Quiescent and Active Hippocampal Neural Stem Cells with Distinct Morphologies Respond Selectively to Physiological and Pathological Stimuli and Aging


Loading.... (view fulltext now)





Full text


Quiescent and Active Hippocampal Neural Stem Cells

with Distinct Morphologies Respond Selectively

to Physiological and Pathological Stimuli and Aging

Sebastian Lugert,1Onur Basak,1Philip Knuckles,1Ute Haussler,2Klaus Fabel,3Magdalena Go¨tz,4,5Carola A. Haas,2

Gerd Kempermann,3Verdon Taylor,1,6,7,*and Claudio Giachino1,7

1Department of Molecular Embryology, Max Planck Institute of Immunobiology, Stubeweg 51, D-79108 Freiburg, Germany 2Neurocenter, University of Freiburg, Breisacher Strasse 64, D-79106 Freiburg, Germany

3Genomics of Regeneration, Center for Regenerative Therapies Dresden, Tatzberg 47-49, D-01307 Dresden, Germany

4Institute for Stem Cell Research, Helmholtz Zentrum Munchen, Ingolsta¨dter Landstrasse 1, D-85764 Neuherberg, Munich, Germany 5Department of Physiological Genomics, Institute of Physiology, Ludwig-Maximilians University Munich, Schillerstrasse 46,

D-80336 Munich, Germany

6Centre for Stem Cell Biology, Department of Biomedical Science, University of Sheffield, Western Bank, S10 2TN Sheffield, UK 7These authors contributed equally to this work

*Correspondence:taylor@immunbio.mpg.de DOI 10.1016/j.stem.2010.03.017


New neurons are generated in the adult

hippo-campus throughout life by neural stem/progenitor

cells (NSCs), and neurogenesis is a plastic process

responsive to external stimuli. We show that

canon-ical Notch signaling through RBP-J is required for

hippocampal neurogenesis. Notch signaling

distin-guishes morphologically distinct Sox2


NSCs, and

within these pools subpopulations can shuttle

between mitotically active or quiescent. Radial and

horizontal NSCs respond selectively to neurogenic

stimuli. Physical exercise activates the quiescent

radial population whereas epileptic seizures induce

expansion of the horizontal NSC pool. Surprisingly,

reduced neurogenesis correlates with a loss of active

horizontal NSCs in aged mice rather than a total loss

of stem cells, and the transition to a quiescent state is

reversible to rejuvenate neurogenesis in the brain.

The discovery of multiple NSC populations with

Notch dependence but selective responses to stimuli

and reversible quiescence has important

implica-tions for the mechanisms of adaptive learning and

also for regenerative therapy.


Self-renewing NSCs drive neurogenesis within discrete regions of the adult mammalian brain. In the subgranular zone (SGZ) of the adult hippocampal dentate gyrus (DG), NSCs (also referred to as Type-1 cells) produce intermediate progenitor cells (IPs, Type-2a cells) that retain expression of stem/progenitor markers including the transcription factor Sox2 (Kempermann et al., 2004; Seri et al., 2001; Steiner et al., 2006). At this stage, neuronal determination becomes apparent, with overlapping expression of the transcription factors Prox1, NeuroD1, and

the structural protein Doublecortin (Dcx) (Type-2b) ( Kemper-mann et al., 2004; Seri et al., 2001; Steiner et al., 2008). These cells generate migratory neuroblasts (Type-3) that proliferate but exit the cell cycle before full maturation into granule neurons. Although recent evidence indicates that Sox2 is important for hippocampal NSCs, activating sonic hedgehog and regulating Wnt activity, the mechanisms controlling DG NSC maintenance and differentiation are not well understood (Favaro et al., 2009; Kuwabara et al., 2009). This is partially due to the fact that NSC identity and the exact neurogenic lineage in the DG has not been fully elucidated. Astrocytes with radial morphology reminiscent of embryonic radial glia are NSCs in the SGZ (Seri et al., 2001), but recent data suggest that a second morpholog-ically distinct NSC population exists (Steiner et al., 2006; Suh et al., 2007). Although both express Sox2 and brain lipid binding protein (BLBP), the exact functions and fates of radial and nonradial putative NSCs in the DG are not known.

Generation of DG granule neurons is a dynamic and regulated process. Under physiological conditions, neurogenesis is controlled at the level of proliferation, differentiation, and survival of newly generated cells and can be modulated by pathologies (Zhao et al., 2008). Whereas physical activity and seizures strongly increase proliferation (Fabel and Kempermann, 2008; Parent and Murphy, 2008), aging is associated with an exponen-tial decrease in DG neurogenesis (Ben Abdallah et al., 2010; Kempermann et al., 1998; Kuhn et al., 1996; Steiner et al., 2008). There is evidence that different neurogenic stimuli and pathological situations affect cells at different stages of neuro-genesis (Kronenberg et al., 2003; Petrus et al., 2009; Steiner et al., 2008). However, whether distinct NSC populations respond differently to these external stimuli, the mechanisms involved and whether they share common molecular pathways for maintenance and differentiation are unknown.

Notch activity is fundamental for maintaining embryonic NSCs in an undifferentiated state (Louvi and Artavanis-Tsakonas, 2006), suppressing proneural gene expression (Lu¨tolf et al., 2002), and supporting progenitor survival ( Androutsellis-Theoto-kis et al., 2006). In addition, Notch receptors regulate cell fate by promoting astrocyte differentiation, inhibiting oligodendrocyte


maturation, and regulating neuroblast migration and neurite morphology (Breunig et al., 2007; Louvi and Artavanis-Tsakonas, 2006; Sestan et al., 1999). These diverse roles of Notch are a combination of canonical and noncanonical pathways. The canonical Notch signal links the transmembrane receptors to a nuclear CSL (RBP-J in mice) transcriptional complex (Mumm and Kopan, 2000). Although both NSCs and IPs respond to Notch activation in the developing neural tube, the canonical pathway (Mizutani et al., 2007) and the Notch target gene Hes5 in particular (Basak and Taylor, 2007) are restricted to embryonic NSCs.

Notch1 has been implicated in adult hippocampal neurogene-sis and modulates dendritic arborization in newborn neurons (Breunig et al., 2007; Sibbe et al., 2009). However, it is unknown whether different pathways downstream of Notch are involved in these diverse effects at specific stages of hippocampal neuro-genesis. Here we show that canonical Notch signaling (Hes5::GFP) is restricted to NSCs functionally segregating the most primitive Sox2 progenitors and distinguishing them from IPs (Type-2 cells), neuroblasts (Type-3 cells), and neurons. Surprisingly, multiple morphologically distinct NSC (Type-1 cells) populations in the DG show dependence on canonical Notch signaling and are differentially regulated by physiological and pathological stimuli. These findings suggest that at least two distinct NSC populations in the SGZ share Notch dependency but show divergent properties in the normal and pathological hippocampus.


Hes5::GFP Expression Distinguishes Early Progenitors from More Differentiated Cells

Transgenic mice expressing GFP under the control of regulatory elements of Hes5 report canonical Notch signaling and selec-tively mark NSCs in the embryo (Basak and Taylor, 2007). Notch1 is expressed by both progenitors and differentiated cells in the DG (Figure S1A available online). We examined

Hes5::GFP+(GFP+) cells in the DG to determine the cellular spec-ificity of canonical Notch signaling. GFP+cells and Hes5 are

localized mainly to the SGZ, consistent with being progenitors within the neurogenic niche (Figure 1A;Figures S1A–S1D;Stump et al., 2002).

