Abstract

We established a model of cortical development that arrests the birth of layer 4 cells by injecting methylazoxymethanol (MAM) on embryonic day 33 (E33) in ferrets. This leads to adult somatosensory cortex with a very thin layer 4. Earlier, we determined the relative absence of layer 4 changed the growth and differentiation of the somatosensory cortex and the growth of thalamic afferents into the cortical plate. To identify other features of cortical organization that might be altered after MAM treatment, we assessed the distribution of selected excitatory and inhibitory receptors in area 3b of ferret somatosensory cortex. Initial screening revealed the distribution of several excitatory receptors (NMDA, AMPA, kainate) in E33 MAM-treated cortex was similar to that in normal adult animals. In contrast, the binding pattern of inhibitory GABAA receptors was altered in MAM-treated cortex. Normally, GABAA receptors densely locate in central layers of cortex. In E33 MAM-treated animals, GABAA receptor binding extended superficially, covering a broader area of cortex. Further experiments using antibodies directed against GABAAα receptors disclosed that pan α GABAA receptors strongly localize to layer 4 in normal area 3b. In E33 MAM-treated cortex, however, GABAAα receptors extend outside and are located above and below the very thin layer 4. The redistribution of inhibitory receptors suggests that layer 4 plays an important role in regulating thalamic terminations and also in the resulting ability to refine processing of incoming stimuli.

Introduction

The pathological mechanisms leading to improper migration of cortical neurons and abnormal formation of cortical layers may cause significant functional alterations. We developed a cortical malformation model of epigenetic origin using an antimitotic agent that interrupts normal development of cortical layers (Noctor et al., 1997, 2001). Methylazoxymethanol (MAM) injected into pregnant ferrets prevents mitosis for a short time. When injected on a single day of corticogenesis it produces highly specific deficits in the normal laminar pattern of the neocortex in newborn ferrets (Cattaneo et al., 1995; Noctor et al., 1997, 2001; Palmer et al., 2001). Noctor et al. (2001) established that injection of MAM on embryonic day 33 (E33) in ferrets interferes with neuronal proliferation and results in a reduction of cortical thickness, especially due to a very thin layer 4. Although many features of cortical organization appear relatively normal after this treatment, such as the somatotopic organization and the dimension and cell size of the remaining cortical layers, the distribution of thalamic afferent terminations is severely disrupted in E33 MAM-treated ferret somatosensory cortex (Palmer et al., 2001). The abnormal thalamic terminations suggest that processes of information transmission are impaired throughout cortical layers.

Because of the likely involvement of excitatory and inhibitory receptors in thalamocortical transmission, we examined whether E33MAM treatment also affected the distribution of markers of excitatory and inhibitory systems in cerebral cortex. The distribution of specific receptors in somatosensory cortex is very similar in many adult species. They concentrate in distinct cortical layers with specific patterns, depending on the receptor (Insel et al., 1990; Chalmers and McCulloch, 1991; Jaarsma et al., 1991; Young et al., 1991; Kumar et al., 1994; Glazewski et al., 1995; Gordon et al., 1996). The precise arrangement of excitatory and inhibitory receptors in cerebral cortex suggests that their specific distribution in cortical laminae may play important roles in the processing of information in each layer and the flow of information through the cortex.

To define whether disruption of layer 4, the main target of thalamic input, alters the distribution of receptors involved in transmission of cortical sensory information, we evaluated the laminar distribution of NMDA, AMPA, kainate and GABAA receptors. We compared this distribution in the somatosensory cortex of normal and MAM-treated ferrets. Although the distribution of excitatory receptors did not vary substantially between normal and MAM-treated animals, the pattern of GABAA receptors was altered in MAM-treated somatosensory cortex and extended outside of layer 4, especially in the upper layers. Our results imply that MAM treatment at embryonic day 33 alters the distribution of inhibitory receptors in somatosensory cortex, leaving excitatory markers unchanged.

Materials and Methods

MAM Treatment and Tissue Preparation

Adult ferrets were obtained from Marshall Farms (New Rose, NY). At E33, pregnant ferrets were anesthetized with 2% halothane and received an i.p. injection of MAM (12 mg/kg). Each animal was monitored carefully until it recovered, when it was returned to its cage. Both normal (n = 12) and MAM-treated (n = 13) animals were used either at various postnatal ages [postnatal day 0 (P0), P14, P28] for immunocytochemistry studies or at P36 for in vitro receptor binding autoradiography. For immunohistochemical experiments, ferrets were deeply anesthetized with pentobarbital (50 mg/kg, i.p.) and perfused with normal saline followed by 4% paraformaldehyde with 4% sucrose in 0.1 M sodium phosphate buffer (pH 7.4) at room temperature. The brains were immediately removed, cryoprotected in 20% sucrose overnight at 4°C, then frozen in isopentane and kept in a –80°C freezer. The brains were cut at 30 µm thickness sections on a cryostat. For the animals used in receptor binding studies, each ferret was anesthetized with pentobarbital (50 mg/kg i.p.) and when insensitive to pain, the brain removed and quickly frozen in isopentane at –35°C. Alternate sections through the somatosensory cortex were cut and saved for Nissl staining or autoradiography.

