3W4DW: My research

[liv]

forestofglory asked to hear more about how your research is going. Goodness only knows I'll take any excuse to talk shop, but I have to be a bit cautious about what I post online regarding work that's currently in progress. So I'm going to take a slightly different tack approaching this question, and instead talk for a bit about why I'm interested in what I'm interested in.

The old fashioned word for the sort of thing I care about is cybernetics. These days cybernetics refers almost exclusively to computers, but in the late 20^th century, it was to do with regarding biological systems in terms of the flow of information. How does an organism, or an organ, or even a cell, receive a whole bunch of inputs from the environment and alter its behaviour appropriately? In informal language I tend to refer to this as decision-making, though it isn't a literal conscious decision process.

I got into cancer because it's one of two sub-fields where approaches to this kind of question are most developed. (The other is embryonic development, which put me off because I'm kinda squeamish about directly performing experiments on mammals and didn't fancy spending my career breeding flies or teeny worms.) And yes, of course I was somewhat motivated by a desire to help find a cure for such a terrible disease, and there were fairly chance elements as there always are in any career path.

So let's say you have a cell in the epithelium of the glands of the breast. Epithelia are the layers at the edges of tissues, so they frequently get sheared off, just by physical abrasion. So that the epithelium doesn't get worn away altogether, it has to produce new cells. Producing new cells means converting nutrients from the diet into the substances of cells, such that the "parent" cell doubles in size. It has to make exactly two copies of its genome, with no extra DNA and no genes left out. Part of this is the semi-conservative replication most people learn about in high school biology, where the two complementary strands of DNA separate and enzymes copy both strands by means of Watson-Crick base-pairing, allowing each strand of the helix to act as a template. That's all fine and good, but the cell still has to "know" when it is time to copy the DNA, and when it has completed the task. Once you have a cell that is twice the size of a typical cell of its type, containing exactly two copies of every chromosome, the cell divides into two "daughters", with each one receiving exactly one copy of each chromosome and (usually) half the cell contents. That's mitosis, which again you may be familiar with.

The thing is, though, the cell can't carry out this process based on purely internal cues, though even that is complicated enough! It needs to divide to make more cells when the epithelium has worn away enough to need replacements, but no more than that as that would lead to excess cells. And excess cells are, well, we have a word for that: a growth or a tumour. In addition, the cell has to change its rate of division based on all kinds of information, such as whether it's in a male or female body, whether it is in an embryo / foetus which needs to undergo some rapid growth, or a child whose breast tissue is growing only in proportion to the rest of its body, or in a girl experiencing puberty whose breast tissue expands rapidly (but not indefinitely!), or indeed whether it's in a pregnant or lactating woman and exactly what are the demands for milk at any given moment. I could tell a similar story for the cells of many other epithelial types, but breasts are a good example.

One way that the cell integrates all these inputs is that chemicals are produced in different parts of the body, hormones or substances produced under the influence of hormones usually, but there are neurological factors as well, and some inputs that respond to conscious experience of the environment. These are carried in the bloodstream to the relevant epithelium, and bind to specialist proteins called receptors either within the cell or more commonly on its surface. Receptors are activated by the binding of these chemicals, and they initiate signalling processes which can alter the cell's behaviour. At any given moment there are going to be several signalling processes going on and the outcome is going to depend on the exact balance between all of them. For example, if somebody is undergoing puberty, and also starving, it's probably better to concentrate resources on survival rather than growing new breast tissue, so the cell is going to have to receive "puberty time" signals as well as "not enough nutrition" signals and act accordingly.

The signalling processes typically involve a cascade of chemical modifications to a series of proteins, the receptor modifies protein A, which in turn is activated to modify protein B, which produces a chemical to activate protein C and so on. I sometimes care what's going on on a chemical level, eg has the charge of the protein changed by the addition of an electrically negative phosphate group, causing it to adopt a new configuration which is now more energetically favourable? But a lot of the time I actually don't care very much; the level of abstraction that I'm interested in is more about regarding each protein in the cascade as an abstract entity, which is either in an active or a repressed state, and how that is going to affect how the cell behaves.

