New & Noteworthy

Mannan From Heaven

February 25, 2015

Once we started eating the yeast in breads and beer, our microbiota adapted to feast on these mannans raining down on them from heaven. Manna reigning from heaven on the Israelites (Exodus 16), from the Maciejowski Bible, c. 1250 C.E.

Thousands of years ago, humans encountered very little yeast in their diet. This all changed with the invention of beer and bread.

Now of course we didn’t include the yeast as a carbon source…we just liked fluffy bread and getting hammered. But it was a different story for the bacteria in our gut. They now found the mannans in yeast cell walls raining down on them like manna from heaven.

Unlike the Israelites, who could eat manna, most of our microbiota probably couldn’t use this newfound carbon source. But at some point, the polysaccharide utilization loci (PULs) of a few species changed so that they could now digest the mannans found in yeast cell walls. And, as a new study in Nature by Cuskin and coworkers shows, it was almost certainly a great boon for them.

They found that at least one gram negative bacterium, Bacteroides thetaiotamicron, can live on just the mannans found in the yeast cell walls. This isn’t just a cool story of coevolution either. It might actually help sick people get better one day.

Glycans, like the mannans found in yeast cell walls, have been implicated in autoimmune diseases like Crohn’s disease. One day doctors may be able to treat and/or prevent diseases like this by providing bacteria that can eliminate these potentially harmful mannans: a probiotic treatment for a terrible disease.

Cuskin and coworkers identified three loci in B. thetaiotamicron, MAN-PUL1, MAN-PUL2, and MAN-PUL3, that were activated in the presence of α-mannan. Deletion experiments showed that MAN-PUL2 was absolutely required for the bacteria to utilize these mannans.

The authors next wanted to determine whether the ability to use yeast cell walls as a food source provided an advantage to these bacteria. For this work they turned to a strain of mice that lacked any of their own gut bacteria.

The authors colonized these gnotobiotic mice with 50:50 mixtures of two different strains of B. thetaiotamicron—wild type and a mutant deleted for all three PULs. They found that in the presence of mannans, the wild type strain won out over the mutant strain and that in the absence of mannans, the opposite was true.

So as we might expect, if you feed mice mannans, bacteria that can break them down do best in the mouse gut. The second result, that mutating the PULs might be an advantage in the absence of glycans, wasn’t expected. This result suggests some energetic cost of having active PULs in the absence of usable mannans.

To better mimic the real world, Cuskin and coworkers also looked at the gut bacteria of these mice when they were fed a high bread diet. In this case, the mutant still won out over wild type but not by as much. Instead of being reduced to 10% as was true in the glycan-free diet, the final proportion of wild type bacteria in guts of mice on the high bread diet was 20%. Modest but potentially significant.

We do not have the space to discuss it here, but the authors next dissected the mannan degradation pathway in fine detail. If you are interested please read it over. It is fascinating.

The authors also analyzed 250 human metagenomic samples and found that 62% of them had PULs similar to the ones found in B. thetaiotamicron. So the majority of the people sampled, but not all, had gut microbiota that could deal with the mannans in yeast cell walls.

Human microbiota have adapted to use the energy from the bits of yeast left in bread and alcoholic beverages. Given how little there is, it might be better to think of it as the filth the peasants collected in Monty Python and the Holy Grail instead of manna. Even though it isn’t much, it has given these bacteria a niche that no one else has (or probably wants).

Still, it has allowed them to at least survive. And if these bacteria can one be repurposed as a treatment for disease like Crohn’s disease, thank goodness they adapted to using these mannans.  

by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics

Categories: Research Spotlight

Tags: gut microbiome, glycans, evolution

Species Can’t Risk the New Coke

January 29, 2015

Genome organization may protect key genes from the ravages of increased mutation rate during meiosis:

Back in 1985, Coca Cola decided to completely rejigger the flavor of their flagship soft drink, calling it the New Coke. This radical change to the product was a colossal failure. Toying with such an essential part of a key product was simply too risky a move. If only they had learned from our favorite beast, Saccharomyces cerevisiae.

If only Coke had protected its essential recipe as well as yeast protects its essential genes! Image via Wikimedia Commons

In a new study in PLOS Genetics, Rattray and coworkers show that the mutation rate is higher during meiosis in yeast because of the double-strand breaks associated with recombination. This makes sense, because any new mutations need to be passed on to the next generation for evolution to happen, and germ cells are made by meiosis. But their results also bring up the possibility that key genes might be protected from too many mutations by being in recombination cold spots. Unlike the Coca Cola company, yeast (and everything else) may protect essential genes from radical change.

