THE JOURNAL OF COMPARATIVE NEUROLOGY 433:101–114 (2001)

Differential Distribution of the

Glutamate Transporters GLT-1 and
GLAST in Tanycytes of the
Third Ventricle
URS V. BERGER* AND MATTHIAS A. HEDIGER
Membrane Biology Program, Department of Medicine, Brigham and Women’s Hospital,
Boston, Massachusetts 02115

ABSTRACT
The ventral one-third of the ventricular lining in the hypothalamus is formed by specialized
ependymal cells called the tanycytes. These cells may serve a neuroendocrine transport function
because of their structural specializations, which include apical microvili on the ventricular
surface and long basal processes that terminate on blood vessels or on the glia limitans. Here, we
describe the expression of mRNA and protein for the glutamate transporters GLT-1 and GLAST
in unique tanycyte populations of the third ventricle in rat brain. Using nonisotopic in situ
hybridization, we demonstrate GLAST mRNA labeling in tanycytes of the ventral floor and
lateral walls in the tuberal and mammillary recess portions of the third ventricle. This GLAST
mRNA labeling had a higher intensity than the labeling intensity observed in regular ependymal
cells throughout the ventricular system. Furthermore, we have identified strong GLT-1 mRNA
labeling in a population of tanycytes situated in the dorsolateral walls of caudal tuberal and
mammillary recess portions. Immunocytochemical staining indicates that both GLT-1 and
GLAST protein are expressed in the tanycyte populations as well. These data corroborate
previous findings that third ventricle tanycytes are functionally heterogeneous. Furthermore, the
GLT-1-expressing tanycytes represent a population of tanycytes that, to date, has not been
recognized as functionally distinct. The strong GLAST expression by the ventral tanycytes in the
hypophysiotropic area suggests a role of tanycyte-mediated glutamate transport in neuroendo-
crine activity. The functional role of GLT-1 in dorsal wall tanycytes remains to be explored. J.
Comp. Neurol. 433:101–114, 2001. © 2001 Wiley-Liss, Inc.

Indexing terms: EAAT1; EAAT2; in situ hybridization

The surfaces of the cerebral ventricles are covered by
ependymal cells, a glial cell type of astrocytic origin. Form-
ing mostly a single-cell layer between the ventricular
spaces and the interstitial areas, the ependymal cells are
primarily cuboid or columnar in form, and they possess
cilia. In the ventrocaudal part of the third ventricle and
also on the surface of the circumventricular organs, the
ependymal cells are replaced by an additional ependymal
cell type, the tanycytes (Flament-Durand and Brion, 1985;
Del Bigio, 1995; Bruni, 1998). Tanycytes are characterized
by their elongated nuclei, apical microvili, lack of cilia,
and single elongated basal process that extends far into
the underlying neuropil. The basal processes enwrap
blood vessels and terminate on neurons or glia. In the
tuberal portion of the third ventricle, encompassing the
infundibular recess and the median eminence, the basal
processes of tanycytes arch downward and terminate ei-
ther on fenestrated capillaries in the median eminence or
on the external glial limitans, which abuts the perivascu-
lar space surrounding the hypophysial portal vessels. Be-
cause of their unique morphology and their position in the
hypophysiotropic area, these tanycytes in the third ven-
tricle are thought to provide a link between the cerebro-
spinal fluid (CSF) and the portal vasculature.
The tanycytes of the median eminence are joined by
tight junctions that restrict free intracellular movement of
Grant sponsor: National Institutes of Health; Grant number: NS 32001.
*Correspondence to: Urs V. Berger, Beth Israel Deaconess Medical Cen-
ter, 330 Brookline Ave., RW861, Boston, MA 02115.
E-mail: uberger@caregroup.harvard.edu
Received 14 July 2000; Revised 1 December 2000; Accepted 18 January
2001

© 2001 WILEY-LISS, INC.

Figure 1
GLUTAMATE TRANSPORTERS IN TANYCYTES
molecules, in contrast to regular ependymal cells, which
are joined by gap junctions (Brightman and Reese, 1969).
Thus, the passage from CSF to interstitial space in the
ventral hypothalamus is restricted. This feature and the
morphological specializations of the tanycytes (see above)
have prompted the hypothesis that the tanycytes and the
CSF are involved in a transport system that serves some
neuroendocrine function (Flament-Durand and Brion,
1985; Bruni, 1998). Investigators speculate that hormones
are secreted into the CSF and then are transported to the
hypophysial-portal vasculature by the tanycytes, in es-
sence paralleling the conventional mode of direct axonal
delivery of hormones. Even though a number of studies
have demonstrated that the median eminence can take up
substances from the CSF (Silverman et al., 1972; Bjelke
and Fuxe, 1993), very little evidence for a tanycyte-
mediated directed transport from CSF to the portal vas-
culature has emerged so far, and the strength of this
hypothesis has been discussed extensively (Pilgrim, 1978;
Flament-Durand and Brion, 1985; Bruni, 1998). Other
postulated functions of tanycytes include the transport of
substances in the reverse direction from the portal vascu-
lature to the CSF (Wittkowsky, 1998) and the modulation
of hormone secretion into perivascular spaces (Flament-
Durand and Brion, 1985; Bruni, 1998).
During our studies of the expression patterns for the
glutamate transporters GLT-1 and GLAST in rat brain
(Berger and Hediger, 1998, 2000), we have noticed unusu-
ally high expression levels of GLT-1 and GLAST mRNA in
what appear to be tanycytes of the third ventricle. Be-
cause GLT-1 expression in tanycytes has not been re-
ported so far, we have performed a detailed mapping
study to determine the extent of the glutamate trans-
porter expression in third ventricle tanycytes. The results
show distinct expression patterns for GLT-1 and GLAST
in tanycytes, providing evidence for functional heteroge-
neity of these ependymal cells.
