Don't Label Me and the Tragedy of the Anticommons.


San Francisco, CA, Feb. 20—To label or not to label genetically modified food. That was the question at the Monday morning Biotechnology Communications symposium at the annual American Association for the Advancement of Science meeting in San Francisco. The non-labeling side was represented by Susanne Huttner, from the University of California at Berkeley. She's been in charge of agricultural biotech research and education for the UC system. The panel's labeler was Edward Groth, III, who is a senior scientist at Consumers Union.

Both agreed that ag-biotech offers a lot of benefits and that the technology is safe. Where they disagreed was over the issue of whether or not foods produced using biotech crops should be labeled. Huttner pointed out that mandatory labeling would mean the creation of two different food handling systems, one for conventional crops and another for biotech crops. That would add greatly to the expense of foods, while not providing any additional safety to consumers.

Groth agreed that the issue is not food safety–biotech crops are safe–but he framed the issue in terms of consumer choice. Biotech foods should be labeled so that consumers can choose to buy them or not. "Consumers have the right to make the choice to avoid biotech foods," said Groth. Groth was keen on achieving what he called "broad social consensus" on biotech crops, but in his presentation he made it clear what he thinks the "consensus" should be–mandatory labeling of biotech foods. Groth complained "We leave things to the market. What we need is a mechanism where we can decide what's best for society." He suggested, "We can learn from Europe" about how to achieve social consensus.

Huttner responded that regulatory agencies such as the Food and Drug Administration, the U.S. Dept. of Agriculture, and the Environmental Protection Agency are inappropriate arenas for trying to arbitrate between competing values. Regulatory agencies should do what they were created to do and that is concentrate on assuring the safety, quality, and efficacy of the products they regulate. "The marketplace is where people get to exercise their values," Huttner declared. "And the market has worked pretty well in that regard."

When questioned, Groth admitted that the new U.S. Department of Agriculture's organic food labeling regulations now give a choice to consumers who want to avoid biotech foods. "But they have to pay more for organic," Groth lamented. Of course, all of us will have to pay more for our food if mandatory biotech food labeling is imposed. Groth also admitted that he thinks that non-biotech conventional food producers will start labeling their foods as non-biotech if consumers start asking for such information. So, if there's a market solution like that, why does Consumers Union want mandatory labeling? "Institutionally, we just think it's better," said Groth.

In keeping with the day's science policy theme, I attended the afternoon symposium on patenting genes. The day before John Doll, who heads the biotech and pharmaceutical division at the U.S. Patent and Trademark Office, described the new guidelines, issued in January, that define utility for gene patents. In order for the PTO to issue a gene patent, the utility of an applicant's claimed invention must first be credible (e.g., no cold fusion or perpetual motion machines allowed). Second, it must be specific (e.g., can't just say the invention is an antibody that binds to an antigen; it has to specify the structure of both). And third, there is a new test that requires that an applicant show that his claimed invention's utility is substantial (e.g., you can't patent a transgenic mouse and say you developed it for snake food. You have to tell the PTO what substantial purpose the mouse will serve).

Stanford law professor Lawrence Lessig explained the general rationale for patents: If there is no legal protection for invention, he argued, then the incentives for inventors will be less than they would optimally be. "Patents are not about property but they're are not perpetual. Patents are government-backed monopolies," said Lessig.

"The patent system was not designed to promote basic research," clarified patent lawyer Kate Murashige. She pointed out that the Founders included patents in the Constitution "to promote the progress of the useful arts." By bringing the commercial world to the lab bench, it shouldn't be a surprise that it sometimes inhibits scientific research.

Brian Wright, an economist from the University of California at Berkeley, believes that patent "holdups" are becoming a problem, particularly in agricultural biotech. Crop biotech innovations are cumulative–that is one company might develop a nifty new pest-resistant seed, another might have a herbicide-resistant gene; while yet another has found a gene for drought tolerance. Naturally, a company will want to include all these features in their seeds. But this means that there must be a great deal of cross-licensing to achieve this. Holdups occur when holders of patents refuse to allow a commercial license that permits other companies to use their inventions.

The other problem is that the cumulative nature of these patents, combined with a lot of pending competing patent applications, make it commercially hazardous for companies to use these technologies. Wright cited the more than 70 different licenses that need to be obtained in the case of Vitamin A-enriched "golden rice" to illustrate the complexity faced by agriculture biotech companies. One way to address the licensing problem is industry consolidation and in fact, most of America's small seed companies have been purchased or merged with larger life sciences companies.

Citing the work of Rebecca Eisenberg, Wright believes that gene patenting right now is experiencing "The Tragedy of the Anti-commons." The classic tragedy of the commons occurs when nobody owns a resource and it is depleted because no one has an incentive to protect it. The tragedy of the anti-commons occurs when everyone claims to own a resource exclusively and refuses to allow anyone else to use it. In this case, according to Wright, a dysfunctional patent system, rather than encouraging innovation, is instead stifling it.

The new PTO utility guidelines might help sort out some of these problems, according to a none-too-optimistic Murashige, but that the devil is in the details of implementation.

Genomics, Socialized Medicine, and How Long Will You Live?

San Francisco, CA, Feb. 19—The best part first: Craig Venter, the head of Celera Genomics, the company that beat the publicly funded International Genome Project to the goal of sequencing the human genome, delivered the plenary speech Sunday evening at the American Association for the Advancement of Science's annual meeting in San Francisco. The contrast between Francis Collins, the head of the IGP, and Venter was made vivid by comparing the tones of their two plenary speeches on the human genome. Collins focused primarily on the problems that sequencing the genome might cause, while Venter focused on the opportunities.

