THE JOURNAL OF COMPARATIVE NEUROLOGY 423:121–131 (2000)
Localization of Glutamate and Glutamate Transporters in the Sensory Neurons of Aplysia JONATHAN LEVENSON,1 DAVID M. SHERRY,2 LAURENCE DRYER,1 JEANNIE CHIN,3 JOHN H. BYRNE,3 AND ARNOLD ESKIN1* 1 Department of Biology and Biochemistry, University of Houston, Houston, Texas 77204-5513 2 College of Optometry, University of Houston, Houston, Texas 77204 3 Department of Neurobiology and Anatomy, University of Texas Medical School at Houston, Houston, Texas 77225
ABSTRACT The sensorimotor synapse of Aplysia has been used extensively to study the cellular and molecular basis for learning and memory. Recent physiologic studies suggest that glutamate may be the excitatory neurotransmitter used by the sensory neurons (Dale and Kandel [1993] Proc Natl Acad Sci USA. 90:7163–7167; Armitage and Siegelbaum [1998] J Neurosci. 18: 8770 – 8779). We further investigated the hypothesis that glutamate is the excitatory neurotransmitter at this synapse. The somata of sensory neurons in the pleural ganglia showed strong glutamate immunoreactivity. Very intense glutamate immunoreactivity was present in fibers within the neuropil and pleural-pedal connective. Localization of amino acids metabolically related to glutamate was also investigated. Moderate aspartate and glutamine immunoreactivity was present in somata of sensory neurons, but only weak labeling for aspartate and glutamine was present in the neuropil or pleural-pedal connective. In cultured sensory neurons, glutamate immunoreactivity was strong in the somata and processes and was very intense in varicosities; consistent with localization of glutamate in sensory neurons in the intact pleural-pedal ganglion. Cultured sensory neurons showed only weak labeling for aspartate and glutamine. Little or no ␥-aminobutyric acid or glycine immunoreactivity was observed in the pleural-pedal ganglia or in cultured sensory neurons. To further test the hypothesis that the sensory neurons use glutamate as a transmitter, in situ hybridization was performed by using a partial cDNA clone of a putative Aplysia high-affinity glutamate transporter. The sensory neurons, as well as a subset of glia, expressed this mRNA. Known glutamatergic motor neurons B3 and B6 of the buccal ganglion also appeared to express this mRNA. These results, in addition to previous physiological studies (Dale and Kandel [1993] Proc Natl Acad Sci USA. 90:7163–7167; Trudeau and Castellucci [1993] J Neurophysiol. 70:1221–1230; Armitage and Siegelbaum [1998] J Neurosci. 18:8770 – 8779)) establish glutamate as an excitatory neurotransmitter of the sensorimotor synapse. J. Comp. Neurol. 423: 121–131, 2000. © 2000 Wiley-Liss, Inc. Indexing terms: mollusc; aspartate; glutamine; glycine; GABA
The synapse between the ventral-caudal (VC) cluster of sensory neurons in the pleural ganglion and their associated motor neurons in the pedal ganglion of Aplysia californica has been used extensively to study the cellular and molecular basis of learning and memory (reviewed in Byrne et al., 1991; Bailey et al., 1996; Abel et al., 1998). Although a great deal is known about regulatory mechanisms affecting transmission at this synapse, the identity of the excitatory neurotransmitter of this synapse remains equivocal. Consequently, it is important to identify the © 2000 WILEY-LISS, INC.
Grant sponsor: Basic Research And Initiatives In Neuroscience (BRAIN) program at the University of Houston; Grant sponsor: National Institutes of Health; Grant numbers: NS28462, NS32748, and NS19895. *Correspondence to: Arnold Eskin, Department of Biology and Biochemistry, University of Houston, 4800 Calhoun Rd, SRII, Houston, TX 772045513. E-mail:
[email protected] Received 9 December 1999; Revised 6 March 2000; Accepted 14 March 2000
122
J. LEVENSON ET AL.
neurotransmitter to further study regulation of neurotransmission at this synapse. The results of previous studies suggested that the excitatory neurotransmitter of the sensorimotor synapse of the pleural-pedal ganglia may be glutamate (Dale and Kandel, 1993). In cocultures of sensory and motor neurons, exogenous glutamate mimicked and a glutamate receptor blocker, DNQX, blocked responses to the endogenous transmitter (Dale and Kandel, 1993). In contrast, Trudeau and Castellucci (1993) observed that glutamate did not elicit an excitatory response in motor neurons in pedal ganglion. More recent studies have shown that glutamate mimics the effect of the endogenous transmitter and DNQX blocks the postsynaptic effects of sensory neuron stimulation (Armitage and Siegelbaum, 1998). To further study the potential role of glutamate at the sensorimotor synapse, we investigated whether the sensory neurons and their terminals contain high levels of endogenous glutamate. We also investigated whether the sensory neurons express a mRNA coding for a highaffinity glutamate transporter, which would indicate that sensory neurons possess a mechanism to terminate glutamatergic transmission. We demonstrate that the sensory neurons and their processes possess high endogenous levels of glutamate and express a gene that likely encodes a high-affinity glutamate transporter, supporting a role for glutamate as the excitatory neurotransmitter of the sensory neurons in the pleural ganglion.
MATERIALS AND METHODS Animals Aplysia californica (100 –150 g) were obtained from Marinus, Inc. (Long Beach, CA) or Alacrity Marine Biological Services (Redondo Beach, CA). Animals were maintained in Instant Ocean (Aquarium Systems, Mentor, OH) at 15°C under light:dark 12:12 hours and were fed romaine lettuce or dried seaweed every 2–3 days. Animals were allowed to adapt to laboratory conditions for 3 days before use and were always used within 2 weeks of arrival.
