Aging and rejuvenation of neural stem cells and their niches

Summary

Aging has a profound and devastating effect on the brain. Old age is accompanied by declining cognitive function and enhanced risk of brain diseases, including cancer and neurodegenerative disorders. A key question is whether cells with regenerative potential contribute to brain health and even brain ‘rejuvenation’. This review discusses mechanisms that regulate neural stem cells (NSCs) during aging, focusing on the impact of metabolism, genetic regulation, and the surrounding niche. We also explore emerging rejuvenating strategies for old NSCs. Finally, we consider how new technologies may help harness NSCs’ potential to restore healthy brain function during physiological and pathological aging.

Introduction

More than any other organ, the brain feels central to us for its ability to coordinate higher-order cognitive functions. But brain functions deteriorate with age. In parallel, neurodegenerative disorders (e.g. Alzheimer’s and Parkinson’s diseases) and brain cancers (e.g. gliomas) surge in the elderly population. While all cell types in the brain are impacted during aging and could contribute to physiological decline and disease, resident neural stem cells (NSCs) in the adult brain have the potential to generate new neurons (neurogenesis), and regenerate aspects of brain function. Thus, maintaining a healthy stem cell pool in the brain throughout aging could be critical to improve overall brain health and reduce the incidence of neurodegenerative diseases and cancer.

The adult mammalian brain contains two primary reservoirs of regenerative NSCs (known as ‘neurogenic niches’) – the subventricular zone (SVZ) of the lateral ventricles and the dentate gyrus (DG) of the hippocampus (Figure 1) (Bond et al., 2015; Silva-Vargas et al., 2013). A third pool of NSCs has more recently been reported in the hypothalamus (Bolborea and Dale, 2013; Pellegrino et al., 2018). This review will primarily focus on the SVZ and hippocampal neurogenic niches but will also discuss specific results from NSCs in the hypothalamus. Neurogenic niches are specialized microenvironments comprising a variety of different cell types, including cells from the NSC lineage, but also endothelial cells (blood vessels) and microglia (Aimone et al., 2014). The head of the NSC lineage are quiescent NSCs (qNSCs), which are normally dormant but can be activated to generate proliferating NSCs. Activated NSCs (aNSCs) in turn give rise to neural progenitor cells (NPCs), which have the potential to differentiate primarily into new neurons, and, in smaller proportions, into astrocytes and oligodendrocytes (Bond et al., 2015). The SVZ and hippocampal neurogenic niches share similarities in their overall cellular composition and neurogenic potential. However, these reservoirs also exhibit differences in their turnover dynamics (symmetric vs. asymmetric divisions) (Calzolari et al., 2015; Encinas et al., 2011; Obernier et al., 2018), their ability to give rise to glia (oligodendrocytes vs. astrocytes) (Bonaguidi et al., 2011; Ortega et al., 2013), and their contribution to brain function (Aimone et al., 2014; Corsini et al., 2009; Dupret et al., 2008; Gao et al., 2018; Gheusi et al., 2000).

Figure 1. Neurogenesis during aging.

The ability of NSCs to proliferate and produce new neurons declines sharply after development and continues to decline during aging while the incidence of neurodegeneration and age-related diseases increases (the diagram shows conceptual trajectories for neurogenesis and neurodegeneration). The adult mammalian brain contains two reservoirs of regenerative neural stem cells (NSCs): the dentate gyrus of the hippocampus and the subventricular zone of the lateral ventricles (teal green). These niches contain quiescent NSCs that can be activated to produce actively proliferating (activated) NSCs. Activated NSCs have the potential to differentiate into neurons, oligodendrocytes or astrocytes.

In the SVZ, NSCs are located within ‘pinwheel’ structures and contact both the brain’s vasculature and the cerebrospinal fluid (CSF) in the lateral ventricle (Mirzadeh et al., 2008; Shen et al., 2008; Tavazoie et al., 2008). SVZ NSCs produce neuroblasts that migrate along the rostral migratory stream and generate new neurons in the olfactory bulb in rodents (Doetsch et al., 1997; Mizrahi et al., 2006) or incorporate into the striatum in humans (Ernst et al., 2014). SVZ neurogenesis is important for various olfactory functions in rodents, such as odor discrimination and olfactory learning and memory (Bragado Alonso et al., 2019; Gheusi et al., 2000; Sakamoto et al., 2011). Interestingly, SVZ NSCs can also be activated upon injury (e.g. stroke) to generate newborn neurons and astrocytes that can help to repair brain injury in mice (Benner et al., 2013; Faiz et al., 2015; Li et al., 2010a).

In the hippocampal niche, NSCs are located at the border of the inner granule cell layer in the DG, and they generate neuroblasts which migrate along the subgranular zone to produce granule neurons (Gage, 2002; Ming and Song, 2011; Sun et al., 2015). Hippocampal neurogenesis contributes to learning and memory formation (Corsini et al., 2009; Dupret et al., 2008; Gao et al., 2018; Guo et al., 2018) and stress resilience in mice (Anacker et al., 2018; Levone et al., 2015). In humans, the level and duration of hippocampal neurogenesis has been subject to debate, but neurogenic regions with regenerative potential are likely present in adulthood in humans (Boldrini et al., 2018; Moreno-Jimenez et al., 2019; Sorrells et al., 2018; Tobin et al., 2019).

During aging, the ability of NSCs to proliferate and give rise to new neurons decreases dramatically. In vivo labeling and microscopy revealed a decline in neurogenesis in both SVZ and hippocampal neurogenic niches during aging (Ben Abdallah et al., 2010; Bondolfi et al., 2004; Enwere et al., 2004; Kuhn et al., 1996; Luo et al., 2006; Maslov et al., 2004). This decline likely involves a number of cellular processes, including increased NSC dormancy, decreased NSC self-renewal, a decline in neuronal fate commitment, and NSC death (Encinas et al., 2011; Obernier and Alvarez-Buylla, 2019; Urban et al., 2016). The decrease in neurogenesis with age is accompanied by poorer performance on learning and memory tasks and by reduced olfactory discrimination (Enwere et al., 2004; Gage and Temple, 2013; McAvoy et al., 2016), suggesting that age-related NSC defects may have broad functional consequences. In humans, the decline in neurogenesis in elderly individuals has been associated with cognitive impairment and neurodegenerative diseases (Moreno-Jimenez et al., 2019; Tobin et al., 2019). Thus, key questions emerging in this field include: How do regenerative regions containing NSCs change over a lifespan? Do the various regenerative niches differ, or do they engage similar mechanisms? Are the cells that divide in the old brain the same cells that had already divided and produced neurons in the younger brain? And how can the regenerative potential of NSCs be leveraged to promote brain homeostasis and repair?

In this review, we discuss recently discovered mechanisms leading to the age-related decline in neurogenesis in vertebrates, focusing on both intrinsic factors (e.g. metabolism and genetic regulation), and extrinsic factors from the niche (e.g. systemic factors and other cell types). We also review the connection between NSCs and brain diseases, as well as emerging strategies to rescue NSC decline during aging. In addition, we speculate on the evolutionary role of vertebrate neurogenesis, with a focus on human neurogenesis. Finally, we present how new technological advances could shape our understanding of NSC aging at the molecular, niche, brain, and organismal level, and provide new avenues to slow, or potentially even reverse, the age-dependent decline in neurogenesis.

Nutrient-sensing pathways, metabolism, and protein homeostasis during NSC aging

Neurogenic niches are sensitive to changes in nutrient availability and signaling, including those occurring during aging. The various cell types in the niche have very different metabolic and protein homeostasis needs, depending on their state (e.g. quiescent NSCs vs. actively proliferating NSCs). In this section, we review recent evidence supporting the role of these processes in NSC aging and rejuvenation.

Nutrient-sensing pathways

Nutrient-sensing pathways, such as the insulin/IGF1-FOXO pathway, are key conserved regulators of aging (Chantranupong et al., 2015) and are also essential for NSC maintenance and function in both the SVZ and the DG. For example, early work revealed that loss of FOXO transcription factors (FOXOs), the downstream effectors of the insulin/IGF1 pathway, leads to premature exhaustion of the NSC pool (Paik et al., 2009; Renault et al., 2009; Yeo et al., 2013). Consistently, loss of the PTEN phosphatase (which is upstream of FOXO) also causes NSC depletion (Bonaguidi et al., 2011), even if PTEN loss initially triggers NSC proliferation (Gregorian et al., 2009). More recent work has shown that suppression of IGF1 signaling in young NSCs via IGF1 receptor (IGF1R) conditional knockout in vivo (which results in FOXO activation), improves neurogenesis in the SVZ, increases spine density in newborn neurons, and enhances olfactory learning (Chaker et al., 2015). Together, these results indicate that suppressing insulin/IGF1 signaling (and subsequently activating FOXO) is beneficial for the long-term maintenance of the NSC pool. Suppression of signaling pathways downstream of the insulin/IGF1 pathway, such as the mTOR pathway, are also likely to enhance NSC maintenance (see below). However, it is important to note that increased IGF1 signaling can be beneficial for neuronal function by promoting neuronal differentiation and organization during neurogenesis in the hippocampus and the olfactory bulb (Hurtado-Chong et al., 2009; Nieto-Estevez et al., 2016). Thus, alternating cycles of low and high insulin/IGF1 signaling might help maintain the stem cell pool while still allowing NSC proliferation and neuronal function/differentiation. Indeed, cycles of a fasting-mimicking diet with ad libitum feeding, which periodically change the levels of insulin and IGF1, increase hippocampal neurogenesis and cognitive performance in old mice (Brandhorst et al., 2015).

