News this Week

Science  31 Jan 2014:
Vol. 343, Issue 6170, pp. 468
  1. Around the World

    1 - Rome
    Italy Faces E.U. Fine Over Animal Testing Directive
    2 - Saint Martin
    Chikungunya Gains Foothold in the Americas
    3 - Tokyo
    Alzheimer's Researcher Strikes Back at Allegations
    4 - Beijing
    Lunar Rover in Peril
    3 - Madrid
    Court Upholds Fines for Unauthorized Clinical Trial


    Italy Faces E.U. Fine Over Animal Testing Directive

    A legislative stalemate over animal research could become very costly for Italy. On 23 January, the European Commission asked the E.U. Court of Justice to impose a fine of more than €4.5 million per month on the country for failing to incorporate a 2010 E.U. directive on animal testing into national laws by 10 November 2012. The directive aims to minimize the use of animals in research and standardize animal protections across the union. Six other countries are also delinquent, but so far only Italy has faced a penalty.

    While Italy awaits the court's decision, a draft law is wending its way through the political system that would impose limitations well beyond those required by the European Union, but would postpone until 2017 some of the most controversial parts of the directive, including bans on drug abuse research involving animals. Scientists have spoken out against the bill, while animal rights activists have launched fierce demonstrations and even personal attacks on Italian researchers. A decision on the law is expected in March.

    Saint Martin

    Chikungunya Gains Foothold in the Americas

    Bad Sign.

    Characteristic rash and vector (Aedes aegypti) of chikungunya.


    The mosquito-borne chikungunya virus appears to have established itself in the Caribbean, raising fears that the disease, which causes high fever and debilitating joint pain, will soon make its way to mainland North and South America. The outbreak represents the first time mosquitoes have spread the virus between people in the Americas. Previously, all chikungunya patients in the region had picked up the virus while traveling in endemic areas in Africa, Europe, and Asia.

    Since the first two locally acquired cases were confirmed on the French section of Saint Martin in December, the virus has been moving fast: Between 16 and 23 January alone, 498 new cases were confirmed in the Caribbean, including on the islands of Dutch Sint Maarten, Martinique, Guadeloupe, Saint Barthélemy, Dominica, and the British Virgin Islands, according to the European Centre for Disease Prevention and Control.

    The chikungunya virus, first detected in Tanzania in 1952, is primarily transmitted by Aedes aegypti mosquitoes, which are also responsible for spreading dengue fever. Symptoms can last for weeks or months, and there is no vaccine or cure.


    Alzheimer's Researcher Strikes Back at Allegations

    Scientists accused of data manipulation in a large Alzheimer's study are staunchly defending the integrity of the data, but acknowledge shortcomings in the project's management. The Asahi Shimbun, a Japanese newspaper, claimed in its 10 January issue that researchers behind the ongoing $31 million Japanese Alzheimer's Disease Neuroimaging Initiative (J-ADNI) changed details of memory tests long after data were collected and that some participants were improperly included (Science, 17 January, p. 234).

    Neuropsychiatrist Takashi Asada of the University of Tsukuba, who oversees the study's clinical side, calls the story "completely fake." After a standard prepublication data check turned up possible inputting errors, researchers double-checked their records and made changes, but did not falsify data, Asada maintains. Patients above the study's 85-year age limit were enrolled by mistake, he says, and will be omitted from analyses and publications. Several U.S. Alzheimer's researchers have voiced support for their Japanese colleagues via Alzforum. The University of Tokyo, where an ethics committee approved the project protocol, has agreed to a health ministry request to investigate the research.


    Lunar Rover in Peril

    Better days.

    Jade Rabbit on the moon last week.


    It may be curtains for China's moon rover. On 25 January, the Chinese defense science agency announced that the Jade Rabbit rover had encountered a mechanical "abnormality" just before entering hibernation mode to wait out the 2-week-long lunar night. The rover simply was unable to move, explains a mission scientist. He says that engineers are holding out hope that the frigid lunar night will "change the status a bit" and allow them to solve the problem.

    A failure to rouse Jade Rabbit after the lunar day dawns next week would scuttle a big chunk of the mission's scientifi c program. The rover began its anticipated 3-month mission on 14 December by probing surface geochemistry and taking ground-penetrating radar images as much as 100 meters deep—a lunar first. But even if sunrise brings bad news, the Chang'e-3 lander will soldier on alone, using its ultraviolet camera to observe Earth's plasmasphere and a near-ultraviolet telescope to make the first detailed observations of stars and galaxies from the moon's surface.


