New & Noteworthy

Redesigning Yeast One Chromosome at a Time

April 03, 2014

Everyone who reads our blog knows how awesome the yeast Saccharomyces cerevisiae is.  Without this little workhorse we would almost certainly not understand ourselves as well as we do now.  It is an indispensable tool in figuring out how eukaryotes work.

And of course yeast is much more than that.  It makes our bread fluffy and our drinks alcoholic.  It can be manipulated into making medicines like artemisinin, a powerful anti-malarial drug, or biofuels or whatever else we can think of.  It is the Swiss Army Knife of useful organisms.

Even with all of this fanfare, everyone knows yeast has its limitations.  It is a powerful tool but it could be improved.  For example, it would be nice if researchers could more easily manipulate its DNA to speed up the introduction of beneficial traits, add new biosynthetic pathways, or to do the kinds of experiments that will help one day cure cancer or Alzheimer’s disease.  This is where Sc2.0 comes in.

Sc2.0 is an idea that has been kicking around for the last decade or so.  First proposed by Ron Davis of Stanford University, the idea is to synthesize artificial yeast chromosomes to make yeast more useful.  Eventually the idea would be to recreate every yeast chromosome and intelligently redesign the genome for our own purposes. And maybe even to add new artificial chromosomes so we can easily add whatever genes we want.

In a new study out in Science, Annaluru and coworkers have taken a major step forward in the Sc2.0 project by replacing all 316,617 base pairs of yeast chromosome III with a 272,871 base pair synthetic version, synIII.   That leaves only 15 chromosomes and around 12.2 million base pairs before we have yeast with completely manmade DNA.

Annaluru and coworkers managed to do this with the help of a bunch of undergraduate students and yeast’s love of homologous recombination.  The first step was to have undergraduates synthesize around 30,000 base pairs each in the “Build a Genome” class at Johns Hopkins.  It took 49 students around 18 months to pull this off for synIII.

Basically they used 60-mer and 79-mer oligonucleotides to PCR up 750 base pair building blocks.  These pieces of DNA were designed so that they could be assembled into 2,000-4,000 base pair minichunks.  The final step was to transform yeast with an average of twelve of these minichunks and to let the yeast use homologous recombination to replace its native DNA sequence with the added DNA.  After 11 rounds of transformation, the yeast now had an artificial chromosome.

As you may have guessed, this chromosome is not exactly the same as the one it replaced.  To eventually free up a codon for repurposing later, all 43 of the TAG stop codons were converted to TAA.  When this is done with all of the chromosomes, researchers will now have a codon they can use to change this yeast’s fundamental genetic code.  This might allow for adding novel amino acids to proteins or even prevent viruses from infecting the new yeast.

Annaluru and coworkers also introduced 98 loxP sites which in the presence of estradiol will cause the yeast to undergo rapid DNA change.  The hope is that scientists will be able to harness SCRaMbLE (synthetic chromosome rearrangement and modification by loxP-mediated evolution), as it has been named, to more quickly evolve useful traits in yeast for both study and biotechnological uses.

As a final step, the researchers cleaned up the chromosome by removing 21 retrotransposons and many introns and by moving 11 tRNA genes to a neochromosome.  They now had created a leaner, meaner chromosome III. 

The next obvious question was whether or not all of these changes affected the yeast.  Despite looking very carefully, Annaluru and coworkers could find little that was different between strains carrying natural and synthetic chromosomes.  They both grew similarly under 21 different conditions in terms of growth curves, colony size, and cell morphology, and had very similar transcription profiles.  But they weren’t identical.

For example, the strain with synIII grew slightly less well in the presence of high sorbitol, and showed differences in expression from wild type in 10 out 6,756 transcripts.  Of these ten, eight were intentionally altered in the creation of synIII and so were expected.  The two unexpected changes were a ~16-fold decrease in the expression of HSP30 on synIII and a ~16-fold increase in the expression of PCL1 on chromosome XIV.

Since all of these changes had such a small effect on the yeast, it is a green light for plowing ahead with creating yeast with completely manmade DNA.  Currently four other chromosomes, II, V, VI, and XII, are nearly done and the design work has been completed for chromosomes I, IV, VII, and XI (see an overview of the project).  It will only be a matter of time before we have a strain of yeast with completely synthetic DNA.  Scientists are making a powerful tool even better…who knows what this new strain will help us discover.

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

Categories: Research Spotlight

Tags: Saccharomyces cerevisiae, synthetic biology, teaching

Silly Sod’s Two Jobs

March 27, 2014

Most SGD users are probably too young to remember Saturday Night Live’s early years.  One very funny commercial parody involved Gilda Radner and Dan Aykroyd arguing over a product called Shimmer.  Gilda argues that it is a floor wax while Dan says it is a dessert topping.  In comes Chevy Chase to tell them that it is both.  Not quite as funny as Bassomatic, but still hilarious.

In a new study, Tsang and coworkers show something similar for the enzyme Sod1p.  Most people know Sod1p as an enzyme that protects the cell and its DNA by directly deactivating harmful reactive oxygen species (ROS) like superoxide.  Turns out that it may also be a transcription factor.

