Increasing evidence suggests that deficits in adult stem cell maintenance cause aberrant tissue repair and premature aging [1]. While the mechanisms regulating stem cell longevity are largely unknown, recent studies have implicated p53 and its family member p63. Both proteins regulate organismal aging [2-4] as well as survival and self-renewal of tissue stem cells [5-9]. Intriguingly, haploinsufficiency for a third family member, p73, causes age-related neurodegen-eration [10]. While this phenotype is at least partially due to loss of the DNp73 isoform, a potent neuronal prosurvival protein [11-16], a recent study showed that mice lacking the other p73 isoform, TAp73, have perturbations in the hippocampal dentate gyrus [17], a major neurogenic site in the adult brain. These findings, and the link between the p53 family, stem cells, and aging, suggest that TAp73 might play a previously unanticipated role in maintenance of neural stem cells. Here, we have tested this hypothesis and show that TAp73 ensures normal adult neurogenesis by promoting the long-term maintenance of neural stem cells. Moreover, we show that TAp73 does this by transcriptionally regulating the bHLH Hey2, which itself promotes neural precursor maintenance by preventing premature differentiation. Results TAp73 Is Necessary to Promote Maintenance of Postnatal Dentate Gyrus Precursors TAp73 2/2 [17] mice display an aberrant hippocampal dentate gyrus (DG), a phenotype of unknown etiology. To understand this, we characterized TAp73 expression and developmental onset of the phenotype. RT-PCR (Figure 1A) and immunostain-ing (Figure 1B) demonstrated that TAp73 is expressed in the newborn mouse hippocampus, where it is predominantly localized to nuclei of cells that coexpress Tbr2, a marker for type 2a precursors [18] (Figure 1B). To determine when the DG first became aberrant, we Nissl stained the postnatal hippocampus (Figure 1C and Figure S1A, available online); TAp73 2/2 and TAp73 +/+ hippocampi were morphologically similar at birth, started to show some differences at postnatal day 6 (P6), and by P16 the dorsal lower blade of the TAp73 2/2 DG was missing. Confirmation that the DG was similar at earlier time points was obtained by immunostaining newborn sections for nestin and prox1, markers for precursors and DG neurons, respectively (Figure S1B). Because the region that becomes aberrant postnatally is comprised of the last-born DG neurons [19], this suggests that TAp73 is necessary for postnatal neurogenesis. To test this idea, we studied ongoing neurogenesis in the adult hippo-campus [20]. Adult TAp73 2/2 and TAp73 +/+ mice were injected with BrdU, and hippocampi were analyzed immunocytochem-ically 24 hr later (Figure 1D). Quantification showed an almost 2-fold decrease in proliferating, BrdU-positive precursors in the TAp73 2/2 subgranular zone (SGZ; the location of the DG precursors) of both the lower and upper DG blades (Figures 1D and 1E). Similarly, doublecortin-positive newly born neurons (Figure 1D) were also reduced approximately 2-fold in the TAp73 2/2 DG (Figure 1F). Thus, TAp73 loss depletes adult DG precursors and decreases neurogenesis. To ask whether this phenotype reflected a cell-intrinsic precursor deficit, we cultured TAp73 +/+ and TAp73 2/2 P3 hippocampal cells in FGF2 and EGF to generate neurospheres [21]. RT-PCR demonstrated that TAp73, DNp73, and p53 mRNAs were expressed in DG neurospheres and that DNp73 and p53 mRNA levels were unaltered by loss of TAp73 (Fig-ure 1G). Immunostaining confirmed that the majority of TAp73 +/+ but not TAp73 2/2 neurosphere cells expressed nuclear TAp73 (Figure 1H). Quantitative analysis at clonal density demonstrated that TAp73 +/+ and TAp73 2/2 neonatal hippocampi contained similar numbers of neurosphere-generating precursors (Figure 1I), consistent with the lack of an in vivo phenotype at birth. However, when sequentially passaged, the TAp73 2/2 neurosphere-forming cells were progressively depleted (Figure 1J), indicating that TAp73 is required for long-term precursor maintenance. In contrast, mean neurosphere diameter (Figure 1K) and Ki67-positive proliferating cells (Figures 1L and 1M) were unchanged, suggesting that TAp73 is not necessary for proliferation of biased progenitors, which comprise the majority of cells in the spheres. Loss of TAp73 Depletes Adult SVZ Precursors and Decreases Olfactory Neurogenesis To ask whether TAp73 is required for maintenance of other adult neural precursors, we examined olfactory neurogenesis, which is ongoing for the life of the animal [20]. Adult TAp73 +/+ and TAp73 2/2 mice were injected five times with BrdU over a 12 hr period. Quantitative immunocytochemical analysis of their olfactory bulbs 30 days later demonstrated an almost 2-fold decrease in BrdU-positive newly born neurons expressing the neuron-specific protein NeuN (Figures 2A and 2B and Figure S2A) in TAp73 2/2 mice. To ask whether this was due to depletion of precursors, we generated clonal neurospheres *Correspondence: fredam@sickkids.ca (F.D.M.), dkaplan@sickkids.ca (D.R.K.) 8 These authors contributed equally to this work