Genome-wide scan reveals that genetic variation for transcriptional plasticity in yeast is biased towards multi-copy and dispensable genes
Introduction
Understanding the genetic bases of complex diseases and adaptive traits are among the fundamental goals of medical and evolutionary genetics. The environment is usually seen as the agent of natural selection, selecting the fittest genotypes in a given population. But the environment is far from constant. Moreover, genetic and environmental effects may interact with each other and may produce complex evolutionary trajectories. For instance, phenotypic changes in response to environmental changes–known as phenotypic plasticity–can themselves be genetically variable. Also, the evolution of robustness against environmental perturbations can lead to the evolution of robustness also against genetic perturbations, consequently redrawing the maps that connect genotype and environment with phenotype and fitness (Waddington, 1942, Hartman et al., 2001, Meiklejohn and Hartl, 2002). Consideration of both genetic and environmental perturbations is therefore necessary for a more complete understanding of evolutionary processes (Levins, 2004).
Genetic variation for phenotypic plasticity emerges from the non-additive interaction between the genotypes of organisms and variation in the environment. In the presence of such interactions, different genotypes respond differently to the same environmental change, so there is genotype-by-environment interaction (GEI). Sensitivity to therapeutic drugs in individuals with some genotypes but not others is a manifestation of GEI in pharmacogenomics. The visualization of GEI can be facilitated through the use of norms of reaction—a mapping of the phenotype onto the genotype as a function of the environment, depicted graphically as a plot of phenotypic values against environmental values (Stearns, 1992, Fig. 1). GEI refers to genetic variation in the shape of those norms of reaction. Different hypothetical norms of reaction are illustrated in Fig. 1, in which each color represents a different genotype. Panel (A) shows a trait with genetic variation but no phenotypic plasticity, and panel (B) shows a trait with phenotypic plasticity but no genetic variation. The trait in panel (C) exhibits both genetic variation and phenotypic plasticity, but there is no genetic variation in phenotypic plasticity: each genotype responds in the same way to each environment. The only trait in Fig. 1 that shows genetic variation for phenotypic plasticity (GEI) is that depicted in panel (D), where the norms of reaction of the different genotypes are not parallel.
The machinery of gene regulation can exhibit an exquisite response to the environment because it has evolved to allow transcription to be induced, repressed, or modulated in response to environmental cues (Suiter et al., 2003). The transcriptional response can be at the level of individual cells, tissues, or the whole organism. In fact, a large fraction of phenotypic diversity within species may result from phenotypic plasticity controlled at the level of differential gene expression (Pigliucci, 1996, Abouheif and Wray, 2002). Studying gene regulation is therefore key to our understanding of the genetic and environmental dissection of complex phenotypes, and this in the perspective of better understanding phenotypic evolution, which is a grand challenge facing evolutionary biology (Singh, 2003). To this end, microarray technologies have become valuable tools because they enable the survey of gene expression at a genomic scale and this quantitatively (e.g., Gibson, 2002).
In order to quantify the extent of genetic variation for transcriptional plasticity, we studied gene expression in Saccharomyces cerevisiae. This unicellular fungus is particularly well suited for such an experiment because the time scale of environmental changes can be much shorter than the generation time (e.g., Gasch and Werner-Washburne, 2002), a condition known to favor the evolution of phenotypic plasticity (Schlichting and Smith, 2002). Additionally, transcriptional responses in microorganisms are at the organismic-level and do not include confounding factors of differential responses in distinct tissues or organs. Finally, the metabolism of microorganisms is relatively tightly linked to genetic regulatory gene networks whose transcriptional state can be estimated by assaying mRNA levels. In yeast, genetic variation in gene expression in a single environment has been documented (e.g. Cavalieri et al., 2000, Townsend et al., 2003), as have changes in gene expression in different environments (e.g., Gasch et al., 2000, Causton et al., 2001). There is also evidence that different strains show different responses to chemical agents (Fay et al., 2004). What is lacking, however, is a large-scale assay for GEI, which would reveal how much genetic variation is present in the transcriptional response to environmental changes and what genetic characteristics of this species favor or limit the diversity of transcriptional responses to environmental triggers. Because genetic variation for phenotypic plasticity emerges from the interaction between genotypes and environments, it can be assayed only by examining the transcriptional response of multiple genotypes across several environments. This can now be carried out on a genomic scale.
