Advertisement

Functional Characterization of Two Novel Mammalian Electrogenic Proton-dependent Amino Acid Cotransporters*

  • Michael Boll
    Affiliations
    From the Molecular Nutrition Unit, Institute of Nutritional Sciences, Technical University of Munich, D-85350 Freising-Weihenstephan, Germany
    Search for articles by this author
  • Martin Foltz
    Affiliations
    From the Molecular Nutrition Unit, Institute of Nutritional Sciences, Technical University of Munich, D-85350 Freising-Weihenstephan, Germany
    Search for articles by this author
  • Isabel Rubio-Aliaga
    Affiliations
    From the Molecular Nutrition Unit, Institute of Nutritional Sciences, Technical University of Munich, D-85350 Freising-Weihenstephan, Germany
    Search for articles by this author
  • Gabor Kottra
    Affiliations
    From the Molecular Nutrition Unit, Institute of Nutritional Sciences, Technical University of Munich, D-85350 Freising-Weihenstephan, Germany
    Search for articles by this author
  • Hannelore Daniel
    Correspondence
    To whom correspondence should be addressed. Tel.: 49-8161-71-3400; Fax: 49-8161-71-3999;
    Affiliations
    From the Molecular Nutrition Unit, Institute of Nutritional Sciences, Technical University of Munich, D-85350 Freising-Weihenstephan, Germany
    Search for articles by this author
  • Author Footnotes
    * 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.The nucleotide sequence(s) reported in this paper has been submitted to the GenBank™/EBI Data Bank with accession number(s) (mPAT1) and (mPAT2).
Open AccessPublished:April 16, 2002DOI:https://doi.org/10.1074/jbc.M200374200
      We cloned two cDNAs encoding proton/amino acid cotransporters, designated as mPAT1 and mPAT2, from murine tissues. They were identified by sequence similarity to the amino acid/auxin permease family member of lower eukaryotes. We functionally characterized both transporters by flux studies and electrophysiology after expression in Xenopus laevis oocytes. Both mPAT1 and mPAT2 induced a pH-dependent electrogenic transport activity for small amino acids (glycine, alanine, and proline) that is altered by membrane potential. Direct evidence for amino acid/H+-symport was shown by intracellular acidification, and a flux coupling stoichiometry for proline/H+-symport of 1:1 was determined for both transporters. Besides small apolar l-amino acids, the transporters also recognize their d-enantiomers and selected amino acid derivatives such as γ-aminobutyric acid. The mPAT1 transporter, the murine orthologue of the recently cloned rat LYAAT-1 transporter, can be considered as a low affinity system when compared with mPAT2. The mRNA of mPAT1 is highly expressed in small intestine, colon, kidney, and brain; the mPAT2-mRNA is mainly found in heart and lung. Phenotypically, the PAT1 transporter possesses the same functional characteristics as the previously described proton-dependent amino acid transport process in apical membranes of intestinal and renal epithelial cells.
      AAAP
      amino acid/auxin permease
      PAT
      proton/amino acid transporter
      GABA
      γ-aminobutyric acid
      VGAT
      vesicular GABA transporter
      LYAAT-1
      lysosomal amino acid transporter1
      EST
      expressed sequence tag
      ORF
      open reading frame
      MES
      4-morpholineethanesulfonic acid
      contig
      group of overlapping clones
      EGFP
      enhanced green fluorescent protein
      One of the largest families of amino acid transporters identified so far is the amino acid/auxin permease (AAAP)1 family (
      • Saier M.H., Jr.
      ,
      • Young G.B.
      • Jack D.L.
      • Smith D.W.
      • Saier M.H., Jr.
      ). This protein family is widespread in nature, with members found in yeast, plants, invertebrates, and mammals. Although their amino acid sequences appear very divergent, a common signature has been identified, and secondary structure predictions show conserved similarities (
      • Young G.B.
      • Jack D.L.
      • Smith D.W.
      • Saier M.H., Jr.
      ). Accordingly, most members of the AAAP family possess 11 membrane-spanning domains with very similar hydropathy profiles. Functionally characterized members as AAP1 and AVT1 found inSaccharomyces cerevisiae and Arabidopsis thalianaoperate as proton symporters for amino acids and selected derivatives such as auxins (
      • Young G.B.
      • Jack D.L.
      • Smith D.W.
      • Saier M.H., Jr.
      ,
      • Boorer K.J.
      • Frommer W.B.
      • Bush D.R.
      • Kreman M.
      • Loo D.D.
      • Wright E.M.
      ,
      • Russnak R.
      • Konczal D.
      • McIntire S.L.
      ). Three subfamilies within the AAAP family that show low although significant similarities (20–30%) have been identified so far in mammalian organisms. They are represented by the VGAT transporter group, the system A/N transporter group, and the LYAAT branch (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ).
      The vesicular γ-aminobutyric acid (GABA) transporter VGAT was the first identified mammalian member of the AAAP family with an assigned function (
      • McIntire S.L.
      • Reimer R.J.
      • Schuske K.
      • Edwards R.H.
      • Jorgensen E.M.
      ). It mediates the uptake of the inhibitory neurotransmitters GABA and glycine into synaptic vesicles (
      • McIntire S.L.
