Experimental Evolution of Yeast Reveals Trade-offs Between Early and Late Stationary Phase

This study demonstrates that experimental evolution of yeast reveals a trade-off between fitness in early versus late stationary phase, where adaptive mutations improve performance in one phase at the expense of the other, while longer stationary intervals drive distinct evolutionary routes and larger fitness effects.

The Gist

The Big Picture: The "Feast and Famine" Cycle

Imagine a group of yeast cells living in a bowl of soup.

1. The Feast: At first, the soup is full of food (sugar). The yeast eat, grow, and multiply rapidly. This is the "exponential growth" phase.
2. The Famine: Eventually, they eat all the food. The soup is now just "spent media." The yeast can't grow anymore, so they stop dividing and enter a survival mode called Stationary Phase.

In nature, most microbes spend the majority of their lives in this "famine" state, waiting for the next meal. Scientists often study how microbes grow fast during the feast, but this paper asks a different question: How do they survive the famine, and does surviving the early days of starvation help them survive the later days?

The Experiment: A "Time-Travel" Gym for Yeast

The researchers set up a gym for yeast to see how they adapt to starvation.

• The Setup: They took a huge population of yeast (each with a unique "barcode" like a social security number) and put them in a non-food environment (glycerol and ethanol).
• The Variable: They changed how long the yeast had to wait before getting a fresh bowl of soup.
  • Some groups waited 2 days (short famine).
  • Some waited 4, 6, 8, or even 10 days (long famine).
• The Goal: They let the yeast evolve for many generations, then picked the "champion" survivors to see what mutations (genetic changes) made them so good at surviving.

Key Discovery 1: The Longer the Wait, The Bigger the Changes

They found that the longer the yeast had to wait in the famine, the more dramatic the changes became.

• Analogy: Imagine a marathon. If you only run for 10 minutes, you might just tighten your shoelaces. But if you have to run for 100 miles, you might need to change your entire diet, buy new shoes, and train your lungs differently.
• The Result: Yeast that faced long starvation periods lost their genetic diversity faster (only the toughest survived) and developed "superpowers" (mutations) that gave them a huge advantage. Short starvation periods resulted in smaller, less dramatic changes.

Key Discovery 2: The "Survival Trade-Off" (The Main Plot Twist)

This is the most important finding. The researchers thought that being good at surviving starvation was a single skill. They were wrong. They found that Stationary Phase is actually two different skills that fight against each other.

• Early Stationary Phase (Days 2–4): This is the "shock absorber" phase. The yeast are just realizing the food is gone and trying to stabilize.
• Late Stationary Phase (Days 6–10): This is the "deep endurance" phase. The yeast are running out of energy reserves and dealing with toxic waste buildup.

The Trade-Off:
The researchers discovered a negative correlation.

• If a yeast mutation makes it great at surviving the first few days of starvation, it usually makes it terrible at surviving the later days.
• Conversely, if a mutation makes it a master of long-term endurance, it often struggles to survive the initial shock.

The Analogy: The Sprinter vs. The Marathoner
Imagine a race where you have to survive a harsh environment.

• Type A Mutant: Is like a Sprinter. They are built to react instantly to the start of the race. They are amazing at the first 2 days. But if the race goes on for 10 days, they burn out and die.
• Type B Mutant: Is like a Marathoner. They are slow to start and struggle in the first few days, but they have incredible stamina. They survive the long haul.
• The Problem: You can't easily be both a world-class sprinter and a world-class marathoner at the same time. The body (or cell) has to choose a strategy.

Key Discovery 3: The "Menu" Doesn't Matter as Much as the "Wait Time"

The researchers also tested if it mattered what the yeast ate before the starvation (Glucose vs. Glycerol/Ethanol).

• The Finding: It didn't matter much. Whether the yeast ate a "sugar diet" or a "fat diet" before starving, the mutations that helped them survive the famine were surprisingly similar.
• The Takeaway: The length of the starvation is the boss. The type of food eaten beforehand is just a side note. The yeast adapt to the duration of the hunger, not the specific menu.

The "SMF2" Gene: A Case Study in Trade-Offs

The researchers found a specific gene called SMF2 that was mutated in many of the long-term survivors.

• These mutants were superheroes at surviving the first few days of starvation.
• But, almost all of them died off quickly once the starvation went on for a long time.
• This perfectly proved the trade-off theory: One gene change made them great at the "Early Game" but doomed them in the "Late Game."

Why Does This Matter?

1. It's Not Just One Thing: Scientists used to think "survival of starvation" was one single trait. This paper shows it's actually a complex timeline with different challenges at different times.
2. Evolution is a Balancing Act: Organisms can't be perfect at everything. If you evolve to be great at surviving short famines, you might lose the ability to survive long ones.
3. Real World Impact: This helps us understand how bacteria and yeast survive in nature (like in the ocean or in our gut), how they age, and how they might develop resistance to drugs that try to starve them.

Summary

Think of the yeast as a group of hikers.

• If the hike is short (2 days), they just need a good pair of boots (Ras/PKA mutations).
• If the hike is long (10 days), they need to change their entire strategy (SMF2, Chromosome 11 duplications).
• The Catch: The hiker who is best at surviving the first 2 hours of a storm is often the worst at surviving the next 8 hours. To survive the whole storm, you have to make a hard choice about which part of the storm you are preparing for.

Technical Summary

1. Problem Statement

Microorganisms, particularly in natural environments, spend the majority of their existence in a non-proliferative "stationary phase" triggered by nutrient exhaustion. While laboratory experimental evolution typically focuses on maximizing exponential growth rates, survival during stationary phase is a critical component of fitness in nature.