In this study, we used cDNA microarrays to examine genetic variation for phenotypic plasticity in genome-wide gene expression in six isolates of S. cerevisiae grown in four different environments, either in standard rich medium or any of three alternative media simulating environmental stress. The specific objective was to reveal which general classes of genes show genetic variation for phenotypic plasticity. We also hoped to identify factors that might constrain or favor the diversity of norms of reaction in the genome of S. cerevisiae.
Section snippets
Genotypes and environments
Six diploid prototrophic strains were studied: Ds288c, EM93, Sgu52, Sgu407, Sgu421, and Sg60. Ds288c is a diploid heterothallic strain derived from a cross of two S288c derivatives isolated in the 1980s. EM93 was originally isolated from a fig in Mercedes, California, in the early days of yeast genetics (Mortimer and Johnston, 1986). The other strains are from the Polsinelli collection of Italian yeast isolates: Sgu52, Sgu407, and Sgu421 were isolated from grapes in the Chianti region, and Sg60
Variation in gene expression across environments
Gene expression was estimated in replicate experiments in which each strain of yeast was grown in four environmental conditions: aerobic growth in YPD medium, a typical rich laboratory condition; anaerobic growth in synthetic wine must (SWM), which mimics wine fermentation conditions; and either 4 or 24 h of anaerobic growth with nitrogen starvation (NS) in medium in which proline was the main source of nitrogen. These environmental conditions simulate the progression of environments that yeast
Acknowledgments
We would like to thank Nadia Aubin-Horth and Cristian Castillo-Davis for comments on an earlier version of the manuscript and members of the Hartl lab for discussions on the work. We would also like thank Ping Ma for helpful discussions about statistical issues and Hao Wu for information about the use of MAANOVA, Keith Morneau for lab assistance and Mario Polsinelli for providing the strains for this work. CRL is a Frank Knox Memorial fellow at Harvard University and is supported by graduate
References (48)
- et al.
Genetic variation in radiation-induced expression phenotypes
Am. J. Hum. Genet.
(2004) - et al.
Unipolar cell divisions in the yeast S. cerevisae lead to filamentous growth: regulation by starvation and RAS
Cell
(1992) - et al.
A single mode of canalization
Trends Ecol. Evol.
(2002) - et al.
The effects of phenotypic plasticity on genetic correlations
Trends Ecol. Evol.
(1991) - et al.
Genomic analysis of essentiality within protein networks
Trends Genet.
(2004) - et al.
Evolution of the gene network underlying wing polyphenism in ants
Science
(2002) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs
Nucleic Acids Res.
(1997)Influence of nitrogen on yeast fermentation of grapes
- et al.
The landscape of genetic complexity across 5700 gene expression traits in yeast
Proc. Natl. Acad. Sci. U. S. A.
(2005) - et al.
GeneMerge-post-genomic analysis, data mining, and hypothesis testing
Bioinformatics
(2003)
Remodeling of yeast genome expression in response to environmental changes
Mol. Biol. Cell
Manifold anomalies in gene expression in a vineyard isolate of Saccharomyces cerevisiae revealed by DNA microarray analysis
Proc. Natl. Acad. Sci. U. S. A.
Statistical tests for differential expression in cDNA microarray experiments
Genome. Biol.
Improved statistical tests for differential gene expression by shrinking variance components estimates
Biostatistics
Preferential duplication of conserved proteins in eukaryotic genomes
PLoS Biol.
Population genetic variation in gene expression is associated with phenotypic variation in Saccharomyces cerevisiae
Genome Biol.
The genomics of yeast responses to environmental stress and starvation
Funct. Integr. Genomics
Genomic expression programs in the response of yeast cells to environmental changes
Mol. Biol. Cell
Microarrays in ecology and evolution: a preview
Mol. Ecol.
Role of duplicate genes in genetic robustness against null mutations
Nature
Comparative genomics of centrality and essentiality in three eukaryotic protein-interaction networks
Mol. Biol. Evol.
Transcriptional regulatory code of a eukaryotic genome
Nature
Principles for the buffering of genetic variation
Science
Prm1p, a pheromone-regulated multispanning membrane protein, facilitates plasma membrane fusion during yeast mating
J. Cell Biol.
Cited by (91)
Plasticity via feedback reduces the cost of developmental instability
2020, Evolution Letters
- 1
Present address: Department of Biological Sciences, Stanford University, Stanford CA, USA.