Exploring the links between chromatin configurations, gene expression, nuclear morphology and cancer.
The DNA in every human cell is about two metres long. Somehow it has to be packed into the tiny cell nucleus. Which presents nature with a problem: how can a thread so long be compacted without excessive tangling and knotting? Furthermore, in order for genes to be read, they need to be exposed to enzymes. That requires the DNA to be continually unraveled and re-packaged in an exquisitely precise and controlled manner. The first level of compaction is understood: the famous double-helix of DNA is wound around little reels made of proteins called histones, like beads on a string, forming what is referred to as chromatin. Many more levels of folding and wrapping produce the structures known as chromosomes, familiar from photographs of cell nuclei.
Listen to Audio Interviews and Read Transcripts
Understanding the organization and dynamics of chromatin is a challenge requiring the expertise of mathematicians, physicists, chemists and biologists. On length scales between chromatin and chromosomes, most of the current picture is guesswork: a mesoscopic, amorphous, partially organized, information-rich and highly dynamic system. In cancer cells, the gross organization of chromatin is visibly altered, with obvious implications for gene expression and suppression. At the lowest level, variant forms of histones incorporated in chromatin have been implicated in triggering cancer. Chromatin structure and organization is thus the perfect intersection of topology and physical science with cancer biology.
This workshop brought together experts from all three disciplines in an attempt to identify ways in which physical interventions in chromatin organization might help us to understand and even control cancer.
Audio Interviews from the Workshop
- Interview with Dorothy Buck, PhD. Department of Mathematics,Imperial College, London. Dorothy Buck is a biomathematician who discusses how the theory of knots applies to chromatin.
More Info: http://www2.imperial.ac.uk/~dbuck | Interview Transcript
- Interview with De Witt Sumners, PhD. Robert O. Lawton Distinguished Professor of Mathematics and member of the Institute of Molecular Biophysics Department of Mathematics, Florida State University. More Info: http://www.math.fsu.edu/~sumners/ | Interview Transcript
- Interview with Jeanne Lawrence, PhD. Professor, Cell Biology at the University of Massachusetts Medical School.
More Info: http://www.umassmed.edu/cellbio/faculty/lawrence.cfm | Interview Transcript
- Interview with Steve Henikoff, PhD. Geneticist at Fred Hutchinson Cancer Institute
More Info: http://www.fhcrc.org/research/profiles/henikoff.html | Interview Transcript
- Interview with Jon Widom, PhD., Professor, Department of Molecular Biosciences, North Western University
More Info: http://www.chem.northwestern.edu/faculty/details?assetID=1465 | Interview Transcript
Interview with Dorothy Buck
Pauline Davies: Now Dorothy, how would you describe yourself?
Dorothy Buck: Erratically! But as a DNA topologist; I’m a mathematician and molecular biologist that looks at DNA protein interactions and looks at how those change one knot type to another.
Pauline Davies: Ok, so this all sounds very intriguing because a topologist is really a pure mathematician, so how does a pure mathematician come to be involved in biology and in cancer biology?
Dorothy Buck: That’s a good question. Infiltrating the enemy camp, as it were! I met a molecular biologist who’s interested in proteins that rearrange the DNA sequence that often as a bi-product form knots and links, like the kind you tie in your shoelace, and he was looking for a hired gun really, a mathematician to come analyze them and see what that said about the mechanism and the pathway of this interaction; so I started talking to him and one thing led to another as it does.
Pauline Davies: Well let’s go back to the beginning, because in a cell, you have chromosomes and DNA and there are times in the cell cycle when the DNA is very, very long, isn’t it? So, how does it pack itself up in a nice neat way and do what it’s supposed to do?
Dorothy Buck: That’s a really good question , and in fact an open question still I think. It’s a hierarchy of compactification and the first levels are well understood but the subsequent higher levels are still very mysterious. The first level though is this idea that it actually unwinds the ladder, the double helix and it does that by introducing writhe like the kind you see in the handset of your phone cord, if you’re old fashioned enough like me to still have handset connected to the base of your phone. So that actually already begins to compactifiy the DNA but the later levels are not well understood.
