SEARCH

SEARCH BY CITATION

Keywords:

  • bicuculline;
  • CACA;
  • glia;
  • interneurons;
  • medium spiny neurons;
  • TPMPA

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Concluding remarks
  7. Acknowledgements
  8. References
  9. Supporting Information

J. Neurochem. (2012) 122, 900–910.

Abstract

GABAergic transmission in the neostriatum plays a central role in motor coordination, in which a plethora of GABA-A receptor subunits combine to modulate neural inhibition. GABAρ receptors were originally described in the mammalian retina. These receptors possess special electrophysiological and pharmacological properties, forming a characteristic class of ionotropic receptors. In previous studies, we suggested that GABAρ receptors are expressed in the neostriatum, and in this report we show that they are indeed present in all the calretinin-positive interneurons of the neostriatum. In addition, they are located in calbindin-positive interneurons and projection neurons that express the dopamine D2 receptor. GABAρ receptors were also located in 30% of the glial fibrillary acidic protein-positive cells, and may therefore also contribute to gliotransmission. Quantitative reverse transcription-PCR suggested that the mRNAs of this receptor do not express as much as in the retina, and that GABAρ2 is more abundant than GABAρ1. Electrophysiological recordings in brain slices provided evidence of neurons expressing a cis-4-aminocrotonic acid-activated, 1,2,5,6-tetrahydropyridine-4-yl methylphosphinic acid-sensitive ionotropic GABA receptor, indicating the presence of functional GABAρ receptors in the neostriatum. Finally, electron-microscopy and immunogold located the receptors mainly in perisynaptic as well as in extrasynaptic sites. All these observations reinforce the importance of GABAρ receptors in the neostriatum and contribute to the diversity of inhibitory regulation in this area.

Abbreviations used:
aCSF

artificial CSF

Act

actin

CACA

cis-4-aminocrotonic acid

cc

corpus callosum

CPu

caudate putamen (neostriatum)

DAPI

4′,6-diamidino-2-phenylindole

Den

dendrite

Drd1-GFP

dopamine D1 receptor–green fluorescent protein coupled

Drd2-GFP

dopamine D2 receptor–green fluorescent protein coupled

GFAP

glial fibrillary acidic protein

ISH

in situ hybridization

LV

lateral ventricle

M

mitochondria

MSNs

medium spiny neurons

PBS

phosphate-buffered saline

qRT-PCR

quantitative reverse transcription-PCR

TM

transmembrane domain(s)

TPMPA

1,2,5,6-tetrahydropyridine-4-yl methylphosphinic acid

The neostriatum is considered the most important connection between the cortical input and the basal ganglia, and it has been associated with motor control, learning, and memory as well as with neurodegenerative disorders such as Parkinson’s and Huntington’s diseases (Kemp and Powell 1971; Alexander and Crutcher 1990; Doig et al. 2010). Neurons in this area are functionally divided into two types: (i) medium spiny neurons (MSNs), which are the most abundant and (ii) a small population of aspiny neurons (interneurons). Almost all of these neurons are GABAergic. However, the MSNs also synthesize neuropeptides, including substance P, dynorphin, and enkephalin (Kawaguchi 1997). In addition, the MSNs express dopamine D1 and D2 receptors; whereas the neurons that express D1 receptors synthesize substance P and dynorphin, those that express D2 receptors express enkephalin (Gerfen et al. 1990; Surmeier et al. 1993).

The interneurons of the neostriatum are aspiny with short axons and are divided into four independent types (Kawaguchi 1993, 1997, Kubota and Kawaguchi 1993; Kawaguchi et al. 1995): (i) those that express acetylcholine but not GABA (Zhou et al. 2002); the other three interneuron types are GABAergic and express either: (ii) parvalbumin (Kita et al. 1990), (iii) calretinin (Bennett and Bolam 1993a; Rymar et al. 2004), or (iv) somatostatin, NADPH, neuropeptide Y, and calbindin (Bennett and Bolam 1993b). These four types are distinguished morphologically by their soma diameter and axon length.

The neostriatum shows strong expression of many types of ionotropic GABA receptors including subunits α1-5, β2-3, γ2, and δ (Fritschy and Mohler 1995; Albrecht et al. 1997). GABAρ receptor expression in the neostriatum was also demonstrated by means of RT-PCR and immunohistochemistry (López-Chávez et al. 2005; Rosas-Arellano et al. 2007). GABAρ receptors possess special electrophysiological and pharmacological properties that set them apart from other GABAA subunits: they desensitize very little upon activation, are insensitive to bicuculline and baclofen (Polenzani et al. 1991), and are activated selectively by the specific agonist cis-4-aminocrotonic acid (CACA) and are antagonized by 1,2,5,6-tetrahydropyridine-4-yl methylphosphinic acid (TPMPA; Ragozzino et al. 1996).