MATERIALS AND METHODS
Animals and chemicals
Male Sprague-Dawley rats (200 –300 g; Taconic Farms,
Germantown, NY) were used. They were kept in a
temperature- and light-controlled room (7 AM to 5 PM light)
with free access to rat chow and water before they were
killed. Rats were anesthetized with pentobarbital (75
Fig. 1. Photomicrographs of serial coronal sections through the
ependyma of rostral third ventricle labeled for the glutamate trans-
porters GLT-1 (A–D) or GLAST (E–H) mRNA using nonisotopic in
situ hybridization. The approximate Bregma levels for the sections
are ⫺2.0 mm (A,E), ⫺2.1 mm (B,F), ⫺2.3 mm (C,G), and ⫺2.6 mm
(D,H). For each photomicrograph, the stretch of ependymal lining
indicated by the arrows is magnified in the inset in the upper right
corner. A, no GLT-1 mRNA labeling; B, weak GLT-1 mRNA labeling
in the ventral wall (arrows; notice different shape of the cells shown in
the inset compared with the cell shapes shown in the inset in A); C,
GLT-1 mRNA labeling farther dorsal in the wall; D, GLT-1 mRNA
labeling extending from the top of the roof of the infundibular recess
(arrowhead) to the area indicated by arrows; E, moderate GLAST
mRNA labeling in cuboidal ependymal cells; F, G, stronger GLAST
mRNA labeling in the ventricle floor and walls; H, strong GLAST
mRNA labeling in the lining of the infundibular recess and moderate
labeling in the central portion of the ventricle floor. Note that the
lining is damaged at the upper end in C. Scale bar ⫽ 100 ␮m.
103
mg/kg) and then killed by decapitation, using procedures
approved by the Harvard Medical School Animal Welfare
Committee. Brains were dissected and snap-frozen in iso-
pentane at ⫺30°C and then either were processed imme-
diately or stored at ⫺80°C. All chemicals, unless noted
otherwise, were obtained from Sigma (St. Louis, MO).
In situ hybridization
The cRNA probes were identical to those used previ-
ously (Berger and Hediger, 1998). The GLT-1 probe con-
tained the whole GLT-1 sequence (1.8 kb; Pines et al.,
1992), whereas the GLAST probe contained nucleotides
58 –2,259 of the GLAST sequence (Storck et al., 1992). The
GLT-1 and GLAST cRNAs were digoxigenin labeled by
transcription from a linearized Bluescript plasmid
(GLT-1) or from a polymerase chain reaction-generated
cDNA fragment with flanking SP6 and T7 polymerase
sites (GLAST). After transcription, all probes were alkali
hydrolyzed to an average length of 250 –500 base pairs
(Schaeren-Wiemers and Gerfin-Moser, 1993). On North-
ern blots of brain total RNA, the GLT-1 and GLAST
probes recognized single bands at approximately 11 kb
and 4.5 kb, respectively (data not shown).
The in situ hybridization procedure was performed es-
sentially as described previously (Berger and Hediger,
1998). Briefly, serial frozen sections (16 ␮m) from approx-
imate levels of Bregma ⫹1.0 mm to ⫺4.6 mm according to
the atlas of Paxinos and Watson (1998) were cut on a
cryostat and picked up on Superfrost plus microscope
slides (Fisher Scientific, Fair Lawn, NJ). In addition, sets
of 2-␮m frozen sections were also collected. Slide-mounted
sections were either processed immediately or first stored
at ⫺80°C. After fixation with 4% freshly depolymerized,
ice-cold paraformaldehyde, the slides were rinsed in
phosphate-buffered saline, pH 7.4 (PBS); acetylated;
rinsed in PBS; and submerged in hybridization buffer
[50% formamide (Life Technologies, Gaithersburg, MD);
5⫻ standard saline citrate (SSC); 2% blocking reagent
(Roche Molecular Biochemicals); 0.1% sodium dodecyl sul-
fate; and 0.02% N-laurylsarcosine, pH 7] in slide mailers
(Berger and Hediger, 1998). After the addition of probe
(100 ng/ml for GLAST; 10 ng/ml for GLT-1), slides were
hybridized overnight or over the weekend at 68°C. Slides
subsequently were removed from the mailers and placed
in coplin jars in 2⫻ SSC for three quick rinses and then in
0.2⫻ SSC for two 30-minute rinses, all at 68°C. Sections
were then blocked in 0.1 M maleate buffer, pH 7.0, con-
taining 1 % blocking reagent for 1 hour at room tempera-
ture and then incubated in sheep antidigoxigenin-Fab
fragments conjugated to alkaline phosphatase (0.15 U/ml;
Roche Molecular Biochemicals) in maleate/blocking buffer
for 1 hour. After several rinses over 1 hour in maleate and
one quick rinse in 0.1 M Tris-HCl and 0.15 M NaCl, pH
9.5, slides were developed in the dark overnight in a
ready-made substrate solution containing 5-bromo-4
chloro-indolylphosphate and nitroblue tetrazolium chlo-
ride (Kierkegard and Perry, Inc., Gaithersburg, MD), to
which 150 mM NaCl had been added. The development
was stopped by incubation in 0.1 M Tris, 0.15 M NaCl, and
25 mM ethylenediamine tetraacetic acid, pH 9.5. Slides
were mounted with glycerol gelatin (Sigma). All of the
sections from the different hypothalamic regions were in-
cubated in identical probe concentrations and were pro-
cessed together to allow direct comparison of labeling in-
Figure 2
GLUTAMATE TRANSPORTERS IN TANYCYTES
tensities. Control incubations used either sense cRNA
probes or no probe at all.
Immunocytochemistry
Sections adjacent to those used for in situ hybridization
as well as 2-␮m frozen sections were stained with an
antisera against glial fibrillary acidic protein (GFAP;
Dako, Carpinteria, CA), which is a marker for astrocytes
and tanycytes (Flament-Durand and Brion, 1985), GLT-1
(B12-26; Lehre et al., 1995), and GLAST (A522-42; Lehre
et al., 1995). Slide-mounted sections were fixed in 4%
paraformaldehyde, incubated in blocking solution (2%
normal goat serum and 0.1% Triton X-100 in PBS) for 30
minutes at room temperature, and then incubated over-
night at 4°C in primary antiserum (1:300 dilution) in
blocking solution. Bound antibodies were detected with
indocarbocyanine (Cy3)-conjugated secondary antibodies
(Jackson Immunoresearch, West Grove, PA) or with bio-
tinylated secondary antibodies and streptavidin-Cy3 (Vec-
tor Laboratories, Burlingame, CA). Control sections were
incubated without primary antibodies.
All photomicrographs were taken with a SPOT digital
camera (Diagnostic Instruments) attached to a Nikon
E-600 microscope (Nikon, Tokyo, Japan) using brightfield
illumination with Nomarski optics or epifluorescence illu-
mination with a rhodamine filter cube. Pictures were com-
bined and processed with Adobe Photoshop software (Ado-
be Systems, Mountain View, CA).
