Identification and characterization of a lysosomal transporter for small neutral amino acids

Abstract

In eukaryotic cells, lysosomes represent a major site for macromolecule degradation. Hydrolysis products are eventually exported
from this acidic organelle into the cytosol through specific transporters. Impairment of this process at either the hydrolysis
or the efflux step is responsible of several lysosomal storage diseases. However, most lysosomal transporters, although biochemically
characterized, remain unknown at the molecular level. In this study, we report the molecular and functional characterization
of a lysosomal amino acid transporter (LYAAT-1), remotely related to a family of H+-coupled plasma membrane and synaptic vesicle amino acid transporters. LYAAT-1 is expressed in most rat tissues, with highest
levels in the brain where it is present in neurons. Upon overexpression in COS-7 cells, the recombinant protein mediates the
accumulation of neutral amino acids, such as γ-aminobutyric acid, l-alanine, and l-proline, through an H+/amino acid symport. Confocal microscopy on brain sections revealed that this transporter colocalizes with cathepsin D, an
established lysosomal marker. LYAAT-1 thus appears as a lysosomal transporter that actively exports neutral amino acids from
lysosomes by chemiosmotic coupling to the H+-ATPase of these organelles. Homology searching in eukaryotic genomes suggests that LYAAT-1 defines a subgroup of lysosomal
transporters in the amino acid/auxin permease family.

Full text

Identification and characterization of a lysosomal
transporter for small neutral amino acids
Corinne Sagne
´*
†‡§
, Cendra Agulhon*
†
, Philippe Ravassard
¶
, Miche
`le Darmon
†
, Michel Hamon
†
, Salah El Mestikawy*
†
,
Bruno Gasnier
储
, and Bruno Giros*
†
*Institut National de la Sante´ et de la Recherche Me´dicale U-513, CHU Henri Mondor, 8 Rue du Ge´ne´ral Sarrail, 94010 Cre´teil Cedex, France; †Institut
National de la Sante´ et de la Recherche Me´dicale U-288, CHU Pitie´-Salpeˆtrie`re, 91 Boulevard de l’Hoˆpital, 75634 Paris Cedex 13, France;
¶Centre National de la Recherche Scientifique UMR 9923, Hoˆpital Pitie´-Salpeˆtrie`re, 83 Boulevard de l’Hoˆpital, 75013 Paris, France; and
储Centre National de la Recherche Scientifique UPR 1929, Institut de Biologie Physico-Chimique, 13 Rue Pierre et Marie Curie,
75005 Paris, France
Communicated by Pierre A. Joliot, Institut de Biologie Physico-Chimique, Paris, France, April 12, 2001 (received for review January 4, 2001)
In eukaryotic cells, lysosomes represent a major site for macromol-
ecule degradation. Hydrolysis products are eventually exported
from this acidic organelle into the cytosol through specific trans-
porters. Impairment of this process at either the hydrolysis or the
efflux step is responsible of several lysosomal storage diseases.
However, most lysosomal transporters, although biochemically
characterized, remain unknown at the molecular level. In this
study, we report the molecular and functional characterization of
a lysosomal amino acid transporter (LYAAT-1), remotely related to
a family of H
ⴙ
-coupled plasma membrane and synaptic vesicle
amino acid transporters. LYAAT-1 is expressed in most rat tissues,
with highest levels in the brain where it is present in neurons. Upon
overexpression in COS-7 cells, the recombinant protein mediates
the accumulation of neutral amino acids, such as
␥
-aminobutyric
acid, L-alanine, and L-proline, through an H
ⴙ
兾amino acid symport.
Confocal microscopy on brain sections revealed that this trans-
porter colocalizes with cathepsin D, an established lysosomal
marker. LYAAT-1 thus appears as a lysosomal transporter that
actively exports neutral amino acids from lysosomes by chemios-
motic coupling to the H
ⴙ
-ATPase of these organelles. Homology
searching in eukaryotic genomes suggests that LYAAT-1 defines a
subgroup of lysosomal transporters in the amino acid兾auxin per-
mease family.
Lysosomes are dense organelles responsible for degrading all
four classes of macromolecules. The degradation products
are then exported to the cytosol through specific transporters
and reused in the cellular metabolism. The physiological
importance of lysosomal metabolite efflux is illustrated by
the existence of a group of lysosomal storage diseases with
transport defects, such as sialic acid storage disorders
and nephropathic cystinosis (1). These inherited diseases
result from defective eff lux of sialic acid and cystine from
lysosomes (2–6), respectively, and they have been linked to
mutations in the membrane proteins sialin and cystinosin (7,
8), which are believed to represent sialic acid and cystine
transporters. However, most lysosomal transporters, although
biochemically characterized (1, 9), remain unknown at the
molecular level.
