SASKit

Senescence-Associated Systems diagnostics Kit for cancer and stroke

The alliance will receive extensive additional funding until mid-2024.
With aging comes cellular senescence, and (multi-)morbidity. Cellular senescence is a key driver of an interconnected disease network including cancer and stroke. We wish to utilize systems modeling and bioinformatics, learning from omics and other lab data, to design and develop a biomarker + software kit with a focus on measuring and interpreting senescence-related signatures for precise (and early) diagnosis, prognosis, and, ultimately, therapy, of pancreatic cancer and ischemic stroke/thromboembolism. We build upon publications describing how cellular senescence and the senescence-associated secretory phenotype are directly involved in the comorbidity of pancreatic cancer, ischemic stroke, and more generally, of cancer and coagulation problems. We are conducting observational human studies for pancreatic cancer and ischemic stroke, measuring senescence markers in particular, preparing the power analyses and the companion diagnostics for larger interventional trials of, e.g., patient-specific natural-compound senolytics such as quercetin. We pre-registered the human study (Henze et al, BMJ Open, 2020), and as specified therein, we expect to perform the main analysis in Summer/Fall 2023. For pancreas, we utilize co-culture studies of cancerous and stellate cells, and a mouse cancer allograft model. For stroke, we study brain slices and stroke recovery in mice. In both cases, to mimic the human cohorts, we study young and old wild-type mice as well as senescence-prone strains already investigated in our past ROSAge GerontoSys project; data and tissues from then provide a valuable reference. High-throughput gene expression and luminex protein quantification data are taken from blood and tissue of mice, and from blood of human, to allow the bioinformatics to extrapolate protein expression and pathway activation for the inaccessible tissue of humans, providing optimized input to machine learning of the best sets of biomarkers. Biomarker learning is also aided by sensitivity analyses based on dynamical models, which are in turn based on integrating mechanistic insights into disease and senescence based on public and consortium data.

^{Project overview. Pancreatic cancer and ischemic stroke are studied in humans and mice, including senescence-prone (“seno”) strains already studied during a predecessor BMBF project (ROSAge). For machine learning and modeling with the aim of a diagnostic/prognostic biomarker kit, human and mouse data are then integrated by the parallelogram approach that extrapolates interspecies data if three of four corresponding data sets are known. Dynamical models are developed specifically with a focus on PAI-1, CDK5 and p16/p21 (CDKN2A/1A). ẞ-gal denotes the senescence associated (SA) form of ẞ-galactosidase. The figure also lists the core measurements done by all subprojects (center), and the measurements done only by some subprojects (below and above the parallelograms).} 

Information on the funding can be found on the BMBF website.