Manipulation of EGFR-Induced Signaling for the Recruitment of Quiescent Neural Stem Cells in the Adult Mouse Forebrain

The ventricular-subventricular zone (V-SVZ) is the principal neurogenic niche in the adult mammalian forebrain. Neural stem/progenitor cell (NSPC) activity within the V-SVZ is controlled by numerous of extrinsic factors, whose downstream effects on NSPC proliferation, survival and differentiation are transduced via a limited number of intracellular signaling pathways. Here, we investigated the relationship between age-related changes in NSPC output and activity of signaling pathways downstream of the epidermal growth factor receptor (EGFR), a major regulator of NSPC activity. Biochemical experiments indicated that age-related decline of NSPC activity in vivo is accompanied by selective deficits amongst various EGFR-induced signal pathways within the V-SVZ niche. Pharmacological loss-of-function signaling experiments with cultured NSPCs revealed both overlap and selectivity in the biological functions modulated by the EGFR-induced PI3K/AKT, MEK/ERK and mTOR signaling modules. Specifically, while all three modules promoted EGFR-mediated NSPC proliferation, only mTOR contributed to NSPC survival and only MEK/ERK repressed NSPC differentiation. Using a gain-of-function in vivo genetic approach, we electroporated a constitutively active EGFR construct into a subpopulation of quiescent, EGFR-negative neural stem cells (qNSCs); this ectopic activation of EGFR signaling enabled qNSCs to divide in 3-month-old early adult mice, but not in mice at middle-age or carrying familial Alzheimer disease mutations. Thus, (i) individual EGFR-induced signaling pathways have dissociable effects on NSPC proliferation, survival, and differentiation, (ii) activation of EGFR signaling is sufficient to stimulate qNSC cell cycle entry during early adulthood, and (iii) the proliferative effects of EGFR-induced signaling are dominantly overridden by anti-proliferative signals associated with aging and Alzheimer’s disease.

EGF-induced NSPC proliferation in vitro and in vivo, and mTOR signaling was necessary for EGF administration to rescue NSPC proliferation in the aging brain (Paliouras et al., 2012;Hartman et al., 2013). However, whether individual downstream branches of EGFR-induced signaling pathways mediate specific aspects of adult NSPC survival, proliferation, and/or differentiation remains unknown.
In the present study, we investigated the contributions and potential of EGFR-triggered signaling pathways for the control of adult NSPC activity in the adult and aging brain. Specifically, we hypothesized that EGFR signaling represents a potential avenue for enhancing NSPC activity in the context of the declining neurogenesis that occurs during normal aging (Eriksson et al., 1998;Curtis et al., 2003;Luo et al., 2006;Shook et al., 2012;Ernst and Frisen, 2015;Daynac et al., 2016) and that is further accelerated in Alzheimer's disease (AD) (Ziabreva et al., 2006;Hamilton et al., 2010Hamilton et al., , 2015Moreno-Jimenez et al., 2019;Scopa et al., 2020). First, we used in vitro pharmacological approaches to better understand how the EGFR-triggered AKT, ERK and mTOR signaling cascades coordinate the processes of NSPC survival, proliferation and/or differentiation. We then used an in vivo electroporation approach to ectopically express and activate EGFR signaling in a subpopulation of ventriclecontacting qNSCs (Joppe et al., 2020) in models of the early adult, middle-aged, and Alzheimer V-SVZ niche. Altogether, this work provides us with a clearer understanding of the potential for manipulating EGFR-induced signaling pathways for enhancing V-SVZ neurogenesis.
Although V-SVZ neurogenesis declined markedly in vivo over these timepoints, the intrinsic capacity for NSPCs to activate EGFR-associated signaling pathways remained unchanged. NSPCs cultures were generated from 2-and 11month-old V-SVZs using EGF-dependent neurosphere cultures ( Figure 1D). Analysis of the neurosphere lysates by Western blotting showed equal expression of proliferating cell nuclear antigen (PCNA) and no significant differences in activation levels of multiple downstream components of EGF-induced signaling pathways, including pAKT, pERK, phospho(p)-mTOR, pS6 or p4EBP1 ( Figure 1E). Hence, in the presence of exogenous mitogen, NSPCs from the young adult and middle-aged V-SVZ have the same intrinsic capacity to activate EGF-induced AKT, ERK, and mTOR signaling pathways.
Thus, there is a sharp drop in V-SVZ neurogenesis during the early adulthood period that is accompanied by selective decreases in certain EGF-induced signaling pathways in vivo.

