Temperature sensitivity of Notch signaling underlies species-specific developmental plasticity and robustness in amniote brains

Abstract

Ambient temperature significantly affects developmental timing in animals. The temperature sensitivity of embryogenesis is generally believed to be a consequence of the thermal dependency of cellular metabolism. However, the adaptive molecular mechanisms that respond to variations in temperature remain unclear. Here, we report species-specific thermal sensitivity of Notch signaling in the developing amniote brain. Transient hypothermic conditions increase canonical Notch activity and reduce neurogenesis in chick neural progenitors. Increased biosynthesis of phosphatidylethanolamine, a major glycerophospholipid components of the plasma membrane, mediates hypothermia-induced Notch activation. Furthermore, the species-specific thermal dependency of Notch signaling is associated with developmental robustness to altered Notch signaling. Our results reveal unique regulatory mechanisms for temperature-dependent neurogenic potentials that underlie developmental and evolutionary adaptations to a range of ambient temperatures in amniotes.

Fig. 1. Inverse temperature dependency of Notch activity in neural progenitors of oviparous animals.

a, b Monitoring Notch activities in turtle (a) and chick (b) neural progenitors cultured at 30 °C or 37 °C for 24 h. A Notch reporter vector (4xCSL-luc) was introduced to isolated neural progenitors prior to culture. c, d Notch reporter activities in turtle (c) and chick (d) neural progenitors cultured at 30 °C or 37 °C (mean + SD, n = 3 biologically independent samples in each group). e DAPT treatment suppressed Notch activity in chicken neural progenitors cultured at 30 °C (means + SD, n = 5 biologically independent samples in each group). f NICD contents of chick neural progenitors cultured at different temperatures. g Schematic illustration of the temperature shift experiment in developing chick embryos. h–k Distribution of Sox2 and/or EdU-positive cells in the developing chick pallium incubated at 30 °C (h, j) or 37 °C (i, k). Insets represent Sox2-positive or negative EdU-labeled cells. l The proportion of Sox2-negative and EdU-positive cells in the chick pallium incubated at different temperatures (means + SD, n = 3 biologically independent samples in each group). m Schematic illustration of DN-MAML1 or NICD overexpression in the developing chick pallium. n Distribution of GFP-positive cells in GFP control, DN-MAML1- or NICD-overexpressing chick pallium incubated at 30 °C or 37 °C. o The proportion of Sox2-negative and GFP-positive cells in controls (30 °C or 37 °C), and DN-MAML1- or NICD-overexpressing chick pallium at 30 °C or 37 °C (means + SD, n = 3 biologically independent samples in each group). Two-sided, unpaired t-test for c, d, e, and l; ordinary one-way ANOVA for o (p-values were adjusted by Tukey test for multiple comparisons). P-values are indicated in each graph. Scale bars: 25 µm.

Fig. 2. Ligand endocytosis mediates temperature-dependent Notch activity.

a NICD-dependent Notch reporter activity in chick neural progenitors cultured at 30 °C or 37 °C (mean + SD, n = 3 biologically independent samples in each group). b Relative expression levels of genes encoding Notch signaling components in chick neural progenitors cultured at 30 °C or 37 °C (mean + SD, n = 3 biologically independent samples in each group). c Overexpression of Notch1 (NotchWT) or mutant Notch (NotchL468A) receptors in developing chick neural progenitors. d Notch reporter activities in control, NotchWT, and NotchL468A overexpressing samples cultured at 30 °C or 37 °C (mean + SD, n = 3 biologically independent samples in each group). e Schematic drawing of Dll1/4 and Dll1/4ΔC; the latter lacks the intracellular domain (ICD) required for ubiquitylation and endocytosis. f, g Notch reporter activity in controls and Dll1ΔC- (f) or Dll4ΔC (g) -overexpressing neural progenitors cultured at 30 °C or 37 °C [mean + SD, n = 3 (f) and 4 (g) biologically independent samples in each group]. h Notch reporter activity in control and dynasore-treated neural progenitors cultured at 30 °C or 37 °C (mean + SD, n = 3 biologically independent samples in each group). i Notch reporter activity in control and dominant-negative dynamin 2 (DN-dynamin 2) overexpressing neural progenitors cultured at 30 °C or 37 °C (mean + SD, n = 3 biologically independent samples in each group). j Notch reporter activity in control and Dll4-overexpressing neural progenitors cultured at 37 °C (mean + SD, n = 4 biologically independent samples in each group). k Schematic illustration of GFP or Dll4 overexpression in the developing chick pallium. I Distribution of Sox2 and/or GFP-positive cells in the developing chick pallium in control and Dll4 electroporated samples. m The proportion of Sox2-negative and GFP-positive cells in controls and Dll4-overexpressing chick pallium at 37 °C (means + SD, n = 3 biologically independent samples in each group). Two-sided, unpaired t-test for a, j, and m (p-values are indicated in each graph), two-sided, negative bimodal distribution for b (p-values were adjusted for multiple comparisons by using Benjamin–Hochberg method, FDR of Dll1 is not true), ordinary one-way ANOVA for d and f–i (p-values were adjusted by Tukey test for multiple comparisons). Scale bar: 25 µm.