    Court Upholds Fines for Unauthorized Clinical Trial

    A Spanish court of appeals has upheld fines for HIV/AIDS researcher Vicente Soriano of the Hospital Carlos III in Madrid. Soriano is liable for €210,000 for three violations: conducting a clinical trial without the Spanish Agency for Medicines and Health Products' approval, failing to obtain insurance for the trial, and informing participants he had his hospital's ethical approval when he did not, according to the decision published 14 January.

    The 2009 study, published in HIV Clinical Trials in 2010, involved 311 patients with undetectable HIV levels undergoing protease inhibitor treatment. Seeking to reduce side effects, Soriano proposed once-daily treatments with the antiretroviral drug raltegravir instead of the approved twice-daily protocol. He appealed the original 2011 decision by a lower court, describing his work as an observational study, not a clinical trial. He says he was the victim of "harassment" from the hospital, where he continues to work.

  2. Random Sample

    Now You See It …


    "One minute we're eating pizza then five minutes later we've helped to discover a supernova." That's how one of four undergraduate students at University College London summed up last week's chance discovery of a type Ia supernova 12 million lightyears from Earth. Named SN2014J, the exploding star is the closest supernova of its kind to be spotted in 4 decades.

    During an evening telescope workshop, the students' professor, Steve Fossey, noticed a bright spot in the galaxy Messier 82 (left) that he hadn't seen before. Other telescopes around the world soon confirmed that it was a type Ia—one of the brightest events of the cosmos. They arise when a white dwarf—the compact, burned-out husk of a solar-mass star—accumulates enough material from a companion star to ignite a runaway explosion. Researchers plan to collect spectral data on SN2014J that should help them better understand the origin and brightness trajectory of type Ias.

    They Said It

    "Are you going to cover the alien rock throwers?"

    —Question submitted via Twitter by Star Trek star William Shatner to a NASA news briefing last week, regarding the mysterious "jelly doughnut" rock photographed by the Mars Opportunity rover.


    Join us on Thursday, 6 February, at 3 p.m. EST for a live chat with experts on combining alternative therapies with conventional medical treatments.

    3D Printouts Reveal How the Clingfish Clings


    While most fish use their fins to move, the aptly named clingfish also uses them to stay put on rocky Pacific shores. Its pelvic fins have formed a suction cup structure that attaches to rocks and holds the fish in place, despite breaking waves. Comparative biomechanist Adam Summers of the University of Washington's Friday Harbor Laboratories couldn't figure out how this nickel-sized structure with 211 bones and muscles generated and maintained the negative pressure needed to stay stuck. So he and Thomas Kleinteich from Kiel University in Germany turned for help to a so-called 3D printer. Based on coordinates from a CT scan of a clingfish suction disc, Kleinteich printed plaster replicas of its bones at 20 times their actual size and pieced them together. Then he recreated the connecting muscles from yarn using a second CT scan of the soft tissue as a guide. The oversized disk proved easier to analyze, revealing specifi c muscles that pump out water to create negative pressure and showing that the backward stroke of a single bone was enough to break the suction, Summers reported earlier this month at the annual meeting of the Society for Integrative and Comparative Biology in Austin. His team has also used the printer to shrink whale shark heads to a workable size and to create dozens of precisely sized shells to test how hermit crabs choose their homes. Once considered a luxury item in biology labs, 3D printers are "now totally in the reach of any lab," Summers says. "Everyone is going to be using them."

    By the Numbers

    54.7 — Average hours worked per week, according to an international survey of 650 early-career researchers by the Global Young Academy.

    1.78 billion — People breathing unhealthy air around the world, according to the latest survey by the Environmental Performance Index, up by 606 million since 2000.

  3. Newsmakers

    Brain Mapper and RNA Researcher Bag Medicine Prize

    Le Bihan



    This year's Louis-Jeantet Prize for Medicine goes to two researchers devoted to studying basic processes: how water moves through the brain and how cells dispose of damaged goods. The Swiss foundation, named for French businessman Louis Jeantet, gives the annual award for fundamental biological research with promising clinical implications. It chose physicist and physician Denis Le Bihan for his role in developing diffusion MRI, a technique that reveals the structure of the brain by tracking how water and other molecules move within tissues. Diffusion MRI has been used to recognize signs of disease in the brain, from cancer and stroke to Alzheimer's and schizophrenia.

