The androgen receptor (AR) is essential for the growth of prostate cancer cells. Here, we report that tyrosine phosphorylation of AR is induced by growth factors and elevated in hormone-refractory prostate tumors. Mutation of the major tyrosine phosphorylation site in AR significantly inhibits the growth of prostate cancer cells under androgen-depleted conditions. The Src tyrosine kinase appears to be responsible for phosphorylating AR, and there is a positive correlation of AR tyrosine phosphorylation with Src tyrosine kinase activity in human prostate tumors. Our data collectively suggest that growth factors and their downstream tyrosine kinases, which are elevated during hormone-ablation therapy, can induce tyrosine phosphorylation of AR and such modification may be important for prostate tumor growth under androgen-depleted conditions.

We previously showed that the 44-kDa serine/threonine kinase Pim-1 (Pim-1L) can protect prostate cancer cells from apoptosis induced by chemotherapeutic drugs (Xie, Y., Xu, K., Dai, B., Guo, Z., Jiang, T., Chen, H., and Qiu, Y. (2006) Oncogene 25, 70 -78). To further explore the mechanisms of Pim-1L-mediated resistance to chemotherapeutic drugs in prostate cancer cells, we employed a yeast two-hybrid screening to identify cellular proteins that were associated with Pim-1L, and we found the ABC transporter BCRP/ABCG2 as one of the potential interacting partners of Pim-1L. We also showed that the expression level of Pim-1L and BCRP was up-regulated in mitoxantrone and docetaxel-resistant prostate cancer cell lines. Pim-1L was co-localized with BCRP on the plasma membrane and induced phosphorylation of BCRP at threonine 362. Knocking-down Pim-1L expression in the drug-resistant prostate cancer cells abolished multimer formation of endogenous BCRP and resensitized the resistant cells to chemotherapeutic drugs suggesting that BCRP phosphorylation induced by Pim-1L was essential for its functionality. This is further corroborated by our finding that the plasma membrane localization and drug-resistant activity of BCRP were compromised by T362A mutation. Our data suggest that Pim-1L may protect prostate cancer cells from apoptosis, at least in part, through regulation of transmembrane drug efflux pump. These findings may provide a potential therapeutic approach by disrupting Pim-1 signaling to reverse BCRP-mediated multidrug resistance.

Protein kinase Pim-1 has been implicated in the development of hematopoietic and prostatic malignancies. Here, we present the evidence that two isoforms, the 44 and 33 kDa Pim-1, are expressed in all human prostate cancer cell lines examined. The subcellular localization of human 44 kDa Pim-1 is primarily on the plasma membrane, while the 33 kDa isoform is present in both the cytosol and nucleus in PCA cells. The 44 kDa Pim-1 contains the proline-rich motif at the N-terminus and directly binds to the SH3 domain of tyrosine kinase Etk. Such interaction leads to the activation of Etk kinase activity possibly by competing with the tumor suppressor p53. This is corroborated by the fact that overexpression of the 44 kDa Pim-1 in prostate cancer cells confers the resistance to chemotherapeutic drugs. Our results suggest that these two isoforms of Pim-1 kinase may regulate distinct substrates and the 44 kDa Pim-1 may play a more prominent role in drug resistance in prostate cancer cells.

The oncogenic serine/threonine kinase Pim-1 phosphorylates and activates the ATP-binding cassette transporter breast cancer resistance protein (ABCG2). The ABC transporter P-glycoprotein (Pgp; ABCB1) also contains a Pim-1 phosphorylation consensus sequence, and we hypothesized that Pim-1 also regulates Pgp. Pgp is exported from the endoplasmic reticulum (ER) as a 150-kDa species that is glycosylated to 170-kDa Pgp, translocates to the cell surface, and mediates drug efflux; alternatively, 150-kDa Pgp is cleaved to a 130-kDa proteolytic product by ER proteases or undergoes ubiquitination and proteasomal degradation. Pim-1 and Pgp interaction was studied in GST pull-down and phosphorylation in in vitro kinase assays. Pim-1 knockdown and inhibition effects on Pgp expression were studied by immunoblotting and flow cytometry and on Pgp stability by immunoblotting after cycloheximide treatment. Pim-1 directly interacted with and phosphorylated Pgp in intact cells and in vitro. Pim-1 knockdown or inhibition decreased cellular and cell surface 170-kDa Pgp, in association with both transient increase in 130-kDa Pgp and increased Pgp ubiquitination and proteasomal degradation. Pim-1 inhibition also decreased expression of 150-kDa Pgp in the presence of the glycosylation inhibitor 2-deoxy-D-glucose. Finally, Pim-1 inhibition sensitized Pgp-overexpressing cells to doxorubicin. Thus, Pim-1 regulates Pgp expression by protecting 150-kDa Pgp from proteolytic and proteasomal degradation and enabling Pgp glycosylation and cell surface translocation and thus Pgpmediated drug efflux. Pim-1 inhibitors are entering clinical trials and may provide a novel approach to abrogating drug resistance.