Despite the accumulated knowledge about GABAρ receptors their specific function in the central nervous system remains largely unknown. These receptors have been located in the amygdala (Delaney and Sah 1999; Fujimura et al. 2005; Flores-Gracia et al. 2010; Rosas-Arellano et al. 2011), in the mitral layer of the olfactory bulb (Chen et al. 2007), Purkinje neurons of cerebellum (Drew and Johnston 1992; Albrecht et al. 1997; Boue-Grabot et al. 1998; Rozzo et al. 2002; Harvey et al. 2006; Mejía et al. 2008; ); in the gray layer of the superior colliculus (Wegelius et al. 1998; Pasternack et al. 1999; Schmidt et al. 2001; Frazao et al. 2007; Born and Schmidt 2008; Wahle and Schmidt 2009), in the oval and pyramidal neurons of layers II-VI of the visual cortex (Wegelius et al. 1998; Wahle and Schmidt 2009; Rosas-Arellano et al. 2011); in the corpus callosum (López-Chávez et al. 2005); in brainstem and spinal cord (Johnston et al. 1975; Wegelius et al. 1998; Rozzo et al. 2002; Zheng et al. 2003; Milligan et al. 2004; López-Chávez et al. 2005; Frazao et al. 2007; Rosas-Arellano et al. 2007). In addition, several reports indicate that the pyramidal neurons of CA1–CA3 areas, dentate gyrus, and subiculum of the hippocampus express GABAρ receptors (Strata and Cherubini 1994; Cherubini et al. 1998; Wegelius et al. 1998; Didelon et al. 2002; Rozzo et al. 2002; Hartmann et al. 2004; Alakuijala et al. 2006; Rosas-Arellano et al. 2011).

Our previous studies disclosed the presence of GABAρ in the neostriatum (López-Chávez et al. 2005; Rosas-Arellano et al. 2007), and we have now studied their distribution in this area. GABAρ receptors were found mainly in calretinin-positive neurons, as well as in a population of calbindin- and Drd2-expressing neurons and in glial fibrillary protein (GFAP)-positive cells. In addition, recordings in brain slices provided evidence of neurons expressing a CACA-activated, TPMPA-sensitive ionotropic GABA receptor, indicating the presence of functional GABAρ receptors in the neostriatum. Finally, electron-microscopy and immunogold located the receptors in extrasynaptic sites.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Concluding remarks
  7. Acknowledgements
  8. References
  9. Supporting Information

Animals

We used the following mouse strains: CD1, NMRI, and two transgenic strains Drd1-GFP, Drd2-GFP. All the animals were handled in accordance with the guidelines of the National Institutes of Health Guide for Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee of the Universidad Nacional Autónoma de México.

RNA isolation and real-time RT-PCR

Total RNA was isolated with TRIzol®Reagent (Invitrogen, Carlsbad, CA, USA) from 100 to 300 mg of retinas or striatum of CD1 mice. RNA was reverse transcribed into cDNA using the SuperScriptTM II Rnase H Reverse Transcriptase (Invitrogen). Expression levels of GABAρ1 and ρ2 were determined and compared for striatum and retina by means of quantitative reverse transcription-PCR (qRT-PCR) using The LightCyclerTM (Roche Diagnostics, Indianapolis IN, USA). To determine the relationship between cycle number (Ct) and expression of each GABAρ mRNA, primers were calibrated by using serial dilutions of cDNA. Data from three independently synthesized samples were collected, and amplifications were carried out in duplicate. Reactions were performed with ‘Light Cycler Fast Start DNA Master SYBR® Green I’ using actin and tubulin as standard. Each reaction included 5 μL of sample cDNA, 2.5 mM MgCl2, 250 μM sense and antisense primers (Table S1), 1 μL of SYBR® Green Taq ReadyMixTM, and 2 μL of water in a total reaction volume of 10 μL. Reaction conditions were 95°C for 10 min for one cycle (hot start), followed by 40 cycles of 95°C for 10 s, 60°C for 10 s, and 72°C for 12 s. The qRT-PCR results were analyzed using the 2−ΔΔCt method and are illustrated as mean ± SE. The data were evaluated and their significance determined by one-way analysis of variance (anova) followed by the Tukey’s test for group comparison.

In situ hybridization

Probes for in situ hybridization (ISH) were isolated by RT-PCR (Table S2). The cDNAs were cloned into pGEM-T-Easy (Promega, Madison, WI, USA) and sequenced. In vitro transcription was performed following the manufacturer’s instructions using digoxigenin-11-UTPs (Invitrogen).

For this analysis, CD1 mice (25–30 g) were anesthetized with pentobarbital and perfused transcardially with saline (0.9% NaCl) and fixative solutions (4% paraformaldehyde in 0.1 M phosphate buffer). Then, the brain and ocular globes were removed and post-fixed, and 150 μm slices of retina and 100 μm slices of whole brain were obtained. Hybridization was performed at 70°C using the method described by the manufacturer (Roche Molecular Biochemicals, Indianapolis, IN, USA). The slices were fixed, mounted on microslides (superfrost/Plus; Daigger, Vernon Hills, IL, USA) with Fluoromont G mounting medium (SouthernBiotech, Electron Micoscopy Sciences, Southern Biotechnology, Birmingham, AL, USA), and observed and photographed in an Olympus BX60 microscope. Sense probes did not show any label and the specificity of the ISH probes was confirmed in mouse retina (not shown).

