Fig 2: Suppression of proneural gene expression in slowly dividing NPCs.a RT and real-time PCR analysis of Ascl1 mRNA in NPCs isolated from the LGE of Rosa-rtTA;TRE-mCMV-H2B-GFP embryos at E16.5 and sorted on the basis of H2B-GFP retention after exposure to 9TB-Dox at E9.5. Data were normalized by the amount of β-actin mRNA, are expressed relative to the corresponding value for GFP+ cells, and are means ± SEM (n = 3 independent experiments), two-tailed paired t test. b, c RT and real-time PCR analysis of Neurog2 and Tbr2 mRNAs in cells positive for CD133 and GFP and negative for CD24 isolated by FACS from the neocortex of E17.5 embryos that had been subjected to in utero electroporation at E14.5 with plasmids for GFP alone (control) or together with p57 (b) or p18 (c). Data were normalized by the amount of β-actin mRNA, are expressed relative to the corresponding value for control, and are means ± SEM (n = 3 independent experiments), two-tailed paired t test. d Immunohistochemical analysis with antibodies to Ascl1 and to Sox2 for sections of the brain of control and Hey1 knockout mice at E14.5. Nuclei were stained with Hoechst 33342. Scale bar, 50 μm. Dashed lines indicate the ventricular surface. e Quantification of the proportion of Ascl1+ cells among Sox2+ cells in the dLGE for sections as in d. Data are means ± SEM (n = 4 embryos), two-tailed Student’s t test. f Model for the division of labor of Notch downstream effectors that can select the fate of “rapidly dividing” (oscillating Hes1/Hes5) versus “slowly dividing” (tonic Hey1) pools of embryonic NPCs, respectively.

Fig 3: Telencephalic region-specific H3K27me3 deposition and Ring1B binding at BMP and Wnt ligand gene loci in early stage NPCs.a CD133+ NPCs isolated from the dorsal midline (DM), CTX, and ventral (V) regions of the telencephalon from WT (ICR) embryos at E11 were subjected to CUT&Tag analysis with antibodies to H3K27me3 or to Ring1B. Biological triplicates (for H3K27me3) or quadruplicates (for Ring1B) whose quality of regional dissection was confirmed by RT-qPCR analysis of region-specific gene expression (Supplementary Fig. 8a) were analyzed. b, c Averages of normalized CUT&Tag signals around Bmp4 (b) or Wnt8b (c) in three (H3K27me3) or four (Ring1B) independent experiments are visualized in the UCSC genome browser. The RefSeq gene models are shown at the bottom of each panel. d, e Averages of normalized CUT&Tag signals for H3K27me3 around (±1 kbp) the transcription start site (TSS) (d) and Ring1B around (±1 kbp) the Ring1B summit (determined as the site in the gene body with the highest Ring1B signals across all samples) (e) of the indicated genes. Data are means ± s.d. for biological triplicates (d) or quadruplicates (e), one-way ANOVA followed by the Benjamini–Hochberg multiple-comparison test. Source data are provided as a Source Data file.

Fig 4: Hey1 is highly expressed in slowly dividing embryonic NPCs.a Scheme for the isolation of slowly dividing and rapidly dividing NPCs from the LGE of Rosa-rtTA;TRE-mCMV-H2B-GFP embryos at E16.5 on the basis of H2B-GFP retention after exposure to 9TB-Dox at E9.5. Embryonic NPCs were defined as CD133+CD24– cells and were subdivided into two populations based on H2B-GFP intensity—top 8% (GFP++) and middle 40% to 65% (GFP+)—by FACS for RT and real-time PCR analysis. b RT and real-time PCR analysis of the indicated Notch target genes in NPCs isolated as in a. Data were normalized by the amount of β-actin mRNA, are expressed relative to the corresponding value for GFP+ cells, and are means ± SEM (n = 4, 4, 5, and 4 independent experiments for Hes1, Hes5, Hey1, and Hey2, respectively), two-tailed paired t test. c Immunohistochemical analysis with antibodies to GFP and to Hey1 for sections of the LGE of Rosa-rtTA;TRE-mCMV-H2B-GFP embryos at E17.5 after exposure to 9TB-Dox at E9.5. Nuclei were stained with Hoechst 33342. Scale bar, 20 μm. Dashed lines indicate the ventricular surface. Arrowheads indicate GFP-retaining slowly dividing cells. d Quantification of Hey1 staining intensity for the indicated numbers of LRCs and non-LRCs (top 15% and bottom 60% of NPCs for H2B-GFP staining intensity, respectively) in the LGE at E17.5 for sections as in c. Data are expressed relative to the corresponding value for non-LRCs. Violin plots show the median (horizontal dashed line) and quartiles (horizontal dotted lines) (n = 1943 and 485 cells for non-LRCs and LRCs, respectively). Eight brain sections were analyzed per embryo, two-tailed Student’s t test. e RT and real-time PCR analysis of Hey1 expression in cells positive for CD133 and GFP and negative for CD24 isolated by FACS from the neocortex of E17.5 embryos that had been subjected to in utero electroporation at E14.5 with plasmids for GFP alone (control) or together with p57 or p18. Data were normalized by the amount of β-actin mRNA, are expressed relative to the corresponding value for control, and are means ± SEM for n = 3 independent experiments for p57 or n = 5 independent experiments for p18, two-tailed paired t test.

Fig 5: Neuroblastoma stem cells have highly enriched GPC2 expression(A and B) GPC2 histograms (A) and GPC2 versus CD133, CD338, CD117, and GD2 flow cytometry plots (B) of neuroblastoma PDXs/xenografts.(C) Fold-change GPC2 MFI plot of GPC2UltraHi versus GPC2Hi cells in neuroblastoma PDXs/xenografts.(D) Fold-change stem cell marker MFI plot of GPC2UltraHi versus GPC2Hi cells in neuroblastoma PDXs/xenografts.(E and F) Gpc2/GPC2 histograms (E) and Gpc2/GPC2 versus CD133, CD338, CD117, and GD2 flow cytometry plots (F) of murine (9464D) and human (SK-N-AS) neuroblastoma cell lines.(G) Fold-change stem cell marker MFI plot of GPC2UltraHi versus GPC2Hi cells in neuroblastoma cell lines.(H) GPC2 (left) and CD133 (right) histograms in SK-N-AS cells after D3-GPC2-PBD ADC treatment.(I) Quantification of CD133- and GPC2-positive SK-N-AS cells after D3-GPC2-PBD ADC treatment (left, ADC × 1; right, ADC × 2).(J) GPC2 western blot of SK-N-SH/NBL-SRPMI/NB cells.(K) Relative growth of SK-N-SH/NBL-SRPMI/NB cells 4 days after treatment with the D3-GPC2-PBD ADC.(L) Fold-change MFI plot of SH-SY5Y-EmptyNB and -GPC2RPMI/NB isogenic cells.Data in (C) and (D) are represented as mean ± SEM of PDXs/xenografts analyzed 2–3 independent times with 4 technical replicates each time for each PDX/xenograft in (C). Data in (G), (I), and (L) are represented as mean ± SEM of experiments done 2–3, 2, and 5 times, respectively. Data in (K) represent mean ± SEM from a representative experiment done in technical triplicate at least two independent times.Neg, negative.*p < 0.05; **p < 0.001; ***p < 0.0001. See also Figure S2.