Isolation and characterization of human anti-VEGF165 monoclonal antibody with anti-tumor efficacy from transgenic mice expressing human immunoglobulin loci

Cancer Letters 273 (2009) 28–34

Contents lists available at ScienceDirect

Cancer Letters journal homepage: www.elsevier.com/locate/canlet

Isolation and characterization of human anti-VEGF165 monoclonal antibody with anti-tumor efficacy from transgenic mice expressing human immunoglobulin loci q Yu Liu a,*, Dong-jie Yun a, Jian-quan Chen b, Jian-yang Zhao b, Si-guo Liu b, Guo-xiang Cheng b,* a b

Department of biochemistry, China Pharmaceutical University, No. 24 Tong Jia Xiang, Nanjing 210009, PR China Shanghai Genon Bioengineering CO., LTD., 88 Cailun Road, Shanghai 201203, PR China

a r t i c l e

i n f o

Article history: Received 17 April 2008 Received in revised form 17 April 2008 Accepted 10 July 2008

Keywords: Five-feature mice Fully human mAb Anti-tumor VEGF165

a b s t r a c t The purpose of this study was to prepare a fully human anti-VEGF (vascular endothelial growth factor) monoclonal antibody with anti-tumor activity from five-feature mice which express human immunoglobin loci. Four hybridomas secreting mAb stably were isolated successfully. Some characters such as isotypes, cross-reactivity, inhibition on the binding of hVEGF to VEGFR-2, dissociation constants and the idiotypic characteristic were determined. Proliferation of T24 and Ls-174-T cell line and nude mice bearing human colorectal cancer were used to evaluate therapeutic effects and safety of this mAb. Pharmacokinetics data shows the half life of this mAb was about 5 days after a single intravenous injection. These results suggest the fully human anti-VEGF mAb maybe safe and efficient for cancer treatment. Ó 2008 Elsevier Ireland Ltd. All rights reserved.

1. Introduction The development of new blood vessels (angiogenesis) is important in the pathogenesis of many disorders, particularly rapid growth and metastasis of solid tumors [1,2]. Vascular endothelial growth factor (VEGF) is an endothelial cell specific mitogen and an angiogenesis inducer released by a variety of tumor cells and expressed in human tumors in situ. K. Jin Kim demonstrated that inhibition of the action of VEGF (mAb specific for VEGF was used) spontaneously produced by tumor cells may suppress tumor growth in vivo [3]. Anti-angiogenesis therapy is unique in targeting tumor vasculature, but not tumor cells themselves, and therefore it is broadly applicable for most solid tumors. However, the immunogenicity of rodent antibodies in humans prevents

q Supported by Natural Science Foundation of China (30672478) and Babraham Institute, Cambridge, England. * Corresponding authors. Tel.: +86 25 8327 1248; fax: +86 25 8327 1249 (Y. Liu); tel.: +86 21 58951015; fax: +86 21 58951012 (G.X. Cheng). E-mail addresses: [email protected] (Y. Liu), [email protected] (G.X. Cheng).

0304-3835/$ - see front matter Ó 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2008.07.022

their application in clinical procedures. To overcome this problem, the humanized and fully human antibodies were constructed. Fully human antibodies could be produced by technologies such as phage display libraries and transgenic mice expressing human Ig genes. It is known that Bevacizumab is a successful humanized anti-VEGF mAb, approved by FDA in 2004, which has been most extensively investigated in a variety of tumors, including non-small cell lung, breast, prostate, renal and colorectal cancers. Bevacizumab in combination with bolus IFL (irinotecan, 5-fluorouracil (5-Fu), leucovorin (LV)) become the first-line therapy of metastatic colorectal cancer [4]. In this study, we presented a successful isolation of fully human anti-VEGF IgM mAb from transgenic humanized mice, i.e., five-feature mice, which were produced by Babraham Institute, Cambridge, England [5]. We immunized several five-feature mice with the recombinant human VEGF expressed by Pichia pastoris. Four of nine specific mAb were isolated and analyzed. The data presented here suggested that anti-VEGF human IgM mAb can be produced in five-feature mice, enabling the generation of a panel of antibodies with therapeutic potential.

Y. Liu et al. / Cancer Letters 273 (2009) 28–34

2. Materials and methods

29

dine tetrahydrochloride plus hydrogen peroxide was added and the reaction was stopped by distilled water addition.

