Netrins Promote Developmental and Therapeutic Angiogenesis

METHODS

Migration Assays

Migration assays were performed as described (3-5). Briefly, 16 h before the assay, 80% confluent 75 cm² flasks (Corning Costar) of human microvessel endothelial cells (HMVEC; Cambrex, Walkersville, MD), human coronary artery endothelial cells (HCAEC; Cambrex), human umbilical artery endothelial cells (HUAEC; Promocell, Heidelburg, Germany), or human umbilical vein endothelial cells (HUVEC; Promocell), were washed with Hank's Balanced Salt Solution (HBSS, Invitrogen) and serum-starved overnight in endothelial basal media (EBM-2, Cambrex) with 0.1% fatty-acid-free BSA (Sigma) and 0.5% fetal calf serum (FCS, Hyclone). The following day cells were lifted with Trypsin/EDTA solution (Promocell), mixed with an equal volume Trypsin Neutralization Solution (Promocell), and washed 3 times in migration media (EBM-2 with 0.1% fatty-acid-free BSA and 0.2% FCS). Cells were resuspended at a density of 1.5×10⁶ cells/ml and were allowed to recover for 1 h at 37°C (5% CO₂). 3.75 × 10⁴ cells were plated into each well of a 48-well Boyden chamber apparatus (NeuroProbe, Cabin John, MD), and the wells were overlayed with an 8 μm pore polycarbonate membrane (NeuroProbe) that had been previously coated with 50 μg/ml human fibronectin (Biomedical Technologies, Inc., Stoughton, MA). Experiments performed with membranes coated with acetylated 1% gelatin from porcine skin (Sigma, St. Louis, MO) gave similar results. The apparatus was assembled and stored inverted at 37°C (5% CO₂) for 2 h. The apparatus was then re-inverted and 52 μl of purified chemoattractants [murine netrin-1 (R&D Systems, Minneapolis, MN), chicken netrin-2 (R&D Systems), murine netrin-4 (R&D Systems), murine netrin-G1a (R&D Systems), human VEGF₁₆₅ (R&D Systems), or control/migration media (EBM-2 with 0.1% fatty-acid-free BSA and 0.2% FCS) were added to the upper chambers, and the migration was allowed to proceed for 2 h at 37°C (5% CO₂). The membranes were then removed, fixed in methanol, stained with a Hema 3 stain set (Fisher Scientific, Pittsburgh, PA), and placed (migrated-side down) onto 50 × 75 mm glass slides. Before 90% mounting medium (in xylenes) and coverslips were applied, the non-migrated cells were removed from the exposed (non-migrated) side of the membrane with a moistened swab. Cells present on the migrated side of the membrane were manually counted (three random 200× fields per well), and data points for each experiment represent the average number of migrated cells from six separate wells (three 200× fields counted per well).

Another method was employed in a separate laboratory to evaluate the effects of the netrins on mouse (MS1) endothelial cells (ATCC, Manassas, VA) using a modified Boyden chamber assay as described previously (6). Briefly, a 5 μm-polycarbonate filter (Poretics) was placed between upper and lower chamber. Cell suspensions (5×10⁴ cells/well) were placed in the upper chamber, and the lower chamber was filled with serum-free medium containing recombinant mouse netrin-1 and -4 at concentrations of 0, 10, 50, 100 and 200 ng/ml. Recombinant mouse VEGF (50 ng/ml) was used as a positive control. The chamber was incubated for 4 hours at 37°C and 5% CO₂. Migration activity was evaluated as the mean number of migrated cells in 3 high power fields (400×) per chamber.

Proliferation Assays

HMVEC (Cambrex), HCAEC (Cambrex), HUAEC (Promocell), and HUVEC (Promocell) were cultured according to manufacturer's recommendations and were used between passages 3−7. A total of 2.5 × 10³ cells were plated into each well of a 96-well plate in complete growth media [EGM-2 MV (Cambrex) for HMVEC and HCAEC, EGM-2 (Promocell) for HUAEC and HUVEC, and allowed to adhere for 4 hours. Complete growth media was then removed, and the cells were washed with HBSS and then serum-starved for 16 h in EBM-2 (Cambrex) containing 0.2% FCS. Factors to be assayed (netrin-1, netrin-2, netrin-4, VEGF₁₆₅, BSA) were then diluted in EBM-2 + 0.5% FCS and added to each well. Assayed factors were re-applied every 24 hours for three days. After 72 h, cells were fixed in Zamboni's Fixative (Newcomer Supply, Middleton, WI). Stained with Gill-1 hematoxylin, overlayed with 80% glycerol, and manually counted. An alternative, more rapid method of assessing final cell number was also utilized. In this method, the number of cells present after 72 hours was quantified according to manufacturer's protocols by using the Cell Counting Kit-8 (Dojindo Molecular Technologies, Gaithersburg, MD) and measuring absorbance at 450 nm with a Thermomax plate reader (Molecular Devices, Menlo Park, CA ). Assays counted manually or with the cell counting kit gave similar results.

