Anti-caveolin-1 antibodies as anti-prostate cancer therapeutics

Abstract

Caveolae are critical cell surface structures important in coordinated cell signaling and endocytosis. One of the major proteins of caveolae is caveolin 1 (Cav-1). Cellular levels of Cav-1 are associated with cancer progression. In prostate cancer cells, levels of Cav-1 are positively correlated with tumor progression and metastasis. Cav-1 can be secreted by prostate cancer cells into the microenvironment and triggers proliferation and anti-apoptosis of the tumor and tumor endothelial cells. Clinical studies have shown increased serum Cav-1 levels in patients with poor prognosis. In tissue culture and animal model experiments, blocking secreted Cav-1 by polyclonal antibodies inhibits tumor cell growth. Cav-1 is therefore a potential therapeutic target for prostate cancer treatment. In this study, we used Cav-1 knock-out mice as hosts to produce monoclonal anti-Cav-1 antibodies. A total of 11 hybridoma cell lines were selected for their ability to produce antibodies that bound GST-Cav-1 but not GST on glutathione-coated ELISA plates. Further screening with ELISAs using GST-Cav-1 fragments on GSH-coated plates classified these antibodies into four groups: N1-31 with five antibodies binds the far N-terminus between amino acids 1 and 31; N32-80 with three antibodies binds between amino acids 32 and 80; CSD with two antibodies potentially bind the scaffolding domain (amino acids 80-101); and Cav-1-C with 1 antibody binds parts of the C-terminal half. Binding affinities (Kd) of these antibodies to soluble Cav-1 ranged from 10(-11) to 10(-8) M. Binding competition experiments revealed that these antibodies recognized a total of six different epitopes on Cav-1. Potency of these antibodies to neutralize Cav-1-mediated signaling pathways in cultured cells and in animal models will be tested. A selected monoclonal antibody will then be humanized and be further developed into a potential anti-prostate cancer therapeutic.

FIG. 1.

Recombinant Cav-1 proteins. (A) Antigens used to induce antibodies. (B) Additional GST-fusion proteins for domain mapping.

FIG. 2.

Coomassie R250-stained SDS-PAGE with 2 μg of purified anti-Cav-1 antibodies. Monoclonal antibody purchased from BD contains BSA.

FIG. 3.

ELISA of purified antibodies against different Cav-1 domains. GSH-coated 96-well plates were incubated with different GST-fusion proteins as indicated on the x-axis. Anti-Cav-1 antibodies were separated into three epitope recognition groups: (A) N1-31; (B) N32-80; (C) CSD and Cav-1-C.

FIG. 4.

Anti-Cav-1 antibodies were tested in immunoblots with cytosolic extracts (20 μg total protein) from Cav-1⁺ DU145 (D) or Cav-1^- LNCaP (L) cells. Biotin-conjugated protein markers (M) were used as migration indicator. The faster migrating band slightly above 20 kDa marker is Cav-1β.

FIG. 5.

Binding affinity of purified anti-Cav-1 antibodies against immobilized GST-Cav-1 on GSH-coated plates. (A) N1-31 group; (B) N32-80 group; (C) CSD and Cav-1-C groups.

FIG. 6.

Binding competition assays. Purified (A) 2A7, (B) 3C12, and (C) 4C9 were labeled with HRP. GST-Cav-1 proteins on GSH-coated plates were first blocked by excess amounts of competing antibodies indicated on the y-axis. After wash, HRP-conjugated antibodies at 1 nM were used for detection. Reactions without competitor were set as 100% and with self-blocking as 0%. Results are the average of four experiments.

FIG. 7.

Anti-Cav-1 antibodies domain mapping.

FIG. 8.

Immunoprecipitation and co-immunoprecipitation of Cav-1α and Cav-1β. (A) Condition medium from overnight cultured DU145 (lane 2) and LNCaP (lane 3) cells in serum-free medium was mixed with 4C9-conjugated Sepharose beads and pull-down Cav-1 was analyzed by immunoblot with 4C9 as primary antibody. DU145 cytosolic extract (10 μg total protein; lane 1) was used as positive control (B), same as A but other than 4C9-conjugated beads (lane 5). 2A7-conjugated Sepharose beads (lane 4) were also used for pull-down assays.