Syndecan-1 Stimulates Adult Neurogenesis in the Mouse Ventricular-Subventricular Zone after Injury

Summary

The production of neurons from neural stem cells (NSCs) persists throughout life in the mouse ventricular-subventricular zone (V-SVZ). We have previously reported that NSCs from adult V-SVZ are contained in cell populations expressing the carbohydrate SSEA-1/LeX, which exhibit either characteristics of quiescent NSCs (qNSCs) or of actively dividing NSCs (aNSCs) based on the absence or the presence of EGF-receptor, respectively. Using the fluorescence ubiquitination cell cycle indicator-Cdt1 transgenic mice to mark cells in G₀/G₁ phase of the cell cycle, we uncovered a subpopulation of qNSCs which were primed to enter the cell cycle in vitro. Besides, we found that treatment with Syndecan-1, a heparan sulfate proteoglycan involved in NSC proliferation, hastened the division of qNSCs and increased proliferation of aNSCs shortening their G₁ phase in vitro. Furthermore, administration of Syndecan-1 ameliorated the recovery of neurogenic populations in the V-SVZ after radiation-induced injury providing potential cure for neurogenesis decline during brain aging or after injury.

Graphical Abstract

Highlights

• •A subpopulation of quiescent NSCs are primed to enter cell cycle
• •The content of primed quiescent NSCs decreases rapidly with age
• •Syndecan-1 favors cell cycle progression of NSCs in vitro and in vivo

Neuroscience; Stem Cell Research

Introduction

The generation of neurons, astrocytes, and oligodendrocytes persists throughout life in specific regions of the mammalian brain and contributes to neural plasticity through rewiring, refreshing of established networks, and maintenance of cognition (Kempermann et al., 2018). Remarkably, two main regions of the forebrain exhibit germinative potentials, namely the subgranular zone of the dentate gyrus in the hippocampus and the ventricular-subventricular zone (V-SVZ) (Obernier and Alvarez-Buylla, 2019). Adult neurogenesis within the V-SVZ is insured by neural stem cells (NSCs), or type B cells, that enter the cell cycle then successively give rise to transit amplifying cells (type C) and neuroblasts (type A) which differentiate into neurons once they have reached the olfactory bulbs (Obernier and Alvarez-Buylla, 2019).

In the adult brain, NSCs are contained in two populations of quiescent NSCs (qNSCs) and active NSCs (aNSCs) with different cell cycle features (Codega et al., 2014; Daynac et al., 2013). Particularly, owing to their high proliferative capacity, proliferating aNSCs are rapidly eliminated in the V-SVZ after anti-mitotic treatment, whereas qNSCs resist and then re-entering the cell cycle to insure progressive recovery of neurogenesis (Codega et al., 2014; Daynac et al., 2013; Llorens-Bobadilla et al., 2015; Mich et al., 2014). The majority of adult NSCs are produced during embryonic days (E13.5 - E15.5) in the mouse and remain largely quiescent until they become reactivated postnatally contributing to neurogenesis (Fuentealba et al., 2015; Furutachi et al., 2015).

Defects of neurogenesis occur during aging and most studies agree that it is related to a progressive reduction in the number of proliferating cells in the V-SVZ (Blackmore et al., 2009; Enwere et al., 2004; Maslov et al., 2004; Tropepe et al., 1997). Cell cycle alterations of NSCs or decline in their number explained the age-related neurogenesis decline and are already visible at 6 months in the adult mouse (Bouab et al., 2011; Daynac et al., 2014, 2016a; Luo et al., 2008). Several factors from the neurogenic niche, including inflammatory factors, have been shown to reduce neurogenesis during aging by altering the cell cycle of NSCs (Daynac et al., 2014; Kalamakis et al., 2019; Pineda et al., 2013; Silva-Vargas et al., 2016). Quiescence of NSCs has also been shown to be triggered by Wnt antagonist in the aging brain (Kalamakis et al., 2019).

