from the mixture or the peptide, but did reduce the pool size of both when taken up

Transcription

1 J. Physiol. (1979), 293, pp With 2 text-figurem Printed in Great Britain FACTORS AFFECTING THE MOVEMENT OF AMINO ACIDS AND SMALL PEPTIDES ACROSS THE VASCULARLY PERFUSED ANURAN SMALL INTESTINE BY C. I. CHEESEMAN From the Department of Physiology, University of Leicester, Leicester, England (Received 7 September 1978) SUMMARY 1. In the vascularly perfused frog small intestine, the characteristics of the washout into the vascular bed of L-leucine and glycine have been investigated following loading either as free amino acids or as the dipeptide glycyl-l-leucine. In both cases, the washout of the two amino acids was biexponential and could be ascribed a fast rate constant K1 and a slow rate constant K2. The K1 for glycine was always smaller than that for L-leucine. 2. The K1 and K2 for L-leucine had a Q1 of 1-71 and 2-66 respectively, while steadystate transfer of the amino acid exhibited a Q1 of Vascular perfusion rate affected the K1 but not the K2 for L-leucine. 4. Na+ replacement in the lumen did not affect K1 and K2 for either amino acid from the mixture or the peptide, but did reduce the pool size of both when taken up as free amino acids. Peptide uptake, as measured by pool size, was not affected by luminal Na+ replacement. 5. If Na+ in the lumen was replaced with Li+, all substitutions of Na in the vascular bed inhibited the exit of L-leucine. 6. These results are discussed in relation to a model proposing that sodium pumping across the basolateral membranes of the epithelial cells to keep the intercellular spaces patent is an important factor in the movement of amino acids from the enterocytes to the blood stream. INTRODUCTION Recently Boyd, Cheeseman & Parson (1975a) have shown that the vascularly perfused small intestine of the frog can provide an excellent system for studying the movement of amino acids from the lumen to the vascular bed. In particular, they demonstrated not only that the movement of L-leucine across the epithelium was influenced by the presence of Na in the solution bathing the luminal aspect of the mucosal cells, but also that there was an inhibition of the movement of the amino acid into the vascular bed when Na was replaced with K in the vascular perfusate. They also presented a washout (perturbation) technique which showed that L-leucine appeared in the portal venous effluent in a biexponential manner, suggesting that the amino acid unloaded from two pools with two different rate constants. Present address and the one to which reprint requests should be sent: Department of Physiology, University of Alberta, Edmonton, Alberta, Canada T6G 2H /79/ $ The Physiological Society

2 458 C. I. CHEESEMAN Subsequently, the same authors proposed a model to explain the effects of Na distribution on the movement of L-leucine across this epithelium (Boyd, Cheeseman & Parsons, 1975b). This paper reports on experiments which have been performed in an attempt to resolve further the question of the involvement of Na in the movement of L-leucine across the frog small intestine and on a more detailed study of the washout of L-leucine into the vascular bed. Furthermore, these results have been compared with the movement of L-leucine across the intestine when the amino acid is initially present as part of the dipeptide glycyl-l-leucine. The data presented here provide additional evidence that the replacement of Na in the vascular bed does influence the movement of amino acids from the epithelium into the capillary bed. Some of these findings amplify earlier observations (Cheeseman, 1977a, b). METHODS Animals Rana pipiens,weight range 25-5 g, were supplied by Xenopus Ltd (Redhill, Surrey, England). Animals were kept at 15-2 'C in tanks with a continuous flow of tap water into a pool. They were fed twice a week with live American house crickets (Xenopus Ltd) or with live locusts. Experimental procedure The technique used was the vascul-irly perfused Anuran small intestine described by Boyd et al. (1975a). The compositions of luminal and vascular fluids are given in Table 1. Single pass flow systems were employed for both the perfusion of the vascular bed and of the lumen. The vascular flow rate was adjusted to fall within the range 3-1 ml.. g dry wt. of gut-' min-' and the luminal flow rate was set to fall within the same range except where indicated. Steady-state experiments The steady-state appearance of L-leuIcine in the vascular bed was measured using the procedures previously described (Cheeseman & Parsons, 1976). Briefly, the lumen was perfused with a particular concentration of substrate for a minimum period of 3 min using a single pass system. The samples of vascular perfusate collected from the portal vein every 5 min were then analysed emzymatically for L-leucine as described below. Washout experiments (perturbation technique) The tissue was loaded from the lumen with radio-labelled substrate using the technique described by Boyd & Parsons (1978). All perfuisions of the lumen were single pass, and to ensure that the introduction of substrate at the start of the loading period and its removal at the end were both very rapid, the following procedure was adopted. Two three-way taps were inserted next to each other in the flow line between the luminal peristaltic pump and the input luminal cannula. A tube with male luer end fittings and a capacity of 16 ml. was pushed into the female luer side arms of the two three-way taps. This tube, or 'side arm', was filled with the appropriate radio-labelled substrate. At the start of a loading period, the two three-way taps were turned to direct the flow of fluid from the pump through the side arm tube before going to the luminal cannula. Similarly, at the end of the loading period, the two three-way taps were turned to their original position and the flow of luminal fluid went directly to the intestine without going through the side arm tube. The loading period lasted 3 min, which ensured that a steady state of transfer was established. Materials Bovine serum albumin, Fraction V, was supplied by Miles Laboratories Ltd (Stoke Poges, Bucks.). Glycine, L-leucine Sigma grade and glycyl-l-leucine were supplied by Sigma London (Poole, Dorset). Bio-Solv BBS3 was from Beckman (Croydon, Surrey) and L-[53H]leucine,

