Fig 2: Sphingomyelin accumulation increases membrane fluidity and induces expression of Nestin in SIRT1 KO mESCs.(A) SIRT1 KO mESCs have an increased membrane fluidity. WT and SIRT1 KO mESCs cultured in ESGRO medium were preincubated with or without 2.5 mM MßCD for 1 hr, then stained with 5 µM di-4-ANEPPDHQ for at least 30 min. The relative ordered fraction in each group was analyzed as described in Materials and methods (n = 30 clones/group, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001). Bars: 10 µm. (B–C) Exogeneous sphingomyelin treatment increases Nestin but not pluripotency markers in mESCs. WT and SIRT1 KO mESCs were treated with indicated concentrations of sphingomyelin (SM) in ESGRO medium for 48 hr. (B) The intensity of AP was analyzed as described in Materials and methods. Bars: 100 µm. (C) The protein abundance of pluripotency marker OCT3/4, Nanog and neuroepithelial stem cell marker Nestin in WT and Sirt1 KO mESCs were determined by immunoblotting. (D) Overexpression of WT but not a catalytic inactive mutant SMPDL3B reduces the expression of Nestin in SIRT1 KO mESCs. WT and SIRT1 KO mESCs transfected with an empty vector (V), a construct expressing WT SMPDL3B protein (SMPDL3B WT), or a construct expressing a catalytic inactive mutant SMPDL3B protein (SMPL3B H135A) were stained for SMPDL3B and Nestin. Scale bars: 20 µm.Figure 6—source data 1.Numerical data for bar graphs in A.Figure 6—source data 2.Uncut immunoblots in C.

Fig 3: Reduced expression of SMPDL3B is partially responsible for impaired in vitro neural differentiation in SIRT1 KO mESCs.(A) Overexpression of SMPDL3B partially rescues gross neuronal morphology in in vitro differentiated SIRT1 KO mESCs. WT and SIRT1 KO mESCs stably infected with lentiviral particles containing empty vector (V) or constructs expressing SMPDL3B protein were subjected to 4 weeks of in vitro neural differentiation. The cell morphology was analyzed using regular light microscopy fixed with ZEISS AxioCamHR camera. Scale bars: 20 µm. (B) Overexpression of SMPDL3B partially rescues neuronal morphology in in vitro differentiated SIRT1 KO cells. WT and SIRT1 KO mESCs expressing vector (V) or SMPDL3B were differentiated for 4 weeks or 6 weeks. Six weeks of differentiation is for a better morphological analysis. The expression of TUBB3 and neuronal morphology were analyzed by immunofluorescence staining. Scale bars: 20 µm. (C) Overexpression of SMPDL3B partially increased the fraction of differentiated cells in in vitro differentiated SIRT1 KO cells. WT and SIRT1 KO mESCs expressing vector (V) or SMPDL3B were differentiated as in (A). The fraction of differentiated cells positive of indicated neural markers were quantified by FACS (n = 3 biological replicates, **p<0.01, ***p<0.001). (D) Overexpression of SMPDL3B partially rescues the expression of neural markers in in vitro differentiated SIRT1 KO cells. WT and SIRT1 KO mESCs expressing vector (V) or SMPDL3B were differentiated as in (A). The expression of indicated neural markers were analyzed by immunofluorescence staining. Scale bars: 20 µm. (E) Overexpression of SMPDL3B partially rescues the expression of neural progenitor markers in in vitro differentiated SIRT1 KO cells. WT and SIRT1 KO mESCs expressing vector (V) or SMPDL3B were differentiated as in (A). The expression of SOX2 and SOX3 were analyzed by qPCR (n = 3 biological replicates, *p<0.05, **p<0.01). (F) Knocking down SMPDL3B in WT mESCs impairs neural differentiation in vitro. WT and SIRT1 KO mESCs with or without stable knockdown of SMPDL3B were in vitro differentiated into neurons for 4 weeks. The expression of neural markers Tau and NEFH were analyzed by immunofluorescence staining. Scale bars: 50 µm. (G) WT but not a catalytic inactive SIRT1 rescues neural differentiation in vitro. WT and SIRT1 KO mESCs expressing vector (V), WT SIRT1, or a mutant SIRT1 lacking catalytic activity (HY) were in vitro differentiated into neurons for 4 weeks. The expression of neural markers SOX1 and Nestin were analyzed by immunofluorescence staining. Scale bars: 50 µm.Figure 8—source data 1.Numerical data for bar graphs in C and E.

