NOVA scienceNOW

PBS Airdate: July 26, 2005
Go to the companion Web site


It wasn't easy. In fact it took two—count them, two—mathematician brothers from Brooklyn, plus their supercomputer running 7.7 quadrillion calculations to solve a problem that looked like it was going to be a piece of cake.

Tonight, David and Gregory Chudnovsky solve "the mystery of the breathing unicorn."

ERIC LANDER (The Broad Institute of MIT and Harvard): The petunia was a big puzzle.

ROBERT KRULWICH: Geneticists took an ordinary purple petunia and wanted to make it purpler.

But when they tried, they got...

RICHARD JORGENSEN (University of Arizona): ...the complete opposite of what we had expected.

ROBERT KRULWICH: purple. Why? The answer to this riddle is so astounding it may lead to treatments for major diseases.

You've just listed cancer and HIV. These are famous, big fat diseases.

GREGORY HANNON (Cold Spring Harbor Laboratory): ...arthritis.

ROBERT KRULWICH: Well, stop listing them and...

Are all these diseases affected by this one discovery?

GREGORY HANNON: Certainly they are.

ROBERT KRULWICH: The dream, of course, is to create a car with an engine so clean, you could drink the exhaust.

TOM MAGLIOZZI : Lick it. Lick it! He's dead.

ROBERT KRULWICH: Because the exhaust...


ROBERT KRULWICH: But how close are we, really, to an affordable hydrogen fuel-cell car?

RAY MAGLIOZZI: I yearn for you, Robert.

ROBERT KRULWICH: You yearn for me, but...

Well, after a short and picturesque chemistry lesson from public radio's car guys, you decide.

Better science makes better innovation a possibility. Better innovation makes better communication a reality—Sprint, making communication better. Sprint is a continuing proud sponsor of NOVA.

Major funding for NOVA ScienceNOW is provided by the National Science Foundation, America's investment in the future; The Howard Hughes Medical Institute, serving society through biomedical research and science education—HHMI.

Additional funding is provided by the Alfred P. Sloan Foundation to enhance public understanding of science and technology, the George D. Smith Fund, and the Kavli Foundation, advancing scientific knowledge at leading universities worldwide.

Major funding for NOVA is also provided by the Corporation for Public Broadcasting, and by PBS viewers like you. Thank you.


ROBERT KRULWICH: Hi, I'm Robert Krulwich, and welcome to NOVA ScienceNOW, where we consider not one, but several science stories. Tonight, they're basically puzzles beginning with a problem, so just...

Come on back. Come on back and...all right, stop. Good. The internal combustion engine, which fouls the air and uses gas which comes from oil—which is getting expensive, involves the Middle East, gets us into all kinds of fights—who wouldn't want to replace this with a more efficient and affordable alternative?

But is there an alternative?

Well, there is this engine we keep hearing about which is supposed to be fabulous. It's coming "soon." But the puzzle is, how soon?

Every year, Detroit unveils, with much to do, a "Car of the Future." And the hoopla here isn't about what this car does. It's about what this car doesn't do. This car doesn't use gasoline, none, because it is powered by a fuel cell.

GEORGE W. BUSH (President of the United States, 2001- ): Fuel cells...

NEWS ANCHOR : ...fuel cells.

NEWS CORRESPONDENT: ...powered by a fuel cell.

GEORGE W. BUSH: Fuel cells are the wave of the future.

NEWS CORRESPONDENT:The wave of the future.

ROBERT KRULWICH: Really? This is the future? Well maybe we should take a closer look.

So I decided to test drive a fuel-cell car, and I invited a couple of friends to come along.

What do we think?

TOM MAGLIOZZI: We don't know anything.

ROBERT KRULWICH: Tom and Ray Magliozzi, the Car Talk guys from National Public Radio.

RAY MAGLIOZZI: What was that whistling noise?

ROBERT KRULWICH: Yeah, what is that whistling?

RAY MAGLIOZZI: What's that about?


TOM MAGLIOZZI: That sounds like F above middle C. I would like to open the hood just to see what's there. Don't shut it off.

RAY MAGLIOZZI: And don't lock the keys in it. I'll leave a window open. I'm dying to see what's under here.

TOM MAGLIOZZI: What the heck is this? See, I knew it. It's not an engine!

ROBERT KRULWICH: Well, it has a box, and it has lots of things connecting to it.

RAY MAGLIOZZI: Yeah, my intuition tells me this is an electric motor.

TOM MAGLIOZZI: This is basically an electric car.

RAY MAGLIOZZI: You know what it sounds like? It sounds like my Norge refrigerator used to sound.

ROBERT KRULWICH: Tom and Ray know a lot about regular cars, the ones with internal combustion engines.

This, on the other hand, is an electric car. But you don't plug it in, instead, it's powered by this mysterious fuel cell thing.

RAY MAGLIOZZI: This is stuff, this is technology that none of us understands.

TOM MAGLIOZZI: No one told us this.

RAY MAGLIOZZI: Do we need help? We need help.

ROBERT KRULWICH: Maybe we should ask...

RAY MAGLIOZZI: We need someone, a passerby, just some ordinary citizen.



ROBERT KRULWICH: Oh, my god. Thank goodness the car comes with its own expert, Dan Kelly, who works for a company that actually makes fuel cells.

RAY MAGLIOZZI: So it has an electric motor. What runs the motor?

DAN KELLY: All the electricity comes from the fuel cell. And the fuel cell is really made up of a collection of these.

ROBERT KRULWICH: This is a fuel cell? It looks like a black plastic license plate.

DAN KELLY: This is one cell. And it's like a sandwich; you just stack them up. You want more power, add more cells. Make a bigger sandwich. And what goes through, if you look at it...

RAY MAGLIOZZI: It's a piece of plastic!

DAN KELLY: ...and the grooves that are in it...

ROBERT KRULWICH: Yeah, what is this?

DAN KELLY: get...Hydrogen runs through these grooves.

ROBERT KRULWICH: So what's really going on inside these little holes? How is this thing making electricity?

Well, inside the fuel cell, there are two sets of tiny passageways separated by a membrane. And if you look very close, the two main sections are kind of like the two sides of a tennis court. On one side is hydrogen and on the other side oxygen, and then there's the membrane that separates them. It's kind of like a net.

