Amino acid fluxes in rat thin limb segments of Henle s loop during in vitro microperfusion

Amino acid fluxes in rat thin limb segments of Henle s loop during in vitro microperfusion OLGA H. BROKL AND WILLIAM H. DANTZLER Department of Physiology, College of Medicine, University of Arizona, Tucson, Arizona 85724-5051 Brokl, Olga H., and William H. Dantzler. Amino acid fluxes in rat thin limb segments of Henle s loop during in vitro microperfusion. Am. J. Physiol. 277 (Renal Physiol. 46): F204 F210, 1999. Amino acids are apparently recycled between loops of Henle and vasa recta in the rat papilla in vivo. To examine more closely papillary amino acid transport, we measured transepithelial fluxes of L-[ 14 C]alanine and [ 14 C]taurine in thin limbs of Henle s loops isolated from rat papilla and perfused in vitro. In descending thin limbs (DTL) in vitro, unidirectional bath-to-lumen fluxes tended to exceed unidirectional lumen-to-bath fluxes for both radiolabeled amino acids, although the difference was statistically significant only for taurine. In ascending thin limbs (ATL) in vitro, unidirectional lumen-to-bath fluxes tended to exceed unidirectional bath-to-lumen fluxes, although the difference was again statistically significant only for taurine. These results are compatible with apparent directional movements of amino acids in vivo. However, none of the unidirectional fluxes was saturable or inhibitable, an observation compatible with apparent reabsorption from the ATL in vivo but not compatible with apparent movement from vasa recta to DTL in vivo. There was no evidence of net active transepithelial transport when concentrations of radiolabeled amino acids were matched on both sides of perfused tubule segments. These data suggest that regulation of amino acid movement in vivo may involve the vasa recta, not the DTL of Henle s loops. The data also suggest that transepithelial movement of amino acids in thin limbs of Henle s loop may occur via a paracellular route. L-alanine; taurine; descending thin limb; ascending thin limb; isolated tubules 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. AMINO ACIDS FILTERED at mammalian renal glomeruli are almost completely reabsorbed along the proximal tubules (10). However, small amounts do remain at the end of the accessible portion of the proximal tubule and these appear to be transported by more distal portions of the nephrons. In fact, for a number of acidic, basic, and neutral amino acids (including the sulfur-containing amino acid, taurine), our previous free-flow micropunctures and continuous microinfusions of ascending loops of Henle in the exposed papilla of the rat kidney indicate that reabsorption can occur beyond the tip of the loop of juxtamedullary nephrons, probably from ascending thin limbs (ATL) (5 8, 11). This reabsorption is generally not saturable, inhibitable (with the possible exception of taurine) (11), or stereospecific (5 8). However, it is not a simple nonspecific leak, because it does not occur for mannitol (6). In addition, our microin- fusions of ascending vasa recta indicate that amino acids can leave the ascending vasa recta and enter ipsilateral papillary tubular structures, probably descending thin limbs (DTL) of Henle s loops, as intact amino acids without first entering the systemic circulation (6 8). This movement from vasa recta to ipsilateral tubules is inhibitable (6 8). It is also stereospecific for some (e.g., glutamate) but not other (e.g., alanine) amino acids (6 8). These observations, together with our micropuncture findings of high fractional deliveries of endogenous amino acids to the tips of Henle s loops (sometimes 1.0), higher amino acid concentrations in the ascending than in the descending vasa recta at the same level of the papilla, and high mean amino acid concentrations in the vasa recta (5), suggest that amino acids can be recycled among the loops and the vasa recta in the papilla (5 8). For some amino acids, such as taurine, which can apparently serve as an osmolyte in the renal medulla (9), this recycling may be important in maintaining appropriate medullary concentrations for this function. Despite this body of in vivo data, the exact sites or characteristics of amino acid movements in the thin limbs of Henle s loops, if indeed they occurred in this region, were not known. To examine these questions more directly