3 AMINO ACID WASHOUT INTO VASCULAR BED [Ul"C]glycine, [U14C]glycyl-L-leucine and glycyl-l-[u14c]leucine were from Radiochemical Centre (Amersham, Bucks). Analytical methods L-leucine was estimated by the L-amino acid oxidase-peroxidase technique (Fujita, Parsons & Wojnarowska, 1972). The standards were made up in the solutions used for infusion into the vascular bed; deproteinization was not necessary. TABLE 1. Luminal and vascular solutions. All values given are final concentrations of constituents in m-mole/l. except where stated. Bicarbonate Ringers were gassed with 95 % 2/5 % C2 gas mixture and phosphate Ringers were gassed with pure 2 NaCl KCl MgSO4 MgCl2 Na2HPO4 NaH2PO4 K2HPO4 KH2PO4 NaHCO3 KHCO3 CaCl2 LiCl Mannitol Choline Cl Albumin (g %) Glucose Lurninal A A -B B C *8 2*15 3 *8 A-9 all Vascular solutions A B C D E F (.LQ x x8 2* * * *5.5 *5 * * Scintillation counting The appearance of L-leucine and glycine in the vascular bed during the perturbation experiments was routinely estimated using scintillation counting. The scintillant was toluene-based and contained 1% Bio-Solv (Beckman Instruments Inc.) which obviated deproteinization of the samples (Carter & van Dyke, 1971). Quenching was measured from the channels ratio using external standard. The counting was performed in a Packard 2425 scintillation counter. Estimation of tissue dry weight The tissue was removed from the animal, opened along its anti-mesenteric border and blotted on moist filter paper. It was then weighed in a tared vessel before and after drying to constant weight in an oven at 1 'C overnight. Analysis of sub8trate within tissue At the end of a 3 min exposure to radio-labelled substrate in the luminal solution, vascular and luminal perfusions were stopped and the tissue was excised and rinsed with ice-cold isotonic choline chloride solution. A portion was retained for wet and dry weight measurements and the remainder was solubilized in a preweighed vial using -5 N-nitric acid. Aliquots of the resulting solution were then counted as described above. No correction was made for extracellular space. Calculation and expression of results All results expressing appearance in the vascular bed were calculated from a knowledge of the rate of arterial inflow and on the assumption of 1 % recovery of the arterial infusate from the portal vein. 459