An alternate method was employed in another laboratory to assess the effects of netrins on proliferation of mouse MS1 endothelial cells (ATCC). The number of viable mouse endothelial cells was determined using a validated non-radioactive cell proliferation assay kit (CellTiter 96; Promega, Madison, WI) as described previously (6). Briefly, cells were seeded in 96-well plates at 80% confluence (1×10⁴ cells/well) and cultured in 10% FBS medium with recombinant mouse VEGF (13ng/ml), netrin-1, or netrin-4 proteins (R&D Systems, 50ng/ml). After 48 hours in culture, cells were incubated with MTS reagent for 4 hours, and proliferation activity was determined by the measurement of absorbance at OD 490 nm.

Tube Formation Assays

Matrigel (200 μl, BD Biosciences) was applied to each well of a 48-well plate and incubated at 37°C for 30 minutes before adding 25 × 10³ endothelial cells per well in serum free media. After 45 minutes at 37°C (5% CO₂), serum free media was removed and replaced with media containing 0.5% FCS and indicated treatments. Tube formation was assessed beginning 4 hours later. Photographs of representative 50 × fields were taken, and endothelial tubes were quantified by counting length and branches using Image J software (NIH). Experiments were repeated three times independently, and statistical analyses were performed using Student's t test. A p value <0.05 was considered statistically significant.

Adhesion

96-well plates (non-tissue culture treated plate, Falcon) were coated with 0.1 % BSA or with 10 μg/ml netrin-1, netrin-2, netrin-4, netrin-G1, or fibronectin for 16 h at 4°C. Wells were washed with PBS and then blocked with PBS containing 5% BSA for 1 h at room temperature. Endothelial cells were harvested, washed three times with, and then resuspended in basal media (serum-free) containing 0.5% BSA. A total of 5 × 10⁴ cells (in 100μl) were added to each well and allowed to attach for 30 min at 37°C. After washing three times with PBS, cells were fixed in Zamboni's fixative, and stained with Hemacolor (EMD chemicals, Gibbstown, NJ). Five 100× fields were manually counted per well. Data points represent mean number of adherent wells for duplicate wells (five 100× fields counted per well).

Migration Following Treatment with Adenosine A2b Receptor Antagonists

HMVEC (Cambrex, passage3−7) were pre-incubated for 2h in endothelial basal media (EBM-2) containing 0.1%FCS and 1.2%DMSO or in endothelial basal media (EBM-2) containing 0.1%FCS, 1.2%DMSO, and the A2b receptor antagonists [1 μM DPSPX (1,3-dipropyl-8-sulfophenylxanthine; Sigma Chemical Co) or 600 μM enprofylline (3-propylxanthine; Sigma Chemical Co)]. Migration towards bovine serum albumin (control), netrin-1, netrin-4, and VEGF, was assessed utilizing a NeuroProbe Boyden Chamber apparatus as described above. DMSO or DMSO-containing A2b inhibitors were also added to the chemotactic factors used during the migration assay. This method was based on that used previously by Stein et al (7).

Migration Following Immunodepletion of Netrin Protein Preparations

Polyclonal goat antibodies (20 μg/ml) against mouse netrin-1 and netrin-4 (R&D Systems) or non-specific goat anti-mouse IgG antibodies (Jackson ImmunoResearch) were incubated with Protein G+ agarose beads (Santa Cruz Biotechnology) in endothelial basal media (EBM-2) containing 0.1% FBS for 14 hours at 4° C. The Protein G+ beads were then pelleted and washed 3 times at 4° C in EBM-2 containing 0.1%FBS. The beads were resuspended in 1 ml EBM-2 containing 0.1%FCS and 50 ng Netrin 1 or Netrin 4 and gently rocked at 4° C for 2 hours. The beads were then pelleted, and the remaining media was added to the appropriate well of a NeuroProbe Boyden Chamber apparatus, and migration was carried out as described above.

