Anti-NRP-1 mAb with a high purity was produced using the hybridoma technique and preserved in the laboratory after freeze-drying (29). 2-(N-morpholino) ethane sulfonic acid (MES), 1-ethyl-3-(3-dimethylaminopropyl) carbodi-imide (EDC), N-hydroxysulfosuccinimide (sulfo-NHS), streptavidin (SA), Sephadex G200 and a Sephadex G200 column (1.5×22 cm) were purchased from Sigma-Aldrich (Merck KGaA). GoldBand 3-color Regular Range Protein Marker (10-180 kDa) was purchased from Shanghai Yeasen Biotechnology Co., Ltd. Endothelial cell medium (ECM) containing human epidermal growth factor was purchased from Shanghai Zhong Qiao Xin Zhou Biotechnology Co., Ltd. DMEM and FBS were purchased from Invitrogen (Thermo Fisher Scientific, Inc.). Mouse IgG isotype (cat. no. 0107-01) was obtained from AmyJet Scientific Inc. Rhodamine B isothiocyanate (RBITC)-conjugated goat anti-mouse IgG (cat. no. D111097), isopropyl-l-thio-B-d-galactopyranoside (IPTG), BSA, Hoechst 33258 powder, and hematoxylin and eosin (H&E) were purchased from Sangon Biotech Co., Ltd. Factors X and VII were obtained from Sigma-Aldrich (Merck KGaA). pET22b (+) plasmid and E. coli BL21(DE3) were purchased from Novagen (Merck KGaA). cDNA encoding for tTF, containing amino acids 1 to 218 of human TF, was used to generate the tTF expression vector tTF-pET22b (+) via PCR using the following primers: Forward, 5′-TCC ATG GGC TCT GGC ACT ACA-3′ and reverse, 5′-GTG CTC GAG TTC TCT GAA TTC C-3′. NcoI and XhoI restriction enzymes were used to insert cDNA into the plasmid. Nickel-nitrilotriacetic acid (Ni-NTA) agarose was purchased from Qiagen, Inc. A B-labeling kit was obtained from Wuhan Elabscience Biotechnology Co., Ltd. Cyanine-5 (Cy5) NHS-ester was purchased from Tiangen Biotech Co., Ltd.

The purity of anti-NRP-1 mAb was first identified via 12% SDS-PAGE. As presented in Fig. 1, the mAb-SA conjugate was synthesized using a coupling method (30). The concentration of mAb was adjusted to 3 mg/ml with reaction buffer (0.1 mol/l MES; 0.5 mol/l NaCl; pH 6.0) according to the improved Bradford method. A total of 3.83 mg (0.02 mmol) of EDC and 4.34 mg (0.02 mmol) of sulfo-NHS were weighed and immediately transferred to the reaction solution (1 ml). The solution was mixed and stirred at room temperature for 15 min. SA concentration was adjusted to 1.2 mg/ml with 0.1 mol/l potassium phosphate buffer (pH 7.5), and 1 ml SA was added to the reaction buffer. Nitrogen gas was purged into the solution for 3 min and the beaker was then sealed. The reaction was subsequently left to proceed at room temperature for 2 h, then purified using a Sephadex G200 column as previously described (31,32), with some modifications. Briefly, Sephadex G200 was swelled in a boiling water bath for 2 h and cooled. Pretreated Sephadex G200 was poured into the column slowly, with no air mixing in, and equilibrated with pure water (pH 7.5) for 40 min. The crude product was then slowly added to the column and eluted with pure water at a rate of 0.2 ml/min. Eluted samples (1.5 ml/tube) were collected and read with a NanoDrop™ 2000 spectrophotometer (NanoDrop Technologies; Thermo Fisher Scientific, Inc.) at 280 nm. All purification steps were conducted at 4°C. Native 8% PAGE was performed to confirm whether conjugated protein had been successfully isolated (33).

HUVECs and the human liver cancer cell line HepG2, which both overexpress NRP-1 (34,35), were obtained from the American Type Culture Collection. HUVECs were used for in vitro experiments, and HepG2 cells were used to establish a mouse tumor model. HepG2 cells were cultured in high-glucose DMEM supplemented with 1% penicillin- streptomycin and 10% FBS. HUVECs were cultured in ECM containing human epidermal growth factor. All cells were incubated in a humidified atmosphere with 5% CO₂ at 37°C.

