Nature Podcast

This is a transcript of the 1st May edition of the weekly Nature Podcast. Audio files for the current show and archive episodes can be accessed from the Nature Podcast index page (, which also contains details on how to subscribe to the Nature Podcast for FREE, and has troubleshooting top-tips. Send us your feedback to


The Nature Podcast, brought to you by An online community devoted to Mini PROTEAN System Users. Explore, Share, Connect and Win at

End Advertisement

Adam Rutherford: Coming up, we reconstruct an animal's internal compass in the lab.

Peter Hore: And this is exciting, because if we can do this in the laboratory, then maybe birds can also do it and use it to detect the earth's magnetic field as an aid to navigation.

Kerri Smith: And after 3 decades, a fabled element of electronics is found.

Stanley Williams: A memristor is essentially a resistor with memory.

Kerri Smith: This is the Nature Podcast. I'm Kerri Smith

Adam Rutherford: And I'm Adam Rutherford. First this week, the eyes have it. The primary function of eyes is seeing, with photoreceptors, that is rods and cones, being the light-sensitive cells that pick up photons, but in recent years, it has emerged that there are photosensitive cells in the retina that have functions unrelated to vision. And in this week's Nature, a team led by Samer Hattar from Johns Hopkins University in Baltimore, has shown that without these cells, mice can't control their circadian rhythms. Samer explained the new results to me. Nature advance online publication 23 April 2008

Samer Hattar: So our new data in a nutshell is that the ability to form images, the ability to use light for complex function as seeing the world around you is completely separable from the ability to detect to light for a simple function of adjusting your activity sleep cycle to the outside solar day. We have discovered that there is a subset of retinal ganglion cells usually the output neuron of the retina that relay the information from rods and cones to the brain, that there is a subset of these ganglion cells, themselves photoreceptors and express a new opsin light molecule, which was named melanopsin.

Adam Rutherford: And these ganglion cells that contain melanopsin, which is a phototransduction molecule, what are they actually doing in the eye?

Samer Hattar: That's a great question. Before the recent paper in Nature, we expected that if you destroy the protein, melanopsin that you may have major defects in the ability of the animal to adjust its biological clock to the outside solar day. But when the melanopsin protein was knocked out, there were some defects, but they were quite minor actually and so we suspected that rods and cones are able to compensate to the absence of the melanopsin protein, but we did not know if they compensate through these unique set of ganglion cells or through other ganglion cells in the retina.

Adam Rutherford: So the function of this cell and the melanopsin in this cell is to help control circadian rhythms and also pupillary reflex, which is how the pupil dilates in response to sensitivity of light. And in the new paper, you've knocked out these cells. Can you tell us how you did that?

Samer Hattar: So we decided to see what happens if we completely kill melanopsin cells instead of just destroying that protein and to do that we simply expressed a very specific toxin that inhibit protein translation, specifically in melanopsin cells using the specific melanopsin promoter. So this way we were able to only kill melanopsin cells, keep all ganglion cells intact in the retina.

Adam Rutherford: We're talking about mice here, so in these mice, their ability to retain normal vision is totally unaffected by knocking out these cells right?

Samer Hattar: You are absolutely right. For the most part, the ability to form images and pattern vision is completely intact in these animals.

Adam Rutherford: But at the same time, they display poor circadian rhythms and their pupillary response is affected.

Samer Hattar: Right, so at the same time and that's why we were very interested in these results despite the fact that they have normal visual functions, they don't have the ability to adjust their circadian clock to the outside solar day. Their circadian clock is completely functional because the clock in endogenously driven. Their pupil responses are highly affected, so that showed us that there is a separation between image formation and non-image-forming functions at the level of the retina. And more interesting for us, even if you can form images in your brain this cannot feed back on the clock centres or the pupil centres in the brain to control their responses to light.

Adam Rutherford: And so what sort of behaviour do you actually see in these mice when they can't control their circadian rhythms?

Samer Hattar: They completely act as if they are in constant dark conditions, so they run with their endogenous clock, with their endogenous circadian period, which is less than 24 hours in mice.

