Myosin 1C isoform A is a novel candidate diagnostic marker for prostate cancer

Abstract

Early diagnosis of prostate cancer is a challenging issue due to the lack of specific markers. Therefore, a sensitive diagnostic marker that is expressed or upregulated exclusively in prostate cancer cells would facilitate diagnostic procedures and ensure a better outcome. We evaluated the expression of myosin 1C isoform A in 5 prostate cell lines, 41 prostate cancer cases, and 11 benign hyperplasias. We analyzed the expression of 12 surface molecules on prostate cancer cells by flow cytometry and analyzed whether high or low myosin 1C isoform A expression could be attributed to a distinct phenotype of prostate cancer cells. Median myosin 1C isoform A expression in prostate cancer samples and cancer cell lines was 2 orders of magnitude higher than in benign prostate hyperplasia. Based on isoform A expression, we could also distinguish clinical stage 2 from clinical stage 3. Among cell lines, PC-3 cells with the highest myosin 1C isoform A level had diminished numbers of CD10/CD13-positive cells and increased numbers of CD29 (integrin β1), CD38, CD54 (ICAM1) positive cells. The surface phenotype of clinical samples was similar to prostate cancer cell lines with high isoform A expression and could be described as CD10-/CD13- with heterogeneous expression of other markers. Both for cell lines and cancer specimens we observed the strong correlation of high myosin 1C isoform A mRNA expression and elevated levels of CD29 and CD54, suggesting a more adhesive phenotype for cells with high isoform A expression. Compared to normal tissue, prostate cancer samples had also reduced numbers of CD24- and CD38-positive cells. Our data suggest that a high level of myosin 1C isoform A is a specific marker both for prostate cancer cells and prostate cancer cell lines. High expression of isoform A is associated with less activated (CD24/CD38 low) and more adhesive (CD29/CD54 high) surface phenotype compared to benign prostate tissue.

Fig 1. Myosin 1C isoform A is overexpressed in prostate cancer specimens on mRNA and protein level.

A–relative normalized mRNA quantity in clinical BPH and PCa specimens and prostate cell lines. Horizontal lines represent range and median; all experiments were conducted in triplicate and the results normalized to YWHAZ, GAPDH, and HPRT1 genes. B—Box plot of myosin 1C isoform A levels at different disease stages. Pairwise comparison of means is indicated with p-values obtained in Conover-Iman post-hoc test with Holm correction. Data shown with median values (thick horizontal line), interquartile ranges (boxes), total ranges (whiskers), and outliers (dots). C–Immunoblotting of myosin 1C isoform A in prostate cancer samples (PCa1-PCa3) and benign prostate hyperplasia samples (BPH1-BPH3). Alpha tubulin was taken as technical control.

Fig 2. Basic immunophenotype of clinical specimens.

Gates were set according to unstained controls. A—Representative immunophenotype of benign hyperplasia. The sample contains a small CD44+/CD24- subpopulation and a more significant CD24+ subpopulation with partial CD44 co-expression. The surface expression of CD133, CD57, and CD90 is low or lacking. B—Representative immunophenotype of “low-expressing” carcinoma. The surface expression of all markers: CD44, CD24, CD133, CD57, and CD90 is low or lacking. C—Percentage of CD24-postitve cells at different disease stages. Plotting as in Fig 1B.

Fig 3. Extended surface phenotypes of model prostate cell lines.

Fig 4. Prostate cells with low and high myosin 1C isoform A expression exhibit different surface phenotypes.

A—Spearman correlation coefficients between myosin 1C isoform A expression and percentages of positive cells for indicated surface markers at different disease stages. Significant correlations (p < 0.05) are marked by color. B–D—Least-squares regression between myosin 1C isoform A and percentage of CD54 (B), CD38 (C) and CD29 (D) positive cells. The approximation and coefficient of determination are included. Dashed line shows the 95% confidence interval.