# News this Week

Science  24 Nov 2006:
Vol. 314, Issue 5803, pp. 1224
1. GLOBAL WARMING

# U.N. Conference Puts Spotlight on Reducing Impact of Climate Change

1. Richard Stone,
2. John Bohannon*
1. The reporting of Stone and Bohannon was supported in part by the Reuters Foundation.

NAIROBI—For the past 6 years, Louis Verchot has had a ringside seat for Lake Victoria's ecological decline. Intense rainstorms pounding down on degraded land have swept in millions of tons of phosphorus-laden sediments from the Nyando River, transforming the lake from a nutrient-limited ecosystem into one with a gross excess of nutrients. On a visit last spring, says Verchot, a soil specialist at the World Agroforestry Centre in Nairobi, the water was so choked with an algal bloom that a glass of it “looked like spinach soup.”

Verchot can't do anything about the torrential rains. But to help communities in western Kenya's Lake Victoria Basin mitigate the damage, he's spearheading a project with the Kenyan Agricultural Research Institute, funded by the Global Environment Facility (GEF), to reforest denuded land with acacias and other indigenous trees and to help farmers switch to sustainable agricultural practices. It will be a long haul, says Verchot, “but we think we will be able to help them out.”

Victoria's downward spiral is a stark example of how climate change—shifting patterns of rainfall in this case—and poor resource management have conspired to create an ecological nightmare. The countries most vulnerable to these effects are also those least able to adapt to the changes, U.N. Secretary-General Kofi Annan told the U.N. Climate Change Conference in Nairobi last week. “Innumerable African communities have suffered climate-related disasters in recent years,” he said. “For them, adaptation is a matter of sheer survival.”

One clear message from the Nairobi meeting is that the need to adapt to climate change is finally being taken seriously on the world stage. Until now, the debate on climate change has been dominated by the epic dispute over how to stem greenhouse gas emissions, says Jon Barnett, an environmental sociologist at the University of Melbourne, Australia. “But we know that even if we completely stopped emissions tomorrow, there are already enough [greenhouse gases] in the atmosphere that more global warming is inevitable,” he says.

Here at the annual U.N. conference of nations that have ratified the landmark 1990 Kyoto Protocol, which binds parties to sharp limits on greenhouse gas emissions, delegates fleshed out an Adaptation Fund that will funnel assistance—eventually amounting to hundreds of millions of dollars—to developing countries that bear the brunt of climate change. But disagreement over who will control the money—GEF or the countries that the fund is designed to help—will delay implementation until next year's meeting at the earliest. “This will be one of the most important debates that the next conference will have,” says Ian Noble of the World Bank.

The fund could be a huge boost to nascent efforts to adapt to climate change. Emerging problems run the gamut from shifting disease patterns and droughts to coastal erosion from rising sea levels. Without adaptation, the World Bank forecasts that climate-change impacts in vulnerable developing countries could cost up to $100 billion per year over the coming decades. One new initiative described at the meeting aims to build climate adaptation into global public health. The World Health Organization (WHO) estimates that climate change is already causing at least 150,000 excess deaths per year. One major killer is malaria. Here in Kenya, some 20 million people are at risk as warmer average temperatures allow the mosquito that transmits malaria to spread into the highlands, says Solomon Nzioka of Kenya's Ministry of Health. “We've established that we have something to be concerned about,” says WHO's Diarmid Campbell-Lendrum. “Now we're at the critical point: telling people what to do about it.” For malaria spread, measures could include more aggressive mosquito control at higher altitudes and stepped-up vaccine R&D. WHO and the U.N. Development Programme have launched a pilot project in seven countries—Barbados, Bhutan, China, Fiji, Jordan, Kenya, and Uzbekistan—with different health vulnerabilities to climate change. Last month, for example, Chinese officials agreed to explore ways to reduce fatalities from heat waves, which are estimated to cause between 225,000 and 890,000 excess deaths per year from strokes and heart attacks in China, says Jin Yinlong, director general of the National Institute for Environmental Health and Engineering in Beijing. “We will be judged on how well we protect people's lives as climate change evolves,” says Campbell-Lendrum. Scores of other projects are getting off the ground. The World Bank is spending about$50 million on adaptation projects, and bilateral programs have committed $110 million to more than 50 projects in 29 countries. Even the United States, which has not ratified the Kyoto Protocol, is getting in on the adaptation action: The U.S. Agency for International Development has promised$2 million for such projects over the next 5 years. Still, “we are orders of magnitude underfunded,” says Alf Wills, South Africa's chief climate negotiator at the conference.

Globe-spanning adaptation efforts are necessary, says Barnett, but there are also immediate priorities on a very local scale. Take the Pacific island nation of Niue, the smallest in the world. Intensification of tropical cyclones and rising sea levels “could wipe the nation off the map within decades in the worst-case scenario,” Barnett says. Luckily, he says, some quick-fix adaptations could make a big difference. “For a start, half the population needs to be relocated to higher ground,” he says. That, along with improvements in infrastructure to help islanders cope with climate-related problems, “comes to a ballpark figure of $60 million.” Considering that what is at stake is an entire nation with its own unique language and culture, says Barnett, “this is incredibly cheap.” 2. DEVELOPMENTAL BIOLOGY # Teams Identify Cardiac 'Stem Cell' 1. Jennifer Couzin* 1. With reporting by Gretchen Vogel. Like many organs, the heart is a patchwork of cell types, from smooth muscle that pulses blood through arteries to endothelial cells lining vessels. These pieces, varied as they are, were long considered distant cousins born of different parent cells. But two new studies have uncovered a primitive type of heart cell in mice that can give rise to the heart's main cell lineages. If the finding holds up, it will make the heart one of very few organs, along with the blood, known to grow largely out of a single type of cell; it may also ease the introduction of embryonic stem cell treatments in cardiac patients. “It's surprising that so much can come from” just one type of heart cell, says Timothy Kamp, who studies cardiovascular regenerative medicine at the University of Wisconsin, Madison. “You have essentially a type of cardiac stem cell.” Although they took different approaches, the two groups that found the heart progenitor cells both identified overlapping genetic markers to define their progenitor population, and both found that the cells could differentiate into cardiac muscle and blood vessel cells, the principal building blocks of the heart. The first paper, led by Gordon Keller, a stem cell biologist at Mount Sinai School of Medicine in New York City, was published earlier this month in Developmental Cell; the second appeared this week in Cell. That work was led by a husband-and-wife team, Karl-Ludwig Laugwitz and Alessandra Moretti, at the Technical University of Munich in Germany, and Kenneth Chien at Massachusetts General Hospital in Boston. The Chien team found that mouse embryonic stem cells developing into heart cells first entered an intermediate state that could be monitored by tracking expression of three different genes. Those intermediates, which the scientists called “triple positive cells,” gave rise only to heart cells. To confirm that these triple positive progenitor cells, grown under artificial conditions, exist in an animal, the researchers examined mouse embryos at different points in their development. Around day 8, they detected them. Although Keller's team did not use all the same markers as Chien's to characterize the cells it found, both groups found that their cells could differentiate into the same cardiovascular cell types. “We're arriving at a similar progenitor,” says Keller, also adding that “it's still pretty early days.” To prove that these progenitor cells can become functioning, specialized heart cells, the scientists need to inject them back into an animal to see whether they give rise to the different cardiac tissue types, Moretti notes. That is also a key experiment to determine whether these master ancestor cells can repair a damaged heart. Keller's group has begun precisely this experiment, inserting the progenitor cells it identified into mice whose hearts resemble those of humans following a heart attack. Chien notes that “we have not formally proven that that cell can make a whole heart.” Still, says Kamp, the work could ease one of the most worrying concerns about using embryonic stem cells in patients: that, left alone to form whatever cell type they fancy, they'll develop into tumors. “If you can have a more committed cell population that can only give rise to limited progeny,” Kamp says, “that's going to dramatically reduce the risk.” And the cells might still be flexible enough to form, say, a coronary artery, which includes different cell types. Still, admits Laugwitz, that “remains to be proven.” Both groups, in the United States and Germany, are working with human embryonic stem cells to see whether the mouse patterns will hold. 3. FUSION # Scientists Reap ITER's First Dividends 1. Daniel Clery Japanese researchers were disappointed when they lost a bid last year to host the$12 billion International Thermonuclear Experimental Reactor (ITER) project. But they should be cheered by the consolation prize: In an agreement due to receive provisional approval this week, some $870 million will be spent on fusion-related facilities in Japan, with equal contributions from Japan and the European Union. European researchers are happy too, as most of Europe's contribution will be in-kind, and the whole effort will speed the work toward a commercial fusion power reactor. The need to compensate the runner-up “has turned necessity into advantage for the fusion program,” says Günter Janeschitz, head of fusion at Germany's Karlsruhe research center. The origins of the deal lie in the frantic diplomacy in 2004 and 2005 during which the then-ITER partners—China, the European Union, Japan, South Korea, Russia, and the United States—tried to decide between sites at Rokkasho in Japan and Cadarache in France. In an effort to win support for their sites, both Japan and the E.U. upped their offers to pay as much as 50% of the total ITER cost if they were host. “There was a lot of money on the table,” says Chris Llewellyn Smith, director of the U.K. Atomic Energy Authority's Culham Laboratory. The idea emerged that this extra money could be used to build the International Fusion Materials Irradiation Facility. IFMIF uses neutrons similar to those inside a fusion reactor to test and validate materials that would be used in a commercial prototype that comes after ITER, dubbed DEMO. Building IFMIF now rather than later would speed the transition to DEMO. Once the ITER site deal was completed in June 2005 and negotiations on what is known as the “Broader Approach” began, there was not as much money on the table—not enough to build IFMIF, anyway. And Japan had other priorities: It wants to rebuild its existing fusion reactor, the JT-60, with superconducting magnets. This would create a mini-ITER where operational scenarios could be tested and refined. Japan also wants to build an International Fusion Energy Research Center at Rokkasho, which will house a supercomputer for simulations and lead the effort to design DEMO. According to an E.U. official involved in the negotiations, Europe's only requirement was that the Broader Approach contain an engineering design effort for IFMIF so that construction could start about 6 years from now. In the agreement presented this week,$190 million is earmarked for IFMIF design. But according to the E.U. official, in an official letter Japan made clear that even though it would lead the design effort, it did not necessarily want to host the machine. The E.U. offered to be the host if no others came forward. “Personally, I'm very happy with this result,” says Llewellyn Smith. “IFMIF is on the road.”

