Dopamine is an important neurotransmitter implicated in the regulation of mood, motivation and movement. We have reviewed here recent data suggesting that dopamine, in addition to being a neurotransmitter, also plays a role in the regulation of endogenous neurogenesis in the adult mammalian brain. In addition, we approach a highly controversial question: can the adult human brain use neurogenesis to replace the dopaminergic neurones in the substantia nigra that are lost in Parkinson's disease?
The recent discovery that the adult mammalian brain has the potential to generate new neurones and to integrate them into existing circuits has caused a shift in our understanding of how the central nervous system functions in health and disease (Alvarez-Buylla and Lim 2004). It has been consistently demonstrated, in two distinct areas of the forebrain, that mature cells in all neural lineages, including neurones, are generated throughout adulthood. Neuroblasts born in the adult subventricular zone (SVZ), subadjacent to the ependyma lining the lateral ventricles, migrate along the rostral migratory stream to the olfactory bulb, where they become interneurones. Neuroblasts born in the adult subgranular zone (SGZ) of the dentate gyrus migrate into the adjacent granular layer, where they become granular neurones. The constitutive neurogenesis that occurs in the SVZ and SGZ is thought to be of functional importance in olfaction, mood regulation and memory processes (Nilsson et al. 1999; Santarelli et al. 2003; Enwere et al. 2004; Kempermann et al. 2004). The factors that govern the generation, migration, differentiation, integration and survival of new neuroblasts in the SVZ and SGZ include diffusible molecules such as neurotransmitters (Hagg 2005; Lledo and Saghatelyan 2005). It is therefore conceivable that an alteration in brain neurotransmitter levels, as occurs in neurodegenerative diseases, would affect adult neurogenesis in the SVZ and SGZ with yet unknown functional consequences. Inversely, the discovery of adult neurogenesis has raised the hope that the SVZ and SGZ, or other regions of the adult brain, might still have the capacity to generate neuroblasts that can replace the neurones lost through disease.
Parkinson's disease (PD) has received much attention in recent years with regard to adult neurogenesis, as the degenerative process is relatively selective for the dopaminergic nigrostriatal projection. We have evaluated the published evidence indicating that (i) dopamine plays a role in the regulation of constitutive neurogenesis in the adult brain, and that (ii) adult neurogenesis can repair the damaged nigrostriatal dopaminergic system.
Dopaminergic control of adult neurogenesis
The neurotransmitter dopamine contributes to the ontogenesis of the mammalian brain by regulating neural precursor cell proliferation. Dopamine (Voorn et al. 1988; Ohtani et al. 2003) and its receptors (Lidow and Rakic 1995; Diaz et al. 1997) appear early during embryonic development in the highly proliferative germinal zones of the brain. Dopamine receptors are classified as either D1-like (D1 and D5) or D2-like (D2, D3 and D4), according to structural homologies and shared second messenger cascades. Dopamine receptors, particularly of the D3 type, are abundantly expressed during brain development in the germinative neuroepithelial zones actively involved in neurogenesis in most basal forebrain structures, thereby supporting the hypothesis that dopamine plays a role in neurogenesis during brain development (Diaz et al. 1997). Indeed, dopamine has been shown to either activate or inhibit, through D1- and D2-like receptors, respectively, the proliferation of precursor cells in the lateral ganglionic eminence, which is the neuroepithelial precursor of the neostriatum, and in the cortical neuroepithelium, the germinal precursor of the dorsomedial prefrontal cortex in the developing brain (Ohtani et al. 2003; Popolo et al. 2004). Although the effect of dopamine on progenitor cell proliferation is well documented, it is not known whether it acts on neural stem cells. In the adult brain, high levels of D3 receptor expression have been shown to persist in the germinal SVZ (Diaz et al. 1997). As a developmental continuum appears to link embryonic stem cells to adult neural precursor cells in the SVZ and SGZ (Alvarez-Buylla et al. 2001), dopamine may also be implicated in adult neurogenesis in the SVZ.