      • Reimer R.J.
      • Schuske K.
      • Edwards R.H.
      • Jorgensen E.M.
      ,
      • Sagne C., El
      • Mestikawy S.
      • Isambert M.F.
      • Hamon M.
      • Henry J.P.
      • Giros B.
      • Gasnier B.
      ). The driving force of VGAT is the electrochemical proton gradient generated by the vesicular H+-ATPase (
      • Forgac M.
      ). Cytoplasmic GABA and glycine can therefore be accumulated within the vesicles via a proton antiport mechanism (
      • McIntire S.L.
      • Reimer R.J.
      • Schuske K.
      • Edwards R.H.
      • Jorgensen E.M.
      ). The system A/N transporter group also belong to the AAAP family. System N plays an important role especially in the homeostasis of glutamine in brain, where it serves as a delivery system for the precursor of the excitatory amino acid glutamate (
      • Chaudhry F.A.
      • Reimer R.J.
      • Krizaj D.
      • Barber D.
      • Storm-Mathisen J.
      • Copenhagen D.R.
      • Edwards R.H.
      ). The system N-mediated glutamine influx is coupled to the cotransport of one Na+ ion in antiport with one H+ ion (
      • Chaudhry F.A.
      • Reimer R.J.
      • Krizaj D.
      • Barber D.
      • Storm-Mathisen J.
      • Copenhagen D.R.
      • Edwards R.H.
      ). Although system A transporters are highly homologous to system N, they show a very different substrate specificity as well as a different ion-substrate flux coupling ratio (
      • Reimer R.J.
      • Chaudhry F.A.
      • Gray A.T.
      • Edwards R.H.
      ). System A transporters prefer alanine and most of the other neutral amino acids and are considered to provide bulk quantities of these amino acids in most cell types. System A carriers act as Na+/amino acid symporters. Although transport by system A is sensitive to low pH, no proton exchanger activity has been shown in contrast to system N (
      • Yao D.
      • Mackenzie B.
      • Ming H.
      • Varoqui H.
      • Zhu H.
      • Hediger M.A.
      • Erickson J.D.
      ).
      Recently, Sagne et al. (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ) report the cloning of the first mammalian proton/amino acid symporter from rat brain, which is also a member of the AAAP family (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ). This lysosomal amino acid transporter, designated as LYAAT-1, when transfected in CV-1 cells mediated the uptake of small amino acids such as alanine and proline and the neurotransmitter GABA. The observation of an enhanced transport activity by low extracellular pH suggested a proton-symport mechanism for amino acid influx. The LYAAT-1 transcripts were identified in glutamatergic and GABAergic neurons in rat brain. The LYAAT-1 protein was localized in the lysosomal membrane in these neurons, although when expressed heterologously, it mediated plasma membrane amino acid uptake (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ).
      The first members of the system A/N subfamily, the SN1 protein, and LYAAT-1 have been identified by data base search for mammalian homologues of the VGAT transporter (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ,
      • McIntire S.L.
      • Reimer R.J.
      • Schuske K.
      • Edwards R.H.
      • Jorgensen E.M.
      ,
      • Chaudhry F.A.
      • Reimer R.J.
      • Krizaj D.
      • Barber D.
      • Storm-Mathisen J.
      • Copenhagen D.R.
      • Edwards R.H.
      ). We searched for mammalian AAAP family members based on the S. cerevisiae protein sequence YKL146wp, which had been identified by genome-scanning analysis as a putative amino acid transport protein. Recently it was shown that YKL146wp indeed codes for a protein that represents a vesicular amino acid transporter, designated as AVT3 (
      • Russnak R.
      • Konczal D.
      • McIntire S.L.
      ). Here we report the identification and characterization of two new members, the murine orthologue of rLYAAT-1, designated as mPAT1, and an additional member of this subfamily, designated as mPAT2. Both transporters, when expressed in Xenopus laevis oocytes, show the functional characteristics of electrogenic H+/amino acid symporters. The mPAT2 transporter displays the characteristics of a high affinity system with a more restricted substrate specificity than mPAT1.

      RESULTS

      Using an in silico cloning strategy, we identified two closely related cDNA sequences from mice based on the yeast protein sequence YKL146wp, designated as mPAT1- and mPAT2-cDNA. The mPAT2-cDNA is represented by an EST clone from a murine 14 days post-coital embryonic cDNA library (IMAGE Clone I.D. 1920302). The complete ORF of mPAT1 has been amplified by reverse transcriptase-PCR using cDNA-specific primers flanking the mPAT1-ORF with RNA purified from murine small intestine. The deduced proteins possess 475 (mPAT1) and 478 (mPAT2) residues (Fig.1). The mPAT1 and mPAT2 proteins show an identity of 69% and a similarity of 80%. Hydropathy analyses predict both mPAT proteins to have 11 membrane-spanning domains (Fig. 1). Within the putative transmembrane domains the amino acid sequences show highest homology, whereas the N terminus, most likely the cytosolic part, is divergent. Three conserved putative N-glycosylation sites were found at the proposed luminal/extracellular site of the mPAT proteins when assuming that the N terminus is directed to the cytosol. The only known mammalian protein that is closely related to the mPAT proteins is the recently identified rat LYAAT-1 (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ). LYAAT-1 is the rat orthologue of the mPAT1 protein with a similarity of 97%. The mPAT proteins and rLYAAT-1 belong to the large eukaryotic AAAP family (
      • Young G.B.
