Serum chemokine (CC motif) ligand 2 level as a diagnostic, predictive, and prognostic biomarker for prostate cancer

Abstract

Prostate-specific antigen (PSA) is regarded as the most sensitive biomarker for prostate cancer. Although androgen/androgen receptor (AR) signaling promotes prostate cancer progression, suppression of AR signaling induces chemokine (CC motif) ligand 2 (CCL2), which enables prostate cancer cells to gain metastatic potential. AR-controlled PSA alone may be an unreliable biomarker for patients receiving androgen deprivation therapy. Therefore, we investigated the validity of CCL2 as a complementary biomarker to PSA for prostate cancer. Our in vitro approach of enriching for prostate cancer cells with higher migration potential showed that CCL2 activated cellular migration. Importantly, we found that CCL2 levels were significantly different between men (n = 379) with and without prostate cancer. Patients with CCL2 ≥ 320 pg/mL had worse overall survival and prostate cancer -specific survival than those with CCL2 < 320 pg/mL. A novel risk classification was developed according to the risk factors CCL2 ≥ 320 pg/mL and PSA ≥ 100 ng/mL, and scores of 2, 1, and 0 were defined as poor, intermediate, and good risk, respectively, and clearly distinguished patient outcomes. CCL2 may serve as a novel biomarker for prostate cancer. The novel risk classification based on combining CCL2 and PSA is more reliable than using either alone.

Keywords: CCL2; androgen deprivation therapy; biomarker; prostate cancer; risk classification.

Figure 1. The effect of CCL2 on human prostate cancer C4-2 cells

A. The migration of C4-2 cells in the absence of suppression of androgen/AR signaling was performed using 24-well transwell inserts and plates with or without treatment using the indicated concentrations of CCL2. B. WB analyses of mesenchymal markers (MMP9 and Snail), an epithelial marker (E-cadherin), and STAT3 in C4-2 cells after treatment with 10 ng/mL of CCL2 were performed and compared with the control. The activation of STAT3 (pSTAT3), which is a key molecule bridging between CCL2 and EMT, was associated with the EMT.

Figure 2. The enrichment of the population of prostate cancer cells with increased migration potential

A. We established an in vitro transwell approach to select prostate cancer cells with increased migration ability. C4-2 cells that migrated from upper transwell inserts to the bottom of lower wells were called mig cells. B. The migration of C4-2 prt and mig cells were compared. C. C4-2 cells attached to the bottom surface after migration were counted, and the morphological change of mig cells into a polygonal shape was observed.

Figure 3. Characteristics of prostate cancer cells with increased migration potential

A.CCL2 concentrations in supernatants of C4-2 prt and mig cells were measured using an ELISA kit. B. mRNA levels of AR expressed by C4-2 prt and mig cells were analyzed using qRT-PCR. C. mRNA levels of TGF-β1 in C4-2 prt and mig cells were analyzed using qRT-PCR. D. WB analyses of N-cadherin (mesenchymal marker), E-cadherin (epithelial marker), and AR expression by C4-2 mig cells were performed, and the results were compared with those for C4-2 prt cells (mig1 and mig2 were independently collected). E. These data indicate that CCL2 induced the migration of prostate cancer cells and the EMT in vitro in the presence or absence of the suppression of androgen/AR signaling and that prostate cancer cells with higher metastatic potential secreted more CCL2 (red arrow).

Figure 4. Analysis of PSA and CCL2 levels as functions of OS, PCaSS, and CFS

A. The scatter plot of PSA (logarithmic scale) and CCL2 levels shows clustering of patients without prostate cancer around small areas of low PSA and CCL2 levels, and their levels in prostate cancer patients were widely scattered. PCa = prostate cancer. B., C. Kaplan-Meier curves of OS and PCaSS in 255 patients with prostate cancer with CCL2 ≥ 320 pg/mL and CCL2 < 320 pg/mL. The 5-year OS rates of patients with CCL2 ≥ 320 pg/mL and CCL2 < 320 pg/mL were 76.9% and 93.9%, respectively. The 5-year PCaSS rates of patients with CCL2 ≥ 320 pg/mL and CCL2 < 320 pg/mL were 84.3% and 97.3%, respectively. D. Kaplan-Meier curves of the CRPC-free survival (CFS) of 102 patients treated with ADT and patients with CCL2 ≥ 320 pg/mL and < 320 pg/mL. The 5-year CFS rates of patients with CCL2 ≥ 320 pg/mL and < 320 pg/mL were 49.5% and 80.9%, respectively.

Figure 5. The prevalence of CCL2 values in each stage of TNM and GS

A. The levels of CCL2 in patients with each T stage were analyzed. The level of CCL2 in patients with T4 was higher than those with T1. B. The levels of CCL2 in patients with each N stage were analyzed. The level of CCL2 in patients with N1 was higher than those with N0. C. The levels of CCL2 in patients with each M stage were analyzed. There was no significant difference in CCL2 levels between patients with M1 and M0 (p = 0.1277). D. The levels of CCL2 in patients with each GS were analyzed. The level of CCL2 in patients with GS ≥ 9 was higher than those with GS = 6.

Figure 6. The novel risk classification

The classification system was designed according to the number of risk factors (CCL2 ≥ 320 pg/mL and PSA ≥ 100 ng/mL) and 2, 1, and 0 were defined as poor, intermediate, and good risk, respectively. A., B.) Kaplan-Meier curves of OS and PCaSS in 255 patients with prostate cancer with poor, intermediate, and good risk. The 5-year OS rates were 52.6%, 83.8%, and 95.6% of patients with poor, intermediate, and good risk, respectively. The 5-year PCaSS rates of patients with poor, intermediate, and good risk were 61.4%, 87.2%, and 99.5%, respectively. C. Kaplan-Meier curves of CFS of 102 patients with prostate cancer treated with ADT. The 5-year CFS rates of patients with poor, intermediate, and good risk were 0%, 66.5%, and 93.0%, respectively.