Parkinson's disease, aging and adult neurogenesis: Wnt/β‐catenin signalling as the key to unlock the mystery of endogenous brain repair

As adult neurogenesis occurring at the SVZ and SGZ levels is a vital ongoing process in the adult brain, its disruption may contribute to various disturbances including reduced neuronal plasticity, deficits in olfaction, cognitive dysfunction, and/or mental health disorders, that may precede and/or accompany PD onset and progression. Therefore, the tight regulation of the sequential steps of adult neurogenesis should be finely orchestrated by a wide panel of transcription factors and epigenetic mechanisms coordinating the progression of neurogenesis (Bond et al., 2015; Hsieh, 2012; Ming & Song, 2005). At the SVZ, located along the ependymal cell layer of the lateral ventricles (Figure 3), four major cell types dynamically interact with each other: slowly dividing SVZ‐glial fibrillary acidic protein (GFAP)‐positive astrocytes (type B cells), rapidly dividing TAPs (type C cells), migrating neuroblasts (type A cells), and ependymal cells (type E cells) (Doetsch, Caillé, Lim, García‐Verdugo, & Alvarez‐Buylla, 1999; Doetsch, García‐Verdugo, & Alvarez‐Buylla, 1997) (Figure 3). Type B cells exhibit a radial morphology and extend a basal process to terminate on blood vessels and an apical process with a primary cilium contacting the cerebrospinal fluid (CSF) in the ventricle (Mirzadeh et al., 2008). B cells give rise to TAPs (Doetsch et al., 1999, 1997), which rapidly divide to become neuroblasts (A cells). Neuroblasts then form a chain and migrate following the rostral migratory stream (RMS) to the olfactory bulb, OB, where they become granular and periglomerular interneurons involved in odor discrimination (Lledo & Valley, 2016) (Figures 1 and 3). In the adult mouse DG, radial glia‐like precursors within the SGZ serve as one type of quiescent NSCs and continuously give rise to both dentate granule neurons and astrocytes (Bonaguidi et al., 2011; Figure 1). Here, different populations of progenitors with different properties have been described, enhancing the complexity of the cellular and molecular mechanisms underlying regulation of adult neurogenesis (Bonaguidi et al., 2011; Encinas et al., 2011; Ming & Song, 2005). Reportedly, the local neurogenic niche both harbours NSCs and regulates their development. Within the SVZ “stem cell niche” (Alvarez‐Buylla et al., 2001; Alvarez‐Buylla & Lim, 2004; Fuentealba et al., 2012; Lim & Alvarez‐Buylla, 2016; Song et al., 2002), DAergic innervation collaborates with a wide variety of factors including growth and neurotrophic factors, morphogens, cytokines, CSF‐derived regulatory molecules and key components of the WβC‐signalling pathway, contributing to SVZ regulation (Adachi et al., 2007; Borta & Holinger, 2007; Kalani et al., 2008; L'Episcopo et al., 2012, 2013; O'Keeffe Barker, & Caldwell, 2009; O'Keeffe, Tyers, et al., 2009b; Pluchino et al., 2008; Silva‐Vargas et al., 2016). Thus, NSCs can sense both central and systemic changes via circulating factors and are targeted by multiple CSF signals (Silva‐Vargas et al., 2016). Among the niche components of the hippocampal SGZ, which include blood vessels, ependymal cells, and mature neurons (Ming & Song, 2005), hippocampal astrocytes have a key role as they instruct the neuronal fate of cultured adult neural progenitors via WβC‐signalling (Lie et al., 2005; Song et al., 2002; Varela‐Nallar & Inestrosa, 2013). Accordingly, lentivirus‐mediated blockade of Wnt signalling reduces the number of immature new neurons in the adult dentate gyrus and impairs hippocampus‐dependent spatial‐ and object‐recognition memory (Jessberger et al., 2009).

Not surprisingly, different physio‐pathological conditions and pharmacological stimuli dynamically modulate both SVZ‐ and SGZ‐NSCs. Among the positive factors, DAergic innervation and DA‐agonists, serotonin, exercise, enriched environment, learning, estrogens, and antidepressant drugs have been variously documented to stimulate NSC neurogenic potential (Baker, Baker, & Hagg, 2004; Hoglinger et al., 2004; Chiu et al., 2015; Ehninger et al., 2011; Kempermann, Kuhn, & Gage, 1997; Kempermann, Kuhn, & Gage, 1998; Kodali et al., 2016; Kohl et al., 2016; van Praag et al., 2005; Salvi et al., 2016; Winner et al., 2009a, 2011 a,b). Conversely, DAergic neurotoxins, aging, inflammation, stress and chronic exposure to certain toxins or drugs decrease the generation of new neurons (Chandel et al., 2016; Das, Gangwal, Damre, Sangamwar, & Sharma, 2014; Hain et al., 2018; Iwata et al., 2010; Klein et al., 2016; L'Episcopo et al., 2011c, 2012, 2013; Marchetti & Pluchino, 2013; van Praag et al., 2005; Seib & Martin‐Villalba, 2015; Sung, 2015; Veena et al., 2011; Villeda et al., 2011). NSCs are extremely vulnerable to a wide range of insults and toxic exposures, such as the neurotoxicants used in experimental models of basal ganglia injury (Cintha et al., 2018; He et al., 2006; He, Uetsuka, & Nakayama, 2008; L'Episcopo et al., 2012; Shibui et al., 2009). Remarkably, exposure to the herbicide paraquat, which is associated with an increased risk of idiopathic PD, results in a pro‐inflammatory senescence‐associated secretory phenotype capable of damaging neuronal, glial and NSC cells, and therefore likely contributing to DAergic neurodegeneration (Chinta et al., 2018). Aging, especially, represents a key driver of neurogenic impairment as a result of a reduced proliferation and differentiation of NSCs both at SVZ and SGZ levels (Ahlenius, Visan, Kokaia, Lindvall, & Kokaia, 2009; Daynac, Morizur, Chicheportiche, Mouthon, & Boussin, 2016; Enwere et al., 2004; Luo, Daniels, Lennington, Notti, & Conover, 2006). NSCs still retain their capacity for proliferation and differentiation into functional neurons, despite lower numbers in the aged brain (Ahlenius et al., 2009). Recent evidence also suggests that mitochondrial dysfunction represents an important cause of the age‐related decline in neurogenesis, as exposure of the SGZ or SVZ to mitogens or enhancement of mitochondrial function restored neurogenesis (Beckervordersandforth et al., 2017). Notably, the mitochondrial transcription factor A (Tfam), a mitochondrial DNA (mtDNA)‐binding protein essential for genome maintenance, has recently gained attention, as its putative dysfunction may play an important role in neurogenesis defects in the aging hippocampus (Beckervordersandforth et al., 2017) and aging‐dependent neurodegeneration (Kang, Chu, & Kaufman, 2018). Tfam has been shown to play a central role in the mtDNA stress‐mediated inflammatory response and recent evidence indicates that decreased mtDNA copy number is associated with several aging‐related pathologies (Kang et al., 2018). Of special mention, besides its crucial role for the development, maintenance and protection of midbrain DAergic neurons, the relevance of the transcription factor Nurr1 for adult hippocampal neurogenesis in PD has been recently highlighthed, and several lines of evidence indicate that pharmacological stimulation of Nurr1 can improve behavioral deficits via an increase in hippocampal neurogenesis (Kim et al., 2015; Kim et al., 2016 and section 6.4.1).

