Effects of truncation of the COOH-terminal region of a Na+-independent neutral and basic amino acid transporter on amino acid transport in Xenopus oocytes.

To determine the role of a neutral and basic amino acid transporter (NBAT) in amino acid transport, we microinjected several COOH-terminal deletion mutants of NBAT cRNA into Xenopus oocytes and measured transport activity for arginine, leucine, and cystine in the presence and absence of sodium. Wild-type NBAT significantly stimulated the uptake of all three amino acids 10-20-fold compared with controls. On the other hand, no mutant, except a Δ511-685 mutant, stimulated the uptake of these amino acids. The Δ511-685 mutant significantly increased the uptake of arginine. In the presence of sodium, the Δ511-685 mutant also increased the uptake of leucine. The Δ511-685 mutant did not stimulate cystine uptake in the presence or absence of sodium. The stimulation of arginine uptake by the Δ511-685 mutant was inhibited by a 100-fold excess of unlabeled leucine in the presence of sodium. Inhibition of L-arginine uptake by L-homoserine was seen only in the presence of sodium, and an increase in the inhibition of L-arginine uptake by L-histidine was seen when the extracellular pH was decreased. Furthermore, an inward current in oocytes injected with the Δ511-685 mutant was recorded electrophysiologically when basic amino acids were applied. Homoserine was also taken up, but sodium was necessary for their transport. These properties of the Δ511-685 mutant correspond to those of the y+ amino acid transporter. If NBAT is a component of the b0,+-like amino acid transport system, it is unlikely that a mutant protein (Δ511-685) is able to stimulate an endogenous y+-like transport system. These results suggest that NBAT functions as a activator of the amino acid transport system in Xenopus oocytes.

To determine the role of a neutral and basic amino acid transporter (NBAT) in amino acid transport, we microinjected several COOH-terminal deletion mutants of NBAT cRNA into Xenopus oocytes and measured transport activity for arginine, leucine, and cystine in the presence and absence of sodium. Wild-type NBAT significantly stimulated the uptake of all three amino acids 10 -20-fold compared with controls. On the other hand, no mutant, except a ⌬511-685 mutant, stimulated the uptake of these amino acids. The ⌬511-685 mutant significantly increased the uptake of arginine. In the presence of sodium, the ⌬511-685 mutant also increased the uptake of leucine. The ⌬511-685 mutant did not stimulate cystine uptake in the presence or absence of sodium. The stimulation of arginine uptake by the ⌬511-685 mutant was inhibited by a 100-fold excess of unlabeled leucine in the presence of sodium. Inhibition of L-arginine uptake by L-homoserine was seen only in the presence of sodium, and an increase in the inhibition of L-arginine uptake by L-histidine was seen when the extracellular pH was decreased. Furthermore, an inward current in oocytes injected with the ⌬511-685 mutant was recorded electrophysiologically when basic amino acids were applied. Homoserine was also taken up, but sodium was necessary for their transport. These properties of the ⌬511-685 mutant correspond to those of the y ؉ amino acid transporter. If NBAT is a component of the b 0,؉ -like amino acid transport system, it is unlikely that a mutant protein (⌬511-685) is able to stimulate an endogenous y ؉ -like transport system. These results suggest that NBAT functions as a activator of the amino acid transport system in Xenopus oocytes.