2.1. Translocus mouse strains and immunogens 2.5. Isotypes and cross-reactivity determination The translocus five-feature mouse strains, which carry human IgM + Igj, k transloci [5], was provided by Babraham Institute (Cambridge). The recombinant human vascular endothelial growth factor (hVEGF165) was expressed in P. pastoris and purified as previously described [6]. 2.2. Immunizations and fusions Groups of 6- to 8-week-old mice were immunized subcutaneously with 100 lg rVEGF165 expressed in P. pastoris. In the primary boost, rVEGF165 was emulsified with Complete Freund’s Adjuvant (CFA) (Sigma). Each mouse was boosted four times at 2-week intervals with the protein in Incomplete Freund’s Adjuvant (IFA) (Sigma). Animals with an anti-hVEGF165 serum response were sacrificed, and their splenocytes were fused with SP2/O myeloma cells in the presence of 50% PEG1500, HAT (Sigma), and HFCS (Hybridoma fusion and Cloning Supplement, Sigma) using standard techniques [7]. 2.3. mAbs purification Stable, positive clones secreting hVEGF-specific IgM mAbs were identified by screening supernatants from hypoxanthine–aminopterin–thymidine (HAT)-selected hybridomas by ELISA on microtiter plates coated with 1 lg of hVEGF165 per ml and developed with horseradish peroxidase (HRP)-labeled goat anti-human IgM (l chain, Sigma). Stable, positive clones were selected by subcloning twice by limiting dilution. Human VEGF165 mAb containing ascitic fluid was prepared by injecting hybridoma cells into the peritoneal cavity of pristine (Sigma–Aldrich Co.)-primed Balb/C mice. mAb concentrations in ascetic fluid were determined by ELISA. Ascitic fluid was collected and centrifuged to remove the oil. At a temperature of 4 °C, MBP–Sepharose 2B (5 ml) was packed into a chromatographic column. Binding and washing steps are performed at 4 °C in 10 mM Tris–HCl (pH 7.4) buffer containing sodium chloride and 20 mM calcium chloride. Elution is made at room temperature in a similar Tris buffer, except that it contains EDTA and is devoid of calcium chloride. The IgM fractions were pooled, and subjected to ultrafiltration against PBS buffer and stored at 20 °C. 2.4. Human anti-VEGF mAbs characterization by SDS–PAGE and Western blotting To further access the reactivity of antibodies to purified hVEGF165, an immunoblotting analysis was performed. Briefly, 20 lg per slot of purified hVEGF165 were applied to a 7.5% SDS–polyacrylamide gel. After electrophoresis, the separated proteins were transferred to PVDF membrane (Millipore) at 100 V for 18 h at 4 °C. The membrane was blocked with 3% BSA for 2 h and developed with HRP-labeled anti-human l chain antibody (Sigma) diluted 1:10,000. After washing, the substrate 3,3-diaminobenzi-

To identify potential murine-human hybrid mAbs and determine the species-specific heavy and light-chain isotypes of all mAbs isolated, two ELISAs were performed. In one ELISA, each mAb was added to microtiter plates (Costar, Corning) coated with 1 lg of recombinant human VEGF165 per ml. Each well was then developed with HRPlabeled goat anti-mouse IgM, IgG1, IgG2a, IgG2b, IgG3, IgA (Southern Biotechnology Associates) or HRP-labeled anti-human IgM (l chain, Sigma). In the second ELISA, the HRP-conjugated mouse antibodies specific for human j, k chain, and murine k chain (Sigma) were used to determine the light-chain isotypes. The specificity of these mAbs was confirmed by ELISA, using as controls several unrelated antigens such as human epidermal growth factor (EGF), basic fibroblast growth factor (bFGF) and human serum albumin. 2.6. The idiotypic characteristic Five-feature male mice, 6-weeks-old, were obtained from Shanghai transgenic center. The mice were normally bred and maintained under specific-pathogen-free conditions, with a constant temperature ranging between 25 and 27 °C, and a constant humidity ranging between 45% and 50%. The mice were injected the MAb V75 3 times at 4 days intervals. At the end of experiment, serum after each injection was got to test its idiotypic characteristic by ELISA. Serum (100 lL) was mixed with 20 lL V75 (500 lg/ml) at 37 °C for 2 h until the equilibrium was reached. The mixture was added to microtiter plates (Costar, Corning) coated with 1 lg of recombinant human VEGF165 per ml. Each well was then developed with HRP-labeled anti-human IgM (l chain, Sigma). Twenty microliters of V75 (500 lg/ml) diluted with 100 lL blank serum was determined as control. Absorbance at 490 nm was observed. 2.7. Inhibition of mAbs on the binding of hVEGF to VEGFR-2 The inhibition effect of mAbs on hVEGF165 binding to VEGFR-2 was observed by ELISA. Microtiter plates were coated with VEGFR-2 (Sigma, diluted 10,000 times) and developed with horseradish peroxidase (HRP)-labeled human mAbs (were made as described [8]) and 1 nM hVEGF165 mixture. Labeled human IgM was determined as control. Absorbance at 490 nm was observed. 2.8. Affinity of mAbs The dissociation constants (Kd) of mAbs were determined by their affinity to hVEGF165 as described by Friguet et al. [9]. mAbs were incubated in solution with different antigen concentration until the equilibrium was reached, then a classical indirect ELISA measured the proportion of mAbs that remained unsaturated at equilibrium. The Kd

