Troy+ brain stem cells cycle through quiescence and regulate their number by sensing niche occupancy

Abstract

The adult mouse subependymal zone provides a niche for mammalian neural stem cells (NSCs). However, the molecular signature, self-renewal potential, and fate behavior of NSCs remain poorly defined. Here we propose a model in which the fate of active NSCs is coupled to the total number of neighboring NSCs in a shared niche. Using knock-in reporter alleles and single-cell RNA sequencing, we show that the Wnt target Tnfrsf19/Troy identifies both active and quiescent NSCs. Quantitative analysis of genetic lineage tracing of individual NSCs under homeostasis or in response to injury reveals rapid expansion of stem-cell number before some return to quiescence. This behavior is best explained by stochastic fate decisions, where stem-cell number within a shared niche fluctuates over time. Fate mapping proliferating cells using a Ki67^iresCreER allele confirms that active NSCs reversibly return to quiescence, achieving long-term self-renewal. Our findings suggest a niche-based mechanism for the regulation of NSC fate and number.

Keywords: cellular dynamics; ki67; modeling; neural stem cells; single-cell sequencing.

Fig. 1.

Troy^GFP population displays NSC characteristics. (A) Schematic representation of the adult SEZ niche. b.v., blood vessels; E, ependymal cells; TA, transient amplifying cell. (B) Endogenous GFP (green) expression in the adult Troy^GFP mouse brain. Blue: DAPI. (C–H) Characterization of the Troy^GFP population using confocal assisted immunohistochemistry. Troy^GFP cells include SOX2+ (C) and GFAP+ (D) progenitors but not DCX+ NBs (E) or S100β+ E cells (F). Some Troy^GFP cells are actively cycling (KI67+, G). (H) Quantification of the results in C–G. (I) FACS-sorted GFP+ cells were assayed for their neurosphere-forming potential. (J) Individual Troy^GFP-high-derived spheres (40/40 spheres, five animals) could be expanded over at least 10 passages and displayed multipotency generating β3-tubulin+ neurons (red, Left), GFAP+ astrocytes (green, Left), and CNPase+ (red, Right) Olig2+ (green, Right) oligodendrocytes upon differentiation. (K) Lineage tracing using Troy^GFPiresCreER Rosa^LacZ mice. X-Gal staining of coronal sections of brains isolated 2 d (Lower Left) or 1 y (Lower Right) after 5 d of Tmx administration (1 × 5 mg each day). Error bars show SD. (Scale bars: B, 1 mm; C and G, 10 μm; D and F, 20 μm; E, 100 μm; J, 100 μm.)

Fig. 2.

Single-cell transcriptome atlas of adult neurogenesis. (A) A t-SNE map showing clusters identified by RaceID2 and expression of key marker genes. (B) Summary of genes differentially expressed in each cluster. Unique (left lane) as well as shared (middle and right lane) genes are shown. See Dataset S2 for a complete list. (C) t-SNE maps displaying the normalized log2 expression of key genes. The color key shows expression values. (D) Distribution of sorted cell populations along pseudotime. Putative cell types are indicated above. Boxes highlight clusters 2, 1, and 6. Colors code for RaceID2 clusters shown in A. (E) Plot displaying the running mean average expression levels of representative (exemplar) genes for selected gene modules. In D and E, cells are ordered on the x axis according to pseudotime; the color bar displays RaceID2 clusters.

Fig. 3.

Analyzing the Troy+ lineage at a clonal level. (A) Schematic representation of the lineage-tracing experiment. (B–E) Representative clones identified as clusters of YFP+ cells (B); migratory NBs were excluded from the analysis. One week after Tmx induction, clones with multiple dividing Troy^GFP+ cells (C), both dividing Troy^GFP+ cells and differentiating progeny (D), as well as large with a quiescent Troy^GFP+ cell and differentiating progeny (E) were visible within the same sample. (F) Comparison of the composition of clones at d 7 to the tissue. (G) Quantification of the number of Troy^GFP+ cells with respect to the clone size over time. Percentages (T, total) on the left and at the bottom show aggregated numbers. S, size. (H) Average composition of different cell types in clones that retain NSCs scored over time. (Scale bar: 60 μm for B, 5 μm for C and D, and 20 μm for E.)

Fig. 4.