One of the more important disorders of proximal tubule transport is human cystinuria (1). This transport defect is characterized as an alteration in the cystine/lysine transporter, a system shared also with the dibasic amino acids ornithine and arginine (2,3). Recently, cDNAs encoding putative renal amino acid transporters have been cloned (designated rBAT and D2H for cDNAs from human kidney, rBAT for cDNA from rabbit kidney, and D2 and NAA-Tr for cDNA from rat kidney) (4 -8). rBAT/D2/D2H/NAA-Tr (which we term neutral and ba-sic amino acid transporter (NBAT) 1 ) functions as a neutral and basic amino acid transporter for system b 0,ϩ (4 -8). The specific mutations in the human NBAT gene have been identified in type I cystinuria patients (9 -14). These mutations nearly abolish the amino acid transport activity induced by NBAT in Xenopus oocytes (9). However, the predicted structure of the NBAT protein is not particularly hydrophobic and contains only one or four transmembrane-spanning domains, suggesting that it functions as a transport activator and a regulatory subunit (4 -6, 8). The predicted NBAT protein is believed to be a type II membrane glycoprotein, as has been shown for 4F2 (15,16). The 4F2 cell surface antigen is a 125-kDa disulfidelinked heterodimer composed of an 85-kDa glycosylated heavy chain and a 41-kDa nonglycosylated light chain (17,18). It was identified originally by the production of a mouse monoclonal antibody against the human T-cell tumor clone HSB-2 (18). Both Bertran et al. (15) and Wells et al. (16) have demonstrated that 4F2 cRNA injection into oocytes results in the stimulation of y ϩ -like transport activity but not system b 0,ϩ (15). The murine receptor for the ecotropic murine leukemia virus has been identified as being responsible for system y ϩ activity, a sodium-independent, cation-preferring amino acid carrier (19,20). The predicted protein for the ecotropic murine leukemia virus receptor contains several membrane-spanning domains and has no homology with 4F2 antigen (15,16). Thus, the function of the 4F2 proteins remains unclear. Furthermore, recent studies demonstrated that NBAT-mediated transport of neutral and dibasic amino acids is associated with net outward and inward currents, respectively, which may be caused by an exchange of neutral with dibasic amino acids (21)(22)(23).
To evaluate whether NBAT functions as a component of amino acid transporters (amino acid exchanger) or as a transport activator, we made several mutated forms of NBAT and assayed their ability to stimulate transport activity. The transport and electrogenic studies suggest that a mutant NBAT (⌬511-685) stimulates an endogenous y ϩ -like transport system in Xenopus oocytes distinct from the b 0,ϩ system activated by wild-type NBAT. These results suggest that NBAT functions as a transport activator. Cloning NBAT from a Human Kidney cDNA Library-RNA was * This work was supported by Grants-in aid for scientific research from the Ministry of Education and Science. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) D82326.
Construction of Mutants for NBAT-All deletion mutants were obtained by linearizing pBAT-11 with XbaI using exonuclease III to remove nucleotides from the 3Ј-end, S1 nuclease digestion of the 5Јoverhangs, and DNA repair synthesis with the Klenow fragment of Escherichia coli polymerase. The truncated fragments were gel purified and then ligated to SmaI-digested pBA, a vector containing a 3Ј-noncoding region and the poly(A) site of pBAT-11, in the presence of an oligonucleotide linker, 5Ј-CTAGACTAGTCTAG-3Ј, which contains stop codons in all three reading frames. Deletion mutants were sequenced at the junction as described (25).
Oocyte Injections, Transport Assays, and Two-electrode Voltage Clamp-Xenopus laevis females were obtained from Hamamatsu Jikken (Sizuoka, Japan) (14). Small clumps of oocytes were treated twice for 90 min with collagenase at 2 mg/ml in a Ca 2ϩ -free solution (ORII solution; 82.5 mM NaCl, 2 mM KCl, 1 mM MgCl 2 , and 10 mM Hepes/Tris, pH 7.5) to remove the follicular layer. Following extensive washing, first with ORII solution and then with modified Barth's solution (88 mM NaCl, 1 mM KCl, 0.82 mM MgSO 4 , 0.4 mM CaCl 2 , 0.33 mM Ca(NO 3 ) 2 , 2.4 mM NaHCO 3 , and 10 mM Hepes/Tris, pH 7.5), the oocytes were maintained in modified Barth's solution overnight at 18°C. Healthy stage V oocytes then were injected with cRNA (dissolved in water at a concentration of 0.5 mg/ml) or water using a manual injector (Narishige, Tokyo, Japan). Following incubation at 18°C for 3 days, transport activity was measured. To measure amino acid uptake, six or seven oocytes from each experimental group were incubated in medium containing L-[ 3 H]arginine, L-[ 3 H]leucine, and L-[ 35 S]cystine (Amersham). Oocytes first were washed for 30 s in solution A (100 mM choline chloride, 2 mM KCl, 1 mM CaCl 2 , 1 mM MgCl 2 , and 10 mM Hepes/Tris, pH 7.5). Uptake assays were performed with 50 M amino acids at 20 Ci/ml at 25°C for 30 min. When L-cystine uptake was measured, 10 mM diamide was added to prevent reduction of disulfide bonds. Inhibition by other amino acids was performed by their addition to the uptake solution at 5 mM. To measure uptake in the presence of sodium, choline chloride (100 mM) was replaced by NaCl (100 mM) in solution A. Following incubation, the oocytes were washed three times with cold Na ϩfree uptake solution containing 5 mM amino acids. Each oocyte then was transferred to a scintillation vial, dissolved in 0.2 ml of 10% sodium dodecyl sulfate, and counted in 5 ml of scintillation fluid. Data of uptake are expressed as pmol/min per oocyte in the basal condition (waterinjected oocytes) and in the stimulated condition (wild-type and mutant NBAT-injected oocytes) or as the increment of uptake due to the injected cRNA (uptake in cRNA-injected oocytes minus uptake in waterinjected oocytes). Student's t test was used for statistical analysis.