Last June, Venter and Collins announced at a White House photo op that draft copies of the human genome had been completed. It appears that Collins' IGP was lagging behind Venter's efforts to complete a draft of the human genome; Collins basically brokered a deal with Venter that he could use the bully pulpit of the White House to announce his success if he would agree to give simultaneous credit to the public project. In a statesmanlike gesture Venter agreed. But Collins, Eric Lander at the Whitehead Institute, and other public project colleagues haven't forgiven Venter for showing them up. The New York Times reported last week that despite the White House agreement that Celera and the public consortium would publish their papers on the human genome simultaneously, the public consortium, behind the scenes, lobbied strongly to block Celera's publication in Science.

At the AAAS conference, populated largely by academic scientists (people who live and die by government largess) and journalists, Collins the bureaucrat is a hero and Venter the entrepreneur is, well, not a hero. For example, after Venter's speech I dined with several prominent theologians who groused about Venter's profit motive and expressed skepticism about his efforts to obtain patents from Celera's discoveries. I asked my dinner companions what a genetic patent actually was, and they responded that patents gave Venter ownership of human genes. The notion of a limited monopoly granted for 20 years for a useful and novel invention seemed beyond them. They didn't appreciate the point of patents is to encourage inventors to disclose information for the public benefit such that other people would be able to duplicate their inventions/discoveries. No trade secrets. The public project leaders complain that Venter used their public genome maps to guide the completion of Celera's version of the genome. They fail to mention that the public project eventually adopted many of the techniques for sequencing that Venter and Celera pioneered.

Anyway, Venter's plenary was an overview of the last 10 years of achievement. He had the temerity to remind the people who had opposed him and his ideas just how wrong they had been. In 1992, with the backing of venture capital, Venter left the National Institutes of Health to found The Institute for Genomic Research (TIGR) as a private research institute to look for genes. In 1994, TIGR submitted a funding proposal to the National Institutes of Health for a project in which TIGR would sequence in one year the entire genome for that laboratory favorite, the E. coli bacterium. The NIH refused to fund the grant on the grounds that TIGR's proposal couldn't possibly work. Venter published the E.coli sequence in Science a few months later.

When Venter approached the fruit fly consortium (another lab standard), the response was different; they enthusiastically embraced the proposal and in 9 months, the full sequence of the fruit fly was available. At the AAAS, Venter mused that given the improvements in technology, Celera could now sequence the fruit fly genome in six weeks and the bacterium H. influenzae (the first free-living organism TGIR–or anyone–ever sequenced) in one morning (down from nine months). Yeast, he said, would take most of a day to do now. He pointed out that more than half of the genomes sequenced so far have been done by him and his company.

Venter reminded the audience of what had been reported about the human genome, but he also went beyond those detail to discuss other findings. He pointed out that the partial genome of the chimpanzee, our closest evolutionary relatives, isn't very useful because "it's like having just another human genome." At this stage of genomic research differences between genomes tell scientists more than do similarities. Preliminary data have the chimp genome differing from ours by 1.3 percent. Venter also pointed out that the genome records the evolutionary history of our species. For example, chromosome 18 appears to be a double of chromosome 20, whose DNA has been added to by four additional events.

Venter noted that a variant version of the gene that codes for CCR5 receptor, which resists HIV infection, appears in 9 percent of Caucasians and in only 0.1 percent of blacks. The higher frequency of this HIV-resistant version in Caucasians may be because this version also helped European ancestors to survive the Black Plague when it swept through Europe 700 years ago.

Celera is now turning its attention to proteomics and is hoping to develop new cancer diagnostics based on detecting tell-tale proteins that would detect cancers at their inception. The night before, when Collins had finished speaking, the crowd leapt to its feet in approbation. When Venter finished, it took a moment but eventually the audience gave him his due, a rousing standing ovation.

That's how the evening ended, but the morning had begun with public policy.

"The ethical, legal, and social implications of the human genome project are profound," declared Rep. Louise Slaughter (D-N.Y.) at a morning session of the Genomics Seminar. Slaughter was at the AAAS meeting to plump for the bill she just introduced in Congress last week, H.R. 602, the Genetic Nondiscrimination in Health Insurance and Employment Act. The bill has some 154 bipartisan sponsors in the House and 17 in the Senate for a companion bill.

Among other things, Slaughter's bill would "prohibit insurers from restricting enrollment or changing premiums on the basis of predictive genetic information or genetic services and ban health plans and insurers from requesting or requiring that an individual take a genetic test, or reveal the results of such a test."

The bill would also "prohibit employers from requiring or requesting disclosure of predictive genetic information, and allow genetic testing only to monitor the adverse effects of hazardous workplace exposures." Both health insurers and employers would be subject to strict civil penalties for violations and insurance clients and employees could sue for damages.

The fact is that every human being has genetic glitches which will have health consequences. Francis Collins estimates that each of us typically has around 40 such glitches. One of the problem with Slaughter's bill that while it prohibits insurers from gaining genetic information it doesn't stop would-be clients from obtaining such information. This unequal information status would allow clients to game the system. If insurance purchasers know through genetic testing that they are likely to come down with a disease they can buy gold-plated insurance coverage at relatively cheap rates. The principle of insurance is that individuals are protecting themselves from unforeseeable losses and misfortunes. Proposals like Slaughter's is an incremental step toward ultimately transforming insurance into a social welfare program.