than was possible in vivo, we undertook the task of isolating and perfusing in vitro DTL and ATL from the renal papilla of rats of the same strain as those used in the in vivo studies. For this initial study, we characterized the thin limbs and examined the movements of the neutral amino acid, L-alanine, and the sulfur-containing amino acid, taurine. The unidirectional fluxes were generally compatible with the in vivo observations, but inhibition data were not. METHODS Animals and dissection of tubules. Male Munich- Wistar rats (80 g average body wt) were treated with furosemide (1.5 mg/100 g body wt) by peritoneal injection to induce diuresis and consequently to lower the medullary osmolality (2). Animals were anesthetized with pentobarbital sodium (0.2 ml/100 g body wt). Their kidneys were removed and bathed on ice in 280 mm sucrose-10 mm HEPES buffer, brought to ph 7.4 with Tris base and gassed with 100% O 2. The medullary portions were dissected from the kidneys, and individual bundles were teased out along the length of the inner medulla under a stereomicroscope. Segments of thin limbs were dissected without the aid of enzymatic agents using reflected light from below the dissection dish. We identified dissected segments by cell type using an inverted microscope with Nomarski differential-interference contrast optics at 400 magnification (2). The DTL with their cells with nuclei protruding F204 0363-6127/99 $5.00 Copyright 1999 the American Physiological Society

AMINO ACID FLUXES IN HENLE S LOOP IN VITRO F205 into the lumen differ markedly in appearance from the ATL with their large, round, flat nuclei (Fig. 1). The upper portion of the DTL within the inner stripe of the outer medulla has a greater outer diameter and a more tortuous lumen than the inner medullary portion as the cells increase in size toward the proximal S3 segment of the outer stripe. Within the inner medulla, the narrowing tubule diameter and the increasing density of cells per unit tubule length distinguish the lower (DTL L ) from the middle (DTL M ) portion of the DTL (Fig. 1). For purposes of perfusion, we initially characterized the DTL M as being from the upper half and the DTL L as being from the lower half of the inner medulla down to the prebend region of the tubule (Fig. 1). In later experiments, we stopped separating the DTL into DTL M and DTL L portions for perfusion. However, most of each segment that we used for perfusion consisted of the DTL M portion. The long ATL also increases somewhat in outer diameter along its length from deep in the inner medulla toward the outer medulla, but it does not change in cell appearance or cell density. At the junction of the inner and outer medulla, the long ATL changes abruptly into the thick ascending limb. For purposes of perfusion, we used ATL segments from the Fig. 1. Differential-interference contrast images of thin limb segments from papillary region of Munich-Wistar rat kidney. Magnification is the same for all images ( 400, as viewed through the microscope). Images are shown next to a diagram of the appropriate regions of the thin limb. DTL M, middle descending thin limb; DTL L, lower descending thin limb; ATL, ascending thin limb. Note nuclei of DTL cells protruding into lumen and round, flat nuclei in ATL cells. top 70% of the inner medulla above the prebend level. The average lengths and outer diameters of DTL and ATL segments used in perfusion are given in Table 1. While studying amino acid fluxes in isolated DTL and ATL, we identified a significant population of nephrons with both DTL and ATL cell types in the same segment (Fig. 2). The regions of DTL-type cells and ATL-type cells varied from a spot region of a few cells of the opposite type to several regional repetitions of alternate cell types to as much as 50% ATL and 50% DTL cell types in a single dissected segment. We used only segments in which all, or almost all ( 90%), of the Table 1. Lengths and outer diameters of tubule segments used for perfusion Tubule Segment Length, µm Diameter, µm DTL M 982 64.0 (17) 20.8 0.65 (17) DTL L 871 72.6 (13) 15.0 0.61 (17) DTL 1082 37.1 (45) 19.6 0.28 (45) ATL 1043 43.6 (47) 19.2 0.34 (47) Values are means SE; numbers in parentheses indicate number of tubules. DTL M, middle descending thin limb; DTL L, lower descending thin limb; DTL, descending thin limb; and ATL, ascending thin limb.

