Volume 66, Issue 2, Pages 205-219 (April 2010) VEGFR2 (KDR/Flk1) Signaling Mediates Axon Growth in Response to Semaphorin 3E in the Developing Brain  Anaïs Bellon, Jonathan Luchino, Katharina Haigh, Geneviève Rougon, Jody Haigh, Sophie Chauvet, Fanny Mann  Neuron  Volume 66, Issue 2, Pages 205-219 (April 2010) DOI: 10.1016/j.neuron.2010.04.006 Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 1 VEGFR2 Protein Is Expressed along the Subiculo-mamillary Axon Tract (A) Schematic representation of the path taken by subiculo-mamillary projections in red, from the subiculum (Sub) through the fimbria (fim), fornix (f), and postcommissural fornix (pf). Before birth, axons in the postcommissural fornix extend beyond the mamillary bodies (mb), which are innervated by interstitial axon collaterals during the first postnatal weeks (Stanfield et al., 1987). (B–J) Immunolabeling of VEGFR2 (B, E, and H) and PECAM-1 (C, F, and I) on coronal sections of wild-type brain at E17.5. Expression of VEGFR2 is apparent throughout the subiculo-mamillary projection, in the fimbria (fim in B and D), fornix (f in E and G), and postcommissural fornix (pf in H and J). (K–M) Immunolabeling of VEGFR2 (K) and the axonal marker L1 (L) on coronal sections of wild-type brain at E17.5, confirming axonal localization of VEGFR2 (M). Inset panel in (M) shows overlap in individual axonal fascicles. (N–S) Immunolabeling of VEGFR2 (N and Q) and PlexinD1 (PlxD1) (O) or Nrp1 (R), two markers of subiculo-mamillary axons, on coronal sections of wild-type brain at E17.5. High magnification views of immunostaining at the level of the fornix show that VEGFR2 is colocalized with PlexinD1 (P) and Nrp1 (S) in subicular projections. Inset panel in (P) and (S) shows overlap in individual axonal fascicles. (T–V) Immunolabeling of VEGFR2 (T) and GFAP (U) on coronal sections of wild-type brain at E17.5. High magnification views of immunostaining at the level of the fimbria show that some VEGFR2-positive cells coexpress GFAP (V). Inset panel in (V) shows overlap in individual GFAP positive cells. Scale bars, 100 μm (B–G, K–V), 60 μm (H–J). Neuron 2010 66, 205-219DOI: (10.1016/j.neuron.2010.04.006) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 2 VEGFR2 Is Necessary for the Development of the Postcommissural Fornix (A–D) Coronal sections through the fornix (f) (A and B) and the caudal hypothalamus (C and D) of E17.5 Vegfr2lox/- control (A and C) and Vegfr2lox/-;Cre mutant (B and D) brains after anterograde DiI tracing of the subiculo-mamillary projections from the subiculum. No labeled axons can be observed in the postcommissural fornix of Vegfr2lox/-;Cre mutant embryos. (E and F) Coronal sections of E17.5 mouse brains showing PlexinD1 immunostaining in the postcommissural fornix in Vegfr2lox/− control (E), but not in Vegfr2lox/-;Cre mutant embryos (F). (G and H) Anti-neurofilament staining of coronal sections of mouse brains at P30. Reduced size of the postcommissural fornix in Vegfr2lox/-;Cre mice (H) as compared to control animals (G). (I and J) High magnification views of cross-sections through the postcommissural fornix at P30. A similar fiber density is observed in Vegfr2lox/-;Cre mice (J) as compared to control animals (I). (K) Quantification of the caudal extension of the postcommissural fornix from the fornix (0 μm) to the caudal hypothalamus (+600 μm) in Vegfr2lox/− (Control), Vegfr2lox/lox;Cre, and Vegfr2 lox/−;Cre mutants at E17.5. (L) Comparison of the cross-sectional surface area of the postcommissural fornix in Vegfr2lox/− (Control), Vegfr2lox/lox;Cre, and Vegfr2 lox/-;Cre mutants at P30. (M) Quantification of the density of NF160 positive axons in the postcommissural fornix of Vegfr2lox/− (Control) and Vegfr2 lox/−;Cre mutants at P30. (N and O) Schematic representation of VEGFR2 function in the prenatal development of the postcommissural fornix. In neural cell-specific Vegfr2 knockout embryos, the development of the subicular projections is massively delayed and no fibers reach the caudal hypothalamus at