Brain Research 1044 (2005) 266 – 271 www.elsevier.com/locate/brainres
Short communication
Learning-induced plasticity of cortical representations does not affect GAD65 mRNA expression and immunolabeling of cortical neuropil Monika Lech, Anna Skibinska, Ewa Siucinska, Malgorzata KossutT Department of Molecular and Cellular Neurobiology, Nencki Institute, 3 Pasteur Street, 02-093 Warsaw, Poland Accepted 24 February 2005 Available online 19 April 2005
Abstract Two forms of glutamic acid decarboxylase (GAD) are present in inhibitory neurons of the mammalian brain, a 65-kDa isoform (GAD65) and a 67-kDa isoform (GAD67). We have previously found that GAD67 is upregulated during learning-dependent plasticity of cortical vibrissal representations of adult mice. After sensory conditioning involving pairing stimulation of vibrissae with a tail shock, the increase in mRNA expression and density of GAD67-immunoreactive neurons was observed in barrels representing vibrissae activated during the training. In the present study, using the same experimental model, we examined GAD65 mRNA and protein levels in the barrel cortex. For this purpose, we used in situ hybridization and immunohistochemistry. No changes in the level of GAD65 mRNA expression were detected after the training. The pattern of GAD65 mRNA expression was complementary to that observed for GAD67. Immunocytochemical analysis found no changes in immunolabeling of neuropil of the barrels representing the vibrissae activated during the training. The results show that, in contrast to GAD67, cortical plasticity induced by sensory learning does not affect the expression of GAD65. D 2005 Elsevier B.V. All rights reserved. Keywords: GABA; Barrel cortex; Mice; Vibrissae
Cerebral cortex has the ability to undergo experiencedependent and learning-dependent modifications of sensory representations (for review, see [2]). Several experimental data show that the inhibitory neurotransmission system is an important component of this process. The level of inhibitory neurotransmitter, g-aminobutyric acid (GABA) was found to be a regulator of beginning of critical period for ocular dominance columns plasticity in mouse visual cortex [11]. It was also suggested that maturation of cortical GABA-ergic transmission contributes towards the termination of early postnatal plasticity [17]. Loss of sensory input downregulates glutamic acid decarboxylase (GAD, GABA synthesizing enzyme) immunoreactivity and the number of GABA-ergic synapses in the affected cortical region [10,16,25], while increased sensory stimulation leads to increases in GAD immunoreactivity [9,26] and increased number of GABA-ergic synapses [13]. T Corresponding author. Fax: +48 22 822 53 42. E-mail address:
[email protected] (M. Kossut). 0006-8993/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2005.02.077
Regulation of GABA synthesis is based upon kinetic features of glutamate decarboxylase, which has two isoforms (GAD65 and GAD67) [12], products of two separate, differently regulated genes [1]. In the central nervous system (CNS), both GAD isoforms are represented in various ratios depending on the structure studied [5] and have different subcellular localization and biochemical properties (for review, see [22]). It was suggested that GAD67 plays primarily a metabolic role, while GAD65 may be specialized for synaptic synthesis of GABA that is subject to rapid regulation; however, both forms were found in axon terminals [6]. It was demonstrated that in the hippocampus, axon terminals are specifically GAD67 or GAD65 immunoreactive [7], with GAD67-containing synapses localized around cell body and GAD65 in apical dendrites. Whisker-to-barrel pathway of rodents is a very good model to study changes in cortical plasticity. The representations of individual whiskers are barrel-shaped clusters of neurons in cortical layer IV. The arrangement of barrels in the somatosensory cortex mimics the arrangement of
M. Lech et al. / Brain Research 1044 (2005) 266–271
whiskers on the snout [28]. This topography makes it feasible to find specific changes in well-defined location in the brain. Previously, using 2-deoxyglucose (2DG) functional brain mapping, we showed that subjecting mice to simple associate learning paradigm: a classical conditioning, in which stimulation of a row of vibrissae (CS) was paired with a tail shock (UCS), causes a plastic change in the barrel cortex–cortical representation of the row of whiskers used during the training is enlarged [20]. This effect was seen 24 h after the end of the training. A band of heavy 2DG uptake, present over the row of barrels representing the row of whiskers stimulated during 2DG incorporation, was about 50% broader and encroached upon the barrels of adjacent rows, if this row of whiskers was activated during the learning task. Within the cortical barrels that receive input from the vibrissae activated during the training, we found increased density of GABA-IR neurons, increased immunolabeling with GAD67 and increased expression of its mRNA [9,21]. The present study complements the previous results— we examined GAD65 mRNA expression by in situ hybridization (ISH) and expression of GAD65 protein with immunocytochemistry in the barrel cortex of mice that were subjected to sensory pairing in a classical conditioning paradigm. We found that after training, there was no significant increase in GAD65 immunodensity in the cortical representation of the btrainedQ row of whiskers (row B of barrels) and no change in mRNA in situ signal. The experiment was performed on 57 young adult (7–8 weeks