Rosas Et Al 2011 GABArho

8/2/2019 Rosas Et Al 2011 GABArho

1/6

Neuroscience Letters 500 (2011) 2025

Contents lists available at ScienceDirect

Neuroscience Letters

journ al h omepage: www.elsevier .com/locate /neulet

The GABA(A) receptors in hippocampal spontaneous activity and theirdistribution in hippocampus, amygdala and visual cortex

Abraham Rosas-Arellanoa, Jorge Parodib, Arturo I. Machuca-Parraa, Adriana Snchez-Gutirrezc,Nibaldo C. Inestrosab, Ricardo Miledia, Atalfo Martnez-Torresa,

a Departamento deNeurobiologaCelulary Molecular,Laboratorio deNeurobiologaMoleculary Celular, InstitutodeNeurobiologa, Campus UNAM-Juriquilla,Quertaro,QRO 76230,

Mexicob Centro de Envejecimiento y Regeneracin (CARE), Departamento de Biologa Celular y Molecular, Facultad de Ciencias Biolgicas, Pontificia UniversidadCatlica de Chile, Santiago,

Chilec Departamento de Neurobiologa del Desarrolloy Neurofisiologa, Instituto de Neurobiologa, Campus UNAM-Juriquilla,Mexico

a r t i c l e i n f o

Article history:

Received 31 March2011

Received in revised form 13May2011

Accepted 31 May2011

Keywords:

GABAPatch-clamp

TPMPA

Amygdala

Visual cortex

a b s t r a c t

A bicuculline-resistant and TPMPA-sensitive GABAergic componentwas identified in hippocampal neu-

rons in culture and in acute isolated brain slices. In both preparations, total GABAergic activity showed

two inactivation kinetics: fast and slow. RT-PCR, in situ hybridization (ISH) and immunohistochemistry

detected expression ofGABA subunits. Immunogoldand electronmicroscopy indicated that the recep-tors are mostly extrasynaptic. In addition, by RT-PCRand immunofluorescencewe found GABA presentin amygdala and visual cortex.

2011 Elsevier Ireland Ltd. All rights reserved.

Hippocampal neurons in culture develop a series of

neurotransmitter-activated ionic conductances mediated mainly

byglutamate and-aminobutyricacid(GABA) receptors. In severalspecies, the functional and pharmacological properties of those

receptors have been studied at different development stages,

including electrophysiological recordings from isolated neurons

in culture and slices e.g. [13,22]. The molecular conformation of

ionotropicGABA receptorsexpressed in hippocampus corresponds

to a combination ofseveral GABA-A subunits, ofwhich the1, 2and 2 are the most abundant, whereas 5 and subunits conferuniquepharmacological propertiesto the receptor [8,25]. Thus, the

physiological characteristics ofthe receptors expressed are largely

determinedby the subunit composition ofthe receptorsexpressed

in the neurons at anygiven time and experimental condition.The ionotropic GABA receptorswerefirst identifiedat the ver-

tebrate retina [20], where they play a central role in modulating

presynaptic inhibition at the axon terminal ofthe ON bipolar cells

[24]. Currently, it is known that GABA receptors are present inseveral areas ofthe brainwhere their functional role is still elusive

[1,14,15,23].

Corresponding author.

E-mail address: [email protected] (A. Martnez-Torres).

Several studies have described GABA responses that areclearly evident at early developmental stages ofthe hippocam-

pus. Cherubinis group [6] elegantly defined the pharmacological

and single channel properties ofthose responses while recording

from neurons isolated from the CA3 area during the early post-

natal development [9]. Further investigations indicate thatGABAreceptorsmay,ormaynot formpart offunctionalsynapses in adult

hippocampus. For example, Chenget al. [4] found that the GABA1subunitexpressedviaadenoviraltransductionis notproperlydeliv-

ered to the synapses in culture hippocampal neurons; whereas,

Alakuijala et al. [1] described that GABA receptors are mainlyextrasynaptic andmostprobablyactivated by GABAspillover from

the synaptic cleft in CA1 neurons.