Immunocytochemistry and antisera Animals were anesthetized with isotonic MgCl2, and ganglia were removed and fixed immediately in a solution of 4% formaldehyde, 1% glutaraldehyde, 25% sucrose, 8.5 g/L NaCl, and 10 g/L sodium bisulfite in 0.1 M PBS, pH 7.4, for 3 hours at room temperature (RT). No amino acid immunofluorescence was detectable if tissue was not fixed immediately upon removal from the animal. After 30 minutes of fixation, the sheath was trimmed and carefully punctured several times with a scalpel, to facilitate penetration of the fixative, and the ganglia were allowed to fix for an additional 2.5 hours. After fixation, the tissue was rinsed in PBS at 4°C, dehydrated in ethanol including an overnight incubation in 100% ethanol, and further dehydrated in propylene oxide at RT. The tissue was infiltrated for 2 days at RT with LX-112/Araldite resin (Ladd Industries, Burlington, VT). For some specimens, infiltration during the last 24 hours was under vacuum. After infiltration, tissue was transferred to fresh resin and polymerized at 55°C overnight. Semithin (1 m) serial sections of pleural-pedal ganglia were cut and dried on gelatin-coated slides at 55°C overnight. All primary antisera were raised in rabbit against glutaraldehyde conjugates of either glutamate (GLU), aspar-
tate (ASP), glutamine (GLN), ␥-aminobutyric acid (GABA), or glycine (GLY) (a generous gift from Signature Immunologics, Salt Lake City, UT). Each primary antiserum was highly specific for the amino acid conjugate it was raised against as determined by preadsorption assays with the appropriate conjugate (data not shown) and previous reports (Kalloniatis et al., 1996; Marc et al., 1998a,b). All antisera were used at a dilution of 1:50 to visualize amino acids in semithin sections. Each primary antiserum was calibrated by the manufacturer to have equal affinity for the appropriate target hapten. Therefore, equivalent concentrations of the target haptens will produce equivalent immunolabeling intensity. These antisera have been used previously to make quantitative measurements of amino acids by using techniques similar to those used in this study (Marc et al., 1995, 1998a,b; Kalloniatis et al., 1996). Substitution of normal rabbit serum for the primary antiserum eliminated labeling, indicating that the methods yielded specific labeling. Postembedding immunofluorescence was performed as previously described (Sherry and Ulshafer, 1992). Sections were etched in sodium ethanolate for 45 minutes at RT, and then rinsed in ethanol and ddH2O. Autofluorescence of sections was reduced by incubation in 1% NaBH4 for 1 minute at RT and then rinsing in ddH2O followed by PBS. Sections were blocked in 2% normal goat serum (Jackson Immunoresearch Laboratories, Westgrove, PA) and 0.1% Triton X-100 in 0.1 M PBS (pH 7.4). The blocking solution was also used to dilute primary and secondary antisera. Primary antisera were applied to sections overnight at 4°C. Primary antisera were removed, slides were rinsed in PBS, blocked and incubated in goat anti-rabbit IgG conjugated to either fluorescein isothiocyanate or Cy3 (Jackson Immunoresearch Laboratories) for 45 minutes at RT. Slides were mounted in a fade-retardant mounting medium and examined (DABCO, Sigma, St. Louis, MO; Johnson et al., 1982). Cultured sensory neurons were fluorescently immunolabeled as previously described (Sherry et al., 1996). Sensory neurons were cultured as previously described (Schacher and Proshansky, 1983; Chin et al., 1999) at a density of three to six sensory neurons per culture dish and were allowed to grow for 3 to 5 days before use. Cultures were rinsed with a 7.5% sucrose, 0.1 M PBS (pH 7.4) solution, then fixed for 45 minutes at RT in 4% formaldehyde, 1% glutaraldehyde in 0.1 M PBS supplemented with 3 mM CaCl2. Cultures were rinsed with PBS and then ddH2O, and autofluorescence was reduced with 0.5% NaBH4. Cultures were rinsed again and blocked with 2% normal goat serum, 0.1% Triton X-100 in 0.1 M PBS for 1 hour at RT. Primary antiserum (1:200) was applied overnight at 4°C. Primary antiserum was removed, and cultures were rinsed in PBS, blocked, and incubated in Cy3conjugated goat anti-rabbit antiserum (1:800). Cultures were mounted in DABCO and examined. The experimenter performed the immunolabeling and analysis of cultured sensory neurons in a manner blind to the primary antisera applied.
Cloning of glutamate transporter Total RNA was isolated from the central nervous systems of several adult Aplysia. Ganglia were dissected, immediately frozen in liquid N2 and stored at ⫺80°C before RNA extraction by using an RNA isolation kit (Ultraspec, Biotecx, Houston, TX). Poly-(A) enriched RNA was
GLUTAMATE AND GLUTAMATE TRANSPORTERS IN APLYSIA isolated by passage of total RNA over an oligo-dT-cellulose column (Ambion) twice. Poly-(A) RNA was reverse transcribed to cDNA by using an anchored oligo-dT primer. Polymerase chain reaction (PCR) was performed on the cDNA to isolate a fragment of the glutamate transporter gene. PCR was performed under standard conditions with a [Mg⫹⫹] of 1 mM and an annealing temperature of 45°C. Degenerate primers were constructed to match peptide sequences of extremely high conservation across all known high-affinity glutamate transporters. The primers corresponded to the peptide sequences: GATINMDG (GG[T/C/A/G] GC[T/C/A/G] AC[T/C/A/G] AT[T/C/A] AA[T/ C] ATG GA[T/C] GG) and AVDWLLD ([A/G]TC [T/C/A/ G]A[A/G] [T/C/A/G]A[A/G] CCA [A/G]TC [T/C/A/G]AC [T/C/A/G]GC). The resulting PCR product was cloned by using a commercially available kit (PCR-Script, Stratagene, La Jolla, CA) into the PCR-Script vector. The clone was sequenced by using an automated sequencing system (ABI Prism, Perkin Elmer, Foster City, CA) in each direction by using existing T3 and T7 sites. The BLAST program was used to search the databases for sequences similar to the clones obtained. Multiple alignments were performed by using the computer program ClustalW.