Short term changes in behaviour can often just involve enzymes which alter the chemical properties of the cell, but longer-term changes, such as committing to division, usually require changes in gene expression and production of new proteins. This is really important because obviously all the cells in the epithelium will have the same genome, yet some will divide and some will not. Not only this, but cells of all the other tissues of the body have the same genome too, and most of them aren't breast epithelium at all, they are brain neurones or kidney filtering cells or immune defence cells or whatever. In order to achieve this, the cells also have the ability to turn genes on and off more or less permanently, as well as in short term response to the current state of signalling. This happens through making chemical modifications or marks on the DNA itself, as well as the structural proteins that hold the DNA together, changing a particular stretch of DNA between a closed state which can't normally be switched on, and an open state which is ready to be expressed if the appropriate combination of signals is present.

To focus down a bit further, I'm particularly interested in a process called apoptosis. From the Greek meaning "leaf fall", this is effectively a fail-safe mechanism by which cells can commit suicide. There are several possible reasons why cell suicide is beneficial to the organism as a whole. One is to remove structures that are needed at one stage and then cease to be needed (eg the lactating mother weans her child), and one is to help maintain exactly the right amount of cells in a tissue, eg if slightly too many are produced for some reason. But another reason for cell suicide is if there is serious damage to the cell; it is far better to remove one cell out of the billions in the body, than to propagate the damage by making more copies of it. This process is one of the key reasons why 2/3 of people never get cancer at all, and most of the rest live 60 years or more before they get it. Even though cells are constantly bombarded with radiation, chemical damage, for some cell types infection by parasites, and all acquire genetic mutations, the great, great majority of the harmful ones trigger apoptosis before the affected cell has a chance to divide out of control and cause a tumour.

During my PhD I worked on a protein called p53, which does a whole bunch of interesting things but the one relevant to me is that it detects cells that have faulty, contradictory signalling and turns on genes that cause apoptosis, before the cell has a chance to turn into a tumour, or if that doesn't work, it can also make cells go into apoptosis before a tumour becomes malignant or metastatic. I was trying to find drugs which would hyper-activate p53 so that tumours would destroy themselves through cell suicide, whereas normal cells would largely be fine because if everything else is normal, one signal alone isn't enough to trigger p53.

As part of that study, I discovered interesting things about p53's monitoring of a part of the cell called the nucleolus. The nucleolus is probably best described as a meta-factory: it makes the ribosomes, which are the parts of the cell machinery in turn responsible for making other proteins. In order for a cell to divide, it needs to make extra ribosomes so that it can make extra proteins in order to increase in size. And if that process is disrupted, its probably a good sign that something is seriously wrong, which will be picked up by p53 and most often the result should be apoptosis, rather than the cell continuing to make incorrect ribosomes and proteins or carry on dividing in a generally messed up situation. So I went off to do a post-doc working on another protein, c-Myc, which again does lots of different things but one of them is regulate this process of ribosome production. As a generalization, c-Myc is activated when the cell receives signals to grow, and if it starts getting activated even when the signals are absent, it may force inappropriate growth leading to tumour formation.

I was interested in finding drugs to prevent that process, but I didn't get very far during my 3 years in Sweden. So I'm still interested in it, and in the interaction between the p53 pathway and the c-Myc pathway, because they're in fact not entirely independent. Right now I don't really have any funding to work on this question, though. What I do have is funding to study a completely new protein, which just to be extra-secure I'm going to call X. X seems to be a master-switch protein that can control whether cells grow and divide, if more cells are needed, or whether cells die by apoptosis, if there are problems or fewer cells are needed. The cell signalling systems involving X probably also involve p53, I'm trying to find out more specifically.