    Biochemist Elena Conti of the Max Planck Institute of Biochemistry in Munich, Germany, studies the structure and activity of cellular mechanisms that recognize faulty RNA molecules and break them down. She will be honored alongside Le Bihan in April at an award ceremony in Geneva. Each will receive roughly $780,000 in prize money, of which nearly $700,000 must be devoted to future research.

  4. The Dangerous Professor

    1. Kai Kupferschmidt

    David Nutt wants to make drug policies science-based and give the world a safe alternative to alcohol. If only politicians would listen, he says.


    Nutt says politicians often have a "primitive, childish" way of thinking about drugs.


    LONDON—David Nutt is trying to develop a new recreational drug that he hopes will be taken up by millions of people around the world. No, the 62-year-old scientist isn't "breaking bad." In fact, he hopes to do good. His drug would be a substitute for alcohol, to create drinks that are just as intoxicating as beer or whiskey but less toxic. And it would come with an antidote to reverse its effects, allowing people to sober up instantly and drive home safely.

    Nutt, a neuropsychopharmacologist at Imperial College London and a former top adviser to the British government on drug policy, says he has already identified a couple of candidates, which he is eager to develop further. "We know people like alcohol, they like the relaxation, they like the sense of inebriation," Nutt says. "Why don't we just allow them to do it with a drug that isn't going to rot their liver or their heart?"

    But when he presented the idea on a BBC radio program late last year and made an appeal for funding, many were appalled. A charity working on alcohol issues criticized him for "swapping potentially one addictive substance for another"; a commentator called the broadcast "outrageous." News-papers likened his synthetic drug to soma, the intoxicating compound in Aldous Huxley's dystopian novel Brave New World. Some of his colleagues dismissed the idea as scientifically unfeasible.

    Nutt wasn't surprised. As a fierce advocate of what he says are more enlightened, rational drug policies, he has been a lightning rod for a long time. Politicians, in Nutt's view, make irrational decisions about drugs that help them win votes but cost society dearly. Drug policy is often based on the moral judgment that people should not use drugs, he says. Instead, it should reflect what science knows about the harms of different drugs—notably that many are far less harmful than legal substances such as alcohol, he says. The plan for a synthetic alcohol alternative is his own attempt to reduce the damage that drug use can wreak; he believes it could save millions of lives and billions of dollars.

    Such views—and the combative way in which he espouses them—frequently land Nutt in fierce disputes. Newspaper commentators have called him "Professor Nutty" or "the dangerous professor." In 2009, he was sacked from his position as chair of the United Kingdom's Advisory Council on the Misuse of Drugs, tasked with giving scientific advice to the home secretary, after he criticized a government decision on cannabis.

    But in November 2013, he received the John Maddox Prize for standing up for science. "In circumstances that would have humiliated and silenced most people," wrote neurobiologist Colin Blakemore, one of the judges, "David Nutt continued to affirm the importance of evidence in understanding the harms of drugs and in developing drug policy."

    Controversial comparisons

    David Nutt does not look like a dangerous professor. Short and heavyset, he has a jovial, round face and an old-fashioned mustache; one could mistake him for a London taxi driver. He limps slightly, has a down-to-earth way of speaking, and laughs a lot when he talks. "He is a real personality," says psychopharmacologist Rainer Spanagel of Heidelberg University in Germany. "You can be in a meeting and almost have a result, then he will come in an hour late, stir everything up, and in the end convince everyone of his position."

    Nutt says he realized at an early age that "understanding how the brain works is the most interesting and challenging question in the universe." When he was a teenager, his father told him a story of how Albert Hofmann, the discoverer of LSD, took a dose of that drug and felt that the bike ride home took hours instead of minutes. "Isn't that incredible, that a drug can change time?" he asks. On his first night as an undergraduate in Cambridge, he witnessed the powers of drugs again when he went drinking with fellow students. Two of them couldn't stop. "I just watched them transform themselves. One of them started wailing and crying and the other became incredibly hostile."

