Supplemental Materials and methods
Mice
C57/B6, SCID and MMTV-PyMT mice were purchased from the Jackson Laboratory. For in situ imaging of metastasis, mice were anesthetized with 2% isoflurane/O2 and injected intraperitoneally with 15mg/ml D-luciferin (firefly, potassium salt in PBS; Caliper LifeSciences) at a concentration of 150mg luciferin/kg body weight. Animals were analyzed at various time points post-injection using a Xenogen apparatus (IVIS Imaging Systems, Caliper LifeSciences).
Cell culture, transfection and siRNA analyses
For transfections, cells were electroporated (0.25 KV, 500 µF; BioRad) and selected in the presence of G418 (1mg/ml; Invitrogen), hygromycin (0.3mg/ml; Invitrogen) or puromycin (1µg/ml; Sigma), depending on the transected plasmid. All cells were maintained at 37°C in 5% CO2.
siRNA oligonucleotides were from Ambion: siRNA Cav1, Cat# 16708A ID 160016 (5'-ccaucaauuuggagacuau-3'); RhoC, Cat#16708A ID 60380 (5'-gguauuugagauggccacu-3'); control siRNAs, siScramble (Dharmacon). Knockdown efficiencies were validated in pilot transfections (data not shown). For siRNA transfections, 6x105 cells were cultured over night in 6-well plates and transfected with 200 pmoles siRNA using Lipofectamine2000 (Invitrogen) according to the manufacturer’s protocol. Transfected cells were analyzed by immunoblotting at 72 hrs post-transfection.
Lentiviral particles for the shRNA against mouse RhoC, Cav1 and RhoA (pLKO.1 vector backbone, Sigma or OpenBiosystems) were produced in Lipofectamine2000 (Invitrogen) transduced 293FT cells according to the manufacturer’s protocol. The transduced cells were selected with 1ug/ml puromycin.
Immunoblotting
Whole cell extracts were prepared in RIPA buffer (50 mM Tris [pH 7.4], 150 mM NaCl, 1 mM EDTA, 1% NP40, 0.25% Na-deoxycolate, 1 mM phenylmethylsulfonyl fluoride, and 2 μg/ml leupeptin and aprotinin). Equal amounts of protein (Bradford assay) were resolved by SDS-PAGE and immunoblotting.
Quantitative alu PCR
Mouse tumor cells in chicken embryonic lung and liver were measured using quantitative PCR to detect the mouse alu sequence as described (Zijlstra et al. 2002; Zijlstra et al. 2008). Briefly, 1:50 diluted mouse DNA extracts (Extract-N-Amp kit, Sigma) from chicken tissue was combined with SYBR green Extract-N-Amp ReadyMix (Sigma) and combined with 0.4 µM of each primer. Primers specific for mouse alu repeats (F: 5'-gggctggtgagatggctcagtgg-3'; R: 5'-cttcagacacaccagaagaggg-3') were used to quantify the numbers of mouse cells and chicken GAPDH primers (F: 5'-gaggaaaggtcgcctggtggatcg-3'; R: 5'-ggtgaggacaagcagtgaggaacg-3') were used to quantify the amount of chicken tissue in each sample. The cT values obtained from the mouse alu amplification were normalized to the chicken GAPDH as previously described (Zijlstra et al. 2002; Zijlstra et al. 2008).
Intravital imaging of cell migration, extravasation and tumor growth
Tumor cells in the chicken embryos were visualized and analyzed the following way: A 100 µm image stack was captured with an 8 µm step size every 15 minutes. Image drift and rotation was corrected using the Stack_Reg plugin (Biomedical Imaging Group http://bigwww.epfl.ch/) running in the open-source software ImageJ (NIH). To quantify the progress of extravasation of individual cells over time, a mosaic imaging and analysis method was employed. Vasculature was labeled with LCA as above, and a 100 µm image stack was captured with an 8 µm step size over an 18 mm circular region of interest (ROI) at time 0 (post-injection), 8 hrs and 18 hrs. At time 0, all of the cells in each ROI were counted (total cells). Individual cells were manually classified as intravascular or extravasated at each timepoint and the total of these two groups was subtracted from the total cells to calculate the number of missing or dead cells at each timepoint. Cells in the process of extravasating were classified as extravasated if more than 50% of the cellular volume was extravascular.
Supplemental Figure Legends
Figure S1 Sequence of the murine Rho-GTPase caveolin-binding domain (CBD) and functional analysis of CBDwt and CBDmut.
(a) IB of ph-p130Cas at Y165 in shRhoA F10 and control cells (shEV).
