Development of Mesencephalic Dopamine Neurons in the Nonhuman Primate

John R. Sladek, Jr., Barbara Blanchard, T. J. Collier, John D. Elsworth, Jane R. Taylor, Robert H. Roth, and D. Eugene Redmond, Jr.

Independent studies have characterized the appearance of dopamine (DA) biochemically and histochemically in mesencephalic neurons in the mouse and rat during early ontogenesis (4, 9, 11, 12, 13, 18, 20, 26, 27, 33, 34, 40). Similar studies have not been performed in comparable detail in nonhuman primates, but considerable data do exist for human brain (5, 19, 21, 23, 24, 29). Our understanding of the temporal events in the development of the mesencephalic DA neurons consequently is based primarily on analysis of rodents. While this information is useful, it may prove difficult to apply to experimental designs involving primates because of major differences in the timing of neurogenesis and the expression of transmitters. For example, DA histofluorescence was first noted during the latter half of gestation in the rat at approximately embryonic day 12.5 (i.e., E 12.5), which corresponds to a surge in the number of labeled cells seen during neurogenesis of the ventral mesencephalic neuron population (10). The presence of tyrosine hydroxylase was first detected in these neurons also at E 12.5, which suggests a temporal relationship that is consistent with the production of dopamine soon after the onset of neurogenesis. DA exists in neurons as they migrate through the deeper layers of the mesencephalon after division in the germinal zone. In fact, dopamine has been reported in neurons while still in the germinal zone, which reflects biosynthesis immediately after mitosis. Thus, dopamine is present during the latter half of gestation in the rat, whereas it is seen during the first trimester of primate development as described in more detail below. The early presence of dopamine in the primate brain may reflect a role for this neurotransmitter in neurogenesis. For example, interruption of the biosynthesis of a monoamine transmitter, serotonin, by prenatal administration of parachlorphenylalanine (PCPA) results in the delay of differentiation of target neurons of the serotonin system, particularly in the hippocampus (12). If similar mechanisms exist in primate brain, then the relatively early appearance of DA could be influential in the differentiation of critical targets such as the striatum and frontal cortex (see also Molecular and Cellular Mechanisms of Brain Development, The Development of Brain and Behavior, and Neurodevelopmental Perspectives on Schizophrenia).

The timing of the expression of DA in mesencephalic neurons in monkeys is understood incompletely and cannot be extrapolated from data generated from either the rat or human brain. DA is seen in mesencephalic neurons early in human gestation, but relatively later in the rat. Our examination of this question, although preliminary at present, is described below as a framework for comparison to other mammalian species, including human. A number of studies have been conducted on aborted human material as described below, but the utility of such information with respect to the design of experiments to test developmental hypotheses in primates is far too limited. Thus, knowledge of similar phenomena in nonhuman primates provides an important database for experimental manipulation of the developing DA systems in order to test theories on transmitter expression or differentiation (for example) which could not be examined in developing human brain.

According to Levitt and Rakic (14), neurogenesis of the substantia nigra occurs in the rhesus monkey between E 36 and E 43. The expression of the DA phenotype in nigral neurons in a closely related Old World macaque, the African green monkey, occurs at a similar time as demonstrated by the presence of the synthetic enzyme tyrosine hydroxylase and shown at day E 40 in , , , ). This is consistent with a generalized "mammalian plan" which results in the biosynthesis of monoamine transmitters early in development, shortly after the initial wave of neurogenesis. The newly formed DA neurons are small and generally rounded, with an average perikaryal length of 9.2 mm (range: 7.4–12.2 mm). Many possess short, dorsally projecting neurites that are about the same length as the perikarya of the mesencephalic neurons (). Others have dorsal neurites that can be traced for distances of 4–5 times (29.6–55.0 mm) the length of the DA perikarya in a single histological section through the mesencephalon. These neurites, at this point in ontogeny, are not recognizable as a rostrally projecting pathway into or toward the striatum or basal forebrain. These migratory neurons occur throughout the depth of the embryonic mesencephalon () with a high concentration dorsally at the germinal zone, as well as ventrally within the developing mesencephalic tegmentum. The density of neurons is exceptionally high, which provides a predominant appearance of cellularity in the E 40 mesencephalon.

The picture of dopamine neurons changes substantially by E 55 in the African green monkey. At this time, the DA neurons form a recognizable cluster in the ventral mesencephalon, although the separation into three distinct groups is not complete. Distinct groups are recognizable at E 61 based on size, shape, and staining density. Also, neurite extension is more extensive at this later stage. At E 55, the individual neurons have increased in size to an average diameter of 14.3 mm (range: 9.8–24.6 mm) and the extension of axons rostrally has proceeded to the level of the striatum (), approximately 2 mm rostral to the ventral mesencephalon. The density of tyrosine-hydroxylase-stained fibers is modest as seen in the ascending DA pathway associated with the medial forebrain bundle () and in the vicinity of the internal capsule where it separates the caudate nucleus from the putamen. However, the ascending monoamine pathways collectively form a prominent bundle at this age, and it is difficult to determine what component is attributable to nigrostriatal axons in the absence of tracer studies or the use of more specific antisera.

