Report

Return to quiescence of mouse neural stem cells by degradation of a proactivation protein

See allHide authors and affiliations

Science  15 Jul 2016:
Vol. 353, Issue 6296, pp. 292-295
DOI: 10.1126/science.aaf4802

Sending neural stem cells back to the garage

In the brain's hippocampus, which modulates memories and emotions, neural stem cells generate new neurons, even during adulthood. How many new neurons are generated, and when, follows from the balance between quiescence and proliferation in the pool of neural stem cells. Urbán et al. asked what signals send proliferating stem cells back into a quiescent state. They found that a key transcription factor that promotes cellular proliferation was degraded through the ubiquitinylation system. This molecular interaction regulated the return to a resting state, but one that was not quite as quiescent as the original state. Stem cells in this resting but primed state sustained the stem cell pool.

Science, this issue p. 292

Abstract

Quiescence is essential for long-term maintenance of adult stem cells. Niche signals regulate the transit of stem cells from dormant to activated states. Here, we show that the E3-ubiquitin ligase Huwe1 (HECT, UBA, and WWE domain–containing 1) is required for proliferating stem cells of the adult mouse hippocampus to return to quiescence. Huwe1 destabilizes proactivation protein Ascl1 (achaete-scute family bHLH transcription factor 1) in proliferating hippocampal stem cells, which prevents accumulation of cyclin Ds and promotes the return to a resting state. When stem cells fail to return to quiescence, the proliferative stem cell pool becomes depleted. Thus, long-term maintenance of hippocampal neurogenesis depends on the return of stem cells to a transient quiescent state through the rapid degradation of a key proactivation factor.

Stem cells contribute to tissue homeostasis by generating new differentiated cells. Adult stem cells can enter a reversible state of quiescence that protects the cells from damage and the population from depletion. Niche signals determine the balance between quiescent and activated states. Excessive quiescence leads to too few differentiated progeny, whereas excessive proliferation exhausts the stem cell population (1).

Neural stem cells (NSCs) in the dentate gyrus (DG) of the mouse hippocampus generate new granule neurons that integrate into the hippocampal circuit to modulate mood and memory (2, 3). Niche signals control expression of the transcription factor Ascl1 (achaete-scute family bHLH transcription factor 1), which in turn directs NSC proliferation (4). To identify factors that regulate Ascl1, we characterized proteins that coimmunoprecipitate with Ascl1 in cultured mouse NSCs using mass spectrometry. We found that Huwe1 (HECT, UBA, and WWE domain–containing 1), a HECT domain E3 ubiquitin ligase associated with idiopathic intellectual disability and schizophrenia (5, 6), interacts with Ascl1 (fig. S1). We generated embryonic telencephalon- and adult hippocampus–derived NSCs in which Huwe1 is expressed and can be inactivated by Cre recombinase (fig. S2) (7). Inactivation of Huwe1 resulted in an accumulation of Ascl1 protein and an extension of its half-life from 38 to 121 min (Fig. 1, A to C, and fig. S2, B, E, and G), whereas proteins destabilized by Huwe1 in other tissues were not affected (fig. S2C) (810). Ascl1 is degraded by the proteasome in NSCs (fig. S2D), and silencing of Huwe1 with short hairpin RNAs (shRNAs) decreased the extent of polyubiquitinylation of Ascl1 (Fig. 1D). Therefore, Huwe1 promotes the proteasomal degradation of Ascl1 protein.

Fig. 1 Huwe1 controls Ascl1 stability in adult hippocampal stem cells.