HUVECs (1×10⁵ cells/ml) were seeded into a 6-well culture plate with one glass cover-slip per well. The cells were then incubated and subsequently washed with PBS (pH 7.4) three times, until 50% cloning efficiency had been reached. The cells were then fixed with 1 ml 4% paraformaldehyde at 4°C for 30 min and washed with PBS three times. A total of 2 ml PBS containing mAb, mAb-SA or a mouse IgG isotype control (1:5,000) was added, and the cells were cultured at 37°C for 1 h. After washing, cells were cultured with a goat anti-mouse RBITC mAb (1:200) at 37°C for an additional 1 h in the dark. Hoechst 33258 was used to stain cell nuclei at 37°C for 5 min and the samples were then examined under a FV1000MPE-B confocal microscope (Olympus Corporation) and photographed. Five random fields per sample were analyzed (magnification, ×600).

Semiquantitative analysis was conducted to further assess the ability of the mAb-SA conjugate to target NRP-1. HUVECs were removed from the culture plate using trypsin and washed with PBS three times. The cells were then fixed with 1 ml 4% paraformaldehyde at 4°C for 30 min and washed with PBS three times. After being resuspended in PBS, cells were incubated with mAb, mAb-SA or the mouse IgG isotype (1:5,000) control at 37°C for 1 h and incubated with the anti-mouse RBITC mAb (1:200) at 37°C for a further 1 h. Each sample of 10,000 cells was analyzed using a CytoFlex S flow cytometer (Beckman Coulter, Inc.). Results were analyzed using CytExpert version 2.0 (Beckman Coulter, Inc.).

E. coli (BL21; DE3) containing tTF expression vector were cultured in Luria broth supplemented with 1% ampicillin. IPTG was added when the germiculture reached 0.6-0.8 at 600 nm to induce the expression of the fusion protein tTF. After stimulation for 6 h, bacterial cells were collected and centrifuged at 12,000 × g at 4°C for 20 min. A total of 5 ml lysis buffer (20 mmol/l Tris/HCl; pH 8.0; 0.5 mol/l NaCl; 2 mol/l urea; 20 ml/l Triton X-100) per gram (wet weight) was subsequently added. After incubating for 90 min, cells were centrifuged at 12,000 × g for 20 min at 4°C. The pellet was resuspended and sonicated (sonication time: 5 sec; interval time: 5 sec) in washing buffer (20 mmol/l Tris/HCl; pH 8.0; 0.5 mol/l NaCl; 2 mol/l urea; 20 ml/l Triton X-100) at 4°C for 30 min. A total of 5 ml solubilization buffer (20 mmol/l Tris/HCl; pH 8.0; 8 mol/l urea; 1 mmol/l β-mercaptoethanol; 20 ml/l Triton X-100) per gram (wet weight) was added to dissolve the inclusion bodies. After incubation at room temperature overnight, the suspension was centrifuged at 12,000 × g for 20 min at 4°C. The supernatant was then purified with a Ni-NTA column according to the protocol of the His-Bind Buffer kit (Novagen; Merck KGaA). The products were analyzed using 12% SDS-PAGE under denaturing conditions. The purified fusion protein tTF was then concentrated using a Centrifugal Filter and freeze-dried for subsequent use.

tTF-B was prepared using a B-labeling kit according to the manufacturer's protocol. A factor X activation assay was performed to assess the coagulation activity of tTF-B, as previously described (36). Various concentrations (0.01-10 µmol/l) of BSA, tTF or tTF-B were incubated with 100 nmol/l factor VII in Tris-buffered saline at 37°C for 10 min. Factor X (5 nmol/l) was then added, and the mixture was incubated for 10 min at room temperature. The reaction was subsequently quenched using 100 mmol/l EDTA. A total of 2 nmol/l Spectrozyme FXa was added to the mixture, and the absorbance was detected at 405 nm after 3 min.

A competitive ELISA was performed to determine the B/SA-binding capacity of mAb-SA/tTF-B. A 96-well plate was coated with tTF-B (100 µl/well; 20 µg/ml) in coating buffer (7.5 mM sodium carbonate; 17.4 mM sodium bicarbonate; pH 9.6) and incubated overnight at 4°C. The plate was then washed three times with PBS-Tween solution (0.01 mol/l, pH 7.4, 0.05% Tween-20), and 10% FBS was then used to block the unbound sites at 37°C for 1 h. After washing the plate five times, HRP-labeled SA (100 µl/well; 200 µg/ml) and was added to each well to serve as the competitor. Serially diluted mAb-SA or SA (100 µl/well; 200-12.5 µg/ml) was then added separately. After incubation for 1 h at 37°C, the enzyme substrate orthophenylenediamine, in 50 mM phosphate-citrate buffer (pH 5.0 with fresh 30% hydrogen peroxide) and at a concentration of 0.4 mg/ml, was added (100 µl/well). The plate was subsequently incubated at room temperature for 20 min, and 1 M H₂SO₄ was added to quench the reaction (50 µl/well). Absorbance values were measured at 490 nm. The inhibition rate was calculated according to a previous study (37) to determine the immunoreactivity of mAb-SA.

