New & Noteworthy
October 7, 2015
In the old days, before the internet, planes or even mass publishing, it was hard to spark a quick, worldwide movement. You simply couldn’t reach out to likeminded people who lived far away.
Nowadays things are very different. With the advent of social media, it is now trivially easy to spread the word. Using Twitter and Facebook, organizers can easily and effectively organize people who live on the other side of the world.
In terms of transcription activation, our friend Saccharomyces cerevisiae seems to be stuck in the old world. Its transcription factors can only turn up nearby genes. This is different from most other eukaryotic beasts, where activation at a distance is routine.
Except maybe yeast isn’t as backward as we think. It may be that yeast has the potential to activate transcription at a distance, but keeps that potential locked away.
In other words, this ability is there but is just prevented from being used. Yeast is keeping social media out of the hands of its genes.
Getting around social media/internet controls is not easy. Pressure might need to be applied at multiple points before people power is finally released through social media. And even when it does happen, it can sometimes be hard to figure out exactly why certain events tipped the balance.
Turns out that both of these are also true for long range transcription activation in yeast. Mutating a single gene was not enough—it took mutations in multiple genes to see any significant effect.
As you can imagine, it would be very tricky to hit all of the right mutations for a polygenic trait in one fell swoop. This is why Reavey and coworkers started their mutant hunt with a strain that could already weakly activate transcription from a distance, a strain in which the SIN4 gene was deleted. Now they just needed additional mutations to make the effect stronger.
In their screen they used a reporter in which the GAL4 upstream activating sequence (UAS) was placed 799 base pairs upstream of the HIS3 reporter. This reporter gives very low levels of activity in a wild type strain. They included a second reporter, the URA3 gene under control of the same upstream sequences, because Reavey and Winston had discovered in previous work using a single reporter that cis acting mutations and chromosomal rearrangements were a frequent source of false positive results.
The researchers put the reporter strain through multiple rounds of mutation with UV light and selection with increasing levels of 3AT, a competitive inhibitor of His3. After each round of selection, they measured mRNA levels for HIS3 and URA3 and chose strains that not only had higher 3AT resistance but also showed more transcription of the reporter genes. In the end they found three strains that survived in the presence of galactose (to turn on the activator) and 10 mM 3AT.
As expected, each strain had multiple mutations. One strain had acquired mutations in the GRR1 and MOT3 genes. To confirm that these were the most important mutations, Reavey and colleagues engineered a fresh strain with just the original sin4 null mutation and the selected grr1-1 and mot3-1 mutations. The fresh strain completely recapitulated the selected strain, showing that these three mutations could unlock yeast’s potential for long-range transcriptional activation.
It makes sense that a grr1 mutation could affect transcriptional activation. Grr1 is a ubiquitin ligase that destabilizes Med3 (also known as Pdg1), a key component of the Mediator complex involved in transcription activation. The researchers provided evidence that this is how the grr1-1 mutation affects the process, by showing that mutating MED3 mimicked the effects of mutating GRR1.
It’s also not too hard to imagine how a mutation in Mot3, a sequence-specific transcriptional activator, could affect transcriptional activation, presumably by changing the expression of a gene under its control.
The results were not so clear-cut for two other strains that were selected. They arose from the same lineage, and each had acquired the ptr3-1 mutation on top of the original sin4 null mutation. One strain went on to further pick up the mit1-1 mutation, while the other got an msn2-1 mutation.
Again it isn’t too surprising that mutations in genes that encode sequence-specific transcriptional activators like Mit1 and Msn2 arose in these strains. But the selection of the ptr3 mutation in these lineages is something of a mystery.
It is hard to imagine how the usual job of Ptr3 in nutrient sensing and transport would be involved in keeping long range transcription activation down. Perhaps the researchers have uncovered a novel function for this gene.
And re-creating these two strains only partially restored the levels of transcription activation at a distance that were seen in the original strains. A little genetic detective work showed that a big reason for this was that both of the selected strains had acquired an extra copy of the chromosome that had the HIS3 reporter, chromosome III.
Reavey and colleagues deleted each of their identified genes to see which ones caused their effect through a loss of function mutation. Deleting GRR1, PTR3, and MSN2 all had the same effect as the original isolated mutations.
The same was not true for MOT3 and MIT1. Deleting either gene actually weakens long range transcription activation, suggesting that these two had their effect through gain of function mutations.
Finally, the researchers showed that the increase in long-distance transcriptional activation was not simply due to a general increase in transcription activation in the selected strains, by showing that their mutants did not have increased activity of a reporter with the GAL4 UAS placed 280 base pairs upstream of HIS3. In fact, if anything, the strains showed decreased activation with this reporter.
