Main

The basal ganglia, bilateral nuclei of subcortical gray matter that include the striatum (caudate and putamen) and pallidum (GPe and GPi), are important in motor and cognitive programs. In a model of Lesch-Nyhan syndrome, early loss of dopaminergic afferents to the basal ganglia causes rats to display biting and self-mutilating behavior(1); similar lesions in adult rats do not produce these behaviors. This and other lines of evidence indicate that dopamine plays a specialized role in brain development(2). In addition, stimulation of dopamine receptors(primarily D1) in rats has been shown to stimulate IEGs, a group of genes that regulates transcription of other genes(3).

The principal subtypes of dopamine receptors in the basal ganglia appear to be D1 and D2. Molecular biologic studies have demonstrated additional, more rare subtypes that fit into D1-like (D5) and D2-like (D3 and D4) families(4). There is little or no mRNA for the D3, D4, and D5 subtypes in the striatum and globus pallidus(5), suggesting that these receptors are not present on neuronal cell bodies in these regions. The fact that there are no selective pharmacologic ligands or antibodies with which to label these receptors(4) limits studies of these receptors at this time.

Little information has been available about dopamine receptors in the basal ganglia of the human infant. D2 receptors have been detected in fetal striatum studies (by HB) as early as 16 wk of gestation(6). Binding to both D1 and D2 receptors has been observed in homogenates of caudate and putamen from children less than 1 y of age (0-0.8 y; HB)(7). However, information has been unavailable in the perinatal brain concerning the intraregional organization of dopamine receptors within the striatum and globus pallidus.

The striatum is organized into discrete patches, or striosomes, embedded within a background matrix(8). The striosomal and matrix compartments have been found to differ with regard to their neurochemical, receptor, connection, and functional characteristics(9). Therefore, it is important to determine whether dopamine receptor subtypes in the human striatum are preferentially localized in subregions and/or compartments during the perinatal period, and whether the regional and/or compartmental preferences differ in the perinatal, compared with adult, striatum. In the rat striatum, rostrocaudal and mediolateral gradients of D1 and D2 receptors have been reported(10).

The globus pallidus also plays a pivotal role in basal ganglia function. The balance of activity between the GPe and GPi is critical for normal control of behavior(9). This intrapallidal balance is controlled, in part, by dopaminergic afferents from the substantia nigra(11). In QA studies in adult humans, D1 receptor levels have been reported to be greater in the GPi than in the GPe(12), whereas the opposite is true for D2 receptors, i.e. D2 density is greater in the GPe than in the GPi(13). A prerequisite for understanding the maturation of the interpallidal balance in humans is a characterization of dopamine receptors in the neonatal human pallidum.

An understanding of the localization of D1 and D2 dopamine receptors in the perinatal basal ganglia is also relevant because developing humans are increasingly exposed to cocaine(14). Cocaine exerts major effects by blocking the re-uptake of dopamine and norepinephrine at the synapse, thus prolonging their actions. In some studies, prenatal cocaine exposure has been found to cause abnormal motor behavior that is evident immediately after birth(15) and abnormal behavioral development that is seen in 2-y(16) and 3-y(15) follow-up studies of infants.

To examine D1 and D2 dopamine receptors, detailed QA was used to examine the distribution of D1 and D2 receptors in the striatum and globus pallidus of the perinatal and adult human. Our finding of strikingly elevated levels of D1 receptors in the perinatal GPi (and, to a lesser extent, the GPe and striatum) suggests that this region may be particularly vulnerable to agents acting through D1 receptors, including cocaine.