The anatomical evidence
The adult SVZ has a unique cytoanatomical and functional organization (Doetsch et al. 1997). A specialized type of glial fibrillary acidic protein+ (GFAP+) astrocytes, so-called B-cells, act as stem cells, which means that they have the potential to self-renew and to give rise to astrocytes, oligodendrocytes and neurones (Doetsch et al. 1999). B-cells generate frequently dividing transit-amplifying C-cells, which can be identified by their expression of the receptor for epidermal growth factor (EGFR) (Doetsch et al. 2002; Höglinger et al. 2004). Asymmetric division of C-cells gives rise to polysialic neural cell adhesion molecule+ (PSA-NCAM+)-restricted neural precursors (A-cells), destined to migrate via the rostral migratory stream to the olfactory bulb, where they integrate as interneurones (Luskin 1993).
Immunohistochemical and electron microscopy studies have show that D2-like dopamine receptors are expressed predominately on C-cells, whereas A-cells express both D1- and D2-like receptors (Höglinger et al. 2004). As highly specific antibodies for the individual receptor types are not available at the present time, a more detailed analysis of the receptor distribution has not yet been possible. The distribution described in the adult SVZ (Höglinger et al. 2004), however, is consistent with the predominant expression of D3 receptors (belonging to the D2-like group) on proliferating precursor cells and D1 receptors (belonging to the D1-like group) on migrating neuroblasts, which has been shown by in situ hybridization in the embryonic brain (Diaz et al. 1997). Furthermore, immunhistochemical studies and both confocal and electron microscopy (Höglinger et al. 2004) have demonstrated that C-cells in the adult SVZ are embedded in a rich network of dopaminergic afferents that form synapse-like structures (Figs 1a and b). Finally, anterograde tracing studies in non-human primates have demonstrated that the dopaminergic fibers in the SVZ originate, at least in part, in the pars compacta of the substantia nigra (Freundlieb et al. 2006). These anatomical observations support the existence of a nigro-subventricular dopaminergic projection terminating on transit-amplifying C-cells, thus raising the question of its effect on function.
The cell biological evidence
In order to assess the functional effect of dopamine on precursor cells in the SVZ, a series of experiments was performed in neurosphere cultures, a culture system enriched in proliferating neuronal stem cells and early precursor cells (Reynolds and Weiss 1992; Morshead et al. 1994). The expression of EGFR in virtually all cells in neurosphere cultures strongly suggests that most cells in these cultures are C-cells (Fig. 1c). Expression of D1- and D2-like dopamine receptors in neurospheres (Fig. 1c) has been demonstrated by immunhistochemistry and PCR (Coronas et al. 2004; Höglinger et al. 2004; Kippin et al. 2005). Treatment of neurospheres with the D2-like agonists bromocriptine and apomorphine, at concentrations typically used for in vitro experiments and for time periods ranging from 12 h to 3 days, significantly increased cell proliferation, as quantified by the percentage of cells incorporating the thymidine analogue 5-bromo-2′deoxyuridine (BrdU) into their nuclear DNA (Coronas et al. 2004; Höglinger et al. 2004). An effect of these D2-like agonists on the rate of cell death in neurosphere cultures was excluded by TUNEL-staining (Coronas et al. 2004; Höglinger et al. submitted). The effect of bromocriptine and apomorphine on cell proliferation was blocked by the D2-like antagonist sulpiride (Coronas et al. 2004; Höglinger et al. 2004), suggesting that it was indeed mediated via the D2-like receptors. No effect of the D1-like agonist SKF 38393 on cell proliferation has been observed (Höglinger et al. 2004). Dopamine, acting on both D1- and D2-like receptors, reproduced the effect of the D2-like agonists (Höglinger et al. 2004). These observations are consistent with the proliferation-inducing effect of D3 receptor activation as described in neuroblastoma cells (Pilon et al. 1994), and suggest that the activation of D2-like receptors on transit-amplifying C-cells in neurosphere cultures stimulates their proliferation (Fig. 2).
