Breaking Science: Science of envy, sugar and kids...

Featuring: Audio Audio

The root of teeth development, cold genetics, the science of complex emotions and does sugar make kids hyperactive?

By: The OpenLearn team (Programme and web teams)

  • Duration 30 mins
  • Updated Sunday 15th February 2009
  • Introductory level
  • Posted under Radio, Breaking Science
Share on Google Plus Share on LinkedIn Share on Reddit View article Comments

The Naked Scientists explore the hottest breakthroughs from the world of research.

Plus in 'Stuff and Non-Science', does sugar make kids hyperactive?


Copyright BBC


Chris Smith: Coming up this week, the true roots of where our teeth come from.

Kat Arney: Where do teeth come from? Now, not just in terms of popping through the gums of a dribbling baby, but in an evolutionary context. And now research published in the journal PloS Biology shows that a common genetic control system is in charge of the development of all known teeth, from the first ever teeth in fish living half a billion years ago to our own pearly whites.

Chris Smith: Kat Arney, who’ll be nibbling through that morsel in just a moment.

Also on the way, a cure for the common cold. Well possibly, because scientists have successfully decoded the genomes of the most common culprits.

Stephen Liggett: If there were regions that were very, very common to all of these viruses then maybe that’s a place where you can target a drug, because that may be telling us that the virus refuses to mutate, or if it does it just dies in that region, so that would be a good place to go.

Chris Smith: Why there’s more to the common cold than we first thought. Stephen Liggett will be here to talk about the viruses responsible for untold annual misery later in the programme. Plus, in this week’s ‘Stuff and Non-Science’, this week’s story of whether sugar really causes hyperactivity.

Hello, I’m Chris Smith and this is Breaking Science, which is produced in association with the Open University.

First with news from across the scientific globe, including parasitic wasps and the brain basis of envy, here’s Kat Arney. And to kick off this week, Kat, it looks there might soon be an easy peasy way to spot prostate cancer.

Kat Arney: Yes. We know that rates of prostate cancer are going up rapidly in Europe and the US, but this isn’t necessarily because more people are getting the disease but it’s because we’re diagnosing it more often. In fact, a large percentage of men in their eighties actually have some kind of prostate cancer but usually that isn’t the thing that kills them. And many more men are having a PSA test, which is a blood test that can indicate that they might have cancer. But often these cancers are slow-growing and can actually be safely left and monitored and will never cause a man significant health problems.

Chris Smith: So the problem is how do you tell these slow-growing cancers from the rapidly-growing aggressive cancers that are going to need urgent treatment?

Kat Arney: Well the dilemma of how to tell these harmless pussy cats from aggressive and dangerous tigers is something that has puzzled researchers for several years, and it’s very important because treatment for prostate cancer can cause impotence and incontinence, so you really only want to be treating the people who have aggressive tumours.

Now researchers from the University of Michigan have discovered a panel of little chemicals which could lead to a simple test that could tell doctors whether a man’s prostate tumour is a tiger or a pussy cat.

Chris Smith: So what have they done?

Kat Arney: Writing in the journal Nature, the scientists led by Arul Chinnaiyan looked at over a thousand chemicals in 262 samples of tissue, blood and urine. Now these chemicals are metabolites, these are the products of the chemical reactions that keep us alive, provide us with energy and help our cells to function normally. So the team compared the metabolite profiles from tissue, blood and urine samples from men with non-cancerous prostates, early stage cancer, advanced prostate cancer and aggressive spreading cancer.

Chris Smith: And what did they find?

Kat Arney: Well the researchers found about ten of these metabolites that were present more often in the cancer samples and especially in the samples from people with advanced cancer. And in particular they found that a chemical called sarcosine was one of the best indicators of advanced prostate cancer. So they found high levels of sarcosine in around eight out of ten samples from aggressive cancers, and around four out of ten in early cancer samples, but they didn’t find it at all in the non-cancerous samples.