Slaughter explained that she had proposed the bill because "we wanted public policy, probably for the first time in the history of the Republic, to keep pace with scientific progress." Of course, the country doesn't seem to be doing too badly because Congress lags behind in responding to scientific advances.

Craig Venter recounted a story from University of Pennsylvania bioethicist Arthur Caplan. Caplan testified before the Pennsylvania legislature on some proposed anti-cloning legislation. Caplan decided to poll the assembled lawmakers and asked them, Where did they think their genomes were located? One third of the solons replied that it was in their brains, one third thought it was in their gonads, and one third had no idea. Can we really trust these sorts of people in Congress to make sensible laws governing fast-moving, new technologies?

Another abiding problem is that smart people like Collins and Venter use their well-deserved scientific authority to endorse public policy proposals in areas in which they have no special expertise. Of course they understand the science and the technologies, but that doesn't give them any insight into how people will use them later–their guesses are no better than anybody else's, and perhaps even worse. With regard to how insurance markets should operate, Collins and Venter are simply not experts. It was particularly disheartening to hear someone as famously entrepreneurial as Venter declare that "universal health insurance seems like a very logical solution" to the specter of genetic discrimination. Universal health insurance would mean the death of the genomics revolution as government rationing and regulation took hold of the health care system.

The historical record shows that people acting through markets have generally been highly successful in protecting their interests. Perhaps genomic information will push health insurance markets into operating something more like life insurance markets do: Purchasers buy policies for long durations. The inception of a new technology is a singularly inappropriate moment for government to interfere with its development. If government gets involved now, it will take decades to unravel the mistakes that it makes . Keeping to the public policy theme of the morning, John Doll, who heads up the federal division that handles biotech and pharmaceutical patents, lectured on DNA patenting. It would have been good for the theologians I dined with to have attended this meeting. Doll clearly explained the new guidelines for patenting biotech inventions and discoveries.

Lee Makowski from the Argonne National Laboratory outlined the U.S. Department of Energy's ambitious plans for determining the three-dimensional "folding" of all proteins The shapes of proteins is how they work. In the past, it took years to figure out these shapes. The academic mantra used to be, one graduate student, one protein, one Ph.D., one career. According to Makowski, getting the three-dimensional structure of a protein used to measured in man-years. The record now at the Argonne Lab is six hours.

In the afternoon, I dropped by a symposium that asked the provocative question, "How Long Can Humans Live?" The dour Leonard Hayflick, from the University of California-San Francisco, kicked off the session. Hayflick made his reputation by discovering that animal cells divide only a set number of times before they senesce and die. Hayflick insists on a useful distinction between lifespan and life expectancy. Lifespan is the longest that any member of species has ever lived. In the case of humans, this appears to be the Frenchwoman Jeanne Calment who lived to be 122 years old. Life expectancy is the age at which 50 percent of any given population cohort has died. Life expectancy in the U.S. is around 76 years today.

Hayflick made the point that if deaths from cancer, heart disease, stroke, and the like were all eliminated, that would increase life expectancy by only 15 years. Hayflick also insisted on a distinction between research on aging and geriatric medicine. Hardly anyone is researching aging; instead they're focusing on diseases that appear in old age, he complained. For example, half of the budget of the National Institute on Aging is spent on Alzheimer's research, claimed Hayflick. Yet if one eliminated Alzheimer's entirely, one would add only 19 days to average life expectancy in the U.S.

Hayflick believes that aging is the result of increasing molecular disorder in cells. Cellular repair mechanisms do a good job of fixing damage until after we pass our reproductive period. Once humans have successfully reproduced, then natural selection has no reason to worry about whether their cells get repaired or not. Thus we age, because from the point of view of natural selection, there's no reason for us to hang around after we've produced progeny.

Hayflick claims that aging is a phenomenon unique to humans and the animals we choose to protect. In the wild, humans and animals died well before they experienced aging. He says that prehistoric remains of humans have not revealed that our ancestors lived beyond age 50. Hayflick doesn't hold out much hope that scientists will be able to conquer ageing any time soon. Which is just fine by him, since he believes that neither individuals nor society would benefit from the prevention of ageing.

Demographer Jay Olshansky from the University of Chicago was not much cheerier. He agreed that if all cancer, cardiovascular disease, diabetes, and the like were eliminated, average life expectancy in the U.S. might reach 90 years, up from 76 now. He pointed out that if every cause of death before age 50 were removed in the U.S., that would boost average life expectancy by only 3.5 years. Mathematically, in order to boost the average life expectancy of a particular cohort to 100 years, 18 percent of that cohort must live to be over 122. So far, that's been achieved by only one person. The only way to really raise life expectancy is to retard or eliminate aging.

George Martin from the University of Washington was a bit more upbeat. It is theoretically possible that as few as 300 genes may be involved with aging and among them a very few might have a big effect. But he agreed with the Hayflick's evolutionary point that deleterious genes that express themselves after age 45 have no effect on reproductive success therefore natural selection doesn't bother to eliminate them. He illustrated this point by looking at the ApoE genes that are involved with Alzheimer's disease. It is now known that carriers of theApoE4 variant of the gene tend to have early-onset dementia. It has not been eliminated by natural selection because dementia begins after 45. Even more interestingly, the ApoE4 variant is far more prevalent among Africans and people descended from Africans. The ApoE4 gene is not very good at delivering lipids to membranes. This is important because the endemic African organism that causes sleeping sickness needs lipids to survive, so the ApoE4 gene protects carriers from sleeping sickness at the expense of causing dementia and cardiovascular disease after their reproductive years are passed.

