L-threonine transport in pig jejunal brush border membrane vesicles. Functional characterization of the unique system B in the intestinal epithelium.

Uptake and inhibitory kinetics of [3H]L-threonine were evaluated in preparations of pig jejunal brush border membrane vesicles. Uptake of [3H]L-threonine under O-trans, Na+ gradient, and O-trans, Na(+)-free conditions was best described by high affinity transport (Km < 0.01 mM) plus a nonsaturable component. The maximal velocity of transport was 3-fold greater under Na+ gradient conditions. 100 mM concentrations of all of the dipolar amino acids and 2-aminobicyclo[2.2.1]heptane-2-carboxylic acid caused complete inhibition of [3H]L-threonine transport under Na+ gradient and Na(+)-free conditions. Imino acids, anionic amino acids, cationic amino acids, and methylamino-isobutyric acid caused significant partial inhibition of L-threonine uptake. Inhibitor concentration profiles for proline and lysine were consistent with low affinity competitive inhibition. The Ki values of alanine and phenylalanine approximated 0.2 and 0.5 mM, respectively, under both Na+ gradient and Na(+)-free conditions. These data indicate that the transport system available for L-threonine in the intestinal brush border membrane (system B) is functionally distinct from other amino acid transport systems. Comparison of kinetics parameters in the presence and absence of a Na+ gradient suggests that both partially and fully loaded forms of the carrier can function to translocate substrate and that Na+ serves to accelerate L-threonine transport by a mechanism that does not involve enhanced substrate binding.

cell types and are often referred to as "ubiquitous" transport systems. System A transports short chain linear dipolar amino acids, excludes phenylalanine, isoleucine, and valine as substrates, and can transport MeAIB' with high affinity (4, 5). System ASC has a high affinity for linear dipolar amino acids, such as alanine, serine, cysteine, and threonine, and does not transport MeAIB, phenylalanine, or histidine (5)(6)(7). Na+independent system L is a broad specificity transport system with a high affinity for leucine, phenylalanine, and methionine and a low affinity for the short chain amino acids glycine, alanine, and serine (5,8). System L transports BCH with high affinity and excludes proline and MeAIB as substrates (5,8,9). Considerable substrate overlap between dipolar amino acid transport systems has been demonstrated often within the same membrane. For example L-threonine is transported by all three systems in Chinese hamster ovary cells (5).
The pathway(s) of dipolar amino acids transport in the intestinal brush border membrane are less well characterized than in nonepithelial cell types. Stevens et al. (10) described a multiplicity of dipolar amino acid pathways in rabbit jejunal brush border membrane vesicles consisting of a broad specificity Na+-dependent "neutral brush border" system that did not transport MeAIB, a Na+-dependent, phenylalanine-preferring system, a Na+-dependent "imino" system, and a Na+independent system with a substrate specificity similar to that of system L as characterized in other cell types. Recently Maenz et al. (11) found that L-glutamic acid could be transported by the high affinity anionic amino acid transport system XAc and by a second low affinity system in rabbit jejunal brush border membrane vesicles. Flux of g glutamic acid through the low affinity system was found to increase with acidification of the media from pH 8 to 6 and could be blocked by excess phenylalanine. The property of pH-dependent conversion of substrate specificity to the uptake of anionic amino acids has been demonstrated for system ASC in several nonepithelial cell types (12,13). This finding implies that either system ASC is present in the intestinal brush border or that another brush border membrane dipolar amino acid transport system has the same functional property of pHdependent conversion to the uptake of anionic amino acids. In his review article of 1990 Christensen (1) speculated that the broad specificity amino acid transport system Bo,+ characterized in oocytes and blastocytes (2,14,15) may be present in the brush border membrane of intestinal and renal epithelial cells. Recently Stevens (16) has proposed that the neutral brush border system be renamed as system B to reflect the similarity in substrate specificity to system Bo,+. In addition to the broad specificity system(s) found in the intestinal brush border a novel broad specificity system (termed system G) able to recognize dipolar as well anionic and cationic amino acids has been found in the brush border membrane of the kidney epithelial cell line MDCK (17).