In Vitro Receptor Binding Autoradiography

Steady-state binding assays were performed by incubating the sections (taken from P36 animals) with the appropriate radioligand to identify NMDA receptors ([3H]MK-801), kainate receptors ([3H]kainic acid), AMPA receptors ([3H]fluorowillardine), or GABAA receptors ([3H]muscimol). NMDA and kainate receptor binding were performed as previously described (Jablonska et al., 1996, 1998). Prior to incubation with the appropriate ligand the brain sections were washed with a specific incubation buffer (see below) at 4°C to remove endogenous ligands, for either 10 min with [3H]MK-801 binding and [3H]muscimol binding or 10 min × 3 for [3H]fluorowillardine and [3H]kainate binding. After this, the sections used to identify NMDA receptors were incubated with 3 nM [3H]MK-801 (32.75 Ci/mmol, Tocris), 5 µM glycine, 5 µM glutamate, 50 µM spermidine in 50 mM Tris–HCl buffer (pH 7.4, 20°C) for 1 h. Non-specific binding was determined in the presence of 100 µM MK801. The slides were subsequently rinsed three times with ice-cold Tris–HCl (pH 7.4) for 15 s, twice in ice-cold water (5 s) and finally rapidly dipped in two changes of acetone containing 2.5% glutaraldehyde for 15 s. Kainate receptors were labeled with 5 nM [3H]kainate (58 Ci/mmol, Tocris) in 50 mM Tris–acetate (pH 7.0, 4°C) for 30 min. Non-specific binding was defined in the presence of 100 µM unlabelled kainate. Following incubation, the sections were washed three times in ice-cold buffer (50 mM Tris–acetate (pH 7.4) and once in ice-cold water over a total wash time of 5 s. Then the slides were dried in a stream of cold air. For the AMPA and GABAA receptor binding assays, sections were incubated in 20 nM [3H]fluorowillardine (29.08 Ci/mmol, Tocris) in 50 mM Tris–HCl buffer (pH 7.4, 4°C) for 40 min or with 20 nM [3H]muscimol (29.50 Ci/mmol, NEN) in the same buffer for 60 min. Non-specific binding was determined in the presence of 0.1 mM fluorowillardine or 0.1 mM muscimol. After incubation with the ligand, the tissue sections were washed in the incubation buffer alone for a total time of either 5 s for [3H] fluorowillardine binding, or 2 min for [3H]muscimol binding, before being blown dry in a stream of cold, dry air. For each radioligand, non-specific binding never exceeded 1% of total binding.

Quantitative Analysis of Autoradiographs

Once the sections were completely dry they were exposed to [3H]-Hyperfilm (Amersham, UK), together with radioactive standards ([3H] microscales, Amersham, UK) in X-ray cassettes for 4 weeks for [3H]MK801, [3H]fluorowillardine and [3H]muscimol binding and 3 months for [3H]kainite binding. Quantitative analysis of the autoradiographs was performed using a computer-assisted optical densitometry system (Image-Pro Plus;1 MediaCybernetics, Carlsbad, CA). The cortical layers were identified in adjacent Nissl-stained sections. Optical density measurements were taken through each cortical layer in the hand region of area 3b. Samplings of each layer were obtained from 5–11 different sections using at least three different animals from each condition and the arithmetic mean (±SEM) was calculated from all readings. The concentration of the bound radioligand was calculated from a standard curve created by using the [3H] microscale references. Receptor density was expressed in femtomoles of specifically bound ligand per milligram of tissue weight (fmol/mg tissue). Differences in binding in each layer between normal and E33 MAM-treated animals were analyzed using two-way t-test.

Immunocytochemistry

To localize GABAA receptors in somatosensory cortex we applied an antibody directed against all α subunits of GABAAα receptors (mouse anti-GABAA Receptor, α-chain monoclonal antibody; Chemicon). To quench endogenous peroxidase, sections were preincubated in Tris–HCl buffer (pH 7.4) with 5% normal horse serum (NHS), 0.3% TritonX-100, 1% H2O2 and 10% methanol for 1 h. To block non-specific binding the slides were transferred to the Tris–HCl buffer (pH 7.4) as above without hydrogen peroxidase for 20 min at room temperature. After overnight incubation with the primary antibody to GABAAα receptor (1:10) sections were washed (3 × 10min in buffer) and then incubated with the secondary antibody (Vectastain Elite ABC kit). The GABAAα staining was visualized using a Vector VIP peroxidase substrate kit. The sections were then mounted on subbed slides and coverslipped. Each experiment included control sections, processed with omission of the primary antibody.

Analysis of Immunoreactivity

The immuno- and Nissl-stained sections were analyzed using a computerized image acquisition and analysis system. Each image was spatially calibrated using an objective scale. To measure the distribution of GABAAα label in sections of normal and MAM-treated animals, profiles of optical density across the cortical layers (from pia to white matter) were measured. Distributions of GABAAα-IR in the somatosensory cortex of ferret were determined by superimposing immunoreacted sections with adjacent Nissl stained sections.

All analysis was done in the region of the hand representation of the somatosensory cortex, area 3b. The location of this region was determined in older animals using cytoarchitecture and previous observations (McLaughlin et al., 1998; Noctor et al., 2001). In younger animals (P0), this region can be identified using established morphologic features, although the cytoarchitecture is not mature at this age (Juliano et al., 1996; Noctor et al., 2001).

Thalamic Injections in Thick Slices of Ferret Somatosensory Cortex

In three brains of normal ferrets obtained at P14, thick slices were prepared using a roughly coronal orientation at an angle that included the ventrobasal thalamus and somatosensory area 3b. Each slice was placed in a chamber prepared to maintain live slices as described previously (Juliano et al., 1996). Injections were made into ventrobasal thalamus using glass pipettes filled with fluororuby (Molecular Probes). The injections were made iontophoretically using positive alternating current at 4 µA through a tip size of ∼20 µm. The slices were maintained in the chamber for 8–10 h to allow for transport of the fluorescent dextran. The slices were then briefly fixed and processed for GABAAα receptor immunoreactivity as described above, except that the secondary antibody was conjugated to Cy2 to allow for easier visualization with the fluorescent dextran injections into the thalamus. After the immunohistochemistry, the thick slices were cut on a vibratome at 30 µm thickness.

Results

Distribution of Excitatory and Inhibitory Receptors in Mature Cortex of Normal and E33 MAM-treated Adult Ferrets

To determine if administration of MAM on E33 affected the distribution of excitatory or inhibitory receptors in ferret somatosensory cortex, a preliminary study was conducted to assess the ligand binding pattern of a battery of receptors, specifically in area 3b of mature cortex. The same region was assessed in each animal, corresponding to a site slightly lateral to the postcruciate dimple, in the hand region of area 3b (McLaughlin et al., 1998). Densitometric readings were obtained separately for each layer. One of the effects of treatment with MAM on E33 is to reduce the thickness of neocortical layers 3 and 4 in area 3b of ferret somatosensory cortex, with layer 4 being drastically thinner than normal. A more detailed and quantitative analysis of the distinctions between normal and E33 MAM-treated cortex can be seen in Noctor et al. (2001).