    During his clinical training, Nutt says he treated many alcoholics but failed "to get anyone interested in how to reduce their addiction to the drug that was harming them." He set out to answer that question, first in the United Kingdom, later as the chief of the Section of Clinical Science at the U.S. National Institute on Alcohol Abuse and Alcoholism, a job he held for 2 years. Today, he runs the department of neuropsychopharmacology at Imperial College, using modern imaging techniques to see what happens in the brain when people take drugs or develop an addiction.

    But his biggest contribution to science, he says, was a discovery he made quite early in his career: that some molecules don't just block receptors in the brain, but actually have the opposite effect of the molecules that normally stimulate them—and in doing so shut down a brain pathway. Nutt called these molecules contragonists, and he has made a second career out of being a bit of a contragonist himself, trying to calm society's overexcited responses to the steady stream of alarming news about drugs.

    Fictional affliction

    In 2009, Nutt published an article in the Journal of Psychopharmacology comparing the harms from ecstasy with those caused by horse riding. Every 10,000th ecstasy pill is likely to hurt someone, he calculated, while an average horse enthusiast can expect a serious accident every 350 hours of riding. The sport, he concluded, was more dangerous than the notorious party drug. That "raises the critical question of why society tolerates—indeed encourages—certain forms of potentially harmful behaviour but not others such as drug use," he added.

    Politicians were not amused, and Nutt's whimsical reference to a fictional affliction he called equine addiction syndrome, or "equasy," did not help. In his book Drugs - Without the Hot Air, Nutt provided his account of a phone conversation he had with U.K. Home Secretary Jacqui Smith after the paper was published. (Smith calls it an "embroidered version" of their talk.)

    Smith: "You can't compare harms from a legal activity with an illegal one."

    Nutt: "Why not?"

    "Because one's illegal."

    "Why is it illegal?"

    "Because it's harmful."

    "Don't we need to compare harms to determine if it should be illegal?"

    "You can't compare harms from a legal activity with an illegal one."

    Nutt says this kind of circular logic crops up again and again when he discusses recreational drugs with politicians. "It's what we would call 'splitting' in psychiatric terms: this primitive, childish way of thinking things are either good or bad," he says.

    He's often that outspoken. He likens the way drug laws are hampering legitimate scientific research, for instance into medical applications for psychedelic compounds, to the church's actions against Galileo and Copernicus. When the United Kingdom recently banned khat, a plant containing a stimulant that's popular among people from the Horn of Africa and the Arabian Peninsula, he compared the decision with banning cats. And he accuses the Russian government of deliberately using alcohol to weaken the opposition. "However miserable they are, however much they hate their government and their country, they will just drink until they kill themselves, so they won't protest," he says.

    But it's his stance on cannabis that got him sacked. In early 2009, ignoring advice from Nutt's advisory council, Smith upgraded cannabis from class C to class B, increasing the maximum penalty for possession from 2 to 5 years in prison. A few months later, Nutt criticized the decision in a public lecture, arguing that "overall, cannabis use does not lead to major health problems" and that tobacco and alcohol were more harmful. When media reported the remarks, Alan Johnson, who succeeded Smith as home secretary in mid-2009, asked him to resign. "He was asked to go because he cannot be both a government adviser and a campaigner against government policy," Johnson wrote in a letter in The Guardian.

    Nutt did not go quietly. With financial help from a young hedge fund manager, Toby Jackson, he set up a rival body, the Independent Scientific Committee on Drugs, "to ensure that the public can access clear, evidence based information on drugs without interference from political or commercial interest." Politics have skewed not just drug laws but research itself, he argues. "If you want to get money from the U.S. government to work on a drug, you have to prove it damages the brain," he says.

    One of his favorite examples is a paper that Science published in September 2002. The study, led by George Ricaurte at Johns Hopkins University, seemed to show that monkeys given just two or three doses of ecstasy, chemically known as MDMA, developed severe brain damage. The finding suggested that "even individuals who use MDMA on one occasion may be at risk for substantial brain injury," the authors wrote. The paper received massive media attention, but it was retracted a year later after the authors discovered that they had accidentally injected the animals not with MDMA but with methamphetamine, also known as crystal meth, which was already known to have the effects seen in the monkeys. Nutt says the mistake should have been obvious from the start because the data were "clearly wrong" and "scientifically implausible." "If that result was true, then kids would have been dropping dead from Parkinson's," he says.

