Currently, few prognostic factors are available to predict the emergence of castration-resistant prostate cancer and no curative options are available. Epigenetic gene regulation has been shown to trigger prostate cancer metastasis and androgen independence. Histone lysine demethylases (KDMs) are epigenetic enzymes that can remove both repressive and activating histone marks. KDM5 family members are capable of removing the histone H3 lysine 4 dimethylation-activating mark, rendering them potential players in the down-regulation of tumor suppressors and suggesting that their activity could repress oncogenes. Here, we systematically investigated KDM5C expression patterns in two independent radical prostatectomy cohorts (822 prostate tumors in total) by immunohistochemistry. Positive nuclear KDM5C staining was significantly associated with a reduced prostate-specific antigen relapse-free survival. Our study confirmed that nuclear KDM5C expression is an independent prognostic parameter. Most strikingly, the prognostic value of nuclear KDM5C expression for progression-free survival was exclusively pronounced for the Gleason group 7. In addition, KDM5C knockdown resulted in growth retardation of prostate cancer cells in vitro and induced regulation of several proliferation-associated genes. Our data indicate that KDM5C is functionally involved in proliferation control of prostate cancer cells and might represent a novel attractive therapy target. Moreover, overexpression of KDM5C is an independent new predictive marker for therapy failure as determined by biochemical recurrence in patients after prostatectomy.

Hereditary axonopathies are frequently caused by mutations in proteins that reside in the endoplasmic reticulum (ER). Which of the many ER functions are pathologically relevant, however, remains to be determined. REEP1 is an ER protein mutated in hereditary spastic paraplegia (HSP) and hereditary motor neuropathy (HMN). We found that HSP-associated missense variants at the N-terminus of REEP1 abolish ER targeting, whereas two more central variants are either rare benign SNPs or confer pathogenicity via a different mechanism. The mis-targeted variants accumulate at lipid droplets (LDs). N-terminal tagging, deletion of the N-terminus, and expression of a minor REEP1 isoform had the same effect. We also confirmed an increase in LD size upon cooverexpression of atlastins and REEP1. Neither wild-type REEP1, LD-targeted HSP variants, nor a non-LD-targeted HMN variant reproduced this effect when expressed alone. We conclude that the N-terminus of REEP1 is necessary for proper targeting to and/or retention in the ER. The protein's potential to also associate with LDs corroborates a synergistic effect with atlastins on LD size. Interestingly, LD size is also altered upon knockdown of seipin, mutations of which also cause HSP and HMN. Regulation of LDs may thus be an ER function critical for long-term axonal maintenance.

No abstract

No abstract

No abstract