One key compartment that is dramatically affected in PD is represented by DAergic innervation of the SVZ and SGZ. Hence, nigrostriatal DAergic neurons originating in the SN project to the SVZ in mice, primates, and humans (Borta & Hoglinger, 2007; Freundlieb et al., 2006; Höglinger, Arias‐Carrión, Ipach, & Oertel, 2014; Hoglinger et al., 2004). Within the SVZ niche, DAergic terminals create a dense network of fibers innervating the SVZ (Figure 3 and Figure S1). Accordingly, DA receptors are widely expressed in the SVZ region and are actively involved in the modulation of neurogenesis (Van Kampen et al., 2004). In the hippocampal SGZ, DAergic neurites remain in close contact with NSCs and make their synaptic connections with granular cells in the DG, and it is thought that DA provides an environment for proliferation and differentiation of NSCs (Höglinger et al., 2014, 2004). Accordingly, a number of studies have shown that DA depletion in animal models of PD reduced the proliferation of NSCs in the SVZ and SGZ (recently reviewed by Winner & Winkler, 2015) . Additionally, besides DA, a dysfunction of the serotonergic system projecting to the hippocampus has been reported to impact on adult neurogenesis and suggested to contribute to early non‐motor symptoms of PD, such as anxiety and depression (Kohl et al., 2016).

NSCs have been identified in the SVZ of the aged human brain (Apple et al., 2017; Berge et al., 2011; Bergman, Spalding, & Frisén, 2015; Donega et al., 2019; Leonard et al., 2009; Radad et al., 2017; van den Berge et al., 2010; Wang et al., 2012; Winner, Vogt‐Weisenhorn, et al., 2009b). NSCs, which are GFAP (δ isoform) and nerve growth factor receptor (NGFR, a.k.a. CD271) positive, were shown to proliferate and differentiate towards neurons and glial cells in vitro, then supporting the potential ability to stimulate these NSCs to regenerate the injured brain (van den Berge et al., 2011; Donega et al., 2019). In post‐mortem PD brains a deregulation of SVZ neurogenesis was revealed, with decreased NSC proliferation correlated with the progression of PD, while L‐DOPA treatment appeared to increase NSC numbers. Also, levels of epidermal growth factor (EGF) and its receptor (EGFR) are decreased in the PD Str and the prefrontal cortex of PD patients, and EGFR⁺‐NSC numbers are decreased in the SVZ of PD patients (O'Keeffe, Barker, et al., 2009; O'Keeffe, Tyers, et al., 2009b). At the mesencephalic level in idiopathic human PD, multipotent NSCs isolated from the SN appeared to lack key factors required for neuronal differentiation as they must be co‐cultured with embryonic stem cell‐derived neural precursors to obtain neurons (Wang et al., 2012). Within the hippocampal SGZ, neurogenesis is severely impaired, which may be associated with hippocampal atrophy. Specifically, Hoglinger et al. (2004) reported a decreased number of DG cells expressing nestin and βIII‐tubulin in three PD patients and five patients suffering from PD with dementia (PDD) compared with three controls, with PDD showing a more severely decreased number of nestin‐expressing cells in the human DG. Additionally, Winner et al. (2012) showed a reduction in the number of SRY box transcription factor 2 (Sox2)‐expressing cells in the DG, in dementia with Lewy bodies, and this was paralleled by an increase in α‐syn‐positive cells (see Miller & Winkler, 2015).

Of special importance, the Str, previously defined as a non‐neurogenic region, was recently found to generate neuroblasts in response to different types of brain injury (Luzzati et al., 2011; Nato et al., 2015). Striatal neurogenesis has also been observed in adult non‐human primates such as the squirrel monkey (Bedard et al., 2002). Remarkably, Ernst et al., using a birth dating approach based on the incorporation of nuclear‐bomb‐test‐derived ¹⁴C in the DNA of proliferating cells, reported that new neurons are generated in the adult Str in humans (Ernst et al., 2014), thereby indicating that the possibility exists to stimulate such a mechanism to sustain DAergic striatal reinnervation in PD.