      • Jack D.L.
      • Smith D.W.
      • Saier M.H., Jr.
      ). In mammals, functionally identified AAAP family members are the vesicular GABA transporter VGAT and the system A/N transporters. The mPAT proteins build a new subfamily beside the VGAT and system A/N subfamily. Analysis of the first version of the human genome sequences and human EST sequences shows that the 2 human orthologue genes are located on chromosome 5q31–33 within a region less than 150 kb on the contig NT_006951.7. The predicted human orthologue proteins share a similarity of 91% (hPAT1) and 88% (hPAT2) to the corresponding murine transporter proteins.
      Figure thumbnail gr1
      Figure 1Multiple sequence alignment of the murine PAT transporters and the rLYAAT-1 protein. The alignment was performed with the CLUSTALW program package. The putative transmembrane domains are shown above the sequences by lines. Identical and conserved amino acid positions are indicated by specific symbols below the alignment. Asterisks, identical residues;colons, conserved substitutions; periods, semi-conserved substitutions. Conserved putativeN-glycosylation sites are indicated by dots above the lines.

      Transport Properties of mPAT1 and mPAT2

      To determine whether the mPAT proteins indeed function as rheogenic amino acid transporters, we applied the two-electrode voltage clamp technique and employed flux studies with radiolabeled amino acids in X. laevis oocytes expressing mPAT1 or mPAT2. Fig.2 A shows the uptake of selected amino acids provided in the medium in concentrations of 100 μm. The uptake rates of glycine, alanine and proline were highly elevated in mPAT1- and mPAT2-expressing oocytes, whereas tyrosine uptake was not significantly enhanced when compared with uptake in water-injected control oocytes. Under voltage clamp conditions (membrane potential −60 mV) glycine, alanine, and proline induced positive inward currents (Fig. 2, B andC). The substrate-evoked currents of the three amino acids were almost equal in mPAT1-cRNA-injected oocytes, whereas mPAT2 produced the highest inward currents with glycine followed by alanine and proline (about 70 and 40% of glycine currents, respectively). As given by the current voltage relationship in Fig. 2 D, glycine (20 mm) evoked currents increased with hyperpolarization of the membrane, suggesting that the membrane voltage is a driving force. This voltage dependence of substrate currents was more pronounced in mPAT1- than in mPAT2-expressing oocytes. Currents generated by mPAT1-expressing oocytes increased 4-fold by shifting the membrane potential from 0 to −120 mV, whereas in mPAT2-expressing oocytes currents increased only about 60%.
      Figure thumbnail gr2
      Figure 2Basic functional characteristics of the mPAT1 and mPAT2 transporters expressed in X. laevisoocytes. A, oocytes (n = 10) were incubated in Na+-free uptake buffer at pH 6.5 for 10 min with 100 μm3H-labeled amino acids as indicated. The uptake rates of Gly, Ala, and Pro were elevated in mPAT1 (black bars)- and mPAT2 (gray bars)-expressing oocytes, whereas tyrosine uptake was not significantly different when compared with uptake in control oocytes (white bars). Applications of Gly, Ala, and Pro (20 mm) induced positive inward currents in mPAT1 (B)- and mPAT2 (C)-expressing oocytes at a holding potential of −60 mV under Na+-free conditions at a pH of 6.5. The upper trace in B represents the current recordings in water-injected control oocytes. D, the glycine-evoked currents in mPAT1- and mPAT2-cRNA-injected oocytes are dependent on the membrane potential. Hyperpolarization of the membrane increased glycine-evoked currents more potently in mPAT1 than in mPAT2 expressing oocytes (n = 3).
      To analyze in detail the underlying mechanisms that caused the transport currents of mPAT1 and mPAT2, we first studied the effect of different cations and anions on the glycine-evoked currents. Replacement of Na+ by choline or K+ and Cl by isethionate had no significant effect on normalized mPAT1 and mPAT2 transport activity (Fig.3 A). On the other hand, lowering the extracellular pH led to a pronounced increase in mPAT1 and mPAT2 mediated l-proline influx (Fig. 3 B). The mPAT1 transport activity was much more strongly influenced by changes of the extracellular pH than that of mPAT2. The uptake rate of 100 μm l-Pro at pH 5.5 was 6-fold higher than that at pH 7.5 in mPAT1-expressing oocytes, whereas its stimulation in mPAT2-expressing oocytes was only increased by 45%. Direct evidence for a proton/amino acid symport mechanism was obtained by measuring intracellular pH changes by a hydrogen ion-selective electrode simultaneously with changes in membrane potential when perfusing oocytes in the absence or presence of 20 mm glycine or proline. As shown in Fig. 3 C, the addition of glycine and proline to the perfusate led to an intracellular acidification as well as to a depolarization of oocytes expressing mPAT1. These effects were completely reversed by washing out the amino acids. The same effect was observed for glycine in mPAT2-expressing oocytes, whereas leucine, which is no substrate of the transporters, failed to induce an acidification and membrane depolarization. Water-injected control oocytes did not respond to the addition of any applied amino acid (data not shown).
