- Email: [email protected]
Localization of the glutamate transporter protein GLAST in rat retina
Brain Research 744 Ž1997. 129–137
Research report
Localization of the glutamate transporter protein GLAST in rat retina Knut P. Lehre ) , Svend Davanger, Niels C. Danbolt Anatomical Institute, UniÕersity of Oslo, POB 1105 Blindern, N-0317 Oslo, Norway Accepted 13 August 1996
Abstract Glutamate is a neurotransmitter in retina. Glutamate transporter proteins keep the resting extracellular glutamate concentration low. This is required for normal neurotransmission and prevents the extracellular concentration of glutamate from reaching toxic levels. Here we describe the light and electron microscopic localization of the glutamate transporter protein GLAST in rat retina using an antibody raised and affinity purified against a peptide corresponding to amino acid residues 522–541. The strongest immunocytochemical labelling was observed in the outer plexiform layer, ganglion cell layer, and optic disc. GLAST was found in Muller cell processes in all retinal ¨ layers, notably ensheathing the photoreceptor terminals in the outer plexiform layer, and in astrocytes close to vessels in the inner retina and optic disc. No labelling was observed in neurons. The electrophoretic mobility of GLAST in retina was similar to that in cerebellum. In conclusion, the findings are in agreement with those reported by Derouiche and Rauen w7x, except that we did not detect any GLAST in the retinal pigment epithelium. Keywords: Glutamate uptake; GLAST; Retina; Pigment epithelium; Glia; Muller cell; Astrocyte; Immunocytochemistry ¨
1. Introduction Glutamate is regarded as a neurotransmitter of both photoreceptor cells, bipolar cells and ganglion cells of the retina w25x. Powerful uptake mechanisms capable of quickly removing extracellular glutamate are therefore required. This is accomplished by glutamate transporter proteins located in the plasma membranes of both neurons and glial cells in the brain Žfor review see w6x. as well as in the retina w2,9x. Several glutamate transporters have been cloned: GLAST w32x, GLT-1 w27x, EAAC1 w17x, EAAT4 w11x. A series of immunocytochemical studies have appeared describing the localization of the GLT-type glutamate transporter in retinal neurons w1,10,12,28x. GLAST has been described in retinal Muller cells, astrocytes and ¨ pigment epithelial cells in albino rats w7x, using an antiserum raised against GLAST protein purified from rat brain. The latter report is in agreement with an in situ hybridization study w26x where GLAST mRNA was de-
tected in Muller cells and astrocytes. With regard to the ¨ pigment epithelium, however, the two reports disagree. The pigment epithelium was found immunoreactive to the GLAST antiserum w7x, while GLAST mRNA was not detected in the epithelium w26x. The possibility exists that the antiserum used w7x could recognize other proteins in addition to GLAST. Data with more defined antibodies are therefore required. Here we demonstrate, light and electron microscopically, the cellular localization of the GLAST protein in retina of albino as well as pigmented rats, using antibodies w21x specifically recognizing a peptide sequence in the C-terminal part of the protein. Since abnormalities have been noted in the retina of albino rats w5x, this study is based mainly on pigmented rats. We confirm the presence of GLAST in Muller cells and astrocytes, and its absence ¨ from neuronal elements. However, the pigment epithelium was immunonegative for GLAST.
2. Materials and methods 2.1. Materials ) Corresponding author. Department of Anatomy, Institute of Basic Medical Sciences, University of Oslo, POB 1105 Blindern, N-0317 Oslo, Norway. Fax: Ž47. 2285 1278; E-mail: [email protected]
Alkaline phosphatase-conjugated anti-rabbit IgG was from Sigma ŽSt. Louis, MO, USA.. Nitrocellulose sheets
0006-8993r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 0 6 - 8 9 9 3 Ž 9 6 . 0 1 0 2 2 - 0
130
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
Ž0.22 m m pores, 100% nitrocellulose. and electrophoresis equipment were from Hoefer Scientific Instruments ŽSan Francisco, CA, USA.. Glutaraldehyde, EM grade, was from TAAB ŽReading, UK. and Durcupan ACM from Fluka ŽBuchs, Switzerland.. Molecular mass markers were the ‘Rainbow markers’ from Amersham ŽBuckinghamshire, UK.. Other reagents were analytical grade or better. 2.2. Antibodies Antibodies to GLAST were prepared as described w21x. Briefly, synthetic peptides corresponding to residues 1–25 ŽMTKSNGEEPRMGSRMERFQQGVRKRC. or 522–541 ŽPYQLIAQDNEPEKPVADSET. of the glutamate transporter GLAST w32x were coupled to hemocyanin with maleimido benzoyl-N-hydroxysuccinimide or glutaraldehyde, respectively, and used to immunize rabbits. The ensuing antisera Žanti-A1 and anti-A522, respectively. were affinity purified against the peptides immobilized by N-hydroxysuccinimide on agarose. The peptides are referred to as A1 and A522, respectively, and the corresponding antibodies as anti-A1 and anti-A522. This report is based Fig. 2. Low magnification light microscopic localization of GLAST in rat retina. GLAST ŽA. 0.1 m grml anti-A522, and preimmune control ŽB. 1 m grml preimmune IgG. Due to narrowing of the iris diaphragm to produce contrast to make the control section visible, photoreceptors can be seen in B, but not in A. Note in A the strong labelling in the outer plexiform and ganglion cell layers, the radiating fibres, and the difference in staining in the inner and outer part of the inner plexiform layer. S, photoreceptor inner and outer segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer. Scale bar, 30 m m.
mainly on the anti-A522 antibodies. The anti-A1 antibodies were only used when stated to verify that the immunoreactivity observed with the anti-A522 was due to GLAST.
Fig. 1. Immunoblot of crude SDS extracts of rat retina ŽR., hippocampal formation ŽH. and cerebellum ŽC. subjected to SDS-PAGE Ž5, 2 and 2 m g proteinrlane, respectively. and reacted with GLAST antibody antiA522 followed by alkaline phosphatase conjugated secondary antibodies. Note the differences in electrophoretic mobility and immunoreactivity of the labelled bands in the three lanes. Molecular mass markers ŽkDa. are indicated.
Fig. 3. GLAST in the optic disc in rat retina. Note in A the strong staining in the inner part of the optic disc where branching vessels Žarrow. are situated and where the inner limiting membrane contains astrocytes. Note in B the labelling of the astroglial columns Ža. and their thin processes Žarrowhead. within the nerve fibre fascicles. Sections from albino rat. Scale bars, 25 m m.
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
2.3. Protein measurement Protein was determined in homogenates dissolved in 1% sodium dodecyl sulfate ŽSDS. as described w24x using bovine serum albumin as standard. SDS was added to give equal concentrations in samples and standards.
131
2.4. Preparation of tissue extracts, electrophoresis, blotting and immunostaining of blots This was done as described w20,21,34x. The gels consisted of 10% acrylamide and were run slowly overnight at about 58C. The samples were incubated Ž2 min, 1008C.
