This study was done to establish and validate a single-pass perfusion method for measuring the absorption of water and electrolytes by the mouse small intestine. The method was then used to study intestinal absorption in mice whose villin gene had been invalidated (v−/−). The single-pass perfusion of the jejunum measures the absorption of water, Cl−, Na+, K+, HCO , and glucose in anesthetized wild-type and v−/− mice in vivo. We measured absorption under basal and stimulated conditions (carbachol, vasoactive intestinal polypeptide, intralumen PGE2). Basal absorption and stimulated secretions were similar to those previously obtained in rats. There was no difference between wild-type and v−/− mice in animals with mixed genetic background or in pure C57BL6 mice. We conclude that this in vivo perfusion method is suitable for studying the absorption/secretion of electrolytes in the mouse intestine and that a lack of villin does not significantly alter basal and secretagogue-stimulated electrolyte movements across the epithelium of the mouse jejunum in vivo.
- intestinal hormones
- brush-border epithelium
transgenic mice are powerful tools for studying the physiological functions of many genes and their products. A foreign gene may alter the function of a native gene and so modify its function. Functions directly or indirectly affected must be screened to detect changes in physiological parameters.
Because mice are small, most published studies have used histological or biochemical tests on fragments of tissues or suspensions of isolated cells to test the effects of the gene of interest. Few data have been obtained in mice by measuring organ function in physiological conditions in vivo, except for general parameters such as blood pressure, body weight, or oxygen consumption.
Tissue and animal models are required to study the functions of epithelial cells, such as absorption and secretion of electrolytes in the small intestine. These functions are well documented in vivo in the rat (5, 21, 22, 24), whereas most studies on mice have used Ussing chamber experiments (6, 10, 13). However, these are more suitable for exploring mechanisms than for establishing physiological secretory conditions.
The present study was done to establish and validate a single-pass perfusion method for measuring the intestinal absorption of water and electrolytes by the mouse small intestine. We then used this method to study intestinal absorption in mice in which the villin gene had been invalidated (v−/− mice) (9).
Villin is an actin-binding protein located in the intestinal and renal brush borders. In vitro, villin presents F-actin bundling and severing activities depending on the Ca2+concentration (3). In vivo pharmacological (carbachol stimulation) or pathophysiological situations that increase the intracellular Ca2+ concentration result in impaired changes in the actin cytoskeleton in v−/− mice (9). The F-actin fragmentation is also necessary for carbachol to inhibit NaCl absorption in the ileum (14). Villin could therefore be implicated in modulating transporter activities by altering the dynamic stability of the actin cytoskeleton. We have measured the basal absorption of water and electrolytes and the effect of hormones that modulate their transport (4, 7), such as carbachol, vasoactive intestinal polypeptide (VIP), and PGE2, in wild-type and v−/− mice.
Each mouse was fasted for 12 h and allowed drinking water ad libitum before it was anesthetized with 1.5 g/kg ip ethylurethane. Each mouse was then placed under a heating lamp to keep its core temperature at 38 ± 0.2°C. It breathed spontaneously, and the normal color of the ears and feet indicated good oxygenation throughout the experiment. The abdominal wall was opened, and the jejunum was ligated 1 cm below the ligament of Treitz. A Silastic perfusion catheter (Silastic 602.205, ID 1.02 mm, OD 2.16 mm) bearing two silicone swellings ∼3 mm apart was inserted into the gut just below the ligature and secured in place by a silk ligature between the two swellings. A collecting cannula (Silastic 602.305, ID 1.98 mm, OD 3.18 mm) was inserted into the lumen of the jejunum 10 cm further down to collect the perfusion fluid (Fig. 1). A third catheter (Silastic 602.155, ID 0.64 mm, OD 1.19 mm) was placed in the peritoneum for intraperitoneal injections. The abdominal wall was closed with three sewing points of 4-0 silk. The isolated segment of jejunum was quickly rinsed with perfusion fluid, the input catheter was attached to a perfusion pump (Minipuls II, Gilson Instruments, speed 80, Technicon Tygon tubes black/black R3607, producing a flow of ∼2 ml/15 min), and the output catheter was placed over a fraction collector set to collect 15-min samples. The total perfusion time was ∼135 min. Each mouse was killed at the end of the perfusion, and the segment of jejunum was removed and its length measured. Macroscopic examination of the abdomen contents indicated no local necrosis or hypoperfusion.
