Pericytes regulate VEGF-induced endothelial sprouting through VEGFR1

Abstract

Pericytes adhere to the abluminal surface of endothelial tubules and are required for the formation of stable vascular networks. Defective endothelial cell-pericyte interactions are frequently observed in diseases characterized by compromised vascular integrity such as diabetic retinopathy. Many functional properties of pericytes and their exact role in the regulation of angiogenic blood vessel growth remain elusive. Here we show that pericytes promote endothelial sprouting in the postnatal retinal vasculature. Using genetic and pharmacological approaches, we show that the expression of vascular endothelial growth factor receptor 1 (VEGFR1) by pericytes spatially restricts VEGF signalling. Angiogenic defects caused by pericyte depletion are phenocopied by intraocular injection of VEGF-A or pericyte-specific inactivation of the murine gene encoding VEGFR1. Our findings establish that pericytes promote endothelial sprouting, which results in the loss of side branches and the enlargement of vessels when pericyte function is impaired or lost.

Fig. 1

Temporally controlled pericyte depletion. a Experimental scheme of tamoxifen and diphtheria toxin (DT) administration to postnatal DTR ^iPC mice. b Confocal images showing P6 retina whole-mounts stained with isolectin B4 (IB4, red), collagen IV (colIV, white) and PDGFRβ (green). Note reduction of PDGFRβ+ cells at the DTR ^iPC angiogenic front. Scale bar, 100 µm. c High-magnification confocal images of the desmin (green) and IB4 (red) stained peripheral vasculature in the P6 DTR ^iPC and littermate control retina. Note extended morphology of desmin+ intermediate filaments in residual DTR ^iPC perivascular cells. Scale bar, 100 µm. d Maximum intensity projections of P6 control and DTR ^iPC retinas stained for NG2 (green) and IB4 (red). Images in the two right columns show higher magnifications of insets representing angiogenic front (AF, white dashed squares) and central plexus (CP, yellow-dashed squares), respectively. Scale bar, 200 µm (images on the left) and 100 µm (higher magnifications). e Quantitation of pericyte (PC) coverage in the control and DTR ^iPC angiogenic front and capillary plexus, as indicated. Error bars, s.e.m. p-values, Student’s t-test

Fig. 2

Altered endothelial sprouting after pericyte depletion. a Isolectin B4-stained P6 control and DTR ^iPC retina whole-mounts. Dashed circles indicate vessel-covered (yellow) and peripheral avascular (white) areas, respectively. Scale bar, 500 µm. b, c Quantitation of P6 control and DTR ^iPC body weight b and radial outgrowth of the retinal vasculature c. Error bars, s.e.m. p-values, Student’s t-test. d Confocal images of IB4 (white/red) and collagen IV (colIV, blue) showing the reduction of branch points but no increase in IB4- colIV+ empty matrix sleeves in DTR ^iPC P6 retinas. Scale bar, 200 µm (top panels) and 50 µm (bottom). e, f Quantitation of branch points e and empty matrix sleeves f in P6 retinas. Error bars, s.e.m. p-values, Student’s t-test. NS, not statistically significant. g Maximum intensity projection of the IB4-stained P6 control and DTR ^iPC retinal angiogenic front illustrating differences in sprout number, morphology and filopodia number after pericyte depletion. Scale bar, 50 µm. h, i Quantitation of sprouts h and filopodia number i in the P6 control and DTR ^iPC angiogenic front. Error bars, s.e.m. p-values, Student’s t-test