Syndecan-1 (SDC1, CD138) is a cell surface heparan sulfate proteoglycan that has been reported to modulate neural progenitor proliferation during embryogenesis through Wnt signaling pathways (Wang et al., 2012). Recently, we reported on a different expression pattern of SDC1 between qNSCs and aNSCs and demonstrated its role in the proliferation of aNSCs in the postnatal V-SVZ (Morizur et al., 2018).

Here, we examined in more detail alterations during aging of NSC proliferation from postnatal V-SVZ in mice particularly with regards to capacity of qNSCs to enter cell cycle and the effects of exogenous SDC1 on NSC proliferation.

Results

LeX^bright qNSCs Are Produced during Embryogenesis

We previously reported on an FACS strategy to sort LeX^bright qNSCs and LeX + EGFR + aNSCs from the adult mouse and early postnatal V-SVZ based on the absence of CD24 expression and the differential expression of the membrane markers EGFR and LeX (Daynac et al., 2013, 2015; Morizur et al., 2018). These LeX^bright qNSCs and LeX + EGFR + aNSCs exhibit strikingly similar molecular profiles to that obtained by other strategies, including anti-GLAST antibody, and/or transgenic GFAP:GFP mice in adult V-SVZ (Beckervordersandforth et al., 2010; Codega et al., 2014; Llorens-Bobadilla et al., 2015; Mich et al., 2014).

Adult NSCs have been reported to be produced during mid-embryonic development in the mouse and remain largely quiescent until they become reactivated postnatally (Fuentealba et al., 2015; Furutachi et al., 2015). Therefore, to mark qNSCs generated during embryogenesis, BrdU was administrated to pregnant mice from E14.5 to E15.5. qNSCs that underwent rare divisions, if any, were characterized by BrdU-label retention in postnatal brains (one month after birth). Immunostaining of SOX2 was used to confirm NSC identity and labeling of the G1 phase cell cycle marker MCM2 was used to detect both slowly and rapidly cycling NSCs (Maslov et al., 2004). LeX^bright qNSCs and LeX + EGFR + aNSCs were sorted from V-SVZ one month after birth. Although they are actively dividing cells in vivo (Morizur et al., 2018), LeX + EGFR + aNSCs were in great majority (72%) BrdU/SOX2/MCM2-triple positive and had a BrdU staining less intense in comparision to LeX^bright qNSCs, suggesting that they derived from the later with few divisions (Figures 1 and S1). On the other hand, the majority of LeX^bright cells (82%) was positive for BrdU confirming their embryonic origin (Figures 1 and S1). Moreover, 56.8% of LeX^bright cells were BrdU/SOX2/MCM2-triple positive cells, i.e. very long-term qNSCs, whereas 25% were BrdU+/SOX2+/MCM2-negative i.e. likely differentiated astrocytes.

Figure 1.

Young Adult LeX^bright qNSCs Are Produced during Embryogenesis

BrdU was given to pregnant mice (from E14.5 to E15.5), and then LeX + EGFR + aNSCs and LeX^bright qNSCs were sorted using a combination of CD24/LeX/EGFR membrane markers one month after birth and BrdU incorporation was quantified together with SOX2 and MCM2. Data are represented as median ± interquartile range from replicate experiments (n = 4) with 2–4 mice per group. See also Figure S1. ^∗P < 0.05.

These data confirmed thus that the LeX^bright qNSCs from the postnatal brain are produced during embryogenesis similarly as previously reported by others using in vivo approaches (Fuentealba et al., 2015; Furutachi et al., 2015).