Resting membrane potential in the current-clamped mode and amino acid-evoked membrane currents were measured using a two-electrode voltage clamp recording system (GeneClamp 500; Axon Instruments, Inc., Foster City, CA). The current and voltage microelectrodes were filled with 1 M KCl and had resistances ranging from 7 to 14 M⍀. A membrane potential stabilization period was observed for 4 -10 min. Oocytes with membrane potentials less negative than Ϫ40 mV were discarded. For electrophysiological studies, oocytes were superfused with a medium containing either 100 mM NaCl or tetramethylammonium chloride and including 2 mM KCl, 1 mM CaCl 2 , 1 mM MgCl 2 , and 10 mM Hepes (pH 7.5) with Tris base (superfusion medium). Amino acids at the indicated concentrations were added to this solution. The temperature of the superfusion medium was controlled at 22 Ϯ 1°C. The membrane potential was then clamped to Ϫ50 mV. A data acquisition system and commercial software were used to send the voltage pulse protocol and simultaneously to record current and voltage signals.
Immunoprecipitation of [ 35 S]Methionine-labeled NBAT in Oocytes-Oocytes were injected with wild-type or mutant NBAT cRNA in a final volume of 50 nl (50 ng). After 30 h, [ 35 S]methionine (1 Ci in 50 nl of diethylpyrocarbonate-treated water; ICN Biochemical, Inc.) was injected, and the oocytes were incubated for 24 h at 18°C in 0.5 ml of modified Barth's solution (14). Single oocytes were triturated in phos-phate-buffered saline using a Gilson P-200 Pipetman. Oocyte plasma membrane "ghosts" were separated from cytoplasmic contents and collected into a pipette. Plasma membranes then were washed by resuspension in phosphate-buffered saline, followed by centrifugation. An intact oocyte ghost contained 2.5-3.0 g of plasma membrane protein (approximately 50,000 cpm). The pellets corresponding to membranes obtained from 15-20 oocytes were solubilized in buffer B (10 mM phosphate-buffered saline, pH 7.4, 0.07% Triton X-100, 0.07% SDS, 0.03% deoxycholate, 0.03% bovine serum albumin, and 1 mM phenylmethylsulfonyl fluoride). Samples were mixed with 10 g of affinity-purified antibody (anti-human NBAT rabbit IgG) and 8 l of protein A-Sepharose. The samples were then mixed at 4°C overnight. The pellets were then washed twice with 600 l of buffer B. Aliquots of these samples were used for electrophoresis. After electrophoresis, gels were dried, and fluorography was performed. For the production of the anti-NBAT antibody, the 15-amino acid sequence (CPRSFKDSDKDGNGD) of the NBAT protein deduced from human NBAT cDNA (amino acids 125-138) was synthesized. The NH 2 -terminal cystine residue was introduced for conjugation with keyhole limpet hemocyanin using m-maleimidobenzoyl-N-hydroxysuccinimide ester. After the peptide (1 mg) was conjugated with keyhole limpet hemocyanin and m-maleimidobenzoyl-N-hydroxysuccinimide ester (1.9 mg) and mixed with a Freund's complete adjuvant, the emulsion (100 g of peptide) was injected subcutaneously into rabbits four times at a 2-week interval. The antiserum after the fourth boost was affinity purified using a gel column (Cellulofine; Seikagaku-Kogyo, Tokyo, Japan) (27). RESULTS cDNA Cloning and Functional Expression-Two cDNA clones (pBAT-1 and pBAT-11) were isolated containing 2283 and 3355 base pairs of insert, respectively. Sequencing data indicated that the two clones have the same open reading frame, encoding a 685-amino acid protein, but have different polyadenylation signals (data not shown). pBAT-11 is 3355 base pairs in length as compared with the shorter pBAT-1, which is 2283 base pairs in length. pBAT-11 is approximately 1.1 kilobases longer in 3Ј-untranslated sequence upstream of its poly(A) signal. pBAT-1 contains an adenine residue at nucleotide 1897, whereas pBAT-11 contains a guanidine residue at that position, resulting in substitution of isoleucine for methionine at amino acid 618. The sequence of pBAT-1 is identical to that of human BAT cDNA reported by Bertran et al. (4). The sizes of two clones (2283 and 3355 base pairs) correspond fairly well to those of the transcripts (2.4 and 3.5 kilobase pairs) seen on Northern blots of kidney mRNA (Fig. 1A). Injection of equal amounts of pBAT-1 and pBAT-11 cDNAs stimulated the uptake of L-arginine and L-cystine to approximately the same degree (data not shown). In the present study, pBAT-11 was used for the construction of deletion mutants, as shown in Fig.  1B. The DNA sequencing of all deletion mutants was performed to confirm that the mutation did not affect other parts of the coding of pBAT-11.