Pauline Davies: And it’s true to say that sometimes the DNA gets itself in knots?
Dorothy Buck: Yes, it does. But it’s quite unusual and when it does it is very bad for the DNA. Usually soon after that the cell commits cell suicide, as it were, if it cannot remove the knots or the links because these knots inhibit all kinds of important metabolic activities.
Pauline Davies: And how do the knots arise?
Dorothy Buck: Oh, so sometimes they arise on purpose and sometimes they arise on accident. On purpose is sometimes when you have circular DNA, the fact that your DNA is a double helix instead of a straight ladder, means when you pull it apart to copy it and have two new DNA molecules, they’re linked together, like your circle and forefinger, two of those link together. So, that’s one way where it happens naturally. Other ways are as I mentioned earlier, where it’s just a nice accident of the product, this long sticky polymer sort of gets in the way of the real reaction and ends up converting these super coils, as they’re called, into knot nodes.
Pauline Davies: And how often does this sort of thing happen?
Dorothy Buck: Well on purpose, these links that happen happen every time a DNA is copied. S o if you have a bacterial infection that’s happening hundreds of thousands an hour, (I’m sorry), but for knots it’s much more rare and especially invivo, in organisms, it’s fortunately not very often at all since, as we said, they’re quite dangerous.
Pauline Davies: On the other hand it seems a bit amazing that when you look at this jumble of DNA it’s just like strands that are thrown down, it looks a real tangle, how is it that you don’t get knots more often?
Dorothy Buck: It is shocking isn’t it? I really don’t know. I think it has to do with this organization, this hierarchy of compactification. That it is being massaged and buffeted and accessed by a whole battery of proteins that are acting as chaperones to kind of nudge it into the right shape. But it is miraculous that it is not more often, particularly at the ends of these very long chromosomal regions.
Pauline Davies: Really another example of nature knowing more tricks than we can even imagine.
Dorothy Buck: Certainly, yes indeed.
Pauline Davies: And if you do get knots, is there any way of untangling them?
Dorothy Buck: Yeah, so there’s these proteins in the cell whose sole function it is to unlock and unlink DNA molecules and they do them extraordinarily efficiently, much more efficient than most computer algorithms we have, so quickly and efficiently.
Pauline Davies: And how is that happening? Do you know?
Dorothy Buck: So they do it in the same way we would do it if we were just given a knotted piece of string and we gave up on trying to un-knot it nicely; we would just cut it like Alexander did with his sword I guess, and have it fall apart but then they reseal it after they do a crossing change. So they attach it a crossing of the knot, they cut one strand, pull it through the other and then reseal. The marvelous thing is if you do it randomly, you don’t usually get something unknotted, you just get another mess, but they are able somehow to drive it to an unknotting or unlinks.
Pauline Davies: It all sounds really amazing, now how are your ideas going down with the traditional biologists?
Dorothy Buck: Well, as I said, I infiltrated some of the enemy camps, so I spent six years in a wet lab so I think it helps that at least I can hold a pipette. I mean, I think that’s one nice thing about biology being really hard, there are all these chimeric teams of people looking at these problems. So for example to understand why these proteins unknot and unlink DNA so well, there’s teams of biochemists, molecular biologists, computer scientists, physicists, mathematicians like myself working in groups, and publishing in groups, to try and understand this process because the biological tools at this point are still just too crude to understand it completely.
Pauline Davies: There seems to be just so much we don’t understand about the inner workings of the cell.
Dorothy Buck: Yes, I think it depends on whether or not you’re a half full or half empty kind of person, but it seems like a great time to be doing molecular biology, that’s for sure.
Interview with De Witt Sumners
De Witt Sumners: I am De Witt Sumners. I am a mathematician at Florida State University and I study knot theory and topology and applications in molecular biology
Pauline Davies: So, how does that combination go together? Topology, that strikes me that’s pure mathematics, isn’t it? And then you have the applied stuff inside the chromosomes in fact, so, how does it marry together?