Examination of the autoradiographs and densitometric readings for excitatory receptors revealed only slight distinctions between the laminar pattern for the normal and MAM-treated brains of NMDA ([3H]MK-801), AMPA ([3H]fluorowillardine) and kainate ([3H]kainate) receptors. The NMDA receptors were located most heavily in the upper layers and specifically higher in layers 2 and 3 (Fig. 1). The AMPA receptors were not as deeply located as the NMDA receptors and were preferentially found in superficial layers for both the normaland MAM-treated animals. In contrast, the kainate receptors were most heavily detected in deeper layers 5 and 6 for both groups (Fig. 1). Although minor variations could be observed between the normal and MAM-treated patterns of excitatory receptors, overall the distributions were similar.

The laminar pattern of inhibitory receptors (GABAA) showed distinctions between the normal and E33 MAM-treated distribution. Generally, GABAA receptors are preferentially located in layer 4 of somatosensory cortex (Schwark et al., 1994; Land et al., 1995), which we observed in normal ferret area 3b. After MAM treatment, layer 4 was most strongly affected. In the E33 MAM-treated cortex, although the diminished layer 4 also contained a relatively high level of muscimol binding, layers 2 and 3 contained the highest level of GABAA receptors, indicating a shift in the overall distribution pattern (Fig. 2).

Distribution of GABAA Receptors in Somatosensory Cortex of Normal and E33 MAM-treated Animals

Because the distribution of inhibitory GABAA receptors in ferret somatosensory cortex of MAM-treated animals was the only distribution that differed from the normal arrangement, we examined the receptor distribution using antibodies directed against GABAA receptors. The receptor binding results were intriguing, but did not reveal enough detail to explain why this pattern differed from the normal. We assessed immunoreactivity to all the α subunits (pan α) of the GABAA receptor. This family of subunits is present in high amounts in the cerebral cortex and confers the action of inhibition (Mohler et al., 1995). First, we assessed the distribution of GABAAα receptors in developing cortex. The same region (i.e. hand region of area 3b) was evaluated.

In normal somatosensory cortex at P0, GABAAα receptor immunoreactivity shows only slight staining in the deeper portion of the cortical plate. A similar staining pattern can be seen in E33 MAM-treated somatosensory cortex (Fig. 3). On P14, the layers are more distinct in normal cortex, although still not fully mature (Noctor et al., 1997). GABAAα receptor immunoreactivity is stronger at this age and in the normal cortex preferentially labels a thin band corresponding to layer 4. In E33 MAM-treated cortex, the distribution is broader and the heaviest label includes a laminar distribution that extends past the boundaries of layer 4, which is very thin in the MAM-treated animals (Fig. 3). In the adult ferret somatosensory cortex, the normal distribution of GABAAα receptor immunoreactivity remains highly localized to layer 4. In the MAM-treated cortex, however, the heavy distribution of immunoreactivity against GABAAα receptors covers a substantially wider region, extending into layers 3 and 5 (Fig. 3).

To more clearly assess the distribution of GABAAα receptors in relation to cortical layers, optical density measurements of the immunoreactivity were obtained across a trajectory from the pia through layer 6. Although a low level of immunoreactivity exists across all layers, these measurements revealed that on P0, the distribution increases slightly in the immature deeper layers (Fig. 4). On P14 and in the adult distributions of immunoreactivity, the greatest optical density values correlate closely with the borders of layer 4 in normal somatosensory cortex, whereas in the MAM-treated brains, the above background GABAAα receptor immunoreactivity extends substantially beyond the layer 4 borders (Fig. 4). In all of these graphs, the optical density values corresponding to layer 1 were omitted. This is because in many sections the immunoreactivity related to layer 1 or the pia were artifactually high.

Cellular Localization of GABAAα Receptors

At higher magnification, the precise distribution of GABAAα immunoreactivity across all layers was more easily accessible (Fig. 5). In normal cortex, layer 1 contained fine-labeled punctate structures in the neuropil and fine processes without a clear orientation; this pattern continued into layer 2, which contained labeled processes and puncta. Layer 3 contained sparsely labeled processes with densely packed small grains. Many of these labeled processes ran in a radial orientation. Layer 4 was the most densely labeled and contained many tightly packed immunoreactive elements occurring at all orientations. In all layers many small round or fusiform shaped neurons were outlined with immunoreactivity (e.g. arrows in Fig. 5).

The details of the distribution were similar in the E33 MAM-treated cortex. In many layers, particularly layers 5 and 6, neurons could be seen outlined with immunoreactivity. The dense immunoreactive product present in normal layer 4 extended in vertical arrays though layer 3 and into layer 2. The fine punctate structures observed in normal layer 1 and upper layer 2 were similar in normal and MAM-treated sections. The punctate distribution of immunoreactivity in layer 1 appeared slightly denser in the E33 MAM-treated cortex.

Relation of Thalamic Afferent Fibers and GABAAα Receptors

An additional set of experiments evaluated the relationship between GABAA receptors in the somatosensory cortex and thalamocortical afferent fibers. In the examples shown in Figure 6, both thalamic afferent fibers and GABAAα receptors are visible. Living slices containing the thalamus and somatosensory cortex were injected with fluororuby in the thalamus and then reacted to detect GABAAα receptors. The images shown in Figure 6 were taken from upper layer 5 of somatosensory cortex and demonstrate that the thalamic axons (shown alone in Fig. 6A,E) are in very close apposition to GABAAα receptors (shown alone in Fig. 6B,F and both fields together in Fig. 6C,D,G,IK). A single cell immunoreactive for GABAAα receptors can be seen in Figure 6H. The label occurs in a punctate form over the somata and extends long distances onto the processes. We infer from the size, that this is a neuron. These images indicate that thalamic afferent fibers are frequently coincident with GABAAα receptors, which are positioned intermittently along thalamic axons arriving in the cerebral cortex. Figure 6G and I (at higher power) demonstrate a thalamic axon that appears to be terminating on a GABAAα immunopositive cell.

