VEGF‐B signaling impairs endothelial glucose transcytosis by decreasing membrane cholesterol content

Figure 1. VEGF‐B regulates endothelial glucose uptake and transcytosis 2‐NBD‐glucose uptake in murine pancreatic endothelial cells (MS1) and primary human umbilical vein endothelial cells (HUVEC) treated for 2 h with VEGF‐B167 (B167) or VEGF‐B186 (B186) compared to control treated cells. Data presented as mean ± SEM from representative experiments performed in triplicates. Statistical evaluation using one‐way ANOVA and Dunnett's multiple comparison test, P‐value: *< 0.05, **< 0.01, ***< 0.001 (compared to untreated control). Scale bar, 20 μm.Trans‐cellular 2‐NBD‐glucose transport in MS1 cells treated for 2 h with VEGF‐B186. Data presented as mean ± SEM from three individual experiments. Statistical evaluation using t‐test, P‐value: **< 0.01 (compared to untreated control).Cardiac glycogen content in 17‐week‐old male Vegfb +/+ (n = 6), Vegfb +/− (n = 6), and Vegfb −/− mice (n = 5) presented as mean ± SEM. Statistical evaluation using one‐way ANOVA and Dunnett's multiple comparison test, P‐value: *< 0.05 (compared to Vegfb +/+).Glucose uptake and accumulation (18F‐DG PET) in hearts of HFD‐fed wild‐type (WT) mice pre‐treated for 1 week with 2H10 antibody (n = 4) compared to control antibody…

Figure EV1. VEGF‐B regulates fatty acid and glucose uptake in endothelial cells of different origin Bodipy‐C12‐FA uptake in murine pancreatic endothelial cells (MS1, panels to the left) and primary human umbilical vein endothelial cells (HUVEC, panels to the right) treated for 2 h with VEGF‐B167 (B167) or VEGF‐B186 (B186). Data presented as mean ± SEM from representative experiments performed in triplicates. Statistical evaluation using one‐way ANOVA and Dunnett's multiple comparison test, P‐value: *< 0.05, **< 0.01 (compared to untreated control). Scale bar, 20 μm.2‐NBD‐glucose uptake (panels to the left) and Bodipy‐C12‐FA uptake (panels to the right) in primary human brain microvascular endothelial cells (HBMEC) treated for 2 h with VEGF‐B167 (B167) or VEGF‐B186 (B186). Data presented as mean ± SEM from representative experiments performed in triplicates. Statistical evaluation using one‐way ANOVA and Dunnett's multiple comparison test, P‐value: *< 0.05, **< 0.01 (compared to untreated control). Scale bar, 20 μm.2‐NBD‐glucose uptake (panels to the left) and Bodipy‐C12‐FA uptake (panels to the right) in primary human cardiac microvascular endothelial cells (HCMEC) treated for 2 h with…

Figure 2. VEGF‐B regulated glucose uptake is mediated by GLUT1, NRP1, and VEGFR1 Relative mRNA expression of the Slc2a glucose transporter (GLUT) family members (1‐13) in cardiac ECs derived from Vegfb +/+ (n = 4) and Vegfb −/− (n = 4) mice. Data represent normalized PLIER values (mean ± SEM) from microarray analysis performed on n = 4+4 RNA samples (one RNA sample/mouse).mRNA expression levels of the SLC2A transporter (GLUT) family members (1‐14) in HUVECs. Data presented as mean ± StDev relative to RPL19 expression from 3 independent experiments.mRNA expression levels of VEGFB, SLC2A1, NRP1, and FLT1 (VEGFR1) in HUVECs. Data presented as mean ± StDev relative to RPL19 expression from 3 independent experiments.2‐NBD‐glucose uptake in response to 2 h treatment with VEGF‐B186 (B186) in HUVECs exhibiting siRNA‐mediated knockdown of FLT1, NRP1, SCL2A1, KDR (VEGFR2), or control siRNA. Data presented as mean ± SEM from three individual experiments. Statistical evaluation using one‐way ANOVA and Sidak′s multiple comparison test, P‐value: *< 0.05 (comparisons made to respective untreated control).Representative images of GLUT1 (left panels) and CD31 (right panels) immunohistochemistry performed…