Chris Smith: Now you mentioned PSA, Prostate Specific Antigen, how does this compare with that, the usual test for prostate cancer?

Kat Arney: Well although this is only a small study, the initial results do suggest that it could be a better marker of advanced cancer than PSA, and also because you can pick it up in urine, this is quite a handy way of testing for it.

Chris Smith: So we might be able to use sarcosine as a test for detecting aggressive prostate cancer, but could we use it for anything else, such as developing new drugs for instance?

Kat Arney: Well the researchers did find that sarcosine’s involved in the biological pathways that control cancer invasion and cancer spread, so targeting the production of sarcosine might actually be a good way to target cancer.

Chris Smith: So where do we go from here, how soon are we going to see this used as a test for cancer?

Kat Arney: Well these are still early results, so we won’t see a test in clinical use in the immediate future, but once the findings are validated in larger samples of patients in clinical trials - that could take several years - then we could see it in use in clinics here. But this is just one chemical and really what we want to see is a whole panel of genes, molecules, chemicals that researchers and doctors can use to test individual men’s prostate cancer to give them a much better picture of what’s going on.

Chris Smith: Certainly encouraging news. Thank you for that, Kat. Now let’s look at another sting in the tail, this time from wasps. But these are a bit specialist, they’re parasite carrying wasps, tell us about these.

Kat Arney: Now if you’ve ever been on the wrong end of a wasp sting you will know how painful it is. But did you know that the paralysing chemicals that some parasitic wasps use to stun their hosts came from virus genes that the wasps picked up around 100 million years ago. Now this new finding answers the conundrum of where they came from.

Chris Smith: Now you say parasitic wasps, these are not the same as the ones that sting us in summer time.

Kat Arney: So this research by Annie Bézier and her colleagues in France and Switzerland was done using wasps called Braconid wasps. These prey on other insects laying their eggs in caterpillars and other larvae, and these wasps use paralysing proteins to immobilise the hosts providing a nice home for their growing wasp babies.

Chris Smith: But how do the viruses come into this?

Kat Arney: Well back in 1967, scientists noticed that these virus-like particles were present in the ovaries of female Braconid wasps and that these were injected into the host larvae when the eggs were laid. Now it’s thought that the virus particles help to suppress any immune response that’s in the host caterpillar which could cause it to reject the wasp grub. These virus particles were found to combine with DNA to make viruses known as poly-DNA viruses, but when these virus-like particles were found in many different types of wasp it posed a problem. Where were these viruses coming from and were they actually viruses at all?

Chris Smith: So what’s the answer?

Kat Arney: So Bézier and her team found that the genes encoding these virus-like particles were related to an ancient type of virus called Nudivirus, but the virus-like protein packages didn’t carry virus DNA as you might expect, but wasp DNA. So somehow the wasp DNA has got mixed up with the virus DNA and has replaced it, and now these virus-like proteins are used to transmit wasp DNA into the parasite’s host.

Chris Smith: It’s absolutely amazing, isn’t it? So how do you think that occurs?

Kat Arney: Well it does sound quite esoteric, but it does shed new light on the relationship between viruses and their hosts, as well as the relationship between these parasitic wasps and their hosts too.

Chris Smith: Well from stings in the tail to a bite from a very sharp tooth, tell us where teeth came from?

Kat Arney: It’s an important question, where do teeth come from? Now not just in terms of popping through the gums of a dribbling baby but in an evolutionary context. And now research published in the journal PloS Biology shows that a common genetic control system is in charge of the development of all known teeth, from the first ever teeth in fish living half a billion years ago, to our own pearly whites.

Chris Smith: It’s a fascinating finding, how did they discover that?

Kat Arney: This is research by Gareth Fraser and his colleagues from the US. And they studied tooth formation in a group of fish that undergo rapid evolution. These are fish known as cichlids that are found in Lake Malawi in Africa.