METHODS

Tissue. Brains were obtained at autopsy and frozen within 24 h of the time of death. When this work was done, there was no institutional requirement for review of studies using postmortem tissue. Three adult (67, 70, and 87 y of age) and four perinatal (0 d (32 wk gestational age), 2 d, 4 d, and 7 wk postnatal) brains were obtained locally; sections from two additional adult brains (70 and 71 y of age) without neurologic disease were obtained from the National Neurologic Research Bank (VA Medical Center West Los Angeles, Los Angeles, CA). The cause of death in each of the perinatal cases was congenital heart disease; cases with gross brain abnormalities and history of HIV or hepatitis were excluded by the neuropathologist, but no further details were available. The causes of death in the adult controls were cardiac arrest (two), cancer, pneumonia, and interstitial lung disease; the local donors were screened for history of neurologic disease, HIV and hepatitis, and recent dopamine agonist or antagonist exposure. The brains were cut into 3-4-cm coronal slabs, frozen in dry ice snow, and stored in aluminum foil at -70 °C until needed. Coronal frozen sections (10 μm) were cut from the full available rostrocaudal extent of the structures, i.e., 6-16 levels in each of the adult human brains and from 5-8 levels in each of the perinatal brains. Sections were collected at 1.5-mm intervals, beginning at the head of the caudate rostral to the putamen and continuing through the pallidum to just rostral to the lateral geniculate nucleus(17), as available. Sections were thaw-mounted onto gelatin-coated slides in the following sequence: D1 nonspecific binding, D1 total binding, [3H]diprenorphine (for striosome labeling in striatal sections), D2 total binding, D2 nonspecific binding. The tissue sections were stored (<1 mo) at -70 °C before assay. Unless otherwise specified, all chemicals were obtained from Sigma Chemical Co. (St. Louis, MO).

Protocol. Labeling of receptors was carried out using a single concentration of radioligand, rather than performing saturation binding analysis, to allow detailed anatomic comparisons between and within structures. “Total” slides were incubated in the presence of the radiolabeled drug in buffer. “Nonspecific” slides were incubated in the same solution used for the “total” slides, but with the addition of a saturating concentration of a nonlabeled drug with high affinity for the receptor of interest(18).

At the time of the assay, the slides to be labeled were removed from the freezer and left at room temperature for 20 min to thaw and dry. Labeling of D1 receptors was carried out in buffer containing 50 mM Tris, 10 mM MgSO4, 2 mM EDTA, 154 mM NaCl, and 10 mg/L BSA, with the pH adjusted to 7.4 at room temperature. Single-concentration binding was carried out with 1.3 nM 3H-SCH-23390 (3H-SCH, Amersham Corp., Arlington Heights, IL). Binding of 3H-SCH to 5-HT2 receptors was blocked with 300 nM ketanserin (Janssen Research Foundation, Beerse, Belgium). Nonspecific binding was defined with 2 μM (+)-butaclamol (Research Biochemicals Inc., Natick, MA). Sections were incubated at 37 °C for 100 min and washed in D1 buffer at 4 °C for 20 min [method adapted from Boyson et al.(10)].

Labeling of D2 receptors was carried out in buffer containing 20 mM N- 2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid, 154 mM NaCl, 5 mM EDTA, and 10 mg/L BSA adjusted to pH 7.5 at room temperature. Single-concentration binding was carried out with 1.3 nM [3H]spiperone(3H-SPIP, Amersham). Binding of 3H-SPIP to 5-HT2 receptors was blocked with 300 nM ketanserin. Nonspecific binding was defined by 2 μM (+)-butaclamol. Sections were incubated at 37 °C for 100 min and washed in D2 buffer at 4 °C for 80 min [method adapted from Boyson et al.(10)].

Striosomes were labeled in a 1.3 nM solution of [H]diprenorphine (a nonselective opiate receptor antagonist, Amersham) made up in D1 buffer that did not contain ketanserin. The sections were incubated at room temperature for 90 min and subsequently rinsed in ice-cold D1 buffer for 20 min [adapted from Ho et al.(19)].

After rinsing in each case, the labeled tissue sections were dipped briefly into ice-cold distilled water to remove buffer salts, air-dried, mounted in x-ray cassettes along with tritium standards (American Radiolabeled Chemicals, Inc., St. Louis, MO), and exposed to tritium-sensitive film (Amersham Hyperfilm) at room temperature for 12 wk. In addition, perinatal sections containing globus pallidus were reexposed for 2 wk to avoid film saturation, and these films were used for analysis.

