February 2007 Archives

Stewart Brand -- “Where are the green biotech hackers?”

| No Comments

Tomorrow's New York Times has a great article on Stewart Brand.  In it, he asks the question, “Where are the green biotech hackers?”  We're coming, Stewart.  It's just that we're still on the slow part of the curves.

It's an interesting question, actually -- when do we get to the fast part?  When does biology start to go really fast?  And what does fast mean?

One answer to the question is the speed and the cost at which we can presently sequence or synthesize an interesting genetic circuit or organism.  Costs for reading genes are halving every 18 months or so, and if the rumors are true, we will hit the Thousand Dollar Genome sooner than my original estimate.  Sequencing is pretty easy at this point, as long as you already have a map to work with, which is the case for an increasing number of organisms.  And if you build the organism yourself, or pay someone else to do it, then you already know both the basic structure of the genome (the map) and the specific sequence.

At the moment, synthesis of a long gene takes about four weeks at a commercial DNA foundry, with a bacterial genome still requiring many months at best, though the longest reported contiguous synthesis job to date is still less than 50 kilobases.  And at a buck a base, hacking any kind of interesting new circuit is still expensive.  As I reported from SB 2.0, the synthesis companies are evidently now using my cost estimates as planning devices, even though that's not why made those estimates in the first place.  They project costs to continue falling by a factor of 2 approximately every year, which means that it will be another 5 years before synthesizing something the size of E. coli from scratch will cost less than US$ 1000, or 1 kilobuck.

The bigger problem, though, is the turnaround time.  No engineer or hacker wants to wait four weeks to see if a program works.  Hit compile, wait for four weeks, no "Hello World."  Start trying to debug the bug, with no debugging tools.  No thanks.  (I've actually had discussions with geneticists/molecular biologists who think even waiting a few days for a synthesis job isn't a big deal.  But what can you say -- biology just hasn't been a hacker culture.  And we are the poorer for it.)

So, Mr. Brand, it will be a few years before green hackers, at least those who aren't supported by Vinod Khosla or Kleiner Perkins, really start to have an impact.  The hackers who are lucky enough to have that kind of support, such as the blokes at Amyris Biotechnologies if their past accomplishments are anything to go by, will probably have something to show for themselves pretty soon.

The article ends with a couple of great paragraphs, which, along with "Science is the only news", are all you need to live by:

“I get bored easily — on purpose,” he said, recalling advice from the co-discoverer of DNA’s double helix. “Jim Watson said he looks for young scientists with low thresholds of boredom, because otherwise you get researchers who just keep on gilding their own lilies. You have to keep on trying new things.”

That’s a good strategy, whether you’re trying to build a sustainable career or a sustainable civilization. Ultimately, there’s no safety in clinging to a romanticized past or trying to plan a risk-free future. You have to keep looking for better tools and learning from mistakes. You have to keep on hacking.

After many, many months of work, Bio Economic Research Associates (Bio-era) today released "Genome Synthesis and Design Futures: Implications for the U.S. Economy".  Sponsored largely by Bio-era and the U.S. Department of Energy, with assistance from Dupont and the Berkeley Nanosciences and Nanoengineering Initiative, the report examines the present state of biological technologies, their applications to genome design, and potential impacts on the biomanufacturing of biofuels, vaccines, and chemicals.  The report also employs scenario planning to develop four initial scenarios exploring the effects of technological development and governmental policy.   Here is a link to the press release; over on the right side of the page are links to a short Podcast with myself and Jim Newcomb describing some of the findings.

It is a giant topic, and even at 180 pages we have really just barely scratched the surface.  The changes we've already witnessed will pale in comparison to what's coming down the pike.  The report deals mostly with science, technology, economics, markets, and policy, and only starts to explore the social and ethical aspects of forthcoming decisions.  Future work will refine the technological and economic analyses, will flesh out the security aspects of the ferment in biological technologies, and will delve into what all this means for our society.  In the preface, Jim Newcomb and Steve Aldrich note:

In presenting this analysis, we are mindful of the limitations of its scope. The arrival of new technologies for engineering biological systems for human purposes raises complex questions that lie at the intersection of many different disciplines. As the historian Arthur M. Schlesinger has written, “science and technology revolutionize our lives, but memory, tradition and myth frame our response.” Because this report is focused on potential economic implications of genome engineering and design technologies for the U.S. economy, there are many important questions that are not addressed here. In particular, we have not attempted to address questions of safety and biosecurity; the likelihood or possible impact of unintended consequences, such as environmental damage from the use of these technologies; or the ethical, legal, and social questions that arise. The need for thoughtful answers to these and related questions is urgent, but beyond the scope of this work. We hope to have the opportunity to investigate these questions in subsequent research.

