Polymerase Pausing: It’s like propping open a molecular door – Recap of talk Dr. Karen Adelman

The Talk:

Regulating signaling networks through pausing of RNA polymerase II

In Plain English:

Cells sometimes keep the RNA-builder enzyme “paused” at the first few base pairs in the gene, which leaves the gene somewhere between “on” and “off”. (It’s kind of like propping a door to avoid getting locked out, except more complicated.)

The Speaker:

Karen Adelman of the NIEHS (National Institute of Environmental Health Science)

The Sponsor:

MIT Biology Colloquium Series

What it covered:

Dr. Karen Adelman is a molecular biologist who studies the behavior of a RNA-building protein called polymerase (RNA Polymerase II, to be exact. It also goes by the nickname “Pol II”) and how it reacts to stress signals that come from outside the cell. Her research is helping to fill a major knowledge gap:

We know quite a bit about how cells regulate their gene expression and about how they sense changes in the surrounding environment.

But we don’t know very much about the chemicals that allow cells to change their gene expression in response to signals they’re getting from outside the cell, which makes it very hard to predict how outside chemicals (like medications and the particles found in smog) will affect cellular health.

There’s a good reason for this omission: Cells are incomprehensibly complicated. They are made of billions of medium-sized molecules called proteins, and these proteins are constantly swirling around and bumping into each other. A lot of people use engineering analogies to explain these relationships, but honestly, I think comparing protein interactions to man-made machines is like comparing Times Square to a “building” made out of duplo blocks. Proteins are constantly shuffling around, getting stuck to each other, blocking each other’s routes, breaking each other apart, and generally causing all kinds of mayhem. Molecular biologists’ job is to observe the jumble of proteins and look for recurring patterns but there are so many proteins that no one person could possibly keep track of all them.

Most molecular biologists focus on one particular signaling pathway (a “signaling pathway” or a “signal cascade” is series of interactions between interlocking proteins causes some kind of effect in the cell). There are scientists who study “extracellular” signaling pathways that happen outside the cell, and there are other scientists who study signal cascades within cells. The problem is that signaling pathways aren’t isolated chains of cause-and-effect; most proteins are involved in multiple pathways, which means that what happens in one signaling pathway can affect dozens of other pathways.

For a long time, we’ve conjectured that extracellular signals associated with stress somehow cause a chain reaction in the cell that affects the proteins responsible for copying genes, but we don’t know very much about the middle steps of the chain reaction.

That’s where Dr. Adelman’s work on polymerase comes in.

Her background is in the behavior of polymerases, the protein complexes that build DNA and RNA strands. RNA polymerase II, her main protein complex of interest, lives in the nucleus and is responsible for transcribing genes. (If you’re not sure what “transcription” is, see genetics 101 refresher below.*) However, she noticed that the exact same extracellular could trigger several different and distinct changes in Pol II activity. For instance, T-regs and macrophages have very different responses to the same stress hormones, even though they’re both white blood cells.

Scientists have noticed that there are many genes where Pol II, which is a very fast-moving enzyme once it gets going, just “hovers” over one spot. In these cases, the DNA is “unzipped”, and the polymerase is already in position on top of the base pairs, but the polymerase doesn’t move forward with transcription.

This type of behavior is pretty strange, because 1) Polymerase is actually protein complex, which means that it’s actually a glob of 4-5 smaller proteins that work together to perform the transcription task. If the polymerase doesn’t start transcribing, it will break apart, leaving only useless pieces behind and 2) If you want a gene to be “turned off”, letting the DNA reseal itself is a much more fool-proof way to shut down the gene. Even when it’s “paused”, Pol II has a hair trigger, and it will start transcribing at the first opportunity.

Dr. Adelman and her colleagues wondered: “What’s causing the polymerase to hover? And what’s the point of going to all the trouble of assembling a polymerase if you just want the gene “switched off”?

In the early 2000s, a team of scientists noticed that one particular protein was always hanging around with the paused polymerases and named it NELF (for “Negative Elongation Factor”) because it binds to Pol II and stops it from adding nucleotides (“elongating”) to the transcript it’s building.