Expression of Netrin Receptors in Endothelial Cells using Real Time RT-PCR

RNA was isolated from multiple primary endothelial cells [human aortic endothelial cells (HAEC, Cambrex), human microvessel endothelial cells (HMVEC, Cambrex), human umbilical vein endothelial cells (HUVEC, Cambrex), and human umbilical artery endothelial cells (HUAEC, Cambrex)] using TRIZOL (Gibco BRL, Gaithersburg, Maryland). Human total brain RNA (BD Biosciences) was used as a positive control. Total RNA was used as template to make random primed cDNA (RetroScript kit, Ambion). Assays for DCC, neogenin, Unc5a, Unc5b, Unc5c, Unc5d, Flk1, PECAM-1 and Robo4 expression MIGHT BE NICE TO INCLUDE SEQUENCES (Assays-on-Demand, Applied Biosystems) were performed according to manufacturer's instructions on an Applied Biosystems 7900HT thermal cycler (University of Utah Genomics Core Facility). Transcripts were normalized in relation to GAPDH expression (human GAPDH, Applied Biosystems). All reactions were performed in triplicate.

Intracellular cAMP Measurement

cDNAs encoding an irrevelant G protein-coupled receptor [arginine vasopressin receptor-2 (mock/control)] and the A2b receptor were cloned into a modified pcDNA3.1 expression construct. Lipofectamine 2000 reagent (Invitrogen) was used to transfect 293T cells with these expression constructs. The following day 100,000 cells were plated into each well of a 96-well plate. At 48 hours after transfection, the cells were treated for 15 minutes with 0−1,000 μM 5′-(N- ethylcarboxamido) adenosine (NECA) or with 0−25 nM (0−2,000 ng/ml) netrin-1 or netrin-4, and cAMP levels were then determined. The experiment was repeated multiple times with lengths of agonist treatment ranging from 5 minutes to 150 minutes, with 25,000−150,000 cells assayed per well, and with and without 3-isobutyl-1-methylxanthine (IBMX). Two different cAMP measurement systems were utilized (Mediomics and Assay Designs); similar results were obtained with each assessment.

Immunoprecipitation of DCC and Unc Receptors with Netrins

Immunoprecipitation was carried out using purified netrin-1 or netrin-4 containing a carboxy-terminal 10 histidine tag (R&D Systems) and purified proteins containing the Fc region of IgG₁ fused to the extracellular region of the deleted in colon carcinoma receptors (DCC and neogenin) or the Unc5 receptors (Unc5a, Unc5b, Unc5c, Unc5d; R&D Systems). 25 μl (50% slurry) Protein A/G PLUS agarose beads (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) were incubated with 2 μg of receptor-Fc fusion proteins and 1μg of His-tagged netrins at 4 °C for 90 min. The protein A/G beads were collected by centrifugation, washed three times with PBS containing 0.1% NP-40, resolved on a 10% acrylamide gel by SDS-PAGE, and analyzed by immunoblotting with an anti-His antibody (Clone BMG-His-1, Roche).

Immunoprecipitation of A2b Receptor with Netrins

A cDNA encoding the A2b receptor was cloned into pcDNA3.1 and into pcDNA3.1-HA. These constructs allowed expression of the A2b receptor with and without a carboxy terminal tag in mammalian cells. These A2b expression constructs were transiently transfected into HEK 293T cells (ATCC) using Lipofectamine 2000 (Invitrogen). Cell lysates were prepared with Lysis Buffer 1 [50 mM HEPES, 0.5 % NP-40, 250 mM NaCl, 2 mM EDTA, 10 % glycerol and Complete Protease Inhibitor (Roche Applied Science, Indianapolis, IN)] (8). The cell lysates were centrifuged in a microcentrifuge at 14,000 rpm for 15 min at 4°C. Supernatants were incubated with 1 μg netrin-1-His or netrin-4-His and then immunoprecipitated with an anti-HA affinity matrix (Roche). Western blot analysis using anti-His antibody was performed. Multiple conditions of immunoprecipitation to detect interaction of netrins with A2b produced similar results.