RESULTS
General
The specificity of our in situ labeling technique and
probes has been established in previous studies (Berger
and Hediger, 1998, 2000). Sense controls produced no
labeling. To increase the sensitivity of labeling for GLT-1
in the ependyma, alkaline phosphatase development pro-
ceeded overnight. This led to a saturation of signal devel-
opment for GLT-1 in gray matter regions of the hypothal-
amus. The coordinates given below are approximations of
the levels shown in the rat atlas of Paxinos and Watson
(1998). The coronal sections were labeled for GLT-1 and
GLAST mRNA in intervals of approximately 72 ␮m. The
description of the findings is separated into anterior, mid-
dle, and posterior portions of the third ventricle.
Fig. 2. Coronal sections through the tuberal portion of the third
ventricle labeled for GLT-1 (A–C), or GLAST (D–F) mRNA at approx-
imate Bregma levels of ⫺3.2 (A,D), ⫺3.5 (B,E), and ⫺3.8 mm (C,F).
For each photomicrograph, the stretch of ependymal lining indicated
by the arrows is magnified in the inset in the upper right corner. A–C:
Throughout the tuberal portion, GLT-1 mRNA labeling is detectable
in the lining from the infundibular recess (A, arrowhead) to the
transition zone with the cuboidal cells. The labeling intensity grows
stronger toward the caudal end. Astrocytes in the internal layer of the
median eminence show strong GLT-1 mRNA labeling, but no GLT-1
labeling is seen in the lining in the pituitary stalk (C, arrowhead).
D–F: The strong GLAST mRNA labeling is present in the ventral
one-third of the ventricle wall. The central portions of the lining over
the median eminence (D, arrowhead) and the lining of the pituitary
stalk (F, arrowhead) show moderate labeling intensity. Scale bar ⫽
100 ␮m.
105
Anterior hypothalamic to tuberal portions
of third ventricle
Figure 1 shows the GLT-1 mRNA labeling patterns (Fig.
1A–D) and the GLAST mRNA labeling patterns (Fig.
1E–H) in the caudal end of the anterior hypothalamic
portion and in the rostral tuberal portion of the third
ventricle. In the anterior hypothalamic portion, extending
from Bregma ⫺0.3 mm to ⫺2.1 mm, the ependymal lining
contains primarily regular, cuboidal, ependymal cells that
express no GLT-1 mRNA. Figure 1A shows the lack of
GLT-1 mRNA labeling in this anterior portion at Bregma
⫺2.0. Starting at around Bregma ⫺2.1 mm and extending
caudally, weak GLT-1 mRNA labeling appears along the
ventral ventricular walls (Fig. 1B). The GLT-1 mRNA-
positive cells have a more elongated cell body, suggesting
that they are tanycytes. At Bregma level ⫺2.3, the
ependymal GLT-1 mRNA labeling is located more dorsally
in the ventricular wall, and it has gained in intensity (Fig.
1C, arrows). At the rostral end of the infundibular recess
(Fig. 1D; Bregma level ⫺2.6), the GLT-1 mRNA labeling in
the ependyma extends from the area indicated by the
arrows to the roof of the recess. The labeling intensity has
intensified further. The inset in Figure 1D shows the
transition zone between the dorsal cuboid ependymal cells
and the ventral tanycytes. Here, two layers can be dis-
cerned: an outer layer of unlabeled cuboidal cells and a
subjacent layer of GLT-1 mRNA-positive tanycytes. Some
of the labeled cells extend processes through the cuboidal
cell layer to the ventricular space. This double layer of
ependymal cells has also been described by Mitro and
Palkovits (1981) and is shown more clearly in Figure 3.
The ependymal cells of the floor and the lateral walls of
the infundibular recess are not labeled for GLT-1 mRNA.
In contrast to GLT-1 and as shown previously (Schmitt
et al., 1997; Berger and Hediger, 1998, 2000), GLAST
mRNA is expressed at moderate levels in ependymal cells
of all ventricles. In the third ventricle, this moderate la-
beling for GLAST mRNA is found in all sections rostral to
Bregma ⫺2.1 mm (Fig. 1E). At around Bregma ⫺2.1 mm,
a more intense labeling appears that is restricted initially
to the most ventral ependymal cells near the ventricle
floor but is then seen more and more along the ventral
walls (Fig. 1F,G), until it stretches to about one-third of
the dorsoventral extension of the ventricle at the begin-
ning of the infundibular recess (Fig. 1H). The ependymal
cells that are situated dorsal to this increased GLAST
mRNA labeling continue to express moderate GLAST
mRNA levels.
Tuberal portion of third ventricle
The tuberal portion of the third ventricle extends from
Bregma ⫺2.1 to ⫺3.6 (Mitro and Palkovits, 1981). GLT-1
mRNA labeling is present in the ventricular walls
throughout this region (Fig. 2A–C). It extends for the most
part from the medial end of the roof of the infundibular
recess (Fig. 2A, arrowhead) to the transition area with the
cuboidal cells (see also Fig. 3). The ventricular floor, the
sides of the roof, and the cuboidal cells are not labeled. The
inset in Figure 2C demonstrates how the labeling inten-
sity tapers off dorsoventrally in the ventricular wall. The
GLT-1 mRNA labeling in the walls intensifies further in
the rostrocaudal direction, particularly in the tanycytes
situated at the dorsal end near the transition zone with
106 U.V. BERGER AND M.A. HEDIGER

Figure 3
GLUTAMATE TRANSPORTERS IN TANYCYTES
the cuboidal cells. No GLT-1 mRNA labeling is present in
the ependyma lining the pituitary stalk (Fig. 2C).
The strong GLAST mRNA labeling in the tuberal por-
tion of the third ventricle is seen primarily in the lateral
sulci, in the roof of the pituitary recess, and along the
ventral walls (Fig. 2D–F). The strong GLAST labeling
extends dorsally also to approximately the transition zone
with the regular cuboidal cells. Only moderate GLAST
mRNA labeling is present in the central portion of the
ventricle floor, where the tanycytes form a monolayer of
relatively flat tanycytes (Mitro and Palkovits, 1981). The
ependymal GLAST labeling extends into the pituitary
stalk (Fig. 2F).
Figure 3 shows a higher magnification view of the
GLT-1 and GLAST ependymal cell mRNA labeling in the
caudal tuberal portion of the third ventricle (approximate
level, Bregma ⫺3.4 mm). The whole dorsoventral exten-
sion is shown and separated into two sections that show
dorsal and ventral labeling for GLT-1 (Fig. 3A,C) and
dorsal and ventral GLAST mRNA labeling in an adjacent
section (Fig. 3B,D). The ependymal cell area a (Fig. 3A,B)
represents the cuboidal ependymal cells, which are la-
beled only for GLAST mRNA and not for GLT-1 mRNA.