Mitochondria

Mitochondria play a central role in maintaining NSC states during aging. Transcriptomic analysis of NSCs isolated from the SVZ or the hippocampus revealed that quiescent NSCs (qNSCs) express genes involved in beta-oxidation, which partly takes place in mitochondria, while aNSCs express genes involved in mitochondrial oxidative phosphorylation (Leeman et al., 2018; Shin et al., 2015). Indeed, mitochondria are morphologically different in quiescent and activated NSCs in the adult hippocampus: mitochondria are thin and elongated in qNSCs, whereas they have mixed globular and tubular shapes in aNSCs (Beckervordersandforth et al., 2017). During aging, mitochondria become more dispersed and less perinuclear in qNSCs of the SVZ niche (Capilla-Gonzalez et al., 2014). Furthermore, aging causes mitochondria to become more densely packed in hippocampal aNSCs, with decreased membrane potential and lower levels of ATP (Beckervordersandforth et al., 2017). Restoring mitochondrial function in old aNSCs with Piracetam (a drug with suggested use for age-related cognitive decline) can improve hippocampal neurogenesis in vivo in old mice (Beckervordersandforth et al., 2017). Similarly, ectopic expression of proliferator-activated receptor gamma coactivator 1 alpha (PGC1a), a factor that increases cellular aerobic capacity by promoting mitochondrial biogenesis and metabolic gene transcription, also improves neurogenesis in the aged SVZ in vivo (Stoll et al., 2015). As both aNSCs and qNSCs rely on mitochondria to optimally control metabolism, restoring mitochondrial function in old individuals could improve cellular function across the NSC lineage, resulting in an overall boost in neurogenesis.

Lipid metabolism

Lipid metabolism is emerging as an important component of NSC regulation, though its role during NSC aging remains unknown. The degradation of lipids via fatty acid oxidation (FAO) is important for the maintenance of quiescence in young hippocampal qNSCs in vivo, and inhibition of FAO (via addition of malonyl-CoA) is sufficient to induce the exit from quiescence and enhance NSC proliferation in vitro (Knobloch et al., 2017). FAO is also a major regulator of quiescence in hematopoietic (Ito et al., 2012), intestinal (Mihaylova et al., 2018) and muscle stem cells (Ryall et al., 2015). Whether FAO changes with age in NSCs has not yet been examined. However, this process decreases with age in intestinal stem cells (Mihaylova et al., 2018), suggesting that maintaining FAO could play an important role in protecting qNSCs during aging. In contrast, young aNSCs upregulate lipid production through fatty acid synthase (FASN)-dependent de novo lipogenesis in the hippocampus (Knobloch et al., 2013). Furthermore, inhibiting FASN in vivo abolishes the beneficial effects of running on hippocampal neurogenesis and cognitive function in mice (Chorna et al., 2013). Proper FASN function is also required for hippocampal NSC proliferation in both mice and humans (Bowers et al., 2020). Thus, the maintenance of lipogenesis in aNSCs could be essential to counter NSC decline during aging. Lipid accumulation in other cells in the niche may also indirectly contribute to NSC aging. For example, dietary supplementation with the mono-unsaturated fatty acid oleic acid leads to lipid droplet formation in ependymal cells, and this is associated with decreased neurogenesis in the SVZ (in the context of an Alzheimer’s disease mouse model, 3xTg-AD) (Hamilton et al., 2015). Similarly, obesity induces the build-up of senescent glial cells containing excessive fat deposits in the SVZ, and genetic ablation of these senescent glial cells can restore neurogenesis in the SVZ niche (Ogrodnik et al., 2019). Future studies will be needed to determine how specific lipids regulate NSC function during aging and whether lipids could have critical roles in processes beyond metabolism, including membrane homeostasis and signaling. Given the malleability of dietary interventions, modulating the composition of lipids in the diet could provide novel ways to ‘rejuvenate’ old neurogenic niches.

Protein homeostasis

Preservation of a pristine proteome is essential for the maintenance of cells such as NSCs that must continuously function over long periods of time. NSCs and their progeny use different strategies to maintain protein homeostasis (proteostasis): qNSCs and differentiated progeny rely mostly on the lysosome-autophagy pathway whereas aNSCs turn to active proteasomes and chaperones (Kobayashi et al., 2019; Leeman et al., 2018; Schaffner et al., 2018; Vonk et al., 2020). During aging, impaired function of the lysosome-autophagy pathway in qNSCs from the SVZ contributes to their reduced ability to exit quiescence and start proliferating (Leeman et al., 2018). However, this decline is not inexorable. Overexpression of a key transcription factor that activates the lysosome-autophagy pathway (TFEB) in primary cultures of qNSCs from old SVZs improves the ability of these old qNSCs to exit quiescence (Leeman et al., 2018). The role of the lysosome-autophagy pathway in quiescence (and its decline with age) has also been reported in the hematopoietic stem cell pool (Ho et al., 2017) and in the nematode germline (Bohnert and Kenyon, 2017), suggesting that targeting this pathway in old qNSCs could improve dormant cell function and reactivation upon stimulation. Autophagy is also important in proliferative NSCs. The FOXO family member FOXO3 functionally regulates the induction of autophagy genes in primary cultures of aNSCs isolated from the SVZ of young adult mice (Audesse et al., 2019). Finally, differentiated progeny (new neurons) rely mainly on the lysosome-autophagy pathway to maintain a healthy proteome (Schaffner et al., 2018). Induction of autophagy via FOXO transcription factors regulates the morphology and spine density of newborn hippocampal neurons in young adult mice (Schaffner et al., 2018). Thus, targeting the autophagy pathway in old individuals could have multiple beneficial effects by improving the function of qNSCs, helping aNSCs, and favoring newborn neurons.

The function of the proteasome deteriorates in aNSCs during aging, which may contribute to the reduced proliferative capacity of these cells. Expression of proteasomal components declines in SVZ aNSCs with age (Zhao et al., 2016). Enhancing the activity of the proteasome, by either overexpression of PSMB5 (a component of the 20S proteasome complex) or 18α-GA (a proteasome activator), leads to increased NSC proliferation in culture and in vivo in young mice (Zhao et al., 2016), though it remains to be determined whether increased proteasome activity also boosts NSCs from old individuals. Similarly, other proliferative stem cell pools, including embryonic stem cells (Vilchez et al., 2012) and induced pluripotent stem cells (Buckley et al., 2012), exhibit high proteasome activity, which may be essential for their ‘immortal’ features. Enhancing proteasome activity in old proliferating stem cell pools could help to ameliorate proteostatic stress and restore their regenerative potential.

The chaperone network also declines with aging, and this deficit impacts aNSCs and their progeny. Chaperones are heavily remodeled during NSC differentiation (Vonk et al., 2020): proliferating NSCs express high levels of the chaperonin TRiC/CCT while differentiated progeny express small heat shock proteins (Vonk et al., 2020). TRiC/CCT protein levels decline in hippocampal aNSCs during aging (Vonk et al., 2020). It will be interesting to determine the relative contributions of the chaperone network to NSC function and differentiation during aging and its interplay with the other branches of the proteostasis network.

Protein aggregates

The decline in proteostasis during aging leads to the accumulation of aggregated proteins. Interestingly, young adult qNSCs already contain protein aggregates whereas young aNSCs have very few aggregates (Leeman et al., 2018). These observations suggest that quiescent stem cells may tolerate more protein aggregates or may require proteins with increased propensity to aggregate for their function, which may have deleterious consequences during aging. Old qNSCs freshly isolated from the SVZ have more aggregates than young ones (Leeman et al., 2018), consistent with a decline in lysosome-autophagy function in these cells. Quiescent NSCs isolated from the hippocampus also harbor protein aggregates (Morrow et al., 2020). Localization of proteasomes to these protein aggregates, mediated by the intermediate filament vimentin, is important for the transition from quiescence to proliferation (Morrow et al., 2020). The difference in aggregate content between activated and quiescent NSCs may be explained by differences in their protein synthesis rates and proteostasis strategies. Proteasomes in aNSCs may be more efficient at eliminating protein aggregates and the combination of proteasome and autophagy induction in aNSCs may help efficiently remove aggregates (Audesse et al., 2019). Furthermore, the high expression levels of the chaperonin TRiC/CCT in proliferative NSCs also contribute to the maintenance of misfolded protein solubility (Vonk et al., 2020). Finally, by virtue of being proliferative stem cells, aNSCs may dilute or differentially segregate protein aggregates during cell division and differentiation. Indeed, when young aNSCs divide, the endoplasmic reticulum (ER) forms a physical barrier that segregates damaged proteins to the non-stem daughter cell, protecting aNSCs from proteostatic stress (Moore et al., 2015). This barrier weakens with age in primary cultures of aNSCs from the hippocampus of old mice, which leads to an equal distribution of damaged proteins between stem and non-stem daughter cells during cell division and an increase in proteostatic burden in aNSCs (Moore et al., 2015). This increase in proteostatic burden could lead to the age-dependent decline in NSC function by either slowing the rate of NSC proliferation or by promoting a more quiescent state. The specific proteins present in these aggregates could also participate in NSC defects, though the composition of the protein aggregates in young and old NSCs remains entirely unknown.

