Mitosis phase enrichment with identification of mitotic centromere-associated kinesin as a therapeutic target in castration-resistant prostate cancer

Abstract

The recently described transcriptomic switch to a mitosis program in castration-resistant prostate cancer (CRPC) suggests that mitotic proteins may be rationally targeted at this lethal stage of the disease. In this study, we showed upregulation of the mitosis-phase at the protein level in our cohort of 51 clinical CRPC cases and found centrosomal aberrations to also occur preferentially in CRPC compared with untreated, high Gleason-grade hormone-sensitive prostate cancer (P<0.0001). Expression profiling of chemotherapy-resistant CRPC samples (n = 25) was performed, and the results were compared with data from primary chemotherapy-naïve CRPC (n = 10) and hormone-sensitive prostate cancer cases (n = 108). Our results showed enrichment of mitosis-phase genes and pathways, with progression to both castration-resistant and chemotherapy-resistant disease. The mitotic centromere-associated kinesin (MCAK) was identified as a novel mitosis-phase target in prostate cancer that was overexpressed in multiple CRPC gene-expression datasets. We found concordant gene expression of MCAK between our parent and murine CRPC xenograft pairs and increased MCAK protein expression with clinical progression of prostate cancer to a castration-resistant disease stage. Knockdown of MCAK arrested the growth of prostate cancer cells suggesting its utility as a potential therapeutic target.

Figure 1. Increased mitosis-phase and centrosomal protein expression in castration-resistant prostate cancer compared with high-grade hormone-sensitive prostate cancer.

Hematoxylin and eosin (H&E)-stained sections showing castration-resistant adenocarcinoma (A), castration-resistant small cell carcinoma (B), and hormone-sensitive high-grade prostate carcinoma (C). Castration-resistant adenocarcinoma immunostained with gamma tubulin (D) and p-histone H3 (G). Castration-resistant small cell carcinoma immunostained with gamma tubulin (E) and p-histone H3 (H). Hormone-sensitive high-grade prostate carcinoma immunostained with gamma tubulin (F) and p-histone H3 (I). Asterisks indicate significantly increased labeling for the mitosis-phase marker p-histone H3 (J) and centrosomal marker gamma-tubulin (K) in castration-resistant prostate carcinoma compared to hormone-sensitive high-grade prostate carcinoma.

Figure 2. Enhancement of mitosis-phase genes and pathways with progression to chemotherapy-resistant disease.

(A) Cell-cycle regulation pathways are significantly upregulated in the CRPC chemotherapy-resistant (TX) group compared with the CRPC chemotherapy-naïve group. (B) Ingenuity Pathway Analysis.(IPA) of the upregulated gene set demonstrates that the canonical pathway of mitotic roles of polo-like kinase is significantly associated with the CRPC chemotherapy-resistant group (P = 0.004). The P value is calculated by Fisher's Exact Test. The genes that are upregulated in the CRPC chemotherapy-resistant group and are involved in this pathway are color-coded in red. The other genes that are included in this pathway but are not in the upregulated gene set are indicated in white. (see text for identification of upreguated gene set in details). (C) Significant overlap of the upregulated gene set in CRPC chemotherapy-resistant tumors with M-phase (n = 27, P = 5.00E-7) compared with the downregulated gene set (n = 9, P = 0.416). (D) GSEA indicated that M-Phase was significantly enriched in the CRPC chemotherapy-resistant group at FDR<10%.

Figure 3. Overexpression of MCAK with clinical progression of prostate cancer.

MCAK-immunostained sections of castration-resistant adenocarcinoma (A), castration-resistant small cell carcinoma (B), and hormone-sensitive prostate carcinoma (C). D) Significantly increased labeling for the mitotic centromere-associated protein MCAK with progression to CRPC, indicated by asterisks.

Figure 4. Growth inhibition of LNCaP and C4-2B cells by si-MCAK.

A) Western blot confirming knockdown of MCAK by si-MCAK#1. Whole-cell lysates from LNCaP cells transfected with si-control (C) or si-MCAK #1 (M) were collected at different time points after transfection, as indicated. Tubulin was used as the loading control. B) MTT cell growth assay. LNCaP cells were treated the same as in A), and after a 24-h transfection, 4,000 cells were seeded into each well of a 96-well plate (n = 10 for each group) and the MTT assay was performed in a 24 h interval. C) Western blot confirming knockdown of MCAK with si-MCAK#1. Whole-cell lysates from C4-2B cells transfected with si-control (C) or si-MCAK #1 (M) were collected at different time points after transfection, as indicated. Tubulin was used as the loading control. D) MTT cell growth assay. C4-2B cells were transfected with si-control and si-MCAK #1, respectively. After a 24-h transfection, 4,000 cells were seeded into each well of a 96-well plate and incubated for different time periods, as indicated, followed by the MTT assay (n = 10 for each group).