The D3 dopamine receptor (D3R) is selectively and transiently expressed in the barrel neurons of the somatosensory cortex (SI) between the first and second postnatal weeks. The D3R expression starts after the initial ingrowth of thalamocortical afferents (TCAs) into the barrel cortex and could be induced or controlled by them. We show that unilateral electrolytic lesion of the thalamic ventrobasal complex immediately after birth leads to a decrease in the D3R mRNA concentration in the lesioned SI 7 days after the lesion, whereas the D3R binding is little affected. Fourteen days after the neonatal thalamic lesion, the D3R binding and mRNA are drastically reduced and the barrel-like pattern of the D3R is absent. Elevation of the D3 binding normally seen between the first and second postnatal weeks does not occur. Thalamic lesion on P6 differentially affects the D3R expression. One day after the lesion, the D3 binding and mRNA are down-regulated, but the effect is transient. Five days after the lesion the concentration of D3 mRNA in the lesioned hemisphere returns to the control level. The typical barrel-like pattern of D3R expression is evident in the lesioned SI, although TCAs are completely absent. Quantitative analysis demonstrated elevated cellular levels of the D3 mRNA in barrel neurons 5 days after the lesion. These higher levels are needed, perhaps, to support the increased production of the D3R protein appropriate for this age. Age-related dynamics of the D3R binding is retained in the lesioned SI, although the concentration of D3R sites remains reduced. These data demonstrate that intact thalamic input is essential for the formation of mechanisms responsible for developmental regulation of the D3R expression in the SI.
One of the most intriguing questions of neurobiology is how individual cortical areas develop their specific characteristics. The rodent primary somatosensory cortex (SI) represents a good model with which to address this question due to its unique cytoarchitectecture. The SI contains a somatotopic map of the body surface that can be visualized by various techniques that stain either cortical cells or thalamocortical afferents (TCAs). Layer IV of the SI contains neuronal aggregates known as barrels (Woolsey and Van der Loos, 1970) that correspond to the pattern of vibrissae on the rodent snout. The posteromedial barrel subfield (PMBSF) contains five rows of large barrels that cor- respond to similar rows of large caudal whiskers called mystacial vibrissae. Smaller, more rostral vibrissae are represented in the SI by a constellation of small barrels rostral to the PMBSF. Tactile information from the vibrissae and oral structures is transmitted in a somatotopic manner via the trigeminal pathway to the ventrobasal complex (VB) of the thalamus and then to the SI (Van der Loos, 1976; Ma, 1991). The cellular and molecular mech- anisms that govern the development of the somatotopic map and establishment and maintenance of the periphery-related pattern in the SI are not yet well understood. TCAs are thought to play an important role because a vibrissae-related pattern of thalamocortical terminals can be visualized in the PMBSF with acetylcholinesterase histochemistry and anterograde labeling with the carbocyanine dye DiI as early as on postnatal day 1 (P1), 3 days before the appearance of cytoarchitectonically defined barrels (Schlaggar and O'Leary, 1994). Such an early patterning of TCAs suggests that they convey patterning information to the SI.
The importance of intact periphery for pattern formation in the SI is supported by a wealth of evidence, including observations that neonatal destruction of hair follicles, dissec- tion of the infraorbital nerve or lesion of the VB all disrupt the development of barrels. The cytoarchitecture of barrels can be altered by disrupting the cortical input during the first few postnatal days (so called ‘critical period’) but remains relatively immune afterwards [for review see Woolsey (Woolsey, 1990) and Kossut (Kossut, 1992)]. However, manipulations of the peripheral input later in development or even in adult animals lead to functional plasticity of barrel neurons without changes in the cytoarchitectural appearance of the barrel cortex [for review see Kossut (Kossut, 1992)].