Integration of nutrient-sensing, metabolism, and proteostasis pathways in NSCs

Overall, the different metabolic and proteostatic strategies used by quiescent and activated NSCs and their progeny are likely necessary to support the fundamental functional differences between these cell types (Figure 2) (see reviews by (Cavallucci et al., 2016; Knobloch and Jessberger, 2017)). Dormant, non-dividing qNSCs have lower metabolic demands and employ mechanisms to produce energy long-term such as the lysosome-autophagy pathway and FAO. In contrast, actively dividing aNSCs have higher metabolic and protein synthesis rates (Baser et al., 2019), so they rely mostly on oxidative phosphorylation – an effective and fast energy production process – and they upregulate lipogenesis. Actively proliferating aNSCs also require chaperones and active proteasomes to remove short-lived proteins, such as cell cycle regulators, as well as damaged or unfolded proteins to prevent them from forming large aggregates that could compromise their function or be transferred on to their progeny. While both metabolic and proteostatic processes ultimately decline with age, this decline can be counteracted. Inducing autophagy can facilitate not only the removal of protein aggregates but also the clearing of defective mitochondria (mitophagy) and the degradation of lipids (lipophagy) to produce free fatty acids for oxidative phosphorylation (Singh and Cuervo, 2011). Similarly, reducing nutrient intake or blocking nutrient-sensing pathways (Insulin/IGF1, mTOR) can improve NSC function by triggering the activation of transcription factors (FOXO, TFEB) that coordinate both metabolism and proteostasis. Given how connected these pathways are, targeting one of these processes in old NSCs may also restore related pathways, resulting in a synergistic effect to improve old NSC function.

Figure 2. Cellular pathways involved in NSC aging.

Cellular and metabolic pathways involved in maintaining NSC and differentiated progeny homeostasis that change during aging.

Nutrient-sensing, metabolism, and proteostasis pathways may also be regulated differentially with age depending on the cell type or even the neurogenic niche. Interestingly, as many aspects of organismal metabolic regulation are non-cell autonomous, it is possible that age-dependent changes in one neurogenic niche could impact other regions. For example, NSCs in the hypothalamus – a region involved in systemic regulation of metabolism – may participate in the regulation of organismal metabolism (Bolborea and Dale, 2013). Understanding the integration of different pathways and their specific actions within neurogenic niches or at a distance in the whole organism will be crucial for identifying strategies to counter NSC aging.

Transcriptional, epigenomic, and cell cycle changes in NSC aging

The decline in neurogenic potential exhibited by aged NSCs is partly mediated by changes to the global transcriptional and epigenetic landscape that regulates stem cell function. Among these global changes, expression changes in key cell cycle regulators have emerged as critical for NSC activation, exhaustion, and senescence. Cell cycle regulators are particularly interesting as they could readily be modulated to improve old NSC function.

Transcriptomic analyses of the aging NSC niche

Analysis of the NSC transcriptome provides a global snapshot of the cellular processes and pathways that are most affected by age. RNA-sequencing (RNA-seq) of freshly isolated astrocytes, qNSCs, aNSCs, and NPCs from the SVZ of young and old mice revealed fundamental transcriptomic differences in aging quiescent and activated NSC states (Leeman et al., 2018). During aging, quiescent cells (astrocytes and qNSCs) undergo more transcriptional changes than proliferative cells (aNSCs and NPCs), including changes in proteostasis pathways (lysosome and proteasome) (Leeman et al., 2018). Consistently, cultured aNSCs from the SVZ of young and old mice displayed few transcriptomic differences with age (Lupo et al., 2018). As aNSCs arise from qNSCs, these results raise intriguing possibilities. The rapid proliferation and differentiation of the transient aNSC population could preclude the accumulation of transcriptional changes. Alternatively, the very process of activation could functionally reset the aging hallmarks of qNSCs, leading to a more ‘youthful’ transcriptional landscape.

One challenge in unravelling how aging affects complex tissues, such as neurogenic niches, is cellular heterogeneity. The rapid expansion in single-cell RNA-seq technologies has helped probe the heterogeneity of cell types and transcriptional profiles in young and old neurogenic niches in the SVZ and the hippocampus (Artegiani et al., 2017; Basak et al., 2018; Dulken et al., 2019; Dulken et al., 2017; Hochgerner et al., 2018; Kalamakis et al., 2019; Llorens-Bobadilla et al., 2015; Luo et al., 2015; Mizrak et al., 2019; Shi et al., 2018; Shin et al., 2015; Zywitza et al., 2018). By capturing the transcriptomes of individual cells along a continuous differentiation lineage, these analyses have identified intermediary cell states during activation and neuronal differentiation (Dulken et al., 2017; Llorens-Bobadilla et al., 2015; Shin et al., 2015), and key genes that change throughout NSC differentiation (Shin et al., 2015). Single-cell transcriptomic analyses of the whole SVZ niche have revealed that the majority of age-related transcriptional changes occur in specific cell populations, namely microglia, endothelial cells, oligodendrocytes, and astrocytes/qNSCs, all of which exhibit strong transcriptional upregulation of interferon signaling pathways with age (Artegiani et al., 2017; Dulken et al., 2019; Kalamakis et al., 2019). Single cell RNA-seq analyses have also shown that both quiescent and activated NSC population numbers are significantly smaller in SVZs from old mice (Dulken et al., 2019; Kalamakis et al., 2019), consistent with previous data by immunofluorescent staining and BrdU incorporation (Luo et al., 2006). Collectively, these single-cell transcriptomic studies have established a global cell atlas of the neurogenic niches, and revealed changes in cell composition (with decreases in NSCs) and in the transcriptome of specific cell populations (e.g. microglia) with aging. It will be interesting to determine which of these age-dependent changes can be reversed by known ‘rejuvenation’ interventions, such as dietary changes.

Chromatin modifiers and epigenomic changes

Analysis of the epigenomic landscape and resulting chromatin states of NSCs during aging can provide key information about how age affects cell identity and function (Klemm et al., 2019). Due to the reversibility of epigenomic marks, identifying the enzymes that are responsible for the epigenomic changes observed during NSC aging is an attractive strategy for therapeutic intervention.

Historically, many studies have been dedicated to understanding the role of DNA methylation, and its upstream modifiers such as DNA methyltransferases, in embryonic and adult neurogenesis (extensively reviewed in (Cui and Xu, 2018; Jobe and Zhao, 2017)). More recently, studies have begun to characterize global DNA methylation changes throughout aging. Global levels of 5-hydroxymethylcytosine (5hmC) decrease with age in whole hippocampus samples and correlate with the age-related decline in neurogenesis (Gontier et al., 2018). In parallel, mRNA levels of TET2, the enzyme that catalyzes the formation of 5hmC from 5mC via oxidation, also decrease in the hippocampus with age. Reducing in vivo TET2 expression in young hippocampal NPCs alone (by injection of Tamoxifen in Tet2^flox/flox; NestinCre-ER^T2 mice) is sufficient to decrease neurogenesis and cognitive performance, suggesting a cell-autonomous effect (Gontier et al., 2018). Conversely, overexpression of TET2 in the DG of young adult mice via in vivo lentivirus administration increases the number of newborn neurons in the hippocampus and improves spatial learning and memory (Gontier et al., 2018). Thus, overexpressing or activating TET2 in middle-aged animals could represent a strategy to enhance both neurogenesis and cognitive performance in aging animals.

Global profiling of histone methylation marks (H3K4me3 and H3K27me3 ChIP-seq) and DNA methylation marks (5-methyl cytosine (5mc) and 5hmC) have uncovered changes in aNSCs cultured from young and old SVZs (Lupo et al., 2018). While there are relatively few epigenomic (and transcriptomic) differences, the regulatory marks associated with the Dbx2 locus change with age (Lupo et al., 2018). The Dbx2 locus encodes a transcription factor that had previously been implicated in spinal cord development (Pierani et al., 1999). Overexpression of DBX2 in young aNSCs in vitro leads to reduced NSC proliferation and to changes in the transcriptome that resemble those occurring in old aNSCs (Lupo et al., 2018). Thus, decreasing DBX2 levels could have potential therapeutic advantages for improving neurogenesis in old individuals.

Several studies have also characterized the global epigenomic differences between quiescent and activated NSCs derived from embryonic stem cells (ESCs) or neonatal mice, although it is not yet known how these epigenomic marks could be affected during aging. Co-localization of H3K27ac (a chromatin mark associated with enhancers) and p300 (a transcriptional co-activator that binds enhancers) has been used to identify quiescent- and activation-specific enhancers genome-wide in cultured murine ESC-derived NSCs (Martynoga et al., 2013). Enhancers in quiescent cells exhibit enrichment for the NFI transcription factor family motif (Martynoga et al., 2013). The loss of the family member NFIX (in Nfix^−/− mice) is sufficient to reduce the proportion of quiescent NSCs in the hippocampus of postnatal mice, implicating NFIX as a key regulator of the quiescent state (Martynoga et al., 2013). Epigenomic assays that profile long-range binding interactions, such as ChIA-PET, have also uncovered how transcription factors bind to regulate NSC maintenance. ChIP-seq and ChIA-PET of murine neonatal forebrain-derived aNSCs identified SOX2 binding sites predominantly located in promoters and enhancers (Bertolini et al., 2019). SOX2 is an important transcription factor for NSC maintenance and its loss causes self-renewal defects (Graham et al., 2003). SOX2 enhancer-promoter interactions are particularly important for the regulation of the gene Socs3, a JAK/STAT signaling inhibitor, and SOCS3 overexpression can rescue the self-renewal defect observed in Sox2^−/− NSCs (Bertolini et al., 2019). However, how NFIX and SOX2, and their target genes, change with age and affect NSC aging is yet to be determined.

Global profiling of epigenomic changes within aging neurogenic niches has identified potential therapeutic targets, such as DBX2 and TET2. However, most of the studies so far have been conducted either on whole tissue samples or using cultured cells. There is a need for single-cell epigenomic assays (such as scATAC-seq (Buenrostro et al., 2015)) to identify heterogeneous epigenetic regulation within and between cell types. In addition, profiling a greater diversity of epigenomic marks on specific, isolated in vivo cell populations in the niche during aging will provide key insights into how to restore this landscape and rejuvenate the niche.