Cell lysates were also prepared in two additional buffers as follows: Lysis Buffer 2 [50 mM Tris-Cl (pH 7.5), 150 mM NaCl, 0.5 % NP-40, 1% Triton X-100, 5 mM EDTA, Complete Protease Inhibitor (Roche Applied Science, Indianapolis, IN)] and Lysis Buffer 3 [50 mM HEPES (pH 7.5), 125 mM NaCl, 5 mM EDTA, 0.1 % NP-40, and Complete Protease Inhibitor (Roche Applied Science, Indianapolis, IN)] (9) followed by sonication. Experiments utilizing Lysis Buffers 2 and 3 yielded identical results to those performed with Lysis Buffer 1.

Cell-Surface Binding of Netrins

COS-7 cells were transfected with a pcDNA3.1 expression construct alone (empty vector) or with pcDNA3.1 constructs containing the Unc5b or A2b receptors using Lipofectamine 2000 (Invitrogen) and were seeded into 12-well culture dishes (Corning Costar). A netrin-1-AP expression construct containing alkaline phosphatase fused to the carboxy terminus of human netrin-1 was kindly provided by Robert P. Kruger and Kun-Liang Guan of the University of Michigan. We replaced netrin-1 from this construct with a mouse netrin-4 cDNA to produce a plasmid encoding alkaline phosphatase fused to the carboxy terminus of netrin-4. Netrin-AP binding was carried out as described (10). Pretreatment (10 times molar excess) with unlabelled netrin before incubation with netrin-AP was performed to demonstrate specificity of interaction. Representative fields were photographed with a Zeiss Axiovert 200 microscope.

Fish Stocks and Embryo Raising

Adult fish were bred according to standard conditions. Heterozygous Tg(fli1:egfp)^y1 transgenic carriers [a gift from Joseph Yost; (11)] were incrossed, and embryos were raised at 28.5°C in E2/GN embryo medium with 0.003% phenylthiourea to inhibit pigment formation and staged according to (12). For in situ hybridization, embryos were fixed in 4% PFA (in PBS) for 1−2 hr at room temperature (RT) or overnight (O/N) at 4°C, washed briefly in PBS, dehydrated, and stored in 100% methanol at −20°C until use.

In Situ Hybridization

Standard in situ hybridization procedures were used. Antisense RNA probe was prepared from a plasmid containing a partial netrin1a clone [a gift from Uwe Strähle; (13)], using T7 polymerase and Dig RNA labeling kit (Boehringer Mannheim). Embryos were rehydrated to PBS, washed briefly with PBST (0.1% Tween-20 in PBS), digested with 0.1% collagenase (Sigma) for 15 min, washed with PBST, postfixed for 10 min in 4% PFA, treated with 0.25% acetic anhydride in 0.1 M triethanolamine (pH 7.0) for 30 min at RT, hybridized with RNA probe O/N at 65°C, washed with SSCT (SSC in 0.1% Tween-20) at 65°C, treated with 10 μg/ml RNase A at 65°C for 30 minutes to digest unbound probe, and incubated for at least 2 hr with 2% Blocking Reagent (Boehringer Mannheim). Embryos were then incubated O/N with alkaline phosphatase-conjugated anti-DIG Fab fragment (1:5000, Boehringer Mannheim), washed with PBST and developed by incubating in AP substrate (Boehringer Mannheim).

Morpholino Oligonucleotide (MO) Sequence and Reverse Transcription

Control MO was a standard control from Gene Tools, sequence 5′-CCTCTTACCTCAGTTACAATTTATA-3′. The netrin1a splice blocking MO, sequence 5′-ATGATGGACTTACCGACACATTCGT-3′, was synthesized by Gene Tools and was a gift from Iain Shepherd. It was designed to bind to the last 11 bp of exon 1 and first 14 bp of intron 1. For RT-PCR analyses, morpholino-injected and wt embryos were placed in Trizol, triturated through a 25 gauge needle and stored at −20°C. Total RNA was extracted using the Qiagen RNeasy kit, treated with DNase, and cDNA prepared using SuperScript™ First-Strand Synthesis (Invitrogen) and random hexamer primers. Primers F1 (5′-CTTTCGGAGACGAAAACGAG-3′) and R1 (5′-GTAGGCGCTTTCCAGAGATG-3′) were used to amplify netrin1a pre-mRNA, while primers F1 and R2 (5′-CTTTGCAGTAGTGGCAGTGG-3′) were used to amplify normal mRNA; primers beta-actin FP (5′-CCCAAGGCCAACAGGGAAAA-3′) and beta-actin RP (5′-GGTGCCCATCTCCTGCTCAA-3′) were used to amplify beta-actin mRNA. PCR conditions: 2 min 94°C; followed by 31 cycles (netrin1a) or 20 cycles (beta actin) of 94°C, 30 s/59°C, 30 s/72°C, 60 s; followed by 5 min 72°C.

