Endogenous Regeneration in Parkinson's Disease: Do We Need Orthotopic Dopaminergic Neurogenesis?

Neurogenesis in the adult mammalian brain is a phenomenon that has long been recognized. Most reports demonstrate the presence of neural stem or precursor cells in the two major neurogenic regions, the dentate gyrus of the hippocampus and the subventricular zone (SVZ) of the lateral ventricles [1, 2], but several recent studies report proliferating precursors in non-neurogenic regions as well, such as spinal cord and cortex [3] or the adult substantia nigra [4]. Adult SVZ precursor cells divide and migrate several millimeters along the rostral migratory stream toward the bulbus olfactorius, where they differentiate mainly into GABAergic granule cells, but 1%–25% of the cells migrate further into the glomerular layer, with a significant number of cells differentiating into dopamine (DA) cells [1, 2]. This DAergic differentiation is exclusively happening within the bulbus olfactorius [1, 2], providing strong evidence for the major importance of the local microenvironment for this terminal differentiation step. To top it off, local cues outside the bulbus olfactorius are not sufficient to fully induce maturation and survival of newborn neurons or—even worse—could have restrictive properties that actively inhibit neurogenesis and in particular DAergic neurogenesis but promote gliogenesis [5–8] (reviewed in [2]). On the other hand, endogenous SVZ precursors seem to be amenable to external modulation, such as intraventricular treatment with epidermal growth factor (EGF), leading to increased proliferation and migration of SVZ precursors [7–10]. Several recent studies have demonstrated a positive influence of brain pathology, such as ischemia/hypoxia, on striatal neurogenesis, which could be even enhanced by striatal application of distinct growth factors [7–9]. This capacity makes adult endogenous neural precursors a potential therapeutic target to regenerate diseased brain cells or functions in numerous acute or chronic neurological diseases, such as Parkinson's disease (PD). In this neurodegenerative condition, the disease process is mostly restricted to the nigrostriatal DA system, and restoring the DA deficit within the striatum by ectopic midbrain DA cells has been shown to be an efficient treatment [11, 12]. Consequently, several recent studies on endogenous DAergic regeneration in PD models aimed at using SVZ neural precursors and driving them to migrate into the striatum and differentiate into DA neurons. However, the DA interneurons generated from SVZ precursors constitute the A16 population of DA neurons, which expresses the transcription factor Pax6 but not the major midbrain-specific transcription factors (Engrailed, Nurr1, Pitx3) [13], uses both DA and GABA as neurotransmitters, and shares only a few functional characteristics with A9 (substantia nigra) DA neurons, which are affected in PD. Indeed, there is as yet no conclusive evidence for generation of midbrain DA neurons within forebrain structures or from forebrain precursor/stem cells either in vitro or in vivo. Isacson et al. [5] extended their previous investigations on effects of intrastriatal infusion of the EGF receptor agonist TGFα by subsequent noggin infusion on adult neurogenesis and precursor phenotype in a PD animal model rising the major question whether in vivo regeneration could be stimulated to restore DA in PD [6]. The authors found that TGFα infusion induced a large population of proliferative cells (resembling C cells) in the adult SVZ that then migrated not only in the bulbus olfactorius but also in the adjacent diseased striatum (a schematic summary of the results is given in Fig. 1) [6]. Differentiation of these migrated cells could be influenced by subsequent noggin infusion increasing the neuroblast-to-astrocyte ratio within the striatum, whereas neurogenic specification was unaltered by noggin, with only a very few Tuj1+ neurons within the striatum. Interestingly, the ectopic striatal neuroblasts did not express the transcription factor Pax6, which is believed a major cell-intrinsic factor for the differentiation of SVZ neuroblasts into the DAergic phenotype of bulbus olfactorius neurons, suggesting that ectopic striatal neuroblasts would differentiate into bulbus olfactorius GABAergic interneurons, rather than DA neurons, without further stimulation. Indeed, the authors did not detect generation of DA neurons within the striatum, and, consistently, there was no behavioral improvement of PD rats [6]. The results [6] nicely demonstrates the current dilemma of regenerative approaches by recruiting endogenous stem cells in PD: enhanced proliferation of endogenous neural precursor cell populations in the SVZ with migration into the DAergic target area of the striatum is possible in PD models, but relevant neurogenesis or even DAergic neurogenesis restoring the DA deficit within the striatum has not yet been conclusively demonstrated. Schematic representation of dopaminergic neurogenesis within the main areas of the central dopaminergic system in PD. (A): shows the bulbus olfactorius; (B) shows the striatal region, with the SVZ of the lateral ventricles, as one neurogenic region of the adult brain; and (C) shows the midbrain region, with the substantia nigra pars compacta. Lateral ventricles are marked in blue, and the third ventricle, aqueduct, and fourth ventricle, without relevant stem/precursor cells populations, are marked in