In recent years, a number of proteins, including several neuro- transmitter receptors, have been found to be somatotopically distributed and more or less specific for the SI and/or barrel neurons. Some of them, such as serotonin (5-HT) uptake sites (Bennett-Clarke et al., 1994; Lebrand et al., 1996) and 5-HT1B receptors (Leslie et al., 1992) are located on TCAs and thus follow the distribution pattern of the terminals. Other proteins are expressed by the barrel neurons themselves, such as the GABAA α5 subunit (Paysan et al., 1997) and α7 nicotinic receptor (Broide et al., 1996). These proteins may play a role in the formation/maintenance of the structure of the barrel cortex. Indeed, 5-HT uptake sites expressed by thalamic neurons and located on TCAs regulate the level of serotonin in the barrel cortex and thus control the barrel formation (Cases et al., 1996) (also Dennis L. Murphy, personal communication).
Cortical development is sensitive to fluctuations of the dopamine (DA) level caused by the use of psychostimulants (Murphy et al., 1997) or genetic disorders (Diamond, 1996; Zagreda et al., 1999). DA has been shown to modulate directly the maturation of a subtype of cortical neurons predominantly via D2-like receptors (Todd, 1992; Reinoso et al., 1996; Porter et al., 1999). Recently, we have identified a unique expression of one member of the D2 DA receptor subfamily, the D3 receptor (D3R), in the whisker barrel cortex (Gurevich and Joyce, 2000). The D3R is found almost exclusively in the SI (low concentrations are also seen in the secondary somatosensory and auditory cortices) and is confined to layer IV. Layer IV of the SI does not express other DA receptors. The D3R binding sites and mRNA delineate precisely various areas of the SI. The expression of D3R is observed immediately after the barrel pattern becomes established: the D3R mRNA can first be detected at P4 and D3R binding sites at P6. The concentration of the D3R mRNA reaches maximum at P7, remains relatively unchanged until P14 and sharply declines afterwards. The expression of the D3R protein lags behind that of the D3R mRNA: D3R binding sites reach their maximum level by P14 then slowly decline. The D3R expression is not completely transient, as a discrete population of layer IV neurons expressing the D3R mRNA can be observed in the adult SI and low levels of D3R binding are also detectable, although the pattern becomes obscure.
It is unclear what factors control the expression of the D3R in the SI. The expression occurs after the initial ingrowth of TCAs into layer IV and thus could be induced or controlled by them. It is also unclear whether intact thalamocortical input is obligatory for the maintenance of the specific developmental profile of the D3R expression. To gain insight into the mode of thalamic control of the D3R expression, we used quantitative in situ hybridization histochemistry and D3 DA receptor autoradiog- raphy to study the D3R expression in the SI following the lesion of the VB at two time points. The lesion was done either shortly after birth, when TCAs begin to arborize in layer IV (Agmon et al., 1993) and when the D3R is undetectable in the SI, or on P6, when the barrel pattern in the SI is well established and the D3R is already expressed by the barrel neurons. The results demonstrate an age-dependent dynamic regulation of the D3R expression by the TCAs.
Materials and Methods
Placement of Lesions
Newborn (P1) rat pups were anesthetized by hypothermia; those aged P6 were anesthetized with Metofane (Pitman-Moore, Mundelein, IL). Unilateral lesions of the VB complex of the thalamus were made using a modified stereotaxic approach. The tip of a stainless-steel electrode, with 10–20 μm exposed at the tip, was placed in the VB complex using the following stereotaxic coordinates: for P1 rat pups, the electrode was inserted 1.5 mm lateral from the midline suture and 3.0 mm beneath the pial surface. For P6 pups, the position was 2.1–2.3 mm lateral from the midline suture and 3.5–3.7 mm beneath the pial surface. Electrolytic lesions were made by passing a 0.9 mA positive current for 20 s. Animals were sutured and allowed to recover from the anesthetic before being returned to the dam. All experiments were carried out to minimize animal suffering; these measures were in accordance with the National Institute of Health Guide for the Care and Use of laboratory animals and were approved by the University of California, Irvine, Animal Resource Committee.