Using well recognized neurotoxic rodent PD models such as the MPTP‐ or 6‐hydroxydopamine (6‐OHDA)‐based models of basal ganglia injury, a number of studies have established that selective nigrostriatal DAergic degeneration in rodents and primate PD models markedly impairs SVZ and/or SGZ neurogenic potential (Chiu et al., 2015; Freundlieb et al., 2006; Hoglinger et al., 2004; L'Episcopo et al., 2012, 2013; Lao, Lu, & Chen, 2013; O'Keeffe, Barker, et al., 2009; O'Keeffe, Tyers, et al., 2009b; Salvi et al., 2016; van Kampen & Eckman, 2006; Van Kampen et al., 2004; Winner, Desplats, et al., 2009a; Winner et al., 2011; Winner, Vogt‐Weisenhorn, et al., 2009b; Winner & Winkler, 2015). In PD, hippocampal non‐motor functions such as spatial learning and memory are impaired, and in several studies the impaired neurogenesis following DA depletion correlated with certain cognitive deficits observed in PD (Das et al., 2014; Klein et al., 2016; Lesemann et al., 2012; Sung, 2015). Other studies, however, showed no difference in proliferation or differentiation of newborn cells in the SGZ of the DG after DAergic lesions (Ermine et al., 2018 and refs therein). As different factors including age, sex, inflammation, the brain region examined, and the post‐injury interval considered, significantly affect the NSC response to injury (Figure 3; and Figure S1) (Khan, Wakade, de Sevilla, Brann, 2015; L'Episcopo, 2013, 2012; L'Episcopo, Tirolo, Serapide, et al., 2018b; Tatar, Bessert, Tse, Skoff, 2013, and following sections), it seems plausible that impairment of SVZ and SGZ neurogenesis in PD may well depend on both DAergic and non‐DAergic related mechanisms.

Another critical factor impacting on adult neurogenesis is the presence of gene mutations. Looking at the overexpression of SNCA (either WT or mutant forms), a number of reports suggest a decrease in the number of newly generated neurons in the DG, OB and SN of the adult brain, including also the DAergic neurons in both the OB and SN (see Le Grand et al., 2015 and refs therein). Transgenic mice overexpressing the LRRK2 G2019S mutation display a significant decrease in the number of proliferating cells in the adult DG, SVZ and RMS, and decreased neurogenesis and DA neurogenesis, as well as decreased survival of newly generated neurons in the OB (Winner, Desplats, et al., 2009a). The combined exposure to the agrichemicals maneb and paraquat or MPTP, powerfully interact with PD mutations and further decrease both SVZ‐ and SGZ‐NSC neurogenic potential, which may contribute to the overall PD pathology (Desplat et al., 2012; Le Grand et al., 2015; Peng & Andersen, 2011; Winner & Winkler, 2015). Conversely, a loss of function mutation of PINK1 (a protein important for mitochondrial quality control via autophagic degradation of damaged mitochondria) negatively impacts on adult neurogenesis. PINK1 deficiency in mice impairs gait, olfaction and serotonergic innervation of the olfactory bulb (Glasl et al., 2012). Recently, Agnihotri and collaborators (2019), reported that PINK1 deficiency is associated with increased deficits of adult hippocampal neurogenesis and lowers the threshold for stress‐induced depression in mice.

As a whole, adult neurogenesis is highly susceptible to multiple “risk factors” for PD, including genetic mutations, DAergic innervation, aging and exposure to environmental toxins. The presence of NSCs in the human PD brain corroborates the potential for developing therapeutic strategies aimed at stimulating the endogenous NSC pool to sustain neurorepair and/or target the pre‐motor symptoms of PD. However, little is known about the molecular and cellular mechanisms of aging and PD‐induced NSC down‐modulation. Based on our own studies and the background literature, the hypothesis being highlighted here relies on the identification of WβC‐signalling as a common hallmark of both extrinsic and intrinsic signals impacting on NSC homeostatic regulation.

Using immunohistochemistry to localize β‐catenin in the SVZ of saline‐ and MPTP‐treated mice to unravel the potential role of WβC‐signalling in SVZ‐NSCs during PD, we first identified a significant proportion of β‐catenin⁺ cells co‐expressing bromodeoxyuridine (BrdU), indicative of proliferation (L'Episcopo et al., 2012). By contrast, MPTP treatment sharply down‐regulated the β‐catenin immunofluorescence signal and β‐catenin⁺ cell proliferation (Figure S2). Since β‐catenin is expressed in type C cells (Adachi et al., 2007) and given that MPTP reduced the proliferation of type C cells, we next addressed the relevance of the SVZ‐WβC‐signalling pathway in MPTP‐induced SVZ neurogenic impairment, by the use of WβC‐signalling activators or inhibitors. Activation of WβC was performed using a specific GSK‐3β antagonist, AR‐AO14418 (AR; Osakada et al., 2007), by either systemic injections or local SVZ administration by intracerebroventricular (icv) infusion. In mice exposed to MPTP, GSK‐3β‐antagonist administration resulted in a sharp increase in β‐catenin expression and cell proliferation in the SVZ ipsilateral to the infusion, therefore counteracting the MPTP‐induced decrease in proliferating cell nuclear antigen‐positive cells and β‐catenin expression in the SVZ (Figure S2). These findings further indicated that the acute exogenous activation of WβC‐signalling during the temporal window of maximal neurogenic impairment may overcome the disrupted neurogenesis observed in the SVZ of MPTP mice in vivo (L'Episcopo et al., 2012). Conversely, WβC antagonism in intact mice using the soluble inhibitor Dkk1, infused via icv, was able to mimic the MPTP‐induced decrease in β‐catenin⁺ cells (L'Episcopo et al., 2012). This inhibitory effect was efficiently counteracted by concomitant activation of the downstream transcriptional effector, β‐catenin, with systemic AR injections (Figure S2). Very recently, the beneficial effect of WβC activation was reported (Singh, Mishra, Bharti, et al., 2018b; Singh, Mishra, Mohanbhai, et al., 2018a) in the 6‐OHDA rat model of PD. Here, activation of WβC‐signalling using the specific GSK‐3β inhibitor SB216763 efficiently counteracted the 6‐OHDA‐induced neurogenic impairment at both the SVZ and SGZ levels. In the hippocampus, GSK‐3β inhibition enhanced dendritic arborization and survival of granular neurons and stimulated NSC differentiation towards the neuronal phenotype in the DG (Singh et al., 2018b; Singh, Mishra, Mohanbhai, et al., 2018a).