In this study we utilized a preparation of purified pig jejunal brush border membrane vesicles with a protocol of measuring rapid uptake kinetics of L-threonine to begin to clarify and characterize the pathways of dipolar amino acid transport in the intestinal brush border membrane.

EXPERIMENTAL PROCEDURES
Preparation and Storage of Jejunal Brush Border Vesicles-Brush border membrane vesicles (BBV) were prepared from pig jejunal mucosal scrapings using the Mg2' precipitation procedure (18) with some modifications. For each batch preparation of BBV, three pigs of 15-kg weight were induced and maintained under halothane anesthesia. 4 feet of proximal jejunum were surgically removed from each pig and rinsed immediately with ice-cold saline. The intestinal mucosa was scraped off with a spatula, and the scrapings were pooled and frozen at -70 "C. The next day the scrapings were thawed, weighed, and resuspended with homogenate media (50 mM mannitol, 2 mM Tris-HC1, pH 7.4) at a ratio of 20 ml of homogenate media per g of mucosal scrapings. The scrapings were homogenized for 60 s at setting 3 with a polytron homogenizer (Brinkman Instruments) and centrifuged for 15 min a t 2, OO X g. MgCl, was added to the supernatant (Sl) to give a final concentration of 10 mM, and the supernatant was stirred on ice for 15 min and then centrifuged for 15 min at 2,400 X g to pellet the precipitant. The resultant supernatant was centrifuged for 30 min at 19,000 X g to generate the crude BBV pellets. The supernatant was poured off, and a total of 6 ml of resuspension media (300 mM mannitol, 50 mM Hepes-Tris, pH 7.0) were added to the centrifuge tubes. The pellets were resuspended by repeated passage through a 26-gauge needle, pooled, divided into 0.25-ml aliquots, and frozen in liquid N2 until the day of use. For a given experiment a suitable number of aliquots of crude BBV were thawed in 70 ml of the resuspension media required for the particular experiment and homogenized in a glass-Teflon homogenizer (6 strokes) prior to centrifugation for 15 min at 500 X g. The resultant supernatant was centrifuged for 30 min at 39,000 X g to generate the final BBV pellets. The final pellets were resuspended with a 26-gauge needle, and a suitable volume of resuspension media was pooled, assayed for protein content, and diluted to give a vesicle concentration of 16 mg protein/ ml. The BBV were then divided into 20-p1 aliquots and frozen in liquid N, until the time of assay.
Uptake Assays-Initial rates ~f [~H ]~-t h r e o n i n e uptake were assayed "by hand" or with an automated fast sampling apparatus constructed by the engineering shops at the University of Saskatchewan following the principles of a similar apparatus first developed at the University of Montreal (19). ' In the by hand procedure an aliquot of frozen BBV was thawed, and a series of single time point uptakes were performed. For each uptake 5 p1 of BBV were rapidly mixed with 120 p1 of uptake media. At the conclusion of the uptake time period 1 ml of ice-cold stop solution (156 mM NaCl, 50 mM Hepes-Tris, pH 7.0,304 mM mannitol, 0.05 mM HgC12) was injected into the mixture. The stopped mixture was filtered through 0.45-pm cellulose acetate filters, the filters were washed twice with 5 ml of ice-cold stop solution, and the ['HI-labeled substrate content of the filters was determined by liquid scintillation counting. "Zero time" BBV substrate content was determined by injecting 5 pl of BBV into a mixture of 1 ml of ice-cold stop solution plus 120 p1 of uptake media. The mixture was filtered and the filters were washed and assayed for [3H]-labeled substrate content.