Area b denotes the transition zone between the outer
GLAST mRNA-positive cuboidal cells and the subjacent
tanycytes, which are labeled strongly for GLT-1 mRNA.
Some of these tanycytes extend processes through the
cuboidal cell layer to the ventricular lumen (Fig. 3A, ar-
rows). The subjacent tanycyte cell layer, which comes to
the ventricular surface in area c, is indicated by the ar-
rows in Figure 3B. These subjacent cells do not appear to
be labeled for GLAST mRNA. Based on the serial adjacent
sections, the subjacent cell layer does not reflect a fold in
the ventricle wall. The GLT-1 mRNA labeling continues
ventrally in the superficial layer through area d and then
stops in area f (Fig. 3C). In contrast, the GLAST labeling
becomes more intense in area d and then continues in the
superficial layers into area f, the roof of the infundibular
recess (Fig. 3D, arrow in area g), the lateral sulcus (not
shown), and the superficial layer of the median eminence
(ventricle floor; Fig. 3D, arrowhead in area g). GLT-1
mRNA labeling is absent from the roof, lateral sulcus. and
ventricle floor.
Mammillary recess of third ventricle
The caudal part of the third ventricle, encompassing the
mammillary recess extending from Bregma ⫺3.9 to ⫺4.6,
Fig. 3. A–D: Higher magnification photomicrographs of ependy-
mal GLT-1 and GLAST mRNA labeling in adjacent sections through
the caudal tuberal portion of the third ventricle (Bregma ⫺3.4 mm).
A,C, GLT-1; B,D, GLAST. In area a, regular cuboidal cells are labeled
for GLAST mRNA but not for GLT-1 mRNA. Area b is the transition
zone between cuboidal cells and subjacent GLT-1-positive tanycytes.
Some GLT-1-positive processes can be discerned that reach the ven-
tricular lumen through the cuboidal cell layer (A, arrows). In area c,
the cuboidal cells stop, and the tanycytes come to the surface. The
arrows in B denote the subjacent tanycyte layer that is unlabeled for
GLAST but labeled for GLT-1. In area d, the GLAST labeling becomes
stronger. In area f, the GLT-1 mRNA labeling stops. The GLAST
labeling continues into the roof of the infundibular recess (arrow in
area g), the lateral sulcus (not shown), and the superficial layer of the
median eminence (ventricle floor; arrowhead in area g). GLT-1 mRNA
labeling is absent from the roof, lateral sulcus, and ventricle floor.
Scale bar ⫽ 25 ␮m.
107
shows the highest labeling intensity for GLT-1 mRNA in
the ventricular walls (Fig. 4A–C,G–I). The strongest la-
beling is observed in the tanycytes in the lateral sulci of
the recess and the dorsal walls. The ventral walls show
moderate-to-weak labeling, and the floor appears unla-
beled. The roof of the mammillary recess, which consists
entirely of regular cuboidal cells, also is unlabeled. A
sharp transition is seen between the strongly labeled dor-
sal tanycytes and the unlabeled cuboidal cells of the roof
(Fig. 4C, arrows and inset). The strong tanycytic labeling
in the lateral sulci extends into the caudal end of the
recess (Fig. 4G–I). The strong GLAST mRNA labeling
continues in the floor and ventral walls of the mammillary
recess (Fig. 4D–F) and then gets weaker toward its caudal
end (Fig. 4J–L). The roof and the dorsal walls exhibit the
regular, moderate labeling intensity of cuboidal cells.
Colocalization of GLT-1 mRNA and GFAP
immunostaining
The tanycytic nature of the cells that express either
GLT-1 mRNA or the stronger GLAST mRNA labeling is
indicated by their elongated cell bodies. Furthermore, pre-
vious studies using scanning electron microscopy have
shown that the ependyma situated in the areas where we
observe GLT-1 and GLAST mRNA labeling, i.e., the ven-
tral parts of the third ventricle walls and the ventricle
floor in tuberal and mammillary recess portions, consist
entirely of tanycytes. A sharp demarcation can be found
between the cilia-carrying cuboidal ependymal cells in the
dorsal walls and the microvili-carrying tanycytes in the
ventral walls (Scott et al., 1974). An additional identifying
characteristic of tanycytes is that they continue to express
GFAP protein into adulthood, whereas the cuboidal
ependymal cells lack GFAP expression (Hajos and Basco,
1984; Bruni, 1998). Thus, to further confirm the tanycytic
cell type, we labeled sections adjacent to those used for in
situ hybridization for GFAP protein throughout the ros-
trocaudal extent of the third ventricle. In addition, to
examine the cells more closely, serial thin sections mea-
suring 2 ␮m were labeled for GLAST and GLT-1 mRNA
and for GFAP protein in the tuberal and mammillary
portions of the third ventricle. The results show that the
GLT-1 mRNA-expressing cells in the lateral walls, as well
as the GLAST mRNA-expressing cells in the ventral walls
and floor, all express relatively high levels of GFAP pro-
tein (Fig. 5), confirming their tanycytic nature. The com-
mon cuboidal ependymal cells situated dorsal to the
GLT-1 mRNA-expressing cells and in the roof of the mam-
millary recess do not express GFAP (Fig. 5A,D,G). Figure
6 shows that an identical result is found farther caudal, at
the beginning of the mammillary recess. The cuboidal cells
of the roof and dorsal walls lack GFAP expression,
whereas the ependymal cells in the walls and infundibular
recess show GFAP staining (Fig. 6C,F).
Correlation of mRNA and protein
expression
To determine whether the GLT-1 and GLAST mRNAs
are translated into protein, we stained frozen sections
through caudal infundibular recess with specific GLT-1
and GLAST antibodies. The results indicate that ta-
nycytes in the dorsal walls also express GLT-1 protein,
whereas tanycytes in ventral walls and floor or cuboidal
cells of the roof do not (Fig. 7A,C,E,G). GLAST immuno-
staining, on the other hand, was present in all ependymal
Figure 4
GLUTAMATE TRANSPORTERS IN TANYCYTES
cells of floor, walls, and roof (Fig. 7B,D,F,H). The intensity
of the GLAST staining in tanycytes of the lateral walls
appeared to be greater compared with the intensity of
staining in cuboidal cells in the roof, although no attempts
of quantification of signal intensity were made.