Rats were decapitated at appropriate ages following the thalamic lesion, and their brains rapidly removed. Cerebral cortices were separated from the brainstem, flattened between two microscopic slides, and then frozen by immersion in isopentane at a temperature below –20°C. For in situ hybridization and binding experiments, 15 μm thick sections from flattened cortices were cut on a cryostat and mounted on Probe On Plus slides (Fisher). Even sections were dried and stored at –80°C to be used in the binding experiments. Odd sections were postfixed in 4% para- formaldehyde for 1 h at 4°C, washed, dehydrated and stored at –80°C until used for in situ hybridization experiments. A tissue block, which included the thalamus, from lesioned brains was frozen and sectioned on a cryostat for lesion verification. Tissue sections (20 μm thick) were mounted onto gelatin-coated slides. Slide-mounted tissue sections for Nissl staining were postfixed with 10% buffered formalin and stained with cresyl violet.
D3 Receptor Binding
[125I] (R)-Trans-7-hydroxy-2-[N-propyl-N-(3′-iodo-2′-propenyl)-amino]tet- ralin ([125I]trans-7-OH-PIPAT) (New England Nuclear, Boston, MA) was used with conditions that allowed for selective labeling of the D3R as described earlier (Gurevich et al., 1999; Gurevich and Joyce, 2000). Briefly, sections were incubated with [125I]trans-7-OH-PIPAT at a concentration of 1.6 nmol (Kd = 0.5 nM) in the presence 100 μM Gpp[NH]p for 1 h at room temperature and washed at 4°C for 3 h with a change of wash buffer (50 mM Tris–HCl with 40 mM NaCl) every hour. 7-OH-DPAT was used to define non-specific binding (10 μM). The slides along with tritium autoradiographic standards were exposed to 3H-Hyperfilm (Amersham, Arlington Heights, IL) for 7 days. Three or four sections for total and two sections for non-specific binding were used for each animal. A higher concentration of the radioligand was necessary to detect low levels of the receptor at early ages, and therefore 1.6 nM (~4 × the Kd) was used for all ages in all experiments. At the nearly saturating concentration of the radioligand used, differences in binding values predominantly reflect changes in Bmax. The [125I]trans-7-OH-PIPAT binding in most cortical regions outside of the SI (with the exception of the secondary somatosensory and auditory cortices) was at the same level as non-specific binding. Visual cortex never displayed specific binding to the D3R, so we measured total [125I]trans-7-OH-PIPAT binding in the visual cortex and treated it as non-specific binding for each section individually. The measurements in the visual cortex for each section were subtracted from the values in the barrel field for the same section to obtain the specific binding values. Adjacent frozen sections were processed for binding with [125I]RTI-55 as described previously (Gurevich and Joyce, 2000). [125I]RTI-55 labels 5-HT transporter, which is localized on thalamocortical terminals (Lebrand et al., 1996). 5-HT uptake sites follow the specific distribution of the thalamocortical terminals in the cortex and are used here to visualize different cortical areas and to appreciate the extent of the thalamic lesion.
In Situ Hybridization Histochemistry (ISHH)
The riboprobe for the D3R and the ISHH procedure have been described previously (Gurevich et al., 1999; Gurevich and Joyce, 2000). Briefly, the D3R riboprobe was 423 bp in length and corresponded to the third cytosolic loop and part of VI TM. The riboprobe was labeled in standard in vitro transcription reaction with [33P]UTP (New England Nuclear). Prior to hybridization, sections were treated with proteinase K (1 μg/ml, 5 min at 37°C) and 0.1 M triethanolamine buffer, pH 8.0, containing 0.25% acetic anhydride, washed, and dried. Approximately 1 × 106 c.p.m. of the probe in 25 μl of hybridization buffer, pH 7.2, containing 75% formamide, 50 mM Tris–HCl, 2.5 mM EDTA, 4 × SSC, 10% dextran sulfate, 1 mg/ml tRNA, 1 mg/ml calf thymus DNA were applied per section. Hybridization was carried out at 55°C for 20 h. Upon completion of hybridization, sections were subjected to RNase digestion and washes of increased stringency. Dried sections were exposed to β-Max film (Amersham) for 1 month. Film was developed with Kodak GBX developer and fixer. Sections were then cleared in xylene, dipped in nuclear emulsion (Ilford, Paramus, NJ) and exposed for 8 weeks at 4°C. The sections were developed, counterstained with Nuclear Fast Red, dehydrated and mounted from xylene with Permount. Labeling with the sense probe in all areas or with the antisense probe in white matter and in most cortical areas outside of the SI was negligible. Three or four sections for antisense probe and two sections for sense probe hybridization were used for each animal. Non-specific hybridization signal was estimated in the same way as for the receptor binding: measurements in the visual cortex were treated as non-specific hybridization and were subtracted from the measurements in the barrel field from the same sections to obtain the values for the specific hybridization in the SI.