What evolution grants, it can take away.

Man or Mouse? Genomically, it doesn't much matter.

San Francisco, CA, Feb. 18—"I live in Washington, D.C.," confessed Science senior editor Barbara Jasny as she opened the "Beyond the Human Genome" seminar at yesterday's session of the Association for the Advancement of American Science's annual meeting in San Francisco. "In Washington, you get used to hype and spin, but this is one time when all the words–landmark, milestone, revolutionary–really do apply."

Jasny's right. But then she showed herself a true denizen of D.C. when she indulged in what is rapidly becoming a standard ritual whenever scientists speak of the human genome. She claimed that "the discovery that it takes far fewer genes than we thought to make a human being is causing people to re-evaluate our relationship to other species." In strict biological terms that is certainly true, but Jasny's utterance is supposed to take us arrogant humans down a peg or two since, after all, it takes only a few thousand more genes to make us than it takes to make a nematode. Rubbish. When a nematode can devise a culture capable of sequencing its own genes, then I'll eat humble pie. But not before.

Most of the morning session of the Genome seminar was devoted to discussing where scientists plan to go from here. That means that there were few new findings to report since the announcement of the publication of the draft sequences of the human genome in Science and Nature earlier this week.

The first presenter was Mark Adams, who is Vice President for Genome Programs at Celera Genomics, the private company that raced the publicly financed International Genome Project consortium to the finish line. Adams gave a quick overview of how Celera had sequenced all 3.2 billion DNA base pairs of the human genome in only 9 months. Adams told the audience that Celera has now sequenced the genomes of 3 different strains of mice and despite the fact that mice and men last shared an ancestor 80 million years ago, they are genetically very similar. "The mouse genome is going to be a tremendous resource for learning about the human genome," said Adams. For example, Adams said that "there are many genome regions that are conserved between mouse and human that do not code for protein." Generally speaking, if a lot of different species share similar segments of DNA, that means that those segments must be important or else natural selection would have eliminated them is some species. Adams thinks that these conserved regions in mice and men probably exercise some sort of regulatory control over other genes. If scientists were unable to compare different animal genomes, these regions would likely not have been found so quickly.

Next up was Eric Green, who works at the National Human Genome Research Institute. He made some suggestions for which animal genomes should be sequenced next. Having other genomes to compare will help uncover the functions of the third of human genes whose functions are currently unknown. "Virtually all mammals have genomes that are roughly 3 billion base pairs," said Green. Why not sequence a lot of different genomes? Cost. Right now it costs between $10 to $15 million to sequence a genome once and one should really do it at least 4 times to make sure that it's accurate, so the costs run to around $50 million per completed animal genome. "Of course, if sequencing technology dramatically improves, then all bets are off," said Green. After the mouse genome, the next one is rat. Sequencing work is going on with various other animal species including chimpanzees, baboons, cows, cats, dogs, chickens, zebrafish and pufferfish. At the end of Green's presentation, a scientist from the audience complained about how hard it is to use the government's gene data banks compared to Celera's much easier system, and asked "Why can't the government do better than government work?"

The afternoon session began with Michael Eisen from Lawrence Berkeley National Laboratory and University of California at Berkeley. "It shouldn't surprise anybody that you can have lots of different things come out of the nearly the same sets of gene repertoires," said Eisen. "After all, you already have very different things stemming from the same set of genes inside yourself now. For example, neural cells don't look very much liver cells." The DNA may be similar, "but the differences come from regulation, and everything in a cell is regulated."

Eisen works with DNA "microarrays" to characterize gene expression in various normal and abnormal tissues. It turns out that normal human tissues taken from different individuals display strikingly similar gene expression patterns. Eisen then compares tissues taken from tumors whose gene expressions are very different than those of the normal tissues in which they arose. But what's interesting is that tumors vary in their gene expressions. Analyzing tissue taken from 600 patients, his lab has identified 5 different types of breast cancer tumors based on their gene expression profiles. Eisen knows the survival rates for patients having each of the 5 types of breast cancer. So using DNA microarrays to determine which kind of tumor a patient has provides information on the prognosis of that patient. "For example, if your tumor expresses keratin genes 5 and 17, then you have a bad prognosis," he explained.

Eisen ended his talk by asking the audience to support his "Public Library of Science" project, in which all scientific papers would be available to anyone online for free. "The permanent archival record of scientific research and ideas should neither be owned nor controlled by publishers, but should belong to the public," he declared. Is the peer review and publication system really broken? And he left unsaid just who would pay for and manage the scientific peer review process. Eisen's project seems to be underwritten by an egalitarian impulse more than anything else. Of course, his proposal was warmly applauded by the audience.

Wendell Weber spoke on the emergence of pharmacogenomics. Pharmacogenomics underpins the idea of truly personalized medicine that would be tailored to patients' particular genomes. Pharmaceutical companies are very aware that different patients have different reactions and outcomes to their drugs. For example, heart patients who have Long QT waves (LQT) in their EKGs are prone to cardiac arrhythmias which can kill them. The problem is that there are two types of LGT patients; one has a potassium ion defect and the other has a sodium ion defect. Treating one as though he were the other could kill a patient. Pharmaceutical companies armed with information coming from the human genome will help them detect differences like the LGT variants and devise better drugs to treat specific genomes.