The Subpopulation of Primed LeX^bright qNSCs Able to Enter the Cell Cycle Rapidly Decreases with Age

Many structural and cellular modifications of the V-SVZ progressively occur during the early postnatal development and later in the adult brain, but data on the cell cycle status of NSCs associated with these changes are still incompletely documented. qNSCs and aNSCs identity in early postnatal V-SVZ was confirmed with GFAP expression, a marker of adult-type B/NSCs (Doetsch et al., 1999) and MCM2 expression. LeX^bright and LeX + EGFR + cells were sorted from early postnatal V-SVZ (10 days). Both LeX^bright qNSCs and LeX + EGFR + aNSCs expressed GFAP (Figure S2). MCM2, which is expressed in actively dividing cells and slowly cycling NSCs (Maslov et al., 2004), was detected on 65% of LeX + EGFR + aNSCs but also 25% of LeX^bright qNSCs (Figure S2C), confirming that a part of qNSCs were primed to enter cell cycle at this developmental stage.

To further examine cell cycle alterations of NSC populations from early postnatal pups to aged mice, we used fluorescence ubiquitination cell cycle indicator (FUCCI) Cdt1-red transgenic mice. FUCCI-Cdt1 system allows the visualization of cells in G1 with the presence of a G1-specific red-Cdt1 reporter (FUCCI^pos), while it is absent in cells during the S/G2/M phases (FUCCI^neg) (Sakaue-Sawano et al., 2008). In addition, FUCCI^high allows the identification of cells that have exited the cell cycle (G0 cells) (Daynac et al., 2014; Roccio et al., 2013).

At each developmental stage, most LeX + EGFR + aNSCs were in G1 (FUCCI^low) and in S/G2/M phases (FUCCI^neg), and fewer had exited cell cycle (FUCCI^high) according to their active proliferating status (Figure 2A and Table S1). Interestingly, we found higher proportions of aNSCs in G0 (FUCCI^high) in the early postnatal brains than in the adult brains, which is reminiscent of an extensive differentiation associated with the brain development after birth. Moreover, the ratio of cells in G1 (FUCCI^low) over S/G2/M phases (FUCCI^neg) was higher in the early postnatal brain than in adult brain indicating a longer G1 phase.

Figure 2.

Cell Cycle Status and Clonogenic Capacities of LeX^bright qNSCs and LeX + EGFR + aNSCs in the Postnatal V-SVZ

(A and B) FUCCI fluorescence was determined by flow cytometry in V-SVZ at various postnatal ages. The percentage of different FUCCI populations (FUCCI^high: G0; FUCCI^low: G1 and FUCCI^neg: S/G2/M) was represented among (A) LeX + EGFR + aNSCs and (B) LeX^bright qNSCs.

(C) The formation of neurospheres, i.e. clonogenic capacity, was determined 7 days after sorting LeX^bright qNSCs and LeX + EGFR + aNSCs from 10 days- to 2 months-old V-SVZ. Data were compared using non-parametric Kruskal-Wallis and Dunn's multicomparison tests and given in Table S1. Data are represented as median ± interquartile. Each dot in C represents individual experiment. See also Table S1 and Figure S3. ^∗∗P < 0.005.

On the contrary, a large proportion of LeX^bright qNSCs had exited the cell cycle (FUCCI^high) at all developmental stages in accordance with their quiescent status (Figure 2B). Nonetheless, we found between 21.6% and 33.2% of LeX^bright qNSCs that were either in G1 (FUCCI^low) or in S/G2/M (FUCCI^neg), whatever the developmental stage, highlighting that a significant and stable fraction of qNSCs are primed to enter cell cycle in the V-SVZ from both early postnatal and adult mouse brain. Nonetheless, a significant fraction of LeX^bright qNSCs were in G1 (FUCCI^low), but few of them were actively cycling (FUCCI^neg) (Figure 2B). Markedly, the percentage of LeX^bright qNSCs in S/G2/M (FUCCI^neg) dropped in the young adult mice (Figure 2B).