30

Y. Liu et al. / Cancer Letters 273 (2009) 28–34

was deduced by linear regression by CurveExpert 1.3 program. 2.9. HMVEC proliferation assay The ability of each hVEGF165-specific mAb to neutralize the effects of purified hVEGF165 was examined in vitro with human dermal microvascular endothelial cells (HMVEC) proliferation assay. HMVEC cells were seeded at 2.5  104/ml in 96-multiwell plates in triplicates in M131 medium (Cascade Biologics), supplemented with Microvascular growth supplement (Cascade Biologics) as recommended by the manufacturer. All culture media were supplemented with 100 U/ml penicillin, 100 lg/ml streptomycin, and 2 mM L-glutamine. After the cells were maintained overnight at 37 °C with 5% CO2 in a humidified incubator, anti-hVEGF mAbs (100 lg/ml) and 5-Fluorouracil (5-Fu, 200 lg/ml) were added in as 10 lL/well respectively. As a control, the inappropriate human IgM was used at the same concentration. After the cells were incubated at 37 °C for 4 days, 10 ll/well MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, Sigma] reagent was added in each well and incubated for another 4 h. The medium was then removed and 150 lL DMSO was added to completely resolve precipitate in the cells. OD values of each well were read at 490 nm and the results were evaluated with the t-test. 2.10. T24 cell and Ls-174-T proliferation assay Human bladder cancer cell line T24 and human colorectal cancer line Ls-174-T was obtained from the Chinese Institute of Biochemistry and Cell Biology (SIBCB). The cells were grown as monolayers in RPMI 1640 supplemented with 10% fetal bovine serum in a 37 °C humidified incubator with a mixture of 95% air and 5% CO2. T24 cells were seeded at 2.5  104/ml and LS-174-T cells were seeded at 1.5  104/ml respectively in 48-multiwell plates (200 ll/ well) as triplicates and maintained overnight at 37 °C with 5% CO2 in a humidified incubator. Twenty microliters per well of anti-hVEGF165 mAbs (100 lg/ml) and 5-Fluorouracil (5-Fu, 200 lg/ml) were added in, respectively. The method of MTT assay was the same as described above. 2.11. Effect on animal bearing tumor model BALB/c nude male mice, 5-weeks-old, were obtained from Shanghai SLAC Laboratory Animal CO., LTD. The mice were normally bred and maintained under specific-pathogen-free conditions, with a constant temperature ranging between 25 and 27 °C, and a constant humidity ranging between 45% and 50%. Ls-174-T cells (0.1 ml) were injected subcutaneously at the back of nude mice (n = 5) at a concentration of 5  107cells/ml to establish a model of tumor-bearing mice. About 30 days post-implanting, the bearing-tumors were extracted. After necrotic tissue and noncancerous tissue of the specimens were removed, the remaining cancerous tissues were cut into small pieces of about 1 mm3 in size. Then, the nude mice models were established by implanting tumor bits directly in the right flank.