The automated fast sampling apparatus was constructed to test the validity of the by hand technique of measuring true initial rates of substrate uptake under steady-state conditions and, of greater significance, to provide an experimental tool that would allow for the precise measurements of steady-state and pre-steady-state uptakes that are impossible to achieve by hand and are necessary for the testing of more sophisticated models of transporter function. Multiple The fast sampling apparatus was constructed with the kind permission and cooperation of Dr. Alfred Berteloot of the Membrane Transport Research Group at the University of Montreal. time point assays were performed by loading the uptake media into the uptake chamber of the apparatus and then loading a positive displacement pipette containing an aliquot of vesicles into the chamber. Uptake was initiated by injecting the vesicles into the media and at each sampling time a pulse of air pressure was automatically imposed over the mixture such that 45 pl were expelled via a needle opening at the base of the chamber. Samples were collected in a series of sample wells which had been preloaded with 1 ml of ice-cold stop solution, and the stopped mixture was filtered through 0.65-pm cellulose acetate filters. The filters were washed five times with l-ml volumes of stop solution and assayed for [3H]-labeled substrate content.
Protein and Enzyme Assays-The purity and recovery of the brush border membrane fraction was determined by marker enzyme assays. Alkaline phosphatase was assayed according to Parkinson et al. (20) and NADPH cytochrome c reductase was assayed following the procedure of Sottocasa et al. (21). Protein was determined using the Sigma microprotein diagnostic kit with human serum albumin as the standard.
Data Analysis-Initial rates of [3H]~-threonine uptake were calculated as the slope of the regression line obtained for vesicle substrate content assayed at multiple time points within the first 2 s from initiation of uptake. The kinetic parameters of substrate uptake or inhibition were determined by nonlinear regression analysis of the nontransformed data using the P.Fit curve fitting program (Biosoft) as described by Malo and Berteloot (22), and the kinetic parameters of the best transport model fit to the data are reported in the figures.

RESULTS
Purity and Stability of Pig Jejunal BBV-The purity of the vesicle preparation was evaluated using marker enzymes known to be localized in specific membranes of the cell. Comparison of specific activities in the homogenate and the final vesicle preparation revealed a 21-fold enrichment in specific activity of alkaline phosphatase while the specific activity of NADPH cytochrome reductase in the vesicles was 0.3 times the specific activity found in the homogenate. These results indicate that the final preparation was highly enriched in brush border membrane with no evidence of significant contamination by basolateral membrane or cell organelles.
An important criteria in fitting models of substrate uptake to data obtained with membrane vesicles is that the preparation is stable and maintains a consistent uptake rate for any given set of conditions over the course of the experiment. Recently, Maenz et al. (23) demonstrated that standard preparations of rabbit and rat intestinal BBV will loose substantial substrate uptake capacity within 6 h from the time of preparation when maintained on ice or at room temperature. Dividing the preparation into aliquots suitable for individual uptake assays and freezing the aliquots in liquid N, until the time of assay resulted in stable uptake rates (23). We chose to follow the protocol of freezing aliquots of the BBV in liquid N, until the time of assay to avoid any possible loss of activity during the course of the experiments. By using this procedure the specific activity of substrate uptake was found to be the same in aliquots of BBV assayed immediately after preparation and in aliquots which had been frozen in liquid NP for 6 h from the time of preparation. In addition, the large batch preparation of partially purified vesicles showed no evidence of any loss of substrate uptake capacity even after 6 months of storage in liquid NZ.
Steady-state Initial Rates Measurements-Kinetic evaluation of transporter function using preparations of purified membrane vesicles are often limited by the crude methods employed in measuring uptake rates under steady-state conditions. The standard by hand procedure involves a single time point assay taken within the first few seconds from initiation of the reaction. Initial rates are subsequently cal-culated using an assumed zero time intercept and assuming that uptake rates are linear over the time span of the procedure. As such, the procedure does not directly measure true initial rates and could potentially contain systematic errors in defining steady state uptake kinetics. These systematic errors could introduce serious artifacts when attempting to fit transport models to the data (22).