Fig 4: SIRT1 promotes the transcription of Smpdl3b through c-Myc in mESCs.(A) SIRT1 KO mESCs have reduced transcription of Smpdl3b. WT and SIRT1 KO mESCs cultured in ESGRO medium were crosslinked and subjected for ChIP-qPCR profiling of PolII, c-Myc, EZH2, and indicated chromatin activation or repression marks near the TSS region of Smpdl3b gene (n = 4 biological replicates, *p<0.05, **p<0.01, ***p<0.001). (B) Association scores of potential transcription factors (TFs) near the TSS of Smpdl3b gene. The association scores of indicated TFs were obtained from a published dataset (Evans et al., 2014). A higher score is suggestive of a higher chance of Smpdl3b gene being targeted by the potential TF. (C) A guide RNA (gRNA) targeting the +528 locus at the TSS region of Smpdl3b gene rescues the expression of this gene. sgRNAs targeting indicated loci near the TSS region of Smpdl3b gene were transfected into WT and SIRT1 KO mESCs stably expressing a dox-inducible dCas9 and BirA-V5 (dCas9 mESCs). The mRNA levels of Smpdl3b were analyzed by qPCR (n = 3 biological replicates, *p<0.05, **p<0.01, ***p<0.001). (D) Inhibition of c-Myc activity reduces the expression of Smpdl3b gene in mESCs. WT and SIRT1 KO mESCs were treated with DMSO or 10 mM 10058-F4 for 48 hr. The mRNA levels of Smpdl3b were analyzed by qPCR (n = 3 biological replicates, *p<0.05, **p<0.01, ***p<0.001). (E) Knocking down c-Myc significantly reduces the expression of Smpdl3b gene in mESCs. WT and SIRT1 KO mESCs were transfected with siRNAs against c-Myc for 48 hr. The mRNA levels of c-Myc and Smpdl3b were analyzed by qPCR (n = 3 biological replicates, *p<0.05, **p<0.01). (F) Overexpression of the KR mutant of c-Myc partially reduces the expression of Smpdl3b gene in SIRT1 KO mESCs. The mRNA levels of Smpdl3b in indicated mESCs were analyzed by qPCR (n = 6 biological replicates, **p<0.01, ***p<0.001). (G) Mutation of the c-Myc binding E-box element on the promoter of Smpdl3b gene abolishes the expression of Smpdl3b luciferase reporter in mESCs. Luciferase reporters containing the basic vector (pGL3-basic), the WT promoter of Smpdl3b gene, or a promoter of Smpdl3b gene with a mutant E-box were transfected into WT and SIRT1 KO mESCs, and the luciferase activities were measured as described in Materials and methods (n = 3 biological replicates, ***p<0.001).Figure 5—source data 1.Numerical data for graphs in A, B, C, D, E, F, and G.

Fig 5: Expression of WT but not catalytically inactive SIRT1 partially rescues the sphingomyelin defect in SIRT1 KO mESCs.(A–B) SIRT1 protein levels in indicated mESCs. WT and SIRT1 KO mESCs were infected with lentiviral particles containing empty vector (PLenti-III-EF1a) or constructs expressing WT or a catalytically inactive mutant SIRT1 (H355Y, HY). The expression of SIRT1 was analyzed by either (A) immunoblotting or (B) immunofluorescence staining. Bars in B: 20 µm. (C–D) Expression of the HY mutant SIRT1 represses the expression of Smpdl3b in WT mESCs, whereas expression of WT but not the HY mutant SIRT1 increases the expression of Smpdl3b in SIRT1 KO mESCs. The expression of SMPDL3B was analyzed by either (C) qPCR or (D) immunoblotting. (n = 3 biological replicates, ***p<0.001). (E–F) Expression of WT but not HY mutant SIRT1 significantly reduces the sphingomyelin levels in both WT and SIRT1 KO mESCs. Indicated WT and SIRT1 KO mESCs cultured in ESGRO medium were labeled with BODIPY FL-labeled sphingomyelin for 30 min at 4 °C then incubated at 37°C for 30 min. The intensity of BODIPY FL-labeled sphingomyelin in cells were analyzed by (E) confocal fluorescence imaging and by (F) quantitative FACS (n = 3 biological replicates, ***p<0.001). Bars in E: 20 µm.Figure 4—source data 1.Numerical data for bar graphs in C and F.Figure 4—source data 2.Uncut immunoblots in A and D.