We had a lot of help along the way, and for my part I would like to thank Drew Endy, Brian Arthur, George Church, Tom Kalil, Craig Venter, Gerald Epstein, Jay Keasling, Brad Smith, Erdogan Gulari, John Beadle, Roger Brent, John Mulligan, Michele Garfinkel, Ralph Baric, and Stephen Johnston, and Todd Harrington. 

Here is web page to buy a hard copy and/or download the PDF.  Just fill out the form (we're trying to track interest), and you will be sent a link to the PDF.

On Indonesia and Distribution of H5N1 Strains

| No Comments | No TrackBacks

News in the last couple of days that Indonesia has decided not to forward homegrown strains of H5N1 to the WHO and instead is dealing directly with Baxter Healthcare for production of vaccines.  The worst bit of this, of course, is that there does not appear to be much cross reactivity elicited by the Vietnamese and Indonesian isolates, where the international reference vaccine is derived from a strain isolated in Vietnam.  Moreover, while Baxter is supposedly making progress in producing influenza vaccines in cell culture (Baxter's Press Releases, CIDRAP's version), this technology is not yet approved for human use; only research contracts, rather than production contracts, have been let by the U.S. Government for cell culture production.  Finally, despite much noise that cell culture is faster/better/cheaper than eggs for producing vaccines, it appears cell culture only beats eggs by a month or two.  (Baxter does have a very comendably decent Influenza information web page, which is here.) 

Here are a few paragraphs from an AP story, "Experts say Indonesian deal on H5N1 virus jeopardizes race for pandemic vaccine", via the IHT:

Indonesia Wednesday signed a memorandum of understanding with U.S. drug manufacturer Baxter Healthcare Corp. to develop a human bird flu vaccine.

Under the agreement, Indonesia will provide H5N1 virus samples in exchange for Baxter's expertise in vaccine production. Other organizations would have access to Indonesian samples provided they agree not to use the viruses for "commercial" purposes, said Siti Fadilah Supari, Indonesia's health minister.

But that is a major departure from the World Health Organization's existing virus-sharing system, where bird flu viruses are freely shared with the global community for public health purposes, including vaccine and antiviral development. Indonesia has not shared any viruses since the beginning of 2007.

Indonesia defended its decision, arguing the system works against poor countries. "The specimens we send to WHO...are then used by vaccine makers who then sell to us (at a profit)," Supari told reporters Wednesday. "This is unfair, we have the virus, we are getting sick, and then they take the virus from WHO — 'with WHO's permission' they say — and make it themselves," said Supari.

There seems to be a bit of confusion among reporters about whether Indonesia now has an official policy of withholding samples from the WHO, but Baxter is making it clear they don't have anything to do with the decision.  From The New York Times' coverage: "A Baxter spokeswoman said the company had not asked Indonesia to stop cooperating with the W.H.O. She added that the agreement under negotiation would not give it exclusive access to Indonesian strains."

In any event, Indonesians feel bent out of shape that they have previously provided strains to the international community, only to be charged for the vaccine when it becomes available.  News reports portray this as something of an IP spat, akin to controversy over biomining.  From the Reuters coverage:

"The specimens we sent to the WHO have been forwarded to their collaborating center. There it has been used for various reasons such as vaccine development ... or research," Supari said.

"Later they sold the discovery to us. This is not fair. We are the ones who got sick. They took the sample through WHO and with WHO consent and they tried to produce it for their own use," she said at a news conference after the signing of the pact with Baxter.

Supari said Australia was producing a human bird flu vaccine using the Indonesian virus strain, but did not give details.

"I was shocked because I never gave permits to Australia to produce a vaccine using our strain," she said.

"We have been working with Baxter since the beginning and are processing intellectual property rights with them. Baxter protects our intellectual property rights," she said.

...Under the memorandum of understanding, Indonesia would have the right to produce and market the bird flu vaccine domestically. It is negotiating to export it to a number of countries.

Production would be carried out by makers appointed by the Health Ministry.

So, in conclusion, the deal appears to put Indonesian isolates of H5N1 out of the reach of governments and firms with other vaccine technologies, at least for the time being.  Finally, in an interesting twist on the distribution of biological technologies, the deal also appears to put Indonesia in a position to become a leader in cell culture production of vaccines, potentially jumping to the head of the pack in the international vaccine market.

