WISP1/CCN4: a potential target for inhibiting prostate cancer growth and spread to bone

Abstract

Prostate cancer (PC) is a leading cause of death in men however the factors that regulate its progression and eventual metastasis to bone remain unclear. Here we show that WISP1/CCN4 expression in prostate cancer tissues was up-regulated in early stages of the disease and, further, that it correlated with increased circulating levels of WISP1 in the sera of patients at early stages of the disease. WISP1 was also elevated in the mouse prostate cancer model TRAMP in the hypoplastic diseased tissue that develops prior to advanced carcinoma formation. When the ability of anti-WISP1 antibodies to reduce the spread of PC3-Luc cells to distant sites was tested it showed that twice weekly injections of anti-WISP1 antibodies reduced the number and overall size of distant tumors developed after intracardiac (IC) injection of PC3-Luc cells in mice. The ability of antibodies against WISP1 to inhibit growth of PC3-Luc cancer cells in mice was also evaluated and showed that twice weekly injections of anti-WISP1 antibodies reduced local tumor growth when examined in xenografts. To better understand the mechanism of action, the migration of PC3-Luc cells through membranes with or without a Matrigel™ barrier showed the cells were attracted to WISP1, and that this attraction was inhibited by treatment with anti-WISP1 antibodies. We also show the expression of WISP1 at the bone-tumor interface and in the stroma of early grade cancers suggested WISP1 expression is well placed to play roles in both fostering growth of the cancer and its spread to bone. In summary, the up-regulation of WISP1 in the early stages of cancer development coupled with its ability to inhibit spread and growth of prostate cancer cells makes it both a potential target and an accessible diagnostic marker for prostate cancer.

Figure 1. WISP1 expression in prostate cancer tissue and in serum from affected patients.

A. Immunohistochemical detection of WISP1 in biopsies from patients with various grades of prostate cancer. Specific staining for WISP1 (LF-185) was found in the stroma and in the epithelium of tumors from lower Gleason Grade (Grade I) scores and diffusely throughout the highly metastatic (Grade IV) neoplastic carcinoma. B. A representative image of a quantitative western blot assay used to measure serum levels of WISP1 in serum from normal subjects (lanes 1–6) and from subjects with prostate cancer (lanes 7–12). C. Quantified bands from 20 normal subjects and 60 subjects with prostate cancer were compared by t-test. D. Samples were stratified by disease stage using the TNM scoring system and compared by an ANOVA test. NL, Normal, pT1, pT2, pT3 lowest to the highest severity, LN⁺/SV⁺, lymph node positive, seminal vesicle positive. E. The levels of WISP1 were stratified by Gleason scores and compared by ANOVA test, 5 is lowest severity, 9 is greatest severity. F. The association of serum WISP1 levels with PSA was analyzed by Spearman correlation (PSA, X-axis, WISP1, Y-axis). Each small circle represents values from a single patient.

Figure 2. WISP1 immunohistochemistry of prostate tissue from TRAMP mice.

A. WISP1 localization at 8 weeks of age in normal prostate tissue (upper panel) and hyperplastic tissue (lower panel, black arrow). B. WISP1 localization at 16 weeks of age in normal prostate tissue (upper panel), hyperplastic tissue (middle panel, black arrow) and carcinoma (lower panel). IgG controls are shown to the right of each panel.

Figure 3. Treatment with WISP1 antibodies reduced spread and establishment of PC3-Luc tumors in immunocompromized mice when injected IC.

A. Total number of tumors developed per mouse treated with IgG or anti-WISP1 (LF-185). **p<0.01 PBS vs. anti-WISP1 treatment, B. Total tumor burden judged by relative luciferease measurements per mouse. ***p<0.001 IgG vs. anti-WISP1 treatment.

Figure 4. Growth of PC3-Luc human prostate cancer cell sub-cutaneous xenografts were reduced by treatment with antibodies specific to human WISP1.

A. Photograph of representative PC3-Luc tumors dissected from immunocompromized mice treated with PBS (left), IgG (middle) or anti-WISP1 (LF-185, right). B. Weight in grams of PC3-Luc cell tumors grown in mice treated with PBS, IgG or anti-WISP1 (LF-185) after 4 weeks. *p<0.05 PBS vs. anti-WISP1, ***p<0.001 IgG vs. anti-WISP. PBS vs. IgG were not significantly different.

Figure 5. PC3-Luc tumor spread and establishment.

A. X-ray of sagittal view of a head from a control mouse devoid of luciferase/PC3-Luc cells (left panel) compared to the head of a mouse with an established tumor (right panel). The lower panels are higher power images of the images shown above them with the arrows pointing to the site of tumor integration below the molars that is more radio-opaque compared to the control. B. Histological sections through a regions of the jaw showing tumor invasion within the bone tissue just below the tooth (incisor). Left, Hematoxylin and eosin stain, middle, staining with TRAP, right immunohistochemistry using anti-WISP1 to show the location of WISP1 relative to the osteoclasts. Upper panels are 50X and the boxed area is shown in the lower panels is 430X magnification. Arrow heads point to the TRAP stained osteoclasts (middle panel), and the localization of WISP1 (brown stain) in the tumor and at the bone tumor/PDL interface.

Figure 6. Cell migration and invasion is blocked by treatment of PC3-Luc with anti- WISP1.

A. Chemotaxis of PC3-Luc cells towards the bottom of a Boyden chamber containing WISP1. **p<0.01 PBS vs. WISP1-protein. B. Relative chemotaxis levels of PC3-Luc pre-treated with PBS, IgG or anti-WISP1 (LF-187) towards a gradient of 5% FBS. C. Relative chemotaxis levels of PC3-Luc pre-treated with PBS, IgG or anti-WISP1 (LF-187) towards 200 ng/ml WISP1 protein. D. Relative invasion of PC3-Luc cells treated with IgG, PBS, or anti-WISP1 (LF-187) through membranes coated with Matrigel™ toward a lower chamber containing WISP1-protein.