The Impact of Collateral Evolution on Optimal Dosing Strategies and Evolution on Paired Fitness Landscapes
Year: 2020
Type: article
Abstract: Drug resistance is an ever-growing threat to successful treatment of bacterial, cancer and viral infections. As pathogens and cancers continue to find evolutionary solutions to the drugs we treat them with, scientists have begun to focus on more evolutionary-based therapies such as drug cycling. These therapies aim to constrain or control evolution in a particular way such that intractable resistance never evolves. In the same vein, recent work has revealed collateral sensitivity as a promising avenue to guide evolution away from untreatable resistance states. Collateral evolution occurs when a population evolves resistance to the selecting drug and this mechanism of resistance confers "collateral" effects to different drugs it is not exposed to. In this work we show how collateral profiles might be used to slow the acquisition to resistance in a simplified laboratory-based evolution experiment. We demonstrate that intuitive cycling protocols often fail over long time periods, whereas mathematically optimized protocols maintain long-term sensitivity at the cost of transient periods of high resistance. We then extend this work to include nonantibiotic stressors such as pH, salt and food preservatives. This extension highlights that more work is necessary to understand the role these common environments have on the development of multidrug resistance. Finally, using the well-known fitness landscape paradigm, we explore how collateral effects influence the evolutionary dynamics of a pair of landscapes with tunable correlations. We show that alternating evolution in highly correlated environments can lead to higher mean fitness than evolution in either landscape alone, while alternating between two anti-correlated landscapes results in a lower mean fitness. We demonstrate this is due to the location and number of shared maxima between the two correlated landscapes, which change as a function of ruggedness (epistasis) and paired landscape correlation. Taken together, these results begin to answer many of the important questions required to translate collateral sensitivity into clinical treatments.
Author: Jeffrey Maltas
Language: English
Subfield: Genetics
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W3111005349