DISCUSSION
The main findings of this study are that the glutamate
transporters GLT-1 and GLAST both are expressed by
tanycytes in the walls of the infundibular and mammillary
recess of the third ventricle and that they show unique,
sequential distribution patterns in these cells (Fig. 8).
GLT-1 is expressed primarily dorsally in tanycytes of the
ventricular wall at the caudal end of the tuberal portion
and throughout the mammillary recess. In contrast,
GLAST is expressed more prominently in tanycytes in the
floor and ventral walls of the tuberal portion and at the
rostral end of the mammillary recess. The tanycytes were
identified based on their positive GFAP immunostaining
and their elongated cell bodies. The expression of GLT-1
mRNA and protein in ependymal cells has not been re-
ported so far, probably because it is restricted to a rela-
tively small area in caudal portions of the third ventricle.
The expression of GLAST mRNA and protein in these cells
has been reported in an earlier study (Schmitt et al.,
1997), although no detailed distribution pattern along the
anterior-posterior axis was provided. These data corrobo-
rate previous findings that the expression patterns for
GLT-1 and GLAST are distinct, even though both trans-
porters are expressed mainly in astrocytic glial cells (Torp
et al., 1994; Chaudhry et al., 1995; Lehre et al., 1995;
Schmitt et al., 1996, 1997; Sutherland et al., 1996; Berger
and Hediger, 1998, 2000).
GLAST mRNA is expressed at moderate levels in all
regular cuboidal ependymal cells in all brain ventricles
(Schmitt et al., 1997; Berger and Hediger, 1998, 2000). In
addition, as shown in the present study, higher GLAST
mRNA levels are found in tanycytes in the ventral walls
and floor of the tuberal and mammillary recess portion of
the third ventricle. The higher mRNA expression in these
tanycytes is correlated with a higher level of protein ex-
pression; both Schmitt et al. (1997) and we have found
more intense immunostaining for GLAST protein in the
tanycytes compared with the cuboidal cells. The higher
mRNA expression starts at approximately Bregma ⫺2.1
mm in floor tanycytes and then appears more and more in
tanycytes of the ventricle walls, including the roof and the
floor of the infundibular recess. At its most dorsal exten-
sion, this strong GLAST mRNA labeling covers about one-
Fig. 4. A–L: Coronal sections through the mammillary recess por-
tion of the third ventricle labeled for GLT-1 (A–C,G–I) or GLAST
(D–F,J–L) mRNA at approximate Bregma levels of ⫺3.9 mm (A,D),
⫺4.1 mm (B,E), ⫺4.2 mm (C,F), ⫺4.3 mm (G,J), ⫺4.4 mm (H,K), and
⫺4.5 mm (I,L). In C and F, the stretch of ependymal lining indicated
by the arrows is magnified in the insets above. The GLT-1 mRNA
labeling is present in the walls and the dorsolateral sulci of the
mammillary recess but not in the cuboidal cells of the roof or the
tanycytes of the ventricle floor. The intensity of the GLT-1 mRNA
labeling is greatest in the sulci next to the transition zone with the
roof. The GLAST mRNA labeling in tanycytes is strong in the floor
and lateral walls in D–F and J but is reduced in intensity at the
caudal end of the ventricle (K,L). Scale bar ⫽ 100 ␮m.
109
third to almost half of the third ventricle lining. This
distribution correlates well with the distribution of ta-
nycytes identified in the third ventricle by scanning elec-
tron microscopy, electron microscopy, Golgi staining, nu-
clear staining, or GFAP staining (Scott et al., 1974; Mitro
and Palkovits, 1981; Bruni et al., 1983; Flament-Durand
and Brion, 1985; Wittkowski, 1998). Furthermore, the
strong tanycytic GLAST mRNA labeling correlates with
the hypophysiotropic area, which is lined by the arcuate
nuclear complex on both sides of the ventricle.
GLT-1 is expressed neither by the regular cuboidal
ependymal cells nor by the ventralmost tanycytes. In-
stead, GLT-1 mRNA and protein are found in tanycytes
primarily in dorsal areas near the transition zone to the
regular cuboidal ependymal cells. The highest expression
levels are observed in the dorsal walls and lateral sulcus of
the mammillary recess, and more moderate levels are
observed anterior to that in the tuberal portion. The ex-
pression levels become successively lower toward rostral
areas of the infundibular recess. In the transition zone to
the cuboidal cells, GLT-1-labeled tanycytes form a subja-
cent layer that extends dorsally below the cuboidal cells.
This subjacent layer of tanycytes has also been described
by Mitro and Palkovits (1981). Ventral to this zone and
also in the mammillary recess, GLT-1 is expressed in all
tanycyte layers. In contrast, GLAST mRNA appears to be
expressed only in the superficial tanycyte layers in dorsal
areas. A schematic view of the localization of GLT-1 and
GLAST mRNA labeling in the third ventricle is shown in
Figure 8. To date, the population of GLT-1 mRNA-positive
tanycytes has not been recognized as functionally distinct.
It is not known whether the expression patterns seen in
this study for GLT-1 and GLAST are also seen with other
ependymal transporters. To our knowledge, the present
study is the first to describe the distribution patterns of a
neurotransmitter transporter in the third ventricle
ependyma in such detail. Previous studies have described
the presence of various transporters in the ependymal
lining, including the ␥-aminobutyric acid transporter
GAT-2, the glucose transporters GLUT-1, the peptide
transporter PEPT2, the monocarboxylate transporter
MCT1, and the urea transporter UT3 (Harik et al., 1993;
Berger et al., 1998; Koehler-Stec et al., 1998; Berger and
Hediger, 1999; Conti et al., 1999; Peruzzo et al., 2000).
However, none of these studies has provided a similar
description of the transporter expression pattern through-
out the whole rostrocaudal axis. Earlier anatomical stud-
ies using Golgi impregnation and enzyme histochemical
stains described differences between tanycytes in differ-
ent areas of the ventricle lining (Luppa and Geustel, 1971;
Akmayev et al., 1973; Akmayev and Popov, 1977; Card
and Rafols, 1978; Seress, 1980). The most conspicuous of
these differences in relation to our data is that tanycytes
in dorsal walls extend their basal processes more or less
perpendicularly into the underlying parenchyma, whereas
the processes of ventral tanycytes curve downward and
end on the pial surface (Millhouse, 1971; Zoli et al., 1995;
Chauvet et al., 1998). It is possible that these straight
basal processes bearing tanycytes correlate with the
GLT-1 expressing tanycytes.