(b) Most murine Rho-GTPase proteins share a FXXXXFXXF CBD amino acid sequence at their N-termini. Shown is the alignment of the CBD amino acids (aa) for RhoA, RhoB and RhoC. *, identical aa; :, divergent aa. Red lettering indicates the aromatic aa in the CBD essential for Cav1 binding.
(c) Structure and sequence of the wt CBD (wild type) (CBDwt; 1st and 2nd rows), the mutated CBD (CBDmut; 2nd row), and the CBD overexpression vectors (3rd row). CBD expression was driven by the CMV promoter in the pcDNA3.1 plasmid.
(d) Analysis of CBDwt-GFP and CBDmut-GFP protein expression levels by Flow Cytometry (FACS). Overexpression of the CBDwt-GFP fusion reduced ph-p130Cas significantly. The control CBDmut-GFP fusion shows not effect on p130Cas phosphorylation.
Figure S2 Effects of disruption of Cav1/Rho-GTPase interaction on migration/invasion in vitro.
(a) Defective wound healing. B16 F10 cells stably expressing either RhoC-specific siRNA (F10siRhoC) or empty vector (F10Ev) were plated onto collagen IV-coated plates. At near-confluence, cultures were wounded by creating a “scratch” in the cell monolayer. At 48 hrs post-wounding, cell migration was examined in fixed cultures by staining with 1% methylene blue in 50% methanol. Representative photographs of three fields/culture are shown. The pictures depict one ‘corner’ of the wound.
(b, c) Reduced migration/invasion capacity. F10Ev and F10-siRhoC cells were cultured overnight in medium containing 0.2% serum and plated onto a matrigel-coated upper Boyden chamber. After 24-36 hrs, cells were fixed and stained as in (a). Stained cells on the upper portion of the matrigel-coated membrane were removed, the membrane was mounted onto a microscope slide, and stained cells on the lower portion of the membrane were photographed. Analysis of transwell-migration and invasion filters. Graph represents percentage of F10siRhoC cells able to move or invade in comparison to control cells.
Figure S3 In vivo effects of disruption of Cav1/Rho-GTPase interaction.
CBDwt expression reduces metastasis in female C57/B6 mice. Top panel, controls. Mice were xenografted with: PBS containing no cells (left), F10Ev cells (middle), and naive F10 cells (right). Middle panel, animals were injected with F10CBDmut cells. Bottom panel, animals were injected with F10CBDwt cells. Mice were imaged at the indicated time points post-injection using a Xenogen apparatus. Mice that received F10CBDwt cells showed markedly reduced metastasis at 4 weeks post-injection compared to mice injected with F10CBDmut cells.
Figure S4 Overexpression of CBDwt alters cellular morphology in vivo.
F10GFP cells expressing CBDwt (F10CBDwtGFP cells), or F10 cells expressing Ev plus tomato (F10Ev-tomato cells), were injected into chicken embryos as for Figure 7. Images revealing cellular morphology were acquired by confocal microscopy. In comparison to F10Ev-tomato cells, F10CBDwtGFP cells showed markedly more filapodia (green arrowheads), and many of the protrusions were abnormally branched.
Supplemental Movie Legends
Movie S1 Intravital videomicroscopy of F10Ev-tomato cell extravasation in the avian embryo chorioallantoic membrane.
A single B16 F10Ev-tomato cell (orange) arrested in the CAM capillary plexus is seen extravasating over a four hour period in a downward direction into the stroma. Vasculature is labeled on the luminal surface with fluorescein lens culinaris agglutinin (green). Each frame (upper panel) is a maximum intensity projection of a 150 µm stack (5 µm sections) taken every 5 minutes. Bottom frame is the orthogonal cross section of the CAM, where the tumour cell is clearly seen migrating out of the vasculature.
Movie S2 Intravital videomicroscopy of F10CBDwtGFP unsuccessfully extravasating in the avian embryo chorioallantoic membrane.
A single B16 F10CBDwtGFP cell (green) arrested in the CAM capillary plexus is seen partially completing extravasation into the stroma before dying. A significantly higher number of F10CBDwtGFP cells were observed unsuccessfully extravasating than F10Ev-tomato cells (see Figure 7b). Vasculature is labeled on the luminal surface with rhodamine lens culinaris agglutinin (orange). Each frame (upper panel) is a maximum intensity projection of a 150 µm stack (5 µm sections) taken every 5 minutes. Bottom frame is the orthogonal cross section of the CAM, where the tumor cell displays deficiency in migrating out of the vasculature and eventually undergoes cell death.
1