In situ hybridization The procedures used were based on those used by Wilkinson et al. (1987). Ganglia were fixed overnight at 4°C in 4% formaldehyde in 0.1 M PBS. After fixation, ganglia were rinsed in PBS, dehydrated through a series of alcohols, and embedded in paraffin. 35S-labelled antisense and sense (control) riboprobes generated with an RNA transcription kit (Stratagene) were hybridized to 10 m sections as described previously (Ressler et al., 1993). Bound riboprobe was visualized by autoradiography. Nuclei were counterstained with Hoechst (Sigma).
Image processing Photographs of sections were transferred to photo CD by a professional photographer. All photographs of immunocytochemistry were taken at identical exposure times and magnifications within each figure. Images of immunocytochemistry were digitally processed to remove all color information. In addition, images were cropped and scaled, and the brightness and contrast were adjusted to highlight specific labeling. All manipulations were performed identically on each image. Photographs of in situ hybridization were digitized by using a color scanner (ScanMaker 5, Microtek). Images were cropped and scaled, color levels were adjusted to enhance Hoechst fluorescence, and brightness and contrast were adjusted to highlight labeling. All image manipulations were performed by using Photoshop (Adobe Systems, Inc., San Jose, CA). Panels were assembled by using Illustrator (Adobe).
RESULTS Glutamate localization in pleural-pedal ganglia To investigate levels of glutamate in the VC sensory neuron cluster in the pleural ganglion, semithin sections were immunolabeled for glutamate. Glutamate immunoreactivity (GLU-IR) was high in the somata of sensory neurons (Fig. 1A). Some other neurons outside of the VC sensory cluster in the pleural ganglion showed moderate
123
levels of GLU-IR (Fig. 1A). However, most neurons in the pleural and pedal ganglia showed weaker GLU-IR than the pleural sensory neurons (Fig. 1A,B). Sections through the neuropil of the pleural and pedal ganglia and pleuralpedal connective showed very strong GLU-IR (Fig. 1C,D) in fibers within the neuropil and connective. This finding is consistent with the anatomic organization of sensory neurons in the pleural ganglion that project axons through the neuropil and connective to the neuropil of the pedal ganglion, where they form synapses with pedal motor neurons. Thus, some GLU-IR in the neuropil is likely to be due to processes from pleural sensory neurons. Substitution of normal rabbit serum for the glutamate antiserum eliminated specific labeling in both the pleural (Fig. 1E) and pedal ganglia (Fig. 1F).
Amino acid immunolocalization in pleuralpedal ganglia To determine the relative distribution of glutamate to other transmitter amino acids and amino acids associated with glutamate metabolism, serial sections were immunolabeled with antisera against GLU, ASP, GLN, GABA, and GLY. Both ASP-IR (Fig. 2C) and GLN-IR (Fig. 2E) were present at levels similar to GLU-IR (Fig. 2A) in the cell bodies of sensory neurons. However, very little ASP-IR (Fig. 2D) or GLN-IR (Fig. 2F) was present in the neuropil or pleural-pedal connective compared with GLU-IR (Fig. 2B). Furthermore, little or no GABA-IR (Fig. 2G,H) or GLY-IR (Fig. 2I,J) was observed in the cell bodies of sensory neurons (Fig. 2G,I), neuropil, or pleural-pedal connective (Fig. 2H,J), consistent with previous reports (Cleary and Li, 1990; Soinila and Mpitsos, 1991; Diaz-Rios et al., 1999). To be certain that the absence of GABA-IR in the pleural-pedal ganglia was not due to technical reasons, sections of buccal ganglia with known GABAergic neurons were immunolabeled for GABA and glutamate. Previous work that used frozen sections (Soinila and Mpitsos, 1991) and whole-mounts (Diaz-Rios et al., 1999) had identified a small cluster of GABAergic neurons in this ganglion. This same population of neurons was identified by our techniques (Fig. 3), verifying that the anti-GABA antiserum accurately detected GABA in our preparation, and supports the observation that sensory neurons did not contain GABA. Furthermore, these results also demonstrate that our techniques are comparable to other immunocytochemical techniques for the detection of small molecule neurotransmitters.
Amino acid immunoreactivity in cultured sensory neurons Intense GLU-IR in fibers within the neuropil and connective of pleural-pedal ganglia was consistent with labeling of the processes of sensory neurons. However, it was not possible to identify the origin of the labeled fibers in semithin sections, although neurites extending from sensory cells were observed (e.g., Fig. 1C). Therefore, cultured sensory neurons were immunolabeled to examine the presence of glutamate and other amino acids in their cell bodies and processes. Each amino acid antiserum was calibrated in the mammalian retina to have equal affinity for the appropriate target hapten (Marc et al., 1995, Marc et al., 1998b; Kalloniatis et al., 1996). The somata, processes and varicosities of cultured sensory neurons showed strong GLU-IR (Fig. 4A), consistent with results
Fig. 1. Levels of glutamate immunoreactivity (GLU-IR) in pleural sensory neurons and neuropil. Glutamate was detected immunocytochemically in semithin sections of pleural-pedal ganglia. A: The large arrow points toward the sensory neuron cluster in the pleural ganglion, which showed high levels of GLU-IR. The adjacent posterior neurons (asterisks) show much lower levels of glutamate immunoreactivity. B: GLU-IR in the pedal ganglion. Although some pedal neurons showed GLU-IR levels comparable to sensory neurons, GLU-IR of pedal neurons was generally much less than in the pleural sensory neurons. C: A section through the sensory neuron cluster and associated neuropil of the pleural ganglion. The large arrow points toward
the sensory neuron cluster of the pleural ganglion. Note the extremely high levels of glutamate in the neuropil (n) and pleural-pedal connective (c). The small arrow within the sensory cluster points toward a proximal process extending from the somata of a sensory neuron. D: A section through the pedal ganglion demonstrates the relatively low levels of GLU-IR in pedal neurons and high levels of GLU-IR in the neuropil and connective. E,F: Serial sections to A and B respectively, immunolabeled with normal rabbit serum. Specific immunolabeling was eliminated by this treatment, indicating the GLU-IR observed was specific. Scale bar ⫽ 40 m in F (applies to all panels).