Fig. 4. High magnification interference contrast light microscopic localization of GLAST in rat retina. The picture was taken from the same section as Fig. 2A. Note the slender stained fibres along vessel wall Žarrow. representing astrocytic processes and the small dot shaped stained structures in transversely cut nerve fibre bundles Žarrowheads. representing Muller cell processes. Scale bar, 15 m m. ¨
132
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
with 4% Žvrv. mercaptoethanol prior to application. The molecular mass markers were run several lanes away from the samples because mercaptoethanol changes the marker mobility w22x. 2.5. Immunocytochemistry Perfusion of rats ŽPVG or Wistar strain, Møllegaard Hansen, Denmark., sectioning and processing of sections were done as described w21x. Retinas or eyecups were embedded in 15–20% Žwrv. gelatine in 0.1 M sodium phosphate buffer pH 7.4 ŽNaPi. and the blocks hardened in the tissue fixative before vibratome sectioning. For labelling of cryostat sections rats were decapitated, anterior poles of the eyes removed in 0.1 M NaPi, posterior eyecups fixed Ž3 h, at room temperature. in 4% freshly depolymerized paraformaldehyde in 0.1 M NaPi, rinsed in 0.1 M NaPi, immersed Ž1–2 days, at 48C. in 30% sucrose in 0.1 M NaPi, and 10–15 m m cryostat sections air dried on slides and stored Ž1–3 days, at y208C. before immuno-
Fig. 6. GLAST in nerve fibre bundle in the ganglion cell layer. Note glial profiles containing reaction product concentrated along the inside of the cell membrane Ž). between unlabelled nerve fibres. Scale bar, 400 nm.
labelling. The anti-A522 antibodies were used at 0.1–0.3 m grml. For negative controls, primary antibodies were omitted, primary antibodies were substituted by preimmune IgG, or primary antibodies were preincubated with the A522 peptide. For light microscopy the sections were observed using a Leitz DM R microscope. Pictures were taken with interference contrast optics when stated. Figures show retinas from pigmented animals unless albino is noted.
3. Results 3.1. Specificity of the antibody and relatiÕe expressions of GLAST in brain and retina
Fig. 5. Interference contrast light micrographs showing lack of GLAST labelling in retinal pigment epithelium of albino rat. The sections are incubated with Triton X-100 to secure penetration of antibodies. A and B are incubated in the same solution, with anti-A522. C is a negative control where the primary antibody has been omitted. Note the unlabelled hexagonal cells in A, a tangential section through the pigment epithelium. Note in B the contrast between the unlabelled pigment epithelium Žarrow. and the heavy staining in the OPL and ONL. Note also that there is no difference between the pigment epithelium in B and C. Without the interference contrast effect the pigment epithelium is invisible in A, B and C. Scale bar, 30 m m.
Even though the specificity of the anti-GLAST antibodies is firmly established in the rat brain w4,21,22x, the possibility existed that retina expresses crossreacting polypeptides. Therefore, the antibody specificity was tested on immunoblots ŽFig. 1. of electrophoretically separated SDS-solubilized whole tissue from rat retina Žpars nervosa., hippocampus and cerebellum. In the retinal extracts ŽLane R, 5 m g protein., the anti-A522 antibodies recognized a broad, somewhat heterogenous band similar to those observed in the cerebellum ŽLane C, 2 m g protein. and hippocampus ŽLane H, 2 m g protein.. Further, in tissue sections, no labelling was observed when they were incubated with preimmune IgG ŽFig. 2B. or with antibodies
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
133
Fig. 7. Perivascular GLAST in the ganglion cell layer. A: astrocyte with evenly pale nucleus Žn. localized next to vessel contains reaction product Žarrows. in the cytoplasm close to the cell membrane. B: unlabelled smooth muscle cells Žs. of ganglion cell layer vessels are surrounded by immunopositive glial processes Ž).. Note that penetration of immunoreactants is secured by mechanical opening of cells in A and B. ŽPatches of subsarcolemmal darkness of smooth muscle cells are not due to reaction product.. Arrowheads, basal laminae of vessel walls. Scale bar, 2 m m.
preabsorbed with A522-peptide Ždata not shown.. The apparent molecular mass of the immunoreactive proteins in retina Ž67–68 kDa. was similar to that in cerebellum Ž67 kDa. but higher than in hippocampus Ž63 kDa.. By comparing the staining intensity and the amounts of protein in each lane, it is concluded that the concentration of GLAST protein in retina is less than half of that in hippocampus. However, the intensity of the immunocytochemical labelling of individual cells in the retina resembled that in the brain w21x, indicating that the concentration of GLAST in those membranes that are immunopositive is similar to that in the brain. 3.2. Immunocytochemical localization 3.2.1. Light microscopy The strongest labelling ŽFig. 2A and Fig. 4. was observed in the ganglion cell layer ŽGCL., the outer plexi-
form layer ŽOPL., and ŽFig. 3A. the inner part of the optic disc. The second strongest labelling ŽFig. 2A and Fig. 4. was found in the inner plexiform layer ŽIPL., the inner nuclear layer ŽINL., and the outer nuclear layer ŽONL.. The outer 2r5 of the IPL Žsublamina A. was weaker than the rest of this layer, and a slightly stronger labelled zone was observed at the border between the INL and IPL. In the part of the retina immediately adjacent to the vitreous body, the staining was present in structures with ramifications penetrating into the retina between bundles of nerve fibres ŽFig. 4.. Within transversely cut nerve fibre bundles small stained dots were present. In more obliquely cut nerve fibre bundles, the stained structures were shaped like longer thin processes separating the unstained nerve fibres. Bundles of nerve fibres passing into the optic disc were separated by stained structures of the same appearance. In the prelaminar portion of the optic disc there was labelling of cells in the astroglial cell columns and thin
Fig. 8. GLAST in the inner plexiform layer. A: labelled glial profiles separate and encircle unlabelled neural processes. B: labelled Ž). glial process contacting unlabelled bipolar terminal Žt.. Scale bars, 400 nm.
134
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
Fig. 9. GLAST in the inner nuclear layer. Labelled glial processes separate cell bodies which appear unlabelled. One of the neuronal cell bodies is opened to the vibratome section surface Žw.. The glial process near the section surface Žarrow. is strongly stained. Glial processes far from the section surface Žarrowhead. show weaker labelling or no labelling. Scale bar, 500 nm.
processes within the nerve fibre fascicles ŽFig. 3B.. In the GCL thin stained processes appeared along the blood vessels ŽFig. 4.. This produced a dark demarcation of the vessels similar to the astrocytic labelling described for S-100 protein w19x. In the IPL, thick radiating fibres in addition to numerous small, thin, laminar processes were stained. There were clearly fewer of these labelled small processes in the sublamina A than in the rest of the IPL. In the INL, a network of thin labelled structures separated apparently unlabelled perikarya. In the OPL, strong labelling was observed in thin processes ensheathing rod spherules. The ONL displayed a similar picture as the INL. Labelling was observed all the way through to the outer limiting membrane. The inner and outer segments of the photoreceptors were unlabelled ŽFig. 4..
No major differences between the labelling of retinas from albino and pigmented rats were observed. However, in some of the albino animals the overall staining intensity in the ONL appeared stronger. The pigment epithelium was studied in vibratome sections from albino rats ŽFig. 5. and cryostat sections from albino and pigmented rats Žnot shown.. At antibody concentrations giving strong labelling of the neuroretina, there was no labelling in adjacent pigment epithelial cells. The absence of pigment epithelial labelling in the presence of neuroretinal labelling was observed with antibodies to the C-terminal of GLAST Žanti-A522; Fig. 5. as well as with antibodies to the N-terminal Žanti-A1; not shown.. These experiments were performed in the presence of Triton X-100 to secure penetration of immunoreagents.