Following this week's initialing of the Broader Approach agreement, both sides will check it through, aiming to sign it by the end of the year. Also up for signing this week is the main ITER agreement, which will mark the creation of the international organization that will build the machine.

4. SHERWOOD BOEHLERT INTERVIEW

# Explaining Science to Power: Make It Simple, Make It Pay

1. Jeffrey Mervis

After 24 years of serving a House of Representatives district in upstate New York, including the last six as chair of the House Science Committee, Sherwood “Sherry” Boehlert will retire next month from the U.S. Congress. A self-proclaimed “cheerleader for science” on a panel that lacks the power of the purse strings, the moderate Republican sought common ground among both conservatives within his party and Democrats across the aisle on a range of issues including tougher environmental standards and undergraduate science education.

The 70-year-old Boehlert is uncertain about his next step—ruling out an afterlife as a Washington lobbyist but hoping to remain active on national science and environmental issues. But before packing up, he sat down last week with Science's Jeffrey Mervis in his Capitol Hill office to reflect on the nature of government and what role scientists can play.

Q: How well do scientists get their message across to politicians?

On the 24 years I've been on the House Science Committee, I'd say they've gone from a D-minus to a solid B. They're beginning to appreciate that politics is a different realm. …

When you talk to Congress, you have to appeal to the interests of the audience that you're dealing with. To talk about some great advance in pure scientific terms isn't enough. What does it do to strengthen the economy, or enhance competitiveness, or provide more jobs?

I'm a typical congressman, with a bachelor's degree in public relations and no science background, yet I ended up on the science committee. And I say that's the perfect place for me because I ask the obvious questions: Why can't we do this? Why won't this work? I make them think in more practical terms.

Q: How important is the economic argument, and does every project have to have one?

You have to remember that this is representative government, and I'm sent here to exercise my best judgment on the important issues of the day. So if you want me to exercise my best judgment, then you have to prove to me that it has some public benefit besides a bunch of Ph.D.s sitting in a laboratory coming up with something that they can publish that no one can understand. I mean, what's the real benefit?

Q: What would it take for scientists to get an A?

You have to do more advocacy, and the people who are good at it have to train their colleagues. … I have a theory that to be an eminent scientist, you have to invest a lot of time and resources in getting a good education, including a Ph.D., and then you publish a lot of papers. Then suddenly, one day, you have arrived, and people who are aware of your vast knowledge begin to beat a path to your door. And they want to listen to you, so the scientists get used to giving tutorials. But then they want to come to Congress and give tutorials. That doesn't work. We don't have time for tutorials. They need to get right to the point: “This is why it's important. I know there are a lot of competing interests, but here's why we should be at the head of the line. And here's what it means for society.”

Q: Some scientists are starting to endorse candidates and raise money for individual campaigns. Good idea?

I don't think that's the way to go. A lot of scientists don't even want to get involved in politics because they think that it's dirty.

I'll bet you that if you look at all the new freshmen, you won't find a single one, from either party, who campaigned on something like the American Competitiveness Initiative, or more resources for NSF [National Science Foundation], or greater investment in science and math education. I'll bet you won't find one. And that's a failure by the scientific community.

Why aren't they more involved? It's not about raising money—although there's certainly a lot of money in politics. Why aren't they visiting candidates and explaining to them, on their home turf at the university in their district, why they should be really interested in their agenda? I tell scientists that their new best friends should be these new congressmen. Don't just visit them in Washington with a lobbyist. Invite them to come to the university in their district, not to a technical presentation that they probably can't understand, but to a general discussion of what's going on and what it means. … I think that the scientific community will be an abject failure if, when these new freshmen start campaigning for reelection, at least a few of them don't have a science component in their platform.

Q: If you became a lobbyist, with professional societies as your clients, what would you tell them to do, and where would you take them?

Of course I would come to the Hill, and to the Science Committee, and to the appropriations committees. But I'd also tell them to get their people back home to come here. Because a person from North Dakota coming to see a congressman from upstate New York is not nearly as persuasive as someone from his district.

Q: What science agencies are most effective at getting their message across, and how do they do it? For example, does it work when the National Science Foundation invites legislators and their staffs to Antarctica?

You're damn right it does. Because there's no substitute for kicking the tires. I've had two trips to Antarctica, and in the last one [January 2006], I was part of a bipartisan group of 10 members. Of that 10, there were probably two who shared my view that global climate change was real and that we damn well better do something about it. The rest were skeptical or neutral. But after we got back, every one of them had a heightened interest in the subject.

Why? Because down at the South Pole, they heard from scientists about how their experiments related to global climate change. The same thing happened at the Great Barrier Reef in Australia, where we heard how this great treasure was being damaged because of something called global change. And the next time there's a floor vote on the budget of some science agency supporting research on climate change—and I won't be around—I'll bet that this group will be a more receptive audience because they've seen it firsthand.

