Brain Research Bulletin, Vol. 57, No. 5, pp. 631– 638, 2002 Copyright © 2002 Elsevier Science Inc. Printed in the USA. All rights reserved 0361-9230/02/$–see front matter PII S0361-9230(01)00757-2 Fos induction in cortical interneurons during spontaneous wakefulness of rats in a familiar or enriched environment Giuseppe Bertini,1* Ze-Chun Peng,1† Paolo F. Fabene,1 Gigliola Grassi-Zucconi2 and Marina Bentivoglio1 1 Department of Morphological and Biomedical Sciences, University of Verona, Verona, Italy; and 2Department of Cell Biology, University of Perugia, Perugia, Italy [Received 28 August 2001; Revised 23 October 2001; Accepted 25 October 2001] ABSTRACT: It has been repeatedly reported that Fos is spontaneously induced in several brain structures, including the cerebral cortex, during wakefulness. To ascertain whether cortical interneurons are involved in this state-dependent oscillation of gene regulation, we combined Fos immunocytochemistry with immunostaining of either parvalbumin or calbindin, known markers of cortical interneurons. Immunopositive neurons were examined in the sensorimotor and cingulate cortex. In rats perfused in basal conditions, a minor proportion (around 8%) of Fos-immunoreactive neurons in the parietal cortex were also parvalbumin- or calbindin-immunoreactive; these double immunostained cells accounted for 13% of the parvalbuminand 34% of the calbindin-labeled neurons. Colocalization of Fos with either calcium-binding protein was instead not observed in the cingulate cortex. In rats stimulated by novel environmental cues during the period of wakefulness preceding perfusion, Fos-positive neurons increased markedly relative to unstimulated animals, and involved the majority of the calbindin- or parvalbumin-labeled cell populations (60 –75% and over 95%, respectively). In the neuronal populations in which Fos was induced by exposure to the enriched environment, the proportion of calbindin- and parvalbumin-labeled cells was larger than in the unstimulated cases, and the increment was statistically significant in the cingulate cortex. The results demonstrate that Fos induction occurring in the cortex during undisturbed wakefulness in a familiar environment involves a minor proportion of interneurons. Furthermore, the findings indicate that the addition of novel environmental stimuli results in an increase of Fos-expressing neurons whose recruitment, at least in the cingulate cortex, involves a higher proportion of interneurons than of projection neurons. © 2002 Elsevier Science Inc. its protein product Fos are induced during wakefulness in the rat cerebral cortex [7,8,18]. These and other studies have provided evidence that gene expression fluctuates considerably in relation to the circadian variation of functional states, as well as to the type of activity that the animal is engaged in [15,19]. However, it remains to be determined whether certain cortical cell populations are preferentially involved in such oscillation. In the present study, we aimed at a characterization of cortical neurons involved in behavioral state-dependent Fos induction. In particular, we sought to ascertain whether Fos is induced during wakefulness in interneurons of the cingulate and parietal cortex, and we examined this issue with a double immunocytochemical approach. In the cerebral cortex, short-axon neurons comprise heterogeneous subsets of inhibitory GABAergic cells, which represent approximately 15–30% of the total neuronal cell population (see [6] for review). On the other hand, cortical projection neurons, which are excitatory and utilize amino acids as neurotransmitters, constitute the largest component (approximately 70 – 85%) of the total population of cortical neurons [6]. A number of different neuroactive molecules, including neuropeptides and calcium-binding proteins of the EF-hand family, have been shown to be colocalized with GABA in cortical interneurons. In particular, cells immunostained by the calcium-binding proteins parvalbumin (Pv) or calbindin D-28k (Cb) represent, taken together, more than 90% of cortical GABAergic interneurons, although with areal and species variations [6]. In the rat cortex, 70% of the GABAergic cells contain Pv [4], while Cb-positive neurons include a much lower percentage of cortical nonpyramidal cells [5,6]. Pv- and Cb-containing cells mostly represent discrete subpopulations of cortical interneurons, and co-localization of the two proteins in the rat somatosensory cortex can be demonstrated in a very limited subset of neurons [20]. In the present study, we selected Pv and Cb immunoreactivity, which is fully compatible with Fos immunostaining (at variance with GABA immunocytochemistry, which requires a different perfusion recipe), to characterize populations of interneurons in the cerebral cortex of rats perfused during KEY WORDS: Calbindin, Parvalbumin, Exploratory behavior, Immediate early genes, Immunocytochemistry. INTRODUCTION The expression of the immediate early gene c-fos has been reported to oscillate during the spontaneous sleep/wakefulness cycle in neuronal populations of the brain (see [1] for review). In particular, it has been repeatedly described that c-fos mRNA and * Address for correspondence: Dr. G. Bertini, Department of Morphological and Biomedical Sciences, Section of Anatomy and Histology, Medical Faculty, Strada Le Grazie 8, 37134 Verona, Italy. Fax: ⫹39-045-802-7163; E-mail: giuseppe.bertini@univr.it † Present address: Brain Research Institute, University of California at Los Angeles, Los Angeles, CA, USA. 631 632 BERTINI ET AL. wakefulness, both in the absence of specific stimulation and after introducing novel environmental cues that trigger exploratory behavior. MATERIALS AND METHODS Animals and Perfusion The study was based on six adult male Wistar rats (250 –300 g) that were housed under a 12 h/12 h light/dark cycle (lights on at 0700h), with controlled temperature and food and water ad libitum, for at least 2 weeks before the experiment. The experimental protocols received institutional approval and authorization by the Italian Ministry of Health. All efforts were made to avoid animal suffering and minimize the number of animals used. In the first three animals (rats 1–3; group F, maintained in a familiar environment during the experiment), epidural electrodes were implanted under barbiturate anesthesia (Nembutal 50 mg/kg, i.p.), and the animals were familiarized with electroencephalogram (EEG) recording for at least 2 h every day for 1 week. On the day of the experiment, EEG was continuously recorded for 3– 4 h starting at lights-off time. The other three animals (rats 4 – 6; group E, exposed to an enriched environment) were not implanted with electrodes and EEG was not recorded. On the day of the experiment, at lights off, environment of the latter animals was enriched by introducing novel objects, such as stones and leaves, as well as oddly shaped pieces of plastic in their home cages. All animals (rats 1– 6) were rapidly anesthetized between 2300h and 2400h and transcardially perfused with saline followed by 4% phosphatebuffered paraformaldehyde. In addition to the EEG monitoring of rats of group F, all animals were individually observed and their behavior manually recorded for the whole 3– 4 h preceding perfusion. Histology and Immunohistochemistry After a brief postfixation and overnight cryoprotection in 30% phosphate-buffered sucrose, brains were cut into 30-␮m-thick serial sections with a freezing microtome. In the rats of group F, every 10th section was processed for Fos immunocytochemistry, whereas the adjacent series was processed for double immunocytochemistry with anti-Fos antibodies, and either anti-Cb (rats 1 and 2) or anti-Pv (rats 2 and 3) antibodies. In the rats of group E, every 10th section was processed for Fos and Cb double immunocytochemistry, and the adjacent series of sections was processed for Fos and Pv double immunocytochemistry. In all cases, one series of sections was stained with cresyl violet for cytoarchitectonic control. For Fos immunohistochemistry, the sections were repeatedly washed in 0.01 M phosphate-buffered saline, pH 7.4 (PBS), soaked in 0.75% H2O2 in PBS and preincubated in 5% normal rabbit serum and 0.2% Triton X-100 in PBS for 1 h. The sections were then