Twenty amino acid transporter families have been identified
thus far (10). In eukaryotes, most of the characterized trans-
porters have been shown to operate at the level of the plasma
membrane (11) or mitochondria. The eukaryotic specific amino
acid兾auxin permease (AAAP) family (12) differs in this respect
because, whereas it was first recognized as a family of H
⫹
兾amino
acid symporters operating at the plasma membrane of plant cells
(13, 14), it was later shown to comprise the transporter respon-
sible for the loading of inhibitory amino acids [
␥
-aminobutyric
acid (GABA) and glycine] into synaptic vesicles of animal nerve
cells (15, 16) through an H
⫹
antiport mechanism. More recently,
additional animal members were identified as the plasma mem-
brane transporters corresponding to system N (17) and system A
(18). In this study, we report the characterization of a novel
mammalian member of the AAAP family that displays the
functional characteristics of an H
⫹
兾amino acid symporter and
that is localized in the lysosomes of brain neurons. This trans-
porter is thus proposed to ensure the export of neutral amino
acids from lysosomes.
Methods
cDNA Cloning. A 495-bp fragment was amplified on the IMAGE
(19) cDNA clone number 45237, corresponding to human
expressed sequence tag y186d11.r1 (accession number H08076)
using primers 5⬘-AAGCTTGGCACGAGGCGTT TCC (sense)
and 5⬘-TAACCAAGAAGGATCT TATCCC (antisense), and
used as a probe to screen a rat hippocampus cDNA library
(lambda ZAPII, number 936518, Stratagene) at high stringency.
Several partial clones were obtained, and one of them was used
to screen again the same library at moderate stringency, thereby
allowing the isolation of a 1.824-bp cDNA clone that was
subcloned into the pcDNA3 expression vector (Invitrogen) for
expression and cRNA synthesis.
Northern Analysis. A Northern blot with 2
␮
g of poly(A)
⫹
mRNA
from different rat tissues (CLONTECH) was hybridized as
recommended by the manufacturer with a
32
P-labeled probe
obtained by PCR amplification on pcDNA3-LYAAT-1 using
primers 5⬘-ACAGGTGATAGAGGCAGCCAACGG (sense)
and 5⬘-CACTGGTGGAATTGGTAGAGGAGT (antisense).
In Situ
Hybridization. Adult male Sprague–Dawley rats were
anesthetized with pentobarbital and perfused transcardially
with 4% paraformaldehyde in PBS (pH 7.4). After dissection,
brains were postfixed2hinthesame solution at 4°C, then
cryoprotected in 15% sucrose兾PBS, frozen, and stored at
⫺80°C until used. Cryosections (14
␮
m) were hybridized either
with
35
S-labeled oligonucleotides or digoxigenin-labeled
cRNA probes. Three
35
S-labeled oligonucleotidic probes, lo-
calized either in the coding (positions 184–219 bp; accession
number AF361239) or the 3⬘noncoding (positions 1574–1634
and 1666–1722 bp) regions of LYAAT-1 cDNA, were hybrid-
ized independently according to a standard protocol and
visualized by exposure to
␤
max x-ray film (Amersham Phar-
macia). The same mRNA distribution was observed with the
three oligonucleotidic probes. Visualization at the cell level
was reached by hybridizing brain sections with a cRNA probe
Abbreviations: GABA,
␥
-aminobutyric acid; AAAP, amino acid兾auxin permease.
Data deposition: The sequence reported in this paper has been deposited in the GenBank
database (accession no. AF361239).
‡Present address: Institut fu¨ r Pharmakologie und Toxikologie, Universita¨ t von Bonn, Re-
uterstrasse 2b, 53113 Bonn, Germany.
§To whom reprint requests should be sent at the *address. E-mail: sagne@im3.inserm.fr.
The publication costs of this article were defrayed in part by page charge payment. This
article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C.
§1734 solely to indicate this fact.
7206–7211
兩
PNAS
兩
June 19, 2001
兩
vol. 98
兩
no. 13 www.pnas.org兾cgi兾doi兾10.1073兾pnas.121183498

labeled with digoxigenin-11-UTP (Promega) using the 1.8-kb
LYAAT-1 insert in pcDNA3 as template. Hybridization was
performed overnight at 65°C in 1 ⫻SSC兾50% formamide兾
10% dextran sulfate兾1 mg/ml rRNA兾1⫻Denhardt’s solution.
Sections were then washed twice with 1 ⫻SSC兾50% form-
amide兾0.1% Tween 20 (at 65°C) and twice with 1 ⫻TBST (at
room temperature). The digoxigenin-labeled hybrids were
detected with an alkaline phosphatase-conjugated anti-
digoxigenin antibody (Roche Molecular Biochemicals).