De Witt Sumners: Well DNA is a long and very thread like molecule and inside the cell it can become entangled, and this entanglement can be in the form of knots, like the knots you tie on your shoelace. And it’s possible for us to study enzymes that make knots in the cell and other enzymes that destroy them that fix that problem if the cell were to happen to generate some, which it does from time to time. So these knots are often diagnostic of molecular processes and you can harness them as reporters and experiments. So I got into it, I was born a pure mathematician, so for the first half of my academic career I did knot theory, pure as the driven snow. But t hen I saw some pictures of DNA knots coming from a the Cossarellin lab at the University of California at Berkeley (http://mcb.berkeley.edu/index.php?option=com_mcbfaculty&name=cozzarellin), and I went out and talked to Nick and I got very interested because what he had pictures of was the stuff I had only seen in textbooks and drawn myself, so here’s mother nature drawing the things that I was working on. So I was very fortunate, because my theoretical work sort of dovetailed with something that was actually practical.
Pauline Davies: And how common are knots in nature whenever you’ve got a long strand?
De Witt Sumners: Well they are very common, in fact, there’s a theorem that I was one of the author’s of, that says that if you’ve got a very long, randomly imbedded curve that its knotted with high probability, it’s got a knot in it. And then the longer it is, the more likely it is to be knotted. And as the link goes to infinity, the knot probability goes to one, exponentially rapidly. And so this is actually useful in some scientific circles because you can work with long molecules and do simulations of knotting and compare that to what you see in the test tube. And it turns out that when you look at viruses, the viruses that infect bacteria whose genomes are DNA, in order to study how the DNA is packed very tightly in the viral capsid, you can look at knots that are produced when the DNA is released from the capsid and the knots are then diagnostic of the geometry of the packing; how it is wrapped around inside that very tightly packed ball.
Pauline Davies: When I think about knots in long pieces of string or even maybe my vacuum cord, there doesn’t seem to be much structure in it and I’m a bit puzzled as to how the geometry of the surroundings would affect the long cord that has got itself in a tangle?
De Witt Sumners: Yeah, well if you were to, say, take your hundred foot extension cord and wind it up very carefully and then connect the ends up and then lay it all out you would probably get a knot that’s fairly simple or it would be a spool-like knot, where as if you were to tell your fifteen year old son to go throw the thing in the garage, that’s randomly packing it, then if you were to put the ends together then you would get a nasty looking random knot. So it’s possible to tell if knots were formed randomly or if they were formed by some cellular process, for example.
Pauline Davies: So if this sort of thing happens in cells and the DNA get’s itself in a twist, how does the organism cope?
De Witt Sumners: Well the organism copes by, there’s an enzyme known as topoisomerase which exists primarily for the purpose of fixing these entanglements, because the entanglements can be lethal to the cell. So these enzymes find the knots and kill them preferentially, it’s not really understood how they act they have to act locally and a knotting is a global phenomenon, so the question is how do they know where to pass the DNA through itself, which is what they do? But they do, and they do it very rapidly and very accurately, so they are very important for cells because without them, the cell will die.
Pauline Davies: So I am still a little bit puzzled because you say the knots have got to be killed, do you mean disentangled?
De Witt Sumners: Disentanglement, turned into open circular form, so with no knots in them at all. Because there is an experiment that has a DNA plasmid , a short, circular piece of DNA, with a gene on it that codes for a resistance to ampicillin, and if you knot that plasmid and then hit the cell with ampicillin, the cell will die because the knot prevents the gene from being expressed.
Pauline Davies: And how do these proteins know where to guide the strand of DNA so to untangle it?
De Witt Sumners: That’s a real interesting question. A lot of people are working on it; it’s not clear. So that’s an open theoretical question and an experimental question, so people are designing experiments to try to figure this out – as to how they know where to do the magic. And people do simulations and there are some indications that it might look for regions of curvature where you’ve got a curved piece of DNA wrapping around a straight piece, maybe that’s it. But the enzymes know how to do it and we don’t. I mean they are a lot smarter than we are that’s the long and the short of it so. They’ve had you know millions of years to evolve so that’s why they are really good at it.
Pauline Davies: So this is an entirely new way of looking at what is going on in a cell, how are biologists taking this?