Female BALB/c nude mice (n=60, 6-8 weeks old, 18-22 g) were purchased from the Experimental Animal Center of Xiamen University, of which 52 were ultimately used for experiments. HepG2 cells were dissociated in a culture plate using 10% trypsin at 37°C for 3 min, and subsequently resuspended in PBS. Each mouse was then subcutaneously injected in the right flank with 100 µl HepG2 cells (1×10⁶). Animals were housed in the Experimental Animal Center under specific pathogen-free conditions: Temperature, 22-25°C; humidity, 50-60%; 12-h light/dark cycle. Mice had free access to water and Purina 5L79 rodent chow. Animal health and behavior were monitored daily. Animal welfare was in accordance with institutional guidelines. When the mean tumor volume [calculated as (length x width²)/2)] reached 150-250 mm³, those mice were used for live imaging, antitumor activity studies and histological analysis after being acclimatized. The largest subcutaneous tumor observed in the present study was 1.9 cm in diameter at its widest point, and no mice exhibited multiple subcutaneous tumors. According to a previous study (38), the humane endpoints were determined based on the mouse weight loss (>20% of total body weight) or mouse activity assessment (hunching, stationary, ruffling and poor grooming), and mice were euthanized via cervical dislocation and dissected. The duration of the animal experiment was ~25 days. Then, all remaining mice were euthanized by cervical dislocation, and no mice were found dead. All animal experiments performed in the present study were approved by the Ethics Committee of Xiamen University.

SA, anti-NRP-1 mAb and mAb-SA were labeled with fluorescein. The concentration of Cy5 NHS-ester used was 1 mg/ml in DMSO. According to the total amount of proteins (SA, anti-NRP-1 mAb or mAb-SA), 0.01 mg fluorescein per mg protein were mixed at room temperature for 2 h. Dialysis was performed to remove unlabeled Cy5 NHS-ester for 4 h at room temperature and subsequently at 4°C overnight with 0.15 mol/l NaCl solution. The whole process was performed in the dark.

To investigate the in vivo distribution of mAb-SA, nude mice with subcutaneous xenografts were randomly divided into four groups (n=3/group). Mice in each group were intravenously injected with 100 µl saline, SA-Cy5, anti-NRP-1 mAb-Cy5, or mAb-SA-Cy5 via the tail vein. After the mice were anesthetized in 1.5% isoflurane, the Cy5 fluoro-chrome was then detected in mice using an Imaging IVIS-200 system (PerkinElmer, Inc.) at 0.5, 3, 6, 12, 24, 36, 48 and 72 h. The nude mice were sacrificed after imaging. Tumor tissues and major organs, including the heart, liver, spleen, lungs, kidneys and brain, were isolated from the mAb-SA-Cy5 treatment group, and their fluorescent signal intensity was measured using Living Image version 4.3 (Caliper Life Sciences, Inc.; PerkinElmer, Inc.).

Nude mice bearing 150±50 mm³ HepG2 tumors were randomly assigned to four groups (n=10/group). Mice in each group received four intravenous administrations of saline, mAb-SA, tTF-B and mAb-SA:tTF-B, at 3-day intervals, at a dose of 5 mg/kg. In the mAb-SA:tTF-B treatment group, tTF-B was injected 24 h after each administration of mAb-SA. Tumor size was measured every second day. Tumor tissues were surgically removed and weighed after 12 days of treatment.

To evaluate the tumor vascular targeting ability of the mAb-SA:tTF-B composite system and its effects on intravascular thrombosis in vivo, tumors from different treatment groups and major organs from the mAb-SA:tTF-B group were fixed in 4% formaldehyde solution at room temperature for 48 h before embedding in paraffin and cutting into 5-7-µm sections for Harris H&E staining. Tissues were stained with hematoxylin and eosin at room temperature for 3 min and 20 min, respectively. Five random fields per sample were examined under a light microscope (magnification, ×200).

All data in the present study were analyzed using Prism 6.0 (GraphPad Software, Inc.) and presented as the mean ± standard deviation. One-way ANOVA with Tukey's multiple comparison post hoc test was performed to assess the differences among groups. P<0.05 was considered to indicate statistically significant differences.