Together, these in vivo findings suggested a disruption of WβC‐signalling associated with decreased proliferation and neuroblast formation in the MPTP‐injured SVZ, and further documented the ability of pharmacological activation of β‐catenin‐mediated signalling to counteract the impaired neurogenic potential of MPTP or Dkk1‐infused mice. Coupled to the recent findings obtained at the SGZ level, these data highlight WβC as a key pathway for the regulation of NSC homeostasis during basal ganglia injury induced by the neurotoxins MPTP and 6‐OHDA, both in the SVZ‐ and hippocampal SGZ‐niches.

Looking at Wnt signalling components using western blot analysis we found a decreased β‐catenin signal in NSCs isolated from MPTP‐ but not saline‐treated mice, whereas the active GSK‐3β signal was sharply increased. We then used quantitative real‐time PCR to show that expression levels of Axin‐2, a direct Wnt target induced by WβC activation (Jho et al., 2002), were down‐regulated in MPTP‐NSCs as compared to Axin‐2 expression levels of saline‐NSCs, thereby supporting MPTP‐induced inhibition of WβC signalling activity in SVZ cells (L'Episcopo et al., 2012).

Next, we took advantage of a small interfering RNA (siRNA) strategy (He & Shen, 2009) to further examine the relationship between the β‐catenin signalling pathway in MPTP‐induced neurogenic impairment, testing the role of β‐catenin and GSK‐3β, in NSCs isolated from MPTP‐ and saline‐treated mice. Transient transfection of NSCs from MPTP‐treated mice with anti‐Gsk3b siRNA resulted in a significant increase in the percentage of BrdU⁺ and microtubule‐associated protein 2a (MAP2a)⁺ cells relative to those treated with a control siRNA, suggesting that MPTP‐induced neurogenic SVZ‐impairment is at least in part mediated by GSK‐3β over‐activation (L'Episcopo et al., 2012). Reciprocally, when NSCs from saline‐treated mice were transiently transfected with anti‐β‐catenin siRNA (He & Shen, 2009), we observed a reduced percentage of BrdU⁺ and MAP2a⁺ cells, thereby corroborating the crucial role of β‐catenin in maintaining SVZ‐NSC neurogenic potential.

To look more deeply into the MPTP‐induced impairment of NSC homeostasis, we addressed the effect of a direct exposure of NSCs to MPP⁺ (the neurotoxic oxidation product of the MPTP prodrug) in vitro and found, here again, that the β‐catenin signal was downregulated in face of a sharp up‐regulation of active GSK‐3β signal. In stark contrast, exogenous activation of WβC‐signalling with the selective GSK‐3β antagonist AR, or the Wnt ligand Wnt1, efficiently reversed both the decreased β‐catenin signal and MPP⁺‐induced decrease in NSC proliferation and neuron differentiation (L'Episcopo al. 2012).

Together, the “in vivo”, “ex vivo” and “in vitro” results clearly established MPTP/MPP⁺‐induced inhibition of WβC‐signalling activity in SVZ‐NSCs as a crucial step in the neurogenic impairment of PD mice.

Within the mechanisms affecting adult neurogenesis in brain diseases, oxidative, and especially nitrosative stress, are likely to play critical roles, given their contribution to the aging process and the development of age‐related diseases (Shetty et al., 2018). In PD especially, glial inflammatory mechanisms have long been recognized to contribute to both nigrostriatal degeneration and self‐repair (see Gao et al., 2011; Hirsh & Hunot, 2009; L'Episcopo, Tirolo, Peruzzotti‐Jametti, et al., 2018a; L'Episcopo, Tirolo, Serapide, et al., 2018b; Marchetti & Abbracchio, 2005; Marchetti, et al., 2005a, 2005b, 2005c; McGeer & McGeer, 2008; Marchetti et al., 2011). A critical role of both central and peripheral inflammation has become evident, with a panel pro/anti‐inflammatory cytokines regulating neurogenesis (Butovsky et al., 2006; Chinta et al., 2013; Ekdhal, Kokaia, & Lindvall, 2009; Jakubs et al., 2008; L'Episcopo et al., 2011a; Monje et al., 2003; Pluchino et al., 2008; Shetty et al., 2018; Tepavcevic et al., 2011; Thored et al., 2009; Villeda et al., 2011; Vucovic et al., 2012; Wallenquist et al., 2012). In our studies (L'Episcopo et al., 2012, 2013), we investigated the inflammatory SVZ response as a function of aging and MPTP exposure, addressing the contribution of astrocytes and microglia in MPTP‐induced SVZ impairment (Figure S3). We explored the nature of the astrocyte and microglial‐derived factors involved, using astrocytes and macrophages/microglia acutely isolated ex vivo, from saline‐ or MPTP‐treated mice (L'Episcopo et al., 2012, 2013). In the same years, by using NSC culture techniques, intraventricular tumor necrosis factor (TNF)‐α infusion and the 6‐OHDA mouse model, mimicking PD‐associated neuroinflammation, Worlitzer et al. (2012) showed significant detrimental effects on SVZ‐NSCs, due to a decreased generation of DCX⁺ neuroblasts, but the signalling mechanism(s) underlying such inflammatory‐mediated detrimental effects were not fully clarified. Then, in our in vivo studies of MPTP injury, we linked early SVZ impairment to WβC down‐regulation in PD mice and the early up‐regulation of microglial phagocyte oxidase (PHOX) and inducible‐nitric oxide synthase (iNOS), generating the highly toxic peroxynitrite fingerprint 3‐nitrotyrosine and defining the exacerbated, proinflammatory, microglial M1 phenotype. Such up‐regulation of oxidative and nitrosative status likely contributed to NSC mitochondrial dysfunction, thus increasing NSC vulnerability to cell death (L'Episcopo et al., 2012, 2013). Additionally, the over‐activation of microglia, coupled with the marked astrocytosis associated with a sharp increase in major proinflammatory cytokines observed in vivo, shifted the SVZ niche to a harmful environment for neuroblast proliferation and differentiation.

The direct effects of the changing glial environment were next studied ex vivo and in vitro, using different glial‐NSC co‐culture paradigms, along with pharmacological antagonism/RNA silencing experiments coupled to functional studies. Hence, we provided the first evidence supporting a critical role of SVZ reactive astrocytes and macrophages/microglia in the remodeling of the SVZ niche of parkinsonian mice, identifying WβC‐signalling as a critical player in NSC‐glia crosstalk.