old during conditioning) Swiss female mice. The animals were cared for in accordance with The European Communities Council Directive (86/609/EEC). The training was done exactly as described in detail by Siucinska and Kossut [20]. The training consisted of stroking the whiskers of row B on the left side of the muzzle (conditioned stimulus, CS). At the last second of stroking, a single tail shock (UCS) was applied (0.5 s, 0.5 mA). The electrical stimulation was discontinued simultaneously with the end of stroking. Pairings were repeated four times per minute for 10 min a day for 3 days. Three groups of experimental animals were used. The first group (CS + UCS) received pairings for 3 days and the mice were sacrificed 1 h or 24 h later. The second group received only conditioned stimulus (CS) at the same schedule. The naive control group received no stimulation. For in situ hybridization, after completion of the training on indicated delay times mice were decapitated, brains rapidly removed and frozen on dry ice. Frozen hemispheres were cut in an oblique coronal plane (558 from the sagittal plane) throughout the barrel field across all barrel rows in parallel to the barrel arcs so that barrels belonging to individual rows could be identified. Slide-mounted and airdried sections were fixed and dehydrated in graded ethanol. Every 6th section was kept unfixed and stained for succinyl dehydrogenase (SDH).
267
Custom synthesized (TIB MOLBIOL) antisense oligonucleotide [5V-GGCGTCCACACTGCAAGGCCTTGTCTCCTGTGTCATAGGACAGGTCAT-3V] was used as a probe according to Ma [15]. Oligonucleotide was labeled by tailing 3Vend with a-tio-ATP (35S) (NEN) using TdT kit (Promega), according to Promega protocol. Probe was purified and the reaction yield was controlled by measuring incorporated radioactivity with liquid scintillation counter. In situ hybridization histochemistry was performed according to Wisden and Morris [27] as described by Gierdalski et al. [9]. Briefly, sections were incubated overnight at 42 8C with mix containing 1 Al labeled oligo probe, 100 Al dminimalistT hybridization buffer and 1 Al 1 M DTT per slide. After incubation, slides were washed, dehydrated in ethanol and air dried. Slides were exposed to mammography medical film (Kodak MH-5) for 4 weeks along with 14C standards (American Radiolabeled Chemicals). In order to achieve cellular resolution, sections after hybridization procedure were covered with nuclear emulsion Ilford K5. Sections were then developed in Kodak D19 developer and thionin stained. Autoradiograms on film were analyzed with a computer-assisted image analyzer MCID (Imaging Research Ontario, Canada). Measurements of optical density (OD) were taken in each section, in individual layer (II/III, IV, V) within cortical columns related to each row of barrels (A–E) and in each layer of cingulate cortex, chosen as a reference structure. Cortical layers and barrels were identified on sections stained for SDH. Data were normalized by calculating a ratio of labeling of the examined region of the barrel cortex to the reference structure. To test the difference between trained row B and other rows, one-way ANOVA test was used. Autoradiographic images obtained from control brains sectioned in oblique coronal plane showed heterogeneous expression of GAD65 mRNA throughout whole thickness of cerebral cortex (Fig. 1). The hybridization signal for GAD65 was concentrated in cortical layers II and V. Labeling of layer VI was weaker than in layer II and V and a little stronger then in layer IV. The cortical layer IV had less GAD65 mRNA signal than other cortical regions. In control animals, in each measured layer, there were no differences between labeling of the cortical columns belonging to different rows of barrels (A–E) or between hemispheres. No clear barrel-like pattern was observed. Control sections incubated with excess of homologous probe gave homogenous, very low signal. In situ hybridization with single cell resolution showed sparsely distributed heavily labeled cells (Fig. 2). Regularly spaced large neurons were present at the border of layer VI and white matter. In the barrel field, the signal was found most frequently in the lower part of layer IV barrels. Analysis of autoradiograms showed that there were no significant changes in mRNA expression level in cortical representation of row B activated during the training. The comparison was made between the barrel rows in the same section, between barrels of left and right hemispheres, as
268
M. Lech et al. / Brain Research 1044 (2005) 266–271
Fig. 1. In situ hybridization to GAD mRNA. Comparison of distribution of radioactive signal after in situ hybridization to GAD65 mRNA in mouse brain slices cut across barrel cortex, oblique coronal sections in mice after (a) conditioning (CS + UCS) and (b) sensory stimulation (CS). Note homogeneity of signal along layers within the barrel cortex. Arrows point to the region of posteromedial barrel sub-field. (c) ISH to GAD67 m RNA, section from the same mouse as (b) Note different distribution of label among cortical layers. Lower panels (e and f) show segments of autoradiograms from the cortex of (b) and (c) in higher magnification, for better comparison of laminar distribution of GAD65 and GAD67 signal. Roman numerals denote the position of cortical layers; (d) autoradiogram of control ISH with excess of homologous probe. Note the absence of labeling, showing specificity of the oligoprobe. GAD67 in situ hybridization was done with identical protocol as GAD65 with oligonucleotide probe described in Gierdalski et al. [8].