In this study we employed a multiple time assay procedure and directly measured true initial rates as the slope of the linear portion of the uptake. The initial experiments were performed by hand with sampling at 0, 1, and 2 s from initiation of uptake. More recently we have made use of an automated fast sampling apparatus, constructed by the engineering shops at the University of Saskatchewan and based on the principles of an existing apparatus at the University of Montreal (19). The apparatus will perform precise and rapid multiple sampling of a vesicle uptake media mixture and thereby provide a direct and accurate measure of true initial rates of substrate uptake. Fig. 1 shows initial rates of ['HH]~-threonine uptake across purified pig jejunal brush border membrane vesicles obtained using the multiple time point assay performed by hand ( A ) or with the fast sampling apparatus ( B ) . Uptake rates were linear over the 2-s time period from initiation with a positive zero time intercept. The intercept is assumed to represent "background substrate association with the filters and/or a rapid association of substrate to the vesicle membrane that is resistant to the washing procedure. In either case the time course of this background binding is within the time scale of the first measurement of vesicle substrate content. True initial rates of substrate uptake were calculated as the slope of the regression line obtained from multiple time point assays of vesicle substrate content. It is worthy of note that indirect initial rate calculations based on a single time point assay with a presumed zero time intercept would have resulted in erroneously high estimates of steady state uptake rates. The intercept and the linear time course for substrate uptake obtained with the fast sampling apparatus is comparable to the results obtained by hand with sampling a t 0, 1, and 2 s from initiation. As such, the results obtained with the apparatus serve to validate the earlier results obtained using the by hand technique. The advantage of the fast sampling apparatus is in the precision of the measurements of initial rates which allows for greater discrimination in model fitting to the data.
Kinetics of PHIL-Threonine Uptake in the Presence of a Nu' Gradient- Fig. 2 shows a direct plot of initial rates of 0.5 p~ ['3H]~-threonine uptake in the presence of an inwardly directed Na' gradient with various concentrations of unlabeled L-threonine incorporated in the uptake media. A model for L-threonine uptake of high affinity transport plus an unsaturable diffusion component converged to the data within the parameters of the computer program. More complex uptake models such as multiple transport pathways with widely separate K, values did not converge. The data suggest that L-threonine transport across the vesicles in the presence of a Na' gradient occurs by a high affinity system(s) with a K,,, of less than 0.1 mM and a V,,, approximating 14 pmol/mg/s with no evidence of a second low affinity route. The inset plot in Fig. 2 shows an Eadie-Hofstee transformation of the transport-mediated component of total uptake rates at nonsaturating concentrations of unlabeled L-threonine. The transformed data approximate a linear relationship between transport rate versus transport rate/substrate concentration and provide no visual evidence of multiple pathways with widely separate K, values. were performed using the fast sampling apparatus. For each uptake two aliquots of vesicles were thawed and pooled, and 33 p1 were taken up by the injection pipette and loaded into the apparatus along with 767 p1 of uptake media. Uptake was initiated by injecting the vesicles into the uptake media. The apparatus automatically sampled the mixture at the indicated times from initiation and vesicle substrate content was determined as described in "Experimental Procedures." Like symbols represent vesicle substrate content taken from the same uptake at the indicated times points.
Inhibitor Specificity under Nu' Gradient Conditions- Fig.  3 shows the effect of 100 mM concentrations of various amino acids and analogs on the initial rate of 0.5 p~ [3H]~-threonine uptake in the presence of a Na' gradient. All of the amino acids and analogs caused a significant inhibition of uptake in comparison to the mannitol control condition. Incorporation of 100 mM unlabeled L-threonine in the uptake media would cause complete inhibition of the transport-mediated component of total ['H]~-threonine uptake (Fig. 2). As such, 100 mM concentrations of all of the dipolar amino acids and BCH caused complete inhibition of [3H]~-threonine transport in the vesicles. MeAIB and the cationic amino acids lysine and arginine caused a slight but significant inhibition that approximated 25% of total [3H]~-threonine transport. The imino acids proline and OH-proline and the anionic amino acids glutamate and aspartate were somewhat more potent partial inhibitors of ['H]L-threonine transport.
