点击显示 收起
【摘要】 Most neutral L -amino acid acids are transported actively across the luminal brush-border membrane of small intestine and kidney proximal tubule epithelial cells by a Na + cotransport system named B 0 that has been recently molecularly identified (B 0 AT1, SLC6A19 ). We show here that the opossum kidney-derived cell line OK also displays a Na + -dependent B 0 -type neutral L -amino acid transport, although with a slightly differing substrate selectivity. We tested the hypothesis that one of the two B 0 AT1-related transporters, SLC6A18 (ortholog of orphan transporter XT2) or SLC6A20 (ortholog of the recently identified mammalian imino acid transporter SIT1), mediates this transport. Anti-sense RNA to OK SIT1 ( o SIT1) but not to OK XT2 ( o XT2) inhibited Na + -dependent neutral amino acid transport induced by OK mRNA injected in Xenopus laevis oocytes. Furthermore, inhibition of o SIT1 gene expression in OK cells by transfection of siRNA and expression of shRNA selectively reduced the Na + -dependent uptake of neutral L -amino acids. Finally, expression of OK cell o SIT1 cRNA in X. laevis oocytes induced besides the transport of the L -imino acid L -Pro also that of neutral L -amino acids. Taken together, the data indicate that in OK cells SIT1 (SLC6A20) is not only an apical imino acid transporter but also plays a major role as Na + -dependent neutral L -amino acid transporter. A similar double role could be envisaged for SIT1 in mammalian kidney proximal tubule and small intestine.
【关键词】 Na + dependent neurotransmitter transporter B AT Xenopus laevis oocytes kidney proximal tubule antisense siRNA
IN THE LAST TWO decades many amino acid transporters have been molecularly identified and their transport function characterized using heterologous expression in cultured cells or Xenopus laevis oocytes ( 2, 18, 25, 26 ). For example, the identifications of gene products corresponding to system B 0 (broad scope Na + -dependent neutral amino acid transport) and to system IMINO (Na + -dependent imino acid transport) have been recently reported ( 1, 11, 22, 24 ). These two transporters are highly expressed in kidney proximal tubule and small intestine. Functionally they have been characterized in the X. laevis expression system and named according to their amino acid transport characteristics B 0 AT1 (SLC6A19) and SIT1 (SLC6A20, XT3s1 in rodents) ( Table 1 ) ( 1, 3, 11, 12, 22, 24 ). They are structurally members of the Na + (and Cl - )-dependent neurotransmitter transporter family SLC6, in which they form together with the still orphan transporter XT2 (SLC6A18) and rodent XT3 (slc6a20) a phylogenetic cluster ( 9, 17, 20, 27 ).
Table 1. Opossum and mammalian transporters of the SLC6 family B 0 -cluster
The question of putative additional transporters, next to B 0 AT1, that transport neutral amino acids across the luminal membrane of small intestine and kidney proximal tubule is still open, as not all cases of neutral amino aciduria (Hartnup disorder) can be explained by mutations in the gene encoding B 0 AT1 (SLC6A19) ( 11, 22 ).
We have shown that the other B 0 -cluster transporters XT2 and SIT1 are also strongly expressed in the brush-border membrane of small intestine and kidney proximal tubule of mice, although with differential axial gradients compared with B 0 AT1 ( 20 ). Because of their brush-border localization and their high degree of structural identity with B 0 AT1, we considered XT2 and SIT1 as possible candidate additional transporters of B 0 -type for which we hypothesized somewhat differential neutral amino acid selectivities.
In the present study, we characterize the B 0 -type neutral amino acid uptake in opossum kidney proximal tubule cells (OK cells) and test the possible role played by endogenous XT2 and SIT1 for this transport. The observed Na + -dependent, broad-range neutral L -amino acid transport resembles, but does not match, that described for mammalian B 0 AT1. Using RNA silencing techniques in OK cells and X. laevis oocyte expression, we show here that the B 0 -like transport of OK cells is mediated, at least in part, by the ortholog of the mammalian IMINO system transporter SIT1 (SLC6A20).
EXPERIMENTAL PROCEDURES
Cell culture. The established cell line from opossum kidney ( Didelphis virginiana ) ( 13 ), clone 3B/2 (kindly provided by Dr. N. Hernando, Zurich) were cultured at 37°C and 5% CO 2 in DMEM/Ham's F-12 medium (1:1) (Life Technologies, Basel, Switzerland) supplemented with 10% heat-inactivated FCS, 5 x 10 4 U/l penicillin, 50 mg/l streptomycin, 2 mM L -glutamine, and 20 mM HEPES as described ( 19 ). Experiments were performed at passage 20 to 30.
Uptake experiments in OK cells. OK cells were seeded on plastic dishes (9.6 or 3.8 cm 2, Corning Costar, Bodenheim, Germany) at 90-100% confluence and cultivated for 6-7 days. Before the uptake, the cell monolayer was washed five times within 1 min with uptake buffer at 37°C. Uptake buffer with 137 mM NaCl (to measure total transport) or 137 mM N -methyl- D -glucamine (NMDG; for Na + -independent transport) contained also (in mM): 10 HEPES, pH 7.4, 1 CaCl 2, 5 KCl, 1 MgCl 2, and 10 glucose. Buffer (0.5 or 0.2 ml) containing L -amino acid ( L -AA) at the indicated concentration and the corresponding radiolabeled tracer (2 µCi/ml of 3 H-AA or 0.2 µCi/ml 14 C-AA) was then applied and cell monolayers were continuously rocked. The uptake was terminated by adding ice-cold uptake buffer, and monolayers were washed five more times. Cells were lysed by adding 1 or 0.5 ml of 0.05% Na + -deoxycholate in 0.1 mM NaOH, and the lysates were neutralized with 100 or 50 µl 1 N HCl. Neutralized lysates (250 or 200 µl) were measured in a beta counter in the presence of 3 ml scintillation fluid (Packard, Meriden, CT). Na + -dependent values were calculated as the difference between total and Na + -independent uptakes. Na + -dependent transport rates were expressed relative to the Na + -dependent L -Phe transport rate. For substrate selectivity measurements, 5 mM L -amino acids were given for 20 s. For competition experiments, the uptake of 5 mM L -Phe (20 s) was performed in the absence (control) or presence of 25 mM competing cold L -amino acid, MeAIB, or 10 mM BCH. For ion dependence experiments, NMDG or Li + was used to substitute for Na +. For amino acid or Na + concentration dependence experiments, five L -Phe or L -Ile concentrations or eight Na + concentrations (NMDG substitution) were used. Curves corresponding to the Michaelis-Menten equation were fitted to the data by nonlinear regression, and transport kinetic parameters ( V max and K 0.5 ) were derived using Prism software (GraphPad Prism Version 4.0, GraphPad, San Diego, CA).