We then compared the capacities of LeX^bright qNSCs and LeX + EGFR + aNSCs sorted from PN10 and young adult V-SVZ to proliferate and to form neurospheres, which are floating clones initiated by actively dividing NSCs (Pastrana et al., 2011). LeX + EGFR + aNSCs had a similar clonogenic capacity in early postnatal and adult brains (Figure 2C). In addition, LeX + EGFR + subpopulations were sorted according to FUCCI (FUCCI^low: G1 and FUCCI^neg: S/G2/M) but presented no difference of clonogenic capacity at both ages (Figure S3). Besides, accordingly with their quiescent status, LeX^bright qNSCs from neonatal brain had a 15.5 times lower clonogenic capacity than their activated LeX + EGFR + aNSCs counterparts (Figure 2C). Clonogenic capacity of LeX^bright qNSCs was further decreased by more than 100-fold in the adult brain. This decrease in clonogenic capacity appeared to reflect the disappearing of clonogenic LeX^brigh cells rather than their positioning within cell cycle since LeX^bright FUCCI^low and FUCCI^neg had similar clonogenicity (Figure S3). This extremely low clonogenic capacity of LeX^bright qNSCs in young adult V-SVZ contrasted with their MCM2 expression (Figure 1), suggesting that most primed LeX^bright qNSCs are not fully activated to generate clones.

Our data revealed an early cell cycle alteration of LeX + EGFR + aNSCs in early postnatal pups most probably coinciding with differentiation during development. Additionally, some LeX^bright qNSCs entered the cell cycle in the early postnatal, whereas they were blocked in G1 in the young adult mice.

A Subpopulation of LeX^bright FUCCI^neg/low qNSCs Maintains the Capacity to Enter Proliferation in the Young Adult V-SVZ

LeX^bright qNSC and LeX + EGFR + aNSC populations were sorted from the V-SVZ of young adult mice and then plated in enriched NSC medium supplemented with BrdU (10 μM) to challenge their proliferation capacities. After 72hr, the great majority of LeX + EGFR + aNSCs (84.2 ± 3.4%) had incorporated BrdU and exhibited an active protein synthesis activity as seen with phosphorylation of S6 ribosomal protein (69.8 ± 3.4%) (Figures 3A and 3B). By contrast, only 27.2 ± 9.8% of LeX^bright qNSCs had incorporated BrdU, and few of them (3.1 ± 1.6%) were positive for phosphoS6 ribosomal protein (Figures 3A and 3B). Both BrdU incorporation and pS6 suggested that a subpopulation of qNSCs was primed to enter cell cycle as previously reported (Codega et al., 2014; Llorens-Bobadilla et al., 2015).

Figure 3.

LeX^bright FUCCI^neg/low qNSCs Divide and Initiate Clones In Vitro but FUCCI^high NSCs Have Definitively Exited the Cell Cycle

(A and B) LeX^bright qNSCs and LeX + EGFR + aNSCs were sorted from young adult mouse V-SVZ and then cultured in the presence of BrdU for 72hr in vitro.

(A) Percentages of BrdU incorporation and (B) pS6-positive cells were quantified in LeX^bright qNSCs and LeX + EGFR + aNSCs from replicate experiments with a total number of 66–144 cells.

(C–E) LeX^bright cells were sorted from young adult V-SVZ according to their FUCCI-Cdt1 fluorescence intensity and were followed by time-lapse videomicroscopy. (C) LeX^bright FUCCI^high cells never divide but (D) the LeX^bright FUCCI^neg/low population showed divisions. (E) Thereafter, formation of clones was rarely observed in the LeX^bright FUCCI^neg/low qNSCs. Scale bars = 10μm.

To further explore the capacity of these LeX^bright qNSCs to re-enter into the cell cycle, we prospectively sorted them from the young adult FUCCI V-SVZ and challenged their proliferation capacities by in vitro time-lapse videomicroscopy.

On one hand, FUCCI^high cells representing the majority of LeX^bright population (interquartile range [67.9%–79.2%]) had apparently exited the cell cycle and presented a bright red fluorescence in vitro, which eventually disappeared when cells died but they never divided (Figure 3C). On the other hand, 20% of the FUCCI^neg/low population of LeX^bright qNSCs, encompassing G1 and scarce S/G2/M cells, had divided during the first 26hr (Figure 3D). Then FUCCI fluorescence showed an increase which became high in some LeX^bright qNSCs. Thereafter, FUCCI^neg/low LeX^bright qNSCs very scarcely generated clones that grew slowly and ultimately gave rise to larger clones (Figure 3E).