Fig. 3. Enrichment of cell surface PE mediates Notch signaling.

a, b PE (a) and PC (b) molecule contents of chick neural progenitors cultured at 30 °C or 37 °C (mean + SD, n = 3 biologically independent samples in each group). c–h Fatty acid composition of PE (32:1, c; 34:1, d; 34:2, e) or PC (32:1, f; 34:1, g; 36:1, h) in chick neural progenitors cultured at 30 °C. i Relative expression levels of enzymes involved in PE synthesis in chick neural progenitors cultured at 30 °C or 37 °C (mean + SD, n = 3 biologically independent samples in each group). j Biosynthesis of PE through the Psd pathway. k Labeling of cell surface PE by SA-Ro treatment in nonpermeable condition. l, m SA-Ro fluorescent labeling of chick neural progenitors cultured at 30 °C (l) or 37 °C (m). Two-sided, unpaired t-test for a, b, two-sided, negative bimodal distribution for i (p-values were adjusted for multiple comparisons by using Benjamin–Hochberg method). P-values are indicated in each graph.

Fig. 4. PE mediates hypothermia-dependent increases in Notch activity.

a Trapping of cell surface PE in cultured chick neural progenitors by treatment with SA-Ro for 12 h. b Notch reporter activity in chick neural progenitors treated with SA-Ro at different temperatures (mean + SD, n = 3 biologically independent samples in each group). c Western blotting of Myc-tagged chick Psd1 in HEK293T cells transfected with siRNAs. d Introduction of the Notch reporter vector and siRNA targeting Psd1 in isolated chick neural progenitors. e Notch reporter activity in chick neural progenitors transfected with control or Psd1 siRNAs (mean + SD, n = 3 biologically independent samples in each group). f Schematic illustration of cPsd1 siRNA introduction in the developing chick pallium. g Distributions of GFP-positive cells in control and cPsd1 siRNA overexpressing chick pallium incubated at 30 °C. h The proportion of Sox2-negative and GFP-positive cells in control and cPsd1 siRNA overexpressing chick pallium at 30 °C (means + SD, n = 3 biologically independent samples in each group). Ordinary one-way ANOVA for b (p-values were adjusted by Tukey test for multiple comparisons), two-sided, unpaired t-test for e, h. P-values are indicated in each graph. Scale bars: 25 µm.

Fig. 5. Temperature robustness of Notch signaling in mouse neural progenitors.

a Schematic illustration of the temperature shift experiment conducted on the developing mouse brain. b, c Distribution of Ki67-positive and/or EdU-positive cells in the developing mouse neocortex cultured at 30 °C (b) or 37 °C (c). d Distribution of GFP-positive cells in the developing mouse neocortex cultured at 30 °C or 37 °C. e, f The proportion of Ki67/EdU-positive cells (e) and Sox2-negative/GFP-positive cells (f) in the developing mouse neocortex cultured at 30 °C or 37 °C (mean + SD, n = 3 biologically independent samples in each group). g Monitoring of Notch activity in mouse neocortical neural progenitors cultured at 30 °C or 37 °C for 24 h. h Notch reporter activities in mouse neural progenitors cultured at 30 °C or 37 °C (means + SD, n = 6 biologically independent samples in each group). i Relative expression levels of genes encoding Notch1 and Dll1 in mouse neural progenitors cultured at 30 °C or 37 °C (means + SD, n = 4 biologically independent samples in each group). j NICD-dependent Notch reporter activities in mouse neural progenitors cultured at 30 °C or 37 °C (means + SD, n = 3 biologically independent samples in each group). k Notch reporter activities in control and dominant-negative dynamin 2 (DN-dynamin 2)-overexpressing neural progenitors cultured at 30 °C or 37 °C (means + SD, n = 3 biologically independent samples in each group). I, m Notch reporter activity in control neural progenitors and neural progenitors overexpressing Dll1 lacking intracellular domain (Dll1ΔC) cultured at 37 °C or 30 °C (mean + SD, n = 3 biologically independent samples in each group). Two-sided, unpaired t-test (e, f, h, j, l, m), ordinary one-way or two-way ANOVA for k and i (p-values were adjusted by Tukey or Sidak test for multiple comparisons). Scale bars: 20 µm.

Fig. 6. Species-specific susceptibility to altered Notch signaling in amniote pallium.

a–f Transient NICD overexpression in the developing mouse neocortex (a–c) and chick pallium (d–f) by in utero or in ovo electroporation. a Schematic illustration of in utero electroporation. b, c Distribution of GFP-positive neurons in the developing mouse neocortex electroporated with the control or NICD expression vector (mean + SD, n = 3 biologically independent samples in each group). d Schematic illustration of the in ovo electroporation experiment. e, f Distribution of GFP-positive neurons in the developing chick dorsal pallium electroporated with control or NICD expression vector (mean + SD, n = 3 biologically independent samples in each group). g Summary of the species-specific temperature responsiveness of Notch signaling, PE synthesis, and the effects of increased Notch signaling in the developing amniote pallium (^¶Refs. ^,). Two-way ANOVA for c, f (p-values were adjusted by Tukey test). Scale bars: 100 µm.