The favorite study object of Stuart Kim, from the Stanford University Medical Center, is the nematode worm. In 1998, the worm C. elegans was the first animal to have its genome fully sequenced C. elegans has some 19,000 genes, but researchers know what only 1,562 of them do. Kim explained to the audience how using DNA microarray technologies combined with some really sweet data visualization tools is helping him and other researchers to figure what all the worm's genes do. Using microarrays, worm researchers are able to tell when genes are turned on and under what circumstances. Combining this information into a "topomap" program that converts the gene data into a three-dimensional surface map on a computer screen, one finds that genes that activate under similar conditions cluster into discreet "mountains." Generally, if two genes turn on at the same time under the same conditions, they are likely to be coregulated and have similar functions. Since Kim knows some of the genes, he looks at what mountain they are on, say Sperm Mountain or Muscle Mountain or Gut Mountain (all terms he uses). If he finds an unknown gene nearby his known gene, he gets an indication about what its function is and can devise further tests to check it out. There is no reason why this technology can be used to help figure out what the unknown one third of human genes do. Besides that, the topomaps are pretty neat looking.

Francis Collins, the head of the National Human Genome Research Institute and coordinator of the International Genome Project, received rock star treatment when he gave the plenary speech in the evening. People crowded around for his autograph or a picture. Collins began by playing a well-produced video about the project called The Secret of Life.

Collins basically reprised the findings from the human genome sequence made public earlier this week which you can find in my column mentioned above. Collins ended his speech by speculating on what the next 30 years of genomics may hold. By 2010, Collins thinks that there will be predictive genetic diagnostic tests for many common diseases including diabetes, heart disease, and many cancers. Pre-implantation diagnosis of genetic diseases carried by embryos will be widely available at fertility clinics. By 2020, gene-based designer drugs will be common, physicians will be able to precisely target treatments for tumors based on their gene expression profiles, and it may be possible to repair single nucleotide defects in embryos so that neither they nor their descendants would ever inherit specific genetic diseases. By 2030, Collins foresees that comprehensive genomics health care will be the norm and that most illnesses will be detected and monitored by molecular surveillance. Average life span will reach 90. He also thinks that public debate will be focusing whether or not human beings should take charge of their own evolution and modify their genomes (he opposes such modifications). Finally, Collins also believes that major anti-technology movements will be active in the U.S. and elsewhere.

The more things change–and improve–the more they stay the same.

Getting High and Having Babies

San Francisco, CA, Feb. 17—Have you ever wondered why you like to get high? Or why most mammals have live births rather than produce them in external eggs (as most sensible genera do)? On Friday, scientists at the American Association for the Advancement of Science's annual meeting had answers to these pressing questions.

First, let's talk about live births. Luis Villarreal, a virologist from the University of California at Irvine began the morning with a standing-room-only lecture titled, "The Role of Persisting DNA Viruses and Retroviruses in Host Evolution" (such a topic would be SRO only at a convention of and for scientists). His first point was that without viruses, eukaryotes would never have evolved. That's important because we're eukaryotes: organisms whose cells have distinct nuclei in which the DNA encoding our genes are enclosed. In contrast, the organisms that dominated for the first 3 billion years of life on earth, bacteria, don't have nuclei. The DNA that encodes their genes just floats randomly around in their cells. It turns out that by comparing gene sequences, Villarreal and his colleagues have found that the genes which allow for the replication of eukaryotic DNA are derived from viruses.

And you thought that viruses just gave you a bad cold or AIDS. Actually, many viruses still do make their livings (if you can call it that) by infecting a host, reproducing rapidly and then infecting another host. Villarreal, however, focused on viruses that don't do a lot of damage to their hosts but instead try to get along (so to speak). Herpes virus and chicken pox virus, for instance, infect and stay with their hosts forever, doing relatively little damage. It is these persistent viruses which, more than 500 million years ago, infected and incorporated their genes for DNA polymerase in a multicellular alga that subsequently used them to replicate its DNA. This is amazing enough, because eukaryotic cells eventually gave rise to complex organisms such as ferns, grass, trees, worms, clams, fish, amphibians, dinosaurs, reptiles, birds, and kangaroos. These viral infections may well be responsible for the Cambrian radiation of species that began around 500 million years ago. The pre-Cambrian world was largely populated by single-cell, blue-green algae. Note that the progeny of all those multicellular animals mentioned above and plants still develop from external propagules (a fancy word for eggs and seeds). The progeny of placental mammals don't.

The human genome, like most mammalian genomes, is loaded with DNA that codes for human endogenous retroviruses (HERVs). They are retroviruses that infected and insinuated their genes into the genomes of our ancestors many millions of years ago. On human chromosome 22 (one of the 23 pairs of chromosomes that make up the human genome), researchers have identified 225 genes and DNA coding for more than 2000 HERVs. Viruses, it turns out, are a great source for the new genes that drive evolution; they mutate frequently and so can devise new combinations quickly. By comparing known genomes, scientists have determined that, in evolutionary terms, only 324 genes derive from the Last Common Ancestor (Luca) of all living things.