GLUTAMATE AND GLUTAMATE TRANSPORTERS IN APLYSIA
125
in semithin sections (Fig. 1A,C). Varicosities were labeled more intensely than cell bodies or processes, consistent with higher glutamate concentrations at synapses than other parts of the neuron (Fig. 4A). ASP-IR and GLN-IR was also present in sensory neuron somata and processes (Fig. 4B,C), but their levels were much lower than GLU-IR (Fig. 4A). Varicosities of cultured sensory neurons did not show appreciable ASP-IR (Fig. 4B) or GLN-IR (Fig. 4C). Cultured sensory neurons showed no appreciable GABA-IR (Fig. 4D) or GLY-IR (Fig. 4E). Normal rabbit serum controls were also immunonegative (Fig. 4F). These
Fig. 3. ␥-Aminobutyric acid-ergic (GABAergic) neurons in the buccal ganglion. By using the same GABA antiserum and immunocytochemical methods as in Figure 2G–H, sections of buccal ganglia were immunolabeled for GABA. The arrows indicate neurons which were immunopositive for GABA. The neuropil (n) and connective tissue sheath (s) had relatively low levels of GABA immunoreactivity. This result indicates absence of GABA from the pleural ganglion is genuine. Scale bar ⫽ 40 m.
results suggest that processes from pleural sensory neurons contribute to the GLU-immunoreactive plexus in the neuropil and connective of the pleural-pedal ganglia (Figs. 1, 2).
Localization of glutamate transporter mRNA Sensory neurons might be expected to possess a highaffinity glutamate uptake mechanism if they are glutamatergic, as high-affinity glutamate transporters are the primary mechanism used for termination of neurotransmission at glutamatergic synapses (reviewed in Kanai and Hediger, 1996). A PCR-based strategy by using cDNA isolated from the Aplysia central nervous system was used to obtain a probe that could be used for in situ hybridization to localize mRNA coding for high-affinity glutamate transporters. Three partial cDNA clones of high-affinity glutamate transporters were obtained. In preliminary experiments, riboprobes made from these three clones were screened for binding to sensory neurons. The first clone
Fig. 2. Levels of glutamate immunoreactivity (GLU-IR) in pleural sensory neurons and neuropil compared with levels of other amino acids. Serial semithin sections were immunolabeled with antisera against GLU, aspartate (ASP), glutamine (GLN), ␥-aminobutyric acid (GABA) or glycine (GLY). Labels to the left side of the photomicrographs indicate the amino acid antiserum to which the sections were exposed. Labels above each column of photomicrographs indicate the field of view. A: A section through the pleural sensory cluster shows that somata of sensory neurons have high levels of GLU-IR (see Fig. 1A). B: A section through the pleural sensory cluster (large arrow) and associated neuropil shows very high levels of GLU-IR in the neuropil (see Fig. 1C). Levels of ASP-IR (C) and GLN-IR (E) in the cell bodies of sensory neurons were similar to levels of GLU-IR (A). However, very little ASP-IR (D) or GLN-IR (F) was present in the neuropil. Sections immunolabeled for GABA (G,H) or glycine (I,J) show very little immunoreactivity in the cell bodies (G,I) of sensory neurons or associated neuropil (H,J). Scale bar ⫽ 40 m in J (applies to all panels).
126
Fig. 4. Levels of glutamate in cultured pleural sensory neurons compared with levels of other amino acids. Sensory neurons isolated from pleural ganglia were grown in culture and then immunolabeled for glutamate (GLU), aspartate (ASP), glutamine (GLN), ␥-aminobutyric acid (GABA), or glycine (GLY). A: GLU immunoreactivity (GLU-IR) in cultured sensory neurons was very high in the cell body (large arrow), processes (arrowhead), and varicosities (small arrows). ASP-IR (B) and GLN-IR (C) was present in the cell body and
J. LEVENSON ET AL.
processes of cultured sensory neurons, but was always weaker than GLU-IR. Neither ASP-IR nor GLN-IR was present in synaptic varicosities. Very little immunoreactivity was seen in sensory neurons immunolabeled for GABA (D) or glycine (E). F: Sensory neurons immunolabeled with normal rabbit serum showed faint background immunofluorescence in the cell body, but none in processes or varicosities. Scale bar ⫽ 80 m in F (applies to all panels).
GLUTAMATE AND GLUTAMATE TRANSPORTERS IN APLYSIA
127
Fig. 5. Partial cDNA clone of an Aplysia glutamate transporter (ApGT1). A: Nucleotide sequence of partial glutamate transporter clone. B: Deduced amino acid sequence of Aplysia glutamate transporter compared with mammalian (EAAT1-5) and other invertebrate glutamate transporter sequences deposited in GenBank database. The Aplysia peptide sequence shares a high degree of identity with
the other sequences. Underlines and numbers indicate putative membrane spanning regions. Black backgrounds indicate identity in at least half of the sequences. Shaded backgrounds indicate a conservative substitution. White backgrounds indicate no identity or conservation with other sequences.