Effects of Several COOH-terminal Deletion Mutants in Transport Activity-Several COOH-terminal deletions (⌬141- FIG. 1. Expression of the human NBAT gene in human kidney and construction of NBAT deletion mutants. A, Northern blot prepared using poly(A) ϩ RNA (10 g) from human kidney obtained from a patient with renal cancer (60 years old) and hybridized to a 32 Plabeled pBAT-11 clone and exposure to film for 3 days. B, all deletion mutants were created as described under ''Experimental Procedures.'' The detailed construction is described under "Experimental Procedures." 685, ⌬261-685, ⌬382-685, and ⌬511-685) in NBAT were made; cRNA was synthesized and injected into Xenopus oocytes; and the uptake of [ 35 S]cystine, [ 3 H]arginine and [ 14 C]leucine was determined. Wild-type NBAT injection into oocytes resulted in the induction of system b 0,ϩ -like amino acid transport; this system is responsible for sodium-independent uptake of L-arginine, L-leucine, and L-cystine (4 -7). As shown in Fig. 2, wild-type NBAT significantly stimulated the uptake of these three amino acids (L-arginine, 10.4 Ϯ 3.6 pmol/min per oocyte; L-leucine, 16.1 Ϯ 2.8 pmol/min per oocyte; and L-cystine, 8.6 Ϯ 2.8 pmol/min per oocyte in the absence of sodium; n ϭ 7), as previously reported by several groups (4,6). On the other hand, of the mutated cDNAs, only ⌬511-685 stimulated Lleucine and L-arginine uptake compared with water-injected oocytes. The ⌬511-685 mutant stimulated L-leucine uptake about 6-fold over control in the presence of sodium (5.6 Ϯ 1.4 pmol/min per oocyte in ⌬511-685 mutant NBAT-injected oocytes and 0.9 pmol/min per oocyte in water-injected oocytes; n ϭ 8). In contrast, in the absence of sodium, ⌬511-685 did not stimulate L-leucine uptake. Furthermore, this mutant could not stimulate uptake of L-cystine, either in the presence or absence of sodium (0.3-0.4 pmol/min per oocyte in ⌬511-685injected and water-injected oocytes). This pattern of ⌬511-685induced uptake is different from that seen in wild-type, NBATinjected oocytes. These observations suggest that the ⌬511-685 mutant stimulates another transport system (the cationspecific y ϩ -like system) but not the neutral and cationpreferring system (b 0,ϩ ). This activity exactly matches 4F2induced Xenopus oocyte amino acid transport activity, as reported by Bertran et al. (15) and Wells et al. (16).
Localization of ⌬511-685 NBAT Protein in Xenopus Oocytes-Next we investigated whether wild-type and mutant NBAT are present in the oocyte plasma membrane. A qualitative impression of the relative amount of the various mutant proteins in the oocyte plasma membrane was obtained by immunoprecipitation. These studies were performed using an affinity-purified antibody raised against a 16-residue synthetic peptide corresponding to the conserved region of human and rabbit NBAT proteins. Oocytes expressing wild-type NBAT and the ⌬511-685 mutant exhibited strong signals on filter membranes (Fig. 3), whereas signals from oocytes expressing other mutants were considerably weaker (data not shown). These data suggest that at least part of the reduced activity of several mutants was caused by a reduced plasma membrane content of the mutant protein, whereas the ⌬511-685 NBAT was present at levels in the plasma membrane similar to those of wild-type NBAT.