Transient transfection of OK cells with small-interference RNA. Approximately 12-24 h before transfection, OK cells were plated on plastic dishes (9.6 or 3.8 cm 2 ) and cultured in complete medium without antibiotics until they reached 90% confluency for the transfection ( 4, 5 ). Transient transfections with 70 nM small-interference (si)RNA or 1 µg short-hairpin (sh) RNA construct were performed according to the manufacturer's instructions (Gene Silencer siRNA transfection reagent, Gene Therapy Systems, San Diego, CA or Lipofectamine 2000, Invitrogen, Carlsbad, CA). siRNA duplex mixed with 4 µl gene silencer reagent and diluted in 1 ml of medium without FCS and shRNA constructs mixed with 4 µl lipofectamine and diluted in 1 ml OPTI-MEN medium were applied to the cell monolayer. Transfection medium was replaced by complete medium after 24 h, and the cells were assayed for Na + -dependent L -Ile uptake 48 h after transfection in the case of siRNA or 72 h in the case of shRNA constructs. Transfection efficiency of shRNA constructs was monitored by viewing enhanced green fluorescence protein fluorescence using an inverted fluorescence microscope (Nikon ECLIPSE TE 300/200).
Selection and production of OSIT 1 si- and sh RNAS. The siRNA sequences used were designed ( http://websoft2.unige.ch/sciences/biologie/bicel/RNAi.html ) and synthesized according to Donzé and Picard ( 6 ). The 21-base small double-strand RNAs from o SIT1 were named according to the position of their first base relative to the first ATG: siRNA 165, siRNA 234, siRNA 424, and siRNA 701. Desalted DNA oligonucleotides were ordered from Microsynth (Balgach, Switzerland). For transcription reactions, 1 nmol of each oligonucleotide was annealed with the T7 sense oligonucleotide in TE buffer (10 mM Tris·HCl, pH 8.0, 1 mM EDTA) by heating at 95°C for 2 min and allowing cooling down slowly. The dsDNA was used as template for RNA synthesis with a T7 transcription kit according to the manufacturer's protocol (MEGAscript, Ambion, Austin, TX). The complementary ssRNA products were annealed (95°C for 5 min followed to 1 h at 37°C), purified, diluted in RNase free water, and used to transfect the OK cells. Negative control OK cells were similarly exposed to the transfection reagent but in the absence of siRNA.
For the shRNA approach, the choice of the oligonucleotide sequences and their annealing and ligation to the vector were performed according to the kit manufacturer (Stratagene, La Jolla, CA). The 29-bp oligonucleotide sequences used were also named according to the position of their first base: shRNA 250, shRNA 420, shRNA 590, and shRNA 830 and ordered from Microsynth. After annealing in 50 mM NaCl, 10 mM Tris·HCl, pH 8.0, 0.1 mM EDTA at 93°C for 3 min and cooling to 37°C, the complementary oligos were ligated to shRNA vector pGE-2hrGFP II (Stratagene) digested with Bam HI and Xba I (Promega, Madison, WI). These shRNA DNA constructs were used to transfect OK cells. The negative control for the experiments was a construct containing a scramble sequence distributed by the manufacturer.
Amplification of the OK cell OXT 2 and OXT 3 sequence. Total RNA from OK cells was isolated with RNeasy Mini Kit according to the manufacturer (Qiagen, Hombrechtikon, Switzerland). One microgram of total RNA was reverse transcribed using Superscript II Reverse Transcriptase (Invitrogen) and the oligos 5'-CDS and SMART II A, provided with a SMART RACE cDNA Amplification Kit (Clontech, Palo Alto, CA). First-strain cDNA was used as a template for 5'-RACE PCR. The SMART RACE cDNA Amplification Kit (Clontech) was used following the manufacturer's instructions. The specific primers 5'-CTT CCC GCT GGA TGC TTT GGT CTG G-3' for o XT2 and 5'-CTG GGT GGC TGC ATT GAT CCA GGC-3' for o SIT1/XT3 were used. Nested PCR was performed using 5% of the first PCR volume with specific nested primers 5'-CGC TCC AGG TAG AGT GAG TCC TCG-3' for o XT2 and 5'-GGC CAT TGG TGG CTC CAT GCA GGG-3' for o SIT1/XT3. The initial amplification of o XT2 and o SIT1/XT3 was done by degenerative PCR and served as a base for the specific primers used for the RACE-PCR. The PCR products were then cloned in the vector pGEMt (Promega) and sequenced using the ABI PRISM BigDye Terminator 3.1 system (Applied Biosystems) with commercial primers. The inserts were subcloned into a pBlueScript modified X. laevis expression vector (KSM) containing both the 5' and 3' UTRs of the -globin gene (kindly provided by Dr. L. V. Virkki). For expression in X. laevis oocytes the opossum o XT2 and o SIT1-KSM plasmids were linearized with Xba I (Promega) and used as template for RNA synthesis from the T3 promoter (mMESSAGE mMACHINE, Ambion). Multiple amino acid sequence comparisons were made using Clone Manager Professional suit (Scientific & Educational Software, Cary, NC).