What are we supposed to do—sit in our offices and read these reports? Like hell. We need to get out in the field and see the facilities. McMurdo Station is not a place I'd suggest as a vacation spot. But we spent 5 days on the ice, and we learned a lot.

Q: Over your career, which science agency heads were the best at getting their message across?

One of the best is Mike Griffin, the current NASA administrator. He understands his audience. I don't need a translator to deal with him, even though I'm a generalist and he's a distinguished scientist.

[Former NSF Director] Erich Bloch is another, without question. In each case, they clearly know their stuff. They know how to make their argument and explain why it's deeply and intensely important to them in a way that is important to the nation. It doesn't do any good if the intended recipient doesn't understand what you're talking about and is looking at their watch, wondering about their next appointment. … To this day, when people think of the ideal NSF director, Erich is who they talk about.

Q: Is the president well-served by his current science adviser, and is science being coordinated effectively across all federal agencies?

Here's the problem. The president has a lot of people vying for his attention. And quite honestly, whether it's this president or Bill Clinton before him, science isn't given the attention it deserves because there's not the sense of urgency that the secretary of defense or the secretary of state bring to the table. And [George W. Bush] has a natural passion for education, which gives the secretary of education an edge. So while we've had capable and fine people as directors of OSTP [Office of Science and Technology Policy], it's not considered a top-tiered adviser to the president, and the director doesn't get the face time that the other secretaries receive. …

So yes, I think that the science adviser should have greater access to the president. But there have been improvements in this Administration. For example, when Mitch Daniels was [Office of Management and Budget] director, for the first time the science adviser was brought into the budget negotiations with all the science agencies. I think that was an important step.

Q: Speaking of budgets, do you think that the next Congress will curb academic earmarks?

I think so. I think you'll see less rather than more, and that trend is good.

Q: Voluntarily?

Are you kidding? You're going to ask the people who benefit from this practice to stop voluntarily? But I think there's general agreement that earmarks have gotten out of hand, and that something needs to be done.

5. U.S. INTELLECTUAL PROPERTY

# Patent Experts Hope High Court Will Clarify What's Obvious

1. Eli Kintisch

Thomas Deuel thought his 1990 discovery of the purified DNA sequences that code for a cellular growth factor called pleiotrophin was sufficiently new and different to deserve a patent. The U.S. Patent and Trademark Office (PTO) respectfully disagreed. Citing “the routine nature of cloning techniques,” PTO concluded that what the cell biologist had done in his lab at Washington University School of Medicine in St. Louis—purify, characterize, and obtain the DNA that codes for a protein—was “prima facie obvious.” But Deuel appealed and won, with a special federal court declaring in 1995 that the patent office's view of what was common knowledge was based on “speculation and an impermissible hindsight.”

Determining what is not obvious—one of the four tests that U.S. inventors must meet to receive a patent—has always been an inexact science, and for nearly 2 centuries, PTO's examiners had wide latitude to disqualify patents on that basis. But in the past 3 decades, the Court of Appeals for the Federal Circuit has restricted their scope with cases such as Deuel's. Next week, the U.S. Supreme Court will hear oral arguments on a landmark case, KSR International Co. v. Teleflex Inc., that could decide whether the current high standard for rejecting a patent based on obviousness should be lowered.

The U.S. high-tech community is deeply divided over the issue. Most computing and technology firms hope the high court will back a broad definition of obviousness, which would give PTO more leeway to reject what the companies consider to be undeserving patent applications. In the past, they argue, such patents have led to expensive court battles and unpleasant business surprises. In contrast, the biotech and pharma sectors want the court to maintain what they see as a continued flow of legitimate innovations to preserve a healthy biomedical industry. Three dozen groups, as diverse as AARP and the Michelin tire company, have filed briefs on one or another side of the debate.

Law professor John Duffy of George Washington University in Washington, D.C., who represents KSR, calls nonobviousness “the heart of what is a patent.” To win patent protection, an idea or object must be new, useful, and properly described. The law also requires that a patentable idea would not have been obvious at the time of invention to a hypothetical “person having ordinary skill in the art.”

Making that call is one of the toughest decisions that an examiner faces. It's not because of ignorance. All of PTO's 282 biotech examiners have advanced science degrees to inform their decisions; 63% have Ph.D.s. Yet federal judges, as in the Deuel case, have steadily narrowed definitions of obviousness, making it harder for the examiners to apply their expertise. “We had been rejecting those kinds of claims,” says Esther Kepplinger, who was a supervisor in the biotechnology examiner corps when Deuel submitted his application. She says that the examiners were “startled that the court would have said this was not obvious.”

## More than common sense

The question before the high court next week began as a standard infringement case. In 2002, Limerick, Pennsylvania-based Teleflex, a manufacturer, sued KSR, an Ontario, Canada-based firm that makes brake pedals, for patent infringement. It won before the federal circuit court, and KSR appealed to the Supreme Court, which decided earlier this year to take the case. At issue is whether Teleflex's 2001 patent, which combines an adjustable and electric pedal, was obvious and should not have been granted.

In a 1966 precedent-setting case involving plow parts, the high court gave examiners the power to “ascertain” or “determine” obviousness without much definition of the term. Patent lawyers say that gave examiners wide latitude to issue rejections. But since its 1982 founding, the federal circuit has established more direct instructions to PTO: An existing specific teaching, suggestion, or motivation for a combination of elements is required to declare a patent claim obvious. “Common sense” does not “substitute for authority,” the court said in 2002. Two years later, a federal court ruled that a patent on a drug combining the painkillers Vicodin and ibuprofen was invalid as obvious. But the federal circuit reversed that decision because there was “no record of evidence … suggesting the enhanced biomedical effect of the combination.”

Critics say such decisions have driven PTO to issue bad patents that hurt consumers and innovators alike. “Anyone who's been sick knows you can put two analgesics together to fight pain,” says Jeffrey Light of Washington, D.C.-based Patients not Patents, which joined with AARP on KSR's side. Such patents, says Light, “lead to higher costs” for consumers and choke competition. And they hurt truly innovative scientists, adds Duffy, who represents KSR: “Follow-on patents can rob the pioneering patents of their just rewards.”

Defenders of the status quo, including the Biotechnology Industry Organization in Washington, D.C., say the high court shouldn't jeopardize a reliance “on factual findings” that has allowed the U.S. research enterprise to flourish. And Kevin Noonan, a patent attorney with McDonnell Boehnen Hulbert & Berghoff LLP in Chicago, Illinois, fears giving examiners, whose expertise varies greatly, too much say in the obviousness call. “Do we really want whether someone gets a patent to be based on what examiner they get by the luck of the draw?” he asks.

The federal circuit itself may even be rethinking the issue. Last month, in what its critics welcome as a new tack, it declared that its obviousness standards are “quite flexible” and require “consideration of common knowledge and common sense.”

Last year, the high court avoided taking any dramatic steps to overhaul the patent system in cases dealing with the patentability of scientific concepts and the legal power of a granted patent. But critics are hopeful that the nine justices will now act forcefully to fix a flaw they think is more central to patent quality. “Obviousness is getting closer to the root of the problem,” says Josh Lerner of Harvard Business School in Boston, an outspoken opponent of the current regime. “KSR is potentially huge.”

6. U.S. INTELLECTUAL PROPERTY LAW

# Government Questions Sequencing Patent

1. Eli Kintisch

A decades-old patent application could rewrite the history of who invented the automated DNA sequencer.