The DA perikarya at this age stain with a higher intensity than at E 40 (). DA neurons are not present in the germinal zone, and migration through the tegmentum appears complete. Some tyrosine-hydroxylase-stained neurons are present in the mesencephalic periaqueductal gray and within the region of nucleus raphe dorsalis and could be misinterpreted as components of the germinal zone. It is likely that they represent the small number of catecholominergic neurons that are present in the adult macaque in the dorsal raphe and periaqueductal gray of the mesencephalon (8).

By E 90 the mesencephalic DA neurons have attained an adult-like appearance with respect to shape, size, neurite outgrowth, and position (, , ). Perikaryal diameters have enlarged to approximately 17.4 mm (range: 11.7–24.3 mm), and individual neurons resemble those seen in adults in all three mesencephalic cell groups (). Neurite extension has progressed over 3.7 mm to reach the striatum where the initial innervation is seen in the form of early patch matrix, or tyrosine-hydroxylase-fiber-rich "islands" () similar to those first described by Tennyson et al. (37, 38) in rabbit neostriatum and by Olson and Seiger (20) in the rat. A prominent bundle of tyrosine-hydroxylase-positive fibers is present rostral to the mesencephalon and appears as a continuous pathway into the striatum ( and ). Dendrite bundles, another feature of the adult substantia nigra, also exist from the zona compacta to the zona reticulata of the substantia nigra ( and ). The neuropil of the ventral mesencephalon reflects highly developed DA systems with considerable interwoven fibers in the local environment ().

Although preliminary at this time (one or two determinations at each time point), we have performed biochemical determinations of DA content in dissected striatum and ventral mesencephalon of fetal African green monkey brains at stages of development comparable to stages reported above. The concentration of homovanillic acid, the main end metabolite of DA in this species, parallels that of DA. At E 47, DA concentration, although relatively low, is higher in the ventral mesencephalon than in the striatum. Later, at about 70–90 days, DA levels in the striatum rise sharply, consistent with synaptogenesis in the striatum. At this time, striatal DA concentration is 53.5 ng/mg protein; by comparison this value is 5.05 at E 70 and 0.415 at E 47.

Transplantation of neurons into the central nervous system has been studied for over 100 years (39). This technique provides a useful tool for the analysis of neuronal development and can be used to test the specificity of connections between transmitter-defined neural systems. For the past decade, neural grafts have been tested as a therapeutic intervention in neurodegenerative disorders, particularly Parkinson's disease (7), 15, 31; see also Parkinson’s Disease). Considerable work in nonhuman primates (1, 25, 32) and with xenografting of human neurons (2, 3, 22, 36) has demonstrated the feasibility of neural transplants in highly ordered primate brain to produce functional, DA-producing grafts that are capable of reversing motor deficits in animal models of Parkinson's disease. Recent reports from several centers suggest that the implantation of embryonic DA neuroblasts can result in some degree of improvement in advanced parkinson patients (6, 16, 17, 35), particularly if the symptoms were caused by the (unknowing) intake of a DA-specific neurotoxin (41). Approximately 300 parkinson patients worldwide have received neural grafts of embryonic mesencephalic neurons, and the recent lifting of a federal restriction on the use of federal support for this experimental surgery will prompt additional attempts at this experimental clinical approach. Yet a number of critical characteristics of the developing DA neurons are incompletely understood such as the optimal embryonic age for survival of grafted neurons in primates, including human.