Now these two jobs aren’t quite as disconnected as a dessert topping and floor wax.  When Sod1p acts as a transcription factor, it is regulating genes that affect a cell’s response to ROS.  It is actually using its two functions to attack the same problem on multiple fronts.

Tsang and coworkers started out by looking at what happens to nuclear DNA under oxidative stress, using the Comet and TUNEL DNA damage assays. They found that endogenous and exogenous ROS caused DNA damage that was much worse in the sod1 null mutant – in other words, Sod1p protected the cells’ DNA. Using immunofluorescence, they also showed that Sod1p quickly went into the nucleus in the presence of ROS.  But if they restricted Sod1p to the cytoplasm by adding a nuclear export signal, the protein no longer protected the DNA.  In fact, it did no better than a strain deleted for SOD1.

In the course of these experiments one of the ways the researchers induced nuclear localization was with a burst of hydrogen peroxide.  But since hydrogen peroxide isn’t a substrate of the enzyme Sod1p, Tsang and coworkers next wanted to figure out how Sod1p got its signal to go nuclear.

Previous work had shown that SOD1 genetically interacted with MEC1, a yeast homolog of ATM kinases which sense oxidative stress.  They deleted MEC1 and found that Sod1p was trapped in the cytoplasm, unable to protect the cell’s DNA from damage.  This result was confirmed in human cells by showing that Sod1p only went nuclear if the cell made ATM kinase.

Tsang and coworkers suspected that this interaction might happen through a protein kinase called Dun1p, whose human homolog is a Mec effector. They confirmed a previous mass spectrometry result that showed Sod1p interacted physically with Dun1p.  And indeed, when DUN1 was deleted, Sod1p was again stranded in the cytoplasm.  Further work showed that Dun1p does its job by phosphorylating Sod1p on two serine residues, S60 and S99. When both these serines are mutated to alanine, preventing phosphorylation, less of the mutant Sod1p makes it into the nucleus. 

Using DNA microarrays, Tsang and coworkers next showed that SOD1 was required to activate 123 genes needed by the cell to respond to hydrogen peroxide.  These genes fell into five categories: oxidative stress, replication stress, DNA damage response, general stress response and Cu/Fe homeostasis.  The final experiment used chromosomal immunoprecipitation (ChIP) to show that in the presence of hydrogen peroxide more Sod1p was bound at the promoters of two of these genes, RNR3 and GRE2, but not the control gene ACT1. 

Of course, the authors have only looked at two of the 123 genes and an obvious next step is to figure out how many of the 123 have more Sod1p bound to their promoters in the presence of hydrogen peroxide.  Still, if these results can be confirmed and expanded they will suggest that Sod1p is able to combat oxidative damage in two completely different ways. 

In the first it uses its enzymatic activity to directly inactivate the ROS superoxide, while in the second it helps the cell respond to other ROS apparently by acting as a transcription factor.  While the jobs themselves are not as different as a floor wax and a dessert topping, how Sod1p goes about getting each job done is.  “Calm down you two, Sod1p is an enzyme AND a transcription factor.”

In addition to these two roles, we’ve written before about yet another regulatory role for Sod1p: it regulates glucose repression by binding to two kinases and stabilizing them. This is truly an overachiever of a protein!

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

Categories: Research Spotlight

Tags: oxidative stress, Saccharomyces cerevisiae, transcription

Guns vs. Butter in Cells

March 19, 2014

Life is a set of tradeoffs for people, countries, and even cells.  For example, governments need to decide how much money to dedicate to defense and how much to economic growth.  Too much on defense and your country fails, because defense spending sucks up so many resources that your country can no longer afford to pay for anything else.  And of course if you spend too little on defense, someone who spent a bit more can come and take you over.

No country lives in a vacuum though—how much to spend on defense and how much on growth depends on the country’s situation.  If you are the Ewoks living next to an Imperial shield generator, you’d better sacrifice some growth for defense.  But once the Death Star has blown up and the Empire is swept away, you probably focus more on growth (until a new Sith lord arrives). 

This guns vs. butter debate plays out at the cellular level too when it comes to protecting DNA from mutations.  If cells expend too much energy to protect their DNA they sacrifice growth, but if they spend too little, they develop too may harmful mutations to survive.  And just like with countries, how much protection a cell’s DNA needs depends on its environment.

If cells need to adapt quickly to a changing environment, a high rate of mutation is favored.  These cells are more likely to develop a mutation that gains them an advantage over their slower mutating brethren.

A new study by Herr and coworkers in the latest issue of GENETICS calculates the upper limit of the rate of mutation in a diploid yeast.  In other words, they figure out how little “spending” on defense these yeast can get away with and survive.

They find that diploid yeast can deal with a 10-fold higher rate of mutation as compared to haploid yeast.  This makes sense, since the extra gene copy afforded by being diploid can mask a recessive lethal mutation, but this study is the first to give this idea hard numbers.