Last week, the U.S. Patent and Trademark Office (PTO) decided that a 1982 application from Enzo Biochem, a small New York biotech company, covers the same invention named in a 1998 patent awarded to former California Institute of Technology biologist Leroy Hood and colleagues. Hood's patent, owned by the California Institute of Technology (Caltech) in Pasadena, covers sequencing using gel electrophoresis—the technology currently underpinning the $7 billion DNA sequencing industry. PTO's decision to begin what's called an interference procedure follows decades of efforts by Enzo's lawyers to win a patent. At stake are presumed millions of dollars in royalty income for Caltech and the fiscal health of sequencing giant Applied Biosystems in Foster City, California, which licensed Hood's technology in a majority of its machines. Applied Biosystems, with fiscal 2006 sequencing-machine revenue of$540 million, has previously fought off other attacks on the intellectual property it owns or licenses.

Attorneys say the announcement itself marks a victory for Enzo, which last fiscal year recorded losses of $15.7 million. But the company's chances of success are hard to determine. Caltech's attorneys, who declined to comment on the matter, are expected to claim that PTO erred in deciding that Enzo's application covers Hood's invention, although a copy of the typed 1982 version does mention the procedure. At some point, the two sides will also bicker over who invented what first—with the answer hinging on yet-to-be-disclosed lab notebooks and calendars. The whole process, which could include a subsequent trial and appeal, could last 5 years or longer, says interference specialist R. Danny Huntington of Bingham McCutchen LLP in Washington, D.C. Caltech's patent expires in 2015. If Enzo wins and receives a patent with a later expiration date, Applied Biosystems would have to pay additional royalties to use the technology. At the same time, a patent on gel electrophoresis could be less important by then, notes George Church of Harvard Medical School in Boston, because scientists are steadily moving toward new methods of sequencing DNA. Techniques include using pores or solid surfaces to cut costs or sequence genes faster (Science, 17 March, p. 1544). 7. U.S. SCIENCE POLICY # Resignations Rock Census Bureau 1. Constance Holden Knowledgeable observers of the U.S. Census Bureau are shaking their heads over the sudden resignations last week of Director Louis Kincannon and his deputy and chief census statistician, Hermann Habermann. It's “time for me to retire,” wrote Kincannon in a 14 November letter to President George W. Bush, who appointed him to the post in 2002. But there are widespread rumors that the men were pushed out. The resignations come amid stepped-up preparations for the 2010 Census, the first one that will use only a short form. The agency is also facing a possible$58 million cut in its 2007 budget, which is still pending in Congress, that would jeopardize the new American Community Survey, ongoing monthly sampling designed to substitute for the old long form in the decadal census.

The 66-year-old Kincannon told Science he's leaving as soon as his successor is in place because he wants to spend more time with his grandchildren in Tennessee. But in other news reports, he noted that his relationship with his bosses at the Department of Commerce had deteriorated since the departure last year of Donald Evans as Commerce secretary. Habermann declined to comment.

Commerce spokesperson Dan Nelson says, “It was mutually agreed that the time was right” for the departures. But Edward Spar, director of the Council of Professional Associations on Federal Statistics, says he is certain that Kincannon was asked to resign and that Habermann, a “consummate statistician” whom he sees regularly, “had no plans to leave January 3 [his stated departure date]. … I still don't understand the actual reason.”

A former Census official who asked not to be quoted by name believes that some Republicans in Commerce and on Capitol Hill are concerned that Democrats will revive efforts to adjust census numbers to make allowance for undercounts of poor people—who are likely to vote Democratic. To counter that attempt, he says, those officials want compliant leadership at the bureau.

But former census director Kenneth Prewitt, now a professor at Columbia University, says those fears are unfounded. “I am absolutely certain that the current [Census Bureau] leadership does not want to adjust the census,” asserts Prewitt. A House Republican staffer told Science he is satisfied that no one wants to revive the idea of an adjustment, which the bureau formally rejected in 2003. Other sources say Habermann, who is responsible for day-to-day operations, was the primary target after resisting pressure to appoint partisans to career posts.

8. CHINESE GENE THERAPY

# Splicing Out the West?

1. Jerry Guo,
2. Hao Xin*
1. Jerry Guo is a writer in New Haven, Connecticut.

Chinese researchers have been the first to put cancer gene-therapy products on the market, but critics question the data behind the success stories

BEIJING—Maria Corina Roman, a Danish surgeon, made international news when she decided to seek treatment for her breast cancer using the world's first commercial gene therapy. Disappointed with standard cancer treatment, Roman flew to China in 2004 to try Gendicine, a Chinese product that contains a virus with a human tumor suppressor gene (p53) spliced into its DNA. Just days after the first injection, Roman reported that she had regained energy and appetite. Gendicine's maker, SiBiono GeneTech Co. in Shenzhen, spread the word. Encouraging reports about this gene therapy appeared in the Financial Times, Business Week, and China Daily.

This fall, however, Roman's tumor has returned, SiBiono acknowledges. The company's chief executive, Peng Zhaohui, says nevertheless that the drug has proved to have “good efficacy,” adding that Roman, SiBiono's most famous client, “should continue to treat with Gendicine.”

Peng's advice is based on more than optimism; it reflects national policy. China's State Food and Drug Administration (SFDA) approved Gendicine for clinical use in October 2003 and licensed its commercial production in spring of 2004. Last year, SFDA approved a second genetically engineered anticancer product: a modified virus, dubbed H101, designed to infect and kill cells containing mutated versions of the p53 gene. The maker, Sunway Biotech Co. in Shanghai, says it expects to strike a licensing deal by the end of this year with Genzyme Corp. in Cambridge, Massachusetts, to run clinical trials of a Genzyme gene-therapy product in China and possibly test H101 in the United States.

As these projects advance in China, gene therapies in North America and Europe are struggling to complete premarket clinical tests. After a U.S. patient died in a 1999 genetherapy trial and two children in French trials developed leukemia in 2002, the U.S. Food and Drug Administration (FDA) tightened controls on experiments, says James Norris, head of the U.K.-based International Society for Cell & Gene Therapy of Cancer. Western companies say they are making progress but have not yet brought a single gene therapy to market.

Some see this as a sign that China is catching up with, or even surpassing, the West. “I think the future of gene therapy will be in China,” says Andre Lieber, a genetherapy researcher at the University of Washington (UW), Seattle. But he warns that recent claims of success should be read with caution. There is a “problem” with interpreting clinical studies done in China, Lieber says. Often the primary data are published only in Chinese—raising a barrier to nonspeakers—and even when they appear in English, critical information may be missing (see sidebar, p. 1233).

Intellectual-property rights may be problematic, too. Some researchers in the West have questioned claims of independent innovations made by Chinese drug companies; this could limit sales outside China. Finally, critics argue that the Chinese regulatory system is not rigorous and that Gendicine, for one, was approved with scant evidence of efficacy. With drugs to treat cancer, “the bar is a lot lower than in the United States to get approval,” says Frank McCormick, director of the University of California, San Francisco, Comprehensive Cancer Center.

## High hopes

On a plot of land in the outskirts of Shenzhen stands an empty building with opaque windows, a site where owners hope a biotech bonanza will blossom. Starting next year, this newly constructed plant will begin producing 1.5 million vials of Gendicine per year, seven times the capacity of SiBiono's current facility, according to SiBiono's Peng. Science visited Peng in his office in May and spoke with him last month by phone.

A hallway at the company's headquarters is plastered with clippings from Chinese and international media describing how Gendicine has helped cancer patients. Peng said SiBiono aims to spearhead the sale of gene-therapy products in China with Gendicine. It was given its Chinese name—jin you sheng, “born again today”—by China's Vice President Zeng Qinghong when he made a ceremonial visit to the company a month before SFDA cleared the drug for market.