F206 AMINO ACID FLUXES IN HENLE S LOOP IN VITRO Fig. 2. Segment of tubule showing both DTL and ATL cell types. Cells in most of the segment are of the DTL type. However, the region between the arrows has cells of the ATL type. cells were of the one specific type for the amino acid flux measurements in the present study. Perfusion and bathing solution. The Ringer solution used for perfusing and bathing the tubules was that initially used by Chou and Knepper (2) in perfusion of thin limbs from chinchilla kidneys. It contained (in mm) 118 NaCl, 25 NaHCO 3, 2.5 K 2 HPO 4, 2 CaCl 2, 1.2 MgSO 4, 5.5 glucose, and 5 urea. The osmolality was about 290 mosmol/kgh 2 O, the ph was adjusted to 7.4, and the solution was continuously gassed with 95% O 2-5% CO 2 to maintain a stable ph and adequate oxygenation. When a high concentration of unlabeled amino acid was added to the perfusate or bathing medium, an equimolar concentration of sucrose was added to the other solution to keep the osmolalities and ionic strengths on the opposite sides of the epithelium equal. Perfusion of tubules. We perfused the dissected tubules in vitro by a technique essentially the same as that originally described by Burg et al. (1) and modified for use in our laboratory (3, 4). Briefly, each tubule was transferred to a special temperature-controlled Lucite bathing chamber with a glass bottom. The ends of the tubule were held in glass micropipettes and the tubule was perfused through a micropipette with its tip ( 7 µm diameter) centered in the tubule lumen. The perfusion rate was normally 15 20 nl/min. The temperature of the bathing chamber was maintained at 37 C during perfusion. To check for leaks and to be certain that there was no net fluid movement across the perfused epithelium, [ 3 H]polyethylene glycol (PEG) (mol wt 4,000) was added to the perfusate or bathing medium as appropriate. Tubules with a PEG leak 0.15% per minute or apparent net transepithelial fluid movement were discarded. The average rate of PEG leak across the epithelium for the experiments used was 0.09 0.03% per minute. Measurement of unidirectional transepithelial fluxes of amino acids. Because there was no net transepithelial fluid movement in these perfused tubules, we determined the lumen-to-bath (reabsorptive) flux of 14 C-labeled amino acids (J lb, fmol min 1 mm tubule length 1 ) from the disappearance of radiolabeled amino acid from the lumen using the following relationship J lb ([ 14 C] p [ 14 C] cf )(V o )/(sp act p )(L) (1) In Eq. 1, [ 14 C] p and [ 14 C] cf are the disintegrations per minute per nanoliter (dpm/nl) of 14 C-labeled amino acid in the initial perfusate and the collected fluid, respectively; V o is the fluid collection rate (nl/min), measured directly as described previously (3, 4); sp act p is the specific activity of the 14 C-labeled amino acid in the perfusate; and L is the length of the tubule as measured directly with an ocular micrometer. Because there was no net transepithelial fluid movement, V o was the same as the initial perfusion rate (V i ), and thus only this simple equation involving V o needed to be used. Because the concentration of amino acid developed in the bathing medium during the course of an experiment was insignificant compared with that in the perfusate, backflux from bath to lumen was negligible and could be ignored. Also, because the bathing medium contained no amino acid, exchange diffusion was eliminated. The bath-to-lumen (secretory) flux of 14 C-labeled amino acid (J bl, fmol min 1 mm tubule length 1 ) was determined from the appearance of radiolabeled amino

AMINO ACID FLUXES IN HENLE S LOOP IN VITRO F207 acid in the collected fluid when radiolabeled amino acid was present in the bath but not in the initial perfusate. The determination was made from the following relationship J bl (V o [ 14 C] cf )(sp act b )(L) (2) In Eq. 2, spact b is the specific activity of 14 C-labeled amino acid in the bath, and the remaining symbols have the meanings given above. In these experiments, the perfusion rate of about 15 20 nl/min kept the concentration of amino acid in the lumen low and prevented significant backflux from lumen to bath or significant transepithelial exchange. In all flux experiments, the collection periods were 5 min in length. A period of 10 min was allowed at the beginning of each experiment and following each change of solutions to permit a steady state to be attained before the start of the experimental collections. Determination of radioactivity. The activity of the 14 C was determined by counting in a liquid scintillation system. The scintillation fluid was Ecolite (ICN Biomedicals). Chemicals. L-[ 14 C]alanine (sp act 163.85 mci/mmol), [ 14 