Supplementary Methods

Transcription

1 Supplementary Methods Antibodies Rabbit polyclonal VEGFR-1, VEGFR-2, Angiopoietin-1, Angiopoietin- 2 and GAPDH specific antibodies were from Santa Cruz Biotechnology. Rabbit polyclonal Akt1, Akt2, ps473 Akt, GSK-3, ps9 GSK-3β and ps1177 enos antibodies were purchased from Cell Signaling. enos antibody was obtained from Transduction Laboratories. Akt 3 antibody was from Upstate Biotechnology. TSP-1 antibodies for immunoblotting and immunohistochemistry were purchased from Calbiochem and Abcam, respectively and rabbit TSP-2 antibody was raised and purified in the lab (P.Bornstein). Laminin, β actin and FITC conjugated SMA specific antibodies were purchased from Sigma. CD31 antibody was purchased from Research Diagnostics Inc. Integrin α v and β 3 antibodies were from Chemicon International. Anti-CD105 was purchased from BD Biosciences. Antibody against SMA, vwf and mouse monoclonal Ki67 were purchased from DAKO Cytomation. Fibrin antibody was purchased from Accurate Chemical and Scientific Corporation. Fluorescence conjugated rat, rabbit, goat and mouse specific antibodies (Alexa 594 and Alexa 488) were obtained from Molecular probes. Electron microscopy Ultrastructure of blood vessels formed in Akt1 / and in WT mice was assessed using electron microscopy approach as previously described 46. Briefly, tumor tissues from two matched pairs of WT and Akt1 / mice were fixed at 4 o C overnight in 0.1 M sodium cacodylate buffer (ph 7.4) containing 2.5% glutaraldehyde and 4% formaldehyde. After washing three times in the same buffer, tissue blocks were post-fixed with 1% aqueous osmium

2 tetroxide for 1 h at 4 0 C. Tissues were washed in sodium cacodylate buffer (ph 7.4), followed by maleate buffer (ph 5.1), and then dehydrated with ascending grades of ethanol and propylene oxide. Samples were embedded in LX-112 medium and polymerized at 70 C for 48 h. Ultrathin sections of 80 nm were stained with 50% saturated uranyl acetate and 0.2% lead citrate, and then examined with a transmission electron microscope (Philips CM120). Vascular permeability assay in vitro EC were seeded on collagen-coated transwells with 0.4 m pore size and allowed to form a monolayer. EC were infected with lentivirus (10 9 particles/well) encoding mouse specific TSP-1 sirna or with an adeno-virus (10 9 particles) with myrakt1. Alternatively, TSP-1 at 1 g/ml was included in EC cultures for at least 24 hours. Permeability was stimulated with VEGF (20 ng/ml) for 30 min followed by addition of HRP (1.5 g/well) into the top chamber for additional 30 min. The amount of HRP in the bottom chambers was determined using 3,3,5,5 -tetramethylbenzidine 44. Vascular permeability assay in vivo Permeability was stimulated using mustard oil (Sigma). After 30 minutes, mice were euthanized and ears were removed and weighed. Evans blue dye was extracted, measured spectrophotometrically at 610 nm and the leakage was expressed as nanograms of dye per milligram of tissue 15. To assess the effect of myrakt1 on vascular permeability in Akt1 / mice, 1 X 10 9 pfu Ad-myrAkt1 in 50 l PBS was injected twice over 4 weeks period into the left ears of 4 weeks old mice Akt1 / mice. The same amount of Ad-GFP as control was injected into the right ears. Vascular permeability in mouse ears was assessed 5 weeks later using a modified Miles assay (no effect