SFDA approved Gendicine as a treatment for head and neck cancer based on small clinical trials showing that more patients had tumors disappear with Gendicine plus radiotherapy (64%) than with radiotherapy alone (19%). Peng has called these “phase II/III” trials, an unusual term that combines safety (phases I and II) with proof of efficacy (phase III).

In 2005, SFDA approved Sunway's H101, also designed for treatment of head and neck cancer, after a 160-patient phase III clinical trial showed that 74% of patients receiving H101 plus chemotherapy experienced a reduction in the size of tumors compared to 40% of patients receiving chemotherapy alone.

Gendicine has now been given to more than 4000 patients to treat not just head and neck tumors but also 50 different cancers, Peng claims. The venture thus far has received about $6 million in grants and government start-up funds as well as$6 million from private investors.

Peng projected in 2004 that 50,000 patients would have received Gendicine treatment by the end of 2006. Demand is far short of that target, but if the drug works—and if patients can afford the high price of treatment, costing $1680 to$3360 per cycle—the market could eventually be huge. “Having 1.3 billion potential patients compared to 300 million in the United States makes a successful drug very lucrative in China,” says Norris.

## Imitation or innovation?

Doubts persist, however, about China's future as a gene-therapy powerhouse. Some U.S. companies allege that China's commercial products are spinoffs of Western inventions with relatively minor modifications. Introgen Therapeutics in Austin, Texas, for example, claims that SiBiono's Gendicine is similar to its own experimental product, a recombinant adenovirus containing the human p53 gene (rAd-p53).

Wei-Wei Zhang, president and CEO of San Diego-based GenWay Biotech, published the first paper on rAd-p53 while working at the University of Texas M. D. Anderson Cancer Center in Houston in 1994. He holds U.S. patents on the viral construct and related processes. M. D. Anderson negotiated a license with Introgen, which has spent more than $70 million to develop a product based on Zhang's rAd-p53, trademarked Advexin. It has been in clinical trials since 1994. The company's ongoing phase III trial using Advexin to treat head and neck cancer is under review for “accelerated approval” by FDA. Introgen's 106-patient phase II trial in 2005 showed a 10% “tumor response rate,” defined by at least 30% reduction in tumor size, in patients who received Advexin alone. Introgen Vice President Robert Sobol says phase III trials are going well. Meanwhile, Introgen CEO David Nance claims that Gendicine is a “derivative” of his company's product. In an August 2006 filing with the U.S. Securities and Exchange Commission, Introgen claims that Gendicine infringes on a 1994 patent filed in China but concedes that “enforcement of patents in China is unpredictable, and we do not know if monetary damages could be recovered from SiBiono.” Peng disputes these statements. In a phone interview, he said that Gendicine is “very different” from Introgen's product, and that the only similarity is the use of p53. Sunway acknowledges that its product, H101, was inspired by U.S. research but says it developed H101 independently—a claim that is not disputed. According to Sunway officials and other observers, H101 is similar to a product called Onyx-015, made by Onyx Pharmaceuticals Inc. in San Francisco. Onyx-015 and H101 both use a modified adenovirus to target probable cancer cells that have a deficient or mutated p53 gene. This so-called oncolytic virus, which has been tested in U.S. phase I and II clinical trials, is designed to replicate in target cells and kill them. Onyx never filed for a patent on Onyx-015 in China. Nevertheless, Sunway CEO Hu Fang says that in developing H101, “we followed almost everything Onyx did in clinical trials. … We modified the virus, very little, for patent purposes.” Although Onyx-015 has shown in phase II trials that it also can achieve local shrinkage of head and neck tumors of about 60% to 70%, McCormick, a co-founder of Onyx, says this was not enough to win FDA approval. Regulators wanted more evidence, specifically data showing that Onyx-015 prolonged survival. Onyx ended a phase III trial when the main backer pulled out in 2005. At this point, Sunway obtained exclusive worldwide rights from Onyx to use the 015 modified virus in H101. “We bought the patent from Onyx because now we want to put our drug in Europe, the United States, and Japan,” says Hu. The distribution network will be ready soon, and Hu expects 2000 patients to sign up in the first year. The company is working on an improved version, H103, that includes a heat shock protein designed to attack metastatic tumors by inducing an immune response. ## Different standards The Chinese government is both an investor in and a regulator of biotech projects such as the ventures that produced Gendicine and H101. Some observers, including Norris, are concerned that the government's dual role could weaken its vigor as an enforcer of standards. He notes that “backers of these companies are high-level government officials.” From 2001 to 2005, the Ministry of Science and Technology (MOST) provided$106 million to innovative drug development, some of which went to SiBiono.

SiBiono's Peng also helped write a regulatory guidebook for SFDA on evaluating cancer gene-therapy products. Leaning forward in his executive chair, Peng proudly shows off a thin pamphlet. “It's the most systematic guidelines in the world, and I was the main framer,” Peng exclaims. There's an appearance of a conflict of interest in this, Norris says, although the government's acceptance of help with regulatory guidelines may reflect a wish to catch up quickly with standards in developed countries.

Peng acknowledges that SiBiono has government support and confirms that the application for Gendicine was sped “through a special channel.” The data from the Gendicine trials were submitted to SFDA in March 2003; the drug was approved 7 months later. Sunway also “pushed” to get its H101 application through in 10 months, Hu confirms. But companies can also apply for accelerated review at the U.S. FDA, and Peng argues that Chinese companies must comply with strict regulations, just like their counterparts in the West.

Yin Hongzhang, SFDA's chief of biological products, says the agency has “special policies” to approve a drug on the fast track if an initial technical review looks fine. “But we would require the manufacturer to do further research and collect more data on efficacy to submit” after approval, he says. Earlier this year, he asked SiBiono to send the required follow-up data; when he spoke with Science he was still waiting for the data.

China's regulatory framework differs in another way. Whereas the U.S. FDA often requires that novel cancer drugs extend the life of the patient to be judged a success, SFDA approved both Gendicine and H101 on the basis of tumor shrinkage.

Sunway's Hu says his company intends to show that H101 increases survival as well as shrinks tumors. “Survival time for patients is very important,” says Hu. In a retrospective study, he says the company has found that H101 can provide a 7-month survival benefit, but the results were not significant. They are now repeating phase III trials with a bigger sample size and more treatment cycles designed to maximize survival benefit.

There is good reason to expect that Chinese biotechnology will have a bright future. Companies in China “have excellent production facilities, a lot of money, and a lot of good people,” says UW's Lieber. Zhang adds that Chinese bioscientists deserve credit for picking up U.S. pioneers' work in cancer gene therapy.

At least a half-dozen Chinese gene-therapy drugs are in clinical trials at the moment, says Savio Woo, past president of the American Society of Gene Therapy. “Before the end of this decade, they should have more drugs. I will be surprised if they didn't,” he says. China also may draw significant outside investment to the field. Genzyme, for example, is negotiating to have Sunway run a phase II genetherapy clinical trial in China. The U.S. company is testing a modified adenovirus construct (Ad2/HIF-1α) to promote angiogenesis in patients with peripheral arterial disease, an immobilizing condition that decreases blood flow to the muscles. Already, Genyzme has enrolled 300 patients in Europe and the United States. “The climate in China is changing, with more innovative companies not just focused on manufacturing,” says Genzyme Vice President Earl Collier Jr. “We want to participate.”

Zhang nevertheless worries about “media hype” that could “mislead patients, officials, and investors and cause significant damage to the further development of China's biotech industry.” He hopes China can avoid repeating the mistakes that set back gene therapy in the West.