C]taurine (sp act 108.5 mci/mmol), and [ 3 H]PEG (sp act 1.51 mci/g) were obtained from Du Pont-NEN. All other chemicals were purchased from standard sources and were of the highest purity available. Statistical analysis. Results are summarized as means SE. The n value is the number of tubules. Each tubule came from a different animal. In unpaired experiments, three to five flux measurements were made in each tubule, and the mean value for these measurements was used as the value for that tubule. Unpaired comparisons of fluxes between two groups of tubules were assessed with Student s t-test. For paired analyses of unidirectional fluxes in single tubules, three to five measurements of one unidirectional flux were made, and the mean value was determined followed by three to five measurements of the other unidirectional flux for which a second mean value was determined. The paired means were then compared for a series of tubules. Similarly, in inhibition studies, three to five control fluxes and three to five experimental fluxes were made in a single tubule, and the means were compared. The significance of the difference between paired means was assessed with a two-tailed t-test for paired observations. Differences for both paired and unpaired analyses were considered statistically significant when P 0.05. RESULTS Unidirectional transepithelial L-alanine fluxes. Initially, we measured either J bl or J lb for L-[ 14 C]alanine (8 µm) in separate segments of DTL and ATL of Henle s loop. In this initial series of experiments, we also divided the DTL into DTL M and DTL L portions, as indicated in METHODS. The results are shown in Table 2. In the DTL M segments, J bl ( 125 fmol min 1 mm 1 ) was significantly greater than J lb ( 45 fmol min 1 mm 1 ). However, in the DTL L segments, J bl and J lb Table 2. Unidirectional L-[ 14 C]alanine fluxes in separate tubules DTL M DTL L ATL J bl (measured), fmol min 1 mm 1 125 25* (7) 45 10 (4) 19 11 (4) J lb (measured), fmol min 1 mm 1 44 7* (4) 45 10 (6) 69 18 (5) J bl J lb (calculated), fmol min 1 mm 1 81 0 50 Values are means SE; numbers in parentheses indicate number of tubules. For J bl J lb, positive value indicates secretion and negative value indicates reabsorption; no SE is given because the unidirectional fluxes were measured in separate tubules and these values are simply the differences between the means. Values in each vertical column that are significantly different from each other by unpaired analysis have the same superscript. J bl, bath-to-lumen flux; J lb, lumen-to-bath flux. were essentially the same ( 45 fmol min 1 mm 1 ). In the ATL segments, J lb ( 70 fmol min 1 mm 1 ) was significantly greater than J bl ( 20 fmol min 1 mm 1 ). These observations (J bl greater than J lb suggesting secretion in the DTL; and J lb greater than J bl suggesting reabsorption in the ATL) are in accord with what appeared to be the dominant direction of the L-alanine fluxes in our in vivo microinfusion studies (7). However, because fluxes were measured in only one direction in any given perfused segment, it appeared possible that the significant differences reflected differences between the segments perfused, not true differences in the direction of the fluxes. Therefore, we measured both J bl and J lb in single perfused segments of DTL and ATL. In this series of studies, we no longer divided the DTL into two portions. However, most of any given DTL segment reflected the area previously represented by DTL M. Because it was difficult to have exactly the same concentration of L-[ 14 C]alanine in each experiment, the concentration varied from 8 to 20 µm in different experiments. However, in each individual experiment, the same solution containing the same concentration of L-[ 14 C]alanine was used for measuring both J bl and J lb. The sequence in which J bl and J lb were measured was alternated from tubule to tubule in these experiments. The results are shown in Table 3. Although J bl was greater on average than J lb in the DTL and J lb was greater on average than J bl in the ATL, the Table 3. Unidirectional L-[ 14 C]alanine fluxes in the same individual tubule DTL ATL J bl (measured), fmol min 1 mm 1 94.1 17.3 (10) 64.3 9.8 (12) J lb (measured), fmol min 1 mm 1 79.6 18.8 (10) 88.6 16.9 (12) J bl J lb (calculated), fmol min 1 mm 1 14.5 8.7 (10) 24.3 15.1 (12) Values are means SE; numbers in parentheses indicate number of tubules. For J bl J lb, positive value indicates secretion and negative value indicates reabsorption; SE is given because unidirectional fluxes were measured in the same tubule and values are means for the paired differences; the mean difference was not statistically different from zero for either limb.