Discussion

Distribution of Glutamatergic and GABAergic Receptors in Somatosensory Cortex

The laminar distribution of excitatory receptors found in this study resembles patterns identified previously in other species. In normal ferret somatosensory cortex, NMDA and AMPA receptors are expressed predominantly in the superficial cortical layers, while kainate binding sites are most intense in layer 5. Receptor binding experiments in somatosensory cortex of mice and rats also find the highest levels of NMDA and AMPA receptors in upper cortical layers (Jaarsma et al., 1991; Blue and Johnston, 1995; Glazewski et al., 1995). The deep cortical localization of kainate receptors agrees with previous autoradiographic data in other areas of cerebral cortex in different species (Miller et al., 1989; Gordon et al., 1991; Kumar et al., 1994; Smith and Thompson, 1994). In the E33 MAM-treated animals, the NMDA, AMPA and kainate receptors were also predominantly located in a pattern similar to the normal distribution. In a different study using rats, Rafiki et al. (1998) found that MAM treatment did not affect the cortical distribution of NR subunits of NMDA receptors.

The prominence of GABAA receptors in layer 4 of normal ferret somatosensory cortex is also similar to observations in rat, cat and monkey, which demonstrate higher levels of [3H]muscimol binding in layer 4 than adjacent in layers (Lidow et al., 1989; Schwark et al., 1994). Relatively intense immunostaining and dense GABAA receptors also occur in layer 4 of rodent somatosensory areas (Land et al., 1995). A high level of GABAAα receptor immunoreactivity was also selectively located in layers 4 and 6 of somatosensory cortex in monkeys (Huntley et al., 1990; Hornung and Fritschy, 1996). Several studies report that the density of GABAA receptors in layer 4 is paralleled by a high density of GABA containing neurons in that site (Beaulieu, 1993; Land et al., 1995).

In the E33 MAM-treated somatosensory cortex, the distribution of GABAA receptors was different. After MAM treatment, layer 4 is greatly diminished in thickness (Noctor et al., 2001); the distribution of GABAA receptors did not parallel this change, however, but expanded over a wider portion of cortex. The heaviest density of GABAA receptors shifted to the upper layers after MAM treatment on E33 and as shown through immunoreactivity, to lower layers as well. The change in distribution reflects a true increase in layers 2 and 3 because in normal animals, these layers demonstrate much lower GABAA immunoreactivity than layer 4, which is replaced by a strong increase after E33 MAM treatment. It is unlikely that the increased label reflects staining on ‘displaced’ layer 4 cells, since our earlier quantification of E33 MAM treatment demonstrates that the cells in layers 2–3 are not similar in size or shape to those in layer 4 and are composed primarily of cells morphologically similar to those normally in supragranular layers.

On What Neuronal Elements are GABAAα Receptors Located?

The CNS contains two prevalent types of GABA receptors, GABAA and GABAB. Our preliminary investigation of receptor binding assessed the distribution of GABAA receptors, because they are largely responsible for mediating postsynaptic inhibition. The GABAA receptors are comprised of several subunits; in the cerebral cortex the most common combinations include alpha, beta and gamma subunits. After finding an altered pattern of GABAA receptor binding in the MAM-treated somatosensory cortex, we evaluated the altered distribution of GABAA receptors in more detail, particularly the alpha subunits (i.e. pan α), since they are predominant in the neocortex and in the somatosensory cortex (Fritschy and Mohler, 1995; Golshani et al., 1997).

The results of our study are not conclusive as to the precise location of these receptors on neural elements. The pattern of immunoreactivity we observed in somatosensory cortex confirms that GABAA receptors are located on axons, somata and dendrites. Several researchers report that GABAA receptors can be found on both pyramidal and non-pyramidal cells, but in non-pyramidal cells the immunoreactivity is localized to somata, whereas in pyramidal cells the label extends into the dendrites (Gu et al., 1993; Land et al., 1995) and is also found on axons (Volk et al., 2002). In our data, the GABAAα immunoreactivity outlines multiple cell types and extends on dendrites, whether cells are pyramidal or not (e.g. Figures 5 and 6). Evidence also exists indicating that GABAAα receptors are present on growth cones (Fukura et al., 1999).

The distribution of GABAAα receptors shows they are in close relation to the thalamic afferent fibers in the cortex. It was somewhat surprising to observe GABAAα receptors studded along the thalamic axons, as well as clearly outlining cell bodies. The intermittent distribution we observed both on cell bodies and on processes, however, resembles the description of GABAA receptors reported by Somogyi et al. (1998), who find multiple types of similarly appearing GABAA receptors distributed throughout the cortex. GABAAα receptors found by these researchers associate with synapses, but are also in extrasynaptic positions. They additionally report axo-axonic synapses acting exclusively through GABAA receptors. Although the findings reported here are not detailed enough to provide the precise relation between thalamic axons and GABAAα receptors and they could either be pre or postsynaptic; in many instances the receptors and thalamic axons are in close relation to each other.

The change in the distribution of GABAAα receptors after E33 MAM treatment is also compatible with several altered features we detected in ferret somatosensory cortex. After MAM treatment that reduces the dimension of layer 4, the termination of thalamic afferents is no longer focused in the central layers, but distributed throughout the cortical thickness (Noctor et al., 2001; Palmer et al., 2001). In concert with this finding, the more diffuse pattern of GABAAα receptor immunoreactivity may be coincident with widespread termination of thalamic afferents. In the MAM model, it is possible that cells originally destined to receive thalamic input are displaced and receive thalamic terminations in unusual locations. Our evidence suggests this is not the case, i.e. thalamic afferents are not terminating on their original misplaced targets, since populations of cells similar to those in normal area 3b (in terms of morphology and cell size) populate the layers remaining after E33 MAM treatment (Noctor et al., 2001). This does not necessarily mean that the thalamic afferents and GABAAα receptors are colocalized, but that they are in similar positions and may influence each other.

What Regulates GABAAα Receptor Expression?

