Conclusions and future perspectives

In the work presented here we transferred a newly discovered function of MK2 as antagonist of Chk1 to the treatment of pancreatic cancer cells with gemcitabine. We could confirm a protective effect of MK2 inhibition to chemotherapy in all cell lines tested. This suggests MK2 as a novel predictor of gemcitabine sensitivity and its activation to be a promising drug target in pancreatic cancers. Interestingly, when we approached the transferability of the antagonism of MK2 and Chk1 to pancreatic cancers, we found the system to work in only one of four tested cell lines. All other cell lines acquired chemoresistance to gemcitabine when Chk1’s kinase function was impaired. This puts the break to the enthusiasm of applying checkpoint kinase inhibitors as chemosensitizers.

In our high throughput approach to find novel modulators of the gemcitabine response we identified the chromatin remodeling factor CHD8, which has been described with contradicting properties in terms of its role in tumor biology and DNA-associated processes.

In our experiments, knockdown of CHD8 led to increased DNA damage signaling, i.e. the induction of γH2AX, but also to increased phosphorylation of the checkpoint kinases Chk1 and Chk2. This was true for UV, gemcitabine as well as for the radiomimetic NCS, suggesting a broad function for CHD8 in the DDR and possibly the involvement of numerous pathways. The effect of CHD8 knockdown was most pronounced when the induction of DNA damage was restricted to a short period of time, and even stronger after the induction of DSBs. As a consequence, we could detect the onset of apoptosis in CHD8-depleted cells,

irrespective of p53. We further detected faster resumption of cell cycle progression, increased incorporation of nucleotides into DNA and a debatable advantage in terms of the recovery after gemcitabine treatment. It is of note that CHD8-deficient cells display diminished levels of wild type as well as oncogenic mutant p53, whose non-targetable abundance remains a challenging task in research on tumor biology. We expand the still small library of mutant-p53 regulating factors by one.

However, other scientists have reported features for CHD8 that apparently contradict our findings. It will thus be of future interest to shed more light on CHD8-regulated processes in the DDR, which we could only roughly outline in this work. To specify CHD8’s role in cell cycle regulation, it would be of interest to determine CDK and Cdc25 activity in a cell cycle-specific manner. If cell cycle progression were resumed faster after the induction of DNA damage by gemcitabine or other agents, it would suggest faster repair of damaged DNA.

Finally, it would be of interest to find the underlying mechanism for the decrease of p53 levels in CHD8-depleted cells, and whether this is attributable to CHD8’s transcriptional function or a secondary effect of general interference of CHD8 with transcription or chromatin reorganization. We further gained ideas for CHD8-regulated processes through our microarray analysis: The downregulation of USP11 as well as the upregulation of Tip60 would promote HR, a DNA repair pathway which we have not analyzed in detail so far. In particular, CHD8 knockdown upregulates MDC1, whose levels and binding to chromatin increase DNA damage signaling. Therefore, the levels and activity of ATM should be determined more directly than via the phosphorylation of Chk2, which we clearly observe in CHD8-depleted cells as first evidence for increased ATM-dependent signaling in our experiments. If this were the case, DNA damage signaling would lead to faster recovery after DNA damage. The preliminary result that CHD8-depleted cells resume proliferation faster after chemotherapy require replication. Further, established markers of eu- and heterochromatin may serve as a readout for the influence of CHD8 on chromatin organization.

Summarized, the data presented in this work does not only establish MK2 as a determinant in the gemcitabine response in pancreatic cancer cells, but challenges the strategy of applying checkpoint inhibitors as cancer therapeutics. Furthermore, our findings demonstrate for the first time a role for CHD8 in the DDR. Mechanistic details need to be analyzed to clarify the contradicting findings for CHD8 in cancer cell biology before one could exploit it as prognostic factor or drug target. Our results rather open a new door for the understanding of a chromatin response to DNA damage, which is well examined in hematological malignancies but is only beginning to be appreciated in terms of cancer promotion in solid tumors. The relaxation and restoration of chromatin is a tightly regulated and ATP-consuming process and therefore susceptible to dysregulation in cancers. It would

be no surprise if a balanced activation and silencing of CHD8 turned out to be a cellular mechanism for resistance as well as for the response to chemotherapy. More generally, instead of inhibiting a single molecule, it seems to be reasonable to target whole pathways and physiological cellular responses in order to account for genetic heterogeneity and the fast acquirement of drug resistance seen in pancreatic cancer. For this purpose, chromatin modifying enzymes are potential candidates as their loss of function broadly affects cellular homeostasis in terms of various signaling pathways.