F208 AMINO ACID FLUXES IN HENLE S LOOP IN VITRO Table 4. Unidirectional [ 14 C]taurine fluxes in the same individual tubule Experiment Fluxes, fmol min 1 mm 1 J bl J lb J bl J lb Descending thin limb 1 509 430 79 2 445 423 22 3 587 527 60 4 375 252 123 5 308 301 7 6 192 128 64 Mean SE 403 58 (6) 344 59 (6) 59 17 (6) Ascending thin limb 1 134 169 35 2 261 310 49 3 127 323 196 4 269 275 6 5 173 304 131 6 400 624 224 Mean SE 227 43 (6) 334 62 (6) 107 37 (6) Mean difference between fluxes was significantly different from zero for both descending and ascending limbs; values in parentheses are numbers of tubules. For J bl J lb, positive values indicate secretion and negative values indicate reabsorption. variation was such that, even with paired analysis, the difference was not statistically significant by the standard set in methods. P 0.1 in both cases. Unidirectional transepithelial taurine fluxes. In the case of taurine, we measured both J bl and J lb for [ 14 C]taurine in single perfused segments of DTL and ATL. The concentration of [ 14 C]taurine varied from 20 to 40 µm in different experiments. However, in each individual experiment, the same solution containing the same concentration of [ 14 C]taurine was used for measuring both J bl and J lb. As in the case of the alanine fluxes, the sequence in which J bl and J lb were measured was alternated from tubule to tubule in these experiments. The results are shown in Table 4. The average values for both unidirectional fluxes for [ 14 C]taurine (Table 4) were significantly larger than those for L-[ 14 C]alanine (Table 2). This was not simply the result of the higher concentration of taurine than L-alanine used in these flux measurements, because it held even when the concentrations of [ 14 C]taurine and L-[ 14 C]alanine were the same ( 20 µm). In the case of the unidirectional [ 14 C]taurine fluxes measured in the same tubule, J bl was significantly greater than J lb in the DTL and J lb was significantly greater than J bl in the ATL (Table 4). These data suggesting secretion in the DTL and reabsorption in the ATL are in agreement with what appeared to be the dominant direction of the taurine fluxes in our in vivo microinfusion studies (7, 11). Attempts to inhibit unidirectional fluxes. We examined possible inhibition of the unidirectional fluxes of L-[ 14 C]alanine by adding unlabeled L-alanine to the bathing medium or perfusion solution. The addition of 10 or 50 mm unlabeled L-alanine to bathing medium containing 20 µm L-[ 14 C]alanine actually tended to increase, not decrease, J bl of L-[ 14 C]alanine in both the DTL and ATL, but the change was not statistically significant (0.10 P 0.20). In four experiments, we also examined the effect of adding 50 mm L-phenylalanine to the bath containing 20 µm L-[ 14 C]alanine on J bl of L-[ 14 C]alanine in the DTL. In proximal tubules, L-phenylalanine inhibits reabsorption of numerous neutral amino acids (10). Moreover, it inhibits the transfer of radiolabeled L-alanine from ascending vasa recta to ipsilateral tubular structures in vivo (7). However, L-phenylalanine also had no significant effect on J bl of L-[ 14 C]alanine (0.20 P 0.30). The addition of 10 mm unlabeled L-alanine to perfusion medium containing 20 µm L-[ 14 C]alanine also had no significant effect on J lb of L-[ 14 C]alanine in either the DTL (0.50 P 0.60) or the ATL (0.30 P 0.40). We also attempted to inhibit the unidirectional fluxes of [ 14 C]taurine by adding unlabeled taurine to the bathing medium or perfusion solution. The addition of 2 mm unlabeled taurine to bathing medium containing 20 µm [ 14 C]taurine had no significant effect on J bl of [ 14 C]taurine in either the DTL (0.20 P 0.30) or the ATL (0.70 P 0.80). Similarly the addition of 2 mm unlabeled taurine to perfusion medium containing 20 µm [ 14 C]taurine had no significant effect on J lb of [ 14 C]taurine in either the DTL (0.10 P 0.20) or the ATL (0.40 P 0.50). In two experiments, we even examined the effect of 100 mm unlabeled taurine on J bl and J lb of [ 14 C]taurine in one DTL and one ATL, but there was no inhibition (data not shown). Direct measurements of net transepithelial fluxes. In a series of experiments, we attempted to measure net movement of L-[ 14 C]alanine and [ 14 C]taurine directly by matching isotope concentrations in the initial perfusate and