But how to explain viviparous birth? How do the immune systems of placental mammalian mothers tolerate their embryos? After all, half of embryonic genes come from fathers which would normally provoke an immune system attack from mothers. Retrorviral genes appear crucial for this step. Villarreal's research finds that a great many HERVs are expressed in early embryonic tissues, especially in the trophectoderm, which is the tissue that eventually becomes the placenta. For example, HERV W seems to be responsible for fusing an the embryonic trophectoderm to the uterine wall and HERV 3 seems to provide for localized immunosuppression so that the mother's immune system doesn't reject the developing embryo. Without retroviruses, we placental mammals wouldn't be possible, concludes Villarreal.

Before getting on to the question of why people like to get high–patience is a virtue!–I sat in on the symposium on the "World's Major Food Crops: Prospects for Feeding 10 Billion People." The good news is that the prospects for feeding 10 billion people by 2040 are excellent, according to the panelists.

Ronald Phillips from the University of Minnesota lectured on the prospects of corn. He pointed out that in 1940 yields were 25 bushels per acre and in 1990 they were 120 bushels per acre. Fifty percent of that increase was due to genetic improvements in the crop and 50 percent was due to improvements in cultural practices, including better weed and pest control, more fertilizer, and better tillage techniques. Phillips is enthusiastic about the prospects of biotech to improve yields, quality, and nutrition. Genomics is making it ever more clear how similar grass species such as corn, sorghum, oats, sugar cane, and rice are genetically. To know how one operates sheds a lot of light on how the others operate.

The first plant genome, Arabidopsis thaliana or mustard weed, was completely sequenced late last year. It consists of 143 million base pairs. Syngenta and Monsanto have both announced that they've sequenced the rice genome, which is around 430 million base pairs. Corn's genome is contained on 10 chromosomes which appear to be essentially two sets of five. Somewhere in its evolution, corn's genome simply doubled itself. Corn's total genome is around 2.5 billion base pairs, making it comparable to the human genome at 3.2 billion base pairs. "Corn is about as smart as humans," quipped Phillips.

Phillips pointed out that 25 percent of the U.S. corn crop is genetically enhanced, which led to an estimated yield increase of 66 million bushels. Growing the equivalent amount of non-biotech corn would have required that an extra 450,000 acres of land be plowed down to produce it. Fungal toxins spread by insects are powerful human carcinogens, but they are 40-fold lower in pest resistant Bt corn.

Gurdev Khush, who is the senior rice breeder at the International Rice Research Institute (IRRI) in the Philippines and a recipient of the World Food Prize, spoke next on improving rice. IRRI was one of the crop research centers that brought about the Green Revolution in the 1960s and 1970s. Twenty-three percent of the calories consumed by the world's people comes from rice, while only 17 percent comes from wheat and 9 percent from corn. Rice yields are up 230 percent and the price of rice is 40 percent lower than it was in the mid-1960s. Khush described various techniques that he believes will boost rice yields by 20 percent or more in the coming years. He also endorsed the new insect-resistant Bt rice varieties and biotech rices which have been engineered to resist bacterial blight. Today, some 27 percent of rice crops are lost to insect pests and 9 percent are lost to diseases. Khush also talked briefly about the efforts of Japanese researchers to move very efficient C4 photosynthesis from corn into rice, which normally employs less-efficient C3 photosynthesis. If the Japanese scientists successfully convert rice from a C3 to C4 cereal, that could greatly boost yields.

Stephen Baenziger from the University of Nebraska was the champion for wheat. Wheat's genome is absurdly long–16 billion base pairs, which makes it more than 5 times longer than the human genome. Modern bread wheat is a hexaploid plant, which means that it is the result of combining the genomes of 3 different progenitor species. Half a million years ago, two progenitor diploid species denominated AA and BB spontaneously combined into the tetraploid species AABB. Then 6,000 years ago, the AABB species combined with another species called DD to create the modern wheat genome which is denominated AABBDD. This genetic variability makes wheat suitable for growing on land far more marginal than can be used for corn and rice.

In the past 30 years, wheat yields have doubled, while corn and rice yields have nearly tripled. Baensziger blamed this on wheat's complicated genetics and the fact that it is typically grown on relatively poor land. But research on wheat, unlike research on corn and rice, is almost entirely supported by public funding. Could this have a negative effect on yield improvement? Baensziger also made the point that there are a total of 900 plant breeders in the United States and two-thirds of them chiefly work on improving corn. It's amazing how much good so few people are able to accomplish.

Randy Shoemaker, from the U.S. Department of agriculture was the expositor for soybeans. Soybeans were brought to Georgia in 1768 from China by Samuel Bowen and were grown as a forage crop in this country until around 1900. As a legume, soybeans produce their own nitrogen fertilizer and are an excellent source of protein. Soybean production in the U.S. rose from 1.2 billion bushels in 1974 to 2.6 billion in 1999. Today, the U.S. produces 49 percent of the world's soybeans.

Almost to why you get high–but first a short detour to Deep Green. Deep Green is a project in which plant scientists are tracing the evolutionary roots of all green plants and placing them in their proper evolutionary relationships. This science is called phylogeny. Deep Green phylogenists believe that blue-green algae (cyanobacteria) began creating the oxygen in the Earth's atmosphere some 2 billion years ago. Green plants emerged as unicellular algae about 1 billion years ago. When oxygen in the Earth's atmosphere reached sufficient concentrations about 500 million years ago, multicellular creatures began to evolve. This began the Cambrian radiation.

