A negative feedback loop at the nuclear periphery regulates GAL gene expression

Abstract

The genome is nonrandomly organized within the nucleus, but it remains unclear how gene position affects gene expression. Silenced genes have frequently been found associated with the nuclear periphery, and the environment at the periphery is believed to be refractory to transcriptional activation. However, in budding yeast, several highly regulated classes of genes, including the GAL7-10-1 gene cluster, are known to translocate to the nuclear periphery concurrent with their activation. To investigate the role of gene positioning on GAL gene expression, we monitored the effects of mutations that disrupt the interaction between the GAL locus and the periphery or synthetically tethered the locus to the periphery. Localization to the nuclear periphery was found to dampen initial GAL gene induction and was required for rapid repression after gene inactivation, revealing a function for the nuclear periphery in repressing endogenous GAL gene expression. Our results do not support a gene-gating model in which GAL gene interaction with the nuclear pore ensures rapid gene expression, but instead they suggest that a repressive environment at the nuclear periphery establishes a negative feedback loop that enables the GAL locus to respond rapidly to changes in environmental conditions.

FIGURE 1:

Determinants of peripheral positioning of the GAL locus. (A) The GAL locus is at the nuclear periphery in the absence of glucose. Wild-type cells expressing the LacO/LacI system and dsRED-HDEL were grown in synthetic medium containing 2% raffinose (SRaf) medium to mid-log phase and then maintained in raffinose or shifted to 2% glucose (SGlu), 2% galactose (SGal), or 2% glycerol (SGly) for 2 h prior to imaging. Percentage of cells with peripheral localization was scored as described in Materials and Methods. (B) GAL locus localization in wild-type and nup1Δ or ada2Δ mutant cells. The percentage of cells with the GAL locus at the nuclear periphery in SGlu, SGal, or SRaf for wild-type, nup1Δ, and ada2Δ cells was determined by microscopy. Cells were grown and scored as described for A. (C) Localization of the GAL locus after transcriptional inhibition. Wild-type and rpb1-1 mutant cells were grown continuously in SGal or SRaf medium at 25°C and then shifted to 37°C. Cells were imaged and scored as described for A following 2 h at 37°C. Error bars represent the SEM for at least three independent experiments in which at least 100 cells were scored for each condition or time point. An unpaired, two-tailed Student's t test was used to determine statistical significance for all microscopy experiments. Statistical significance is as follows: p > 0.05 is not significant (NS); *0.05 > p > 0.01; **0.01 > p > 0.001; ***p < 0.001. p values represent a comparison between percentage of cells at the nuclear periphery in glucose to other selected media in wild-type or mutant cells.

FIGURE 2:

The nuclear periphery inhibits GAL1 mRNA expression. (A) Changes in GAL1 mRNA expression in yeast with disrupted peripheral localization of the GAL locus. Wild-type, ada2Δ, or nup1Δ mutant cells were grown in YP plus 2% glucose (YPD) to mid-log phase. GAL1 mRNA expression was induced by galactose. GAL1 mRNA levels were measured at the indicated times by quantitative RT-PCR. GAL1 mRNA levels were normalized to levels of a control gene, TFC1, and the fold change in expression was calculated relative to the baseline expression at the zero time point for each strain. Error bars represent SEM for three independent experiments. (B) GAL1 mRNA turnover in galactose in wild-type and ada2Δ mutant cells. rpb1-1 and rpb1-1 ada2Δ cells were grown in YPGal medium at 25°C and then shifted to 37°C to inactivate transcription. GAL1 mRNA levels were measured at the indicated time points by qRT-PCR and plotted as a function of time following the shift to the nonpermissive temperature. GAL1 mRNA levels were normalized to the RNA pol III transcript SCR1 and expressed relative to the level of transcript at the zero time point, defined as 1.0. (C) Gal1 protein levels increase in ada2∆ cells. Cells were grown in raffinose medium and induced with 2% galactose for the indicated time, and Gal1-GFP protein levels were analyzed using flow cell cytometry. Mean GFP intensity of the population (in arbitrary units) is plotted as a function of time in galactose. Error bars represent SEM for three independent experiments.

FIGURE 3:

The nuclear periphery specifically regulates expression of galactose-induced genes. (A) Schematic of the GAL locus and adjacent genes on chromosome II, including the integration site of 256 copies of the LacO repeats. (B) mRNA expression levels of GAL locus genes (GAL1, GAL7, GAL10) and neighboring genes (FUR4 and KAP104) in yeast lacking the gene–periphery tether. Wild-type, ada2Δ, or nup1Δ mutant cells were grown in YPD medium to mid-log phase and shifted to YPGal medium for 1 h. mRNA levels were measured by qRT-PCR, and the expression of each gene was normalized to the control gene, TFC1. Fold change of mRNA levels for all genes in ada2Δ or nup1Δ strains was calculated relative to expression in the wild-type strain, defined as 1.0. (C) Kinetics of GAL1 expression with constitutive peripheral tethering. GAL1 gene expression, measured by qRT-PCR in wild-type and nup1Δ strains containing the LacO repeats integrated near the GAL locus with and without the Nup2-LacI gene tether. Cells were grown in YPD medium, and GAL1 expression was induced with 2% galactose for the indicated times. GAL1 mRNA levels were normalized to levels of the control gene TFC1, and the fold change in expression was calculated relative to the baseline expression at the zero time point for each strain. (D) mRNA expression levels of GAL locus genes and neighboring genes with the Nup2-LacI gene tether. mRNA levels were detected by qRT-PCR following 1 h induction with galactose. The expression of each gene was normalized to the control gene TFC1, and the fold change of mRNA levels for all genes was calculated relative to expression in the GAL::LacO strain, defined as 1.0. Error bars represent SEM for three independent experiments.

FIGURE 4:

Repression of GAL1 is delayed with a disrupted gene–periphery tether. Wild-type, nup1Δ, and ada2Δ strains were grown at room temperature in YP plus 2% raffinose to mid-log phase. GAL1 mRNA expression was induced by galactose addition for 2 h and then inhibited by the addition of glucose for the indicated time. GAL1 mRNA levels were monitored by qRT-PCR and normalized to the control gene, ACT1. The fold change is calculated relative to the transcript levels for each strain at the zero time point, defined as 1.0.