Now these fish have two sets of teeth, they have one set in their jaws in their mouths and they have another set back in their throat. Now these two sets of teeth are very different in evolutionary and developmental terms; teeth set back in the throat are a much older invention than teeth in your mouth. But the researchers were surprised to find that there was a link between the number of teeth in the mouth of these fish and in their throats, suggesting that there was some kind of genetic link between the two.

Chris Smith: But that could just be a coincidence couldn’t it, how do you prove that it’s down to the same genes?

Kat Arney: Well the researchers used a technique called in-situ hybridisation, which allows scientists to precisely reveal the genetic patterns of activity of specific genes. So using this they found a common set of genes controls the teeth in the cichlids’ mouths and also in their throats. These genes are hox genes which are involved in patterning many of the body’s structures. For example, some of our own hox genes give us the regular pattern of vertebrae in our spines and give us five fingers and toes. Now the scientists found that a precise pattern of hox gene activity, along with other related genes controlled by the hox genes, was needed for developing both teeth in the throat and teeth in the mouth. So it’s likely that the earliest fishes also used hox genes to generate teeth in their throats, and later on in evolution this same genetic pathway got co-opted and hijacked with a few tweaks to create teeth in the mouth.

So the researchers basically think that every tooth made throughout evolution probably uses this core set of genes. And not only that, they think that probably similar pathways are at work in the kind of pattern structures we see in other animals like hair and feathers too.

Chris Smith: Now Kat, just to finish off, we don’t expect you to be of the jealous kind or the jealous persuasion, but scientists have discovered what goes on in the brain when we get envious of someone.

Kat Arney: Yes. Life can seem very unfair sometimes. You put on your new jeans, then you get splashed by a bus driving through a puddle. But how sweet it is to see someone else get splashed. Now that’s an emotion known as schadenfreude or taking pleasure from someone else’s pain or discomfort. Envy is also another common social emotion, but now scientists in Japan have used a technique called functional MRI scanning, this is a technique that highlights the areas of activity in the brain, to map the areas involved in these complex emotions.

Chris Smith: So how did they do it and what did they find?

Kat Arney: Well, the researchers used FMRI to peer inside the brains of 19 healthy volunteers and looked at how they responded to schadenfreude or envy in various situations, and they found that envy was reflected by activity in a region of the brain called the dorsal anterior singulated cortex, that’s the same region that’s involved in our response to physical pain, but when it came to schadenfreude the team found that this actually activated a region called the ventral striatum, which is involved in reward, and the feelings of reward were more intense when bad things happened to someone else that the volunteer envied.

Chris Smith: So what does this actually mean, why have we evolved a system that enables us to do this?

Kat Arney: Well it does seem in this case that someone else’s pain is interpreted as gain, but when you’re envious someone else’s gain is your pain. And it does show that social emotions are treated more like physical experiences by the brain than we previously thought. It may also be that while we experience lack of food, drink or shelter as physical pains, there may be some social benefit in interpreting complex emotions as pain and it’s probably evolved as a result of humans living in complex social communities as we do.

Chris Smith: So when your neighbour buys a better car than yours, the pain is real. But unlike a headache, it can only be cured by a trip to the dealership. At least that’s what I’ll be telling my wife in future. Thank you, Kat. That was Kat Arney from the Naked Scientists with a look at some of this week’s top science news stories. And incidentally, if you’d like to follow up on any of those items, the full stories and the references are all on the Open University’s website at

In just a moment, how scientists are dealing with the tricky problem of making robots that can cope with walking over sand. Apparently it’s not as easy as a crab makes it look, but therein lies the solution. That’s on the way. But first to a family that you might not have heard of but you almost certainly will be very familiar with; the rhinoviruses, also known as the cause of the common cold.

Whilst most colds are just an annoyance to many people, for some they can have life-threatening consequences, and that includes provoking asthma attacks. Ideally we’d like to find a vaccine or a drug to prevent these pests but until now no-one’s succeeded. But that could be about to change because US scientist Stephen Liggett has applied to the problem the power of modern genetics to find out what makes these viruses tick.