Significantly, GSK‐3β/β‐catenin disruption appeared to be a potential candidate mediator of MPTP‐microglia‐induced NSC impairment. Such a hypothesis was supported by different lines of evidence, with the pharmacological inhibition of microglial reactive oxygen species (ROS) and reactive nitrogen species (RNS): (i) normalizing GSK‐3β activity; (ii) affording a significant reversal of NSC impairment; and (iii) up‐regulating β‐catenin levels in NSCs, corroborating crosstalk between inflammatory and WβC‐signalling components in SVZ‐NSCs (Figure 4).

We next addressed whether the age‐dependent dysfunctional status of the SVZ microenvironment might influence the prototypical components of Wnt/β‐catenin pathway in the aging SVZ. Using real‐time PCR we identified that young adult SVZ‐NSCs harbor most Fzds, but Fzd1 was the most abundant within the transcripts studied. On the other hand, while NSCs from the aged SVZ exhibit small changes in Fzd transcript levels, after MPTP challenge a significant decrease in Fzd1 was observed in aging, but not young, NSCs. In keeping with this result, western blotting showed a sharp Fzd1 down‐regulation only in MPTP‐aged NSCs, indicating that neurotoxin exposure may impair the ability of aged NSCs to respond to Fzd1 ligands. Both aging and neurotoxin exposure exerted a synergic inhibition of canonical WβC activation in SVZ cells, as reflected by decreased CTNNB1 (β‐catenin) and AXIN2 transcript levels, associated with a sharp upregulation of Gsk3b. Immunohistochemistry showed reduction of β‐catenin⁺ cells associated with reduced BrdU incorporation in the SVZ of aged compared to younger mice, thereby supporting aging‐induced dysregulation of WβC‐mediated signalling both in the VM (L'Episcopo et al., 2011a) and SVZ (L'Episcopo et al., 2012).

The development of anti‐inflammatory drugs targeting inflammatory molecules to preserve adult neurogenesis during PD neurodegeneration‐induced inflammation has been addressed by an increasing number of studies (see L'Episcopo, Tirolo, Serapide et al., 2018b; Peruzzotti‐Jametti, et al., 2018a). The studies of Worlitzer et al. (2012) showed that pharmacological inhibition of neuroinflammation in the 6‐OHDA mouse model by the semi‐synthetic tetracycline derivative minocycline led to increased NSC proliferation in the SVZ (Worlitzer et al., 2012). Also, in the 6‐OHDA minocycline‐treated group the proportion of neuroblasts that had migrated deeply (50 μm and more) into the lesioned Str was increased 2–4‐fold compared to all other groups, however no newly generated striatal neurons could be detected (Worlitzer et al., 2012). Four months after surgery, in the presence of minocycline therapy, the authors observed increased oligodendrogenesis in the lesioned Str which was associated with a significant behavioral improvement in PD symptoms, thus indicating oligodendrocytes and oligodendrogenesis as having an important role to play in the pathology of PD (Worlitzer et al., 2012). In our own studies (L'Episcopo et al., 2012) we addressed the potential of anti‐inflammatory drug treatment in the modulation of SVZ neurogenesis in MPTP mice, and its ability to reverse WβC down‐modulation both in young and aging mice, using the mixed cyclooxygenase inhibitor HCT1026. HCT1026 is endowed with a favourable safety profile and shown to downregulate microglial activation via the inhibition of PHOX and iNOS‐derived RNS, resulting in a significant protection of nigrostriatal DAergic neurons in PD rodent models, in both young and aging mice (L'Episcopo, Tirolo, Caniglia, et al., 2010b; L'Episcopo et al., 2011c; L'Episcopo, Tirolo, Testa, et al., 2010a). HCT1026 efficiently increased β‐catenin protein and β‐catenin⁺ staining in SVZ of HCT1026‐MPTP mice of both ages back to normal untreated control levels in the face of a sharp downregulation of active GSK‐3β, as compared to MPTP mice fed with a control diet. These effects of HCT1026 were associated with a significant increase in the number of BrdU⁺ and DCX⁺ expressing cells in the striatal SVZ, correlating with a strong inhibition of reactive microglial cell density in the Str and SVZ (L'Episcopo et al., 2013). Treatment of aging mice with HCT1026 resulted in a severe down‐regulation of microglial pro‐oxidant and pro‐inflammatory mediators in the SVZ, including Nos2 (iNOS) and TNFα, and a reversal of MPTP‐induced up‐regulation of inflammatory mRNA species, as opposed to aged mice fed with a control diet. This effect was accompanied by the normalization of redox/inflammatory balance in aged‐HCT1026 mice, as revealed by a robust up‐regulation of Nrf2 and Hmox1 transcripts associated with microglial inhibition, thus confirming HCT1026‐induced up‐regulation of an SVZ anti‐inflammatory response. Furthermore, a significant degree of DAergic re‐innervation followed HCT1026 treatment of MPTP mice, in line with a significant protection of midbrain cell bodies, confirming our previous report (L'Episcopo, Tirolo, Caniglia, et al., 2010b; L'Episcopo et al., 2011c).

Finally, we uncovered a crosstalk between Wnt/β‐catenin pathway and the pivotal phosphoinositide 3‐kinase (PI3K)/protein kinase B (Akt)/GSK‐3β signalling cascades, that finely control the transcriptional activator β‐catenin, which in turn represents a point of convergence to direct proliferation/differentiation/survival in the SVZ stem niche. The “rejuvenation” of the SVZ may have beneficial consequences for DAergic neuroprotection, and vice versa (L'Episcopo et al., 2013).

Targeting Nrf2 as a promising therapeutic avenue in neurodegeneration has been recently reviewed in the work of Abdalkader et al. (2018), underscoring its important role in modulating neurodegenerative processes, including its ability to suppress mitochondrial dysfunction, likely underlying its crucial role in efficiently protecting and driving an anti‐stress response within the major neurogenic niches.