9. CHINESE GENE THERAPY

# Gendicine's Efficacy: Hard to Translate

1. Hao Xin*
1. With reporting by Jerry Guo.

Clinical data supporting China's advances in gene therapy often appear in Chinese-language journals—which are inaccessible to many Western readers. To bridge the gap, James Wilson, editor of Human Gene Therapy (HGT), last year solicited a review in English summing up published clinical evidence behind China's first gene-therapy product, Gendicine, by Peng Zhaohui, CEO of SiBiono GeneTech in Shenzhen, the company that put Gendicine on the market (see main text).

Peng's review in the September 2005 issue of HGT has been cited at least a dozen times by experts as a definitive view of Chinese clinical trial results. However, Marshall Posner, medical director of the Head and Neck Oncology Program at the Dana-Farber Cancer Institute in Boston, says that, after reading translations of the original reports, the findings are hard to evaluate. The trials “were not done with a high degree of structure, and it is not clear what protocols were followed or how patients were randomized,” Posner says. Others question the quality of the data.

Comparing Peng's summary with original Chinese-language reports, Science found that the summary did not include some information in the originals. For example, Peng described patients in a phase I (safety) clinical trial of Gendicine as having “advanced” cancers. But a Chinese-language report said seven of the 12 participants in this trial had limited primary tumors that had not spread to lymph nodes. Although the original paper reported that all 12 patients received surgery along with gene therapy, Peng's summary of therapeutic effects mentioned only treatment with Gendicine, noting that 11 patients who received it had a remission of cancer lasting more than 3 years. In a telephone interview, Peng said that he had inadvertently omitted data on the surgeries.

Peng's review discussed so-called phase II/III trials of Gendicine in 2001–2002, citing three primary publications. But the primary papers reported only phase II trials—relatively modest ones that had enrolled a total of 124 patients. (Phase III trials are larger and demonstrate efficacy.) Another flaw, says Anthony Chan, chair and chief of service of the Department of Clinical Oncology at Prince of Wales Hospital in Hong Kong, is that these trials—which compared Gendicine plus radiotherapy to radiotherapy alone for head and neck cancer—is that “the definition of complete response … was not provided,” even though it is especially difficult to define in such cases.

China's State Food and Drug Administration (SFDA) approved Gendicine for production in 2004 without data from a standard phase III trial. Peng's explanation: SFDA did not require such trials for new drug approvals before May 1999, and because “our clinical trials were approved before 1999, we were not required to do phase III trials.” Peng adds that this is “okay” because “the SFDA approved our drug on safety and efficacy.”

10. # HERVÉ THIS PROFILE: The Joy of Evidence-Based Cooking

1. Martin Enserink

Molecular gastronomist Hervé This is trying to demystify cooking in a country whose cuisine is famous worldwide

PARIS—Is it true that pears turn red in covered copper pans lined with tin? Do you always have to whip cream in the same direction? Does the skin of suckling pigs really get more crackling when the head is cut immediately after roasting? What of the old French wisdom that mayonnaise, a delicate emulsion of oil and water, will fail when prepared by menstruating women?

Such are the questions that occupy the mind of French celebrity scientist Hervé This, who studies the science of cooking. This (pronounced “Teess”), who has dual appointments at the National Institute for Agronomic Research (INRA) and the Collège de France, wants to know whether common rules of cooking are science-based or just bogus. (The answers to the above questions, in case you are wondering, are no, no, yes, and no, respectively.)

This is the most prominent spokesperson of a small but growing research field known as “molecular gastronomy,” or, as famed food science writer Harold McGee from Palo Alto, California, puts it, “the science of making delicious things.” He studies what happens in pots, pans, and ovens to create that divine flavor and texture. And in the process, he's trying to give cooking a more solid scientific basis, which means getting rid of some age-old wisdoms.

That may seem like a hard sell in a country where tradition reigns, especially in matters relating to food. Yet This has been remarkably successful. A series of books, columns, and TV appearances, as well as his close ties to some famous chefs, have made him a household name in France; his efforts to introduce science into culinary schools and to acquaint children with science through cooking have met with enthusiasm. Even those who criticize his scientific output concede that This has been a remarkably effective spokesperson for both science and culinary innovation.

Although trained as a physical chemist, This, 51, started his career in 1981 as an editor at Pour la Science, a popular science magazine. But he was crazy about cooking, had his own lab at home, and very often wrote about food. In 1995, chemist and Nobel laureate Jean-Marie Lehn asked This to join his chemistry lab at the Collège de France, a job This initially combined with his work at the magazine. But when he was offered a job at INRA as well in 2000, he quit his editing job to become a full-time researcher.

Although the science of cooking has existed for centuries, the field matured, and unmistakably picked up cachet, thanks to a series of now-legendary annual gatherings between 1992 and 2003 at a resort in Erice, Sicily. This organized the meetings with physicist Nicholas Kurti, a pioneer in cooking research at Oxford University who died in 1998. Participants would discuss the science behind food preparation, occasionally cook, and invariably eat and drink well for about 4 days. “It was a place where Nobel scientists and three-star chefs came together, indulging in a hobby, if you will,” says Anthony Blake, a retired flavor expert who attended several times.

Kurti and This coined the term molecular gastronomy as they prepared the first meeting, in part because it sounded modern and sexy. Since then, the name has stuck as a way to distinguish the small group of researchers who study restaurant and home cooking from the larger, older, and less glamorous f ield of industrial food chemistry. But McGee—another frequent guest at Erice—considers it a misnomer, because scientists in this field don't study the interaction of individual molecules like molecular biologists do; it's just food chemistry, he says. (This disagrees.)

To add to the confusion, the term molecular gastronomy is also widely used to describe the cuisine at some creative top restaurants that have their own labs, such as elBulli, 2 hours from Barcelona, which was named the world's best restaurant by Restaurant magazine this year. Actually, elBulli chef Ferran Adrìa has invented most of his revolutionary techniques—such as the use of hydrocolloids and agar-agar to create new textures—without the help of scientists, says McGee. And Adrìa resa resents the fact that so many press stories link him to the scientific field; scientific curiosity is just one of the many elements of his cooking, his says.

## Deconstructing stock

On a recent afternoon at his Collège de France lab, one of This's co-workers was making a carrot stock. Stocks may be commonplace in the kitchen, This explains, but they are still something of a scientific mystery. This has studied exactly which compounds come out of the carrot to give the liquid its flavor—sugars and amino acids, mostly—but he also wants to know how this happens. Are they released as cells in the carrot burst open? Or do they simply diffuse out of the channels in the carrot? And does it make a difference whether you simmer for 2 or 20 hours?

One of This's obsessions is that chefs, despite knowing so little about science, have developed such elaborate laws. Over the years, he has meticulously collected more than 25,000 instructions, called précisions in French, from cookbooks, many of which are useless, he says. So where do they come from? “Our parents love us. Why are they teaching us all these rules that make no sense?” His hypothesis: Cooks, using trial and error, remembered the circumstances in which they created a successful dish, even if they were irrelevant, and made them part of the recipe.

If that's true, he says, then dishes prone to fail—such as mayonnaise—should have accumulated more précisions than the easy ones; in other words, there should be an inverse relation between what This calls the recipe's “robustness” and the number of précisions. Testing the theory for a number of different dishes, This did indeed find the predicted relation—although there was one outlier, meat stock, which is hard to blow yet surrounded with précisions. (This chalks it up to stocks' extraordinary importance in French culinary culture.)

This's ambition is to do away with all unnecessary instructions and the wasted time they entail. If each of France's 500 culinary schools tested four précisions a year, an idea he is now promoting, the job could be done in just over 10 years, he says. Not everybody is equally fascinated. “I'm not sure I'd spend so much time studying misunderstandings of the past,” says McGee. But food scientist Erik van der Linden of Wageningen University in the Netherlands says investigating these old wisdoms is “hugely important” because it can lead to new scientific questions.