Fig. 3

Pericytes control endothelial behavior in angiogenesis. a Confocal images showing proliferating cells (EdU, blue), IB4 (red) and EC nuclei (Erg1, green) in the P6 control and DTR ^iPC retinal angiogenic front, as indicated. Images on the right show higher magnification of boxed insets in left column. Dashed lines outline the angiogenic leading vessel from where the sprouts emerge (second column) or vessel shape (fourth column) at the edge of the vascular plexus. Scale bar, 100 µm (left panels) and 50 µm (higher magnifications). b–d Quantitation of total b and regional EC proliferation c as well as EC density d within the peripheral vascular plexus of P6 control and DTR ^iPC retinas. Error bars, s.e.m. p-values, Student’s t-test. NS, not statistically significant. e High magnification of IB4 (red) and Erg1 (green) stained sprouts highlighting the accumulation of EC nuclei in P6 DTR ^iPC but not control sprouts. Scale bar, 50 µm. f Confocal images showing IB4 (red), NG2+ pericytes (green) and vessel lumen (ICAM2, white) in the control and DTR ^iPC angiogenic front of P6 retinas. Images on the right show higher magnification of boxed insets in the second column. Note that IB4 signal covers a larger area than the apical ICAM2 immunostaining (arrowheads in third column). Scale bar, 100 µm (left panels) and 50 µm (higher magnifications)

Fig. 4

Endothelial changes after pericyte depletion. a–f Maximum intensity projection of confocal images from control and DTR ^iPC P6 retinas stained for IB4 (red) in combination with VEGF-A a, VEGFR2 b, VEGFR3 c, Tie2 d, Esm1 e, and Dll4 f (all in white), as indicated. Note local increase of VEGFR2, VEGFR3, and Esm1 (arrowheads in b, c, e) but not Tie2 or VEGF-A at the edge of the vessel plexus. Dll4 expression in DTR ^iPC sprouts is increased in some regions (arrowheads) but absent in others (arrows). Scale bar, 100 µm. g–j Quantitation of VEGF-A immunosignals area and intensity g, signal intensity for VEGFR2 h and VEGFR3 i and proportion of Esm1+ area with respect to vascular area j in the P6 control and DTR ^iPC angiogenic front. Error bars, s.e.m. p-values, Student’s t-test

Fig. 5

Vascular alterations after intraocular VEGF-A injection. a Morphology of IB4-stained P6 wild-type retinal vessels at 4 h after administration of human VEGF-A165 (0.5 µl at a concentration of 5 μg μl⁻¹). Note blunt appearance of the vessel front after VEGF-A injection but not for vehicle (PBS) control. Scale bar, 200 µm. b Quantitation of sprouts and filopodia at the front of the P6 vessel plexus after injection of VEGF-A165 or vehicle control. Error bars, s.e.m. p-values, Student’s t-test. c PDGFRβ+ (green) pericytes are unaffected by short-term VEGF-A administration, whereas VEGFR2 immunosignals (white) are increased in IB4+ (red) ECs (arrowheads). Images shown correspond to insets in a. Scale bar, 100 µm. d Quantitation of VEGFR2 immunosignals intensity in the peripheral plexus of P6 retinas after injection of VEGF-A165 or vehicle control. Error bars, s.e.m. p-values, Student’s t-test. e Confocal images showing increased Esm1 immunostaining (white) in IB4+ (red) ECs in the peripheral plexus (arrowheads) after VEGF-A injection in P6 pups. Scale bar, 200 µm. f VEGF-A165 injection-mediated increase of Esm1 immunosignals (normalized to IB4+ EC area) in the peripheral capillary plexus but not at the edge of the angiogenic front in comparison to PBS-injected controls at P6. Error bars, s.e.m. p-values, Student’s t-test. NS, not statistically significant. g Short-term VEGF-A165 administration leads to clustering of Erg1+ (green) and IB4+ (red) ECs, as indicated, in thick sprout-like structures of P6 retinas. Panels in the center and on the right (scale bar, 20 µm) show higher magnification of the insets on the left (scale bar, 100 µm). Dashed lines in panels on the right outline IB4+ vessels. h Quantitation of EC density in the leading front vessel and emerging sprouts of the P6 angiogenic front after injection of VEGF-A165 or vehicle control. Error bars, s.e.m. p-values, Student’s t-test