Because it's a completely new protein, only recently discovered (by my collaborator as well as some other independent researchers) as something that seems to be important in cancer, almost anything we find out about it is going to be interesting. I want to know why it sometimes makes cells divide faster and makes them continue dividing even when you poison them and they ought to be committing apoptosis, yet sometimes X itself makes cells halt division and die by apoptosis even when you haven't otherwise stressed them. I want to know what other proteins X interacts with, what it signals to and what signals to it. I have a very faint thread of possibility that X itself may be involved in switching genes on or off, but if it doesn't do that directly it likely affects other proteins which will change gene expression. I have some friendly surgeons who are going to lend me the tumours they cut out of patients (the ones who give permission for us to use their removed tumours for research), so I'm going to find out if X is mutated or changes its activity in some way in cancer. And it might turn out that X is useful in genetic tests, perhaps for cancer diagnosis or for making decisions about which treatment is useful. Or else we might possibly be able to look for drugs which change X from the cancer-promoting state which makes cells divide and evade apoptosis, to the cancer-preventing state which makes cells halt division and die. But that's kind of a long term aim, we'll need a whole lot more information first.

I am going to have to keep the exact details fairly vague, but with that caveat I'm totally happy to answer any questions. These sorts of essays always end up being both too technical and over-simplified, especially when I just type them as the thoughts occur to me, without very much planning.

Comments

[forestofglory]

Thanks! I'd still like to hear more about how you feel your research is going. I know scientists aren't supposed to have feelings, but I want to know how you are doing as well as what you are doing. Feel free to decide that's too personal for this context.

I'm sure I've said this to you before, but I wish cancer research wasn't framed as being about cures, but as being about preventing people form dying and suffering form cancer. I think if we better understand how cancer starts (which is part of what your research is about) we will be able to better understand what type of environmental conditions, lifestyle factors, and other things contribute to cancer, and maybe even do something about them.

[liv]

Totally fair question, and you're right to pull me up on talking about facts instead of feelings. I am planning a follow-up post where I talk about being a supervisor, and that will have much more feelings in it. I'm still mulling how to do that without inappropriately discussing my student on the internet, though. I have no problem talking about my own feelings, that's not the issue.

I think you're absolutely right about cancer as a public health and environmental issue. That's not the sort of thing I'm particularly trained to do, but I entirely agree there should be more of it.

[forestofglory]

Well when you are writing for the whole internet it's harder to talk about feelings than if the two of us where chatting over tea. (I want to know things like how are you doing grant money wise? are you finding enough time to get research done? Are you finding the things you are doing right now interesting or tedious? But I'm not sure it's ok to ask in this context.)

I think you frequently fall into the trap of writing about cancer research as though finding a cure was the only goal for example in this post you say " And yes, of course I was somewhat motivated by a desire to help find a cure for such a terrible disease..." So maybe in the future when you frame your research you could use a different phase like "reduce people's suffering form cancer." (Very Pedantic here.)

[lethargic_man]

Ten years ago, you told me your Ph.D. supervisor believed strongly that cancer would be a solved problem before the end of his career, or at any rate within his lifespan. I said at the time I thought this grossly overoptimistic, but who was I to argue with a world expert.

How do you think the field is looking up against this goal now (and do you have any ideas how he feels about it now)?

[liv]

Haven't spoken to my former boss in a while, but from my own personal impression: it seems to me like the pendulum has swung back in the direction of cancer as multiple different diseases. We're in a phase of gathering amounts of data that would have been unimaginable, both practically and computationally, 10 years ago, and looking very much at tailored treatments.

The actual survival rates for people diagnosed with cancer have been improving at a measurable rate for the past 15-20 years, whereas before that that bottom line wasn't improving very much. But it feels to me like we're getting the low-hanging fruit, hormone-sensitive breast cancers, blood cancers with defined genetic defects, that kind of thing. It's not obvious that the positive trend is going to continue to include rarer and intractable cancers.

[ptc24]

Decision making: if you like etymological puns, you might talk about how cells etc. govern themselves. Or possibly not if (as seems likely) there's a relevant pre-existing technical meaning of the word.

[liv]

Heh, I like that. We usually tend to talk about regulating rather than governing, but more engineering puns sounds like a good plan to me.

