Apical size and deltaA expression predict adult neural stem cell decisions along lineage progression

The maintenance of neural stem cells (NSCs) in the adult brain depends on their activation frequency and division mode. Using long-term intravital imaging of NSCs in the zebrafish adult telencephalon, we reveal that apical surface area and expression of the Notch ligand DeltaA predict these NSC decisions. deltaA-negative NSCs constitute a bona fide self-renewing NSC pool and systematically engage in asymmetric divisions generating a self-renewing deltaAneg daughter, which regains the size and behavior of its mother, and a neurogenic deltaApos daughter, eventually engaged in neuronal production following further quiescence-division phases. Pharmacological and genetic manipulations of Notch, DeltaA, and apical size further show that the prediction of activation frequency by apical size and the asymmetric divisions of deltaAneg NSCs are functionally independent of Notch. These results provide dynamic qualitative and quantitative readouts of NSC lineage progression in vivo and support a hierarchical organization of NSCs in differently fated subpopulations.

The PDF file includes: Figs. S1 to S6 Table S1 Legends for movies S1 and S2 Other Supplementary Material for this manuscript includes the following:

Movies S1 and S2
Supplementary Fig. S1 (related to Fig. 1).Apical area and deltaA expression by cell type and state.
S1A. Boxplots showing the distribution of cell anisotropies, perimeters, AAs and numbers of neighbors as a function of cell types and states (qNSCs, aNSCs and aNPs).n=4 independent hemispheres, Dm region, same samples as in Fig. 1G.

Figure S2
S2B. 2D projection methods.Four different projection methods were tested on the timelapses to project in 2D the ZO1-mKate2 signal: Maximum projection and local Z projector (both available on Fiji), preMosa (available source code on the preMosa GitHub) and CARE (python code available on the CSBdeep GitHub).Zoom on a very low-resolution area in the image to compare the four projection methods.We selected CARE as the best method to resolve the ZO1 staining and improve the segmentation afterwards.
S2C. Closer look at deltaA:GFP expression intensities by color-coding eGFP intensity (FIRE lookup tables).Manual correction of the deltaA signal is necessary to ensure a correct assignment of deltaA expression to individual NSCs, because the apical surface and the corresponding underlying cell cytoplasm, which is GFP-positive, are not always in perfect register.Thus, many NSCs negative for deltaA are wrongly classified as positive if neighboring a balloon-shaped deltaA pos cell.Green arrows show examples where the GFP signal from one cell invades a neighbouring AA in 2D.
S2D. Manual scoring of the deltaA:GFP signal to add quantitative information to the segmentation of each cell, based on the fire LUTs and using visual and temporal criteria (see Materials & Methods).Examples of NSCs with no deltaA (0), weak deltaA (1), medium deltaA (2), and strong deltaA (3) levels.For analyses using quantitative deltaA values (Figs.S3C, S4B, S5A), deltaA expression was calculated using the manual segmentation.The value of deltaA:GFP expression for NSCs classified as "no deltaA" is set at 0 and the values for other NSCs are calculated as the sum of the pixel intensity normalized by the apical area.
S2E.Comparison of the exact quantitation of deltaA:GFP expression (normalized by the area of each cell, see 2D) with manually assessed intensity scores for all cells in one time-lapse.
S2F.Alignment of live and fixed images.We are able to trace back previously live-imaged NSC apical surfaces on 3-mpf double transgenic Tg(gfap:hZO1-mKate2);Tg(mcm5:gfp) fish after fixation and IHC for SOX2, ZO1 and PCNA.Left: live-imaged tissue; right: fixed and immunostained tissue.The images on the right panels show the segmentation of the same group of cells.
S2G.Comparison of AAs in live and fixed/immunostained samples.Scatter plot showing AA before fixation as a function of AA after fixation for the same cell.Linear regression line with 95% CI (slope = 0,89 and R squared = 0,92).The average AA ratio between fixed AA and live AA is calculated as follows: AA ratio fixed/live = (Fixed AA -Live-imaged AA) / (Liveimaged AA) x 100.In average, fixed AA are 5% larger than live-imaged AA.The statistical difference between the two segmented regions was assessed by a two-tailed non-parametric t-test (Mann-Whitney) (n.s.p-value = 0,6667) (n = 95 cells on 2 fish).Figure S3 S3D.Representation of all dividing tracks (n= 194, from 828 NSCs tracked in 3 fish), classified according to deltaA expression of MCs, and showing deltaA expression (deltaA pos NSCs: blue, deltaA neg NSCs: black) and AA (y axis, in µm 2 ) as a function of time (x axis, in days) (each dot is an imaging time point).Red arrowheads mark the different DCs for each track (division events post-quiescence phases, note that there can be several such events along the same track when DCs enter quiescence before dividing again).Green arrowheads mark the DCs from reiterative divisions (successive division events that occur without a quiescence phase in between).