      Figure thumbnail gr3
      Figure 3Ion dependence of mPAT1- and mPAT2-mediated transport activity. A, the equimolar replacement of 100 mm NaCl by choline chloride, Na+ isethionate, or KCl had no significant effect on glycine-evoked currents under voltage clamp conditions (−60 mV, 20 mm Gly, pH 6.5,n = 6 oocytes). Currents are normalized to the glycine-evoked currents in NaCl-containing medium. B, oocytes (n = 10) were incubated in Na+-free uptake buffer for 10 min with 100 μm l-[3H]proline at different medium pH values (pH 5.5–8.5). At pH 8.5 there was hardly any stimulation of proline uptake in cRNA-injected oocytes, whereas the uptake at pH 5.5 was stimulated 26-fold (mPAT1) and 11-fold (mPAT2) when compared with the uptake of water-injected oocytes. Intracellular pH changes in mPAT1 (C)- and mPAT2 (D)-expressing oocytes were recorded with an hydrogen ion-selective electrode (upper traces) simultaneously with the changes in membrane potential (lower traces). The time periods when oocytes were perfused with Na+-containing buffer at pH 7.4 (white boxes) and pH 6.5 (checkered boxes) and when 20 mm of the amino acid was present (black lines) are indicated. In mPAT1-expressing oocytes, the addition of 20 mm Gly and l-Pro led to a membrane depolarization and a concomitant intracellular acidification. The same effect of Gly but not of l-Leu was observed in mPAT2-expressing oocytes.
      The H+/l-proline-coupling stoichiometry was directly determined by comparing thel-proline-dependent charge (QPro) with the concomitant accumulation rate ofl-[3H]proline in individual oocytes voltage-clamped at −60 mV (Fig.4 A). QProcorrelated with the 3H accumulation in a linear fashion with a slope of 0.97 (mPAT1) and 1.28 (mPAT2). The H+/l-proline-coupling coefficients (n) were 1.22 ± 0.04 (mPAT1) and 1.28 ± 0.08, and the mean QPro values were not significantly different from the 3H accumulation rates in a paired ttest (Fig. 4 B). This convincingly shows that the H+-coupled l-proline transport via mPAT1 and mPAT2 occurs with 1:1 stoichiometry.
      Figure thumbnail gr4
      Figure 4H+l-proline-coupling coefficient determined byl-[3H]proline uptake under voltage clamp conditions. Oocytes were clamped at −60 mV, and 500 μm l-[3H]proline was applied for 5 min at a medium pH of 6.5. A, the proline-dependent charge (QPro) was compared with the concomitant tracer accumulation rate in the same oocytes expressing mPAT1 (squares) or mPAT2 (triangles). Tracer uptake in control oocytes (42.2 ± 3.8 pmol·5 min−1, n = 5 oocytes) was first subtracted. The ratio of QPro:3H was 1.22 ± 0.04 for mPAT1 and 1.28 ± 0.08 for mPAT2 (mean ± S.E.,n = 6 oocytes), and there was no significant difference between QPro (open bars) and 3H (filled bars) accumulation (B) according to Student's paired t test.
      Both transporters are very restrictive regarding substrate recognition. We tested all 20 proteinogenic l-amino acids for their ability to induce positive inward currents in mPAT1- and mPAT2-expressing oocytes at concentrations of 20 mm. Only amino acids with short side chains such as glycine, alanine, proline, and to a much smaller extent serine were able to interact with the transporters substrate binding sites (Fig.5, A and B). No other proteinogenic amino acid was able to induce inward current significantly above background oocyte activity. On the other hand, the transporters are not very enantioselective, as shown by the inward currents induced by d-alanine, d-proline, andd-serine in Fig. 5, C and D. In oocytes expressing mPAT1 d-alanine- andd-proline-evoked currents were almost as high as those of glycine, whereas in the case of mPAT2, all d-enantiomers induced less than 30% of the glycine currents. In addition, β-alanine and the neurotransmitter GABA induced positive inward currents more potently in mPAT1- than in mPAT2-expressing oocytes (Fig.5, C and D).
      Figure thumbnail gr5
      Figure 5Substrate specificity of the mPAT1 and the mPAT2 transporter. Oocytes expressing mPAT1 (A) or mPAT2 (B) were held at a membrane potential of −60 mV and perfused with Na+-free buffer at pH 6.5 in the absence or presence of 20 mm indicated amino acid (one-letter code). The amino acid-evoked currents are displayed relative to glycine-induced currents (I Gly = 1856 ± 171 and 623 ± 40 nA for mPAT1 and mPAT2, respectively) as the mean ±S.E. (n = 6 oocytes). Amino acid-evoked currents in water-injected oocytes were, when detectable, subtracted. Thelower panel shows representative current traces obtained in oocytes expressing mPAT1 (C) and mPAT2 (D) when perfused with selected d-amino acids, β-alanine, and GABA at concentrations of 20 mm in Na+-free buffer at pH 6.5. The time period when the compound was present in the perfusate is marked by the black lines.