well as between trained and control mice. Similar results were found in the group of animals that received only conditioned stimulus (CS). In the cortical representation of row B, we found no changes in mRNA expression level, neither as compared between btrainedQ and bcontrolQ hemispheres, nor as compared with control animals. The results for the CS + UCS group and CS group did not differ, except for a tendency towards increased signal in the btrainedQ barrels 24 h after the end of conditioning in the CS + UCS group, which, however, did not reach statistically significant value (Fig. 3). The lack of changes was found in all cortical layers examined. GAD65 immunohistochemistry was performed at one time-point after training. Approximately 24 h after the end of 3 days of the conditioning training (CS + UCS), animals were anaesthetized with Nembutal (intraperitoneally) and immediately perfused via the ascending aorta with 1 M phosphate-buffered saline, followed by 4% paraformaldehyde in 0.1 M PBS, pH = 7.4. The brains were removed, post-fixed and cryoprotected. Left and right hemispheres were dissected. Each of them was cut into 35 Am thick sections in an oblique coronal plane (558 from the sagittal plane) on the cryostat and collected in PBS. Hemispheres from one control brain were flattened before freezing and sectioned tangentially to cortical surface. Every second section was used for cytochrome oxydase staining (CO) in order to visualize the barrels. The polyclonal antibody (Chemicon, Temecula, CA) was used for localization of the GAD65. To block non-specific
binding, the free-floating sections were kept in PBS buffer (pH 7.4) with 10% normal goat serum (NGS) for 1 h at room temperature. After 24 h of incubation with the primary antibody to GAD65 (1:1000), sections were incubated in biotinylated anti-rabbit IgG at a dilution of 1:200 in PBS (pH 7.4) for 2 h at room temperature, rinsed in PBS (pH 7.4) and then incubated in avidin–biotin complex reagent (ABC Kit; Vector Laboratories, Burlingame, CA) and DAB. Sections were analyzed using Nikon Eclipse microscope with Image ProPlus computerized system. Barrels from rows B and D from experimental and control hemispheres were examined under 20 magnification and measurements of optical density were taken from the entire barrels and from the barrel hollows. Ratios from B/D labeling in a single section were averaged within hemispheres. Optical density profiles were done across barrel rows (A to E) and across cortical layers from pia to white matter. GAD65 immunoreactivity in the barrel cortex was the strongest in layer IV; layers II and III were also heavily stained (Fig. 4a). Less staining was found in layer VI, the poorest immunostaining was observed in cortical layer V. GAD65 protein was detected in the cerebral cortex mostly as immunoreactive puncta representing probably GABAergic axon terminals and GABA-ergic processes cut across. There were also a few cell bodies covered with immunoreaction product (Fig. 4b) and the puncta-ring structures composed of GAD65 IR elements surrounding the soma and proximal dendrites of neurons (Fig. 4c). Semiquantitative analysis of immunolabeling revealed that intensity of row B
M. Lech et al. / Brain Research 1044 (2005) 266–271
269
Fig. 2. In situ hybridization of GAD65 mRNA on nuclear emulsion. Upper panel-dark field montage of the section 10 magnification; cells with high accumulation of silver grains visible throughout cortical depth and in the hippocampus. Roman numerals denote position of cortical layers. Lower panel—100 magnification of a layer IV barrel, light field micrograph.