Altogether, these data show that some of the LeX^bright qNSCs were primed to re-enter cell cycle in vitro and were contained within the FUCCI^neg/low population.^.

Recombinant Syndecan-1 Shortens Cell Cycle Progression In Vitro

We have shown that SDC1 is highly expressed in proliferating LeX + EGFR + aNSCs from adult mice while it is found at low levels in LeX^bright qNSCs (Morizur et al., 2018). Here, we reported that SDC1 mRNAs were expressed at a similar level in LeX + EGFR + aNSCs from neonatal and adult mice (Figure S4). Although the expression of SDC1 mRNAs was low in LeX^bright qNSCs from neonates, it showed a further decline in adult mice (Figure S4) coinciding with the drop in clonogenic capacities. We have previously reported that SDC1 knockdown reduces proliferation of LeX + EGFR + aNSCs (Morizur et al., 2018). Likewise, addition of chondroitinase ABC (50 and 100 U/mL) which degrades chondroitin sulfate proteoglycans, including SDC1, reduced the formation of neurospheres initiated by LeX + EGFR + aNSCs (data not shown). Conversely, deglycanation of SDC1 with heparanase (4 ng/mL) which transforms SDC1 into a highly selective surface-binding protein (Ma et al., 2006) increased BrdU incorporation in proliferating NSCs (data not shown).

Therefore, we tested the effects of exogenous recombinant SDC1 (recSDC1) on proliferation of NSCs freshly sorted from young adult mice. Two days after plating, LeX + EGFR + aNSCs gave rise to small colonies containing 4-6 cells in control conditions and 4-8 cells in the presence of 2.5μg/ml of recSDC1. This increase was further observed on neurosphere size at 6 days in the presence of 2.5 μg/mL of rec SDC1 (Figure 4A). This increase in neurosphere size inversely mirrored the data we previously reported after silencing SDC1 (Morizur et al., 2018). Besides, the clonogenic efficacy of LeX + EGFR + aNSCs appeared to be almost significantly increased in the presence of recSDC1 (Figure 4B).

Figure 4.

Recombinant SDC1 Increases the Neurosphere Size of aNSCs and Hastens the Division of aNSCs and qNSCs

(A and B) LeX + EGFR + aNSCs were sorted from young adult V-SVZ. After 6 days in culture of LeX + EGFR + aNSCs in the absence, or the presence, of 2.5μg/ml recSDC1: (A) the neurosphere area (μm²) and (B) clonogenic capacities were determined.

(C and D) LeX + EGFR + aNSCs were sorted from FUCCI neonates and were followed by time-lapse for 26-30hr in the absence, or the presence, of 2.5μg/ml recSDC1, then (C) the time for the first division and (D) the length of the subsequent G₁ were recorded.

(E and F) LeX^bright FUCCI^neg/low qNSCs were sorted from adult FUCCI mice, then (E) the percentage of dividing cells/field and (F) the time for the first division were recorded. Data are represented as the median ± interquartile range of replicate experiments with 2–4 mice per group. See also Figure S5. ^∗P < 0.05; ^∗∗P < 0.005.