The authors had previously generated a number of mutations in POL3, the yeast gene for DNA polymerase δ, that affect its ability to find and/or fix any mistakes made during DNA replication.  The study first focused on two mutations affecting accuracy, pol3-L612G and pol3-L612M, and one mutation affecting proofreading, pol3-01.  The accuracy mutations caused about a 10-fold increase in the mutation rate, while the proofreading mutation caused anywhere from a 20-100-fold increase.  Neither was enough to seriously affect a diploid’s growth.

The next step was to combine accuracy and proofreading mutations into the same gene to figure out if the combination resulted in a higher mutation rate.  The authors suspected that it did when they discovered that even though the heterozygotes were fine, their spores were inviable.  The POL3/pol3-01,L212M and POL3/pol3-01,L212G strains sporulated just fine, but none of the spores could germinate and grow. 

One way to explain this was that the double mutation increased the error rate to the point that it would kill off haploids but not diploids.  By looking at mutations in the hemizygous CAN1 gene they could see that the mutation rate in these diploids was indeed at around the haploid threshold. In terms of the CAN1 gene, this mutation rate was around 1X10-3 can1 mutations/cell division.

They next determined the mutation rate by sequencing the genomes of each mutant as well as the wild type.  They found a single T-G mutation in the wild type, 1535 point mutations in POL3/pol3-01,L212M and 1003 mutations in POL3/pol3-01,L212G.  From this they calculated a mutation rate of around 3-4X10-6/base pair/generation. 

Even though this level of mutation kills haploids but not diploids, this does not mean the diploids escaped unscathed.  When the heterozygous diploid colonies were subcloned the resulting colonies were variable in size, indicating that their higher mutation rate was catching up with them.  This high mutation rate was making them sick. 

Given this result, it wasn’t surprising that diploid homozygotes of each double mutant could not survive—the mutation rate was now too high.  The strains homozygous for pol3-01,L212M managed to get to around 1000 cells before petering out.  Strains homozygous for pol3-01,L212G did even worse—they only made it to around 10 cells.

In a final set of experiments Herr and coworkers used a variety of other mutations to tweak these mutation rates to find the threshold at which diploids fail to survive.  Some of these mutations were in POL3 while others were deletions of the MSH2 and/or DUN1 genes.  After testing many different combinations, they found that these yeast did pretty well up to around 1X10-3 can1 mutations/cell division (the haploid threshold rate).  Then, from 1X10-3 to 1X10-2 can1 mutations/cell division there began a rapid drop off with little to no growth at the end. 

So as might be expected, diploids can deal with a significantly higher mutation rate than can haploids.  But even though they can, wild type yeast in the lab still have a very low mutation rate.  It is like they are living near the Imperial city planet of Coruscant.  They are willing to expend the energy to keep their DNA protected.

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

Categories: Research Spotlight

Tags: DNA replication, mutation, Saccharomyces cerevisiae

A Heartfelt Need for Copper

March 06, 2014

Imagine the heater at your house is run by a homemade copper-zinc battery.  You are counting on a delivery of a copper solution that will keep the thing going.  Unfortunately it fails to come, which means the battery doesn’t work and you are left out in the cold. 

Turns out that something similar can happen in cells too.  The respiratory chain that makes most of our energy needs copper to work.  In a recent study, Ghosh and coworkers showed that if Coa6p doesn’t do its job delivering copper to the respiratory chain, the cell can’t make enough energy.

This isn’t just interesting biology.  In this same study, the researchers showed that mutations in the COA6 gene cause devastating disease in humans and zebrafish. And their discovery that added copper can cure the “disease” in yeast just might have therapeutic applications for humans.

The respiratory chain is a group of large enzyme complexes that sit in the mitochondrial inner membrane and pass electrons from one to another during cellular respiration. This process generates most of the energy that a cell needs.  Hundreds of genes, in both the nuclear and mitochondrial genomes, are involved in keeping this respiratory chain working.

Yeast has been the ideal experimental organism for studying these genes, because it can survive just fine without respiration. If it can’t respire for any reason, yeast simply switches over to fermentation, generating the alcohol and CO2 byproducts that we know and love.

Human cells aren’t as versatile though. Genes involved in respiration can cause mitochondrial respiratory chain disease (MRCD) when mutated. This is one of the most common kinds of genetic defect, with over 100 different genes known so far that can cause this phenotype.

Ghosh and colleagues wondered whether there were as-yet-unidentified human genes involved in maintaining the respiratory chain. They reasoned that any such genes would be highly conserved across species, because they are so important to life, and that the proteins they encoded would localize to mitochondria.

One of the candidates, C1orf31, caught their eye for a couple of reasons.  First, some variations in this gene had been found in the DNA of a MRCD patient.  And second, the yeast homolog, COA6, encoded a mitochondrial protein that had been implicated in assembly of one of the respiratory complexes, Complex IV or cytochrome c oxidase.

They first did some more detailed characterization of COA6 in yeast.  They were able to verify that the coa6 null mutant had reduced respiratory growth because it had lower levels of fully assembled Complex IV.

They also looked to see what happens in human cell culture.  When they knocked down expression of the human homolog, they also saw less assembly of Complex IV. This suggested that the function of this protein is conserved across species.