Data analysis. The autoradiographs were analyzed using the DUMAS Image-Processing System (Tretiak et al. Drexel University, Philadelphia, PA). Before tissue analysis, standard curves were generated from the film images of the tritium standards. The following striatal regions on each “total” and “nonspecific” section were analyzed: total caudate, caudate quadrants (delimited by vertical and horizontal lines so that approximately 25% of the area was in each quadrant), total putamen, putamen quadrants, and the striatal compartments (striosome/matrix) in both the caudate and putamen. Striosomal measurements were made by 1) capturing the D1 or D2 total or nonspecific section to the first image buffer, 2) aligning the adjacent “striosome” slide([3H]diprenorphine) in the second image buffer, 3) drawing outlines of the striosomes on this second image, and 4) transferring these outlines to the matching D1 or D2 image to be analyzed. Matrix measurements were made by excluding all striosomal regions in a section and analyzing the remaining area as a single compartment. Striosomal/matrix measurements were made on four of the five adult and three of the four perinatal brains. An average of 10-13 striosomes in each section of caudate and 6-7 striosomes in each section of putamen were analyzed. Quadrant measurements in the caudate were made on two of the five adult and three of the four perinatal brains, whereas quadrant measurements in the putamen were made on three of the five adult and three of the four perinatal brains. Quadrant analysis on the remaining adult and perinatal brains was not feasible due to morphologic constraints. Data from only four adult brains are reported for the GPi due to the absence of this segment in the tissue available for one adult case.

Statistics. The data were analyzed by analysis of variance with post hoc Scheffe F test (adult versus perinatal and quadrant comparisons) and paired t test (striosome versus matrix comparisons; Statview, Brainpower, Inc., Calabasas, CA), except for the D1 results in Table 3. Analysis of variance was not valid for Table 3 because the data were not normally distributed as a whole. In the adult versus perinatal comparisons, these data required the use of the nonparametric Mann-Whitney rank sum test; the parametric t test was valid for the within-brain comparisons (Sigmastat, Jandel, San Rafael, CA).

Table 3 D1 dopamine receptor binding in human striatum
Full size table

A calculated, true ratio of D1:D2 receptors in a given structure can be determined by taking into consideration the fractional occupancy of receptors. This is necessary because binding was carried out at nonsaturating radioligand concentrations (to decrease nonspecific binding). The true or “maximal” density of receptors (Bmax) can be calculated using the equation Bmax = (measured value)·(L + Kd)/L, where L = the radioligand concentration [1.3 nM for both 3H-SCH and3 H-SPIP in this study; see Limbird(18)]. The Kd values can be estimated from those obtained in rat caudate-putamen and reported by Boyson et al.(10) (D1, 1 nM; D2, 0.5 nM). This yields a correction factor of 1.77 for D1 receptors and 3.0 for D2 receptors. The reported values should be multiplied by the correction factor to obtain a calculated Bmax for each receptor, and these calculated values then divided to obtain a true D1:D2 ratio. Thus, the calculated ratio of D1:D2 receptors in adult caudate is 0.30.

RESULTS

Globus Pallidus

D1 receptors. The mean density of D1 receptors in the perinatal GPe was nearly 7-fold greater than the mean density of D1 receptors in the adult GPe (Fig. 1; Table 1). An even more striking discrepancy between perinatal and adult values was observed in the GPi. In this structure, the density of D1 receptors in the perinatal cases was 15 times greater than the D1 receptor density in the adults. In Figure 3, it is evident that the 7-wk D1 values are closer to the adult than to the younger perinatal cases. When the 7-wk values were excluded from analysis, only the values in the GPi became statistically more significant(p = 0.0001). Within each age group, D1 receptor levels varied across the two pallidal components, with the GPi exhibiting higher D1 receptor levels than the GPe by 4 (adult)-to 8 (perinatal)-fold.

Figure 1
figure 1

D1 and D2 dopamine receptors in adult and perinatal globus pallidus. Pseudocolor-enhanced images of representative coronal sections through the globus pallidus of adult (a, b, e, f) and perinatal (c, d, g, h) human brains. C, caudate;P, putamen; Ge, GPe; Gi, GPi. These images depict binding to D1 (total binding, a and c; nonspecific binding, b and d) and D2 (total binding, e and g; nonspecific binding, f and h) receptors in the pallidum (and caudal striatum) of the adult and perinatal cases. Color-bar scales at the right of each image indicate density of labeled receptors, in fmol/mg protein. e and f show a relatively rostral section, which is not representative of the entire pallidum; see Figure 4.