Previous work in the Strathern lab had suggested that when double strand breaks (DSBs) in the DNA are repaired, one result is an increased mutation rate in the vicinity. The major culprit responsible for the mutations appeared to be DNA polymerase zeta (Rev3p and Rev7p).

To test whether the same is true for the DSBs that happen during the first meiotic prophase, Rattray and coworkers created a strain that contained the CAN1 gene linked to the HIS3 gene. The idea is that mutants in the CAN1 gene can be identified as they will be resistant to canavanine. The HIS3 gene is included as a way to rule out yeast that have become canavanine resistant through a loss of the CAN1 gene. So the authors were looking for strains that were both resistant to canavanine and could grow in the absence of histidine.

The first things the authors found was that the mutation rate during meiosis was indeed increased as compared to mitosis in diploids. For example, when the reporter cassette was inserted into the BUD5 gene, the mitotic mutation rate was 5.7 X 10-8 while the meiotic mutation rate was 3.7 X 10-7, a difference of around 6.5 fold.

This effect was dependent on the DSBs associated with recombination, since the increased mutation rate wasn’t seen in a spo11 mutant; the SPO11 gene is required for these breaks. Using a rev3 mutant, the authors could also conclude that at least half of the increased mutation rate is due to DNA polymerase zeta. This all strongly suggests that the act of recombination increases the local mutation rate.

If recombination is associated with the mutation rate, then areas on the genome that recombine more frequently should have a higher rate of mutation during meiosis. And they do. The authors inserted their cassette into a known recombination hotspot between the BUD23 and the ARE1 genes and saw a meiotic mutation rate of 1.77 X 10-6  as compared to a rate of 4.9 X 10-7 when inserted into a recombination coldspot. This 3.6 fold increase provides additional evidence that recombination is an important factor in meiotic recombination.

This may be more than just an unavoidable side effect of recombination. It could be that yeast and perhaps other beasts end up with their genes arrayed in such a way as to protect important genes by placing them in recombination dead zones.

And perhaps genes where lots of variation is tolerated or even helpful are placed in active recombination areas. In keeping with this, recent studies have shown that essential S. cerevisiae genes tend to be located in recombination cold spots, and that this arrangement is conserved in other yeasts.

It is too early to tell yet how pervasive this sort of gene placement is.  But if this turns out to be a good way to protect essential genes, Coca Cola should definitely have left the Coke formula in a part of its genome with little or no recombination. Mutating that set of instructions was as disastrous as mutating an essential gene!

by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics

Categories: Research Spotlight

Tags: meiosis, recombination, evolution, Saccharomyces cerevisiae

Lots of Ways to Get to the Same Place

November 13, 2014


An advantage of taking one route to the post office might be that you get to see the elephant topiary in front of the zoo. For a yeast cell, taking a roundabout route to a wild-type phenotype might confer a big advantage in a different environment. Image from Wikimedia Commons

How Bad Mutations Can Help Yeast Thrive in New Environments:

If you’ve ever asked for directions from more than one person, you know there are many ways to get to the same place. But not all routes are created equal. You might be trying to get to the post office, but on some routes you’ll pass by the zoo, while on others you might pass the museum or the bus station.

Turns out that something like this may be happening with individuals in a population too. Each may be well adapted for its environment, but each may have arrived there in different ways. And although they may seem similar, they might actually be more different than they look on the surface.

In a new study in PLOS biology, Szamecz and coworkers show that a lot of these different routes to the same place happen in yeast because of bad mutations. They found that when a yeast gets a deleterious mutation, it is sometimes able to mutate its way back to being competitive with wild type again. But the evolved strain is genetically distinct from the original wild type strain. Similar phenotype, distinct genotype.

And this isn’t just an interesting academic exercise either. As any biologist knows, bad mutations are much more common than are good ones. This means that populations may often be evolving to overcome the effects of these bad mutations. This process may help to explain the wide range of diverse genotypes seen in any wild population. 

The authors started out by focusing on 187 yeast strains in which a single gene had been deleted. Each strain grew more poorly than wild type under the tested conditions.

They then took 4 replicates of each mutant along with the wild type strain and grew them for 400 or so generations. They looked for strains that had evolved to overcome the growth defect caused by the mutation. 

To take into account the fact that every strain would probably evolve a bit to grow better in the environment, they only looked for those that had gained more growth advantage than the wild type had. Around 68% of the strains showed at least one replicate that met this criterion.

So as we might expect, it is possible for a strain that grows poorly to mutate its way closer to a wild type growth rate. The next question was whether these mutant strains had mutated back to something close to wild type or to something new.

The authors decided to answer this question by doing a gene expression analysis of the wild type, the eight mutant strains, and a corresponding evolved line from each of these eight. After doing transcriptome analysis, they found that for the most part the evolved lines did not simply revert back to the original gene expression pattern of the wild type strain. Instead, they generated a novel gene expression pattern to deal with the consequences of having lost the original gene. And in the next set of experiments, the authors showed that this matters when the evolved strains are put in a new environment.