Radial astrocytes spanning the granule cell layer are putative DG NSCs (Filippov et al., 2003; Seri et al., 2001). However, SGZ GFP+cells have both radial (54%) and nonradial (46%) morphol-ogies, indicating that the Hes5+ population is heterogeneous

(Figures 1B and 1C). Recent data indicate that Sox2+cells in the SGZ with horizontal processes also display NSC characteris-tics (Steiner et al., 2006; Suh et al., 2007). In support, most

Hes5::GFP+cells express the progenitor markers Sox2 (82%), BLBP (74%), and Nestin and the astroglial protein glial fibrillary acid protein (GFAP; 66%) and this independent of their morphology (Figures 1D–1H;Figures S1B–S1D andTable S1). Thus, Hes5::GFP+cells express markers in common with NSC and display radial and horizontal morphologies suggesting at least two populations of progenitors in the SGZ. Interestingly, GFP+cells constituted only a fraction of the SGZ Sox2-, BLBP-,

and GFAP-expressing cells, indicating restriction of Hes5 expres-sion within these populations (Figure 1I;Figures S1B and S1C).

Transcription factor expression in the DG, including the pro-neural gene Mash1 (Ascl1), defines the progression from stem cells to IPs (Kim et al., 2008; Roybon et al., 2009). Canonical Notch signaling represses proneural gene expression via Hes proteins (Louvi and Artavanis-Tsakonas, 2006), and Hes5::GFP and Mash1 expression in the DG are largely complementary. Only a few horizontal cells coexpressed Mash1 and display low levels of GFP, suggesting either weak expression or GFP stability (Figures 1J and 1K). Mash1+IPs still express Sox2 (not shown). These data suggest that high-level Mash1 expression defines a narrow transition window from the Hes5+NSC state to IPs, possibly controlling progression of horizontal cells through the neurogenic program.

Supporting that Hes5::GFP marks putative NSCs in the granule neuron lineage, Dcx+ neuronal precursors (Type-2b and Type-3 cells) and NeuN+neurons do not express Hes5::GFP (Figure 1L;Table S1). However, Dcx+neuroblasts were often clustered with GFP+progenitors (Figure 1L). GFP+cells in the SGZ do not express the mature astrocyte protein S100b and only rare GFP+S100b+cells can be found at the granule/molec-ular layer interface (Figure 1M, asterisk). Collectively, these data indicate that Hes5 expression in the SGZ marks morpholog-ically distinct subpopulations of early progenitor cells and distin-guishes them from IPs, neuroblasts, and neurons (Figure 1N).

Hes5::GFP Expression in Early SGZ Progenitors Subdivides Sox2+Progenitor Cells into Defined Pools

Ninety-six percent of the proliferating cells (PCNA+) in the SGZ

express Sox2 and 70% BLBP (Figures 2A–2C). However, although >80% of Hes5::GFP+cells are Sox2+(Figure 1H), only

some of these are in the cell cycle (PCNA+) and incorporate BrdU during a 1 day pulse; the rest are quiescent (PCNA ) (Figures 2D and 2E). We examined the proliferating GFP+cells and found that they account for 1/3 of the total mitotic SGZ cells (Figure 2F;Table S2) and express Sox2 and BLBP. In contrast, proliferating GFP cells constitute partially overlapping Sox2+, BLBP+, and Dcx+ populations (Figure 2G; Table S2). These

data suggest that Hes5::GFP marks a more restricted and undif-ferentiated population than Sox2 and BLBP (Figure 2G).

To follow the progeny of mitotic cells, we analyzed newborn cells at different time points after BrdU administration. A 2 hr BrdU pulse confirmed that GFP+cells are a significant proportion of the dividing cells in the DG, whereas proliferative Dcx+cells represent a smaller population and NeuN+neurons do not

prolif-erate. After a BrdU pulse and 5 day chase, the proportion of BrdU+GFP+cells decreased to 17% and to 7% after a 30 day

chase, at which point most BrdU-labeled cells had become neu-roblasts and neurons (Figure 2H;Table S3). These data support the hypothesis that GFP+progenitor cells are at the beginning of the neurogenic lineage and give rise to more committed cells.

Hes5::GFP+Cells Are Self-Renewing and Multipotent NSCs

DG NSCs self-renew and expand in vitro (Babu et al., 2007; Suh et al., 2007). We sorted adult DG cells based on GFP expression and found that they are <6% of the total (Figure 2I), which is in agreement with our in vivo analyses. GFP+cells grew in culture

and could be expanded >30 passages, demonstrating extended self-renewal capacity, and most expressed Sox2, Nestin, and


BLBP (Figures 2I–2M and 2P). Sorted GFP cells adhered to the culture substrate but did not expand even when plated at high density (Figure 2J; insert). Under differentiating conditions, GFP+cells gave rise to b-TubulinIII+neurons and GFAP+

astro-cytes (Figures 2N and 2O). Thus, Hes5::GFP+ cells include self-renewing DG NSCs with multipotent potential in vitro.

Hes5::GFP Expression and DG Neurogenesis Depend on Canonical Notch Signaling

GFP+cells display NSC characteristics in vivo and in vitro, sug-gesting that canonical Notch signaling is critical for DG NSCs. RBP-J is an indispensable mediator of canonical Notch signaling and Hes5 expression (de la Pompa et al., 1997). We inactivated

RBP-J in glial cells, including SGZ NSCs, by expressing CreERT2

from the astrocyte-specific glutamate transporter (GLAST) locus. Three weeks after RBP-J ablation, Hes5::GFP+ cells were almost undetectable in the SGZ and this was independent

of radial or horizontal morphology. Moreover, proliferation in the SGZ had been abolished (Figure S2). By following the progeny of recombined progenitors, we found a dramatic reduc-tion in neurogenesis (data not shown). This is in agreement with the previously reported role of Notch1 in the DG (Breunig et al., 2007) and confirms the specificity of the Hes5::GFP as a Notch reporter and marker of NSCs.

Hes5::GFP+Cells Include Active and Quiescent NSCs

Our results indicate that DG NSCs can be distinguished by the expression of Hes5 and both radial and horizontal NSCs depend on Notch/RBP-J signaling. Previous work suggests that radial progenitors in the SGZ are mostly quiescent (Steiner et al., 2006; Suh et al., 2007); however, it is not clear whether all hori-zontal progenitors are dividing or whether they can also be quiescent. Only 13% of GFP+cells express PCNA and incorpo-rated BrdU during a 1-day pulse (Figures 3A–3D;Table S2) and

Figure 1. Hes5::GFP Expression Identifies a Subpopulation of Progenitor Cells (A) In adult Hes5::GFP transgenic mice, GFP+

cells are found in the SGZ of the DG.

(B) SGZ Hes5::GFP+

cells have both radial (arrow) and horizontal (arrowhead) morphologies. (C) Percentage of radial versus horizontal Hes5:: GFP+


(D–G) Hes5::GFP+ cells express the progenitor markers Sox2 (D), BLBP (E), GFAP (F), and Nestin (G). Both radial (arrow) and horizontal (arrowhead) Hes5::GFP+

cells express Sox2 (D). (H) Most Hes5::GFP+

cells express progenitor markers.

(I) Only60% of the progenitor pool in the SGZ expresses Hes5::GFP, whereas the rest of the progenitor cells are GFP .

(J and K) Hes5::GFP and Mash1 expression (arrow) are largely complementary. (J) Most GFP+cells are Mash1-negative. Occasionally Hes5::GFP and Mash1 colocalize in horizontal (arrowhead) (J) but not in radial cells, and usually corresponds to Mash1+

GFP low expression (K).