    Some resent this combative style. "He is a polarizing figure and the drug policy area is polarized enough," says Jonathan Caulkins, a professor of public policy at Carnegie Mellon University in Pittsburgh, Pennsylvania. But Jürgen Rehm, an epidemiologist at the Centre for Addiction and Mental Health in Toronto, Canada, says Nutt has helped stimulate debates that were long overdue. "You don't get to be on the front page of The Lancet and The New York Times unless you sharpen your arguments a little bit," Rehm says. "I can live with that."

    Ranking the drugs

    In 2010, Nutt sparked a new firestorm when he published another comparison: a Lancet paper ranking drugs according to the harm they cause. Nutt and other experts scored a long list of drugs on 16 criteria, nine related to the user, such as death from an overdose or wrecked relationships, and seven related to society, such as drug-fueled violence and economic costs. In the end, every drug was given a score between 0 and 100 to indicate its overall harm. Alcohol came out on top, ahead of heroin; mushrooms and ecstasy were at the low end (see graphic).

    Critics said the study's methodology was flawed because it didn't address drug interactions and the social context of drug use. "For instance, the number of fatalities caused by excessive alcohol use is going to depend in part on gun control laws," says Caulkins, who calls the whole idea of expressing drug harm as a single number "embarrassing."

    Caulkins adds that even if a perfect ranking of drug harms were possible, it wouldn't mean that politicians should put the tightest control measures on the most harmful drugs. Suppose drug A is more harmful to the individual and society than drug B, he says, but impurities in drug A, when illegally produced, can lead to potentially fatal organ failure while they just taste bad in drug B. If you were going to prohibit only one of the two drugs, it should be drug B, he says, even though it causes less harm per se, because criminalizing drug A would lead to a more dangerous product and more deaths. Nutt's ranking of drugs, he says, is "a pseudoscientific exercise which is trying to take control of the policy process from a technocratic perspective in a way that isn't even sound."

    Other scientists defended the paper. Using Nutt's harm scales, "flawed and limited as they may be, would constitute a quantum leap of progress towards evidence-based and more rational drug policy in Canada and elsewhere," two Canadian drug scientists wrote in Addiction. Regardless of its quality, the paper has been hugely influential, Rehm says. "Everyone in the E.U. knows that paper, whether they like it or not. There is a time before that paper and a time after it appeared."

    Nutt says his comparisons are an essential first step on the way to more evidence-based drug policies that seek to reduce harm rather than to moralize. The best option would be a regulated market for alcohol and all substances less harmful to the user than alcohol, he argues.

    That scenario, under which only heroin, crack cocaine, and methamphetamine would remain illegal, seems unlikely to become a reality. But Nutt says he can already see more rational policies taking hold. Recently, Uruguay and the U.S. states of Colorado and Washington legalized the sale of recreational cannabis, going a step further than the Netherlands, which stopped enforcing laws on the sale and possession of small amounts of soft drugs decades ago. Nutt was also happy to read President Barack Obama's recent comment that cannabis is less harmful than alcohol. "At last, a politician telling the truth," he says. "I'll warn him though—I was sacked for saying that."

    New Zealand, meanwhile, passed a law in 2013 that paves the way for newly invented recreational drugs to be sold legally if they have a "low risk" of harming the user. Nutt, who has advised the New Zealand government, is delighted by what he calls a "rational revolution in dealing with recreational drugs." The main problem now, he says, is establishing new drugs' risks—which is difficult because New Zealand does not allow them to be tested on animals—and deciding what "low risk" actually means. "I told them the threshold should be if it is safer than alcohol," he says. "They said: 'Oh my god, that is going to be far too dangerous.'"

    Safer substitute

    Nutt agrees that alcohol is now one of the most dangerous drugs on the market—which is why he's trying to invent a safer substitute. The World Health Organization estimates that alcohol—whose harms range from liver cirrhosis, cancer, and fetal alcohol syndrome to drunk driving and domestic violence—kills about 2.5 million people annually. "When I scan the brains of people with chronic alcohol dependence, many have brains which are more damaged than those of people with Alzheimer's," Nutt says.

    In a paper published this month in the Journal of Psychopharmacology, Nutt and Rehm summarize the top six interventions that governments should consider to reduce the harms of alcohol, such as minimum prices and restrictions on the places that can sell hard liquor. They also argue that governments should support the development of alternatives. Nutt points to e-cigarettes—devices that heat and vaporize a nicotine solution—as a model. "In theory, electronic cigarettes could save 5 million lives a year. That is more than [the death toll from] AIDS, malaria, tuberculosis, and meningitis put together," he says. "I would argue that the e-cigarette is going to be the greatest health invention since vaccination."