Some genes are more likely to be “paused” than others, so Dr. Adelman’s team decided to create a line of mutant flies with a broken NELF gene and see how the absence of NELF would affect frequently-paused genes. They expected that since NELF stops the polymerase from transcribing that these genes would be over-expressed in the NELF knock-out flies.

However, they found that the genes that were usually paused where under-expressed in the NELF-less flies. In fact, since a lot of the frequently-paused genes are important for regulating tissue development, the NELF-less flies died as embryos because the frequently-paused proteins were missing. They had to create a second set of mutant flies, where the NELF was only missing in certain types of cells (like the immune cells) so that they could actually study NELF’s behavior in a living organism.

Dr. Adelman told us, “NELF is like a puppy that follows Pol II everywhere it goes,” but I thought it sounded more like a built-in leash. NELF grabs Pol II and forces it to “pause” before it can get very far in its transcript. So the polymerase ends up tethered to one spot by the NELF and can’t get its work done until NELF leaves.

Fortunately, NELF has another protein buddy called P-TEFb (Positive Transcription Elongation Factor-b) that can bind to NELF and force it to let go of Pol II. There are several different types of P-TEFb proteins, and they are drawn to specific regions in the genome. If a P-TEFb wanders over to its favorite genomic hang-out and sees a NELF clinging to a Pol II, it will bind to NELF, setting the polymerase free. However, if there aren’t any P-TEFbs nearby, the NELF just keeps hold on to Pol II. By adjusting which types of P-TEFb it produces, the cell can engineer “casual run-ins” between NELF and P-TEFb, which allows the cell to control which transcripts are “paused”.


[I was kind of sleep-deprived while I was doodling up this visual…I feel like I owe NELF an apology for drawing it as Rick Astley, but I couldn’t a good diagram of what the protein actually looked like. Here is the model of RNA polymerase II that I based its character on.]

When NELF is completely knocked out, the polymerase simply copies the gene without pausing. The problem is that unless another polymerase forms over the promoter immediately after the first polymerase leaves, the DNA will close up the “transcription bubble” and wrap itself around a protein called a histone. Once that happens, you need a whole slew of proteins to get the DNA unwrapped again.

However, if there is a “paused” polymerase sitting on the DNA strand, the DNA won’t close up and the polymerase can begin transcription at a millisecond’s notice. All you need to do is get the P-TEFb to show up and then– boom!

Dr. Adelman still wasn’t sure why a cell would want to “pause” a gene in the first place. She couldn’t think of any proteins in the nucleus that would alter the expression or behavior of PTEF-b, and if the PTEF-b wasn’t susceptible to change, what was the point of all of this?

“That’s when my lab had an intervention,” she quipped. “They had to remind me that there are these proteins out in this thing here– it’s called the cytoplasm– that are also part of the gene ontology complex.”

Her team had noticed that many frequently-paused genes were involved in the flies’ immune systems, and after running a protein analysis, they found that there were more NELF proteins floating around in cells from flies that were fighting bacterial infection.

Regulating immune response is a tricky business. Macrophages and killer T-cells are among the most dangerous cells in your body; their job is to kill anything that looks like it might be harmful. During infection, you want your immune response to be quick and lethal, but the rest of the time, you want your immune cells to be as calm and controlled as possible. Otherwise they will start killing bystander cells.

There are a lot of highly influential extracellular signals that tell immune cells how to behave, where to go, and when to die. Cells have to be able to produce these distress signals at a moment’s notice, but if cells overproduce these signals, other cells around them will start freaking out, too. And then they’ll start overproducing stress signals. If enough cells are overproducing stress signals, the immune cells will freak out, and pretty soon you have a full-blown autoimmune disease.

Dr. Adelman’s team hypothesized that pausing might be a tool cells can use to leave these crucial signals “partially on” without running the risk of overproducing these signals. If a cell is actually in danger, it would be relatively easy for the cell to up its production of PTEF-bs and turn off the NELFs, so pausing would probably be a good way to ensure rapid-but-not-excessive immune response.