Drug resistance remains a clinical challenge in cancer treatment due to poor understanding of underlying mechanisms. We have established several drug-resistant prostate cancer cell lines by long-term culture in medium containing chemotherapeutic drugs. These resistant lines displayed a significant increase in side population cells due to overexpression of drug efflux pumps including ABCG2/BCRP and MDR1/Pgp. To uncover potential mechanisms underlying drug resistance, we performed microarray analysis to identify differentially expressed genes in 2 drug-resistant lines. We observed that POU5F1/OCT4, a transcription factor key to regulating pluripotency in embryonic stem cells, was upregulated in drug-resistant lines and accompanied by transcriptional activation of a set of its known target genes. Upregulation of OCT4 in drug-resistant cells was validated by RT-PCR and sequencing of PCR products as well as confirmation by Western blot and specific shRNA knockdown. Analysis of the regulatory region of POU5F1/OCT4 revealed a reduction of methylation in drug-resistant cell lines. Furthermore, these drug-resistant cells exhibited a significant increase in tumorigenicity in vivo. Subcutaneous inoculation of as few as 10 drug-resistant cells could initiate tumor formation in SCID mice, whereas no detectable tumors were observed from the parental line under similar conditions, suggesting that these drug-resistant cells may be enriched for tumor-initiating cells. Knocking down OCT4 expression by specific shRNAs attenuated growth of drug-resistant cells. Our data suggest that OCT4 re-expression in cancer cells may play an important role in carcinogenesis and provide one possible mechanism by which cancer cells acquire/maintain a drug-resistant phenotype.

Aberrant elevation of JARID1B and histone H3 lysine 4 trimethylation (H3K4me3) is frequently observed in many diseases including prostate cancer (PCa), yet the mechanisms on the regulation of JARID1B and H3K4me3 through epigenetic alterations still remain poorly understood. Here we report that Skp2 modulates JARID1B and H3K4me3 levels in vitro in cultured cells and in vivo in mouse models. We demonstrated that Skp2 inactivation decreased H3K4me3 levels, along with a reduction of cell growth, cell migration and malignant transformation of Pten/Trp53 double null MEFs, and further restrained prostate tumorigenesis of Pten/Trp53 mutant mice. Mechanistically, Skp2 decreased the K63-linked ubiquitination of JARID1B by E3 ubiquitin ligase TRAF6, thus decreasing JARID1B demethylase activity and in turn increasing H3K4me3. In agreement, Skp2 deficiency resulted in an increase of JARID1B ubiquitination and in turn a reduction of H3K4me3, and induced senescence through JARID1B accumulation in nucleoli of PCa cells and prostate tumors of mice. Furthermore, we showed that the elevations of Skp2 and H3K4me3 contributed to castration-resistant prostate cancer (CRPC) in mice, and were positively correlated in human PCa specimens. Taken together, our findings reveal a novel network of SKP2- JARID1B, and targeting SKP2 and JARID1B may be a potential strategy for PCa control.

Matrix metalloproteinase (MMP) is defined as an endopeptidase in the extracellular matrix (ECM), which plays essential roles in physiological processes such as organogenesis, wound healing, angiogenesis, apoptosis and motility. MMPs are produced and assembled in the cytoplasm as proenzymes with a cytoplasmic domain and require extracellular activation. MMPs can degrade receptors, extracellular matrix proteins, PARPs and release apoptotic substances. MMPs have been found in the cytosol, organelles and extracellular compartments and recently many types of MMPs have been found in the nucleus. However, the mechanisms and roles of MMPs inside the cell nucleus are still poorly understood. Here we summarized the nuclear localization mechanisms of MMPs and their functions in the nucleus such as apoptosis, tissue remodeling upon injury and cancer progression. Most importantly, we found that nuclear MMPs have evolved to translocate to membrane and target ECM possibly through evolution of nuclear localization signal (NLS), natural selection and anti-apoptotic survival. Thus, the knowledge about the evolution and regulation of nuclear MMPs appears to be essential in understanding a variety of cellular processes along with the development of MMP-targeted therapeutic drugs against the progression of certain diseases.

The type III receptor tyrosine kinase fms-like tyrosine kinase 3 (FLT3) is expressed on both normal hematopoietic stem cells and acute myeloid leukemia (AML) cells and regulates their proliferation. Internal tandem duplication (ITD) mutation of FLT3 is present in a third of AML cases, results in constitutive activation and aberrant signaling of FLT3, and is associated with adverse treatment outcomes. While wild-type (WT) FLT3 is predominantly a 150 kDa complex glycosylated cell surface protein, FLT3-ITD is partially retained in the endoplasmic reticulum as a 130 kDa underglycosylated species associated with the chaperones calnexin and heat shock protein (HSP) 90, and mediates aberrant STAT5 signaling, which upregulates the oncogenic serine/threonine kinase Pim-1. FLT3 contains a Pim-1 substrate consensus serine phosphorylation site, and we hypothesized that it might be a Pim-1 substrate. Pim-1 was indeed found to directly interact with and serine-phosphorylate FLT3. Pim-1 inhibition decreased the expression and half-life of 130 kDa FLT3, with partial abrogation by proteasome inhibition, in association with decreased FLT3 binding to calnexin and HSP90, and increased 150 kDa FLT3 expression and half-life, with abrogation by inhibition of glycosylation. These findings were consistent with Pim-1 stabilizing FLT3-ITD as a 130 kDa species associated with calnexin and HSP90 and inhibiting its glycosylation to form the 150 kDa species. Pim-1 knockdown effects were similar. Pim-1 inhibition also decreased phosphorylation of FLT3 at tyrosine 591 and of STAT5, and expression of Pim-1 itself, consistent with inhibition of the FLT3-ITD-STAT5 signaling pathway. Finally, Pim-1 inhibition synergized with FLT3 inhibition in inducing apoptosis of FLT3-ITD cells. This is, to our knowledge, the first demonstration of a role of Pim-1 in a positive feedback loop promoting aberrant signaling in malignant cells.