E17.5. f, fornix; pf, postcommissural fornix, Sub, subiculum; mb, mamillary bodies. Scale bars, 100 μm (A–D), 150 μm (E and F), 70 μm (G and H), 5 μm (I and J). See also Figure S1. Neuron 2010 66, 205-219DOI: (10.1016/j.neuron.2010.04.006) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 3 VEGFR2 Is Required for the Axon Growth-Promoting Effect Induced by Sema3E (A) RT-PCR analysis of Vegfr1-3 mRNAs expression in cultured neurons from the subiculum. PECAM-1 mRNA was used to control the absence of endothelial cells. GAPDH mRNA was used as an internal control. (B) Immunolabeling of cultured neurons from the subiculum and ventrolateral cortex with an anti-VEGFR2 antibody. (C) Typical images of dissociated subicular neurons cultured in the presence or absence of 10 μg/ml polyclonal anti-VEGFR2 and 5nM Sema3E. Blockade of VEGFR2 prevents stimulation of subicular axon growth by Sema3E. (D) Quantification of the results illustrated in (C). Histograms of Sema3E effects in cultures of subicular dissociated cells. Data are presented as mean axonal length ± SEM (n = 3) and are normalized to 100% for values obtained in control conditions. (E) Typical images of dissociated neurons cultured in the presence or absence of inhibitors of VEGFR2 (0.1 μM SU1438 or 0.1 μM Ki8751) and of 5 nM Sema3E. Blockade of VEGFR2 affects the growth-promoting effect of Sema3E on subicular axons. (F) Quantification of the results illustrated in (E). (G) Typical images of dissociated subicular neurons electroporated with GFP-expressing vector together either control siRNA, VEGFR2 siRNAs 1 and 2, or VEGFR3 siRNA (see Experimental Procedures) cultured in the presence or absence of 5 nM Sema3E. Knockdown of VEGFR2, but not VEGFR3, abolishes the growth-promoting effect of Sema3E. (H) Quantification of the results illustrated in (G). (I) Dissociated subicular neurons from Vegfr2lox/- (Control) and Vegfr2 lox/−;Cre mutants at E17.5 were cultured in the presence or absence of 5 nM Sema3E. Neurons from Vegfr2 mutants demonstrated a complete loss in their growth response to Sema3E. (J) Quantification of the results illustrated in (I). (K) Typical pattern of axonal outgrowth from E17.5 subicular explants from Vegfr2lox/− (Control) and Vegfr2 lox/−;Cre mutants cocultured with control and Sema3E-expressing HEK293T cells. (L) Quantification of the axonal guidance response of subicular neurons to control and Sema3E-expressing HEK293T cells. Data are expressed as P/D ratio, where P and D are the mean lengths of axons in the quadrant proximal and distal to the cell aggregate. A P/D ratio close to 1 indicates radial outgrowth. Subicular axons from Vegfr2lox/− (Control) are attracted by a source of diffusible Sema3E, whereas subicular axons from Vegfr2 lox/−;Cre mutants show no response. (M) Dissociated neurons of the ventrolateral cortex from Vegfr2lox/− (Control) and Vegfr2 lox/−;Cre mutants at E17.5 were cultured in the presence or absence of 2 μg/ml Nrp1-Fc and 5 nM Sema3E. Soluble Nrp1-Fc switches the growth response of control cortical neurons to Sema3E from inhibition to stimulation. The growth-promoting response induced by Sema3E in the presence of Nrp1-Fc is abolished in cortical neurons from Vegfr2 lox/−;Cre mutants. (N) Quantification of the results illustrated in (M). ∗∗∗significantly different with p < 0.001. Scale bars, 25 μm (B, C, E, G, I, and M), 100 μm (K). See also Figure S2. Neuron 2010 66, 205-219DOI: (10.1016/j.neuron.2010.04.006) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 4 VEGF Does Not Modulate Sema3E Axonal Growth Promotion (A) Typical images of dissociated subicular neurons cultured in the presence or absence of 5 nM Sema3E and of 10 μg/ml polyclonal anti-VEGFA164 and VEGFA120 antibody. (B) Quantification of the results illustrated in (A). Blockade of VEGF has no effect on axon growth promotion by Sema3E. Data are presented as mean axonal length ± SEM (n = 3) and are normalized to 100% for values obtained in control conditions. (C) Typical images of dissociated subicular neurons cultured