Western blot and immunohistochemistry

The specificity of the antibodies was tested by western blot (for a complete list of antibodies used in this study see Tables S3 and S4) and by their ability to locate the receptor in retina, cerebellum, and Secreting Tumor Cells -1 (not shown). For western blot, ocular globes and neostriatum of CD1 mice were processed as previously reported (Miledi et al. 2002). Proteins were resolved in a 12% polyacrylamide gel electrophoresis and anti-GABAρ1, anti-ρ2 or anti-actin were probed against the nitrocellulose membrane. Immunohistochemistry was based on our previous reports (Rosas-Arellano et al. 2007, 2011). To assess the expression of GABAρ in MSNs we used male mice (25–30 g) of the bacterial artificial chromosome transgenic mice strains Drd1-GFP and Drd2-GFP that express GFP in neurons that present either D1 or D2 dopamine receptors (Gong et al. 2003; Doig et al. 2010). 16 slices from three different brains of each strain were processed for immunofluorescence. To determine expression of the receptor in interneurons, four males of the CD1 strain were used for each marker making a total of 96 slices. Mice were perfused as described in the previous section, cryoprotected in sucrose gradients, and 30 μm sagittal sections of whole brain were obtained in a cryostat (Leica CM1850). Sections exposing the neostriatum in a lateral view were placed on slides (Superfrost®/Plus by Daigger) and stored at 4°C until processing by immunoflourescence or double immunofluorescence.

For immunoflourescence, the tissue was washed in phosphate-buffered saline (PBS)-Tween20, and the non-specific sites were blocked with a solution containing 2% donkey serum. The sections were incubated at 4°C in the primary antibody solution [anti-GABAρ1 or ρ2 (Tables S3 and S4), 0.05% thimerosal, and 0.1% Tween20 in TBS], washed in PBS-Tween20, and incubated at 4°C in the secondary antibody conjugated to Alexa 594. Finally, the tissues were washed in PBS-Tween20 and treated as described at the end of this section.

For double immunofluorescence, the primary antibodies included anti-GABAρ1 [from goat or from rabbit (for double localization GABAρ1/ρ2)], anti-GABAρ2 in combination with anti-calretinin, anti-choline acetyltransferase, anti-somatostatin, anti-calbindin, anti-parvoalbumin. After several experiments, we observed that some processes labelled by the antibodies resembled those of astrocytes; thus, we performed double immunofluorescence with GABAρ1 or GABAρ2 with goat IgG anti-glial fibrillary protein (GFAP).

The tissue was counter-stained with 4′,6-diamidino-2-phenylindole (DAPI), dehydrated in alcohol gradients, and mounted with Vectashield H-1000 (Vector Laboratories, Burlingame, CA, USA). For imaging we used a Zeiss LSM510 Meta confocal microscope; 561 nm was used for excitation of Alexa 594, 488 nm for Alexa 488 and GFP, and 750 nm for DAPI.

For image quantitative analysis, the z-stack images (4 or 5 consecutive confocal sections of 512 × 512) were obtained every 5 μm with stack size of X: 450 μm and Y: 450 μm and processed in Aim Image Examiner. The population of cells labelled by the different fluorescent probes was measured independently and contrasted with the number of DAPI labelled nuclei for each confocal section. NIH/ImageJ analysis software was used for quantifying simple and double immunolabeling and is reported at the total of immunolabeling cells per nuclei labelled with DAPI. The statistical analysis was performed as described previously by Griffiths and Lovick (2005). The abundance and spatial distribution of GABAρ1 and GABAρ2 in the neostriatum were determined by using one-way anova followed by Fisher’s Protected Least Significant Difference post hoc test. Differences were considered significant at p < 0.05.

Neostriatal brain slices preparation and electrophysiological recordings

Slices were prepared from 8- to 10-day-old NMR1 mice (Charles River) because neuron access and identification for patch-clamp recording is easier than in older animals. Slices, artificial CSF (aCSF), internal solutions, and patch micropipettes were prepared and used as previously described (Reyes-Haro et al. 2010).

The neostriatum was recognized using light microscopy, and neurons in the dorsal area of rostral coronal brain slices were recorded with the patch-clamp technique using the whole-cell recording configuration (Hamill et al. 1981). Current signals were amplified with a triple EPC10 (HEKA), filtered at 3 kHz, sampled at 10 kHz, and recorded using the TIDA software (5.19). Chemicals were obtained from Sigma-Aldrich (St Louis, MO, USA) or Tocris Cookson (Ballwin, MO, USA) if not otherwise indicated. Slices were superfused with oxygenated aCSF with 1 μM Tetrodotoxin to minimize the indirect effect of neuronal electrical activity. Likewise, 25 μM 6-cyano-7-nitroquinoxaline-2,3-dione, 50 μM D-2-amino-5-phosphonovaleric acid, and 1 μM strychnine were added to the aCSF to block ionotropic glutamate receptors and glycine receptors. In general, the number of experiments (n) refers to the number of GABA- or CACA-evoked responses. A small hyperpolarizing voltage command (10 mV) was given during the experiment to monitor access resistance (5–20 Mohm). Access resistance was monitored continuously, and experiments were abandoned if changes > 20% were detected. No cell capacitance, series resistance, or liquid junction potential compensations were made. Three to five puffs of the agonist were applied as control responses on each recorded neuron and then applied in the presence of the antagonist (TPMPA or bicuculline) that was added to the aCSF and pre-incubated > 1 min before agonist puffs. Statistical analysis was performed using Origin7.0 software (Origin Laboratories, LLC, Pasadena, CA). The results are expressed as mean ± SEM if not otherwise stated. When experiments included a control and more than one test group, data were statistically evaluated with the Tukey’s test. We used the Student’s t-test to compare two groups (control and test) within an experiment. The p values of < 0.05 were considered significant.