The researchers took 237 evolved lines that grew nearly as well as the original wild type strain and tested how well they each did in 14 different environments. In other words, they tested genetically distinct, phenotypically similar strains in new environments.

They found that even though the original mutant strains grew poorly in all the environments tested, the evolved ones sometimes did better. Fitness improved in 52% of the strains and declined in 8%. What is even more interesting is that a few stumbled upon genotypes that were significantly better than the evolved wild type in a particular environment.

A couple of great examples are the rpl6b or atp11 deletion mutant strains. Strains evolved from either mutant did around 25% better than the evolved wild type strain in high salt, even though both of the original mutants did significantly worse than the wild type strain. By suffering a bad mutation, the evolved strain had been rerouted so that it now grew better than wild type. 

So it looks like getting a bad mutation may not be all bad after all. It might just give you that competitive edge you need when things change. Sometimes the best way to get from point A to point B is not a straight line. 

by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics

Categories: Research Spotlight

Tags: evolution, Saccharomyces cerevisiae

If Yeast Can’t Stand the Heat, Our World May Be in Trouble

October 23, 2014

In case there were still any doubters, the world is definitely heating up. April to September 2014 were the warmest these months have ever been since we started keeping records in 1880. And about 35,000 walruses were forced ashore this summer in Alaska because there wasn’t enough room left for them on the sea ice where they normally hang out. Unfortunately, the trend looks to be more record breaking heat for as long as we can see in the future.

Like the X-men, mutated yeast might one day be the superheroes that save us from the worst effects of global warming. Image by Gage Skidmore via Wikimedia Commons

This is bad news for us, and even worse news for nature.  But all is not lost! Our hero yeast may be able to swoop in and make us biofuels to replace gasoline.  This won’t stop global warming, but it might at the very least slow it down.

But yeast in its current form probably couldn’t make enough biofuels to make a difference without using too much precious food in the process. Like the X-men, it will need a helpful mutation or two to increase its efficiency to the point where it can make a real dent in slowing down the relentless rise in global temperatures. And ironically, one new trait yeast needs to pull this off is to be able to survive better in the heat.

To efficiently turn the parts of the plants we can’t use (mostly anything to do with cellulose) into biofuels, we need for yeast to grow in cultures where enzymes are chopping cellulose into bite sized pieces at the same time.  These enzymes work best at temperatures that are a real struggle for wild type yeast.

In a new study in Science, Caspeta and coworkers isolated mutants better able to tolerate high temperatures by simply growing Saccharomyces cerevisiae at around 39.5° C for over 300 generations.  They did this in triplicate and ended up with strains that grew on average about 1.9 times faster than wild type yeast at these temperatures.

The next step was to use genome sequencing and whole-genome transcription profiling to figure out why these mutant strains were heat tolerant. This is where things got really interesting.

They found many genes that were affected, but identified ERG3 as the most important player. All of the thermotolerant strains that they selected contained a nonsense mutation in ERG3. And in fact, when they introduced an erg3 nonsense mutation into wild type yeast, they found that this engineered strain grew 86% as well as the original mutant strain at high temperatures. In other words, most of the heat tolerance of these strains came from mutations in the ERG3 gene.

This makes sense, as membrane fluidity is a key factor in dealing with higher temperatures, and ERG3 codes for a C-5 sterol desaturase important for membrane composition.  When Caspeta and coworkers looked at the membranes of the mutant strains, they found a buildup of the “bended” sterol fecosterol.  Since “bended” sterols have been shown to protect the membranes of plants and Archaea from temperature swings, this could be the reason that the mutant strains dealt with the heat so well.

So as we might predict, changing the composition of the cell membrane affects how the yeast respond to temperature. What we couldn’t predict is that in order to get to this point via artificial selection, the yeast had to have a second mutation in either the ATP2 or the ATP3 genes that actually make yeast less able to grow at higher temperatures.

Because ATP2 and ATP3 are needed for yeast to grow on nonfermentable carbon sources, the authors hypothesized that perhaps thermotolerance could not evolve while oxidative respiration worked at full speed. Consistent with this, Caspeta and coworkers found that their evolved thermotolerant strains were more susceptible to oxidative stress and could not grow on nonfermentable carbon sources. This was not true of the strain they engineered to only have the mutation in the ERG3 gene—it grew well at 40° C and could use nonfermentable carbon sources.