Concentration Profiles for Lysine and Proline Inhibition of L-Threonine Uptake-Partial inhibition of ['H]L-threonine uptake by excess concentration of lysine and proline could represent a low affinity for inhibitor interaction with the 'stinal Brush Border System B transport system or a block of substrate uptake through a second lysine and proline sensitive system. In the case of a low affinity of the inhibitors for the transport system one would expect to find no evidence of saturation of the inhibitory effect with increasing concentrations of inhibitor in the media. A saturation of the partial inhibitory effect would be consistent with blockage of substrate flux through a second system with a high affinity for the inhibitors. Fig. 4 shows the effects of increasing concentrations of lysine and proline on the initial rate of 1. ml of frozen P,, and initial rates of substrate uptake were determined using the fast sampling apparatus as described in Fig. 3. The uptake media contained the indicated concentration of test agent + mannitol to give a final concentration of 104 mM. The points represent the mean and standard deviation obtained from four replicate initial rate determinations.

L-Threonine Transport by
Intestinal Brush Border System B 22083 threonine uptake versus log L-threonine concentration in the absence of Na+. The best model fit to the data is that of high affinity transport plus diffusion. The parameters of K, and diffusion are comparable to those obtained in the presence of a Na+ gradient while the V,,,, approximated 5 pmol/mg/s. An Eadie-Hofstee transformation of the transport-mediated component of total uptake provided no visual evidence of more complex transport models involving multiple systems with widely separate K , values.
Inhibitor Specificity under Nu+-free Conditions-The specificity of inhibition of ['H]~-threonine uptake under Na+-free conditions by amino acids and analogs is shown in Fig. 6. The results are comparable to those obtained under Na+ gradient conditions in that all of the dipolar amino acids and BCH caused complete inhibition of the transport-mediated component of total substrate uptake. Under these conditions, MeAIB and the cationic amino acids lysine and arginine had no significant effects on the rate of ["H]~-threonine uptake. The imino acids, proline and OH-proline, and the anionic amino acids, glutamate and aspartate, did cause significant partial inhibition of ["H]L-threonine uptake.
Kinetics of Alanine Inhibition of PH]~-Threonine Uptake- Fig. 7 shows the rates of ["H]L-threonine uptake in the presence of varying concentrations of L-threonine with 0 , 0.05, or 0.50 mM alanine incorporated in the media under 0trans, Na+-gradient conditions. In the absence of alanine the best model fit to the data was that of high affinity transport system plus a nonsaturable component. Incorporating alanine in the uptake media caused an apparent increase in the K, with no change in the maximal velocity of transport. These results are consistent with a model of competitive inhibition for substrate binding to the transport system. A model of high affinity ["H]L-threonine transport with competitive inhibition by alanine did converge to the data within the 95% confidence All procedures were the same as described for Fig. 3 except that the uptake media contained 156 mM ChCl in place of NaCl. The bars show the mean and standard deviation obtained from four uptakes and a common number indicates no statistically significant difference in mean uptake rates ( p = 0.05). L-threonine uptake under 0-trans, Na+-free conditions. All procedures were the same as described in Fig. 8 except that phenylalanine replaced alanine as the competitive inhibitor of [3H]~-threonine transport and the uptake media contained 156 mM ChCl in place of NaCl. mM and 0.610 k 0.024 mM obtained for 0.1 mM and 1.0 mM phenylalanine, respectively.

DISCUSSION
A detailed characterization of transport mechanisms using membrane vesicles requires a stable preparation of purified membrane and an accurate methodology for measuring initial rates of substrate uptake under steady-state conditions. The purity of intestinal brush border vesicle preparations prepared by standard divalent cation precipitation procedures can be considered as adequate for transport experiments given the enrichment in brush border marker enzyme activities combined with little apparent contamination by other cellular organelles (24, 25). The results of marker enzyme assays in this and in other studies (26) using pig jejunal brush border membrane vesicles are consistent with the purifications obtained using similar procedures in other species (24, 25). In our experiments brush border vesicles were prepared from a partially purified batch preparation, divided into aliquots suitable for transport assays, and then frozen in liquid NP until the time of assay. This procedure ensures a stable uptake activity for any given set of conditions and thus avoids any possibility of artifact resulting from a decline in the specific activity of substrate uptake as has been shown to occur in rabbit and rat intestinal vesicles maintained "on-ice'' from the time of preparation (23). Our approach in evaluating the kinetics of steady-state substrate uptake across the vesicles was to directly calculate true initial rates as the slope of the regression line obtained from vesicle substrate content assayed at multiple times within the first 2 s from initiation of the reaction. This procedure avoids any assumptions as to the zero time intercept and the time course of the reaction that are inherent in single time point assays of initial rates (22). The recent development of an automated fast sampling apparatus in our laboratory allows for precise and rapid sampling of a vesicle/uptake media mixture. The improved precision of initial rate determination allows for greater discrimination in fitting models of substrate uptake to the data (19,22).