NSPC Activity Is Differentially Regulated by EGF-Induced Signaling Modules
We performed pharmacological loss-of-function experiments to better understand how modulations of sub-branches of EGFR-induced signaling might impact NSPC proliferation, survival, and differentiation. For these experiments, we focused on two receptor-proximal pathways, the phosphatidylinositol-3-kinase/protein kinase B (PI3K/AKT) and mitogen-activated protein kinase/extracellular-regulated kinase (MAPK/ERK) cascades, and the more downstream mammalian target of rapamycin (mTOR) complexes, mTORC1 and mTORC2 (Figure 2A).
We performed TUNEL assays to determine whether the effects of these inhibitors on NSPC proliferation were secondary to cell death. In contrast to their roles in NSPC proliferation, the PI3K/AKT and MEK/ERK modules did not appear to be key mediators of NSPC survival: 1.2 ± 0.3% of NSPCs were TUNEL + when treated with DMSO control, and this was not statistically affected by treatment with LY294002, PD184352 or KU0063794 ( Figure 3F). In contrast, the proportion of TUNEL + cells increased to 9.1 ± 2.7% with the mTORC1 inhibitor Rapamycin (p = 0.01, one-way ANOVA), indicating that mTORC1 inhibition, but not PI3K or MEK inhibition, increases sensitivity to cell death.
NSPCs treated with the MEK inhibitor PD184352 showed strong increases in both the astrocyte marker GFAP and the neuronal marker βIII-tubulin (Figures 3G-I). This was not observed with LY290042, KU0063794 or Rapamycin, which instead reduced expression of βIII-tubulin, GFAP and the oligodendroglial marker CNPase (Figures 3G-J). We performed immunocytochemistry on NSPCs treated under the same conditions to test whether the changes observed by Western blotting reflected changes in the numbers of cells expressing these markers. Immunocytochemistry confirmed that NSPC cultures yielded tri-lineage differentiation into GFAP + astrocytes, βIII-tubulin + neurons, and CNPase + oligodendrocytes (Supplementary Figure 1A). Upon PD184352 treatment, quantifications revealed a twofold increase of GFAP + cells (DMSO, 35.56 ± 2.70% of Hoechst; PD184352, 73.46 ± 2,87% of Hoechst; n = 4 per condition) and fourfold increase of βIII tubulin + cells (DMSO, 14.93 ± 1.62% of Hoechst + cells; PD184352, 62.45 ± 9.59% of Hoechst + cells; n = 4 per condition), consistent with the Western blot observations (Supplementary Figures 1B,C). Interestingly, induction of βIII tubulin + cells in the PD184352 condition occurred mainly within the expanded population of GFAP + cells, highlighting transition from a GFAP + precursor at this early 3-day differentiation timepoint. Since NSPC proliferation was reduced by all inhibitors (Figures 3E,K) but differentiation markers were stimulated only by PD184352, only the MEK/ERK module acts as an EGF-induced repressor of NSPC differentiation. Conversely, activity of the PI3K/AKT and mTORC1 modules is necessary for appropriate NSPC differentiation.
These data indicate that the PI3K/AKT, MEK/ERK, and mTOR1/2 modules have both overlap and selectivity in the biological functions of NSPCs that they regulate: all three modules are implicated in NSPC proliferation, the mTORC1 module contributes to NSPC survival and the MEK/ERK module represses NSPC differentiation.