I’m condensing years of research and dozens of painstaking experiments into a few paragraphs, but it looks like this hypothesis is holding up. NELF-deficient tissues have serious problems with autoimmune attacks.

When they knocked out the NELF genes in human lung cells (in a petri dish, not an actual living human), they ended up with lung cells that became highly inflamed at the slightest exposure to a foreign compound. In these petri-dish-based lung tissues, they found macrophages that were so confused, they were eating each other. If this tissue belonged to an actual person, they would not have been able to breathe.

Research into NELF’s behavior is still in its early stages, but it looks like NELF may be a promising drug target for anti-autoimmune drugs. Very few proteins are expressed as widely as NELF is, and it’s very possible that some autoimmune disorders are at least partially caused by deficiencies in NELF. We also know that PTEF-b is frenemies with NF-kB, a protein that plays a key role in altering transcription pathways during infection (and has one of the longest wikipedia pages I’ve ever seen for a protein). It looks like all three of these proteins are key power players in determining transcriptional response to stress signals, which means scientists want to know a lot more about them.


*Quick genetics 101 refresher: “Transcription” is the process of copying the base pairs in the DNA onto an RNA strand. Polymerases are very powerful players in gene expression, but they rely on signals from other nucleus-dwelling proteins to tell them which genes they need to transcribe.

This step is crucial, because each gene codes for just one protein, and each RNA-copy of the gene codes for just one unit of the protein. So if you need 100 copies of Protein-X, you would need at least 100 RNA-transcripts of the gene that codes for Protein-X. (Some RNA-copies get broken or lost or kidnapped on their way to the ribosome, so you would probably want to slightly overproduce RNA-transcripts to account for those losses.) There are a whole slew of things that can affect a gene’s transcription rate: If another protein or chemical wanders into the nucleus and happens to get stuck to the gene, then the polymerase can’t transcribe it. If the helicase, aka “the enzyme that unzips DNA” gets distracted or unzips the wrong section of the DNA, then the polymerase won’t be able to transcribe the “target gene”. And if the proteins that tell polymerase where to go are messed up, then the polymerase could very easily end up making bazillions of copies of the wrong protein.

My Personal Take:

I thought that this talk was really cool, but even as I was frantically scribbling notes, I was thinking, “This is fantastic. But if I tried to explain why this research is important to someone without a biology degree, it would probably take like an hour.

It took me almost three hours of writing to come up with an analogy I liked: Pausing the polymerase is like propping the door to your dorm so that it doesn’t lock. You want the door mostly closed, but getting locked out is a huge hassle. When the nucleosome forms, that’s the cellular equivalent of getting locked out of a particular gene. The lucky polymerase gets to be the door prop. I promptly forgot about this analogy as soon as I started describing the protein interactions.

Autoimmune diseases are a huge issue. They’re becoming more prevalent, they have insanely complicated prognosises, and even the world leaders in biomedicine are only beginning to figure out what’s going on in chronically dysfunctional immune systems. You would think that a protein like NELF, which has potentially huge implications for health and medicine, would get a lot of attention.

But it doesn’t. Molecular biology is so full of acronyms and oddly specific technical terms that it’s almost a separate language. You have to practice listening to molecular biology speak to be able to understand it, but hardly anyone has time to do that. It doesn’t translate into 300-word news briefs very well. Especially when we’re dealing with a topic like polymerase pausing, where almost all of the key players are proteins that most people have never heard of (and never will).

The average person on the street doesn’t need to know very much about molecular biology (and honestly, it’s hard to keep track of more than a few crucial pathways), but I do think people should understand just how hard molecular biologists have to work to predict all the possible side effects of medications and how tough it is for environmental toxicologists to pinpoint a particular chemical that’s responsible for a spate of health problems. These are Herculean feats.