Using a protocol of measuring true initial rates of [ 3 H ]~threonine uptake across a stable preparation of purified pig jejunal brush border vesicles we have begun to characterize the pathway(s) available for dipolar amino acid transport in the intestinal brush border membrane. In evaluating initial rates of [3H]~-threonine uptake with varying concentrations of unlabeled L-threonine in the media, a model of high affinity transport (K,,, < 0.1 mM) plus a nonsaturable component provided the best model fit to the data under 0-trans, Na+ gradient, and 0-trans, Na+-free conditions. An evaluation of transport kinetics in itself cannot be used to discriminate between one versus two or more transport systems. Substrate uptake rates obtained in vesicle systems containing multiple high affinity transport routes with similar K,,, values would converge with models of uptake defined by a single transport system. These results do however tend to exclude significant transport of ['HIL-threonine by low affinity routes in the intestinal brush border membrane.
The broad specificity of inhibition of threonine uptake by excess concentrations of amino acids and analogs suggests that the high affinity transport route for L-threonine shows marginal selectivity among amino acid substrates. All of the dipolar amino acids and BCH caused complete inhibition of transport while the anionic and cationic amino acids, the imino acids, and MeAIB caused significant partial inhibition under Na' gradient conditions. One could speculate that threonine could function as a high affinity substrate for the anionic amino acid system (ll), the "imino" system (lo), and lysine-preferring, Na+-independent pathway (10) described in the brush border membrane. However the inhibition profiles obtained with varying concentrations lysine and proline are not consistent with saturable high affinity partial inhibition of threonine' flux through these alternate pathways. As such, available evidence indicates that anionic and cationic amino acids and imino acids likely function as low affinity competitive inhibitors of threonine transport. In rabbit jejunal brush border vesicles L-glutamic acid is transported by a specific high affinity system and a second low affinity phenylalanine-sensitive system (11).
In this study the K , values obtained for ['H]~-threonine transport were found to be remarkably similar under Na+ gradient and Na+-free conditions. This finding is compatible with either of two transport models. In the first model distinct high affinity Na+-dependent and Na+-independent transport systems function to catalyze threonine uptake in the vesicles. However in this study the kinetics of inhibition of threonine uptake by amino acids and analogs were remarkably similar under Na+ gradient and Na+-free conditions, and as such we found no functional evidence of two distinct transport systems. An alternative model is that of a single high affinity system functioning under Na' gradient and Na+-free conditions. If L-threonine is transported across the intestinal brush border membrane via a single high affinity system, this transporter could operate via an ordered or a random substrate binding mechanism. A strictly ordered Na' first, L-threonine second mechanism is unlikely given that L-threonine transport occurs in the absence of Na' (27). An ordered L-threonine first, Na' second mechanism with slippage (translocation of the partially loaded form of the carrier) is consistent with the data. Alternatively L-threonine and Na+ could bind to the transporter via a random mechanism with slippage of the partially loaded form of the carrier (27, 28). A generalized "preferred random" model for Na+-amino acid transport systems with provision for slippage of the partially loaded carrier has recently been put forward by Stevens (16, 28). Further experiments under equilibrium exchange conditions in the presence and absence of Na' may help to distinguish between a random versus ordered substrate binding mechanism (27).
