Studying disease-linked phenotypes using haploid genetics

Abstract

Although genes are unequivocally important for the development of both common and rare human diseases, the connection between the genotype (an individual’s genetic makeup) and phenotype (an individual’s observable traits) is often ill-defined. Even genetic disorders caused by a defect in only a single gene often manifest themselves differently in affected patients, indicating that the precise consequence of genetic mutations is often difficult to predict. This phenotype-genotype conundrum is a result of genes not functioning in isolation, but in complex genetic networks where gene function can be modulated by the action of other genes or the environment. A powerful approach to understand the genetic contribution to a phenotype is to generate mutations in model organisms. Large-scale mutation studies in yeast have illuminated some of the complex genetic architecture, but owing to technical constraints mutagenesis of human cells has long been unachievable. Central to this work is the application of haploid human mutant cells to study gene-environment and gene-gene interactions and their contribution to disease-relevant phenotypes. We describe a method for mapping mutations in pools of mutant cells following a phenotypic selection; typically mutants that survive exposure to otherwise noxious stimuli. This is applied to study the interactions between genes and the environment, focusing on a number of bacterial toxins that exploit the function of human genes in order to enter and cause damage to cells. In addition, the mechanism of resistance to multiple clinically-used drugs is examined. This same method is applied to identify host factors required by Rift Valley Fever Virus for infecting human cells. Further optimizations to this approach have made it possible not only to identify mutants which are favored under particular conditions, but also those which are under negative selection. We have applied this to define a set of around 2.000 genes required for cultured human cells to grow. This includes a number of genes which were uncharacterized, despite their essential function in basic cellular processes. The ability to identify essential genes furthermore enables examining gene essentiality specific to particular genetic backgrounds, referred to as genetic interactions. Mapping the essential genes in cell lines deficient for different genes uncovered a synthetic lethality network focused on the human secretory pathway. Comparable to yeast, human genes frequently engage in genetic interactions, which implies that many non-essential genes will become essential in the absence of other genes. The ability to efficiently map mutations in populations of fixed haploid cells is further leveraged to examine the genetic regulators of intracellular phenotypes. For a dozen of diverse biological traits, such as oncogenic pathway activation or organelle size, we have identified hundreds of genes that impact the measured phenotype. Based on this, we construct a preview of a phenotypic map of a human cell, and observe that pleiotropy is frequent for human genes, as close to half of the identified regulators affect multiple phenotypes. By iterative screens in genetic backgrounds, suppressors can be identified which restore the affected intracellular phenotypes. This enables the systematic dissection of a cellular process to gain mechanistic understanding.