bathing medium and measuring the change in concentration of radiolabeled substrate between the initial perfusate and the collected fluid. Unfortunately, it proved extremely difficult, even using the same original solution, to be absolutely certain that the concentration of radiolabeled substrate in initial perfusate and bathing medium were identical. Therefore, the data were limited in reliability by scatter in the apparent fluxes. Nevertheless, in the case of L-[ 14 C]alanine, the apparent flux did not differ from zero in four DTL (4.30 1.82 fmol min 1 mm 1 ; 0.10 P 0.20) and four ATL (3.52 2.05 fmol min 1 mm 1 ; 0.20 P 0.30). Similarly, in the case of [ 14 C]taurine, the apparent flux did not differ from zero in six DTL (3.42 20.0 fmol min 1 mm 1 ; 0.80 P 0.90) and six ATL (20.8 13.5 fmol min 1 mm 1 ; 0.20 P 0.30). DISCUSSION In the present study, we were able to isolate papillary thin limbs of Henle s loops from the kidneys of Munich- Wistar rats without the aid of enzymatic agents and to perfuse segments of these thin limbs in vitro. Therefore, we were able to examine amino acid transport in vitro in tubule segments of the same type and from the same strain of rats as we had used for our micropuncture and continuous microinfusion studies in vivo (5 8, 11). In these initial in vitro studies, we chose to examine the transport of L-alanine, a prevalent neutral

AMINO ACID FLUXES IN HENLE S LOOP IN VITRO F209 amino acid, and taurine, a neutral -amino (sulfonic) acid that can exist in very high concentrations in the papillary vasa recta (5) and that may be important as a cellular osmolyte in the renal medulla (9). For both radiolabeled amino acids, there was a tendency for the flux into the DTL to dominate and for the flux out of the ATL to dominate. These observations are in agreement with the in vivo data that indicate that radiolabeled amino acid continuously microinfused into papillary ATL is reabsorbed to a significant extent and that radiolabeled amino acid continuously microinfused into papillary ascending vasa recta appears to enter the ipsilateral DTL (although this tubule site of entry was not shown directly in vivo) (6 8, 11). However, when the unidirectional fluxes were examined in the same tubule segments so that a paired analysis could be performed, the domination of J bl in the DTL and J lb in the ATL, although appearing to hold for both amino acids, was statistically significant only for taurine. Statistical significance only reached the 10% level for L-alanine. It is possible that the lack of statistical significance reflected variability in the tubule segments dissected (e.g., deep versus superficial) since only superficial segments were microinfused in vivo. Such variability may not have shown up to the same extent for taurine where the unidirectional fluxes and the differences between them were much larger than for L-alanine. In the present study with perfused thin limbs, we were unable to saturate or inhibit any of the unidirectional fluxes of the radiolabeled form of the amino acids with high concentrations of the unlabeled form of the amino acids. Although we had no reason to expect that active net transepithelial transport of these amino acids could occur in the thin limbs of Henle s loop, especially in view of the failure of inhibition of the unidirectional fluxes, we attempted to examine this possibility directly by looking for net transepithelial transport when the concentrations of radiolabeled amino acid in the bathing medium and initial perfusate were identical. Unfortunately, it is extraordinarily difficult to be certain that these low concentrations of radiolabeled amino acids are exactly matched, and this leads to considerable variability in the results. Nevertheless, even taking the scatter into account, we found no evidence of active net transepithelial transport. The failure to saturate or inhibit J lb in the ATL in vitro is in agreement with the in vivo microinfusions of the ATL for alanine and most other amino acids (6, 7). However, reabsorption of radiolabeled taurine during in vivo microinfusions of the ATL is saturable and inhibitable, although this may only reflect the inhibition that clearly occurs in the ascending thick limb of Henle s loop (11). The lack of saturation or inhibition of J bl in the DTL does not fit with the less direct observations