Full size image
Figure 3
figure 3

D1 dopamine receptors in adult and perinatal globus pallidus. D1 receptor density in successive rostrocaudal sections of perinatal and adult GPe and GPi. In the perinatal brains, the sections ranged from the level of the mammillary bodies to the rostral tip of the lateral geniculate body; in the adult brains, the sections ranged from the decussation of the anterior commissure to the rostral tip of the lateral geniculate body(18). The 7-wk-old case values are represented by inverted triangles (▾, the bottom curve) in the perinatal panels. There were no statistically significant rostrocaudal gradients in any of these regions for D1 receptors.

Full size image

There was no significant rostrocaudal gradient for D1 receptors (Fig. 3) in the adult GPe or GPi. In the perinatal GPi, there appeared to be an inverted U-shaped rostrocaudal gradient of D1 receptor densities; there was no clear gradient in the GPe.

D2 receptors. Although the D2 densities did not differ significantly between the adult and perinatal cases (Fig. 1; Table 2), the distribution of D2 receptors across the two pallidal segments was different for the two age groups. D2 receptor densities were similar in both pallidal regions perinatally, whereas D2 receptors were twice as dense in the GPe as in the GPi in the adults. In contrast to the D1 receptors, the D2 receptors did not differ in the 7-wk from the perinatal cases (Fig. 4).

Figure 4
figure 4

D2 dopamine receptors in adult and perinatal globus pallidus. D2 receptor density in successive rostrocaudal sections of perinatal and adult GPe and GPi. Levels and symbols are comparable to those in Figure 3. There was a significant, decreasing rostrocaudal gradient for D2 receptors in the perinatal GPi (by linear regression, p = 0.03). There was a significant, decreasing rostrocaudal gradient in the adult GPe, even when the data for which there were only a few data points per brain (n = 2) were excluded(p = 0.007). There were no significant gradients in the perinatal GPe or adult GPi.

Full size image

There was a decreasing rostrocaudal gradient of D2 receptors in the perinatal GPi, as well as in the adult GPe (Fig. 4). (This decreasing gradient in the perinatal GPi accounts for the apparent discrepancy between data shown in Figure 1g, a relatively rostral section where D2 receptors are more plentiful in the GPe than GPi, and data in Table 2, where there is no significant difference in the D2 densities in perinatal GPe and GPi.) There was a very consistent density of D2 receptors at all levels in the perinatal GPe. No rostrocaudal gradient was detected in the adult GPi despite D2 receptor variability within this region.

Table 2 D2 dopamine receptor binding characteristics
Full size table

Striatum

D1 receptors. The mean density of D1 receptors was similar in the adult caudate and putamen (Fig. 2; Table 3). Although the D1 receptor density was 37% higher in the putamen than in the caudate perinatally, this difference was not quite significant (p = 0.056). Comparable D1 receptor values were observed in perinatal and adult caudate and putamen.

Figure 2
figure 2

D1 and D2 dopamine receptors in adult and perinatal striatum. Black and white or pseudocolor-enhanced images of representative coronal sections through adult (a-c, g-i) and perinatal (d-f, j-l) neostriata. Cd, caudate;Pt, putamen. The black and white images depict striatal striosomal labeling with [3H]diprenorphine (to label opiate receptors). The pseudocolor images depict binding to D1 (total binding, b and e; nonspecific binding, c and f) and D2(total binding, h and k; nonspecific binding, i and l) receptors in the adult and perinatal human striatum. Color bar scales at the right of each pseudocolor-enhanced image indicate density of labeled receptors, in femtomoles/mg of protein.

Full size image

Despite the overall similarity, D1 receptor density was modestly higher in the striosomal compartment, compared with the matrix, in both the perinatal and adult caudate. A similar trend was not significant in the putamen at either age. Neither quadrant (data not shown) nor rostrocaudal(data not shown) analysis revealed significant gradients at either age.