While recording the synaptic activity in mouse hippocampalpyramidal neurons in culture, it came to our attention that a

small component of spontaneous activity was not blocked even

in the presence of a combination of tetrodotoxin (TTX), bicu-

culline, Mg2+ and 6-cyano-7-nitroquinoxaline-2, 3-dione (CNQX).

Thisreportshows thatthis GABAergiccomponentcanbeattributed

to the expression ofGABA receptors. Additionally, we show thata slow-inactivating GABA-component is also present in hippocam-

pal slices. BymeansofRT-PCR, ISHandimmunohistochemistry;we

detected the GABA subunits in this area. Furthermore, electronmicroscopy disclosedtheir presence in specific synaptic regions. In

addition, we localized the receptors in amygdalaandvisual cortex.

0304-3940/$ see frontmatter 2011 Elsevier Ireland Ltd. All rights reserved.

doi:10.1016/j.neulet.2011.05.235

8/2/2019 Rosas Et Al 2011 GABArho

2/6

A. Rosas-Arellano et al. / Neuroscience Letters 500 (2011) 2025 21

In the present study, we used CD1 and C57BL/J6 mice. All the

animals were handled in accordance with the NIH Guide for Care

and Use ofLaboratory Animals and approved by the Institutional

Animal Care and ethic committee ofthe INB-UNAM.

Hippocampal neurons were obtained from 18 day embryos

(C57BL/J6). Whole-cell patch clamp recordings of12DIV neurons

were performed as in previous report [19]. Hippocampal slices

were prepared according to previously published standard pro-

cedures from 22 to 30 day old C57BL/6 mice and used ACFS for

external solution [3]. Neuron membrane currents were recorded

by whole-cell patch clamp using the Axopatch-200B amplifier

(Axon Instruments, Inc., Burlingame, CA). The membrane poten-

tial was clamped at 60mV and the currents were recorded

every 50s and filtered at 2kHz. TTX (100nM), was added to theexternal solution to block evoked synaptic activity, thus allow-

ing the detection ofthe spontaneous miniature synaptic currents

(mIPSCs and mEPSCs); CNQX (4M) and bicuculline (10M)blocked glutamatergic and most of the GABAergic components,

respectively. In this study 5M, of 1,2,5,6-tetrahydropyridin-4-

yl-methylphosphinic acid (TPMPA) was used for blocking GABAtransmission.

RNA was isolated from retina, hippocampus, visual cortex and

amygdala using TRIzol reagent (Invitrogen), followingmanufac-

turers instructions. Total RNA was reverse transcribed into cDNA

and PCR was performed using gene-specific primers (Table S1).

Probes for ISHwere isolatedbyRT-PCR(Table S2). The cDNAswere

cloned into pGEM-T-Easy (Promega) and sequenced. In vitro tran-

scriptionwas performed following themanufacture specifications

using digoxigenin-11-UTPs (Invitrogen).

For ISH, CD1 mice (2530g) were anesthetized with pen-

tobarbital, perfused transcardially with 0.9% NaCl, and 4%

paraformaldehyde in phosphate buffer saline (PBS). The brain

and ocular globes were removed, postfixed, cryoprotected and

then 150m slices were obtained in cryostat (LeicaCM 1850).Hybridization was performed using the method described by the

manufacturer (Roche). The slices were placed (Superfrost/Plus

DAIGGER), mounted (Fluoromont G SouthernBiotech, Electron

Micoscopy Sciences), observed and photomicrographed (Olympus

Fig. 1. (A) Culture hippocampal neurons recordings. Miniature spontaneous activity recording in the presence ofTTX and Mg2+ . GABA ionotropic receptors are eliminated

with bicuculline (Bic) and TPMPA. GABA-activity is revealed with TTX, Mg2+ and CNQX. Most spontaneous activity is wiped out with bicuculline; however some activity