So not only have these authors found a mutant yeast with the superpower of growing at higher temperatures, but they also showed that sometimes engineering works better than natural selection at creating the mutant they want. Evolving sometimes requires passing through an intermediate that makes the final product less useful than it could have been. In this case it worked best to use a combination of artificial selection and engineering to build a better mutant.   

Even this super engineered yeast mutant won’t stop global warming in its tracks, but it might help us to slow it down enough so that natural systems have a chance to adapt. A superhero indeed… 

by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics

Categories: Research Spotlight

Tags: biofuel, evolution, thermotolerance, Saccharomyces cerevisiae

RNase P, Unmasked

August 28, 2014

No matter how fancy, all masks hide the identity of a wearer. And no matter how fancy an RNase P is, all it likely does is trim tRNA precursors. Images from Wikimedia Commons

Masks for a masquerade party come in a dazzling array of shapes and sizes.  And yet they all pretty much serve the same purpose — they hide the identity of the wearer.

Biology sometimes has its own dazzling array of cellular machines all doing the same thing.  One of the best examples of this is RNase P.   This enzyme trims tRNA precursors into mature tRNAs and has pretty much been around in one form or another since there were cells.  And yet, despite this common heritage and its one apparent job, it seems that no two are exactly alike.

In bacteria, RNase P is a piece of RNA that serves as the enzymatic component, complexed with a single protein.  Most Archaea and eukaryotes kept the RNA and added a varying number of protein subunits to make some wildly complex enzymes.  But in a few eukaryotes, the RNA has been dropped completely and a single protein substituted to provide the enzymatic activity. 

A new study out in PLOS Biology by Weber and coworkers shows that, despite this structural diversity, all the different forms of RNase P pretty much do the same thing.  Just like someone can hide who they are with any old mask, a cell can trim its tRNA precursors with any old RNase P.  Well, at least the simple RNase P of Arabidopsis thaliana, comprised of a single enzymatic protein subunit, can replace the enzymatic RNA and at least one protein subunit from the much more complex RNase P of our friend Saccharomyces cerevisiae!

This suggests that evolution has done something weird here.  It took what most likely started out as an RNA enzyme and made various changes to it over time.  Despite these changes, the enzyme kept doing the same thing: trimming tRNA precursors.  It is as if the enzyme went through a bewildering set of evolutionary changes and ended up at nearly the same place doing the same thing.   

How did Weber and coworkers arrive at this startling finding? Yeast RNase P consists of nine protein subunits and an RNA component that comes from the RPR1 gene.  The first thing Weber and coworkers did was to show that the lethal phenotype of a rpr1 knockout could be rescued by the single-subunit RNase P from either the plant Arabidopsis thaliana or the trypanosome Trypanosoma brucei.  The RNase P in these beasts consists of only a single polypeptide.

The authors next integrated the RNase P gene of A. thaliana into the genome of a yeast cell lacking both RPR1 and one of the protein subunits of RNase P, Rpr2p, and put it through a set of rigorous tests.  To their surprise, they found that this strain does a perfectly fine job of processing tRNA precursors.  There was no buildup of intermediates and, if anything, the A. thaliana RNase P proved to be a bit more efficient at trimming these tRNA precursors.

Of course just because the simpler RNase P can substitute for the RNA subunit of the more complex RNase P, that does not mean the two do the exact same thing.  It could be that the more complex form of RNase P has a broader set of functions, but that the only function absolutely required for life is the trimming of tRNA precursors.  But this does not appear to be the case.

Previous research showing that unprocessed forms of other RNAs accumulate at the restrictive temperature in an rpr1-ts mutant had suggested that yeast RNase P also processes a number of other RNAs besides tRNAs.  Since Weber and coworkers didn’t see these unprocessed forms accumulating in their strain, either the simple A. thaliana RNase P was able to process those other RNAs, or they’re actually not RNase P substrates. 

By analyzing the phenotypes of several different RNase P mutants, they showed that the other RNAs aren’t RNase P substrates; apparently their accumulation in the rpr1-ts mutant is an indirect effect.  All in all, these results show that the added complexity of yeast RNase P did not arise so that the enzyme could also process these other RNAs.

The authors next set out to see if there was any subtle difference between the two strains.  In other words, does replacing the RNA component of yeast RNase P with the catalytic protein subunit from A. thaliana have any effect on the yeast whatsoever? 

Weber and coworkers tested this by comparing the growth of the two strains under a wide range of conditions.  They saw no significant effects in any of the over 30 conditions tested.  If the yeast RNase P has any added features over the A. thaliana one, they are very, very subtle.

Pushing to see if they could find any differences, they even set the two up in direct competition to see which was the best suited for survival.  They did this by adding GFP to one or the other strain so that they could follow it, putting the two strains together, and growing them for many generations to see if one routinely outcompeted the other.  Neither did…it was a draw.  There appears to be no advantage to having the yeast RNase P despite its complexity!