4 46 C. I. CHEESEMAN The rates of appearance of L-leucine and glycine are expressed as molee. g dry wt.-' hr-1 or have been normalized where indicated against steady-state control rates of appearance set at 1. This procedure permits comparison between individual animals which show widely varying transport capacities (Cheeseman & Parsons, 1976). Groups of data were compared using the Student's t test (paired or unpaired) and the 5% level of significance was used as a criterion for rejecting the null hypothesis. Washout kinetics To determine the rate constants and initial pool sizes of the two washout components, the data were first plotted graphically. If the curves were smooth, indicating that the preparation was in a steady state for the entire washout period, a curve-fitting procedure was adopted as follows. The rates of appearance and times of observation were processed using a programme kindly supplied by Dr S. Petersen. The data were converted to semilogarithmic form and a regression line was fitted to the last three data points by a 'least-squares' method. The component contributed to the preceding values by this line was then subtracted and a second regression line was fitted to the remaining data. Next, a X2 value for these two lines was calculated. This procedure was repeated using the last four data points for the first regression line, and so on until all the data points were included in the first line. Finally, the pair of lines with the smallest X2 value was chosen as the best fit. Any set of data which did not give a significant X2 estimate was rejected. The two straight lines obtained by the above procedure could each be described by the equation log z = log (KSO) Kt, (1) where is the rate of appearance of substrate in the vascular bed, K is the rate constant for the washout, So is the size of the pool contributing to the washout at time zero and t represents time. Eqn. (1) is derived by differentiation from the expression log St = log So--434Kt, (2) where St is the size of the pool at time t after the start of the washout. In order to distinguish between the two components contributing to the washout the following subscripts have been used. Sol and K1 refer to the fast component and S2 and K2 refer to the slow component. t is expressed in minutes. It should be noted that estimates of So and S2 from the experimental data using the above expressions assume that the two pools empty independently of each other. If they empty in series or are coupled in some way, then these estimates of the pool sizes may be quite inaccurate (Huxley, 196). RESULTS The rate of appearance of L-leucine or glycine in the vascular bed decreased with time and this washout can be described by the sum of two exponential with appropriate rate constants of K1 for the fast component and K2 for the slow component (Fig. 1). Assuming these two components to be independent of each other, the size of the two pools was estimated for L-leucine in four separate experiments. Analysis of the acid extracts from the tissue also enabled a direct estimate of the leucine within the tissue to be made at the end of the 3 min loading period (legend to Fig. 1). The direct estimate of L-leucine in the tissue was the same order of magnitude for either the fast component pool (So,) or the slower component pool (So2). In order to attempt to distinguish further between these two components, the effect of temperature and vascular flow rate on the washout was investigated.

5 AMINO ACID WASHOUT INTO VASCULAR BED E Time (min) Fig. 1. The washout of L-leucine into the vascular bed. This is a semilogarithmic plot of the results of a typical experiment. The tissue was loaded from the lumen with [14C]Lleucine ( 1 mm) for 3 min at the end of which time the luminal perfusate was abruptly changed to a solution containing no substrate. The concentration of L-leucine was measured in fractions of the portal vein effluent collected over 3 min periods for the first 3 min and thereafter over 5 min periods. These data were then expressed as a rate of appearance in the vascular bed in Ezmole. g dry wt. of tissue-'. hr-1 and plotted semilogarithmically. The two lines were fitted to the points by the procedure given in the text and their slopes represent the rate constants for the two components which appear to make up the washout. The term K2 refers to the slow component and K, refers to the faster ' stripped ' component. At the end of four such experiments the intestine was loaded with [14C]L-leucine (l mm) for a further 3 min period and after rapid excision the tissue was extracted with -5 N-nitric acid. Aliquots of the acid extract were then analysed for their L- leucine content by liquid scintillation counting. This estimate of the total tissue content of L-leucine was 6-53 ± 1-52, while the indirect estimate made from the intercept and the slopes of the lines for the two pools contributing to the washout gave values of 4-53 ± 1-43 for the So, and for the So2. These values are expressed in mole.g dry wt. of tissue-' + S.E. of mean. TABLE 2. The effect of temperature on washout rate constants and rate of steady-state transfer of L-leucine. I MM-L-leucine in luminal solution A (Table 1) was perfused through the lumen for 3 min, during which time the steady state rate of appearance in the vascular bed (transfer) was measured. Values given are in molee. g dry wt.-' hr-11 + s.e of mean. The estimates of the washout rate constants K, and K2 for L-leucine were made in four animals and are expressed in min-" ± s.e. of mean. The preparation was stabilized for I hr at each temperature before any measurements were made Temperature KjL K2 Transfer 15 C ± ± 11 * 25 (C Q ± ± -2

