Gerber et al. 10.1073/pnas.0611492104. |
Fig. 5. Structural requirements for VEGF binding to different antibodies. (A) Structural representation of the footprint of the bevacizumab-Fab bound to human VEGF (green). The 10 aa mutated from mouse to human are shown in orange. Ser-88 is the only nonconserved amino acid located within the area of ligand/antibody interaction and colocalizes with the VEGF receptor 1-binding site (data not shown). (B) Molecular stick representation of murine VEGF-A (yellow), including Ser-77 (red), with the additional space taken by serine relative to the glycine (human) highlighted in blue. The presence of the additional CH2 groups in serine led to interference with bevacizumab Fab. (C) Footprint of G6-31-Fab binding to hVEGF-A (green), with Ser-88 (blue) overlapping (D) Footprint of B20-4 Fab binding to hVEGF-A (green). Atoms of VEGF that are 4.7A or closer to the respective Fab and nonconserved are colored blue (Lys/Arg 111).
Fig. 6. Schematic representation of the targeting vectors to generate hum-I and hum-X KI mice. Mutations were introduced in exons 3-5 of the targeting vectors, resulting in mice expressing the hum-I or hum-X VEGF form, designated in the figure as Mut I and Mut X, respectively. Hum-X VEGF protein consists of the following mutations: muVEGF-R26H, A57G, A64G, S71E, S87G, S99N, R100K, T110A, K111R, P112Q. This nomenclature starts from the mature sequence.
Fig. 7. Growth curves (A) and terminal weight (B) of Calu-6 tumors after administration of low-dose anti-VEGF Mabs. Calu-6 cells were implanted in RAG2 KO; hum-X VEGF KI. Treatment with anti-VEGF Mabs, at the dose of 0.5 mg/kg twice weekly, was initiated 48 h after tumor cell inoculation. Tumor weight was determined at day 42 in the control group and day 70 in the anti-VEGF Mab groups.
Fig. 8. Hepatic changes in RAG2 KO; hum-X VEGF KI double-homozygous mice treated with anti-VEGF-A antibodies (human Fc framework). Animals treated (5 mg/kg, two times weekly for 54 days) with antibodies having increasing affinity for VEGF-A have increased murine VEGF-A staining of sinusoidal endothelium (A-E), and increased numbers of F4/80- and MAC-2-positive mononuclear cells (macrophages and/or Kupffer cells) adherent to central veins (F-J).
Table 1. Binding of muVEGF164, huVEGF165, and hum-X-VEGF to bevacizumab, Y0317, G6-31, and B20.4 Mabs. Kd (Koff/Kon), nM, was determined at 37°C
Bevacizumab | Y0317 | B20-4.1 | G6-31 | |
huVEGF165 | 4.3 | 0.01 | 1.7 | 0.3 |
Hum-X VEGF | 3.2 | 0.02 | 2.3 | 0.5 |
muVEGF164 | NB | 479 | 1.1 | 0.2 |
Each measurement represents an average of three independent assays that vary <20%. NB, no binding.
Table 2. Inhibition of VEGF-stimulated bovine retinal capillary endothelial cell proliferation by bevacizumab, Y0317, B20-4.1 or G6-31 Mabs
Bevacizumab IC50 (ng/ml) | Y0317 IC50 (ng/ml) | B20-4.1 IC50 (ng/ml) | G6-31 IC50 (ng/ml) | |
muVEGF164 | NA | NA | 500 | 3.8 |
huVEGF165 | 32 | 0.75 | 52 | 4.9 |
hum-X VEGF | 55 | 2.7 | 76 | 6.1 |
IC50 of various antibodies to interfere with proliferation of bovine retinal capillary endothelial cells (ACE) stimulated for 5 days by various forms of VEGF. Data shown are means from triplicate experiments that varied by <20%.
SI Text
Construction of the Targeting Vector and Generation of ES Cells and hum-X VEGF KI Mice.