Resistance from the culinary world can be strong, however: For instance, several chefs balked when This told them that it's useless to throw cooked haricots verts into ice water to preserve the fresh green color. “They thought that the cold fixated the chlorophyll,” says This. “Chemically, that doesn't mean anything.”

In another attempt to bring rigor to the messy process of cooking, This has developed a system for “classification of dispersed systems,” which describes each dish as a formula, based on the state of its ingredients (gas, liquid, or solid) and the preparation process. (In this system, puff pastry becomes ((S1/S2)0.5σ ((W/O)/S3)0.5)σ729.) The formulas—a bit like those Lavoisier developed to describe chemical reactions—can be used not only to classify dishes, This says, but to invent new ones as well. “He's the first one ever to try that, and it's something to be proud of,” says Van der Linden.

Although he says he's more interested in research than in cooking, This does have close ties with a three-star chef, Pierre Gagnaire of the eponymous restaurant in Paris. Every month, This sends him an idea from the lab—for instance, an egg cooked at 65°C, which is far less rubbery than those cooked at 100°—which Gagnaire then turns into a recipe. (The entire collection is available on Gagnaire's Web site.)

Meanwhile, This is tirelessly campaigning to promote his f ield. His CV lists 600 interviews and press conferences—until he stopped keeping track. His lectures are enormously popular—“I've always thought of him more as a showman than a scientist,” Blake says—and his columns are published in 11 journals and magazines in France and abroad. At the request of former culture minister Jack Lang, This developed a science and cooking class for schoolchildren in 2001, which is still running. (“A great way to make them love chemistry,” he says.) He has just started a Foundation for Food Science and Culture at the prestigious Académie des Sciences.

“He is really effective and wonderful as a popularizer, and that's very important,” says McGee. And if more chefs follow This's lead and become a tad less loath to forgo tradition, he adds, France might have less trouble fending off newcomers such as Spain and the United Kingdom that are threatening its position as the world's best country for eating.

11. SCIENCE FUNDING

# Italy's Research Crunch: Election Promises Fade

1. Susan Biggin*
1. Susan Biggin is a writer in Trieste, Italy.

Critics say no-growth agenda could leave Italian science isolated in Europe

Italy's researchers are bracing for a tough year ahead. The 2007 national finance bill, which is creeping through the legislature and is due for signature by 31 December, would provide no growth for cash-starved universities and research centers. Indeed, some centers are facing cuts as deep as 13%. But the bill does create new research jobs and makes small allocations to selected research budgets, mainly in response to protests. Also included are administrative “reforms,” which have been greeted with both hope and suspicion.

Researchers are feeling the pinch because the center-left government of Prime Minister Romano Prodi, elected in May, is under pressure to reduce the country's deficit. The chief of the university and research ministry, Fabio Mussi, who says he is trying to avert a crunch, has warned already that 2007 will be “a lean year for everyone.” Appeals for more funds are coming from students, university rectors, institute heads, and eminent scientists. In a widely reported plea for new research positions, Nobel laureate Rita Levi-Montalcini said during a recent debate: “Italy is poor in raw material but rich in human capital. If it's destroyed, Italy can't help but sink.” In response, the government came up with a small hike to cash already earmarked for 2000 new posts over the next 3 years.

But the finance bill is a huge disappointment to scientists. During the election, Prodi's team campaigned on a pledge to hike research spending from the current level of 1.1% of gross domestic product to 3% by 2010. Such a boost would have put Italy in line with European Union (E.U.) goals for creating a knowledge economy (Science, 7 April, p. 37). Mussi has now set his sights lower: “reaching 1.5% within 5 years.” Fabio Pistella, head of the National Research Council (CNR), says that the “incredible” cuts of 13% he is facing will mean the council can't even cover salaries. “Italian research runs the risk of being completely left out of the E.U.'s Framework 7 initiatives,” he warns, and Piero Benvenuti, chief of the National Institute for Astrophysics, fears the loss of “the predominant role that Italian astrophysics has created for itself in the world.”

The institutional reforms in the bill, Mussi insists, are designed to improve transparency and remove “party politics.” A U-turn would restore autonomy to institutions such as the National Institute for the Physics of Matter, which was incorporated into CNR by former science minister Letizia Moratti. Another change has already separated the education and research ministries, reversing a merger carried out by the previous government.

The bill also includes a radical measure to remove research institution heads—viewed by many as political appointees—and set up committees to search for replacements on merit. Some scientists grumble that this measure would only increase government control. But others are encouraged. Carlo Bernardini, a physicist at the University of Rome “La Sapienza,” says the measures are a “gulp of oxygen” that could help science recover from the “business mentality” of the previous government. Along with other scientists, he is pleased that the government is backing a shift toward autonomy and accountability in research institutions. The Italian Space Agency is already being overhauled (Science, 10 November, p. 903).

In pushing for new research posts, Mussi recognized that the workforce needs rejuvenation, not just expansion. The average age of a newly appointed university ricercatore (researcher in the first career step) is almost 36, whereas the average age of all ricercatori is about 50. Only half the nation's estimated 108,000 academic staff have tenure, and 30,000 will be retiring in the next few years.

The academic appointment system itself is in for overhaul too. Currently, selection competitions—known as the concorsi—are run by individual universities. Under the new regulations, universities would still advertise their posts, but evaluation would follow national criteria. Procedures would be established to ensure transparency and speed up selections, and members of selection committees would be drawn from outside a university making an appointment. Successful candidates would be cleared for specific universities only on the go-ahead of a new assessment agency, ANVUR. This long-debated independent organ would have broad authority to evaluate the merit of research produced by institutions as the basis for distribution of new resources. Academic leaders are wary. Mussi has only sketched out his plan; members of the Accademia dei Lincei, an independent scholarly society, want to see the details. They are concerned about delays and paperwork inherent in centralized systems of review.

What's missing in the bill, says Aldo Schiavone, law faculty head at Florence University, is “a plan or list of priorities” for reforming the universities, a sentiment echoed by head of state Giorgio Napolitano, who has called for a “courageous reform” of the entire university system. But that's not in the cards this year.

12. EVOLUTION

# Two Rapidly Evolving Genes Spell Trouble for Hybrids

1. Elizabeth Pennisi

Evolutionary geneticists are pinning down pairs of genes that help promote speciation; these genes are rapidly evolving, but not in response to ecological pressures

New species arise when populations become separated and evolve along different paths until, eventually, their members can no longer breed successfully with each other. That was Darwin's revolutionary insight, and it has shaped our understanding of the natural world. But the underlying mechanism has been hard to pin down. Why, for example, do even closely related species have difficulty producing viable offspring? Hybrids, if they survive at all, tend to be less fit than their parents. And therein lies the crux of speciation.

Now, one group has nailed down a 70-year-old theory about why hybrids are usually doomed to failure. On page 1292, Daniel Barbash, a geneticist at Cornell University, and his colleagues report the identification of a pair of genes that are key to making two closely related fruit fly species reproductively incompatible. Other groups are closing in on genes that cause problems for hybrids in monkeyflowers and marine invertebrates called copepods. In each case, the genes appear to be evolving rapidly, implying that they are under selective pressure. It's the “beginning of a new phase in speciation research, where we can get at both the specific genetic mechanisms and [the] interactions underlying one of the most fundamental questions in evolutionary biology,” says Mohammed Noor of Duke University in Durham, North Carolina.

