Flotillin proteins recruit sphingosine to membranes and maintain cellular sphingosine-1-phosphate levels

Figures

• 

  Fig 1. Loss of flotillins causes a reduction in cellular and tissue sphingosine-1-phosphate. A. Quantitative lipid mass spectrometry using calibration with defined standards, of WT and flotillin knockout MEFs. 1x106 cells were analysed in each sample.; Cer, Ceramide; SM, sphingomyelin; Sph, sphingosine; S1P, sphingosine-1-phosphate. Data from 4 biological replicates, each data point being derived from a separate experiment. Bars are mean. Statistical test is one-way ANOVA with Dunnett’s multiple comparisons test. B. S1P levels in mouse tissues and serum as shown, determined by quantitative lipid mass spectrometry. Bars are means and SD, each point represents a single sample from a different animal. Student’s t-test. C. Western blot to show reduction in flotillin 2 expression in cells treated with siRNAs against flotillin 1 and flotillin 2. Full scans of the blot are included in S3 Data File. D. siRNAs against flotillin 1 and flotillin 2 cause a reduction in S1P levels in HeLa cells as determined by quantitative lipid mass spectrometry. Data are normalised so that the mean of the control samples = 1, the data shown are from two separate experiments, in each experiment 3 biological…

• 

  Fig 2. Deletion of flotillin genes replicates a known cellular phenotype of reduced S1P. A Histone H3K9 acetylation in WT, Flot1-/-, Flot2-/-, SphK1-/- and SphK2-/- primary MEFs was analysed by Western blotting. B. Blots from 3 separate experiments quantifying H3K9 acetylation as A were quantified by densitometric scanning, and values normalised so that intensity in the WT sample = 1. Bars are SD. The same normalisation procedure is carried out in all quantitative graphs in this Fig. Significance is assessed using one-way ANOVA and Dunnett’s multiple comparisons test. C Histone expression and acetylation in WT and Flot1-/-, Flot2 primary MEFs was analysed by Western blotting using the antibodies shown. D. Blots from 3 separate experiments quantifying histone acetylation as C were quantified by densitometric scanning. Bars are SD. E. Isolated nuclei were treated with 20 μM Sph or S1P as shown. Samples were then analysed by Western blotting to detect total histone H3 and histone H3K9 acetylation. Blots representative of 4 experiments are shown. F. Quantification of H3K9 acetylation as E, Bars SD. Significance is assessed using one-way ANOVA and Dunnett’s multiple comparisons test to…

• 

  Fig 3. Isg15 levels are reduced in flotillin knockouts, and after perturbation of SphK activity. A. Isg15 protein expression in WT and Flot1-/-, Flot2-/- primary MEFs was analysed by Western blotting using antibodies against the indicated proteins. B. Densitometric quantification of Isg15 expression as A, Bars SD. One-way ANOVA and Dunnett’s multiple comparisons test. C. Isg15 protein expression in WT, Flot1-/- Flot2-/-, Sphk1-/-, and Sphk2-/- primary MEFs was analysed by Western blotting using antibodies against the indicated proteins. D. Quantification of Isg15 expression as C, Bars SD. One-way ANOVA and Dunnett’s multiple comparisons test. E. Isg15 expression levels in WT MEFs cells treated with SphK1 inhibitor PF543 (10 μM) and with SphK2 inhibitor ABC294640 (50 μM) for 24h as shown. F. Quantification of Isg15 expression as E, Bars SD. Student’s t-test. G. Isg15 expression in HeLa cells transfected with siRNAs to reduce expression of SphK1 and SphK2. H. Quantification of Isg15 expression as G, Bars SD. One-way ANOVA and Dunnett’s multiple comparisons test. Full scans of all blots are shown in S3 Data File.

• 

  Fig 4. Flotillins bind to 3H-Sph. A. 3H-Sph distribution into cytosolic and membrane franctions in WT and Flot1-/-, Flot2-/- MEFs. Data are means with SD, pooled data from 2 separate experiments, each point is a separate replicate, Student’s t-test. B. As A, but lipid extracts from membrane and cytosolic fractions were further fractionated by thin-layer chromatography before isolation of the specific Sph band. C. 3H-Sph binding to in vitro translated and affinity purified flotillin-1-GFP and flotillin-2-GFP proteins. 3H-Sph binding was competed with excess of cold Sph, ceramide (Cer) or S1P. Each point is a separate replicate, pooled data from 3 separate experiments. Data are means with SD, Student’s t-test. D. 3H-Sph binding to in vitro translated and affinity purified flotillin-1-GFP and flotillin-2-GFP truncation mutants. The graphical representations of flotillin 1 and flotillin 2 sequences show in black regions of hydrophobicity > 0.25 using Kyte–Doolittle values averaged over a window of 15 amino acid residues. Precise sizes of all constructs are given in the Methods section. Statistical comparisons apply one way ANOVA with Tukey’s multiple comparison test.

• 

  Fig 5. Flotillins recruit 3H-Sph to specific cellular membrane domains. A. Detergent resistant membrane (DRM) isolation by sucrose gradient centrifugation of 3H-Sph labelled WT and Flot1-/- primary MEFs. 3H-Sph levels in fractions from the gradient were analysed by liquid scintillation counting. Data are means with SEM, N = 3 separate experiments. Western blots below the graph shows caveolin 1 distribution in the gradient fractions. Full scans of all blots are shown in S3 Data File. B. 3H-cholesterol distribution into DRMs in WT and Flot1-/- MEFs. Data are means with SEM. Western blots below the graph show flotillin 1 distribution in the gradient fractions. C. Overexpression of flotillins increases Sph accumulation in HeLa cells. HeLa cells were transfected with GFP-tagged proteins and the overexpressing cells were FACS sorted and incubated with 3H-Sph. Radioactivity was measured with liquid scintillation. Data are means with SEM, N = 3 separate experiments. One way ANOVA with Dunnett’s multiple comparisons test, Flot-GFP expressing cells compared to all other groups. D. Ectopic re-localisation of flotillins to mitochondria. Flotillin-1-GFP is co-expressed with either flotillin 2 or…