Next they turned to a sequencing study of an MRCD patient who had, sadly, died of a heart defect (hypertrophic cardiomyopathy) before reaching his first birthday. The sequence showed a mutation in a conserved cysteine-containing motif of COA6. To see whether this might be the cause of the defect, they created the analogous mutation in yeast COA6. The mutant protein was completely nonfunctional in yeast.

To nail down the physiological role of COA6 in a multicellular organism, they turned to zebrafish. The embryos of these fish are transparent, so it’s easy to follow organ development. Given the phenotype, the fact that they can live without a functional cardiovascular system for a few days after fertilization was important too.

When the researchers knocked down expression of COA6 in zebrafish, they found that the embryos’ hearts failed to develop normally and they eventually died. The abnormal development of the fish hearts paralleled that seen in the human MRCD patient carrying the C1orf31/COA6 mutation. And reduced levels of Complex IV were present in the fish embryos.

Going back to yeast for one more experiment, Ghosh and colleagues decided to see whether Coa6p might be involved in delivering copper to Complex IV. They knew that Complex IV uses copper ions as a cofactor, and furthermore Coa6p had similarities to several other yeast proteins that are known to be involved in the copper delivery.

They tested this by supplying the coa6 null mutant with large amounts of copper. Sure enough, its respiratory growth defect and Complex IV assembly problems were reversed.  The delivery of copper kept the energy flowing in these cells. And this result showed that Coa6p is involved in getting copper to Complex IV.

These experiments showcase the need for model organism research even in the face of ever more sophisticated techniques applied to human cells. The mutation in human C1orf31/COA6 was discovered in a next-generation sequencing study, but yeast genetics established the relationship between the mutation and its phenotype. The zebrafish system allowed the researchers to follow the effects of the mutation in an embryo from the earliest moments after fertilization. And the rescue of the yeast mutant by copper supplementation offers an intriguing therapeutic possibility for some types of MRCD. Just another testament to the awesome power of model organism research!

YeastMine now lets you explore human homologs and disease phenotypes.  Enter “COA6” into the template Yeast Gene -> OMIM Human Homolog(s) -> OMIM Disease Phenotype(s) to link to the Gene page for human COA6 (the connection between COA6 and disease is too new to be represented in OMIM).  To browse some diseases related to mitochondrial function, enter “mitochondrial” into the template OMIM Disease Phenotype(s) -> Human Gene(s) -> Yeast Homolog(s).

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

Categories: Research Spotlight, Yeast and Human Disease

Tags: respiration, Saccharomyces cerevisiae, yeast model for human disease, zebrafish

Passing the Hog: How a Long Noncoding RNA Helps Yeast Respond to Salt

February 25, 2014

Most people know that Incans relied on human runners to get messages across their empire.  Basically they had runners stationed at various places and one runner would hand the message off to the next.  This relayed message could then quickly travel across the country.

As shown in a new study by Nadal-Ribelles and coworkers, it turns out that something similar happens in yeast when the CDC28 gene is turned up in response to high salt.  In this case, the runner is the stress activated protein kinase (SAPK) Hog1p and it is stationed at the 3’ end of the gene.  When the cell is subjected to high salt, the message is relayed from the 3’ end of the CDC28 gene to its 5’ end by the Hog1p kinase.  The end result is about a 2-fold increase in the amount of Cdc28p made, which allows the cell to enter the cell cycle more quickly after the salty insult.

Unlike the Incans who had their paths all set up in front of them, poor Hog1p has to build its own path.  It does this by activating a promoter at the 3’ end of the CDC28 gene that produces an antisense long noncoding RNA (lncRNA) that is needed for the transfer of the Hog1p.  It is as if our Incan runner had to build a bridge over a gorge to send his message.

This mechanism isn’t peculiar to the CDC28 gene either.  The authors in this study directly show that something similar happens with a second salt sensitive gene, MMF1.  And they show that a whole lot more lncRNAs are induced by high salt in yeast as well.

Nadal-Ribelles and coworkers started off by identifying coding and noncoding regions of the yeast genome that respond positively to high salt.  The authors found that 343 coding regions and 173 noncoding regions were all induced at 0.4 M NaCl.   Both coding and noncoding regions required the SAPK Hog1p for activation. 

The authors next focused on CDC28 and its associated antisense lncRNA.  After adding high salt, Nadal-Ribelles and coworkers found that Hog1p was both at the start and end of the CDC28 gene – as would be expected, since both CDC28 and the antisense lncRNA required this kinase for transcriptional activation. 

Things got interesting when they were able to prevent the lncRNA from being made.  When they did this, Hog1p was missing from both the 5′ and 3′ ends of the CDC28 gene and as expected, activation was compromised.  But Nadal-Ribelles and coworkers showed that expressing the lncRNA from a plasmid did not allow for CDC28 activation. It appears that where the lncRNA is made is just as important as whether it is made.