The expression of GABAAα receptors appears to be regulated both by intrinsic and extrinsic mechanisms. Cells in culture display a neuron specific pattern of GABAA receptor subtype expression and a precise spatial pattern of this receptor is expressed prenatally, suggesting a degree of intrinsic regulation (Paysan et al., 1994; Penschuck et al., 1999). On the other hand, GABAAα receptor expression is influenced by external factors such as alcohol consumption and a number of disease conditions (Chen et al., 1998). GABAAα receptors also seem to be regulated by activity dependent mechanisms. Huntsman and Jones (1998) report that the mRNA of several GABAA receptor subunits in visual cortex is differentially regulated by monocular deprivation, suggesting activity driven by the thalamus regulates GABAA receptor expression (Jones, 1997; Huntsman and Jones, 1998). Several other studies support this idea by demonstrating the GABAA receptors are specifically reduced in layer 4 after reduction of input to the somatosensory cortex (Paysan et al., 1997; Wellman et al., 2002). Poluch (2002) also found that GABAAα receptors are located on thalamocortical fibers, providing a precedent for an intimate linkage between GABAA receptors and thalamic terminations.

Effects of Altered GABAA Receptor Distribution on Cortical Transfer of Information

In MAM-treated cortex, the pattern of excitatory receptors was similar to normal ferret somatosensory cortex distribution, but the inhibitory pattern was altered. This suggests that the transmission of excitatory impulses remains relatively intact, while inhibitory processing is disrupted. There is substantial evidence that GABAergic mechanisms are important in the transfer of information through the cortex and in the initial sculpting of a response. Inhibitory influences act differently at each level of the cerebral cortex and are crucial at the initial transfer of information in layer 4 (Kyriazi et al., 1996, 1998; Brumberg et al., 1999). GABAergic cells in layer 4 are often targets of axons arriving from the thalamus. In subsequent processing, the neural tuning initiated in layer 4 is projected through other cortical layers. This transmission of information results in electrical coupling between GABAergic cells, which leads to patterns of synchronous activity (Beierlein et al., 2000; Tamás et al., 2000). In MAM-treated cortex, the initial transfer of information from the thalamus is altered because layer 4 is largely missing and the thalamic terminations are not focused on layer 4 but distributed through many cortical layers (Noctor et al., 2001; Palmer et al., 2001). The GABAA receptors are also widespread. The abnormal arrangement of thalamic projections and GABAA receptors probably leads to axons from the thalamus that terminate on cells different from their normal objectives. As a result, the normal projection pattern from layer 4 is probably lost, as is the resulting synchronous pattern of activity. Taken together, these breakdowns are likely to lead to disruption in the cortical response and loss of ability to appropriately process stimulation.

This work was supported by NIH-NINDS RO1-NS24014 (S.L.J.). We thank Donna Tatham for outstanding technical assistance and Amanda Dempster for excellent assistance with the receptor binding studies.

Address correspondence to Dr Sharon L. Juliano, Department of Anatomy, Physiology and Genetics, USUHS, 4301 Jones Bridge Rd, Bethesda, MD 20814, USA. Email: sjuliano@usuhs.mil.

Figure 1. Shown here are digitized autoradiographs of 3H-ligand binding to specific excitatory receptors and histograms indicating optical density values obtained from each layer of somatosensory cortex. In the left column are autoradiographs of sections taken from normal ferret somatosensory cortex with accompanying histograms of mean optical density values obtained through each layer; in the right column are autoradiographs and histograms obtained from somatosensory cortex of ferrets treated with MAM on E33. The error bars represent the standard error; the mean optical density values were obtained from at least three different animals. Illustrated are the distribution of binding patterns to NMDA ([3H]MK-801), AMPA ([3H]fluorowillardine) and kainate receptors ([3H]kainate). The histograms were made by obtaining optical density values (converted to fmol/mg tissue) through each cortical layer (indicated on the x-axis); in the MAM-treated cortex, a value for layer IV was obtained through the thin remaining layer; the layers are indicated with numbers. The autoradiographs and histograms indicate that distributions of specific excitatory receptors through layers of somatosensory cortex are similar when comparing normal and MAM-treated sections. Although the E33 MAM-treated brains show shallower sulci and gyri, the overall pattern of receptor binding does not vary substantially when comparing the two conditions. The histograms also indicate that a similar pattern of density values occurs through each layer of somatosensory cortex. Although the density values are not exactly the same for normal and E33 MAM-treated cortex, they present comparable overall distributions. Scale bar = 1500 µm.

Figure 1. Shown here are digitized autoradiographs of 3H-ligand binding to specific excitatory receptors and histograms indicating optical density values obtained from each layer of somatosensory cortex. In the left column are autoradiographs of sections taken from normal ferret somatosensory cortex with accompanying histograms of mean optical density values obtained through each layer; in the right column are autoradiographs and histograms obtained from somatosensory cortex of ferrets treated with MAM on E33. The error bars represent the standard error; the mean optical density values were obtained from at least three different animals. Illustrated are the distribution of binding patterns to NMDA ([3H]MK-801), AMPA ([3H]fluorowillardine) and kainate receptors ([3H]kainate). The histograms were made by obtaining optical density values (converted to fmol/mg tissue) through each cortical layer (indicated on the x-axis); in the MAM-treated cortex, a value for layer IV was obtained through the thin remaining layer; the layers are indicated with numbers. The autoradiographs and histograms indicate that distributions of specific excitatory receptors through layers of somatosensory cortex are similar when comparing normal and MAM-treated sections. Although the E33 MAM-treated brains show shallower sulci and gyri, the overall pattern of receptor binding does not vary substantially when comparing the two conditions. The histograms also indicate that a similar pattern of density values occurs through each layer of somatosensory cortex. Although the density values are not exactly the same for normal and E33 MAM-treated cortex, they present comparable overall distributions. Scale bar = 1500 µm.