    Scoring drugs.

    Nutt and colleagues at the Independent Scientific Committee on Drugs in London ranked 20 drugs according to how harmful they are to the individual user and to others, expressed as a number between 0 and 100. They deemed crack cocaine the most harmful drug to the user and alcohol most harmful to U.K. society.


    Can an alcohol alternative do the same? "I think that idea is utopian," says Spanagel, the German psychopharmacologist. One reason is that researchers have recently developed a much more complex picture of what ethanol, as chemists call it, actually does. Twenty years ago, they thought that once it reached the brain, alcohol elicited its many effects by infiltrating the membranes of neurons there and changing their properties. "Now we know that's nonsense. You would have to drink 5 liters of schnapps for that to happen," Spanagel says.

    In fact, scientists have learned that alcohol, like other drugs, interacts with the receptors for certain neurotransmitters. But unlike other drugs, it acts on a wide range of them, including receptors for GABA, NMDA, serotonin, and acetylcholine. That will make it hard to find a substance to emulate most of alcohol's wanted effects while avoiding the unwanted ones, Spanagel predicts.

    Nutt is concentrating on the GABA system—the most important inhibitory system in mammalian brains. Alcohol activates GABA receptors, effectively quieting the brain and leading to the state of relaxation many people seek. Nutt has sampled some compounds that target GABA receptors and was pleasantly surprised. "After exploring one possible compound I was quite relaxed and sleepily inebriated for an hour or so, then within minutes of taking the antidote I was up giving a lecture with no impairment whatsoever," he wrote in a recent article.

    But he wants to go one step further. "We know that different subtypes of GABA mimic different effects of alcohol," he says. Nutt combed the scientific literature and patents for compounds targeting specific GABA receptors, and, in an as-yet unpublished report that he shared with Science, he identifies several molecules that he says fit the bill. Compounds targeting subtypes of the GABAA receptor called alpha2 and alpha3 are particularly promising, he says. Some of these molecules were dropped as therapeutic drug candidates precisely because they had side effects similar to alcohol intoxication.

    Gregg Homanics, an alcohol researcher at the University of Pittsburgh, is skeptical that another substance could mimic all the positive effects of alcohol. "You could come up with a drug that might make you feel good. But is it going to be the same good feeling as alcohol? I doubt that." Such a drug might have downsides of its own, warns Andreas Heinz, an addiction researcher at Charité University Medicine Berlin. It could still turn out to be addictive or to harm a small proportion of the population. "There is an advantage when you have known drugs for hundreds of years and you know exactly what they do," he says.

    Still, Nutt's appearance on the BBC radio program attracted new investors, ranging "from Ukrainian brewers to American hedge funds," he says, and Imperial Innovations, a company that provides technology transfer services, is working with him "to consider a range of options for taking the research forward," a spokesperson says. "We think we have enough funding now to take a substance all the way to the market," Nutt says—in fact, he hopes to be able to offer the first cocktails for sale in as little as a year from now.

    Even a very good alcohol substitute would face obstacles. Many people won't forsake drinks they have long known and loved—such as beer, wine, and whiskey—for a new chemical, Spanagel says. The idea will also trigger all kinds of political and regulatory debates, Rehm says. "How will such a new drug be seen? Will you be able to buy it in the supermarket? In the pharmacy? Will society accept it?"

    Whatever the outcome, Nutt's quest for a safer drink has already made people think about alcohol in a new way, Rehm adds. "It's provocative in the best sense of the word." Much the same could be said of the scientist who thought it up.

  5. Cryptography

    Quantum Spy Games

    1. Adrian Cho

    Someday a quantum computer may unlock all our secrets—if it isn't first stymied by simpler technologies.


    When the news broke earlier this month that the U.S. government is trying to build a superfast quantum computer, reports suggested the demise of private information as we know it. “[T]he National Security Agency is racing to build a computer that could break nearly every kind of encryption used to protect banking, medical, business and government records around the world,” The Washington Post said in a front-page story on 3 January.

    The end may not be so nigh, however.