So this experimental strategy allowed Reavey and coworkers to identify some of the key genes involved in keeping transcription activation at a distance under control in yeast. In particular, they found compelling evidence that the Mediator complex is an important player. But there is still plenty of work to do. For example, which of the genes regulated by Mit1, Msn2, and Mot3 are important in long range activation? And what on Earth is Ptr3 doing in all of this?
The success of this approach also confirms that doing repeated rounds of selection in yeast is a viable way to select multiple mutants and study polygenic traits. This strategy may prove a boon for studying the many human diseases that are the result of polygenic traits.
Not only can we use yeast to uncover its activation potential, but we can also now potentially use it to uncover new treatments for human disease. Unleashing another awesome yeast power…
by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics
September 30, 2015
Stinging wasps get our attention, and with good reason—getting stung hurts a lot! If you see wasps going into a nest within the walls of your house, you’ll likely try to block their access.
But this could backfire: instead of being able to peacefully go into their nest, a swarm of angry wasps could be buzzing around looking for trouble. It might get so bad that you’ll need to call in an exterminator to take care of the problem.
Mitochondrial proteins might seem a lot less scary than wasps, but it turns out that they can also cause trouble if they can’t get into mitochondria. In two new letters to Nature, Wrobel and colleagues and Wang and Chen used complementary approaches to ask what happens when dysfunctional mitochondria aren’t able to import all of the proteins that are waiting to get in.
What they both found was that these piled up proteins cause real problems for the cells. In desperation, the cells slow down protein synthesis to reduce the excess, and also turn to their own exterminator, the proteasome, to keep these proteins under control.
This is a paradigm shift in thinking about how poorly functioning mitochondria cause disease. In the past, almost everyone focused on how damaged mitochondria couldn’t make enough energy for a cell. Now it looks like there are other ways for a nonworking mitochondria to do a cell in. And new targets for scientists to go after in treating mitochondrial disease.
As we all know, mitochondria are the powerhouses of the cell, where energy is generated, and they’re also the site of many other essential biochemical reactions. They’re composed of about 1,000 proteins, and nearly all of those are synthesized in the cytoplasm and then imported into mitochondria by an intricate system of transporters.
In order to find out what happens when these 1,000 proteins don’t get into mitochondria as efficiently as they should, both groups created strains whose mitochondrial import was impaired.
Wrobel and colleagues used the temperature-sensitive mia40-4int mutation, affecting an essential component of the mitochondrial import system. Wang and Chen started with the aac2-A128P mutation in PET9 (which is also known as AAC2), an ADP/ATP carrier of the mitochondrial inner membrane. Overexpression of the aac2-A128P allele causes mitochondrial dysfunction and eventual cell death.
Wrobel and colleagues decided to get a comprehensive look at what happens in the mia40-4int mutant by assaying its transcriptome and proteome, using RNA-seq and stable isotope labelling by amino acids in cell culture (SILAC), respectively. Surprisingly, one of the biggest differences from wild type that they saw in the import-defective mutant was a decrease in cytoplasmic translation. Whether they looked at the mRNAs encoding ribosomal proteins, the proteins themselves, or the polysome content and translational activity of the cells, everything pointed to down-regulation of translation. And at the same time, the proteasome—the molecular machine that breaks down unwanted proteins—was activated.
To verify that what they were seeing wasn’t peculiar to the mia40-4int mutant, Wrobel and colleagues slowed down mitochondrial import in several other ways: using different mia40 mutant alleles, other import mutants, or treatment with a chemical that destroys the mitochondrial membrane potential required for import. Under these different conditions causing the accumulation of mitochondrial precursor proteins in the cytoplasm, they still saw decreased cytoplasmic translation and increased proteasome activity.
Wang and Chen took a different approach, looking to see whether over-expression of any other genes could compensate for the lethality of overexpressing the aac2-A128P allele. The researchers transformed the mutant with a library of yeast genes on a multicopy plasmid, and found 40 genes whose expression could keep it alive.
The suppressor genes found by Wang and Chen were all involved in some aspect of synthesis or degradation of cytoplasmic proteins, just like the genes found by Wrobel and colleagues whose expression was altered in the mia40 mutant. And Wang and Chen also verified that these suppressors weren’t specific to the aac2-A128P mutation: they suppressed a variety of other mutations that decreased import.
Both groups observed precursors of mitochondrial proteins accumulating in the cytosol of the mutant strains they studied. Wang and Chen saw a couple other very interesting proteins increase in abundance: Gis2 and Nog2. These proteins are involved in regulating ribosome function, and the researchers speculate that their stabilization during this stress response contributes to the translational down-regulation. Intriguingly, their human orthologs are implicated in neuromuscular degenerative disease.
So, using orthogonal approaches, the two groups converged on the same model: a newly discovered cellular pathway that regulates cytosolic translation and protein degradation in order to deal with the stress of inefficient mitochondrial import. Wrobel and colleagues have named it UPRam, for Unfolded Protein Response activated by mistargeting of proteins, while Wang and Chen call it mPOS, mitochondrial Precursor Over-accumulation Stress.