incubated for 36 – 40 h in sheep polyclonal antibodies that recognize Fos (Cambridge Research Biochemicals, Cambridge, UK), diluted 1:4000 in the preincubation solution to which 0.1% bovine serum albumin and 0.05% sodium azide had been added. The sections were then incubated for 2 h in biotinylated rabbit anti-sheep secondary antibodies (Vector, Burlingame, CA, USA), diluted 1:200 in 2% normal rabbit serum and 0.2% Triton X-100 in PBS. The sections were finally reacted with avidin-biotin peroxidase (ABC kit, Vector) using a heavy metal intensification of 3-3⬘-diaminobenzidine (DAB; Sigma, St. Louis, MO, USA): 0.025% DAB, 0.1% nickel ammonium sulfate, and 0.03% H2O2 in PBS. Following Fos immunostaining, sections destined to double immunohistochemistry were repeatedly washed in PBS, and then treated with mouse monoclonal antibodies that recognize either Pv (Sigma; diluted 1:10,000) or Cb (Swant, Bellinzona, Switzerland; diluted 1:5000). The same procedure outlined above was followed, except that normal horse serum and biotinylated horse anti-mouse IgG (Vector) as secondary antibodies were used. A previously described protocol [17] was adopted in the last step of the procedure. Briefly, sections were reacted in a freshly prepared solution of 0.05% ␣-naphtol, 0.1% ammonium carbonate, and 0.003% H2O2, and then incubated in 0.1% pyronin B in 0.1 M phosphate buffer, pH 7.4. This recipe results in a pink-purple color of the reaction products. All sections were then mounted on gelatinized slides, dehydrated through the alcohol series, cleared in xylene, and coverslipped with Entellan. Data Analysis The material was studied at the microscope under bright-field illumination with the aid of an X-Y digital plotting system connected to the microscope stage by means of transducers, and of a JVC CCD KY-F58 digital camera connected to the microscope and the image analysis software ImagePro Plus 4.1 (Media Cybernetics, Silver Spring, MD, USA). Analysis of the areal and laminar distribution of labeled cells in the cases of group F was performed by charting double and single Fos-immunoreactive (ir) and Cb- or Pv-ir neurons in the cerebral cortex, using sample sections at the level of the frontoparietal cortex. Due to the very high number of Fos-ir cells in the rats of group E, only double and single Cb- or Pv-ir neurons were instead charted in sample sections through the frontoparietal cortex of these animals. Quantitative analysis of the material was performed by counting single-immunostained and double-immunostained neurons in selected areas of the parietal and cingulate cortex. A strategy of unbiased section and field sampling was adopted in order to obtain samples at consistent locations across animals. First, using a 25⫻ objective we acquired digitized images of two entire coronal sections for each experimental animal (Fig. 1). With the aid of the image analysis software, we then measured for each section the perimeter of the cortex on the coronal plane between the rhinal sulcus and the indusium griseum (Fig. 1, arrowheads), which served as reference points. We then placed 300 ⫻ 1500 ␮m frames spanning all cortical layers (from the pial surface to the border with the white matter) in the parietal and in the cingulate cortex of one hemisphere (Fig. 1), at the same relative distance from the two reference points across animals. Based on anatomical landmarks and cytoarchitecture of adjacent cresyl violet-stained sections, all the frames were located within the Par1 and Cg1 fields, as defined by Zilles [21]. Immunostained cells found within the frames’ boundaries were directly observed with 25⫻ and 40⫻ objectives (at final magnifications of 250⫻ and 400⫻) and individually counted through all focal planes. For statistical evaluation, data were summarized by averaging together cell counts for the two rostro-caudal levels and for the different animals in each group. The effects of exploratory behavior on Fos induction in different classes of neurons were analyzed on the basis of 2 ⫻ 2 contingency tables of Fos-labeled neurons. In particular, we compared the mean number of Fos-positive cells that contained either