Research interests

The DNA damage response in chemotherapy

The basic strategy in chemotherapy mostly consists in the induction of DNA damage. In general, tumor cells are more susceptible towards such damage than normal cells. As a consequence, they respond with cell death or permanent cell cycle arrest more often. The mechanistic reasons for this phenomenon are not fully clarified, but they appear to consist in the signaling cascades induced by DNA damage. We are trying to describe such signaling cascades, enabling their therapeutic manipulation. Thus, our goal is to predict the outcome of chemotherapy more exactly than currently possible, and to increase the efficacy of chemotherapy by interfering pharmacologically with DNA damage-induced signals.

DNA Fibre Assay The tumor suppressor p53 is at the center of our   
 previous investigations. P53 is posttranslationally modified 
 and activated in response to DNA damage. Thus, p53 
 represents an integrator of DNA damage signaling, 
 potentially inducing cell death. Since p53 acts as a 
 transcription factor, we first characterized the activating 
 mechanisms for p53-inducible genes. For instance, we   
 found that a polymorphic microsatellite that emerged recently
 during evolution serves as a p53-binding promoter element 
 (Contente et al., 2002 and 2004). In addition, we identified 
 the p53 target gene p21 as a central mediator of p53- dependent transcriptional repression (Löhr et al., 2004).

Initially, p53 appeared primarily as an effector of programmed cell death upon DNA damage. This implied that the loss of intact p53 in tumors led to therapeutic resistance in any case. However, we found that the pharmacological activation of p53 can be protective for a cell. This is especially true regarding the use of nucleoside analogues as therapeutics, e. g. gemcitabine and cytosine arabinoside. Possibly, this principle might help to avoid unwanted side effects in future therapeutic regimens. In contrast, tumor cells with mutant p53 are not expected to be influenced in their response by a pharmacological p53 activator (Kranz et al., 2008). Thus, the tumor suppressor p53 does not uniformly act as a trigger of cell death, but appears to secure the survival of cells in some cases. The choice of suitable chemotherapeutics may thus be based on the p53 status, in addition to other parameters.

Targeted manipulation of replicative stress for selective elimination of tumor cells

Tumor cells display the signs of DNA damage (e. g. the accumulation of phosphorylated histone 2AX) even at their earliest detectable stages. The reason appears to consist in a tendency of those cells to undergo a transition to S phase prematurely, resulting in replication errors. Normal cells show much less of such spontaneous DNA damage, even when proliferating fast. This can be taken advantage of in cancer therapy. Many conventional chemotherapeutics increase replicative stress. This is most obvious when using nucleoside analogues, since those are incorporated in replicating DNA almost exclusively during S phase, thus leading to the collapse of replication forks. However, additional chemotherapeutics, e. g. topoisomerase inhibitors and alkylating agents, are most potent in cells that are trying to replicate their damaged DNA.

We are conducting siRNA screens to investigate the interference of cellular signaling pathways with replicative stress. To this end, we are transfecting cells in a 96 well format, knocking down signaling mediators, such as kinases, ubiquitin ligases, or parts of the DNA damage response complexes. Transfection is carried out using an automated, robot-supported pipetting system, to ensure high throughput and equal assay conditions. Subsequently, we expose the cells to replicative stress, e. g. by treating them with nucleoside analogues or by ultraviolet irradiation. As readout, we are using quantitative immunofluorescence, supported by automated microscopy and image analysis. In this way, we are quantifying e. g. DNA damage induced histone phosphorylation of ca. 1000 individual cells per well. Regarding siRNAs per gene and ca. 500 genes investigated, this corresponds to the evaluation of 1.5 million single cells. The distribution of the DNA damage response intensity for each knockdown allows us to define the detailed influence of the corresponding gene to chemosensitivity. This approach is completed by the use of non-invasive translucent microscopy on living cells, measuring cell proliferation and cell death at different time points upon treatment. In this way, we determine how the expression of single genes determines short- and long-term survival (“clonogenic survival”). The gene products that contribute to the level of replicative stress and cell death represent promising targets for chemosensitization. We have already identified modulators of chemoresponse by these approaches, and we are currently working at their mechanistic basics.

microRNAs as effectors and regulators of tumor suppression

microRNAs represent a novel class of gene products. They do not encode proteins but can regulate the synthesis of other proteins, e. g. by interaction with the corresponding mRNA molecules. We have investigated how microRNAs form part of the tumor suppression machinery. Firstly, we identified microRNAs that can be induced by the tumor suppressor p53. Besides miR-34a, that was also identified by other investigators, we found microRNAs 192, 194, and 215 as p53-inducible. These microRNAs contribute to p53-induced cell cycle arrest (Braun et al., 2008; Georges et al., 2009). In a similar approach, we investigated microRNAs that are increased by the oncogenic transcription factor E2F1. This was found to be true in case of miR-449. However, miR-449 does not represent an oncogenic microRNA. Instead, miR-449 can mediate apoptosis, in a manner analogous to miR-34a. Apparently, the oncogenic activity of E2F1 can be limited in this way. In the presence of excess E2F1 activity, the overexpression of miR-449 appears to induce cell death. (Lizé et al., 2010).

p53-homologues as tumor suppressors in the germ line

p53 is not the only tumor suppressor of its kind. Rather, the cell synthesizes two additional paralogues to p53, i. e. p63 and p73. In earlier studies, we found that the latter two gene products, in contrast to p53, are not inactivated by certain oncoviral proteins (e. g. Roth et al., 1998, Dobbelstein et al., 1999 und 1998). This functional difference is understandable in the light of the fact that p63 and p73 are frequently found in aminoterminally truncated isoforms, lacking the principal transactivation domain. Thus, those deltaN proteins bind to promoters but generally cannot activate the adjacent genes. They do not act as tumor suppressors but can even contribute to oncogenesis.

It was particularly surprising to investigate the p63 gene structure. In humans and great apes (hominids), we found a long terminal repeat from the endogenous retrovirus ERV9, integrated upstream of the coding sequence. This LTR has strong promoter activity and mediates the expression of a novel, transactivating gene product, especially in testicular germ cells. The function of this testicular p63 protein apparently consists in fortifying cell death upon DNA damage. This increase in genomic surveillance may have contributed to the evolution of the complex genome organization in hominids. These investigations also revealed that the vast majority of testicular cancers loose p63 expression. However, pharmacological inhibition of histone deacetylases (HDACs) can rescue the expression of proapoptotic p63 isoforms. Possibly, the inhibition of HDACs, in combination with conventional chemotherapy, may represent a means to optimize the therapy of testicular carcinomas (Beyer et al., 2011).