(A and B) Huwe1 inactivation in embryonic telencephalon-derived cultured NSCs increases levels of Ascl1 protein [(A) Western blot; Actin B (Actb) is used as loading control] but not of Ascl1 mRNA [(B) quantitative polymerase chain reaction analysis of empty (Ctrl) or Cre recombinase (cKO)–expressing adenovirus-transduced cells]. (C) Cells were treated with cycloheximide so as to stop protein synthesis for different times and processed for Western blot to determine Ascl1 half-life; n = 4 independent experiments. (D) Ubiquitynilated Ascl1 (top) and total Ascl1 levels (bottom) were determined by immunoblotting with an antibody to hemagglutinin (HA) after transfection with HA-Ascl1 and control or Huwe1 shRNA. (E to H) Huwe1 was inactivated in adult DG NSCs by means of five injections of tamoxifen at P60 followed by analysis at P90. (F) Scale bar, 10 μm. The number of Ascl1-positive NSCs, identified by their position in the subgranular zone and the presence of a GFAP+ radial process [(F) and (G)], and the intensity of Ascl1 immunolabeling per cell (H), were quantified; (H) n = 4 mice per condition. (G) n = 27 Ascl1-positive cells from four mice (control) and 42 Ascl1-positive cells from four mice (Huwe1cKO). Yellow arrows in (F) point to Ascl1-positive cells.

Huwe1 is expressed throughout the brain, including in hippocampal NSCs and their progeny in the subgranular zone of the DG (fig. S3). To study Huwe1 function in these cells, we generated mice in which administration of the small molecule tamoxifen inactivates the Huwe1 gene and initiates yellow fluorescent protein (YFP) expression in hippocampal NSCs [Huwe1fl;GLAST-CreERT2; Rosa-Stop-YFP mice (11), called Huwe1cKO mice hereafter]. One month after tamoxifen administration, the intensity of Ascl1 immunolabel was enhanced in cells of the subgranular zone of Huwe1cKO mice compared with controls (Fig. 1, E, F, and G; and fig. S4). The number of GFAP+ (glial fibrillary acidic protein–positive) radial NSCs expressing Ascl1 was also increased (Fig. 1H). We observed no difference in the expression of other known Huwe1 substrates (fig. S5). Thus, Huwe1 regulates Ascl1 stability in hippocampal NSCs. Because Ascl1 promotes NSC activation in the hippocampus (4), up-regulation of Ascl1 in Huwe1cKO mice might stimulate NSC proliferation. Indeed, a higher proportion of NSCs in the DG of Huwe1cKO mice were cycling at postnatal day 90 (P90) (Fig. 2, A and B). Thus, Huwe1 suppresses hippocampal NSC proliferation in wild-type mice.

Fig. 2 Huwe1 inactivation promotes hippocampal stem cell proliferation and blocks progenitor differentiation.

(A and B) Hippocampal stem cell proliferation was assessed by means of Ki67 staining (B) and BrdU incorporation after a 2-hour pulse [(A) and (B)]. The total number of NSCs remained the same (fig. S6B). Yellow arrowheads point to BrdU-negative NSCs, and yellow arrows point to BrdU-positive NSCs. (A) Scale bar, 20 μm; n = 3 mice (BrdU) and 6 mice (Ki67) per condition. (C) The generation of neuronal precursors was assessed by counting the numbers of Tbr2-positive intermediate progenitors and of DCX-positive neuroblasts; n = 3 mice per condition.

Huwe1cKO mice also had too few intermediate progenitors and neuroblasts, and the remaining cells ectopically expressed Ascl1 (Fig. 2C and fig. S6). The deletion of Huwe1 did not induce a switch toward gliogenesis, and intermediate progenitors were most likely eliminated through apoptosis (figs. S7 and S8). We suggest that persistence of Ascl1 protein in progenitors lacking Huwe1 maintains the proliferative state of NSCs and prevents differentiation of early intermediate progenitors.

To study the role of the interaction between Huwe1 and Ascl1 in the regulation of quiescence, we labeled quiescent NSCs by means of prolonged exposure to BrdU followed by a chase (label-retention assay) (12). We then inactivated Huwe1 and analyzed the mice 3 weeks later (Fig. 3A). The numbers of BrdU-retaining progenitors were not significantly different in Huwe1cKO and control mice, indicating that the loss of Huwe1 did not lead to premature activation of quiescent stem cells, which would result in BrdU dilution (Fig. 3B and fig. S9, A to F). Thus, Huwe1 is not required to maintain NSCs in quiescence.

Fig. 3 Adult hippocampal stem cells fail to return to quiescence in Huwe1cKO mice.

