Mutator hormesis

Project leader: Berit Jungnickel / Helmut Pospiech
Doctoral candidate: Gleice Borges

Background and previous work

DNA replication in normal human cells occurs with exceptional accuracy, mainly due to the preferred chemical bonding of complementary bases, the proofreading function of replicative polymerases  (Pols) δ and ε as well as the post-replicative mismatch repair (MMR) system^1–3. Conversely, many human cancers exhibit tens of thousands of alterations in their nuclear DNA sequences, indicating that they express a mutator phenotype⁴. Accordingly, mutations in any gene involved in the genetic stability maintenance would create a series of alterations throughout the genome, which could confer cancer cells selective advantages and allow adaptation and further clonal expansion⁵. However, the considerable number of mutations in these cells would also turn them more prone to death when mutation rate is increased, characterizing the error catastrophe^5,6. On the other hand, decrease in mutation rate could delay tumor evolution¹.

Mutations in the POLE gene that encodes for Pol ε have been associated with hypermutative cancers such as melanomas and certain colorectal carcinomas^7,8. Genetic alterations responsible for this phenotype are frequently found within or regulating the exonuclease domain of Pol ε, an important region for its proofreading function⁷. In our previous work, we identified increased mutation rate in the exonuclease-deficient Pol ε and Pol δ when compared to WT cells by performing fluctuation assays (not published). Similar results were seen with two cancer-predisposing point mutations (P286R, L474V) in the exonuclease domain of Pol ε. Moreover, by performing colony formation assay, we detected increased number of colonies at low concentrations of fluorouracil (5-FU) in exonuclease-deficient Pol ε when compared to heterozygous cells. However, at higher concentrations, a decrease in colonies was observed. Additionally, it was early observed that cancer cells with extreme high mutation rates (due to defects in MMR factors or Pol proofreading) show surprisingly high chromosomal stability⁹. Taken together, the concept to regard a mutator phenotype in malignant cells as a hormetic response is an interesting novel approach that also opens new avenues for the treatment of many human cancers. Once these cells exceed the “hormetic dose of mutagenesis”, they may rely on certain metabolic pathways for standing such toxically high rates of mutation within their genomes.

Specific aims and working program

We would like to identify and characterize genes involved in metabolic pathways beneficial or detrimental for survival of cancer cells exhibiting a hypermutator phenotype. To this end, by a genome-wide knock-out screen of a human cancer cell model, we will identify potential genes and underlying pathways beneficial or detrimental for hypermutating cells. We will take advantage of the human Genome-scale CRISPR Knockout (GeCKO) pooled library to perform a positive/negative selection screen comparing the nearly haploid human HAP1 cell lines with a WT and exonuclease-deficient Pol ε context, respectively. In addition, we will study the drift of targeted genes for both cell populations at different time points after the genome-wide knockout. We will also study the effect of the loss of the identified candidate genes in terms of mutation rate of the exonuclease-deficient Pol ε versus WT cells, and will characterize the most prominent pathways increasing or reducing survival of the hypermutated cells compared to the WT in detail.