Kippin et al. (2005) reported that a chronic, but not an acute, treatment of adult rats with the D2-like antagonist haloperidol led to an increase in the number of primary neurospheres obtained from the SVZ, and that passaging of primary neurospheres in the presence of either dopamine or the D2-like agonist quinpirole led to the formation of fewer subsequent neurospheres. From these observations they concluded that dopamine inhibits the proliferation of B-cells. However, as no dopamine receptors have been identified by electron microscopy on B-cells in vivo, and as there is no evidence for an alteration of B-cell proliferation by dopamine depletion in vivo (Höglinger et al. 2004), the idea that B-cells are indeed modulated by dopamine remains controversial.
The evidence from integrated in vivo systems
A substantial volume of work has been carried out to assess the effects of dopamine on precursor cell proliferation in the SVZ in vivo. Experimental ablation of the dopaminergic innervation of the forebrain in mice and rats using the neurotoxins MPTP and 6-hydroxydopamine (6-OHDA) has been repeatedly shown to reduce global cell proliferation in the SVZ by 30–45% (Baker et al. 2004; Höglinger et al. 2004; Winner et al. 2006). More detailed analysis showed that this was a result of selective reduction in the proliferation of transit-amplifying C-cells (Höglinger et al. 2004), consistent with the results of the in vitro experiments described above. The inhibition of dopaminergic transmission in adult rats in vivo using the D2-like antagonist haloperidol has led to inconsistent results (Wakade et al. 2002; Kippin et al. 2005). However, systemic treatment of either normal or dopamine-depleted rats with the D2-like agonists ropinirole or 7-hydroxy-N,N-di-n-propyl-2-aminotetralin (7-OH-DPAT) significantly increased precursor cell proliferation in the SVZ (Höglinger et al. 2004; Van Kampen et al. 2004). Levodopa also had a significant, although less pronounced, stimulatory effect on SVZ cell proliferation in the dopamine-depleted rats (Höglinger et al. 2004). In unlesioned adult control mice, 7-OH-DPAT did not significantly stimulate SVZ cell proliferation (Baker et al. 2005), perhaps because the dopaminergic stimulation of cell proliferation was already maximal, as dopamine depletion effectively reduced cell proliferation in the SVZ (Baker et al. 2004; Höglinger et al. 2004). Together, these data suggest that in the integrated adult rodent SVZ in vivo, the predominant effect of D2-like receptor activation is the stimulation of C-cell proliferation and, consecutively, the production of migrating neuroblasts (A-cells), whereas dopamine depletion has the opposite effect (Fig. 2).
The fact that there is increased gliogenesis in the striatum of dopamine-depleted rats (Mohapel et al. 2005), mice (Kay and Blum 2000; Mao et al. 2001; Chen et al. 2002, 2004) and primates (Tandéet al. 2006) does not invalidate this conclusion. This phenomenon appears to result from a reaction of gliogenic precursor cells intrinsic to the striatum to neurotoxin-induced neurodegeneration, rather than from the recruitment of neural precursor cells in the SVZ. Within 12 h of MPTP administration in mice (Höglinger et al. unpublished), proliferating cell nuclear antigen-immunoreactive nuclei suggestive of cell proliferation were distributed homogeneously throughout the striatum, as were newborn glial cells in MPTP-treated primates (Tandéet al. 2006), but there was no evidence of a cell density gradient that would be observed if the cells were migrating from the SVZ.
Also in the SGZ of mice, dopaminergic fibres were observed in the vicinity of precursor cells (Höglinger et al. 2004), and precursor cell proliferation was reduced after MPTP treatment. It cannot be concluded, however, on the basis of the scarce data available, that dopamine also regulates precursor cell proliferation in the adult SGZ.
