Research ArticleGENETICS

Synchronization of stochastic expressions drives the clustering of functionally related genes

See allHide authors and affiliations

Science Advances  09 Oct 2019:
Vol. 5, no. 10, eaax6525
DOI: 10.1126/sciadv.aax6525
  • Fig. 1 Schematics explaining the hypothesis that coordination of stochastic expressions drives the clustering of functionally related genes.

    Genes 1 and 2 encode two enzymes catalyzing successive reactions in a metabolic pathway, where the intermediate metabolic of the two reactions is toxic. (A) When the two genes are unlinked, their stochastic expressions are uncoordinated, leading to a high variation of their expression ratio that causes the accumulation of the toxic intermediate. (B) When the two genes are closely linked, their common chromatin environment coordinates their stochastic expressions, stabilizing their expression ratio, which lessens the accumulation of the toxic intermediate. (C and D) Expected concentrations of the intermediate metabolite under different levels of correlation between the concentrations of enzyme 1 and enzyme 2, determined by numerical simulations under steady states (C) or non–steady states (D). Each circle shows the average from 100,000 simulations in (C) or 10,000 simulations in (D). Error bars show SEs.

  • Fig. 2 Testing advantageous coordination of stochastic expressions of GAL1 and GAL10 in the GAL gene cluster.

    (A) Chromosomal locations of GAL cluster genes in yeast. Arrows indicate transcriptional directions. Each small triangle indicates a UAS. (B) Yeast galactose catabolism pathway, showing metabolic reactions catalyzed by enzymes encoded by the GAL gene cluster (solid arrows) and a reaction catalyzed by aldose reductase (dashed arrow). The skull-and-crossbones symbol indicates toxicity. Galactitol is probably toxic (28, 29). (C) Expected concentrations of galactose-1-P under different levels of correlation between the concentrations of Gal1 and Gal10, determined by non–steady-state numerical simulations. The correlation coefficients in the concentration between Gal1 and Gal7 and that between Gal7 and Gal10 are fixed at various levels indicated on the top of the figure. Each circle indicates the average from 2000 simulations. (D) Schematics showing cis- and trans-tagging strains. Blue and yellow arrows, respectively, represent the cyan (CFP) and yellow (YFP) fluorescence protein genes. (E) The CV of the Gal1-yfp/Gal10-cfp protein level ratio among cells is significantly lower for the cis- than trans-tagging strains. Each circle indicates the mean from eight biological replicates, each with 5000 cells. (F) Schematics showing cis- and trans-deletion strains. A red cross indicates gene deletion. (G and H) Cellular concentrations of galactose-1-P (G) and galactitol (H) are significantly lower in the cis- than in the trans-deletion strains. In (G) and (H), each circle represents the mean from four biological replicates, each having three technical repeats. a.u., arbitrary units. (I) Schematics showing cis- and trans-deletion strains used for fitness estimation via strain competition. Green and yellow cells, respectively, indicate cells that express a green fluorescence protein gene (GFP) and a YFP gene at the HO locus. (J) The fitness of the cis-deletion strain relative to the trans-deleting strain in (I) is substantially greater than 1 in a galactose medium (YPGal) but is slightly below 1 in a glucose medium (YPD). (K) Schematics showing cis- and trans-deletion strains with reciprocal fluorescence markers used for fitness estimation. (L) The fitness of the cis-deletion strain relative to the trans-deleting strain in (K) is substantially greater than 1 in YPGal but is not different from 1 in YPD. In (J) and (L), each circle represents the mean from eight biological replicates, and statistical tests of the null hypothesis that the relative fitness equals 1 are performed. In all panels, error bars show SEs. Significance levels are indicated as follows: NS, P ≥ 0.05; *P < 0.05; **P < 0.01, ***P < 0.001.

  • Fig. 3 Testing advantageous coordination of stochastic expressions of GAL7 and GAL10 in the GAL gene cluster.

    (A) Schematics showing cis- and trans-tagging strains and cis- and trans-deletion strains. Blue and yellow symbols, respectively, represent the CFP and YFP genes. A red cross indicates gene deletion. (B) The CV of the Gal7-yfp/Gal10-cfp protein level ratio among cells is significantly lower for the cis- than for the trans-tagging strains. Each circle is the average from six replicates, each with 5000 cells. (C and D) Cellular concentrations of galactose-1-P (C) and galactitol (D) are significantly lower in the cis- than in the trans-deletion strains. In (C) and (D), each circle represents the average from four biological replicates, each having three technical repeats. (E) The fitness of the YFP-marked cis-deletion strain relative to the GFP-marked trans-deletion strain is substantially greater than 1 in a galactose medium (YPGal) but is slightly below 1 in a glucose medium (YPD). (F) The fitness of the GFP-marked cis-deletion strain relative to the YFP-marked trans-deletion strain is substantially greater than 1 in YPGal but is not different from 1 in YPD. In (E) and (F), each circle represents the average from eight biological replicates, and statistical tests of the null hypothesis that the relative fitness equals 1 are performed. In all panels, error bars show SEs. Significance levels are indicated as follows: NS, P ≥ 0.05; *P < 0.05; **P < 0.01, ***P < 0.001.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/10/eaax6525/DC1

    Fig. S1. Simulation of the GAL pathway.

    Fig. S2. Comparison of protein expression levels of GAL genes between cis- and trans-tagging strains.

    Fig. S3. Flowchart showing the construction of cis- and trans-deletion strains for the pair of GAL1 and GAL10 genes.

    Fig. S4. Ln(cell number ratio between the cis- and trans-deletion strains) as a function of competition time.

    Table S1. Parameter values used in the simulation of the linear pathway with two reactions.

    Table S2. Parameter values used in the simulation of the GAL pathway.

    Table S3. Primers for strain construction.

    Table S4. gRNA target sequences.

    Table S5. Doubling times for various trans-deletion strains.

    References (4651)

  • Supplementary Materials

    This PDF file includes:

    • Fig. S1. Simulation of the GAL pathway.
    • Fig. S2. Comparison of protein expression levels of GAL genes between cis- and trans-tagging strains.
    • Fig. S3. Flowchart showing the construction of cis- and trans-deletion strains for the pair of GAL1 and GAL10 genes.
    • Fig. S4. Ln(cell number ratio between the cis- and trans-deletion strains) as a function of competition time.
    • Table S1. Parameter values used in the simulation of the linear pathway with two reactions.
    • Table S2. Parameter values used in the simulation of the GAL pathway.
    • Table S3. Primers for strain construction.
    • Table S4. gRNA target sequences.
    • Table S5. Doubling times for various trans-deletion strains.
    • References (4651)

    Download PDF

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article