calcium-binding protein to those that did not (dependent variable), in the sections obtained from groups F and E (independent variable). Tables were obtained for each of the two studied calcium-binding proteins and for each of the two cortical areas analyzed, and were statistically evaluated with chi-square (␹2) tests. FOS EXPRESSION IN THE AWAKE RAT CORTEX 633 Immunocytochemistry FIG. 1. The drawing illustrates the location of the cortical regions of interest (shaded rectangles) for quantitative analysis; sections were located ⫺0.3 mm (A) and ⫺2.8 mm (B) from bregma [21]; reference landmarks for positioning of the regions of interest are indicated by the arrowheads; the empty arrowheads point to the rhinal sulcus, and the filled ones to the indusium griseum. RESULTS Behavior Records of behavioral observation indicated that the rats of group F were active (grooming, eating, walking, and rearing) for most of the time preceding perfusion, with short periods of rest lasting few minutes. Systematic EEG scoring in these rats confirmed that the animals were almost always awake during the 3– 4 hours preceding perfusion, with short episodes of slow-wave sleep, and a single episode of paradoxical sleep, lasting less than 2 min, observed in case 3 (Table 1). The rats of group E showed substantially more intense activity than those of group F, and displayed patterns of exploratory behavior consisting of systematic inspection, accompanied by constant whisking, of each of the objects recently added to the cage. These systematic exploratory routines were spontaneously repeated several times. TABLE 1 PROPORTION OF TIME SPENT IN WAKEFULNESS (W), SLOW-WAVE SLEEP (SS), AND PARADOXICAL SLEEP (PS) DURING THE HOURS PRECEDING PERFUSION BY EACH OF THE 3 ANIMALS OF GROUP F, STUDIED IN THE FAMILIAR ENVIRONMENT Rat # W SS PS 1 2 3 99.6 95.8 90.7 0.4 4.2 8.5 0.0 0.0 0.8 The patterns of Pv and Cb immunostaining in all examined cases were consistent with those previously reported in the literature. Briefly, Pv-positive cells were found throughout the neocortex, and exhibited a relatively uniform distribution in layers II–VI. Neurons containing Cb were overall less numerous and, although present in all layers, more irregularly distributed. In all rats of group F, numerous Fos-immunostained cells were seen in the cortex, concentrated in the cingulate and sensorimotor areas. In terms of laminar distribution, Fos-ir cells were mainly located in a more superficial band (layers II–III) and in a deeper band (mostly layer V) (Fig. 2). Fos immunoreactivity was consistently bilateral, and labeled neurons were often densely clustered. Foci of dense labeling were observed in several different cortical fields (including Cg1, Par1, Fr1, Pir, Te1, Oc2 [21]) and exhibited a mirror distribution in the two hemispheres. In the rats of group E, Fos-ir cells were extremely numerous throughout the cerebral cortex and in all layers. The bilaminar distribution of labeled cells, observed in the previous cases, was still evident but less marked, and many Fos-ir neurons were also seen in layers IV and VI. As previously described [17], in the sections processed for double immunostaining the different colors of the reaction products and their subcellular distribution enabled a clear distinction between single Fos-ir neurons and single Pv-ir or Cb-ir ones. Thus, dark brown-black Fos immunoreactivity was confined to the neuronal nuclei, whereas bright pink-purple immunostaining filled the perikarya and proximal dendrites of Cb-ir or Pv-ir neurons (Fig. 3). Double immunostained neurons displayed Fos immunoreactivity in the nucleus and calcium-binding protein immunoreactivity in the cytoplasm (Fig. 3B, C). Neuropil staining was rather high in the sections processed for Pv immunocytochemistry, but not in those processed with anti-Cb antibodies (Fig. 3D). In the latter sections, however, a relatively high background was observed in the superficial cortical layers, as reported in previous studies on the distribution of Cb in the rat cortex [5]. In the cases of group F, Fos was found to be colocalized with either calcium-binding protein in a relatively small