Adam Rutherford: Okay, so I know you are in Baltimore and it is 6 o'clock in the morning there. We're in London. Do these results have any impact on human conditions, such as troubled sleep?

Samer Hattar: That's a great question and we really think that this will have a major impact on sleep studies because if you have a problem in your melanopsin system, even though you may have a normal image function, you may not be able to adjust to a new time zone when you travel across time zones. You may not be able to adjust your daily rhythms to the outside solar day as well as it should, so we really believe that melanopsin system defects can cause may be some sleep problem and we also think may be affected in seasonal affective depression people and in fact, you know, when you talk to human researchers about these results, they get very excited about the prospect that may be newer, more easier tests should be developed to measure these light-dependent non-image dependent functions in humans.

Adam Rutherford: Samer Hattar of Johns Hopkins University in Baltimore. And for more neuroscience, tune into the new series of NeuroPod. In the current issue, we have teenage brains, stressed brains, and gambling brains all presented by our very own Kerri. Find it at, with all the rest of our shows.

Kerri Smith: And we stick with the retina in our next report as Charlotte Stoddart finds out about the protein behind animal magnetism.

Charlotte Stoddart: Migrating birds use the sun, stars and sometimes even the moon to find their way, but scientists think that these birds and many other animals including mammals, reptiles, fish, and insects are also able to orientate themselves using the earth's magnetic field. This internal compass seems to work through a protein in the retina of the eye. The idea is that the magnetism modulates a light-sensitive chemical reaction there. But the earth's magnetic field is very weak, just 50 micro-Teslas, so can such a tiny force really drive a chemical compass? Researchers at the University of Oxford believe it can. They've tested a model system in their lab and demonstrated that it is indeed possible to use a chemical reaction to detect the direction of a very weak magnetic field. Here's lead author Peter Hore. Nature advance online publication (30 April 2008)

Peter Hore: It has been known since 1960s that amongst various other mechanisms birds use the earth's magnetic field as an aid to navigation, in particular migratory birds. The more popular mechanism by which people think they may do this is based on crystals of magnetite – magnetic iron oxide, but it was suggested in the late 1970s that perhaps there could be a chemical mechanism. This idea was largely ignored until 2000 when Thorsten Ritz revived it and suggested that a molecule called cryptochrome might be responsible in the bird's retina for detection of the earth's magnetic field using a chemical mechanism.

Charlotte Stoddart: The earth's magnetic field is very weak, so it is pretty incredible to think that birds could use this to navigate and in your lab you have used a molecule that's similar to cryptochrome to test this theory. So tell me some more about your model system.

Peter Hore: We chose this particular molecule, because it was known to have the sort of properties that we thought might make it suitable for detecting weak magnetic fields. Its chemistry is that when excited with light, it forms long-lived radical pair states and these radical pair states responds to weak magnetic fields because they are able to undergo different chemical reactions depending on the direction of the magnetic field.

Charlotte Stoddart: And when you tested this molecule in magnetic fields comparable to that of the earths', what did you find?

Peter Hore: We found that the lifetime of the transient radical pair state changed when we applied a magnetic field of similar size or in fact weaker than the earth's magnetic field.

Charlotte Stoddart: And this is a sort of proof of principle; you've demonstrated that a photochemical reaction can act as a magnetic compass. What's the next step? What does this tell us about how birds navigate?

Peter Hore: Right, the next step is that we'll want to do similar sorts of experiments on cryptochromes, which are the proteins that have been suggested as the possible magneto receptors in the bird's retina. If we could establish that cryptochromes responded to magnetic fields in the same sort of way as this model compound, then that I think would be a major step forward.

Charlotte Stoddart: This cryptochrome you are talking about, this is a kind of retinal receptor that's also involved in circadian rhythms, which is interesting that it might be involved in that and also this magnetic compass.