6 CHEESEMAN The effect of temperature on washout rate constants Paired experiments in the same animal at 15 and 25 'C showed that both rate constants for the washout were affected. The Q1 for the fast rate constant was 1P71 and that for the slow component was Individually, these effects are similar to that of temperature on the over-all steady-state transfer of L-leucine from the lumen to the vascular bed, which showed a Q1 of 2-45 (Table 2). 2 - ' 15 - C o -1 X.~~~~~~- *,' c '. Fast component C.<- y= -94 x M o Slow component, -5 _ = 7x + 27 o D_ ~ - -' o Vascular flow rate (ml. g dry wt.-' min' ) Fig. 2. The effect of vascular flow rate on the washout rate constants K1 and K2 for L-leucine. The filled circles represent the fast component rate constants K1 and the open circles represent the slow component rate constants K2. The straight lines through the points were fitted using regression analysis. Vascular flow rate The washout of L-leucine into the vascular bed was compared at three different arterial infusion rates in individual experiments, and it was found that only that 'fast' component was affected by changing the flow rate. Fig. 2 shows the pooled data from several experiments and it is clear that K1 is sensitive to vascular flow rate, whereas K2 appears to be almost completely unaffected. Comparison of L-leucine and glycine washout The tissue was loaded with both amino acids using either an equimolar mixture of the two amino acids or with the dipeptide glycyl-l-leucine to compare their efflux into the vascular bed. Both amino acids showed biexponential washouts when loaded either as the amino acid mixture or as the dipeptide, and in both cases the slow rate constant K2 for the two amino acids was very similar. The fast rate constant K1 for leucine was the same whether the amino acid was administered as the mixture or as the peptide and the K1 for glycine was nearly 5 % smaller than that for leucine in both cases (Table 3).

7 AMINO ACID WASHOUT INTO VASCULAR BED 463 Cqi o oq 6co 6 o 66 eco O aq * o o 1 CO CO - o (6 *-H 6 +i.4 Oa O _ o - CX _- CO _ co 11 co eq 1 UZ 11 C eq 1 11 t r 11 r-66 cq t- * C) ;.4-1 * 4 -o W41 C4 - CO 4iU 1 CC C eoo aq 1 CO O C1 to 6 6 o D_4.o 1 CO Q1U: to CO co - 1: ti II 11 Cqo II 11 1 w _w cm e o o6 aq o co ew o aq m aq t- OX ce r W- - o *C) +O+ ++ I- + O C+

8 464 C. I. CHEESEMAN Influence of luminal sodium upon efflux rates Comparison of the washout rate constants for L-leucine and glycine showed that the replacement of Na with K in the solution flowing through the lumen had no significant effect (Table 3). The slow rate constants were again the same for both amino acids and the fast rate constant for L-leucine was still apparently greater than that for glycine. TABLE 4. Comparison of steady-state transfer of L-leucine into the vascular bed from either 5 mm-l-leucine in luminal solution A (Table 1) or 5 mm-glycyl-l-leucine in luminal solution B (Table 1). The control rate for leucine transfer was taken as 1 for each animal (the mean rate of transfer is given in parentheses and is expressed in molel. g dry wt.-i hr-1 + S.E. of mean) and the other values obtained were normalized against that figure to allow for variation between individual animals (Cheeseman & Parsons, 1976). The different vascular solutions are as listed in Table 1 and the main substituting ion present in them is also given. The vascular bed was perfused for 3 min in every hour with solution A and during the remaining 3 min of each hour with one of the other solutions B-F. The luminal perfusion rate was in the range ml. g dry wt.-' min-'. The leucine in the vascular effluent was estimated enzymatically Vascular solution A B C D E F Na+ K+ K+ Li+ Mannitol Choline 5 mm-l ± ± 5 leucine + luminal ( ) Na+ n mM ± glycyl-l-leucine (8-5 ± 5-9) no luminal Na+ n Estimates of tissue pool sizes of amino acids Table 3 also shows the estimated pool sizes for the fast (So,) and slow (So2) components of the washout for L-leucine and glycine when loaded into the tissue either as free amino acid or as a peptide. This was done either in the presence or absence of Na in the lumen which was replaced with K. When loaded as the free amino acids the So1 for L-leucine was always greater than that for glycine while there was no obvious trend for the size of the S2. Replacement of Na in the lumen reduced the size of Sol for both L-leucine and glycine. However when the two amino acids were loaded as the peptide glycyl-l-leucine the Sol for glycine appeared to be larger than the Sol for L-leucine. Also the replacement of Na in the lumen had no significant effect upon the size of either of these pools. The pool sizes for the slow component (SO2) were all smaller than the 'fast component pools' and again Na substitution in the lumen did not appear to affect their size. Vascular Na replacement Steady-state transfer of 5 mm-l-leucine in frog Ringer is reduced by 16 % when the NaCi is substituted in the vascular perfusate by KCl (vascular solution C). But