beige. In the PD state, the major pathology includes a degeneration of DA neurons in the substantia nigra pars compacta (C), with a subsequent decrease of SVZ stem or precursor cells (B). In contrast, numbers of DA cells of the forebrain phenotype are increased in parkinsonian models and PD patients in both the striatum (B) and the glomerular cell layer of the bulbus olfactorius (A). Stimulation of precursor cell proliferation with transforming growth factor-α or other stimulants results in an increase of SVZ stem/progenitor and neuroblasts within the SVZ and the striatum (B) and bulbus olfactorius (A) but does not influence the numbers of DA neurons within these areas (details are given in text). Abbreviations: DA, dopamine; PD, Parkinson's disease; SVZ, subventricular zone. At first glance, the study by Fallon et al. [16] seems to be contradictory to the above-mentioned studies, showing increased amounts of DA neurons within the striatum after intrastriatal TGFα infusion in 6-hydroxydopamine-lesioned PD rats in combination with improved behavior [7]. Worth mentioning, however, is the fact that only very few DA neurons were detected, and no correlation between behavioral improvement and the amount of newly generated DA neurons was shown. Furthermore, behavioral improvement was rather low and variable, including 10% of animals even worsening after TGFα infusion. Finally, there was no evidence for relevant numbers of newborn DA cells (colabeled with the proliferation marker 5-bromo-2′-deoxyuridine [BrdU]), most likely pointing toward a phenotypic shift of resident striatal neurons into DA nerve cells [14, 15]. This type of cellular plasticity has already been reported in brains of PD patients and of PD animal models (Fig. 1): in monkeys, nigrostriatal degeneration using 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) intoxication causes a threefold increase in tyrosine hydoxylase (TH+) DA cells in the striatum, which was similarly shown for rodents (reviewed in [2]). Furthermore, the number of TH+ cells was also reported to be markedly increased in the striatum of PD patients [15]. It is believed that the increase in TH+ cells is a compensatory response to restore DA levels due to the absence of DAergic inputs. Careful analyses of such TH+ neurons within the striatum, however, revealed no coexistence of BrdU+ and TH+ cells in mice or macaques, suggesting a phenotypic shift of striatal neurons rather than newly generated DA nerve cells [14]. This observation and the phenotypic differences between forebrain and hindbrain DA neurons must be considered carefully in future investigations of DAergic neurogenesis in PD models. Moreover, it is still completely unclear whether ectopic DAergic neurogenesis within the striatum could actually be sufficient to provide relevant long-term behavioral effects without (motor) complications, as observed in striatal DAergic transplantations [11, 12], or whether we need orthotopic DAergic neurogenesis in the right place within the substantia nigra to allow the cells to functionally integrate into neuronal circuitries. Thus, is there any evidence for dopaminergic neurogenesis in the adult midbrain? Recent studies have revealed contradictory results: Zhao et al. mentioned a continuous turnover of nigral DA nerve cells, with a partial restoration of the nigrostriatal pathway, both under physiological conditions and in parkinsonian models [4]. Furthermore, using Nestin-LacZ transgenic mice, basal levels of neurogenesis and increased proliferation and DAergic differentiation after MPTP treatment have been suggested [16]. Worth mentioning is that in both studies, numbers of newborn DA neurons were extremely low. Newer studies could not reproduce these data and raised some technical problems within these older studies [14, 17–19]. It is now widely accepted that there is no evidence for newborn DA neurons within the substantia nigra, either under physiological conditions or in PD animal models [14, 17–19]. This restricted regenerative potential in vivo might not, however, reflect the whole potential of the remaining precursor cells of the adult midbrain. Lie et al. showed the isolation of neuroprecursors out of the midbrain region, which could be proliferated and differentiated in vitro into neurons [18]. After transplantation of these cells into the neurogenic region of the dentate gyrus, they became neurons, whereas when they were transplanted back to the substantia nigra, they acquired only a glial phenotype [18]. However, the authors did not achieve DAergic differentiation of these adult midbrain precursor cells [18]. Recently, we could show for the first time the induction of DA neurons from adult precursors in vitro using a multistep differentiation protocol, providing the proof of principle for adult DAergic neurogenesis in vitro [20]. Interestingly, we succeeded only when using adult midbrain-derived precursors/stem cells, and we failed when using adult hippocampal stem cells [20]. In addition, these DA neurons express a phenotype similar to that observed in primary DA neurons from midbrain (A9/A10 region). Subsequently, Shim et al. demonstrated the in vitro generation of mature DA neurons from cells isolated from SVZ and white matter of adult rodents by genetically overexpressing the key fate-determining transcription factor for the midbrain DA phenotype Nurr1. After these engineered SVZ precursors were transplanted into the striatum of parkinsonian rats, the cells