Live-cell analysis of IMPDH protein levels during yeast colony growth provides insights into the regulation of GTP synthesis

Abstract

The purine nucleotides ATP and GTP are made from the common precursor inosine monophosphate (IMP). Maintaining the correct balance of these nucleotides for optimal cell growth is controlled in part by the enzyme IMP dehydrogenase (IMPDH), which catalyzes the first dedicated step of GTP biosynthesis. The regulation of IMPDH mRNA and protein levels in the yeast S. cerevisiae grown in liquid culture has been studied in some detail, but regulation of IMPDH protein under conditions of cellular crowding on a solid substrate has not been examined. Here, we report real-time, live-cell analysis of the accumulation of the Imd2 isoform of IMPDH in yeast cells forming a monolayer colony in a microfluidic device over a 50-hour time course. We observe two distinct phases of increased Imd2 accumulation: a guanine-insensitive phase early in outgrowth and a guanine-sensitive phase later, when cells become crowded. We show that the IMPDH inhibitor mycophenolic acid enhances both phases of increase. Deletion of a transcription attenuator upstream of the mRNA start site that decreases Imd2 mRNA synthesis in the presence of high GTP increases the baseline level of Imd2 protein 10-fold and abolishes guanine-sensitive but not guanine-insensitive induction. Our results suggest that at least two mechanisms of yeast Imd2 regulation exist, the known GTP-dependent attenuation of RNA polymerase II elongation and a GTP concentration-independent pathway that may be controlled by cell growth state. Live-cell analysis of IMPDH protein levels in a growing yeast colony confirms a known mechanism of regulation and provides evidence for an additional mode of regulation.

Importance: This study used live-cell microscopy to track changes in the level of a key enzyme in GTP nucleotide biosynthesis, inosine monophosphate dehydrogenase (IMPDH), during growth of a brewers yeast colony over 2 days in a microfluidic device. The results show that feedback regulation via transcription attenuation allows cells to adapt to nutrient limitation in the crowded environs of a yeast colony. They also identify a novel mode of regulation of IMPDH level that is not driven by guanine nucleotide availability.

Keywords: GTP biosynthesis; microfluidics; mycophenolic acid; nucleotide metabolism; transcription attenuation.

Fig 1

IMD2 mRNA synthesis is regulated by GTP-dependent start site selection and transcriptional attenuation. Imd2 catalyzes the first committed step of GTP synthesis from IMP, which is also a precursor of ATP. IMD2 has multiple transcription start sites (TSS, bent arrows). Upstream TSS are used when cellular GTP is sufficient, producing transcripts containing a Nrd1-Nab3-Sen1 (NNS) terminator that terminate prematurely and are degraded by the nuclear exosome. When cellular GTP is deficient, a downstream TSS is preferentially used, resulting in full-length mRNA [adapted from reference (8)].

Fig 2

Single-cell measurement of Imd2-GFP protein accumulation over 50 hours with and without guanine supplementation. (A) Differential interference contrast (DIC) images of a representative viewing field after 50 hours of perfusion. Left panel: viewing field imaged with a 40× objective. Center panel: image with a 10× objective of the edge of the yeast cell monolayer adjacent to the media inlet channels. Cells that were pushed into the inlet by growth of the monolayer have formed a three-dimensional colony. Right panel: 40× objective image of yeast cells entering the media inlet from the monolayer. (B) DIC, Imd2-GFP, or Nhp6a-TagRFP images of cells after 0, 10, 20, 30, 40, or 50 hours of perfusion with synthetic complete media with no additions. Approximately one-quarter of the standard viewing field is shown. Very bright cells in the RFP channel are likely dead (see main text). (C) Single-cell measurement of Imd2-GFP protein accumulation over 50 hours with and without guanine supplementation at 0 hours. Top panel: Imd2-GFP fold-change versus hours of perfusion with untreated media (black and green lines) or media supplemented with 400 µM guanine (red and blue lines). Both treatments were performed in biological replicate. The solid lines indicate population mean per cell GFP fluorescence, and shading indicates one standard deviation above and below the mean. Dotted lines indicate the time at which guanine supplementation stops the increase in Imd2-GFP levels and the percent confluence of the cells at that time. Bottom panel: the percentage of the viewing field occupied by cells over time.