Fig. 10. GLAST in the outer plexiform ŽA. and outer nuclear ŽB. layer. A: photoreceptor terminals Žt. appear unlabelled. Labelling Ž). can be observed along the cell membranes and in the adjacent cytoplasm of the thin glial processes interposed between the terminals. B: glial profiles in between photoreceptor cell perikarya and processes are labelled. No labelling appears in the photoreceptors. Scale bars, 300 nm.
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
3.2.2. Electron microscopy Immediately next to the vitreous body staining was present on the plasmalemma and adjacent cytoplasm of structures bordering on the inner limiting membrane Ždata not shown.. In nerve fibre bundles, the nerve fibres appeared unlabelled, but the glial processes running in between them clearly contained reaction product ŽFig. 6.. Cell bodies of ganglion cells were unlabelled, but occasionally labelled glial processes could be seen next to the cell bodies Ždata not shown.. Cells situated adjacent to vessels in the GCL with evenly pale nuclei and filaments in their perikarya and processes Ži.e. astrocytes. contained reaction product where exposed to the surface of the vibratome section ŽFig. 7A.. Glial processes contacting the basal lamina of endothelial cells and smooth muscle cells were labelled ŽFig. 7B.. At higher magnification some of these labelled processes could be seen to contain filaments typical of astrocytes. In the IPL, glial profiles surrounding synapses and axons were labelled ŽFig. 8A.. None of the observed nerve terminals, including those containing synaptic ribbons Žbipolar terminals., appeared stained ŽFig. 8B.. In both nuclear layers, stained glial processes could be seen to separate unstained neuronal cell bodies and processes ŽFigs. 9 and 10B.. In the OPL the photoreceptor terminals appeared unlabelled. This was true even when the terminals were cut open by the vibratome knife, securing access of the anti-A522 antibody to its epitopes, which are on the inside of cell membranes w21x. Labelling was observed along the cell membranes and in the adjacent cytoplasm of the thin glial processes interposed between the photoreceptor terminals ŽFig. 10A.. Glial processes far from the surface of the vibratome sections ŽFig. 9., and glial processes in the control sections were unlabelled.
4. Discussion The labelled structures bordering on the inner limiting membrane are interpreted as Muller cell endfeet, although ¨ astrocytes proper occasionally border directly on the inner limiting membrane. Several investigators have noted that rat retinal astrocytes are not found between ganglion cell axons w13,23,31x. At this point, the retina of the rat, and mouse w16x, differ from that of several other species, as nerve fibres show a strong tendency to attract astrocytes in the cat w3,15,31x, rabbit w31x and monkey retina w3,8x. Thus, the labelled glial processes in nerve fibre bundles that we observed in the rat retina most likely represent Muller cell ¨ processes. Stone and Dreher w31x have noted that astrocytic processes do not invest the somata of ganglion cells in the retina of rat and cat. This has been confirmed for the cat w15x, and shown to be true also for the monkey retina w8x. Thus, the labelled glial processes contacting ganglion cells represent Muller cell processes. All the glial labelling ¨ described in the outer part of the retina represents Muller ¨ cell labelling, notably the strong labelling of the thin glial
135
processes interposed between the photoreceptor terminals. It is concluded that the anti-A522 GLAST antibodies label Muller cell processes in all retinal layers. ¨ The presence of GLAST in astrocytes proper was confirmed both at the light and electron microscopic levels: as the optic disc contains no Muller cells, all the glial la¨ belling observed in the disc represents astrocytes. This labelling is in agreement with the observation that GLAST mRNA is found in about 90% of cells in the optic nerve head, where more than 90% of cells express glial fibrillary acidic protein mRNA w26x. The glial investments of blood vessels in the inner retina of the rat are formed by astrocytic processes w14,19,31x. Processes containing numerous filaments, often found along the blood vessel basal laminae, were seen to emerge from cells with evenly pale nuclei situated close to blood vessels in the innermost part of the retina. These cell bodies and processes often could be seen to be labelled, and were interpreted as GLAST positive astrocytes. Sarantis and Attwell w30x have shown that glutamate evokes an inward membrane current in Muller cells from ¨ the rabbit retina. This glutamate uptake current is dependent on extracellular sodium and intracellular potassium. This is in agreement with the ion cotransport characteristics of brain glutamate transporters, including the GLAST protein w18x. The findings of Otori et al. w26x indicate that Muller ¨ cells and astrocytes in the rat retina express GLAST mRNA. Since GLT does not appear to be present in retinal glial cells w28x, GLAST most probably is responsible for glutamate transport in retinal glia. This is in contrast to the brain, where GLT and GLAST coexist in astrocytes w21x. Labelling of the pigment epithelium was neither observed with antibodies to the C-terminal part of GLAST Žresidues 522–541. nor with antibodies to the N-terminal Žresidues 1–25.. The lack of labelling of the pigment epithelium is in agreement with the in situ hybridization study by Otori et al. w26x where no GLAST mRNA was detected at this site. However, the presence of GLAST in the retinal pigment epithelium has been reported in another immunocytochemical study w7x. This study w7x is based on an antiserum raised against the GLAST protein purified from rat brain. Except for the labelling of the pigment epithelium, the results of these authors are in agreement with ours. It is legitimate to speculate why this antiserum labels the pigment epithelium. One possibility is that the GLAST protein expressed in the pigment epithelium is different from the GLAST protein expressed in the neuroretina and the brain, and that the antiserum w7x recognizes epitopes common to the two versions. Another possibility is that the antiserum used in the cited study w7x could contain some antibodies to a retinal epithelial protein different from GLAST. The in situ hybridization signal for GLAST mRNA outlining the ventricles of the rat brain reported by our group w33x should not, as suggested w7x, be taken as support for the presence of GLAST in the pigment epithelium.
136
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
This signal has not been interpreted as ependymal w33x, but has been explained by the presence of a subependymal glial plexus immunoreactive to GLAST w21x. Observations using antibodies to synthetic peptides corresponding to the EAAC glutamate transporter protein w17x suggest that this protein is expressed in the rat pigment epithelium ŽK. Ullensvang and K.P. Lehre, unpubl. obs... This may explain the glutamate uptake reported in rat pigment epithelium w29x. In conclusion, GLAST is expressed in neuroretina essentially as described previously w7x. However, the glutamate uptake observed in pigment epithelium is not, as suggested w7x, due to the GLAST protein w32x found elsewhere in the CNS, but to another glutamate transporter.
w12x
w13x
w14x
w15x
w16x
w17x
Acknowledgements w18x
We thank Jon Storm-Mathisen for discussions and critical reading of the manuscript and Gunnar F. Lothe for technical assistance. This work was supported by the Norwegian Research Council, Anders Jahres Fond, Nansenfondet, Schreiners fond and EU BIOMED2 Žcontract no. BMH4-CT95-0571..