So here we are in a tennis court. I, of course, you'll notice now, represent the atom oxygen, and for those of you who are oxygen atoms yourselves, you'll, of course, recognize that I have eight electrons. This is a fuel cell, so oxygen is opposite hydrogen. There are two hydrogen atoms, and if you'll rotate please, you'll see that they too each have one electron. Now it is in the nature of this kind of chemistry that hydrogen and oxygen are attracted to each other.

RAY MAGLIOZZI: I yearn for you, Robert.

ROBERT KRULWICH: You yearn for me, but there is a membrane between us. Now the rules of the fuel cell are you will try to come to me—try to come to me—but you'll have to go under or through the membrane. Try under the membrane.

TOM MAGLIOZZI: Try under. Okay.

ROBERT KRULWICH: Down they go, the two hydrogen atoms approaching the oxygen with ardor, but notice now...

TOM MAGLIOZZI: We can do it! Robert, I love you!

ROBERT KRULWICH: ...through the membrane...

TOM MAGLIOZZI: We made it, Robert! We made it! Oh, my god, finally.

ROBERT KRULWICH: Now, but—and I use the word "but" advisedly—could you please rotate? So show the audience your butts at this moment. You'll notice that this has got no electron; no electron here. The question has to be asked, "Where are the electrons?"

Remember when the hydrogen had to come through the membrane? Well, when it did that, its electrons were stripped off. Electrons simply aren't allowed through the membrane in a fuel cell.

But they do want to come to the other side. So very cleverly, they go around the membrane, right at the edge of the fuel cell, over and over and over again. And moving electrons, well, that's electrical current. That's what lights up a light bulb, or, in this case, what powers your car.

And here's what makes the fuel cells so clean: after the electrons get to the other side, they rejoin their old friends, hook up with the oxygen, and before you know it, you've got H2O—water—pretty neat.

So in these cars, when you check the tailpipe, instead of exhaust, what you get...


ROBERT KRULWICH: Oh, how do you know?

RAY MAGLIOZZI: 'Cause it looks like water.

TOM MAGLIOZZI: Take a drink. Lick it. Lick it!

He's dead! I didn't like him anyway.

ROBERT KRULWICH: The very idea that a car's motor could be this clean has an enormous appeal, especially to certain politicians. They act like it's going to be easy—well, sometimes they do—but as it turns out, there's a catch, actually, a bunch of catches. Fuel cells are still very expensive to make, they wear out more quickly. And, oh, yeah, there's another thing...

DANIEL NOCERA (Massachusetts Institute of Technology): A fuel cell needs fuel, so we've been talking about hydrogen and oxygen as our fuel. There's lots of oxygen. But where are we going to get the hydrogen in the first place?

ROBERT KRULWICH: MIT chemistry professor, Dan Nocera, says, "Remember, you've got to have pure hydrogen, all by itself, on one side of the membrane to get things going."

So where do you get pure hydrogen? Well, there's plenty of hydrogen on earth; it's just not pure. It's stuck to other stuff, like oxygen, in water, of course. And hydrogen can be found in fuels like natural gas, you know, hydrocarbons.

But if you take it out of there—and that's where most hydrogen comes from today—you do get a waste product, carbon dioxide. And that's one of the bad guys in global warming.

So what's the answer?

DANIEL NOCERA: I think water is the key for the future.

ROBERT KRULWICH: Every high school chemistry student knows how to split water into hydrogen and oxygen. You just run electricity through it. That's electrolysis.

But hydrogen and oxygen are so cozy and comfortable together, you use up so much electricity prying them apart, it could cost a fortune. So we are facing a significant technical problem here: how do we find a cheap, clean source of hydrogen?

Oh, and there is another issue...

RAY MAGLIOZZI: Excuse us for a minute.

TOM MAGLIOZZI: We'll be right back.

ROBERT KRULWICH: Yeah. Don't we worry a little bit about the danger problem?

RAY MAGLIOZZI: What about the Hindenburg?

ROBERT KRULWICH: Yeah, what about the Hinden...the huge balloon the Germans...

RAY MAGLIOZZI: Yeah. Oh, the humanity...the whole thing.

TOM MAGLIOZZI: We've got to ask him.

RAY MAGLIOZZI: Yeah, Come on.


RAY MAGLIOZZI: What about the Hindenburg? Remember the Hindenburg?

TOM MAGLIOZZI: Yeah, what about the Hindenburg? Yeah.

DAN KELLY: The Hindenburg?

ROBERT KRULWICH: The Hindenburg pretty much ruined hydrogen's reputation. But after the explosion, the fire here, the continuous flames that you see, some say they came from the canvas blimp which was coated with highly flammable shellac.

DAN KELLY: Are you comfortable getting back in now?

RAY MAGLIOZZI: Not really, no. I'll be walking back.

ROBERT KRULWICH: One moment, one moment.

Of course, gasoline is flammable, too, and we drive around with gallons of that in our cars.

It's something else about hydrogen...

NATHAN LEWIS (California Institute of Technology): Well, hydrogen's a gas.


NATHAN LEWIS: That means most of the space between the hydrogen molecules is not useful to make energy. There's nothing there.

ROBERT KRULWICH: So says Nate Lewis, a scientist at Caltech.

Getting enough hydrogen into a car is a challenge, because it likes to spread out, and you've got to squeeze a lot of it into a small place.

You can't use an ordinary gas tank to hold it, because it would burst open. And you wouldn't want that.

Did I run a stop sign?



So the hydrogen tanks have to be super strong, to hold hydrogen squeezed in at pressures up to five to ten thousand pounds per square inch. But even with that special tank, it's tough to get enough hydrogen in the car. For instance, to drive 300 miles, you might need a gas tank four times the size of the one you've got now.

Which leads Professor Lewis to suggest the best place for hydrogen fuel might be in power stations, to light up cities and factories.

After all, why rush to put this fuel out on the road, when you can easily store plenty of it in a factory basement?

So you're a put-it-in-the-basement guy?

NATHAN LEWIS: Put it in the basement first. And then, if we ever figure out how to use it to move us around, that will be great. But your car is the last place that you want to put hydrogen.

ROBERT KRULWICH: But the dream of a car that spews out nothing but water is so appealing. And if we could use water to fuel the car, that would be even better.