Characterization of NBAT Mutant ⌬511-685-We next tested whether stimulation of amino acid uptake by the ⌬511-685 mutant (sodium-independent uptake of L-arginine and sodium-dependent uptake of L-leucine) is due to stimulation of a y ϩ -like transport system. This transport system has been reported to be a major component of sodium-independent uptake of cationic amino acids in Xenopus oocytes (28). In addition, system y ϩ -like transport for cationic amino acids is inhibited by neutral amino acids, for which it has a higher affinity in the presence of sodium. To confirm that the amino acid uptake stimulated by ⌬511-685 corresponds to activation of a y ϩ -like transport system in Xenopus oocytes, we tested inhibition of uptake by various amino acids. As shown in Fig. 4, sodiumindependent L-arginine uptake in the presence of the amino acids tested was identical to the control activity in waterinjected oocytes and to the activity stimulated by the ⌬511-685 mutant. This profile is distinct from the y ϩ transport system in mammalian cells but is specific for a cationic transporter (y ϩlike system) in Xenopus oocytes (28). Methylaminoisobutyric acid, an amino acid analogue that is a specific substrate for the sodium-dependent neutral amino acid transport system A, had no effect on the uptake of L-arginine. We also observed that cystine, which is transportable by system b 0,ϩ and is not taken up by water or mutant NBAT (⌬511-685)-injected oocytes, caused a 60% decrease in L-arginine uptake but had no effect on L-leucine uptake. The profile of inhibition by cystine agrees with that reported by Wells et al. (16) for 4F2 antigen. In the presence of sodium in the incubation medium, both basal Larginine uptake (in water-injected oocytes) and uptake stimulation due to ⌬511-685 cRNA were completely inhibited (95%) by a 100-fold excess of unlabeled L-leucine. These results are consistent with stimulation by ⌬511-685 of a component of the transport pathway shared by L-arginine and L-leucine.
We also studied the sodium dependence of inhibition of ⌬511-685 activity by L-homoserine (Fig. 5). It has been demonstrated that the activity of system y ϩ is not inhibited by L-homoserine (5 mM), in the absence of sodium, whereas it is largely inhibited by this amino acid in the presence of sodium. In contrast, system b 0,ϩ interacts with L-homoserine in both the presence and absence of sodium. Fig. 5 shows that the basal L-arginine uptake of oocytes and the L-arginine uptake stimu-lated by ⌬511-685 are both inhibited by L-homoserine only in the presence of sodium. (L-Arginine uptake was inhibited only 11-20% by L-homoserine in the absence of sodium but was inhibited 70 -75% by L-homoserine in the presence of sodium.) We next measured L-histidine uptake inhibition as a function of extracellular pH (Fig. 6). Uptake of histidine by the y ϩ -like dibasic amino acid transport system is pH dependent, with increased uptake as the pH is lowered from neutral toward the pK a of histidine. The majority of L-histidine molecules are cationic, with a pK a of 6.1, and are transported by the y ϩ -like system. As shown in Fig. 6, L-histidine inhibition of L-arginine uptake in oocytes injected with the ⌬511-685 mutant in-   creased as the pH decreased (15). Furthermore, we determined the kinetics of arginine uptake in oocytes injected with the ⌬511-685 mutant. Uptake of arginine, as shown on a Michaelis-Menten plot, was saturable. The K m of ⌬511-685 mutant-induced arginine uptake was 31 M, similar to the K m (39 M) for water-injected oocytes (data not shown).