We thus investigated the effects of recSDC1on cell cycle length of LeX + EGFR + aNSCs by time-lapse videomicroscopy using FUCCI mice. Interestingly, the time for the first division of LeX + EGFR + aNSCs from neonates was significantly shortened in the presence of 2.5μg/ml of recSDC1 in comparison to control (Figure 4C). A similar shortening was observed in young adult V-SVZ (Figure S5). The length of the G1 phase was measured for the subsequent cycle taking advantage of the FUCCI-Cdt1 fluorescence. A significant shortening of the G1 phase was observed in LeX + EGFR + aNSCs (Figure 4D). We then investigated the effects of recSDC1 on LeX^bright FUCCI^neg/low qNSCs, i.e. primed qNSCs. The percentage of dividing LeX^bright FUCCI^neg/low appeared not significantly increased in the presence of SDC1 (Figure 4E). Strikingly, the time for the first division of LeX^bright FUCCI^neg/low qNSCs was significantly shortened in the presence of 2.5μg/ml of recSDC1 (7h36) in comparison to control (9h12) (Figure 4F). However, LeX^bright FUCCI^neg/low cells remained FUCCI-Cdt1 positive at the end of the videomicroscopy (28-30hr) and did not re-enter in another division cycle even in the presence of recSDC1. By contrast, FUCCI^high LeX^bright cells did not divide even in the presence of SDC1 (data not shown). Strikingly, 25% (first interquartile) of LeX^bright FUCCI^neg/low and LeX + EGFR + NSCs exposed to SDC1 had divided before 5h and 7h for adults and neonates, respectively, suggesting that numerous aNSCs and qNSCs might be blocked in G₂.

Altogether, our data show that SDC1 favored cell cycle progression of primed qNSCs and aNSCs.

Recombinant Syndecan-1 Accelerates Recovery of Neurogenic Populations In Vivo

Subsequently, we addressed the capacity of recSDC1 to ameliorate V-SVZ recovery in vivo after brain irradiation in young adult FUCCI-Cdt1 transgenic mice. After such radiation-induced injury, highly proliferating cells, including LeX + EGFR + aNSCs, are rapidly eliminated, whereas LeX^bright qNSCs are radioresistant and enter cell cycle 48hr after exposure, which is followed by the progressive recolonization of the V-SVZ by LeX + EGFR + aNSCs, EGFR + progenitors and CD24 + EGFR + young neuroblasts (Daynac et al., 2013). Collection of V-SVZ cells from young adult FUCCI-Cdt1 mice 48hr after irradiation demonstrated the elimination of the majority of LeX + EGFR + aNSCs, EGFR+ and CD24 + EGFR + cells while leaving a significant amount of LeX^bright FUCCI^neg/low qNSCs, i.e. progressing throughout the cell cycle (Figure 5). Recombinant SDC1 was administrated intraventricularly while mice received BrdU through systemic route and recovery of neurogenic populations was analyzed three days later. The absolute counts of the neurogenic populations revealed that recSDC1 favored primarily the recovery of EGFR+ and CD24 + EGFR + cells (Figures 5 and S6A). Interestingly, the increase in these EGFR+ and CD24 + EGFR + neurogenic populations with recSDC1 was observed in the FUCCI^low and FUCCI^neg cycling populations (Figures 5C and 5D) and was associated with BrdU incorporation, i.e. proliferation (Figure S6B). On the other hand, the presence of recSDC1 did not significantly improve the recovery of LeX^bright qNSCs and LeX + EGFR + aNSCs (Figures 5A and 5B) while BrdU incorporation indicated their proliferation (Figure S6B). Part of the LeX + EGFR + aNSCs had even exited cell cycle, i.e. FUCCI^high (Figure 2B). Nevertheless, this would be expected from asymmetric division of NSCs and their subsequent loss instead of EGFR+ and CD24 + EGFR + which undergo symmetric division.

Figure 5.

Exogenous SDC1 Favors Recovery of Neurogenic Populations after Radiation-Induced V-SVZ Injury

For a Figure360 author presentation of this figure, see https://doi.org/10.1016/j.isci.2020.101784.

(A–D) Two-months-old adult FUCCI-Cdt1 mice were irradiated (4 Gy/head only) and received recSDC1 intraventricularly. The absolute numbers of (A) LeX^bright qNSCs, (B) LeX + EGFR + aNSCs, (C) EGFR + progenitors, and (D) CD24 + EGFR + proliferating neuroblasts and FUCCI fluorescence were determined by flow cytometry within FUCCI subpopulations (FUCCI^high: G0; FUCCI^low: G1 and FUCCI^neg: S/G2/M). Data are represented as median ± interquartile range. Each plot represents an individual mouse (the number of mice is indicated within bars). Data were compared using non-parametric Kruskal-Wallis and Dunn's multicomparison tests. Comparisons with the Ctrl group are given upper the boxes and comparisons between PBS and recSDC1 over the brackets (° for FUCCI^high; # for FUCCI^low and ∗ for FUCCI^neg). See also Figure S6.