Apparently algae began their conquest of the land near the beginning of the Cambrian period as well. For years, biologists believed that marine algae were the progenitors of today's terrestrial plant species, but genetic research has uncovered that most land plants arose from a freshwater alga similar to today's alga, Mesostigma. Today there are more than 500,000 green plants, of which 250,000 are flowering plants.

All right, so why do you get high? Because you learned to like it, according to the panelists at the symposium titled "Addiction is a Brain Disease: How Drugs Change Your Mind."

The consensus of the panel, which included the current director of the National Institute of Mental Health Steve Hyman, is that addiction, defined as "out of control" drug use, occurs when drug use hijacks learning circuits in the brain. As Hyman put it, drugs induce the "inappropriate recruitment of normal molecular mechanisms responsible for associative learning." The brain is supplied with various receptors for neurotransmitters such as dopamine and serotonin. When an organism learns something useful, like sex is fun or food tastes good, it gets a burst of pleasurable neurotransmitters that help it remember this important fact by associating pleasurable feelings with the conduct. Drugs increase the quantity of neurotransmitters or mimic them, thus causing the user to associate pleasure with ingesting the drugs even if they have other, deleterious effects. Brain scans indicate that drugs do activate many of the same areas of the brain that are associated with pleasurable experiences such as sex and satiation. Experiencing pleasure is, in evolutionary terms, supposed to reinforce behaviors that lead to survival.

"The brain didn't evolve for drugs. Drugs capture brain circuits that normally serve as motivational and pleasurable reward systems," declared Hyman. Very interesting science.

Hyman also claimed that "the use of these drugs [such as cocaine, heroin, and alcohol] go well beyond the limits of the adaptive behaviors that humans have evolved to experience."

Charles O'Brien, of the Veterans Administration Medical Center in Pennsylvania, talked about a number of medications (dare we say, drugs) that disrupt an addict's pleasurable experiences with drugs, including naltrexone for alcohol and bupropion for nicotine. Hyman pointed out that long-term memories are fixed through protein synthesis in the brain, including the memories for pleasurable drug experiences. When you administer a protein-synthesis inhibitor in an addicted rat's brain when it's recalling a memory, it apparently disrupts the protein that encodes the memory and weakens it, according to Hyman. One could conceivably do this to human addicts. He ended by saying, "Whether that's good or scary, I leave to the audience."

Small Science: Nanotechnology Gets Big

San Francisco, CA, Feb. 16 -The American Association for the Advancement of Science kicked off its 167th national meeting at the Hilton Hotel in San Francisco yesterday with a seminar on nanoscience and nanotechnology. The keynote speaker on the subject was Harvard biochemist George Whitesides, commonly regarded as one of the United States' leading scientific statesmen. Whitesides declared that nanotechnology is "a legitimate scientific frontier." If he says so, then it must be so. He added: "Nano's going to be a contender with genomics to be the technology which changes the world."

Whitesides began, prosaically enough, by defining nanoscience as science that deals with lengths measured in nanometers. To give the audience some idea of just how tiny a nanometer is, Whitesides explained that viruses are between 100 and 10 nanometers, and biological molecules come in at around 1 nanometer; the width of DNA's double helix is about 2 nanometers. Twenty gold atoms strung together are just about 10 nanometers long. In other words, Whitesides is talking about really, really small objects.

Whitesides told the packed hotel ballroom that "there really isn't any nanotechnology right now"–though there sure is a lot of exciting nanoscience. Researchers are exploring all kinds of interesting properties and fabrication techniques. What's driving this research? First, scientists have gotten a number of neat new tools, like the scanning tunneling microscope (STM), which allows them to measure and manipulate single atoms. An STM is nothing like a light microscope. Instead, it's essentially a very fine point of some material, usually silicon or a metal, which is attached to a readout device. The sensitive tip of an STM is dragged across a surface or electrified so that it interacts with molecules. An STM is so sensitive that it can measure the waves of electrons that slosh back and forth like ocean waves over the surface of materials.

Commercial necessity is also driving the development of nanotechnology, according to Whitesides. Following "Moore's Law," electronics companies have been miniaturizing transistors and computer circuits at an exponential rate for nearly 40 years. The electronics companies fear that they will reach the limits of micron-scale miniaturization allowed by silicon technologies in the next 10 to 15 years, at which point they will need new manufacturing techniques, new designs, and new materials to keep the miniaturization process going.

Whitesides is a Big Picture thinker, and like many of that breed, he sometimes gets lost in his own canvas. So he worries that nanotech has "the potential for technology dislocations in several areas" and he is concerned about how ubiquitous nanotech might affect society. He illustrated the type of dislocations he fears by citing what happened to floppy disk makers: The 12-inch floppy disk makers were replaced by 8-inch disk makers, who were replaced by 5.25-inch disk makers, who were replaced by 3.5-inch diskette makers, according to Whitesides history lesson. The new memory storage disks using magnetoresistive materials "have simply displaced a $1 billion industry," he lamented. Without getting into details (such as the fact that the disk manufacturers initiated and profited from many of the format changes), we might well ask, So what? Evidently, Whitesides is not a fan of Schumpeterian creative destruction.

In a similar vein, Whitesides deplored the pitiful amounts of tax money that the federal government is "investing" in nanotech—a mere $116 million annually compared to Japan's $120 and Europe's $128 million. Whitesides justified federal spending on nanotech on national defense grounds and the competitive need to "protect the U.S.'s dominant position in electronic and information technologies."