survived, integrated, and differentiated into DA neurons in vivo and reversed the behavioral deficits of the PD rats [21]. Recently, Papanikolaou et al. showed in vitro differentiation of SVZ-derived adult stem cells into neurons showing a phenotype of forebrain-like DA neurons lacking the expression of the midbrain-specific transcription factors Engrailed, Nurr1, and Pitx3 [13]. Together, these data suggest that only midbrain stem cells harbor cell-intrinsic factors, such as Nurr1 or Pitx3, which enables them to differentiate into DA neurons of the midbrain phenotype. These midbrain stem cells can be isolated from the adult brain [20], most likely deriving from the ventricular system of the midbrain regions, as recently shown by Golmohammadi et al. [22]. On the other hand, data on adult DAergic neurogenesis in vitro [20] strongly point to cell-extrinsic yet unknown factors within the midbrain microenvironment (particularly the substantia nigra) that, similar to those in the striatum, inhibit (DAergic) neurogenesis. However, the potential existence of a neurogenic niche also in the midbrain/hindbrain SVZ region including stem or precursor cells with DAergic differentiation potential should initiate detailed analyses of the SVZ in midbrain/hindbrain regions compared with the SVZ of the lateral ventricles as one major neurogenic region. In particular, the differences between these two regions would be of great interest in analyzing potential endogenous restrictors of adult neurogenesis/stem cell proliferation in the midbrain and investigating in situ stimulation approaches of endogenous regeneration in the midbrain, as Cooper and Isacson [5] and de Chevigny et al. [6] showed for the striatum. In light of the data discussed here, endogenous restoration of the DAergic nigrostriatal projections by inducing orthotopic DAergic neurogenesis from endogenous stem or precursor cells in the midbrain with targeted reinnervation of the striatum is a major challenge. Intranigral transplantation of DA cells in rodents [23, 24], primates [25, 26], and humans [27] demonstrated graft integration, survival, and partial functional recovery, providing impetus for the existence of factors guiding targeted axonal outgrowth from the substantia nigra to the striatum. However, spontaneous nigrostriatal reconnection seems to be age-related and is observed only in early postnatal rodents, not in adult rodents or humans [23, 27]. Furthermore, fiber outgrowth appeared to be partially undirected, also showing axonal outgrowth in many different brain regions [24]. Therefore, ectopic generation of sufficient numbers of DA neurons in the striatum to restore the local DA deficit by either recruitment of endogenous progenitors from the SVZ or stimulation of resident cells in the striatum to shift their phenotype toward functional DAergic properties seems a feasible alternative. A couple of candidate growth factors have already been tried, though; together, they show induction of proliferation of SVZ precursors and partial migration in the striatum, such as platelet-derived growth factor, brain-derived neurotrophic factor, glial cell line-derived neurotrophic factor, or TGFα with or without subsequent noggin application (reviewed in [2]). As discussed, all studies fail at the point of terminal differentiation and survival of newborn DA neurons in the healthy or diseased striatum. Interestingly, the neurotransmitter DA itself has been identified as possible mitogen for SVZ precursors [28]. Furthermore, stimulation of dopamine D2/D3 receptors has been shown not only to increase cell proliferation in the SVZ and striatum of rats but also to significantly increase proliferation in the lining of the third ventricle, as well as the substantia nigra. A subset of BrdU-labeled cells adopted a neuronal phenotype or differentiated into TH+ cells [29]. Moreover, improvement in locomotor function and striatal innervation from these nigral cells was shown. Whether these were direct effects of DA receptor stimulation or rather the result of changing environmental signals (such as increasing, for example, the release of ciliary neurotrophic factor for example [30]) remains to be elucidated. Adult dopaminergic neurogenesis most likely depends on both cell-intrinsic and cell-extrinsic signals. Current data support the hypothesis that adult stem or precursor cells derived from the midbrain/hindbrain harbor the specific cell-intrinsic factors for differentiation into functional DA neurons of the midbrain phenotype. In vitro data also show that local cell-extrinsic signals within both the striatum and the substantia nigra under physiological and pathological conditions restrict the differentiation of adult precursors into DA neurons. To be able to influence or even overcome these local commitments, we unambiguously have to understand the cell-intrinsic differences between stem/precursor cells from the different brain regions to elucidate potential targets for manipulation. Identification of the factors inhibiting neurogenesis of SVZ cells within the striatum and the midbrain or, on the other hand, the factors inducing differentiation of DA neurons within the bulbus olfactorius is urgently warranted. In our opinion, understanding and overcoming these restrictions of forebrain/hindbrain cues is the future direction in the field of endogenous regeneration in PD and would dramatically change the therapeutic spectrum of PD. The authors indicate no potential conflicts of interest.