Transport studies in X. leavis oocytes. Expression studies and uptake measurements in X. laevis oocytes were performed as described previously ( 15 ). Briefly, oocytes were injected with the indicated RNA and incubated for 72 to 144 h at 16°C in ND96 buffer (in mM: 96 NaCl, 2 KCl, 1 MgCl 2, 1.8 CaCl 2, and 5 HEPES, pH 7.4, adjusted with Tris, supplemented with 50 mg/l tetracycline). Before uptake measurements, oocytes were preincubated for 2 min at 22°C, and the respective uptake buffer containing amino acids (0.02 µCi/ml) was added. Uptake was stopped after 6 or 30 min by washing the oocytes five times with 3 ml of ice-cold buffer. Oocytes were then distributed to individual vials, lysed in 2% SDS, and the radioactivity was determined by liquid scintillation. As for cells, total uptake was measured in buffer containing NaCl (100 mM), Na + independence was measured in buffer containing NMDG-Cl (100 mM), and in both conditions the buffer contained (in mM): 2 KCl, 1 MgCl 2, 1.8 CaCl 2, and 10 HEPES, pH 7.4, adjusted with Tris. Gluconate (Glc) salts were used when chloride was substitute; the solution was also buffered with HEPES·Tris, pH 7.4. Concentration dependence of L -Ile and L -Pro uptake was measured for 6 min, and selectivity measurements were performed with 1 mM L -amino acids for 30 min. For competition experiments, the uptake of 0.1 mM L -Pro or L -Ile was performed in the absence (control) or presence of 40 mM cold L -Ile or 2 mM L -Pro, MeAIB, or BCH. The transport in the presence of the inhibitor was normalized to that of L -Pro or L -Ile alone (control) and expressed as percentage.
Hybrid depletion. The oligonucleotides were designed within the amplified sequences of o XT2 and o SIT1 ( http://scitools.idtdna.com/AntiSense /), and named according to the position of their first base (related to the first ATG): AS-S 446-466 and 881-900 for o XT2, and AS-S 71-89 and 222-242 for o SIT1. Total RNA was extracted from OK cells using TRIzol reagent (Life Technologies, Basel, Switzerland), and poly(A) + mRNA prepared with PolyATtract R mRNA Isolation Systems (Promega) following the manufacturer's protocol. Oocytes were injected with poly(A) + mRNA (25 ng/oocyte) or not injected. Hybrid depletion experiments were performed as described previously ( 16 ). OK cell poly(A) + mRNA was denaturated at 65°C for 5 min in the presence of the sense or antisense oligonucleotide complementary to opossum o XT2 or o SIT1 (100 ng oligonucleotide in 50 mM NaCl, RNase free) and further incubated at 42°C for 30 min. The oocytes were injected with poly(A) + mRNA alone or depleted (25 ng/oocyte) and incubated for 72-144 h. Na + -dependent transport was calculated and normalized to the poly(A) + mRNA alone and to its respective sense control.
Statistics. All data are shown as means ± SE, and n is the number of cell monolayer dishes or number of oocytes. Experimental protocols were repeated at least twice. Comparison of means was done by t -test with a Mann-Whitney U -test as a nonparametric posttest ( P < 0.05), Wilcoxon Signed Rank Test against a hypothetical value, or ANOVA with Dunnett's test as a posttest ( P < 0.05).
RESULTS
OK cells exhibit system B 0 - TYPE L -neutral amino acid transport. L -Amino acid transport was assayed in confluent OK cells monolayers grown on plastic dishes in the presence and absence of Na +. Under these culture conditions, OK cells differentiate into a polarized phenotype with delineated apical and basolateral membranes ( 8 ). Total transport of L -amino acids (in the presence of Na + ) was always larger than Na + -independent transport (in the absence of Na + ) and the difference between these values was considered as the Na + -dependent part of the transport. Na + -dependent L -Phe (5 mM) uptake being linear for 1 min (data not shown), further amino acid uptake experiments were performed for 20-s periods, in particular also for the determination of the substrate concentration dependence of L -Phe and L -Ile ( Fig. 1 A ). Na + -dependent apparent affinities ( K 0.5 ) derived from the fitted Michaelis-Menten curves amounted to 1.9 and 1.1 mM for L -Phe and L -Ile, respectively. Furthermore, replacement of Na + with NMDG or Li + showed that these substitutions are not tolerated ( Fig. 1 B ). Also, the Na + concentration dependence of L -Ile uptake (5 mM) was measured by replacing Na + with NMDG and the derived K 0.5 for Na + amounted to 2.4 mM ( Fig. 1 C ).