TOM MAGLIOZZI: When the technology is really there, you'll be able to open up that little gas fill and fill it up with water like my brother used to do when he was a kid.

DAN KELLY: Yeah. So the car will run...

ROBERT KRULWICH: But the technology to get hydrogen from water efficiently and affordably? That technology doesn't exist yet.

Well, is there anything that we know of that does this regularly, that breaks apart water and...?

DANIEL NOCERA: We do. It's the leaf.

ROBERT KRULWICH: The, the what?


ROBERT KRULWICH: The leaf. It turns out the leaf is a pro at splitting water. In photosynthesis, leaves take water and break it down. The oxygen goes into the air—thank goodness, because that's what we breathe—and the hydrogen hooks up with carbon to make carbohydrates, which the plant needs for fuel. And the energy to do all this comes from the sun.

So does Dan Nocera want to "make like a leaf?"

DANIEL NOCERA: So what we've been doing is trying to actually, not duplicate what a leaf does, but we're saying, "Can we generate hydrogen and oxygen by entirely new ways?"

ROBERT KRULWICH: In his lab, lasers stand in for the sun.

DANIEL NOCERA: What we do is we bring this laser beam in, and now you see, actually, here's a compound that's capturing that green light.

ROBERT KRULWICH: So this is your version of sunshine, and this is your version of something that sunshine is acting on?

The laser shoots through a liquid mixture concocted in Nocera's lab. He's trying to design a chemical, something like this pink peapod thing, that does a special trick, so when light hits it, it will pull apart the H from the O.

Have you ever taken a droplet of water and found a way to just release hydrogen?

DANIEL NOCERA: We haven't done water, but we've done hydrochloric acid and we've been able to make hydrogen.

ROBERT KRULWICH: So that's a "no." But there are labs all over the country now working on hydrogen. Some work with algae, some with solar collectors. And these guys are making hydrogen from water and sunlight.

But at least for now, it's very expensive. And just last year, the National Academy of Engineering and the National Research Council reported there are "major hurdles" on the path to a hydrogen economy, and that clearing them "will not be simple."

So even though the President is saying we could have hydrogen cars for today's generation...

GEORGE W. BUSH: The first car driven by a child born today could be powered by hydrogen and pollution-free.

ROBERT KRULWICH: If the car's really going to be pollution-free, the hydrogen in the tank will have to come from a clean source, and so far, when it comes to splitting water, we're way behind the leaf.

Dan and other scientists are trying to catch up and will keep trying, but the secret, he thinks, may be very subtle.

So how long did it take for the leaf to figure out how to separate hydrogen from oxygen?

DANIEL NOCERA: That took between two and four billion years.

ROBERT KRULWICH: But how much time do you have?

DANIEL NOCERA: I'm guessing around 20 more years.

ROBERT KRULWICH: And so, whatever the politicians may say, learning to "make like a leaf" could take a while.

Okay, so we do have time then for a very short puzzle? This one lasts fifty seconds, exactly, unless, of course, you happened to guess the answer earlier.

MAN: The key ingredient of this substance was originally extracted from the sapodilla tree. Nowadays, the formulation is coated with fine particles of sucrose and enclosed by a thin layer of aluminum seven microns thick. It contains synthetic materials blended with natural products, such as leche caspi and masuranduba. On contact with the enzyme amylase, secreted by glands in the buckle cavity, its texture is transformed to a malleable compound with elastic properties. After several minutes grinding in a heterodont interface, oral fatigue often leads to removal of the sticky molecular structure and adhesion to a suitable surface.

ROBERT KRULWICH: Well, if you got that one, we do have another one at the end of program. It involves a filter-feeding sedentary metazoan. We think you'll probably want to stick around for that.


Whenever a museum wants to really examine a masterpiece, get in close enough to see the hidden details, the brushstrokes, the really teeniest of gestures, what they can do is they can divide the painting into imaginary squares, small, small squares. And then they can take detailed digital photographs of each segment. And then they can patch all the photos together into an exquisitely detailed whole for scholars all over the world to study. This isn't hard. It's actually kind of easy.

But suppose instead of a painting, suppose you're photographing a unicorn stitched out of thousands and thousands of fibers. We're talking here about a tapestry. Now to put the unicorn pictures back together is a puzzle of such enormous difficulty you would need a supercomputer to do this right. And not just a supercomputer, you'd need a couple of guys we know in Brooklyn.

Correspondent Chad Cohen reports.

GREGORY CHUDNOVSKY ( Polytechnic University): And of course, we cannot go back to just one bit, and another is to have two bits at a time, so we least 50 percent or more significance.

CHAD COHEN (Correspondent): The sound of equations blends with the hum of New York City traffic, as these two world class mathematicians make their way to their lab at the Brooklyn Polytechnic University.

Gregory and David Chudnovsky seem to have a never ending conversation about their passion for numbers. It's been that way since their childhood days in Kiev, where younger brother, Gregory, developed myasthenia gravis, a crippling neuromuscular disease.

David has always been there to help, and since emigrating from Ukraine in 1978, the pair have been inseparable.


DAVID CHUDNOVSKY (Polytechnic University): No.

CHAD COHEN: Oh, for the paper.

DAVID CHUDNOVSKY: Yeah, paper consists of already grades...

CHAD COHEN: Between their math-speak and Russian accents, conversation with the brothers can be a challenge. Thoughts come to them rapidly, so rapidly they often complete each other's sentences.

GREGORY CHUDNOVSKY: It simply becomes nothing. You take a difference...

DAVID CHUDNOVSKY: Nothing or a small, small...

GREGORY CHUDNOVSKY: ...small something.

DAVID CHUDNOVSKY: Small discussion.

CHAD COHEN: Their passion for the bigger, deeper meanings of numbers demanded the calculating power of a supercomputer. And since they could never put the millions of dollars together to buy one, they built one from mail-order parts.

DAVID CHUDNOVSKY: It's a good advertisement for Home Depot, because PVC pipe, $2.50 for 12 feet.

GREGORY CHUDNOVSKY: Just a little computer device...

CHAD COHEN: Tom Morgan is their doctoral candidate.