(L) Hes5::GFP never colocalizes with Dcx or NeuN, but Dcx+

neuroblasts cluster next to GFP+

cells. (M) Hes5::GFP does not colocalize with the astrocytic marker S100b in the SGZ. Some GFP+


cells were found at the border between the granule cell layer and the molecular layer (asterisk).

(N) Summary of the marker expression analysis. Hes5::GFP expression identifies an early subpop-ulation of progenitor cells. Type-1 refers to cells expressing stem/progenitor markers, indepen-dent of their morphology. Type-2a cells retain progenitor markers but downregulate GFAP and Hes5::GFP and upregulate Mash1. Type-2b cells show neuronal determination markers like Dcx but retain low levels of the progenitor markers Sox2 and BLBP. Type-3 cells are Dcx+


SGZ, subgranular zone; GCL, granule cell layer; ML, molecular layer.

Error bars indicate SD. Numbers are listed inTable S1. Scale bars represent 50 mm in (A) and 10 mm in (B), (D)–(G), (J), (L), (M).


proliferation correlated preferentially with a horizontal mor-phology (16% dividing and 84% quiescent; Figure 3C). In contrast, very few GFP+ radial cells proliferate under normal conditions (4%; Figure 3C). Thus, we have identified three Hes5+NSC populations: (1) a largely quiescent radial population; (2) a quiescent horizontal population; and (3) an actively prolifer-ating horizontal population (Figure 3E).

S phase label retention has been used to identify stem cells with prolonged cell cycle reentry times and is often referred to as a stem cell characteristic (Cotsarelis et al., 1990; Tumbar et al., 2004; Zhang et al., 2003). We performed a cumulative 15 day BrdU labeling and found that BrdU-labeled cells increased to 26% of the GFP+cells, indicating that mitotically inactive cells

must have entered the cell cycle during this period (Figures 3A, 3F, and 3G;Table S2). The proportion of the GFP+ cells that

were PCNA+BrdU+ did not change during this time and GFP+BrdU+PCNA cells accumulated (Figure 3G; Table S2),

so mitotically active (PCNA+) GFP+cells divide but can exit the cell cycle to become PCNA and quiescent. To confirm that these

cells were indeed dividing slowly, we administered BrdU for 15 days followed by a 30 day chase. GFP+label-retaining cells were found in the SGZ. Some were PCNA+ and had either remained in the cell cycle but divided infrequently with a pro-tracted cycle length or exited the cell cycle and reentered later (Figure 3H). Hence, Hes5::GFP+NSCs can be active or quiescent.

Even slow dividing NSCs should go through multiple rounds of cell cycle. To check for S phase reentry, we used a double thymidine analog incorporation assay and administered chloro-deoxyuridine (CldU) and iodochloro-deoxyuridine (IdU) 6 days apart (Figure S3A;Breunig et al., 2007). CldU IdU double-positive cells indicated that label-retaining cells divide slowly or infrequently, entering S phase on day 0 and day 6 (Figure S3B). Most GFP+CldU-retaining cells had a horizontal morphology (92%); however, only a few had reentered the S phase on day 6 and incorporated IdU (Figures S3C and S3D). Confirming our previous findings, radial CldU+cells that also incorporated IdU

or expressed PCNA on day 6 were extremely rare, precluding quantification (Figures S3E and S3F). Hence, some Hes5::GFP+

Figure 2. Hes5::GFP-Expressing Cells Pro-liferate In Vivo and Are Self-Renewing and Multipotent Colony-Forming Cells In Vitro (A and B) Most proliferating cells (PCNA+

) express Sox2 (A) and BLBP (B).

(C) Most proliferating cells in the SGZ express Sox2 and BLBP and only1/4 express Dcx. (D–F) Some Hes5::GFP+

cells proliferate, but the majority of the proliferating SGZ progenitors are GFP . Some Hes5::GFP+

cells express PCNA (D) and incorporate BrdU (E). To label fast-dividing cells in vivo, BrdU was administered for 1 day in the drinking water prior to sacrifice (asterisk) (F). Only1/3 of the proliferating cells express Hes5 (F). Arrow and arrowheads point to proliferating radial and horizontal cells, respectively. Note that not all GFP+

cells are proliferating (D and F). (G) Summary of the marker expression analysis. Hes5::GFP expression identifies a restricted and undifferentiated population among the prolifer-ating SGZ cells.

(H) Percentage of newly generated cells that express Hes5::GFP, Dcx, and NeuN, 2 hr, 5 days, and 30 days after BrdU administration. BrdU labeling paradigms and the point when mice were sacrificed (asterisk). Mice received 1 or 3 intraperitoneal injections of BrdU or BrdU was administered for 15 days in the drinking water. Note that newly generated cells downregulate GFP and upregulate neuronal markers during differentiation.

(I) Scheme of the sorting of DG Hes5::GFP+

cells and the colony-forming assay. FACS-sorted GFP+cells comprised 5.8% of total cells and included all of the colony-forming cells. (J) Hes5::GFP+

cells can be expanded and passaged (P) in vitro.

(K–M) Sorted Hes5::GFP+

cells maintain progen-itor marker expression over multiple passages in vitro.

(N and O) Hes5::GFP+

cells can differentiate into both neurons and astrocytes in vitro.

(P) Three independent lines of sorted DG progenitors (isolation 1–3) derived from Hes5::GFP+

cells show comparable expression of progenitor markers. Control, progenitor cell cultures from unsorted DG cells.


NSCs undergo more than one round of cell divisions (at day 0 and day 6) without diluting the CldU label.

We analyzed the kinetics of GFP+cells that leave or reenter the cell cycle over time. Proliferating cells were labeled with CldU and allowed to exit S phase during a 1 day chase period ( Fig-ure 3I). Subsequently, we followed the mitotic status of the GFP+cells via a cumulative IdU label for 10 days and PCNA

expression. Only 0.3% of radial GFP+ cells had incorporated CldU+at day 2, confirming their relative quiescence. By contrast,

5% of the horizontal GFP+cells had incorporated CldU (Figures 3I–3L). We focused on horizontal cells in the subsequent anal-yses. The proportion of CldU+GFP+cells decreased with time, indicating that these cells either underwent multiple rounds of cell cycle, diluting the CldU over time, or differentiated and became GFP (Figure 3L;Table S4). However, some horizontal GFP+cells retained CldU even 11 days after labeling (0.8% at

day 11;Figure 3L). This suggests that they had either left the cell cycle or divided infrequently. To examine the dynamics of cell cycle exit, we quantified CldU+ cells in combination with PCNA expression. All CldU-retaining GFP+cells were PCNA+

and still in the cell cycle after 2 days, but most were PCNA by day 11 (Figures 3J and 3L;Figure S3G). Thus, Hes5::GFP+cells

can go from an active proliferating to a quiescence state. These data provided snapshots of the mitotic status of GFP+ cells. We quantified S phase reentry by examining CldU IdU double-labeling. 70% of the CldU+GFP+ cells underwent a second S phase between day 1 and day 2 (3.3% of all horizontal cells; Figures 3K and 3L). CldU+IdU+GFP+ cells decreased between day 2 and 11 whereas IdU+GFP+cells increased up

to day 11 (17%). This suggests that some GFP+cells continued to proliferate and diluted out the CldU (Figure 3L;Figure S3I). However, some GFP+ CldU label-retaining cells incorporated IdU between day 6 and day 11 (Figure 3L), indicating that they divided, although infrequently enough not to dilute the CldU.