remains (inset).Remainingactivitywas eliminatedpartiallywith TPMPA, and totallywithTPMPA-bicuculline. (B) Frequencyofspontaneous events due to GABAor glutamate

receptors; inset: traces aresample traces (withoutfiltering)offast andslowcomponents.(C) GABA subunitsin hippocampal slices.Evidenceofthe recordingsofslowandfastGABA-components and partial (TPMPA or bicuculline) and total blocking (TPMPA and bicuculline). (D) Frequencies ofsynaptic after blocking the glutamatergic component

(bar GABA) evidencing the GABA receptors activity. TPMPA reduced to 61.32.1% ofthe GABA-currents. The inset shows two sample-currents ofeither fast or slow decay

time.

8/2/2019 Rosas Et Al 2011 GABArho

3/6

22 A. Rosas-Arellano et al. / Neuroscience Letters 500 (2011) 2025

BX60 microscope). Sense probes did not show any label, the anti-

sense probes specificity was tested in retina (not shown).

Immunohistochemistrywasperformedas previouslydescribed,

in 30m slices ofwhole brain and ocular globes [23]. For imag-ing, weused Zeiss LSM510 Metaconfocal microscope;wavelength

of 561nm (Alexa 594) and 750nm (DAPI) were used for excita-

tion ofAlexa 594 andDAPI, respectively. Thez-stack imageswere

obtained and processed in Aim Image Examiner. The antibodies

specificity waspreviously tested in retinaandSTC-1cells (data not

shown).Finally, immunogold labelingwas performedas previously

described [15].

Spontaneous synaptic activitywas recorded from12 pyramidal

neurons thatweremaintained in vitro for 12 days. Thefrequencyof

spontaneous miniature synaptic currents was about 1.8Hz (2min

ofrecording per cell) when TTX (100nM) and Mg2+ (2mM), were

added to themedium(Fig. 1A). After addingCNQX (4M) to blockglutamate receptors, the spontaneous miniature synaptic activ-

ity due to GABA receptors was evidenced. This GABA activity was

mostly abolished when 10M bicuculline was also included inthe bath to inhibit GABA-A receptors. Nevertheless, a detailed

analysis ofthe recordings revealed that in 9 neurons some spon-

taneous activity remained (0.29Hz), standard error of the mean

(S.E.M.0.11). Whenwe combined TTX, Mg2+, CNQX, TPMPA and

bicuculline the spontaneous activity was eliminated. The totalGABAergic synaptic activity was generated by two different com-

ponents (Fig. 1B, inset): (a) a frequent component that made up

to the 90% ofthe GABA-activity and that showed a fast inactiva-

tionconstant (=494ms) and (b)a slow inactivating component

(=11421ms) that correspond to 111% ofthe GABA-activity

(Fig. 1B).

Pyramidal cells ofhippocampal slices exposed to TTX (100nM)

and Mg2+ (2mM) showed classical miniature synaptic activity

(1.90.15Hz) (Fig. 1C). Addition ofTTX (100nM), Mn2+ (2mM)

and CNQX (1M) to the recording bath revealed clearly the

GABAergic component which exhibited two kinetics: (1) a fre-

quent fast inactivating component (=304.2ms) that made up

about 83% ofthe GABAergic transmission. (2) A slow inactivating

(=12018ms) that formed the remaining17%. These two kinet-

ically different components resemble those observed in cultured

pyramidal neurons (c.f. 1B) andsample tracesofthese twocompo-

nents and their frequencies are comparedwith the total miniature

synaptic activity (Fig. 1D). Additionally, 5M of TPMPA wipedout 17% ofthe GABA activity (0.50.1Hz), which was recovered

after washing out the compound and substitution ofTPMPA with

10M bicuculline eliminated most of the GABAergic componentand evidenced solely GABA activity (0.110.02Hz). CombiningTPMPAandbicuculline inCNQXabsenceshowed theglutamatergic

activity (1.20.17Hz). Finally, the presence ofCNQX in combi-

nation with GABA-A and GABA blockers eliminated most of theminiature synaptic activity. Comparative analysis of the ampli-