De Witt Sumners: Well they think it is very interesting because they have come up with these problems and as mathematicians, they’re great problems. They aren’t easy, so they are very interesting and so it’s a fertile ground for interaction between mathematicians and biologists, so biologists are very receptive to it.
Interview with Jeanne Lawrence
Jeanne Lawrence: I’m Jeanne Lawrence. I’m a professor in the Department of Cell Biology at the University of Massachusetts Medical School and I study chromosomes, genome regulation and nuclear structure.
Pauline Davies: And from what I was picking up from your talk you’re saying that cancer is caused when the regulators of chromatin don’t behave in the way they should behave, is that right?
Jeanne Lawrence: Well, I was saying that cancer can develop for a lot of reasons. OK, so some cancers would develop because there is a specific mutation in a tumor suppressor. But I was suggesting that if you have defects in the epigenetic regulation of a genome then this would produce so many different expression profiles that any one of them has the potential to become neo-plastic, and then that would be selected for because those survive the best. So basically I was suggesting, specifically, that heterochromatin – regions of repressed DNA – if you have a defect in maintaining that repression, you could really contribute to a neo-plastic progression. I didn’t say something which I should have, which is, it’s known that silencing of tumor suppressor genes inappropriately is part of the mechanism of some cancers, but that actually often occurs in the context in loss of silencing of other regions. So it’s really about this imbalance in the epigenetic state, not necessarily all repressive or all activating, but the wrong things being repressed and the wrong things are being activated.
Pauline Davies: And what is the epigenetic state?
Jeanne Lawrence: Well, I guess my definition of the epigenetic state would be the status of the genome in terms of what genes are being expressed and what genes are not which will vary with cell type.
Pauline Davies: But what causes that; what causes the expression of these genes?
Jeanne Lawrence: Well, we’ve had a lot of discussion about that and that’s not a simple answer, but it’s an interplay of transcription factors, accessibility of chromatin based on chromatin modifications, andin my case, I am quite interested in the role of non-coding RNA’s and helping to define that epigenetic state. So what I tell my students is that your genome is kind of like the hardware of your computer, and the epigenome is like the software that interprets the information or programs in it in certain ways; so I don’t know if that makes sense.
Pauline Davies: It makes, well it does make sense, and in terms of helping us understand cancer and eventually preventing and curing cancer, are we any way forward?
Jeanne Lawrence: Well, gee that’s a big question. I’m sure that we are forward, we’ve made advances because there’s a lot of specific genes like p53 and BRCA that we know have specific roles, and they have roles in quite a number of cancers, particularly like p53 or MIC, we’ve learned a lot about those, but I think part of the difficulty and why we haven’t been more successful, is that there are so many ways to get cancer and that’s what I was trying to say. As long as you have a diversity of aberrant expression profiles, maybe 99 percent of them will be selected against and those cells will not grow, but you just need a small fraction of cells to have a growth advantage and you can generate a cancer. So cancer can be derived in many ways and I think we need to not think of it as one thing, I am certainly not the only person who says that, but it’s many different kinds of things that can lead to abnormal cell growth.
Pauline Davies: Yes, I think the public misunderstands that, they tend to lump everything together and call it cancer and think it is one disease.
Jeanne Lawrence: Yeah, and it really isn’t. and that’s what I was saying; that if you’re trying to bring physics into it, and physics likes to look for unifying principles – I mean maybe the unifying principle is that there Is no unifying mechanism you know, there’s just a host of different mechanisms – but certainly one of them could be to change the epigenetic program of the cell and that might be one that has lagged a little behind looking for gene mutations. More and more people are studying that now.
Pauline Davies: Have you learned anything from this meeting?
Jeanne Lawrence: Well actually the last talk about the DNA sensitivity was quite interesting – I didn’t know about that work and that was quite interesting. And I think some ideas that I’ve wondered about before, about what exactly silences a gene, hearing from experts like Steve Heinkoff where he’s talking about it isn’t any particular chromatin modification, it could be the compaction of DNA, is something I’ve thought about before and it was interesting to hear him talk about that.
Pauline Davies: Terrific, and what about the topologists, did that make any impression on you or did you think that wasn’t relevant?