Fig. 4. Immunocytochemistry of GAD65. (a) Barrel cortex, oblique coronal section. Note well visible barrels of rows E–A. Roman numerals denote cortical layers. (b) Tangential section through the barrel field, high magnification. A few cell bodies show IR product in the cytoplasm. Arrows point to outlines of row B barrel, delineated by unlabeled somata. (c) High magnification 100 showing immunoreactive fibers and puncta in the hollow of barrel.
Fig. 3. GAD65 mRNA expression in barrels of row B in control and experimental hemispheres expressed as a ratio of labeling of row B (the site of the plastic change) to mean labeling of DE (away from the plastic change) barrel rows.
labeling was slightly higher that in row D, both in trained mice in the experimental and control hemisphere and in naive controls. No changes were observed when OD reading for barrels B was compared to OD readings for other barrels in the same brain section (Fig. 5b). No changes in the number of the IR cells were seen, but their number was very low. No changes of immunoreactivity following sensory conditioning were found in any cortical layer. Comparison of present results with our previous in situ data for GAD67 mRNA in the mouse barrel cortex reveals that the mRNAs for the two GAD isoforms have a very
270
M. Lech et al. / Brain Research 1044 (2005) 266–271
Fig. 5. Optical density profiles from immunolabeled sections. y axis— relative optical density units (ROD), x axis—cortical distance. (a) Optical density profile across cortical layers. Width of the reading window— 100 Am. (b) Optical density profile across the five rows of barrels in layer IV. Width of the reading window—100 Am. A–E mark position of barrels.
different laminar distribution in the barrel cortex (Fig. 1). The highest density of hybridization signal of GAD67 mRNA was observed in layer IV. Similar result was illustrated by Zhang et al. [29]. On the other hand, GAD65 mRNA expression was high in layers II/III and V, but very low in layer IV. This complementary distribution pattern was visible not only in the barrel cortex, but throughout the cortex down to the rhinal sulcus at the examined anteroposterior location. Cellular distribution of the in situ signal was also different for GAD65 mRNA than it was described for GAD67, with many more heavily labeled neurons in extragranular layers in the case of GAD65 mRNA, while with GAD67 mRNA, layer IV had the highest density of labeled neurons [8]. This result corresponds well with the recent data from Lopez-Bendito et al. [14], who found that GAD65 neurons concentrate in upper cortical layers and do not express parvalbumin, which frequently colocalizes with GAD67 [24]. Thus, layer IV of the somatosensory cortex may contain a subset of interneurons that have only or predominantly one isoform of the enzyme—GAD67. GABA-ergic interneurons of layer IV are mostly basket cells and most of them contain parvalbumin [23]. GAD 65 IR laminar pattern was different from the GAD65 mRNA. The strongest GAD65 IR was seen in the barrels of layer IV (Fig. 4a), where the density of symmetrical synapses is the highest [4]. Concentration of immunolabeling in the upper cortical layers is similar to the
distribution observed in the parietal cortex of rats [5] and in visual cortex of cats [19]. The immunoreaction product was concentrated in neuropil, very few neuronal cell bodies showed immunolabeling (Fig. 4b). IR-labeling of the barrels of row B (representing the trained row of vibrissae) was not significantly higher than of row D in the same barrel field. Similar slight gradient was found in control mice. The semiquantitative estimations of the colored immunoreaction product do not allow more precise measurement. This raises the question of sensitivity of our measurement. Some validation of our measuring technique comes from examining laminar differences of immunolabeling and comparing them with EM data of synaptic density. The optical density profile across the barrels showed a clear difference in labeling between the septa and the barrels, but no systematic differences in immunolabeling of barrels of the btrainedb row B and barrels of the other rows (Fig. 5b). At the same time, the optical density profile across cortical layers showed clear differences in labeling density, with the peak in the middle of cortical layer IV (Fig. 5a), corresponding to the barrel hollow. The EM data on symmetrical synapses in mouse barrels find that the density of symmetrical synapses in barrel hollows is about 10% higher than in the septa and sides [18]. Thus, changes of the order of 10% could have been observed with our measurement technique. Similar comparison was done by Silver and Stryker [19] for GAD65 immunocytochemistry and numerical density of synapses in the visual cortex and a good fit of immunocytochemical fluorescence and synaptic density as estimated by EM was obtained. The lack of changes in immunolabeling of neuropil with anti-GAD65 antibody observed in our learning-induced cortical plasticity model resembles the lack of changes in GAD65 IR seen by Silver and Stryker in visual cortex after monocular deprivation [19]. But it presents a striking contrast with the data we obtained for GAD67, which showed upregulation of mRNA expression and increased immunoreactivity in barrels of row B following training. It is likely that the observed differences can be explained on the basis of structural features of GAD. The activity of this enzyme depends on association with cofactor: pyridoxal 5Vphosphate. It has been demonstrated that GAD67 is present in neurons as an active form of enzyme, bound with cofactor (holoenzyme). GAD65 is present in axon terminals as an apoenzyme—an inactive form of enzyme, and it needs cofactor for activation [3]. Consequently, activity of GAD65 can be regulated by accessibility of cofactor, while increase of GAD67 can be achieved only by synthesis of additional GAD67 molecules. We demonstrated that learning-induced plasticity of cortical vibrissal representation is not accompanied by an increase in the level of GAD65 in neuropil of layer IV barrels or by changes in GAD65 mRNA expression. The results of this study show also that the distribution of GAD65 mRNA expression in the somatosensory cortex forms a pattern complementary to that observed for
M. Lech et al. / Brain Research 1044 (2005) 266–271
GAD67 mRNA expression, suggesting that not all inhibitory interneurons contain both isoenzymes.