Regulation of cell cycle and senescence

The ability of NSCs to efficiently enter the cell cycle declines with age. Genes involved in regulating the cell cycle, including tumor suppressors, play a key role in the maintenance of NSC quiescence and activation. These same cell cycle regulators are involved in the initiation of cellular senescence, which could diminish the functionality of the stem cell pool during aging. We next discuss how regulation of gene expression programs that control the balance between quiescence and proliferation/commitment is critical to guarantee lifelong neurogenesis and avoid premature stem cell exhaustion and senescence.

p16^INK4a is a negative cell cycle regulator and a prominent biomarker for cellular senescence, a cell state characterized by irreversible cell cycle arrest and distinct morphological and transcriptional changes (reviewed in (Hernandez-Segura et al., 2018)). The expression of p16^INK4a is strongly induced with age in the SVZ and may contribute to the accompanying decline in neurogenesis through cell cycle suppression of aNSCs and NPCs (Molofsky et al., 2006). In middle-aged mice, p16^INK4a suppresses hippocampal NSC proliferation in response to exercise (Micheli et al., 2019). This suppression could be mediated by the ability of p16^INK4a to induce senescence. Indeed, p16^INK4a upregulation (driven by removal of the monocytic leukemia zinc finger protein (MOZ)) leads to replicative senescence in ESC-derived NSCs (Perez-Campo et al., 2014). In cultured NSCs derived from the hypothalamus, expression of an upstream repressor of the p16^INK4a locus, the non-coding RNA Hnscr, declines with age (Xiao et al., 2020). Knocking down Hnscr in vivo in hypothalamic NSCs increases cellular senescence and results in decreased cognitive performance (i.e. novel object recognition) (Xiao et al., 2020). Consistently, the senescence phenotype of primary NSCs cultured from the SVZ of an accelerated senescence mouse model (SAMP8 mice) could be prevented by decreasing the mRNA levels of p19^ARF, another tumor suppressor encoded within the same locus as p16^INK4a, through epigenetic regulation of the locus by histone acetyltransferases (Soriano-Canton et al., 2015). Collectively, these studies suggest that increased expression of negative cell cycle regulators, such as p16^INK4a and p19^ARF, drives senescent phenotypes in NSCs and seems to be at least partially responsible for the decrease in neurogenic potential observed in NSCs during aging.

Conversely, expression of positive cell cycle regulators, such as the Polycomb family member BMI-1 (PCGF4) and cyclin dependent kinases (CDKs), can promote NSC expansion and lead to cognitive improvement. BMI-1 is a known transcriptional repressor of cell cycle inhibitors, including p19^ARF/Mdm2/p53 (Gu et al., 2014), and upregulation of BMI-1 increases self-renewal in hippocampal NSCs (Zelentsova-Levytskyi et al., 2017). Furthermore, simultaneously overexpressing CDK4 and cyclin D1 in hippocampal NSCs in older mice (16 months old) by lentiviral delivery of a transgene in vivo increased hippocampal neurogenesis and rescued aspects of age-related cognitive impairment (Berdugo-Vega et al., 2020).

Surprisingly, however, low-level overexpression of negative regulators of the cell cycle (p16^INK4a/p19^ARF/p53) in mice increases NSC number in both the SVZ and DG of old individuals and improves cognitive performance (Carrasco-Garcia et al., 2015). The beneficial effects of low-level overexpression of negative cell cycle regulators might be due to their ability to prevent premature exhaustion of the stem cell pool. Thus, the timing and levels of cell cycle regulator expression are key to maintaining NSC fitness throughout lifespan.

Overall, the critical nature of cell cycle regulators such as p16^INK4a, p19^ARF, p53, and BMI-1 in regulating NSC viability and neurogenesis has positioned them as potential therapeutic targets to rejuvenate aged NSCs. However, these pathways are delicately regulated. Future work remains to be done to determine the correct dosage and timing of manipulation that will improve NSC health without inadvertently causing premature stem cell exhaustion or senescence.

The role of the niche and inflammation in NSC aging

NSCs reside in complex and specialized microenvironments within the brain that contain different cell types and are influenced by a variety of local and systemic factors. NSCs can integrate signals from the niche to couple their activation state and fate decisions to the tissue demands. The cellular composition of NSC niches, their response to local or systemic cues, and even their physical properties such as stiffness (Segel et al., 2019) change during aging. This section examines recent advances in our understanding of how age-related changes in the niche (e.g. systemic or local extracellular cues, other cell types) can influence and contribute to the decrease in neurogenesis during aging (summarized in Figure 3). While many of the mechanisms that underlie the NSC response to these extrinsic cues are not yet known, they are likely to impinge on the intrinsic pathways described above to impact NSC function. In the future, it will be interesting to better understand the interplay between extrinsic and intrinsic regulators of NSC aging.

Figure 3. The role of the niche and inflammation in NSC aging.

Changes that occur in the NSC niche (SVZ is depicted) during aging (left: young, right: old). Inflammation increases in the niche, highlighted by the increase in inflammatory cytokines, activated microglia, and T cell infiltration.

Systemic blood factors and local factors

Systemic factors in the blood play an important role in NSC aging. Heterochronic parabiosis studies, in which the circulatory system of a young and an old mouse are joined, have revealed positive and negative effects of young and old blood on neurogenesis in both the DG (Rebo et al., 2016; Villeda et al., 2011) and SVZ neurogenic niches (Katsimpardi et al., 2014). A variety of systemic factors and signaling pathways mediating these effects have been identified, many of which are immune-related. CCL11/Eotaxin (Villeda et al., 2011), an inflammatory cytokine, and β₂-microglobulin (Rebo et al., 2016; Smith et al., 2015a), a component of MHC class I, are both elevated in old blood and have negative effects on neurogenesis and cognition in young animals. In contrast, GDF11, a circulating transforming growth factor, appears to be present at higher levels in young blood and to have positive effects on old SVZ NSCs (Katsimpardi et al., 2014), though the extent of GDF11 rejuvenating properties remain unclear (Egerman et al., 2015; Smith et al., 2015b). TIMP2, a metalloproteinase inhibitor found in human umbilical cord plasma, improves synaptic plasticity and cognition when injected systemically in old mice (Castellano et al., 2017). However, the number of newborn neurons in the hippocampus are unchanged upon TIMP2 injection, suggesting that this factor acts independently of neurogenesis, perhaps by changing the properties of the extracellular matrix (Castellano et al., 2017). It also remains to be determined whether these factors mediate their effects on neurogenesis by acting directly on niche cells or indirectly by shifting the niche milieu towards a less inflammatory state. In addition to systemic factors, local morphogens, including Notch, Wnt, sonic hedgehog, and bone morphogenic proteins (BMPs), are critical during embryonic development, and they also regulate NSCs throughout life (reviewed in (Bond et al., 2015; Ming and Song, 2011). Upregulating Wnt signaling in the hippocampus can counteract the age-related decrease in neurogenesis and its associated cognitive decline (Seib et al., 2013). Notch signaling also plays an important role in maintaining NSC quiescence and activation in both SVZ and hippocampal niches (Basak et al., 2012; Blomfield et al., 2019; Engler et al., 2018; Zhang et al., 2019), and could be critical during aging.

Choroid plexus and cerebrospinal fluid

The choroid plexus is a monolayer of epithelial cells in the brain ventricles that produces cerebrospinal fluid (CSF) (Johanson et al., 2011) and constitutes the blood-CSF barrier, integrating signals from both systems. Analysis of the effects of the lateral ventricle choroid plexus secretome on the growth of freshly purified NSCs in vitro, revealed that aNSCs from the SVZ are very sensitive to age-related changes in secreted factors from the choroid plexus (Silva-Vargas et al., 2016). Specifically, factors that can induce NSC proliferation, such BMP5 and IGF1, are depleted from the choroid plexus secretome of old mice (Silva-Vargas et al., 2016). The aged choroid plexus also exhibits an expression profile corresponding to type I interferon response (IFN-I; i.e. IFNα and IFNβ) (Baruch et al., 2014). Notably, blocking IFN-I signaling by delivery of antibodies against the IFNα receptor into the CSF improves hippocampal neurogenesis and cognitive function in old mice (Baruch et al., 2014). Hypothalamic NSCs can secrete exosomes into the CSF at the third ventricle (Zhang et al., 2017). Thus, NSCs in the hypothalamic niche may not only be influenced by the CSF, but may also contribute to systemic changes that impact the aging process (Zhang et al., 2017). The choroid plexus and CSF play a crucial role relaying changes in the brain environment, including inflammatory signals during aging, to the NSC niche.