EGFR-CA Can Promote Neurogenic Activity of qNSCs in vivo
Recently, we showed that adult brain electroporation can be used to genetically target a subpopulation of ventricle-contacting, GFAP-expressing cells within the qNSC population (Joppe et al., 2020). In that study, lineage-tracing showed these electroporated precursors produced small numbers of olfactory bulb neurons in vivo and did so independently of the neurosphere-forming aNSC lineage. Here, we used this adult electroporation strategy to test whether this highly quiescent qNSC population can be recruited via activation of EGFR-induced signaling.
Since qNSCs are reportedly EGFR-negative, we first confirmed that the electroporated subpopulation of qNSCs is not responsive to exogenous EGF. When we implanted 3-day intraventricular osmotic pumps containing EGF, total EdU incorporation in the V-SVZ markedly increased ( Figure 4B) but the electroporated cells did not increase in number or show an increase in EdU incorporation (Figures 4C-E).
We then electroporated 3-month-old Rosa26-stop-EYFP reporter mice with hGFAP-cre plasmids that had been mixed with either an empty vector (EV) control or a constitutively active form of EGFR (EGFR-CA) ( Figure 4F). To assess the responses of the electroporated qNSCs, we performed fluorescent immunostaining and quantified YFP + cells that were positive for EGFR, pS6 and EdU. At 1 week postelectroporation, the electroporation of EGFR-CA led to a 35fold increase of electroporated qNSCs expressing the EGFR receptor (EV, 0.5 ± 0.2% of YFP + cells versus EGFR-CA, 18.1 ± 1.9% of YFP + cells; Mann-Whitney, p = 0.0079; n = 5 per group) (Figures 4G,H). Levels of pS6, which is a downstream readout of EGFR-induced mTOR signaling and a marker of quiescent stem cells that are "primed" for division (Rodgers et al., 2014), was likewise increased (EV, 2.1 ± 0.6% of YFP + cells versus EGFR-CA, 22.8 ± 1.4% of YFP + cells; Mann-Whitney, p = 0.0079; n = 5 per group) (Figures 4I,J). Notably, EdU incorporation was significantly increased upon EGFR-CA electroporation (EV, 2.9 ± 0.2% of YFP + cells versus EGFR-CA, 5.0 ± 0.3% of YFP + cells; Mann-Whitney, p = 0.0079; n = 5 per group) ( Figure 4K). It is highly unlikely these YFP + EdU + cells incorporated EdU during DNA repair in the absence of division, as control experiments with the DNA repair marker, p-γH2AX, showed that the vast majority of EdU + p-γH2AX + cells in the V-SVZ indeed express the proliferation marker PCNA (98.33 ± 0.65% of EdU + p-γH2AX + cells are PCNA +, n = 3) (Supplementary Figure 2). Interestingly, and despite only transient expression of the EGFR-CA plasmid, analysis of a cohort of electroporated mice at 4 weeks post-electroporation revealed trends for EGFR-CA to eventually decrease YFP + cells within the V-SVZ and increase YFP + cells in the OB (Figures 4L-P). In line with this, when we processed electroporated V-SVZs for neurosphere cultures, we found that EGFR-CA-induced activation of qNSCs did not lead to generation of labeled neurospheres (Figures 4Q,R); this is consistent with the electroporated qNSCs being separate from the self-renewing, neurosphere-forming aNSC lineage (Joppe et al., 2020).
These data reveal that genetic activation of EGFR signaling pathways in 3-month-old mice is sufficient to prime qNSCs and shift them into the dividing subpopulation, but these electroporated qNSCs remain distinct from neurosphere-forming aNSCs.