These data show that SDC1 favored recovery of neurogenic populations in the V-SVZ after injury.

Discussion

Different FACS strategies have been reported for the isolation of qNSCs and aNSCs (Beckervordersandforth et al., 2010; Codega et al., 2014; Daynac et al., 2013; Llorens-Bobadilla et al., 2015; Mich et al., 2014). Briefly, anti-GLAST or anti-LeX antibodies or GFAP:GFP are used as NSC markers in combination with other markers. Interestingly, qNSCs and aNSCs obtained from these different methods have strikingly similar molecular profiles. Particularly, qNSCs being characterized by the expression of several genes as Prom1, Aldh1l1, Gjb6, CD9, Sox9, Id2 and Id3, whereas aNSCs expressing high level of Ascl1, Egr1, Fos, Sox4, and Sox11 (Beckervordersandforth et al., 2010; Codega et al., 2014; Daynac et al., 2013; Llorens-Bobadilla et al., 2015; Mich et al., 2014). In addition, we have shown that LeX^bright qNSCs are produced during embryogenesis similarly as previously reported by for type B/NSCs in vivo (Fuentealba et al., 2015; Furutachi et al., 2015).

A pool of NSCs remains largely dormant until qNSCs become reactivated postnatally contributing to neurogenesis for brain homeostasis or repair after injury (Daynac et al., 2013). Using the FUCCI-Cdt1 system, we reported, here, on the prospective isolation of primed qNSC. Interestingly, we demonstrate that the capacity of qNSCs to enter cell cycle declines as early as in the young adult mouse brain. However, we have shown that cell cycle progression of LeX^bright qNSCs is favored by addition of exogenous SDC1.

Adult neurogenesis declines with aging due to the depletion and functional impairment of neural stem/progenitor cells. In this context, previous reports have indicated that the number of NSCs in the mouse V-SVZ declines by mid-age (10-12 months), with an additional reduction in older mice (22 months) reviewed in (Lupo et al., 2019). While prior studies have demonstrated a reduction of NSC content and have noted a paradoxical increase in the NSC division rate in old animals (Luo et al., 2008; Shook et al., 2012), we observed a decline in cell cycle activity of aNSCs at mid-age (6-12 months) but with stable contents (Daynac et al., 2014, 2016a). These seemingly divergent data are reconciled by the fluctuation of NSC proliferation with a decline between 2 and 18 months, and then unexpectedly a reversal faster cell cycle at 22 months (Apostolopoulou et al., 2017). On the other hand, increasing NSC quiescence is predicted to contribute to the age-related decline of neurogenesis (Bast et al., 2018; Kalamakis et al., 2019). NSCs begin undergoing quiescence-associated changes at mid-adulthood in the mouse V-SVZ (Bouab et al., 2011).