Figure EV2. VEGF‐B regulated glucose uptake is mediated by GLUT1, NRP1, and VEGFR1 2‐NBD‐glucose uptake in response to 2 h treatment with VEGF‐B186 (B186) in murine pancreatic endothelial cells (MS1) exhibiting stable shRNA‐mediated knockdown of Flt1 (VEGFR1), Nrp1, or control shRNA (left panel). Data presented as mean ± SEM from a representative experiment performed in triplicates. Statistical evaluation using one‐way ANOVA and Sidak's multiple comparison test, P‐value: *< 0.05 (compared to respective control). ShRNA‐mediated knockdown of Flt1 (VEGFR1) and Nrp1 expression (right panel).2‐NBD‐glucose uptake in response to 2‐h treatment with VEGF‐B186 (B186) in the absence or presence of isopropyl β‐D‐1‐thiogalactopyranoside (IPTG) in murine pancreatic endothelial cells (MS1) exhibiting inducible shRNA‐mediated knockdown of Slc2a1 (Glut1) (left panel). Data presented as mean ± SEM from a representative experiment performed in triplicates. Statistical evaluation using one‐way ANOVA and Sidak's multiple comparison test, P‐value: *< 0.05 (compared to respective control). ShRNA‐mediated knockdown of GLUT1 protein expression (right panel).2‐NBD‐glucose uptake in response to 2 h treatment with…

Figure EV3. VEGF‐B signaling does not change endothelial energy substrate use or energy balance A, BMeasurements of oxygen consumption rate (OCR, panel A) and extracellular acidification rate (ECAR, panel B) in primary human umbilical vein endothelial cells (HUVEC) stimulated or not with VEGF‐B186 (B186) for 2 h using the Seahorse® live‐cell metabolic assay platform. Glucose, oligomycin and rotenone + antimycin A (Rot/AA) was added at the indicated time points. Data presented as mean ± SEM from a representative experiment performed with n = 43–44 wells/condition.C, DComparison of oxidative (mitochondrial) versus glycolytic ATP production rate after VEGF‐B stimulation during basal (glucose‐free) or glucose‐induced conditions. Data presented as mean ± SEM from a representative experiment performed with n = 43–44 wells/condition.

Figure EV4. VEGF‐B signaling does not change expression of genes involved in endothelial cell metabolism RNA from primary human brain microvascular endothelial cells (HBMEC) stimulated with VEGF‐B167, VEGF‐B186, or vehicle (Control) for 6 h (n = 3 cell dishes/condition) were subjected to microarray hybridization using Affymetrix GeneChip® Human Gene 1.0 ST chips. Shown is a heat map over selected genes involved in intermediary metabolism (listed in Appendix Table S3). The three differentially expressed genes in response to VEGF‐B stimulation are indicated with arrow heads.

Figure 3. VEGF‐B signaling changes microdomain localization of GLUT1 in the plasma membrane Immunofluorescence labeling of GLUT1 (white), early endosomes (EEA1, green), and nuclei (blue) in HUVECs treated with VEGF‐B186 for 10, 30 min, or 2 h. Scale bar, 10 μm.GLUT1 protein expression in HUVECs treated with VEGF‐B186 for 15 min or 2 h. Data presented as mean ± StDev of 3–4 independent experiments relative to β‐Actin expression. Statistical evaluation using t‐test revealed no significant changes. Western blots shown in right panel.Sucrose gradient (15–50% sucrose) fractions of HUVEC cellular lysates after 2 h VEGF‐B186 stimulation were analyzed with Western blots shown in Appendix Fig S2. Different cellular sub‐compartments were identified by accumulation of specific protein markers: EEA1 for early endosomes, LIMPII for late endosomes/lysosomes, flotillin for plasma membrane/lipid rafts, and calnexin for endoplasmic reticulum (ER) (left panels). Amount and subcellular distribution of VEGFR1, NRP1, and GLUT1 protein in control or VEGF‐B186 stimulated HUVECs shown in panels to the right. Data represent % protein relative to total protein of interest measured from a representative…

Figure 4. VEGF‐B reduces membrane cholesterol prior to reduction of glucose uptake and increased FA uptake ARepresentative confocal (gray) and epifluorescence (blue) images of filipin staining in HUVECs in response to 2 h VEGF‐B186 stimulation (left panels). Cellular cholesterol content measured by quantification of fluorescent filipin staining. Data presented as mean ± SEM from a representative experiment performed in triplicates. Statistical evaluation using t‐test, P‐value: *< 0.05. Scale bars, 10 μm.B–DMeasurement of cholesterol content (B), glucose uptake (C), and FA uptake (D) in HUVECs after VEGF‐B186 stimulation for different time points. Data presented as mean ± SEM from representative experiments performed in triplicates. Statistical evaluation using t‐test, P‐value: *< 0.05, **< 0.01, ***< 0.001 (compared to respective control).EColorimetric quantification of non‐esterified cholesterol extracted from HUVECs stimulated with VEGF‐B186 for 20 min. Data are presented as mean ± SEM of three individual experiments. Statistical evaluation using t‐test, P‐value: *< 0.05.FUptake of a fluorescent cholesterol tracer (TopFluor) in HUVECs after VEGF‐B186 stimulation for 20 min or 2 h. Data…