Figure 2. Example of autoradiographs of binding to inhibitory receptors ([3H]muscimol) from normal (left) and E33 MAM-treated somatosensory cortex. This pattern of receptor binding was different in normal and MAM-treated cortex. In normal cortex, the muscimol binding concentrated most heavily in layer 4. In MAM-treated cortex, however, the highest concentration of GABAA receptors shifted so that layers 2–3 and the remaining layer 4 had high receptor binding, causing the distribution of GABAA receptors to be more spread out than in normal cortex. The density of binding was significantly greater in layers 2/3 of the MAM-treated animals (*P < 0.001). The histograms are prepared as in Figure 1, the error bars represent standard error. Scale bar = 1500 µm.

Figure 2. Example of autoradiographs of binding to inhibitory receptors ([3H]muscimol) from normal (left) and E33 MAM-treated somatosensory cortex. This pattern of receptor binding was different in normal and MAM-treated cortex. In normal cortex, the muscimol binding concentrated most heavily in layer 4. In MAM-treated cortex, however, the highest concentration of GABAA receptors shifted so that layers 2–3 and the remaining layer 4 had high receptor binding, causing the distribution of GABAA receptors to be more spread out than in normal cortex. The density of binding was significantly greater in layers 2/3 of the MAM-treated animals (*P < 0.001). The histograms are prepared as in Figure 1, the error bars represent standard error. Scale bar = 1500 µm.

Figure 3. These are examples of immunoreactivity against GABAAα receptors in ferret somatosensory cortex at different ages. In the left column are sections taken from normal animals and in the right column are sections taken from E33 MAM-treated animals. The approximate laminar locations are indicated in the figure. It should be noted that in the E33 MAM-treated cortex on the right, layer 4 is present, but very thin, so is not given a specific designation. At P0, the pattern of immunoreactivity looks similar in both normal and MAM-treated sections, with very light label occurring in deeper portions of the immature cortex that corresponds to the developing cortical plate (CP). The upper portion, corresponding to the marginal zone (MZ) has almost no immunoreactivity (top panels). By P14 (middle panels) the immunoreactivity is strongly localized to layer 4 in normal cortex, while in E33 MAM-treated cortex, the GABAAα receptor immunoreactivity extends outside the boundaries of layer 4. In normal adult somatosensory cortex, GABAAα receptors are preferentially located in layer 4, while in E33 MAM-treated cortex, the immunoreactivity covers a much wider area, extending into layers 3 and 5, although layer 4 is much thinner than normal in E33 MAM-treated somatosensory cortex. Scale bar = 350 µm

Figure 3. These are examples of immunoreactivity against GABAAα receptors in ferret somatosensory cortex at different ages. In the left column are sections taken from normal animals and in the right column are sections taken from E33 MAM-treated animals. The approximate laminar locations are indicated in the figure. It should be noted that in the E33 MAM-treated cortex on the right, layer 4 is present, but very thin, so is not given a specific designation. At P0, the pattern of immunoreactivity looks similar in both normal and MAM-treated sections, with very light label occurring in deeper portions of the immature cortex that corresponds to the developing cortical plate (CP). The upper portion, corresponding to the marginal zone (MZ) has almost no immunoreactivity (top panels). By P14 (middle panels) the immunoreactivity is strongly localized to layer 4 in normal cortex, while in E33 MAM-treated cortex, the GABAAα receptor immunoreactivity extends outside the boundaries of layer 4. In normal adult somatosensory cortex, GABAAα receptors are preferentially located in layer 4, while in E33 MAM-treated cortex, the immunoreactivity covers a much wider area, extending into layers 3 and 5, although layer 4 is much thinner than normal in E33 MAM-treated somatosensory cortex. Scale bar = 350 µm

Figure 4. Shown are graphs of optical density values taken from a trajectory through somatosensory cortex from layer 1 to layer 6 in normal (left column) and E33 MAM-treated brains (right column) at different ages. The values on the x-axis correspond to distance from the pia (in micrometers), the y-axis values correspond to optical density values. The bars colored gray coincide with regions identified cytoarchitectonically as layer 4 (using Nissl staining on adjacent sections). At P0, the density values indicate slight increases in a region corresponding to deeper cortical layers. At P14 and in the adult, the optical density values increase in the central part of the cortex, corresponding to layer 4 in normal cortex, and extending well beyond cytoarchitectonic layer 4 in E33 MAM-treated cortex. In these sections, optical density (OD) values for layer 1 are omitted (see text).

Figure 4. Shown are graphs of optical density values taken from a trajectory through somatosensory cortex from layer 1 to layer 6 in normal (left column) and E33 MAM-treated brains (right column) at different ages. The values on the x-axis correspond to distance from the pia (in micrometers), the y-axis values correspond to optical density values. The bars colored gray coincide with regions identified cytoarchitectonically as layer 4 (using Nissl staining on adjacent sections). At P0, the density values indicate slight increases in a region corresponding to deeper cortical layers. At P14 and in the adult, the optical density values increase in the central part of the cortex, corresponding to layer 4 in normal cortex, and extending well beyond cytoarchitectonic layer 4 in E33 MAM-treated cortex. In these sections, optical density (OD) values for layer 1 are omitted (see text).

Figure 5. These are higher power views of immunoreactivity against GABAAα receptors in normal and MAM-treated adult ferret somatosensory cortex. Shown are details of different cortical layers. In layers 1 and 2 immunoreactive processes can be seen coursing in many directions, and immunoreactivity outlines several small cells. The label also appears punctate (arrows). In layer 3, immunoreactive processes take on a more vertical orientation (arrows), but it is not easy to discriminate if they are axons or dendrites. In the MAM-treated section, the vertical pattern of immunoreactivity is substantially denser and more extensive than in the normal section. Only a small portion of layer 4 is shown, which contains numerous densely labeled processes and outlined cells. Layer 5 shows processes running in many directions and pyramidal cells outlined with immunoreactivity (arrows). In layer 6 fusiform cells can be seen outlined with GABAAα immunoreactivity (arrows). Scale bar = 100 µm.