Using various approaches based on FUCCI-Cdt1, BrdU incorporation, clonogenicity, which showed consistency, we have shown that proliferation capacities of NSCs decreased early in the postnatal V-SVZ. Particularly, aNSCs showed a G1 lengthening in the early postnatal V-SVZ associated with cell cycle exit which is in accordance with differentiation during brain development (Salomoni and Calegari, 2010). Grippingly, we have identified a population among LeX^bright qNSCs that are characterized by a low FUCCI-Cdt1 fluorescence which are primed to enter the cell cycle. The population of LeX^bright is able to re-entry into cell cycle after V-SVZ irradiation accordingly to what was previously reported (Daynac et al., 2013) and similarly to what was shown for other brain injuries (Codega et al., 2014; Llorens-Bobadilla et al., 2015; Mich et al., 2014). Noticeably, we have shown a drop of primed LeX^bright FUCCI^neg qNSCs entering the cell cycle in the young adult which was associated to the decline in their clonogenic capacities. Although expressing a significant level of SOX2 and MCM2, marking slow dividing NSC (Maslov et al., 2004), LeX^bright cells divided once, then stopped in G₁ but very scarcely formed neurospheres in neonates and even more rarely when isolated from young adults. This might indicate that experimental conditions are not adequate for NSC proliferation in vitro and/or that high LeX expression interfered with NSC proliferation as previously reported (Luque-Molina et al., 2017). These early events are reminiscent to the human V-SVZ which shows a drop in neurogenesis several months after birth (Sanai et al., 2011). This continuum of NSC proliferation decline starting at birth is linked to postnatal brain development. Also, this suggests that qNSCs are recruited, through their activation, early during the postnatal brain development. Therefore, NSC quiescence appears as a programmed developmental process and a consequence of molecular aging processes.

Loss of the ability of qNSCs to activate and/or proliferation defect during aging are related to alterations of transcriptomic or lysosomal activities (Leeman et al., 2018; Lupo et al., 2018). Raise in inflammatory factors within the neurogenic niche has also been shown to reduce neurogenesis by altering the cell cycle or quiescence of NSCs during aging (Daynac et al., 2014, 2016a; Engler et al., 2018; Kalamakis et al., 2019; Pineda et al., 2013; Silva-Vargas et al., 2016). Several signaling pathways such as WNT or Sonic Hedgehog have been involved in the regulation of NSC quiescence and proliferation (Chavali et al., 2018; Daynac et al., 2016b). Interestingly, activation of LeX^bright qNSCs after radio-induced injury was associated with the upregulation of several genes belonging to WNT signaling (Table S2). Conversely, the niche-derived WNT antagonist sFRP5 has been shown to induce quiescence in the aging brain (Kalamakis et al., 2019).

SDC1 plays a role in NSC proliferation during embryogenesis and in postnatal NSCs (Morizur et al., 2018; Wang et al., 2012). SDC1 regulates proliferation in part by modulating the ability of neural progenitors to respond to WNT ligands (Wang et al., 2012). Here, we have shown that addition of exogenous SDC1 favors proliferation of aNSCs in vitro through the reduction of the G1 phase. Morever, exogenous SDC1 hastened the time for the first division of primed qNSCs. Interestingly, numerous NSCs divided early after exogenous SDC1 treatment suggesting that they are paused in G₂, as reported in the drosophila (Otsuki and Brand, 2018). Whether these proliferation effects involved WNT signaling needs to be further investigated. Grippingly, exogenous SDC1 also demonstrated in vivo activity by favoring recovery of neurogenic populations after V-SVZ injury. Therefore, SDC1 might be used to counteract brain injury. If validated in older animals, administration of SDC1, or of an agonist, might provide the means to control the proliferation of NSCs and to counteract the neurogenesis decline during aging which is still a major concern.

Limitations of the Study

This study has some limitations that should be kept in mind when interpreting the relevance of the findings to regeneration of neurons. Primed LeX^bright qNSCs were able to make only a single division suggesting that unknown factor(s) are further required to insure subsequent divisions. Nevertheless, molecular characterization of this rare population might be helpful to identify such factor(s). The administration of recombinant Syndecan-1 ameliorated the recovery of neurogenic populations in vivo, but the target cells have to be further characterized. In addition, the production of neurons and their functionality remain to be analyzed. Finally, the transfer to clinic has to take into account that Syndecan-1 might be hijacked by cancer cells to proliferate.

Resource Availability

Lead Contact

Further requests should be directed to and will be fulfilled by the Lead Contact: Marc-André Mouthon (marc-andre.mouthon@cea.fr).

Materials Availability

The study did not generate any unique reagent.

Data and Code Availability

This published article includes all datasets generated or analyzed during this study.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.

Published: December 18, 2020