in the presence or absence of 5 nM Sema3E and of VEGFA164 (10 to 100 nM). (D) Quantification of the results illustrated in (C). At high concentration, exogenous VEGFA164 partially diminishes the axonal growth response to Sema3E. ∗∗∗significantly different with p < 0.001. Scale bars, 25 μm. Neuron 2010 66, 205-219DOI: (10.1016/j.neuron.2010.04.006) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 5 Activation of VEGFR2/PlexinD1/Nrp1 Receptor Complex by Sema3E (A and B) Coimmunoprecipitation experiments showing that VEGFR2 can form a complex with PlexinD1 and Nrp1 proteins in both the absence (A) and the presence (B) of Sema3E. HEK293T cells were transfected with the indicated combinations of VSV-tagged PlexinD1, Flag-tagged VEGFR2, and HA-tagged Nrp1, and treated or not with Sema3E (10 nM). (C) Immunoprecipitation of ligand-receptor complexes (see Experimental Procedures) showing that Sema3E precipitates all three PlexinD1, Nrp1 and VEGFR2 proteins. (D) Binding of AP-Sema3E and AP-VEGFA164 to HEK293T cells expressing PlexinD1 or/and Nrp1 or/and VEGFR2. Binding of AP-Sema3E is observed on cells expressing PlexinD1, alone or in combination with Nrp1 and/or VEGFR2. Binding of AP-VEGFA164 to cells expressing VEGFR2 and/or Nrp1 confirms cell-surface expression of these proteins. (E) Tyrosine phosphorylation of VEGFR2 in HEK293T cells transfected with cDNA encoding Flag-VEGFR2 or/and VSV-PlexinD1 or/and Nrp1. Cells were stimulated with 10 nM Sema3E for 7 min. Lysates were immunoprecipitated with anti-Flag antibodies and immunoprecipitates were blotted with anti-phosphotyrosine antibodies directed against Tyr1173/1175 and Tyr1212/1214 residues in VEGFR2. A representative blot from three experiments is shown. (F and G) Quantification of Sema3E-induced phosphorylation of VEGFR2 at Tyr1173/1175 (F) and Tyr1212/1214 (G). This panel represents densitometric analysis of Sema3E-induced VEGFR2 phosphorylation from three individual experiments. Data are expressed as Rstim/Rcont ratio, where Rstim and Rcont are the mean phosphorylated VEGFR2 level in presence or absence of Sema3E, respectively, normalized to the total amount of VEGFR2 (see Experimental Procedures). A Rstim/Rcont ratio close to 1 indicates no variation in VEGFR2 phosphorylation following Sema3E stimulation. Enhancement of the Rstim/Rcont ratio indicates an increased phosphorylation upon Sema3E treatment. VEGFR2 phosphorylation at Tyr1173/1175 and Tyr1212/1214 is enhanced in responses to Sema3E in cells coexpressing PlexinD1, Nrp1, and VEGFR2. Scale bar, 100 μm. See also Figure S3. Neuron 2010 66, 205-219DOI: (10.1016/j.neuron.2010.04.006) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 6 VEGFR2 Is the Signal Transducing Component of the Growth Promoting Receptor Complex for Sema3E (A) Dissociated mouse subicular neurons were electroporated with control siRNA or siRNA to mouse Nrp1 (see Experimental Procedures) together with expression vectors encoding GFP, rat Nrp1, or rat Nrp1 deleted of the cytoplasmic domain (rNrp1ΔC) and cultured in the presence or absence of 5nM Sema3E. Knockdown of Nrp1 causes subicular neurons to switch their response to Sema3E from growth promotion to inhibition. Rat Nrp1 rescues the growth promoting response to Sema3E even in absence of the cytoplasmic domain. (B) Quantification of the results illustrated in (A). Data are presented as mean axonal length ± SEM (n = 3) and are normalized to 100% for values obtained in control conditions. (C) Typical images of dissociated subicular neurons cultured in the absence or presence of 5 nM Sema3E after electroporation with control siRNA or siRNA to mouse PlexinD1 together with expression vectors encoding GFP, human PlexinD1 (hPlxD1), or human PlexinD1 deleted of the cytoplasmic domain (hPlxD1Δc). Knockdown of PlexinD1 inhibits the response of subicular neurons to Sema3E. Human PlexinD1 rescues the growth promoting response to Sema3E even in absence of the cytoplasmic domain. (D) Quantification of the results illustrated in (C). (E) Typical images of