Human T24 bladder carcinoma and SW1222 colon carcinoma cells (American Type Culture Collection, Manassa, VA) were maintained in Dulbecco's modified Eagle's medium (DMEM, Life Technologies, Paisley, UK), 10% foetal calf serum (FCS) and glutamine (2 mM) in a humidified incubator, 37°C, in 5% CO2/air. Human umbilical vein endothelial cells from pooled donors (HUVECs, TCS CellWorks, Botolph Clayton, UK) were cultured on gelatine coated dishes in M199 media (Invitrogen, Paisley, UK) supplemented with 20% FCS, 4 mM L-glutamine, 20 μg/ml endothelial cell growth supplement (First Link, Birmingham, UK), and 80 μg/ml heparin. Only cells between the first and fourth passages were used for experiments. Human dermal microvascular endothelial cells (HDMECs, Clonetics, Cambrex, Wokingham, UK) were cultured in EBM-2 media (Clonetics) containing 5% FCS, hydrocortisone, hFGF-B, VEGF, IGF-1, ascorbic acid, hEGF and GA-1000 as supplied by the manufacturer (Clonetics).
Cells were treated and analysed when they either reached 40–50% confluence (sparse) or 80–90% confluence (dense). Sparse cells were actively dividing, whereas dense cells approached quiescence as determined by fluorescence activated cell sorting (FACS) analysis (results not shown). Densely growing endothelial cells were used to mimic quiescent, and possibly more mature vessels, whereas sparsely seeded cells were used as simple models to mimic the angiogenic phase of blood vessels. In a similar way, we studied sparsely and densely seeded tumour cell lines, as cell density has been implicated in gene expression studies in the past .
To achieve hypoxia and induce HIF-1, cells were plated in 6 cm oxygen impermeable Permanox dishes (Nalge, Nunc, Loughborough, UK) at 1–5 × 105/dish (4–18 × 103/cm2). Hypoxic conditions were achieved by placing the dishes in an anaerobic glove cabinet (in an atmosphere of 5% CO2, 5% H2, 90% N2, with palladium catalyst; Don Whitley Scientific LTD, Shipley, UK). The media pH was monitored and found to be the same in the aerobic incubator and the anaerobic glove cabinet. There was no significant difference is cell viability between aerobic and hypoxic cells over a 24 h period (results not shown).
Alternatively, the hypoxia-mimetic agent, CoCl2 (Sigma Aldrich, Gillingham, UK), was used at 100 μM concentration for varying periods of time 0–24 h . Maximum HIF-1 accumulation was reached within 4 h of hypoxia or CoCl2 treatment, and stayed at that level for up to 24 h, and hence 4 h was used as the standard time for subsequent experiments.
CA-4-P was synthesised in house (Dr. M Naylor et al., Gray Cancer Institute; ) and used at 0.001, 0.01, 0.1, or 1 μM in complete medium, avoiding excessive exposure to light. The concentrations of CA-4-P used are clinically relevant (CA-4 peak plasma concentration at a dose of 68 mg/m2 reached 2.26 μM; CA-4 concentration stayed above 0.01 μM in plasma for over 10 h in those patients ) and were shown to induce both morphological and functional changes in endothelial cells in vitro . CA-4-P is the more soluble, phosphated, form of the active compound CA-4. CA-4-P is commonly used for in vitro and in vivo studies, where it is cleaved to the natural form by endogenous phosphatases and taken up into cells .
The Rho kinase inhibitors Y27632 (10 μM; Welfide Corporation (Osaka, Japan)) or HA1077 (10 μM; Calbiochem (Nottingham, United Kingdom)) were added to HUVECs for 1 hour prior to treatment with CA-4-P. HUVECs were incubated with the Rho A inhibitor C3 exoenzyme (10 μg/ml; Upstate (Dundee, Scotland)) for 12 hours before washing and treatment with CA-4-P. We have previously used these concentrations and exposure times, and shown them to be effective in preventing CA-4-P-induced cytoskeletal changes .
Cells were fixed and stained essentially as described previously . Briefly, formalin-fixed and Triton-X-permeabalised cells were stained with primary antibody (Ab) (anti-HIF-1α at 1/250 (BD Biosciences, Cowley, UK), followed by secondary Ab (biotinylated anti-mouse, Vector Laboratories, Peterborough, UK) and FITC – Avidin (Vector laboratories) with DAPI (DNA stain, Vector Laboratories). Slides were visualised on a Nikon Eclipse TE200 Microscope.
For HIF-1 detection, denatured total cell lysates were prepared as described previously . Care was taken to include both adherent and floating cells, although few floating cells were observed. For NFκB studies, cellular fractions were isolated using a NE-PER kit (Perbio, Cramlington, UK) according to manufacturer's instructions. For Western blot analysis, equal amounts of protein (Pierce Micro BCA kit, Rockford, IL) were separated on NuPAGE 4–12% Bis-Tris Gels (Invitrogen) and transferred to nitrocellulose membranes. Anti-HIF-1α (primary Ab at 1/250, BD Biosciences) and secondary anti-mouse (HRP-labelled, Dako, Ely, UK) antibodies, or goat anti-human anti-NFκB (p65) antibody (Dako) and secondary anti-goat (HRP-labelled, Dako) were used. Immunoreactive bands were visualised by enhanced chemiluminescence (ECL, Amersham-Biosciences, Chalfont St Giles, UK). Equal protein loading was confirmed using EZBlue gel stain (Sigma). Normalisation using actin staining was not possible in this study, as CA-4-P acts directly on the actin cytoskeleton .
Electrophoretic mobility shift assay (EMSA)
EMSAs were performed on nuclear fractions using an EMSA kit for NFκB (Panomics, Cambridge Biosciences, Cambridge, UK). Competition assays were performed in the presence of 100-fold excess of cold (unlabelled) NFκB probe. Supershift assays for p50 and p65 components were performed by incubation of samples with antibodies against p50 or p65 for 30 minutes prior to the addition of probe.
Semi-quantitative real-time RT-PCR
RNA was extracted using Trizol (Invitrogen, Carlsbad, CA) following manufacturer's instructions. Total RNA (1 μg) was DNase-1 pre-treated (Invitrogen) immediately prior to reverse-transcription using a SuperScript™ First-Strand Synthesis System for RT-PCR (Invitrogen) according to manufacturer's instructions. Polymerase chain reaction (PCR) was carried out in a 25 μl reaction volume containing 1 μl of prepared cDNA, 10 μl RealMasterMix™ Probe (Eppendorf, Hamburg, Germany), 200 nM of each LUX FAM-labelled primer pair for the gene of interest (Invitrogen) and 100 nM of a commercial 18S LUX JOE-labelled primer (Invitrogen). Real-time RT-PCR was carried out in a Corbett Rotor-Gene 3000 thermal cycler (Mortlake, Australia). Cycling parameters consisted of an initial denaturation step at 95°C for 1 minute, followed by 45 cycles of denaturation at 95°C for 20 s, primer annealing at 63°C for 20 s with data acquisition, and extension at 68°C for 20 s, followed by melt curve analysis. Primer sequences for VEGF were as follows: Reverse primer (5'-3') CGA AGC CAT GAA CTT CAC CAC TT [FAM] G; forward primer (5'-3') GCT CTA CCT CCA CCA TGC CAA G.
The mean levels were compared between groups using ANOVA. Where significant differences between groups were identified, these were further explored using Fisher's Least significant difference tests. A value of p < 0.05 was considered significant.