Fig. 1. Apical system B 0 -like Na + -dependent amino acid transport in opossum kidney (OK) cells. A : concentration dependence of Na + -dependent L -Ile and L -Phe uptake in OK cells. Na + -dependent transport was calculated as the difference between uptake in the presence (total transport) and in the absence of Na + (NMDG substitution). Curves corresponding to Michaelis-Menten kinetics were fitted to Na + -dependent uptake values obtained for 0.05, 0.5, 1.67, 5, and 16.7 mM L -Phe and L -Ile. Data are from 2 independent experiments (means ± SE, n = 6). B : ion dependence of L -amino acid transport in OK cells. L -Phe (5 mM) uptakes were performed in buffer containing NMDG or Li + as Na + substitutes. Absolute values were normalized to the Na + -independent L -Phe uptake (NMDG), * P < 0.01, n = 10. C : Na + concentration dependence of L -Ile transport in OK cells. Uptake of L -Ile (5 mM) was performed in buffer containing 0, 2.5, 5, 12.5, 25, 50, 75, 100, and 137 mM NaCl using NMDG as substitute. Michaelis-Menten equation was fit to the Na + -dependent L -Ile transport. The data are from 2 independent experiments (means ± SE, n = 6). D : substrate selectivity of Na + -dependent L -amino acids ( L -AA) transport. For each amino acid, 2-3 independent experiments were pooled after normalizing the values to the Na + -dependent uptake of L -Phe. Data represent means ± SE, * P < 0.01 compared with L -Phe, n = 6-9. E : competition of Na + -dependent L -Phe uptake by L -amino acids. The uptake of 5 mM L -Phe was competed with 25 mM nonlabeled amino acid, MeAIB, or 10 mM BCH but not with L -Tyr (nd). For each amino acid, results obtained in 2-3 independent experiments were pooled after normalizing them to the Na + -dependent uptake of L -Phe. Data represent means ± SE, * P < 0.01 compared with no competitor, n = 6-9.
The selectivity of the Na + -dependent amino acid uptake of OK cells was assayed for almost all proteogenic L -amino acids, and the uptake rates are shown normalized to that of L -Phe (12.27 ± 0.68 nmol·dish -1 ·20 s -1; Fig. 1 D ). Only neutral L -amino acids were found in these conditions to be transported in a Na + -dependent manner and the most efficiently transported ones were those with branched side chains followed by aromatic side chains and L -alanine.
To test whether the observed Na + -dependent uptake of the different amino acids was potentially via the same transporter(s) as the uptake of L -Phe and to gain some information about relative substrate affinities, uptake of L -Phe (5 mM) was competed by a fivefold excess (25 mM) of the different L -amino acids ( Fig. 1 E ). As expected from the uptake experiments, L -Phe transport was most efficiently inhibited by branched chain amino acids (% inhibition: L -Ile 88, L -Val 86, L -Leu 77) and then by aromatic L -amino acids and L -alanine ( L -Phe 67, L -Trp 41, L -Tyr not tested, L -Ala 56). From the other neutral L -amino acids that were shown to be transported relatively well ( L -Gln, L -Asn, and L -Thr), only L -Gln competed efficiently for Na + -dependent L -Phe transport (% of inhibition: L -Gln 72, L -Asn 14, L -Thr 3). These results suggest that the branched chain and aromatic amino acids as well as L -Ala and L -Gln are to a large extent transported via the same Na + -dependent transporter(s) as L -Phe. In contrast, it appears that L -Asn and L -Thr could be transported by other Na + -dependent pathways.
A surprising observation for which we have yet no explanation is the fact that the presence of 25 mM Gly stimulated the Na + -dependent uptake of L -Phe. It is conceivable that this effect might be indirectly due to the functional interaction of Gly with other transport systems of OK cells rather than to a direct effect at the level of L -Phe transporters. Another surprising observation is the fact that amino acids that were only little or not transported Na + dependently inhibited L -Phe uptake. This was to some extent the case for L -Arg (30%) and L -Pro (34%) and more strikingly for L -Cys, L -His, and L -Met (83, 71, and 85%). Finally, the two tested amino acid analogs BCH and MeAIB did not compete substantially with Na + -dependent L -Phe uptake. Based on these results that revealed L -Ile as the most efficient substrate for Na + -dependent transport and competitor for L -Phe uptake, we used this branched amino acid as leading substrate for further functional experiments.
OK cell OXT 2 and OSIT 1 have a high degree of identity with their mammalian orthologs. To test the role of the opossum o XT2 and o SIT1 gene products in the Na + -dependent transport of neutral L -amino acids in OK cells, corresponding cDNAs were first amplified from OK mRNA by RACE PCR using degenerate primers based on mammalian XT2 and SIT1 nucleotide sequences. The cDNA open reading frames isolated by this method were 1,854 and 1,782 nucleotides, respectively, and the translated proteins predicted to have 618 and 594 amino acids each. During the course of this study, opossum ( Monodelphis domestica ) genome project sequences corresponding to XT2 and SIT1 (XT3) appeared in the Ensembl database ( http://www.ensembl.org /). The protein encoded by our o XT2 cDNA obtained from the opossum D. virginiana OK kidney cell line is 95% identical to that encoded by M. domestica entry Q96N87 (gene product #5) and our OK SIT1 ( o SIT1) is 96% identical to entry 08876 (gene product #2). o XT2 and o SIT1 share 45% identical residues, whereas their identity with mammalian orthologs amounts to 70 and 85%, respectively ( Table 1 ).
Inhibition of OSIT 1 gene expression reduces NA + - DEPENDENT transport of L -I LE in X. laevis oocytes and OK cell monolayers. As a first experiment to test the putative role of o XT2 and o SIT1 in the Na + -dependent transport of neutral L -amino acids observed in OK cell monolayers, the uptake of L -Ile was assayed in X. laevis oocytes expressing poly(A) + RNA isolated from intact OK cell monolayers with and without hybrid depletion of o XT2 or o SIT1 mRNA. Poly(A) + RNA of OK cells induced L -Ile transport, 50% of which was Na + dependent ( Fig. 2 A ). The hybrid depletion of o XT2 or o SIT1 mRNA from OK cell poly(A) + RNA was attempted using 19- to 21-base-long antisense oligonucleotides (AS) and their corresponding sense (S) oligonucleotides as controls. Figure 2 B shows that o SIT1 mRNA depletion, and not o XT2 mRNA depletion, decreased the Na + -dependent L -Ile transport induced in X. laevis oocytes (65.5 and -0.3% inhibition, respectively).