The evidence in primates
We performed a series of studies to determine whether the stimulatory effect of dopamine on SVZ precursor cell proliferation is conserved in primates. We demonstrated a close anatomical relationship between afferent dopaminergic fibres and proliferating cells to be preserved also in the SVZ of non-human primates (Freundlieb et al. 2006). We also found EGFR+ C-cells in the SVZ of adult humans (Fig. 1d) in close proximity to dopaminergic afferents (Höglinger et al. 2004). Experimental dopamine depletion in non-human primates using MPTP significantly reduced precursor cell proliferation in the SVZ (Freundlieb et al. 2006). Finally, in postmortem brain tissue from patients with PD, a disease characterized by dopamine depletion, the number of proliferating precursor cells in the SVZ was reduced (Höglinger et al. 2004). Thus, dopamine appears to stimulate precursor cell proliferation in the SVZ in adult primates including humans.
As in parkinsonian rodents, we found reduced numbers of precursor cells in the SGZ of PD patients (Höglinger et al. 2004), but there is still not enough evidence to suggest that dopamine also controls precursor cell proliferation in the primate SGZ.
The functional impact
The predominant neurochemical alteration in PD is the loss of dopamine. PD patients are typically diagnosed when the cardinal motor symptoms akinesia, rigidity and tremor, which result from the loss of dopamine, are detected. These symptoms usually start when more than 70% of the striatal dopamine is aready lost, and are thus rather late clinical manifestations of the disease. Interestingly, a characteristic set of non-motor symptoms typically precedes the onset of the motor symptoms. These non-motor symptoms include olfactory dysfunction (Berendse et al. 2001), impaired spatial memory (Pillon et al. 1997) and depression (Oertel et al. 2001). These symptoms do not respond to pharmacological dopamine replacement therapy, suggesting that they result from a brain dysfunction outside of the nigrostriatal dopaminergic track; however, their precise pathological substrate is not known. Although the function of adult neurogenesis is still a matter of debate, it is striking that experimenatal inhibition of adult neurogenesis in the SVZ and SGZ has been suggested to impair olfaction (Enwere et al. 2003), spatial memory (Nilsson et al. 1999) and depression (Santarelli et al. 2003).
As we have found evidence of reduced neurogenesis in the SVZ and SGZ in the postmortem brains of PD patients (Höglinger et al. 2004), this might underlie some of the non-motor symptoms in PD. There is as yet no direct experimental evidence to support this hypothesis. Should it be confirmed, however, the stimulation of adult neurogenesis in the SVZ, with D3-like agonists for example, might improve some of these non-motor symptoms.
Repair of the nigrostriatal dopaminergic system by adult neurogenesis
PD is a neurodegenerative disease characterized by a progressive loss of the dopaminergic neurones in the substantia nigra, pars compacta in the midbrain. Although pharmacological dopamine replacement strategies provide temporary symptomatic relief, there are at present no therapeutic methods for stopping the progressive neuronal cell loss. Both neuroprotective and cell-replacement strategies are being explored (Lindvall et al. 2004). The hypothesis that neurogenesis from endogenous neural stem cells in the adult brain could repair the nigrostriatal dopaminergic system has received much attention in recent years, but remains highly controversial.
Evidence that nigral dopaminergic neurogenesis occurs in adults
Cells expressing PSA-NCAM, a sensitive marker of immature, migrating neuroblasts (A-cells) in the SVZ and SGZ, have been detected in the postmortem substantia nigra of human PD patients (Yoshimi et al. 2005). However, PSA-NCAM is expressed not only in neuroblasts, but also in reactive astrocytes (Oumesmar et al. 1995) and in neurites of mature neurones undergoing plastic changes (Charles et al. 2002). Furthermore, the PSA-NCAM+ cells in the human substantia nigra were identified in the pars reticulata, but not in the pars compacta, where the degenerating dopaminergic neurones reside. Thus, although the observations of Yoshimi et al. (2005) provide evidence for some kind of plasticity in the diseased brain, they do not support the hypothesis that dopaminergic neurones are generated in the adult human substantia nigra.