w19x
w20x w21x
References w1x Brandstatter, J.H., Greferath, U., Euler, T. and Wassle, H., Co¨ ¨ stratification of GABA A receptors with the directionally selective circuitry of the rat retina, Vis. Neurosci., 12 Ž1995. 345–358. w2x Brew, H. and Attwell, D., Electrogenic glutamate uptake is a major current carrier in the membrane of axolotl retinal cells, Nature, 327 Ž1987. 707–709. w3x Bussow, H., The astrocytes in the retina and optic nerve head of ¨ mammals: a special glia for the ganglion cell axons, Cell Tissue Res., 206 Ž1980. 367–378. w4x Chaudhry, F.A., Lehre, K.P., van Lookeren Campagne, M., Ottersen, O.P., Danbolt, N.C. and Storm-Mathisen, J., Glutamate transporters in glial plasma membranes: highly differentiated localizations revealed by quantitative ultrastructural immunocytochemistry, Neuron, 15 Ž1995. 711–720. w5x Creel, D., Inappropriate use of albino animals as models in research, Pharmac. Biochem. BehaÕ., 12 Ž1980. 969–977. w6x Danbolt, N.C., The high affinity uptake system for excitatory amino acid neurotransmitters in the brain, Prog. Neurobiol., 44 Ž1994. 377–396. w7x Derouiche, A. and Rauen, T., Coincidence of L-glutamaterL-aspartate transporter ŽGLAST. and glutamine synthetase ŽGS. immunoreactions in retinal glia: evidence for coupling of GLAST and GS in transmitter clearance, J. Neurosci. Res., 42 Ž1995. 131–143. w8x Distler, C., Weigel, H. and Hoffmann, K.-P., Glia cells of the monkey retina. I. Astrocytes, J. Comp. Neurol., 333 Ž1993. 134–147. w9x Eliasof, S. and Werblin, F., Characterization of the glutamate transporter in retinal cones of the tiger salamander, J. Neurosci., 13 Ž1993. 402–411. w10x Euler, T. and Wassle, H., Immunocytochemical identification of ¨ cone bipolar cells in the rat retina, J. Comp. Neurol., 361 Ž1995. 461–478. w11x Fairman, W.A., Vandenberg, R.J., Arriza, J.L., Kavanaugh, M.P.
w22x
w23x
w24x
w25x w26x
w27x
w28x
w29x
w30x
w31x
and Amara, S.G., An excitatory amino-acid transporter with properties of a ligand-gated chloride channel, Nature, 375 Ž1995. 599–603. Grunert, U., Martin, P.R. and Wassle, H., Immunocytochemical ¨ ¨ analysis of bipolar cells in the macaque monkey retina, J. Comp. Neurol., 348 Ž1994. 607–627. Hildebrand, C. and Waxman, G., Regional node-like membrane specializations in non-myelinated axons of rat retinal nerve fiber layer, Brain Res., 258 Ž1983. 23–32. Holash, J.A. and Stewart, P.A., The relationship of astrocyte-like cells to the vessels that contribute to the blood-ocular barriers, Brain Res., 629 Ž1993. 218–224. Hollander, H., Makarov, F., Dreher, Z., van Driel, D., Chan-Ling, T. ¨ and Stone, J., Structure of the macroglia of the retina: sharing and division of labour between astrocytes and Muller cells, J. Comp. ¨ Neurol., 313 Ž1991. 587–603. Huxlin, K.R., Sefton, A.J. and Furby, J.H., The origin and development of retinal astrocytes in the mouse, J. Neurocytol., 21 Ž1992. 530–544. Kanai, Y. and Hediger, M.A., Primary structure and functional characterization of a high-affinity glutamate transporter, Nature, 360 Ž1992. 467–471. Klockner, U., Storck, T., Conradt, M. and Stoffel, W., Electrogenic ¨ L-glutamate uptake in Xenopus-laevis oocytes expressing a cloned rat brain L-glutamate-L-aspartate transporter ŽGLAST-1., J. Biol. Chem., 268 Ž1993. 14594–14596. Kondo, H., Takahashi, H. and Takahashi, Y., Immunohistochemical study of S-100 protein in the postnatal development of Muller cells ¨ and astrocytes in the rat retina, Cell Tissue Res., 238 Ž1984. 503–508. Laemmli, U.K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature, 227 Ž1970. 680–685. Lehre, K.P., Levy, L.M., Storm-Mathisen, J., Ottersen, O.P. and Danbolt, N.C., Differential expression of two glial glutamate transporters in the rat brain: quantitative and immunocytochemical observations, J. Neurosci., 15 Ž1995. 1835–1853. Levy, L.M., Lehre, K.P., Walaas, S.I., Storm-Mathisen, J. and Danbolt, N.C., Down-regulation of glial glutamate transporters after glutamatergic denervation in the rat brain, Eur. J. Neurosci., 7 Ž1995. 2036–2041. Ling, T., Mitrofanis, J. and Stone, J., Origin of retinal astrocytes in the rat: evidence of migration from the optic nerve, J. Comp. Neurol., 286 Ž1989. 345–352. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 Ž1951. 265–275. Massey, S.C. and Redburn, D.A., Transmitter circuits in the vertebrate retina, Prog. Neurobiol., 28 Ž1987. 55–96. Otori, Y., Shimada, S., Tanaka, K., Ishimoto, I., Tano, Y. and Tohyama, M., Marked increase in glutamate-aspartate transporter ŽGLASTrGluT-1. mRNA following transient retinal ischemia, Mol. Brain Res., 27 Ž1994. 310–314. Pines, G., Danbolt, N.C., Bjøras, ˚ M., Zhang, Y., Bendahan, A., Eide, L., Koepsell, H., Seeberg, E., Storm-Mathisen, J. and Kanner, B.I., Cloning and expression of a rat brain L-glutamate transporter, Nature, 360 Ž1992. 464–467. Rauen, T. and Kanner, B.I., Localizatiion of the glutamate transporter GLT-1 in rat and macaque monkey retinae, Neurosci. Lett., 169 Ž1994. 137–140. Salceda, R. and Saldana, M.R., Glutamate and taurine uptake by retinal pigment epithelium during rat development, Comp. Biochem. Physiol., 104C Ž1993. 311–316. Sarantis, M. and Attwell, D., Glutamate uptake in mammalian retinal glia is voltage- and potassium-dependent, Brain Res., 516 Ž1990. 322–325. Stone, J. and Dreher, Z., Relationship between astrocytes, ganglion cells and vasculature of the retina, J. Comp. Neurol., 255 Ž1987. 35–49.
K.P. Lehre et al.r Brain Research 744 (1997) 129–137 w32x Storck, T., Schulte, S., Hofmann, K. and Stoffel, W., Structure, expression, and functional analysis of a Na q-dependent glutamateraspartate transporter from rat brain, Proc. Natl. Acad. Sci. USA, 89 Ž1992. 10955–10959. w33x Torp, R., Danbolt, N.C., Babaie, E., Bjøras, ˚ M., Seeberg, E., Storm-Mathisen, J. and Ottersen, O.P., Differential expression of
137
two glial glutamate transporters in the rat brain: an in situ hybridization study, Eur. J. Neurosci., 6 Ž1994. 936–942. w34x Towbin, H., Staehelin, T. and Gordon, J., Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications, Proc. Natl. Acad. Sci. USA, 76 Ž1979. 4350–4354.