(A and B) Mice received BrdU in the drinking water for 5 days, followed by Huwe1 inactivation. Analysis was performed 3 weeks later; n = 3 (control) and 6 (Huwe1cKO) mice. An additional BrdU-retention experiment is shown in fig. S9, D to F. (C and D) BrdU was administered after Huwe1 inactivation, when more Huwe1cKO NSCs than control NSCs proliferate (Fig. 2B). Analysis was performed 3 weeks later; n = 5 (control) and 3 (Huwe1cKO) mice. (E to G) EdU was injected 24 hours before analysis. EdU+ Ki67+ cells continue proliferating, whereas EdU+ Ki67 cells have exited the cell cycle. No EdU+ NSCs expressed the astrocytic marker S100β (fig. S11, D to I). The yellow arrowhead points to an EdU+ Ki67 NSC, and the yellow arrow points to an EdU+ Ki67+ NSC; n = 47 (control) and 38 (Huwe1cKO) EdU+ NSCs from 6 and 5 mice, respectively. (H to K) Analysis was performed 5 months after Huwe1 inactivation. The overall number of NSCs was not changed, but fewer NSCs proliferated in Huwe1cKO than control mice; n = 4 mice per condition. Scale bars, 20 μm (G) and 50 μm (I).

To determine whether Huwe1 is required for proliferating NSCs to return to quiescence, we marked cells exiting the cell cycle in the absence of Huwe1 by first inactivating Huwe1 and then performing a BrdU label–retention assay (Fig. 3C). BrdU-retaining radial cells in the subgranular zone of control mice were quiescent NSCs and not astrocytes because they did not express the astrocytic marker S100β (fig. S9H). There were fewer BrdU-retaining NSCs in Huwe1cKO mice than in control mice (Fig. 3D and fig. S9, I and J), indicating that without Huwe1, fewer NSCs returned to quiescence. We could not directly examine the divisions of Huwe1cKO NSCs by means of in vivo clonal analysis (13) because the low dose of tamoxifen required for this analysis was not sufficient to delete the Huwe1fl mutant allele (fig. S10). To directly assess whether Huwe1 is required in proliferating NSCs for their return to quiescence, we marked instead a cohort of proliferating cells with a pulse of EdU and identified the fractions of NSCs that had either exited or reentered the cell cycle 24 hours later by double labeling for EdU and Ki67 (Fig. 3E and fig. S11). In control mice, 23.4% of EdU+ NSCs were negative for Ki67, suggesting that they had returned to quiescence after cycling and incorporating EdU (Fig. 3, F and G). In Huwe1cKO mice, only 2.6% of EdU+ NSCs were negative for Ki67, indicating that almost all Huwe1 mutant NSCs had reentered the cell cycle (Fig. 3, F and G). Thus, elimination of the proactivation factor Ascl1 from proliferating NSCs by Huwe1 in wild-type mice drives the cells into quiescence.

The long-term consequence of excessive proliferation of hippocampal NSCs in Huwe1cKO mice was examined 5 months after Huwe1 deletion, at P210 (Fig. 3H). The overall number of NSCs was unchanged, confirming that Huwe1 is not required for the maintenance of the predominant quiescent NSC population (Fig. 3, I and J). In contrast, the number of proliferating NSCs was reduced (2.4 ± 0.1% Ki67+ NSCs in control mice; 0.3 ± 0.3% in Huwe1cKO mice) (Fig. 3K), indicating that Huwe1 is required for the long-term maintenance of the proliferative NSC population in the hippocampus. This result also shows that stem cells that have proliferated and returned to quiescence are required to replenish the proliferative stem cell pool (fig. S12).

Ascl1 activates the transcription of several cell-cycle regulators in NSCs (4, 14). Huwe1-deficient NSCs showed higher expression of CcnD1 (Cyclin D1) and CcnD2 (Cyclin D2), two targets of Ascl1 (figs. S13A and S14). The elevation of CcnD1 and CcnD2 expression in Huwe1-mutant NSCs was due to the accumulation of Ascl1 because it was abolished after Ascl1 knockdown or deletion (figs. S13D and S14A). The increase in CcnD1 expression in Huwe1cKO mice was seen in quiescent NSCs and to a greater extent in proliferating NSCs (Fig. 4, B to E, and fig. S14F). Thus, stabilization of Ascl1 in NSCs lacking Huwe1 promotes cell cycle reentry by inducing the expression of CcnD genes.