MO Injections

Lyophilized MO was diluted and stored at 4°C; concentrations were periodically checked by spectrophotometry (A265 in 0.1N HCl, as suggested by Gene Tools). Working solutions were diluted in water and phenol red (0.5% final concentration), and 1 nl was pressure-injected into the cell of 1-cell embryos or high yolk of 2-cell embryos using a Picospritzer II or an ASI MPPI-2 injector. The injected volume was calculated using a calibrated eyepiece micrometer. netrin1a SBMO was injected at 5.6 ng/embryo, control MO at 6.0 ng/embryo.

Microscopy and Image Analysis

Live GFP-positive embryos were embedded laterally in 1.5% low-melt agarose in E2/GN on a Petriperm dish (Greiner Bio-One). An Olympus Fluoview confocal microscope was used to acquire confocal z-stacks at the level of somites 7−12, focusing through both the near and far sides of the trunk. Presence or absence of PAV in each hemisegment was scored after examining all slices of each z-stack. Fluoview software was used to create confocal z projections, and Adobe Photoshop and Adobe Illustrator were used to compose figures.

Zebrafish Antibody Staining

Stored 32hpf embryos were rehydrated in PBS, incubated in 1% H₂O₂ for 1 hr to block endogenous peroxidase, permeabilized with 0.1% collagenase for 15 min, washed in PBST, blocked with NCST (10% heat-inactivated newborn calf serum, 0.1% Tween-20, 1% DMSO, in PBS) for 1.5 hr, and incubated with rabbit anti-GFP pAb (1:400, Molecular Probes) and either 4D9 (1:4, Zebrafish International Resource Center) or F59 (1:5, Developmental Studies Hybridoma Bank) O/N at 4°C. Embryos were then extensively washed with PBST, incubated with anti rabbit Alexa 488 (1:200, Molecular Probes) and either anti mouse HRP (1:100, Molecular Probes) or anti mouse Cy3 (1:100, Jackson Immunoresearch) in NCST respectively at RT for 3.5 hr. Embryos incubated with the anti mouse HRP secondary were then treated with 0.5% DMSO in PBS, and developed in Alexa-568 tyramide (1:100, Molecular Probes) in 0.0015% H₂O₂ in PBS.

Murine Ischemic Hindlimb Model

All procedures were performed in accordance with the Institutional Animal Care and Use Committee of Caritas St. Elizabeth's Medical Center. Ischemia was induced in 8- to 12-week-old male FVB/NJ mice as described previously (1). VEGF, netrin-1, netrin-4 and empty vector plasmid (50 μg/100μl saline in each group) were locally injected to the right gastrocnemius muscle immediately, 7, 14 and 21 days after surgery.

Physiological Assessment of Treated Animals

Laser Doppler perfusion imaging (Moor Instrument, Wilmington, DE) was used to record serial blood flow (LDBF) measurements over the course of 4 weeks postoperatively, as previously described (2). Before initiating laser scanning, mice were placed for 15 min on a heating plate kept at 37°C to minimize body temperature changes. After scanning, the stored images were subjected to computer-assisted quantification of blood flow, and the average flow of the ischemic and nonischemic feet was calculated. The LDBF value was expressed as the ratio of right (ischemic) to left (nonischemic) limb blood flow to avoid data variations caused by ambient light and room temperature.

Neurophysiological measurements

Sciatic nerve conduction velocity was measured using standard orthodromic surface recording techniques and a Teca TD-10 (Oxford Instruments, Chicago, IL) portable recording system in all mice at baseline (before treatment) and then at 1 week, 2 weeks, 3 weeks and 4 weeks after treatment as described previously (6). Briefly, motor nerve conduction velocity (MCV) was calculated by dividing the distance between stimulating electrodes by the average latency difference between the peaks of the compound muscle action potentials (CMAPs) evoked from two sites (sciatic notch and ankle). Sensory nerve conduction velocity (SCV) was calculated by dividing the distance between stimulating and recording electrodes by the latency of the signal from the stimulation artifact to the onset of the peak signal. For each nerve, maximal velocities were determined bilaterally. All measured data from both sides were averaged.