A restricted niche regulates adult NSC numbers. (A) Average number of active (aNSC) and quiescent (qNSC) NSCs per NSC-containing clone scored over time alongside model predictions (shaded areas indicate 95% plausible intervals). (B) Density of NSC-retaining clones over time. (C) Distribution of the number of qNSCs per NSC-containing clone averaged over time points between d 14–112 and model prediction (95% plausible intervals). (D) Schematic representation of the model of niche regulation of NSC numbers. qNSCs become activated at rate $\alpha$, while aNSCs return to quiescence at rate $\mu$ or divide at rate $\lambda$. When an aNSC divides in a niche containing a total of $k$ NSCs (active or quiescent), it undergoes symmetric cell duplication with a probability $p_k$ and symmetric differentiation with a probability $1-p_k$, with $p_k=1/k^x$ and $x=0.9\pm 0.1$ (see SI Theory for details). (E) Sample plots comparing the fraction of clones with given aNSC and qNSC composition estimated by simulation of the model (bars) to collected data (red bars) at various time points.

Fig. 5.

aNSCs return to long-term, reversible quiescence. (A) Mouse alleles used for lineage tracing. (B) Characterization of the KI67^iresCreER Rosa^tdTomato lineage tracing d 1, d 7, 1 mo, and 1 y after a single injection of 250 mg/kg Tmx. (C) Quantification of cell types shown in B. (D) As depicted (schematic), tracing starts from aNSCs, TA cells, and some NBs. Contact to the ventricular surface is visualized by β-catenin (at the surface). Differentiation status is evaluated using KI67 and DCX (4 μm below the surface). (E) Density (number of clones per mm²) and size distribution of tdTomato+ clones contacting the ventricles. (F) Quantification of D displaying active fraction (KI67+/tdTomato+) of tdTomato+ cells in pinwheels of a given size. (G) Comparison of the frequency of tdTomato+ cells per pinwheel in KI67^{iresCreER +/HET} Rosa^{tdTomato +/HET} mice (red) with the steady-state distribution (d 14 onward) of Troy^GFP+ cells in clones (clonal distribution, dark green) and the number of Troy^GFP+ cells per pinwheel (light green) in Troy^{GFPiresCreER +/HET} Rosa^{YFP +/HET} mice. (H) Optical sections showing contact of Troy^GFP+ (green) GFAP+ (red) NSCs to the ventricular surface. (Scale bars in B, D, and H, 10 μm.)

Fig. 6.

Clonal dynamics of deep quiescent Troy+ NSCs activated during regeneration. (A) Injection of 5-FU abolishes proliferating cells in the SEZ. Recombination in Troy^{GFPiresCreER +/HET} Rosa^{YFP +/HET} mice was induced with Tmx 2 d (Tmx, −d 2) before 5-FU injection (d 0). Mice were collected at 2 (d 2), 4 (d 4), 7 (d 7), and 14 (d 14) d after 5-FU injection. Troy^GFP+ NSCs and DCX+KI67− NBs survive the treatment. (B) Quantification of the clonal analysis at d 7 and d 14. S, size; T, total percentage for each row or column. (C) Examples of Troy^GFP+ activated NSCs at d7. At this stage, clones are exclusively formed of Troy^GFP+ NSCs. (D) Examples of activated (Left) as well as quiescent (Right) NSCs within clones at d 14. (E) Clonal distribution of Troy^GFP+ NSCs 14 d after Tmx induction is similar between unperturbed and injured (5-FU) conditions. (F) Treatment with 5-FU in Ki67^{iresCreER +/HET} Rosa^{tdTomato +/HET} mice. (Left) Treatment with 5-FU (d 0) 1 d before Tmx induction (d 1) kills the majority of recombined cells (analyzed at d 2). (Right) When 5-FU is treated 1 d after Tmx induction, some of the tdTomato+DCX+ NBs and tdTomato+KI67− cells contacting the ventricles survive. (G) Quantification of F. [Scale bars: A, 50 μm; C and D, 20 μm (Upper Images) and 10 μm (Lower Images); F, 50 μm (Upper) and 20 μm (Lower)].

Fig. 7.

A closed niche model of adult neurogenesis. Schematic representation of the NSC niche at selected time points. The lower plot depicts the result of a numerical simulation of the model dynamics with the inferred parameters showing changes in the number of quiescent and aNSCs as well as the production of TA cells at given time points. These simulations reveal a pattern of stochastic dynamics in which the sporadic entry of qNSCs into the cycle leads to a burst of proliferative activity leading to TA cell production before a return to quiescence.