Through a set of clever experiments, the authors showed that not only does the lncRNA need to be made in the right place, but it needs to be activated in the right way.  When they set up a system where the lncRNA was induced in the right place using a Gal4-VP16 activator, CDC28 was not induced by high salt.  A closer look showed that this was most likely due to a lack of Hog1p at the start of the CDC28 gene.

The situation was different when they activated the lncRNA with a Gal4-Msn2p activator which uses Hog1p to increase expression.  In this case, CDC28 now responded to high salt and Hog1p was present at both the start and end of the CDC28 gene.  But this activation went away if they added a terminator which prevented the full length lncRNA from being made. 

Phew, that was a lot!  What it means is that for there to be a Hog1p at the business end of the CDC28 gene, there needs to be one at the 3’ end.  It also means that for the Hog1p to get to the start of the CDC28 gene, the antisense lncRNA needs to be made.

This would all make sense if maybe the lncRNA was involved in DNA looping, which could get the Hog1p from the end of CDC28 to the start where it can do some good.  Nadal-Ribelles and coworkers showed that this indeed was the case, as CDC28 activation required SSU72, a key looping gene.  When there was no Ssu72p in a cell, salt induction of CDC28 was severely compromised.

So it looks like an antisense lncRNA in yeast is being used as part of a looping mechanism to provide the cell with a quick way to start dividing once it has dealt with its environmental insult.  The authors show that yeast that can properly induce their CDC28 gene enter the cell cycle around 20 minutes faster than yeast that cannot induce the gene.  The cells are poised for a quick recovery.

And this is almost certainly not merely a yeast phenomenon.  Some recent work in mammalian cells has implicated lncRNAs in recruiting proteins involved in controlling gene activity through a looping mechanism as well (reviewed here).  Now that the same thing has been found in yeast, scientists can bring to bear all the powerful tools available to dissect out the mechanism(s) of lncRNA action.  And that’s far from a loopy idea…

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

Categories: Research Spotlight

Tags: DNA looping, lncRNA, Saccharomyces cerevisiae, transcription

Signaling in a Crowd

February 18, 2014

There are two very different kinds of sports in the Winter Olympics (and in all sporting competitions really).  In one set, it is the athletes alone out on the ice or sliding down the slope, trying to get the best time they can.  They can only use themselves as the motivator.

In another set of sports, like speed skating, athletes compete directly with one another.  Here they can use each other to push themselves to go faster, farther, etc.

The key to each is obviously the proximity of other athletes.  If there are a bunch of athletes around you, you will all do better by feeding off each other’s signals.  If you are by yourself, then only you can produce the signals to motivate yourself to go faster.

Youk and Lim show in a new study that the same sort of thing happens in cells that can both secrete and sense the same signal.  If there aren’t a lot of cells around they tend to signal themselves, but in a crowded place, they are all signaling each other. 

This may seem a bit esoteric but it really isn’t.  These sorts of “secrete-and-sense” systems are common in biology.  Cell types from bacteria to our own T cells have them, and they allow for a surprisingly wide range of responses.  Understanding how these systems work will explain a lot of biology and, perhaps, help scientists create new sensing systems for bioengineered beasts.

Youk and Lim used our favorite organism Saccharomyces cerevisiae to study this widespread signaling system.  They created a bevy of strains that can either secrete and sense alpha factor or that can only sense the pheromone.  They grew varieties of these two strains together under various conditions to determine when the “secrete-and-sense” strains could also signal to the “sense only” strains.  Like our athletes, the cell concentration was important.  But so too were the levels of alpha factor and receptor.

The authors first created a strain that senses the presence of alpha factor with the Ste2p receptor and in response turns on GFP through the FUS1 promoter.  (The strain is deleted for FAR1 to prevent cell cycle arrest.) As expected, increasing amounts of alpha factor resulted in increased levels of GFP.

It is from this strain they created their “secrete-and-sense” and “sense only” strains.  The “secrete-and-sense” strain included a doxycycline inducible promoter driving the alpha factor gene.  The more doxycycline, the more alpha factor it makes, resulting in more GFP.  To tell the two strains apart in experiments, they added a second reporter, mCherry, under a constitutive promoter to the “sense only” strain.  Now in their experiments they can distinguish between the strains that glow only green and those that glow red and, sometimes, green.

The first experiment was simply to see what effect differing cell and alpha factor concentrations had on the two strains’ ability to glow green.  At low cell and doxycycline concentrations, only the “secrete-and-sense” strain glowed green.  This makes sense, as too little alpha factor was made to get to the relatively distant neighbors.  At high cell and doxycycline concentrations, both glowed green almost indistinguishably.  Here the system was flooded with enough alpha factor for everyone to respond.

The results were less binary at either low cell and high doxycycline concentrations or high cell and low doxycycline concentrations.  Under either of these conditions, the “sense only” strain did glow green although at a much slower rate.

Youk and Kim didn’t stop there.  They also tested whether the amount of receptor affected these results.  When the two strains expressed high levels of receptor, the amount of alpha factor didn’t matter at low cell concentrations—only the “secrete-and-sense” strain glowed green.  This makes sense as the strain can quickly suck up any amount of alpha factor it makes.  Again at high cell concentrations the differences disappear.