Figure 5. Cholesterol content and glucose uptake are tightly linked in ECs Measurement of cellular cholesterol content (upper panel) and glucose uptake (lower panel) in HUVECs treated with VEGF‐B186 for 1 h or after 15‐min treatment with cholesterol‐extracting methyl‐beta‐cyclodextrin (MbCD). Data are presented as mean ± SEM from a representative experiment performed in triplicates. Statistical evaluation using one‐way ANOVA and Dunnett's multiple comparison test, P‐value: ***< 0.001 (compared to untreated control).Measurement of cellular cholesterol content (upper panel) and glucose uptake (lower panel) in HUVECs treated with VEGF‐B186 for 15 min alone or supplemented with soluble methyl‐beta‐cyclodextrin complexed with cholesterol (5, 10 or 20 μg/ml cholesterol). Data presented as mean ± SEM from a representative experiment performed in triplicates. Statistical evaluation using one‐way ANOVA and Dunnett's multiple comparison test, P‐value: *< 0.05, **< 0.01, ***< 0.001 (compared to control w/o cholesterol).Data information: See also Fig EV5.

Figure EV5. Fatty acid uptake by VEGF‐B is cholesterol‐ and LDLR‐independent Bodipy‐C12‐FA uptake (left panel) and corresponding measurement of cellular cholesterol content (right panel) in primary human umbilical vein endothelial cells (HUVEC) treated for 2 h with VEGF‐B186 (B186) or 15 min with cholesterol‐extracting methyl‐beta‐cyclodextrin (MbCD). Data presented as mean ± SEM from representative experiments performed in triplicates. Statistical evaluation using one‐way ANOVA and Dunnett's multiple comparison test, P‐value: **< 0.01, ***< 0.001 (compared to untreated control).Bodipy‐C12‐FA uptake (left panel) and corresponding measurement of cellular cholesterol content (right panel) in primary human umbilical vein endothelial cells (HUVEC) treated for 2 h with VEGF‐B186 or cholesterol‐extracting methyl‐beta‐cyclodextrin (MbCD). Data presented as mean ± SEM from representative experiments performed in triplicates. Statistical evaluation using one‐way ANOVA and Dunnett's multiple comparison test, P‐value: *< 0.05, **< 0.01, ***< 0.001 (compared to untreated control).Knockdown of LDLR and SCARB1 (SRB1) mRNA expression following siRNA targeting in primary human umbilical vein endothelial…

Figure 6. VEGF‐B regulates endothelial cholesterol handling in an LDL receptor‐dependent manner AmRNA expression analysis of lipoprotein receptors in HUVECs. Data presented as mean ± StDev relative to RPL19 expression from 3 independent experiments.B, CCellular cholesterol content (B) and glucose uptake (C) in HUVECs treated with control siRNA or siRNA targeting LDLR or Scavenger receptor B1 (SCARB1, SRBI) followed by 15 min stimulation with VEGF‐B186. Data presented as mean ± SEM from representative experiments performed in triplicates. Statistical evaluation using one‐way ANOVA and Dunnett's multiple comparison test, P‐value: *< 0.05, **< 0.01, ***< 0.001 (compared to respective control).DCellular cholesterol content of ECs freshly isolated from Ldlr +/+ and Ldlr −/− mice stimulated for 15 min with VEGF‐B186. Data presented as mean ± SEM from a representative experiment performed in triplicates. Statistical evaluation using one‐way ANOVA and Tukey's multiple comparison test, P‐value: ***< 0.001 (compared to Ldlr +/+ control).ECellular cholesterol content (left panel) and glucose uptake (right panel) of HUVECs treated with de novo cholesterol synthesis inhibitor simvastatin. Data…

Figure 7. VEGF‐B regulates cardiac cholesterol content in vivo Quantification of non‐esterified cholesterol content by means of filipin staining in the blood vessel compartment (podocalyxin+) in hearts of 15‐week‐old male C57BL/6 Vegfb +/− (n = 5) and Vegfb −/− mice (n = 7), normalized to wild‐type (Vegfb +/+) mice (n = 8). Data presented as box‐and‐whisker plots. Boxes represent lower/upper quartiles with the median values indicated with a horizontal line. Whiskers represent min–max values. Statistical evaluation using one‐way ANOVA and Dunnett's multiple comparison test, P‐value: *< 0.05, **< 0.01 (compared to Vegfb +/+ mice).Quantification of non‐esterified cholesterol content by means of filipin staining in hearts of 15‐week‐old male C57BL/6 Vegfb +/− (n = 5) and Vegfb −/− (n = 7) mice, normalized to wild‐type (Vegfb +/+) mice (n = 8). Data presented as box‐and‐whisker plots. Boxes represent lower/upper quartiles with the median values indicated with a horizontal line. Whiskers represent min–max values. Statistical evaluation using one‐way ANOVA and Dunnett's multiple comparison test, P‐value: **< 0.01 (compared to Vegfb +/+ mice).Quantification of non‐esterified cholesterol content by…