Georgios Karras
Assistant Professor
The University of Texas MD Anderson Cancer Center
Department of Genetics
Research Interests protein homeostasis and disease predisposition aging and proteotoxic stress genome instability and biological robustness cancer therapy resistance A single mutation can increase the risk of disease hundred-fold. Mutations in central DNA repair genes such as BRCA1 predispose individuals to cancer by dramatically increasing the rate at which new mutations accumulate in normal cells. Given that mutations are generally detrimental to biological function, it remains unclear how extensive mutational burdens can be tolerated and even fixed in tumors and populations. Understanding how cancer cells minimize the cost of mutation and harness the acquired genetic variation to form tumors, metastasize and resist therapies will lead to more effective ways to combat cancers. We demonstrated that the molecular chaperone HSP90, a sentinel of protein homeostasis, “buffers” deleterious mutations in human cells. Proteins with HSP90-buffered mutations retain normal function at basal conditions but lose activity when the HSP90 function is reduced. HSP90 buffering may allow mutations to accumulate within tumors as well as in populations to shape the course of disease and aging. We are seeking strong candidates for two high-impact cross-disciplinary PhD projects aiming to understand the role of HSP90 buffering in tumor evolution and aging. Question 1: Which mutations are buffered by HSP90? A paradigmatic example of HSP90-buffering we identified involved a case of identical twins with Fanconi Anemia, which is a severe cancer predisposition and premature aging syndrome (Karras et al., Cell 2017). More recently, we showed that HSP90 buffers mutations in the major tumor suppressor BRCA1 (Gracia, Zhang et al., Mol Cell 2025). HSP90-buffered mutations in tumors can be targeted using combination therapies based on HSP90 inhibition. In addition, we used our high-throughput LUMIER platform to predict the severity of BRCA1 mutations based on the binding of the encoded protein variants to chaperones HSP90 and HSP70 (Gracia et al., Cell Rep 2024). It is currently unclear why HSP90 buffers certain mutations but not others. The candidate will use established functional genomics and cutting-edge proteomics methods to decipher the selectivity of HSP90 buffering in the context of genome instability and map the identified HSP90-buffered variation to disease phenotypes through mining publicly available genomics databases. Project 2: Which ecological stressors uncover HSP90-buffered variation? This project explores the molecular underpinnings of HSP90’s unusual environmental sensitivity.
Seemingly benign environmental proteotoxic stressors uncover HSP90-buffered mutations thereby inducing mutant-specific phenotypes. We showed that HSP90 has driven yeast adaptation to industrial niches by mediating ecologically relevant gene-environment interactions (Condic et al., Science 2024). At industrially relevant concentrations, ethanol, but not other fermentation-associated stressors, uncovers HSP90-buffered genetic variation in metabolism and strongly selects for gene duplications characteristic of domesticate beer and bread yeasts. We are investigating the role of HSP90 buffering in predisposition to alcohol-related cancers. Using budding yeast as a discovery platform, we identified diverse cancer-related metabolic and genetic factors that modulate HSP90’s buffering capacity. The candidate will study the mechanisms shaping the environmental sensitivity of HSP90 buffering and their evolutionary value in the tumor microenvironment and in normal tissues during aging. For this, we have established an approach integrating chemical genetics and spatial multi-omics in clinically relevant animal models.
Education & Training
Ph.D., Ludwig Maximilian University of Munich, 2010
Research Info
Protein Homeostasis and Cancer Evolution