Fig. 2. Inhibition of o SIT1 by hybrid depletion or RNA interference reduces Na + -dependent L -Ile transport. A : expression of OK cell poly(A) + RNA in Xenopus laevis oocytes induces L -Ile transport. X. laevis oocytes injected with 25 ng OK cell poly(A) + RNA were tested after 72 h. Total uptake of L -Ile (0.5 mM) was measured in the presence of Na + (gray bar), Na + -independent transport with NMDG as Na + substitute (open bar), and calculated Na + -dependent transport (filled bar). Data are from 2 independent experiments (means ± SE, n = 13). B : hybrid depletion of o XT2 and o SIT1 mRNA in OK poly(A) + RNA expressed in X. laevis oocytes. OK poly(A) + mRNA was incubated with sense or antisense oligonucleotide before injecting X. laevis oocytes and measuring L -Ile uptake as in A. Data (means ± SE) are from 4-5 independent experiments. Open bars: o SIT1 antisense ( n = 38) or sense ( n = 35), ** P < 0.01; gray bars: o XT2 antisense ( n = 32) or sense ( n = 28). C : transfection of o SIT1 siRNA reduces Na + -dependent transport of L -Ile in OK cell monolayers. In vitro synthesized siRNA was transfected at a final concentration of 70 nM, whereas control cell monolayers were exposed to transfection reagent only. Na + -dependent L -Ile uptake was essayed after 48 h (as in Fig. 1 ) and normalized to the uptake measured in control monolayers. Data represent 3 independent experiments (means ± SE, n = 9), *** P < 0.001. D : expression of o SIT1 shRNA reduces Na + -dependent transport of L -Ile in OK cell monolayers. OK cells were transfected with 1 µg of the shRNA construct or the negative control vector. They were assayed for L -Ile uptake 72 h later (as in Fig. 1 ) and normalized to the uptake measured in control monolayers. The data represent 3 independent experiments (means ± SE, n = 9), ** P < 0.01.
To test more directly in OK cell monolayers the putative role of o SIT1 for the Na + -dependent transport of neutral amino acids, endogenous o SIT1 expression was inhibited using two RNA interference methods. On the one hand, in vitro synthesized siRNA was directly transfected into OK cell monolayers and, on the other hand, OK cells were transfected with a plasmid that controls the production of shRNA. Four different siRNAs and four different shRNA constructs were transiently transfected into confluent OK cell monolayers. Figure 2 shows the inhibition of Na + -dependent L -Ile uptake measured with one of the oligonucleotides from each method, namely, with siRNA 424 in Fig. 2 C and with shRNA 420 in Fig. 2 D. Actually, all four siRNA reduced Na + -dependent L -Ile uptake between 55.68 ± 24.85 and 92.59 ± 14.46% (siRNA 424) relative to their control. Similarly, but to a lesser extent, two shRNA constructs reduced Na + -dependent L -Ile transport 25.34 ± 17.11 and 61.3 ± 12.47%, in the case of shRNA 420. The fact that the Na + -dependent transport of inorganic phosphate was not affected by shRNA 420 expression (P i transport in control cells: 2.48 ± 0.23 nmol·9.6 cm 2 dish -1 ·10 min -1; with shRNA 420: 2.85 ± 0.14 nmol·9.6 cm 2 dish -1 ·10 min -1, n = 3) essentially excludes the possibility that shRNA 420 interferes with a more general mechanism that would affect apical Na + -dependent transports. In summary, the results of all three experimental series with different antisense oligonucleotides and RNA interference approaches strongly suggest that o SIT1 but not o XT2 plays an important role in the Na + -dependent neutral amino acid uptake in OK cell monolayers.
OK SIT 1 expressed in the X. laevis oocytes promoted Na + -dependent L -amino and imino acid transport. o SIT1 function was then directly characterized using the X. laevis oocytes expression system. The induced L -amino acid transport was assayed in the presence of 100 mM NaCl 72-120 h after injection of its cRNA and the uptake rate measured in noninjected oocytes was subtracted (generally <25%). Time dependence measurements using L -Ile and L -Pro (1 mM) as substrates indicated that the uptake rate was constant for 10 min (data not shown), and thus concentration dependence was determined for uptakes lasting 6 min. Substrate concentrations inducing half-maximal transport rates ( K 0.5 ) were derived from fitted Michaelis-Menten curves and amounted for L -Pro to 0.25 ± 0.12 mM and for L -Ile to 2.6 ± 0.90 mM. The V max calculated for L -Ile and L -Pro were 1,875 ± 236 and 613 ± 54 mM, respectively ( Fig. 3 A ). The Na + -dependent transport of amino and imino acids ( L -Ile and L -Pro) by o SIT1 expressed in X. laevis oocytes was entirely dependent on the presence of Cl - ( Fig. 3 B ). The substrate selectivity of o SIT1 for all proteogenic L -amino acids was assayed by comparing the uptake of each L -amino acid given at a 1 mM concentration for 30 min ( Fig. 3 C ). L -Val, L -Ile, and L -Pro were taken up the most efficiently via o SIT1 in these conditions. The amount of L -Leu and L -Phe taken up in this time was 50% while the transport of L -Gln, L -Met, L -Ala, L -Tyr, L -Trp, L -Thr, L -Ser, L -Cys, L -Asn, Gly, and L -Lys was 20% that of the best substrates. L -Arg, L -Glu, L -Asp, and L -His were not transported.