      Dose-response experiments for currents in voltage-clamped oocytes injected with mPAT1- or mPAT2-cRNA were performed for selected amino acids and derivatives. Net substrate-evoked currents at five different concentrations were calculated, and after Eadie-Hofstee transformation and linear regression analysis, the apparent affinities (K m values) were determined. TableI summarizes the obtained apparentK m values. The proteinogenic amino acid with the highest affinity was l-Pro, with apparentK m values of 2.8 mm (mPAT1) and 0.12 mm (mPAT2). In general, mPAT2 was found to have a more than 10-fold higher affinity for all three l-α-amino acids when compared with the affinities of mPAT1. The very low affinity ofl-serine of >40 mm for both transporters shows the very strict specificity, as the introduction of the polar hydroxy group almost abolishes affinity. d-Alanine has almost the same affinity for both transporters (K m values of 6.3 and 6.5 mm for mPAT1 and mPAT2, respectively), again demonstrating that mPAT2 is more enantioselective than mPAT1. In addition, GABA is a very good substrate of mPAT1 (K m = 3.1 mm) but a poor substrate of mPAT2 (K m = 30.9 mm). In summary, mPAT2 therefore represents a high affinity-type PAT transporter and is more restrictive in substrate recognition when compared with the low affinity type mPAT1.
      Table IApparent affinities of selected amino acids and derivatives of proton/amino acid cotransporters
      SubstrateK m valueIn vitro Km value
      mPAT1mPAT2Caco-2Renal BBMV
      mmmm
      Glycine7.0 ± 0.70.59 ± 0.045.3 ± 0.4 (13)3.9 (17)
      l-Alanine7.5 ± 0.60.26 ± 0.055.7 ± 0.7 (14)4.4 (18)
      l-Proline2.8 ± 0.10.12 ± 0.029.2 ± 0.9 (15)4.5 (19)
      l-Serine69 ± 543 ± 4NDND
      d-Alanine6.3 ± 0.76.5 ± 1.1ND7.86 ± 0.38 (20)
      GABA3.1 ± 0.230.9 ± 0.11.95 ± 0.78 (16)ND
      Substrate-dependent transport kinetics of mPAT1 and mPAT2 were determined by two-electrode voltage clamp for the substrates indicated (5 concentrations) in Na+-free uptake solution (pH 6.5). Substrate-evoked currents were transformed according to Eadie-Hofstee, and K m values were calculated by linear regression analysis. Data represent the mean ± S.E. of six oocytes. For comparison, K m values taken from the literature (in parentheses) on PAT-like transport activity, as found in the apical membranes of Caco-2 cells and in renal brush border membrane vesicles (BBMV) from pars convoluta, are shown. ND, not determined.

      Tissue Distribution of the mPAT Subtypes

      Northern blot analyses revealed that the mPAT mRNAs are differentially expressed in murine tissues (Fig. 6 A). The 5.0-kb mPAT1 transcript could be detected at higher levels in small intestine, kidney, brain, and colon. After longer exposure times, faint bands were also visible in all other tissues tested, excluding testis (data not shown). The mPAT2 mRNA (2.5 kb) was most abundantly expressed in lung and heart. Weaker signals could also be detected in kidney, testis, muscle, and spleen.
      Figure thumbnail gr6
      Figure 6Tissue distribution of the mPAT mRNA subtypes and subcellular localization of the mPAT2-EGFP fusion protein. A, the Northern blot was performed on murine Poly(A)+ RNA isolated from different tissues using mPAT subtype-specific cDNA probes. As a control for the RNA integrity, the blot was also probed for β-actin. B, after transfection of a mPAT2-EGFP construct in HeLa cells, most of the green fluorescence appeared in the perinuclear region (asterisks) but also partly in the plasma membrane (arrows, upper left). Cells were stained with the LysoTrackerRed dye, which accumulates in acidic organelles, mainly lysosomes (upper right). In the superimposed image hardly any colocalization is observed (lower image). Scale bars are shown in the upper images.

      Subcellular Localization of an mPAT2-EGFP Fusion Protein in Transfected HeLa Cells

      Although functional expression in oocytes suggested the transporters to be integral plasma membrane proteins, we used the EGFP-tagging approach to study the cellular localization of mPAT2. The mPAT2 protein, when tagged C-terminally with the EGFP protein was obviously translocated to the plasma membrane of HeLa cells (Fig. 6 B). Substantial fluorescence was also found in the perinuclear compartment, most likely represented by the endoplasmic reticulum and the Golgi apparatus. However, currently we do not have any direct experimental evidence for this assumption. Because the rat orthologue of mPAT1, the rLYAAT-1 protein, has been colocalized in rat brain to lysosomes in neurons (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ), we performed costaining experiments of the mPAT2-EGFP fusion protein in transfected HeLa cells with the LysoTrackerRed dye, which accumulates in acidic compartments, mainly lysosomes. When fluorescence of both dyes was merged, no significant colocalization was observed, suggesting that mPAT2 is not an integral protein of lysosomal membranes, at least not when overexpressed in HeLa cells (Fig. 6 B).

      DISCUSSION

      We identified by homology screening in expressed sequence tag databases two closely related murine transporter proteins, designated as mPAT1 and mPAT2, which belong to the eukaryotic AAAP family (
      • Young G.B.