Fig. 4 CcnD genes are abnormally up-regulated in Huwe1cKO hippocampal stem cells.

(A to E) EdU was added to the drinking water for 48 hours before the analysis to mark cells that progressed through S-phase during this period. Colabeling for EdU and CcnD1 identifies cells that have proliferated and still express CcnD1, which is required for proliferation of adult hippocampal stem cells (17, 18). Pie charts in (E) show the percentage of EdU+ NSCs that maintain CcnD1 expression. In (A), yellow arrows point to CcnD1+ NSCs; in (D), yellow arrowheads point to EdU+ CcnD1, and yellow arrows point to EdU+ CcnD1+ NSCs. The elevation of CcnD1 and CcnD2 expression was not due to an increase in proliferation of Huwe1-mutant NSCs [(C) and fig. S13, B and C]. (C) n = 6 mice per condition and (E) n = 33 (control) and 47 (Huwe1cKO) EdU+ NSCs from 6 mice per condition. Scale bars, 20 μm.

Posttranscriptional regulation controls stem cell activity, alongside transcriptional and epigenetic mechanisms (15). In the embryonic nervous system, Huwe1 promotes cell cycle exit and neuronal differentiation of progenitors by destabilizing N-myc (7). In the adult brain, we show here that Huwe1 targets the proactivation factor Ascl1 to promote the return of proliferating hippocampal NSCs to a resting state. Regulation of Ascl1 alone is not sufficient to promote quiescence exit, suggesting that additional signals are required to stimulate stem cell activity. Most NSCs continue to divide once activated and are eventually lost, thus contributing to the rapid attrition of the stem cell population over time (16). However, Huwe1 promotes the return to a resting state of a minority of dividing NSCs, which is essential for the long-term maintenance of the diminishing pool of proliferating stem cells (fig. S12). Our results suggest that proliferating stem cells that return to quiescence form a pool of temporarily resting cells that is distinct from the main dormant pool and is the main contributor to neurogenesis in the adult hippocampus.

Correction (17 March 2017): Report: "Return to quiescence of mouse neural stem cells by degradation of a proactivation protein " by N. Urbán (15 July 2016, p. 292). The first sentence of the Acknowledgments has been revised to add A. Lasorella: “We gratefully acknowledge L. Chen, A. Achimastou, and E. Lavedeze for technical support; M. del Mar Masdeu, R. Subramaniams, and J. Mor for managing the mouse colony; A. Lasorella, A. Iavarone, and M. Götz for providing transgenic mice; and I. Crespo, R. Kageyama, and members of the Guillemot laboratory for discussions.”

Supplementary Materials

www.sciencemag.org/content/353/6296/292/suppl/DC1

Materials and Methods

Figs. S1 to S14

Table S1

References (1935)

References and Notes

  1. Acknowledgments: We gratefully acknowledge L. Chen, A. Achimastou, and E. Lavedeze for technical support; M. del Mar Masdeu, R. Subramaniams, and J. Mor for managing the mouse colony; A. Lasorella, A. Iavarone, and M. Götz for providing transgenic mice; and I. Crespo, R. Kageyama, and members of the Guillemot laboratory for discussions. The study was conceived and the manuscript was written by N.U. and F.G. Most of the work was performed by N.U.; D.L.C.B., J.A.A.D., and C.H. performed the mass spectrometry analysis; J.A. helped with experiments; and A.F. and O.A. analyzed the polyubiquitination of Ascl1. N.U. was supported by fellowships from the Medical Research Council (MRC) and the Francis Crick Institute, D.L.C.B. was supported by fellowships from the Federation of European Biochemical Societies and the Francis Crick Institute, J.A.A.D. was financed by the Netherlands Organisation for Scientific Research (NWO) project (184.032.201), and A.F. was financed by the Fondation de France. This work was supported by grants from the Wellcome Trust (106187/Z/14/Z), the Biotechnology and Biological Sciences Research Council (BB/K005316/1), the MRC (U117570528), and the Francis Crick Institute (BZ10089) to F.G. The authors declare no conflicts of interest. Additional data is available in the supplementary materials.
View Abstract

Stay Connected to Science

Navigate This Article