Saccharomyces cerevisiae is known as the Baker's Yeast. It has helped us make bread, beer, and wine since before recorded history. These days we also use it to make fuel, pharmaceuticals, and for basic biology research. With the innumerable industrial, food, and research purposes we use it for, it is a thoroughly domesticated organism.
With the various mammals we've domesticated, researchers have identified a "domestication syndrome"; a set of features common across domesticated animals. They have shorter faces, milder temperaments, reduced weaponry (teeth, horns, claws), and color changes. In short, they've become cuter. To some degree these are traits that could have been actively selected for, but it turns out that if we only select on temperament, all of the other traits come along for free because all those traits are mediated by the action of neural crest cells throughout the body.
Now, yeast don't have neural crest cells, but they're still domesticated. It didn't evolve to have a more amenable temperament, but it did evolve to grow rapidly in the amenable conditions we provide for them. There's a different sort of "domestication syndrome" that it would have developed along the way. Any trait or ability it needed to live as a wild yeast, but did not need to live under our care, would be lost. This would happen because any lineage that dispensed with those traits would be able to grow faster without the energy drain they represent.
So. What traits would yeast lose under domestication? It's not entirely clear. We can't just look at the cells and see a difference. Nor do we exactly have the wild progenitor yeast around to make comparisons with.
Here we're going to take a bit of diversion.
My first major project in grad school was to figure out how to use flow cytometry to determine the genome size of a different yeast called Candida albicans. In the past, This analysis had proven difficult to do with this yeast for others. This difficulty had been generally blamed on the organism's ability to grow either as independent yeast cells or as elongated hyphal cells that get all tangled up in each other.
I started with protocols developed for S. cerevisiae. At three months in, I was testing yet another protocol variation and the data that came out of the experiment looked like the figure at right. Previous data had much broader, indistinct peaks. (I'm sure I have some of those early figures around somewhere, but I'm not going to spend a bunch of time digging for them.) I was amazed and quickly set up a repeat of the exact same experiment. It failed miserably.
I had made a mistake somewhere in the protocol which made things work. Because it was a mistake, it wasn't written down in my lab notes. You can only write down what you know you're doing.
It took me another frustrating month to figure out what it was I had done wrong. I had used way too much EDTA in the buffers for processing the cells. With this improved protocol, I could get good flow cytometry data from even the most difficult hyphal-growing strains of C. albicans. This disproved the previous theory as to why this species was difficult to work with while doing this assay.
Subsequently, the protocol proved effective with every random yeast species I was able to acquire for testing. I never tested them with the original S. cerevisiae protocol for comparison. In retrospect, I consider this to be an oversight.
I pretty quickly developed a working theory about what was going on. EDTA binds to divalent cations (Ca2+ and Mg2+) in solution, locking them up so other enzymes don't have access to them. Many enzymes require certain levels of these ions to function normally. For whatever reason, the endogenous nucleases of C. albicans were much less sensitive to low levels of divalent cations than those found in S. cerevisiae. Now, I couldn't think of any way to test this theory. I wasn't in a biochemistry or structural biology lab, so the techniques that would have been useful were well outside our wheelhouse.
This uncertainty has stuck with me for the roughly seven years since then. Just a couple days ago, I developed an idea that in some sense explains the results. Domestication.
S. cerevisiae is a thoroughly domesticated species. It hasn't had to fight for what it needs, so it could very well have evolved enzymes that are used to easier environments with more consistent levels of necessary ions. I strongly suspect the flow cytometry protocol for S. cerevisiae only works because of the domestication syndrome of traits found in S. cerevisiae.
I'm not sure how one would test this theory, but it sure seems to make sense of the observations so far.
References:
With the various mammals we've domesticated, researchers have identified a "domestication syndrome"; a set of features common across domesticated animals. They have shorter faces, milder temperaments, reduced weaponry (teeth, horns, claws), and color changes. In short, they've become cuter. To some degree these are traits that could have been actively selected for, but it turns out that if we only select on temperament, all of the other traits come along for free because all those traits are mediated by the action of neural crest cells throughout the body.