This work supports a theory first proposed in 1937 by Theodosius Dobzhansky and independently a few years later by Hermann Joseph Muller. They suggested that the root cause of hybrid failure is that pairs of genes whose proteins interact with each other—for instance, an enzyme and the protein it breaks down—evolve along different paths after populations split. In each population, the gene pairs evolve in concert so that their protein products continue to work together. But, said Dobzhansky and Muller, eventually the proteins in the individuals in one population will have changed so much that they no longer work properly with their former partners in the other population. When mixed back together in hybrids, these proteins are incompatible—an enzyme from one population will no longer break down the target protein from the other, for example—sand potentially lethal problems arise: Hybrids may be sterile or may not survive at all.

The Dobzhansky-Muller model gained wide acceptance. “It's really our best general model of how mutations can accumulate to cause reproductive isolation,” says Hopi Hoekstra, an evolutionary biologist at the University of California, San Diego. Confirming the details, however, has been challenging. “The problem is really, really hard because what you are trying to do is genetics between species,” says H. Allen Orr, an evolutionary geneticist at the University of Rochester in New York.

Over the years, researchers have found evidence supporting parts of the Dobzhansky-Muller model but not all of it. Typically, researchers find one gene but not its putative partner. For example, for decades, researchers have known that crossing two aquarium fish—a platyfish and a swordtail—has dire consequences. The offspring develop large black spots, and crossing the hybrid back to a parent often results in lethal skin tumors. Cancer researcher Manfred Schartl of the University of Würzburg, Germany, tracked down a causative gene, Xmrk2, on the X chromosome. He knows that it interacts with a “suppressor” gene that keeps Xmrk2 in check and suspects that Xmrk2 and the suppressor have diverged across the two species so they no longer interact effectively. However, to this day, the true identity of the suppressor remains unknown.

Drosophila researchers were also stumped for a long time. They could produce offspring by mating D. melanogaster with D. simulans, D. mauritiana, or D. sechellia, but too few offspring survived for researchers to carry out additional breeding experiments. Takao Watanabe came to the rescue in the 1970s when he discovered a mutant strain of D. simulans that could hybridize quite successfully with D. melanogaster. Watanabe, a geneticist at the National Institute of Genetics in Mishima, Japan, surmised that somewhere in its genome, the D. simulans strain carried a mutant gene that interacts successfully with a partner in D. melanogaster. He called the unidentified gene for “lethal hybrid rescue.” The finding “jump-started the field,” says Barbash.

In the late 1980s, Michael Ashburner and Pierre Hutter of the University of Cambridge uncovered evidence for a similar gene in D. melanogaster, calling it hmr for “hybrid male rescue.” They didn't know the exact location or identity of this gene, but crosses between hmr mutant strains and D. simulans worked just fine. With these strains in hand, researchers were able to produce viable hybrids, and they began modifying the genomes of the parents further to track down the specific genes involved in hybrid sterility and lethality.

## On to the genes

Barbash picked up where Watanabe and Ashburner and their colleagues left off. In 2003, he and his colleagues pinpointed and sequenced the D. melanogaster hmr gene and discovered that it was a transcription factor. A year later, he and Philip Awadalla of North Carolina State University in Raleigh and colleagues demonstrated that the hmr genes had indeed diverged functionally between the two species. When they put an intact copy of D. melanogaster hmr into the hmr mutant strain, hybrids with D. simulans died as larvae. But when they repeated the experiment with an intact hmr from D. simulans, hybrids survived, Barbash reported. When they compared the differences in 250 genes between the two species, they found that hmr was one of the most rapidly evolving.

With one gene that fulfilled Dobzhansky and Muller's expectations in hand, Barbash began to chase down its partner. He focused on lhr, as several earlier studies suggested that and hmr worked as a pair. The rough location of lhr was already known but not its identity. With the help of the newly generated genome sequence data for D. simulans, Nicholas Brideau, Jun Wang, and Heather Flores in Barbash's lab looked for genes whose sequence indicated that their proteins could interact with the hmr protein. They concentrated on one that had not only diverged quite a bit from its counterpart in D. melanogaster but is also mutated in Watanabe's D. simulans strain.

Brideau, Wang, and Flores designed an ingenious experiment to test whether they had the correct gene. They put the candidate lhrm gene from D. simulans into D. melanogaster and mated the resulting fruit flies with Watanabe's D. simulans strain. If the candidate gene was indeed lhr, its presence in D. melanogaster should override the mutant lhr in D. simulans and result in dead hybrids. It did. Barbash's group has confirmed that the lhr and hmr proteins interact. “We don't understand the mechanistic or molecular basis of the interaction,” Barbash says, “but both genes in combination are required to kill the hybrid.”

Scores of other incompatible gene pairs have likely evolved over the millions of years that fruit flies have diverged. Daven Presgraves, an evolutionary geneticist at the University of Rochester, is well on his way to pinning down a second pair. In 2003, after devising a way to screen for hybrid lethality genes, he turned up with one called Nup96, which codes for a protein that is part of the nuclear pore in eukaryotes. To begin to track down Nup96's partner, he and Wolfgang Stephan of the University of Munich, Germany, took a close look at five of the 30 other fruit fly pore proteins to see how they differed between D. melanogaster and D. simulans. To their surprise, all five are evolving quite fast, they reported online 20 October in Molecular Biology and Evolution. The screen Presgraves used to identify Nup96 detects only those genes whose interacting partner is on the X chromosome. Only one of the five other pore proteins, called Nup153, has that genomic address. “We are certainly hot on the trail” of pinning down Nup96's incompatible partner, says Presgraves.

Although much of the progress in identifying Dobzhansky-Muller gene pairs comes from fruit fly studies, researchers are starting to track down these genes in other species. In the monkeyflower, for instance, they have narrowed the search to relatively small chromosomal regions. In other cases, such as copepods, two genes are in hand, but their relationship is known primarily through test-tube studies and not through genetic analyses.

While a graduate student with John Willis at Duke University, Andrea Sweigart tracked down the cause of hybrid sterility in two closely related species of monkeyflower. One, Mimulus guttatus, is pollinated by insects, while the other, M. nasutus, is self-fertilizing. Both species occur in western North America but tend to grow in different habitats. Hybrids do form where they coexist, but the species maintain distinct identities, says Sweigart, now at the University of Rochester.

In 2001, Lila Fishman, now at the University of Montana, Missoula, and Willis showed that second-generation hybrids suffer from male sterility, suggesting genetic incompatibilities were at work. From extensive breeding and genetic mapping studies between the two species and between hybrids and the parental lines, Sweigart and Willis identified two places in the genome, called hms1 and hms2, where the incompatible genes are located, they reported in the April issue of Genetics.

Ronald Burton of the Scripps Institution of Oceanography in San Diego, California, has found two interacting genes that may be helping to isolate different populations of copepods, a Californian intertidal invertebrate. He and his students have found that the gene for the protein cytochrome c, which is important for electron transport and energy generation, varies across copepod populations. Test-tube studies indicate that these variations affect the efficiency of the protein's reaction with cytochrome c oxidase, suggesting that these two could be genetically incompatible in hybrids.

## Selective pressures

These new findings have thrown up some surprises. In particular, the genes behind hybrid lethality are evolving and adapting at an unusual pace compared to the rest of the genome. “Almost all these genes have a strong signature of natural selection,” says Hoekstra. Yet the genes seem unlikely candidates for rapid evolution. The lhr protein is associated with heterochromatin, the parts of chromosomes containing lots of repetitive DNA, and nuclear pores are conserved from yeast to humans. “You just don't expect those genes to evolve rapidly,” says Presgraves.

The fact that nucleoporin genes evolved quickly in species that are widely separated geographically suggests that ecological factors are not at the root of those gene changes, Presgraves adds. Indeed, notes Jerry Coyne, an evolutionary biologist at the University of Chicago in Illinois, “where the action is going to be is to [learn] what kind of natural selection is acting on these genes.” The answer is unlikely to take another 70 years.