Fig 3

Dose-dependent induction of Imd2-GFP by MPA. (A) Top panel: population mean fold change in Imd2-GFP (solid lines) with one standard deviation above and below the mean (shaded areas) for cells treated with no MPA (black), 0.15 µg/mL (blue), 1.5 µg/mL (red), or 15 µg/mL MPA (green). Bottom panel: percentage of the viewing field occupied by cells over time. (B) Violin plots showing the distribution of Imd2-GFP expression across the cell population. Solid dots represent the population mean Imd2-GFP.

Fig 4

Guanine-specific and dose-dependent suppression of Phase 2 induction of Imd2-GFP in the presence of MPA. (A) Top panel: population mean Imd2-GFP fold change (solid lines) with one standard deviation above and below the mean (shaded area) for 1.5 µg/mL MPA alone (black) or in the presence of 100 µM (green), 200 µM (blue), or 400 µM guanine (purple). Bottom panel: percentage of the viewing field occupied by cells over time. (B) Same as panel A, but with the indicated concentrations of adenine in place of guanine.

Fig 5

Imd2-GFP expression is not homogenous across the viewing field after confluence, and cell growth emanates from areas with the highest Imd2-GFP levels. Still images showing the central 20% of the viewing field (vertically) for the Imd2-GFP channel of cells treated with 1.5 µg/mL MPA at 0–14 hours after confluence. Nutrient flow is from right to left. Selected cells, identified by their brightness in the RFP channel, are circled in yellow, and their change in horizontal position over time is highlighted by yellow arrows. Corresponds to data shown in Fig. 4A and Video S6.

Fig 6

Overexpression of catalytically active IMD2 from a plasmid decreases endogenous Imd2-GFP levels in confluent cells. (A) Schematic genotype of strains used in this experiment. TDH3pro: promoter region from the TDH3 gene; PGK1pA: 3′-UTR and polyA site from the PGK1 gene; 2 micron: high-copy yeast plasmid origin. (B) Top panel: levels of Imd2-GFP in cells treated with 1.5 µg/mL MPA and containing the pG1(LEU2) plasmid with no insert (black), IMD2-WT (purple), or the catalytically inactive IMD2-C335A allele (blue). Solid lines indicate population mean per cell GFP fluorescence and shading indicates one standard deviation above and below the mean. Bottom panel: the percentage of the viewing field occupied by cells over time. (C) Still images of the Imd2-GFP channel when cells expressing the pG1(LEU2)-IMD2-WT or pG1(LEU2)-IMD2-C335A plasmids first reached confluence.

Fig 7

Addition of MPA shortly before or after confluence results in rapid, guanine-sensitive Imd2-GFP induction. Top panel: population mean Imd2-GFP fold induction (solid lines) with one standard deviation above and below the mean (shaded area) for cells perfused for 10 or 20 hours with untreated media before addition of 1.5 µg/mL MPA without or with 400 µM guanine. Colored arrows indicate the time that MPA or MPA + guanine was added to the cells (see key). Bottom panel: the percentage of the viewing field occupied by cells over time.

Fig 8

Deletion of nonproductive “G” transcription start sites from IMD2 increases basal Imd2-GFP levels 10-fold and abolishes guanine sensitivity. (A) Imd2-GFP expression in the attenuator mutant strain (ELS107) compared with wild-type cells (KES002) after 0.5 hours of perfusion with untreated media based on seven experiments and at least two technical replicates per experiment. Imd2-GFP levels were normalized to the average Imd2-GFP expression of wild-type cells during the first hour of perfusion with untreated media. Error bars represent standard error of the mean. Based on a two-tailed t-test, P value = 4 × 10⁻¹². (B) Images showing the Imd2-GFP channel after 0.5 hours of perfusion with untreated media for wild-type KES002 cells (left) or attenuator mutant ELS107 cells (right). (C) Top panel: population mean Imd2-GFP fold induction (solid lines) with one standard deviation above and below the mean (shaded area) for attenuator mutant (ELS107) and wild-type (KES002) cells perfused with untreated media or 400 µM guanine (see key). Bottom panel: the percentage of the viewing field occupied by cells over time. (D) Population mean Imd2-GFP fold-induction averaged over three separate experiments (solid lines) with one standard deviation above and below the mean (shaded area) for attenuator mutant (ELS107) and wild-type (KES002) cells perfused with 1.5 µg/mL MPA or MPA with 400 µM guanine (see key).