dissociated subicular neurons cultured in the absence or presence of 5 nM Sema3E after electroporation of expression vectors encoding GFP, human VEGFR2 (hVEGFR2) or human signaling-defective VEGFR2 mutant (hVEGFR2Y1175F) and mouse VEGFR2 siRNA 1 (see Experimental Procedures). Knockdown of VEGFR2 blocks the response to Sema3E. Wild-type human VEGFR2, but not hVEGFR2Y1175F, rescues the growth promoting response to Sema3E. (F) Quantification of the results illustrated in (E). ∗∗∗significantly different with p < 0.001. Scale bars, 25 μm. See also Figure S4. Neuron 2010 66, 205-219DOI: (10.1016/j.neuron.2010.04.006) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 7 Sema3E Growth-Promoting Effect Is Mediated by Activation of the PI3K/Akt/GSK-3β Pathway via VEGFR2 (A) Typical images of dissociated subicular neurons cultured in the presence or absence of 5 nM Sema3E and 5 μM LY294002 or of 50 nM wortmannin (Wort). Both inhibitors abolish the growth response of subicular neurons to Sema3E. (B) Quantification of the results illustrated in (A). Data are presented as mean axonal length ± SEM (n = 3) and are normalized to 100% for values obtained in control conditions. (C) Typical images of dissociated subicular neurons cultured in the presence or absence of 5 nM Sema3E after electroporation with Akt siRNA. Knockdown of Akt causes subicular neurons to inhibit their response to Sema3E. (D) Quantification of the results illustrated in (C). (E) Typical images of dissociated subicular neurons cultured in the presence or absence of 5nM Sema3E after electroporation of expression vectors encoding GFP, a dominant-negative form of PI3K (DN PI3K), a dominant-negative form of Akt (DN Akt), or a constitutive active form of GSK-3β (CA GSK3). Blockade of the PI3K/AKT/GSK-3β pathway inhibits the ability of subicular neurons to respond positively to Sema3E. (F) Quantification of the results illustrated in (E). (G) Immunolabeling of growth cones from growing subicular neurons using anti-phosphotyrosine antibody to Tyr1173/1175 in VEGFR2 (αpY1175VEGFR2) or anti-phospho-Akt antibodies (αpAkt) or anti-phospho-GSK-3 antibodies (αpGSK3). Growth cones were treated 5 min with 10 nM of Sema3E in presence or absence of 10 μM VEGFR2 inhibitor (Ki8751). Sema3E induces phosphorylation of VEGFR2, Akt and GSK-3, in a VEGFR2-dependent manner. (H) Quantification of the immunostaining illustrated in (G). The values represent the mean pixel intensity of αpY1175VEGFR2, αpAkt, and αpGSK3 staining within growth cones (see Experimental Procedures). ∗∗∗significantly different with p < 0.001. Scale bars, 25 μm (A, C, and E), 7 μm (G). See also Figure S5. Neuron 2010 66, 205-219DOI: (10.1016/j.neuron.2010.04.006) Copyright © 2010 Elsevier Inc. Terms and Conditions

Figure 8 Reconstruction of Functional Receptor Complexes for Sema3E (A–C) Typical images of dissociated thalamic neurons cultured in the presence or absence of 5 nM Sema3E after electroporation with expression vectors encoding GFP, in combination with different expression vectors for VEGFR2, signaling-defective VEGFR2 mutant (VEGFR2Y1175F), PlexinD1 (PlxD1), cytoplasmic domain-deleted PlexinD1 mutant (PlxD1Δc), Nrp1, and cytoplasmic domain-deleted Nrp1 mutant (Nrp1ΔC). (D) Quantification of the results illustrated in (A)–(C). Data are presented as mean axonal length ± SEM (n = 3) and are normalized to 100% for values obtained in control conditions. (E–H) Molecular models for axonal growth responses to Sema3E. Sema3E inhibits axonal growth of neurons which express PlexinD1 alone (E). In contrast, axons of neurons expressing VEGFR2, PlexinD1, and Nrp1 show a growth promoting response to Sema3E (F). The intracellular domain of VEGFR2 is required to mediate the growth response to Sema3E (G and H), indicating that VEGFR2 is the signal transducing subunit of the complex. ∗∗∗significantly different with p < 0.001. Scale bars, 20 μm (A–C). See also Figure S6. Neuron 2010 66, 205-219DOI: (10.1016/j.neuron.2010.04.006) Copyright © 2010 Elsevier Inc. Terms and Conditions