• The Gap: Previous studies often treated stationary phase as a monolithic trait or focused on specific "Growth Advantage in Stationary Phase" (GASP) phenotypes. However, the death rate of stationary phase cells is not constant over time, suggesting that performance in early versus late stationary phase may involve distinct physiological mechanisms.
• The Question: How does the duration of stationary phase influence evolutionary dynamics, genomic adaptation, and the trade-offs between performance in different temporal intervals of nutrient limitation? Does adaptation to one carbon source (glucose) translate to another (glycerol/ethanol), and are there trade-offs between surviving the first few days of starvation versus the later stages?

2. Methodology

The authors employed a high-throughput experimental evolution approach using Saccharomyces cerevisiae (baker's yeast) with barcoded lineages to track evolutionary trajectories.

• Experimental Design:

  • Strains: Haploid barcoded yeast pools were evolved in triplicate.
  • Media: Two carbon sources were used: Glucose (fermentable) and a non-fermentable mix of 0.5% Glycerol + 0.5% Ethanol (Gly/Eth).
  • Evolution Regime: Populations were serially transferred with varying intervals between transfers to manipulate the time spent in stationary phase: 0, 2, 4, 6, 8, and 10 days.
  • Population Size: 5 × 10⁷ cells transferred per cycle (8 generations per cycle).
  • Duration: 16–25 transfers (128–200 generations).

• Genomic Analysis:

  • Barcode Sequencing: Used to track lineage frequencies and estimate fitness trajectories (using FitMut2).
  • Clone Isolation: 480 uniquely barcoded adaptive clones were isolated based on trajectory data.
  • Whole Genome Sequencing (WGS): Performed on the 480 clones to identify specific mutations.

• Fitness Assays:

  • The 480 clones were pooled with ancestral strains and evolved glucose-adapted clones.
  • Competition Assays: The pool was propagated in 9 different conditions (varying carbon sources and transfer intervals).
  • Performance Metrics: By comparing the fitness of clones in cycles with different stationary phase durations (e.g., 2-day vs. 10-day cycles), the authors mathematically inferred the specific fitness contribution of specific time intervals (e.g., Days 2–4 vs. Days 6–10).

3. Key Results

A. Impact of Stationary Phase Duration on Evolutionary Dynamics

• Diversity Loss: Populations with longer stationary phase intervals lost barcode diversity more rapidly, indicating stronger selection and faster sweeps.
• Mutation Effect Size: Adaptive mutations in longer stationary phase regimes had larger fitness effects compared to those in short regimes.
• Distinct Genomic Signatures: The specific mutations selected depended heavily on the duration of stationary phase:
  • 0–2 Days: Dominated by Ras/PKA pathway mutations (loss of function in negative regulators), similar to glucose-evolved clones.
  • 4 Days: Low parallelism; diverse adaptive routes.
  • 6 Days: High parallelism driven by Chromosome 11 duplications (increasing UTH1 copy number for oxidative stress resistance) and SMF2 mutations (metal ion transport).
  • 8 Days: Enrichment of heterozygous missense mutations in FZF1 (sulfite metabolism transcription factor).
  • 10 Days: A mix of Ras/PKA, Chromosome 11 duplications, SMF2, and novel mutations in ACE2 (cell division/septum) and RAV1 (vacuolar transport).

B. Carbon Source Independence

• Correlation: Mutations that improved stationary phase performance following glucose exhaustion also improved performance following Gly/Eth exhaustion.
• Conclusion: The underlying traits conferring stationary phase survival are largely independent of the specific carbon source consumed during the growth phase, suggesting a shared physiological mechanism for surviving nutrient depletion.

C. The Trade-off: Early vs. Late Stationary Phase

• Negative Correlation: The study's most significant finding is a negative correlation between performance in early stationary phase (Days 2–4) and late stationary phase (Days 6–10).
• Evidence: Mutants that excelled in the first few days of starvation often performed poorly in later days, and vice versa.
• Gene-Specific Trade-off: This trade-off was observed even within a single gene. SMF2 mutants (the most frequent adaptation in 6-day Gly/Eth conditions) showed a massive increase in early stationary phase performance but a significant decrease in late stationary phase performance.
• Robustness: This trade-off held true across different genetic backgrounds and carbon sources, even when controlling for diploidy and specific mutation types.

4. Significance and Contributions

1. Decoupling Stationary Phase: The study challenges the view of stationary phase as a single, uniform state. It demonstrates that "early" and "late" stationary phase represent distinct fitness phenotypes with antagonistic evolutionary requirements.
2. Evolutionary Trade-offs: It provides empirical evidence that adaptation to immediate starvation stress (early phase) can be detrimental to long-term survival (late phase), likely due to resource allocation conflicts or physiological constraints (e.g., SMF2 mutants may scavenge resources aggressively early on but fail to maintain homeostasis later).
3. Context-Dependent Adaptation: The genomic route to adaptation is highly sensitive to the duration of the stressor. A mutation beneficial in a 2-day cycle may be neutral or deleterious in a 10-day cycle, and vice versa.
4. Ecological Relevance: These findings suggest that in natural environments where nutrient pulses are irregular, microbial communities may maintain diversity through these trade-offs, preventing a single "super-mutant" from dominating all temporal niches.

5. Conclusion

The authors conclude that stationary phase in yeast is a dynamic, multi-phasic process. Adaptation to this state is not a singular optimization but a balancing act between competing physiological demands. The trade-off between early and late survival implies that microbial life-history strategies are constrained by the temporal structure of their environment, and that "fitness" in stationary phase must be measured as a function of time, not just a single endpoint.