Acknowledgments The research was supported by grant from Foundation for Polish Science 4/2000 and The Wellcome trust Collaborative grant to MK/SG and grant KBN State Committee for Scientific Research grant #23023 to E.S. We thank Mrs Renata Zakrzewska for skilled technical assistance and Dr. Stanislaw Glazewski for helpful discussion and comments on the manuscript.
References [1] D.F. Bu, M.G. Erlander, B.C. Hitz, N.J. Tillakaratne, D.L. Kaufman, C.B. Wagner-McPherson, G.A. Evans, A.J. Tobin, Two human glutamate decarboxylases, 65-kDa GAD and 67-kDa GAD, are each encoded by a single gene, Proc. Natl. Acad. Sci. U. S. A. 89 (1992) 2115 – 2119. [2] M.B. Calford, Dynamic representational plasticity in sensory cortex, Neuroscience 111 (2002) 709 – 738. [3] C.-H. Chen, G. Battaglioli, D.L. Martin, S.A. Hobart, W. Colon, Distinctive interactions in the holoenzyme formation for two isoforms of glutamate decarboxylase, Biochim. Biophys. Acta 1645 (2003) 63 – 71. [4] J. De Felipe, P. Marco, A. Fairen, E.G. Jones, Inhibitory synaptogenesis in mouse somatosensory cortex, Cereb. Cortex 7 (1997) 619 – 634. [5] M. Esclapez, N.J.K. Tillakarante, D.L. Kaufman, A.L. Tobin, C.R. Houser, Comparative localization of two forms of glutamic acid decarboxylase and their mRNAs in rat brain supports the concept of functional difference between the forms, J. Neurosci. 14 (1994) 1834 – 1855. [6] S. Feldblum, M.G. Erlander, A.J. Tobin, Different distributions of GAD65 and GAD67 mRNAs suggest that the two glutamate decarboxylases play distinctive functional roles, J. Neurosci. Res. 34 (1993) 689 – 706. [7] T. Fukuda, Y. Aika, C.W. Heizmann, T. Kosaka, GABAergic axon terminals at perisomatic and dendritic inhibitory sites show different immunoreactivities against two GAD isoforms, GAD67 and GAD65, in the mouse hippocampus: a digitized quantitative analysis, J. Comp. Neurol. 395 (1998) 177 – 194. [8] M. Gierdalski, B. Jablonska, A. Smith, J. Skangiel-Kramska, M. Kossut, Deaferentation induced changes in GAD67and GluR2 mRNA expression in mouse somatosensory cortex, Mol. Brain Res. 71 (1999) 111 – 119. [9] M. Gierdalski, B. Jablonska, E. Siucinska, M. Lech, A. Skibinska, M. Kossut, Rapid regulation of GAD67 mRNA and protein level in cortical neurons after sensory learning, Cereb. Cortex 11 (2001) 806 – 815. [10] S.H. Hendry, E.G. Jones, Reduction in number of immunostained GABAergic neurones in deprived-eye dominance columns of monkey area 17, Nature 320 (1986) 750 – 753. [11] Y. Iwai, M. Fagiolini, K. Obata, T.K. Hensch, Rapid critical period induction by tonic inhibition in visual cortex, J. Neurosci. 23 (2003) 6695 – 6702.