(ApGT1) specifically hybridized to the sensory neurons and some glia (Fig. 6B; Levenson et al., 1999). The second clone (ApGT2) seemed to hybridize to all neurons, whereas the third clone (ApGT3) seemed to hybridize only to glia (Levenson et al., 1999). We have used ApGT1 in the studies reported here (Fig. 5A). The partial sequence of the ApGT1 clone includes the last three membrane spanning regions of the protein, as well as a putative substrate binding region (Fig. 5B; Seal and Amara, 1998). The predicted peptide sequence of ApGT1 shares a high degree of identity (75– 85%) with other vertebrate and invertebrate high-affinity glutamate transporters (Fig. 5B). ApGT1 has essentially no sequence similarity to transporters for other amino acid or neurotransmitter transporters. A full-length ApGT1 riboprobe was used to localize expression of glutamate transporters in the pleural-pedal ganglia by in situ hybridization. Strong hybridization was observed in the pleural ganglion, the pleural-pedal connective, neuropil, and other pedal connectives (Fig. 6A). The VC-cluster of sensory neurons in the pleural ganglion showed strong hybridization to the ApGT1 riboprobe (Fig. 6B). Most other neurons of the pleural and pedal ganglion
did not express this mRNA (Fig. 6A,B). However, some small, presumably glial cells surrounding neurons in the pedal ganglion (Fig. 6A,B), showed hybridization to the riboprobe. This ring-like hybridization pattern surrounding some cells, designated by asterisks, is in contrast to the uniform hybridization observed in the sensory neurons. Thus, the pattern of hybridization associated with the sensory neurons is, therefore, most likely due to neuronal and not glial hybridization. The comparatively low level of glutamate transporter mRNA in the neurons of the pedal ganglion (Fig. 6A) corresponds to the low levels of GLU-IR detected in pedal neurons (Fig. 1B). Interestingly, the riboprobe also hybridized to some glial cells in the connective (Fig. 6C) and neuropil (Fig. 6D) of the pleural-pedal ganglia. These glia could represent a subpopulation of glia that are associated with glutamatergic synapses. An antisense riboprobe to ApGT1 did not hybridize (Fig. 6E), indicating the hybridization of the glutamate transporter probe was specific. These results indicate that a subset of the neurons and glia of the pleural and pedal ganglia, including the sensory neurons, express
128
J. LEVENSON ET AL.
Figure 6
an mRNA that may encode a high-affinity glutamate transporter. To investigate whether the glutamate transporter gene cloned in this study was a good marker for glutamatergic neurons, we also investigated the pattern of hybridization
in the buccal ganglion. Specifically, we were interested in whether the riboprobe recognized neurons B3 or B6 in the rostroventral cluster of motor neurons in the buccal ganglion, because these neurons are known to be glutamatergic (Fox and Lloyd, 1997, 1999). Neurons, which appeared
GLUTAMATE AND GLUTAMATE TRANSPORTERS IN APLYSIA to be B3 and B6, exhibited very strong hybridization (Fig. 6F). In general, few other neurons in the buccal ganglion hybridized to ApGT1. Strong hybridization associated with glia was seen in the buccal-buccal connective (Fig. 6F). This result indicates that ApGT1 hybridization may be specific for neurons that use glutamate as their neurotransmitter.
129
Dale and Kandel, 1993; Klein et al., 1998; Fox and Lloyd, 1999).
Localization of glutamate
Glutamate is used as a neurotransmitter of the CNS by several molluscs, including Aplysia. Biochemical and anatomic studies have shown that molluscan brains possess high levels of glutamate (Osborne et al., 1971), and that both neurons and glia can accumulate radiolabeled glutamate (Turner and Cottrell, 1978). Similar to other molluscs, Aplysia ganglia have been shown to possess relatively high amounts of endogenous glutamate (Iliffe et al., 1977). Furthermore, levels of glutamate in some individual Aplysia neurons are greater than other amino acids (Borys et al., 1973). All of these studies indicate that the molluscan CNS accumulates glutamate at levels higher than would be expected if glutamate were used for protein synthesis alone. In support of this observation, several neurons in Aplysia have been subsequently shown to use glutamate as a neurotransmitter (Yarowsky and Carpenter, 1976; Taraskevich et al., 1977; Sawada et al., 1984; King and Carpenter, 1987, King and Carpenter, 1989;
In this study, all neuronal somata possessed detectable GLU-IR. This result was expected, because glutamate is required for protein synthesis and other metabolic functions in all cells. However, the pleural sensory neurons expressed much higher levels of glutamate in their cell bodies compared with most other pleural and pedal neurons. This finding is consistent with the observation that glutamatergic neurons often show very high levels of glutamate (Yudkoff et al., 1989, 1990; Marc et al., 1998a; reviewed in Erecinska and Silver, 1990; Schousboe et al., 1993). One striking observation was the extremely high levels of glutamate in the neuropil of pleural-pedal ganglia (Fig. 1C,D), and in processes and varicosities of cultured sensory neurons (Fig. 4A). This finding is in contrast to the distribution of aspartate, glutamine, glycine, and GABA, which were not present at such high levels in the neuropil of the pleural or pedal ganglia (Figs. 2D,F,H,J, 4B–E). In many vertebrate glutamatergic neurons, somatic glutamate concentrations are approximately 1 mM, and increase dramatically in processes, reaching 10 mM at the terminal and approaching 100 mM within synaptic vesicles (reviewed in Erecinska and Silver, 1990; Kanai and Hediger, 1996). Glutamate immunoreactivity in Aplysia pleural-pedal ganglia shows this same subcellular distribution, with the highest glutamate concentrations in the synaptic region of the neuropil (Fig. 1C,D). Furthermore, this pattern of immunoreactivity was specific to glutamate, as this pattern was not observed for the other amino acids that were investigated (Fig. 2). To determine whether some of the immunoreactivity seen in the neuropil could be due to processes originating from pleural sensory neurons, we investigated the immunoreactivity of cultured sensory neurons. Glutamate immunoreactivity was high in the cell body and processes, and very high in varicosities (Fig. 4A), indicating sensory neurons are glutamatergic and that high levels of immunoreactivity in the neuropil is due in part to sensory neurons. However, high levels of glutamate in processes of other neurons, glia, or both, may also contribute to the high levels of positive immunolabeling in the neuropil.