Endothelial cells and pericytes

The brain’s vasculature (endothelial cells and surrounding pericytes) forms a tightly regulated interface between the circulatory system and the brain parenchyma known as the blood brain barrier. Recent studies have shown that the transcriptome of endothelial cells in the neurogenic niche changes dramatically with age, shifting towards an inflammatory transcriptomic profile (Dulken et al., 2019; Kalamakis et al., 2019; Leeman et al., 2018; Yousef et al., 2019). Endothelial cells and pericytes can secrete factors, such as placental growth factor 2 (PlGF-2), that promote the proliferation of NSCs isolated from the SVZ of young mice (Crouch et al., 2015). However, in middle-aged mice, endothelial cells start producing increased levels of transforming growth factor-β (TGF-β), an inflammatory cytokine that triggers NSC apoptosis via TGF-β and SMAD3 signaling (Pineda et al., 2013). Consistently, genetic or pharmacological attenuation of the TGF-β pathway can rescue neurogenesis in old mice (Yousef et al., 2015). Endothelial cells also display focal upregulation of the vascular cell adhesion molecule (VCAM1) in old brains (Yousef et al., 2019). Decreasing VCAM1 levels in endothelial cells, either genetically or via systemic anti-VCAM1 antibodies, reverses the negative effects of old plasma in the young brain and increases the number of NSCs in the hippocampus of old mice (Yousef et al., 2019). While VCAM1 is normally known for facilitating vascular-immune cell interactions (Osborn et al., 1989; Schlesinger and Bendas, 2015), during aging VCAM1 may allow tethering of blood cells to endothelial cells, but not their transport, and this may induce chronic inflammation of endothelial cells (Yousef et al., 2019). The inflamed state of endothelial cells could in turn lead to microglia activation (see below), inhibition of NSC proliferation, and cognitive impairment (Yousef et al., 2019). Thus, restoring endothelial cell function in old brains may help to reduce inflammation, ameliorate NSC decline, and improve brain function.

Microglia

Microglia are the resident immune cells of the brain. These cells patrol the brain parenchyma, maintaining homeostasis through the engulfment and degradation of extracellular materials via phagocytosis. The phagocytic activity of microglia helps to maintain homeostasis in the adult hippocampal neurogenic niche by removing newborn neural progenitor cells that undergo apoptosis before becoming neuroblasts (Sierra et al., 2010). Furthermore, the secretome of phagocytic microglia limits the production of new neurons both in hippocampal NSC cultures and when injected directly into the hippocampus of young adult mice (Diaz-Aparicio et al., 2020). However, during aging, microglial phagocytosis is impaired (Marschallinger et al., 2020; Pluvinage et al., 2019) which may contribute to the accumulation of debris and aggregates in the niche and may perturb NSC niche homeostasis. Microglia also normally support neurogenesis in the hippocampus and SVZ through the production of growth factors and cytokines in young animals (Shigemoto-Mogami et al., 2014; Ziv and Schwartz, 2008). Sustained inflammation during aging leads to increased microglial activation (i.e. proliferative microglia secreting inflammatory cytokines) and subsequent reduction in NSC proliferation in both the SVZ and the hippocampus (Bachstetter et al., 2011; Monje et al., 2003; Solano Fonseca et al., 2016). Therefore, improving microglial function in old brains might not only boost the function of other cells (e.g. NSCs) in the niche, but also actively promote clearance and reduce infiltration of immune cells in the niche.

Infiltration of T cells

The brain has been long regarded as an immune-privileged organ, especially under physiological conditions. Interestingly, however, immune cells, such as T cells, infiltrate the brain of old individuals in mice and humans (Dulken et al., 2019; Mrdjen et al., 2018; Ritzel et al., 2016). Single-cell RNA sequencing of the entire SVZ niche revealed the infiltration of CD8+ T cells in the old SVZ, which was confirmed by immunocytochemistry in both aged mice and elderly humans (Dulken et al., 2019). T cell receptor sequencing showed that brain T cells are clonally expanded (Dulken et al., 2019), suggesting that they may recognize specific antigens in the brain rather than passively diffuse through a disrupted aged blood-brain-barrier (Montagne et al., 2015). These T cells may be attracted to the old neurogenic niche by a combination of brain-specific antigens and chemokines. An exciting possibility is that they may recognize neo-antigens derived from aggregated proteins in old NSCs (Leeman et al., 2018). This is the case in both Parkinson’s disease and multiple sclerosis, where T cells recognize α-synuclein (Sulzer et al., 2017) and β-synuclein (Lodygin et al., 2019), respectively.

Infiltrating T cells are the main cellular source of IFNγ in the SVZ niche (Dulken et al., 2019). IFNγ is known to act as an antiproliferative agent in the adult neurogenic niche suggesting that the accumulation of T cells might contribute to age-related neurogenic decline (Li et al., 2010b; Pereira et al., 2015). Indeed, interferon response is associated with a decline in NSC proliferation in the SVZ (Dulken et al., 2019). Interestingly, inhibition of IFN signaling (using IFNα and IFNγ receptor knock-out mice) results in higher proliferating NSC numbers in the SVZ of old animals (Kalamakis et al., 2019). Similarly, inhibition of the chemokine CXCL10, which is released in response to IFNγ, decreases the number of quiescent NSCs and increases the production of neuroblasts in old SVZ neurogenic niches, providing a causal link between inflammation and the inability for NSCs to activate in the old brain (Kalamakis et al., 2019). Thus, while infiltrating T cells may have beneficial effects during development (hippocampal neurogenesis and spatial learning abilities are significantly decreased in T-cell deficient mice in early adulthood (Wolf et al., 2009; Ziv et al., 2006)), they may have a detrimental effect on NSCs during aging in part by secreting inflammatory cytokines.

It will be interesting to determine the relationship between infiltrating T cells and NSCs during development and aging, and the potential shift from a protective role during development to a detrimental, possibly cytotoxic, effect of T cells in the inflamed old brain. It will also be important to explore how the different neurogenic niches respond to immune cell infiltration and inflammatory signals. For example, the proinflammatory pathway IKKβ/NF-κB impairs survival, proliferation, and differentiation of adult NSCs in the hypothalamic neurogenic niche, and results in obesity and pre-diabetes in mice (Li et al., 2012). Thus, blocking T cells or reducing inflammation levels in old neurogenic niches could be an effective strategy to restore old brain function.

Meningeal lymphatic vessels and the glymphatic system

The brain contains dorsal and basal lymphatic vessels in the meninges that assist in the drainage of CSF components and meningeal immune cells into the cervical lymph nodes (Ahn et al., 2019; Da Mesquita et al., 2018; Louveau et al., 2015). In addition, the ‘glymphatic’ system – a glial-associated lymphatic system (Aspelund et al., 2015) – is likely connected to lymphatic vessels (Louveau et al., 2017), and has been shown to be involved in β-amyloid clearance (Xie et al., 2013). Meningeal lymphatic vessels and the glymphatic system provide a new potential link between the CNS and the peripheral immune system. Interestingly, the meningeal lymphatic system also declines functionally during aging (Ahn et al., 2019; Da Mesquita et al., 2018), and disrupting meningeal lymphatic vessels with a photodynamic drug (visuadyne) results in cognitive impairments in young mice (Da Mesquita et al., 2018). Although enhancing meningeal lymphatics improves aspects of learning and memory in old mice, it does not increase the number of dividing NSCs in the hippocampus of these mice (Da Mesquita et al., 2018). Nonetheless, it will be exciting to determine how meningeal lymphatic vessels and the glymphatic system interact and integrate with the neurogenic niche to influence the function of NSCs and other niche cells during aging.

Ependymal cells

Ependymal cells form a layer of epithelial multiciliated cells that line the brain’s ventricular walls. The coordinated beating of ependymal cilia contributes to CSF dynamics, which is crucial for exposure of NSCs to trophic factors and metabolites and for clearance of waste and toxins from the brain (Spassky and Meunier, 2017). While quiescent NSCs and ependymal cells both derive from a common radial glial cell progenitor during development (Ortiz-Alvarez et al., 2019; Spassky et al., 2005) and share phenotypic markers (SOX2, SOX9, NESTIN, and CD133), these two cell types have different morphologies and functions in the adult brain. Ependymal cells isolated by FACS using α-smooth muscle actin (SMA) as a marker are transcriptionally different from NSCs and do not proliferate in vitro or in vivo under pro-growth conditions that normally induce NSCs proliferation (Shah et al., 2018). Single cell RNA-seq confirmed that ependymal cells do not have neural stem cell function (Shah et al., 2018). However, the modulation of CSF dynamics by ependymal cell cilia can impact NSC function and neurogenesis, and this could be critical during aging. For example, the beating of ependymal cilia is required for the directional migration of neuroblasts toward the olfactory bulb in mice (Sawamoto et al., 2006). During aging, ependymal cells accumulate intermediate filaments, dense bodies and lipid droplets, and exhibit fewer cilia (Capilla-Gonzalez et al., 2014). Furthermore, ependymal cell cilia tufts are more tangled in old mice, resulting in extended ventricular areas completely devoid of cilia (Capilla-Gonzalez et al., 2014). These age-related changes in ependymal cells could contribute to neurogenic decline and negatively affect neuroblast migration, although this has not been tested directly. Thus, uncovering exactly how ependymal cells relay signals from the ventricle and choroid plexus to the SVZ niche, and the changes they undergo during aging, may help identify strategies to counteract NSC and brain function decline.

Neuronal inputs

The neurogenic niche is embedded in and interacts with neural networks in the brain. In the hippocampus, NSC quiescence is maintained by tonic GABAergic innervations from parvalbumin-positive interneurons (Song et al., 2012), which are themselves regulated by long-range projections from the medial septum (Bao et al., 2017). Hippocampal mossy cells, which are glutamatergic neurons innervating both mature granule cells and local interneurons (Scharfman, 1995), also contribute to this circuitry by promoting NSC quiescence via local interneurons and NSC activation through direct glutamatergic innervation of NSCs (Yeh et al., 2018). Likewise, the anterior ventral SVZ is innervated by proopiomelanocortin (POMC) neurons from the hypothalamus that promote NSC proliferation and deep granule neuron generation (Paul et al., 2017). Interestingly, hunger and satiety states could affect NSC proliferation in a POMC neuron dependent manner (Paul et al., 2017). Although all these studies were carried out in young animals, they suggest that modulating the neural circuitry of the old NSC niche, either by electrical stimuli or indirect input from feeding via the hypothalamus, may rescue NSC function and neurogenesis.