Peter Hore: That's right. Cryptochromes have a number of functions. In plants for example, they regulate growth and recently Margaret Ahmad in Paris has found effects of weak magnetic fields on plant growth, which she attributes to field sensitivity of a cryptochrome. So that provides a bit more evidence that these cryptochromes may be used in magneto-reception by animals.

Kerri Smith: Peter Hore talking to Charlotte. You are listening to the Nature podcast.


Adam Rutherford: Coming up in just a moment, the missing electronic gizmo, the memristor, is found. But first the Podium. Nature is running a special package of features on science education this week from, why we should teach molecular biology in European classrooms, to teaching artists the science of colour. Martin Rees the Astronomer Royal and Professor of Cosmology and Astrophysics at the University of Cambridge, weighs it on why we need to get more students to stick with science.

Martin Rees: Like me, many scientists probably owe their initial impetus to an inspiring teacher, but today many pupils never encounter an expert and enthusiastic science teacher at all. Too few new specialist teachers are joining the profession, especially in physics and maths to replace those retiring. Further measures are needed. Biology teachers could be given extra expertise in physics, mature professionals should be encouraged to move into teaching from research, industry or the armed forces and we must encourage scientists in universities to spend time in schools and vice versa. Very young children have a natural interest in science whether focussed on space, dinosaurs or tadpoles, but we are bad at converting youthful enthusiasms into sustained engagement for science in their 11 to 16-age range. And in England's unduly specialized education system, those who are turned off science by 16 can drop it fore closing most options of ever studying at university. No country can achieve success in technology, unless science attracts a good share of the talents at university entrance. But as it also matters, what careers these people end up in. In my Cambridge College, I asked a group of final year engineering students what their career plans were. Only one was going to be an engineer, the rest were heading for the city. My message to them was that's fine as long as they don't call it the real world. Manipulating financial derivatives is further from anything that matters than being a teacher, a scientist or an engineer. This flight of talent from academia and manufacturing should concern us. At a standing in the Saturday premiere league and our economy would be threatened if staff quality isn't sustained. Anecdotal evidence suggests that many large manufacturers don't offer high fliers enough scope early in their career to make a mark, to show initiative and to achieve a financial premium in the way that the financial sector does. To stem the internal brain drain from academia, pay should surely keep pace with the public services at least, and we should resist the erosion of the distinctive features of academic life that compensate us for modest pay. Relative autonomy and the prospect without undue hassle of gaining basic funding for the research one chooses to do. That's what's available at the best universities in the UK and of course in the US as well. Top universities are highly valued assets because of the expertise of their faculty and the consequent quality of the graduates; they feed into all walks of life. The offer optimum opportunities of curiosity driven research and they excel at it Moreover, each university is embedded in a cluster of research laboratories, small companies and NGOs to symbiotic benefit. Indeed a recent report showed that the scale and success of such clusters in the UK is steeply correlated with the research strength of the embedded university. In the clusters of great universities attract around them thousands attract talents and big companies too. Success breeds success and failure is accepted as a step toward later success. A dynamic and interactive community developed that offers in a word to the financial times a low risk place to do high-risk things. The most effective knowledge transfer is rather the movement of people. We don't know what would be the 21st century counterparts of the electron, quantum theory, the double helix and the computer. Nowhere the great innovators of the future will get their formative training and inspiration, but as a Brit, I hope that the 21st century would be influenced by the creative ideas that germinate in these small islands.

Kerri Smith: Martin Rees there and there are more science education features at If you've got any comments for us, please do get in touch by e-mailing Finally this week, a new report is set to change the way we teach people about electronics. Mike Hopkin reports.