Perfusion Fluid and Protocol
The perfusion fluid was a Krebs-Ringer-bicarbonate solution (KRB) containing (in mM): 120 NaCl, 4.5 KCl, 0.7 Na2HPO4, 1.5 NaH2PO4, 1.2 CaCl2, 0.5 MgCl2, and 10 glucose. Perfusion began with an equilibration period of 45 min; these samples were discarded. Three 15-min samples were then collected (t= −45 to t = 0 min) to obtain basal absorption/secretion. One of the following was administered attime 0: carbachol (20 μg/kg ip), VIP (50 nmol/kg ip), or PGE2 (10 μg/ml) in the KRB perfusing fluid for 45 min. PGE2 was dissolved in ethanol to give 1% ethanol in the perfusing fluid. The gut was first perfused with KRB containing 1% ethanol for 30 min before adding PGE2, which was then perfused for 45 min.
The volume of each sample was determined by weight (assuming that 1 ml = 1 g). The actual input volume was determined before each perfusion by averaging the weight of three 15-min samples directly from the pump. The absorbed or secreted volume was calculated as collected volume minus perfused volume and expressed per centimeter of perfused bowel. Negative values indicate absorption, and positive values indicate secretion.
Na+, K+ (ion electrodes), Cl−(interaction with mercury sulfocyanide and subsequent photometry of iron sulfocyanide), total CO2 (enzymatic method with phosphoenol pyruvate carboxylase), and glucose (hexokinase) were measured by standard established clinical methods using a laboratory autoanalyzer (Kone Specific Supra 4.4, Kone Instruments, Evry, France). The net fluxes of ions and glucose were calculated from the concentrations in the influx and efflux fluids and the flow volume.
v−/− Mice were obtained as previously described (9). The following experiments were performed on siblingv−/− mice and wild-type animals with either a mixed genetic background (M2) or on mice obtained by multiple crosses with the C57BL6 genetic background, which gave pure C57BL6 wild-type andv−/− mice (B5). The Curie Institute has approved the protocol used for animal testing.
Groups were compared by two-way ANOVA, which separated the effects of treatment vs. control and v−/− vs. wild-type mice (Statview software). A P value of <0.05 was considered significant.
Basal Absorption in M2 and B5 Wild-Type Mice
The average values of all of the variables (fluid, Na+, K+, Cl−, total CO2, glucose) under basal conditions were obtained in four series of experiments before any hormone stimulation in M2 (3 groups of 5, 6, and 6 animals, respectively) and B5 mice (n = 7). The basal values for the two genetic backgrounds were essentially the same (Table 1), except for total CO2, which was slightly less well absorbed in M2 mice than in B5 mice (P < 0.001).
Hormone Stimulation of Wild-Type Animals
Carbachol decreased fluid absorption by ∼ during the initial 15 min of its administration in both M2 (Table2, Fig. 2) and B5 (−26.57 ± 4.46 μmol · cm−1 · 15 min−1;n = 7) animals. Cl− (B5: −2.35 ± 0.58 μmol · cm−1 · 15 min−1; n = 7) and Na+ (B5: −1.77 ± 0.60 μmol · cm−1 · 15 min−1; n = 7) absorptions followed similar patterns. K+ transport was not significantly altered, remaining near 0 (B5: −0.03 ± 0.03 μmol · cm−1 · 15 min−1;n = 7). Total CO2 absorption decreased in both M2 and B5 mice (B5: −0.22 ± 0.44 μmol · cm−1 · 15 min−1;n = 7) and became zero or a small secretion in M2 mice (0.03 ± 0.08 μmol · cm−1 · 15 min−1; n = 5). The loss of glucose from the gut lumen did not significantly change in response to carbachol (B5: −1.04 ± 0.06 μmol · cm−1 · 15 min−1;n = 7).
VIP (50 nmol/kg ip) produced a decrease in water absorption in M2 wild-type mice (n = 6) (Table 2, Fig.3) that was larger than that produced by carbachol. The increase in secretion peaked in the first (0–15 min) fraction collected after VIP injection and returned to the preinjection level 45 min after VIP injection (Fig. 3). The other variables measured were also shifted toward secretion after VIP injection (P < 0.001), except for glucose absorption.