Immunogold and electron microscopy

This was performed as described previously (Mejía et al. 2008) with several modifications. Four animals were anesthetized as described above, and perfused with 155 mM NaCl, 2 mM KCl, 3.4 mM NaH2PO4, 2 mM CaCl2, 1 mM MgCl2, 30 mM HEPES, and 0.08 g/100 mL heparin sodium salt, pH 7.4. The fixative solution contained 10 mM HEPES, 2 mM CaCl2, 120 mM NaCl, 4% paraformadehyde, and 0.5% glutaraldehyde. After an impregnation lapse, careful dissection of the whole brain was performed. After post-fixing, the region corresponding to the neostriatum was carefully dissected. The tissue was cut in two sections (dorsal and ventral regions); only the dorsal region was placed again in fresh fixative solution. Then it was washed in 0.1 M sodium cacodylate, pH 7.0, with 2.5% sucrose; dehydrated in ethanol gradients, infiltrated in LR white resin, and embedded for thermal curing in gelatin capsules. Finally, 500 nm slices were obtained with a glass knife, and from them 60 nm slices were obtained in an ultramicrotome (MRC-MTXL) using a diamond point knife; these sections were placed on nickel grids covered with Formvar (by IACSA).

The samples were washed with deionized and distilled water and PBS, and then incubated in 0.5 M glycine in PBS; non-specific sites were eliminated with 2% rabbit serum, and antibody anti-GABAρ1 or anti-GABAρ2 was added. After rinsing in PBS, the secondary antibody was added (anti-goat 10-nm gold conjugate). Finally, the samples were dyed with 2% uranyl acetate, placed on 2% lead citrate, rinsed, and observed under the electron microscope (JEOL JEM 1010). For quantitative analysis, the number of nano gold particles was determined with the NIH/ImageJ software, and statistically analyzed as described previously by Gundersen et al. (1998). Briefly, micrographs were taken of four brains of the dorsal neostriatum at 50 000× magnification, and the number of gold particles determined in images digitally zoomed at 80 000×. GABAρ1 and ρ2 immunoreactivity was quantified in pre- and post-synaptic regions and in extra, peri and synaptic sites. To determine if differences exist in the number of gold particles labelling each subunit, as well as their spatial distribution, the results were statistically evaluated performing the non-parametric Kruskal–Wallis test and Mann–Whitney U as post hoc test. To assess the co-localization of GABAρ1 and GABAρ2, their distribution in ultrathin sections was analyzed in the axon terminals, dendritic processes and soma, applying the same tests described above.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Concluding remarks
  7. Acknowledgements
  8. References
  9. Supporting Information

GABAρ receptors in the mouse neostriatum

Previous findings indicated the expression of GABAρ receptors in the bovine caudate nucleus, and this prompted us to analyze in more detail their expression pattern in the homologous and predominantly GABAergic structure of the mouse (neostriatum). First, to ensure that in this species the expression of GABAρ is preserved, we performed RT-PCR. This analysis detected the three GABAρ genes in RNA isolated from retina (as control), whereas only GABAρ1 and ρ2 were detected in the neostriatum (Fig. 1a). qRT-PCR revealed that GABAρ mRNAs are not as strongly expressed in the neostriatum as in retina, with levels of GABAρ mRNA in striatum < 10% of those found in the retina (Fig. 1b; Figures S1 and S2). Nevertheless, western blots revealed the presence of GABAρ1 and ρ2 in extracts from neostriatum, and showed that the receptors were indeed less abundant in the neostriatum than in the retina (Fig. 1c).

image

Figure 1.  mRNA and protein expression of GABAρ in neostriatum and retina. (a) RT-PCR of retina (r) and neostriatum (ns). The synthesized cDNA products were mixed with the primers for actin (Act) and the three GABAρ subunits. The PCR products were separated by agarose gel electrophoresis containing ethidium bromide and standard molecular weight markers to determine the fragment size. The PCR products of GABAρ1, ρ2 and ρ3 were 174, 180 and 197 bp respectively. The gel shows the expression of the three GABAρ subunits in the retina and only GABAρ1 and ρ2 in the neostriatum. (b) qRT-PCR of GABAρ1 and ρ2 subunits. Expression of GABAρ1 and GABAρ2 in the neostriatum corresponded to 1.2 and 4.6% of that found in the retina (Left). Data were analyzed by the 2−ΔΔCt method and expressed in percentage to the relative expression of each subunit in the retina. Bars represent the mean ± SE of three independent experiments (p < 0.05). ***Significantly different from control group. (c) A representative western blot of extracts of neostriatum and retina, that distinguished the presence of GABAρ1 and GABAρ2 (∼50 kDa). Consistently, the expression of the receptors was more abundant in retina than in neostriatum. C, control without cDNA; Act, actin.