    “People are not actually shaking in their shoes so much,” says Lance Fortnow, a computer scientist and blogger at the Georgia Institute of Technology (Georgia Tech) in Atlanta. The $80 million National Security Agency (NSA) initiative came to light in documents leaked by former NSA contractor Edward Snowden, but experts say they're not surprised that NSA is working on quantum computing. “It would be shocking if they weren't,” Fortnow says. But a full-fledged quantum computer is likely decades away.

    More important, even if a quantum code cracker can be built, it might be defeated by encryption algorithms already in the works—or by another technology, called quantum key distribution, that relies on quantum mechanics itself for security. “In the future, we believe our adversaries will have better computers and better algorithms, but they won't be able to break the laws of physics,” says Richard Hughes, a physicist at Los Alamos National Laboratory in New Mexico.

    Which raises a practical question: If such countermeasures rob a quantum computer of its widely perceived killer app, will anybody put up the huge sums of money likely needed to make a quantum computer a reality?

    The threat

    Be it data file or love letter, information crosses the Internet in the form of numbers, written in binary strings of 0s and 1s. Cryptography is the mathematics of scrambling a numerical message so that it can be unscrambled only by the intended recipient, at least in a reasonable amount of time. A quantum computer could crack current algorithms for so-called asymmetric or public key encryption, the technique typically used to begin secure Internet communications.

    In such schemes, a sender, Alice, scrambles a message to the recipient, Bob, using a numerical key that Bob makes public. Bob also possesses a private key that mates with the public key to unscramble the message. In principle, an eavesdropper, Eve, could deduce the private key from the public key. In practice, that problem is so hard it would take nearly forever to solve.

    For example, in the widely used RSA algorithm, the public key is a huge number that factors into two prime numbers known only to Bob. To encode a message, Alice multiplies the binary message by itself a number of times—that is, raises it to a power—that's based on those factors and specified by Bob. She divides the result by the public key and takes the remainder—an operation called modding out. (For example, 33 mod 15 equals 3.) She sends the remainder to Bob as the coded message. To unscramble the message, Bob raises it to another power that depends on the two primes. That power is his private key. He then mods out by the public key. Voilà! The original message pops out.

    Confusing? It's supposed to be. But basically it's as if Alice has a counter that resembles a car's odometer, except that it rolls over to zero when the count equals the public key. She dials in the original message and scrambles it by advancing the counter according to a certain recipe, letting the counter roll over however many times it will. Bob then advances the counter in a way that makes the original message roll around again.

    The algorithm is secure because, without knowing how many times the counter has rolled over, Eve can't readily reverse Alice's manipulations. She may have a better shot at figuring out Bob's private key—the power to which he's going to raise the coded message. But that requires factoring the public key. As the key is huge—typically 2048 bits—that task would overwhelm an ordinary computer.

    But it would be easy for a quantum computer. An ordinary computer employs “bits” that can be set to either 0 or 1; a quantum computer would use “qubits”—perhaps spinning ions or little loops of superconductor—that could be set to 0, 1, or, thanks to the weirdness of quantum mechanics, 0 and 1 at the same time. The qubits' state would be described by quantum waves that can interfere to reinforce or cancel one another, making new algorithms possible. In 1994, Peter Shor, a computer scientist at the Massachusetts Institute of Technology (MIT) in Cambridge, reported an algorithm that uses such “quantum interference” to factor huge numbers in far fewer steps than an ordinary computer needs. Shor's algorithm could crack not just RSA, but all current public key schemes.

    Building a quantum computer, however, is easier said than done. Researchers must keep the qubits isolated enough to maintain their delicate quantum waves, yet make them interact enough to perform a calculation. That's a monumental challenge. Quantum computing has been a hot topic since the 1990s, but physicists are still struggling to manipulate a handful of qubits. The best anyone has done with Shor's algorithm has been to factor 21.

    One company, D-Wave of Burnaby, Canada, sells a more limited type of supposedly quantum computer that cannot run Shor's algorithm. But for the most part, quantum computing remains a basic research subject, says Addison Snell, an industry analyst with Intersect360 Research in Mountain View, California. “If I were a big company like IBM, I'd have a couple of guys in the back room working on this,” he says.


    Even if a quantum computer remains a remote threat, some researchers are already working to counter it. That's because it's not just future communications that might be jeopardized, but also messages we're sending today, which could be cracked after a quantum computer becomes a reality. “How do you know that somebody hasn't recorded all those communications?” says Georgia Tech's Fortnow. “The Internet doesn't forget so fast.”