PGE2 (10 μg/ml) was dissolved in KRB containing 1% ethanol and perfused into the jejunum lumen (n = 6). The ethanol in the perfusing fluid produced a slight, transient increase in the volume absorbed during the first 15-min period (−30 to −15 min; Fig. 4). The volume absorbed returned to the preethanol values in the −15- to 0-min period (Fig.4). PGE2 produced a progressive and significant (Table 2, Fig. 4) decrease in absorption that affected all of the variables measured (even glucose absorption was slightly decreased after PGE2). Total CO2 net flux turned to clear secretion during PGE2 infusion. The peak secretion for all variables occurred during the 15- to 30-min period, and this secretion remained constant during the 30- to 45-min period of PGE2infusion (Fig. 4).
Basal Absorption in M2 and B5 v−/− Mice
Basal absorption was measured in four independent series ofv−/− animals, three groups of M2 mice (of 5, 6, and 6 animals, respectively), or one group of B5 animals (n = 8). The average values of all the variables (fluid, Na+, K+, Cl−, total CO2, glucose) in M2 and B5 mice are given in Table 3. The basal values for the two genetic backgrounds, M2 and B5, were similar, as for wild-type mice, except for the CO2, which was lower in M2 mice than in B5 mice. There was no apparent difference between wild-type and v−/− animals.
Hormone Stimulation in v−/− Mice
The stimulated values for wild-type (Table 2) and v−/−mice were not significantly different (Table4; Figs. 2-4). Carbachol stimulation decreased the absorption of fluid in M2 mice (Table 4, Fig. 2). Cl− and Na+ absorptions followed similar patterns, and CO2 absorption became a small secretion. The loss of glucose from the gut lumen did not significantly change after carbachol. Similar patterns were observed in B5 mice (volume: −26.30 ± 5.48 μl · cm−1 · 15 min−1; Na+: −1.00 ± 0.74 μmol · cm−1 · 15 min−1; K+: 0.00 ± 0.04 μmol · cm−1 · 15 min−1; Cl−: −1.87 ± 0.65 μmol · cm−1 · 15 min−1; glucose: −1.20 ± 0.08 μmol · cm−1 · 15 min−1;n = 8). The difference in CO2 absorption in B5 and M2 wild-type mice also occurred in v−/− (B5: −0.16 ± 0.09 μmol · cm−1 · 15 min−1; n = 8). VIP (50 nmol/kg ip;n = 6) decreased water absorption (Table 4, Fig. 3). The other variables were also shifted toward secretion (P < 0.001), except for glucose absorption. PGE2 (10 μg/ml, n = 6) produced a progressive, significant decrease in absorption that affected all variables, including glucose (Table 4, Fig. 4).
Our studies on the absorption of water and electrolytes by the mouse jejunum gave basal and hormone-stimulated values of the same order of magnitude as those found for the rat. We found no major differences between the values for v−/− mice and their wild-type littermates for both mixed and pure (C57BL6) genetic backgrounds.
This is the first report, to our knowledge, of intestinal transport in vivo in mice measured by using a perfusion method. Our unpublished studies on Wistar rats under similar conditions provided reference basal values of intestinal absorption. The basal absorption of water in Wistar rats was greater per centimeter of intestine than the absorption in mice (−45.9 ± 2.9 to −64.8 ± 5.6 μl · cm−1 · 15 min−1 from 7 series of 6 rats each). This may be due to differences in the absorptive capacity of the jejunal epithelium or to differences in the absorptive surfaces and/or in the rate of perfusion (2 ml/15 min in mice vs. 4 ml/15 min in rats). This perfusion flow rate was chosen to take into account the smaller diameter of the mouse jejunum. Rats absorbed the various electrolytes [basal values from 7 different series of experiments including 5–7 Wistar rats each (in μmol · cm−1 · 15 min−1); Na+: −3.31 ± 0.65 to −6.84 ± 0.65; K+: −0.08 ± 0.08 to −0.31 ± 0.06; Cl−: −5.65 ± 0.26 to −7.07 ± 0.34; and glucose −1.45 ± 0.15 to −1.72 ± 0.13] at about the same rates as mice, except that HCO (−1.43 ± 0.15 to −1.74 ± 0.19 μmol · cm−1 · 15 min−1) seems to be less readily taken up by the mouse jejunum, and this mainly in M2 mice.