Ten days post-operation, when the diameter of the tumor reached about 0.3–0.5 cm, mice were divided into a control group and three experiment groups (n = 10 for each group). Each mouse in the experiment groups was injected via the tail vein with V2, V75, or 5-Fu (respective 10, 10, and 100 mg/kg every other day for 14 days). Mice in control group were administrated intravenously (i.v) with 0.2 ml 0.9% normal saline (NS). The experiments were terminated and mice were sacrificed by cervical decapitation 15 days after treatment. Transplanted tumors harvested at autopsy were weighed and sectioned, and fresh tissue samples were wrapped in aluminum foil and immediately frozen in liquid nitrogen. All animal procedures were performed according to standard protocol and in accordance with the standard recommendations for the proper care and use of laboratory animals. The general conditions of the mice were observed every day. Nude mice were weighted and the weight of tumors was measured at the end of experiment. Other organ metastases were detected with autopsy when the mice were killed. The tumor growth inhibition rate (IR) was calculated according to the weight of the tumor: IR (%) = (1  weight of experiment group/weight of control group)  100%. 2.12. Pharmacokinetics of human anti-VEGF mAb in rats 2.12.1. Radiolabeling of anti-VEGF mAb The mAb V75 was radioiodinated with 125I (China Atomic Energy Institute, Beijing, China) with Iodogen method [10]. 125I-VEGF-mAb was separated and purified on a Sephadex G50 column equilibrated with nitrogen purged 0.01 mol/l PBS at pH 7.4, containing 2% human serum albumin. The radiochemical purity was measured with trichloroaceticacid precipitation [11]. Assessment of the immunoreactivity was performed using a direct binding assay, and the binding efficiency was calculated according to the following equation: Binding efficiency ð%Þ ¼ Radioactivity of depositionðcpmÞ  100%. Total radioactivityðcpmÞ 2.12.2. Pharmacokinetics analyses Pharmacokinetics analyses for anti-VEGF mAb were conducted using SD rats. All rats were administrated 1% potassium iodide in their drinking water for 4 days before the experiment, in order to reduce the exposure of their salivary glands to unwanted radiation. Eight rats in each group (180–220 g, four males and four females per group) received a single-dose intravenous injection of 14.0, 7.0, and 3.5 mg/kg body weight of V75 mAb, and the injection radioactivity dose of three groups was 1.77  106Bq kg1. Serum samples were collected 0 (predose), 0.25, 0.5, 1, 2, 4, 8 h and 1, 2, 3, 5, 7, 9, 12, and 15 days after injection. Antibody concentrations in sera were determined by gamma counter. 2.13. Statistical analysis Data were processed using Microsoft EXCEL version 2003. Pharmacokinetics data were analyzed using Drug And Statistics for Windows (DAS ver2.0) program.

31

Y. Liu et al. / Cancer Letters 273 (2009) 28–34

Measurement data were expressed as mean ± SD and Student’s t-test for two groups comparison was used.

3. Results 3.1. Generation of fully human anti-VEGF mAb in transgenic mice The presence of non-human sequences has been a major factor in limiting the utility of monoclonal antibodies as human therapeutics. To overcome this problem, transgenic mice expressing human Ig genes provide an alternative method for isolating human mAb of potential therapeutic value. Five-feature mouse, a human IgM producing transgenic mouse strain [1], were immunized with hVEGF165 expressed in P. pastoris in a soluble native conformation. Serum titers were tested by ELISA at 2-week intervals and were typically in the range of 1:104–105 after three immunizations. At this point, the mice received an additional injection of hVEGF165 and were then sacrificed. Hybridomas were produced from splenic B cells using standard techniques, and a panel of 77 hybridomas producing fully human IgM anti-hVEGF165 mAbs was generated. After being subcloned 4–5 times, 4 hybridomas (V2, V50, V71, and V75) which secrete human IgM anti-hVEGF165 monoclonal antibodies stably were chosen by ELISA screening. The binding of mAb to the intact hVEGF165 was then tested by ELISA (Fig. 1). The anti-VEGF165 mAb was purified to homogeneity by precipitation with ammonium sulfate followed by affinity chromatography on MBP IgM Purification column (Amersham Pharmacia Biotech). The purity and the molecular weight of the V2 and V75 were determined by reduced SDS–PAGE (Fig. 2). One major band at 80 kDa (heavy chain) and one major band at 30 kDa (light chain) were observed. These results verified that the anti-VEGF165 antibody was homogeneous. These two mAbs were characterized by Western blotting compared with mouse ascites (Fig. 3). The cross-reactivity of different IgM antibodies was determined using ELISA. As shown in Fig. 1, all four antibodies bound human VEGF165 with high affinity. But V50 has some cross-reactivity with bFGF (p < 0.05).

Fig. 3. Both purified human IgM V2 (lane 2), V75 (lane 3), and mouse ascite (lane 1) were characterized by Western blotting.