      • Jack D.L.
      • Smith D.W.
      • Saier M.H., Jr.
      ). The two deduced proteins could be functionally characterized as rheogenic proton/amino acid symporters with a specificity for amino acids with small side chains. Together with the rat LYAAT-1 transporter (the orthologue of PAT1), the mPAT proteins build a new subfamily next to the VGAT and system A/N transporter subfamilies within the phylogenetic tree of the AAAP family. Similar to mPAT1/rLYAAT-1 (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ), a human orthologue of mPAT2 can be identified in the first version of the human genome. That the PAT2 gene is indeed expressed in human tissues in vivo is demonstrated by the presence of cDNA sequence fragments in the public EST data base (e.g.GenBankTM accession numbers BE501426 and BF511146).
      Both mPAT transporters display similar functional properties when expressed in X. laevis oocytes. We show that the PAT carriers increase specifically the influx of small amino acids and demonstrate for the first time the electrogenic nature of the transport mechanism. Because all substrate amino acids are zwitterionic in the pH range studied, net charge movement across the plasma membrane is due to H+ influx stoichiometrically coupled to amino acid import. This was directly proven by a substrate-specific intracellular acidification induced by mPAT transport activity. No evidence was found for interactions of other tested cations or anion (Na+, K+, Cl), suggesting that protons are the only ionic species cotransported with amino acids by mPAT1 and mPAT2. Transport activity of mPAT1 was highly dependent on membrane voltage and on extracellular pH in contrast to mPAT2, found to be only slightly influenced by both factors. This suggests that activity of mPAT1, but less so mPAT2, could be regulated by changes in membrane potential and pH within the physiological range. Whether these factors influence only the transport velocity or also substrate or H+ affinities has to be determined.
      Both transporters display a narrow substrate specificity, preferentially accepting amino acids with small apolar side chains, such as glycine, alanine, and proline, whereas l-Ser is only poorly recognized. The mPAT2 protein has much higher affinities to the amino acids than mPAT1, with affinities 12–29-fold higher for Gly,l-Ala, and l-Pro under the same experimental conditions. The introduction of a hydroxy group (l-Ser) in the side chain dramatically diminishes the substrate affinity for both transporters (10- and >70-fold) in comparison to theK m observed for l-Ala. Moreover, the introduction of a sulfhydryl group as in l-Cys abolishes substrate binding completely. The mPAT2 transporter shows a more restrictive substrate recognition pattern and a more pronounced enantioselectivity with only small neutral l-α-amino acids serving as high affinity substrates, whereas GABA and β-alanine are only poorly transported. Besides these functional differences, the mRNA transcripts of the transporters also show differential expression patterns in murine tissues. The mPAT1 mRNA has been detected in most of the examined tissues, with abundant expression in small intestine, colon, kidney, and brain, whereas highest signal levels of the mPAT2 mRNA were obtained in heart and lung. The physiological significance of the differential distribution of proton-dependent small amino acid transporters with different functional properties in the organism has to be investigated.
      The only functionally known and characterized mammalian homologue of the new murine PAT transporters is the lysosomal amino acid transporter (rLYAAT-1) from Rattus norvegicus. It most likely represents the first lysosomal proton/amino acid symporter found in mammals (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ). rLYAAT-1, which has been functionally characterized after transient transfection in the fibroblast cell line CV-1, displays a very similar substrate specificity and shows similar kinetics as mPAT1, which represents the murine orthologue. Only two minor differences were obtained regarding functional properties. First, l-serine did not inhibit GABA uptake into rLYAAT-1-transfected CV-1 cells at a concentration of 0.5 mm (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ), whereas in our experiments 20 mm l-serine induced significant inward currents in mPAT1-expressing oocytes. The most simple explanation might be the very low affinity of this amino acid to mPAT1, with an apparentK m value of 69 mm, which may prevent inhibition. Second, the GABA affinity for rLYAAT-1 has been determined with a K m value of 0.5 mm (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ), whereas GABA displayed a K m of 3.1 mm in the case of mPAT1. This difference, however, may be explained by the different experimental conditions such as pH (5.5 versus6.5) and membrane potential (open circuit versusvoltage-clamped) or species differences (rat versusmouse).
      Regarding the physiological importance of the PAT carriers in mammals, it may be considered that the two proteins described here are as well lysosomal transporters responsible for the export of small amino acids from the lysosomal lumen into the cytoplasm. However, as already stated by Sagne et al. (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ), no direct functional evidence for such systems has been found in lysosomal preparations. Moreover, the functionally characterized lysosomal small amino acid transport systems f and p have significantly different substrate specificities and affinities and appear not to act as proton symporters (
      • Pisoni R.L.
      • Flickinger K.S.
      • Thoene J.G.
      • Christensen H.N.
      ). On the other hand, in comparing our functional data with amino acid influx studies in cell lines and membrane vesicle preparations, we found a striking similarity between mPAT1 and the proton-dependent amino acid transport systems previously identified in the apical membrane of the intestinal Caco-2 cells (
      • Thwaites D.T.
      • McEwan G.T.
      • Simmons N.L.
      ,
      • Thwaites D.T.
      • McEwan G.T.