In a final set of experiments the authors created positive feedback loops and signal degradation systems, which are both very common in nature.  The positive feedback loop was created by putting the doxycycline activator, rtTA, under the control of doxycycline, and a signal degradation system was engineered using Bar1p, a protease that degrades alpha factor.  Using these systems they were able to show that at low cell concentration, low Bar1p expression, and strong positive feedback, individual cells were either on or off.  This sort of activity may be important in nature, where under certain conditions a response may be beneficial and in others a response may not.  This bet hedging means that the population can survive under both sets of conditions.

It is amazing that such a simple set of conditions can lead to so many different responses, almost as varied as the performances of Olympic athletes.  These findings not only help to explain how these deceptively simple systems work and why they are so common in nature, but might also be incredibly useful in setting up synthetic secrete-and-sense circuits for biotechnology applications.  

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

Categories: Research Spotlight

Tags: pheromone, Saccharomyces cerevisiae, signal transduction

Studying the Ballistics of Yeast Mutagenesis

February 06, 2014

Fireworks shells all pretty much look the same from the outside. They definitely all make the same boom when they’re launched. But when they burst in the air, each different kind creates a different shimmering pattern.

It turns out that the same is true for yeast strains carrying mutator alleles.
These are mutant alleles of genes that normally stop mutations from happening. When these genes are disabled, a strain eventually accumulates lots of extra mutations.

Mutator strains tend to look similar from the outside; many are deficient in DNA replication and repair pathways. But, in a new paper in GENETICS, Stirling and coworkers show that like different firework shells, each strain ends up with a distinct pattern of secondary mutations bursting across their respective genomes. Not only is this fascinating information about how yeast maintains its genomic integrity, but it may also provide valuable insights into how cancers progress.

Mutator genes have been found previously using the knockout collection of mutations in nonessential genes. But, not surprisingly, many genes required for genome maintenance are essential to life. So the first step by Stirling and coworkers was to expand the list of mutator genes by screening conditional mutant alleles of essential genes.

Using an assay for mutation frequency that counts canavanine resistance mutations arising in the CAN1 gene, they came up with 47 alleles in 38 essential genes that caused a mutator phenotype.  But this standard assay for mutator phenotype has its limitations: the only mutations that can be detected are those that fall in or near the CAN1 gene, and inactivate it. So that they could look at the full spectrum of mutations arising in the mutator strains, Stirling and coworkers decided to use whole genome sequencing instead to detect them.

The researchers chose 11 mutator alleles of genes representative of different processes such as homologous recombination, oxidative stress tolerance, splicing, transcription, mitochondrial function, telomere capping, and several aspects of DNA replication. They grew these strains for 200 generations and then did whole-genome sequencing of 4 to 6 independent progeny of each to find all the resulting mutations.

Under these conditions, wild-type yeast accumulated 2-4 mutations per genome. In contrast, the mutator strains ended up with 2- to 10-fold more mutations. And most every type was represented: single-nucleotide variants, structural variants (showing altered chromosome structure), copy-number variants (amplification of certain regions or entire chromosomes), and insertions or deletions.

However, while all of the mutator strains had accumulated mutations, the different types of mutation were in different proportions. For example, a mutant in the Replication Factor C subunit gene, rfc2-1, tended to give rise to transition mutations (changing a pyrimidine to a pyrimidine, or a purine to a purine).  The same was true for the telomere-capping protein mutant, stn1-13. 

But the pol1-ts DNA polymerase mutant instead showed more transversions (changing a purine to a pyrimidine or vice versa). And a deletion of the nonessential RAD52 gene, encoding a recombinase, tended to cause mutations in the transcribed strand of genes, suggesting that transcription-associated recombination was compromised in those cells and this affected DNA repair. 

Positions of the accumulated mutations also differed between strains. The stn1-13 and pol1-ts mutants preferentially accumulated mutations in subtelomeric regions. Some of the alleles gave rise to clusters of mutations, while others did not. And, as has been seen in cancer cells, many of the mutator strains had mutations in regions of the genome that replicate late in DNA replication.

Even though this work generated a huge amount of data (much more than we can discuss here), one conclusion reached by the authors is that even more mutant progeny of mutator strains, arising under a variety of different conditions, need to be analyzed using whole-genome sequencing to give a truly comprehensive picture of the mutational spectrum associated with each allele.

But another conclusion is clear: that different mutator alleles do result in characteristic patterns of mutations. Given that some of these same genes have been found to be mutated in cancer cells, this work may help other scientists predict what mutations a cancer will develop. And that would really give us a bang for our research buck!

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

Categories: Research Spotlight

Tags: mutator phenotype, Saccharomyces cerevisiae, yeast model for human disease

Yeast, Smarter than a Train Wreck

January 30, 2014

Imagine you run a railroad that has a single track. You need for trains to run in both directions to get your cargo where it needs to go.

One way to regulate this might be to have the trains just go whenever and count on collisions as a way to regulate traffic. Talk about a poor business model! Odds are your company would quickly go bankrupt.