As presented in Fig. 2A, the purity of anti-NRP-1 mAb reached ~95%. To separate the desired synthetic product from any impurities, including the cross-linked byproduct, any unreacted antibodies and the unreacted SA, the crude products were purified using a Sephadex G200 gel-filtration column, which was washed using pure water (pH 7.5). To avoid dissociation to monomeric species, the conjugate was analyzed using native 8% PAGE under a non-boiled condition. The bands shown in Fig. 2B correspond to (from left to right) unreacted SA, unreacted mAb and conjugate products after purification. The final purified conjugate was well-defined and clear.

Confocal immunofluorescence and flow cytometry were performed to determine whether mAb-SA maintained biological targeting activity. As presented in Fig. 3A, red fluorescence of RBITC was observed on the surface membrane of HUVECs treated with mAb or mAb-SA. This result indicated that mAb-SA was able to bind with NRP-1 on the cell surface. The results of flow cytometry (Fig. 3B) were similar to those of the confocal experiment. The anti-NRP-1 mAb and mAb-SA groups exhibited significant HUVEC-binding activity. The NRP-1-binding capacity of mAb-SA was almost equivalent to that of mAb.

To optimize the culture conditions, various concentrations (0-1 mmol/l) of IPTG were used. 12% SDS-PAGE was performed to evaluate the induced fusion protein yield (Fig. 4). A maximum yield was observed when the IPTG concentration reached 0.8 mmol/l. The fusion protein was expressed in the form of inclusion bodies. The inclusion bodies were washed with washing buffer and subsequently dissolved in denaturation buffer. A nickel affinity column was used to purify the soluble products. As presented in Fig. 4, the objective fusion protein was ~34 kDa with ~99% purity.

A factor X activation assay was performed to evaluate the retention of blood coagulation activity in tTF-B. As presented in Fig. 5, tTF and tTF-B exhibited similar factor X activity at the corresponding concentrations, indicating that tTF-B retained the tTF moiety when inducing blood coagulation.

A competitive ELISA was used to detect the in vitro binding ability of tTF-B and mAb-SA (Fig. 6). The results revealed that the EDC reaction did not impair the B-binding ability of SA. The tTF-B-binding capacity of the mAb-SA conjugate was 73% compared with that of unmodified SA. Therefore, the in vitro use of the mAb-SA:tTF-B system may also be of value for in vivo studies.

Live imaging was performed to assess the distribution of mAb-SA in tumor-bearing mice. A variety of Cy5-labeled reagents and saline (blank control) were intravenously injected, and the fluorescent signals were detected at various time points. As presented in Fig. 7A, Cy5-labeled mAb and mAb-SA were detected in the tumor area at 3 h following intravenous injection. The peaks in fluorescent signals were similar at 24 h; however, no SA-Cy5 fluorescent signal was observed in tumor tissues. The fluorescent signals in the mAb and mAb-SA-Cy5 treatment groups remained strong at 72 h (Fig. 7B and C). Weak Cy5 fluorescence signals were detected in the major organs; the fluorescence intensity ratio of tumor to non-tumor tissue was ~8.83-29.80 in mice injected with mAb-SA-Cy5. The results of the present study suggested that the mAb-SA conjugate was able to selectively target tumor tissues, but was not retained by normal organs.

The anti-tumor activity of the mAb-SA:tTF-B system was evaluated in HepG2 xenograft-bearing nude mice. Tumor growth curves are presented in Fig. 8A. Overall, tumors in mice treated with mAb-SA:tTF-B exhibited decreased growth compared with other treatment groups. Tumor regression was also only observed following the third treatment on day 6. This indicated that the two-step coagulation approach was able to effectively inhibit the growth of HepG2 xenografts and may result in tumor regression. In addition, the results revealed that mAb-SA alone could markedly affect tumor growth rate (mAb-SA vs. saline, P<0.05). These results are supported by the final tumor weights (Fig. 8B) of mice in different treatment groups after treatment for 12 days.

H&E staining was performed to observe histological changes in the tumor tissues of mice from the different treatment groups. Extensive thrombosis (estimated at ~80%) in blood vessels and necrosis were observed in the tumor tissues of mice treated with mAb-SA:tTF-B (Fig. 9A). In contrast, decreased intravascular embolization was exhibited in the other treatment groups. These results confirmed that the mAb-SA conjugate retained tumor vasculature-targeting ability and induced intravascular thrombosis in vivo by directing tTF-B to the vascular endothelial cell surface, thereby causing inhibition of tumor growth and promoting tumor regression. Major organs from the mAb-SA:tTF-B treatment group did not exhibit any visible ectopic embolism or overt tissue injury (Fig. 9B).