The Ability of EGFR-CA to Recruit qNSCs Is Overridden by the Contexts of Aging and Alzheimer's Disease Lastly, we asked whether EGFR-CA expression in electroporated qNSCs would enable them to overcome the anti-proliferative influences of aging and/or AD on the V-SVZ. When EGFR-CA was overexpressed in 3-versus 6-month-old mice (Figure 5A), electroporated qNSCs showed an equal ability to upregulate EGFR-CA-induced pS6 signaling (3 m EV, 22.8 ± 1.4% versus 6 m EV, 21.7 ± 1.9%, pS6 + YFP + /YFP + cells) (Figure 5B). Interestingly, however, when we examined EdU incorporation, we found that 6-month-old mice had a significantly lower baseline EdU incorporation (3 m EV, 2.9 ± 0.2% versus 6 m EV, 0.4 ± 0.2%, EdU + YFP + /YFP + cells, p < 0.0001,  Note that YFP + cells in 3-and 6-month-old mice equally upregulated EGFR-induced pS6 but their division significantly increased only in the 3-month-old mice [two way ANOVA: age factor, F(1,16) = 141.1, p < 0.0001; EGFR-CA factor, F(1,16) = 19.62, p < 0.0083; interaction, F(1,16) = 9.072, p = 0.0083]. In (D), the proportion of primed hGFAP + cells that incorporate EdU decreases between 3 and 6 months of age (p = 0.0317, Mann-Whitney test). N = 5 per condition and age. Three month data is from Figure 4. (E-H) Baseline and EGFR-CA-induced proliferation of GFAP + qNSCs in 3xTg-AD mice and their WT control strain. (E) Paradigm for adult brain electroporation with hGFAP-GFP plasmids, mixed with either EV or EGFR-CA plasmids, in WT or 3xTg mice. (F-H) 1-week post-electroporation, quantification of the percentage of YFP + recombined cells that are (F) primed (GFP + pS6+) or (G) dividing (GFP + EdU+). Note that GFP + cells in WT and 3xTg mice were equally primed by EGFR-CA but only showed a significant increase in EdU incorporation in the WT mice [two way ANOVA: strain factor, F(1,16) = 16.82, p = 0.0008; EGFR-CA factor, F(1,16) = 16.52, p < 0.0009; interaction, F(1,16) = 9.245, p = 0.0078]. In (H), the proportion of primed GFP + cells that incorporate EdU shows a tendency to decrease in the 3xTg model of Alzheimer's disease (p = 0.12, Mann-Whitney test). N = 5 per condition and age. On graphs, white bars represent EV electroporations, gray bars represent EGFR-CA electroporations. In (C,G) striped bars represent EdU + pS6+. Two-way ANOVA, * p ≤ 0.05, * * p ≤ 0.01, * * * p ≤ 0.001. Mann-Whitney test, * p = 0.0317. two-way ANOVA) and that, unlike their younger counterparts, the 6-month-old animals failed to show an increase in EdU incorporation in response to EGFR-CA ( Figure 5C). Triple colocalization of YFP, EdU, and pS6 confirmed that the increased cell cycle entry observed in 3-month-old mice was restricted to the pS6 + population ( Figure 5C) and showed that the propensity of pS6 + cells to incorporate EdU in response to EGFR-CA was decreased at 6 months of age ( Figure 5D). Thus, despite retaining the ability to overexpress EGFR and activate downstream signaling, the proportion of hGFAP + qNSCs that enters the cell cycle declines between 3 and 6 months of age and is not rescued by genetic activation of EGFR signaling.
We also analyzed the effects of EGFR-CA overexpression in the 3xTg transgenic mouse model of familial AD, which was developed by Oddo et al. (2003) and carries human familial AD mutations in Presenilin 1, Amyloid Precursor Protein, and Tau. Since 3xTg mice and their WT control strain (B6;129) do not permit Cre-mediated lineage tracing, we used hGFAP-myrGFP reporter plasmids to identify hGFAP + qNSCS cells ( Figure 5E). GFP + qNSCs cells of WT and 3xTg mice showed no difference in baseline expression of pS6, and EGFR-CA increased GFP + pS6 + cells equally in both WT and 3xTg mice ( Figure 5F). However, an EGFR-CA increase in EdU incorporation was only stimulated in the WT strain (WT EV, 0.1 ± 0.1% versus WT EGFR-CA, 1.1 ± 0.2%, EdU + GFP + /GFP + cells, p = 0.0006, two-way ANOVA) ( Figure 5G). Consistent with this, the proportion of pS6 + cells that incorporated EdU showed a trend toward a decrease ( Figure 5H). Thus, while EGFR-CA can stimulate downstream signaling in qNSCs cells of 3xTg mice, AD-associated signals prevent these cells from undergoing division ( Figure 5H).
Together these finding reveal that anti-proliferative signals in the middle-aged and familial AD contexts are dominant over the ability of EGFR-CA to promote qNSC division.