Na' could serve as an activator of L-threonine transport by increasing the affinity for substrate binding, by increasing the overall rate of transporter cycling in the membrane, or by a combination of both kinetic mechanisms. The similarity in K,,2 and K, values coupled with the %fold difference in maximal velocity under Na+ gradient versus Na'-absent conditions provides good evidence that Na' does not stimulate L-threonine transport through an allosteric effect on the substrate binding site. These results imply a velocity effect whereby the overall rate-limiting step in the process of transporter recycling is accelerated in the presence of a 0-trans, Na+ gradient relative to 0-trans, Na+-free conditions. Inhibition of threonine transport by phenylalanine, histidine, isoleucine, and valine and marginal inhibition by MeAIB indicate that the Na'-dependent dipolar amino acid transport systems A and ASC do not function to transport threonine in the intestinal brush border membrane. This conclusion concurs with the earlier work of Stevens et al. (10) who found no functional evidence of systems A and ASC activity in rabbit jejunal brush border membrane vesicles. Recently Christensen (1) speculated that the broad specificity neutral brush border system originally described by Stevens et al. (10) may be similar if not identical to the broad specificity Na+-dependent system B".' found in oocytes and blastocytes (2, 14, 15) and the broad specificity system G (general) described in the kidney epithelial cell line MDCK (17). A characteristic of system B".' in mouse blastocytes and of system G in MDCK cells is a high affinity for dipolar and cationic amino acids. In this study lysine functioned as a very low affinity competitive inhibitor of threonine transport suggesting that systems Bo.+ and G do not catalyze threonine uptake in the intestinal brush border membrane. Recently Stevens (16) has renamed the neutral brush border system as system B to reflect the ontogenetic and substrate specificity relationship to system BO,' . In this study the high affinity transport system for L-threonine was similar in substrate specificity to system Bo.' with the notable exception that lysine was a poor inhibitor of Lthreonine transport. Hence the term system B does serve to indicate the existence of a unique dipolar amino acid transporter in the intestinal brush border and yet reflect the similarity of this system to system Bo.+ as characterized in oocytes and blastocytes.
L-Threonine transport under Na+-free conditions in pig jejunal brush border vesicles was completely inhibited by excess concentrations of the full spectrum of dipolar amino acids and BCH while proline and MeAIB were poor inhibitors of threonine transport. This pattern of inhibition is consistent with the known substrate specificity of Na+-independent system L found in most metabolically active cells (5, 8, 9), and it has been suggested that system L is functional in rabbit jejunal brush border vesicles (10). It is interesting to speculate that the Na+-independent pathway for dipolar amino acid transport in the intestinal brush border membrane that has been attributed to system L may represent substrate flux through system B cycling via translocation of the partially loaded form of the carrier. This hypothesis would be consistent with the inhibitor specificity of threonine transport found in this study and the substrate specificity of the broadspectrum Na+-independent dipolar amino acid transporter described in rabbit jejunal brush border vesicles (10).
Transport of L-threonine in the kidney epithelial cell brush border membrane shows no similarity to L-threonine transport in the intestinal brush border. Kuhlmann and Vadgama (29) found functional evidence for high and low affinity Na+dependent transport systems plus two Na+-independent systems with inhibitory specificity indicative of systems L and ASC. Boerner et al. (17) describe dipolar amino acid transport in the apical membrane of early confluent cultures of the kidney epithelial cell line MDCK as occurring through systems A, ASC, L, and the novel broad specificity system G. In this study we found no evidence for such a complex spectrum of pathways for L-threonine transport in intestinal brush border vesicles.
This study and that of Stevens et al. (10) indicate that the intestinal system B is functionally distinct from any other amino acid transport system found in other cell types. L-Threonine would appear to be an ideal substrate for the further characterization of system B in that we found no evidence of L-threonine transport by additional system such as the "Phe" and "Imino" systems of the intestinal brush border membrane (10). Our work suggests that system B has a broad specificity for dipolar amino acids and BCH and may function via a mechanism that includes provisions for translocation of both the fully and partially loaded forms of the carrier. In future more exacting kinetic evaluations of transport rates with varying concentration of Na' and substrate under 0-trans and equilibrium exchange conditions will be used to further characterize the transport mechanism of system B in the intestinal brush border membrane.