9 AMINO ACID WASHOUT INTO VASCULAR BED there is no inhibition when the Na concentration in the vascular bed is reduced to 28 mm by the replacement of NaCl with LiCl, choline chloride or isotonic mannitol (vascular solutions D, E and F). Also, the K-associated inhibition is not significantly increased when the Na is completely replaced with K salts using vascular solution B (Table 4). These observations were then extended by repeating the vascular Na substitution experiments with the additional replacement of Na in the luminal solution by Li. The replacement of luminal Na would have reduced the uptake of L-leucine very considerably, so the L-leucine was presented to the tissue as the dipeptide glycyl-l-leucine. The uptake of this peptide is not Na+-dependent (Cheeseman & Parsons, 1976). The earlier experiments had already shown that the exit of L-leucine from the tissue behaved in the same way whether the amino acid was loaded as peptide or as free amino acid, so it was considered that the two groups of experiments could be compared. In the absence of luminal Na+, the substitution of Na+ in the vascular solution with ions other than K+ also inhibited the L-leucine transfer (Table 4). DISCUSSION The biexponential nature of amino acid washout into the vascular bed confirms the earlier observations of Boyd et al. (1975a) and is clearly a similar phenomenon to that found for sugars by Boyd & Parsons (1978). A change in temperature affects both the fast and slow washout rate constants equally and the magnitude of the effect suggests that they do not just represent diffusion. However, flow dependence is only exhibited by the fast component rate constant K1 and the relationship over a wide range of vascular flow rates appears to be linear. Boyd & Parsons (1978) demonstrated a similar effect of vascular flow rate on the washout of the sugar 3-- methylglucose. Comparison of L-leucine and glycine washout Both amino acids have almost identical slow component washout rate constants when they are loaded into the tissue as an equimolar mixture, but the fast rate constant for glycine is only half that found for L-leucine. So it would appear that glycine cannot move as rapidly as L-leucine into the vascular bed from the epithelium, although glycine has a molecular weight only half that of the other amino acid. This difference between two neutral amino acids in their ability to leave the epithelium lends support to the proposal of Newey & Smyth (1962) that there is an exit mechanism for amino acids across the basolateral membranes. Munck & Schultz (1969) came to the conclusion that the ability of lysine to exit across the serosal surface of the rabbit ileum indicated the presence in the basolateral membrane of a system with a maximum rate of transport for the amino acid, and Danisi, Tai & Curran (1976) showed that the serosal exit of alanine in the same tissue was also a saturable process. Additional evidence for the existence of exit transport systems has come in a report from Boyd (1978) which showed that in the frog small intestine there is competition for exit into the vascular bed between the sugar x-methyl-d-glucoside and only neutral hydrophobic amino acids. This raises the possibility that there may be two or more such exit systems with different specificities for amino acids in the basolateral membranes of the intestinal mucosal epithelial cells. If this is the case, 465