This is weird.  It is almost like round trip evolution.  RNase P starts out as a single RNA that processes tRNA precursors.  Then as it moves around the tree of life, it picks up various bells and whistles and occasionally is even replaced by a protein.  And yet in the end, all RNase P’s are strangely equivalent.  As if all of that evolving was for naught!

Obviously there are still plenty of unanswered questions.  Why did yeast build up this complexity if there is seemingly no advantage?  And is the protein subunit superior to the RNA subunit?  If so, this last question would at least explain why a few beasts evolved away from the RNA catalytic subunit to the protein one – but still wouldn’t answer why all those proteins are glomming onto the perfectly adequate RNA that probably predates proteins.  More studies in yeast may help us “unmask” the answer to this fundamental question.

by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics

Categories: Research Spotlight

Tags: tRNA processing, RNase P, evolution, Saccharomyces cerevisiae

Slipping Through Haldane’s Sieve

May 22, 2014

Just as harsh panning can uncover hidden gold nuggets, so too can loss of heterozygosity reveal beneficial new recessive mutations. Image via Wikimedia Commons

Imagine you are panning for gold in a river and there are two kinds of nuggets.  One type is naked gold while the other is gold hidden inside of normal rock.  Pretty easy to figure out which nuggets you’ll gather first!

Now imagine instead that the process of panning is a rough one that knocks the shell off of the second type of nugget revealing the gold inside.  Now there won’t be any difference between the two.  You will be just as likely to keep both types of nuggets.

The same sort of situation applies to new beneficial mutations in a changing environment.  Back in 1927, J. B. S. Haldane predicted that the more dominant a mutation, the more likely it was to help a diploid beast adapt to a new environment.  The naked gold was more likely to be taken over the covered gold.

Gerstein and coworkers show in a new study that at least in the yeast Saccharomyces cerevisiae, Haldane’s sieve (as it is called) may not always apply.  The process of adapting to a new environment can strip away the dominant older allele, revealing the recessive one.  Loss of heterozygosity (LOH) uncovers the hidden gold of the recessive phenotype.

The authors had previously identified haploid mutants that were able to survive in the presence of the fungicide nystatin.  They mated these mutants to create either heterozygotes or homozygous recessive mutants and compared these to wild-type diploids growing either in the presence or absence of nystatin. 

Gerstein and coworkers found a wide range of effects of these mutations in the absence of nystatin.  Sometimes heterozygotes grew better than either homozygote, sometimes homozygous recessive strains did best, and sometimes wild type grew best.   Phenotypes were all over the map.

The story was very different in the presence of nystatin where only the homozygous recessives managed to grow.  This appears to contradict Haldane’s sieve.  Here there were no dominant mutations that allowed for survival.

Gerstein and coworkers found that some heterozygote replicates started to grow after a prolonged lag period.  A closer look at the heterozygotes that grew showed that they had lost the dominant allele so that they could now show the recessive phenotype and survive.  LOH had broken Haldane’s sieve. 

The authors found that the lower the nystatin levels, the more likely a population was to break through Haldane’s sieve.  They postulate that the populations survive longer at lower levels of nystatin, which increases the chances that a LOH will happen.  It is a race between survival and eliminating the dominant allele that keeps them from growing. 

The next step was to determine if LOH was common enough that populations with a small percentage of heterozygotes could survive.  They found that even in populations where only 2% were heterozygotes, around 5% of the 96 replicate populations managed to lose an allele and grow.  So even at low levels, a recessive mutation can give a population the advantage it needs to adapt and survive.

Combining the awesome power of yeast genetics with cheap sequencing is allowing scientists to test fundamental models of genetics that will unearth how populations adapt and survive in new environments.  We are finding those nuggets of scientific knowledge that have remained hidden.

Now of course, not every diploid is as numerous or as genetically flexible as yeast.  Cows, chickens, lizards, and people may all still be slaves to Haldane’s sieve.  We will need more studies to see if our recessive treasures can be uncovered in time to save us. 

by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics

Categories: Research Spotlight

Tags: antifungal resistance, evolution, Saccharomyces cerevisiae

How Yeast Populations Make the Cut

April 15, 2014

Kids like these can overcome some physical limitations with lots of hard work and practice. But yeast needs to stumble upon the right mutations to win out over its peers. Image from the U.S. Navy via Wikimedia Commons

Imagine that your dream is to be a professional basketball player.  Unfortunately for you, you are only five feet six inches tall and you can’t jump very high.  No matter how much you practice and work out, it is exceedingly unlikely you will be a starter for the Miami Heat.