    The easiest safeguard would be to improve cryptography. A quantum computer couldn't defeat every type of encryption. In symmetric or private key systems, Alice and Bob share a secret key and a public algorithm that, for example, makes two different chunks of coded message look like random bits when Eve compares them. In that case, Eve can't hack a transmission by looking for correlations in the stream of bits. Such schemes seem capable of withstanding a quantum attack and already encrypt most of the data transmitted in secure Internet transactions.

    New public key algorithms might also fend off a quantum computer. For example, Chris Peikert, a cryptographer at Georgia Tech, and colleagues are working on so-called lattice methods. A lattice is a repeating array of dots like the atoms in a crystal. In three dimensions, each distinct crystal lattice can be characterized by three arrows or “basis vectors” that can be added to and subtracted from one another repeatedly to reconstruct the structure. Describing a lattice in 1000 dimensions requires 1000 basis vectors.

    Given only the basis vectors for a 1000-dimensional lattice, which two points in the lattice are closest together? That is, what is the shortest vector that can be made from the basis vectors? That problem appears to be hard, even for a quantum computer, Peikert says. So he's developing schemes in which the basis vectors serve as the public key to encode a message, and the shortest vector serves as the private key to decode it. “In terms of having something that is real that can be measured and optimized, we are either there or very close to there,” Peikert says.

    Perhaps the best defense against a quantum computer is another quantum strategy: quantum key distribution. In that technology, Alice passes Bob a secret key encoded in individual photons. The photons can be polarized—say, horizontally for 0 and vertically for 1. Thanks to quantum theory, Eve can't measure the photons without altering them and revealing herself.

    A handful of companies already sell quantum key distribution systems. For example, ID Quantique in Geneva, Switzerland, has roughly 100 systems in use; its clients include the Canton of Geneva, which protects election results with the technology. Current systems cost about $100,000 for two stations linked by an optical fiber. And they're not yet ideal: A message sent across a network must be decoded at each station along the way, where it might be intercepted.


    As a code cracker, a quantum computer might be foiled by simpler quantum key distribution.

    To overcome that weakness, researchers hope to use the photons not simply to send the key, but to create a mysterious quantum connection called entanglement between qubits in 0-and-1 states at either end of a link. Alice and Bob could then generate a secret key by measuring their qubits: Alice would measure hers, causing its state to “collapse” into either 0 or 1. She'd know that Bob's had collapsed into the same state. Entanglement could be passed from node to node, so that Alice and Bob could share the key without danger of spying at the nodes between them.

    That vision requires a device called a quantum repeater at each node. Such a thing does not yet exist, but it should be easier to make than a quantum computer, says Gregoire Ribordy, co-founder and CEO of ID Quantique: “Which is good, because it means we will be able to do [fully] quantum communications before encryption is broken.”

    More immediate dangers

    Even without a quantum computer, public key systems are under pressure because of improvements in algorithms and increases in computing power. When RSA was invented in 1977, mathematician and writer Martin Gardner challenged readers to crack a message encoded with a 426-bit key. Ronald Rivest, a computer scientist at MIT and one of the inventors of RSA, estimated that it would take 40 quadrillion years to do so. Seventeen years later, 600 volunteers decoded the message in 8 months. To fend off such challenges, Internet companies such as Google and Microsoft are now phasing out 1024-bit RSA keys in favor of tougher 2048-bit ones.

    But going further and switching to new algorithms would be hugely expensive. Widespread quantum key distribution would require rewiring networks with optical fibers. Without a dramatic threat, the technologies may not be worth the hassle. “The old joke is that one of the uses for quantum computing is to create a market for quantum key distribution,” says Scott Aaronson, a theoretical computer scientist at MIT.

    The economic calculus could also cut the other way, however. If quantum-proof algorithms and quantum key distribution take off, then they might blunt the desire of governments and private companies to build a quantum computer. “It could happen,” Ribordy says. “It's a good argument.”

    Aaronson argues that a quantum computer will have other uses and that the real reason to build one is to prove it can be done. “That's a harder case to make to many business people, but I think that's the honest case,” he says. It also doesn't sound like a goal that NSA would spend $80 million to reach.

Log in to view full text

Via your Institution

Log in through your institution

Log in through your institution