DISCUSSION
Neural stem/progenitor cells in the adult V-SVZ are continuously exposed to a variety of niche signals, which modulate neurogenic output by regulating proliferation, differentiation, and survival of these precursor cells (Lim and Alvarez-Buylla, 2016). Transduction of niche signals occurs via multiple families of receptors, such as RTKs, Eph receptors, interleukin receptors and G-protein coupled receptors for example, which ultimately converge onto a relatively limited number of intracellular signaling pathways (Obernier and Alvarez-Buylla, 2019). Here, we focused on EGFR and three of its downstream signaling pathways that are shared with other RTKs: the PI3K/AKT, MEK/ERK and mTOR cascades (Oda et al., 2005;Lemmon and Schlessinger, 2010;Annenkov, 2014). Our findings support 3 main conclusions. First, NSPCs from the young adult and middle-aged V-SVZ are equally able to activate these pathways in culture but show differences in vivo. Second, the PI3K/AKT, MEK/ERK and mTOR signaling cascades exhibit differential impacts on NSPC proliferation, survival and differentiation. Third, ectopic expression of an activated EGFR receptor in qNSCs stimulates their activation and division in the young adult brain, but aging and AD mutations dominantly suppress EGFR-triggered qNSC division (Figures 6A,B).

Activation of RTK-Induced Signaling Pathways in NSPCs During Early-and Mid-Adulthood
Age-related declines in adult neurogenesis occur in virtually all mammals examined, including humans (Maslov et al., 2004;Luo et al., 2006;Bouab et al., 2011;Ernst et al., 2014;Boldrini et al., 2018). Our time-course analysis revealed a biphasic decline of neurogenesis within the V-SVZ of C57Bl6 mice: a steep slope of decline in proliferation, TAPs and neuroblasts until approximately 3 months of age followed by a more gradual rate of decline until the latest 11-month timepoint. Declining neurogenesis is likely to result from changes to both the intrinsic properties of aging NSPCs and extrinsic properties of the aging niche (Hamilton et al., 2013;Capilla-Gonzalez et al., 2015;Lupo et al., 2019). Our comparison of the response of NSPCs from 2-and 11-month-old mice to EGF supports the idea that the intrinsic proliferative ability of NSPCs is maintained over this period (Ahlenius et al., 2009;Bouab et al., 2011;Shook et al., 2012;Apostolopoulou et al., 2017). Notably, however, multiple downstream components of the mTORC1 and mTORC2 pathways were selectively decreased in the V-SVZ niche between 1 and 4 months of age (pAKT and p4EBP1, with a strong trend toward decrease of pS6). Interestingly, recent studies have implicated mTOR in the regulation of lysosome function in NSPCs, with lysosomemediated elimination of protein aggregates being necessary In vitro, the PI3K/AKT, mTORC1/2 and, to a lesser extent, MEK/ERK, modules all contribute to EGF-induced NSPC proliferation. Only the mTORC1/2 modules showed evidence of contributing to NSPC survival, while only the MEK/ERK module mediated EGF-induced repression of NSPC differentiation. The effect of mTORC1 and mTORC1/2 modules did not significantly differ for these parameters, so we present only mTORC1. In each table, the main effect of each module is represented in red. (B) Effect of ectopic activation of EGFR-induced signaling in electroporated qNSCs. In young adult mice (left side of diagram), electroporation with EGFR-CA upregulates levels of pS6, a downstream marker of EGFR signaling and of "primed" qNSCs (Rodgers et al., 2014) and promotes their cell cycle entry. EGFR-CA electroporation in qNSCs of mice at middle age or in mice carrying familial Alzheimer's disease (AD) mutations (right side of diagram) remains sufficient to activate EGFR-induced pS6, but these qNSCs do not undergo cell division. Thus, the antiproliferative signals associated with aging and AD are dominant to the pro-proliferative effects of EGFR-induced signaling pathways. and sufficient for NSCs to transition from the quiescent to activated state (Leeman et al., 2018;Morrow et al., 2020). It remains to be demonstrated whether the observed impairments in mTOR signaling are causally related to age-related reductions in proteostasis (Leeman et al., 2018).

PI3K/AKT, MEK/ERK, and mTOR Signaling Cascades Differentially Regulate NSPC Proliferation, Survival, and Differentiation
We used a pharmacological loss-of-function approach to study the individual contributions of the PI3K/AKT, MEK/ERK, and mTOR pathways to proliferation, survival, and differentiation in NSPC cultures. In control experiments, we verified that there was no cross inhibition between the PI3K inhibitor, LY294002, and the MEK inhibitor, PD184352. Besides their own signaling modules, these inhibitors also both inhibited the downstream mTOR pathway, but to different extents: MEK inhibited partially and PI3K inhibited completely the phosphorylation of mTORC1 targets. NSPC proliferation in response to EGF, as measured by Ki67 staining, was compromised by all four of the inhibitors, indicating that the PI3K/AKT, MEK/ERK, and mTORC1 signaling modules are all important in this process. Rapamycin and KU0063794 had equivalent effects, suggesting that mTORC2 signaling is not involved. Previous studies have also found that NSPC proliferation was inhibited by Rapamycin (Paliouras et al., 2012;Wang et al., 2016), LY294002, or U0126, another inhibitor of MEK/ERK signaling (Yuan et al., 2015).
Interestingly, when the inhibitors were used in neurosphere cultures, neurosphere growth was completely arrested by the PI3K inhibitor, LY294002, and the MEK inhibitor, PD184352, and only partially affected by the mTOR inhibitors, suggesting different modes of action. This may be related to our finding that only the mTOR inhibitors promoted cell death. This was surprisingly, given that mTOR signaling was also repressed upon PI3K inhibition, and may be due to the fact that the PI3K/AKT module has other targets besides mTOR. MEK/ERK inhibition did not increase cell death and, thus, is dispensable for NSPC survival. Our data indicate that only the mTOR pathway is involved in the process of EGFmediated NSPC survival.
Differentiation in NSPC cultures is normally blocked in the presence of EGF. However, treatment with PD184352 significantly increased the expression of the differentiation markers GFAP and βIII tubulin. This effect was not observed by inhibiting either the PI3K/AKT or mTOR modules, suggesting that EGF-induced repression of NSPC differentiation is mediated specifically by the MEK/ERK signaling module. A previous study reported a similar observation with embryonic NSPCs cultured from the E15 telencephalon, where NSPCs treated with U126 showed reduced proliferation and increased neuronal differentiation (Wang et al., 2009); however, in that study, MEK/ERK inhibition reduced astrocytic differentiation, which was strongly increased here.
We conclude that the PI3K/AKT, MEK/ERK and mTOR1/2 modules play overlapping but dissociable roles in the biological effects of EGF on NSPCs: all 3 modules are implicated in NSPC proliferation (with the PI3K/AKT and mTOR modules being prominent), only the mTORC1 module contributes to NSPC survival and only the MEK/ERK module represses NSPC differentiation.