D2 receptors. In the adult, the D2 receptor density was 39% higher in the putamen than in the caudate (Fig. 2; Table 4). The density of D2 receptors in the perinatal caudate was not significantly different from that of D2 receptors in the perinatal putamen if the data from all four cases were combined. However, the D2 receptor density in the striatum of the 7-wk case was already within the adult range (caudate = 4266 fmol/mg of protein, putamen = 7028 fmol/mg of protein); these were 2-3 times the mean striatal D2 density of the remaining three younger cases (0, 2, and 4 d old; caudate = 2030 ± 63 fmol/mg of protein, putamen = 2355 ± 54 fmol/mg of protein). When the 7-wk data were excluded from the analysis on the basis of this disparity, the density of D2 receptors in the perinatal caudate was found to be significantly lower (p = 0.05) than that of the perinatal putamen, similar to the pattern in both the 7-wk infant and adult.

In both the adult and perinatal caudate, the density of D2 receptors was significantly greater in the striosomes than in the matrix (by approximately the same amount as for the D1 receptors). D2 receptor levels were comparable in the two compartments of both the perinatal and adult putamen. As was true for the D1 receptors, no significant gradients were found in quadrant (data not shown) or rostrocaudal (data not shown) analysis.

DISCUSSION

Technical Considerations

Interpretation of the data must be undertaken in light of some potentially limiting factors. The small sample sizes and large variations among the individual cases in some regions may have resulted in type II errors(i.e. missing a real difference between the two populations) during statistical analysis. A wide scatter of D1 and D2 receptor levels across a broad range of ages has been previously reported (using HB)(7). However, small differences between populations may attain statistical significance without having biologic significance.

The degree of quenching of tritiated ligands is proportional to the amount of myelin in a brain region, and thus may be a complicating factor in comparing the perinatal and adult brains. Myelination is at an early stage in the perinatal basal ganglia [degree 1 (of 4) of Brody et al.(20) and our unpublished personal observation], compared with the adult. In rats, the amount of tritium emission quenching due to myelin is twice as high in the adult as in the 5-d-old pup in all basal ganglia regions(21). If these quench coefficients were valid for our human data, for D1 receptors in the GPi, e.g. the ratio of perinatal:adult values would drop from 15 to 8-a still remarkable difference.

The density of dopamine receptors observed in the present study may have been affected by a postmortem delay in freezing (<24 h in all cases). A delay of up to 24 h has been reported to result in an approximately 10% (QA)(22) to 30% (HB)(23) decrease in D1 receptors and a 10% (HB) decrease in D2 receptors(23) in rat striatum. However, Seeman et al.(7) (with HB) reported personal observations of no loss of D1 or D2 dopamine receptors in human striatum over this time period. Therefore, the levels of dopamine receptors reported in the present study may be as much as 10-30% lower than the true values.

Dopamine receptor levels in the perinatal cases may also have been affected by tissue hypoxia, as the cause of death in each case was congenital heart disease. Rat models indicate that D1 receptor densities decrease 24%(QA)(24) to 64% (QA)(25) at 24 h after severe hypoxic-ischemic insult to the basal ganglia. D2 receptor levels have been reported as either unchanged (QA)(24) or decreased by 22% (QA)(25) at 24 h. Note that adult donors, too, may have been exposed to hypoxia-ischemia in extremis. These data suggest that all dopamine receptor densities reported in the present study could underestimate the true values.

The D1 and D2 receptor levels in our perinatal cases may also be of concern because some of these babies may have been treated with pressor agents premortem, given that the cause of death for all was congenital heart disease. The drug of concern would almost certainly have been dopamine, as this was the agent of choice in the donor hospital at the time of specimen collection (Elizabeth Thilo, M.D., personal communication). In adults, dopamine does not cross the blood-brain barrier(26), but this barrier develops gradually in children until about 6 mo of age(27). Dopamine (and dobutamine) is rapidly inactivated by catechol O-methyltransferase, which is widely distributed in the body and brain, including in intercellular spaces(26). Dopamine is also rapidly taken up by the presynaptic neuron of a dopaminergic synapse, so it is unlikely that much exogenous dopamine could have reached the dopamine receptors of interest in the basal ganglia.

Another concern that might be raised with the present study is the age at death of our adult specimens, ranging from 67 to 87 y. Seeman et al.(7) found a wide range of values in 247 control striata(up to age 93) in an HB study, with an average decline after age 20 of 3.2% per decade for D1 receptors and 2.2% per decade for D2 receptors. If age 30 or 40 were preferred as the adult comparison age, these modest rates of decline would not change our conclusions for the striatum, although the p values would be lower in some cases. No such data are available for the globus pallidus.