As BrdU incorporation during 24 hr and PCNA label the same pool of cells (Figures 3A and 3B), mitotically active GFP+cells

enter S phase within a 24 hr period. Interestingly, the proportion of GFP+cells that had incorporated IdU+at day 2 was higher than

the CldU+ proportion at this point (Figure 3L). Therefore, the IdU+CldU GFP+ cells were quiescent during the CldU pulse

Figure 3. Hes5::GFP Cells Are Segregated into Quiescent and Active NSCs

(A) Different BrdU labeling paradigms were used to identify fast or slowly dividing populations and label-retaining cells. Asterisks indicate time point of sacrifice.

(B) Only a subpopulation of Hes5::GFP+

cells proliferate.

(C) Most proliferating Hes5::GFP+cells have a hori-zontal morphology.

(D) Examples of radial (arrow) and horizontal (arrowhead) Hes5::GFP+

proliferating cells. (E) Hes5::GFP+

cells can be subdivided into quies-cent radial, quiesquies-cent horizontal, and active hori-zontal subpopulations. Numbers are expressed as percent of the total Hes5::GFP+

population. (F and G) 26% of Hes5::GFP+

cells are labeled by BrdU administration for 15 days. 10% of Hes5:: GFP+cells that incorporated BrdU exit the cell cycle and 16% still express PCNA.

(H) 30 days after BrdU administration, some Hes5::GFP+

BrdU label-retaining cells reenter the cell cycle and express PCNA.

(I) CldU, IdU labeling paradigm to address cell cycle dynamics. Asterisks indicate time point of sacrifice.

(J and K) Some CldU-labeled Hes5::GFP+

cells remain in the cell cycle (PCNA+

) (J) and reenter S phase to incorporate IdU (K).

(L) Analysis of Hes5::GFP cell cycle dynamics. Some GFP+

cells retain CldU and can exit and reenter the cell cycle (CldU+

/PCNA+ and CldU+ / IdU+ ). IdU+ GFP+

cells are more numerous than CldU+GFP+cells at day 2.

(M) Infusion of the antimitotic AraC drastically reduces cell proliferation in the SGZ. Proliferation is restored after AraC removal.

(N and O) Proliferation of radial (N) and horizontal (O) Hes5::GFP+after AraC removal.

(P) Horizontal Hes5::GFP+

constitute the majority of proliferating cells during regeneration. Error bars indicate SD. Numbers are listed in Tables S2 and S4. Scale bars represent 10 mm.


and entered the cell cycle between day 1 and day 2 (Figure 3L). Together, all these data and the observation that GFP+cells can incorporate both CldU and IdU administered with a 6 day temporal separation (Figures S3A–S3E) suggest that some

Hes5::GFP+cells can exit the cell cycle and reenter later.

A functional readout of slow-dividing cells in the adult brain is the ability to survive treatment with antimitotic drugs such as Cytosine b-D-arabinofuranoside (AraC) (Doetsch et al., 1999; Seri et al., 2001). Intracranial administration of AraC (Figure S3L) reduced BrdU incorporation in the DG, indicating that actively dividing cells had been eliminated (Figure 3M). Radial and hori-zontal GFP+cells survived the treatment, confirming a long cell

cycle or quiescence (Figure S3M). Upon removal of AraC, GFP+cells entered the cell cycle within 12 hr but their density did not change significantly during regeneration (Figures 3N and 3O;Figure S3M). Thus, GFP+cells participate in regenera-tion of the neurogenic SGZ and their number remains constant, probably undergoing asymmetric divisions to generate GFP cells. Intriguingly, although both radial and horizontal cells survived the AraC treatment, most of the GFP+cells that prolifer-ated during regeneration had a horizontal morphology ( Fig-ure 3P). Therefore, the horizontal GFP+population in the SGZ includes slowly dividing regenerative NSCs. Radial GFP+NSCs

Figure 4. Running Induces Activation of Quiescent Radial Cells

(A) The established protocol used for exercise-induced hippocampal neurogenesis. Mice had access to a running wheel for 12 days and were then sacrificed (asterisk).

(B and C) Representative images of Hes5::GFP+

cells and PCNA+

proliferating cells in control and running mice. The density of proliferating cells is increased after running (arrows).

(D) The density of dividing cells significantly increases in response to exercise.

(E) The density of dividing Hes5::GFP+

cells signif-icantly increases in running mice.

(F) The density of dividing radial, but not horizontal, Hes5::GFP+

cells significantly increases after exercise.

(G) Representative image of a radial Hes5::GFP+

proliferating cell (arrow). (H) The density of Hes5::GFP+

cells does not significantly change after running.

(I) The proportion of horizontal versus radial Hes5::GFP+

cells does not change.

(J) The density of horizontal and radial Hes5::GFP+ cells is not affected by exercise.

t test: *p < 0.05, **p < 0.01. Error bars represent SD. Numbers are listed inTable S5. Scale bars represent 50 mm in (B) and (C) and 10 mm in (G).

remain mostly quiescent after AraC treat-ment and are not activated by injury.

Quiescent Radial NSCs Respond to Physiological Neurogenic Stimuli

Adult neurogenesis is dynamically regu-lated by a variety of physiological and pathological factors (Zhao et al., 2008). However, it is not always clear whether changes in neurogenesis are a NSC response or rather adaptations in proliferation and survival of other cell types or a combination of these effects. Because of the overlapping expression of current lineage markers in the DG (Sox2, GFAP, Nestin, Mash1, Dcx, and NeuN;Figure 1N) and the unresolved identity of the NSC, these questions remain open. We tested whether the Notch-depen-dent NSCs in the SGZ respond directly to known modulators of adult neurogenesis and whether quiescent and active NSC subpopulations behave differently.

Physical activity enhances generation of DG neurons by inducing proliferation of radial Sox2+ progenitor cells and neuronal precursors (Kronenberg et al., 2003; Steiner et al., 2008; Suh et al., 2007; van Praag et al., 1999b). In previous experiments, we focused on Nestin::GFP expression and did not find a specific effect on radial Type-1 cells after exposure to physiological stimuli (Steiner et al., 2008). Here we focused on the Hes5+subpopulations and, in agreement with previous data, found that PCNA is increased and more GFP+cells

prolif-erate in runners (Figures 4A–4E). Thus, Notch-dependent NSCs proliferate in response to physical activity (Figure 4E;

Figure S4A andTables S5 and S6). Interestingly, the different GFP+ NSC subpopulations responded differently to exercise.


The proportion of PCNA+horizontal Hes5+cells did not change

but running induced more quiescent radial Hes5+NSCs to enter the cell cycle (Figure 4F;Figure S4B). In contrast to sedentary animals, the majority of the proliferating GFP+cells after physical exercise are radial (Figure 4F; Figure S4B) and the ratio of radial to horizontal or the density of GFP+cells did not change,

suggesting that running promotes asymmetric cell division of radial NSCs to generate GFP progeny (Figures 4H–4J). This suggests that running specifically recruits quiescent radial GFP+NSCs to the active pool without affecting the horizontal


Quiescence of Horizontal NSCs Accounts for Impaired Neurogenesis in Aging

Hippocampal neurogenesis is greatly reduced in aging (Ben Abdallah et al., 2010; Jessberger and Gage, 2008; Kuhn et al., 1996) and this might correlate with some age-associated cogni-tive deficits. It is not known whether the decrease in neurogen-esis with age is the result of a loss of NSCs. Proliferation is dramatically decreased in the SGZ of aged (1–2 years) versus young (2 months) mice (Figures 5A–5D), and proliferation of GFP+cells is also heavily impaired, indicating that NSCs are

directly affected (Figure 5E;Figure S5A andTables S5 and S6). Proportionally, proliferation of radial GFP+NSCs is significantly

reduced in old mice even compared to the already low levels in young mice (Figure 5F;Figure S5B). Proliferation of horizontal GFP+cells drops 4-fold with age, which accounts for the overall reduced proliferation of NSCs (Figure 5F; Figure S5B). We

addressed whether GFP+NSCs might be lost with increasing

age, but the density of GFP+cells in the SGZ of aged animals is only slightly reduced compared to the young (Figures 5H and 5J;Table S5). Although most radial and horizontal GFP+cells remain in old mice and still express Sox2 (Figure 5C), the prolif-erative subpopulations are almost completely lost (Figure 5F). Thus, reduced neurogenesis with age is not due to a complete loss of Notch-dependent NSCs but rather their transition to a quiescent state.

