New & Noteworthy

Don’t Ignore the Scullery Maid, Hobbits, or Pol III

February 13, 2018

In the Disney story, Cinderella is ignored as the scullery maid. Most of the attention falls on her step-sisters.


If the prince ignored Cinderella, he would not have a happily ever after. Nor might we if we ignore Pol III. (strwalker)

This is, of course, a mistake. The best person by far is Cinderella and she should not be ignored because of her position in life.

The RNA polymerases in a eukaryotic cell are similar except that one is the star and two are the scullery maids. RNA polymerase II (Pol II), the reader of protein encoding genes, gets most of the attention. Pol I and Pol III, the enzymes that make RNAs that aren’t usually translated into proteins, are mostly relegated to supporting roles.

In a study in Nature, Filer and coworkers show that ignoring Pol III can be as big a mistake as ignoring Cinderella. If you want to live a longer life that is.

Just as there were hints of Cinderella’s specialness, so too there were hints that perhaps Pol III plays an important role in longevity. That hint goes by the name of TORC1.

Previous work had shown that inhibiting TORC1 can increase the lifespan of beasts ranging from yeast to mice. Given that TORC1 is a strong activator of Pol III, Filer and coworkers set out to determine if inhibiting Pol III could do the same thing.

As any good researcher should, they started out by looking in our favorite model organism, the yeast Saccharomyces cerevisiae.

Pol III is pretty complicated—it is made up of 17 different subunits encoded by 17 different essential genes. These authors went after the largest subunit, C160, encoded by the RPO31 (RPC160) gene.

To control its level of expression, they fused an auxin-inducible degron to the protein. This fusion protein is degraded in the presence of indole-3-acetic acid (IAA) and E3 ubiquitin ligase from the rice Oryza sativa.

This means that there should be less Pol III in a cell when IAA is around. Western blot analysis confirmed that this was indeed the case.

The authors found that yeast with less Pol III had a longer chronological lifespan. In other words, each yeast cell survived longer. Having less Pol III did not significantly increase replicative lifespan though; they did not bud more often over their lifetime.

These authors did not get as significant a result when they targeted the largest subunit of Pol II, RPO21 (RPB220). The strain appeared to do better in the presence of IAA compared to the control strain, suggesting that targeting Pol II had a small effect on longevity; but it was nothing like was seen with Pol III.

They next turned to a couple of model organisms with more cells—nematodes, also known as C. elegans, and fruit flies, or Drosophila.

When Filer and coworkers used RNA interference (RNAi) to knock down the amount of the largest subunit of Pol III, rpc-1, in C. elegans, the worm lived longer. And when they generated a strain of Drosophila which had only one working copy of dC53 (CG5147), a gene that encodes a Pol III subunit, these fruit flies lived longer too. Lowering the amount of Pol III is a good thing if you want to live longer!


Sauron would still be alive if he didn’t ignore those hobbits. We might live longer too if we stop ignoring Pol III! (Colony of Gamers)

Previous work had shown that the gut often controls longevity in C. elegans. These authors showed that the same holds true for Pol III reduction. Targeting just the Pol III in the gut was sufficient to give this little worm a longer life.

The same turned out to be true in the fruit fly. Using RNAi against two different Pol III subunits in the Drosophila gut was enough to have extend these flies’ lives. Targeting Pol III in the fat body did not have the same effect, and targeting Pol III in neuronal cells extended life only a little.

So reducing Pol III in the gut is good enough for flies and worms to live longer. In fact, reducing Pol III activity had as much effect on longevity as does inhibiting TORC1. Scientists ignore this hidden gem at their own risk!

Since Pol III is ignored, but dangerous for lifespan, perhaps a better analogy than Cinderella might be Sauron and the hobbits of the Shire from the Lord of the Rings Trilogy. Sauron ignores the hobbits who hold the key (or the ring) to Sauron’s shortened lifespan.

If Sauron had paid attention, he’d still be lording it over Middle Earth at the end of the third book. Just as we might live longer if we start paying more attention to that hidden danger—excessive Pol III activity, possibly in the gut.

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

Categories: Research Spotlight

Tags: Longevity , TORC1 , RNA polymerase III

Strung Out Chromosomes and Coyotes

January 29, 2018

As Wile E. Coyote knows, when setting a bear trap, you need to make sure the tension is just right. As you can see in this video, if there isn’t enough, then the Road Runner gets away:

And of course, Wile E. ends up getting caught in the trap!