Coupled with the previous sections, evidence may suggest exogenous activation of β‐catenin signalling, or pharmacological mitigation of a microglial proinflammatory phenotype up‐regulating β‐catenin in the SVZ, successfully rescues NSC proliferation and neuroblast formation. These findings implicated modulation of the Wnt/β‐catenin/Nrf2/Hmox1 axis, either directly or indirectly via inflammation‐dependent SVZ modulation, as a potential strategy for DAergic neuroprotection/self‐repair (Marchetti & Pluchino, 2013) (Figure 4).

Studies of the last few years have identified the Aq‐PVR as a novel Wnt/β‐catenin responsive brain region and addressed the properties of Aq‐PVR‐NSCs in MPTP‐induced PD (L'Episcopo et al., 2014a). Looking at the factors/mechanisms regulating the behaviour of these Aq‐PVR‐NSCs in vitro, we addressed the neurogenesis‐promoting and inhibitory conditions during the process of aging and MPTP‐induced nigrostriatal injury, investigating the potential to activate these progenitors to rescue DAergic plasticity in aged‐MPTP mice. To this end, we also took advantage of transgenic BATGAL mice expressing nuclear beta‐galactosidase under the control of the β‐catenin‐activated transgene (BAT) promoter (Maretto et al., 2003), together with both in vivo experimental PD young and older mice models, coupled to ex vivo/in vitro cell cultures of midbrain‐Aq‐PVR‐NSCs (mNSCs).

We first characterized mNSC properties during in vitro clonal expansion supporting their neurogenic potential as indicated by the expression of proliferation, precursor, pro‐neural, and astrocyte cell markers, and very low forebrain (distal‐less homeobox 2, Dlx2) and hindbrain (homeobox D3, Hoxd3) markers, as opposed to the high expression of the midbrain marker engrailed 2 (En2) (L'Episcopo et al., 2014a). We next addressed the changes of mNSC response as a function of age and PD, by assessing the proliferative, differentiation and survival properties of mNSCs from 2–5 (young, Y) and 9–12 (aged, A) month‐old mice, in basal conditions and during MPTP‐induced injury and recovery. As observed in SVZ‐NSCs, we then showed that with the aging process, the percentages of both BrdU⁺ incorporation and Tuj1⁺ (β‐III‐tubulin) cell formation significantly decreased (L'Episcopo et al., 2014a). While Y‐mNSCs isolated 7 days post‐MPTP treatment responded with only a transient decrease of their proliferation and neuronal differentiation potential that returned back to pre‐MPTP levels by 45 days post‐treatment, in A‐mNSCs MPTP did not change the low levels of A‐mNSC proliferation and neuron formation after in vitro expansion.

These findings, coupled to the increased Caspase‐3‐like activity of A‐ as compared to Y‐counterparts, suggested reduced mNSC survival capacity as a result of aging and MPTP, underscoring reduced mNSC plasticity, a phenomenon correlating with the failure of TH⁺ neurons to recover from MPTP insult.

We thus addressed the potential to overcome aging‐induced loss of Wnt1‐mediated activation of mNSCs in the Aq‐PVR, by pharmacologically targeting this niche in situ via intracerebral injection of the GSK‐3β‐antagonist AR: direct WβC‐signalling activation in aged‐MPTP mice resulted in a marked stimulation of proliferation and differentiation of Aq‐PVR‐Nurr1⁺/TH^‐ precursors into fully differentiated TH⁺/Nurr1⁺ DAergic neurons, thus contributing to both the histopathological and functional restoration of nigrostratal DAergic neurons (L'Episcopo et al., 2014a) (Figure 6). Recently, using Nestin‐CreER^TM::ROSA26‐LacZ mice and a CD133‐Promoter2‐Cre plasmid construct in 6‐OHDA intranigral injected mice, Xie et al. (2019) traced the origin of SN newborn DAergic neurons. The authors showed a significant increase in the SVZ‐derived NSCs of the third ventricle (3V) and Aq‐surrounding regions. Hence, the SN newborn DAergic neurons were mainly derived from the migration and differentiation of the NSCs in the SVZ of 3V‐ and Aq‐adjacent regions (Xie et al., 2017), thus supporting our indication that quiescent neuroprogenitors of mNSCs can be activated not only in vitro (L'Episcopo et al., 2011c), but more importantly, in vivo, upon SNpc‐DAergic lesions with PD neurotoxins (L'Episcopo et al., 2014b).

After our first studies demonstrating dysregulated WβC‐signalling in the MPTP mouse model of PD, both in vivo and in vitro (L'Episcopo et al., 2011a;b), more recent reports corroborated the key role of WβC in the physiopathology of PD (see Marchetti, 2018 for an up‐to‐date review). Among others, Singh, Mishra, Bharti, et al. (2018b) used short hairpin (sh)RNA‐lentiviruses to genetically knockdown AXIN2 to up‐regulate WβC‐signalling in the SNpc, in 6‐OHDA‐induced parkinsonism rats. Not surprisingly, genetic knockdown of AXIN2 (a WβC‐pathway suppressor, see Section 2), up‐regulated WβC‐signalling with beneficial consequences for DAergic neurogenesis (Singh, Mishra, Bharti, et al., 2018b).

Coupled to the bulk of results obtained in vivo in the two β‐catenin reporter lines (L'Episcopo et al., 2014b; L'Episcopo, Tirolo, Peruzzotti‐Jametti, et al., 2018a), and in conjunction with literature findings, these studies documented and corroborated that in young mice the MPTP‐induced DA neuron death (a) activates endogenous Wnt/β‐catenin signalling in the Aq‐PVR niche; (b) β‐catenin is highly expressed in postmitotic Nurr1⁺/TH^‐ neurons with an ongoing β‐catenin signalling activation both in Nurr1⁺/DAT^‐ pre‐DAergic and recovering DAT⁺ neurons during the neurorescue process of young MPTP mice; and (iii) it is possible to reactivate the aged Aq‐PVR niche via both in situ infusion of Wnt/β‐catenin activators or nigral NSC transplantation (L’Episcopo et al., 2011a).