Research report
Localization of the glutamate transporter protein GLAST in rat retina Knut P. Lehre ) , Svend Davanger, Niels C. Danbolt Anatomical Institute, UniÕersity of Oslo, POB 1105 Blindern, N-0317 Oslo, Norway Accepted 13 August 1996
Abstract Glutamate is a neurotransmitter in retina. Glutamate transporter proteins keep the resting extracellular glutamate concentration low. This is required for normal neurotransmission and prevents the extracellular concentration of glutamate from reaching toxic levels. Here we describe the light and electron microscopic localization of the glutamate transporter protein GLAST in rat retina using an antibody raised and affinity purified against a peptide corresponding to amino acid residues 522–541. The strongest immunocytochemical labelling was observed in the outer plexiform layer, ganglion cell layer, and optic disc. GLAST was found in Muller cell processes in all retinal ¨ layers, notably ensheathing the photoreceptor terminals in the outer plexiform layer, and in astrocytes close to vessels in the inner retina and optic disc. No labelling was observed in neurons. The electrophoretic mobility of GLAST in retina was similar to that in cerebellum. In conclusion, the findings are in agreement with those reported by Derouiche and Rauen w7x, except that we did not detect any GLAST in the retinal pigment epithelium. Keywords: Glutamate uptake; GLAST; Retina; Pigment epithelium; Glia; Muller cell; Astrocyte; Immunocytochemistry ¨
1. Introduction Glutamate is regarded as a neurotransmitter of both photoreceptor cells, bipolar cells and ganglion cells of the retina w25x. Powerful uptake mechanisms capable of quickly removing extracellular glutamate are therefore required. This is accomplished by glutamate transporter proteins located in the plasma membranes of both neurons and glial cells in the brain Žfor review see w6x. as well as in the retina w2,9x. Several glutamate transporters have been cloned: GLAST w32x, GLT-1 w27x, EAAC1 w17x, EAAT4 w11x. A series of immunocytochemical studies have appeared describing the localization of the GLT-type glutamate transporter in retinal neurons w1,10,12,28x. GLAST has been described in retinal Muller cells, astrocytes and ¨ pigment epithelial cells in albino rats w7x, using an antiserum raised against GLAST protein purified from rat brain. The latter report is in agreement with an in situ hybridization study w26x where GLAST mRNA was de-
tected in Muller cells and astrocytes. With regard to the ¨ pigment epithelium, however, the two reports disagree. The pigment epithelium was found immunoreactive to the GLAST antiserum w7x, while GLAST mRNA was not detected in the epithelium w26x. The possibility exists that the antiserum used w7x could recognize other proteins in addition to GLAST. Data with more defined antibodies are therefore required. Here we demonstrate, light and electron microscopically, the cellular localization of the GLAST protein in retina of albino as well as pigmented rats, using antibodies w21x specifically recognizing a peptide sequence in the C-terminal part of the protein. Since abnormalities have been noted in the retina of albino rats w5x, this study is based mainly on pigmented rats. We confirm the presence of GLAST in Muller cells and astrocytes, and its absence ¨ from neuronal elements. However, the pigment epithelium was immunonegative for GLAST.
2. Materials and methods 2.1. Materials ) Corresponding author. Department of Anatomy, Institute of Basic Medical Sciences, University of Oslo, POB 1105 Blindern, N-0317 Oslo, Norway. Fax: Ž47. 2285 1278; E-mail: [email protected]
Alkaline phosphatase-conjugated anti-rabbit IgG was from Sigma ŽSt. Louis, MO, USA.. Nitrocellulose sheets
0006-8993r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 0 6 - 8 9 9 3 Ž 9 6 . 0 1 0 2 2 - 0
130
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
Ž0.22 m m pores, 100% nitrocellulose. and electrophoresis equipment were from Hoefer Scientific Instruments ŽSan Francisco, CA, USA.. Glutaraldehyde, EM grade, was from TAAB ŽReading, UK. and Durcupan ACM from Fluka ŽBuchs, Switzerland.. Molecular mass markers were the ‘Rainbow markers’ from Amersham ŽBuckinghamshire, UK.. Other reagents were analytical grade or better. 2.2. Antibodies Antibodies to GLAST were prepared as described w21x. Briefly, synthetic peptides corresponding to residues 1–25 ŽMTKSNGEEPRMGSRMERFQQGVRKRC. or 522–541 ŽPYQLIAQDNEPEKPVADSET. of the glutamate transporter GLAST w32x were coupled to hemocyanin with maleimido benzoyl-N-hydroxysuccinimide or glutaraldehyde, respectively, and used to immunize rabbits. The ensuing antisera Žanti-A1 and anti-A522, respectively. were affinity purified against the peptides immobilized by N-hydroxysuccinimide on agarose. The peptides are referred to as A1 and A522, respectively, and the corresponding antibodies as anti-A1 and anti-A522. This report is based Fig. 2. Low magnification light microscopic localization of GLAST in rat retina. GLAST ŽA. 0.1 m grml anti-A522, and preimmune control ŽB. 1 m grml preimmune IgG. Due to narrowing of the iris diaphragm to produce contrast to make the control section visible, photoreceptors can be seen in B, but not in A. Note in A the strong labelling in the outer plexiform and ganglion cell layers, the radiating fibres, and the difference in staining in the inner and outer part of the inner plexiform layer. S, photoreceptor inner and outer segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer. Scale bar, 30 m m.
mainly on the anti-A522 antibodies. The anti-A1 antibodies were only used when stated to verify that the immunoreactivity observed with the anti-A522 was due to GLAST.
Fig. 1. Immunoblot of crude SDS extracts of rat retina ŽR., hippocampal formation ŽH. and cerebellum ŽC. subjected to SDS-PAGE Ž5, 2 and 2 m g proteinrlane, respectively. and reacted with GLAST antibody antiA522 followed by alkaline phosphatase conjugated secondary antibodies. Note the differences in electrophoretic mobility and immunoreactivity of the labelled bands in the three lanes. Molecular mass markers ŽkDa. are indicated.
Fig. 3. GLAST in the optic disc in rat retina. Note in A the strong staining in the inner part of the optic disc where branching vessels Žarrow. are situated and where the inner limiting membrane contains astrocytes. Note in B the labelling of the astroglial columns Ža. and their thin processes Žarrowhead. within the nerve fibre fascicles. Sections from albino rat. Scale bars, 25 m m.
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
2.3. Protein measurement Protein was determined in homogenates dissolved in 1% sodium dodecyl sulfate ŽSDS. as described w24x using bovine serum albumin as standard. SDS was added to give equal concentrations in samples and standards.
131
2.4. Preparation of tissue extracts, electrophoresis, blotting and immunostaining of blots This was done as described w20,21,34x. The gels consisted of 10% acrylamide and were run slowly overnight at about 58C. The samples were incubated Ž2 min, 1008C.