Download figure to PowerPoint

In situ hybridization

To obtain a detailed spatial expression pattern of the mRNAs of GABAρ1 and ρ2, in situ hybridization was performed in sagittal sections of the neostriatum followed by histological analysis. The results indicate that they were expressed sparsely in cells with soma of diameters close to 20 μm (Fig. 2a and b). Histological analysis revealed the expression localized mostly to the dorso-caudal region, some was found in rostral and ventral regions (fundus striati), and scarcely any in other regions (i.e. central regions) (not shown). It is interesting to indicate that we located the expression concentrated mostly in cells with somas of ≤ 10 μm of diameter (Fig. 2c–e and f–h).

image

Figure 2. In situ hybridization and immunohistochemistry. ISH analyses in sagittal sections of the expression of (a) GABAρ1 and (b) GABAρ2. The arrows show the sparsely distributed cells in the dorsal region of the neostriatum. Panels (c–e) and (f–h) show cells (arrows) and somas with diameters ≤ 10 μm (arrowheads). (i) Diagrams of brain sagittal sections showing the localization of GABAρ distribution at several levels along the neostriatum. Note that both subunits were located mainly in rostral, dorsal, and ventral regions, as indicated in circles in light- and dark gray. Consistently, the label was most abundant in dorsal areas. (j) Relative amounts of the two GABAρ subunits as percent of the total along five levels of the neostriatum, as indicated in the text and figure. Notice that GABAρ2 is more predominant except at 2.28 mm. Numbers below the figures indicate the distance to the brain midline. Scale bars: (a, b) = 100 μm; (c–e) and (f–h) = 20 μm. cc, corpus callosum; LV, lateral ventricle; CPu, caudate putamen (neostriatum).

Download figure to PowerPoint

Distribution in neostriatum

Immunohistochemistry and immunofluorescence in slices of CD1, transgenic Drd1, and Drd2 mice revealed that the regional localization of the receptor was consistent with that of their mRNAs. In photomicrographs of 450 × 450 μm of 4–5 confocal sections, the label of GABAρ2 was more abundant than GABAρ1 in regions lateral to the midline (Fig. 2i–j): at 3.00 mm, GABAρ1 was present in 33.3%, whereas GABAρ2 was found in 66.7% of the cells; at 2.28 mm, the relative amounts were 54.8% and 45.2%; at 1.44 mm, 32.3% and 67.7%; at 1.32 mm: 19.1% and 80.2%; finally at 0.84 mm, 32% and 68%, respectively (Fig. 2i–j). Thus, the analysis of the distribution suggests that GABAρ2 is more abundant than GABAρ1 in each plane analyzed except at 2.28 mm (anova and Fisher’s Protected Least Significant Difference p < 0.05).

In Drd1-GFP and Drd2-GFP and CD1 mice, the interneurons were found to represent only about 3.3% of the total cell population (848 vs. 25,455 of DAPI signal), a number similar to that previously determined for the rat neostriatum (Kawaguchi et al. 1997), whereas the vast majority corresponded to MSNs expressing either D1 or D2 receptors. GABAρ2 was found only in a few MSNs that express Drd2-GFP (4.1%; Fig. 3a–c) and was not detected in MSNs that express Drd1-GFP, in contrast GABAρ1 was not found in either Drd1-GFP or Drd2-GFP neurons (not shown). That is to say, only a small population of MSNs express exclusively GABAρ2.

image

Figure 3.  Expression of GABAρ2 in Drd2-GFP mouse neurons and double-immunolabeling of GABAρ1 or GABAρ2/calbindin- and calretinin-positive cells in sagittal sections of neostriatum. (a) Representative immunofluorescence detection of GABAρ2-positive cells stained with Alexa 594 (red) in Drd2-GFP mice. (b) Drd2-GFP-positive cells (green). (c) Merge, colocalization of GABAρ2 in Drd2-GFP-positive cells (yellow); note that not all GABAρ2-positive cells (arrows) are GFP-positive, whereas several D2-GFP-positive neurons express GABAρ2 (arrowheads); nuclei were stained with DAPI. (d–f) Co-localization of calretinin (green) with GABAρ1 and GABA ρ2 (red in g–i). In this case, all of the calretinin-expressing cells were also positive for the receptors. Co-localization of calbindin-positive cells immunolabeled with Alexa 488 (green) and GABAρ1 (j–l) or GABAρ2 (red label in m–o). Merge images (yellow in l and o) showed that not all calbindin-positive cells are GABAρ-positive (arrowheads). Scale bars: 20 μm.