TOM MORGAN (Polytechnic University): You wake up one morning and we have a unicorn emergency. The day before, you're working on some study of mathematics, and the next morning you're working to try to correct a problem with a unicorn.

CHAD COHEN: Yeah, he did say unicorn, the Unicorn Tapestries, to be precise, the prized possessions of the famous Cloisters Museum in Manhattan.

These 500-year-old fabrics are some of the most celebrated medieval art in the world. The masterfully woven interlace of wool, silk and silver threads, tens of thousands of them in each tapestry, took years to complete.

The artworks presented a mathematical challenge that would end up in the Chudnovskys' lab and started a few years ago when all six of the priceless tapestries were taken down during a building renovation.

When the tapestries were removed from their galleries at the Cloisters, they arrived in an unmarked truck here at the Metropolitan Museum. They were taken to a room in the basement known as the Wet Lab. You won't find it on any visitor's guide. There, they were laid out to be cleaned and photographed.

The museum had just begun to use new digital photography to document its two and a half million piece collection. And the giant tapestries demanded a special approach for the head of the Met photo studio, Barbara Bridgers.

BARBARA BRIDGERS (The Metropolitan Museum of Art): What we really wanted was for each individual digital file to reveal almost each and every thread.

CHAD COHEN: So they built a special rig to suspend the camera over each tapestry, and carefully scanned three foot by three foot sections that would be pieced together later. But there was an unexpected problem: the digital files were so enormous, none of the museum's computers could handle them. So they searched for a solution.

Five years later, enter the brothers Chudnovsky.

BARBARA BRIDGERS: They called me and explained that, in their spare time—for fun, I think, which is what great mathematicians do in their spare time—they thought it would be interesting to create a software program that would stitch image files together outside of any visual clues, based purely on the zeros and ones and the mathematics underlying the file.

GREGORY CHUDNOVSKY: The problem looked amazingly simple from a mathematical point of view.

CHAD COHEN: Their challenge: to take the 30 tiles, 30 digital snapshots of the unicorn in captivity, and precisely stitch them together. For two talented mathematicians? Piece of cake.

GREGORY CHUDNOVSKY: It was a technical challenge.

DAVID CHUDNOVSKY: But you think that you can really work over it, and in two weeks, you help people. You have fun. You use hardware, and you walked out, and everybody feels...


CHAD COHEN: And you're on to the next thing.

DAVID CHUDNOVSKY: And unfortunately...

GREGORY CHUDNOVSKY: never works this way.

CHAD COHEN: When they loaded the images into their supercomputer, something was very, very wrong. The tiles just wouldn't line up.

GREGORY CHUDNOVSKY: And here, this tile, it simply doesn't match. But it's very peculiar it does not match. Some pieces actually match and some pieces don't. But this one does. It's actually absolutely maddening.

CHAD COHEN: Right, so you would think you'd just take this and move it over.

GREGORY CHUDNOVSKY: Right, but look at this flower. The flower is perfect.

CHAD COHEN: So therein lies the rub. If you move this here, this flower's off.

BARBARA BRIDGERS: So they said, "This just is not working. And it's very clear to us, from the numbers, that someone has manipulated these files." And I said, "No," you know? "That's not the case."

CHAD COHEN: What was the case was that the brothers' math did reveal some kind of motion in the tapestry. Perhaps the photo staff shifted the fabric accidentally, or maybe the camera was jolted during the long exposures. So David and Gregory created a vector displacement map—that's these little arrows here—to track how the tapestry was moving.

GREGORY CHUDNOVSKY: As these were moving forward, the other pieces actually were moving, in some places, backwards.

CHAD COHEN: By examining the direction of 15,000 vector arrows, the brothers realized the motion had nothing to do with the photographers at all. The tapestry had moved all by itself.

After centuries of hanging on a wall, when the 500-year-old threads were cleaned and placed on the floor, they started to relax and shift randomly. And it was all captured in the pictures.

DAVID CHUDNOVSKY: It's basically like photographing a surface of a, very slowly moving water. It's never the same. And so you photograph one piece, and even if you photograph it, say, a week later, it would be totally different shape.

GREGORY CHUDNOVSKY: Of course, this is an afterthought. When you go...

DAVID CHUDNOVSKY: Yeah, it's a good after...

GREGORY CHUDNOVSKY: It's a good afterthought.

CHAD COHEN: Now they had a problem worthy of a supercomputer and a team of mathematicians. The brothers had to create what are called "warping transformations." That's the kind of math that finds recognizable patterns among features which are similar, but not identical. It's also used to solve problems in handwriting analysis and speech recognition. For the tapestries, it took a gigantic set of equations to optimize the position of every pixel, all 240 million of them.

TOM MORGAN: You start to pull out heavier and heavier artillery. And at some point in there you just say, "Well, I'm not going to let this thing defeat me. We're going to make this image one way or another."

CHAD COHEN: Week after week, Team Chudnovsky fed subsets of the tapestry into the supercomputer, which had to perform some 300 million operations per pixel to digitally weave each thread back together.

GREGORY CHUDNOVSKY: It was a very interesting challenge, and it's a very damn complicated thing. Unfortunately, there is an issue of pride.

CHAD COHEN: After four months of number crunching, the Chudnovskys entered the final instructions, and for 30 straight hours the supercomputer ground out not a billion, not a trillion, but 7.7 quadrillion calculations.

Finally, "Unicorn in Captivity" was reborn in digital form. It was flawless. A version of it now hangs in the Brooklyn Polytechnic Library not far from the brothers' lab.

BARBARA BRIDGERS: After they successfully put the one tapestry together, I said, "Well there are six more." And they said, "We're busy. We're booked."

CHAD COHEN: Well, they didn't stay booked very long. This summer, the brothers will be working once again with the museum's photo team, most likely on this Vermeer painting. Their goal is to use mathematics, of course, to break down the pattern of brushstrokes, perhaps uncovering the sequence of how Vermeer laid them down. And with a little luck, maybe this time the artwork will sit still.

DAVID CHUDNOVSKY: Never happens.

GREGORY CHUDNOVSKY: There will be something else; I do not know what, but there will be something else happening here. It just...otherwise, again, it will be too good to be true.

DAVID CHUDNOVKSY: We know not to touch tapestries, but may be after that we will know not to touch paintings...