3.2. The result of idiotypic characteristic Three five-feature mice were injected with purified V75 3 times. More than 95% of the anti-hVEGF165 reactivity remained after incubation of V75 compared with control. There was no detectable anti-idiotype reactivity when the anti-V75 sera were incubated with human monoclonal antibody V75. These properties demonstrate that the serum from those three five-feature mice can not block the binding of the MAb V75 to its antigen hVEGF165 (Table 1). 3.3. Blockade of VEGFR-2 binding To determine the ability of the anti-VEGF165 antibodies to block binding of VEGF165 to its receptors, a VEGF receptor2 binding assay were performed using ELISA. It has been previously reported that hVEGF165 binds to VEGFR2 with high affinity (Kd = 400–800 pm) [12]. As shown in Table 1, the inhibition is antigen-specific because the control antibody human IgM had no effect. 3.4. Inhibition of HMVEC proliferation by the fully human anti-VEGF mAbs This assay was used to determine the effects of the antibody on the interactions of growth factors with their receptor. Since VEGF/VEGFR interactions are thought to play a role in proliferation of endothelial cells, Table 1 Result of V75 on idiotypic characteristics (mean ± SD, n = 3)

Five-feature mice serum Control serum

1 Time

2 Times

3 Times

1.712 ± 0.324 1.788 ± 0.208

1.733 ± 0.256 1.803 ± 0.312

1.688 ± 0.365 1.821 ± 0.334

1 time: Serum was got after injection of V75 1 time.

Fig. 1. Binding of mAbs to hVEGF165 and other unrelated antigens were tested by ELISA. mAb solution (100 ll) was added into ELISA wells respectively. Values represent the mean ± SD for three independent experiments. We can see V2 has cross-reactivity with bFGF from the figure.

Fig. 2. Reduced SDS–PAGE analysis of V2 (lane 2) and V75 (lane 3) purified by MBP affinity. Lane1 was molecular mass marker proteins. The proteins were stained with Coomassie Blue.

Fig. 4. Inhibition of the VEGF-induced HMVEC proliferation with VEGFspecific antibodies (mean ± SD) n = 3. Each column represents the mean of three individual experiments. *Significant decrease the number of living HMVEC cells in mAbs (p < 0.05).

32

Y. Liu et al. / Cancer Letters 273 (2009) 28–34

HMVEC cells which express VEGFR-3 can be used to demonstrate inhibitory effects of biospecific antibody molecules. After the HMVEC cells were cultured in the presence of VEGF165 (2 ng/ml) overnight at 37 °C with 5% CO2, 10 lL/well of anti-hVEGF165 mAbs (100 ng/ml) and 5-Fu (200 ng/ml) were added, respectively. The results presented in Fig. 4 demonstrate that V50 and V75 reduced the number of living cells significantly (p < 0.05).

Table 2 Blockade of VEGFR-2 binding

3.5. Inhibition of T24 and Ls-174-T cell proliferation by the fully human antiVEGF mAbs

scopic examination at the time of sacrifice, implanted tumor survived in all nude mice (40/40). The tumors were clearly limited from adjacent tissue. At the end of treatment tumors were removed from mice to determine organ weights. Relative tumor weight was reduced to 61.8% by treatment with the V2 (157.2 ± 115.3 mg, n = 10), V75 (215.8 ± 129.1 mg, n = 10) compared with saline-treated control animals (411.5 ± 112.7 mg, n = 10) (Table 1). 5-Fu treatment resulted in significant reduction of relative tumor weight (12.3 ± 5.9 mg, n = 10) with 100% treatment response (p < 0.001). But weights of mice in 5-Fu group were significantly decreased compared with saline control group. And the weights of mice were not significantly different between the saline control and mAbs groups (p > 0.05, Table 3).

The growth of T24 and Ls-174-T cells cultured in the presence of 100 lg/ml of mAbs was measured after 1, 2, and 3 days. The results presented in Fig. 5 and Fig. 6 demonstrated that V2 and V75 reduced the number of living cells significantly compared with the control group (p < 0.05). 3.6. Inhibition of tumor growth in tumor bearing nude mice To evaluate the potency of the fully human anti-VEGF mAbs in vivo, we developed a tumor model in male nude mice with Ls-174-T cells. The Ls-174-T cell line was chosen because it is more sensitive to our mAbs than the T24 cell based on the results described above. The tumor implantation rate was 100% (40/40). According to the autopsy and micro-