Seizures Activate Both Radial and Horizontal Quiescent NSCs

Acute seizures dramatically induce abnormal production of DG neurons, which has been suggested to contribute to chronic epilepsy (Parent, 2007; Scharfman and Gray, 2007). All progen-itor cell populations in the SGZ seem to be affected by seizures, including the radial (Type-1) cells (Hu¨ttmann et al., 2003; Steiner et al., 2008) and disproportionately the neuroblasts (Type-3 cells) (Jessberger et al., 2005). We addressed whether the Hes5::GFP+ NSCs subpopulations are responsible for seizure-induced increases in neurogenesis.

We administered kainic acid (KA) to Hes5::GFP+ mice, an

established experimental model of temporal lobe epilepsy, and observed increased proliferation in the SGZ 4 days later (Figures 6A–6D). Compared to controls, a greater proportion of the GFP+cells were proliferating, indicating that activation of Hes5+

NSCs contributes to increased neurogenesis after seizures (Figure 6E; Figure S6A and Tables S5 and S6). We analyzed

Figure 5. Loss of Horizontal Active NSCs during Aging

(A and B) Representative images of Hes5::GFP+

cells and PCNA+

proliferating cells in young and aged mice. The density of proliferating cells is decreased during aging (arrows).

(C) Hes5::GFP+

cells express the progenitor marker Sox2 in aged mice.

(D) The density of dividing cells is dramatically decreased in aged mice.

(E) The density of dividing Hes5::GFP+

cells is also dramatically decreased in the SGZ of aged mice. (F) Proliferating radial and horizontal Hes5::GFP+

are decreased with age. Loss of proliferating horizontal cells accounts for the overall reduced proliferation of GFP+


(G) Representative image of a horizontal Hes5:: GFP+

proliferating cell.

(H) The density of Hes5::GFP+cells is slightly but significantly reduced in the SGZ of aged mice. (I) The proportion of horizontal versus radial Hes5::GFP+

cells does not change with age. (J) There is a nonsignificant trend to reduction in density of horizontal and radial Hes5::GFP+cells in the SGZ of old mice.

t test: *p < 0.05, ***p < 0.001. Error bars represent SD. Numbers are listed inTable S5. Scale bars represent 50 mm in (A) and (B) and 10 mm in (C) and (G).


the subpopulations of GFP+cells and found that a fraction of radial cells had begun proliferating. Surprisingly, and in contrast to the running paradigm, seizures dramatically increased prolif-eration of horizontal GFP+cells (Figure 6F;Figure S6B). Thus,

although both radial and horizontal cells responded and divided more frequently, horizontal cells still constituted the majority of proliferating GFP+NSCs after KA (Figure 6F). Interestingly, the density of GFP+cells in the SGZ increased after KA (Figure 6H). We interpret this to be a symmetric expansion of the NSC pool that remains to be shown. Moreover, the ratio of horizontal to radial cells increased significantly (Figures 6I and 6J). Thus, seizures increase horizontal Notch-dependent NSCs, recruiting quiescent cells to expand the active horizontal subpopulation.

Quiescent NSCs Can Be Activated in the Aged Hippocampus

The degree of plasticity of the aged DG to reactivate neurogen-esis is not clear. The DG of aged animals may be able to selec-tively respond to different neurogenic stimuli (Hattiangady et al., 2008; Kempermann et al., 2002; Rao et al., 2005, 2008; van Praag et al., 2005). We showed that reduced neurogenesis in old mice did not correlate with a complete loss of SGZ NSCs but a transition to a quiescent state. Thus, we addressed whether the remaining quiescent NSCs could be reactivated.

Figure 6. Expansion of the Active NSC Pool after Seizures

(A) To induce seizures, KA was injected i.p. into mice. Seizures developed within 45 min after injec-tion and spontaneously stopped within 2–3 hr. Mice were sacrificed 4 days later (asterisk). (B and C) Representative images of Hes5::GFP+

cells and PCNA+

proliferating cells in control and KA-induced mice. The density of proliferating cells is increased after seizures (arrows).

(D) The density of dividing cells is significantly increased in the SGZ of KA-induced mice. (E) The density of dividing Hes5::GFP+

cells is also significantly increased by KA induction. (F) The density of dividing radial and horizontal Hes5::GFP+

cells significantly increases after seizures. The horizontal proliferating population is 3-fold larger than the radial proliferating popula-tion after seizures.

(G) Representative image of radial and horizontal Hes5::GFP+

proliferating cells in KA-induced mice. (H) The density of Hes5::GFP+

cells is significantly increased in the SGZ of KA-induced mice. (I) The proportion of horizontal versus radial Hes5:: GFP+

cells is shifted toward horizontal cells by KA induction.

(J) The density of horizontal but not radial Hes5::GFP+

cells is increased after seizures. t test: *p < 0.05, **p < 0.01, ***p < 0.001. Error bars indicate SD. Numbers are listed inTable S5. Scale bars represent 50 mm in (B) and (C) and 10 mm in (G).

Because KA induced seizures activate both radial and horizontal NSCs in young mice, we tested this paradigm in aged mice (Figure 7A). Four days after KA injection, proliferation in the aged SGZ increased to the levels measured in young mice (Figures 7B–7D) and many more GFP+cells were proliferating compared to

con-trol animals (Figure 7E;Figure S7A andTables S5 and S6). Both radial and horizontal GFP+cells responded to seizures, entering

a proliferative state (Figure 7F;Figure S7B). Interestingly, like in young mice, horizontal cells constituted the vast majority of proliferating NSCs cells after KA treatment, supporting that horizontal cells are a more plastic population than radial cells in aged mice (Figure 7F). However, and in contrast to young mice, neither the density of GFP+cells nor the density of GFP+ horizontal cells in the aged SGZ increased after seizures (Figures 7H–7J). These data suggest that the quiescent state of Notch-dependent NSCs in the aged mammalian brain is reversible but that NSCs fail to expand as efficiently as in the young brain.


Neurogenesis in the adult brain has been demonstrated in several mammalian species, including humans (Amrein and Lipp, 2009), but the precise identity of adult NSCs and many of the molecular mechanisms regulating their fate have not been fully elucidated. Previous work demonstrated the requirement of Notch1 in multiple steps of adult hippocampal neurogenesis (Breunig et al., 2007). Accordingly, Notch1 receptor expression


and cleavage was found both in progenitors and neurons of the DG (Breunig et al., 2007). By using a transgenic approach, we show that NSCs in the adult DG depend on canonical Notch/ RBP-J signaling and express Hes5 whereas IPs and neuronal precursors do not. Canonical Notch/RBP-J signaling and Hes5 expression label a fraction of the Sox2+progenitors in the DG, and based on their in vivo and in vitro characteristics, we define the Hes5::GFP+cells as NSCs.

