The tumor suppressive transcription factor p53 is frequently inactivated in cancer cells by missense mutations that cluster in the DNA binding domain. 30% hit mutational hotspot residues, resulting in a complete loss of transcriptional activity and mutant p53-driven chemotherapy resistance. Of the remaining 70% of non-hotspot mutants, many are partial loss-of-function (partial-LOF) mutants with residual transcriptional activity. The therapeutic consequences of a partial-LOF have remained largely elusive. Using a p53 mutation engineered to reduce DNA binding, we demonstrate that partial-LOF is sufficient to enhance oncogene-driven tumorigenesis in mouse models of lung and pancreatic ductal adenocarcinoma and acute myeloid leukemia. Interestingly, mouse and human tumors with partial-LOF mutations showed mutant p53 protein accumulation similar as known for hotspot mutants. Different from the chemotherapy resistance caused by p53-loss, the partial-LOF mutant sensitized to an apoptotic chemotherapy response and led to a survival benefit. Mechanistically, the pro-apoptotic transcriptional activity of mouse and human partial-LOF mutants was rescued at high mutant protein levels, suggesting that accumulation of partial-LOF mutants enables the observed apoptotic chemotherapy response. p53 non-hotspot mutants with partial-LOF, therefore, represent tumorigenic p53 mutations that need to be distinguished from other mutations because of their beneficial impact on survival in a therapy context.

Background Impaired p53 function is one of the central molecular features of a tumor cell and even a partial reduction in p53 activity can increase the cancer risk in mice and men. From a therapeutic perspective it is noteworthy that tumor cells often become addicted to the absence of p53 providing a rationale for developing p53 reactivating compounds to treat cancer patients. Unfortunately, many of the compounds that are currently undergoing preclinical and clinical testing fail to fully reactivate mutant p53 proteins, raising the crucial question: how much p53 activity is needed to elicit a therapeutic effect? Methods We have genetically modelled partial p53 reactivation using knock-in mice with inducible expression of the p53 variant E177R. This variant has a reduced ability to bind and transactivate target genes and consequently causes moderate cancer susceptibility. We have generated different syngeneically transplanted and autochthonous mouse models of p53-deficient acute myeloid leukemia and B or T cell lymphoma. After cancer manifestation we have activated E177R expression and analyzed the in vivo therapy response by bioluminescence or magnetic resonance imaging. The molecular response was further characterized in vitro by assays for gene expression, proliferation, senescence, differentiation, apoptosis and clonogenic growth. Results We report the conceptually intriguing observation that the p53 variant E177R, which promotes de novo leukemia and lymphoma formation, inhibits proliferation and viability, induces immune cell infiltration and triggers cancer regression in vivo when introduced into p53-deficient leukemia and lymphomas. p53-deficient cancer cells proved to be so addicted to the absence of p53 that even the low-level activity of E177R is detrimental to cancer growth. Conclusions The observation that a partial loss-of-function p53 variant promotes tumorigenesis in one setting and induces regression in another, underlines the highly context-specific effects of individual p53 mutants. It further highlights the exquisite sensitivity of cancer cells to even small changes in p53 activity and reveals that changes in activity level are more important than the absolute level. As such, the study encourages ongoing research efforts into mutant p53 reactivating drugs by providing genetic proof-of-principle evidence that incomplete p53 reactivation may suffice to elicit a therapeutic response.

Spermatogenesis in Drosophila melanogaster is characterized by a specific transcriptional program during the spermatocyte stage. Transcription of thousands of genes is regulated by the interaction of several proteins or complexes, including a tTAF-containing TFIID variant, tMAC, Mediator, and chromatin interactors, e.g., bromodomain proteins. We addressed how distinct subsets of target genes are selected. We characterized the highly similar proteins tPlus3a and tPlus3b, which contain a Plus3 domain and are enriched in the testis, mainly in spermatocytes. In tPlus3a and tplus3b deletion mutants generated using the CRISPR/Cas9 system, fertility was severely reduced and sperm showed defects during individualization. tPlus3a and tPlus3b heterodimerized with the bromodomain protein tBRD-1. To elucidate the role of the tPlus3a and tPlus3b proteins in transcriptional regulation, we determined the transcriptomes of tplus3a-tplus3b and tbrd-1 deletion mutants using next-generation sequencing (RNA-seq) and compared them to that of the wild-type. tPlus3a and tPlus3b positively or negatively regulated the expression of nearly 400 genes; tBRD-1 regulated 1,500 genes. Nearly 200 genes were regulated by both tPlus3a and tPlus3b and tBRD-1. tPlus3a and tPlus3b activated the Y-chromosomal genes kl-3 and kl-5 , which indicates that tPlus3a and tPlus3b proteins are required for the function of distinct classes of genes. tPlus3a and tPlus3b and tBRD-1 repress genes relevant for seminal fluid and heat shock. We hypothesize that tPlus3a and tPlus3b proteins are required to specify the general transcriptional program in spermatocytes.

Background In vivo gene editing of somatic cells with CRISPR nucleases has facilitated the generation of autochthonous mouse tumors, which are initiated by genetic alterations relevant to the human disease and progress along a natural timeline as in patients. However, the long and variable, orthotopic tumor growth in inner organs requires sophisticated, time-consuming and resource-intensive imaging for longitudinal disease monitoring and impedes the use of autochthonous tumor models for preclinical studies. Methods To facilitate a more widespread use, we have generated a reporter mouse that expresses a Cre-inducible luciferase from Gaussia princeps (GLuc), which is secreted by cells in an energy-consuming process and can be measured quantitatively in the blood as a marker for the viable tumor load. In addition, we have developed a flexible, complementary toolkit to rapidly assemble recombinant adenoviruses (AVs) for delivering Cre recombinase together with CRISPR nucleases targeting cancer driver genes. Results We demonstrate that intratracheal infection of GLuc reporter mice with CRISPR-AVs efficiently induces lung tumors driven by mutations in the targeted cancer genes and simultaneously activates the GLuc transgene, resulting in GLuc secretion into the blood by the growing tumor. GLuc blood levels are easily and robustly quantified in small-volume blood samples with inexpensive equipment, enable tumor detection already several months before the humane study endpoint and precisely mirror the kinetics of tumor development specified by the inducing gene combination. Conclusions Our study establishes blood-based GLuc monitoring as an inexpensive, rapid, high-throughput and animal-friendly method to longitudinally monitor autochthonous tumor growth in preclinical studies.