The different distribution patterns of GLT-1 and
GLAST mRNA in the tanycytes add to the already consid-
erable amount of evidence that the expression of these two
transporters is regulated by distinct mechanisms. GLT-1
and GLAST have different developmental expression pat-
Figure 5
GLUTAMATE TRANSPORTERS IN TANYCYTES
Fig. 6. Labeling for GLT-1 mRNA, GLAST mRNA, and GFAP
protein at the level of the infundibular recess in three adjacent 16-␮m
sections. A: GLT-1 mRNA is expressed in dorsal walls but not in the
cuboidal cells (D, arrows) of the roof. B: GLAST mRNA is expressed
throughout in walls, the lateral parts of the floor, and in cuboidal cells
(E, arrows). C: GFAP staining is present in the walls and floor but not
terns (Shibata et al., 1996; Sutherland et al., 1996; Furuta
et al., 1997; Ullensvang et al., 1997) and different regional
distributions (Torp et al., 1994; Lehre et al., 1995; Berger
and Hediger, 1998, 2000; Lehre and Danbolt, 1998), and
they are regulated differentially in cell culture (Kondo et
al., 1995; Gegelashvili et al., 1996, 1997, 2000; Eng et al.,
1997; Sonnewald et al., 1997; Swanson et al., 1997; Schlag
et al., 1998; Duan et al., 1999; Stanimirovic et al., 1999;
Zhou and Norenberg, 1999). At this point, we can only
speculate about the possible functional significance of the
differential glutamate transporter expression in the ta-
nycytes of the third ventricle. The strong GLAST expres-
Fig. 5. A–M: Closely spaced, thin (2 ␮m) coronal sections through
the third ventricle at the level of anterior hypothalamus (rostral to
infundibular recess) labeled for GLT-1 mRNA (A,D,H,K), GLAST
mRNA (B,F,I, L), and glial fibrillary acidic protein (GFAP) protein
(C,G,J,M). For A–C, areas labeled by lowercase letters a– c are mag-
nified in D–G, H–J, and K–M, respectively. Cuboidal cells in dorsal
walls (area a) express GLAST mRNA (F) but not GLT-1 mRNA (D)
and express very little GFAP protein (G, arrows; the arrowhead points
to an ependymal cell expressing a small amount of GFAP protein).
Tanycytes in ventral walls (area b) express GLT-1 mRNA (H), GLAST
mRNA (I), and GFAP protein (J). Tanycytes in the ventralmost walls
and ventricle floor (area c) express little GLT-1 mRNA (K) and express
high amounts of GLAST mRNA (L) and GFAP protein (M). Scale
bars ⫽ 25 ␮m in C (also applies to A,B); 10 ␮m in B (also applies to
D,F,H–M).
111
in cuboidal cells of the roof (F, arrows). D–F: Higher magnification
photomicrographs of the cuboidal ependyma indicated by arrowheads
in A–C, respectively. In F, strongly stained subependymal astrocytes
extend processes through the cuboidal cell layer to the lumen. Scale
bar ⫽ 100 ␮m.
sion in tanycytes of the hypophysiotropic area suggests
that glutamate transport mediated by GLAST in these
ventral tanycytes is related to the neuroendocrine activity
of this region. Glutamate is an important regulator of the
release of hypothalamic releasing and inhibitory factors
(Brann and Mahesh, 1997), and, as such, the control of its
levels in extracellular and CSF may be especially critical.
Possibly, GLAST-mediated glutamate uptake by ta-
nycytes contributes to the regulation of these extracellular
glutamate levels, a task that, otherwise, is performed by
astrocytes. Alternatively, the tuberal tanycytes may have
higher metabolic rates because of their involvement in
neuroendocrine activity, and they may need to take up
glutamate for metabolic purposes. It is interesting to note
that glutamate dehydrogenase, which is a metabolic en-
zyme of glutamate, has been detected in these tanycytes
as well (Schmitt and Kugler, 1999). A third possibility
that needs to be considered is that the ventral tanycytes
indeed fulfill a transport function between CSF and the
portal vessels. Possibly, glutamate is taken up at one end
of the tanycyte and released at the other end.
The functional significance of GLT-1 expression in dor-
sally situated tanycytes is less obvious. These dorsal ta-
nycytes are not connected as intimately to the hypophys-
iotropic region, and their processes extend for the most
part into the neuropil of arcuate nuclei. Also, there are no
tight junctions between the tanycytes that would prevent
diffusion into the underlying tissue. Thus, a role of GLT-1
112 U.V. BERGER AND M.A. HEDIGER

Figure 7
GLUTAMATE TRANSPORTERS IN TANYCYTES
Fig. 8. Schematic presentation of the distribution of ependymal
GLT-1 and GLAST mRNA labeling in the third ventricle. The dark-
ness of the shading reflects the intensity of the mRNA labeling. Weak
GLT-1 mRNA labeling appears at the beginning of the tuberal portion
in tanycytes of the ventral walls. It becomes more and more intense
toward caudal levels, until it reaches its highest intensity in the
in a directed transport of glutamate is less likely. It is
possible, however, that GLT-1 expressed by the tanycytes
plays an important role in the control of extracellular
glutamate in both the ventricle and the arcuate nuclei.
Indeed, knockout studies have demonstrated that GLT-1
is more predominant in the removal of extracellular glu-
tamate than GLAST (Tanaka et al., 1997; Watanabe et al.,
1999). In any case, our data have identified a hitherto
unknown tanycyte population that appears to have a spe-
cial function in the caudal third ventricle.
Fig. 7. Immunofluorescence staining for GLT-1 and GLAST pro-
tein in the caudal infundibular recess. A: Positive GLT-1 staining is
present in dorsal tanycytes but not in ventral tanycytes or cuboidal
cells. B: Positive GLAST staining is present throughout the ependy-
mal lining. C,D: Higher magnification photomicrographs of the cuboi-
dal cell layer (arrows) stained for GLAST (D) but not for GLT-1 (C).
E,F: Both GLT-1 and GLAST staining is present in the dorsal ta-
nycytes (arrows). G,H: In ventral tanycytes (arrows), GLT-1 staining
is weak to absent (G), whereas GLAST staining is strong (H). In the
median eminence, GLT-1 immunostaining is localized mostly to the
internal layer, corresponding to the distribution of astrocytes (see also
GLT-1 mRNA labeling in the median eminence in Fig. 6A), whereas
GLAST immunostaining is located throughout the median eminence
thickness, corresponding to the distribution of tanycytes and astro-
cytes. Scale bars ⫽ 100 ␮ m in B (also applies to A); 25 ␮m in H (also
applies to C–G).