Neurogenesis in the adult DG is a multistep process dynami-cally regulated by several factors (Jessberger and Gage, 2008; Zhao et al., 2008). Each of these modulators may act at one or multiple levels of the neurogenic lineage. There are hints that some regulate proliferation of early progenitors including NSCs (Hu¨ttmann et al., 2003; Steiner et al., 2008; Suh et al., 2007). We demonstrate that Hes5::GFP+ NSCs directly respond to modulatory stimuli, showing that reactive neurogenesis acts very early in the neurogenic lineage (Graphical Abstract available online). This does not exclude that further regulation occurs in more differentiated cell types. Consistent with the theory, we show that the NSC pool labeled by Hes5::GFP is rather quiescent (PCNA and does not incorporate S phase label), with only a fraction of cells proliferating at any given time. However, the magnitude of the active NSC population dynamically changes according to demand. This can be achieved through different mechanisms in different paradigms, as a result of recruitment

of quiescent cells to the active pool or symmetric expansion of the active pool itself (Graphical Abstract). A physiological regulator of neurogenesis like exercise activates some quiescent radial Hes5+NSCs. This is intriguing because, based on their proliferation status, these cells seem not to contribute majorly to neurogenesis in the DG under normal conditions. These find-ings support the view that radial NSCs are a reserve pool of cells that can be called upon to increase the neurogenic process in response to changes in conditions, but this remains to be shown by genetic labeling.

On the other hand, seizures activate quiescent NSCs and also expand the active neurogenic population. Thus, we propose that the NSC pools themselves are subjected to dynamic regulation and expansion. These dramatically different modes of regulating neurogenesis could be exploited to increase neurogenesis to repair brain damage or neuronal loss. In this respect, a funda-mental question is: how much plasticity does the aged brain retain? Because DG neurogenesis actively contributes to memory and cognitive performance, which are heavily impaired during aging (Clelland et al., 2009; Garthe et al., 2009; Shors et al., 2001), it is critical to understand the models and mecha-nisms of generating new neurons in situ. It is likely that a combi-nation of factors such as reduced proliferation, impaired neuronal differentiation, and alterations in the neurogenic niche are responsible for loss of neurogenesis. Although data suggest

Figure 7. Activation of the Quiescent NSC Pool in Aged Mice after Seizures

(A) To induce seizures, KA was injected i.p. into aged mice. Mice were sacrificed 4 days later (asterisk).

(B and C) Representative images of Hes5::GFP+

cells and PCNA+

proliferating cells in control and KA-induced old mice. The density of proliferating cells in the SGZ is increased after seizures (arrows).

(D) The density of dividing cells significantly increases in aged KA-induced mice.

(E) The density of proliferating Hes5::GFP+

cells significantly increases in KA-induced old mice. (F) The density of dividing radial and horizontal Hes5::GFP+

cells significantly increases after seizures. The horizontal proliferating population is 3-fold larger than the radial proliferating popula-tion after seizures.

(G) Representative image of horizontal Hes5:: GFP+

proliferating cells in aged KA-induced old mice (arrowheads).

(H) In contrast to younger mice, the density of Hes5::GFP+cells does not change in KA-induced old mice.

(I) The proportion of horizontal versus radial Hes5::GFP+

cells is not significantly shifted in KA-induced old mice.

(J) The density of horizontal and radial Hes5::GFP+ cells does not change after seizures during aging. t test: *p < 0.05, **p < 0.01. Error bars indicate SD. Numbers are listed inTable S5. Scale bars repre-sent 50 mm in (B) and (C) and 10 mm in (G).


that a net increase in new neurons cannot be induced in the aged brain, even by strong activators of neurogenesis like injury and seizures (Hattiangady et al., 2008; Rao et al., 2008), other reports indicated that generation of new neurons can be partially restored by physical activity and environmental enrichment (Kempermann et al., 2002; van Praag et al., 2005). However, it is not known whether alterations or loss of the NSCs pool is one of the factors that accounts for reduced neurogenesis (Hattiangady and Shetty, 2008). Our data show that reduced proliferation of NSCs has a major effect on neurogenesis in aging, and the reduction in NSCs is mostly attributable to loss of the active pool whereas the majority of quiescent NSCs are still present (Graphical Abstract). Based on continued Notch activity in the SGZ NSCs, it is not a lack of canonical Notch signaling that causes reduced neurogenesis but an imbalance between maintenance and differentiation.

Based on our thymidine analog/PCNA labeling experiments, active NSCs have a cell cycle time <24 hr (Figures2F,3B, and 3G) and imply that active NSCs can leave and reenter the cell cycle at a later stage (Figures 3F–3L). Therefore, the mitotic status of neurogenic NSCs is not fixed and they can dynamically shuttle between the active and quiescent pools. Recent retroviral labeling data suggested that proliferating cells predominantly generate neurons and rarely give rise to quiescent radial cells (Suh et al., 2007) and therefore, we propose that horizontal NSCs retain their morphological characteristics and just change their proliferative status. Because of the consistent and con-served activity of Notch in both quiescent and actively dividing NSCs in young animals, it is unlikely that Notch signaling alone is modulating the proliferative state of NSCs in the DG. This is reminiscent of skeletal muscle where satellite stem cells remain in old mice but fail to be activated because of imbalance between Notch and TGF-b signaling resulting in ineffective regeneration of muscle fibers after lesion (Carlson et al., 2008).

We show that seizures can activate quiescent NSCs and induce them to proliferate, indicating that they can be targets to reinstate neurogenesis. However, reactivated quiescent NSCs fail to expand symmetrically in the aged DG and inducing prolifer-ation is not sufficient for productive neuron formprolifer-ation (Rao et al., 2008). Thus, additional factors promote neuronal differentiation and survival and these are also altered in the aged brain. It will be critical to determine which pathways are modulated or lost with age and those that are activated by seizures.

We confirm with independent approaches that a Sox2+

hori-zontal cell population in the SGZ contains active NSCs. Whether distinct NSC populations have different requirements for their maintenance and differentiation and different morphologies truly reflect different functions is unclear. We provide evidence that NSCs with radial and horizontal morphologies share a common dependence on canonical Notch signaling and both express

Hes5. Unexpectedly, and despite these similarities, radial and

horizontal NSCs behave strikingly differently in different situa-tions. Horizontal NSCs are the active neurogenic population in the SGZ, probably giving rise to Mash1+IPs by asymmetric cell division and subsequently neuroblasts (Figure 1N). Hes5+

horizontal NSCs represent about 1/3 of all proliferating cells and this population is lost or exhausted with age.

A similar situation of dual NSC populations exist in the adult olfactory epithelium where basal cells can be recruited under

lesion conditions (Leung et al., 2007). A similar mechanism may be present in the adult DG, with horizontal active cells being the major neurogenic population under normal conditions and quiescent NSCs being selectively recruited to increase or replenish the active pool. In the future, it will be fundamental to elucidate the relationship between radial, horizontal, and hori-zontal active cells and whether these subpopulations are within the same lineage or act independently in parallel to produce differentiated progeny.


Animals, Administration of Thymidine Analogs, Tissue Preparation, and Immunhistochemistry

Mice were maintained according to institutional regulations under license numbers H-05/01, 0-06/02, G-09/18, G-09/19, and G-08/26 (Ethical Commis-sion Freiburg, Germany). BrdU, CldU, and IdU (Sigma) were administered in the drinking water or by i.p. injection at equimolar ratios (Supplemental Exper-imental Procedures). For tissue collection and histology, mice were deeply anesthetized and perfused with ice-cold 4% paraformaldehyde (PFA). Immu-nostaining of coronal sections was performed as described previously and antibodies used are listed inSupplemental Experimental Procedures.