Autoradiographic images of [125I]trans 7-OH PIPAT binding and in situ hybridization were analyzed using BRAIN for Macintosh software (Drexel University, Philadelphia, PA). Only the PMBSF was analyzed. The optical density values derived from the autoradiograms of [125I]trans 7-OH PIPAT binding were converted to fmol/mg protein using 3H standards calibrated for 125I. For the D3R mRNA autoradiography, optical density values were converted to c.p.m. per 10 000 sq pixels using 14C standards calibrated for 33P. Emulsion-coated sections were analyzed essentially as previously described (Gurevich and Joyce, 2000). Briefly, images were collected at 40× magnification with the aid of C-Imaging systems software Simple 32 (Compix Inc., Cranberry Township, PA). The PMBSF was outlined with the help of images of the [125I]RTI-55 binding and D3 mRNA. To determine the exact position of the PMBSF in lesioned hemispheres, where the PMBSF pattern of the [125I]RTI-55 binding was absent, we used remaining signals of the D3 binding or mRNA and/or positions of other areas of the SI that remained intact. The area of the PMBSF was measured in the postero-medial to antero-lateral direction, and 10 roughly equidistant viewfields were selected on each section in a stepwise fashion using a motorized, computer-controlled stage. Manual adjustments were only made if an automatically selected viewfield happened to be in the area of the septae, which contains few, if any, D3 mRNA-positive cells (Gurevich and Joyce, 2000). In such cases, the stage was moved to the nearest barrel. In lesioned hemispheres, where the barrel cytoarchitecture was often absent or indistinct, only automatic veiwfield selection was applied. Automatic counting within two or three sections per hemisphere for each animal allowed for sampling of 1500–3500 cells per hemisphere (depending on age) per animal. Background labeling was determined by grain counting in the visual cortex individually for each section (five viewfields per section). Grain data were averaged for all cells of all background viewfields for each section. Cells that had a grain number exceeding the mean ± 3 SD were excluded (<4% of all cells). The resulting mean number of grains was used in all subsequent calculations as the background.
SuperANOVA software (Abacus Concept, Inc., Berkeley, CA) was used for the statistical analysis. Autoradiographic binding and ISHH data were analyzed by repeated-measures ANOVA with Treatment (Control versus Lesioned hemisphere) as the within-group factor. The difference was considered significant for P < 0.05. The data from the emulsion-coated sections were analyzed in two ways. First, the number and percentage of positive cells (the cells with at least twice the background number of grains) was calculated for each viewfield for each animal. Second, the number of grains per positive cell was calculated for each viewfield for each animal. The data were analyzed statistically with two-way ANOVA with Treatment (Control versus Lesion) as the main factor and Animal as a blocking factor. Using Animal as a blocking factor helped to account for individual variability including variability caused by litter and sex differences.
Emulsion-coated sections were viewed and photographed with a Nikon Optiphot microscope equipped with Olympus DP-10 digital photo- camera. Images acquired with the camera were transferred to Adobe Photoshop 3.0 (Macintosh) for labeling and assembly. Autoradiograms of the D3R binding and mRNA were scanned directly in Adobe Photoshop 3.0 (Macintosh) at 300 pixels/in. resolution in the inverted mode and not otherwise modified.