Earlier studies on PD mouse models first established that NSCs “possess an intrinsic capacity to rescue dysfunctional neural pathways” via multiple homeostatic effects (see Madhavan, Daley, Paumier, & Collier, 2009; Madhavan, Daley, Sortwell, & Collier, 2012; Madhavan, Ourednik, & Ourednik, 2008; Ourednik et al., 2009; Redmond et al., 2007). Several studies carried out in rodents and non‐human primate models of PD have shown the ability of transplanted NSCs to survive, proliferate, migrate and produce functional effects via the production of neuroprotective, neurogenic, angiogenic, and anti‐inflammatory factors in the PD brain. Notably, NSCs grafts can also establish neurogenic niches in nonneurogenic regions, and are capable to generate neurons and glia, thus improving the microenvironment, the plasticity and function of the aged hippocampus (Shetty & Hattiangady, 2016).

However, very little is known on NSC ability to promote neuroprotection in the aged PD brain. Recently, using SVZ‐derived adult NSCs transplanted into the aged MPTP‐injured SN, we showed a remarkable time‐dependent endogenous nigrostriatal DAergic neurorestoration program orchestrated by astrocyte‐derived Wnt1 (L'Episcopo, Tirolo, Serapide, et al., 2018b). Here, within the aged microenvironment, NSC grafts survived and proliferated within the SN of both sides in situ. A significant fraction of transplanted cells acquired an astrocytic phenotype both at the SN level and at the midbrain Aq‐PVRs, with a robust migration of NSCs to the Wnt‐sensitive midbrain DAergic niche. NSC grafts in the aged SNpc promoted the expression of Wnt1 in both NSC‐derived and reactive endogenous astrocytes (RAS), leading to Fzd1/β‐catenin signalling in SNpc‐TH⁺ neurons. NSC grafts did not change the mRNA levels for a number of growth and neurotrophic factors, including glial cell‐derived growth factor (Gdnf), insulin‐like growth factor 1 (Igf1), nerve growth factor (Ngf), transforming growth factor α (Tgfa), Tgfb and Fgf8. By contrast, several DAergic‐specific developmental factors including En1/En2, and the direct WβC‐signalling targets, Lef1, LIM Hhomeobox transcription factor 1 α (Lmx1a) and Fgf20, as well as the indirect Wnt1/β‐catenin targets pituitary homeobox 3 (Pitx3) and brain‐derived neurotrophic factor (Bdnf) (see Zhang et al., 2015, and references therein), were up‐regulated of about 2.5–3.5‐fold in MPTP + NSC grafts versus MPTP + PBS sham controls, with Wnt1 being the most up‐regulated DAergic‐specific neurodevelopmental and pro‐survival factor amongst those studied (L'Episcopo, Tirolo, Serapide, et al., 2018b). Likewise, looking at 19 WβC‐signalling members, the canonical Wnts components were sharply down‐regulated in the aged MPTP brain, whereas major Wnt‐antagonists, including the Dkk and sFrp families of Wnt inhibitors, and Gsk3b showed marked up‐regulation of 6‐to‐14‐fold in aged‐MPTP tissues. In stark contrast, NSC grafts, again, reversed the MPTP‐induced up regulation of Dkks, sFrps and Gsk3b (L'Episcopo, Tirolo, Serapide, et al., 2018b).

At the Aq‐PVRs, such an up‐regulation of WβC‐signalling promoted the accumulation of β‐catenin‐immunofluorescence and β‐catenin⁺/DAPI⁺ cells, as unveiled in BATGAL/NSC‐grafted MPTP mice (L'Episcopo, Tirolo, Serapide, et al., 2018b). Here, astrocyte‐derived Wnt1 expression recapitulated a genetic Wnt1‐dependent DAergic developmental programme inciting the acquisition of the mature Nurr1⁺/TH⁺ neuronal phenotype in post‐mitotic DAergic precursors, thus increasing the endogenous pool of midbrain DAergic neurons contributing to the reversal of MPTP‐induced nigrostriatal toxicity (L'Episcopo, Tirolo, Serapide, et al., 2018b). As a functional outcome of this Wnt‐mediated Aq‐PVR neuroprogenitor activation, TH⁺ cell bodies with different morphologies and size were seen coursing from the Aq‐PVR and along the midline down to the ventral tegmental area (VTA) of NSC grafted mice, mimicking TH⁺ neurons trafficking from the Aq to the SNpc observed in younger MPTP mice during SNpc neurorepair (L'Episcopo et al., 2014a; L'Episcopo, Tirolo, Serapide, et al., 2018b).

Corroborating the role of WβC‐signalling in Dkk1‐treated NSC‐grafted MPTP mice, we observed no TH⁺ neurons in Aq‐PVRs. These data, thus, demonstrated that NSC grafts activate β‐catenin transcription, inducing a marked increase of TH⁺ cells within the Aq‐PVR and VTA, whereas this process is blocked by WβC‐antagonism, finally resulting in the inhibition nigrostriatal histopathological and functional repair (L'Episcopo, Tirolo, Serapide, et al., 2018b).

Together, these novel findings supported the potential of nigrostriatal restoration by activating WβC‐signalling in Wnt‐responsive niches, through either pharmacological and cellular approaches aimed at activating/recruiting endogenous progenitors and rescuing the imperiled/diseased DAergic neurons. Moreover, these data suggested that WβC‐signalling activation by NSC grafts at the SNpc and Aq‐PVR is required for NSC‐promoted DAergic functional restoration of aged PD mice. Thus, harnessing WβC‐signalling represents a potential means to boost the endogenous self‐repair/regenerative capacity of the aged PD brain (Figure 6).

