An evolutionarily conserved prion-like element converts wild fungi from metabolic specialists to generalists

Abstract

[GAR(+)] is a protein-based element of inheritance that allows yeast (Saccharomyces cerevisiae) to circumvent a hallmark of their biology: extreme metabolic specialization for glucose fermentation. When glucose is present, yeast will not use other carbon sources. [GAR(+)] allows cells to circumvent this "glucose repression." [GAR(+)] is induced in yeast by a factor secreted by bacteria inhabiting their environment. We report that de novo rates of [GAR(+)] appearance correlate with the yeast's ecological niche. Evolutionarily distant fungi possess similar epigenetic elements that are also induced by bacteria. As expected for a mechanism whose adaptive value originates from the selective pressures of life in biological communities, the ability of bacteria to induce [GAR(+)] and the ability of yeast to respond to bacterial signals have been extinguished repeatedly during the extended monoculture of domestication. Thus, [GAR(+)] is a broadly conserved adaptive strategy that links environmental and social cues to heritable changes in metabolism.

Figure 1. [GAR⁺] is common in wild strains of S. cerevisiae

(A) Diverse wild strains of S. cerevisiae have the capacity to acquire the heritable ability to grow on GLY + GlcN. (B). Scatter plot of the frequency of [GAR⁺] appearance among S. cerevisiae strains from different ecological niches (See also Fig. S1, Table S1).

Figure 2. Soil isolates are naturally [GAR⁺]

(A) Three soil isolates grew robustly on GLY/GlcN even after many generations of nonselective propagation. (B) These same isolates lost this trait after transient reduction in Hsp70 function from a dominant negative plasmid. (See also Table S2).

Figure 3. Prion-based reversal of glucose repression occurs in diverse fungi

Variants of N. castellii. C. glabrata, and D. bruxellensis that (A) could grow on on GLY/GlcN medium were stable through (B) multiple passages on non-selective medium, but could be eliminated by (C) transient chemical inhibition of Hsp70 (shown here after three passages on GLY plates containing 50 µM myricetin). (See also Fig. S2).

Figure 4. Evolutionary conservation of [GAR⁺] signaling networks across the fungal lineage

(A) The species tree for fungi studied. (B) Presence and absence of homologs for key proteins involved in the [GAR⁺] phenotype. Color indicates the degree of sequence conservation relative to S. cerevisiae (see also Table S3, S4). (C). Protein network involved in the [GAR⁺] phenotype in S. cerevisiae and predicted consequences of Std1 and Mth1 loss in S. pombe.

Figure 5. Epigenetic switches enable a generalist strategy

(A) Schematic of generalist vs. specialist strategies for growth and survival. (B) Normal [gar⁻] cells pursue a specialist strategy: a narrow range of resource conditions with high fitness, whereas [GAR⁺] cells are generalists: fitter across a wider range of conditions. Similar effects are seen with the [GAR⁺]-like states from diverse other fungi. Resource conditions were quantified by the fraction of galactose to total amount of carbon: starting with a carbon source of no glucose and 2% galactose to 2% glucose, and no galactose. Fitness was measured as the ratio of final biomass yield to doubling time in exponential phase. Error bars represent the SD from three independent experiments.

Figure 6. Metapopulation model to test for conservation of the bet-hedging properties of the [GAR⁺] prion

(A) Upper panel: a [gar−] cell can switch to [GAR⁺] at a rate of m_[GAR+], but this state is reversible, as the same cells can switch back to the [gar⁻] state. [gar⁻] cells can also rarely reverse glucose repression via mutation. We define the rate at which this irreversible change occurs as m_mimic. Lower panel: Schematic of the population genetics model we employed (Lancaster and Masel, 2009). A sub-population of cells within the metapopulation can ‘bet-hedge’ to survive environmental changes by switching to the [GAR⁺] state when the environment favors reversal of glucose repression (purple) and by switching back to [gar⁻] when the environment returns to a state that favors this response [GAR⁺] (blue). Alternatively if the cells circumvent glucose repression via a genetic mutation, they may be eliminated when the environment changes because they cannot return to the [gar⁻] state. (The equal time the subpopulation spends in both kinds of environments is for illustrative purposes only: in the model, environmental changes occur independently and stochastically within each subpopulation). (B) Inference of strong selection pressure for [GAR⁺] switching. We define Ω as the rate of environmental change for which the [GAR⁺] phenotype is adaptive. For an effective population size of N_e=5×10⁶ (see Extended Experimental Procedures), the contour plot depicts the inferred strength of selection (measured by the product N_e Ω) as a function of m_[GAR+] m_mimic. For illustrative purposes we have divided the selection landscape into three regions: weak selection (1<N_eΩ< 5; colored in cream); moderate selection (5<N_e Ω< 50, colored in orange); and strong selection (N_eΩ >50, colored in red). Superimposed on the contour plot are the maximum-likelihood estimates for m_[GAR+] for each of the S. cerevisiae strains and for the other species. Mimic rates appear to be very small, however, we estimated an upper limit to the uncertainty of this parameter of 3.15–7.88×10⁻⁶. The depicted m_mimic therefore ranges from the very low (10⁻¹⁰) to this upper limit, and we placed strains equispaced on the vertical axis across this range. See Extended Experimental Procedures and Table S5 for more explanation and details of computations. See also Fig. S3 where we validated the robustness of these analyses to a wide range of uncertainty in the parameters.

Figure 7. Domestication extinguishes the capacity of bacteria to secrete a [GAR⁺]-inducing signal and the capacity of yeast to perceive it

A) ‘Wild’ bacteria (e.g. E. coli strain MG1655) are better able to induce yeast (S. cerevisiae strain W303) to grow on GLY + GlcN medium than domesticated bacteria (e.g. E. coli strain W3110). B) ‘Wild’ yeast (e.g. S. cerevisiae strain UCD2780) are better able to be induced to acquire [GAR⁺] than domesticated yeast strains (e.g. S. cerevisiae strain 74D). (See also Table S6).