Fig. 3. Expression of o SIT1 in X. laevis oocytes induces Na + -dependent L -amino and L -imino acid transport. A : L -Ile and L -Pro concentration dependence of o SIT1 mediated transport in oocytes. Curves corresponding to Michaelis-Menten kinetics were fitted to Na + -dependent uptake values obtained for 0.01, 0.03, 0.1, 1, 3, 5, and 10 mM L -Ile or 0.01, 0.03, 0.1, 1, 2, 3, 4, and 5 mM L -Pro in 6 min. Data are from 5 independent experiments (means ± SE, n = 14-36). B : transport of L -Ile and L -Pro induced by o SIT1 in X. laevis oocytes is Na + and Cl - dependent. Na + was replaced by NMDG and Cl - by gluconate (Glc). Data are from 2 independent experiments and values obtained for noninjected oocytes are subtracted (means ± SE, * P < 0.01, n = 14-17). C : substrate selectivity of o SIT1 expressed in X. laevis oocytes. Uptakes of L -amino acids (1 mM for 30 min) were performed as above 72 to 120 h after injection of 50 ng of cRNA. Data are from 2-5 independent experiments (means ± SE, * P < 0.01 compared with L -Ile, n = 12-38) and values obtained for noninjected oocytes are subtracted. D : transport of L -Ile by o SIT1 is competed by L -Pro and vice versa. L -Ile or L -Pro transport (0.1 mM) was competed by concentrations of L -Ile (40 mM) and L -Pro (2 mM) corresponding approximately to 20-fold of their apparent affinity (see A ). The amino acid analogs MeAIB and BCH were used at 2 mM. Data were normalized to the transport of L -Ile or L -Pro alone and the means ± SE (* P < 0.01 and ** P < 0.001, n = 11-14) of 2 independent experiments are shown.
The uptake of 0.1 mM L -imino ( L -Pro) or amino acid ( L -Ile) was measured in the presence of an excess of L -Ile (40 mM 20 x K 0.5 L-Ile) or L -Pro (2 mM 20 x K 0.5 L-Pro), and in the presence of 2 mM MeAIB or BCH ( Fig. 3 D ). L -Pro, MeAIB, and BCH efficiently competed L -Ile transport (% inhibition: L -Pro 100, MeAIB 100, BCH 47). Similarly, L -Ile and MeAIB efficiently competed the uptake of L -Pro (percent inhibition: L -Ile 66, MeAIB 83), whereas BCH (2 mM) did not (-6% inhibition).
DISCUSSION
Characterization of neutral L -amino acid transport in polarized OK cells. In this study, we first characterized the Na + -dependent transport of L -amino acids in polarized OK cell monolayers, a much used kidney proximal tubule cell model ( 8 ). OK cells have been shown previously to functionally express several proximal kidney tubule amino acid transporters, as for instance the apical and basolateral heterodimeric exchangers b 0,+ AT-rBAT and LAT2-4F2hc ( 7 ). They were also shown earlier to display an apical Na + -dependent, electrogenic transport of some neutral L -amino acids ( L -Phe, L -Ala, L -Val, and L -Gln) ( 21 ). Here, we demonstrate that they display a B 0 -type transport with apparent affinities ( K 0.5 ) for L -Phe and L -Ile in the lower millimolar range (1.9 and 1.1 mM, respectively; Fig. 1 A ). This Na + -dependent transport showed an apparent Na + affinity of 2.4 mM and a Hill coefficient of 1.00 ± 0.096, suggesting a 1:1 Na + - L -amino acid cotransport ( Fig. 1, B and C ). Although this transport exhibits a broad substrate selectivity range as expected for system B 0 ( Fig. 1 D ), it differs to some extent from the substrate selectivity previously determined for mouse and human B 0 AT1, suggesting that (an)other transporter(s) besides B 0 AT1 might be involved.
The hypothesis that more than a single transporter is involved in Na + -dependent neutral amino acid uptake in OK cell monolayers is supported by the complex competition pattern of L -Phe uptake by various L -amino acids. In particular, whereas most substrates displaying an efficient Na + -dependent transport also efficiently competed with L -Phe transport, some well-transported substrates did not compete efficiently and some poorly transported substrates competed quite efficiently ( Fig. 1 E ). Additionally, neither the amino acid analog BCH, known to compete with amino acid transport via system L, B 0,+ and B 0, nor MeAIB, known to compete transport via SIT, inhibited the Na + dependent uptake of L -Phe in OK cells ( 1, 14, 23, 24 ). Many of these discrepancies could be explained by the presence of multiple transport systems with overlapping and complementary substrate selectivities and affinities that would be expressed in OK cells. In summary, OK cell monolayers display a B 0 -type Na + -dependent neutral amino acid transport that is suggested to be mediated by more than one B 0 -like transporter, a possibility that might also be considered for mammalian kidney proximal tubule and small intestine.
Hybrid depletion and RNA silencing suggest a major role for OSIT 1 in Na + -dependent L -I LE transport in OK cells. After having verified that orthologs of all three SLC6 family B 0 -cluster transporters B 0 AT1 (SLC6A19), XT2 (SLC6A18), and SIT1 (SLC6A20) are expressed in OK cells by RT-PCR using degenerate primers (data not shown), we investigated the putative role of both o XT2 and o SIT1 in the transport of neutral L -amino acids. Using as a first approach exogenous OK cell mRNA expression in X. laevis oocytes, coupled with oligonucleotide-mediated hybrid depletion, we obtained results that suggested a role for o SIT1 but not for o XT2 in Na + -dependent neutral amino acid uptake. A similar conclusion was suggested by the use of two RNA interference approaches directly in OK cell monolayers. All three approaches suggested that a very high fraction of the Na + -dependent L -Ile uptake in OK cells is mediated by o SIT1. It is, however, difficult to quantify precisely this fraction due to several reasons. First, we focused only on the "leading substrate" L -Ile. Second, we used it at a high concentration (5 mM) that favors high-capacity/low-affinity transporters like B 0 AT1 and SIT1 over high-affinity and low-capacity ones. Third, the absence of competitors is nonphysiological, and, fourth, the inhibition of o SIT gene expression was not monitored at the protein level (lack of anti- o SIT antibody). Despite these technical limitations, the fact that three different approaches led to the same conclusion strongly supports the hypothesis that o SIT1 plays a major role for the Na + -dependent apical uptake of neutral amino acids in OK cells. In view of the similarity of OK cell o SIT1 and mammalian SIT1 and of the localization of SIT1 in small intestine and kidney proximal tubule, it can be proposed that SIT1 also could play an important role as additional L -amino acid uptake pathway besides B 0 AT1 in mammalian kidney and intestine.
