Exploring PARP biology and therapy response via metabolic flux analysis and a novel chemical proteomics workflow

点击翻译按钮获取中文摘要

Poly-ADP-ribose polymerases (PARPs) are NAD-dependent enzymes catalysing the formation of ADP-ribose modifications on proteins and are critical for DNA damage response, chromatin remodelling and with emerging roles in RNA splicing (Matveeva et al., 2016). Across the 17 PARPs currently identified we lack a detailed understanding of the subset of the proteome preferentially modified by each enzyme partly due to the analytical challenge of defining the global PARylome. PARP (I/II/III) inhibitors are synthetically lethal in HR-deficient cancer and are becoming an important maintenance therapy in ovarian cancer. The efficacy of PARP inhibitors is closely linked to the extent to which PARP is 'trapped' on damaged DNA (Murai et al., 2012). The NAD cofactor is essential for PARP activity and removal of trapped PARP by autoPARylation; however the impact of metabolic reprogramming in cancer cells and NAD availability on the PARylome or PARP inhibitor response has not been systematically explored We (Keun) have observed that KRAS mutant ovarian cancer cells are selectively sensitive to NAD depletion via chemical inhibition (FK866) of the NAD salvage enzyme NAMPT (Figure 1). This is consistent with KRAS-driven metabolic reprogramming leading to a classic Warburg state with a high requirement for both NAD+ in the cytosol to maintain a high glycolytic rate and NADPH for biosynthesis and antioxidant protection. We hypothesise that extra demands on NAD supply in KRAS mutant ovarian cancer cells alter the global PARylome, and subsequently the downstream function of PARylated proteins. 4 Using a novel quantitative chemical proteomics workflow involving metabolic incorporation of multiple clickable NAD+ precursors combined with TMT isobaric tagging developed in the DiMaggio lab (in review, Nature Chemical Biology), we unexpectedly observed that RNA splicing factors showed the greatest reduction in PARylation with PARP inhibition in breast cancer cells (Figure 2). Therefore we also hypothesise that NAD metabolism may have a previously uncharacterised role in regulating function of the spliceosome via PARP activity. This hypothesis is supported by a recent CRISPR screen that revealed ribonucleases are novel synthetically lethal determinants of PARP inhibitor response (Zimmermann et al., 2018). In this study a total of 73 high-confidence genes were identified across three cell lines that when mutated resulted in increased sensitivity to PARP inhibitors. A number of RNA splicing factors, such as DDX46 and SRSF11, were among these genes but not investigated further. Notably, DDX46 and SRSF11 were also identified by our chemical proteomics workflow as one of the most de-PARylated proteins after treatment with PARP inhibitors in MDA-MB-231 cells .