GREGORY CHUDNOVSKY: ...touch paintings either.


ROBERT KRULWICH: Correspondent Chad Cohen.

What if I told you that recently, scientists made a discovery that's so surprising and so powerful, not only are we about to know much, much more about how all these diseases work—Alzheimer's and asthma and arthritis and cancer and HIV and all the others—there's a chance, it's a real chance, that we can treat many of these diseases much more effectively? All because of this one discovery called RNAi, with a little "i" at the end, which I'll explain later. You don't hear much about it, but RNAi is a really big deal.

And the curious thing about it is the discovery of RNAi was an accident. It was a puzzle that appeared in a petunia. It was a purple petunia. But to fully appreciate this tale, let's back up a bit.

Every creature, and you know this from high school, is made from a recipe that comes from its DNA, spelled out in chemicals: "A"s and Cs and Ts and Gs, inside the famous double helix. Every creature has its own DNA, different for mice than for whales and for flowers. But to go from a chemical recipe, "A"s, Gs, and Ts, to a real creature that squeaks or soars through the air or turns gloriously pink, that requires RNA. RNA is the thing that turns you from a chemical code to a real, pulsing, living creature. RNA builds life.

That's big; so big, that to RNA researchers like Greg Hannon, RNA is more important than DNA.

GREGORY HANNON: DNA really works for RNA and proteins really work for RNA.

ROBERT KRULWICH: Would you get an argument, by the way, from somebody else?

GREG HANNON: Oh, undoubtedly. Sure.

ROBERT KRULWICH: So how do you get from DNA to become a real creature?

Well, let's take one of those fantastic voyages and we'll show you. We're going to find DNA in, well, we'll make it a typical cell. So we're going have to fly in and then go off to the nucleus of the cell, which we'll make a beautiful castle, the headquarters.

And there's the DNA, the master code, inside the nucleus. "DNA," says Greg, "never leaves the nucleus."

GREG HANNON: You ever meet one of those mean librarians? You know, the...


GREG HANNON: ...Special Reserve section?

ROBERT KRULWICH: The ones that go, "Pow"

GREG HANNON: Right, you can take the thing, you can copy it, but you can't take the book, 'cause somebody else might need it.

ROBERT KRULWICH: So if DNA is locked in the nucleus, how do we get the information out to build our creature?

Well, that's what RNA does. That scribe copying recipes out of the cookbook and throwing them out the window, out to the cytoplasm sea that makes up most of the cell, all those recipes floating through the air? They are RNA.

And to finish up, in that sea, you see hundred of thousands of, well, we've made them into little guys with chef hats.

GREG HANNON: Those would be ribosomes. And in your world, they're chefs who are using the recipes that are written in the RNA.

ROBERT KRULWICH: And whenever a recipe lands on a chef, whatever it is, he cooks it?

GREG HANNON: Whatever it is, he cooks it.

ROBERT KRULWICH: And each recipe is for a protein. Proteins build cells: bone cells, brain cells, all cells. So, all these chefs are basically building you. You are made of proteins.

And because of RNA, we can copy, we can distribute, and we can cook up you and me. And RNA has been doing this for more than three billion years.

But there was something spectacular about RNA that nobody knew till just a few years ago, and they learned about it, as we told you, by accident.

RICHARD JORGENSEN: Here's a good one. Maybe this...

ROBERT KRULWICH: In 1986, geneticist Rich Jorgensen was working at a biotech startup company in California. He was asked to create a spectacularly dazzling flower...

RICHARD JORGENSEN: That looks good.

ROBERT KRULWICH: attract investors.

RICHARD JORGENSEN: So that we could convince venture capitalists, investors, to give us more of the green stuff, more money.

ROBERT KRULWICH: Still, back in 1986, geneticists didn't know how to work that easily with, say, roses. And so...

RICHARD JORGENSEN: We began with a simple plant, regular, garden-variety petunias—petunias being a plant that were easy to introduce genes to in 1986.

ROBERT KRULWICH: And so they decided to create a very, very, very purple petunia.

Rich knew which gene produced purple. He knew how to sneak an extra copy of that gene into the plant's DNA—the master text to be copied by that monk-like scribe.

RICHARD JORGENSEN: It will be transcribed by the monk the same as any other gene. He'll throw the transcript out the window, into the cytoplasm, where the chef will be able to pick it up and use it.

ROBERT KRULWICH: Rich thought that if he added more purple recipes, he'd get a purpler petunia. So he did it, and he waited.

And what happened?

RICHARD JORGENSEN: We produced, instead, white flowers.

ROBERT KRULWICH: White flowers?

RICHARD JORGENSEN: The complete opposite of what we had expected, completely white flowers. We lost pigmentation completely. Our initial reaction was that something must have been wrong with the gene that we had engineered, introduced to the plant.


RICHARD JORGENSEN: A mistake. So we checked everything out, and there were no mistakes that we could find.

ROBERT KRULWICH: So why didn't the petunias turn purple? What happened?

ERIC LANDER: The petunia was a big puzzle. Nobody understood why, when you add an extra gene for purple, you should not get more purple, but less purple. It took a decade of brilliant scientists working on petunias and fruit flies and worms and other organisms to finally work out what was going on.

ROBERT KRULWICH: And what was going on is, quite by accident, Rich had discovered a secret inside living cells.

Cells, from time immemorial, have had a mortal enemy called the virus. So let's imagine that the virus is a pirate ship. It lands; it then sends the invaders inside the cell to shower recipes down to those cooks. But some of those recipes, you'll notice, look a little different. And what's in these recipes is not good for the cell.

GREGORY HANNON: No, it's decidedly not good for the cell, because the sole purpose of that virus is to make additional copies of itself, and to the point that the entire cell is filled up with this. And the cell explodes, releasing these viruses to go and then infect whatever other cells they can find.

ROBERT KRULWICH: So the theory is that long ago, cells developed a secret defense system, which we will call the Cop.

What the Cop does is, when viruses invade and create a shower of murderous recipes, the Cop looks and thinks, "Hmm, some of these have a very fishy shape."

It's a chemical difference, which comes down to some of the viral recipes are two pages instead of one, and one side is a mirror image of the other. But the point is, to the Cop there's something not right about this shape.

So when they see it in that shape?