Ten amino acids within the genomic targeting vector for VEGF-A consisting of exon 3, 4 and 5 of mouse VEGF-A (1) were mutated from mouse to human sequences. For site directed mutagenesis of the residues located within exons 3, 4 and 5, the following oligonucleotides were used: Exon3-R/H: AGCGAAGCTACTGCCATCCGATTGAGACC, Exon3-A/G,A/G: TGATGCGCTGTGGAGGCTGCTGTAACGATGAAGGCCTG, Exon3-A/G,S/E: TGTAACGATGAAGGCCTGGAGTGCGTGCGTGCCCACGGAAGAGAGCAAC. For exon 4: Exon4-S/G: ATCAAACCTCACCAAGGCCAGCACATAGGAGAGATG, Exon4-S/N, R/K: TGAGCTTCCTACAGCACAACAAATGTGAATGCAGGTG, Exon5-T/A,K/R,P/Q: TGCAGACCAAAGAAAGACAGAGCACGGCAAGAAAAGTAAGTGG. The corresponding amino acids are: muVEGF-R26H, A57G, A64G, S71E, S87G, S99N, R100K, T110A, K111R, P112Q. Correct recombination events in ES cell were identified by PCR analysis and confirmed by Southern blot as described previously (1). Briefly, in correctly targeted ES cells, the neomycin resistance marker flanked by Lox-P sites was deleted by transient expression of Cre recombinase. Correct genomic recombination products were identified by genomic PCR and confirmed by Southern Blotting of the 3' and 5' flanking regions. ELISA experiments confirmed binding of A4.6.1 to hum-X VEGF protein present in conditioned media of targeted ES cells (data not shown). In addition, the genomic DNA isolated from selected ES cell clones was digested with EcoRI and analyzed by Southern blotting as described (1) and by genomic sequencing to test for correct recombination events. One derivative of three different parental ES cell clones containing the floxed VEGF allele was used to generate chimeric mice by microinjection into the blastocoele cavity of 3.5-day C57BL/6N blastocysts (2). Chimeric males were mated with C57BL/6N females and agouti offspring were screened for germline transmission by PCR analysis for VEGF alleles containing the loxP-1 and loxP-3 sites as described previously. For generation of hum-X VEGF.ki.Rag2.ko double homozygous animals, hum-X VEGF het females (B6.129) were mated to Rag2.ko males [B6 (H2b) (#RAGN12-M; Taconic)]. Double heterozygous animals were interbred to produce double homozygous hum-X VEGF.ki;Rag2.ko animals. The strain is maintained as double-mutant breeding sets.Determination of Antibody Binding Affinities.
For binding affinity measurement, surface plasmon resonance (SRP) measurement with a BIAcore-3000 (BIAcore, Inc., Piscataway, NJ) was used. Carboxymethylated dextran biosensor chips (CM5; BIAcore Inc.) were activated with N-ethyl-N'- (3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC) and N-hydroxysuccinimide (NHS) according to the supplier's instructions. Human or murine VEGF-A was immobilized to achieve »60 response units (RU). Two-fold serial dilutions of IgG (0.78-500 nM) were injected in PBS with 0.05% Tween 20 (PBST) at 37°C at a flow rate of 25 ml/min. Association rates (kon) and dissociation rates (koff) were calculated using one-to-one Langmuir binding model (BIAcore Evaluation Software version 3.2) by simultaneous fitting the association and dissociation sensorgram. The equilibrium dissociation constant (Kd) was derived as the ratio koff/kon.Endothelial Cell Proliferation Assays.
Bovine retinal microvascular endothelial cells were seeded at a density of 500 cells per well in 96-well plates in growth medium (low-glucose DMEM supplemented with 10% calf serum, 2 mM glutamine, and antibiotics). For inhibition assay, antibodies were added at indicated concentrations first. After 0.5-1 h, hVEGF-A, mVEGF-A or MutX were then added to a final concentration of 6 ng/ml. After 6-7 days, cell growth was assayed with the use of alamar Blue (BioSource). Fluorescence was monitored at 530-nm excitation wavelength and 590-nm emission wavelength. IC50 values from quadruplicate determinations were calculated using KaleidaGraph.Long-Term Treatment with anti-VEGF-A Mabs.
Eight- to 9-month-old hum-X VEGF-KI mice were treated twice weekly, i.p., with 10 mg/kg of the indicated antibody for the duration of 90 days. Alternatively, 5 mg/kg, IP, once weekly was administered. Body weights were assessed weekly, serum was harvested via retro-orbital bleeding and submitted for PK and blood chemistry analysis. Mice were killed when changes in body weights exceeded 20% and/or ascites formation was prominent.Histology and Immunohistochemistry (IHC).