3 was observed after shorter treatment). Vascular leakage of Evans blue was stimulated by mustard oil. MyrAkt1 treatment did not affect the basal level of vascular permeability; however, in response to a pro-inflammatory agent, mustard oil, vascular leakage was significantly lower in ears treated with myrakt1 adenovirus compared to control GFP adenovirus (n = 6 mice per group). Vascular leakage assay in vivo This approach is based on the specific binding of ricin to the sites of luminal surface of blood vessels where endothelial lining was lost, and the basal membrane became exposed. A characterization of microvessel morphology and leakage sites in ricin lectin-stained whole mounts of ear skin was performed using a previously described protocol 15. In the absence of stimulation, only weak staining was observed in both types of mice. However, when vascular permeability was induced with mustard oil, ricin bound strongly to the sites of the exposed basement membrane in microvasculature. Re-expression of TSP-1 and TSP-2 Primary syngeneic cells were isolated from digested lungs of Akt1 +/ mice (littermates of Akt1 / mice used in this experiment). This cell population was comprised of ~70% of endothelial cells and ~30% of fibroblasts/smooth muscle cells based on FACS analysis. Cells were cultured for 1 week and transfected with murine TSP-1 and TSP-2 using nucleofection according manufacturer's protocol from Amaxa (Germany) for endothelial cells. This procedure resulted in an increase in levels of both TSP-1 and TSP-2 in cultured cells prior to their injection into mice (shown in Fig.6c). Cells over-expressing TSP-1 and TSP-2 (1x10 5 cells per mouse) were subcutaneously injected into chimeric Akt1 / mice transplanted with wild type

4 bone marrow (see Methods for bone marrow transplantation for details). Two weeks later, 8x10 5 B16F10 melanoma cells combined with 1x10 5 TSPoverexpressing cells were inoculated into the same areas, and tumors were collected after 1 week. As a result of cell based gene transfer, TSP-1 and TSP-2 levels in lysates of collected tumors were increased as evidenced by Western blot presented in Fig.6c. Supplementary Figures Supplementary Fig. 1 a, Western blot analysis and band densitometry of different isoforms of Akt (Akt1, Akt2 and Akt3) in skin (i) and tumor (ii) of WT and Akt1 / mice b, Western blot analysis of Akt isoforms expression in EC from lung (i), aorta (ii) and matrigel implants (iii) from WT and Akt1 / mice c, Migration of WT (black bars) and Akt1 / (white bars) EC toward fibrinogen was stimulated by 25 and 100 ng/ml of VEGF as indicated. d, Analysis of soluble ligand binding by WT (filled circles) and Akt1 / (open circles) aortic EC was performed using FITC-fibrinogen. Isolated EC were incubated in suspension with increasing concentrations of FITC-fibrinogen in the presence of 0.5 mm of MnCl 2 and Mean of Fluorescence Intensity (MFI) was determined by FACS. Note an impaired binding of extracellular matrix to Akt1 / EC. Supplementary Fig. 2 a, Ultrastructure of blood vessels developed in Akt1 / and in WT mice using electron microscopy (magnification: 3800 X) as described in Supplementary Methods online. Abbreviations: EC - Endothelial cells, RBC - red blood cells. b, Quantification of laminin levels in lysates of tumors implanted

5 in WT vs Akt1 / mice as compared to CD31 levels. Results of three independent experiments are shown. c, Comparison of laminin and CD31 levels in tumors from WT and Akt1 / mice was performed by Western blot. Note that increased levels of CD31 in Akt1 / mice (due to the enhanced vascularization) were associated with a decrease in laminin. Protein loading was normalized using two markers, GAPDH and β-actin. d, Representative photographs of vwfstained (brown color) skin sections showing increased number of blood vessels in Akt1 / compared to WT mice in response to Ad-VEGF-A. Scale bars: 50 m. e, Co-staining for SMA (brown) and vwf (red) in skin of WT and Akt1 / mice upon stimulation with Ad-VEGF-A. Arrows indicate positively stained blood vessels. Note increased number of SMA-positive blood vessels in WT compared to Akt1 / mice. Scale bars: 20 m. f, Angiopoetin levels in skin and in B16F10 tumors implanted in Akt1 / and WT mice were measured by Western blots. Note increased expression of angiopoietin-1 and angiopoietin-2 in skin lysates of Akt1 / compared to WT mice (top panel). No significant difference in angiopoetins levels was observed in lysates of tumors developed in Akt1 / compared to WT mice (bottom panel). Levels of control protein, β-actin are also shown. Supplementary Fig. 3 Absence of Akt1 results in impaired collagen matrix organization. a, Collagen in skin was visualized using Masson s trichrome stain. Collagen fiber density (blue color) in skins of Akt1 / mice is markedly lower than that of WT mice and the dermis is slightly expanded in order to compensate for the defect. Scale bars: 20 m. b, Comparison of collagen content in skins of WT