Now, yeast don't have neural crest cells, but they're still domesticated. It didn't evolve to have a more amenable temperament, but it did evolve to grow rapidly in the amenable conditions we provide for them. There's a different sort of "domestication syndrome" that it would have developed along the way. Any trait or ability it needed to live as a wild yeast, but did not need to live under our care, would be lost. This would happen because any lineage that dispensed with those traits would be able to grow faster without the energy drain they represent.
So. What traits would yeast lose under domestication? It's not entirely clear. We can't just look at the cells and see a difference. Nor do we exactly have the wild progenitor yeast around to make comparisons with.
Here we're going to take a bit of diversion.
I started with protocols developed for S. cerevisiae. At three months in, I was testing yet another protocol variation and the data that came out of the experiment looked like the figure at right. Previous data had much broader, indistinct peaks. (I'm sure I have some of those early figures around somewhere, but I'm not going to spend a bunch of time digging for them.) I was amazed and quickly set up a repeat of the exact same experiment. It failed miserably.
I had made a mistake somewhere in the protocol which made things work. Because it was a mistake, it wasn't written down in my lab notes. You can only write down what you know you're doing.
It took me another frustrating month to figure out what it was I had done wrong. I had used way too much EDTA in the buffers for processing the cells. With this improved protocol, I could get good flow cytometry data from even the most difficult hyphal-growing strains of C. albicans. This disproved the previous theory as to why this species was difficult to work with while doing this assay.
Subsequently, the protocol proved effective with every random yeast species I was able to acquire for testing. I never tested them with the original S. cerevisiae protocol for comparison. In retrospect, I consider this to be an oversight.
The flow cytometry protocol has since then been used in numerous papers from several separate labs. The flow cytometry protocol and analysis tools I developed become the second chapter in my thesis. The idea of wrapping up the material into a paper did come up after I graduated, but I really didn't have the time/energy to dedicate to the process. Researchers should probably cite that chapter, but I know that thesis chapters tend to only get cited rarely. If you are interested in all the details, you are welcome to have a read.
I pretty quickly developed a working theory about what was going on. EDTA binds to divalent cations (Ca2+ and Mg2+) in solution, locking them up so other enzymes don't have access to them. Many enzymes require certain levels of these ions to function normally. For whatever reason, the endogenous nucleases of C. albicans were much less sensitive to low levels of divalent cations than those found in S. cerevisiae. Now, I couldn't think of any way to test this theory. I wasn't in a biochemistry or structural biology lab, so the techniques that would have been useful were well outside our wheelhouse.
This uncertainty has stuck with me for the roughly seven years since then. Just a couple days ago, I developed an idea that in some sense explains the results. Domestication.
S. cerevisiae is a thoroughly domesticated species. It hasn't had to fight for what it needs, so it could very well have evolved enzymes that are used to easier environments with more consistent levels of necessary ions. I strongly suspect the flow cytometry protocol for S. cerevisiae only works because of the domestication syndrome of traits found in S. cerevisiae.
I'm not sure how one would test this theory, but it sure seems to make sense of the observations so far.
References:
- Domestication Syndrome: http://www.genetics.org/content/197/3/795
- Fox domestication: https://blogs.scientificamerican.com/guest-blog/mans-new-best-friend-a-forgotten-russian-experiment-in-fox-domestication/
- PhD thesis of Darren Abbey: https://conservancy.umn.edu/handle/11299/185630
- Papers with C. albicans flow cytometry analysis:
- https://www.nature.com/articles/nature11865
- https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5908203/
- http://www.genetics.org/content/209/3/725.long
- https://www.ncbi.nlm.nih.gov/pubmed/28195309
- https://www.ncbi.nlm.nih.gov/pubmed/28543785
- https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.1001815
- ... probably many others.
No comments:
Post a Comment