Plasmalemmal Vesicle Associated Protein (PLVAP) as a therapeutic target for treatment of hepatocellular carcinoma

Abstract

Background: Hepatocellular carcinoma (HCC) is a malignancy with poor survival outcome. New treatment options for the disease are needed. In this study, we identified and evaluated tumor vascular PLVAP as a therapeutic target for treatment of HCC.

Methods: Genes showing extreme differential expression between paired human HCC and adjacent non-tumorous liver tissue were investigated. PLVAP was identified as one of such genes with potential to serve as a therapeutic target for treatment of HCC. A recombinant monoclonal anti-PLVAP Fab fragment co-expressing extracellular domain of human tissue factor (TF) was developed. The potential therapeutic effect and toxicity to treat HCC were studied using a Hep3B HCC xenograft model in SCID mice.

Results: PLVAP was identified as a gene specifically expressed in vascular endothelial cells of HCC but not in non-tumorous liver tissues. This finding was confirmed by RT-PCR analysis of micro-dissected cells and immunohistochemical staining of tissue sections. Infusion of recombinant monoclonal anti-PLVAP Fab-TF into the main tumor feeding artery induced tumor vascular thrombosis and extensive tumor necrosis at doses between 2.5 μg and 12 μg. Tumor growth was suppressed for 40 days after a single treatment. Systemic administration did not induce tumor necrosis. Little systemic toxicity was noted for this therapeutic agent.

Conclusions: The results of this study suggest that anti-PLVAP Fab-TF may be used to treat HCC cases for which transcatheter arterial chemoembolization (TACE) is currently used and potentially avoid the drawback of high viscosity of chemoembolic emulsion for TACE to improve therapeutic outcome. Anti-PLVAP Fab-TF may become a viable therapeutic agent in patients with advanced disease and compromised liver function.

Figures

Figure 1

Changes of tumor blood flow and tumor histology at 2, 4, 24, 48 and 72 hours after treatment with 10 μg MECA32-Fab-TF. Tumor blood flow was monitored using 3D power Doppler sonography (upper panel). The histology sections were stained with hematoxylin and eosin (lower panel). There were two mice at each time point. Results were the same between two mice at each time point. Only result from one of the two mice studied at each time point is shown. Upper panel shows change of tumor blood flow before and after treatment. White arrows point at blood flow signal in tumors. Blood flow signal disappeared at 2 hours and persisted up to 72 hours. Lower panel shows that fibrin thrombi (balck arrows) in blood vessels became evident at 2 hours after treatment and persisted throughout the study period. Tumor tissue became morphologically degenerated at 24 hours. Frank necrosis became evident at 48 hours. Photomicrographs were taken at 100x magnification.

Figure 2

Blood supply and tumor growth in Hep3B tumor xenografts after intra-arterial infusion of 20 μg MECA32 mAb chemically conjugated with human tissue factor (MECA32-TF) into a tumor feeding femoral artery. Control mice were infused with 20 μg MECA32 mAB. A: Power Doppler was performed 48 hours before and after the treatment. Red signals in tumors represent blood flow, which were significantly diminished in mice after treating with MECA32-TF (white arrow) but not in those treated with control MECA32 mAb. B: Tumor growth before and after treatment. Solid circles (•) are control mice and crosses (x) are mice treated with MECA32-TF. †: Death.

Figure 3

Differential expression of PLVAP between paired HCC tissue and adjacent non-tumorous liver tissue. A: Differential expression of the PLVAP gene according to microarrays of 18 pairs of HCC and adjacent non-tumorous liver tissue. PN: paired non-tumorous liver; PHCC: paired HCC tissue. B: Relative quantities of PLVAP mRNA in the same 18 tissue pairs. One non-tumorous liver tissue sample was chosen as a reference control (relative quantitative expression = 1). C: Immunohistochemical (IHC) staining of PLVAP in four randomly selected HCC cases. IHC staining was performed using GY5 murine anti-human PLVAP monoclonal antibody. Endothelial cells lining blood vessels of HCC showed positive staining for PLVAP in brown color (arrows). IHC staining (panel C) showed that PLVAP was not expressed by the endothelial cells of hepatic central vein (C-II right panel), hepatic sinusoid (CI-IV right panels), and hepatic arterioles (portal tract) (C-III right panel) in the adjacent non-tumorouse liver tissues. The large empty space in the right panel of C-II was lumen of a hepatic central vein which showed absence of PLVAP expression in the lining endothelial cells. We also stained HCC sections including adjacent non-tumorous liver with anti-human CD34 monoclonal antibody. Endothelial cells of hepatic central vein and hepatic areteriole were stained positively for CD34 expression (data not shown). Liver sinusoidal endothelial cells did not express CD34 as expected.

Figure 4

Tumor necrosis 72 hours after infusion of different doses of MECA32-Fab-TF. The results of two different studies are shown here. The largest tumor cross sections were submitted for histology and studied. Necrotic tumors and viable residual tumors were outlined as areas of pink and blue, respectively. The relative size of necrotic and viable tumor tissue was measured based on two dimensional areas. Percentages shown in the figure represent relative necrotic area in tumor sections. In study I, all three control tumors at right showed no necrosis (0%). In study II, photomicrographs of residual viable tumor and adjacent necrotic tumor tissue are shown at a higher magnification of 12.5x on the right. A 40x magnification to show few layers of residual viable tumor cells is shown in the inset.

Figure 5

Tumor growth after infusing MECA32-Fab-TF or control MECA32 mAb into a tumor feeding artery. The results of two different studies were shown here. In study A, tumor bearing mice were treated with 5 or 10 μg MECA32-Fab-TF or 10 μg MECA32 mAb. All mice were euthanized 24 days after treatment. The growth rates between the treatment groups and the control group were compared using a linear mixed-effects model. Significant differences in tumor growth between controls and 5 μg or 10 μg treatment groups were noted (p = 0.003 and 0.001). In study B, tumor bearing mice were treated with 10 μg MECA32-Fab-TF (n = 4) or control MECA32 mAb (n = 2). Mice were sacrificed when tumors grew large enough to interfere with movement and food intake. The average numbers of days required to reach a tumor size of 1600 mm3 for control and treatment groups were 9.8 and 51.8 days, respectively. Different rates of tumor growth were noted between experiments and between mice within the same experiments. Therefore, effort was made to match tumor sizes between control and treatment groups in each study.

Pre-publication history