The newborn neurons generated from the SVZ and hippocampal niches have to functionally integrate into the existing neural circuitry to survive. In the olfactory bulb, new neurons integrate into the granule cell and glomerular layers, where they function as interneurons (Whitman and Greer, 2007). Long-term in vivo imaging showed that newborn neurons do not compete for integration in the olfactory bulb, but are non-selectively added (Platel et al., 2019). In the hippocampus, immature newborn neurons migrate to the granule cell layer of the DG and differentiate into dentate granule cells. Freshly integrated neurons provide new substrates for learning and might facilitate the formation of new memories. Excitingly, favoring adult-born dentate granule cell integration by transient overexpression of Kruppel-like factor 9 (KLF9), a negative transcriptional regulator of dendritic spines in mature granule cells, can rejuvenate the DG with new neurons and improve cognitive function in old mice (McAvoy et al., 2016). However, it remains to be determined whether newborn neurons replace older, less functional neurons in the brain or whether they replace other neurons, especially in the DG (Murray et al., 2020). In addition, as new cells integrate, pre-existing circuits could also be altered, which could make established memories harder to access at later time points (Frankland et al., 2013; Gao et al., 2018). Thus, ensuring that newborn neurons integrate into the correct circuits will be equally essential to rejuvenate old brain function.

Similarities and differences between neurogenic niches

While many molecular mechanisms governing the regulation of NSC populations are shared between the SVZ and the hippocampus niches, some signaling pathways have antagonistic effects on NSC regulation depending on the neurogenic region. For example, IFN signaling early in life seems to be important to maintain neurogenesis in the hippocampus (Baruch et al., 2014), whereas over-activation of this pathway during aging negatively impacts the neurogenic niche in the SVZ (Dulken et al., 2019). Similarly, reducing p38 MAPK signaling (an important signaling pathway that coordinates inflammation, proliferation, and apoptosis) attenuates the age-dependent decline in neurogenesis in the hippocampus (Cortez et al., 2017), while inhibiting p38 signaling in cultured NSCs isolated from the SVZ of old mice increases their neurogenic potential (Moreno-Cugnon et al., 2020). Furthermore, NSCs from the hippocampus and SVZ differ in their long-term turnover dynamics. Hippocampal NSCs largely undergo asymmetric division resulting in the generation of neurons and decline in NSCs (Encinas et al., 2011), while NSCs in the SVZ undergo both symmetric self-renewing and consuming divisions allowing for simultaneous self-renewal and neuronal production later in life (Calzolari et al., 2015; Obernier et al., 2018). A more complete understanding of the similarities and differences between SVZ and hippocampal niches and their age-dependent regulatory mechanisms would provide new insights into generalized vs. specific strategies to improve regeneration in old brains.

NSCs and diseases: link with aging and translational potential

Because of their regenerative abilities, therapeutic interventions involving NSCs have the potential to improve pathological brain phenotypes. Consequently, how NSCs are affected by disease states and how they might be harnessed to improve pathological states is an exciting current field of study.

Neurodegenerative diseases

Aging is accompanied by both a decline in neurogenesis and an increased risk of neurodegenerative diseases, including Alzheimer’s and Parkinson’s diseases. A recent study showed that Alzheimer’s disease progression in humans is accompanied by a decline in the number and maturation level of newborn neurons in the hippocampus, which is more precipitous than the decline during non-pathological aging (Moreno-Jimenez et al., 2019). This is consistent with the notion that reduced neurogenesis may partially underlie aspects of Alzheimer’s disease pathology. Thus, the capacity of NSCs to generate functional newborn neurons has sparked interest in either transplanting NSCs or activating pre-existing NSCs in diseased brains for therapeutic purposes. Transplantation of human NSCs near the hippocampus in the APP/PS1 Alzheimer’s mouse model can reduce amyloid plaque load and improve hippocampal-dependent cognition (McGinley et al., 2018). In an Alzheimer’s mouse model, transplantation of fetal murine NSCs improves cognition, reduces amyloid processing, increases anti-inflammatory cytokine secretion, and restores synaptic impairment (Kim et al., 2015). However, this transplantation therapy is only effective when administered prior to advanced disease progression, indicating that the timing of stem cell transplants is essential for delivering therapeutic benefit (Kim et al., 2015). Similar proof-of-concept transplantation studies have been performed for other neurodegenerative diseases such as Parkinson’s Disease (PD), Amyotrophic Lateral Sclerosis (ALS), Huntington’s Disease (HD), and Multiple Sclerosis (MS) (reviewed in (Reekmans et al., 2012; Tang et al., 2017). A potentially more efficient strategy could be to identify ways to stimulate resident NSCs to improve neurodegenerative pathologies. Simultaneously increasing levels of hippocampal neurogenesis and elevating BDNF levels in an Alzheimer’s mouse model is sufficient to improve cognition, although increasing neurogenesis alone does not confer cognitive benefits (Choi et al., 2018). It is also important to note that NSC-derived newborn neurons are functionally different from those that die during disease progression. Thus, further research is needed to assess whether the observed beneficial effects of NSC transplantation or stimulation are a consequence of the integration of newborn neurons into the neuronal circuits damaged by disease (e.g. dopaminergic neurons in PD) or whether they result from the remodeling of the existing circuitry (i.e. by production of specific factors).

Stroke

The regenerative potential of NSCs is activated during stroke, when regions of the brain are deprived of oxygen (ischemia) due to arterial occlusion or rupturing. Following ischemic injury, a common stroke model, resident NSCs receive cues to activate and differentiate into glial cells which form a ‘glial scar’ (Adams and Gallo, 2018) and neurons which migrate to the site of injury and can improve neuropathology in rodents (Kernie and Parent, 2010) and humans (Jin et al., 2006). However, the regenerative potential of NSCs drastically declines with age, limiting their ability to repair injury following stroke. Transplantation of exogenous NSCs has the potential to improve recovery from ischemic injury in elderly stroke patients. Injection of human NSCs (derived from fetal brain tissue or iPSCs) into the ipsilesional hippocampus of a mouse model of stroke, results in infarct volume reduction, decreased expression of pro-inflammatory factors, and improvement of stroke-induced behavioral deficits (Eckert et al., 2015; Huang et al., 2014). Co-administration of human NSCs (derived from fetal brain tissue) with a small molecule neuroprotectant (3K3A-activated protein C) in mice with ischemia following arterial occlusion results in increased neuronal production, promotion of synaptic circuit repair, and functional improvement of post-ischemic recovery (Wang et al., 2016). Significant interest is also being given to stimulating endogenous NSCs instead of transplanting them following ischemic injury. For example, treating rats post-stroke with cerebrolysin, a mixture of neurotrophic peptides, increases SVZ neurogenesis and oligodendrogenesis, and improves neurological function (Zhang et al., 2013). Similarly, injections of GDF11 after stroke increase NPC numbers in the SVZ and improve neuronal regeneration in mice (Lu et al., 2018). However, sustained activation of NSCs after injury can also lead to stem cell exhaustion, depleting the brain’s lifelong regenerative reservoir. Thus, finding a balance between NSC activation and exhaustion will be essential. There is additionally concern that transplantation of exogenous NSCs could have deleterious effects due to improper functional integration of newborn neurons. Indeed, post-stroke hippocampal neurogenesis can be detrimental to contextual and spatial memory performance, and newborn neurons can improperly integrate into the pre-existing hippocampal circuitry (Cuartero et al., 2019). In this context, inhibiting hippocampal neurogenesis (using either temozolomide or with a genetic model (Nestin-Cre ERT2/NSE-DTA)) actually improves memory retention (Cuartero et al., 2019). These results highlight the need to better understand neuronal circuit integration and synaptic remodeling to harness the regenerative potential of NSCs for stroke patients.

Cancer

The high proliferative capacity of NSCs increases their risk for mutagenesis and makes NSCs more prone to become precursors to certain brain cancers such as gliomas and glioblastomas. Indeed, glioblastoma-initiating cells share remarkable similarities with non-pathological NSCs (comprehensively covered in other reviews such as (Goffart et al., 2013)). For example, acyl-CoA-binding protein (ACBP), a proliferative factor expressed exclusively in astrocytes, NSCs, and NPCs, is highly expressed within glioblastoma cells and drives tumorigenesis (Duman et al., 2019). Furthermore, deep sequencing of human glioblastoma samples showed that NSCs in the SVZ contained low-level driver mutations that matched the mutational landscape of the corresponding primary glioblastoma, suggesting that NSCs could be the cell of origin for human glioblastoma (Lee et al., 2018). Thus, understanding the processes underlying NSC activation and proliferation could help uncover the molecular triggers of tumorigenesis and provide new treatments for brain cancers. Furthermore, it will help ensure the safety of translational NSC therapies, such as transplantation and endogenous activation, by preventing potential glioblastoma formation.

Interventions to rejuvenate old neurogenic niches

While NSC levels and neurogenic potential decline during aging with detrimental functional consequences, this decline is not inexorable. As described above, targeting specific molecular pathways and metabolic organelles in NSCs, such as TFEB/lysosomes (Leeman et al., 2018), transcription factors (Carrasco-Garcia et al., 2015), and mitochondria (Beckervordersandforth et al., 2017) can improve NSC function in the old. Despite the promise of such molecular manipulation, an interesting parallel avenue of research involves using lifestyle interventions to improve NSC and brain function. Thus far, three interventions have shown the greatest potential for rejuvenating old neurogenic niches: diet, exercise, and systemic blood factors (Figure 4).

Figure 4. Interventions to rejuvenate old neurogenic niches.