Effects of Thalamic Lesion at P1 on the D3R Expression
Histological verification of lesion placement revealed that the lesion was limited to the VB in all cases (Fig. 1A,B), and its extent varied from that limited to the central VB to complete destruction of the VB. The lesion led to the expected massive reduction in [125I]RTI-55 labeling of 5-HT transporters located on TCAs, so that the SI was virtually devoid of labeling (Fig. 1C,D). The lesion of the VB at P1 resulted in profound alterations in the D3R expression that was apparent macroscopically and micro- scopically. In the group that survived for 7 days following the lesion and was analyzed at P8, the concentration of D3 binding sites in the barrel cortex, which is normally quite low at this age, was not significantly affected by the lesion [F(1,7) = 0.7, not significant] (Figs 2A,B and 3). Conversely, the concentration of the D3 mRNA in the lesioned barrel cortex was reduced by 42.8% as compared to the control hemisphere [F(1,7) = 9.65, P < 0.02] (Fig. 2C,D and 3). In normal development, the D3 mRNA expression reaches its peak at this age, so the lack of a distinct barrel like pattern in the lesioned hemisphere was easily discernible although the signal was appropriately positioned (Fig. 2D). When examined at the microscopic level, the appear- ance of the barrel neurons in the control hemisphere was typical for 8 days of age (Fig. 4A). Barrels were easily identifiable on counterstained, emulsion-coated sections as clusters of cells, most of which were labeled for the D3R mRNA. As previously described (Gurevich and Joyce, 2000), cells within the barrel center are most heavily labeled. In the lesioned hemisphere, barrels were not easily identified at low magnification. At high magnification, clusters of barrel cells, although present, were less dense and the general concentration of the D3R mRNA appeared reduced (Fig. 4B). Quantitative cellular analysis showed a small reduction in the total number of cells per viewfield [by 4.5%, F(1,374) = 18.2, P < 0.001] (Fig. 5A). The analysis of the D3 mRNA expression confirmed the significant reduction in the proportion of D3 mRNA-positive cells [by 15.2%, F(1,374) = 85.1, P < 0.001] (Fig. 5B) and in average grain density per cell [by 11.8%, F(1, 10 000) = 335.2, P < 0.001] (Fig. 5C).
The effect of the P1 VB lesion in the animals surviving for 14 days following the lesion (and analyzed at P15) differed from that in the animals that survived for 7 days. As appropriate for this age, the control hemisphere exhibited a high concentration of D3Rs, and these were organized in a well-defined, barrel-like pattern (Fig. 2E). The expression of the D3 mRNA was lower than at P8, although the concentration remained relatively high and the pattern quite distinct (Fig. 2G). At this time point, the impact of the lesion on the D3 binding was quite pronounced with many animals exhibiting very low binding, and the typical barrel-like arrangement of the D3R was absent (Fig. 2F). The average concentration of D3 binding sites on the lesioned side was decreased by 62% [F(1,6) = 33.3, P < 0.002] (Fig. 3). In addition, the normal developmental profile of the D3R binding was disturbed. In the control hemisphere, the concentration of D3 binding sites in the P15 group (14 days survival after P1 lesion) was more than two times higher than in the P8 group (7 days survival), reflecting normal postnatal dynamics of D3R expression. No such increase was observed in the lesioned hemisphere, as the binding remained virtually the same at P15 as at P8 (Fig. 3).
Expression of D3 mRNA was also strongly affected at 14 days after the P1 lesion. The concentration of the D3 mRNA in the lesioned hemisphere was about half (reduced by 53.5%) of that on the control side [F(1,5) = 7.89, P < 0.05] (Figs 2H and 3). The effect of the lesion on the D3 mRNA was exacerbated by longer survival time. The reduction in the D3 mRNA concentration was greater at 14 than 7 days after the lesion as compared to the age-appropriate level in the intact hemisphere. In most cases, the somatotopic pattern of the D3R expression was absent on the lesioned side; instead the D3 mRNA signal, like that of binding sites, had a diffuse appearance (Fig. 2F,H). In the rat, the barrels at this age (P15) are difficult to detect based on cytoarchitecture alone. In the intact hemisphere, overall cell density decreased as compared to P8, and cell clustering and the difference in cell density between the barrels and septae became less conspicuous (data not shown). On the lesioned side, the barrels were not identifiable, and there was an obvious reduction in the labeling for the D3 mRNA. Cells intensely labeled for the D3 mRNA typical for the intact hemisphere were rarely seen on the lesioned side (Fig. 4C,D). Quantitative analysis of grains confirmed these observations. As in the previous group, we found small but significant decrease in total cell density [by 6.2%, F(1,378) = 26.5; P < 0.001] (Fig. 5A).The proportion of positive cells [F(1,378) = 142.9, P < 0.001] (Fig. 5B) and number of grains per cell [F(1,10 000) = 621.5, P < 0.001] (Fig. 5C) were significantly reduced by 28.5 and 24.5% respectively.