Knowledge gained from basic developmental studies can be applied to questions of neuronal survival following transplantation. For example, if the outgrowth of neurites is considered as a potential determinant of viability following the extirpation of DA neurons from embryonic brain, then the observation of neurite extension from the region of the developing substantia nigra to the striatum in the E 55 monkey described above suggests that neurons from this age are too well developed to survive the obligatory axotomy that accompanies the extirpation. In fact, we have observed between one-tenth and one-twentieth as many DA neurons following grafting of mesencephalon from later-stage embryos such as E 55 in monkey in comparison to younger donor grafts (). Moreover, the "window" for the best survival probably is quite narrow. Ten times as many DA neurons were seen following grafting of E 44 tissue in comparison to E 49 in African green monkey (30), and our recent observations suggest that maximal DA neuron survival is achieved from donor tissue during a 48-hr period, at E 41–42. If we consider that gestation in the monkey is approximately 6 months, as compared to 9 months for the human, then the window for human tissue might be 50% longer and perhaps span a 3-day period at a critical point in neurogenesis or neuronal differentiation. The determination of that critical point is an essential next step for neural grafting experiments in humans. This is particularly crucial because a relatively wide range of embryonic ages has been used in the extant human experiments in parkinson patients (6, 16, 17, 35)). The predictive nature of developmental studies in human material is underscored by the observations described below that attempt to characterize transmitter expression, neurite extension, and the growth of terminal plexuses of DA neurons as a template for transplantation. Thus, the presumption is that survival is influenced by the degree to which axons have extended from the perikarya; that is, extension past a specific distance renders neurons incapable of surviving grafting if axons are severed during the dissection. In this case, only those neurons or neuroblasts that appear as rounded cells with no or short processes will survive grafting. This pool would include that percentage of neuroblasts that are "born" immediately prior to grafting. Neurogenesis takes place over at least 1 week in monkeys, and Freeman et al. (5) reported that tyrosine-hydroxylase-positive neuroblasts were present in the germinal zone in humans as late as 10 weeks, which is 3.5 weeks after they were first seen in this region. This suggests that neurogenesis of the human substantia nigra may occur over a considerably longer period than that in monkeys and that a certain percentage of neurons are capable of surviving grafting at any point during this 3.5-week period. This has been verified in our studies in monkeys where survival of DA neurons was observed in late-stage tissue (27, 32), and predictably the number of grafted tyrosine-hydroxylase-positive neurons was exceptionally low (i.e., often fewer than 100/animal).

Factors other than axon elongation may play a critical role in determining cell survival following transplantation. The development of intracellular messengers, DA autoreceptors, transporter mechanisms, availability of trophic factors, and gene regulation of transmitters and associated enzymes all could influence the ability of embryonic neuroblasts to survive neural transplantation. Systematic analysis of these, as well as other, potential variables is needed to provide a clear understanding of this question in human as well as nonhuman primate brain.

A number of investigators have examined events in the development of human monoaminergic neuron systems, particularly with histofluorescence and immunohistochemical techniques as indices of transmitter expression. The first description by Olson et al. (21) noted (a) the relatively early appearance of monoamines in human neuroblasts of the mesencephalon and (b) the gradual growth of fluorescent ascending axons and striatal patches of catecholamine histofluorescence. At 7 weeks, fluorescence attributed to serotonin and catecholamines was seen within small, round neuroblasts that generally possessed few processes (0–2/cell). By 10 weeks of gestation the neurons were situated in the ventral mesencephalon in groups, but were not subdivided into the A8–A10 subgroups. Some fluorescent axons were present in the striatum, although the development of islands was not seen until 12.5 weeks. The more caudally placed putamen received a DA innervation prior to that of the caudate, which is consistent with the directional course of the ascending DA fiber bundle which approaches the striatal complex in a caudal-to-rostral direction. The mesencephalic complex of DA neurons was "heavily subdivided" by 15.5 weeks, and the striatum was more heavily supplied with a dense patchy pattern of terminals of nigrostriatal origin. Thus, over the relatively short period of 10 weeks the mesencephalic DA neurons are born, grow into the host striatum, and develop a terminal pattern reminiscent of adults. By comparison, the mesocortical pathways were reported by these investigators to follow a somewhat later time course with respect to terminal innervation, although the histofluorescence technique was less capable of demonstrating cortical monoamine fibers than were later versions of the method as described below. A parallel study by Nobin and Björklund (19) basically confirmed these findings and noted that the overall stage of development of the monoamine systems in the human fetus was comparable to that of the rat during the second week after birth, which raises questions about the potential role of the relatively early appearing DA in the human brain.

More recent investigations of human gestation by Freeman et al. (5) using immunohistochemistry noted the presence of tyrosine hydroxylase in mesencephalic neuroblasts at 6.5 weeks and neurite extension at 8 weeks. Tyrosine hydroxylase fibers were reported first in the putamen at 9 weeks, and, remarkably, putative DA neurons were seen adjacent to the ventricle, presumably in the germinal zone, at 10 weeks of gestation. A similar study by Silani et al. (28) reported a few tyrosine hydroxylase cells in the germinal zone of the mesencephalon as early as 5.5 weeks in human material. Thus, the systems apparently begin to produce synthetic enzyme in advance of the first reported demonstration of DA fluorescence, and the presence of DA is an early event that begins during and shortly after neurogenesis. The role of DA at these early stages has been the subject of investigation as described earlier and will undoubtedly continue to be examined carefully, for example, with molecular techniques as interest on the development of monoamine neurons, particularly those that utilize DA, continues to focus on neurologic and psychiatric disorders that may in part be due to abnormal development.

This research was made possible by the contributions of the staff of St. Kitt's Biomedical Research Institute. This study was funded by NIH grants PO1 NS 24032 and RSA MH 00643 (D.E.R.) and by the Axion Research Foundation.

published 2000