Comparison of three promising interventions to rejuvenate old neurogenic niches – the subventricular zone (SVZ) and hippocampus dentate gyrus (DG). Dietary interventions are divided into three sub-categories (caloric restriction (CR), intermittent fasting (IF), and fasting-mimicking diets (FMD)). Exercise is divided into moderate and strenuous groupings. Systemic factors are divided between those in blood or plasma. The neuroprotective effects of these interventions were assessed in various diseases and injuries: Alzheimer’s Disease (AD), Parkinson’s Disease (PD), Multiple Sclerosis (MS), Traumatic Brain Injury (TBI), Stroke, and Epilepsy. A question mark (“?”) signifies conflicting or insufficient evidence to support the claim.

Dietary interventions

Dietary interventions, such as dietary restriction and intermittent fasting, are arguably the most well-established longevity interventions to date and have been shown to confer multiple health benefits and extend lifespan in mice (Mitchell et al., 2019). Studies have begun to investigate whether dietary interventions can protect against the age-dependent neurogenic decline or even have a rejuvenating effect on neurogenesis. For example, a 40% reduction in caloric intake in mice from young adulthood (4 months old) onwards protects the SVZ from the age-dependent decline in neurogenesis and increases olfactory memory performance (Apple et al., 2019). Under basal conditions, three months of intermittent fasting can increase the proliferation and survival of newborn neurons in the DG and SVZ of middle-aged mice (Manzanero et al., 2014). Intermittent fasting also limits the extent of cell death associated with ischemic injury and results in physiological improvements such as diminished sensorimotor deficits (Manzanero et al., 2014). However, intermittent fasting fails to boost the generation of newborn neurons in the context of ischemic injury, suggesting that the observed neuroprotective effects are a result of reduced neuronal cell death rather than increased neurogenesis (Manzanero et al., 2014). Other intermittent fasting regimes, such as alternate-day feeding, are accompanied by a thicker pyramidal cell layer in the DG and result in cognitive improvements, such as improved learning and memory (Li et al., 2013), suggesting a possible link between fasting and neurogenesis. As long-term dietary restriction is difficult to adhere to, other dietary interventions, such as the fasting-mimicking diet (FMD), have been developed to recapitulate the health benefits of caloric restriction while improving regime adherence and compliance (Brandhorst et al., 2015). Four days on an FMD diet twice per month from 16 months onwards increases the mean lifespan of mice by 11.3% and ameliorates performance on cognitive tests evaluating short-term memory, long-term memory, and learning (Brandhorst et al., 2015). This was also accompanied by an increase in hippocampal neurogenesis (as measured by BrdU incorporation and DCX staining in 23-month old animals) (Brandhorst et al., 2015). Thus, dietary interventions have the potential to improve neurogenesis and confer neuroprotective effects upon insult or injury, as well as provide overall health and longevity for the organism.

The molecular mechanisms underlying the benefits of dietary interventions on NSCs remain to be determined, but nutrient sensing pathways, including mTOR and insulin-IGF signaling, likely play a key role given their importance in regulating cellular metabolism and maintaining NSC homeostasis. It will be crucial to determine whether the benefits of these interventions are due to the fasting period, when processes such as autophagy are activated, or whether they are due to an overall reduction in the intake of specific ‘damaging’ nutrients such as glucose or free oleic acid. Specific nutrients may also play an important role in preserving or rejuvenating NSCs. For example, both dietary restriction (DR) and ad libitum low-protein, high-carbohydrate (LPHC) diets positively affect hippocampal biology and cognition in old mice (Wahl et al., 2018). Similarly, a ketogenic diet can recapitulate the systemic longevity and healthspan benefits of DR regimes, and even improve memory in aged animals (though the specific effects on the neurogenic niches were not investigated in these studies) (Newman et al., 2017; Roberts et al., 2017). Thus, modulating the levels of certain nutrients in the diet may be as effective as the drastic removal of food in improving NSC function and overall brain health.

Exercise

Exercise is beneficial to combat age-dependent cognitive decline, as measured by scores in short and long-term memory as well as executive function (reviewed by (Hotting and Roder, 2013)). Voluntary aerobic exercise increases adult hippocampal neurogenesis in mice (van Praag et al., 1999; van Praag et al., 2005) and rats (Nokia et al., 2016), and promotes the development and integration of newborn granule cell neurons in the aging hippocampus (Trinchero et al., 2017). However, not all forms of exercise have an equal impact on improving brain health. The benefits of exercise for adult hippocampal neurogenesis in rats appear to be greater when the animals perform voluntary endurance training compared to resistance training (Nokia et al., 2016). Furthermore, while both moderate and fatiguing exercise increase neuronal differentiation and migration in the hippocampus of young adult mice (2 months old), only moderate exercise improves cell proliferation and survival (So et al., 2017). Moderate exercise also results in a greater increase in hippocampal neurogenesis compared to fatiguing exercise accompanied by improved cognitive benefits, such as enhanced spatial discrimination, that were absent in the fatigued group (So et al., 2017). Thus, moderate aerobic exercise may have the most rejuvenating effects in neurogenic niches when compared to other forms of exercise.

Excitingly, exercise can also be a neuroprotective intervention in the context of certain neurodegenerative diseases such as Parkinson’s disease (Alvarez-Saavedra et al., 2016) and Alzheimer’s disease (Choi et al., 2018). In transgenic mice with an ataxic phenotype to model Parkinson’s disease, voluntary running improves lifespan and ameliorates motor dysfunction. These physiological improvements were accompanied by an increase in cerebellar oligodendrogenesis and subsequent de novo myelination mediated by VGF secretion (Alvarez-Saavedra et al., 2016). In an Alzheimer’s mouse model, exercise increases hippocampal neurogenesis, improves cognition, and reduces amyloid-beta load (Choi et al., 2018). These beneficial effects of exercise can be recapitulated when neurogenesis and BDNF levels are boosted together suggesting that exercise increases the levels of both systemic factors (such as BDNF) and neurogenesis to ameliorate neurodegenerative disease pathology (Choi et al., 2018).

While the molecular mechanisms underlying exercise-induced benefits remain largely unknown, particularly in old animals, recent evidence suggests that hippocampal neurogenesis and spatial memory improvements induced by running are dependent on the secretion of Cathepsin B, a proteolytic enzyme known to be expressed in humans and monkeys upon exercise (Moon et al., 2016). Systemic factors such as activated platelets have also been implicated as mediators of the neurogenic benefits of exercise (Kratz et al., 2006; Leiter et al., 2019). Exercise-induced activated platelets promote NSC proliferation in the hippocampus of young adult mice (8–10 weeks old) and increase subsequent neurogenesis in vivo. This NSC boost is likely mediated by Platelet Factor 4, which is sufficient to increase levels of DCX+ immature neurons in the dentate gyrus (Leiter et al., 2019). It will be interesting to determine how Platelet Factor 4 affects old animals. While acute exercise leads to increased inflammation in response to muscle damage, regular aerobic exercise exerts its beneficial effects at least in part by reducing systemic inflammation reflected in lower levels of inflammatory cytokines such as TNFα and IFNγ (Woods et al., 2012). The beneficial neurogenic effects of running in the hippocampus of both young (8 weeks) and middle-aged (8 months) mice are inversely correlated with the density of microglia, cells that normally contribute to age-related neuroinflammation (Gebara et al., 2013). Given the inflamed state of the old neurogenic niche, it is tantalizing to speculate that the positive effects of exercise on neurogenesis in old individuals may also result from a decrease in the inflammatory state of the cells and milieu in the old niche.

Blood factors

Another promising regenerative strategy involves administering young blood, or the factors therein, to an aged animal to improve declining neurogenesis and cognitive performance. As described above, impaired neurogenesis and cognitive function in the aged hippocampus and SVZ can be improved by young blood during heterochronic parabiosis (Katsimpardi et al., 2014; Villeda et al., 2014). Both dilution of factors in old blood and presence of factors in young blood may underlie the beneficial effects of heterochronic parabiosis. In fact, injection of young blood plasma is sufficient to improve cognitive performance in contextual fear conditioning as well as spatial learning and memory (Villeda et al., 2014), although it is not clear whether these are due to improved neurogenesis or neuronal function per se. Despite promising results from small human clinical trials to assess the safety and feasibility of routine blood injections (Sha et al., 2019), the clinical challenges (and potential ethical concerns) of routinely administering young blood to elderly individuals has prompted ongoing work to identify the specific factors within young blood that are capable of independently conferring health benefits and increasing neurogenesis. More needs to be done to identify specific factors from young blood that could rejuvenate the old neurogenic niche and to determine the effector cells.

Given the neurogenic benefits of these different interventions, one intriguing hypothesis is that combinatorial application of two or more of these strategies could have additive or even synergistic beneficial effects (Fabel et al., 2009; Hutton et al., 2015). Additional ‘rejuvenating’ strategies, such as in vivo partial reprogramming (Han et al., 2020; Ocampo et al., 2016) or microbiome transfer (Barcena et al., 2019; Smith et al., 2017) could also be exploited to rejuvenate old NSCs and improve age-related sensorimotor decline. While further studies will be required to determine the specific mechanisms of action of each strategy, the potential synergistic benefits of dietary interventions, exercise, and blood factors hold promise as a strategy to rejuvenate the brain for both physiological and pathological aging.