113
dorsolateral sulcus of the mammillary recess. GLAST mRNA labeling
is present at moderate intensity in all cuboidal cells. In addition,
GLAST mRNA labeling of stronger intensity is present in tanycytes of
lateral floor and walls throughout the tuberal portion and into the
mammillary recess portion.
ACKNOWLEDGMENTS
This study was supported by grant NS 32001 to M.A.H.
from the National Institutes of Health. We thank Dr. J.E.
Bruni for helpful discussions and Niels C. Danbolt for
providing us with the antibodies to GLT-1 and GLAST.
LITERATURE CITED
Akmayev IG, Popov AP. 1977. Morphological aspects of the hypothalamic-
hypophyseal system. VII. The tanycytes: their relation to the hypophy-
seal adrenocorticotrophic function. An ultrastructural study. Cell Tis-
sue Res 180:263–282.
Akmayev IG, Fidelina OV, Kabolova ZA, Popov AP, Schitkova TA. 1973.
Morphological aspects of the hypothalamic-hypophyseal system. IV.
Medial basal hypothalamus. An experimental morphological study. Z
Zellforsch Mikrosk Anat 137:493–512.
Berger UV, Hediger MA. 1998. Comparative analysis of glutamate trans-
porter expression in rat brain using differential double in situ hybrid-
ization. Anat Embryol (Berl) 198:13–30.
Berger UV, Hediger MA. 1999. Distribution of peptide transporter PEPT2
mRNA in the rat nervous system. Anat Embryol (Berl) 199:439 – 449.
Berger UV, Hediger MA. 2000. Distribution of the glutamate transporters
GLAST and GLT-1 in rat circumventricular organs, meninges, and
dorsal root ganglia. J Comp Neurol 421:385–399.
Berger UV, Tsukaguchi H, Hediger MA. 1998. Distribution of mRNA for
the facilitated urea transporter UT3 in the rat nervous system. Anat
Embryol (Berl) 197:405– 414.
Bjelke B, Fuxe K. 1993. Intraventricular beta-endorphin accumulates in
DARPP-32 immunoreactive tanycytes. Neuroreport 5:265–268.
Brann DW, Mahesh VB. 1997. Excitatory amino acids: evidence for a role
114
in the control of reproduction and anterior pituitary hormone secretion.
Endocrine Rev 18:678 –700.
Brightman MW, Reese TS. 1969. Junctions between intimately apposed
cell membranes in the vertebrate brain. J Cell Biol 40:648 – 677.
Bruni JE. 1998. Ependymal development, proliferation, and functions: a
review. Microsc Res Tech 41:2–13.
Bruni JE, Clattenburg RE, Millar E. 1983. Tanycyte ependymal cells in the
third ventricle of young and adult rats: a Golgi study. Anat Anz 153:
53– 68.
Card JP, Rafols JA. 1978. Tanycytes of the third ventricle of the neonatal
rat: a Golgi study. Am J Anat 151:173–189.
Chaudhry FA, Lehre KP, van Lookeren Campagne M, Ottersen OP, Danbolt
NC, Storm-Mathisen J. 1995. Glutamate transporters in glial plasma
membranes: highly differentiated localizations revealed by quantitative
ultrastructural immunocytochemistry. Neuron 15:711–720.
Chauvet N, Prieto M, Alonso G. 1998. Tanycytes present in the adult rat
mediobasal hypothalamus support the regeneration of monoaminergic
axons. Exp Neurol 151:1–13.
Conti F, Zuccarello LV, Barbaresi P, Minelli A, Brecha NC, Melone M.
1999. Neuronal, glial, and epithelial localization of gamma-
aminobutyric acid transporter 2, a high-affinity gamma-aminobutyric
acid plasma membrane transporter, in the cerebral cortex and neigh-
boring structures. J Comp Neurol 409:482– 494.
Del Bigio MR. 1995. The ependyma: a protective barrier between brain and
cerebrospinal fluid. Glia 14:1–13.
Duan S, Anderson CM, Stein BA, Swanson RA. 1999. Glutamate induces
rapid upregulation of astrocyte glutamate transport and cell-surface
expression of GLAST. J Neurosci 19:10193–10200.
Eng DL, Lee YL, Lal PG. 1997. Expression of glutamate uptake transport-
ers after dibutyryl cyclic AMP differentiation and traumatic injury in
cultured astrocytes. Brain Res 778:215–221.
Flament-Durand J, Brion JP. 1985. Tanycytes: morphology and functions:
a review. Int Rev Cytol 96:121–155.
Furuta A, Rothstein JD, Martin LJ. 1997. Glutamate transporter protein
subtypes are expressed differentially during rat CNS development.
J Neurosci 17:8363– 8375.
Gegelashvili G, Civenni G, Racagni G, Danbolt NC, Schousboe I, Schousboe
A. 1996. Glutamate receptor agonists up-regulate glutamate trans-
porter GLAST in astrocytes. Neuroreport 8:261–265.
Gegelashvili G, Danbolt NC, Schousboe A. 1997. Neuronal soluble factors
differentially regulate the expression of the GLT1 and GLAST gluta-
mate transporters in cultured astroglia. J Neurochem 69:2612–2615.
Gegelashvili G, Dehnes Y, Danbolt NC, Schousboe A. 2000. The high-
affinity glutamate transporters GLT1, GLAST, and EAAT4 are regu-
lated via different signalling mechanisms. Neurochem Int 37:163–170.
Hajos F, Basco E. 1984. The surface-contact glia. Adv Anat Embryol Cell
Biol 84:1–79.
Harik SI, Hall AK, Richey P, Andersson L, Lundahl P, Perry G. 1993.
Ontogeny of the erythroid/HepG2-type glucose transporter (GLUT-1)
in the rat nervous system. Brain Res Dev Brain Res 72:41– 49.
Koehler-Stec EM, Simpson IA, Vannucci SJ, Landschulz KT, Landschulz
WH. 1998. Monocarboxylate transporter expression in mouse brain.
Am J Physiol 275:E516 –524.
Kondo K, Hashimoto H, Kitanaka J, Sawada M, Suzumura A, Marunouchi
T, Baba A. 1995. Expression of glutamate transporters in cultured glial
cells. Neurosci Lett 188:140 –142.
Lehre KP, Danbolt NC. 1998. The number of glutamate transporter sub-
type molecules at glutamatergic synapses: chemical and stereological
quantification in young adult rat brain. J Neurosci 18:8751– 8757.