ERIC LANDER: They say, "Virus!" They say...


ERIC LANDER: ..."Uh oh."

ROBERT KRULWICH: And the Cop destroys the recipe.

Now when you say it "destroys," is this, should we think like a kung fu kind of thing? Is it like a "Hyeh!" sort of deal?

ERIC LANDER: Yeah, a little enzymatically, a little thermodynamics. Things like that.

ROBERT KRULWICH: Enzymatically?

ERIC LANDER: Enzymatic kung fu maybe, yeah.

ROBERT KRULWICH: The Cop destroys not only the oddly shaped version. Whenever he sees that recipe, oddly shaped, regular shaped, that recipe in any form must be destroyed to defeat the virus.

And the interesting thing is, until 1998, nobody knew that cells had this defense mechanism.

ERIC LANDER: We had no idea it was there. That's what's so amazing, is...this whole mechanism had been sitting there, where cells were able to tell that something was very funny when they saw mirror image messages and start not just destroying the messages, but destroying anything that looked like that message. They'd worked out this whole defense system against viral RNA, and we then accidentally stumbled into using it.

ROBERT KRULWICH: The accident was Rich Jorgensen's purple petunia. The question, remember, was, when Rich tried to make his petunia more purple, why did it turn white?

Well, the answer, it turns out, was that Rich, by accident, discovered the Cop. When Rich invaded the petunia cell and inserted his make-more-purple instructions, he didn't know it, but his purple instructions happened to have that suspicious viral shape.

So when the Cop saw the recipe, the Cop thought "Virus!" and destroyed every recipe for purple in the cell.

RICHARD JORGENSEN: So there's no possibility anymore of producing the purple pigment, because the purple transcripts are gone.

ROBERT KRULWICH: If there are no recipes for purple, the chefs don't cook purple.

RICHARD JORGENSEN: And because there's no purple pigment produced, the flowers will be white.

ROBERT KRULWICH: And that's how Rich and his petunias helped discover what we now call RNAi.

ERIC LANDER: RNAi means RNA interference, because the Cop is interfering with RNA messages, with the recipes in the cell.

ROBERT KRULWICH: And when scientists realized that every plant and animal cell has RNAi—a way to turn off the recipes, turn off genes—they thought, "Hmm, maybe we can use these Cops to work for us."

MARTY RUSSELL: Okay, Trevor?

ROBERT KRULWICH: Which brings us to Marty. Seventy-eight years old, she and her husband used to spend lots of time here, at their daughter's nursery.

MARTY RUSSELL: Thank you, Rosie.

ROBERT KRULWICH: Years ago, she enjoyed doing lots of things.

TREVOR RUSSSELL: Her passion was reading.


TREVOR RUSSELL: She would read everything...



MARTY RUSSELL: I loved to play golf, bridge.

ROBERT KRULWICH: But then Marty began losing her sight.

MARTY RUSSELL: Couldn't see. And I'd probably get the peppers in with the zucchinis, and there'll be big problems then.

ROBERT KRULWICH: So she went to her doctor, who told her...

MARTY RUSSELL: "You have macular degeneration."

ROBERT KRULWICH: A degenerative disease caused by too many blood vessels growing in the eye underneath the retina.

PETER KAISER, M.D. (Cleveland Clinic Cole Eye Institute): As these blood vessels grow, they leak out fluid and blood in the center of her vision, and it's as if you're looking through a very dirty windshield, essentially.

MARTY RUSSELL: I went home; I was just devastated.

ROBERT KRULWICH: So Marty volunteered to be a candidate for RNAi therapy, something so new, she's kind of a pioneer.

PETER KAISER: She was probably one of the first to get it for any disease whatsoever, specifically for macular degeneration.

Hey. How are you?

MARTY RUSSELL: Hello, Dr. Kaiser. How are you?

PETER KAISER: Good to see you. This is the V.I.P. room.

MARTY RUSSELL: I feel very honored.

ROBERT KRULWICH: The reason Marty has so many blood vessels growing in her eye, clouding her vision, is there's probably a mistake in her DNA, in a gene that produces too many recipes that say, "Make more blood vessels." So the chefs cook up proteins for more, and she ends up with too many blood vessels.

Her doctor wants her to have fewer blood vessels, but how do you get the chefs to make fewer blood vessels?

ERIC LANDER: It was pretty easy. You want to shut down a gene? Put in a copy of the gene with its mirror image...signals the cell, "Better shut this thing down."

PETER KAISER: We inserted a needle after numbing her up.

ROBERT KRULWICH: So the doctors put—literally injected—RNA recipes into Marty's eyes that said, "Make more blood vessels." But they made those recipes look dangerous, like viral recipes, hoping the Cop in Marty's cells would leap to it and destroy lots of recipes for more blood vessels, leaving Marty with fewer blood vessels.

They wanted the Cop to turn off Marty's disease.

Did it work?

PETER KAISER: Marty's vision has improved. It's a very promising result.

MARTY RUSSELL: I can play bridge now.

PETER KAISER: Which is very important.

MARTY RUSSELL: I'm not great, but it's part of my life.

ROBERT KRULWICH: She can see flowers again.

MARTY RUSSELL: Oh, some of them are just gorgeous.

ROBERT KRULWICH: So, apparently, they did trick the Cop in Marty's cells to reduce vein production, although not completely.

MARTY RUSSELL: I see the yellow. The inside is just a little cloudy, but I can see it.

PETER KAISER: There's a lot of questions still that need to be answered. This is not a treatment that is proven.

ROBERT KRULWICH: Can we deliver recipes to the right cells?


ROBERT KRULWICH: Does the treatment last?

MARTY RUSSELL: That's beautiful.

ROBERT KRULWICH: All these are big questions. Still, in mice, RNAi has been effective with Huntington's Disease, Lou Gehrig's disease, hepatitis, breast cancer.

"So," says Greg, "if we ever work this out in humans..."

GREGORY HANNON: Any sort of disease that you can imagine becomes fair game.

ROBERT KRULWICH: All the diseases which would be helped if you shut off a gene?

GREGORY HANNON: Cancer, HIV, for example...

ROBERT KRULWICH: Wait a minute. Are you this because you're just an RNA buff that you're've just listed cancer and HIV. These are famous, big, fat diseases.