10 466 C. I. CHEESEMAN then the movement of amino acids across the intestinal epithelium would appear to involve not only the well documented entry systems found in the brush border membrane (Wiseman, 1968) but also several exit systems in the basolateral membrane of the epithelial cells. It should also be noted that if the two amino acids were presented to the tissue in the form of the dipeptide glycyl-l-leucine their washout rate constants were very similar to those observed for the amino acid mixture. Thus, although the entry mechanisms for dipeptides and amino acids are distinct from one another (Cheeseman & Parsons, 1976), once inside the cell the amino acids appear to enter the same pools and exit to the vascular bed via common pathways. Analysis of pool sizes These experiments cannot distinguish whether the unloading of each of the two pools Sol and So2 is coupled or independent, and so any measure of their size using the rate constant can only be considered as an estimate (Huxley, 196). However, comparison of their sizes does conform with previous observations made using steady state techniques. When the tissue is loaded with a mixture of leucine and glycine, it appears that the Sol for leucine is larger than that for glycine, and Cheeseman & Parsons (1976) have previously observed that leucine is more readily taken up by the Anuran intestinal epithelium than glycine. But when the tissue is loaded with the dipeptide glycyl-l-leucine, the size of the Sol for glycine is increased relative to the So, for L-leucine. This would suggest that the two amino acids do not compete with each other for entry into the cell when they are presented to the brush border in the form of the dipeptide. Also, under most conditions, the rate constant for exit (K1) of glycine from the cells is significantly lower than that of L-leucine. This would account for the apparently higher concentration of glycine than of L-leucine found in the tissue during the steady-state transfer of both amino acids from the dipeptide. When Na is replaced in the luminal solution with Li+ but the vascular solution is unchanged, the washout of the amino acids remains unchanged compared to control values, but the pool sizes for the free amino acid mixtures are reduced, confirming the Na+ dependence of the entry systems (Schultz & Curran, 197). The loading of the tissue with dipeptide was not affected by the Na+ replacement in the lumen, con-. firming that peptide uptake in the Anuran small intestine is not Na+ dependent (Cheeseman, 1977a). Vascular Na+ replacement Steady-state transfer of L-leucine is reduced by substitution of K+ for Na+ in the vascular solution when Na+ is present in the lumen, but replacement with Li, choline or mannitol has no such effect. This implies that the inhibition is a consequence of the high K+ concentration rather than the reduced Na levels in the vascular bed. However, when Na is also absent from the luminal solution, then apparently Li and choline are also unable to substitute for Na and the exit of the amino acid is inhibited. Both sets of observations can be explained on the basis of the model proposed by Boyd et al. (1975b). If the tissue is largely depleted of Na by replacing the ion in both the vascular and luminal solutions, this would cause a reduction in the Na pumping across the basolateral membranes, and as a consequence a collapse of the

11 AMINO ACID WVASHOUT INTO VASCULAR BED lateral intercellular spaces. Recently, Gupta, Hall & Naftalin (1978) have shown that in the rabbit there is a high concentration of Na in the lateral intercellular spaces during solute (and solvent) transport which would produce a flow of water into these spaces. A reduction in water flow through the lateral intercellular spaces and their subsequent collapse would considerably hinder the movement of substrates from the cells to the vascular bed. Danisi et al. (1976) found no influence of Na in the incubation medium on the movement of L-alanine across the serosal border of the rabbit ileum. However, the movement was considerably inhibited by the addition of ouabain. The latter observation would support the view that Na pumping, which will keep the lateral intercellular spaces distended, is an important factor in the movement of substrates from the epithelium to the vascular bed. However, their inability to show an effect when Na is substituted in the bathing medium could be a consequence of the large unstirred layer at the serosal pole, as suggested by Boyd et at. (1975b). When Na is removed from the vascular bed but is still present in the lumen, there would still be a large enough supply of the ion to maintain intracellular levels, and so pumping could continue and keep the lateral intercellular spaces patent. But high concentrations of K+ in the capillaries will load the mucosal epithelial cells with this ion and cause cell swelling, which could in turn collapse the lateral intercellular spaces and produce the inhibition of transfer of substrate. Thus, replacement of Na in the vascular bed with K could be explained in terms of an effect of the substituting ion (K+) rather than the removal of the Na. Boyd & Parsons (1978) have suggested that during the washout, the fast component for sugars represents unloading of the enterocytes and their associated luminal unstirred layer, whereas the slow component involves the unloading of the underlying muscle. The data presented here would support such a contention inasmuch as the fast component could represent unloading of L-leucine from the epithelial cells through the lateral intercellular spaces into the vascular bed. However, there is no satisfactory explanation for the slow component of the amino acid washout. Conclusions Amino acids entering the intestinal epithelium either as free amino acids or as dipeptides exhibit the same characteristics for exit into the vascular bed. L-leucine appears to make this transfer more readily than glycine. The ionic composition of the solution in the vascular bed can affect the unloading of amino acids, and it is probable that this is a consequence of changing the intracellular ion concentrations, which will affect cell size or ion pumping rates, both of which could in turn alter the degree of opening of the lateral intercellular spaces. I would like to thank Dr S. Petersen for providing the curve stripping computer programme, and Drs N. Standen and P. Stanfield for helpful suggestions during the course of this work. I am indebted to N. Longford and D. Lilge for their valuable technical assistance, and to Dr M. Poznansky for his critical comments on the manuscript. Part of this work was supported by the University of Alberta Medical Research Fund. 467