Activation of EGFR Signaling Pathways to Promote qNSC Activation
Since NSPC proliferation required activity of all three branches of EGFR signaling that we examined, we tested whether upregulation of EGFR signaling would be sufficient to trigger proliferation of qNSCs in vivo. We recently found that a population of quiescent, ventricle-contacting neural precursors that produces small numbers of olfactory bulb neurons can be genetically targeted by adult brain electroporation (Joppe et al., 2020). We therefore used this approach to transfect these quiescent precursors in 3-month-old adult male mice with a constitutively active EGFR construct (EGFR-CA). This led to detectable EGFR protein expression, and downstream mTOR expression in 20-25% of the electroporated qNSCs. After 1 week, their proliferation was increased, and after 1 month, they showed a trend for an increase in labeled cells within the OB. Thus, EGFR-CA is sufficient to recruit these quiescent precursors and potentially increase their neurogenic output in the young adult V-SVZ.
V-SVZ neurogenesis declines with age and this is further reduced in models of AD, including 3xTg mice (Hamilton et al., 2010. We therefore tested whether forced intrinsic expression of EGFR-CA would be sufficient to enable NSPCs to overcome the inhibitory influence of age and/or AD. Since neurogenesis in the V-SVZ is markedly decreased between 3 and 6 months of age, we first electroporated 3-versus 6-monthold wildtype mice with EGFR-CA. While electroporated mice of both ages increased expression of the mTORC1 readout, pS6, only electroporated cells in the 3-month-old mice increased their proliferation. Similarly, when we electroporated 3xTg mice and their strain controls, both increased pS6 expression but only the strain controls increased proliferation. Thus, activation of EGFR-induced signaling pathways promotes the activation and proliferation of qNSCs, but this is insufficient to allow qNSCs to overcome the inhibitory effects of age or AD on proliferation. Previous studies have reported that aging in the V-SVZ is associated with a reduced expression of pro-mitotic factors such as FGF-2 and VEGF (Shetty et al., 2005;Bernal and Peterson, 2011). Conversely, there is also an increase of antiproliferative factors from microglia (Solano Fonseca et al., 2016), astrocytes (Clarke et al., 2018), and the blood (Villeda et al., 2011;Katsimpardi et al., 2014). Our findings suggest declining neurogenesis during normal or pathological aging is primarily due to the presence of anti-proliferative factors present in the microenvironment rather than a loss of mitogens and indicate that such anti-proliferative factors are dominant over the effects of EGFR-induced signaling pathways.

Conclusion
In order to utilize the neurogenic capacity of adult NSPCs to prevent and/or reverse the age-related defects in the brain, it is essential to better understand the relationship between intrinsic and environmental mechanisms that control their activity. In this work, we show that EGFR-mediated activation of the PI3K/AKT, MEK/ERK, and mTOR cascades play dissociable roles in the control of NSPC proliferation, differentiation and survival in vitro. Forced activation of EGFR signaling can recruit quiescent NSPCs into division in vivo but is not by itself able to overcome dominant antiproliferative signals associated with normal or Alzheimertype aging. This suggests that optimal NSPC recruitment during aging is likely to be achieved only through combined NSPC stimulation and neutralization of antiproliferative extrinsic factors.

Contact for Reagent and Resource Sharing
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, KF (karl.fernandes@usherbrooke.ca).