Transport Assay. Expression of LYAAT-1 in CV-1 cells using the
vaccinia virus兾bacteriophage T7 system for transport assays at
pH 7.5 was performed ac cording to Povlock and Amara (20). For
other experiments, COS-7 cells were transfected with Lipofectin
(Life Technologies, Grand Island, NY). Briefly, 1 day before
transfection, 50,000 cells/well were plated in 24-well dishes. The
day of transfection, cells were washed once with 0.5 ml of
serum-free medium and then incubated 16 –20 h with 250
␮
lof
serum-free medium containing a complex formed by 3
␮
lof
Fig. 1. Alignment of LYAAT-1 to a subgroup of the AAAP family. (a) Multiple alignment of the LYAAT-1 sequence with those of D. melanogaster putative
proteins CG16700, CG3424, CG13384, CG6327, CG1139, and CG7888, C. elegans putative proteins Y43F4B.7 and T27A1.5, and S. cerevisiae predicted proteins
YNL101w and YKL146w was performed with the PILEUP software of the GCG 10.0-UNIX package. Black boxes indicate identical residues, and gray boxes indicate
conservative substitutions. Potential transmembrane domains and sites for N-glycosylation are indicated above alignments by lines and stars, respectively. (b)
Dendrogram for the 13 members of the AAAP family in C. elegans, as well as the four functionally characterized mammalian members (in italics). LYAAT-1 shows
a closer homology to the worm proteins T27A1.5, Y43F4B.7, Y38H6C.17, H32K16.1, C44B7.6, and F59B2.2.
Sagne´ et al. PNAS
兩
June 19, 2001
兩
vol. 98
兩
no. 13
兩
7207
BIOCHEMISTRY

Lipofectin and 1
␮
g of pcDNA3 or pcDNA3-LYAAT-1. One
milliliter of medium supplemented with FCS was added on the
following day, and transport assays were performed 36–48 h
after the beginning of transfection. Cells were washed twice with
0.5 ml of Krebs–Ringer (KR) phosphate buffer (146 mM NaCl兾3
mM KCl兾1 mM CaCl
2
兾1 mM MgCl
2
兾10 mM KH
2
PO
4
兾K
2
HPO
4
)
adjusted at pH 7.5 and then incubated for 15 min at 26°C in KR
buffer adjusted at pH 5.5, supplemented with 0.5–1
␮
Ci of
[
3
H]GABA and 100
␮
M GABA unless stated otherwise in the
text. Reaction was terminated by two washes with ice-cold KR
buffer at pH 7.5. Cells were lysed in 0.1 N NaOH, and their
radioactivity was measured after neutralization by scintillation
counting in Aquasol (Packard). [
3
H]GABA, L-[
3
H]glutamine
and L-[
3
H]glutamate were purchased from Amersham Pharma-
cia; L-[
3
H]proline and L-[
3
H]alanine were purchased from NEN.
Immunocytochemistry. A polyclonal antibody raised against the
peptide SSTDVSPEESPSEGLGC (amino acid residues 15–30)
coupled to keyhole limpet hemocyanin was produced in rabbit
(Eurogentec, Brussels). For immunological detection of
LYAAT-1 on brain sections, the antibody was affinity-purified
on the peptide linked to Affigel.
Transfected COS-7 cells were grown on glass coverslips, fixed
for 10 min in PBS containing 4% paraformaldehyde, rinsed in
PBS, blocked and permeabilized in blocking buffer (PBS兾0.2%
Triton X-100兾5% donkey serum), incubated for at least1hwith
anti-LYAAT-1 serum diluted at 1:500 in the same buffer,
washed in PBS, incubated with Cy3-conjugated donkey anti-
rabbit IgG (Jackson Immunochemicals) at 1:1,600 in blocking
buffer, and rinsed in PBS. Coverslips were mounted on glass
slides with Fluoromount-G solution (Southern Biotechnology
Associates). All steps were performed at room temperature.
Immunoautoradiographic detection of LYAAT-1 was per-
formed on 14-
␮
m sections of frozen brains from rats anesthe-
tized with pentobarbital and perfused transcardially with 100 ml
of 0.9% NaCl兾0.1% NaNO
2
. Sections were fixed for 5 min at 4°C
in 4% paraformaldehyde兾PBS, preincubated for1hinPBS
containing 3% BSA and 1% donkey serum (B1 buffer), and
incubated overnight at 4°C with affinity-purified anti-LYAAT-1
antibody at 1:250 in B1 buffer. On the following day, sections
were washed in PBS, incubated in B1 buffer containing anti-
rabbit [
125
I]IgG (0.2
␮
Ci/ml, 750-3000 Ci/mmol; Amersham
Pharmacia), rinsed in PBS, dried, and exposed for 4 –5 days.