Before this work, it was unknown whether cytosolic pathways were even affected by mitochondrial dysfunction. Now we know that the cell has a specific response when mitochondrial precursor proteins begin swarming in the cytosol, unable to get into their home: it slows down the production of those proteins and calls in the proteasome exterminator to take care of them.
We usually think of mitochondrial disease symptoms as being caused by the reduced energy generation of sick mitochondria, or by the lack of other key events that happen in mitochondria—for example, the synthesis of the iron-sulfur clusters that some vitally important enzymes need. Now, these findings raise the possibility that proteostatic stress on the cell caused by the accumulation of mitochondrial precursors could also lead to impaired cell function and disease.
Perhaps drugs that inhibited cytoplasmic translation, or activated the proteasome exterminator, would be helpful in reducing the buzzing swarm of mitochondrial precursor proteins. Wouldn’t it be wonderful if this knowledge suggested new avenues of treatment to take some of the sting out of human mitochondrial disease?
by Maria Costanzo, Ph.D., Senior Biocuration Scientist, SGD
September 23, 2015
In the classic Dr. Seuss tale Horton Hears a Who, the elephant Horton thinks he hears voices coming from a speck of dust. He gets into all sorts of trouble over this until all the Whos in Whoville prove they are alive when they all shout at once. Now Horton’s jungle compatriots believe him and Horton can hang out with his new friends.
Horton’s companions never get to hear an individual Who. They are not blessed with Horton’s big elephant ears and so have to just hear all the Whos shouting at once.
Up until recently, we have been in the same situation as the kangaroo and everyone else in the jungle when it comes to transcription in a cell. We can use all sorts of tools to get at what goes on when RNA polymerase II (pol II) gets ready and then starts to transcribe a gene, but we can only get an aggregate picture of lots of cells where it is happening. We can’t hear the Mayor of Whoville amidst all of the other Who voices.
In a new study out in Nature, Fazal and coworkers use the equivalent of elephant ears, optical tweezers, to study the initiation of transcription by purified pol II machinery from Saccharomyces cerevisiae on single molecules. And what they find is that at least for one part of the process, our having looked at things in the aggregate may have fooled us about how the process worked. It was important that we be able to pick out individual voices from the cacophony of the crowd.
Not surprisingly, transcribing a gene is tricky work. It is often split into three steps: initiation, elongation, and termination. And each of these can be subdivided further.
Fazal and coworkers focused on transcription initiation. Previous work had suggested that the process goes something like this:
Top image via Wikimedia Commons
Basically, an alphabet soup of general transcription factors and pol II sit down on a promoter. This complex then pries open the DNA and looks for a signal in the DNA to start transcribing. The polymerase then transcribes short transcripts until it shifts into high gear when it escapes the promoter and enters elongation phase.
This theory comes from the study of transcription in bulk. In other words, it derives from looking at many cells all at once or many promoter fragments in a test tube.
Fazal and coworkers set out to look at how well this all holds up when looking at single genes, one at a time. To do this they used a powerful technique called optical tweezers.
Optical tweezers can “see” what is going on with moving enzymes by measuring the change in force that happens when they move. For this study, the preinitiation complex bound to a longish (2.7 kb) piece of DNA was attached to one bead via pol II, the moving enzyme. The other end of the DNA was attached to a second bead. Each bead is then immobilized using lasers (how cool is that!) and the DNA is stretched between the two beads. Watch this video if you want more details on the technique.
Depending on where you attach the DNA to the bead, you can either track polymerase movement or changes in DNA by precisely measuring changes in the forces keeping the beads in place. Using this technique the researchers found that the bulk studies had done pretty well for most every step. Except for the initial transcribing complex.
The earlier studies had suggested that an open complex of around 15 nucleotides was maintained until elongation began. This study showed that in addition to the 15 base pairs, an additional 32 to 140 base pairs (mean of about 70 base pairs) was also opened before productive elongation could begin. And that this whole region was transcribed.
This result paints a very different picture of transcription initiation. Rather than maintaining a constant amount of open DNA, it looks like the DNA opens more and more until the open DNA collapses back down to the 12-14 base pair transcription bubble seen during elongation.
It turns out that this is consistent with some previous work done in both yeast and fruit flies. Using KMnO4, a probe for single stranded DNA, scientists had seen extended regions of open DNA around transcription start sites but had interpreted it as a collection of smaller, opened DNA. In other words, they thought they were seeing different polymerases at different positions along the DNA.
These new results suggest that they may have actually been seeing initial transcribing complexes poised to start processive elongation. Seeing just one complex at a time changed how we interpreted these results.
Fazal and coworkers were also able to see what happened to some of the 98% of preinitiation complexes that failed to get started. Around 20% of them did end up with an extensive region of open DNA of around 94 +/- 36 base pairs but these complexes were independent of transcription, as they didn’t require NTPs.