      • Brown C.D.
      • Hirst B.H.
      • Simmons N.L.
      ,
      • Thwaites D.T.
      • McEwan G.T.
      • Cook M.J.
      • Hirst B.H.
      • Simmons N.L.
      ,
      • Thwaites D.T.
      • Basterfield L.
      • McCleave P.M.
      • Carter S.M.
      • Simmons N.L.
      ) and renal brush border membrane vesicle preparations (
      • Roigaard-Petersen H.
      • Jessen H.
      • Mollerup S.
      • Jorgensen K.E.
      • Jacobsen C.
      • Sheikh M.I.
      ,
      • Jorgensen K.E.
      • Sheikh M.I.
      ,
      • Roigaard-Petersen H.
      • Jacobsen C.
      • Jessen H.
      • Mollerup S.
      • Sheikh M.I.
      ,
      • Jessen H.
      • Vorum H.
      • Jorgensen K.E.
      • Sheikh M.I.
      ). These transport systems in apical membranes of intestinal and renal tubular epithelial cells have been well characterized, and it was convincingly shown that they are secondary active systems for the same small amino acids (Gly, Ala, Pro) and selected derivatives. The substrate specificity and kinetics of mPAT1 is identical to the transport system described in Caco-2 cells, which preferentially transports small apolar amino acids, selectedd-amino acids (e.g. d-Ala andd-Ser), the β-amino acid β-alanine, and the neurotransmitter GABA (
      • Thwaites D.T.
      • McEwan G.T.
      • Simmons N.L.
      ). Moreover, the stoichiometry of the transport system was proposed to be 1:1 (
      • Thwaites D.T.
      • McEwan G.T.
      • Brown C.D.
      • Hirst B.H.
      • Simmons N.L.
      ), as found for mPAT1. The substrate affinities, as available from the literature, are in the same range as determined for mPAT1 (see Table I for K m values; Refs.
      • Thwaites D.T.
      • McEwan G.T.
      • Simmons N.L.
      ,
      • Thwaites D.T.
      • McEwan G.T.
      • Brown C.D.
      • Hirst B.H.
      • Simmons N.L.
      ,
      • Thwaites D.T.
      • McEwan G.T.
      • Cook M.J.
      • Hirst B.H.
      • Simmons N.L.
      ,
      • Thwaites D.T.
      • Basterfield L.
      • McCleave P.M.
      • Carter S.M.
      • Simmons N.L.
      ,
      • Roigaard-Petersen H.
      • Jessen H.
      • Mollerup S.
      • Jorgensen K.E.
      • Jacobsen C.
      • Sheikh M.I.
      ,
      • Jorgensen K.E.
      • Sheikh M.I.
      ,
      • Roigaard-Petersen H.
      • Jacobsen C.
      • Jessen H.
      • Mollerup S.
      • Sheikh M.I.
      ,
      • Jessen H.
      • Vorum H.
      • Jorgensen K.E.
      • Sheikh M.I.
      ,
      • Pisoni R.L.
      • Flickinger K.S.
      • Thoene J.G.
      • Christensen H.N.
      ). Minor differences in affinities may again be due to different experimental conditions such as medium pH or membrane potential. The mPAT1 transcript is in addition abundantly expressed in the small intestine and kidney, giving further evidence for its potential role as a transport system in the apical membrane of enterocytes and kidney tubular cells. In addition, when performing reverse transcriptase-PCR with Caco-2 cell RNA and human PAT1-specific primers, we amplified a PCR product with the predicted size (data not shown).
      Expression of mPAT1 and mPAT2 in oocytes and of rLYAAT in CV-1 cells clearly demonstrated that the proteins serve as plasma membrane carriers; however, this could be an artifact of overexpression. Nevertheless, our transfection experiments with the mPAT2-EGFP fusion protein did not reveal evidence for a lysosomal localization of this protein in HeLa cells. Based on this circumstantial evidence we have chosen the more general designation PAT for the two new proton amino acid cotransporters to indicate that they are not a priorilysosomal carriers, although the rLYAAT-1 protein was exclusively expressed in vivo in neuronal lysosomes of the rat brain (
      • Sagne C.
      • Agulhon C.
      • Ravassard P.
      • Darmon M.
      • Hamon M., El
      • Mestikawy S.
      • Gasnier B.
      • Giros B.
      ). Immunohistochemical studies in other tissues than brain have so far not been performed. Consequently, nothing is known currently about the subcellular localization of the mPAT1/LYAAT-1 proteins in tissues such as small intestine and kidney. Nevertheless it might be that the mPAT1/LYAAT-1 protein is targeted in a cell-specific manner to reach either the plasma membrane or the lysosomes. Such cell type-specific targeting has already been shown for other membrane proteins including the epithelial peptide transporter PEPT1, which is mainly found in the apical plasma membrane of intestinal and renal epithelial cells (
      • Bockman D.E.
      • Ganapathy V.
      • Oblak T.G.
      • Leibach F.H.
      ,
      • Ogihara H.
      • Saito H.
      • Shin B.C.
      • Terado T.
      • Takenoshita S.
      • Nagamachi Y.
      • Inui K.
      • Takata K.