Effects of the Thalamic Lesion at P6 on the D3R Expression
The lesion of the thalamic VB complex performed at P6 produced effects different from those of the lesion at P1. In the group that survived 1 day after the lesion, a drastic reduction in the concentration of D3 binding sites (Fig. 6A,B) and mRNA (Fig. 7A,B) was observed, in some cases to barely detectable levels. The concentration of D3 binding sites was decreased in the lesioned hemisphere by 53.4% [F(1,9) = 15.8, P < 0.01] (Fig. 8). The concentration of the D3 mRNA was down-regulated by 38.5% [F(1,12) = 31.5, P < 0.001] (Fig. 8). On the microscopic level, the barrel arrangement of neurons on the lesioned side was not evident, with barrels and septae indistinguishable (data not shown). Typical clustering of cells with high concentrations of grains for D3R mRNA was not visible on the lesioned side (Fig. 4E,F). There was a statistically significant reduction in the total number of cells per viewfield [by 12.6%, F(1,514) = 223.0, P < 0.001] (Fig. 9A). Quantitative measurements confirmed that the proportion of cells positive for the D3R mRNA was reduced by 30.3% [F(1,514) = 821.9, P < 0.001] (Fig. 9B) and the average grain density by 28.8% [F(1,10 000) = 899.6, P < 0.001] (Fig. 9C).
Unlike the situation with the P1 lesion, longer survival times after the P6 lesion lead to restoration of the D3R expression in the SI, including the age-appropriate somatotopic pattern. A barrel-like patterning of D3R binding sites and mRNA was evident in the lesioned hemisphere 5 (Figs 6C,D and 7C,D) and 11 days after the lesion (Figs 6E,F and 7E,F). In fact, the concentration of the D3 mRNA was virtually the same in the lesioned and control hemispheres at 5 [F(1,6) = 3.72, P = 0.1 NS] and 11 days [F(1,7) = 2.37, P = 0.17 NS] after the P6 lesion (Fig. 8). In contrast, the D3R binding was reduced by 33.5% at 5 days [F(1,7) = 13.9, P < 0.01] and by 49.4% at 11 days [F(1,7) = 11.6, P < 0.02] after the lesion (Fig. 8). Even though the thalamic lesion on P6 reduced the levels of D3R binding sites, it did not alter the age-related dynamics of D3R expression. The D3 binding increased between P7 (survival for 1 day) and P11 (5 days survival) and then decreased by P17 (11 days survival) in the lesioned as well as in the control hemisphere.
The cytoarchitectural appearance of barrel neurons in the lesioned hemisphere was quite normal 5 and 11 days after the P6 lesion. In contrast with the group that survived for 1 day, there was no reduction in the total number of cells per veiwfield 5 or 11 days after the lesion (Fig. 9A). Five days after the lesion, at P11, the barrels were well defined, with clusters of cells densely labeled for the D3 mRNA evident in both hemispheres (Fig. 4G,H). Quantitative analysis demonstrated that, although the proportions of D3 mRNA-positive cells (Fig. 9B) were similar in the control and lesioned hemisphere, the average grain density per positive cell was slightly but significantly increased in the lesioned hemisphere as compared to control [6.2%, F(1,10 000) = 11.7, P < 0.001] (Fig. 9C). Eleven days after the lesion, at P17, there was an age-related decrease in overall cell density and in the concentration of the D3 mRNA. There were no apparent differences between the control and lesioned hemispheres in cytoarchitecture or the cellular level of the D3 mRNA expression. Quantitative analysis confirmed that the proportion of D3 mRNA positive cells and the number of grains per cell were similar in both hemispheres (Fig. 9B,C).