subset of neurons of the sensorimotor cortex. In particular, both Fos/Pvlabeled neurons and Fos/Cb-labeled ones were observed in the parietal cortex, with a slight prevalence for layers II/III (Fig. 2), and the findings appeared consistent across animals. At the quantitative evaluation, these double-labeled cells were found to represent a minor proportion of the Fos-ir cells in the parietal cortex, i.e. approximately 8% for both Pv and Cb (Table 2 and Fig. 4). With respect to the cell population expressing either calciumbinding protein, the double-labeled cells accounted for approximately 13% of the total Pv-ir neurons and for 34% of the Cb-ir neurons in the parietal cortex. No double-labeled cells were instead observed in the cingulate cortex (Table 2 and Fig. 4). In the rats of group E, which had been stimulated by novel environmental cues, double-labeled cell populations were instead commonly observed in both the parietal and cingulate cortex. The distribution of double-labeled neurons in these rats was consistent with that of calcium-binding proteins in the cases of group F, with a prevalence in layers II/III and layer V, especially for the Pvlabeled cells (Fig. 2). At the quantitative analysis, over 95% of the Pv-ir neurons were found to express also Fos in both the examined cortical fields, while the proportion of Fos/Cb-positive cells accounted for 60% of the Cb-ir cell population in the cingulate cortex and for 75% of the same cell population in the parietal cortex (Table 2). The quantitative analysis also revealed that the Fos/Pv-ir neurons accounted for an average proportion of about 14% of the total Fos-ir cell population in the parietal cortex, and the Fos/Cb-ir 634 BERTINI ET AL. FIG. 2. Graphic representation of the distribution of immunoreactive neurons in the parietal cortex. Panels represent the regions of interest from sections processed with Fos/Pv ((Fos ⫹ Pv, left-side panels) and Fos/Cb (Fos ⫹ Cb, right-side panels) double immunocytochemistry. Comparison of the immunoreactivity in the samples obtained from rats exposed to familiar vs. enriched environment demonstrates the markedly larger population of Fos-ir neurons, as well as the substantial increase of double-immunostained neurons, in the latter cases. Fos-ir cells are represented by small black dots, whereas neurons containing either calcium-binding protein are shown in red and double-stained (Fos/Pv or Fos/Cb) cells are blue. For clarity of the graphic representation, each dot represents two labeled neurons of each cell population. FIG. 3. (Next page) The microphotographs illustrate double Fos/Pv immunostaining (A–C) and Fos/Cb immunostaining (D–F) in the somatosensory cortex of a rat of the first group (rat 3), perfused in basal conditions after 3 h of wakefulness documented by electroencephalographic recording. Fosimmunopositive neurons exhibit brown labeling of the cell nuclei, whereas Pv or Cb immunostaining is revealed by the purple-pink labeling. (A) Low-power view, illustrating the density of both labels and their laminar prevalence. (B) and (C) show at higher power Pv/Fos double-labeled neurons (arrowheads) intermingled with single labeled ones. (D) Low-power view of layers II/III showing Cb-positive and Fos-positive cells, the intermingling of which is illustrated at higher power in (E) and (F); note that the labels are evident in distinct neuronal populations. Scale bars: A, 140 ␮m; C, 25 ␮m (applies also to B); D, 80 ␮m; F, 45 ␮m (applies also to E). FOS EXPRESSION IN THE AWAKE RAT CORTEX 635 636 BERTINI ET AL. TABLE 2 MEAN NUMBER AND STANDARD DEVIATION OF IMMUNOREACTIVE NEURONS IN SAMPLES PROCESSED WITH FOS/PV AND FOS/CB DOUBLE IMMUNOCYTOCHEMISTRY IN THE TWO GROUPS OF CASES (GROUP F, FAMILIAR ENVIRONMENT AND GROUP E, ENRICHED ENVIRONMENT) Parietal cortex Cingulate cortex Familiar Enriched Familiar Enriched Pv Fos/Pv Fos 85.50 ⫾ 7.78 12.50 ⫾ 0.71 138.00 ⫾ 4.24 3.75 ⫾ 3.10 84.00 ⫾ 6.06 525.75 ⫾ 27.29 84.00 ⫾ 4.24 0.00 ⫾ 0.00 132.00 ⫾ 9.90 3.25 ⫾ 1.89 90.25 ⫾ 6.70 499.50 ⫾ 26.25 Cb Fos/Cb Fos 26.00 ⫾ 8.52 13.25 ⫾ 2.63 159.50 ⫾ 20.17 13.25 ⫾ 2.50 38.25 ⫾ 3.86 590.50 ⫾ 36.86 26.75 ⫾ 14.24 0.00 ⫾ 0.00 141.50 ⫾ 