Fig. 6. Pleural sensory neurons express mRNA encoding for a putative glutamate transporter. Sections through pleural-pedal ganglia were hybridized with a radiolabeled ApGT1 riboprobe for a putative high-affinity glutamate transporter. Nuclei were counterstained with Hoechst and appear blue. Micrographs were taken with darkfield microscopy, making hybridization appear as yellow grains. A: Strong hybridization was seen in the ventral-caudal sensory neuron cluster of the pleural ganglion (to right and above solid white curve). Strong hybridization of the riboprobe was also seen in the pleural ganglion, pleural-pedal connective (arrowhead), pedal connectives (PC and P9) and neuropil (N). Little to no hybridization was observed in neurons outside of the sensory neuron cluster (to left and below solid white curve) or in neurons of the pedal ganglion. Note the rings of hybridization around some of the pedal neurons denoted by red asterisks (asterisks). These rings most likely represent hybridization of the riboprobe to glial cells. The lack of hybridization to pedal neurons correlates with the relatively low glutamate immunoreactivity observed in these neurons (see Fig. 1B,D). B: A higher magnification of a portion of the image in A (field of view is denoted by dashed line in
A) shows strong hybridization of ApGT1 in sensory neurons (above white curve). Adjacent posterior neurons (below white line, region denoted with large white asterisk) showed little to no hybridization to ApGT1. Note the hybridization associated with the sensory neurons is uniformly distributed about the nuclei, in contrast to the rings of hybridization associated with pedal neurons (small red asterisks). C: A cross-section through the pleural-pedal connective (approximate level of section denoted by arrowhead in B) shows that the hybridization in the connective is always associated with glia (arrows indicate several glia). D: A section of neuropil also demonstrates that hybridization in the neuropil is associated with glia as well (several glia denoted by arrows). E: Riboprobes constructed by using the sensestrand of the clone showed no hybridization in the pleural ganglion, indicating that the hybridization observed in the previous panels was specific. F: Cells that appeared to be B3 and B6 (arrows) in the rostroventral motoneuron cluster of the buccal ganglion showed strong hybridization to the probe. Strong hybridization of the riboprobe in the buccal ganglion was also seen in the buccal connective (arrowhead).
DISCUSSION We have provided additional evidence that glutamate is a neurotransmitter of the sensorimotor synapse of Aplysia. The cell bodies of the VC cluster of sensory neurons in the pleural ganglion contain high levels of endogenous glutamate relative to other neurons in the pleural-pedal ganglia. Even higher levels of glutamate were observed in the neuropil and connective. Levels of glutamate also were high in the somata of cultured sensory neurons, and higher in their varicosities. Other amino acids were not localized in the neuropil or pleural-pedal connective at levels comparable to glutamate. Furthermore, the sensory neurons expressed a mRNA that likely encodes a highaffinity glutamate transporter, a trait expected of glutamatergic neurons.
Aplysia use glutamate as a neurotransmitter
130
J. LEVENSON ET AL.
Localization of other amino acids
CONCLUSIONS
Immunoreactivity for aspartate and glutamine was also investigated in cultured sensory neurons. Both ASP-IR and GLN-IR were high in the cell body, but lower in processes and varicosities (Fig. 4B,C). Furthermore, ASP-IR and GLN-IR were never as high as GLU-IR, suggesting that they may be used as precursors for glutamate synthesis. Indeed, many glutamatergic neurons use glutamine as a precursor for glutamate biosynthesis (reviewed in Torgner and Kvamme, 1990) and aspartate can be converted to glutamate by means of the enzyme aspartate aminotransferase (reviewed in Kugler, 1993). Another striking observation was the near complete absence of GABA and glycine immunoreactivity in the pleuralpedal ganglia (Fig. 2G–J). GABA-IR has been characterized previously in the pleural-pedal ganglia and has been shown to be restricted to a small population of neurons located in the pedal ganglion (Cleary and Li, 1990; DiazRios et al., 1999). These neurons project only a few processes in both the pleural and pedal ganglia (Cleary and Li, 1990; Diaz-Rios et al., 1999). Therefore, our results are consistent with previous observations. The distribution of glycine in the pleural ganglion has not been characterized before. We found basal levels of glycine in the sensory neurons (Fig. 2I), but little or no glycine in the neuropil of the pleural ganglion (Fig. 2J). This finding suggests that glycine may not be a major neurotransmitter of the pleural ganglion.
This study provides additional evidence that glutamate is the excitatory neurotransmitter used at the sensorimotor synapse. The sensory neurons of the VC cluster express higher levels of glutamate in their cell bodies, processes, and synaptic terminals than other neurons. The sensory neurons also appear to express a mRNA for a high-affinity glutamate transporter, which is also expressed by known glutamatergic neurons B3 and B6. Physiologic data also support glutamate as a transmitter of the sensory neurons. Previous studies have demonstrated that glutamate mimics the postsynaptic effects of the sensory neuron and an inhibitor of glutamate receptors eliminates the postsynaptic response (Dale and Kandel, 1993; Lin and Glanzman, 1994; Armitage and Siegelbaum, 1998). Although release of glutamate from the sensory neurons has not been shown directly, taken together, these results strongly suggest that glutamate is an excitatory transmitter at the sensorimotor synapse. Other sensory neurons in Aplysia also appear to be glutamatergic (Klein et al., 1998). It is possible that most, if not all, of the mechanosensory neurons of Aplysia are glutamatergic, although this has not been established to date. Identification of the neurotransmitter used by the sensorimotor synapse in the pleural-pedal ganglia should prove an important step to uncovering further cellular, molecular, and biochemical processes involved in plasticity at this synapse. Roles of regulation of glutamate uptake, synthesis, and release mechanisms can begin to be investigated, as well as regulation of the postsynaptic receptors (Trudeau and Castellucci, 1995; Conrad et al., 1999). We have initiated experiments to study glutamate uptake in sensory neurons and found that it is regulated during long-term sensitization (Eskin et al., 1999).