Like our super-genius coyote friend, Saccharomyces cerevisiae cells need to make sure there is the right amount of tension sometimes, too. One of those times, for example, is tension on their chromosomes at the metaphase stage of cell division.

If the chromosomes aren’t attached to the spindle in just the right way, the cell will not let mitosis move from metaphase to anaphase. The cell won’t put down the birdseed until it “knows” the trap is set properly.

While Wile E. Coyote probes the tension by eyeballing it, yeast does it in a much more thorough way. Two key pieces of the sensing machinery in yeast are the tension sensing motif (TSM) from histone H3, 42KPGT, and the Shugoshin 1 protein, Sgo1p.

The current model is that the TSM of histone H3 acts as an anchoring point or a landing pad for Sgo1p, the protein responsible for sensing that there is the right amount of tension between chromosomes and the spindle/kinetochore. When the TSM is mutated with, for example, a G44S allele, the idea is that Sgo1p can no longer bind causing chromosomes to sometimes segregate incorrectly.

Previous work had shown that mutating Gcn5p can partially rescue the effects of the G44S allele. In a new study in GENETICS, Buehl and coworkers provide evidence that the reason it can has to do with the tail of histone H3. Basically, this tail can act as a secondary anchor if the TSM, the first anchor point, is mutated. But the tail can only work if it is not acetylated by Gcn5p.


Bad things can happen when plans go astray for Wile E. Coyote and for chromosome segregation in yeast. (Wikimedia)

It is as if Wile E. Coyote’s bear trap is broken so he has to fashion a new trigger to get it to snap. When the TSM of H3 is gone, yeast can fashion a new sensor by eliminating the acetylation on the tail of the H3 histone.

Buehl and coworkers focused on the tail because it is a known target of Gcn5p acetylation. In particular, there are four lysine residues in the tail—K9, K14, K18, and K23—that are acetylated by Gcn5p to varying degrees. The most acetylated is K14 followed by K23 and K18. K9 is hardly acetylated at all.

When they deleted the histone tail, they found that the effects of the TSM mutation, G44S, were more severe. But when they mutated the four lysines all at once, the effects of G44S weren’t as bad. This is consistent with the unacetylated tail serving as a landing pad for Sgo1p in the absence of the TSM. No tail, no binding, leading to chromosome mis-segregation.

As might be expected if this was linked to Gcn5p, each individual lysine mutant had effects that corresponded to their level of acetylation by Gcn5p. So K14A did the best job of partially rescuing the G44S allele, K18A and K23A did an okay job and K9A had very little effect.

Buehl and coworkers provide three additional sets of experiments to show that G44S causes chromosome mis-segregation and that K14A can partially rescue it. There is only time to discuss one of these, but you can read the paper for the other two.

This strategy uses the idea that diploids cannot mate because of a repression function on each copy of chromosome III (chrIII). If a diploid loses one of its chrIII’s, then it will be able to mate.

As expected, the authors found that both wild type and the K14A mutant had very few diploids that could mate. But it was a different story for the G44S mutant—it had many more diploids able to mate, presumably due to the loss of one chrIII. The double mutant K14A/G44S had more diploids mate than wild type but less that G44S on its own.

This is what we would expect if K14A can only affect tension sensing in a TSM mutant. It would do little on its own but could partially rescue the TSM mutant G44S.


If Wile E. Coyote had a backup plan like yeast does for chromosome segregation, he might get to dine on the Road Runner. (Rosenfeld Media)

The authors next used chromatin immunoprecipitation (ChIP) assays to directly show that the K14A mutation in histone H3 allowed Sgo1p to bind when the TSM was mutated. There was one hitch though—Sgo1p bound most everywhere and not just at the pericentromeres where it usually does.

In a final set of experiments, the authors used GST-pulldowns to show a direct interaction between the H3 tail and Sgo1p. To their surprise, they found that acetylating the tail did not affect Sgo1p binding. They suggest that perhaps the acetylated tail interacts with something that blocks Sgo1p from binding.

Yeast has a backup just in case its main way of sensing tension between the spindle/kinetochore and chromosome, the TSM of histone H3, breaks down. In that case, it can use the tail of histone H3, assuming it is not acetylated.

If only Wile E. Coyote included backups like this in his crazy plans! Then he could finally dine on that tasty Road Runner he is always chasing.

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

Categories: Research Spotlight

Tags: acetylation , Histone H3 , chromosome segregation , metaphase

Exploding Yeast to Protect the Wild

January 22, 2018

In the first Kingsman movie, Samuel L. Jackson plants little bombs into the necks of people who will follow him to his “ark”. If they disobey him, Jackson can simply blow up their heads.