Now imagine instead that you are six feet tall with a reasonable vertical jump.  Here, with enough effort you have a shot at beating out the guy with the genetic advantage of being six foot six inches high who doesn’t work as hard as you do.  Keep practicing and you might be passing the ball to LeBron James instead of him! 

In a new study in GENETICS, Frenkel and coworkers show that something similar can happen in yeast too.  If a population of yeast has some overwhelming advantage over a second population, the first will quickly outcompete the second every time.  But if the first population is just a bit better than the second, then the second can sometimes end up with a mutation that gives it an even better advantage than the first.  Now the first population is outcompeted and the second takes over.

Of course, when presented in a general way this is sort of obvious.  But Frenkel and coworkers set up their experiments in such a way that they got some hard numbers for just how much of an advantage one population needs to overcome to have a chance at winning.  If six feet is tall enough, what about five feet eleven inches?

The first step was to generate a number of mutants with different measured fitness advantages.  They selected mutant populations with advantages of 3, 4, 5, or 7%.  These populations were all tagged with a fluorescent marker.

They then seeded these mutants individually into 658 replicate reference populations that were tagged with a different fluorescent marker.  The mutants were seeded at a high enough level to prevent genetic drift from wiping them out.  The authors then followed each population for hundreds of generations by determining the levels of each population every 50 or so generations.

Their first finding was that mutants with a 7% advantage won out every time.  The reference population had no chance at getting a good enough mutation to beat it out.  No one is going to beat LeBron James out for his starting position with the Miami Heat.

Once the advantage was only 5%, around 16% of the time the second population won out.  As the advantage got smaller and smaller, the second population won out more and more often.   Even a genetically less gifted player has a shot at beating out the 12th guy on the Heat’s roster!

These results can tell us quite a bit about the mutational landscape of haploid Saccharomyces cerevisiae.  For example, from these data Frenkel and coworkers figured out that only populations that get mutations that give at least a 2% advantage have a chance at outcompeting other populations.  By assuming a mutation rate of 4X10-3, around 1 in 1000 mutations fit this bill, which might seem surprisingly high but is consistent with previous studies.  With a bit more hand waving, the authors hypothesize that disruption of something like 1 in 100 yeast genes is actually beneficial!

So yeast have a surprisingly level playing field.  Unless they are up against the equivalent of Kobe Bryant or Michael Jordan, they have a good shot at stumbling on a mutation that gives them an edge over their peers.    

by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics

Categories: Research Spotlight

Tags: mutation, population genetics, evolution, Saccharomyces cerevisiae

Gene Knockouts May Not Be So Clean After All

November 18, 2013

Imagine you have the instructions for building a car but you don’t know what any of the specific parts do.  In other words, you can build a working car but you don’t understand how it works.

If a cell were a car and you removed its radiator, it might adapt by evolving an air cooled system. If it happened soon enough, you might never figure out what the radiator did. Image by Joe Mazzola obtained from Wikimedia Commons.

One way to figure out how the car works would be to remove a part and see what happens.  You would then know what role that part played in getting a car to run.

So if you remove the steering wheel, you’d see that the thing runs into a wall.  That part must be for steering.  When you take out the radiator, the car overheats so that part must be for cooling the engine.  And so on.

Sounds like a silly way to figure out how the car works, but this is essentially one of the key ways we try to figure out how a cell works.  Instead of parts, we knock out genes and see what happens. A new study by Teng and coworkers is making us rethink this approach.

See, one of the big differences between a machine and a cell is that the cell can react and adapt to the loss of one of its parts.  And in fact, it not only can but it almost certainly will.

Each cell has gone through millions of years of evolution to adapt perfectly to its situation.  If you tweak that, the cell is going to adapt through mutation of other genes.  It is as if we remove the radiator from the car and it evolves an air cooling system like the one in old Volkswagen Bugs.

Teng and coworkers decided to investigate whether or not knocking out a gene causes an organism to adapt in a consistent way.  In other words, does removing a gene cause a selection pressure for the same subset of mutations that allows the organism to deal with the loss of the gene.  The yeast knockout (YKO) collection, which contains S. cerevisiae strains that individually have complete deletions of each nonessential gene, gave them the perfect opportunity to ask this question.

There have long been anecdotal reports of the YKO strains containing additional, secondary mutations, but the authors first needed to assess this systematically. They came up with an assay that could detect whether secondary mutations were occurring, and if so, whether separate isolates of any given YKO strain would adapt to the loss of that gene in a similar way.  The assay they developed had two steps.

The first step was to fish out individual substrains from a culture of yeast that started from a single cell in which a single gene had been knocked out.  This was simply done by plating the culture and picking six different, individual colonies.  Each colony would have started from a single cell in the original culture.