GREGORY HANNON: ...arthritis...

ROBERT KRULWICH: Well, stop listing them and tell me, is this a prejudice that you're telling me or is this true? I mean, these are all candidates for this kind of therapy?

GREGORY HANNON: Certainly they are.

ROBERT KRULWICH: And finally, we have saved the best for last. The true power of RNAi goes even deeper than finding cures to terrible diseases. Because what RNAi does...remember, the Cop's job is to turn off information, turn off genes.

ERIC LANDER: The big problem of understanding, say, the human genome is you have 20,000 genes. How in the world are you supposed to know what each one does? Well, one very good way to start would be to turn off gene number one and see what went wrong.

ROBERT KRULWICH: So you could go through all the genes that make up a human or, for that matter, make up a petunia, and turn off each gene one at a time. If you trick the Cop to turn off gene number one, no color.

So gene number one is involved in color production. Try gene number two, no petals—gene number two, involved with petals. And so on.

GREGORY HANNON: You could make too many leaves; they could curl up; they could be upside down. Almost anything could happen.

ROBERT KRULWICH: But getting rid of the gene tells you what the gene does when it's working?

GREGORY HANNON: That's right.

ERIC LANDER: The RNAi discovery is just amazing. Ten years ago, when we were sitting around talking about what we would really need to understand the human genome, we all said we would need some magic way that you could turn off any gene at will, just based on knowing it's sequence. And what's happened is this discovery by scientists about RNAi has given us exactly that. It turns out that nature already had a way to turn off any gene at will.


ROBERT KRULWICH: And now, with RNAi as their key, scientists will have the means to decode every living thing, to identify the genes that allow us to grow, that allow us to move, that give us beauty and color.

RNA is a modest little molecule, but what it gives us is the world.

So now we go from the profound to the creepy. Imagine, if you will, a glacier, one of those vast mountains of ice that hang off mountains or off coastlines. When you read about glaciers, they tell you that every year glaciers move. They don't sprint, exactly. They just creep along at a rate of, well, we did look it up, and if this block of ice right here were a typical glacier, it would move about ten inches every day.

So if I am standing, as I am right now, 20 inches from—let's make this an incoming glacier—in two days, it would move from where it is now to here.

So that's an average glacial speed. It's less than one foot every day. Now here's the puzzle. We're going to take you to a real glacier. It's a big one that, for some reason, says our correspondent, Peter Standring, is behaving very badly.

PETER STANDRING (Correspondent): Take a look around. This is Greenland's largest glacier, the Jakobshavn. It's more than four miles wide and over 1,000 feet thick. It looks solid and indestructible, but looks can be deceiving.

Like any glacier, the Jakobshavn is actually a river of ice; it's moving. Fed by the massive Greenland ice sheet, it's now the fastest moving glacier in the world.

Jay Zwally, a glaciologist from NASA, knows exactly how fast the ice is moving.

JAY ZWALLY (NASA): So right now this ice is moving. You can't feel it moving. It's moving this far in one day. In a month, it'll be five feet past that pole over there.

PETER STANDRING: Instead of moving one foot a day, which is normal for glaciers, in the last five years the speed of the Jakobshavn has increased to an astounding 113 feet every day, and no one knows why. It's a mystery buried deep within the arctic ice.

Here, you can see the consequence of all that speeding up, where the glacier meets the sea and enormous icebergs break off. This glacier is not just speeding up; it's shrinking, eight miles in just the last five years, as huge pieces of ice fall into the ocean.

To give you a better idea of just how much ice, take a look at this iceberg here. We're estimating it to be about 50 feet high. In terms of thickness, that's how much ice the Jakobshavn Glacier is losing each year. The result is that this glacier alone is responsible for four percent of all sea level rise worldwide. So why is this happening?

At this research station set up on the Greenland ice sheet, scientists use a variety of tools to answer that question.

Hello, there. How are we doing? Are we keeping warm?

KONRAD STEFFEN (University of Colorado): Yes.

PETER STANDRING: Konrad Steffen is the camp's founder. He's spent the last 31 winters out on the Arctic ice.

And do you guys manage to go about your business and your experiments without freezing to death out here?

KONRAD STEFFEN: Oh, I have never lost a grad student here on the ice.

PETER STANDRING: About 10 miles from camp is this GPS receiver, which gives precise measurements of the position of the ice. This allows scientists to track changes in its speed and when they happen.

JAY ZWALLY: It's always exciting to take a look and turn the unit on and see what it's recorded throughout the year.

PETER STANDRING: Around the clock, weather stations measure the height of the snow pack, the wind speed, humidity and temperature. These instruments collect volumes of information, and here is where they put it all together. These red tents serve as both kitchen and laboratory.

KONRAD STEFFEN: Good morning.

MAN: Good morning.

KONRAD STEFFEN: I heard you have Starbucks here.

PETER STANDRING: A warm place to ponder the mystery of what's happening to the Jakobshavn Glacier.

KONRAD STEFFEN: That's minus 5.8.

PETER STANDRING: And is that Fahrenheit or Celsius?

KONRAD STEFFEN: That's all in Celsius. We measure everything in Celsius.

PETER STANDRING: Koni Steffen says that the main culprit is rising temperatures.

KONRAD STEFFEN: The summer temperature have increased by about one and a half to two degrees over the 15 years, but the winter temperature have increased by six degrees.

PETER STANDRING: That temperature increase has led to a longer melt period, more days each year when the ice melts rather than freezes.

Aerial and satellite observations show that the melt water, that blue area on the ice, doesn't stay on surface. So where does it go?

I know it's hard to imagine, but I'm actually standing at the bottom of a big lake. It measures three miles by three miles. By the end of July, enough of this ice and snow will have melted to fill this lake with 60 feet of water. Eventually, all of that water finds a hole and makes its way down to the bottom of the ice sheet.

Glaciologists thought that water would freeze before it ever made its way down through 4,000 feet of ice. And normally it would, if it weren't for one thing. The incredible pressure caused by the huge volume of increased melt water on the surface forces it to the bottom by sheer weight.

JAY ZWALLY: What people didn't know before was where that water went. They thought maybe it stayed in the glacier, thought maybe it went to the bottom, stayed in a tunnel. But our results show that it goes to the bottom and spreads out, actually lifts the glacier up a little bit, and allows it to flow partly on a bed of water.