In an amazing scene from the movie, our heroes get a hold of the “detonator” and blow up the heads of every one of Jackson’s followers. As you can see below, the director uses cool colors to make it more fabulous than gruesome:

In a new study out in Nature Communications, Maselko and coworkers explode misbehaving yeast instead of people’s heads. (Click here for an incredibly cool video of it happening…so cool!)

These researchers have created a strain of yeast that can only form diploids within the same strain. If instead it mates and forms a diploid with a yeast outside of its strain, the resulting diploids explode.

These authors aren’t being strainist—they are creating a proof of principle organism that could potentially solve the problem of gene transfer from genetically modified organisms (GMOs) to wild organisms.

Assuming a similar idea works in plants, then the idea is that a GMO like an herbicide-resistant GM plant could not pass its herbicide-resistance to its wild cousin. The resulting plants would explode (or perhaps die in a bit less spectacular fashion). The herbicide-resistant gene would not escape into the wild population because none of the resulting plants could survive.

And that isn’t this approach’s only possible uses. It might be used to kill off disease carrying mosquitoes or controlling invasive species. Watch out sea lamprey!

The basic idea is that the engineered species makes a “poison” that it is immune to. If it mates with a wild species, it passes that poison on to the resulting “children.” These progeny are not immune and so die.

The poison in this case is a CRISPR/Cas9-inspired transcription activator. This activator turns a gene up so much that it kills the yeast. But the catch is that the activator’s binding site is only found in the wild type genome and not the engineered genome.

So the engineered strain carries around this deadly poison that does not affect it. But when this yeast mates with a strain that has the binding site, the resulting diploid yeast has the activator binding site, meaning that the silent killer can now bind and kill the diploid.

For their proof of principle experiments, Maselko and coworkers used a well-tested and available CRISPR/Cas9 transcriptional activator that involves dCas9-VP64 and a single guide RNA aptamer binding MS2-VP64. Previous work has shown that this is a very powerful activator.

Next they turned to the Saccharomyces Genome Database (SGD) to find genes that have been shown to be lethal when overexpressed. They created at least one guide RNA for each of the 8 genes and found that one of the guide RNAs that targeted the ACT1 promoter was lethal. The yeast made actin until it exploded (think the man in Monty Python’s Meaning of Life who eats so much he explodes).

Now they were ready to create their engineered strain. They used CRISPR/Cas9 to engineer a single nucleotide change such that their activator could no longer bind near the ACT1 gene. This strain survived when the activator was introduced.

But when they mated this strain with a wild type strain, they got almost no survivors. The resulting diploids grew fat on overproduced actin and exploded!

Note that there were a few survivors…as Jeff Goldblum famously said in Jurassic Park, “Life will find a way.” The survivors appeared at a frequency of 4.83 × 10−3 with the most common reason found being a mutation in the activator binding site.

This survival rate is one area of obvious concern. Given the genetic variability of wild populations, there may be a few individuals that happen to have a mutation that is resistant to the activator. These immune individuals would then go on to found a new population.


Like the engineered yeast in this strain, Hawat can survive only because he has both the poison and the antidote. (Dunepedia)

The authors suggest that targeting a species that has gone through a recent bottleneck to decrease genetic variability might help. Another possibility might be to target multiple genes making the likelihood of an individual having and/or developing multiple mutations exceedingly unlikely. (This is the strategy used to target HIV, a virus with a high mutation rate.)

These authors have created an elegant system reminiscent of the one created by Baron Harkonnen in the novel Dune. In the book, a character named Thufir Hawat is given a slow acting poison by the Baron. Hawat can be controlled because he needs the antidote each day to survive. Once the antidote is removed, he dies.

Here the researchers have given the poison and engineered the antidote into their new strain. When this strain mates with a strain lacking the antidote, like a wild strain, most of the progeny die. Like Hawat, the yeast die for the good of mankind. And hopefully those pesky weeds can be prevented from picking up transgenes too.

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

Categories: Research Spotlight

Tags: synthetic biology , horizontal gene transfer , CRISPR

Moonlighting Proteins Do Double Duty

January 11, 2018

While most people only know her as the friendly waitress at the local diner, it is by night that Louise gets to live her passion. At night, she is a mime.