OK cell OSIT 1 is a relatively high-affinity transporter for L - PROLINE and a lower-affinity transporter with higher capacity for neutral L -amino acids. We have called the SLC6A20 ortholog of OK cells o SIT1, because of its similarity with human SIT1. In the evolution of rodents, an additional duplication of this gene leading to a tandem arrangement has taken place and the products of these two genes are provisionally called XT3 and XT3s1 ( Table 1 ) ( 9, 10 ). It should be noted that human SIT1 and o SIT1 resemble more XT3s1, the mRNA of which is more abundant in small intestine and in brain than that of XT3 ( 9, 10, 20 ).
The functional characterization of o SIT1 in X. laevis oocytes shows that this transporter has characteristics that highly resemble those of mammalian SIT1 ( Fig. 3 ). Indeed, mammalian o SIT1 from human, mouse, and rat was recently functionally characterized and shown to transport L -Pro with a relatively high apparent affinity of 200 µM and also to transport related compounds such as MeAIB, hydroxiproline, methilproline, pipecolate, sarcosine, and D -proline ( 12, 24 ). Based on these transport characteristics and on the mRNA tissue distribution, this transporter was identified as the previously functionally described system IMINO. The two involved laboratories also reported that the transport of L -Pro by SIT1 was to some extent competed by neutral L -amino acids but that these were not efficiently transported.
We characterized the transport of L -neutral amino acids by o SIT1 in more details, always comparing it with the transport of the imino acid L -proline. Importantly, the apparent affinity ( K 0.5 ) of L -Ile was 10-fold higher (2.6 mM) than that of L -Pro (0.25 mM) ( Fig. 3 A ), a difference that can explain why neutral L -amino acid transport was not investigated further in the first studies on mammalian SIT1. Indeed, when we compared the transport rate of L -Pro with that of neutral amino acids at a similar low concentration (100 µM), L -Pro transport was 3.5-fold higher. However, it has to be noted that the V max of SIT1 is about threefold higher for L -Ile than for L -Pro and that its apparent affinity for L -Ile is similar to that of the dedicated neutral amino acid transporter B 0 AT1. Direct competition experiments performed in oocytes expressing o SIT1 confirmed that L -Pro and L -Ile are transported by the same transporter. This was confirmed by the fact that MeAIB competed both L -Ile and L -Pro uptake.
At first sight, these latter functional results and the little Na + -dependent L -Pro transport measured in OK cell monolayers ( Fig. 1 D ) do not appear to support the RNA interference-based hypothesis that a large fraction of Na + -dependent L -neutral amino acid transport of OK cells is mediated by o SIT1. An explanation for this apparent discrepancy is that the Na + -independent transport of L -Pro was very high in OK cell monolayers at the elevated substrate concentrations tested (5 mM), whereas the V max of o SIT1 for Na + -dependent L -Pro transport is quite low, compared with its V max for L -Ile ( Fig. 3 A ). Together, these observations can explain why in our experimental conditions the Na + -dependent transport of L -Pro into OK cells was to a large extent masked.
In summary, the present results demonstrate that o SIT1, and by extrapolation probably also its mammalian orthologs, transport neutral L -amino acids besides L -imino acids. This suggests the possibility that SIT1, besides efficiently (re)absorbing L -Pro, also participates next to B 0 AT1 to the Na + -dependent (re)absorption of neutral L -amino acids in the kidney proximal tubule and in the small intestine.
GRANTS
This project was supported by grants to F. Verrey from the Swiss National Science Foundation 31-59141.99/02 and 31-108021/1 and the European FP6 project EUGINDAT.
【参考文献】
Broer A, Klingel K, Kowalczuk S, Rasko JE, Cavanaugh J, and Broer S. Molecular cloning of mouse amino acid transport system B0, a neutral amino acid transporter related to Hartnup disorder. J Biol Chem 279: 24467-24476, 2004.
Broer S and Wagner CA. Structure-function relationships of heterodimeric amino acid transporters. Cell Biochem Biophys 36: 155-168, 2002.
Camargo SM, Makrides V, Virkki LV, Forster IC, and Verrey F. Steady-state kinetic characterization of the mouse B(0)AT1 sodium-dependent neutral amino acid transporter. Pflugers Arch In press.
Dalby B, Cates S, Harris A, Ohki EC, Tilkins ML, Price PJ, and Ciccarone VC. Advanced transfection with Lipofectamine 2000 reagent: primary neurons, siRNA, and high-throughput applications. Methods 33: 95-103, 2004.
Di Sole F, Cerull R, Babich V, Quinones H, Gisler SM, Biber J, Murer H, Burckhardt G, Helmle-Kolb C, and Moe OW. Acute regulation of Na/H exchanger NHE3 by adenosine A 1 receptors is mediated by calcineurin homologous protein. J Biol Chem 279: 2962-2974, 2004.
Donze O and Picard D. RNA interference in mammalian cells using siRNAs synthesized with T7 RNA polymerase. Nucleic Acids Res 30: e46, 2002.