Fig. 4. High magnification interference contrast light microscopic localization of GLAST in rat retina. The picture was taken from the same section as Fig. 2A. Note the slender stained fibres along vessel wall Žarrow. representing astrocytic processes and the small dot shaped stained structures in transversely cut nerve fibre bundles Žarrowheads. representing Muller cell processes. Scale bar, 15 m m. ¨
132
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
with 4% Žvrv. mercaptoethanol prior to application. The molecular mass markers were run several lanes away from the samples because mercaptoethanol changes the marker mobility w22x. 2.5. Immunocytochemistry Perfusion of rats ŽPVG or Wistar strain, Møllegaard Hansen, Denmark., sectioning and processing of sections were done as described w21x. Retinas or eyecups were embedded in 15–20% Žwrv. gelatine in 0.1 M sodium phosphate buffer pH 7.4 ŽNaPi. and the blocks hardened in the tissue fixative before vibratome sectioning. For labelling of cryostat sections rats were decapitated, anterior poles of the eyes removed in 0.1 M NaPi, posterior eyecups fixed Ž3 h, at room temperature. in 4% freshly depolymerized paraformaldehyde in 0.1 M NaPi, rinsed in 0.1 M NaPi, immersed Ž1–2 days, at 48C. in 30% sucrose in 0.1 M NaPi, and 10–15 m m cryostat sections air dried on slides and stored Ž1–3 days, at y208C. before immuno-
Fig. 6. GLAST in nerve fibre bundle in the ganglion cell layer. Note glial profiles containing reaction product concentrated along the inside of the cell membrane Ž). between unlabelled nerve fibres. Scale bar, 400 nm.
labelling. The anti-A522 antibodies were used at 0.1–0.3 m grml. For negative controls, primary antibodies were omitted, primary antibodies were substituted by preimmune IgG, or primary antibodies were preincubated with the A522 peptide. For light microscopy the sections were observed using a Leitz DM R microscope. Pictures were taken with interference contrast optics when stated. Figures show retinas from pigmented animals unless albino is noted.
3. Results 3.1. Specificity of the antibody and relatiÕe expressions of GLAST in brain and retina
Fig. 5. Interference contrast light micrographs showing lack of GLAST labelling in retinal pigment epithelium of albino rat. The sections are incubated with Triton X-100 to secure penetration of antibodies. A and B are incubated in the same solution, with anti-A522. C is a negative control where the primary antibody has been omitted. Note the unlabelled hexagonal cells in A, a tangential section through the pigment epithelium. Note in B the contrast between the unlabelled pigment epithelium Žarrow. and the heavy staining in the OPL and ONL. Note also that there is no difference between the pigment epithelium in B and C. Without the interference contrast effect the pigment epithelium is invisible in A, B and C. Scale bar, 30 m m.
Even though the specificity of the anti-GLAST antibodies is firmly established in the rat brain w4,21,22x, the possibility existed that retina expresses crossreacting polypeptides. Therefore, the antibody specificity was tested on immunoblots ŽFig. 1. of electrophoretically separated SDS-solubilized whole tissue from rat retina Žpars nervosa., hippocampus and cerebellum. In the retinal extracts ŽLane R, 5 m g protein., the anti-A522 antibodies recognized a broad, somewhat heterogenous band similar to those observed in the cerebellum ŽLane C, 2 m g protein. and hippocampus ŽLane H, 2 m g protein.. Further, in tissue sections, no labelling was observed when they were incubated with preimmune IgG ŽFig. 2B. or with antibodies
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
133
Fig. 7. Perivascular GLAST in the ganglion cell layer. A: astrocyte with evenly pale nucleus Žn. localized next to vessel contains reaction product Žarrows. in the cytoplasm close to the cell membrane. B: unlabelled smooth muscle cells Žs. of ganglion cell layer vessels are surrounded by immunopositive glial processes Ž).. Note that penetration of immunoreactants is secured by mechanical opening of cells in A and B. ŽPatches of subsarcolemmal darkness of smooth muscle cells are not due to reaction product.. Arrowheads, basal laminae of vessel walls. Scale bar, 2 m m.
preabsorbed with A522-peptide Ždata not shown.. The apparent molecular mass of the immunoreactive proteins in retina Ž67–68 kDa. was similar to that in cerebellum Ž67 kDa. but higher than in hippocampus Ž63 kDa.. By comparing the staining intensity and the amounts of protein in each lane, it is concluded that the concentration of GLAST protein in retina is less than half of that in hippocampus. However, the intensity of the immunocytochemical labelling of individual cells in the retina resembled that in the brain w21x, indicating that the concentration of GLAST in those membranes that are immunopositive is similar to that in the brain. 3.2. Immunocytochemical localization 3.2.1. Light microscopy The strongest labelling ŽFig. 2A and Fig. 4. was observed in the ganglion cell layer ŽGCL., the outer plexi-
form layer ŽOPL., and ŽFig. 3A. the inner part of the optic disc. The second strongest labelling ŽFig. 2A and Fig. 4. was found in the inner plexiform layer ŽIPL., the inner nuclear layer ŽINL., and the outer nuclear layer ŽONL.. The outer 2r5 of the IPL Žsublamina A. was weaker than the rest of this layer, and a slightly stronger labelled zone was observed at the border between the INL and IPL. In the part of the retina immediately adjacent to the vitreous body, the staining was present in structures with ramifications penetrating into the retina between bundles of nerve fibres ŽFig. 4.. Within transversely cut nerve fibre bundles small stained dots were present. In more obliquely cut nerve fibre bundles, the stained structures were shaped like longer thin processes separating the unstained nerve fibres. Bundles of nerve fibres passing into the optic disc were separated by stained structures of the same appearance. In the prelaminar portion of the optic disc there was labelling of cells in the astroglial cell columns and thin
Fig. 8. GLAST in the inner plexiform layer. A: labelled glial profiles separate and encircle unlabelled neural processes. B: labelled Ž). glial process contacting unlabelled bipolar terminal Žt.. Scale bars, 400 nm.
134
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
Fig. 9. GLAST in the inner nuclear layer. Labelled glial processes separate cell bodies which appear unlabelled. One of the neuronal cell bodies is opened to the vibratome section surface Žw.. The glial process near the section surface Žarrow. is strongly stained. Glial processes far from the section surface Žarrowhead. show weaker labelling or no labelling. Scale bar, 500 nm.
processes within the nerve fibre fascicles ŽFig. 3B.. In the GCL thin stained processes appeared along the blood vessels ŽFig. 4.. This produced a dark demarcation of the vessels similar to the astrocytic labelling described for S-100 protein w19x. In the IPL, thick radiating fibres in addition to numerous small, thin, laminar processes were stained. There were clearly fewer of these labelled small processes in the sublamina A than in the rest of the IPL. In the INL, a network of thin labelled structures separated apparently unlabelled perikarya. In the OPL, strong labelling was observed in thin processes ensheathing rod spherules. The ONL displayed a similar picture as the INL. Labelling was observed all the way through to the outer limiting membrane. The inner and outer segments of the photoreceptors were unlabelled ŽFig. 4..
No major differences between the labelling of retinas from albino and pigmented rats were observed. However, in some of the albino animals the overall staining intensity in the ONL appeared stronger. The pigment epithelium was studied in vibratome sections from albino rats ŽFig. 5. and cryostat sections from albino and pigmented rats Žnot shown.. At antibody concentrations giving strong labelling of the neuroretina, there was no labelling in adjacent pigment epithelial cells. The absence of pigment epithelial labelling in the presence of neuroretinal labelling was observed with antibodies to the C-terminal of GLAST Žanti-A522; Fig. 5. as well as with antibodies to the N-terminal Žanti-A1; not shown.. These experiments were performed in the presence of Triton X-100 to secure penetration of immunoreagents.