for amino acids during in vivo microinfusions of the ascending vasa recta (6 8). For almost every radiolabeled amino acid examined during these in vivo studies (including both L-alanine and taurine), movement from the vasa recta into ipsilateral tubular structures could be inhibited (6 8). Although we could be certain of the transfer of intact radiolabeled amino acids from vasa recta to ipsilateral tubular structures and of the inhibition of such transfer in these in vivo studies, we could not be certain of the exact site in the tubules where transfer occurred (6 8). Indeed, it is possible that the inhibition observed in vivo could have involved the exit step from the ascending vasa recta. We formerly thought that saturable, inhibitable carrier-mediated transport was more likely to occur in the epithelia of the thin limbs than in the vasa recta (6 8). However, in view of the present results, the involvement of the vasa recta needs to be reconsidered. In summary, the unidirectional fluxes for radiolabeled L-alanine and taurine observed in isolated thin limbs of Henle s loop in vitro reveal a pattern similar but certainly not exactly equivalent to that observed in vivo. The general lack of inhibition of amino acid fluxes during the current in vitro studies suggests that the transepithelial movement of these amino acids may occur via a paracellular route. However, in the case of taurine, the paired unidirectional fluxes indicate that this would have to be a rectifying pathway, and it is difficult to imagine such rectification in the absence of active transport. In view of our inability to be certain of the exact site of tubular transfer during the in vivo studies, the present in vitro results could be compatible with the in vivo observations, especially if the vasa recta provide a controlling barrier to transport. We thank Diane Abbott for assistance with data analysis. This study was supported in part by National Institutes of Health Grant DK-16294 and Training Grants HL-07249, NS-07309, and GM-08400, and by Grant ES-06694 to the Southwest Environmental Health Sciences Center. Address for reprint requests and other correspondence: W. H. Dantzler, Department of Physiology, College of Medicine, University of Arizona, Tucson, AZ 85724-5051 (E-mail: dantzler@u.arizona.edu). Received 23 December 1998; accepted in final form 30 March 1999. REFERENCES 1. Burg, M., J. Grantham, M. Abramow, and J. Orloff. Preparation and study of fragments of single rabbit nephrons. Am. J. Physiol. 210: 1293 1293, 1966. 2. Chou, C.-L., and M. A. Knepper. In vitro perfusion of chinchilla thin limb segments: segmentation and osmotic water permeability. Am. J. Physiol. 263 (Renal Fluid Electrolyte Physiol. 32): F417 F426, 1992. 3. Dantzler, W. H. Characteristics of urate transport by isolated, perfused snake proximal renal tubules. Am. J. Physiol. 224: 445 453, 1973. 4. Dantzler, W. H. PAH transport by snake proximal renal tubules: differences from urate transport. Am. J. Physiol. 226: 634 641, 1974. 5. Dantzler, W. H., and S. Silbernagl. Amino acid transport by juxtamedullary nephrons: distal reabsorption and recycling. Am. J. Physiol. 255 (Renal Fluid Electrolyte Physiol. 24): F397 F407, 1988. 6. Dantzler, W. H.,and S. Silbernagl. Amino acid transport: microinfusion and micropuncture of Henle s loops and vasa recta. Am. J. Physiol. 258 (Renal Fluid Electrolyte Physiol. 27): F504 F513, 1990.

F210 AMINO ACID FLUXES IN HENLE S LOOP IN VITRO 7. Dantzler, W. H., and S. Silbernagl. Specificity of amino acid transport in renal papilla: microinfusion of Henle s loops and vasa recta. Am. J. Physiol. 261 (Renal Fluid Electrolyte Physiol. 30): F495 F504, 1991. 8. Dantzler, W. H., and S. Silbernagl. Basic amino acid transport in renal papilla: microinfusion of Henle s loops and vasa recta. Am. J. Physiol. 265 (Renal Fluid Electrolyte Physiol. 34): F830 F838, 1993. 9. Garcia-Perez, A., and M. B. Burg. Renal medullary organic osmolytes. Physiol. Rev. 71: 1081 1115, 1991. 10. Silbernagl, S. The renal handling of amino acids and oligopeptides. Physiol. Rev. 68: 911 1007, 1988. 11. Silbernagl, S., K. Völker, H.-J. Lang, and W. H. Dantzler. Taurine reabsorption by a carrier interacting with furosemide in short and long Henle s loops of rat nephrons. Am. J. Physiol. 272 (Renal Fluid Electrolyte Physiol. 41): F205 F213, 1997.