Supplementary Material

1

Supplemental Figure 1. Structure and purity of netrin proteins. (A) Amino acid comparison of netrins-1, -2, and -4. Netrins-1 and -2 are more closely related (overall 52% amino acid identity) than are netrins-1 and -4 (overall 31% amino acid identity). All three netrins are secreted, diffusible proteins. (B) Silver stain analysis of netrin protein preparations. To assess purity of the recombinant netrins used in our in vitro assays, 1 μg of mouse netrin-G1a, mouse netrin-1, chicken netrin-2, and mouse netrin-4 were analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis followed by silver staining. Molecular weight standards were electrophoresed in the well on the left. A single band of molecular weight 65−70 kDa for netrin-G1a and of 75−80 kDa for netrins-1, -2, and -4 was observed with no other significant contaminating proteins. (C) Immunodepletion of protein preparations using antibodies against netrin-1 or netrin-4 removes the promigratory activity. Polyclonal goat antibodies (20 μg/ml) against mouse netrin-1 (αN1 IgG) and netrin-4 (αN4 IgG) or non-specific goat anti-mouse IgG antibodies (NS IgG; 20 μg/ml) bound to Protein G+ agarose beads were incubated with purified, recombinant netrin-1 and netrin-4 preparations (50 ng/ml) for 2 hours at 4° C. Following removal of the beads, the protein preparations were loaded into Boyden chambers for migration assays using human microvessel endothelial cells (HMVEC). Immunodepletion using antibodies against netrin-1 and netrin-4 removed essentially all of the promigratory activity previously contained in these protein preparations. Nonimmunodepleted netrin-1 and netrin-4 (50 ng/ml) stimulated approximately threefold increased migration, as did an equimolar amount of vascular endothelial growth factor (VEGF; 13 ng/ml), compared to 0.1% BSA control (Ctl).

Supplemental Figure 2. Netrins promote endothelial cell migration, proliferation, and tube formation but not adhesion. (A) Netrin-1 (N1) and netrin-4 (N4) stimulate migration human microvessel endothelial cells (HMVEC), human umbilical vein endothelial cells (HUVEC), human umbilical artery endothelial cells (HUAEC), and human coronary artery endothelial cells (HCAEC). We obtained similar results for MS1 murine endothelial cells and for murine lung endothelial cells (data not shown). Migration towards 50 ng/ml netrin-1 and netrin-4, as well as towards an equimolar amount of VEGF (13 ng/ml), was compared to 0.1% BSA control (Ctl). (B) Proliferation of HMVEC after 72 hours in the presence of 10−200 ng/ml netrins-1 and -4 or 13 ng/ml VEGF. Equimolar amounts of the netrins and VEGF (50 ng/ml netrin or 13 ng/ml VEGF) produced similar increases in proliferation. Data points represent the mean ± SEM of 24 separate determinations (8 individual wells from each of three separate experiments). (C) Proliferation of HUVEC after 72 hours in the presence of 10−200 ng/ml netrins-1 and -4 or 13 ng/ml VEGF. (D) Netrin-1 and -4 (200 ng/ml) as well as VEGF (13 ng/ml) stimulate tube formation in HUVEC as compared to control treatment (0.1% BSA). Quantification of line length in formed tubes demonstrates that the stimulatory effect of netrins-1 and -4 on tube formation is similar in magnitude to that of VEGF. Data points represent the mean value ± SEM. (E) Endothelial cells do not adhere to netrins. Quantification of HUVEC adhesion to 96-well plates treated overnight at 4° C with 0.1% BSA (control) or 10 μg/ml netrin-1, netrin-2, netrin-4, or fibronectin. Each data point represents the mean ± SEM of combined data from two separate experiments where cells in 6 low power (10×) fields from duplicate wells were manually counted. (F) HUVEC adhesion to 96-well plates treated overnight at 4° C with 0.1% BSA (control) or 10 μg/ml netrin-1, netrin-2, netrin-4, or fibronectin. Each panel represents a 10× light microscope field, and the scale bar in the left panel represents 100 μm.