We have analyzed in detail the effects of electrolytic lesion of the thalamic VB complex and consequent loss of TCAs on transient D3R expression the barrel cortex. Because the D3R expression is quite tightly developmentally controlled, the lesion may have affected not only the level of expression at any given point, but also the regulatory mechanisms themselves. Our results indicate that the effect of the lesion depends on its timing and the survival period after the lesion. The P1 lesion of the thalamic input to the cortex led to profound alterations in the expression of the D3R. The D3 mRNA was significantly reduced 7 days and further decreased 14 days after the lesion. The D3R binding was little affected 7 days after the lesion but substantially reduced after longer survival periods. The effect of the thalamic lesion at P6 on D3R expression in the SI differed from the effect of the neonatal thalamic lesion. The D3 mRNA concentration dimin- ished considerably 1 day after the lesion, but later returned to normal levels. Five days after the lesion, barrel neurons of the lesioned hemisphere even increased their cellular concentration of the D3 mRNA as compared to the intact side, perhaps as a compensatory measure necessary to support the elevated pro- duction of the D3R protein evident at this age. However, the D3R binding remained down-regulated during the entire period of transient D3R expression in spite of normal mRNA levels.
What role do the TCAs play in specification of the cytoarchi- tectonical and neurochemical identity of the somatosensory cortex and in regulating the expression of its markers such as the D3R? Formation of the barrels is known to depend on the integrity of the pathways from peripheral receptors through trigeminal and thalamic relay nuclei to the cortex. It is a well-established fact that neonatal destruction of the VB results in failure of the barrels to form (Wise and Jones, 1978; Broide et al., 1996; Paysan et al., 1997). Neonatal lesion of the whisker follicles or dissection of the infraorbital nerve have similar effects (Jeanmonod et al., 1981; Jensen and Killackey, 1987; Vos et al., 1990). These data, together with transplantation studies (Shlaggar and O'Leary, 1991), led to the conclusion that TCAs convey ‘patterning’ information to the developing cortex. Recent data obtained in mice carrying mutations that produce ‘barrelless’ phenotypes support the idea that proper topo- graphic targeting and arborization of the thalamocortical axons is essential for normal barrel formation (Cases et al., 1996; Welker et al., 1996; Abdel-Majid et al., 1998; Maier et al., 1999).
There is little information as to what happens to layer IV neurons themselves when they are deprived of appropriate thalamic input and fail to segregate into barrels. The barrel neurons seem to survive the perinatal destruction of thalamo- cortical terminals (Wise and Jones, 1978; Windrem and Finlay, 1991). The minor reduction in cell density following the thalamic lesion observed in this study seems more likely due to our sampling procedure than to actual cell loss. Theoretically, the absence of cytoarchitectonically defined barrels does not per se mean that layer IV neurons should be unable to express their intrinsic markers. For example, layer IV neurons somatotopically express a specific transgenic marker H-2Z1 from P2 with the highest expression during the second postnatal week (Cohen-Tannoudji et al., 1994). The transgene is expressed normally in mice that have no barrels due to an excess of serotonin [monoamine oxidase A knockout mice (Cases et al., 1996)] or following upper lip coagulation (Gitton et al., 1999a). Thus, barrel neurons must have a certain degree of functional autonomy from the thalamic input, but the extent of this self- determination is unknown. It is possible that thalamic afferents, even mistargeted and unsegregated, are able to support certain aspects of barrel neuron physiology as long as they actually arborize and make synaptic contacts within layer IV. There is growing evidence that many aspects of early cortical regional- ization and lamination are largely independent of thalamic influence relying instead on intrinsic cues (Nothias et al., 1998; Miyashita-Lin et al., 1999; Rubenstein et al., 1999; Bishop et al., 2000). It is unclear, however, whether postnatal neurochemical markers normally associated with defined barrels could be expressed without any thalamic input. Expression of the transgenic marker H-2Z1 by barrel neurons when grafted heterotopically (Cohen-Tannoudji et al., 1994) or in the prenatal cortex explant culture (Gitton et al., 1999b) is the only evidence to date demonstrating autonomous expression of a marker for barrel neurons. The H-2Z1 is not expressed in vivo until P2, but, evidently, the