13.92 11.50 ⫾ 8.54 17.25 ⫾ 4.79 553.00 ⫾ 47.25 neurons accounted for 6% of the total Fos-ir cells in the same area (Table 2 and Fig. 4). In addition, as mentioned above, in the rats of group E double-immunolabeled cells were also seen in the cingulate cortex (Table 2). In the latter area, Fos/Pv-ir neurons accounted for an average proportion of 15%, and Fos/Cb-ir cells for 3%, of the total Fos-ir cell population (Fig. 4). The above data indicate that the increment of calcium-binding protein-containing neurons that expressed Fos after exposure to the enriched environment was larger than the total increment of Fos-ir cells. Thus, while the cases of group E showed on average four times as many Fos-ir neurons in the parietal cortex than those of group F (Table 2), the proportion of Pv-ir cells among Fosexpressing neurons in group E was 7.5 times larger than in group F. The latter finding raised the question as to whether recruitment of Fos-expressing neurons as a consequence of exploratory behavior preferentially involved calcium-binding protein-containing interneurons rather than being equally distributed among neuronal populations. This issue was addressed with chi-square tests of the already described contingency tables. In the cingulate cortex, the analyses indicated that both the Pv-ir and Cb-ir cell populations were involved by the environ- ment-dependent increase of Fos expression significantly more than neurons that did not contain either calcium-binding protein, and that this effect was highly significant for Pv-expressing neurons (Pv: ␹2 ⫽ 23.09, p ⬍ 0.0001; Cb: ␹2 ⫽ 4.38, p ⬍ 0.05). In the parietal cortex, a similar (although nonsignificant) trend was revealed for the Pv-ir cell population, while the effect was absent for Cb-ir neurons (Pv: ␹2 ⫽ 3.26; Cb: ␹2 ⫽ 0.57). DISCUSSION Consistent with previous reports of circadian oscillations in the pattern of c-fos expression in the brain (reviewed in [1]), our results show that the Fos protein is spontaneously induced in neuronal subsets of the cerebral cortex during wakefulness even in the absence of any specific stimulation. Previous studies have also shown that Fos expression in the cerebral cortex during sleep is instead very low, especially in relation to the relative amount of slow-wave sleep [7,8]. Altogether these findings indicate that the induction of the c-fos gene in cortical neurons is a specific feature of wakefulness-dependent behavior. A variety of studies of stimulus-induced Fos expression have FIG. 4. The charts illustrate the differences in the average proportion of calbindin- or parvalbumin-immunoreactive (ir) neurons relative to the total Fos-ir cell populations (i.e., Fos/parvalbumin and Fos/calbindin double-labeled neurons) in the two groups of cases, examined after wakefulness in a familiar environment and in an enriched environment, respectively. FOS EXPRESSION IN THE AWAKE RAT CORTEX demonstrated that the protein becomes detectable in the neuronal nucleus 1–2 h following an effective stimulus and peaks shortly afterwards (reviewed in [9]). Naturally, a reliable assessment of “spontaneous” expression requires particular methodological care. Our experimental procedure ensured that the observed pattern of Fos expression reflected brain activity during a period of undisturbed wakefulness, physiologically triggered by the onset of darkness, and spanning approximately 3 to 1 h prior to perfusion. The high number of Fos-ir cells observed when the animals had been stimulated by novel environmental cues also indicated that Fos induction during wakefulness parallels the animal’s activity and exploratory behavior. Our experiment, however, does not allow a differentiation between sensorimotor activity-related Fos expression and gene induction concurrent with intrinsically generated activation states. The present study demonstrates that in animals examined at the end of a period of wakefulness in a familiar environment Fos is induced in a minor proportion of local circuit neurons that express Pv or Cb in the parietal cortex, whereas no Fos induction was found in Cb- or Pv-positive interneurons of the cingulate cortex. Our data