Jeanne Lawrence: I don’t know a lot about DNA topology so I was on a learning curve there and you know it’s useful background for me to have. It didn’t directly link to anything I’m doing, in part because the studies were mostly on naked DNA and in vitro, which is a very far cry from the actual genome inside a cell, so I wasn’t quite sure how to connect it; but it was interesting to know that background.
Pauline Davies: Well, thank you very much.
Jeanne Lawrence: Ok!
Interview with Steve Henikoff
Steve Henikoff: I’m Steve Henikoff. I’m a geneticist at the Fred Hutchinson Cancer Research Center.
Pauline Davies: And Steve you were talking about that centromere in this meeting and its importance in cancer development, can you explain the thrust of your argument?
Steve Henikoff: Well first of all, you have to understand what the centromere is. The centromere is the point of attachment of a chromosome to pull to the poles at every cell division. It’s central to segregation of chromosomes. So, every chromosome has one centromere, and if it has zero or two, you’re dead. Centromeres have to do everything nearly perfectly and the amazing thing about centromeres is that you would think that they would use DNA sequence in order to do this perfect process, but apparently they don’t. They are maintained independently of DNA sequence and we’re trying to understand that. And it’s turning out that it’s the basic sub unit of chromatin, called a nucleosome, that’s very very different. A nd it’s so different that what we’ve been finding is that it actually wraps DNA just the opposite of the way it’s wrapped in the rest of the genome. So, some fundamental difference in how you wrap the DNA around proteins at centromeres seems to be what’s responsible for this remarkable behavior and we’re trying to work that out.
Pauline Davies: Why would it matter whether or not the chromatin is wrapped left or right?
Steve Henikoff: Well, we can only speculate on this but here’s an interesting possibility. Let’s think about what centromeres do; they have to be pulled upon so there’s all this force on the centromere that pulls it to the poles, I mean when you actually watch chromosomes you see them go very quickly to the poles. So, there’s a lot of force that’s being exerted on it. Now, if you actually calculate how much force there is pulling on it, you have to worry a little bit about popping out nucleosomes and losing them. So what we think is going is that, it’s been shown that actually if you pull on DNA, just pull on it from the ends, it actually starts to twist, it actually winds up, somewhat counter-intuitively, before it eventually unwinds – up to about thirty piconewtons of force, that’s kind of a lot of force. So, what happens is, we think, that when you pull on it, it tightens up, rather than pops out. So we think it’s inherently better for staying in there and resisting the force.
Pauline Davies: It sounds like you’ve got a very good grasp about what’s going on, but yet there’s still lots of research to do in this area?
Steve Henikoff: What I’m telling you is still pretty much speculative, a lot of what I’ve told you, we haven’t even written up for publication – it’s the way we’re thinking about it and we’re testing various models. But I’m trying to give you a sense of how we think it all works. Some of what I told you we’ve actually shown in both in vitro and in vivo but of course only in a very simple model organism and it’s still unclear whether what I’m telling you about the wrapping is true in general. We think it is but there’s a lot to do there, even if we are right about it.
Pauline Davies: It sounds like very exciting work.
Steve Henikoff: Oh yeah, I’ve been doing this kind of thing for over 25, maybe 30 years and this is to me is the most exciting science I’ve ever done.
Pauline Davies: Final question, what do you think was the most valuable thing about this meeting?
Steve Henikoff: Well, I think getting together and talking about the problems in the field and learning that we actually have similar and related ideas and fitting them together. For example when John Stempolopolis, he basically showed some of his newest data and it was quite spectacular and it made it clear to me that some of the ideas that we’ve been thinking about with respect to how dynamic nucleosomes are, might be actually correct. But it came from looking at results he presented, which were also new results. So sharing those kinds of observations and thinking about it and putting them together with your own observations made this quite a valuable day for me.
Interview with Jon Widom
Jon Widom: I’m Jon Widom. I’m a professor at Northwestern University and my background was chemistry in college and biochemistry in graduate school and structural biology as a post doc and I began my life as a physical chemistry professor and now I’m a molecular biology professor, mostly.