6 and Akt1 / mice. Total collagen was quantified in skin samples by measuring hydroxyproline in skin hydrolysates as described 47. The concentration of hydroxyproline in each tissue was converted to collagen concentrations by multiplying the values by 8. c, The matrix is known to physically restrain tumor growth by generating solid stress, which, in turn, affects cellular packing density. Tumors grown in WT but not in Akt1 / mice were densely packed and showed signs of solid stress as the individual cells were shaped differently. The number of nuclei per field was significantly higher in WT tumors (WT: 477 ± 39 vs. Akt1 / : 292 ± 21, p = 1.15E-05, n = 6). Micrograph shows loosely arranged tumor cells in Akt1 / mice (right panel) compared to tightly packed cells in tumors grown in WT mice (left panel). Scale bars: 50 m. Supplementary Fig. 4 Tumor vasculature and TSP-1 expression are inversely related. a, Immunohistochemical staining of 6 micron paraffin sections of tumors implanted into WT mice for TSP-1 and CD31 is shown in left and right panels, respectively. Note that the majority of blood vessels are present on the periphery of tumor, whereas TSP-1 is expressed in the center. Scale bar: 100 m. b, A costaining for TSP-1 (visualized using second antibodies labeled with FITC) and CD31 (Texas Red) in 6 micron frozen sections of tumors grown in WT mice. Left panel shows peripheral part of the tumor and right panel shows central area. Scale bar: 20 m. c, Comparison of surface expression of TSP-1 and TSP-2 on EC isolated from lungs and aortas of WT and Akt1 / mice. Supplementary Fig. 5 Basal level of enos phosphorylation is reduced in Akt1 / endothelial cells (EC). a, Western blots of WT and Akt1 / EC lysates show

7 basal levels of enos phosphorylated at Ser1177 (this site is known to be phosphorylated by Akt). No changes in total levels of enos were observed. Note substantial increase in phosho enos (penos) upon VEGF treatment in WT as well as in Akt1 / EC. Also note similar levels of phospho enos in skin of WT and Akt1 / mice (left panel) suggesting a possibility of a compensatory mechanism. b, Systemic decrease in enos activity often results in hypertension. However, no significant difference in the blood pressure was observed between WT and Akt1 / mice. Blood pressure measurement was performed using RTBP1001 rat-tail blood pressure system for rats and mice (Harvard Apparatus) according to the manufacturer s protocol. Supplementary Fig. 6 In order to examine the role of enos in the regulation of neovascularization, we assessed tumor-induced angiogenesis in enos / mice. Wild type and enos / mice on C57BL/6 background were obtained from Jackson Laboratories. B16F10 melanoma cells were injected subcutaneously into flanks of mice and 7d later tumors were excised. Frozen tissue sections were fixed in methanol at -20 o C for 2 min and incubated with 0.1% Triton X-100 in PBS for 20 min. Sections were incubated with anti-cd31 antibodies overnight at 4 o C, followed by Texas Red-conjugated secondary antibodies and rat anti-mouse αsma (directly FITC-conjugated). The slides were mounted in media containing DAPI. Slides were immediately analyzed using confocal microscopy. a, Fluorescent micrographs of frozen sections of tumors from WT and enos / mice stained with CD31 (red) and SMA (green) are shown. Note that proportion of blood vessels stained for both CD31 and SMA is substantially decreased in

8 enos / mice. Thus, the tumor blood vessels in enos / mice are significantly less mature than in WT, with an apparent lack of pericyte recruitment, therefore, resembling the phenotype of Akt1 / mice. Scale bar: 50 m. A micrograph of SMA positive blood vessel at higher magnification is shown in inset. Scale bar: 10 m. b, Figure showing an increase in number of blood vessels (identified by laminin staining) in tumor sections from enos / mice (right panel) compared to WT (left panel). Scale bar: 50 m. c, Quantification of the data shown in figure 6b demonstrates that vascular area of tumors developed in enos / mice was similar to that of WT mice. d, Quantitative analysis of SMA positive blood vessels in enos / tumor compared to WT suggesting a decrease in blood vessel maturation in enos / tumors compared to WT. However, in contrast to Akt1 / mice, no changes in the expression of TSPs were observed in enos / mice (not shown).