Immunofluorescence detection of LYAAT-1 was performed
on brains from adult rats transcardially perfused with 4%
paraformaldehyde in 100 mM potassium phosphate buffer (pH
7.4), postfixed in the same solution for 4 h and cryoprotected in
10% sucrose at 4°C for 2 days. All of the following steps were
performed at room temperature. Sections (14
␮
m) were blocked
for 30 min in PBS containing 0.2% gelatin兾0.25% Triton X-100
(B2 buffer) and incubated overnight with affinity-purified anti-
LYAAT-1 (1:500) and goat anti-cathepsin D antibodies (1:100;
Santa Cruz Biotechnology) in B2 buffer. Sections were then
washed in B2 buffer, incubated for 90 min with Cy3-conjugated
donkey anti-rabbit IgG (1:1,600) and Alexa 488-conjugated
donkey anti-goat IgG (1:1,000; Molecular Probes) in B2 buffer,
washed in PBS, and mounted with Fluoromount-G solution.
Sections were observed under a laser-scanning confocal micro-
scope (Leica TCS 400), and images were generated with Adobe
Photoshop 5.0.
Results
cDNA Isolation. Homology searching (21) in public databases
using the amino acid sequence of the synaptic vesicle inhibitory
amino acid transporter [VIAAT (15) or VGAT (16)], a member
of the AAAP protein family (12), revealed several homologous
human brain expressed sequence tags. The IMAGE clone (19)
corresponding to one of these expressed sequence tags (acces-
sion no. H08076) was used to isolate from a rat hippocampus
cDNA library a 1,824-bp cDNA containing a 1,425-bp ORF, with
an in-phase STOP codon 120-bp upstream of the initiation
codon. The corresponding 475-aa protein, named LYAAT-1, is
predicted to comprise 11 transmembrane domains, a cytosolic
N-terminal domain, and 3 consensus N-glycosylation sites in
predicted extracytosolic loops (Fig. 1a). Pairwise sequence align-
ment, using the Besfit software of the GCG 10.0-UNIX package,
indicated that LYAAT-1 is distantly related to previously char-
acterized members of the AAAP family, such as the mouse
VIAAT (15) (24% identity over 408 aa residues), the rat plasma
membrane glutamine transporters SN1 (17) (22% over 190) and
GlnT (18) (24% over 402), and the plant amino acid permease
AAP1 (13) (26% over 457). Interestingly, homology searching in
totally sequenced eukaryotic genomes revealed that LYAAT-1
represents a mammalian homologue for a subgroup of function-
ally uncharacterized members of the AAAP family, as illustrated
in Fig. 1bfor the nematode worm Caenorhabditis elegans. The six
putative proteins of this C. elegans AAAP family subgroup
Fig. 2. Northern analysis, in situ hybridization, and immunocytochemistry
show that LYAAT-1 is predominantly expressed in the adult rat brain, where
it is present in glutamatergic and GABAergic neurons. (a) Northern blot
analysis of poly(A)⫹mRNA with a radiolabeled probe against LYAAT-1 shows
a 5.2-kb transcript in all tissues tested, except testis and muscle, where dou-
blets around 2.4 kb and 6 kb are detected, respectively. (band c)35S-labeled
antisense oligonucleotide probe (b) and the corresponding sense probe (c)
were hybridized with brain parasagittal sections. LYAAT-1 mRNA is widely
distributed throughout the brain, with higher expression in the cortex, thal-
amus, pyramidal cells of the hippocampus, and Purkinje cells of the cerebel-
lum, with no significant hybridization by the sense probe. Cb, cerebellum; Cx,
cortex; Hip, hippocampus; Pn, pontine nuclei; Th, thalamus. ( fand g) A rabbit
anti-LYAAT-1 antibody detected by immunofluorescence (red) a protein
mainly expressed in intracellular compartments in COS-7 cells transfected with
pcDNA3-LYAAT-1 (f) but not in mock-transfected cells (g). (d) Immunoauto-
radiographic detection of LYAAT-1 on parasagittal brain sections using the
anti-LYAAT-1 antibody revealed that the protein is enriched in regions ex-
pressing high levels of the corresponding mRNA. (e) Preincubation of the
antibody with the peptide used to produce this antibody abolishes the im-
munoreactivity on an adjacent parasagittal section. (h–k) Comparison of
LYAAT-1 mRNA and protein at the cellular level showed that LYAAT-1 is
mainly expressed in the somatodendritic domain of neurons. Cellular distri-
bution of LYAAT-1 mRNA was detected with a digoxigenin-labeled antisense
probe (hand i), and LYAAT-1 protein was detected by immunofluorescence
and laser-scanning confocal microscopy (jand k) in pyramidal cells of Ammon’s
horn (CA3) of the hippocampus (hand j) and pyramidal cells of the frontal
cortex (iand k). Scale bars in e,g,j, and kindicate 5 mm, 35
␮
m, 105
␮
m, and
42
␮
m and apply to b–e,fand g,hand j, and iand k, respectively.