mAb

V2

V50

V71

V75

A490 nm

0.033 ± 0.012

0.029 ± 0.010

0.042 ± 0.020

0.027 ± 0.016

3.7. Pharmacokinetics in the SD rats Because the low cross-reaction and the high effect of V75 was observed from experiments above, pharmacokinetics studies of human anti-VEGF mAb were carried out using V75 in SD rats. SD rats (five females and five males per group) received a single intravenous injection of V75 at 3.5, 7.0, and 14.0 mg/kg (Table 4). The elimination profiles for 15 days after injection are shown in Fig. 7. No significant difference was observed in the profiles between male and female rats. The average serum V75 concentration at each time point was calculated and analyzed using a three-compartment model. As shown in Table 3, the tissue distribution half-life for the a phase was 0.4 ± 0.1, 0.3 ± 0.1, and 0.7 ± 0.9 h for 14.0, 7.0, and 3.5 mg/kg treated group, respectively. The tissue distribution half-life for the b phase was 7.3 ± 2.2, 5.9 ± 3.3, and 16.1 ± 24.1 h and the terminal elimination half-life for the c phase was 135.6 ± 12.6, 203.4 ± 41.9, and 156.2 ± 23.2 h, respectively.

4. Discussion

Fig. 5. Effect of human mAbs on T24 cell proliferation (mean ± SD) n = 3. Each column represents the mean of three individual experiments. * Significant decrease in mAbs on the second and the third day (p < 0.05).

Currently, angiogenesis is considered as one of major factors that affect tumor growth. Anti-angiogenesis therapy has appeared as another promising method for tumor treatment. VEGF was first discovered by Ferrara in 1989, and it was shown that it can stimulate proliferation of vas-

Fig. 6. Effect of human mAbs on Ls-174-T cell proliferation (mean ± SD) n = 3. Each column represents the mean of three individual experiments. The open bars represent the 1 day group, and the black bars represent the 2 days group. *Significant decrease in mAbs after cultured one and two days (p < 0.05).

Fig. 7. Pharmacokinetics profiles of V75 in SD rats. Serum concentrations of human anti-VEGF mAb V75 in SD rats were measured after 3.5, 7.0, and 14.0 mg/kg body weight single-dose intravenous bolus injection. Blood was drawn at appropriate time points, and antibody concentrations in sera were measured using gamma counter. Triangles and squares represent 14.0 and 7.0 mg/kg groups, respectively, and circles represent the 3.5 mg/kg group.

33

Y. Liu et al. / Cancer Letters 273 (2009) 28–34

Table 3 The tumor growth inhibition rate (IR) of mAbs on Ls-174-T transplant tumor tissues and the weight of nude mice at the beginning and end of experiment (mean ± SD) (n = 10) Groups

Dose

Saline control V2 V75 5-Fu **

p < 0.05,

***

0.2 ml 10 mg/kg 10 mg/kg 100 mg/kg

Weight of mice (g) Beginning

End

18.4 ± 1.9 18.9 ± 1.2 18.1 ± 1.3 18.7 ± 0.8

25.4 ± 2.9 24.8 ± 3.2 23.7 ± 3.5 13.4 ± 0.9**

Weight of tumors (mg)

IR of tumor weight (%)

411.5 ± 112.7 157.2 ± 115.3** 215.8 ± 129.1** 12.3 ± 5.9***

/ 61.8 47.6 100

p < 0.01 (vs saline control).

Table 4 Pharmacokinetics of human anti-VEGF antibody in SD rats Dose (mg/kg)

T1/2a (h)

T1/2b (h)

T1/2c (h)

AUC (mg/L/h)

Cmax (mg/L)

CL (L/h/kg)

14.0 7.0 3.5

0.4 ± 0.1 0.3 ± 0.1 0.7 ± 0.9

7.3 ± 2.2 5.9 ± 3.3 16.1 ± 24.1

135.6 ± 12.6 203.4 ± 41.9 156.2 ± 23.2

908.95 ± 116.01 483.74 ± 76.60 279.25 ± 37.95

78.79 ± 8.30 30.46 ± 8.32 15.72 ± 3.61

0.014 ± 0.002 0.011 ± 0.001 0.010 ± 0.002

Abbreviations: T1/2 = half life; AUC = area under the curve; Cmax = the maximum concentration; CL = plasma clearance.