But since this opening did require dATP, they propose that it was due to the general transcription factor TFIIH, a helicase. It looks like in these failed complexes, TFIIH is opening the DNA without the polymerase being present.
A clearer picture of what might be going on at the promoter of genes starts to emerge from these studies. Once around 15 base pairs of DNA are pried open to form the appropriately named open complex, TFIIH unwinds an additional 70 or so base pairs. The polymerase comes along, transcribing this entire region. The whole 85 or so base pairs stays open during this process.
Eventually the polymerase breaks free and the opened DNA collapses back down to around 12-14 base pairs. Now the polymerase can merrily elongate to its heart’s content. Until of course something happens and it stops…but that is another story.
September 9, 2015
Looking at the phenotypes—the observable characteristics—of creatures that have mutations in various genes can give important clues to scientists trying to figure out what those genes do. And the ability to systematically mutate thousands of genes in our favorite organism, Saccharomyces cerevisiae, has made it an awesomely powerful genetic system.
But this awesome system has something of an Achilles heel. If you delete an essential gene, defined as a gene necessary for life under standard growth conditions, you end up with an experimental dead end: a dead cell. How then can you study mutant phenotypes for essential genes, which are nearly 20% of the genome?
In a new paper in G3, Kofoed and colleagues address this problem by creating conditional mutations in 600 essential genes. The mutant genes function normally at standard temperatures (25-30° C) but can be inactivated by raising the temperature. This mutant collection covers about half of S. cerevisiae essential genes and gets us much closer to being able to do mutant screens that are truly genome-wide, helping us to discover unexpected connections between genes and pathways or processes.
One popular approach to studying essential genes has been to put them under control of promoters that can be turned down when you add a particular chemical or carbon source. But it can be hard to tease apart the phenotype of down-regulating transcription of your gene of interest from the effects of the other changes that you need to make in order to regulate expression of the construct.
A solution that avoids some of these issues is to use temperature-sensitive (ts) alleles of essential genes. These mutations make the resulting proteins unstable at high temperatures. If you shift a ts strain to high temperature, the mutant protein will stop functioning; by growing the cells at an intermediate temperature, you can often produce a partially active protein that allows slow growth of the strain.
Temperature-sensitive alleles are not nearly as straightforward to create as are deletion mutations. Although newer technologies are helping, it’s still a lot of work. But Kofoed and colleagues took on this challenge. Building on their previous collection of 250 genes, they targeted genes without existing ts alleles to create a mutant collection including a total of 600 genes, about half of yeast’s essential genes.
The researchers used error-prone PCR to introduce mutations into DNA fragments encoding the genes, then transformed the mutagenized fragments into heterozygous diploid cells containing one wild-type copy of the gene and one deletion allele. The flanking sequences of the mutagenized genes targeted them to integrate in place of the deletion allele. Kofoed and colleagues could then sporulate the diploids and screen the haploid progeny for temperature sensitivity.
The genetic background of these strains is S288C, the widely used strain from which the reference genome sequence is derived. Each strain contains a “barcode”, a sequence that can be used to uniquely identify it. In addition to creating new alleles, the authors also used these methods to transfer some previously isolated ts alleles into this same background.
Kofoed and colleagues validated and stored multiple ts alleles for each gene on their list. Although they chose one allele per gene for inclusion in the collection, the other alleles are available on request and could be very informative for researchers studying those particular genes.
As a test to see whether their collection of 600 genes was biased in any way, the researchers did Gene Ontology (GO) enrichment analysis. This compares the GO terms, representing molecular function, biological process, and subcellular localization, that are associated with the genes in a set. If the genes share related GO terms, that’s an indication that they may be involved in a common process or share the same location.
Most of the time, when scientists do GO enrichment on sets of genes they’ve come up with in an experiment, they’re hoping the genes have significantly shared terms. But in this case, the researchers were happy to find no shared terms, meaning that their collection represented a wide variety of places and roles in the cell.
The mutant strains may be studied individually in classical genetics experiments or the whole set can be tested using robotic manipulation of all the strains at once. Alternatively, the strains can be pooled and grown together in a competitive situation, like a microscopic Survivor show. When the strains are forced to compete for survival, some will prosper and others will die out. Their unique barcode identifiers allow researchers to figure out which strains were the survivors and which got voted out of the chemostat.
So this collection represents an important step in our ability to survey the phenotypes of virtually all genes in the genome. In addition to creating this resource, Kofoed and colleagues provide a summary (in Table S6) of all currently available mutant collections for essential genes. The yeast research community is now poised to turn up the heat on genome-wide mutant screens and make new discoveries about the roles of these essentially important genes.
by Maria Costanzo, Ph.D., Senior Biocuration Scientist, SGD