Also, on a personal level, I just really like learning about molecular biology because I like looking at the cause-and-effect patterns. Don’t get me wrong: I had almost no idea what was happening during the first month of my cancer biology seminar (where we were constantly reading papers about topics like this), but once I got a little bit more fluent in molecular biology, it became one of my favorite classes. Some of the words and acronyms are a little bit ridiculous, but I love learning about all the little proteins that hardly anybody knows about. (“NELF is still underground, man. It hasn’t gone commercial yet.”)

I also thought it was a nice example of a possible mechanism for rheostat-style transcription control. I’m always hearing cell biologists talk about genes in terms of “light switch” genes (which are always either “on” or “off”) and “rheostat” genes (which are expressed at varying levels), but only a few mechanisms have been proposed to account for the varying expression levels associated with rheostat-control. The ones that I’ve heard are very confusing and full of “regulatory feedback loops”, which are basically signal cascades where the proteins at the downstream end can re-activate the upstream proteins and start the whole process over again.

So it’s nice to hear about a molecular mechanism for controlling immune system activity that sounds pretty drug target-able.

Biggest Misconception to Avoid:

When a polymerase is “paused”, it’s usually paused for less than 15 minutes. (That’s an entire lifetime in transcription protein time.) And in the vast majority of cases, the polymerase does get released and is able to complete its transcript. NELF is good at its job, but so is P-TEFb.

Best One-Liner:

“I love NELF, but it’s dumb. It just goes everywhere polymerase goes. P-TEFb is the smart part of the complex, and it’s the one responsible for targeting.”

Best Audience Question:

Audience member: If I’m understanding you correctly, the pause happens in one of two ways: Either the polymerase is paused until release (meaning it goes forward and copies the gene) or the polymerase is paused until it turns over (meaning the polymerase falls off of the promoter region and breaks apart, in which case the promoter region would have to recruit another polymerase to get the gene transcribed). How much turn-over is there in polymerase?

Dr. Adelman: (translating for the undergrads in the audience) So the question is: To what extent is the polymerase complex that gets paused the one that makes the full transcript? Versus how often does it fall off, leaving the promoter region to recruit “another, better polymerase”?

(now answering the question) Well, these are in vivo dynamics, so I don’t think anyone is completely certain.

But we did an experiment where we added triptolide, a compound which blocks the helicases from reaching the promoter region (Helicases are the enzymes that unzip the DNA so that the polymerase can land on it. So if the helicases are blocked, the polymerases can’t do their own thing and end up just floating around until they break apart) and then we watched the decay rate of the paused complexes. The decay rate was inversely correlated with the number of polymerase complexes.

Now we were only looking at 15 samples, which is a pretty small n, but it looked like the polymerases were completing the gene transcription rather than terminating. However, some genes are more likely to terminate than others, so it could vary a lot depending on where you are in the genome.

Key Terms:

  • Polymerase = a protein complex that builds a strand of DNA or RNA. RNA polymerase II is the enzyme that builds mRNA in eukaryotic cells.
  • Protein complex = a group of conjoined proteins that work together to accomplish a task
  • Nucleosome = a piece of DNA wrapped around a histone. Once a nucleosome forms, it’s much harder to initiate transcription. A chromosome is a chain of nucleosomes.
  • Paused polymerase = when a polymerase is prevented from releasing and completing its transcript after binding to the DNA’s promoter region
  • Signaling pathway = a series of protein interactions that causes something to happen
  • Signal transduction = the process of an extracellular molecule interacting with a receptor in the cell’s outer membrane and causing a protein cascade that alters an aspect of cell function.

Tl;dr: A protein called NELF sometimes grabs the proteins that create RNA-transcripts and forces it to “pause” at the promoter region. NELF’s behavior allows cells to express genes at low levels without completely turning off transcription.


  1. Patrick Sweetman

    Readable from the POV of a recent graduate, but you need to differentiate your writing more for different audiences, and say at the beginning what level of reader you expect.

    I like that spell out the enormity of our ignorance regarding processes within cells.

Leave a Reply

Your email address will not be published. Required fields are marked *

Website Protected by Spam Master