Download figure to PowerPoint

In neurons, most of the label for GABAρ1 and ρ2 was located in interneurons, and especially in calretinin-positive cells, 100% of which showed positive label for GABAρ1 and GABAρ2 (Fig. 3d–i). GABAρ1 was expressed in 8.7% and GABAρ2 in 36.7% of the calbindin-immunoreactive cells (Fig. 3j–o), whereas the choline acetyltransferase-, somatostatin- and parvalbumin-positive neurons merged with neither GABAρ1 nor ρ2 (not shown).

In all the slices, we observed that the GABAρ receptors were also found in cells with large and numerous processes, similar in shape to astrocytes. Therefore, we performed a double immunofluorescence with anti-GABAρ1/anti-GFAP and anti-GABAρ2/anti-GFAP. The results showed that GABAρ receptors are expressed in a considerable fraction of GFAP-positive cells (33.9%, 791 vs. 2,332 of DAPI signal; e.g. Fig. 4).

image

Figure 4.  GABAρ receptors in GFAP-positive cells. (a, c) Sagittal sections through the neostriatum showed immunodetection of GABAρ1 and ρ2 (in red) in cells with large and numerous processes. (b, d) Double immunofluorescence revealed that those cells were also GFAP-positive, indicating that they are glial cells. Yellow indicate the merged image of GABAρ receptors and GFAP, nevertheless not all GFAP-positive cells (green) expressed GABAρ. Nuclei were stained with DAPI (blue). Scale bars: 5 μm.

Download figure to PowerPoint

Functional expression of GABAρ

Our results showed that GABAρ1 and ρ2 subunits were present in a subset of interneurons and in a small population of MSNs Drd2-GFP. To assess if the receptors were functional in these cells, we recorded GABA responses from neurons of the neostriatum in situ. Puffs of 10 μM GABA generated currents with an amplitude of 756 ± 147 pA (n = 37). However, only a fraction of the neurons responded to 100 μM TPMPA (7 of 18 neurons), which partially blocked the GABA-current (from 756 ± 147 to 421 ± 105; n = 19; p < 0.01) (Fig. 5b and d); in contrast, the GABA-currents recorded in the rest of the neurons (11 of 18) were not significantly affected by TPMPA (from 603 ± 122 to 753 ± 140; n = 18; p = 0.42) (Figure S3). We observed that the GABA response was abolished when 200 μM bicuculline was co-applied with 100 μM TPMPA in the aCSF (n = 5) (Fig. 5c and d). We also tested CACA, a highly selective agonist of GABAρ, and found that 500 μM CACA evoked currents with an amplitude of 134 ± 7 pA (n = 8) that were reduced by 37% in the presence of 100 μM TPMPA (2 out of 5 neurons; p < 0.05) (Figure S4a and b). These results indicate that GABAρ subunits are functionally expressed and participate in the GABA -evoked currents.

image

Figure 5.  Functional expression of GABAρ in striatal neurons. (a) A representative neostriatal neuron patched in a brain slice and filled with Alexa 594. (b) GABA (10 μM) mediated response was reduced when TPMPA (100 μM) was added to the aCSF (n = 19 GABA mediated responses in 7 neurons of 18). (c) GABA-mediated responses were abolished when TPMPA (100 μM) and bicuculline (200 μM) were added to the aCSF (n = 5 responses from five neurons). (d) Summary of experiments with GABA, GABA + TPMPA and GABA + TPMPA + Bicuculline. Data are mean ± SEM. Asterisks represent significant differences: **p < 0.01 or ***p < 0.001.

Download figure to PowerPoint

Immunogold and electron microscopy

To determine the synaptic distribution of GABAρ in the neostriatum, we performed immunogold labeling and electron microscopy. Most label was found at extrasynaptic and perisynaptic sites, and little or nothing in the synaptic clefts (Fig. 6a–d). The GABAρ1 label was located as follows: 57.1% extrasynaptic, 33.3% perisynaptic, and 9.5% synaptic (an average of 4.5 of gold particles per 500 nm2); for GABAρ2 the label was 54.1%, 43.7%, and 2.0%, respectively (an average of 13.2 of gold particles per 500 nm2). The Mann–Whitney U test, two tails showed (p < 0.05) showed that GABAρ2 is more abundant than GABAρ1. In addition, when we compared the distribution of each subunit, the Kruskal-Wallis test (p < 0.05) indicated differences in the distribution of the number of receptors along the three compartments observed (extra-, peri- or synaptic), and the non-parametric post hoc test (Mann–Whitney U test, two tails) showed that the distributions in extrasynaptic versus synaptic, and perisynaptic versus synaptic are different (p < 0.05). In total, only one of ∼50 post-synaptic densities was labelled for GABAρ1, GABAρ2, or both.

image

Figure 6.  Ultrastructural localization of GABAρ. (a–d) GABAρ1 (a and c) and ρ2 (b and d) were observed on the pre-synaptic and post-synaptic sides of asymmetric synapses (arrows), arrowheads point to axon terminals. (e) Schematic representation of the localization GABAρ, illustrating that the receptors were found in extrasynaptic and perisynaptic regions, and only in a few cases were detected at the synapse. (f–h) Co-localization of GABAρ1 and ρ2. (f) GABAρ1 stained with Alexa 488 (green) and (g) GABAρ2 stained with Alexa 594 (red). (h) Merge images showed the expression of both subunits (arrows) in the same cells, whereas a few cells expressed only GABAρ2 (arrowhead). Den, dendrite; M, mitochondria; Nf, neurofilaments: E, extrasynaptic; S, synaptic; P, perisynaptic. Scale bars: (a, b), 200 nm; (c, d), 100 nm; (f–h), 20 μm.