Michael Hopkin: Anyone who did high school physics knows that electronic circuits are made up of building blocks called resistors, inductors, and capacitors, but what they didn't teach you at school is that there is a fourth fundamental element called a memory resistor or memristor for short. It was first described in theory in 1971, but only now has it become reality. Thanks to the nanoscale efforts of electronics engineers at Hewlett-Packard's California lab. Now they are set to transform electronics promising computers that can remember what you were doing before you switch them off. It all sounds pretty important, doesn't it, which is why I asked lead researcher, Stan Williams why we've never heard of memristors before and for that matter, what they actually do? Nature 453, 80–83 (1 May 2008)

Stanley Williams: A memristor is essentially a resistor with memory. What that means is that the actual resistance of the memristor changes depending upon the amount of voltage and the time for which that voltage has been applied to the device. Even though it was first predicted to exist in 1971, there hasn't been an example of one until we built it in the lab and showed that it worked as predicted. Memristance and a memristor are something that emerges at very small length scales. Memristors actually get better as they get smaller and it wasn't until we were really working in the nanoscale regime that memristance really became important enough to be noticeable.

Michael Hopkin: It has taken the team about 6 years to make a working memristor, since first stumbling across the 1971 work by Leon Chua, the electronics expert, who first described the devices in principle and it turns out that to see how memristors work in practice you have to make them very small indeed as Chua explains.

Leon Chua: We are talking about technology that is made up only a few layers of atoms together. We're talking about a 5-nanometer kind of film for the purpose of the audience; a nanometre is best, sort of, visualized as the size of a sugar molecule. So we are talking about 5 sugar molecules, that kind of thickness and when you get to that level, and when you've technology that can control it to that level, which is what we have now, this effect becomes dominant, as we were shown by Dr. Williams and I was very excited because I never thought that's possible, never thought that I would live to see this happen.

Michael Hopkin: Stan Williams reckons that another reason why the idea lay dormant for 37 years was because the maths involved is pretty complicated stuff.

Stanley Williams: The original prediction in the papers in which the prediction appeared was very heavy mathematically, and so it required a very significant investment in order to read those papers and really understand what the mathematics was saying.

Michael Hopkin: Leon Chua disagrees. He reckons the idea took a while to find its feet because it is so weird as to be almost theoretical.

Leon Chua: It's not really that difficult, it's more being sort of theoretic, you know, nobody would believe that this is, you know, the case and so it's unnatural in some sense.

Michael Hopkin: Now that they are here, memristors look set to revolutionize computing as Williams explains.

Stanley Williams: We see a huge number of possible uses for memristors; one is in the area of non-volatile random access memory. Because these memory resistors have this property of remembering their past, they're ideal for storing information, so you can think of essentially putting a voltage in and switching a memristor from a state that you define as a 0 to a state you define as a 1 and then if you put the opposite voltage on the device, you can switch it back. The big advantage that we would see there would be that you wouldn't ever have to boot your computer up again, because the memory would remember were it was the last time you were when you turned computer off, the next time you turn your computer on, you'd immediately start up exactly where you left off.

Michael Hopkin: And that's not all. Memristors might even be able to do things that human brain can do that today's computers can't. So one day, you might have a conventional computer to do your number crunching and a memristic one for tasks like phase recognition or super sophisticated web searching and if even that's not enough, Chua predicts that memristors might even be able to help you out with that old bugbear, the battery life of your mobile phone.

Leon Chua: The memristor is unusual in the sense that if you would pull the plug it actually doesn't go dead, it actually would remember just prior to the plug what it was in and so that makes it a wonderful sort of device that would allow you to memory that is non-volatile in the sense that you pull the plug and you don't have to power it. So some day I would imagine that you don't have to charge your cell phones or your laptops every so often, you know, you would see a new power of course but you would tremendously lengthen the life of your batteries when it comes to remembering important data.

Kerri Smith: That's all for this week's show. Join us next week when we will be featuring the genome of possibly the most ridiculous animal on the planet.

Adam Rutherford: Our sound of science this week is a campaign song from a UK radio station that broadcast near the Jodrell Bank Observatory. The telescope on site is threatened by funding cuts and a group of students penned this ditty to petition the government to save the scope. I'm Adam Rutherford.

Kerri Smith: And I'm Kerri Smith, that's all for now.

[Sound of Science Plays]


The Nature Podcast is sponsored by Bio-Rad, at the centre of scientific discovery for over 50 years, on the web at

End Advertisement