Tumor tissues were fixed in 10% neutral buffered formalin for 12-16 h before paraffin embedding. Histologic sections 4-5 microns thick were stained with hematoxylin and eosin. Murine VEGF-A was detected by using 0.5 mg/ml goat polyclonal antibody from R & D Systems (AF-493-NA); rehydrated paraffin-embedded tissues were treated with Target retrieval solution (S1700; DAKO) at 99° C for 20 min, followed by 20 min at room temperature. Primary antibody was detected with biotin-conjugated rabbit anti-goat, avidin-biotin complex (Vectastain ABC Elite; Vector Labs) and metal-enhanced diaminobenzidine (Pierce). Complement factor C3 was detected by direct immunofluorescence on frozen sections using FITC-conjugated anti-complement F(ab')2 (Cappel Labs). Anti-VEGF monoclonal antibodies were detected by direct immunofluorescnce using FITC-conjugated rabbit anti-human Fc (Jackson Immunoresearch). Methacrylate-embedded 1-mm-thick sections were stained with toluidine blue or Jones silver stain for basement membrane. Ultrathin sections were stained with uranyl acetate/lead citrate and examined on a Philips CM12 transmission electron microscope.Purification of Recombinant hum-X VEGF.
Pellets from bacterial cells expressing hum-X VEGF were resuspended in 10 volumes of 25 mM Tris, 5 mM EDTA, pH 7.5, with a Polytron homogenizer. Cells were lysed by passing the cell suspension through a Microfluidizer (Microfluidics International) and the solution was clarified by centrifugation. The pellet was resuspended in extraction buffer containing 7 M urea, 50 mM Hepes, and 10 mM DTT, pH 8, and the solution was stirred at room temperature for 1 h. The solution was centrifuged at 33,000 ´ g for 30 min to remove insoluble cell debris and the supernatant containing denatured and reduced hum-X VEGF was diluted tenfold into refolding buffer (1 M urea, 50 mM Hepes, 15 mg/L dextran sulfate 8000, 0.05% Triton X-100, pH 8.2). The refolding mixture was stirred overnight at room temperature and then centrifuged to remove precipitated protein. Ammonium sulfate was added to 1 M concentration before loading the mixture onto a Phenyl TSK column equilibrated in 1 M ammonium sulfate, 25 mM Tris, pH 7.5; the hum-X VEGF was eluted with a decreasing ammonium sulfate gradient in this buffer to 0 M. hum-X VEGF containing fractions were pooled and further purified on a preparative C4 reversed phase column (Vydac). Fractions containing the dimeric hum-X VEGF were pooled and lyophilized.Anti-VEGF-A Antibody ELISA.
VEGF-A coat format to determine free anti-VEGF-A antibodies. Maxisorp 96-well ELISA plates (Nunc, Roskilde, Denmark) were coated overnight with 0.5 mg/ml VEGF-A165 in 50 mM sodium carbonate, pH 9.6, at 100 ml per well. Plates were washed with PBS containing 0.05% polysorbate 20 and blocked with 150 ml/well of 0.5% BSA, 10 ppm Proclin 300 (HyClone, Logan, UT) in PBS at room temperature for 1 h. Two-fold serial dilutions of standards (0.0625 -8 ng/ml of anti-VEGF mouse IgG2a, anti-VEGF human IgG1, or trap-human IgG1) in 0.05% BSA, 0.2% bovine m-globulins (Sigma, St. Louis, MO), 0.25% CHAPS, 5 mM EDTA, 0.35M NaCl, and 0.05% polysorbate 20 in PBS, pH 7.4 (samples buffer), and samples (minimum 1:20 dilution) were added to the plates at 100 ml per well. Plates were incubated at room temperature for 2 h and washed. Bound mouse IgG2a antibodies and human IgG1 anti-VEGF-A antibodies were detected by adding 100 ml per well of anti-mouse IgG2a-HRP (Pharmingen, San Diego, CA) and anti-human FcHRP (Jackson ImmunoResearch, West Grove, Pennsylvania), respectively. After a 1-hour incubation, plates were washed, and the substrate 3,3',5,5'-tetramethyl benzidine (Kirkegaard and Perry Laboratories, Gaithersburg, MD) was added (100 ml per well). The reaction was stopped by adding 1M H3PO4 (100 ml per well). The absorbance was read at 450 nm by using a SpectraMax 250 microplate reader (Molecular Devices Corp., Sunnyvale, CA). The titration curves were fit by using a four-parameter regression curve-fitting program (KaleidaGraph; Synergy Software, Reading, PA). Data points within the range of the standard curve were used for calculating the anti-VEGF-A antibody concentrations in samples.1. Gerber HP, Hillan KJ, Ryan AM, Kowalski J, Keller GA, Rangell L, Wright BD, Radtke F, Aguet M, Ferrara N (1999) Development (Cambridge, UK) 126:1149-1159.
2. Hogan B, Beddington R, Constantini F, Lacy E (1994) Manipulating the Mouse Embryo: a Laboratory Manual (Cold Spring Harbor Lab Press, Plainview, NY).