Download figure to PowerPoint

Finally, we performed double immunoflourescence and double immunogold analyses for GABAρ1/ρ2, in which we found that both subunits are co-expressed in the same cells, except for a few neurons that expressed only GABAρ2 (Fig. 6f–h).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Concluding remarks
  7. Acknowledgements
  8. References
  9. Supporting Information

The GABAρ receptor has been found in several areas of the CNS (see Introduction section) including the bovine caudate nucleus, where GABAρ receptors were located in pyramidal and fusiform cells (López-Chávez et al. 2005; Rosas-Arellano et al. 2007). In the present study, we confirm and extend our previous reports; by means of RT-PCR, qRT-PCR, and Western blot we determined that GABAρ receptors are expressed in the mouse neostriatum. ISH and immunofluorescence analyses showed that neurons and glial cells in rostral, dorsal, and ventral regions display the receptors. Additionally by means of electrophysiology we demonstrated that GABAρ subunits are functional in the neostriatum.

GABAρ receptors were found in all interneurons that express the calcium binding protein, calretinin. Calretinin-expressing neurons constitute a small population of cells distributed along the neostriatum, but that show a higher presence towards the rostral regions; these cells have diameters between 9 and 17 μm, and their soma can be either round, oval or fusiform (Bennett and Bolam 1993a). These characteristics are consistent with the cells observed in this study to express GABAρ. The electrophysiological profile of these neurons remains mostly unknown and thus, future molecular and functional studies are required to determine the role of GABAρ in these interneurons.

We observed also that a small population of calbindin-positive cells express GABAρ. These neurons are found mostly in ventral regions and have somas with diameters of 10 to 20 μm, which fit well the description reported in the rat neostriatum (Bennett and Bolam 1993a,b; Kawaguchi et al. 1997). It was precisely in this region where we detected a significant concentration of GABAρ-positive cells containing calbindin, but in contrast to calretinin-positive cells, not all of the calbindin-positive cells express GABAρ. At least two classes of calbindin containing cells have been described: spiny and aspiny neurons; further studies are needed to determine the identity and connectivity of the GABAρ-calbindin subpopulation. Again, the role of GABAρ receptors in these cells remains to be explored, but they may well confer tonic inhibition to these interneurons. In addition, we observed expression of GABAρ2 in a small population of MSNs that express Drd2-GFP. We speculate that tonic inhibition of these projection neurons may be mediated by GABAρ, suggesting an important role in selective signalling to the remaining MSNs.

Glial cells are known to express GABAA receptors both in culture and in situ, including in Bergmann cells of the cerebellum, astrocytes of the spinal cord, optic nerve, retina, hippocampus and pituitary gland (i.e. Berger et al. 1992; Müller et al. 1994; Verkhratsky and Steinhäuser 2000), whereas GABAρ2 receptors are known to be expressed in cerebellar astrocytes in culture (Martínez-Delgado et al. 2011). Here, we report that both the GABAρ1 and ρ2 GABA subunits are expressed in a population of GFAP-positive cells of the neostriatum; immunolabeling for GABAρ was identified mainly in processes and in smaller proportion in somas. What is the role of this receptor in the neostriatal glia? It may be that, similar to other GABAA subunits expressed in astrocytes, they modulate the release of neurotransmitter substances from glial cells and serve to coordinate neuronal electrical activity in response to GABA release. The GABAρ receptors may also be involved in the induction of morphological changes of the astrocytes, or in stimulating dendrite development (Matsutani and Yamamoto 1997; Matsutani and Yamamoto 1998; Mong et al. 2002).

The main special properties of GABAρ receptors are a high affinity for GABA, slow desensitization, activation by CACA, blockage by TPMPA, and insensitivity to bicuculline, which is a classic GABAA antagonist (Polenzani et al. 1991; Ragozzino et al. 1996). While recording from neurons in neostriatal slices, we observed some non-desensitizing GABA-currents which were sensitive to TPMPA, and CACA-activated currents that were also blocked by TPMPA. In some electrophysiological assays of CACA-currents, we recorded a small component that was TPMPA-insensitive. We presume that this current is due to the presence of receptors made up of both GABAρ subunits assembled independently or in combination with GABAA subunits, forming heteromeric complexes with mixed pharmacology (Pan et al. 2000; Hartmann et al. 2004; Milligan et al. 2004; Harvey et al. 2006; Frazao et al. 2007). However, this needs to be investigated further by a more elaborated pharmacological analysis in combination with subunit selective co-immunopreciptation.