The second step involved coming up with a way to distinguish differently adapted substrains.  The first approach was to see how well each substrain responds to increasing temperatures.  To do this, they looked for differences in growth at gradually increasing temperatures using a thermocycler.

They randomly selected 250 YKO strains and found that 105 of them had at least one substrain that reproducibly responded differently from the other substrains in the assay.  In contrast, when they looked at 26 isolates of several different wild type strains, including the background strain for the YKO collection, there were no differences between them. This tells us that the variation they saw in the knockout substrains was due to the presence of the original knockout.

So this tells us that strains can pretty quickly develop mutations but it doesn’t tell us that they are necessarily adapting to the knocked out gene.  To see if parallel evolution was indeed taking place, the authors chose to look at forty strains in which the same gene was independently knocked out.  They found that 26 of these strains that had at least one substrain with the same phenotype, and fifteen of those had mutations that were in the same complementation group.  So these 15 strains had evolved in similar ways to adapt to the loss of the same gene.

Teng and coworkers designed a second assay independent of the original heat sensitivity assay and tested a variety of single knockout strains.  They obtained similar results that support the idea that knocking out a gene can lead cells to adapt in similar ways.  This is both good and bad news.

The bad news is that it makes interpreting knockout experiments a bit trickier.  Are we seeing the effect of knocking out the gene or the effect of the secondary mutations that resulted from the knockout?  Are we seeing the loss of the radiator in the car or the reshaping that resulted in air cooling?  We may need to revisit some earlier conclusions based on knockout phenotypes.

The good news is that not only does this help us to better understand and interpret the results from yeast and mouse (and any other model organism) knockout experiments, it also gives us an insight into evolution and maybe even into the parallel evolution that happens in cancer cells, where mutations frequently co-occur in specific pairs of genes.  And while we may never be able to predict if that knock you hear in your engine really needs that $1000 repair your mechanic says it does, we may one day be able to use results like these to predict which cells containing certain mutated genes will go on to cause cancer and which ones won’t.   

by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics

Categories: Research Spotlight

Tags: deletion collection, evolution, Saccharomyces cerevisiae

When One Spark Isn’t Enough

August 07, 2013

A single mutation, just like a single spark, is more likely to fizzle out.

As anyone who has ever tried to start a fire with flint knows, a single spark is rarely enough.  You need to get a bunch of sparks all working at once to end up with that roaring campfire.  And with the wrong kindling or wood, even lots of powerful sparks just can’t get it done.

A new study in the yeast S. cerevisiae by Lang and coworkers suggests that evolution may be similar.  A single helpful DNA change may not be enough to give an individual yeast that leg up it needs to spread through the population.  Turns out that more often than not it needs something like 5-7 mutations.   And again, even that may not be enough if the rest of its DNA isn’t up to par.  A set of powerful sparks on soggy wood still won’t light a fire.

Lang and coworkers followed 40 different yeast populations for a thousand generations as the yeast adapted to a new environment (rich medium).  They sequenced each population to 100-fold depth at 12 different time points.  Not only did this allow the researchers to watch mutations rise and fall over time, it also let them screen out sequencing errors.  Real mutations will correlate over time, sequencing errors won’t.

The key finding of their research was that mutations that increased in the population over time almost always came in bunches (or cohorts) and that not all the mutations were beneficial.  Neutral mutations invariably hitchhiked along with strongly beneficial ones.

A great example of this involves the ELO1 and GAS1 genes.  These two mutations arose together in a yeast population but when the researchers looked at each individually, only GAS1 was beneficial.  ELO1 appeared to go along for the ride.

Another key point of this study is that mutations do not happen in a vacuum…beneficial mutations only “catch” in the context of a good background.  This is clearly shown in one of the populations they followed. 

In this population, yeast with a mutation in the SPC3 gene began to spread through the population.  After about 300 generations, though, a second yeast with mutations in the WHI2 and ROT2 genes began to outcompete the SPC3 mutant.  If things stayed like this, the SPC3 mutation would disappear from the population even though it was obviously helpful.

What happened instead was that a yeast with the SPC3 mutation developed a useful mutation in the YUR1 gene.  This combination was strong enough for this yeast to stay in the game until one of them developed a third mutation in the WHI2 gene.  This triple threat proved too much for the yeast with the mutations in the WHI2 and ROT2 genes – they were driven to extinction.

No wonder people refer to evolution as a dynamic process!  This example shows just how tumultuous it actually is.  Even helpful mutations like the ones in ROT2 and WHI2 can disappear over time if they happen in a weaker background.  And presumably even potentially harmful mutations can spread if they hitchhike along with a cohort of strongly beneficial mutations.