7208
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.121183498 Sagne´ et al.

exhibit 29–36% identity to LYAAT-1 over 379 – 469 aa. In the
fruit f ly Drosophila melanogaster, seven predicted proteins
(CG16700, CG3424, CG13384, CG6327, CG1139, CG7888, and
CG8785) are 36–46% identical to LYA AT-1. In the yeast
Saccharomyces cerevisiae, two of the seven AAAP family mem-
bers, YNL101w (40% over 311) and YKL146w (32% over 466),
Fig. 3. LYAAT-1 is a small amino acid兾proton symporter. (a) CV-1 cells transiently expressing LYAAT-1 (black bars) using the vaccinia virus expression system
accumulate significantly more [3H]GABA (57 nM) but not L-[3H]glutamine (91 nM) or L-[3H]glutamate (208 nM) than mock-transfected cells (dashed bars). (b)
Accumulation of [3H]GABA into COS-7 expressing LYAAT-1 (
䊐
) strongly increases at acidic pH, whereas transport into mock-transfected cells is not affected (
E
).
pH was controlled by 10 mM K2HPO4兾KH2PO4in the uptake buffer. (c) At pH 5.5, the replacement of NaCl by an equivalent concentration of either lithium chloride
(LiCl) or choline chloride (ChoCl) in the transport buffer had no effect on the 3H-amino acid specific accumulation. By contrast, preincubation of the cells for 10
min with 5
␮
M nigericin in a potassium-free KR buffer strongly inhibited the LYAAT-1-mediated accumulation of GABA into COS-7 cells. Results are expressed
as percent of controls performed in standard (LiCl, ChoCl) or potassium-free (nigericin) KR buffer. (d) Specific LYAAT-1-mediated GABA accumulation
(mock-subtracted) into COS-7 cells at pH 5.5 is saturable. (e) Transport of GABA is inhibited by L- and D-proline, glycine, and L-alanine. Transport of GABA (0.5
␮
Ci [3H]GABA ⫹30
␮
M unlabeled GABA) was measured in the absence (100%) or the presence of 0.5 mM amino acid at pH 5.5. Amino acids, named with the
one letter code, were of the L-form except D-proline (D-P). GA, GABA; Cy, cystine. (f)L-Proline and L-alanine are substrates of LYAAT-1. At pH 5.5, COS-7 cells
expressing LYAAT-1 accumulated specifically 238 ⫾14, 322 ⫾23, and 112 ⫾17 pmol/well of 100
␮
M[
3H]GABA, L-[3H]proline, and L-[3H]alanine, respectively. Each
panel shows one representative (a,b,d, and f) or the mean (cand e) of at least three experiments performed on cells from independent transfections. Transport
measurements were performed in triplicate on LYAAT-1-transfected wells and on paired mock-transfected wells. Error bars represent the SEM. ***,P⬍0.0001;
**,P⬍0.002; *,P⬍0.02.
Sagne´ et al. PNAS
兩
June 19, 2001
兩
vol. 98
兩
no. 13
兩
7209
BIOCHEMISTRY

display closer homology to LYAAT-1. Alignment of sequences
from this AAAP family subgroup in several species, including rat
LYAAT-1, reveals a high conservation in putative transmem-
brane domains 1, 3, 6, and 10, as well as in the putative cytosolic
loop connecting transmembrane domains 7 and 8 (Fig. 1a).
Analysis of high throughput genomic sequence databases shows
a cluster of human contigs (accession nos. AC025433, AC008520,
AC034205, AC011337, AC008552, AC008385, and AC011391)
located at chromosome 5q31-33, which allows reconstitution,
within ten exons, of a human protein sharing 83% identity and
94% similarity with the rat LYA AT-1.
Distribution of LYAAT-1 mRNA and Protein. Northern blot analysis
revealed that LYAAT-1 is largely expressed in rat tissues and is
most abundant in the brain (Fig. 2a). In situ hybridization on
brain sections showed that LYAAT-1 mRNA is abundantly
found in hippocampus, cerebral cortex, cerebellum, and the
thalamic and pontine nuclei, regions known to be rich in neurons
using either glutamate or GABA as neurotransmitter (Fig. 2b).