      ), but it is also found in lysosomes of pancreatic acinar cells (
      • Shen H.
      • Smith D.E.
      • Yang T.
      • Huang Y.G.
      • Schnermann J.B.
      • Brosius III, F.C.
      ). Therefore, it could be that the PAT proteins serve as both proton-dependent plasma membrane import transporters and lysosomal exporters for small amino acids.
      In conclusion, we have identified, cloned, and functionally characterized two mammalian electrogenic H+/amino acid symporters with a high selectivity for amino acids with apolar and small side chains as well as related compounds. Although their cellular localization has not yet been finally determined, they may serve as plasma membrane and or lysosomal carriers. The characteristics, in particular of mPAT1, resemble in view of the mode of operation, substrate specificity, substrate affinity, and enantioselectivity the functions of apical membrane carriers identified in intestine and kidney for which the protein entities had not been identified until now.

      REFERENCES

        • Saier M.H., Jr.
        Microbiology. 2000; 146: 1775-1795
        • Young G.B.
        • Jack D.L.
        • Smith D.W.
        • Saier M.H., Jr.
        Biochim. Biophys. Acta. 1999; 1415: 306-322
        • Boorer K.J.
        • Frommer W.B.
        • Bush D.R.
        • Kreman M.
        • Loo D.D.
        • Wright E.M.
        J. Biol. Chem. 1996; 271: 2213-2220
        • Russnak R.
        • Konczal D.
        • McIntire S.L.
        J. Biol. Chem. 2001; 276: 23849-23857
        • Sagne C.
        • Agulhon C.
        • Ravassard P.
        • Darmon M.
        • Hamon M., El
        • Mestikawy S.
        • Gasnier B.
        • Giros B.
        Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 7206-7211
        • McIntire S.L.
        • Reimer R.J.
        • Schuske K.
        • Edwards R.H.
        • Jorgensen E.M.
        Nature. 1997; 389: 870-876
        • Sagne C., El
        • Mestikawy S.
        • Isambert M.F.
        • Hamon M.
        • Henry J.P.
        • Giros B.
        • Gasnier B.
        FEBS Lett. 1997; 417: 177-183
        • Forgac M.
        J. Biol. Chem. 1999; 274: 12951-12954
        • Chaudhry F.A.
        • Reimer R.J.
        • Krizaj D.
        • Barber D.
        • Storm-Mathisen J.
        • Copenhagen D.R.
        • Edwards R.H.
        Cell. 1999; 99: 769-780
        • Reimer R.J.
        • Chaudhry F.A.
        • Gray A.T.
        • Edwards R.H.
        Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 7715-7720
        • Yao D.
        • Mackenzie B.
        • Ming H.
        • Varoqui H.
        • Zhu H.
        • Hediger M.A.
        • Erickson J.D.
        J. Biol. Chem. 2000; 275: 22790-22797
        • Kottra G.
        • Daniel H.
        J. Physiol. 2001; 536: 495-503
        • Thwaites D.T.
        • McEwan G.T.
        • Simmons N.L.
        J. Membr. Biol. 1995; 145: 245-256
        • Thwaites D.T.
        • McEwan G.T.
        • Brown C.D.
        • Hirst B.H.
        • Simmons N.L.
        J. Membr. Biol. 1994; 140: 143-151
        • Thwaites D.T.
        • McEwan G.T.
        • Cook M.J.
        • Hirst B.H.
        • Simmons N.L.
        FEBS Lett. 1993; 333: 78-82
        • Thwaites D.T.
        • Basterfield L.
        • McCleave P.M.
        • Carter S.M.
        • Simmons N.L.
        Br. J. Pharmacol. 2000; 129: 457-464
        • Roigaard-Petersen H.
        • Jessen H.
        • Mollerup S.
        • Jorgensen K.E.
        • Jacobsen C.
        • Sheikh M.I.
        Am. J. Physiol. 1990; 258: F388-F396
        • Jorgensen K.E.
        • Sheikh M.I.
        Biochem. J. 1987; 248: 533-538
        • Roigaard-Petersen H.
        • Jacobsen C.
        • Jessen H.
        • Mollerup S.
        • Sheikh M.I.
        Biochim. Biophys. Acta. 1989; 984: 231-237
        • Jessen H.
        • Vorum H.
        • Jorgensen K.E.
        • Sheikh M.I.
        Biochim. Biophys. Acta. 1988; 942: 262-270
        • Pisoni R.L.
        • Flickinger K.S.
        • Thoene J.G.
        • Christensen H.N.
        J. Biol. Chem. 1987; 262: 6010-6017
        • Bockman D.E.
        • Ganapathy V.
        • Oblak T.G.
        • Leibach F.H.
        Int. J. Pancreatol. 1997; 22: 221-225
        • Ogihara H.
        • Saito H.
        • Shin B.C.
        • Terado T.
        • Takenoshita S.
        • Nagamachi Y.
        • Inui K.
        • Takata K.
        Biochem. Biophys. Res. Commun. 1996; 220: 848-852
        • Shen H.
        • Smith D.E.
        • Yang T.
        • Huang Y.G.
        • Schnermann J.B.
        • Brosius III, F.C.
        Am. J. Physiol. 1999; 276: F658-F665