Figure 5. These are higher power views of immunoreactivity against GABAAα receptors in normal and MAM-treated adult ferret somatosensory cortex. Shown are details of different cortical layers. In layers 1 and 2 immunoreactive processes can be seen coursing in many directions, and immunoreactivity outlines several small cells. The label also appears punctate (arrows). In layer 3, immunoreactive processes take on a more vertical orientation (arrows), but it is not easy to discriminate if they are axons or dendrites. In the MAM-treated section, the vertical pattern of immunoreactivity is substantially denser and more extensive than in the normal section. Only a small portion of layer 4 is shown, which contains numerous densely labeled processes and outlined cells. Layer 5 shows processes running in many directions and pyramidal cells outlined with immunoreactivity (arrows). In layer 6 fusiform cells can be seen outlined with GABAAα immunoreactivity (arrows). Scale bar = 100 µm.

Figure 6. Examples of thalamic afferent fibers that are labeled with fluorescent dextran alone (A,E) and the same cortical field labeled with GABAAα receptor immunoreactivity (B,F) in layer 5 of ferret somatosensory cortex. C and G are the two fields superimposed. H is an example of a cell immunoreactive for GABAAα receptor antibody. The cell in H and those in B and F indicate that immunoreactivity occurs on the cell body and extends onto cell processes in the form of small, localized clusters. When the two images are superimposed, many thalamic afferents are studded with clusters of GABAAα receptor immunoreactivity. These regions are indicated with arrows in C and G and shown in higher power in D, I–K. Scale bar = 25 µm.

Figure 6. Examples of thalamic afferent fibers that are labeled with fluorescent dextran alone (A,E) and the same cortical field labeled with GABAAα receptor immunoreactivity (B,F) in layer 5 of ferret somatosensory cortex. C and G are the two fields superimposed. H is an example of a cell immunoreactive for GABAAα receptor antibody. The cell in H and those in B and F indicate that immunoreactivity occurs on the cell body and extends onto cell processes in the form of small, localized clusters. When the two images are superimposed, many thalamic afferents are studded with clusters of GABAAα receptor immunoreactivity. These regions are indicated with arrows in C and G and shown in higher power in D, I–K. Scale bar = 25 µm.

References

Beaulieu C (
1993
) Numerical data on neocortical neurons in adult rat, with special reference to the GABA population.
Brain Res
 
609
:
284
–292.
Beierlein M, Gibson JR, Connors BW (
2000
) A network of electrically coupled interneurons drives synchronized inhibition in neocortex.
Nat Neurosci
 
3
:
904
–910.
Blue ME, Johnston MV (
1995
) The ontogeny of glutamate receptors in rat barrel field cortex.
Brain Res
 
Dev Brain Res
 
84
:
11
–25.
Brumberg JC, Pinto DJ, Simons DJ (
1999
) Cortical columnar processing in the rat whisker-to-barrel system.
J Neurophysiol
 
82
:
1808
–1817.
Cattaneo E, Reinach B, Caputi A, Cattabeni F, Di Luca M (
1995
) Selective in vitro blockade of neuroepithelial cells proliferation by methylazoxymethanol, a molecule capable of inducing long lasting functional impairments.
J Neurosci Res
 
41
:
640
–647.
Chalmers DT, McCulloch J (
1991
) Selective alterations in glutamate receptor subtypes after unilateral orbital enucleation.
Brain Res
 
540
:
255
–265.
Chen F, Rezvani A, Jarrott B, Lawrence AJ (
1998
) Distribution of GABAA receptors in the limbic system of alcohol-preferring and non-preferring rats: in situ hybridisation histochemistry and receptor autoradiography.
Neurochem Int
 
32
:
143
–151.
Fritschy JM, Mohler H (
1995
) GABAA-receptor heterogeneity in the adult rat brain: differential regional and cellular distribution of seven major subunits.
J Comp Neurol
 
359
:
154
–194.
Fukura H, Kitani Y, Komiya Y, Igarashi M (
1999
) GABAA receptor in growth cones: the outline of GABAA receptor-dependent signaling in growth cones is applicable to a variety of α-subunit species.
J Neurosci Res
 
58
:
407
–416.
Glazewski S, Kossut M, Skangiel-Kramska J (
1995
) NMDA receptors in mouse barrel cortex during normal development and following vibrissectomy.
Int J Dev Neurosci
 
13
:
505
–514.
Golshani P, Truong H, Jones EG (
1997
) Developmental expression of GABA(A) receptor subunit and GAD genes in mouse somatosensory barrel cortex.
J Comp Neurol
 
383
:
199
–219.
Gordon B, Daw N, Parkinson D (
1991
) The effect of age on binding of MK-801 in the cat visual cortex.
Brain Res
 
Dev Brain Res
 
62
:
61
–67.
Gordon B, Pardo D, Conant K (
1996
) Laminar distribution of MK-801, kainate, AMPA, and muscimol binding sites in cat visual cortex: a developmental study.
J Comp Neurol
 
365
:
466
–478.
Gu Q, Perez-Velazquez JL, Angelides KJ, Cynader MS (
1993
) Immunocytochemical study of GABAA receptors in the cat visual cortex.
J Comp Neurol
 
333
:
94
–108.
Hornung JP, Fritschy JM (
1996
) Developmental profile of GABAA-receptors in the marmoset monkey: expression of distinct subtypes in pre- and postnatal brain.
J Comp Neurol
 
367
:
413
–430.
Huntley GW, de Blas AL, Jones EG (
1990
) GABAA receptor immunoreactivity in adult and developing monkey sensory-motor cortex.
Exp Brain Res
 
82
:
519
–535.
Huntsman MM, Jones EG (
1998
) Expression of alpha3, beta3 and gamma1 GABA(A) receptor subunit messenger RNAs in visual cortex and lateral geniculate nucleus of normal and monocularly deprived monkeys.
Neuroscience
 
87
:
385
–400.
Insel TR, Miller LP, Gelhard RE (
1990
) The ontogeny of excitatory amino acid receptors in rat forebrain — I. N-Methyl-d-aspartate and quisqualate receptors.
Neuroscience
 
35
:
31
–43.
Jaarsma D, Sebens JB, Korf J (
1991
) Localization of NMDA and AMPA receptors in rat barrel field.
Neurosci Lett
 
133
:
233
–236.
Jablonska B, Kossut M, Skangiel-Kramska J (
1996
) Transient increase of AMPA and NMDA receptor binding in the barrel cortex of mice after tactile stimulation.
Neurobiol Learn Mem
 