Indeed, we confirmed at the cellular level that LYAAT-1
transcript is expressed in glutamatergic neurons such as pyra-
midal cells in the cerebral cortex and hippocampus, thalamic and
pontine nuclei neurons, and GABAergic neurons such as Pur-
kinje cells in the cerebellum (Fig. 2 hand i, and data not shown).
Immunodetection of LYAAT-1 on brain sections using an
antibody directed against a peptide located in the N-terminal
domain of the transporter revealed that the protein, although
widely present throughout the encephalon, is more abundant in
regions expressing high levels of LYAAT-1 mRNA (Fig. 2d). At
the cellular level, immunocytochemistry confirmed that
LYAAT-1 is mainly localized in the somatodendritic domain of
neurons (Fig. 2 jand k). By contrast, neither LYAAT-1 mRNA
nor protein could be detected in glial cells, suggesting that
LYAAT-1 expression in brain is restricted to neurons.
Functional Characterization. To investigate the functional proper-
ties of LYAAT-1, uptake experiments were performed using
transiently transfected fibroblasts. Although most recombinant
protein was expressed in the intracellular compartment (Fig. 2f),
intact cells overexpressing the transporter accumulated twice as
much [
3
H]GABA as mock-transfected cells at neutral pH (Fig.
3a). However, no specific uptake of L-[
3
H]glutamine or
L-[
3
H]glutamate could be observed under the same conditions
(Fig. 3a). LYAAT-1 mediated accumulation of [
3
H]GABA into
transfected fibroblasts was time-dependent, remaining linear for
⬇15 min (data not shown) and saturable (Fig. 3d) with a K
m
of
499 ⫾135
␮
MandaV
max
of 2.5 ⫾0.5 pmol/min/
␮
g protein
(Eadie–Hofstee analysis, n⫽3). LYAAT-1 activity was strongly
pH-dependent: a 5-fold increase in [
3
H]GABA specific uptake
was observed at an extracellular pH of 5.5 compared with 7.5
(Fig. 3b). Disruption of the artificially imposed pH gradient with
the ionophore nigericin, which exchanges extracellular protons
for internal potassium ions (22), abolished the specific uptake of
[
3
H]GABA, suggesting a direct role of protons as cosubstrate
(Fig. 3c). By contrast, replacement of sodium ions by either
lithium or choline in the uptake buffer had no effect on
LYAAT-1-mediated [
3
H]GABA accumulation, showing that
LYAAT-1 does not depend on external Na
⫹
(Fig. 3c). Among
natural amino acids, only glycine, L-alanine and L-proline sub-
stantially inhibited LYAAT-1-mediated [
3
H]GABA accumula-
tion, suggesting that they represent additional substrates of the
transporter (Fig. 3e). D-Proline competed with [
3
H]GABA as
efficiently as did L-proline (Fig. 3e). Accumulation of L-[
3
H]pro-
line and L-[
3
H]alanine into fibroblasts overexpressing LYAAT-1
at acidic pH directly confirmed that the transporter is able to
accept several small neutral amino acids as substrate (Fig. 3f).
Colocalization with a Lysosomal Marker. Because LYAAT-1 ap-
peared as a proton兾amino acid symporter mostly active at acidic
pH, we reasoned that it might be involved in the efflux of small
neutral amino acids from acidic organelles. Lysosomes, which
produce amino acids by proteolysis, thus appeared as choice
candidate organelles for LYAAT-1 function. To test this hy-
pothesis, we compared the intracellular distribution of native
LYAAT-1 to that of cathepsin D (EC 3.4.23.5), a lysosomal
aspartyl protease (23). Laser-scanning confocal microscopy on
brain sections revealed that LYAAT-1 immunoreactivity mas-
sively colocalized with that of cathepsin D (Fig. 4). Quantifica-
tion of LYAAT-1-immunoreactive puncta indicated that 95%
and 92% of them coexpressed cathepsin D in pyramidal neurons
of hippocampus and cortex, respectively (analysis of 529 and 451
puncta on 20 randomly selected neuronal somata in each re-
gion). Conversely, 97% and 94% of cathepsin D-positive puncta
also expressed LYAAT-1 (analysis of 516 and 438 puncta on 20
hippocampal and cortical pyramidal neurons, respectively).