Experimental Model and Subject Details
Animal work was conducted in accordance with the guidelines of the Canadian Council of Animal Care and were approved by the animal care committees of the University of Montreal and the Research Center of the University of Montreal Hospital (CRCHUM) (protocol N16002KFs, approval from January 2016). We used 42 male C57BL/6 (C57BL/6NTac, obtained from Taconic) for the time course experiment cultures and microdissections, 45 male Rosa26-stop-EYFP [B6.19 × 1-Gt(ROSA)26Sor tm1(EYFP)Cos /J, stock number: 006148, obtained from the Jackson Laboratory] for the infusion, electroporation and culture experiments, 10 female 3xTg-AD [B6;129-Psen1 tm1Mpm Tg(APPSwe,tauP301L)1Lfa/Mmjax; stock number: 034830-JAX, obtained from the Jackson Laboratory] and 10 female of their wildtype control strain (B6129SF2/J; stock number: 101045, obtained from the Jackson Laboratory) for electroporation experiments. Mice were socially housed (up to 5 mice/cage) on a 12 h light-dark cycle with free access to water and food.

Surgical Procedures
Mice were provided with water supplemented with acetaminophen as an analgesic (1.34 mg/ml, Tylenol) for 4 days, starting 1 day prior to surgery (Foley et al., 2019), in accordance with the animal committee requirements (protocol N16002KFs, approval from January 2016). Surgeries were performed under isoflurane general anesthesia (Baxter) and Bupivacaine local anesthesia (1 mg/kg, Hospira).

Electroporation
For the electroporations in young adults, we used 3-month old male Rosa26-stop-EYFP mice (n = 5 per condition). For the electroporations followed by neurosphere cultures experiment, we used 3-month old male Rosa26-stop-EYFP mice (n = 4 per group). For the electroporation in early middle-aged animals, we used 6-month old male Rosa26-stop-EYFP mice (n = 5 per condition). For the electroporation in an Alzheimer model, we used 6-month old female 3xTg mice and their WT control strain (n = 5 per group). Adult brain electroporation was conducted as described previously (Barnabe-Heider et al., 2008;Joppe et al., 2015;Joppe et al., 2020). Plasmids (Supplementary Table 1) were amplified by using an endotoxin-free 40 min Fast Plasmid Maxiprep Kit (Biotool), then purified and concentrated by ethanol precipitation. Intracerebroventricular plasmid injections were performed using a 10 µl Hamilton syringe into the left ventricle at coordinates: 0 mm anteroposterior (AP), +0.9 mm mediolateral (ML), -1.5 mm dorsoventral (DV) to Bregma, using a stereotaxic apparatus (Stoelting Co.), based on The Mouse Brain in Stereotaxic Coordinates, Compact 3rd Edition (Academic Press). Animals received an ICV injection of 10 µg of total DNA in 2 µl, delivered over 2 min, following by five pulses at 50 ms intervals at 200 V applied with 7 mm platinum Tweezertrodes (Havard Apparatus) and an electroporator (ECM 830, Havard Apparatus). Plasmids are listed in Supplementary Table 1.

Osmotic pump infusions
For the electroporation followed by osmotic pump infusion experiment, we used 3-month old male Rosa26-stop-EYFP (n = 5 per condition). Mice were under isoflurane general anesthesia (1% oxygen, 2% isoflurane) during the whole procedure. For electroporations, intracerebroventricular plasmid injections were performed using a 10 µl Hamilton syringe into the left ventricle at coordinates: 0 mm anteroposterior (AP), +0.9 mm mediolateral (ML), -1.5 mm dorsoventral (DV) to Bregma using a stereotaxic apparatus (Stoelting Co.), based on The Mouse Brain in Stereotaxic Coordinates, Compact 3rd Edition (Academic Press). Immediately after the electroporation, we proceeded with the ICV infusions. The ICV infusions were performed using 3-day osmotic pumps (Alzet, model 1003D, Durect) attached to brain infusion cannulae (Alzet, Brain infusion kit 3, Durect). Pump cannulae were implanted contralateral to the electroporation (right ventricle) procedure using a stereotaxic apparatus (Stoelting Co.) at coordinates: 0 mm anteroposterior (AP), -0.9 mm mediolateral (ML), -1.5 mm dorsoventral (DV) to Bregma, based on The Mouse Brain in Stereotaxic Coordinates, Compact 3rd Edition (Academic Press). Cannulae were fixed to the skull using dental cement (Co-oral-ite Dental Mfg). Osmotic pumps were filled with EGF diluted in a vehicle solution (0.1% BSA in PBS) and infused at 400 ng/day for 3 days, then animals were euthanized by intraperitoneal injection of ketamine/xylazine (347/44 mg/kg, Bimeda-MTC/Boehringer Ingelheim Canada Ltd) for immediate analysis.
For the biochemical analysis of 1-vs. 4-month old V-SVZ, we used microdissections from C57BL/6 (n = 4/age group) (Figure 1), microdissected V-SVZ and striatal samples were obtained from freshly dissected mouse brains as follows. Brains were placed in a brain mold and straight-edge razor blades were used to cut a 2-mm-thick coronal section through the forebrain (corresponding to the region between 3-and 5-mm posterior to the anterior edge of the olfactory bulbs). Using a dissecting microscope, fine-tipped tungsten needles were then used to dissect out a block of V-SVZ tissue, by first making lateral cuts through the corpus callosum (dorsally) and above the anterior commissures (ventrally), and then tracing the graywhite boundaries at the V-SVZ/striatum and V-SVZ/septum borders. The V-SVZ region was defined by the cell-dense tissue directly located around the ventricle. The Striatum region was defined by its location adjacent to the ventricle and its fibrous aspect. A block of striatal tissue of similar size was cut from the center of the adjacent striatum for biochemical comparison and to provide a control for SVZ microdissection purity. A schematic of the microdissection procedure is presented in Supplementary Figure 3.