FACS Sorting and Precursor Cell Cultures, RNA Isolation, RT-PCR, and PCR

The DG of 8-week-old Hes5::GFP mice were micro-dissected and mechani-cally dissociated. Cells were sorted by FACS and gating on GFP (wild-type levels) and GFP+

populations. The cells were cultured and RNA isolated and PCR performed as described in Supplemental Experimental Procedures. Three independent lines of hippocampal neural progenitors were derived from Hes5::GFP+

mice and analyzed.

Intracranial AraC Infusion for Regeneration, Running, and Induction of Epileptic Seizures

AraC treatment of 3-month-old mice is described inSupplemental Experi-mental Procedures. The running paradigm has been described previously (van Praag et al., 1999a, 1999b). Seizures were induced by i.p. administration of KA (Tocris Bioscience) (30 mg/kg body weight). Mice were sacrificed 4 days after KA injection.

Quantification and Statistical Analysis

Randomly selected, stained cells were analyzed with fixed photomultiplier settings on a Zeiss LSM510 confocal microscope (Zeiss). Data are presented as average percentages of colabeled cells. The number of Hes5::GFP+

and PCNA+cells in the SGZ was estimated with a 633 magnification objective. The area of the granule cell layer was measured with ImageJ software and used to estimate the number of labeled cells per mm2

. Statistical comparisons were conducted by two-tailed unpaired Student’s t test. Significance was established at p < 0.05. In all graphs, error bars are standard deviation (SD).


Supplemental Information includes Supplemental Experimental Procedures, seven figures, and six tables and can be found with this article online at doi:10.1016/j.stem.2010.03.017.


We thank Drs. T. Honjo and P. Soriano for mice, S. Jessberger and J. Gilthorpe for comments on the manuscript, members of the Taylor lab for helpful discus-sions, and Frank Sager for excellent technical assistance. S.L. and P.K. are IMPRS-MCB PhD students of the Faculty of Biology, University of Freiburg, Germany. This work was supported by the Deutsche Forschungsgemein-schaft (SPP1109; TA-310-1; TA-310-2).


Received: October 13, 2009 Revised: February 15, 2010 Accepted: March 18, 2010 Published: May 6, 2010


Amrein, I., and Lipp, H.P. (2009). Adult hippocampal neurogenesis of mammals: Evolution and life history. Biol. Lett. 5, 141–144.

Androutsellis-Theotokis, A., Leker, R.R., Soldner, F., Hoeppner, D.J., Ravin, R., Poser, S.W., Rueger, M.A., Bae, S.K., Kittappa, R., and McKay, R.D. (2006). Notch signalling regulates stem cell numbers in vitro and in vivo. Nature 442, 823–826.

Babu, H., Cheung, G., Kettenmann, H., Palmer, T.D., and Kempermann, G. (2007). Enriched monolayer precursor cell cultures from micro-dissected adult mouse dentate gyrus yield functional granule cell-like neurons. PLoS ONE 2, e388.

Basak, O., and Taylor, V. (2007). Identification of self-replicating multipotent progenitors in the embryonic nervous system by high Notch activity and Hes5 expression. Eur. J. Neurosci. 25, 1006–1022.

Ben Abdallah, N.M., Slomianka, L., Vyssotski, A.L., and Lipp, H.P. (2010). Early age-related changes in adult hippocampal neurogenesis in C57 mice. Neurobiol. Aging 31, 151–161.

Breunig, J.J., Silbereis, J., Vaccarino, F.M., Sestan, N., and Rakic, P. (2007). Notch regulates cell fate and dendrite morphology of newborn neurons in the postnatal dentate gyrus. Proc. Natl. Acad. Sci. USA 104, 20558–20563. Carlson, M.E., Hsu, M., and Conboy, I.M. (2008). Imbalance between pSmad3 and Notch induces CDK inhibitors in old muscle stem cells. Nature 454, 528–532.

Clelland, C.D., Choi, M., Romberg, C., Clemenson, G.D., Jr., Fragniere, A., Tyers, P., Jessberger, S., Saksida, L.M., Barker, R.A., Gage, F.H., and Bussey, T.J. (2009). A functional role for adult hippocampal neurogenesis in spatial pattern separation. Science 325, 210–213.

Cotsarelis, G., Sun, T.T., and Lavker, R.M. (1990). Label-retaining cells reside in the bulge area of pilosebaceous unit: Implications for follicular stem cells, hair cycle, and skin carcinogenesis. Cell 61, 1329–1337.

de la Pompa, J.L., Wakeham, A., Correia, K.M., Samper, E., Brown, S., Aguilera, R.J., Nakano, T., Honjo, T., Mak, T.W., Rossant, J., and Conlon, R.A. (1997). Conservation of the Notch signalling pathway in mammalian neurogenesis. Development 124, 1139–1148.

Doetsch, F., Garcı´a-Verdugo, J.M., and Alvarez-Buylla, A. (1999). Regenera-tion of a germinal layer in the adult mammalian brain. Proc. Natl. Acad. Sci. USA 96, 11619–11624.

Fabel, K., and Kempermann, G. (2008). Physical activity and the regulation of neurogenesis in the adult and aging brain. Neuromolecular Med. 10, 59–66. Favaro, R., Valotta, M., Ferri, A.L., Latorre, E., Mariani, J., Giachino, C., Lancini, C., Tosetti, V., Ottolenghi, S., Taylor, V., and Nicolis, S.K. (2009). Hippocampal development and neural stem cell maintenance require Sox2-dependent regulation of Shh. Nat. Neurosci. 12, 1248–1256.

Filippov, V., Kronenberg, G., Pivneva, T., Reuter, K., Steiner, B., Wang, L.P., Yamaguchi, M., Kettenmann, H., and Kempermann, G. (2003). Subpopulation of nestin-expressing progenitor cells in the adult murine hippocampus shows electrophysiological and morphological characteristics of astrocytes. Mol. Cell. Neurosci. 23, 373–382.

Garthe, A., Behr, J., and Kempermann, G. (2009). Adult-generated hippo-campal neurons allow the flexible use of spatially precise learning strategies. PLoS ONE 4, e5464.

Hattiangady, B., and Shetty, A.K. (2008). Aging does not alter the number or phenotype of putative stem/progenitor cells in the neurogenic region of the hippocampus. Neurobiol. Aging 29, 129–147.

Hattiangady, B., Rao, M.S., and Shetty, A.K. (2008). Plasticity of hippocampal stem/progenitor cells to enhance neurogenesis in response to kainate-induced injury is lost by middle age. Aging Cell 7, 207–224.

Hu¨ttmann, K., Sadgrove, M., Wallraff, A., Hinterkeuser, S., Kirchhoff, F., Steinha¨user, C., and Gray, W.P. (2003). Seizures preferentially stimulate proliferation of radial glia-like astrocytes in the adult dentate gyrus: Functional and immunocytochemical analysis. Eur. J. Neurosci. 18, 2769–2778. Jessberger, S., and Gage, F.H. (2008). Stem-cell-associated structural and functional plasticity in the aging hippocampus. Psychol. Aging 23, 684–691. Jessberger, S., Ro¨mer, B., Babu, H., and Kempermann, G. (2005). Seizures induce proliferation and dispersion of doublecortin-positive hippocampal progenitor cells. Exp. Neurol. 196, 342–351.