cular endothelial cell and induce vascular leakage, and therefore it was named vascular endothelia growth factor (VEGF). Four isoforms, i.e., VEGF206, 189, 165, and 121, have been identified. It has been confirmed that VEGF165 was found in many human tumors, suggesting that most solid tumors produce and secrete VEGF165 for tumor angiogenesis. Therefore VEGF165 was chosen as a specific antigen in this study. VEGF has become an efficient target in anti-tumors research now. A successful example is AVASTIN, a humanized anti-VEGF monoclonal antibody approved by FDA in 2004 in clinical tumor therapy. V2, V50, V71, and V75 produced in this study are fully human anti-VEGF monoclonal antibody which functioned as a tumor anti-angiogenic agent. We hypothesized that the fully human IgM might achieve strong therapeutic anti-tumor effect. Immunoneutralizing mAb is an inhibitor and many anti-VEGF mouse mAbs were reported to have strong inhibitory activity on the growth of human tumors. But because of their HAMA reaction, mouse mAbs have limited use in tumor therapy. At present, many mouse mAbs were humanized by gene engineering technology to solve this problem. With the development of transgenic technology, mice carrying transloci bearing human Ig genes in germline configuration and endogenous mouse Ig loci being silenced by gene-targeting techniques have been achieved. Antigen-specific fully human mAbs can be produced by routine hybridoma production using transgenic mice. Several mouse strains harboring human Ig loci have been generated over the last few years as an alternative strategy for the generation of human mAb of therapeutic value. In this report, we present our first successful generation of fully human anti-VEGF165 mAb using five-feature translocus mice. Four anti-VEGF IgM mAbs (V2, V50, V71, and V75) were isolated which can secrete IgM stably. These were fully human Ig and could interact with the intact hVEGF165. All the four human mAbs strongly recognized the intact hVEGF165 and was capable of inhibiting hVEGF165-induced HMVEC proliferation, suggesting that they could neutralize the hVEGF165 in medium. Although IgM antibodies are gen-

erally considered as poor inducers of antigenic modulation, such an effect has been previously reported [13,14]. The heavy chain and light chain isotypes of the isolated IgM were characterized by ELISA. All the mAbs we generated were found to have a human l heavy chain and a human k chain. None of them has mouse Ig(H+L) (Data not shown). These results suggested that all those mAbs were fully human anti-VEGF mAbs. This study also discovered that the human IgM, and several murine antibody isotypes (such as murine IgG) were found in the serum of five-feature mice using ELISA (data not shown). It can be explained by the fact that the endogenous Igk locus was not disrupted [15]. In conclusion, four fully human mAb specific for the native hVEGF165 expressed by P. Pastoris (V2, V50, V71, and V75) were produced from the five-feature transloci mice, and had strong binding ability. To determine the biological activities of the fully human anti-VEGF mAbs, V2, V50, V71, and V75 antibodies were selected and tested in this study. In addition to blocking hVEGF165 binding to its receptor VEGFR2, four mAbs were capable of inhibiting T24 cell proliferation. Except for V50, the other three mAbs inhibited the proliferation of Ls-174T cells significantly. Both T24 and Ls-174-T cell line can secrete hVEGF165 to stimulate its proliferation, so the inhibition effect could be observed in vitro. This anti-tumor effect was further verified in an in vivo transplantation human tumor model in nude mice. Nude mice have been used as a standard method to establish rodent models of human cancer [16]. Systemic administration of V2 and V75 led to significant suppression of human Ls-174-T transplanted tumor model in nude mice (Table 2). The ability of antiVEGF mAbs to decrease the tumor weight suggests that it is feasible to treat cancer by systemic administration. V2 at a dose of 10 mg/kg achieved the same effect as V75 at the same dose. Both mAbs inhibited the tumor growth in nude mice significantly. Although the inhibition rate was not as high as the 5-Fu group, mAbs were found to be more tolerated and safer than 5-Fu. While the animal weight of 5-Fu group decreased significantly and some of them were dead during or after the experiment, the general condition

34

Y. Liu et al. / Cancer Letters 273 (2009) 28–34

of mAb group animals has no difference with the saline control group. These data implied that V2 and V75 might be effective and safe on tumor therapy (see Table 3). The pharmacokinetics analysis of V75 is conducted to determine the pharmacokinetics of a fully human mAb derived from five-feature transloci mice in rats. Since V75 is a human anti-VEGF IgM antibody, it should have a half life in humans similar to the endogenous IgM (5 days). The results showed that the half life of V75 was similar as control endogenous human IgM. It confirmed in another way that V75 is a fully human IgM antibody. Although the half life is much shorter than IgG monoclonal antibody (20 days), it was found that V75 can target the gastric tissue (data not shown). The discovery of the targeting fragments will benefit the potential therapeutic effect of this mAb on gastric cancer in future. Acknowledgements This work was funded by an NSFC (National Natural Science Founding of China). We are grateful for the expertise and advice of Wu Hailong, who helped to make the antibodies. We also thank Cai Yinfeng, Fan Jianhua for antibody production studies.