Fig. 2. GABAmRNA expression. (A) RT-PCR GABA expression in retina and hippocampus by (B and H) ISH for GABA1 and 2 respectively, in subiculum (S), CA1-3, anddentate gyrus (DG). (CandI) GABA1 and 2 are present in subiculum. (DF)GABA1 andJL.GABA2mRNAs inCA1-3 in pyramidal layer (Py), oriens layer(Or),and stratum

radiatum (Rad). (G and M) In DG are localized in granular (GrDG), polymorph (PoDG), andmolecular layer (Mol).

8/2/2019 Rosas Et Al 2011 GABArho

4/6

A. Rosas-Arellano et al. / Neuroscience Letters 500 (2011) 2025 23

tudes ofthe fast and slow kinetics were not significantly different

in neurons (524pA and 43.77 for the fast and slow com-

ponent, respectively, ANOVA p>0.05) or in slices (536pA for

the fast and 427pA for the slow, ANOVA p>0.05). Consider-

ing that the remaining synaptic activity suggested the presence

of a slow-inactivating TPMPA-sensitive, we therefore tried to

determine which of the GABA subunits are expressed in thehippocampus.

RNA was isolated from 20 retinas and from 20 hippocampi of

30 days old mice. In independent preparations, we found that all

the GABA subunits were amplified (Fig. 2A). Even though thisassaywasnot quantitative it seemed that the GABA3 subunitwasless represented in the hippocampus, and we therefore decided

to concentrate on describing the GABA1 and GABA2 distribu-tion.

Four brains in sagittal sliceswere used for ISH. Wedetected the

expression ofGABA1 and GABA2 in CA1-3 areas, dentate gyrusaswell as subiculum(Fig. 2B and H). In the subicullum, theexpres-

sion ofboth subunits was found in sparsely cells (Fig. 2C and I).

In CA1, CA2andCA3the label was foundmostly in pyramidal neu-

ronswhereasa smallpopulationofinterneuronsalso showed some

mark (Fig. 2DF and JL); whereas, GABA2 was also observed in

the processes ofpyramidal cells ofCA1. Finally, in dentate gyrus

most of the label was detected towards the granular, molecular

and polymorph layers (Fig. 2G and M). In four brains, immunoflu-

orescencerevealed thatGABA1was detected in CA1CA2,mostly

Fig. 3. GABA immunolocalization. (A and B)Distribution in pyramidal layer (Py), stratum radiatum (Rad)andoriens layer(Or) in CA1.(C and D) In CA2 in PyandRad.(E andF) In CA3 GABA1 in Or andRad, GABA2 in Pyand Or. (Gand H)Expressionin subiculum. (I andJ) In dentate gyrus(DG) bothsubunits are present in polymorph(PoDG), andmolecular (Mol) layers. (K) At ultrastructural localizationGABA1 was foundmainly in extrasynaptic sites (arrowheads). The arrows point towards postsynaptic densities.

(L) GABA1 in synaptic cleft (arrowheads). (M) GABA2 (arrowheads) outside ofsynaptic density.Abbreviations: At, axon terminal; Den, dendrite.

8/2/2019 Rosas Et Al 2011 GABArho

5/6

24 A. Rosas-Arellano et al. / Neuroscience Letters 500 (2011) 2025

Fig. 4. GABA expression and localization in amygdala and visual cortex. (A) RT-PCR ofGABA expression. (B, C, E and F) Immunolocalization ofboth subunits in the samenuclei ofamygdala. (D) Immunofluorescence ofGABA1 was detected mostly in somas. (G) Whereas GABA2 was localized in soma and processes in II and III layers, andsparsely distributed in V and VI layers (not shown).Abbreviations: BM, basomedial nucleus; PMCo,posteromedial cortical nucleus.

in the neuron processes, several somas ofinterneurons and pyra-

midal layer ofCA1, CA2 (Fig. 3A and C), while only a few cells and

processeswere positive in CA3 (Fig. 3E). The GABA2 was detected

mostly in somas andsome processesofpyramidal layer ofCA1 andCA2. In CA3 its distribution waswider that GABA1 (Fig. 3B, D andF).