Like a waitress who moonlights as a mime, some proteins have more than one job too. (Max Pixel)

If she loses her voice, Louise will struggle as a waitress but do fine as a mime. It is a different story when she gets walloped with a nasty flu and can’t get out of bed. Now she can’t wait tables by day or be stuck in an imaginary box by night.

In a new study out in GENETICS, Espinosa-Cantu and coworkers get to explore which enzymes in our devoted pal Saccharomyces cerevisiae are like Louise in that they do two unrelated jobs. They set out to systematically identify enzymes that have both a catalytic job and a job unrelated to its catalytic function. In other words, they set out to find enzymes that moonlight at second jobs.

Other researchers have stumbled upon these moonlighting enzymes in the course of their regular studies. Something along the lines of stumbling upon Louise struggling against an imaginary wind at a local park.

The difference here is that Espinosa-Cantu and coworkers systematically looked for dual-function enzymes. They did this by comparing the effects of knocking out the entire protein vs. just knocking out the catalytic activity of the enzyme with a single point mutation. They compared Louise having laryngitis vs. her being bedridden with illness.

They found that 4 out of the 11 enzymes they looked at did indeed have additional jobs unrelated to their catalytic function. These enzymes were both mimes and waitresses.

They started out looking at enzymes involved in amino acid biosynthesis mostly because it is easy to test them for loss of function. The yeast can’t grow in the absence of the amino acid the enzyme helps synthesize!

Next, like every good yeast researcher should, they turned to the Saccharomyces Genome Database (SGD) to identify 86 genes that when knocked out, caused the mutant yeast strain to not grow in the absence of an amino acid. They winnowed this list to 18 by focusing only on those enzymes with well-characterized catalytic sites and that were annotated with a single molecular function.

Then, they needed to have versions of these enzymes that were essentially identical to wild type except that they could not catalyze their reaction. They focused on single amino acid mutants that wiped out catalytic activity. The authors got this to work in 15 of their candidate enzymes.

They were now ready for the big experiment. They started out with a strain deleted for one of the enzymes and then added back either a wild type version or a catalytically inactive mutant version on a low copy (CEN-ARS) plasmid. They also included a third strain that had just an empty plasmid (their knockout strain).

They compared the growth of these three strains under 19 different growth conditions and looked for conditions where the knockout strain had a larger effect on growth compared to the strain with the catalytic mutation. The idea here is that if an enzyme does more than one job, knocking it out should have a larger effect than just disabling its enzymatic activity.


If Penny the waitress from Big Bang Theory is sick, both the Cheesecake factory and her movie set will suffer. (Wikia)

Before analyzing their results, the authors wanted to make sure that their wild type strain, the knockout strain with the deleted gene added on a CEN-ARS plasmid, grew equivalently to a true wild type strain. This was only true in 11/15 of the strains and so their analysis focuses on these 11 enzymes.

Espinosa-Cantu and coworkers found no difference in growth rate with all three strains for 38.6% of the growth rate comparisons. Under these conditions, the enzyme isn’t doing much for the growth rate of the cells.

In 233 growth conditions, the enzyme did matter for growth and in 70% of these, there was no difference between the knockout and catalytically inactive strains. Presumably under these conditions the enzymes have a single job—their catalytic function.

But this left 30% of cases where there was a difference suggesting that under these conditions, the enzymes have more than one job. Most of these involved 4 of the 11 genes—ALT1, BAT2, ILV1, and ILV2.

They decided to focus in more detail on one of these, ILV1. To figure out what else Ilv1p might be doing besides its threonine deaminase job, they generated two separate strains of double knockouts with 3,878 non-essential genes. The first set of 3,878 strains had a deletion of ILV1 while the second set had a catalytically inactive form of ILV1.

When they looked at the cases where the ILV1 knock out strain was more severely affected compared to the catalytically inactive strain, they found that 187 out of 345 negative interactions had at least some non-catalytic component.  The non-catalytic interactions were enriched for gene regulation with chromatin modification being one of the most striking.

These interactions put the non-catalytic role of ILV1 in a category more similar to stress response, protein sorting and chromatin remodeling genes than to other metabolic genes. This suggests that Ilv1p is both involved in amino acid metabolism and stress response. Quite a different moonlighting gig compared to its day job!

So researchers need to be careful when drawing conclusions from their knockout experiments. Sometimes the effect you are seeing is unrelated to the function you are studying. Sometimes when a waitress falls ill, the mime performance is affected as well.

No this isn’t a show about ILV1.

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

Categories: Research Spotlight

Tags: protein moonlighting , pleiotropy , systems genetics