When it comes to diet, yeast cells seem to hedge their bets
Responding to environmental stimuli, molecular networks have evolved functions that are triggered automatically. Circadian rhythms, for example, are encoded within regulatory circuits and coordinated by cues such as light. But additionally, the speed and uniformity with which these simple networks transmit signals can also be optimised, suggesting a form of intelligent adaptability. Because this so-called stochastic or noisy behaviour is less predictable, it can seem uncoordinated and random. “This could be bet hedging where organisms adapt to situations where events, such as availability of food, happen or don’t,” explains Kevin Verstrepen, YEASTMEMORY project coordinator from the VIB-KU Leuven Center for Microbiology, the project host. The EU-supported YEASTMEMORY wanted to know more about the timing of an organism’s response to stimuli and which behaviours proved advantageous. Studying how yeast cells responded to glucose availability, YEASTMEMORY found that responses varied between genetically identical cells in the same population and were influenced by past experiences. They identified the genetic component likely responsible, using their new protocol for single-cell sequencing in yeast.
The genetic basis of different response behaviours
“This project grew out of a conversation 15 years ago with a brewer; not so much over a beer, but about beer,” says Verstrepen. Beer is made by exposing maltose, a barley sugar, to fermenting yeast cells. Beforehand, brewers sometimes grow their yeast in a lab using glucose. This brewer wanted to know why his yeast cells did not ferment maltose more efficiently after several days cultivation on glucose. “Accounting for the speed yeast cells divide, I saw that newborn cells behaved differently, depending on the previous generations’ diet, and that this behaviour was very varied between them. An epigenetic phenomenon seemed likely responsible,” notes Verstrepen. Fast-forward 15 years to YEASTMEMORY. The team found that after growing in glucose for at least 10 hours, most yeast cells then needed at least 6 hours to shift to feeding on maltose, with some genetically identical cells taking over 20 hours and some never transitioning.
Differences were also found between yeast strains.
The key factor, potentially creating a bottleneck, was found to be the transition from fermentative to respiratory metabolism. This major change relies on the synthesis of many new proteins and complexes and so is a slow and energy-intensive process for cells. “It makes sense that cells avoid responding too quickly and in unison, risking investing time and energy into changing diet from glucose to maltose, only to find glucose available again,” adds Verstrepen. To investigate this, the team hunted down the parts of yeast DNA, so-called QTL, that determine this behavioural difference. Their newly developed single-cell RNAseq protocol allowed the team to discover that changes in one specific uncharacterised gene, YLR108C, could explain much of the difference. Cells with a more active form of YLR108C are slower to transition but seem better at growing in more stable conditions. “This is one of the first clear examples of how a relatively simple, naturally occurring mutation in one gene dictates the behaviour of a regulatory system. “We can now investigate the trade-offs of each strategy – the slower and more stochastic generalist or quicker and more uniform specialist – to better understand the evolutionary pressure that favours specific behaviours,” remarks Verstrepen. A patent has already been secured to use YLR108C variants to optimise the performance of yeast in industrial fermentations. Alongside a European Research Council (ERC) proof of concept grant, the team is already in discussions with industrial partners regarding potential applications.
Keywords
YEASTMEMORY, yeast, beer, regulatory circuits, molecular networks, glucose, maltose, RNAseq, epigenetic, cells, gene