Lehre KP, Levy LM, Ottersen OP, Storm-Mathisen J, Danbolt NC. 1995.
Differential expression of two glial glutamate transporters in the rat
brain: quantitative and immunocytochemical observations. J Neurosci
15:1835–1853.
Luppa H, Geustel G. 1971. Location and characterization of hydrolytic
enzymes of the 3rd ventricle lining in the region of the recessus infun-
dibularis of the rat. A study on the function of the ependyma. Brain Res
29:253–270.
Millhouse OE. 1971. A Golgi study of third ventricle tanycytes in the adult
rodent brain. Z Zellforsch Mikrosk Anat 121:1–13.
Mitro A, Palkovits M. 1981. Morphology of the rat brain ventricles,
ependyma, and periventricular structures. Bibl Anat 21:1–110.
Paxinos G, Watson C. 1998. The rat brain in stereotaxic coordinates. 4th
ed. San Diego: Academic Press.
Peruzzo B, Pastor FE, Blazquez JL, Schobitz K, Pelaez B, Amat P, Rodri-
U.V. BERGER AND M.A. HEDIGER
guez EM. 2000. A second look at the barriers of the medial basal
hypothalamus. Exp Brain Res 132:10 –26.
Pilgrim C. 1978. Transport function of hypothalamic tanycyte ependyma:
how good is the evidence? Neuroscience 3:277–283.
Pines G, Danbolt NC, Bjoras M, Zhang Y, Bendahan A, Eide L, Koepsell H,
Storm-Mathisen J, Seeberg E, Kanner BI. 1992. Cloning and expres-
sion of a rat brain L-glutamate transporter [published erratum appears
in Nature (1992) 360(6406):768]. Nature 360:464 – 467.
Schaeren-Wiemers N, Gerfin-Moser A. 1993. A single protocol to detect
transcripts of various types and expression levels in neural tissue and
cultured cells: in situ hybridization using digoxigenin-labelled cRNA
probes. Histochemistry 100:431– 440.
Schlag BD, Vondrasek JR, Munir M, Kalandadze A, Zelenaia OA, Roth-
stein JD, Robinson MB. 1998. Regulation of the glial Na⫹-dependent
glutamate transporters by cyclic AMP analogs and neurons. Mol Phar-
macol 53:355–369.
Schmitt A, Kugler P. 1999. Cellular and regional expression of glutamate
dehydrogenase in the rat nervous system: non-radioactive in situ hy-
bridization and comparative immunocytochemistry. Neuroscience 92:
293–308.
Schmitt A, Asan E, Puschel B, Jons T, Kugler P. 1996. Expression of the
glutamate transporter GLT1 in neural cells of the rat central nervous
system: non-radioactive in situ hybridization and comparative immu-
nocytochemistry. Neuroscience 71:989 –1004.
Schmitt A, Asan E, Puschel B, Kugler P. 1997. Cellular and regional
distribution of the glutamate transporter GLAST in the CNS of rats:
nonradioactive in situ hybridization and comparative immunocyto-
chemistry. J Neurosci 17:1–10.
Scott DE, Kozlowski GP, Sheridan MN. 1974. Scanning electron micros-
copy in the ultrastructural analysis of the mammalian cerebral ven-
tricular system. Int Rev Cytol 37:349 –388.
Seress L. 1980. Development and structure of the radial glia in the post-
natal rat brain. Anat Embryol 160:213–226.
Shibata T, Watanabe M, Tanaka K, Wada K, Inoue Y. 1996. Dynamic
changes in expression of glutamate transporter mRNAs in developing
brain. Neuroreport 7:705–709.
Silverman AJ, Knigge KM, Peck WA. 1972. Transport capacity of median
eminence. I. Amino acid transport. Neuroendocrinology 9:123–132.
Sonnewald U, Westergaard N, Schousboe A. 1997. Glutamate transport
and metabolism in astrocytes. Glia 21:56 – 63.
Stanimirovic DB, Ball R, Small DL, Muruganandam A. 1999. Developmental
regulation of glutamate transporters and glutamine synthetase activity in
astrocyte cultures differentiated in vitro. Int J Dev Neurosci 17:173–184.
Storck T, Schulte S, Hofmann K, Stoffel W. 1992. Structure, expression,
and functional analysis of a Na(⫹)-dependent glutamate/aspartate
transporter from rat brain. Proc Natl Acad Sci USA 89:10955–10959.
Sutherland ML, Delaney TA, Noebels JL. 1996. Glutamate transporter
mRNA expression in proliferative zones of the developing and adult
murine CNS. J Neurosci 16:2191–2207.
Swanson RA, Liu J, Miller JW, Rothstein JD, Farrell K, Stein BA, Longue-
mare MC. 1997. Neuronal regulation of glutamate transporter subtype
expression in astrocytes. J Neurosci 17:932–940.
Tanaka K, Watase K, Manabe T, Yamada K, Watanabe M, Takahashi K,
Iwama H, Nishikawa T, Ichihara N, Kikuchi T, Okuyama S, Ka-
washima N, Hori S, Takimoto M, Wada K. 1997. Epilepsy and exacer-
bation of brain injury in mice lacking the glutamate transporter GLT-1.
Science 276:1699 –1702.
Torp R, Danbolt NC, Babaie E, Bjoras M, Seeberg E, Storm-Mathisen J,
Ottersen OP. 1994. Differential expression of two glial glutamate
transporters in the rat brain: an in situ hybridization study. Eur
J Neurosci 6:936 –942.
Ullensvang K, Lehre KP, Storm-Mathisen J, Danbolt NC. 1997. Differen-
tial developmental expression of the two rat brain glutamate trans-
porter proteins GLAST and GLT. Eur J Neurosci 9:1646-1655.
Watanabe T, Morimoto K, Hirao T, Suwaki H, Watase K, Tanaka K. 1999.
Amygdala-kindled and pentylenetetrazole-induced seizures in gluta-
mate transporter GLAST-deficient mice. Brain Res 845:92–96.
Wittkowski W. 1998. Tanycytes and pituicytes: morphological and func-
tional aspects of neuroglial interaction. Microsc Res Tech 41:29 – 42.
Zhou BG, Norenberg MD. 1999. Ammonia downregulates GLAST mRNA
glutamate transporter in rat astrocyte cultures. Neurosci Lett 276:
145–148.
Zoli M, Ferraguti F, Frasoldati A, Biagini G, Agnati LF. 1995. Age-related
alterations in tanycytes of the mediobasal hypothalamus of the male
rat. Neurobiol Aging 16:77– 83.