Stephen Liggett: So we wanted to understand more about the diversity of the rhinovirus. I think I’ve got an analogy or two that might help with this idea. If someone told you you had a piece of fruit, that’s not very descriptive but we would all recognise that it’s a piece of fruit but there are multiple types of fruit; oranges, apples, pears, etc. But even then, if you were told that you had an apple, there are actually multiple types of apple and they have different appearances, different tastes and different properties, so you can see how knowing more about that apple, or in this case the virus, knowing more about that virus, especially its sequence, would tell us more about its properties.

Chris Smith: Why is it that when we get a cold, because we’ll assume that most colds are down to one member of this family of viruses, the rhinoviruses, why is it we get them again and again and again?

Stephen Liggett: Well there are several possibilities. One is that the immune response that the body has is probably mostly generated from the mucus membranes in the nose, which is where a lot of the major symptoms are, and that type of immunity doesn’t hold very well. The second possibility is that the virus itself may mutate as it infects people in a large population and by the time it gets back around to you again it has mutated to the point that whatever immunity you may have built up is no longer capable of fighting it and you get it again.

Chris Smith: So when you did this study, were you going out and getting samples of virus from the population at large in order to screen through those viruses and see what their genetic signature was?

Stephen Liggett: Actually that was a smaller part of what we’re doing, but it’s certainly going to be a big follow-up. There’s an institute in the United States that has frozen away the 99 so-called reference human rhinoviruses out of frozen capsules, so we thought it would be best to go to that repository and sequence all of the viruses that they have. And then we did get ten isolates from patients as well just to sort of begin to augment that large load of viruses that we wanted to sequence there.

Chris Smith: So I guess what you’re doing is producing a sort of roadmap using the known strains, the identified 99 strains of rhinoviruses, and then by going to the community, you can then begin to ask well do these actually fit into this set of categories that we’ve drawn up arbitrarily, 99 different types of rhinoviruses, and what’s the sort of mutation rate in the general population.

Stephen Liggett: That’s correct, and in fact that roadmap, I would call it better a family tree. And so in a family tree which deals with viruses there can be multiple branches, first of all we need to find out are there some major branches that we don’t even know about. It will be important to bring in that next group of viruses in the wild, so to speak, which we won’t have any trouble collecting, as I’m sure you can imagine.

Chris Smith: When you screened across this 99 viruses and started to assemble the family tree, what patterns emerged, what did you see?

Stephen Liggett: We first saw that there were three major families and probably a fourth major family, and they’re not very imaginatively called HRV A, B, C and our new family, D. Within the Bs, which have been known for quite some time, there were probably four sub-families within B, and then within the As there were probably ten sub-families, and these sub-families were different enough from each other that maybe that’s why we have not been able to come up with a good drug to treat the common cold in terms of killing the virus. And so I think we now understand the molecular basis for our failure to be able to have a one-drug fits all rhinovirus infection situations.

Chris Smith: But given that we have failed so far, but you’ve now got these new insights which people didn’t have when they were trying to conquer the common cold that way, are there any insights that have emerged from what you’ve done that may give us a cure for the common cold?

Stephen Liggett: Yes, there are. If there are regions that were very, very common to all of these viruses then maybe that’s a place where you can target a drug because that may be telling us that the virus refuses to mutate or if it does it just dies in that region, so that would be a good place to go. And then the second alternative is you can find places that are fairly flexible in their mutations and so you would certainly avoid those but then there would still be other places where there might be some commonalities within these small groups.

Chris Smith: So potentially some common targets in the common cold. That was Stephen Liggett of the University of Maryland School of Medicine who’s published that work in this week’s edition of the journal Science.

Well from catching bugs to working out how they crawl now. Here’s Dan Goldman.