Another, more sane possibility is to somehow keep the trains from running into each other. Maybe you schedule them so their paths never cross. Or maybe you have small detours where a train can wait while the other passes. Anything is better than regulation by wreckage!

Turns out that at least in some cases, nature is a better business person than many people previously thought. Instead of trains on a track, nature needs to deal with nearby genes that point towards one another, so-called convergent genes. If both genes are expressed, then the RNA polymerases will barrel towards one another and could collide.

A new study in PLoS Genetics by Wang and coworkers shows just how big a deal this issue is for our favorite yeast Saccharomyces cerevisiae. An analysis of this yeast’s genome showed that not only did 20% of its genes fit the convergent definition but that in many cases, each gene in a pair influenced the expression of the other gene. Their expression was negatively correlated: when one of the pair was turned up, the other went down, and vice versa.

One way these genes might regulate one another is the collision model. When expression of one gene is turned up and a lot of RNA polymerases are barreling down the tracks, they would crash into and derail any polymerases coming from the opposite direction. A prediction of this model is that orientation and location matter.  In other words, the negative regulation would work only in cis, not in trans.  Surprisingly, the authors show that this is clearly not the case.

Focusing on four different gene pairs, Wang and coworkers showed that if the genes in a pair were physically separated from one another, their expression was still negatively correlated.  This was true if they just flipped one of the genes so the two genes were pointed in the same direction, and it was still true if they moved one gene to a different chromosome.  Clearly, collisions were not the only way these genes regulated one another.

Using missense and deletion mutation analysis, the authors showed that neither the proteins from these genes nor the coding sequence itself was required for this regulation.  Instead, the key player was the overlapping 3’ untranslated regions (UTRs) of the transcripts.  The authors hypothesize that the regulation is happening via an anti-sense mechanism using the complementary portions of the 3’ UTRs.

This anti-sense mechanism may be S. cerevisiae’s answer to RNAi, which it lost at some point in its evolutionary history.  Given the importance of RNA-mediated regulation of gene expression in other organisms, perhaps it shouldn’t be surprising that yeast has come up with another way to use RNA.  

Instead of RNAi, it relies on genomic structure and overlapping 3’ UTRs to regulate genes.  This may be a bit more cumbersome than RNAi, but at least yeast came up with a more clever system than polymerase collisions to regulate gene expression.  

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

Categories: Research Spotlight

Tags: RNA polymerase II, Saccharomyces cerevisiae, transcription, UTR

Yeast, the New Fountain of Youth

January 23, 2014

Ponce de Leon searched the New World for the fountain of youth.  Turns out that if he had some of the tools at our disposal, he wouldn’t have even had to leave Europe.  He just needed to go to the local bakery or brewery and look inside the yeast he found there.  Of course, then he wouldn’t have found Florida…

Using in silico genome-scale metabolic models (GSMMs) in yeast, Yizhak and coworkers identified GRE3 and ADH2 as two genes that significantly increased the lifespan of yeast when knocked out.  Even more importantly, their method also allowed them to identify the mechanism behind this increased lifespan—the mild stress of increased reactive oxygen species (ROS).  This last finding may help scientists identify drug targets that they can target to increase the lifespan of people too.  If only Ponce de Leon had lots of -omics data and a powerful computer or two!

After constructing an in silico starting state, Yizhak and coworkers entered two sets of data from previous work that had been done on aging in yeast.  They next used gene expression profiling to identify which metabolic reactions were different and which were the same in young and old yeast.  They then systematically tested the effect of knocking out these reactions one at a time in their computer model to identify those that could potentially transform yeast from old to young with minimal side effects. 

Their first finding was that many of their best hits, like HXK2, TGL3, and FCY2, had already been identified as important in prolonging a yeast cell’s life.  They decided to look at seven genes that had not been previously identified as being involved in aging. 

When two of these seven, GRE3 and ADH2, were knocked out, these yeast strains lived significantly longer with minimal side effects.  For example, the strain lacking GRE3 lived ~100% longer than the wild type strain.

Figuring out why these yeast probably lived longer was made simpler because they used metabolic models to identify the genes.  The hormesis model of aging suggests that mild stress, like that found in caloric restriction, can lead to increased life span.  With this model in mind, the authors focused in on the possibility that knocking out GRE3 and/or ADH2 would lead to increased stress through the production of increased levels of ROS.  When they looked, they found that the two knockout strains did indeed have higher levels of two common forms of ROS, hydrogen peroxide and superoxide. 

Of course none of us is particularly interested in extending the life of a yeast!  But these results could suggest new drug targets to go after that might mimic the effects of caloric restriction without us having to starve ourselves.  And these same methods can be used on human cells to find key pathways to target in people.  In fact, the authors have started to use their computer models to investigate aging in human muscle cells and found that like in yeast, many of the genes they have identified are consistent with previous work on human aging. 

Now we probably shouldn’t get too far ahead of ourselves here.  This is a promising first step but it really isn’t much more than Ponce de Leon boarding his ship to begin his trip to the New World.  We still have a long voyage ahead of us before we find the fabled fountain of youth.  