Lie et al. (2002) demonstrated that the substantia nigra of adult rats contains progenitor cells with the potential to differentiate into astrocytes, oligodendrocytes and neurones, and Hermann et al. (2006) reported that neural stem cells isolated from the mouse adult tegmentum had the capacity to generate dopaminergic neurones in vitro. It has not been demonstrated, however, that these cells indeed produce dopaminergic neurones in vivo. According to Lie et al. (2002), suppressive cues in the microenvironment of the adult substantia nigra in vivo might prevent endogenous precursor cells from forming new neurones.
However, the progressive appearance of tyrosine hydroxylase+ (TH+) neurones that incorporated BrdU, a marker of cell division, in the substantia nigra of normal adult mice (Zhao et al. 2003) was thought to show that the entire population of dopaminergic neurones in the substantia nigra could be replaced by adult neurogenesis once in the lifetime of a mouse. These authors also reported that the rate of generation of nigral BrdU+ TH+ neurones in adult mice was significantly enhanced by the MPTP-induced degeneration of the nigrostriatal dopaminergic projection (Zhao et al. 2003), arguing that nigral neurogenesis is regulated to compensate for cell loss in the damaged brain. However, the existence of TH+ BrdU+ neurones was not convincingly documented by confocal microscopy in all three orthogonal optical planes; the two markers might therefore be located in adjacent satellite cells. Similar data presented more recently by Shan et al. (2006), can be criticized on the same grounds. Furthermore, in this study, BrdU immunoreactivity in the nucles was weaker than in cytoplasmic storage vesicles, and might therefore reflect DNA repair rather than DNA replication during mitosis. Unfortunately, no control for the BrdU incorporation during mitosis, such as BrdU+ nuclei in proliferating glial cells, was shown.
In an attempt to identify molecular factors that enhance neurogenesis in the adult substantia nigra, Van Kampen and Robertson (2005) chronically infused the D3 dopamine receptor agonist 7-OH-DPAT into the lateral ventricle of adult rats. In vehicle-infused control rats 5–10% of the nigral BrdU+ cells were reported to express the dopaminergic marker TH. In 7-OH-DPAT-treated rats, this percentage increased to more than 20%. The authors thus concluded that chronic D3 receptor stimulation with 7-OH-DPAT triggers a strong induction of cell proliferation in the rat substantia nigra and promotes the expression of a dopaminergic phenotype in some of these newly generated cells (Van Kampen and Robertson 2005). To illustrate their observations, the authors demonstrate an overview photomicrograph showing a triple labelling with the dopaminergic marker TH, the neuronal marker neuronal nuclear antigen (NeuN) and BrdU. To demonstrate the congruence of the TH+ cytoplasm with the BrdU+ nucleus, they showed serial sections through two cells taken on a confocal microscope. Unfortunately, however, the BrdU-channel in the confocal detail series shows the picture that is referred to as NeuN-channel in the overview picture. Thus, a convincing evaluation of their observations is unfortunately not possible to the reader with the photomicrographs provided in the publication (Van Kampen and Robertson 2005).