Fernandez E, Torrents D, Chillaron J, Martin Del Rio R, Zorzano A, and Palacin M. Basolateral LAT-2 has a major role in the transepithelial flux of L -cystine in the renal proximal tubule cell line OK. J Am Soc Nephrol 14: 837-847, 2003.
Hernando N, Sheikh S, Karim-Jimenez Z, Galliker H, Forgo J, Biber J, and Murer H. Asymmetrical targeting of type II Na-P i cotransporters in renal and intestinal epithelial cell lines. Am J Physiol Renal Physiol 278: F361-F368, 2000.
Kiss H, Darai E, Kiss C, Kost-Alimova M, Klein G, Dumanski JP, and Imreh S. Comparative human/murine sequence analysis of the common eliminated region 1 from human 3p21.3 Mamm. Genome 13: 646-655, 2002.
Kiss H, Kedra D, Kiss C, Kost-Alimova M, Yang Y, Klein G, Imreh S, and Dumanski JP. The LZTFL1 gene is a part of a transcriptional map covering 250 kb within the common eliminated region 1 (C3CER1) in 3p21.3. Genomics 73: 10-19, 2001.
Kleta R, Romeo E, Ristic Z, Ohura T, Stuart C, Arcos-Burgos M, Dave MH, Wagner CA, Camargo SR, Inoue S, Matsuura N, Helip-Wooley A, Bockenhauer D, Warth R, Bernardini I, Visser G, Eggermann T, Lee P, Chairoungdua A, Jutabha P, Babu E, Nilwarangkoon S, Anzai N, Kanai Y, Verrey F, Gahl WA, and Koizumi A. Mutations in SLC6A19, encoding B0AT1, cause Hartnup disorder. Nat Genet 36: 999-1002, 2004.
Kowalczuk S, Broer A, Munzinger M, Tietze N, Klingel K, and Broer S. Molecular cloning of the mouse IMINO system: an Na + - and Cl - -dependent proline transporter. Biochem J 386: 417-422, 2005.
Koyama H, Goodpasture C, Miller MM, Teplitz RL, and Riggs AD. Establishment and characterization of a cell line from the American opossum ( Didelphys virginiana ). In Vitro 14: 239-246, 1978.
Lynch AM and McGivan JD. Evidence for a single common Na + -dependent transport system for alanine, glutamine, leucine and phenylalanine in brush-border membrane vesicles from bovine kidney. Biochim Biophys Acta 899: 176-184, 1987.
Meier C, Ristic Z, Klauser S, and Verrey F. Activation of system L heterodimeric amino acid exchangers by intracellular substrates. EMBO J 21: 580-589, 2002.
Mora C, Chillaron J, Calonge MJ, Forgo J, Testar X, Nunes V, Murer H, Zorzano A, and Palacin M. The rBAT gene is responsible for L -cystine uptake via the b0,(+)-like amino acid transport system in a "renal proximal tubular" cell line (OK cells). J Biol Chem 271: 10569-10576, 1996.
Nash SR, Giros B, Kingsmore SF, Kim KM, el-Mestikawy S, Dong Q, Fumagalli F, Seldin MF, and Caron MG. Cloning, gene structure and genomic localization of an orphan transporter from mouse kidney with six alternatively spliced isoforms. Receptors Channels 6: 113-128, 1998.
Palacin M, Estevez R, Bertran J, and Zorzano A. Molecular biology of mammalian plasma membrane amino acid transporters. Physiol Rev 78: 969-1054, 1998.
Pfister MF, Lederer E, Forgo J, Ziegler U, Lotscher M, Quabius ES, Biber J, and Murer H. Parathyroid hormone-dependent degradation of type II Na + /Pi cotransporters. J Biol Chem 272: 20125-20130, 1997.
Romeo E, Dave MH, Bacic D, Ristic Z, Camargo SM, Loffing J, Wagner CA, and Verrey F. Luminal kidney and intestine SLC6 amino acid transporters of B 0 AT-cluster and their tissue distribution in Mus musculus. Am J Physiol Renal Physiol 290: 376-383, 2006.
Schwegler JS, Heuner A, and Silbernagl S. Electrogenic transport of neutral and dibasic amino acids in a cultured opossum kidney cell line (OK). Pflügers Arch 414: 543-550, 1989.
Seow HF, Broer S, Broer A, Bailey CG, Potter SJ, Cavanaugh JA, and Rasko JE. Hartnup disorder is caused by mutations in the gene encoding the neutral amino acid transporter SLC6A19. Nat Genet 36: 1003-1007, 2004.
Stevens BR, Ross HJ, and Wright EM. Multiple transport pathways for neutral amino acids in rabbit jejunal brush border vesicles. J Membr Biol 66: 213-225, 1982.
Takanaga H, Mackenzie B, Suzuki Y, and Hediger MA. Identification of mammalian proline transporter SIT1 (SLC6A20) with characteristics of classical system imino. J Biol Chem 280: 8974-8984, 2005.
Verrey F, Closs EI, Wagner CA, Palacin M, Endou H, and Kanai Y. CATs and HATs: the SLC7 family of amino acid transporters. Pflügers Arch 447: 532-542, 2004.
Verrey F, Ristic Z, Romeo E, Ramadan T, Makrides V, Dave MH, Wagner CA, and Camargo SM. Novel renal amino acid transporters. Annu Rev Physiol 67: 557-572, 2005.
Wasserman JC, Delpire E, Tonidandel W, Kojima R, and Gullans SR. Molecular characterization of ROSIT, a renal osmotic stress-induced Na + -Cl - -organic solute cotransporter. Am J Physiol Renal Fluid Electrolyte Physiol 267: F688-F694, 1994.
作者单位：1 Institute of Physiology and Center for Integrative Human Physiology, University of Zürich, Zürich, Switzerland; and 2 Department of Biochemistry and Molecular Biology, University of Barcelona, Barcelona, Spain