Nuclear cathepsin L and SERPINA5

In this project, we want to investigate how a protein (protease inhibitor), which can be internalized by cells, and an enzyme normally localized in lysosomes, react with each other in the nucleus, and which impact this interaction has on the nuclear protein profile. SERPINA5 (= protein C inhibitor) is a secreted glycoprotein of the superfamily of serpins (serine proteases inhibitors), which inhibits a variety of extracellular proteolytic enzymes by formation of stable complexes. In humans, SERPINA5 is found in most body fluids. Recent studies indicate that SERPINA5 may have a preventive effect against malignant tumor growth. Besides many other proteases, SERPINA5 inhibits cathepsin L, a cysteine protease, which occurs mainly intracellularly in lysosomes. SERPINA5 binds to heparin-like glycosaminoglycans and phospholipids. These molecules, which are also components of the cell membrane, modulate the inhibitory activity of SERPINA5 and support the internalization of SERPINA5 by cells. Internalized SERPINA5 can translocate to the nucleus. So far the exact (unconventional) mechanism of SERPINA5 internalization and the function of SERPINA5 in the nucleus have not been defined. Our own preliminary data obtained with lymphoma cells suggest that SERPINA5 interacts with cathepsin L in the nucleus. This finding seems important since nuclear localization of cathepsin L is mainly seen in tumor cells, but not in their normal counterparts. In nuclei of tumor cells cathepsin L proteolytically processes several proteins involved in cell cycle regulation and DNA repair, which could contribute to the malignant behavior of these cells. In the proposed project we want to analyze by immunocytochemistry and by proximity ligation assays, under which conditions SERPINA5 and cathepsin L are co-localized in the nucleus and if they interact in living cells. Furthermore we want to analyze if exogenously added and internalized SERPINA5 can inhibit cathepsin L in the nucleus. If this is the case this should have an impact on the nuclear protein profile. Using proteomics techniques (2D-DIGE, mass spectrometry) we will compare the nuclear protein profile of cells pretreated with SERPINA5 with that of untreated cells. In controls we will use SERPINA5 mutants that either cannot be internalized, or cannot translocate to the nucleus, or have no protease inhibitory activity, respectively. We will also compare the nuclear protein profile of SERPINA5 treated cells to that of cells treated with a cell permeable, small cathepsin L inhibitor. After validation of the results with 1- and 2- dimensional Western blots we will perform functional studies (e.g. cell proliferation assays, apoptosis assays) to get insights into the biological relevance of our findings.

This project aimed to investigate the influence of the inhibition of the protease cathepsin L on a cell's functions and protein composition. Cathepsin L (CTSL) is increasingly formed in tumour cells, and the increased occurrence in the cell nucleus promotes the growth of tumour cells. SERPINA5 is a secreted serine protease inhibitor that has been shown to block the enzymatic activity of CTSL. In this study, an inverse correlation between SERPINA5 and CTSL expression as a function of malignancy potential was demonstrated, with SERPINA5 being expressed at elevated levels in non-malignant RWPE-1 cells but only at low levels in prostate cancer DU145 cells. We observed high CTSL expression and almost no SERPINA5 in DU145 cells. Based on these observations, we investigated how CTSL deficiency affects cell physiology by knocking out the CTSL gene in DU145 cells using the CRISPR/Cas method. The protein repertoire of the CTSL-knockout DU145 cells was examined in comparison to its wild type using two complementary proteomics technologies: bottom-up "label-free shotgun" and top-down two-dimensional fluorescence in-gel differential gel electrophoresis "2D-DIGE". Proteomic profiling of cytoplasmic and nuclear extracts of WT and CTSL KO DU145 cells revealed that 156 out of 1690 totally identified proteins were only present in nuclear while 517 out of 2045 proteins were only present in cytoplasmic CTSL KO cells. For example, the top-down proteomics method found significant amounts of proteoforms of cathepsins with an increased amount of the 22 kDa cathepsin B cleavage product and the 27 kDa cathepsin D cleavage product. In parallel, we also detected increased expression of the sequence-based precursor of CTSB in our CTSL-ko-DU145 cells by shotgun proteomics. Interestingly ,an decreased amount of a 42 kDa cleavage proteoform of the tumor-specific variant of the pyruvate kinase, PKM2, were also abundant in the CTSL KO DU145 cells. This 42 kDa proteoform of PKM2 was previously found in pancreatic tumour cells and was identified as an enzymatic cleavage product of cathepsin B. The 42 kDa cleavage product of PKM2 may be responsible for its reduced enzymatic activity and, thus, in part, for the reduced citric acid cycle-mediated oxidative phosphorylation and thus rewiring of cell metabolism to glucose, known as the Warburg effect in tumour cells. Cell cycle analysis showed that DU145 cells are predominantly in S-phase, where CTSL activity is also high but still increased at G2. When we analyzed DU145 CTSL knock cells compared to WT cells, we found a decreased percentage of cells in the G2 phase, whereby an increase in G1 was associated with accelerated glycolysis. Overall, the cell cycle experiments confirmed our proteomic results for CTSL-regulated PKM2 activity, thus pointing to an essential role of CTSL in regulating energy metabolism and cell cycle control in DU145 cells.