Western blotting
Cultures analyzed by Western blotting were lysed in Ripa Buffer as previously described (Bouab et al., 2011). Protein samples for Western blotting were prepared as described previously (Bouab et al., 2011;Paliouras et al., 2012), and primary antibody information is provided in Supplementary Table 2. Ten micrograms of protein from each sample were run per lane. Membranes were blocked for 2 h with 5% milk in TBST. Primary antibodies diluted in TBST were applied at 4 • C overnight. Secondary antibodies were applied at room temperature for 45 min and washed in TBST. Signals were revealed using the Clarity kit (Bio-Rad), detected using ChemiDoc (Bio-Rad). Membranes were subsequently stripped with Re-Blot Plus Mild (Millipore), reblocked, and reprobed appropriately. Quantifications were performed using Image Lab 4.1 software (Bio-Rad). Normalized densities were obtained by taking the ratio of the optical density of the protein of interest and dividing by the optical density of Actin. This method was used for the 1 vs. 4 months experiments using C57BL/6 (n = 4/age group), 2 vs. 11 months experiments using Rosa26-stop-EYFP (n = 3/group) and cell culture experiments using Rosa26-stop-EYFP (n = 3-4/group).
Neurosphere numbers were quantified by plating cells in 24-well plates at the clonal densities indicated above (minimum of 8 wells/treatment/N). Neurosphere sizes were quantified by measuring the diameter of at least 100 neurospheres/condition using Fiji software (version 1.52p v, NIH, United States) and graphed using GraphPad Prism, Version 8.4 (GraphPad Software, Inc).

Quantification and Statistical Analyses
Immunostained tissue sections were examined using a motorized Olympus IX81 fluorescence microscope, an Olympus BX43F light microscope, or a TCS-SP5 inverted microscope (Leica Microsystems, Exton, United States). All quantifications were performed by a blinded observer using coded slides and 40X or 60X objectives. For quantification of total Ki67, Mash1, DCX, EGFR, or p-S6 cells, 3-6 V-SVZ sections/animal were analyzed. For quantification of electroporated cells and their progeny, 6-12 sections/animal were used for the V-SVZ and 5-15 sections/animal for the OB. Counts in the V-SVZ were limited to the DAPI-defined subventricular zone. Counts in the OB (Figures 4N,O) were performed by scanning the OB sections for positive cells at 32x objective magnification. All YFP positive cells in the V-SVZ or OB were confirmed for the presence of a DAPI-stained nucleus. No electroporated animals were excluded from this study.
All statistical analyses were achieved using GraphPad Prism, Version 8.4.3 (GraphPad Software, Inc). The statistical analyses that were performed are indicated in the respective figure legends and included either parametric tests (twotailed unpaired Student's t-test, one-way or two-way ANOVA with Tukey's or Dunnett's post-test) or the nonparametric Mann-Whitney test (the latter in the case of numbers of electroporated cells, which might follow a nonnormal distribution). Error bars represent mean ± standard error of the mean (SEM). Significance level was set at p ≤ 0.05.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

ETHICS STATEMENT
The animal study was reviewed and approved by Animal Care Committees of the University of Montreal and the Research Center of the University of Montreal Hospital (CRCHUM).

AUTHOR CONTRIBUTIONS
LC and KF: conceptualization. LC, SJ, and L-CL: methodology. LC, SJ, L-CL, FP, and AA: investigation. LC and KF: manuscript writing and visualization. KF: supervision and funding acquisition. All authors contributed to the article and approved the submitted version.