The two subunitswere also found in the subiculum(Fig. 3G and

H). In dentate gyrus, both subunits are distributed in somas ofthe

polymporph and molecular layers (Fig. 3I and J); we did not find

evidence of their presence in neither the somas or processes of

the granular neurons, where we observed the expression oftheir

mRNAs.

The fluorescence was localized to puncta in the hippocam-

pal neurons, suggesting that the receptors are at the synapses.

To examine this question we used immunogold and electron

microscopy. We found most of the label at extrasynaptic sites,

either in the plasma membrane or near the cell surface, and only

3 out ofmore of20 postsynaptic densities showed some label foreither GABA1 (Fig. 3K and L) or GABA2 (Fig. 3M).

Additionally, we found some label for GABA receptors byimmunofluorescence in anterior and posterior areas ofthe baso-

medial nucleus, as well as in the posteromedial cortical nucleus

in the amygdala (Fig. 4B, C , E and F), and towards the mediomedial

area ofthe secondaryvisual cortex (V2MM) (Fig.4D andG). RT-PCR

assays confirmedtheexpressionofthe three subunits (Fig. 4A).ISH

confirm thismRNA expression in the same areas ofamygdala and

in visual cortex (datanot shown).

As in previous studies we found very little spontaneous synap-

tic activity in mice pyramidal neurons exposed to antagonists of

GABA-A and glutamate receptors [13,18]. At first, we presumed

that this activity was due to an incomplete block ofthe receptors

expressedbytheneuronsinculture.However,an alternativeexpla-

nation immediately arose, when we obtained partial blockage of

the GABA-currents by including TPMPA, a good blocker ofGABAreceptors [21]. This result suggestedthatthe spontaneoussynaptic

activitywhichremainedafter thecombinedexposureto bicucullineand CNQXwas due to the action ofGABA receptors.

Protocols ofstimulus-evoked postsynaptic potentials have sug-

gested the presence of GABA in perisynaptic regions of thehippocampus CA1 area [1]. It is also known that GABA-A recep-

tors containing the5 subunitmediate thephasic inhibition inCA1[26,29]. However, sincewe foundthatpyramidalneuronsgenerate

a TPMPA-sensitive component, complementary to themore abun-

dant bicuculline-blocked GABA-A receptors, a role for GABA intonic inhibition should not be discarded.

The presence ofGABA receptors in CA1 perisynaptic regionshas been documented [1]. These observations are partially con-

sistent with our findings, since immuonogold revealed most of

the label at extrasynaptic sites; although additional label was also

found in postsynaptic densities. This suggests a potential role forGABA receptors in fast synaptic transmission either as indepen-dent receptors or forming heteromeric complexes with others

GABA-A subunits, such as those reported in brainstem [16], and

in CA1 [13]. On the other hand, althoughwe observed the expres-

sion ofthe GABA mRNAs along the granular layer of DG, theimmunofluorescence did not detect the receptors, thus suggest-

ing the possibility that the proteins are expressed at much lower

levels [16]. It is worth mentioning that the GABA-A 1 subunit hasbeen found in both synaptic and extrasynaptic sites [2], ofpyra-

midal neurons ofCA1 and this subunit is known to associatewith

GABA1 [11,13,16].The functional propertiesofthe GABA receptors in the subicu-

lum need to be investigated further. We have not found evidence

in the literature of electrophysiological recordings indicating a

8/2/2019 Rosas Et Al 2011 GABArho

6/6