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

Categories: Research Spotlight

Tags: aging, metabolic model, Saccharomyces cerevisiae

Cutting Down on the ChIPs

January 16, 2014

We all know that potato chips are delicious.  But we also know that eating too many of them isn’t very good for our arteries or our waistlines. And apparently these aren’t the only chips that can be too much of a good thing.

Chromatin immunoprecipitation (ChIP) is an incredibly valuable technique that lets us see where a particular protein binds in a genome. It can show us the target genes of a particular transcription factor, the distribution of RNA polymerases as they transcribe genes, the places where silencing proteins bind to turn off expression of particular regions, and lots more.

But just like potato chips, more ChIP results aren’t always better. Teytelman and coworkers, publishing in Proceedings of the National Academy of Sciences, and Park and coworkers, publishing in PLoS ONE, have discovered that highly transcribed regions of the genome consistently give false positive ChIP results. In other words, very active regions of the genome look like everything is binding there even when it almost certainly is not.  Teytelman and colleagues call these regions “hyper-ChIPable”. 

Far from being a reason to despair, though, the discovery of this artifact explains some puzzling previous results and inspires the creation of new, more reliable ChIP methods. This is exactly what Kasinathan and coworkers have done, in a recently published paper in Nature Methods. 

The idea behind the ChIP technique is that if you want to know all of the places across the genome where your protein of interest binds, you can lyse cells, shear the DNA into relatively short fragments, and immunoprecipitate your protein from the mixture. Usually the protein and DNA are cross-linked before immunoprecipitation, to strengthen their bond during the rest of the procedure.

After immunoprecipitation, the DNA fragments associated with the protein can be identified using a variety of methods. Finally, mapping the sequences of the fragments to the genomic sequence shows us all the sites that the protein occupies.

Teytelman and colleagues used ChIP-seq to ask whether the silencing complex (Sir2p, Sir3p, and Sir4p) ever binds to non-silenced regions of the genome. They thought they might see some binding, but they were astounded to find significant binding of the complex at 238 distinct euchromatic (non-silenced) loci. This didn’t really make sense, since the yeast Sir proteins are extremely well-studied and there were no biological hints that they have such a large presence at non-silenced genes. 

As a control, they looked at previously published ChIP data on the locations of two unrelated proteins, Ste12p and Cse4p, and found that their binding was enriched at the same 238 loci. Finally, they did a ChIP study using green fluorescent protein (GFP) alone. Sure enough, the ChIP data showed that this jellyfish protein apparently bound strongly to chromatin at those 238 sites! The common denominator shared by these loci: they were all very highly expressed.

Meanwhile, Park and coworkers were embarking on a similar journey. They found using ChIP-seq that several unrelated transcription factors seemed to have common targets, which didn’t make biological sense. Control experiments looking at binding sites of Mnn10p (a cytoplasmic protein not expected to have any contact with DNA), or even using nonspecific antibodies that didn’t recognize any yeast proteins, still gave the same set of ChIP targets. Again, these targets were all highly expressed genes.

Each group found several factors contributing to this artifact, although all the reasons why highly expressed regions yield false positives may not yet be uncovered.  But whatever the reasons, this finding helps explain some previously perplexing results – such as binding of Mediator complex all over the genome, or the paradoxical binding of silencing regulator Sir3p to the GAL1–GAL10 regulatory region under conditions where transcription is activated, not silenced.

In response to these issues, many researchers are actively trying to improve the ChIP technique. Kasinathan and colleagues have devised a method that they call ORGANIC (Occupied Regions of Genomes from Affinity-purified Naturally Isolated Chromatin) that eliminates crosslinking and substitutes micrococcal nuclease treatment for sonication (to shorten the DNA fragments).  In a pilot project, they mapped binding sites for the transcription factors Reb1p and Abf1p. The method looks to be both accurate and sensitive. Most binding locations that they found contained the binding motif sequence for that transcription factor, and also correlated with in vivo occupancy as determined by Dnase I footprinting – both of which support their biological relevance. Importantly, the technique shows no bias towards highly expressed regions.

The lesson for researchers is that ChIP results for highly expressed genes, particularly those done using older protocols, need to be viewed cautiously.  And of course this artifact could be an issue for organisms other than yeast. ChIP experiments are used across species, and have been valuable in elucidating the targets of disease-related proteins like the tumor suppressor p53.

The fact that yeast genetics and molecular biology have so well established the roles of certain chromatin-associated proteins was a key part of this puzzle, helping to point out the artifactual nature of some of the ChIP results. Just as a new recipe for potato chips could allow us to eat more of them while staying healthy, yeast research has led the way to a new recipe for more accurate ChIP studies.

Aside from the molecular biology behind this work, it is quite interesting from a sociological point of view as well. What is it like to make a discovery that calls into question a routinely-used technique and a lot of published results? Lenny Teytelman’s blog post on this topic provides a fascinating glimpse into this situation.

Categories: Research Spotlight

Tags: chromatin immunoprecipitation, Saccharomyces cerevisiae

Next