Neurogenesis in the vertebrate world

Evolution of vertebrate neurogenesis

Adult neurogenesis has been observed in all vertebrate species studied, but there is a wide range of neurogenic capacities in different species. For example, amphibians and teleosts have many neurogenic regions in their brains and display a remarkable capacity for neurogenesis and brain regeneration (reviewed in (Alunni and Bally-Cuif, 2016)). In contrast, mammals, including humans, display a relatively low degree of neurogenic potential among vertebrates, with limited regenerative potential and fewer neurogenic niches, even though they are still capable of producing newborn neurons throughout their entire lifespans (reviewed in (Alunni and Bally-Cuif, 2016)). While the generation of newborn neurons is necessary during embryonic and juvenile development, it is not immediately obvious why evolutionary pressures would have selected for the ongoing generation of newborn neurons in the adult brain, especially given the associated increased risk of cancer formation. In fact, it was long assumed that adult neurogenesis was an evolutionary relic without functional benefits (reviewed in (Kempermann, 2015)). However, it seems increasingly likely that adult neurogenesis is under evolutionary positive selection, not only for its regenerative benefits in the event of brain injury but also to increase brain plasticity and improve an organism’s ability to integrate novel information throughout life (reviewed in (Konefal et al., 2013)). This is perhaps best exemplified by the songbird brain, an important and historic model for the study of adult neurogenesis (Goldman and Nottebohm, 1983; Nottebohm, 1985). Adult male starlings, among other species, exhibit dramatic seasonal plasticity in their ability to learn new songs, a mating behavior integral to the species’ sexual selection (De Groof et al., 2009). Adult neurogenesis may also play an important role in rodent mating behaviors as the generation of olfactory bulb neurons from SVZ NSCs might contribute to mate selection via olfactory cues. In fact, disrupting olfactory bulb neurogenesis in female mice using focal irradiation of the SVZ results in abnormal social interactions with males but not females, perhaps due to a reduction in the ability to detect male-specific odors resulting in impaired mate selection (Feierstein et al., 2010).

Human neurogenesis: controversy, challenges and way forward

While the human brain was long regarded as an entirely post-mitotic organ, over two decades ago a seminal study showed dividing cells in the brains of cancer patients (Eriksson et al., 1998). Since then, many studies have tried to address this question, typically relying on carbon dating (Spalding et al., 2013) and immunostaining techniques (Knoth et al., 2010) in postmortem human brains. NSCs and neuroblasts have been observed lining the walls of the lateral ventricle in humans. However, while new neurons born from the SVZ are added to the olfactory bulb in mice and other mammals (Lois and Alvarez-Buylla, 1994), the generation of new neurons in the olfactory bulb is practically negligible in humans (Bergmann et al., 2012; Sanai et al., 2011; Wang et al., 2011). Instead, new neurons born from the human SVZ integrate in the adult human striatum, adjacent to the lateral ventricle wall (Ernst et al., 2014). Whether these newborn neurons derive from bona fide NSCs in the SVZ, and whether they have a functional role, has not been tested directly. Thus, more work will be needed to better understand the importance of neurogenesis in the adult SVZ in humans, and how it is impacted by aging.

A key question, which has been subject to much debate over the years, is whether neurogenesis takes place in the human DG of the hippocampus, a region known to be important for learning and memory (reviewed by (Ernst and Frisen, 2015; Kempermann et al., 2018; Paredes et al., 2018; Tartt et al., 2018)). While some studies have reported a lack of evidence for neurogenesis in humans after adolescence (Cipriani et al., 2018; Sorrells et al., 2018), others have found the opposite - persistent human neurogenesis in adulthood with a small decline during aging (Boldrini et al., 2018; Moreno-Jimenez et al., 2019; Tobin et al., 2019). The number of neuroblasts also correlated with cognitive status, with higher numbers present in individuals with better cognitive performance (Tobin et al., 2019). Though correlative, these results suggest that neurogenesis may be linked to cognitive function in humans.

Additional work will be needed to reconcile opposing outcomes from different studies (Flor-Garcia et al., 2020). For example, comparing staining protocols, preservation of samples from human donors, and post-mortem collection times will help address these discrepancies. Importantly, advances in single-cell RNA-seq technologies should help move the field forward by providing more specific cell markers for immunological identification of NSCs as well as transcriptomic signatures that could be compared across species. Indeed, a recent single-cell RNA-seq dataset from adult human olfactory neuroepithelium revealed cells with characteristics of NSCs and neural progenitors (Durante et al., 2020). It will be interesting to assess neurogenesis between basal conditions (as done in most of these studies) and conditions where neurogenesis should be elevated (e.g. after exercise or even brain injury in humans). Together, analyses in humans could help provide novel avenues to specifically boost neurogenesis and improve old brain function in humans.

Conclusion and frontiers in the aging neural stem cell field

NSCs have the potential to rejuvenate old brains and ameliorate age-related neurodegenerative diseases. In the past few years, intrinsic and extrinsic changes that occur in NSCs and their niches during aging have been uncovered. In parallel, interventions that improve health and/or lifespan (e.g. dietary restriction, exercise, blood factors, etc.) have been shown to improve neurogenesis in old individuals. Combining the knowledge gained from these two areas should accelerate the development of specific and targeted interventions to recover brain function in aged individuals. The advent and refinement of new technologies (e.g. single-cell analyses, multi-’omics’, genome-wide screens, and innovative imaging) should further improve our understanding of how aging influences NSCs and their niche, and how NSCs interact with brain circuity and systemic/organismal signals (Figure 5).

Figure 5. Frontiers in the aging neural stem cell field.

The advent and refinement of new technologies (right) should further improve our understanding of how aging influences NSCs and their niche, and how NSCs interact with brain circuity and systemic and organismal signals.

Multiplexing of next-generation sequencing techniques at the single-cell level, such as single-cell RNA-seq, single-cell ATAC-seq or single-cell proteomics (Stuart and Satija, 2019), will capture a snapshot of the genomic, epigenomic, proteomic, and metabolomic landscape of individual cells in a single assay. The resolution of these techniques should help refine our understanding of intermediate stages in NSC activation and differentiation in vivo. A next frontier will be to integrate all this molecular and cellular information in situ to provide both spatial and “omics” resolution of the niche using technologies such as STARmap (Wang et al., 2018) or Slide-seq (Rodriques et al., 2019).

Visualizing NSC activation and differentiation in situ will provide valuable insights into the dynamics of neurogenic niches during aging. Intravital imaging can be performed to visualize the fate of NSCs in situ, including their symmetric vs. asymmetric divisions and the integration of newborn neurons into pre-existing circuits (Breton-Provencher et al., 2016; Pilz et al., 2018; Wang et al., 2019). Lineage tracing is another powerful tool that can be leveraged to understand the cell fate decisions of NSCs, notably the extent of self-renewal vs. differentiation (Encinas et al., 2011; Obernier et al., 2018). Advanced lineage tracing tools with CRISPR-Cas9 or Cre recombination (Hwang et al., 2019; Marx, 2018; Spanjaard et al., 2018) will help reveal how the fate of cells in the NSC lineage changes over time during normal aging and disease, and in response to interventions that boost neurogenesis. These advances in lineage tracing and imaging will help address outstanding questions in the field, such as to what degree individual NSCs persist and continue to produce newborn neurons, and the extent and functional importance of neuronal turnover in the aging brain.

High-throughput functional genetic screening using different flavors of the CRISPR technology could uncover new molecular mechanisms of NSC aging. In addition, mechanistic studies of NSCs during aging could be enhanced by the ability to directly reprogram fibroblasts from old healthy or diseased individuals into NSCs (Ring et al., 2012). Finally, adapting the human brain organoid system (Lancaster et al., 2013) to study brain aging and related diseases could allow for a deeper investigation of the neurogenic niche and how it becomes altered with time.

Many mechanisms of aging in the neurogenic niche remain mysterious. For example, the interaction between the neurogenic niche and the surrounding neural networks in the brain, and especially how this is affected with age, is still largely unknown. Tissue optical clearing techniques (Wan et al., 2018) paired with super-resolution imaging could help uncover the neuronal inputs and outputs from the NSC niche. Optogenetics or related methods (Rajasethupathy et al., 2016) could also be exploited to modulate neuronal activity and determine the impact on the niche. In humans, initiatives such as the mapping of the human connectome may help to uncover the interactions between the neural circuitry and the NSC niche and how they are perturbed with age and disease. Finally, it will be crucial to further investigate the crosstalk between neurogenic niches and organism-wide systemic aging. For example, understanding the systemic effects of circadian rhythms (Malik et al., 2015), sleep (Kumar et al., 2020), reproductive endocrine status (Oboti et al., 2015), or microbiome composition (Ogbonnaya et al., 2015) on neurogenic function during aging could be critical to understand long-range cues that could indirectly affect NSC regeneration over time.

At the population level, inter-individual variability is emerging as a defining aspect of old age, which could reflect differences in aging trajectories and brain function (Li et al., 2017). Individuals are exposed to different environments and behavioral activities that continually alter the brain throughout life (Kempermann, 2019). Interestingly, differences in spatial memory function between individual old rats are positively correlated with newborn neuron numbers in the hippocampus (Drapeau et al., 2003). Likewise, inbred mice living together in one large enriched environment develop increasing individual differences in exploratory behavior which correlate with differences in hippocampal neurogenesis (Freund et al., 2013, 2015; Korholz et al., 2018). Understanding the sources and consequences of these differences between individuals could provide new strategies to rejuvenate the old neurogenic niche and lead to personalized and precision medical therapies.

Together, these technological and conceptual advances provide an unprecedented opportunity to develop a better understanding of the molecular mechanisms regulating NSC aging with the ultimate goal of developing novel regenerative therapeutics. As the global demographic shifts towards becoming increasingly geriatric, longevity strategies that promote healthy brain function rather than just extend lifespan will become crucial to combat the devasting consequences of brain aging.