Different conditions, cell therapies and/or pharmacological treatments are being studied for their potential to modulate endogenous neurogenesis to favour neuroprotection, neurorepair and immunomodulation (Wenker & Pitossi, 2019), with an increasing number of manipulations being associated to either a direct or indirect ability to up‐regulate the WβC‐signalling cascade (Tables 1 and 2, Figure 7). Owing to its crucial function in stem cell maintenance and tissue homeostasis, there are potential risks and concerns for modulation of WβC‐signalling, regarding both the safety and selectivity, underscoring “Dr. Jekyll'‐Mr Hyde's” dual facet of WβC‐signalling activation/inhibition (Banerjee et al., 2019; Chen et al., 2019; Herrera‐Arozamena, Martí‐Marí, Estrada, Fuente Revenga, & Rodríguez‐Franco, 2016; Huang, Tang, et al., 2019a; Huang, Yan, et al., 2019b; Janda et al., 2017; Kahn, 2014; Mahmood, Bhatti, Syed, & John, 2016; Maiese, 2015; Narcisi et al., 2016; Nusse & Clevers, 2017). Therefore, indirect WβC modulation appears as an attractive strategy to up‐regulate endogenous neurogenesis. Environmental enrichment, physical exercise, the administration of hormones, anti‐oxidant and anti‐inflammatory molecules, pharmacological, pharmacogenetic and/or epigenetic strategies, optogenetic and neural stimulation, or deep brain stimulation, and stem cell therapies, are all being explored for their potential to reverse neurogenic impairment, incite neurorepair and/or reverse cognitive impairment, apparently through activation of WβC signalling, in experimental models of NDs. Also, a certain number of herbal derivates, primarily from the Traditional Chinese Medicine, endowed with pharmacological properties (including anti‐cancer, anti‐bacterial, and anti‐oxidant activities) are increasingly being studied for their ability to modulate WβC‐signalling (recently reviewed by Liu D et al., 2019), with interesting therapeutic potentials for NDs including PD (Table 1 and 2 and Figure 7).

In recent years, druggable molecular targets and signalling pathways involved in neurogenic processes have been identified, and as a consequence, different drug types have been developed and tested in neuronal plasticity. The field of small molecules as potential tools to selectively activate or inhibit WβC is increasingly recognized with a number of both established and novel modulators, including Wnt3a‐like agonists, siRNAs and inhibitors targeting GSK‐3β, Axin‐LRP5/6 or transcription factor complexes (Table 2, Figure 6). The manipulation of WβC‐signalling has become an attractive strategy to ameliorate in vitro differentiation protocols for increasing the fraction of midbrain DAergic neurons (see Kirkeby et al., 2017; Toledo et al., 2017; Brodski et al., 2019).

Harnessing endogenous pro‐neurogenic mechanisms to counteract the decline of adult neurogenesis and promote DAergic plasticity in the aged brain represents a major goal in the PD physiopathological field. We herein discussed the critical role of WβC‐signalling in rebuilding a regenerative microenvironment and in promoting neuronal differentiation of endogenous or exogenous NSCs, which is pivotal for the recovery of neurologic functions in the aged PD brain.

There is compelling evidence that WβC‐signalling plays a vital role in adult neurogenesis in PD, with the important collaboration of glial pathways. Hence, astrocyte‐derived Wnt1 and WβC activation can engage DAergic neurogenesis/neurorepair in the affected PD brain, as uncovered in transgenic β‐catenin reporter PD‐injured mice. An early downregulation of Wnt signalling in endogenous niches starts by middle age, which is associated to down‐regulation of NSC proliferation and differentiation. Moreover, an up‐regulation of the endogenous sFRP and Dkk families of Wnt‐antagonists, together with the overexpression of the pivotal β‐catenin destruction complex protein, GSK‐3β, in synergy with crucial risk factors, particularly inflammation and MPTP exposure, act in concert to impair Wnt signalling both inside and outside the neurogenic niches, with harmful effects for adult neurogenesis and nigrostriatal neurorestoration. Hence, aged NSCs face a harmful niche microenvironment, and its exacerbation upon MPTP‐induced PD, as a result of a failure of the “WβC‐inflammatory connection” which provides a robust homeostatic regulatory mechanism for NSC survival, proliferation, differentiation and integration.

Herein we provided an overview of the molecular mechanism(s) underlying the crosstalk between Wnt‐signalling and NSCs involving astrocyte and microglial mediators, with a crucial role of astrocytic‐microglia crosstalk, including modulation of the Nrf2/Hmox1/WβC‐axis, critical for protection against exacerbated inflammation, oxidative stress and mitochondrial dysfunction. As WβC‐signalling also plays a prominent role in hippocampal SGZ neurogenesis, and in light of the marked decline of WβC‐signalling components in the aged hippocampus, a comparable “WβC‐inflammatory (dis)connection” might well be at play in PD‐injured SGZ‐niche.

Remarkable potential exists to revert some of these age‐dependent changes, targeting WβC‐signalling components either directly or indirectly. Pharmacological and cellular therapies, in particular NSC‐grafts, and immunomodulation were documented to ameliorate the aged microenvironment, promoting endogenous neurogenesis, and ultimately boosting a neurorestoration program in the aged PD brain via the up‐regulation of WβC‐signalling.

A number of novel molecules and conditions were reviewed for their potential to activate WβC for translational applications in regenerative medicine. Hence, targeting Axin‐LRP6 with phenanthridine derivatives, PP2A modulators, sFRP inhibitors, and surrogate Wnt‐agonists, showed their ability to promote cell survival/protection and/or immunomodulation. Also, targeting GSK‐3β with small inhibitors, such as CHIR or BIO, can promote neuroprogenitor homeostasis and neural induction, and restore WβC‐signalling in iPSC‐derived neuronal progenitor cells from PD patients. Emerging studies at the interface between NSC biology and tissue engineering are being exploited for innovative therapeutic applications in brain repair/regeneration therapies, including optogenetics, neural stimulation, and micro‐and nanocarriers releasing multiple biomolecules involved in WβC activation. These interventions are aimed at mobilizing NSCs efficiently from their niches, and, in combination with sustained release of therapeutic agents, can be envisaged as a promising approach to induce neuronal differentiation of NSCs, down‐regulating the exacerbated pro‐inflammatory microenvironment and promoting neurorepair in the injured aged PD brain.

In conclusion, the continuous investigations and further deepening of our knowledge on WβC‐signalling and its role on endogenous adult stem cell biology, NSC crosstalk within the PD injured microenvironment, the response of NSCs to different pharmacological/cellular strategies, as well as its implication will translate into therapeutic breakthroughs and novel applications. Specifically, harnessing their synergistic interactions may lead to the optimization of cell‐based therapies for PD.