271
[12] D.L. Kaufman, C.R. Houser, A.J. Tobin, Two forms of the gammaaminobutyric acid synthetic enzyme glutamate decarboxylase have distinct intraneuronal distributions and cofactor interactions, J. Neurochem. 56 (1991) 720 – 723. [13] G.W Knott, C. Quairiaux, C. Genoud, E. Welker, Formation of dendritic spines with GABAergic synapses induced by whisker stimulation in adult mice, Neuron 34 (2002) 265 – 273. [14] G. Lopez-Bendito, K. Sturgess, F. Erdelyi, G. Szabo, Z. Molnar, O. Paulsen, Preferential origin and layer destination of GAD65-GFP cortical interneurons, Cereb. Cortex 14 (2004) 1122 – 1133. [15] W. Ma, T. Behar, L. Chang, J.L. Barker, Transient increase in expression of GAD65 and GAD 67 mRNAs during postnatal development of rat spinal cord, J. Comp. Neurol. 346 (1994) 151 – 160. [16] K.D. Micheva, C. Beaulieu, An anatomical substrate for experiencedependent plasticity of the rat barrel cortex, Proc. Natl. Acad. Sci. U. S. A. 92 (1995) 11834 – 11838. [17] C. Rozas, H. Frank, A.J. Heynen, B. Morales, M.F. Bear, A. Kirkwood, Developmental inhibitory gate controls the relay of activity to the superficial layers of the visual cortex, J. Neurosci. 21 (2001) 6791 – 6801. [18] Y. Sadaka, E. Weinfeld, D.L. Lev, E.L. White, Changes in mouse barrel synapses consequent to sensory deprivation from birth, J. Comp. Neurol. 457 (2003) 75 – 86. [19] M.A. Silver, M.P. Stryker, Distributions of synaptic vesicle proteins and GAD65 in deprived and nondeprived ocular dominance columns in layer IV of kitten primary visual cortex are unaffected by monocular deprivation, J. Comp. Neurol. 422 (2000) 652 – 664. [20] E. Siucinska, M. Kossut, Short-lasting classical conditioning induces reversible changes of representational maps of vibrissae in mouse SI cortex—a 2DG study, Cereb. Cortex 6 (1996) 506 – 513. [21] E. Siucinska, M. Kossut, M.G. Stewart, GABA immunoreactivity in mouse barrel field after aversive and appetitive classical conditioning training involving facial vibrissae, Brain Res. 843 (1999) 62 – 70. [22] J.J. Soghomonian, D.L. Martin, Two isoforms of glutamate decarboxylase: why? TIPS 19 (1998) 500 – 505. [23] J.F. Staiger, K. Zilles, T.F. Freund, Distribution of GABAergic elements postsynaptic to ventroposteromedial thalamic projections in layer IV of rat barrel cortex, Eur. J. Neurosci. 8 (1996) 2273 – 2285. [24] N. Tamamaki, Y. Yanagawa, R. Tomioka, J. Miyazaki, K. Obata, T. Kaneko, Green fluorescent protein expression and colocalization with calretinin, parvalbumin, and somatostatin in the GAD67-GFP knockin mouse, J. Comp. Neurol. 467 (2003) 60 – 79. [25] E. Welker, E. Soriano, H. Van der Loos, Plasticity in the barrel cortex of the adult mouse: effects of peripheral deprivation on GADimmunoreactivity, Exp. Brain Res. 74 (1989) 441 – 452. [26] E. Welker, E. Soriano, J. Dorfl, H. Van der Loos, Plasticity in the barrel cortex of the adult mouse: transient increase of GADimmunoreactivity following sensory stimulation, Exp. Brain Res. 78 (1989) 659 – 664. [27] W. Wisden, B.J. Morris, In situ hybridization with synthetic oligonucleotide probes, in: W. Wisden, B.J. Morris (Eds.), In Situ Hybridization Protocols for the Brain, Academic Press, London, 1994, pp. 9 – 34. [28] T.A. Woolsey, H. Van der Loos, The structural organization of layer IV in the somatosensory region (SI) of mouse cerebral cortex. The description of a cortical field composed of discrete cytoarchitectonic units, Brain Res. 17 (1970) 205 – 242. [29] L. Zhang, D.R. Rubinow, W. Ma, J.M. Marks, A.N. Feldman, J.L. Barker, T.A. Tathan, GABA receptor subunit mRNA expression in brain of conflict, yoked control and control rats, Mol. Brain Res. 58 (1998) 16 – 26.