12 468 C. I. CHEESEMAN REFERENCES BOYD, C. A. R. (1978). A classification of systems available for amino acid exit from small intestine into the blood. J. Physiol. 276, 52-53P. BOYD, C. A. R., CHEESEMAN, C. I. & PARSONS, D. S. (1975a). Amino acid movements across the wall of Anuran small intestine perfused through the vascular bed. J. Physiol. 25, BOYD, C. A. R., CHEESEMAN, C. I. & PARsoNs, D. S. (1975b). Effects of sodium on solute transport between compartments in intestinal mucosal epithelium. Nature, Lond. 256, BOYD, C. A. R. & PARSONS, D. S. (1978). Effects of vascular perfusion on the accumulation, distribution and transfer of 3--methyl-D-glucose within and across the small intestine. J. Physiol. 274, CARTER, G. W. & VAN DYE, K. (1971). A superior counting solution for water soluble tritiated compounds. Clin. Chem. 17, CHEESEMAN, C. I. (1977a). The role of sodium in the intestinal transport of small peptides. J. Physiol. 272, 63-64P. CHEESEMAN, C. I. (1977b). Studies on the fluxes of amino acids across the vascularly perfused Anuran small intestine. In Proceedings of the International Union of Physiological Sciences, vol. xiii, p CHEESEMAN, C. I. & PARSONS, D. S. (1976). The role of some small peptides in the transfer of amino nitrogen across the wall of vascularly perfused intestine. J. Physiol. 262, DANISI, G., TAi, Y.-H. & CURRAN, P. F. (1976). Mucosal and serosal fluxes of alanine in rabbit ileum. Biochim. biophys. Acta 455, FUJITA, M., PARSONs, D. S. & WOJNAROwsKA, F. (1972). Oligopeptidases of brush border membranes of rat small intestine mucosal cells. J. Physiol. 227, GUPTA, B. L., HALL, T. A. & NAFTALIN, R. J. (1978). Microprobe measurement of Na, K and Cl concentration profiles in epithelial cells and intercellular spaces of rabbit ileum. Nature, Lond. 272, HUXLEY, A. F. (196). Compartmental methods of kinetic analysis. In Mineral Metabolism, chap. 1, ed. COMMAR, C. L. & BRONNER, F. O., pp New York: Academic. MUNCK, B. G. & SCHULTZ, S. G. (1969). Lysine transport across isolated rabbit ileum. J. gen. Physiol. 53, NEWEY, H. & SMYTH, D. H. (1962). Cellular mechanisms in intestinal transfer of amino acids. J. Physiol. 164, SCHULTZ, S. G. & CURRAN, P. F. (197). Coupled transport of sodium and organic solutes. Physiol. Rev. 5, WISEMAN, G. (1968). Absorption of amino acids. In Handbook of Physiology, section 6, vol. 3, ed. CODE, C. F., pp Washington, D.C.: American Physiological Society.