PETER STANDRING: Literally sliding it into the sea. And although this effect takes place mostly in the summer, recent satellite measurements show that even in the sub-zero Arctic winter, when there is no melt effect, the Jakobshavn is showing little signs of slowing down. Jay Zwally places the blame for this on global warming.

JAY ZWALLY: The really disturbing thing is the pace of things that have been happening in the last 10 years. And almost all of the scientists in the world, all the respectable scientists, believe that man is having a definite impact. And we're seeing this more in the Arctic than anywhere else.

PETER STANDRING: Just as unsettling is data from some of Greenland's other glaciers that show that they, too, are starting to undergo the same dramatic changes. Where and when or even if this will stop, no one knows.

But what scientists do know is that changes they once thought occurred over hundreds or thousands of years are today taking place right in front of their eyes.

ROBERT KRULWICH: Correspondent Peter Standring. Before we leave you tonight, we're going to offer you one last puzzle.

We are thinking of an object which, like a glacier, is found in regions familiar to hydrographers, where it shows remarkable properties of both compression and evulsion. Can you guess what I'm thinking of?

MAN: Discrete spicules fuse together to form the skeleton of this organism. Its life revolves around pumping large volumes of fluid through the core tissue. Specialized flagellated cells drive the current in a single direction through a system of complex canals and chambers. Construction of these labyrinthine corridors can take years. It is a sedentary filter-feeding metazoan at home in the sea and in the bath.

ROBERT KRULWICH: Well, we're going to leave you now until our next episode, but NOVA ScienceNOW continues online. You can watch any part of this broadcast again. You can e-mail your questions to scientists who are working on fuel cells, RNAi, glaciers... We also have a new section where we discuss and investigate new story ideas. We'd love your help there, so please, please tell us what you think about this broadcast. You can find us at

I'm Robert Krulwich. Goodbye.

NOVA is a production of WGBH Boston.

Better science makes better innovation a possibility. Better innovation makes better communication a reality—Sprint, making communication better. Sprint is a continuing proud sponsor of NOVA.

Major funding for NOVA ScienceNOW is provided by the National Science Foundation, America's investment in the future; The Howard Hughes Medical Institute, serving society through biomedical research and science education—HHMI.

Additional funding is provided by the Alfred P. Sloan Foundation to enhance public understanding of science and technology, the George D. Smith Fund, and the Kavli Foundation, advancing scientific knowledge at leading universities worldwide.

Major funding for NOVA is also provided by the Corporation for Public Broadcasting, and by PBS viewers like you. Thank you.


NOVA scienceNOW: July 26, 2005

Fuel Cells

Edited by
Doug Quade & Dick Bartlett

Written, Produced and Directed by
Julia Cort


Edited by
Win Rosenfeld

Co-produced by
Kyla Dunn

Produced by
Vincent Liota

Fastest Glacier

Edited by
David Small & Stephen Mack

Produced and Directed by
Peter Doyle

The Brothers Chudnovsky

Edited by
Stephen Mack

Co-produced by
Mary Robertson

Produced and Directed by
Dean Irwin

Executive Producer
Samuel Fine

Executive Editor
Robert Krulwich

Senior Series Producer
Vincent Liota

Supervising Producer
Andrea Cross

Development Producer
Kyla Dunn

Associate Producer
Win Rosenfeld

Program Editor
David Small

Unit Manager
Candace White

Senior Researcher
Jason Spingarn-Koff

Production Secretary
Hong Jung

Rob Morsberger

NOVA scienceNOW Series Animation

Associate Producers
Marty Johnson
Anthony Manupelli
Mary Robertson

Yunsik Noh

Edgar Boyles
Joel Coblenz
Brian Dowley
Anthony Forma
Mark Gunning

Sound Recordists
Ted Gesing
Mike Karas
David Pickner
Erik Reisner
Merce Williams

Aerial Production Coordinator
Billy Richards

Audio Mix
John Jenkins

Mitch Butler

Production Assistant
Robbie Gemmel

Maxwell Foxman
Jonathan Lung

Special Thanks
Robin Abbott
Waleed Abdalati
California Fuel Cell Partnership
Ted Driscoll
Maria Ghirardi
The Metropolitan Museum of Art
National Renewable Energy Laboratory
Polytechnic University
Simon Stephenson
United Technologies Company
Brian Von Herzen

Howard Hughes Medical Institute Investigator
Gregory Hannon, Cold Spring Harbor Laboratory

Archival Material
ABCNEWS VideoSource
Avalanche, Inc.
BBC Motion Gallery
Corbis Corporation
Getty Images

NOVA Series Graphics
yU + co.

NOVA Theme Music
Walter Werzowa
John Luker
Musikvergnuegen, Inc.

Additional NOVA Theme Music
Ray Loring

Post Production Online Editor
Spencer Gentry

Closed Captioning
The Caption Center

NOVA Administrator
Dara Bourne

Adina Barnett
Lisa Cerqueira
Jonathan Renes
Diane Buxton
Olivia Wong

Senior Researcher
Barbara Moran

Production Coordinator
Linda Callahan

Unit Manager
Lola Norman-Salako

Richard Parr

Legal Counsel
Susan Rosen Shishko

Post Production Assistant
Alex Kreuter

Associate Producers, Post Production
Nathan Gunner
Patrick Carey

Post Production Supervisor
Regina O'Toole

Post Production Editor
Rebecca Nieto

Post Production Manager
Maureen Barden Lynch

Supervising Producer
Stephen Sweigart

Producer, Special Projects
Susanne Simpson

Coordinating Producer
Laurie Cahalane

Senior Science Editor
Evan Hadingham

Senior Series Producer
Melanie Wallace

Managing Director
Alan Ritsko

Senior Executive Producer
Paula S. Apsell

NOVA scienceNOW is a trademark of the WGBH Educational Foundation

This material is based upon work supported by the National Science Foundation under Grant No. 0229297. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

© 2005 WGBH Educational Foundation

All rights reserved

NOVA scienceNOW
Watch online

The hour-long program is available to view online here.


About NOVA | NOVA Homepage | Support NOVA

© | Created September 2006