Characterization of mutant strains with different Gdh1p noise and abundance profiles indicates that fitness in different ammonia conditions correlates with noise in Gdh1p expression. (A) Schematic of the genetic construct for building the GDH1 promoter mutant library. A region 500 nt upstream of the GDH1 coding sequence was amplified by mutagenic PCR, assembled with a selectable LEU2 marker, and reintegrated to construct a mutant library. (B) A random set of clones from the GDH1 promoter mutant library exhibit different values for Gdh1p mean abundance and noise. Noise is calculated as the square of the coefficient of variation (CV2). Sets of mutants with low, wildtype, and high mean abundances are indicated as blue, red, and green, respectively. Wildtype is indicated in yellow. All errors are within 5% of the reported values. (C) Relative fitness in high and low ammonia concentrations correlates with noise in Gdh1p expression (correlation coefficient = 0.83, R2 = 0.69). Relative fitness for select constant abundance-different noise mutant sets is reported as Wenv, or the ratio of fitness in 556 mM ammonia to fitness in 17 mM ammonia using the competition fitness assays. The wildtype relative fitness is equal to 1. Low Gdh1p noise is correlated with high fitness in ammonia-poor conditions and low fitness in ammonia-rich conditions (Wenv < 1). High Gdh1p noise is correlated with low fitness in ammonia-poor conditions and high fitness in ammonia-rich conditions (Wenv > 1). (D) Relative fitness in high and low ammonia concentrations does not correlate with Gdh1p abundance (correlation coefficient = 0.079, R2 = 0.0062). Relative fitness (Wenv) for a set of GDH1 promoter mutants with similar Gdh1p noise values (approximate CV2 = 0.65) and different abundance values is reported.