An Automated System For High-Throughput Longevity and Healthspan Discovery in Caenorhabditis elegans

Abstract

Over the last century, the study of aging biology has primarily been advanced through the development and use of animal models that share common molecular hallmarks with human aging. One major model system that has been widely used during the last several decades and through the present day is the roundworm Ceanorhabditis elegans. Its short lifespan, easy husbandry, and genetic tractability have allowed it to be easily adapted for studying the molecular biology of aging. For instance, research using C. elegans was used to discover the interplay between insulin signaling and rate of aging. However, while C. elegans research proved orders of magnitude more efficient (at the cost of a larger evolutionary gap) than using vertebrate models, such as mice or non-human primates, it’s use was still hamstrung by the necessity for manually collected lifespans. During the last decade, the lowering cost and increasing capabilities of digital microscopy and computer vision have led to researchers attempting to automate the most basic aging related experiment in C. elegans: the measurement of an individual’s lifespan relative to the population within which it resides. Some examples of this include the Automated Lifespan Machine, The WormMotel, as well as several different microfluidic platforms such as the NemaLife chip. These platforms sought to semi-automate or even fully-automate the collection and recording of lifespan measurements but often had large bottlenecks in their pipeline that precluded them from achieving the scale necessary to robustly advance the field of aging biology. This thesis will present the development and use of a novel robotics platform that, when paired with a suite of AI-powered computer vision, fully automates lifespan analysis of C. elegans.