lamina- and region-specific manner of its expres- sion is determined as early as at E12.5 (Cohen-Tannoudji et al., 1994; Gitton et al., 1999b). Our data suggest that the expression of D3R in vivo can be initiated in the absence of thalamic input but it cannot be sustained. Most of the available data stress the importance of intact periphery for the in vivo expression of postnatal somatosensory markers. Neonatal lesion of the whisker follicles interferes with normal down-regulation of the number of somatostatin-immunoreactive cells (Parnavelas et al., 1990) and leads to a loss of GABA immunoreactivity (Kossut et al., 1991) in the SI. Neonatal lesion of the VB thalamic complex reduces the expression of α5 subunit of the GABAA receptor (Paysan et al., 1997) and the α7 nicotinic acetylcholine receptor (Broide et al., 1996). Surprisingly, in contrast to embryonic cortical explants that independently initiate the expression of the barrel-specific H-2Z1 transgene in vitro, the expression does not occur in early postnatal cortical explants (Gitton et al., 1999a). Additionally, in vivo, neonatal lesion of the thalamic VB nucleus eliminates the H-2Z1 expression in the SI (Gitton et al., 1999a). The authors suggest that an abrupt change in the developmental properties of the cortex takes place around birth that places the previously independent regionalization process under control of thalamic afferents. The present data indicate that in vivo expression of the DA D3R, a very selective and natural marker of the SI, depends heavily upon integrity of the thalamic input. Apparently, the mechanisms that ensure the maintenance of its specific age-related pattern of expression are established before or at the time of initiation of the D3R expression. Intact thalamocortical input seems indispensable for the formation of these mechanisms.
For the destruction of the periphery to be effective in inducing malformations of the barrel cortex, it must be done during the first few days after birth, the time known as the ‘critical period’. The same manipulations done later than P4 do not affect the barrel pattern of SI (Jensen and Killackey, 1987; McCasland et al., 1992). Whether or not the thalamic input still exerts an important influence on the barrel cortex after the critical period is unclear. Section of the infraorbital branch of the trigeminal nerve at P7 does not alter the formation of barrels but prevents intracortical connections from maturing (McCasland et al., 1992). It has also been shown that lesion of the VB in 7–8 day old rats delayed the development of the normal pattern of parvalbumin and calbindin expression in the barrel cortex (Alcantara et al., 1996). The effect of the VB lesion at P6 on the expression of the α7 nicotinic acetylcholine receptor is pro- found but transient, as the expression of both mRNA and protein is back to normal in 3–4 days (Broide et al., 1996). This is similar in many respects to what we observed for the DA D3R, but differences do exist, particularly in the vulnerability of the receptor protein production to the loss of TCAs. Generally, there is no reason to suppose that the requirements for thalamic input are the same for all markers expressed by the barrel neurons or for all aspects of their function. Nonetheless, our data add credence to the hypothesis that TCAs regulate important aspects of maturation of the SI long after the barrel formation.
Our data demonstrate that there is indeed a critical period shortly after birth when the spatial and temporal characteristics of D3R expression in the SI are established. TCAs appear to play an essential role in this process. Ablation of the thalamic input during this period does not prevent the initiation of D3R expression but precludes its transient elevation found in the intact cortex between the first and second postnatal week. Lesion of TCAs after the barrels develop results in a transient loss of the D3R. Several days after the lesion, however, a recovery process takes place leading to normal or even exaggerated expression of the D3 mRNA and restoration of the somatotopic pattern. Normal age-dependent dynamics of the D3 mRNA and binding sites is preserved, although D3 binding sites remain reduced during the entire period of the transient D3R ex- pression. Evidently, some kind of thalamic control is necessary even during the second postnatal week to support normal production of the D3R protein. Alternatively, immaturity of local intracortical circuits resulting from the removal of thalamic input (McCasland et al., 1992) may be the culprit.
This work was supported by NIH grants MH 56284 and AG 09215.
Address correspondence to Jeffrey N. Joyce, Sun Health Research Institute, 10515 West Santa Fe Drive, Sun city, AZ 85351, USA. Email: email@example.com.