on Fos induction in cortical inhibitory cells during relatively “quiet” wakefulness are in line with previous evidence, obtained in different experimental paradigms of cortical activation such as epilepsy [10,14] and thalamic stimulation [2], demonstrating c-fos induction in relatively small proportions of cortical interneurons, as defined by their colocalization with calciumbinding proteins. In the animals that had been exposed to an enriched environment, however, Fos induction was detected in a much higher number of cortical neurons than in the behaviorally unstimulated cases. This finding is in agreement with the recent report [19] of a significant increase of the protein products of immediate early genes, including c-fos, in the barrel cortex of rats placed overnight in an enriched environment. Interestingly, in the present study we could determine that the general increase of cortical cells in which Fos was induced by the exploratory behavior was also accompanied by a substantial increment of Fos-ir interneurons, which did not simply parallel an overall increase in Fos expression in the corresponding cortical areas. Rather, cortical interneurons, at least in the cingulate cortex, were significantly more recruited during behaviorally relevant sensorimotor activity than projection neurons. It is also interesting to note that such relative increase of Fos-ir cortical interneurons in actively exploring rats, compared to animals habituated to their environment, was especially marked in the Pv-containing cell population. Taken together, these results indicate that meaningful activity, such as exploration of environmental novelty, is accompanied by selective recruitment of functionally different types of interneurons involved in regulatory actions on pyramidal cells of distinct cortical areas. In this respect, it should be pointed out that Pv- or Cb-containing cortical interneurons have different physiological properties. In particular, nonpyramidal cells that express Pv have been identified as fast spiking cells in layers II–III and V of the rat frontal cortex [3,11–13]. On the other hand, Cb-containing cells were characterized as low-threshold spike cells in layer V of the frontal cortex [12], but the physiological characteristics of neocortical nonpyramidal neurons that express Cb were also found to be more heterogeneous and complex [3]. On the other hand, the very high proportion of Fos-positive cells that did not express Pv or Cb in our material provides indirect evidence that the majority of cells expressing Fos in basal conditions are represented by projection neurons. In addition, the laminar distribution of Fos-positive cortical cells during spontaneous wakefulness indicated that the oscillation of gene expression involved both corticocortical projection neurons and neurons of the 637 deep layers that innervate subcortical targets. It should be pointed out, however, that Fos induction was observed also in layer IV neurons of the animals exposed to novel environmental stimuli. The latter findings support previous data on marked Fos induction in neurons of layer IV and supragranular layers of the primary visual and somatosensory cortices of rats exploring a novel environment, suggesting that the activation of these layers may be specifically related to the attention paid to environmental stimuli [15]. c-fos is a transcription factor involved in the signal transduction cascade that links extracellular events to long-term intracellular adaptations, and has been reported to be induced in neurons by a variety of stimuli [9,16]. The significance of the spontaneous fluctuation of gene expression in wake/sleep conditions is still unknown, but the present findings point out that distinct subsets of interneurons are involved in the cortex in intrinsic regulatory mechanisms that occur in state-dependent behavior. Our data also indicate that the neural correlates of the processing of novel environmental stimuli involve a transcriptional response of select populations of cortical interneurons. ACKNOWLEDGEMENT The study was supported by grants of the Italian Ministry of University and Research. REFERENCES 1. Bentivoglio, M.; Grassi-Zucconi, G. Immediate early gene expression in sleep and wakefulness. In: Lydic, R.; Baghdoyan, H., eds. Handbook of behavioral state control. Boca Raton, FL: CRC Press; 1998: 235–253. 