Supplemental Figure 3. Netrin-induced endothelial cell migration does not require activation of the adenosine A2b receptor. (A) COS-7 cells overexpressing the Unc5b receptor or A2b were incubated with 2 nM netrin-1-alkaline phosphatase (netrin-1AP) or netrin-4-alkaline phosphatase (netrin-4AP) fusion proteins. The cells on the left side of each panel were first pretreated with a 10× molar excess of unlabelled netrins to demonstrate specificity of interaction. The cells were then washed and exposed to precipitating substrate. Representative light microscopy fields are shown. Control transfections with empty vector (pcDNA3.1) demonstrated no baseline interaction of netrins with COS-7 cells, whereas netrin-1AP binds to Unc5b-overexpressing COS-7 cells, and pretreatment with unlabeled netrin-1 blocks this interaction. Similar binding of Netrin-4AP to Unc5b-expressing cells was not observed. Neither netrin-1AP nor netrin-4AP binds to A2b overexpressing cells. Scale bar, 50 μm. (B) Human microvessel endothelial cells (HMVEC) were incubated in basal media containing 1.2% DMSO (control, netrin alone, and vascular endothelial growth factor) or in basal media containing 1 μM DPSPX (1,3-dipropyl-8-sulfophenylxanthine) or 600 μM enprofylline (3-propylxanthine) and 1.2% DMSO. DPSPX and enprofylline are antagonists of the adenosine A2b receptor. DPSPX and enprofylline at concentrations well in excess of their K_i values have no significant effect on HMVEC migration induced by netrin-1 and netrin-4, a result similar to that of Stein et al for axon outgrowth.

Supplemental Figure 4. Trunk morphology is preserved in netrin1a morphants. Antibody staining shows that slow muscle cells (F59) and engrailed-positive cells (4D9), which include both muscle pioneers and a subset of fast muscle cells, are not disrupted in netrin1a morphants (A, C respectively) when compared to uninjected controls (B, D respectively). Confocal z-projections of 32hpf fli:egfp embryos; anti-GFP staining in green. 4D9 and F59 in magenta. (E) Netrin1a splice-blocking morpholino greatly reduces normal mRNA. Positions of netrin1a splice-blocking MO and RT-PCR primers relative to the exon/intron genomic structure of netrin1a (not to scale). (F) Total RNA was isolated from uninjected embryos and netrin1a morphants (5.6 ng MO), DNase-treated, and RT-PCR performed using primer pairs that amplify unspliced netrin1a pre-mRNA (F1+R1); properly spliced netrin1a mRNA (F1+R2); and beta-actin mRNA as a loading control (beta-actin). Normal mRNA seems to be completely abolished by injection of the netrin1a MO, while unspliced pre-mRNA levels are slightly increased. Reactions lacking reverse transcriptase (-RT) confirm that amplified bands reflect RNA and not genomic DNA.

Supplemental Figure 5. Local delivery of vascular endothelial growth factor (VEGF), netrin-1, or netrin-4 expression constructs promotes revascularization, reduces scar formation, and increases the number of vessels with smooth muscle α-actin in mice with surgically-induced hindlimb ischemia. (A) Ratio of blood flow in the ischemic compared to the normal limb as assessed by laser Doppler perfusion imaging. Hindlimb ischemia was surgically induced in FVB/NJ mice by ligation and excision of the proximal femoral artery as described previously (1). VEGF, netrin-1, netrin-4 and empty vector DNA expression plasmids (50 μg DNA in 100 μl saline in each group) were locally injected into the right gastrocnemius muscle immediately, 7, 14 and 21 days after surgery. Laser Doppler perfusion imaging was used to record serial blood flow measurements over the course of 4 weeks postoperatively, as previously described (2). Ischemic hindlimbs injected with VEGF, netrin-1, or netrin-4 expression constructs demonstrated improved perfusion compared to those injected with empty vector (EV) beginning at day 7 following surgical induction of ischemia and plasmid injection. This improvement continued through day 28. (B) FITC-isolectin B4 staining of sections of gastrocnemius muscle 28 days after surgical induction of ischemia and initiation of plasmid injection is shown in the left panel. On the right is seen quantification of capillary density per high power field as determined by isolectin B4 staining in ischemic muscle 28 days after induction of ischemia. (C) The left panel demonstrates Masson's Trichrome staining of muscle cross section from animals on day 28 after surgical induction of ischemia and local injection of empty vector, VEGF, netrin-1, or netrin-4 DNA expression constructs. Areas of scar stain blue. Scar quantification is shown on the right, calculated as percentage area of the muscle bundle. Animals injected with VEGF, netrin-1, or netrin-4, had significantly less scar compared to animals injected with empty vector. (D) Staining of muscle cross sections with FITC-smooth muscle α-actin on day 28 after surgical induction of ischemia is seen on the left. VEGF, netrin-1, netrin-4 and empty vector DNA expression plasmids (50 μg DNA in 100 μl saline in each group) were locally injected into the right gastrocnemius muscle immediately, 7, 14 and 21 days after surgery. On the right is displayed the quantification of vessels positive for staining with smooth muscle α-actin. Animals injected with netrin-1 plasmid had a greater density of smooth muscle α-actin-positive vessels.