Most published studies on mice have used in vitro methods in an Ussing chamber, measuring the short-circuit current (I sc), which reflects global electrogenic ion transport across the mucosa. I sc is usually believed to mainly reflect Cl− secretion by intestinal crypts (13). The basal values ofI sc in the mouse duodenum and jejunum are on the order of 30–50 μA/cm2 of mucosa (6, 13,22), with large individual variations, when measured in KRB solution without glucose in the mucosal compartment. There are few reports of individual ion transport: a net Na+ influx of 5–6 μeq · cm−2 · h−1was reported in wild-type mouse jejunum with KRB, corresponding to neutral NaCl transport (6). Our in vivo data of ∼12 μeq · cm−2 · h−1 are in the same range, assuming an equivalent Ussing chamber area of 1 cm/cm2 of jejunum length and a doubling of Na+ absorption in the presence of 10 mM glucose.
Carbachol, VIP, and PGE2 all stimulated secretion in vivo, as indicated by studies on the mouse jejunal mucosa in Ussing chambers (7) and in the rat jejunum in vivo (21, 23). The amplitude and duration of the responses, however, depend greatly on the conditions of drug administration. Carbachol and VIP were given as single intraperitoneal injections, and their effects lasted only 15–30 min. PGE2 was added to the lumen perfusion fluid and had a progressive effect that peaked 30 min after the onset of perfusion and lasted as long as PGE2 remained in the perfusion fluid. A clear secretory effect was observed with 50 nmol/kg VIP (166 μg/kg ip), whereas effects of this order of magnitude were obtained with intravenous doses of 30–100 μg · kg−1 · h−1 of VIP in the rat (23). PGE2 was given as a steady-state perfusion, whereas carbachol and VIP were given as bolus injections. Thus their secretory and possible motor effects changed the steady state conditions, so that the absolute absorption values obtained should be considered with caution. However, the results for the wild type and v−/− mice were always parallel. Hence, it is unlikely that there was any obvious difference between the two groups of mice.
Villin is mainly expressed in absorptive epithelial cells, such as those of the small and large intestine and the kidney proximal tubule (2, 16, 18). Villin belongs to a large family of Ca2+-regulated actin-binding proteins that are structurally and functionally similar. In vitro, villin presents F-actin bundling and severing activities, depending on the Ca2+concentration (3). Villin-severing capacity might be involved under certain physiological conditions, in bacterial infection, or in response to fasting and feeding, which are believed to modulate intracellular Ca2+ (11, 15, 17). Carbachol activates basolateral muscarinic receptors, leading to an increase in intracellular Ca2+ (8). We have shown that carbachol placed directly on jejunum loops isolated in situ causes the brush-border actin to fragment, which was not observed inv−/− mice (9).
Studies on rabbit designed to analyze the inhibition of NaCl absorption by carbachol suggest that villin is involved in intestinal absorption (14). The authors proposed that carbachol inhibits NaCl absorption by fragmenting actin due to the severing activity of villin when the intracellular Ca2+ concentration is increased. Nevertheless, the stabilization of F-actin filaments by jasplakinolide only slightly increased the carbachol-abolished NaCl transport: NaCl absorption remained very low, ∼15% of the basal absorption. The present study shows no differences in water and electrolyte transport by the wild-type and v−/− mice under both basal and stimulated conditions. This suggests that the transport of these ions is not mainly correlated with villin-induced actin fragmentation. As our previous data showed that F-actin is stable in the absence of villin (9), the present results suggest that the fragmentation of F-actin is not a major parameter in NaCl absorption. This conclusion is supported by the small inhibitory effect of jasplakinolide (14). However, studies on the intact animal level point out the complexity of the interactions that maintain a normal intestinal function. The present studies on mice were not intended to specifically study Na+/H+ exchange, which might be slightly different in wild-type and v−/−mice, but to simulate physiological in vivo conditions and so determine whether the small differences found by in vitro methods were important in vivo. Invalidation of the villin gene is apparently not sufficient to cause significant absorptive changes under our conditions, probably because of the complexity of the many parameters that interact to maintain normal intestinal function.
In conclusion, we have developed an in vivo method for studying intestinal transport in mice. The absence of any difference between wild-type and v−/− mice confirms the complexity of the interactions between proteins at the cell, tissue, organ, and animal levels. The exact role of villin under pathological conditions remains to be further investigated.
Address for reprint requests and other correspondence: S. Robine, Laboratoire de Morphogénèse et Signalisation Cellulaires, Institut Curie, UMR 144, 26, rue d'Ulm, 75248 Paris cedex 05, France (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
First published January 2, 2002;10.1152/ajpgi.00327.2001
- Copyright © 2002 the American Physiological Society