66
:
36
–43.
Jablonska B, Smith AL, Kossut M, Skangiel-Kramska J (
1998
) Development of laminar distributions of kainate receptors in the somatosensory cortex of mice.
Brain Res
 
27
:
325
–329.
Jones EG (
1997
) Cortical development and thalamic pathology in schizophrenia.
Schizophr Bull
 
23
:
483
–501.
Juliano SL, Palmer SL, Sonty RV, Noctor S, Hill GF II (
1996
) Development of local connections in ferret somatosensory cortex.
J Comp Neurol
 
374
:
259
–277.
Kumar A, Schliebs R, Bigl V (
1994
) Postnatal development of NMDA, AMPA, and kainate receptors in individual layers of rat visual cortex and the effect of monocular deprivation.
Int J Dev Neurosci
 
12
:
31
–41.
Kyriazi H, Carvell GE, Brumberg JC, Simons DJ (
1996
) Quantitative effects of GABA and bicuculline methiodide on receptive field properties of neurons in real and simulated whisker barrels.
J Neurophysiol
 
75
:
547
–560.
Kyriazi H, Carvell GE, Brumberg JC, Simons DJ (
1998
) Laminar differences in bicuculline methiodide’s effects on cortical neurons in the rat whisker/barrel system.
Somatosens Mot Res
 
15
:
146
–156.
Land PW, de Blas AL, Reddy N (
1995
) Immunocytochemical localization of GABAA receptors in rat somatosensory cortex and effects of tactile deprivation.
Somatosens Mot Res
 
12
:
127
–141.
Lidow MS, Goldman-Rakic PS, Gallager DW, Geschwind DH, Rakic P (
1989
) Distribution of major neurotransmitter receptors in the motor and somatosensory cortex of the rhesus monkey.
Neuroscience
 
32
:
609
–627.
McLaughlin DF, Sonty RV, Juliano SL (
1998
) Organization of the forepaw representation in ferret somatosensory cortex.
Somatosens Mot Res
 
15
:
253
–268.
Miller KD, Chapman B, Stryker MP (
1989
) Visual responses in adult cat visual cortex depend on N-methyl-d-aspartate receptors.
Proc Natl Acad Sci USA
 
86
:
5183
–5187.
Mohler H, Benke D, Benson J, Luscher B, Fritschy JM (
1995
) GABAA-receptor subtypes in vivo: cellular localization, pharmacology and regulation.
Adv Biochem Psychopharmacol
 
48
:
41
–56.
Noctor SC, Scholnicoff NJ, Juliano SL (
1997
) Histogenesis of ferret somatosensory cortex.
J Comp Neurol
 
387
:
179
–193.
Noctor SC, Palmer SL, McLaughlin DF, Juliano SL (
2001
) Disruption of layers 3 and 4 during development results in altered thalamocortical projections in ferret somatosensory cortex.
J Neurosci
 
21
:
3184
–3195.
Palmer SL, Noctor SC, Jablonska B, Juliano SL (
2001
) Laminar specific alterations of thalamocortical projections in organotypic cultures following layer 4 disruption in ferret somatosensory cortex.
Eur J Neurosci
 
13
:
1559
–1571.
Paysan J, Bolz J, Mohler H, Fritschy JM (
1994
) GABAA receptor alpha 1 subunit, an early marker for area specification in developing rat cerebral cortex.
J Comp Neurol
 
350
:
133
–149.
Paysan J, Kossel A, Bolz J, Fritschy JM (
1997
) Area-specific regulation of gamma-aminobutyric acid type A receptor subtypes by thalamic afferents in developing rat neocortex.
Proc Natl Acad Sci USA
 
94
:
6995
–7000.
Penschuck S, Paysan J, Giorgetta O, Fritschy JM (
1999
) Activity-dependent regulation of GABAA receptors.
Ann N Y Acad Sci
 
868
:
654
–666.
Poluch S (
2002
) Expression precoce de recepteurs de l’AMPA dans le neocortex embryonnaire chez le rat. Thesis, Université Montpellier II, Neurosciences, Biologie Santé.
Rafiki A, Chevassus-au-Louis N, Ben-Ari Y, Khrestchatisky M, Represa A (
1998
) Glutamate receptors in dysplasic cortex: an in situ hybridization and immunohistochemistry study in rats with prenatal treatment with methylazoxymethanol.
Brain Res
 
782
:
142
–152.
Schwark HD, Li J, Fuchs JL (
1994
) Regional distribution of GABAA receptor binding sites in cat somatosensory and motor cortex.
J Comp Neurol
 
343
:
362
–369.
Smith AL, Thompson ID (
1994
) Distinct laminar differences in the distribution of excitatory amino acid receptors in adult ferret primary visual cortex.
Neuroscience
 
61
:
467
–479.
Somogyi P, Tamas G, Lujan R, Buhl EH (
1998
) Salient features of synaptic organisation in the cerebral cortex.
Brain Res
 
Brain Res Rev
 
26
:
113
–135.
Tamás G, Buhl EH, Lorincz A, Somogyi P (
2000
) Proximally targeted GABAergic synapses and gap junctions synchronize cortical interneurons.
Nat Neurosci
 
3
:
366
–371.
Volk DW, Pierri JN, Fritschy JM, Auh S, Sampson AR, Lewis DA (
2002
) Reciprocal alterations in pre- and postsynaptic inhibitory markers at chandelier cell inputs to pyramidal neurons in schizophrenia.
Cereb Cortex
 
12
:
1063
–1070.
Wellman CL, Arnold LL, Garman EE, Garraghty PE (
2002
) Acute reductions in GABAA receptor binding in layer IV of adult primate somatosensory cortex after peripheral nerve injury.
Brain Res
 
954
:
68
–72.
Young AB, Sakurai SY, Albin RL, Makowiec RL, Penney JB (
1991
) Excitatory amino acid receptor distribution: quantitative autoradiographic studies. In: Excitatory amino acids and synaptic transmission (Wheal H, Thomson A, eds), pp.
19
–32. London: Academic Press.