.W.1987.Intestinal water and electrolyte transport. I n : L.R. Johnson (ed.) Physiology ojthe Gastrointestinal Tract, 2nd edn, pp. 1267-1305, Raven Press, New York. POWELL, D.W., BINDER, H.J. & CURRAN, P.F. 1973. Active electrolyte secretion stimulated by choleragen in rabbit ileum in vitro. Am 3 Physiol 225,
HCBEL,K..k 1967. Bicarbonate secretion in the rat ileum and its dependence on intraluminal chloride. Am 3 Physiol 213, 1409-1413. IiUBEL, K.A. 1968.The ins and outs of bicarbonate in the alimentary tract. Gastroenterology 54, 64745I . HUBEL,K.A. 1969.Effect of luminal chloride con781-787. centration on bicarbonate secretion in rat ileum. POWELL,D.W., SOLBERG,L.I., PLOTKIN,G.R., .4m 3 Physiol 217,40-45. S.B. CATLIN,D.H., MAENZA, R.M. & FORMAL, HUBEL,K.A. 1973.Effect of luminal sodium con1971. Experimental diarrhea. 111. Bicarbonate centration on bicarbonate absorption in rat jejunum. transport in rat salmonella enterocolitis. Gastro7 Clin Inrest 52, 3172-3179. enterology 60, 10761086. L ~ I T C HG.J. , & BURROWS, W. 1968.Experimental R.A. & NELLANS,H.N. cholera in the rabbit ligated intestine: ion and water SCHULTZ,S.G.,FRIZZELL, 1974.Ion transport by mammalian small intestine. accumulation in the duodenum, ileum and colon. 3 Ann Rev Physiol36, 51-91. Infect Dis 118, 349-359. H.E. & FIELD, M. 1977.Ileal mucosal cyclic LIEDTKE, C.M.& HOPFER,U. 1982.Mechanism of SHEERIN, AMP and CI secretion : serosal vs. mucosal addition C 1 ~translocation across small intestinal brushof cholera toxin. Am 3 Physiol 232, E210-Ez15. border membrane. 11. Demonstration of P., BUTLER,T., KABIR,I., ALI, A. & C1 -OH-exchange and CI- conductance. Am 3 SPEELMAN, BANWELL, J. 1986. Colonic dysfunction during Ph,ysiol 242, G272-Gz80. cholera infection. Gastroenterology 91, I 164-1 170. LOVE,A.H.G. 1969.Water and sodium absorption by SWALLOW, J.H. & CODE, C.F. 1967. Intestinal the intestine in cholera. Gut 10. 6347. transmucosal fluxes of bicarbonate. Am 3 Physiol LCBCKE,R., HAAG,K., BERGER, E., KNAUF,H. & 212, 717-723. GEROK,W. 1986. Ion transport in rat proximal TAI,Y. & DECKER,R.A. 1980. Mechanisms of colon in vivo. Am 3 Physiol 251, G 1 3 2 4 1 3 9 . MCEWAN,G.T.A., DANIEL, H., FETT,C., BURGESS, electrolyte transport in rat ileum. Am 3Physiol238, Gzo8-Gz I 2. M.N. & LVCAS, M.L. 1988.The effect of Escherichia L.A., FORDTRAN, J.S., CARTER, N.W. & roli STa entrotoxin and other secretagogues on TURNBERG, RECTOR JR, F.C. 1970.Mechanism of bicarbonate mucosal surface pH of rat small intestine in vivo. absorption and its relationship to sodium transport Proc R SOCLand 234, 219-237. in human jejunum. 3 Clin Invest 49, 548-556. MOOREJR, W.L., BIEBERDORF, F.A., MORAWSKI, S.G., L. 1957.Acid-base changes FINKELSTEIN, R.A. & FORDSTRAN, J.S. 1971.Ion WILSON,T.H. & KAZYAK, across the wall of hamster and rat intestine. Biochim transport during cholera-induced ileal secretion Biophys Acta 24, 124-132. in the dog. 3 Clin Invest 50, 312-318. NORRIS,H.T., CURRAN, P.F. & SCHULTZ, S.G. 1969.