Finally, this discussion will reveal that our results do not always agree with those of other investigators. Several factors may be considered to account for these differences. Tissue sampling could well have been different in each of the cited studies. For HB, a core of tissue from a region is often taken to be representative, thus excluding border zones of the structure. In QA, although a variably limited number of sections is analyzed, very precise regions may be outlined for quantitation. Tissue from brain banks may be circumscribed, and it is not always blocked in the same manner from bank to bank. In the present study, we attempted to obtain even sampling of 16 sections (six in the perinatal cases) throughout the entire rostrocaudal, dorsoventral, and mediolateral extents of the basal ganglia, although we, too, were more limited in those tissues obtained from a brain bank (two adults), especially for the globus pallidus. Differences in assay constituents and conditions may also contribute to variable results.

Globus Pallidus

D1 receptors. The most surprising and unique finding of the present study was the great excess of D1 receptors in both the GPe and GPi in the perinatal cases, compared with the adults. The perinatal profusion of pallidal D1 receptors is almost entirely due to the elevated D1 levels in the three younger cases, as the D1 receptor density in the 7-wk postnatal case was within the adult range in each pallidal segment. In the present study, D1 receptor levels were greater in the GPi than in the GPe in both perinatal and adult cases, in concordance with observations by Cortés et al.(12) (using QA) in the adult human pallidum. Interestingly, Barks et al.(28) reported enhanced binding of [3H]glutamate in the basal ganglia of two fetuses(18 and 21 wk of gestation) relative to that observed in the adult, suggesting that transient elevations in receptor levels could be a common occurrence in brain development.

The most likely location of pallidal D1 receptors is on striatopallidal terminals, which use GABA as a neurotransmitter(9). D1 receptor mRNA has not been detected in the pallidal-equivalent subdivisions of the rat(29), indicating that the D1 receptors are not on intrinsic pallidal neurons. A loss of D1 receptor binding (QA) has been observed in the globus pallidus (homolog of human GPe) and entopeduncular nucleus (homolog of human GPi) of the adult rat after striatal neuron lesioning with quinolinic acid(30). Finally, stimulation of D1 receptors on striatopallidal terminals has been demonstrated to increase GABA release in the pallidum in the 6-hydroxydopamine-lesioned rat(31).

If the elevated D1 receptors in the perinatal human are indeed located on striatopallidal terminals, dopaminergic stimulation of these terminals could lead to a marked increase in pallidal (especially GPi) GABA release. A subsequent reduction in the output of the GPi to the thalamus would then be expected, according to recent models of basal ganglia circuitry(9, 32). Behaviorally, this could contribute to the choreiform movements exhibited by normal human infants.

D2 receptors. In adults, D2 receptor densities in the GPe were found to be two times as great as those in the GPi, a pattern similar to that reported in previous studies in adult humans (QA)(13, 33). In contrast, the density of D2 receptors was similar in the GPe and GPi of the perinatal cases. In the rat, D2 receptor mRNA has been found in scattered cells in the globus pallidus (human GPe), with somewhat higher levels in the entopeduncular nucleus (human GPi)(29), suggesting that at least some of the D2 receptors may be present on cell bodies in the pallidum, in contrast to D1 receptors. The functional significance of the perinatal difference from the adult pattern is unknown.

A new finding was the decreasing rostrocaudal gradient of D2 receptors in both the perinatal GPi and the adult GPe. These gradients could be related to the topography of afferent and efferent connections.

Striatum

D1 receptors. In the present study, comparable densities of D1 receptors were observed in the caudate and putamen of both the adult and perinatal human striatum. These results agree with previous studies reporting similar levels of D1 receptors in adult humans [QA(12) and HB(7)].

D1 receptor densities were modestly greater in the striosomal than in the matrix compartment of the caudate in both adult and perinatal cases, although the biologic significance of an 11-17% difference is questionable. No significant compartmental differences in D1 receptor levels were seen at either age in the putamen. [Joyce et al.(34) found no quenching differences between the two compartments.] Another study of human (dorsal) striatum (QA)(35) also found a striosomalpredominant distribution of D1 receptors. In the rat, there is a transient pattern of early segregation of D1 receptors in the striosomal compartment of the immature striatum (QA)(36, 37). The quadrantic and rostrocaudal analyses of the striatum were undertaken because significant regional gradients for both D1 and D2 receptors have been found in the rat (QA)(10) that appear to relate to functional divisions (e.g. motor to the dorsolateral striatum; QA)(38). Our lack of a finding of regional gradients for D1 receptors may reflect the different orientation and/or greater degree of functional specialization of the caudate and putamen (see below) in the human, compared with the rat.