Fig. 10. GLAST in the outer plexiform ŽA. and outer nuclear ŽB. layer. A: photoreceptor terminals Žt. appear unlabelled. Labelling Ž). can be observed along the cell membranes and in the adjacent cytoplasm of the thin glial processes interposed between the terminals. B: glial profiles in between photoreceptor cell perikarya and processes are labelled. No labelling appears in the photoreceptors. Scale bars, 300 nm.
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
3.2.2. Electron microscopy Immediately next to the vitreous body staining was present on the plasmalemma and adjacent cytoplasm of structures bordering on the inner limiting membrane Ždata not shown.. In nerve fibre bundles, the nerve fibres appeared unlabelled, but the glial processes running in between them clearly contained reaction product ŽFig. 6.. Cell bodies of ganglion cells were unlabelled, but occasionally labelled glial processes could be seen next to the cell bodies Ždata not shown.. Cells situated adjacent to vessels in the GCL with evenly pale nuclei and filaments in their perikarya and processes Ži.e. astrocytes. contained reaction product where exposed to the surface of the vibratome section ŽFig. 7A.. Glial processes contacting the basal lamina of endothelial cells and smooth muscle cells were labelled ŽFig. 7B.. At higher magnification some of these labelled processes could be seen to contain filaments typical of astrocytes. In the IPL, glial profiles surrounding synapses and axons were labelled ŽFig. 8A.. None of the observed nerve terminals, including those containing synaptic ribbons Žbipolar terminals., appeared stained ŽFig. 8B.. In both nuclear layers, stained glial processes could be seen to separate unstained neuronal cell bodies and processes ŽFigs. 9 and 10B.. In the OPL the photoreceptor terminals appeared unlabelled. This was true even when the terminals were cut open by the vibratome knife, securing access of the anti-A522 antibody to its epitopes, which are on the inside of cell membranes w21x. Labelling was observed along the cell membranes and in the adjacent cytoplasm of the thin glial processes interposed between the photoreceptor terminals ŽFig. 10A.. Glial processes far from the surface of the vibratome sections ŽFig. 9., and glial processes in the control sections were unlabelled.
4. Discussion The labelled structures bordering on the inner limiting membrane are interpreted as Muller cell endfeet, although ¨ astrocytes proper occasionally border directly on the inner limiting membrane. Several investigators have noted that rat retinal astrocytes are not found between ganglion cell axons w13,23,31x. At this point, the retina of the rat, and mouse w16x, differ from that of several other species, as nerve fibres show a strong tendency to attract astrocytes in the cat w3,15,31x, rabbit w31x and monkey retina w3,8x. Thus, the labelled glial processes in nerve fibre bundles that we observed in the rat retina most likely represent Muller cell ¨ processes. Stone and Dreher w31x have noted that astrocytic processes do not invest the somata of ganglion cells in the retina of rat and cat. This has been confirmed for the cat w15x, and shown to be true also for the monkey retina w8x. Thus, the labelled glial processes contacting ganglion cells represent Muller cell processes. All the glial labelling ¨ described in the outer part of the retina represents Muller ¨ cell labelling, notably the strong labelling of the thin glial
135
processes interposed between the photoreceptor terminals. It is concluded that the anti-A522 GLAST antibodies label Muller cell processes in all retinal layers. ¨ The presence of GLAST in astrocytes proper was confirmed both at the light and electron microscopic levels: as the optic disc contains no Muller cells, all the glial la¨ belling observed in the disc represents astrocytes. This labelling is in agreement with the observation that GLAST mRNA is found in about 90% of cells in the optic nerve head, where more than 90% of cells express glial fibrillary acidic protein mRNA w26x. The glial investments of blood vessels in the inner retina of the rat are formed by astrocytic processes w14,19,31x. Processes containing numerous filaments, often found along the blood vessel basal laminae, were seen to emerge from cells with evenly pale nuclei situated close to blood vessels in the innermost part of the retina. These cell bodies and processes often could be seen to be labelled, and were interpreted as GLAST positive astrocytes. Sarantis and Attwell w30x have shown that glutamate evokes an inward membrane current in Muller cells from ¨ the rabbit retina. This glutamate uptake current is dependent on extracellular sodium and intracellular potassium. This is in agreement with the ion cotransport characteristics of brain glutamate transporters, including the GLAST protein w18x. The findings of Otori et al. w26x indicate that Muller ¨ cells and astrocytes in the rat retina express GLAST mRNA. Since GLT does not appear to be present in retinal glial cells w28x, GLAST most probably is responsible for glutamate transport in retinal glia. This is in contrast to the brain, where GLT and GLAST coexist in astrocytes w21x. Labelling of the pigment epithelium was neither observed with antibodies to the C-terminal part of GLAST Žresidues 522–541. nor with antibodies to the N-terminal Žresidues 1–25.. The lack of labelling of the pigment epithelium is in agreement with the in situ hybridization study by Otori et al. w26x where no GLAST mRNA was detected at this site. However, the presence of GLAST in the retinal pigment epithelium has been reported in another immunocytochemical study w7x. This study w7x is based on an antiserum raised against the GLAST protein purified from rat brain. Except for the labelling of the pigment epithelium, the results of these authors are in agreement with ours. It is legitimate to speculate why this antiserum labels the pigment epithelium. One possibility is that the GLAST protein expressed in the pigment epithelium is different from the GLAST protein expressed in the neuroretina and the brain, and that the antiserum w7x recognizes epitopes common to the two versions. Another possibility is that the antiserum used in the cited study w7x could contain some antibodies to a retinal epithelial protein different from GLAST. The in situ hybridization signal for GLAST mRNA outlining the ventricles of the rat brain reported by our group w33x should not, as suggested w7x, be taken as support for the presence of GLAST in the pigment epithelium.
136
K.P. Lehre et al.r Brain Research 744 (1997) 129–137
This signal has not been interpreted as ependymal w33x, but has been explained by the presence of a subependymal glial plexus immunoreactive to GLAST w21x. Observations using antibodies to synthetic peptides corresponding to the EAAC glutamate transporter protein w17x suggest that this protein is expressed in the rat pigment epithelium ŽK. Ullensvang and K.P. Lehre, unpubl. obs... This may explain the glutamate uptake reported in rat pigment epithelium w29x. In conclusion, GLAST is expressed in neuroretina essentially as described previously w7x. However, the glutamate uptake observed in pigment epithelium is not, as suggested w7x, due to the GLAST protein w32x found elsewhere in the CNS, but to another glutamate transporter.
w12x
w13x
w14x
w15x
w16x
w17x
Acknowledgements w18x
We thank Jon Storm-Mathisen for discussions and critical reading of the manuscript and Gunnar F. Lothe for technical assistance. This work was supported by the Norwegian Research Council, Anders Jahres Fond, Nansenfondet, Schreiners fond and EU BIOMED2 Žcontract no. BMH4-CT95-0571..