Discussion
Homology searching in expressed sequence tag databases al-
lowed us to identify a novel mammalian member of the AAAP
Fig. 4. Double immunofluorescence labeling and laser-scanning confocal
microscopy on brain sections show that the native LYAAT-1 colocalizes with
the lysosomal protease cathepsin D. In pyramidal neurons of frontal cortex
(a–c) and of Ammon’s horn (CA3) (d–f), LYAAT-1 immunoreactivity (red) is
shown in aand d, and cathepsin D immunoreactivity (green) is shown in b
and e. In the superimposed images (cand f), the yellow color indicates the
colocalization of the two markers (arrows); arrowheads indicate cathepsin
D-positive, LYAAT-1-negative puncta. Scale bar indicates 5
␮
m and applies
to a–f.
7210
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.121183498 Sagne´ et al.

family, named LYAAT-1. This 475-aa protein belongs to a
subgroup of the AAAP family, consisting of noncharacterized
putative transporters revealed by the sequencing of eukaryotic
genomes. LYAAT-1 is predominantly expressed in the brain,
where both mRNA and protein could be detected in neurons but
not in glial cells. When overexpressed in fibroblasts, the protein
is able to accumulate some neutral amino acids into the cells,
through an H
⫹
兾amino acid symport driven by an artificially
imposed pH gradient. Finally, the colocalization of LYAAT-1
with cathepsin D, a lysosomal protease, in brain revealed that the
native protein is localized on lysosomes. Taken together, these
data support a physiological role of LYAAT-1 as an amino acid
transporter involved in the efflux of lysosomal proteolysis prod-
ucts, such as L-proline, L-alanine, or glycine, from the organelle
lumen to the cytosol. Each of these amino acids is present at a
⬇50
␮
M concentration in the lumen of liver lysosomes (24), thus
amounting to an overall substrate concentration of 150
␮
M,
which is in good agreement with a K
m
value of ⬇500
␮
M (Fig.
3dand e).
Its H
⫹
symport mechanism (Fig. 3c) implies that LYAAT-1-
mediated amino acid efflux is actively driven by the lysosomal
H
⫹
-ATPase, possibly to allow net f lux against high concentra-
tions of these amino acids in the cytosol. It should be noted that
our functional assay, which takes advantage of a partial expres-
sion of LYAAT-1 at the plasma membrane of transfected
fibroblasts, allows measurement of LYAAT-1-mediated trans-
port in the same direction as the proposed LYAAT-1-mediated
lysosomal efflux because the extracellular medium is topologi-
cally equivalent to the lysosomal lumen. Two transport systems
for L-proline, systems f and p, have been biochemically described
in lysosomes purified from fibroblasts (25), but none seems to
correspond to LYAAT-1, as they appeared to be pH-
independent and stereospecific for L-proline. LYAAT-1 may
thus be absent from fibroblasts, as it is from glial cells in the
brain. GABA, which should be considered as a model substrate
rather than an actual physiological substrate, may be recognized
by LYAAT-1 because it adopts a cyclic conformation structur-
ally similar to proline, as proposed in the case of another member
of the AAAP family, ProT2, which transports L-proline, D-
proline, and GABA with similar efficiencies (26).
LYAAT-1 shares no sequence homology with the putative
lysosomal transporters sialin or cystinosin, and it appears as
the first characterized member of a novel subgroup of proteins
intheAAAP family, present in diverse eukaryotic species (Fig.
1). The identification of LYAAT-1 could thus open an avenue
for the isolation of other lysosomal amino acid transporters
and, possibly, for the characterization of novel lysosomal
storage diseases.
The restricted expression of LYAAT-1 in neurons indicates
that lysosomal transporters, beyond their general role in cellular
metabolism, may be involved in specialized cellular functions. A
previously documented example is lysosomal transport system h,
which recognizes and recycles monoiodotyrosine in thyroid
epithelial cells, thereby contributing to a more efficient synthesis
of thyroid hormones (27). It will thus be interesting to examine
in the future whether LYAAT-1 also participates to specific
neuronal functions, including neurotransmission.
Note Added in Proof. After submission of the manuscript, a study
reported the functional characterization of S. cerevisiae AAAPs YNL
101w and YKL 146w (28).
We thank P. Gaspar for help with immunof luorescence on brain sections,
B. Goud and L. Johannes for allowing access to the vaccinia v irus facility,
J. L. Popot for prediction of the secondary structure, H. Boenisch for
fruitful discussions, and C. Betancur for careful reading of the manu-
script. This work is supported in part by The Association France
Parkinson and the European Economic Community (to C.S.), Hoechst
Marion Roussel fellowship (to C.A.) and funding (to B. Giros), the
Fondation Singer Polignac (to C.A.), Institut National de la Sante´etde
la Recherche Me´dicale (to B. Giros and M.H.), and Centre National de
la Recherche Scientifique (to B. Gasnier and P.R.).
1. Mancini, G. M., Havelaar, A. C. & Verheijen, F. W. (2000) J. Inherited Metab.
Dis. 23, 278–292.