Dan Goldman: The problems that we’re trying to solve are really to understand how organisms like crabs and lizards can move across sand seemingly so effortlessly. To that end we’re using models of the organisms, or you could call them robots, to really explore how limbs interact with complicated ground like sand for which we don’t have the physics equations yet.

Chris Smith: Why sand, Dan?

Dan Goldman: Well sand is sort of an interesting material. One, it’s of practical importance in biological systems, lots of organisms live in beaches or deserts, and in fact lots of robots ultimately may have to now and may have to go in materials like sand, and I should say other materials, unconsolidated materials like rubble and debris which have interesting features such that when limbs hit those materials, the material can behave a solid, it can sort of hold the limb for a little bit, and then as a fluid it can flow and slip as the limb pushes too hard or too rapidly.

Chris Smith: And so what you’re saying is that organisms by their evolution have come up with strategies that mean that they can traverse these surfaces and they don’t have to think about doing it, it just happens.

Dan Goldman: Yes, we don’t speculate on what the organisms are thinking, at this stage, but certainly organisms like Zebra Tailed Lizards out in the desert south west or Ghost Crabs - in the desert south west of the United States, I should say, or Ghost Crabs on the eastern coast of the US are quite nimble over a variety of granular surfaces. And if we look to kind of human-made devices on that scale, small robots, we don’t tend to have devices which have that mobility.

Chris Smith: Why do they tend to fail, the human-made devices?

Dan Goldman: Well we’re not sure, completely, why they tend to fail, and one of the problems is that we don’t yet have fundamental understanding of how feet interact with the ground. So if this was the analogous problems, if I want to understand why a plane flies or how a propeller works or a rotor works in air and water, I can, in principle, write down the equations of fluid flow and then try to design my device to optimise some sort of lift or thrust or drag, and we can’t yet do that for limbs of robots, much less for limbs of organisms to try to understand why they fail.

So the approach we’re taking is to use a robot, there’s a robot called Sandbot, which is the smaller cousin of a robot called Rex, who is a six-legged robot which is modelled after a cockroach, and on hard ground it bounces around on the ground at up to speeds of a couple of metres per second but when it moves onto sand Rex and, what we subsequently found its cousin, Sandbot, tend to have more trouble.

Chris Smith: And what bogs them down?

Dan Goldman: What we’re finding is that when the robot moves into the sand, if we don’t tune how it moves its limbs just right it tends to sort of dig holes in the material and embed itself deep into the material and ultimately get stuck in a kind of slow swimming mode. But no-one had yet systematically investigated performance on granular materials, on things like sand.

And our collaborators had observed over ten years that if they changed how it moved its limbs a little bit, and I’ll explain that, that they could get some increased performance. And when we did so the robot was suddenly able to go at 30cm/second across the ground.

Chris Smith: What was the key difference that you made to make that dramatic improvement in performance?

Dan Goldman: In fact though, in this robot, three limbs fired at a time and then the other three limbs fired at a time, it’s called an alternating tripod gait inspired by cockroaches, and what the engineers put into this was a sort of biologically inspired limb motion, such that the limbs could have a slow phase when they were on the ground and a more rapid phase as they whipped around for the next cycle. And what we found was that if we made the onset of the slow phase right when the limbs contacted the ground that seemed to have the best effect. Thus the limbs had to come in sort of gently into the material. But it seems that with a study we’ve done, a comparative study of lizards and crabs and some cockroaches and even some hatchling sea turtles that in fact these animals seem to sort of do this very thing, they kind of bring their limbs in, in a kind of gentle way to keep the material as solid as it can be for as long as it can be, and they’re effectively pushing off the solid properties of the material. They’re letting the material solidify a little bit.

Chris Smith: It’s intriguing to think that you started this inspired by biology but then when you start doing the experiments you actually begin to get insights into how the real world is working around you in nature.