[5]

[6]

[7] [8]

[9]

[10]

[11]

[12]

[13]

References [14] [1] Donald Ingber, Takeshi Fujita, Shoji Kishimoto, et al, Synthetic analogues of fumagillin that inhibit angiogenesis and suppress tumor growth, Nature 348 (1990) 555–557. [2] N. Weidner, J.P. Semple, W.R. Welch, J. Folkman, Tumor angiogenesis and metastasis correlation in invasive breast carcinoma, Engl. J. Med. 324 (1991). [3] K.J. Kim, B. Li, J. Winer, et al, Inhibition of vascular endothelial growth factor-induced angiogenesis suppresses tumor growth in vivo, Nature (London) 362 (1993) 841–844. [4] Fairooz F. Kabbinavar, Julie Hambleton, Robert D. Mass, Herbert I. Hurwitz, Emily Bergsland, Somnath Sarkar, Combined analysis of efficacy: the addition of Bevacizumab to fluorouracil/leucovorin

[15]

[16]

improves survival for patients with metastatic colorectal cancer, J. Clin. Oncol. 23 (16) (2005) 3706–3712. I.C. Nicholson, X. Zou, A.V. Popov, G.P. Cook, E.M. Corps, S. Humphries, C. Ayling, B. Goyenechea, J. Xian, M.J. Taussig, M.S. Neuberger, M. Bruggemann, Antibody repertoires of four- and fivefeature translocus mice carrying human immunoglobulin heavy chain and happa and lambda light chain yeast artificial chromosomes, J. Immunol. 163 (1999) 6898–6906. Yu Liu, Guo-Xiang Wu, Jian-yang Zhao, Tian-Hua Yan, Zi-Long Shen, Guo-Xiang Chen, Pichia expression of recombined human VEGF(165) and activity determination, J. China Pharm. Univ. 35 (1) (2004) 82. G. Galfre, C. Milstein, Preparation of monoclonal antibodies: strategies and procedures, Methods Enzymol. 73 (1981) 3–46. Ramadan A. Abuknesha, Connie Y. Luk, Hannah H.M. Griffith, Alexandra Maragkou, Despina Iakovaki, Efficient labelling of antibodies with horseradish peroxidase using cyanuric chloride, J. Immunol. Methods 306 (1–2) (2005) 211–217. B. Friguet, A.F. Chaffotte, L. Djavadi-Ohaniance, et al, Measurement of the true affinity constant in solution of antigen-antibody complexes by enzyme-linked immunosorbent assay, J. Immunol. Methods 77 (1985) 305–319. P.R.P. Salacinski, C. McLean, J.E. Sykes, W. Clement-Jones, P.J. Lowry, Iodination of proteins, glycoproteins, and peptides using iodogen, Anal. Biochem. 117 (1981) 136–140. T. Lindmo, E. Boven, F. Cuttitta, J. Fedorko, P.A. Bunn Jr., Determination of the immunoreactive fraction of radiolabeled monoclonal antibodies by linear extrapolation to binding at infinite antigen excess, J. Immunol. Med. 72 (1984) 77–89. Michelle D. Tallquist, Philippe Soriano, Richard A. Klinghoffer, Growth factor signaling pathways in vascular development, Nature 18 (55) (1999) 7917–7932. B. Collet, A. Martin, G. Merdrignac, B. Genetet, L. Toujas, Modulation of expression of mouse macrophage surface antigens by monoclonal antibodies, Immunology 56 (1985) 543–548. K. Sugita, O. Majdic, H. Stockinger, W. Holter, R. Bruger, W. Knapp, Induction of human complement activation without cytolysis by mouse monoclonal antibodies to human leukocyte antigens, Transplantation 43 (1987) 570–574. S. Magadan, M. Valladares, E. Suarez, I. Sanjuan, A. Molina, C. Ayling, S.L. Davies, X. Zou, G.T. Williams, M.S. Neuberger, M. Bruggemann, F. Gambon, F. Diaz-Espada, A. Gonzalez-Fernandez, Production of antigen-specific human monoclonal antibodies: comparison of mice carrying IgH/Kappa or IgH/Kappa/lamda transloci, Biotechniques 33 (2002) 680. 682, 684 passi. C. Manzotti, R.A. Audisio, G. Pratesi, Importance of orthotopic implantation for human tumors as model systems: relevance to metastasis and invasion, Clin Exp Metastasis 11 (1993) 5–14.