[ptc24]

It's more than just an engineering pun. Let's borrow from wikipedia because of orthography: "The term cybernetics stems from Ancient Greek κυβερνήτης (kybernētēs), meaning "steersman, governor, pilot, or rudder" (the same root as government)."

[403]

That is so cool. If/when you publish, would you mind pointing me to the paper(s)?

[liv]

I will PM you; I'm not linking directly to the aspects of this that are published because that would break confidentiality on the aspects that are still underway. But you're certainly welcome to have a look.

[lilacsigil]

This is really fascinating! I donated my tumour for research and it's really interesting to hear what directions that research is taking.

[liv]

Yay, I'm glad you found it interesting. And I'm sure you've already been thanked for donating your tumour but add me to the list of people who are grateful.

[aphenine]

That was really clear and very interesting. Thank you.

[liv]

Thank you, I am glad you found it clear, with your investigations of physics teaching I know you've thought a lot about communicating technical stuff to a general audience.

[kerrypolka]

Thanks for this - I haven't trained my brain to stay focused on technical science text and I was worried I would switch off a paragraph in, but this was very clear and interesting (and exciting!). I particularly like the poetry of classical science etymologies - it makes the world feel a little more unified for some reason.

[liv]

Thank you, that's a really high compliment. If I can communicate my research to a professional journalist and make it sound exciting then I'm doing something right. And yes, I'm totally etymology girl, I'm always telling the students the origins of all the medical technical terms they have to use :-)

[simont]

The old fashioned word for the sort of thing I care about is cybernetics. These days cybernetics refers almost exclusively to computers, but in the late 20^th century, it was to do with regarding biological systems in terms of the flow of information. How does an organism, or an organ, or even a cell, receive a whole bunch of inputs from the environment and alter its behaviour appropriately?

Ooh, that's really interesting. I had a friend who studied cybernetics at university, and I don't think he got taught that aspect of it either – it was all robots for him (and he turned out not to like it much anyway).

I wanted to ask some kind of a clarifying question along the lines of "does cybernetics concern itself more with the mechanism of how behaviour 'decisions' are made, or with the policy of what decisions should be made in what circumstances to best achieve the strategic objective?", but then I thought, probably that's a false dichotomy and precisely the interesting part of the discipline is where the two collide – where the strategically best decision policies are computationally or mechanistically infeasible given the limited resources, and conversely the decision policies which can be implemented easily are no good, so what's needed (whether by demanding it of a designer or by rewarding it evolutionarily) is to push the boundaries of both just a little.

It's especially interesting that your involvement with cancer arises from a more fundamental interest in that sort of thing. Do you know if there are any completely different career paths that the same underlying interest might just as plausibly have taken you into? (Or is that an unanswerable question?)

[liv]

You really do ask the best questions, thank you! Mechanism or policy, I hadn't thought of it in that way. I think you're right, it's a matter of where it's possible to get to via evolution rather than what might the absolute optimum way of making these decisions. Some people have a view that there's a tension between the evolutionary interests of each individual cell (to proliferate as much as possible, one might suppose) and the evolutionary interests of the organism (not getting cancer is a good start). And a lot of what's going on with cell fate regulation is the result of a balance between those two evolutionary pressures. As well as what's actually chemically feasible, of course. Cells are fairly tightly constrained by the need to keep all the complex machinery going, let alone changing behaviour in response to stimuli.

Different career paths, I'm not sure. I suppose I might have found a similar sort of interest in economics or something like that, looking at a really complex system and how it responds to different inputs. If I'd had a little bit more exposure to compsci and a little less to traditional biology I could well have ended up in systems biology, which is much more directly about information. I am a very, very detail-focused person, very much a splitter, what sorts of careers do people go into when they love specifics and get bored with theories and generalizations?

[whereisirisnow]

This is so fascinating! Thanks for this lovely post about your work! You explained everything very well, so I now really have the feeling that I kind of know (in a very simplistic kind of way) what you are doing. :D