These results not only shed light on how evolution works, but could also spark other discoveries on how cancer progresses, how bacteria become resistant to antibiotics, and how viruses deal with our immune system, just to name three.  And that could help kindle a brighter future.  

by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics

Categories: Research Spotlight

Tags: evolution, Saccharomyces cerevisiae

Flip-Flopping Yeasts

June 20, 2013

We all have certain things we can’t live without. But what’s essential to one person may be completely trivial to another. For example, a teenager who can’t live without his video games is fine without that antique tea set,  while the opposite may be true for grandma. 

Just like all of the pieces of this set are essential to grandma, so too are all of the proteins in an essential complex usually essential to yeast (and maybe to us too!).

In a new article in Genome Biology and Evolution, Ryan and coworkers find that the same holds true for different yeast species.  One gene that is essential in one yeast is dispensable in another.  And furthermore, they tend to be essential or nonessential in sets. Just like grandma’s tea set and the teenager’s games.

It’s been seen in S. cerevisiae that the genes that encode the proteins in a complex tend to be either mostly essential or mostly nonessential.  It is like the teapot and the cups and saucers all being essential to grandma or all being nonessential to the teenager.  This is called modular essentiality.

Ryan and coworkers found that if some protein complex is essential in one yeast, most or all of those genes will be essential in that species.  On the other hand, if that protein complex is dispensable in a different yeast, then most or all of those genes will be nonessential.  The genes encoding the entire complex flip together from essential to nonessential – again, just like the tea set.  It isn’t as if one of the tea cups happens to stay essential when the teenager gets ahold of the set! 

To do this work, Ryan and coworkers started out by looking at S. cerevisiae.  As all of us here at SGD know, this was an excellent choice!  But not just because we work on it…

Because the S. cerevisiae genome sequence has been available for quite a while, we know which genes are essential to life, which genes interact genetically, and which proteins interact physically with each other. We also have a very good list of the protein complexes that exist in yeast and what their subunits are.

Ryan and coworkers used updated data to confirm that modular essentiality exists in S. cerevisiae.  Most of the proteins in an essential complex tend to come from essential genes and most of the proteins in nonessential complexes come from nonessential genes.  There is very little overlap…the tea set does not often contain a video game!

Next the authors asked whether this modular essentiality is found in other species too. At the moment, Schizosaccharomyces pombe, or fission yeast, is the only other eukaryote with complete data on the essentiality of genes. Although it’s also a single-celled yeast, S. pombe is about as far away from S. cerevisiae as you can get and still be a yeast. The two are thought to have diverged as much as 400 million years ago.

Even though it is so different, S. pombe also shows modular essentiality. And using an incomplete set of data from knockout mice, the authors see a similar pattern! So it looks like modular essentiality is at least conserved across fungi, and may be universal.

Next they asked if complexes that have flipped from essential to nonessential over time still maintain their modular essentiality.  Do all the tea cups become nonessential, or just some of them? 

When grandma tidies up the video games, none of them will seem important to her. The same thing happens when a complex switches from essential to nonessential as a species evolves.

Most (83%) of the genes that are present as one-to-one orthologs in both yeasts are either essential in both or nonessential in both. Ryan and coworkers focused on the other 17%, where a gene was essential in one species but not in the other.

In the cases where essentiality is “flipped” between the species, whole protein complexes tend to flip as a unit. The subunits of a complex that is nonessential in budding yeast are mostly nonessential, while the subunits of the analogous complex in fission yeast are mostly essential.

An example of this is the large subunit of the mitochondrial ribosome. Mitochondrial translation is optional for S. cerevisiae, but obligatory for S. pombe. In keeping with this, almost all the proteins that make up the S. cerevisiae large mitochondrial ribosomal subunit are nonessential. In S. pombe, the situation is flipped.

So the essentiality of a complex mirrors the lifestyle of its owner, just like the teenager and his grandmother.  The two yeasts, with their different lifestyles, place different importance on the mitochondrial ribosome. This wasn’t a big surprise, since this lifestyle difference was already known. But other complexes that are flipped between the species may point to things that we don’t yet know about their physiology.

These results support the idea that modular essentiality is universal, which would mean that in various organisms we can expect that mutants in subunits of a complex will share the same phenotype, disease association, and drug sensitivity. Obviously there are important implications here for antifungal drug design or for disease treatment: if you want to stop a complex from working, any of its subunits (or perhaps several at the same time) might prove to be good targets.

But another bigger point is how much we can learn from a deep understanding of an organism’s genome.  By teasing apart what is essential and what isn’t we can learn a lot about the beast we’re studying.  And someday, maybe a lot about ourselves.

by Maria Costanzo, Ph.D., Senior Biocurator, SGD

Categories: Research Spotlight

Tags: essentiality, evolution, Saccharomyces cerevisiae