Ignoring such noisome detours into dubious industrial-policy concerns, Whitesides discussed a number of exciting nanotech possibilities. On the immediate horizon, bar codes on products may soon be replaced by product coding devices that consist of around 10,000 transistors to which are attached tiny antennae. These product coding devices would simply be printed onto groceries, clothing, whatever and cost less than a penny a piece. As shoppers passed through the store's exit, the devices would be activated, note the products, and debit the shoppers' bank accounts in seconds–no more waiting at the check out counter. Nanotechnology should also make it possible to create memory storage devices the size of a wrist watch which could hold the information now found on 1000 CDs, which Whitesides noted "is probably more information than you know, plus or minus a little bit." Nanotech devices could also read and store a person's complete genome, making individual genomics a reality

"Sociology and psychology might become predictive sciences," suggested Whitesides. How? Information collected for commercial purposes would also be available to social science researchers and ubiquitous nanotech will make collecting and analyzing very big sets of personal data easy, according to Whitesides.

Whitesides put up a slide showing a schematic of one of the molecular motors designed by the godfather of nanotechnology, Eric Drexler. While appreciative of Drexler's pivotal role in nanotech, Whitesides was at pains to disagree with many of his ideas. "I personally don't think that these kinds of things are going to work," said Whitesides. "We already have biological motors." He then put up another slide of a bacterial flagellum showing its rotary motor elements and bushings. Whitesides pointed out that researchers already know how these biological motors work and that they are investigating them for nonbiological uses. In fact, some experimenters have already attached these bacterial motors to nonbiological surfaces and fed them ATP (the energy molecule found in cells), and they worked just fine. By studying these biological motors Whitesides thinks that we will be able to make similar small motors of nonbiological materials, perhaps in 20 years or so.

Similarly, Whitesides dismissed Drexler's notion of nanotech assemblers designed to manufacture objects one atom at a time. How would one power an assembler, asked Whitesides. How would one get it the information it needs to know what to do? And would it be really be strong enough to break atomic bonds? Whitesides added that if one did decide to try to build an assembler, one might decide to fuel it with glucose and code its instructions on DNA, but then one would be well on the way to reinventing bacteria. He pointed to ribosomes, which are the organelles in cells that make proteins as already existing biological assemblers. "The sophistication of ribosomes is so far ahead of anything we could think of building today," he declared. Nevertheless, Whitesides says that he is "personally very optimistic about biomimetic nanotechnology." By studying how the protein machines in cells work, scientists will be able to develop ways to mimic those processes in nonbiological materials.

It would have been more interesting had the organizers of the seminar invited nanotech visionaries like Eric Drexler of the Foresight Institute to participate. After all, Whitesides acknowledged that "Drexler is one of the people who deserves a great deal of credit for bringing nanotechnology to the public's attention." One has the suspicion that the visionaries were not invited because they don't hold chairs at prestigious universities.

Whitesides closed his remarks by noting that "in the last year, you are beginning to see lots of venture capital interest in nanotech."

After Whitesides came some of the leading academic nanotech researchers in the world. Up first was Dan Eigler from IBM, who described his elliptical "quantum corral." That's a space where, if you place an iron atom at one focus of the ellipse, it causes an electronic "mirage" to appear at the other focus. This discovery could be basis for a new way to transfer information on a nanoscale. "At IBM we're very interested in understanding the behavior of electrons in very small structures," said Eigler. "That is the future of our hardware technologies."

Uzi Landman from Georgia Tech described his work with nanojets. The fact is that pushing a few atoms of a fluid through a nanonozzle takes a surprisingly large amount of energy, but he hopes to use the jets to write nanocircuitry and to inject genes into cells.

Mara Prentiss of Harvard University is researching atomic optics. The wavelengths of atoms are more compact than light and atoms can be guided by magnetic fields. She is working on creating atomic mirrors, beam splitters, polarizers, and wave guides.

Michael Roukes from Caltech described how he uses nanoscale magnets for extremely sensitive magnetic resonance imaging. These devices would be able to detect the conformational changes that many biomolecules go through as they interact.

University of Wisconsin researcher Franz Himpsel is working on nanoscale memory storage. He noted that the famous physicist Richard Feynman once calculated that if he could use 125 atoms to code each bit of information, then he could store everything ever printed in a cube 1/200th of inch on each side. Himpsel showed a slide of a surface which could theoretically store a million times more information per square inch than current DVDs do. However, getting a readout of information at a reasonable rate up from a nanoscale to the macroscale where we live is a big problem.

The nanotech session ended with Harold Craighead, who described his work on microfluidics. He is creating devices that can quickly measure DNA. Using a system of tiny nanochannels, Craighead was able to obtain data about DNA strands in only 10 minutes which would have taken 11 hours using conventional techniques. He is also adapting his techniques for analyzing proteins which would be very useful in this post-genomic era.

The evening was closed out with a speech by the new president of the AAAS, Mary Good, a professor at the University of Arkansas who served for four years as Undersecretary for Technology and in the U.S. Department of Commerce. I wish I could say her speech was inspiring, but it is clear that Prof. Good sees her role as chief lobbyist for science. As such, her speech was a predictable plea for more government largess–and a call for the creation of a Cabinet-level Office of Science and Technology Policy.

Ah well. Scientists are only human, after all.