Localization of glutamate transporters Termination of neurotransmission is vital to the proper functioning of a synapse. At glutamatergic synapses, termination of neurotransmission is accomplished in large part by the action of high-affinity glutamate transporters (reviewed in Lerner, 1987; Kanai et al., 1993; Kanai and Hediger, 1996). Indeed, the presence of uptake for a transmitter is often regarded as evidence for use of that transmitter at a synapse. The localization of a mRNA encoding an Aplysia high-affinity glutamate transporter was investigated, to provide further evidence that the pleural sensory neurons were glutamatergic. Strong hybridization of the riboprobe was observed specifically in the pleural sensory neurons and in some glia (Fig. 6B). Furthermore, known glutamatergic neurons of the buccal ganglion also showed strong hybridization to the riboprobe (Fig. 6D), consistent with the assertion that ApGT1 is associated with glutamatergic neurons specifically. The localization pattern in our study is similar to the localization patterns of high-affinity glutamate transporters in the mammalian brain, in which glutamatergic neurons and associated glia express high levels of mRNA encoding high-affinity glutamate transporters (Rothstein et al., 1994; Torp et al., 1997; Berger and Hediger, 1998; Kugler and Schmitt, 1999). The strong hybridization associated with putative glial cells may represent glia that are associated with glutamatergic synapses. Thus, hybridization of ApGT1 to glia may be useful for the localization of terminals of sensory neurons. It should also be noted that glutamate transporters may play roles in the nervous system apart from their presumed synaptic roles. Thus, presence of glutamate transporters in the connectives of the pleural and pedal ganglia may suggest a role in providing glial cells with basal levels of glutamate required for metabolism.
ACKNOWLEDGMENTS The authors thank X. Ren for technical assistance. The authors also acknowledge the generosity of Signature Immunologics for their kind gift of the amino acid antisera used in this study. J.L. was supported by a graduate research fellowship from the Basic Research And Initiatives In Neuroscience (BRAIN) program at the University of Houston and A.E., L.D., and J.B.H. were supported by the NIH.
LITERATURE CITED Abel T, Martin KC, Bartsch D, Kandel ER. 1998. Memory suppressor genes: inhibitory constraints on the storage of long-term memory. Science 279:338 –341. Armitage BA, Siegelbaum SA. 1998. Presynaptic induction and expression of homosynaptic depression at Aplysia sensorimotor neuron synapses. J Neurosci 18:8770 – 8779. Bailey CH, Bartsch D, Kandel ER. 1996. Toward a molecular definition of long-term memory storage. Proc Natl Acad Sci USA 93:13445–13452. Berger UV, Hediger MA. 1998. Comparative analysis of glutamate transporter expression in rat brain using differential double in situ hybridization. Anat Embryol (Berl) 198:13–30. Borys HK, Weinreich D, McCaman RE. 1973. Determination of glutamate and glutamine in individual neurons of Aplysia californica. J Neurochem 21:1349 –1351. Byrne JH, Baxter DA, Buonomano DV, Cleary LJ, Eskin A, Goldsmith JR, McClendon E, Nazif FA, Noel F, Scholz KP. 1991. Neural and molecular bases of nonassociative and associative learning in Aplysia. Ann N Y Acad Sci 627:124 –149.
GLUTAMATE AND GLUTAMATE TRANSPORTERS IN APLYSIA Chin J, Angers A, Cleary LJ, Eskin A, Byrne JH. 1999. TGF-1 in Aplysia: role in long-term changes in the excitability of sensory neurons and distribution of TR-II-like immunoreactivity. Learn Mem 6:317–330. Cleary LJ, Li T. 1990. Distribution of GABA in the pleural-pedal ganglia of Aplysia. Soc Neurosci Abstr 16:549. Conrad P, Wu F, Schacher S. 1999. Changes in functional glutamate receptors on a postsynaptic neuron accompany formation and maturation of an identified synapse. J Neurobiol 39:237–248. Dale N, Kandel ER. 1993. L-glutamate may be the fast excitatory transmitter of Aplysia sensory neurons. Proc Natl Acad Sci USA 90:7163– 7167. Diaz-Rios M, Suess E, Miller MW. 1999. Localization of GABA-like immunoreactivity in the central nervous system of Aplysia californica. J Comp Neurol 413:255–270. Erecinska M, Silver IA. 1990. Metabolism and role of glutamate in mammalian brain. Prog Neurobiol 35:245–296. Eskin A, Endo S, Kategaya LS, Byrne JH, Levenson J. 1999. Glutamate uptake is increased by long-term sensitization in Aplysia pleural-pedal ganglia. Soc Neurosci Abstr 25:862. Fox LE, Lloyd PE. 1997. Serotonin and the small cardioactive peptides differentially modulate two motor neurons that innervate the same muscle fibers in Aplysia. J Neurosci 17:6064 – 6074. Fox LE, Lloyd PE. 1999. Glutamate is a fast excitatory transmitter at some buccal neuromuscular synapses in Aplysia. J Neurophysiol 82:1477– 1488. Iliffe TM, McAdoo DJ, Beyer CB, Haber B. 1977. Amino acid concentrations in the Aplysia nervous system: neurons with high glycine concentrations. J Neurochem 28:1037–1042. Johnson GD, Davidson RS, McNamee KC, Russell G, Goodwin D, Holborow EJ. 1982. Fading of immunofluorescence during microscopy: a study of the phenomenon and its remedy. J Immunol Methods 55:231–242. Kalloniatis M, Marc RE, Murry RF. 1996. Amino acid signatures in the primate retina. J Neurosci 16:6807– 6829. Kanai Y, Hediger MA. 1996. High-affinity glutamate transporters: physiological and pathological relevance in the central nervous system. In: Brann DW, Mahesh VB, editors. Excitatory amino acids: their role in neuroendocrine function. Boca Raton: CRC Press. p 103–131. Kanai Y, Smith CP, Hediger MA. 1993. A new family of neurotransmitter transporters: the high-affinity glutamate transporters. FASEB J 7:1450 –1459. King WM, Carpenter DO. 1987. Distinct GABA and glutamate receptors may share a common channel in Aplysia neurons. Neurosci Lett 82: 343–348. King WM, Carpenter DO. 1989. Voltage-clamp characterization of Cl⫺ conductance gated by GABA and L- glutamate in single neurons of Aplysia. J Neurophysiol 61:892– 899. Klein AN, Weiss KR, Cropper EC. 1998. Glutamate is the fast excitatory neurotransmitter of peptidergic mechanosensory neuron B21. Soc Neurosci Abstr 24:359. Kugler P. 1993. Enzymes involved in glutamatergic and GABAergic neurotransmission. Int Rev Cytol 147:285–336. Kugler P, Schmitt A. 1999. Glutamate transporter EAAC1 is expressed in neurons and glial cells in the rat nervous system. Glia 27:129 –142. Lerner J. 1987. Acidic amino acid transport in animal cells and tissues. Comp Biochem Physiol B 87:443– 457. Levenson J, Dryer L, Byrne JH, Eskin A. 1999. Identification and characterization of high-affinity glutamate transporters in Aplysia. Soc Neurosci Abstr 25:862. Lin XY, Glanzman DL. 1994. Hebbian induction of long-term potentiation of Aplysia sensorimotor synapses: partial requirement for activation of an NMDA-related receptor. Proc R Soc Lond B Biol Sci 255:215–221. Marc RE, Murry RF, Basinger SF. 1995. Pattern recognition of amino acid signatures in retinal neurons. J Neurosci 15:5106 –5129.