2. Bubser, M.; de Brabander, J. M.; Timmerman, W.; Feenstra, M. G.; Erdtsieck-Ernste, E. B.; Rinkens, A.; van Uum, J. F.; Westerink, B. H. Disinhibition of the mediodorsal thalamus induces Fos-like immunoreactivity in both pyramidal and GABA-containing neurons in the medial prefrontal cortex of rats, but does not affect prefrontal extracellular GABA levels. Synapse 30:156 –165; 1998. 3. Cauli, B.; Audinat, E.; Lambolez, B.; Angulo, M. C.; Ropert, N.; Tsuzuki, K.; Hestrin, S.; Rossier, J. Molecular and physiological diversity of cortical nonpyramidal cells. J. Neurosci. 17:3894 –3906; 1997. 4. Celio, M. R. Parvalbumin in most ␥-aminobutyric-acid-containing neurons of the cat cerebral cortex. Science 231:995–997; 1986. 5. Celio, M. R. Calbindin D-28k and parvalbumin in the rat nervous system. Neuroscience 35:375–382; 1990. 6. DeFelipe, J. Neocortical neuronal diversity: Chemical heterogeneity revealed by colocalization studies of classic neurotransmitters, neuropeptides, calcium-binding proteins, and cell surface molecules. Cereb. Cortex 3:273–289; 1993. 7. Grassi-Zucconi, G.; Giuditta, A.; Mandile, P.; Chen, S.; Vescia, S.; Bentivoglio, M. c-fos spontaneous expression during wakefulness is reversed during sleep in neuronal subsets of the rat cortex. J. Physiol. Paris 88:91–93; 1994. 8. Grassi-Zucconi, G.; Menegazzi, M.; De Prati, A. C.; Bassetti, A.; Montagnese, P.; Mandile, P.; Cosi, C.; Bentivoglio, M. c-fos mRNA is spontaneously induced in the rat brain during the activity period of the circadian cycle. Eur. J. Neurosci. 5:1071–1078; 1993. 9. Herrera, D. G.; Robertson, H. A. Activation of c-fos in the brain. Prog. Neurobiol. 50:83–107; 1996. 10. Hiscock, J. J.; Mackenzie, L.; Willoughby, J. O. Laminar distribution of Fos/calcium-binding protein and Fos/neurofilament protein-labeled neurons in rat motor and sensory cortex after picrotoxin-induced seizures. Exp. Neurol. 149:373–383; 1998. 11. Kawaguchi, Y. Physiological subgroups of nonpyramidal cells with specific morphological characteristics in layer II/III of rat frontal cortex. J. Neurosci. 15:2638 –2655; 1995. 12. Kawaguchi, Y.; Kubota, Y. Correlation of physiological subgroupings of nonpyramidal cells with parvalbumin- and calbindinD28k-immuno- 638 13. 14. 15. 16. 17. BERTINI ET AL. reactive neurons in layer V of rat frontal cortex. J. Neurophysiol. 70:387–396; 1993. Kawaguchi, Y.; Kubota, Y. GABAergic cell subtypes and their synaptic connections in rat frontal cortex. Cereb. Cortex 7:476– 486; 1997. Mihaly, A.; Szente, M.; Dubravcsik, Z.; Boda, B.; Kiraly, E.; Nagy, T.; Domonkos, A. Parvalbumin- and calbindin-containing neurons express c-fos protein in primary and secondary (mirror) epileptic foci of the rat neocortex. Brain Res. 761:135–145; 1997. Montero, V. M. c-fos induction in sensory pathways of rats exploring a novel complex environment: Shifts of active thalamic reticular sectors by predominant sensory cues. Neuroscience 76:1069–1081; 1997. Morgan, J. I.; Curran, T. Stimulus-transcription coupling in neurons: Role of cellular immediate-early genes. Trends Neurosci. 12:459 – 462; 1989. Peng, Z.-C.; Chen, S.; Bentivoglio, M. A sensitive double immuno- 18. 19. 20. 21. staining protocol for Fos-immunoreactive neurons. Brain Res. Bull. 36:101–105; 1995. Pompeiano, M.; Cirelli, C.; Tononi, G. Immediate-early genes in spontaneous wakefulness and sleep: Expression of c-fos and NGFI-A mRNA and protein. J. Sleep Res. 3:80 –96; 1994. Staiger, J. F.; Bisler, S.; Schleicher, A.; Gass, P.; Stehle, J. H.; Zilles, K. Exploration of a novel environment leads to the expression of inducible transcription factors in barrel-related columns. Neuroscience 99:7–16; 2000. van Brederode, J. F. M.; Helliesen, M. K.; Hendrickson, A. E. Distribution of the calcium-binding proteins parvalbumin and calbindinD28k in the sensorimotor cortex of the rat. Neuroscience 44:157–171; 1991. Zilles, K. The cortex of the rat. Berlin: Springer-Verlag; 1985.