In a more recent publication, Van Kampen and Eckman (2006) used a rat model of PD based on the intrastriatal administration of the neurotoxin 6-OHDA, a paradigm that is known to induce a subchronic retrograde degeneration of the nigral dopaminergic neurones, where the degeneration process can last several weeks (Sauer and Oertel 1994). They observed cells that were triple-stained for TH, NeuN and BrdU in the lesioned substantia nigra, the number of which increased significantly in 7-OH-DPAT-treated animals. Approximately 28% of all BrdU+ cells in the substantia nigra ipsilateral to the 6-OHDA lesion in 7-OH-DPAT-treated rats were immunoreactive for both NeuN and TH, indicating that they were dopaminergic neurones (Van Kampen and Eckman 2006). After a period of 8 weeks of 7-OH-DPAT treatment, the percentage of TH+ neurones in the lesioned substantia nigra increased by 55%, suggesting that chronic D3 receptor stimulation can repair the adult substantia nigra, in a rat model of PD, by triggering the genesis of new dopaminergic neurones (Van Kampen and Eckman 2006). This time the dimensional reconstruction of the confocal images of triple-stained TH+ NeuN+ BrdU+ was convincing. However, both TH immunoreactivity, which is normally cytoplasmic, and BrdU staining, which is normally nuclear, were present in the same cellular compartment. This is suggestive of a breakdown of the nuclear membrane, which is a classical early feature of apoptotic cell death (Vila and Przedborski 2003). In this case, nuclear BrdU-immunoreactivity might have been the consequence of an aberrant re-entry into the cell cycle that has been demonstrated to occur during apoptotic cell death in the subchronic 6-OHDA rat model of PD (El-Khodor et al. 2003). It is therefore possible that the recovery of behavioural deficits in the partially lesioned 6-OHDA rats after chronic 7-OH-DPAT treatment, reported by Van Kampen and Eckman (2006), was a result of the induction of axonal sprouting or other plastic events rather than of neurogenesis.
Evidence that nigral dopaminergic neurogenesis does not occur in the adult
The effect of growth factors on dopaminergic neurogenesis in the adult substantia nigra has also been investigated. No evidence for the generation of BrdU+ TH+ neurones was found after chronic infusion of glia-derived neurotrophic factor (GDNF) into the striatum of adult rats, although the number of nigral TH+ neurones increased by 32%, which was most likely the result of the induction of a dopaminergic phenotype in pre-existing neurones (Chen et al. 2005). Similarly, no TH+ BrdU+ double-labelled neurones were observed in either the lesioned or the unleasioned substantia nigra of 6-OHDA-treated rats after 4 weeks of the intrastriatal infusion of transforming-growth factor α (TGFα) (Cooper and Isacson 2004). Intraventricular infusion of either brain-derived neurotrophic factor (BDNF) or platelet-derived neurotrophic factor (PDGF) for 10 days was also ineffective in generating BrdU+ TH+ neurones in the substantia nigra of adult rats either with or without 6-OHDA lesions (Frielingsdorf et al. 2004; Mohapel et al. 2005), as was the intrastriatal infusion of liver growth factor (LGF) (Reimers et al. 2006).
Conclusion and perspectives
In the last two years a large body of converging data have consistently shown that the neurotransmitter dopamine stimulates endogenous adult neurogenesis in the SVZ by activating D2-like receptors on transit-amplifying progenitor cells. It remains to be studied in detail, however, whether the decreased neurogenesis in diseases such as PD, which are characterized by dopamine depletion, has clinically relevant functional consequences. Nevertheless, as precursor cells in the adult mammalian brain are pharmacologically accessible to the systemic administration of dopamimetic drugs, the stimulation of endogenous neurogenesis appears to be a potential strategy for a cell replacement therapy of the brain in diseases in which SVZ-derived cells appear to contribute to repair processes (Arvidsson et al. 2002; Picard-Riera et al. 2002; Curtis et al. 2003; Jin et al. 2004).
In contrast, the existence of neurogenesis in vivo in the adult mammalian substantia nigra remains controversial, although progenitor cells that can differentiate into neurones are present (Lie et al. 2002). The species of rodent studied, the manner in which BrdU was administered, or the delay between mitotic labelling and analysis might explain some discrepancies. Other confounding factors are summarized in Table 1. Because of the chronic progression of neuronal cell loss in PD, the therapeutic interest of stimulating adult neurogenesis in the substantia nigra is great, but the genesis in vivo of dopaminergic neurones in the substantia nigra from precursor cells in the adult mammalian brain has yet to be unequivocally demonstrated, dampening for the present our enthusiasm for this therapeutic approach.
This work was supported by the German Federal Ministry of Education and Research Network ‘Stem Cells in PD’ (grant 01 GN0513) and the Peter Hofmann Research Project. We thank Merle Ruberg for critically revising this work.