H5N1 is back in the U.K.

| No Comments | No TrackBacks

The headlines are today loudly announcing the return of H5N1 to the United Kingdom (CNN, New York Times) at a Turkey farm near Lowestoft.  Though nobody can say for sure, the virus probably arrived via migrating birds.  It appears that the likelihood of transmission by migrating bird or smuggled poultry has a geopolitical dependence.

Last month, Kilpatrick, et al., published a paper in PNAS ("Predicting the global spread of H5N1 influenza") that looked at a variety of factors to classify historical outbreaks and predict new ones.  The abstract does a decent job of summarizing the paper, so here it is:

The spread of highly pathogenic H5N1 avian influenza into Asia, Europe, and Africa has resulted in enormous impacts on the poultry industry and presents an important threat to human health. The pathways by which the virus has and will spread between countries have been debated extensively, but have yet to be analyzed comprehensively and quantitatively. We integrated data on phylogenetic relationships of virus isolates, migratory bird movements, and trade in poultry and wild birds to determine the pathway for 52 individual introduction events into countries and predict future spread. We show that 9 of 21 of H5N1 introductions to countries in Asia were most likely through poultry, and 3 of 21 were most likely through migrating birds. In contrast, spread to most (20/23) countries in Europe was most likely through migratory birds. Spread in Africa was likely partly by poultry (2/8 introductions) and partly by migrating birds (3/8). Our analyses predict that H5N1 is more likely to be introduced into the Western Hemisphere through infected poultry and into the mainland United States by subsequent movement of migrating birds from neighboring countries, rather than from eastern Siberia. These results highlight the potential synergism between trade and wild animal movement in the emergence and pandemic spread of pathogens and demonstrate the value of predictive models for disease control.

Of course, the only way to know if the model really works is, alas, to wait for more outbreaks.  Anyway, it seems the U.S. is safe from poultry smuggling, which we have a chance of intercepting, but susceptible to migrating birds, a pathway that almost certainly resists any defensive measures.


H5N1 Influenza and Countermeasures Update

| No Comments | No TrackBacks

There is an excellent news piece in last week's Science, where here the definition of excellent is both "informative" and "highly unsettling".  Dennis Normile and Martin Enserink write:

An upsurge in H5N1 bird flu outbreaks in poultry across Asia is driving home the message that even countries that have eliminated the virus once shouldn't become complacent. The continuing high death toll in humans, including two recently detected cases of infection with a Tamiflu-resistant strain in Egypt, is also a grim reminder of how devastating the virus might be if it acquires the ability to spread easily among humans.

...Over the past 3 weeks, Thailand and Vietnam reported their first H5N1 outbreaks among poultry in 6 months. Japan, which seemed to have dodged the bullet since its cluster of outbreaks in 2004, confirmed that the virus hit one farm on 11 January and probably a second farm on the 23rd. South Korea, which last November suffered its first outbreak since containing the virus in 2004, reported that the virus had turned up on a fifth poultry farm. Several wild birds found dead in Hong Kong tested positive for H5N1. And Indonesia on 20 January reported its fifth human death from the virus in just 10 days, bringing its death toll to 62, by far the most of any country.

The increase in outbreaks in the Northern Hemisphere follows what has become an established pattern. The reason for the seasonality is still not well understood, says Les Sims, a veterinarian based in Manunda, Australia, who advises the U.N.'s Food and Agriculture Organization (FAO). It is likely to be some complex interaction among several factors, including cooler temperatures enabling the virus to survive longer in the environment, greater poultry trade in preparation for winter festivals, and movements of wild birds.

The recurrence of the virus in South Korea and Japan is particularly notable. In both the winter of 2003-'04 and this year, outbreaks in South Korea were followed 4 to 6 weeks later by outbreaks in Japan. "The outbreaks in Japan and South Korea suggest to me free-flying birds as the most likely origin," says Sims. Both countries are trying to determine how the virus was reintroduced.

So it seems unlikely we will be rid of the virus through culling programs, the primary mechanism thus far employed for biosecurity.  That the virus seems to be spread by wild birds in these cases is interesting, but this isn't the only pathway for reintroduction into poultry or people.

Last week's issue of New Scientist revisits the notion that "Deadly H5N1 may be brewing in cats".  (Most of the relevant text is available here at ProMed.)  Felines may be serving as a mammalian host that might enable the virus to adapt to mammalian biology and thereby become more dangerous to humans.  This is something I started wondering about after cats started dying in Europe so soon after the virus arrived there last year.  The New Scientist provides corroborating evidence that cats are important in the dynamics of the virus in Indonesia.  The story reports some surprise on the part of scientists doing the field work with regard to the prevalence of the virus in cats in Indonesia:

In the first survey of its kind, an Indonesian scientist has found that in areas where there have been outbreaks of H5N1 in poultry and humans, 1 in 5 cats have been infected with the virus, and survived. This suggests that as outbreaks continue to flare across Asia and Africa, H5N1 will have vastly more opportunities to adapt to mammals than had been supposed.