GABAρ1 is found alone or in hetero-oligomeric complexes with GABAAα1 in rat brainstem, where the receptor is found extrasynaptically in the post-synaptic region, (Milligan et al. 2004). In the rat retina, GABAρ1 receptors have been found in both post-synaptic and extrasynaptic regions (Koulen et al. 1997) and extrasynaptic in the pre-synaptic region of Purkinje neurons of cerebellum (Mejía et al. 2008). However, the specific localization of GABAρ2 remains largely unknown. In the present study we show that this subunit is relatively more abundant than GABAρ1 both in pre- as well as post-synaptic regions, and both subunits are located in the same places (extra- and perisynaptic, but little or none in synaptic regions). Thus, due mostly to the perisynaptic location of GABAρ receptors as well as to their peculiar functional properties, it may be suggested that they play a central role during tonic inhibition. Tonic inhibition mediated by extrasynaptic GABAA subunits, such as those that include the α5, β3 and δ subunits, permits cells to sense the environmental levels and spillover of GABA released from synapses, and could protect or reduce excitotoxic injury and cell death through a persistent inhibition, allowing the regulation of neural network excitability and information processing (Barnard et al. 1998; Pirker et al. 2000; Schwarzer et al. 2001; Ade et al. 2008; Santhakumar et al. 2010; Janssen et al. 2011).

Concluding remarks

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Concluding remarks
  7. Acknowledgements
  8. References
  9. Supporting Information

The present results show that GABAρ receptors are functional in the mouse neostriatum and suggest that calretinin-positive, calbindin-positive, and Drd2 neurons as well as GFAP-positive cells express GABAρ receptors. These receptors are sparsely distributed, but with some preference for the rostral, ventral and dorsal regions. Inferences about their role in the neostriatum are supported merely by their known electrophysiological properties; however, we localized the population of cells that possess the receptor, thus providing the basis for further functional characterization.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Concluding remarks
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank A. E. Espino Saldaña, E. Ruiz Alcíbar, and I. Martínez-Dávila for their excellent technical assistance. M. L. Palma Tirado and E. N. Hernández Ríos provided support in electron and confocal microscopy, and A. Antaramian and A. Sánchez for DNA sequencing. Transgenic mice were donated by Dr V. Álvarez (NIH/NIAAA). A. Sánchez-Gutiérrez provided in situ hybridization support. We are indebted to Dr R. Gutiérrez (CINVESTAV) and Dr R. Arellano-Ostoa (INB-UNAM) for recommendations that contributed to this study; to Dr A. Cárabez-Trejo, Dr A. Varela-Echavarría, Dr J. Larriva-Sahd (INB-UNAM), and members of their laboratories for the facilities and suggestions; to Dr J. Heuser (Washington University, St Louis, MO, USA) for advice on immunogold techniques; and to Dr H. Kettenmann (Max Delbrück Center, Berlin, Germany) for the facilities to perform the electrophysiological recordings. The authors thank Dr D. D. Pless (INB-UNAM) for editing the manuscript. This work was supported by grants from PAPIIT-UNAM 202609-21 and 205308-21 (AM-T and RM), and CONACYT 101851 (AM-T). AR-A (189290) and AIM-P (210374) are recipients of fellowships from CONACYT-México. We thank the INB-UNAM, and Programa de Doctorado en Ciencias Biomédicas. AM-T acknowledges support from the A. & W. Shedid Fund.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Concluding remarks
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Concluding remarks
  7. Acknowledgements
  8. References
  9. Supporting Information

Table S1. Primer sequences for qRT-PCR.

Table S2. Primer sequences for cloning In situ hybridization probes.

Table S3. List of primary antibodies.

Table S4. List of secondary antibodies.

Figure S1. qRT-PCR amplification curves for GABAρ1 gene. Representative qRT-PCR amplification curves showing the PCR products for Actin, Tubulin and GABAρ1 in retina and neostriatum. Samples are indicated in colors.

Figure S2. qRT-PCR amplification curves for GABAρ2 gene. Representative amplification curves of the qRT-PCR showing the PCR products for Actin, Tubulin and GABAρ2 in retina and neostriatum. Samples are indicated in colors.

Figure S3. Representative traces of GABA (10 µM) evoked responses that were sensitive (A) or not sensitive (B) to TPMPA (100 µM). C. and D. Summary of experiments from A and B respectively. Data are mean ± S.E.M. Asterisks represent significant differences: P<0.01 (**)

Figure S4. A. CACA (500 µM) evoked currents were partially blocked by TPMPA (n = 8 responses from 2 neurons out of 5). B. Summary of experiments with CACA and the partial block of TPMPA. Data are mean ± S.E.M. Asterisk represent significant differences: P<0.05 (*).

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JNC_7621_sm_FigS1-S4-TableS1-S4.pdf399KSupporting info item
JNC_7621_sm_FigS1.tif442KSupporting info item
JNC_7621_sm_FigS2.tif406KSupporting info item
JNC_7621_sm_FigS3.tif279KSupporting info item
JNC_7621_sm_FigS4.tif75KSupporting info item
JNC_7621_sm_TableS1-S4.doc57KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.