Pauline Davies: Wow, all sorts of things. Can you describe what’s going on in the cell and in the nucleus and all that important stuff?
Jon Widom: Well the short answer is no. I can’t and nobody else can either.
Pauline Davies: Because I’m totally amazed that we can actually see pictures of nucleosomes which consist of proteins and then they are wrapped around like DNA, a bit like beads on a string, we can see that and that’s the really fundamental building block of the chromosomes. And then at the other end of the spectrum we can see the chromosomes and that’s been familiar to us for decades and decades, but in between it’s all a big mystery.
Jon Widom: Yeah, there are two puzzles; one is, what happens in between? And then also, with regard to the most compact chromosome, yes you can see it but we really have no idea what its structure is, in part because we don’t know how you go from the chain of nucleosomes to the chromosome. Every problem like this on these size scales in cell biology is hard. So the example that I like to use is the Golgi apparatus which was discovered, I don’t know, a hundred and fifty years ago or something like that. And we still today do not know if it’s one thing or a set of unconnected things. And it’s because problems on this size scale are hard; the structures often have order but they are not regular in the geometric sense, and that makes structural studies much more difficult. And it’s a problematic size and it’s small for a lot of conventional optical microscopy,big for electron microscopy and there are terrible problems with radiation damage, super resolution optical microscopy can help, but there is the pesky third dimension. And so all objects in cell biology, or more or less all objects with a few notable exceptions, are proving extremely problematic. Many of us think that some combination of fancy electron microscope tomography and super resolution optical microscopy will eventually solve the problems, but they are in no danger of finishing that today.
Pauline Davies: So, tell me how the smallest objects are visualized?
Jon Widom: Yeah, the smallest objects turn out to be easy because you can chop them up, you can liberate individual nucleosomes by chopping up chromatin fibers using enzymes that munch the DNA that isn’t wrapped. So, you can biochcemically purify nucleosomes and then they’re molecules and you can crystallize them, and then the crystal of the nucleosomes is ordered and x ray crystallographic methods allow you to solve the structures. And then these days you can build nucleosomes yourself from purified components to make really homogeneous nucleosomes and really nice crystals and get really good high resolution structures.
Pauline Davies: And to actually see the DNA wrapped around them, I think that is quite amazing.
Jon Widom: Yeah, that was a shocking proposal by Roger Kornberg originally a very long time ago in advanced of the definitive structural studies and then confirmed brilliantly in a series of studies leading to the first crystal structure in 1984.
Pauline Davies: So, all this stuff in between, we can see the chromosomes as you explained, the stuff in between is so hard, is it because people just haven’t thought of developing suitable microscopes for that sort of size?
Jon Widom: No, people have thought lots and lots. Again, a lot of the action in cell biology occurs on these intermediate size scales bigger than a molecule or bigger than a single protein, but small compared to the cell. So people are desperately developing new methods to try to understand these problems.
Pauline Davies: Is it a matter of money?
Jon Widom: No, it’s tough physics and ugly details. So an example of an ugly detail is electron microscopes for decades have had more than sufficient resolution to obtain atomic resolution images but the problems are two-fold so it’s a three dimensional problem so you need to do EM tomography, and then the problems are – in no particular order. There is almost no contrast. Cells are filled with stuff all built with light atoms, so it all looks the same to the electrons and then worse yet, the radiation damage causes the samples to vaporize long before you’ve taken enough pictures for high resolution tomography. So those are just ugly details but very real ones, they are physical limits, that are always with us and so we have to live with them and try to extract the most information as possible despite the physical limits.
Pauline Davies: So all that stuff that we see in the textbook, we see the pictures of the nucleosomes with the DNA wrapped around it and then we see all sorts of structure of the chromatin condensing and piling itself up nice and neatly, and then wrapping itself around itself and eventually turning into chromosomes, are you saying all that is make believe?
Jon Widom: It is strictly fantasy. There are few aspects of reality, for example, loops are shown at some level down the figure, and there, in certain respects, those demonstrably exist, but all the other things that are shown are strictly fantasy. And so, yeah, that just highlights what we don’t know.