Dan Goldman: That’s exactly right. It’s a model and, you know, as a model it’s not a perfect copy of an organism, we think it embodies the principles of an organism, and one principle we seem to be finding at least in the robot is that solidification helps when moving on this kind of stuff.

Chris Smith: Dan Goldman from the School of Physics at the Georgia Institute of Technology, he’s helping robots to find their feet in sand and in turn work out how animals do it in the natural world. He’s published that work in this week’s edition of the journal PNAS.

You’re listening to Breaking Science with me, Chris Smith, and it’s time now for this week’s ‘Stuff and Non-Science’, where we massacre myths. And going for a glucose fix, here’s Diana O’Carroll.

Diana O’Carroll: For this week’s non-science, I’ve been stuffing some sugar. They say that too much sugar can make the youthful of us hyperactive. Maybe not. Here’s David Benton.

David Benton: I’m not sure where this idea comes from. I suspect it’s little more than people assume that when you consume sugar that the level of your blood glucose rises rapidly, that’s a source of energy and given you’ve got energy you become more active. But if you do consume sugar there are two potential mechanisms. One is a biological mechanism, that is your body reacts adversely to sugar in some way, or alternatively there’s the psychological mechanism which is that you believe that is true and it becomes a self-fulfilling prophecy.

So that does mean that we really need a mechanism of teasing the biology from the psychology. And the way that this is done are the so-called double blind trials, and the results are absolutely unequivocal, there is no effect of consuming sugar on the behaviour of children, or at least not groups of children just chosen because they have the diagnosis of ADHD, attention deficit hyperactivity disorder.

On the other hand, if you take some mothers and then you tell them, “Your child is drinking sugar,” in fact they’re not drinking sugar, it’s a lie, on occasions when they think their child’s drinking sugar their behaviour to the child changes and induces changes of behaviour in the child. So the risk really is that there are parents out there that are teaching their child to respond to the food rather than the food is causing a problem.

Diana O’Carroll: So sugar won’t make you charge about, demanding to play British Bulldogs, and that was David Benton, who’s Professor of Psychology at Swansea University.

We’ll have another bit of ‘Stuff and Non-Science’ next week, and you can suggest your own by emailing me,

Chris Smith: And I hear that it’s better to eat a potato if you want some serious energy. Thank you, Diana. That was Diana O’Carroll with this week’s ‘Stuff and Non-Science’. That’s it for this time. We’ll be back next week with another round-up of global science news.

Breaking Science is produced in association with the Open University and you can follow up on any of the items we’ve discussed in the programme via the OU’s website. That’s at Alternatively, you can follow the links from the Five Live Up All Night web pages.

The production this week was by Diana O’Carroll from and I’m Chris Smith. Until next time, goodbye.


Does it sound good? Why not get the Breaking Science podcast, and receive new episodes as they're released?


These are the sources used by the Naked Scientists in making the show:

In the news

'Metabolomic profiles delineate potential role for sarcosine in prostate cancer progression'
by Arun Sreekumar et al.
in Nature, Vol. 457, No. 7231

'When Your Gain Is My Pain and Your Pain Is My Gain: Neural Correlates of Envy and Schadenfreude'
by Takahashi et al
in Science (2009) vol 323, pp937-939

'Polydnaviruses of Braconid Wasps Derive from an Ancestral Nudivirus'
by Bezier et al
in Science (2009) vol 323, pp 926-930

'An ancient gene network is co-opted for teeth on old and new jaws'
by Fraser GJ, et al. (2009)
in PLoS Biol 7(2): e1000031.


Stephen Liggett for 'Sequencing and Analyses of All Known Human Rhinovirus Genomes Reveals Structure and Evolution', by Ann C. Palmenberg et al in Science

Dan Goldman for 'Sensitive dependence of the motion of a legged robot on granular media', by Chen Li, Paul B. Umbanhowar, Haldun Komsuoglu, Daniel E. Koditschek, and Daniel I. Goldman in PNAS

David Benton for 'Stuff and Non-Science'

More like this