Chairul Anwar Nidom of Airlangga University in Surabaya, Indonesia, told journalists last week that he had taken blood samples from 500 stray cats near poultry markets in four areas of Java, including the capital, Jakarta, and one area in Sumatra, all of which have recently had outbreaks of H5N1 in poultry and people.

Of these cats, 20 per cent carried antibodies to H5N1. This does not mean that they were still carrying the virus, only that they had been infected - probably through eating birds that had H5N1. Many other cats that were infected are likely to have died from the resulting illness, so many more than 20 per cent of the original cat populations may have acquired H5N1.

This is a much higher rate of infection than has been found in surveys of apparently healthy birds in Asia. "I am quite taken aback by the results," says Nidom, who also found the virus in Indonesian pigs in 2005. He plans further tests of the samples at the University of Tokyo in February.

The data explicitly contradicts prior statements from the WHO downplaying the role of cats in harboring or spreading the virus, which I wrote about here.  I continue to be fascinated by the extent to which the behavior of the virus in the wild contradicts the expectations and public statements of "officials" in various organizations around the world.  H5N1 is clearly evolving in ways that are both surprising and worrying.

The New Scientist and Science stories both note that two people in Egypt who recently died from H5N1 infections were carrying strains of the virus evidently resistant to  Tamiflu.  It is unclear whether the virus carried the relevant mutations before it infected these patients, or whether it evolved during their illness because they were treated with Tamiflu in the hospital.  Either way, it seems that many people infected with H5N1 are diagnosed after the window in which antivirals are most effective, in part because diagnosis is both difficult and slow.  This phenomenon is described in two articles and a commentary in the 26 November, 2006, issue of The New England Journal of Medicine that report disturbing analyses of human H5N1 outbreaks in Indonesia and Turkey last year.

In a New York Times article about the NEJM papers, Donald Mcneil, reports the following:

Rapid tests on nose and throat swabs failed every time, and in Turkey, so did all follow-up tests known as Elisas. The only tests that consistently worked were polymerase chain reaction tests, or PCRs, which can be done only in advanced laboratories and take several hours.

''It'll be a disaster if we have to use PCRs for everybody,'' said Dr. Anne Moscona, a professor of pediatrics and immunology at Weill Cornell Medical College. ''It just isn't available at a whole lot of places.''

If the A(H5N1) flu mutates into a pandemic strain, rapid tests ''will be really key,'' she said.

What the NYT didn't report is that the patients were infected on average 5 days prior to the appearance of symptoms, outside the window recognized for effective use of antiviral drugs.  Robert Webster and Elena Govorkova have an excellent Perspective piece accompanying the NEJM articles, and they note that in the Indonesian cases, "...Treatment [with oseltamivir] began 5 to 7 days after initial infection.  Such delayed administration of the drug limits its value in decreasing the viral load and might lead to the selection of resistant strains."  It isn't clear from the paper describing the Turkey outbreak when oseltamivir was administered, but those patients did not experience symptoms for an average of 5 days after exposure to the virus, and then received antibiotics for the first 3-7 days of hospitalization before transfer to a unit that treated them for influenza.  In summary, it appears the virus is often being exposed to oseltamivir after the period when the drug is expected to be effective, enhancing the probability of selecting for resistant mutants.

Finally, in a slight change of direction, in the 21 December issue of Nature, John Oxford has a review of a new book on influenza, "Bird Flu: A Virus of Our Own Hatching", by Michael Greger.  You may recall that Oxford is primarily responsible for the hypothesis that the 1918 flu emerged at a British army camp near Etaples, a tale I wrote about a couple of years ago (The Spanish Flu Story).  Oxford notes that:

I am often kicked around by American authors in books about influenza. How dare a Limey suggest that the Spanish influenza A H1N1 virus arose in a gas-infected, pig-ridden and bird-infected army camp of 100,000 people in France in 1916, when the whole world knows it started in Dorothy's home state, Kansas? But I felt less bruised than usual. Perhaps I am getting used to it.

I still find Oxford's version of the origin of the Spanish Flu to be the most compelling, in part because it describes a situation of close contact between animals and people, where those animals were killed and prepared as food by soldiers on a daily basis in conditions not so dissimilar to those in many developing countries where H5N1 is present today.