D2 receptors. A modest regional difference between the caudate and putamen was observed: D2 receptors were about 30% higher in the latter structure in both age groups. Functionally, although not spatially, this agrees with the rat study cited above(38), as the human putamen is more specialized for control of movement than is the caudate. The actual density of D2 receptors in each of the striatal structures in the perinatal cases was half that of the adults.

In humans, reports of D2 receptor densities' being higher in the caudate [QA(13) and HB(7)], higher in the putamen [QA(39) and HB(40)], and comparable in the two structures [QA)(34)] have all appeared in recent literature. None of the intrastriatal differences was greater than 50%, however. Our results in the perinatal human are in general agreement with those in the infant baboon (QA)(41).

D2 receptor densities were modestly greater in the striosomal than in the matrix compartment of the caudate in both the adult and perinatal cases in the present study. As with the D1 receptors, the biologic significance of a 10-13% difference is debatable. Again, significant compartmental differences in the putamen were not found. This is in apparent contrast to studies in the adult human where the density of D2 receptors was approximately twice as high in the “D2-rich zones” as in the “D2-poor zones” (which were said to have 96% agreement with the matrix and striosomes, respectively) in both caudate and putamen (QA)(33, 34).

With regard to the compartmental distribution of D2 receptors in immature animals, D2 receptor levels were greater in striosomes than in the matrix in both the caudate and putamen in the baboon between birth and 6 mo of age (QA)(41), in general agreement with our data.

Developmental Implications

There is broad evidence for a role of dopamine in both vertebrate and invertebrate embryogenesis [reviewed in Lauder(2)]. Our observation of very high levels of D1 receptors in the perinatal globus pallidus and, to a lesser extent, in the striatum raises a question as to their possible role in brain maturation. Stimulation of D1 receptors by D1 agonists, or by the indirectly acting dopamine agonist cocaine, results in an induction of several IEGs in both adult and immature rats(3, 42, 43). Because the protein products of IEGs act to modulate transcription of other genes(3), IEGs may participate in the control of normal developmental programs. If stimulation of D1 receptors is important for CNS development (which also assumes that the receptors are functionally coupled and active), then some of the abnormalities observed in cocaine-exposed babies may arise from an amplification of the normal D1-mediated regulation of gene expression. Fetal or perinatal exposure to other drugs that increase synaptic dopamine(e.g. amphetamines and some antidepressants) or block dopamine receptors (neuroleptics) may also have long lasting effects on the basal ganglia, especially.

Future work should more completely determine the specific regional maturational time course of dopamine receptors so as to define the period of possible vulnerability to effects of cocaine or other dopaminergic agonists or antagonists on the developing brain. Even at the earliest age we studied, a 32-wk gestational age stillborn, there were significant levels of both D1 and D2 receptors in the basal ganglia. One HB study found evidence of D2 receptors in the striatum as early as 16 wk of gestation(6). The adult-like pattern of D1 receptors seen in the 7-wk infant (Fig. 3) raises the question of whether receptor levels may change rapidly at the time of birth. Does birth, itself, induce such major changes? It is possible that there are age-specific transient peaks [similar to those for D1 and glutamate(28) receptors in the pallidum] or nadirs in other receptor levels, so further fetal and childhood studies would be worthwhile. Future studies should also include the other subtypes of both D1 and D2 receptor families that have been defined by molecular biology, even though these have been thought to be less important in the adult basal ganglia(4, 5). Finally, developmental studies of dopaminergic drug-exposed children should include long-term assessment for parkinsonism, tremor, fidgety dyskinesias, dystonia, tics, obsessive-compulsive disorder, and eye-movement dyscontrol-problems that are typically associated with the basal ganglia.

Table 1 D1 dopamine receptor binding characteristics
Full size table
Table 4 D2 dopamine receptor binding in human striatum
Full size table