Kempermann, G., Kuhn, H.G., and Gage, F.H. (1998). Experience-induced neurogenesis in the senescent dentate gyrus. J. Neurosci. 18, 3206–3212. Kempermann, G., Gast, D., and Gage, F.H. (2002). Neuroplasticity in old age: Sustained fivefold induction of hippocampal neurogenesis by long-term environmental enrichment. Ann. Neurol. 52, 135–143.

Kempermann, G., Jessberger, S., Steiner, B., and Kronenberg, G. (2004). Milestones of neuronal development in the adult hippocampus. Trends Neuro-sci. 27, 447–452.

Kim, E.J., Battiste, J., Nakagawa, Y., and Johnson, J.E. (2008). Ascl1 (Mash1) lineage cells contribute to discrete cell populations in CNS architecture. Mol. Cell. Neurosci. 38, 595–606.

Kronenberg, G., Reuter, K., Steiner, B., Brandt, M.D., Jessberger, S., Yamaguchi, M., and Kempermann, G. (2003). Subpopulations of proliferating cells of the adult hippocampus respond differently to physiologic neurogenic stimuli. J. Comp. Neurol. 467, 455–463.

Kuhn, H.G., Dickinson-Anson, H., and Gage, F.H. (1996). Neurogenesis in the dentate gyrus of the adult rat: Age-related decrease of neuronal progenitor proliferation. J. Neurosci. 16, 2027–2033.

Kuwabara, T., Hsieh, J., Muotri, A., Yeo, G., Warashina, M., Lie, D.C., Moore, L., Nakashima, K., Asashima, M., and Gage, F.H. (2009). Wnt-mediated activa-tion of NeuroD1 and retro-elements during adult neurogenesis. Nat. Neurosci. 12, 1097–1105.

Leung, C.T., Coulombe, P.A., and Reed, R.R. (2007). Contribution of olfactory neural stem cells to tissue maintenance and regeneration. Nat. Neurosci. 10, 720–726.

Louvi, A., and Artavanis-Tsakonas, S. (2006). Notch signalling in vertebrate neural development. Nat. Rev. Neurosci. 7, 93–102.

Lu¨tolf, S., Radtke, F., Aguet, M., Suter, U., and Taylor, V. (2002). Notch1 is required for neuronal and glial differentiation in the cerebellum. Development 129, 373–385.

Mizutani, K., Yoon, K., Dang, L., Tokunaga, A., and Gaiano, N. (2007). Differen-tial Notch signalling distinguishes neural stem cells from intermediate progen-itors. Nature 449, 351–355.

Mumm, J.S., and Kopan, R. (2000). Notch signaling: From the outside in. Dev. Biol. 228, 151–165.

Parent, J.M. (2007). Adult neurogenesis in the intact and epileptic dentate gyrus. Prog. Brain Res. 163, 529–540.

Parent, J.M., and Murphy, G.G. (2008). Mechanisms and functional signifi-cance of aberrant seizure-induced hippocampal neurogenesis. Epilepsia 49 (Suppl 5), 19–25.

Petrus, D.S., Fabel, K., Kronenberg, G., Winter, C., Steiner, B., and Kempermann, G. (2009). NMDA and benzodiazepine receptors have synergistic and antagonistic effects on precursor cells in adult hippocampal neurogenesis. Eur. J. Neurosci. 29, 244–252.

Rao, M.S., Hattiangady, B., Abdel-Rahman, A., Stanley, D.P., and Shetty, A.K. (2005). Newly born cells in the ageing dentate gyrus display normal migration, survival and neuronal fate choice but endure retarded early maturation. Eur. J. Neurosci. 21, 464–476.

Rao, M.S., Hattiangady, B., and Shetty, A.K. (2008). Status epilepticus during old age is not associated with enhanced hippocampal neurogenesis. Hippocampus 18, 931–944.

Roybon, L., Hjalt, T., Stott, S., Guillemot, F., Li, J.Y., and Brundin, P. (2009). Neurogenin2 directs granule neuroblast production and amplification while NeuroD1 specifies neuronal fate during hippocampal neurogenesis. PLoS ONE 4, e4779.


Scharfman, H.E., and Gray, W.P. (2007). Relevance of seizure-induced neurogenesis in animal models of epilepsy to the etiology of temporal lobe epilepsy. Epilepsia 48 (Suppl 2), 33–41.

Seri, B., Garcı´a-Verdugo, J.M., McEwen, B.S., and Alvarez-Buylla, A. (2001). Astrocytes give rise to new neurons in the adult mammalian hippocampus. J. Neurosci. 21, 7153–7160.

Sestan, N., Artavanis-Tsakonas, S., and Rakic, P. (1999). Contact-dependent inhibition of cortical neurite growth mediated by notch signaling. Science 286, 741–746.

Shors, T.J., Miesegaes, G., Beylin, A., Zhao, M., Rydel, T., and Gould, E. (2001). Neurogenesis in the adult is involved in the formation of trace memo-ries. Nature 410, 372–376.

Sibbe, M., Fo¨rster, E., Basak, O., Taylor, V., and Frotscher, M. (2009). Reelin and Notch1 cooperate in the development of the dentate gyrus. J. Neurosci. 29, 8578–8585.

Steiner, B., Klempin, F., Wang, L., Kott, M., Kettenmann, H., and Kempermann, G. (2006). Type-2 cells as link between glial and neuronal lineage in adult hippocampal neurogenesis. Glia 54, 805–814.

Steiner, B., Zurborg, S., Ho¨rster, H., Fabel, K., and Kempermann, G. (2008). Differential 24 h responsiveness of Prox1-expressing precursor cells in adult hippocampal neurogenesis to physical activity, environmental enrichment, and kainic acid-induced seizures. Neuroscience 154, 521–529.

Stump, G., Durrer, A., Klein, A.L., Lu¨tolf, S., Suter, U., and Taylor, V. (2002). Notch1 and its ligands Delta-like and Jagged are expressed and active

in distinct cell populations in the postnatal mouse brain. Mech. Dev. 114, 153–159.

Suh, H., Consiglio, A., Ray, J., Sawai, T., D’Amour, K.A., and Gage, F.H. (2007). In vivo fate analysis reveals the multipotent and self-renewal capacities of Sox2+ neural stem cells in the adult hippocampus. Cell Stem Cell 1, 515–528. Tumbar, T., Guasch, G., Greco, V., Blanpain, C., Lowry, W.E., Rendl, M., and Fuchs, E. (2004). Defining the epithelial stem cell niche in skin. Science 303, 359–363.

van Praag, H., Christie, B.R., Sejnowski, T.J., and Gage, F.H. (1999a). Running enhances neurogenesis, learning, and long-term potentiation in mice. Proc. Natl. Acad. Sci. USA 96, 13427–13431.

van Praag, H., Kempermann, G., and Gage, F.H. (1999b). Running increases cell proliferation and neurogenesis in the adult mouse dentate gyrus. Nat. Neu-rosci. 2, 266–270.

van Praag, H., Shubert, T., Zhao, C., and Gage, F.H. (2005). Exercise enhances learning and hippocampal neurogenesis in aged mice. J. Neurosci. 25, 8680– 8685.

Zhang, J., Niu, C., Ye, L., Huang, H., He, X., Tong, W.G., Ross, J., Haug, J., Johnson, T., Feng, J.Q., et al. (2003). Identification of the haematopoietic stem cell niche and control of the niche size. Nature 425, 836–841. Zhao, C., Deng, W., and Gage, F.H. (2008). Mechanisms and functional implications of adult neurogenesis. Cell 132, 645–660.