Supplemental Figure 6. Local delivery of vascular endothelial growth factor (VEGF), netrin-1, or netrin-4 expression constructs into db/db animals increases capillary density of the sciatic vasa nervorum. (A) BS1 lectin (endothelial) staining of longitudinal and cross sections of sciatic nerves and BrdU staining of adjacent cross sections of sciatic nerves from db/db mice at 28 days following initiation of injection of empty vector, netrin-1, netrin-4, or VEGF. White arrowheads indicate representative structures that stained with both B1 lectin and BrdU, indicating endothelial structures with active cell division. Blue arrowheads indicate representative structures that stained with BrdU and not BS1 lectin. S100 staining of adjacent sections suggest that these structures are Schwann cells. Scale bar, 200 μm (longitudinal sections), 100 μm (cross sections). (B) Quantification of capillary density in the vasa nervorum of the sciatic nerves of db/db mice as assessed by BS1 lectin staining demonstrates increased capillary density in animals injected with the netrin-1, netrin-4, or VEGF constructs as compared to those injected with empty vector. (C) The number of cells per sciatic nerve cross section that stain for both BS1 lectin and BrdU is significantly greater (P<0.05) in db/db animals injected with netrin-1, netrin-4, or VEGF compared to animals injected with empty vector. (D) The number of cells per sciatic nerve cross section that stain for both BrdU and S100 (Schwann cell) is significantly greater in db/db animals injected with netrin-1 than in animals injected with VEGF, netrin-4, or empty vector.

Supplemental Figure 7. Generation of Unc5b mutant mice. (A) Targeting strategy. Homologous recombination generates the Unc5b^Neo allele. FRT sites flank the neomycin cassette. Cre (Hprt-Cre) recombinase deletes exons 4−13, the region between lox P sites, resulting in the Unc5b deleted allele (Unc5b-). The transmembrane domain encoded by exon 9 is marked by a black bar. (B and C) Southern blot and PCR analysis of offspring from chimeras confirm germline transmission. The Southern probe and detected DNA fragments are indicated in (A). The PCR fragments amplified from wild-type and targeted alleles are approximately 100 bp and 150 bp, respectively. (D) PCR genotyping of offspring from Unc5b +/− and Unc5b +/− mating. For the Unc5b deleted allele (Unc5b −), an approximately 300 bp PCR fragment is amplified by the D2F and 2R primer set. For the Unc5b + allele, a 150 bp PCR product is amplified by the 2F and 2R primer set. (E) RT-PCR analysis confirms the lack of Unc5b transcripts in Unc5b −/− embryos but presence in Unc5b +/+ or +/− embryos. Actin was used as a loading control. No product is amplified without reverse-transcriptase (- RT).

Supplemental Figure 8. Vascular patterning of Unc5b +/+ and Unc5 −/− embryos are similar. (A) Vascular patterning of Unc5b +/+ and Unc5 −/− embryos are indistinguishable by PECAM staining at E10.5-E11.5. Unc5b −/− embryos do not show any significant vascular patterning defects in the head, trunk region (side and dorsal views) and yolk sac compared to littermate controls. (B) Vascular patterning is not significantly altered before or after mice die at E12.0-E12.5. Unc5b mice were crossed to Robo4-AP mice to label vascular endothelial cells using alkaline phosphatase (AP) activity. At E12.0, Unc5b −/−;Robo4-AP/+ mice are stained for AP activity and compared to the same staining of Unc5b +/+;Robo4-AP/+ mice. Vasculature of yolk sac and tails of embryos in E12.0 are similar in Unc5b +/+ and Unc5b −/− mice. Mice die in utero between E12.0 and E12.5. AP staining of yolk sac and tails from E12.5 Unc5b −/−;Robo4-AP/+ and Unc5b +/+;Robo4-AP/+ embryos do not show significant vascular patterning differences, although Unc5b −/− embryos are necrotic. (C) IsolectinB4 staining at E10.5-E11.5. IsolectinB4 staining of hindbrain vessels at E10.5 and E11.5 shows similar patterning in mutant and wildtype mice. Scale bars, 0.2 mm.