Electrophysiological Experiments-Wild-type or ⌬511-685 mutant NBAT cRNA-injected oocytes show different currents than control oocytes when immersed in Barth's solution. In the presence of 50 M L-leucine, outward currents of 30.2 Ϯ 1.9 nA are invariable seen at Ϫ50 mV in NBAT-expressing oocytes but not in control oocytes. These currents are independent of Na and Cl (data not shown). Similar currents were seen on exposure to alanine, isoleucine, serine, and glutamine. In contrast, addition of 50 M arginine to Barth's solution caused an inward current of 25 Ϯ 1.2 nA at Ϫ50 mV (Table I). No amino acidinduced currents were seen on application of proline, aspartic acid, or glutamic acid. System y ϩ is a sodium-independent transport for basic amino acids described above. In oocytes expressing the ⌬511-685 mutant, the current induced by arginine is also sodium independent (Table I). However, in the absence of sodium, outward currents were not seen on exposure to L-leucine (Table I). One characteristic of system y ϩ is that the transport of basic amino acids is inhibited by some neutral amino acids, such as homoserine, but only in the presence of sodium, as shown in Fig. 5. Homoserine itself induced an inward current in mutant NBAT-injected oocytes, but this was completely dependent on the presence of extracellular sodium (Table I). Thus, mutant NBAT transports basic amino acids in a sodium-independent manner and some neutral amino acids by a sodium-dependent process. These results indicate that mutant NBAT functions as the sodium-independent basic amino acid uptake system generally known as y ϩ .

DISCUSSION
To investigate whether NBAT functions as a component of transporter systems or as a transport activator, we made several deletion mutants of NBAT and analyzed their effects on amino acid transport in Xenopus oocytes. Most of the mutants failed to stimulate L-arginine, L-cystine, or L-leucine transport. One COOH-terminal deletion mutant (⌬511-685), with a deletion in a leucine zipper motif, stimulated L-leucine transport in the presence of sodium and the sodium-independent uptake of L-arginine. Mutant NBAT (⌬511-685) cRNA-stimulated, sodium-independent uptake of L-arginine was inhibited completely by L-leucine in the presence of sodium. The stimulation both of sodium-dependent uptake of L-arginine and L-leucine induced by mutant NBAT (⌬511-685) cRNA was inhibited completely by basic L-amino acids. In voltage clamp experiments, dibasic amino acids induced inward currents in Xenopus oocytes injected with the ⌬511-685 mutant, but neutral amino acids did not induce outward currents. These results demonstrate that mutant NBAT (⌬511-685) may stimulate a cation-preferring amino acid transport system identical to the y ϩ -like transport already present in oocytes. If NBAT is a component of the amino acid transport system, it is unlikely that mutant NBAT (⌬511-685) stimulates other amino acid transport systems. These results strongly suggest that NBAT functions as a transport activator.
As shown in Fig. 8A, Mosckovitz et al. (29) have demonstrated that NBAT has four transmembrane domains using site-directed polyclonal antibodies directed against NBAT. Two mutations (M467T and M467K), which so far are the most common cystinuric mutations in the Spanish and Italian populations, are located in the third transmembrane domain (9,12,13). These mutations abolish the amino acid transport activity induced by NBAT in oocytes (9), suggesting that the transmembrane domain of NBAT is important to stimulate the amino acid transport activity. However, the ⌬511-685 mutant, which is truncated at the fourth transmembrane domain, can activate the y ϩ -like system in oocytes. If wild-type NBAT has four membrane-spanning domains, the truncation, which at most deletes one of the four transmembrane segments, causes a different topology of the rest of the protein, which is highly unlikely. In addition, NBAT protein and 4F2 have a very similar localization of the putative transmembrane domains within their sequences (15,16). Both Bertran et al. (15) and Wells et al. (16) have found significant amino acid sequence homology between 4F2 and NBAT, with 26% identity and 45% FIG. 7. Structural homologies between 4F2 heavy-chain (4F2hc) and NBAT. The aligned NBAT and 4F2hc protein sequences show 26% identity (45% similarity) in their overall amino acid sequence, as described by Bertran et al. (15). Four amino acid sequence fragments shaded black (10 -18 amino acid residues long) are highly conserved (67-80% identity) in the three proteins. M1, transmembrane domains; H1-H4, highly conserved regions. similarity overall between the two amino acid sequences. As shown in Fig. 7, four regions in the amino acid sequences are highly conserved, with homologues of 67-80% between the two proteins (15). The ⌬511-685 mutant has a deletion in one of these highly conserved regions the leucine zipper motif of NBAT. However, the structure of mutant NBAT (⌬511-685) cRNA still may be similar to that seen with the 4F2 antigen. We therefore suggest that NBAT has one transmembrane domain, as shown in Fig. 8B. Furthermore, the COOH-terminal (mutant 511-685) of NBAT is important for determining its specificity as a transport activator. Further work is needed to clarify the transport systems associated with members of the type II membrane glycoprotein family.