131
Marc RE, Murry RF, Fisher SK, Linberg KA, Lewis GP. 1998a. Amino acid signatures in the detached cat retina. Invest Ophthalmol Vis Sci 39: 1694 –1702. Marc RE, Murry RF, Fisher SK, Linberg KA, Lewis GP, Kalloniatis M. 1998b. Amino acid signatures in the normal cat retina. Invest Ophthalmol Vis Sci 39:1685–1693. Osborne NN, Briel G, Neuhoff V. 1971. Distribution of GABA and other amino acids in different tissues of the gastropod mollusc Helix pomatia, including in vitro experiments with 14C glucose and 14C glutamic acid. Int J Neurosci 1:265–272. Ressler KJ, Sullivan SL, Buck LB. 1993. A zonal organization of odorant receptor gene expression in the olfactory epithelium. Cell 73:597– 609. Rothstein JD, Martin L, Levey AI, Dykes-Hoberg M, Jin L, Wu D, Nash N, Kuncl RW. 1994. Localization of neuronal and glial glutamate transporters. Neuron 13:713–725. Sawada M, Gibson D, McAdoo DJ. 1984. L-glutamic acid, a possible neurotransmitter to anterior aorta of Aplysia. J Neurophysiol 51:375–386. Schacher S, Proshansky E. 1983. Neurite regeneration by Aplysia neurons in dissociated cell culture: modulation by Aplysia hemolymph and the presence of the initial axonal segment. J Neurosci 3:2403–2413. Schousboe A, Westergaard N, Hertz L. 1993. Neuronal-astrocytic interactions in glutamate metabolism. Biochem Soc Trans 21:49 –53. Seal RP, Amara SG. 1998. A reentrant loop domain in the glutamate carrier EAAT1 participates in substrate binding and translocation. Neuron 21:1487–1498. Sherry DM, Ulshafer RJ. 1992. Neurotransmitter-specific identification and characterization of neurons in the all-cone retina of Anolis carolinensis. I: Gamma-aminobutyric acid. Vis Neurosci 8:515–529. Sherry DM, St. Jules RS, Townes-Anderson E. 1996. Morphologic and neurochemical target selectivity of regenerating adult photoreceptors in vitro. J Comp Neurol 376:476 – 488. Soinila S, Mpitsos GJ. 1991. Immunohistochemistry of diverging and converging neurotransmitter systems in mollusks. Biol Bull 181:484 – 499. Taraskevich PS, Gibbs D, Schmued L, Orkand RK. 1977. Excitatory effects of cholinergic, adrenergic and glutaminergic agonists on a buccal muscle of Aplysia. J Neurobiol 8:325–335. Torgner I, Kvamme E. 1990. Synthesis of transmitter glutamate and the glial-neuron interrelationship. Mol Chem Neuropathol 12:11–17. Torp R, Hoover F, Danbolt NC, Storm-Mathisen J, Ottersen OP. 1997. Differential distribution of the glutamate transporters GLT1 and rEAAC1 in rat cerebral cortex and thalamus: an in situ hybridization analysis. Anat Embryol (Berl) 195:317–326. Trudeau LE, Castellucci VF. 1993. Excitatory amino acid neurotransmission at sensory-motor and interneuronal synapses of Aplysia californica. J Neurophysiol 70:1221–1230. Trudeau LE, Castellucci VF. 1995. Postsynaptic modifications in long-term facilitation in Aplysia: upregulation of excitatory amino acid receptors. J Neurosci 15:1275–1284. Turner JD, Cottrell GA. 1978. Cellular accumulation of amines and amino acids in the central ganglia of a gastropod mollusc, Planorbis corneus: an autoradiographic study. J Neurocytol 7:759 –776. Wilkinson DG, Bailes JA, Champion JE, McMahon AP. 1987. A molecular analysis of mouse development from 8 to 10 days post coitum detects changes only in embryonic globin expression. Development 99:493– 500. Yarowsky PJ, Carpenter DO. 1976. Aspartate: distinct receptors on Aplysia neurons. Science 192:807– 809. Yudkoff M, Zaleska MM, Nissim I, Nelson D, Erecinska M. 1989. Neuronal glutamine utilization: pathways of nitrogen transfer studied with [15N]glutamine. J Neurochem 53:632– 640. Yudkoff M, Nissim I, Hertz L. 1990. Precursors of glutamic acid nitrogen in primary neuronal cultures: studies with 15N. Neurochem Res 15:1191– 1196.