Fig. 6

Expression of sFlt1 in pericytes at the angiogenic front. a Maximum intensity projections of confocal images from P6 retinas of the Hey1-GFP transgenic reporter mouse model stained for GFP (green), PDGFRβ+ (white) and IB4 (red). Images of the first row show enrichment of Hey1-GFP+, PDGFRβ+ perivascular cells in the angiogenic front in comparison to mural cells covering the remodeling central plexus around veins (middle row) and arteries (bottom row). Note strong expression of Hey1-GFP reporter in arterial ECs (bottom row). Scale bar, 50 µm. b, Quantitation of Pdgfrb expression by qPCR in P6 PDGFRβ+ retinal pericytes sorted based on GFP expression in comparison to whole-retina single-cell suspension (input). Note significant enrichment of Pdgfrb in both (GFP+ and GFP−) pericyte fractions and higher expression in the Hey1-GFP+ subset. Error bars, s.e.m. p-values, Kruskal–Wallis and Dunn’s multiple comparison test. NS, not statistically significant. c Quantitation of sFlt1 expression by qPCR in sorted P6 retinal pericytes in comparison to whole-retina single-cell suspension (input). Note significant enrichment of sFlt1 expression in Hey1-GFP+ pericytes in comparison to input and GFP- pericytes. Error bars, s.e.m. p-values, one-way ANOVA and Tukey’s multiple comparison test. NS, not statistically significant

Fig. 7

Inactivation of Flt1 in PDGFRβ+ cells. a Experimental scheme of tamoxifen administration for the generation of Flt1 ^iPC mutants. b P6 control, Flt1 ^iPC/+ and Flt1 ^iPC retinas stained with isolectin B4 (IB4). Dashed circles indicate vessel-covered (yellow) and peripheral avascular (white) areas in the overview pictures (top). Scale bar, 500 µm. c Quantitation of body weight and radial outgrowth of the retinal vasculature in control, Flt1 ^iPC/+ and Flt1 ^iPC P6 pups. Error bars, s.e.m. p-values, one-way ANOVA. NS, not statistically significant. d Confocal images of the IB4-stained P6 control, Flt1 ^iPC/+ and Flt1 ^iPC retinal angiogenic front illustrating differences in sprout number and morphology. Scale bar, 100 µm. e Quantitation of sprouts and filopodia in P6 control, Flt1 ^iPC/+ and Flt1 ^iPC retinas. Error bars, s.e.m. p-values, one-way ANOVA and Tukey’s multiple comparison test. NS, not statistically significant. f Confocal images of IB4 (red), Erg1 (green) and VEGFR2 (white) stained P6 retinas highlighting the accumulation of EC nuclei and enhanced VEGFR2 immunosignals (arrowheads) in Flt1 ^iPC sprouts. Vessels are outlined by dashed lines on the right panel. Scale bar, 100 µm. g Quantitation of EC proliferation (EdU+ Erg1+) at the angiogenic front, EC density in sprouts and leading front vessel and VEGFR2 immunosignals intensity in the angiogenic front of control and Flt1 ^iPC P6 retinas. Error bars, s.e.m. p-values, Student’s t-test. h Esm1 (white) expression (arrowheads) in the angiogenic front (IB4+, red, first two columns) and detection of desmin+ pericytes (green, third column) in P6 control and Flt1 ^iPC retinas. Scale bar, 100 µm. i Quantitation of Esm1+ proportion relative to vascular area (IB4+) in the angiogenic front of control and Flt1 ^iPC P6 retinas. Error bars, s.e.m. p-values, Student’s t-test. j Confocal images of P6 retinas stained for NG2 (green) and IB4 (red) showing no significant changes in pericyte coverage in the front (first two columns) or the remodeling plexus around veins (v) or arteries (a) (last two columns). Scale bar, 100 µm. k, l Quantitation of pericyte coverage k and relative gene expression by qPCR on whole lysates l in control and Flt1 ^iPC P6 retinas. Error bars, s.e.m. p-values, Student’s t-test. NS, not statistically significant