w19x
w20x w21x
References w1x Brandstatter, J.H., Greferath, U., Euler, T. and Wassle, H., Co¨ ¨ stratification of GABA A receptors with the directionally selective circuitry of the rat retina, Vis. Neurosci., 12 Ž1995. 345–358. w2x Brew, H. and Attwell, D., Electrogenic glutamate uptake is a major current carrier in the membrane of axolotl retinal cells, Nature, 327 Ž1987. 707–709. w3x Bussow, H., The astrocytes in the retina and optic nerve head of ¨ mammals: a special glia for the ganglion cell axons, Cell Tissue Res., 206 Ž1980. 367–378. w4x Chaudhry, F.A., Lehre, K.P., van Lookeren Campagne, M., Ottersen, O.P., Danbolt, N.C. and Storm-Mathisen, J., Glutamate transporters in glial plasma membranes: highly differentiated localizations revealed by quantitative ultrastructural immunocytochemistry, Neuron, 15 Ž1995. 711–720. w5x Creel, D., Inappropriate use of albino animals as models in research, Pharmac. Biochem. BehaÕ., 12 Ž1980. 969–977. w6x Danbolt, N.C., The high affinity uptake system for excitatory amino acid neurotransmitters in the brain, Prog. Neurobiol., 44 Ž1994. 377–396. w7x Derouiche, A. and Rauen, T., Coincidence of L-glutamaterL-aspartate transporter ŽGLAST. and glutamine synthetase ŽGS. immunoreactions in retinal glia: evidence for coupling of GLAST and GS in transmitter clearance, J. Neurosci. Res., 42 Ž1995. 131–143. w8x Distler, C., Weigel, H. and Hoffmann, K.-P., Glia cells of the monkey retina. I. Astrocytes, J. Comp. Neurol., 333 Ž1993. 134–147. w9x Eliasof, S. and Werblin, F., Characterization of the glutamate transporter in retinal cones of the tiger salamander, J. Neurosci., 13 Ž1993. 402–411. w10x Euler, T. and Wassle, H., Immunocytochemical identification of ¨ cone bipolar cells in the rat retina, J. Comp. Neurol., 361 Ž1995. 461–478. w11x Fairman, W.A., Vandenberg, R.J., Arriza, J.L., Kavanaugh, M.P.
w22x
w23x
w24x
w25x w26x
w27x
w28x
w29x
w30x
w31x
and Amara, S.G., An excitatory amino-acid transporter with properties of a ligand-gated chloride channel, Nature, 375 Ž1995. 599–603. Grunert, U., Martin, P.R. and Wassle, H., Immunocytochemical ¨ ¨ analysis of bipolar cells in the macaque monkey retina, J. Comp. Neurol., 348 Ž1994. 607–627. Hildebrand, C. and Waxman, G., Regional node-like membrane specializations in non-myelinated axons of rat retinal nerve fiber layer, Brain Res., 258 Ž1983. 23–32. Holash, J.A. and Stewart, P.A., The relationship of astrocyte-like cells to the vessels that contribute to the blood-ocular barriers, Brain Res., 629 Ž1993. 218–224. Hollander, H., Makarov, F., Dreher, Z., van Driel, D., Chan-Ling, T. ¨ and Stone, J., Structure of the macroglia of the retina: sharing and division of labour between astrocytes and Muller cells, J. Comp. ¨ Neurol., 313 Ž1991. 587–603. Huxlin, K.R., Sefton, A.J. and Furby, J.H., The origin and development of retinal astrocytes in the mouse, J. Neurocytol., 21 Ž1992. 530–544. Kanai, Y. and Hediger, M.A., Primary structure and functional characterization of a high-affinity glutamate transporter, Nature, 360 Ž1992. 467–471. Klockner, U., Storck, T., Conradt, M. and Stoffel, W., Electrogenic ¨ L-glutamate uptake in Xenopus-laevis oocytes expressing a cloned rat brain L-glutamate-L-aspartate transporter ŽGLAST-1., J. Biol. Chem., 268 Ž1993. 14594–14596. Kondo, H., Takahashi, H. and Takahashi, Y., Immunohistochemical study of S-100 protein in the postnatal development of Muller cells ¨ and astrocytes in the rat retina, Cell Tissue Res., 238 Ž1984. 503–508. Laemmli, U.K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature, 227 Ž1970. 680–685. Lehre, K.P., Levy, L.M., Storm-Mathisen, J., Ottersen, O.P. and Danbolt, N.C., Differential expression of two glial glutamate transporters in the rat brain: quantitative and immunocytochemical observations, J. Neurosci., 15 Ž1995. 1835–1853. Levy, L.M., Lehre, K.P., Walaas, S.I., Storm-Mathisen, J. and Danbolt, N.C., Down-regulation of glial glutamate transporters after glutamatergic denervation in the rat brain, Eur. J. Neurosci., 7 Ž1995. 2036–2041. Ling, T., Mitrofanis, J. and Stone, J., Origin of retinal astrocytes in the rat: evidence of migration from the optic nerve, J. Comp. Neurol., 286 Ž1989. 345–352. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 Ž1951. 265–275. Massey, S.C. and Redburn, D.A., Transmitter circuits in the vertebrate retina, Prog. Neurobiol., 28 Ž1987. 55–96. Otori, Y., Shimada, S., Tanaka, K., Ishimoto, I., Tano, Y. and Tohyama, M., Marked increase in glutamate-aspartate transporter ŽGLASTrGluT-1. mRNA following transient retinal ischemia, Mol. Brain Res., 27 Ž1994. 310–314. Pines, G., Danbolt, N.C., Bjøras, ˚ M., Zhang, Y., Bendahan, A., Eide, L., Koepsell, H., Seeberg, E., Storm-Mathisen, J. and Kanner, B.I., Cloning and expression of a rat brain L-glutamate transporter, Nature, 360 Ž1992. 464–467. Rauen, T. and Kanner, B.I., Localizatiion of the glutamate transporter GLT-1 in rat and macaque monkey retinae, Neurosci. Lett., 169 Ž1994. 137–140. Salceda, R. and Saldana, M.R., Glutamate and taurine uptake by retinal pigment epithelium during rat development, Comp. Biochem. Physiol., 104C Ž1993. 311–316. Sarantis, M. and Attwell, D., Glutamate uptake in mammalian retinal glia is voltage- and potassium-dependent, Brain Res., 516 Ž1990. 322–325. Stone, J. and Dreher, Z., Relationship between astrocytes, ganglion cells and vasculature of the retina, J. Comp. Neurol., 255 Ž1987. 35–49.
K.P. Lehre et al.r Brain Research 744 (1997) 129–137 w32x Storck, T., Schulte, S., Hofmann, K. and Stoffel, W., Structure, expression, and functional analysis of a Na q-dependent glutamateraspartate transporter from rat brain, Proc. Natl. Acad. Sci. USA, 89 Ž1992. 10955–10959. w33x Torp, R., Danbolt, N.C., Babaie, E., Bjøras, ˚ M., Seeberg, E., Storm-Mathisen, J. and Ottersen, O.P., Differential expression of
137
two glial glutamate transporters in the rat brain: an in situ hybridization study, Eur. J. Neurosci., 6 Ž1994. 936–942. w34x Towbin, H., Staehelin, T. and Gordon, J., Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications, Proc. Natl. Acad. Sci. USA, 76 Ž1979. 4350–4354.