2. Gahl, W. A., Tietze, F., Bashan, N., Steinherz, R. & Schulman, J. D. (1982)
J. Biol. Chem. 257, 9570–9575.
3. Havelaar, A. C., Mancini, G. M., Beerens, C. E., Souren, R. M. & Verheijen,
F. W. (1998) J. Biol. Chem. 273, 34568–34574.
4. Mancini, G. M., Verheijen, F. W. & Galjaard, H. (1986) Hum. Genet. 73,
214–217.
5. Renlund, M., Tietze, F. & Gahl, W. A. (1986) Science 232, 759 –762.
6. Tietze, F., Seppala, R., Renlund, M., Hopwood, J. J., Harper, G. S., Thomas,
G. H. & Gahl, W. A. (1989) J. Biol. Chem. 264, 15316–15322.
7. Town, M., Jean, G., Cherqui, S., Attard, M., Forestier, L., Whitmore, S. A.,
Callen, D. F., Gribouval, O., Broyer, M., Bates, G. P., et al. (1998) Nat. Genet.
18, 319–324.
8. Verheijen, F. W., Verbeek, E., Aula, N., Beerens, C. E., Havelaar, A. C., Joosse,
M., Peltonen, L., Aula, P., Galjaard, H., van der Spek, P. J. & Mancini, G. M.
(1999) Nat. Genet. 23, 462– 465.
9. Pisoni, R. L. & Schneider, J. A. (1992) in Mammalian Amino Acid Transport,
eds. Kilberg, M. S. & Ha¨ussinger, D. (Plenum, New York), pp. 89 –99.
10. Saier, M. H., Jr. (1999) Microbiol. Mol. Biol. Rev. 64, 354 –411.
11. Palacin, M., Estevez, R., Bertran, J. & Zorzano, A. (1998) Physiol. Rev. 78, 969–1054.
12. Young, G. B., Jack, D. L., Smith, D. W. & Saier, M. H., Jr. (1999) Biochim.
Biophys. Acta 1415, 306–322.
13. Frommer, W. B., Hummel, S. & Riesmeier, J. W. (1993) Proc. Natl. Acad. Sci.
USA 90, 5944–5948.
14. Ortiz-Lopez, A., Chang, H. & Bush, D. R. (2000) Biochim. Biophys. Acta 1465,
275–280.
15. Sagne´, C., El Mestikawy, S., Isambert, M. F., Hamon, M., Henr y, J. P., Giros,
B. & Gasnier, B. (1997) FEBS Lett. 417, 177–183.
16. McIntire, S. L., Reimer, R. J., Schuske, K., Edwards, R. H. & Jorgensen, E. M.
(1997) Nature (London) 389, 870– 876.
17. Chaudhry, F. A., Reimer, R. J., Krizaj, D., Barber, D., Storm-Mathisen, J.,
Copenhagen, D. R. & Edwards, R. H. (1999) Cell 99, 769–780.
18. Varoqui, H., Zhu, H., Yao, D., Ming, H. & Erickson, J. D. (2000) J. Biol. Chem.
275, 4049– 4054.
19. Lennon, G., Auffray, C., Polymeropoulos, M. & Soares, M. B. (1996) Genomics
33, 151–152.
20. Povlock, S. L. & Amara, S. G. (1998) Methods Enz ymol. 296, 436 –443.
21. A ltschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. (1990) J.
Mol. Biol. 215, 403– 410.
22. Pressman, B. C. (1968) Fed. P roc. 27, 1283–1288.
23. Whitaker, J. N., Terry, L. C. & Whetsell, W. O., Jr. (1981) Brain Res. 216,
109–124.
24. Vadgama, J. V., Chang, K., Kopple, J. D., Idriss, J.-M. & Jonas, A. J. (1991)
J. Cell Physiol. 147, 447–454.
25. Pisoni, R. L., Flickinger, K. S., Thoene, J. G. & Christensen, H. N. (1987)
J. Biol. Chem. 262, 6010–6017.
26. Breitkreuz, K. E., Shelp, B. J., Fischer, W. N., Schwacke, R. & Rentsch, D.
(1999) FEBS Lett. 450, 280–284.
27. Andersson, H. C., Kohn, L. D., Bernardini, I., Blom, H. J., Tietze, F. & Gahl,
W. A. (1990) J. Biol. Chem. 265, 10950–10954.
28. Russnak, R., Konzcal, D. & McIntire, S. L. (March 26, 2001) J. Biol. Chem.,
10.1074兾jbc.M008028200.
Sagne´ et al. PNAS
兩
June 19, 2001
兩
vol. 98
兩
no. 13
兩
7211
BIOCHEMISTRY