5 July 2002

THE DEVIL IN THE DETAIL: THE TECHNICAL ARGUMENT BEHIND THE NATURE RETRACTION (PT 2 OF MEXICAN MAIZE MADNESS)

EXCERPT:

Although the idea of transgenes moving around freely in a host genome is not generally accepted, [Former Calgene research head Robert Goodman] says a lot is not understood about what happens to transgenes once they are inserted.

"It is not surgically precise," he said. "It's a bit of a fling at the genome. Odd things happen."

"Why some lines don't continue to perform are not understood," he said. "There are so few people who have looked at those kinds of questions."

***

Mexican Maize Madness

http://www.abc.net.au/science/slab/mexicanmaize/part_two.htm
Anna Salleh, ABC Science Online, 4 July 2002
Part II - The Devil in the Detail: the technical argument behind the retraction

Last year, Nature published a paper claiming  transgenes had been found in native Mexican maize. Four months later  the paper was retracted. In Part 1 of Mexican Maize Madness, Anna Salleh explores the issues behind the retraction. In Part 2, she delves into the murky world of technical argument behind the retraction and the broader issues it raises. Much has been said about the commercial and other interests of the players involved in the debate, but what about the science?

Part II - The Devil in the Detail: the technical argument behind the retraction

The Findings

Quist and Chapela's original paper in Nature reported two findings.

Firstly, they used a technique called Polymerase Chain Reaction (PCR) to amplify the presence of transgenic DNA in the genome of Mexican landrace varieties of maize.

Secondly, they used a technique called inverse Polymerase Chain Reaction (i-PCR) to amplify sequences on either side of the transgenic DNA. This was to try and work out where it had positioned itself in the native maize genome.

They reported finding transgenic DNA in multiple locations of the genome. In some cases the transgenic DNA was fragmented and Quist and Chapela argued that it could have been fragmented prior to crossing with the Mexican maize, but also that it could have become fragmented after it had crossed.

The idea of transgenic DNA fragmenting as it moves from one generation to the next implies that it is not stable - a concept which is quite alien to genetic engineers.

"The attitude of most people is that the initial transgene insertion is a random event but once it's in it's stable," says former research director of biotechnology company Calgene, Professor Robert Goodman.

The referees' opinion

Traditionally when a paper is peer reviewed there are several 'referees'. One of the three referees who studied the exchanges between Quist and Chapela and their critics stated that none of the data was sufficient to warrant any of the conclusions and demanded retraction of the manuscript unless further data was supplied.

As Nature's editorial note suggests, even after the further data was provided, disagreements persisted.

The other two referees, however, said that Quist and Chapela's first conclusion - that transgenic corn is growing in Mexico and crossing with local varieties - remained unchallenged.

These referees did accept that there were legitimate criticisms in the letters relating to Quist and Chapela's second conclusion - that transgenes were moving around in the Mexican maize genome - and one recommended that Quist and Chapela publish a correction.

For the prosecution

Kaplinsky and colleagues argued the findings were most likely artefacts of the technique used. They argued that the i-PCR primer (short stretches of DNA which pair up with parts of the longer sequence to be amplified) used could easily pair up in the wrong place and amplify the wrong bit of DNA. They said the bits of DNA amplified were more likely to be "junk DNA" present elsewhere in the maize genome, rather than sequences adjacent to the transgenic DNA.

Matthew Metz's letter also argued the results were artefacts of a careless application of the i-PCR method.

Fragmentation, said the critics, went against all previous evidence that showed transgenes were stable entities and not in the habit of splitting up and moving around as they pass through generations. The idea was "unprecedented", "revolutionary", "totally ludicrous" and "more mysticism than science".

For the defence

In their rebuttal, Quist and Chapela acknowledged two of their sequences could have been artefacts, but not the others, which they described as have a "revealing pattern of discontinuity...indicating the integration junction between the transonic DNA and the native host genome."

They also cited four references to show transgenic construct fragmentation is "not such a mind boggling idea".

However, Quist and Chapela were accused of mis-citing the references, which actually refer to fragmentation during transformation (insertion of the transgenic DNA into the original genetically modified organism), not further fragmentation once the GMO has been established.

UC Davis plant geneticist, Paul Gepts, comments that although Quist and Chapela's evidence is consistent with fragmentation during insertion, it is insufficient to really demonstrate it. And as for their claim about further fragmentation, he says they really should have provided significantly more evidence and explanation before jumping to such a novel and significant conclusion.

"The acknowledgement by Quist and Chapela that two of their sequences might be artifacts does not inspire confidence in their methodology," said Gepts.

Unstable genes?

Quist and Chapela's idea that transgenes are fragmenting and moving around in the host genome may go against the beliefs of most working genetic engineers, however Gepts stresses these are important questions to investigate and yet the idea has never been tested in a systematic way.

"Again, we just don't know. There is little research done on the stability of transgenic DNA," he said.

He says it has been known since the days of Barbara McClintock, that the plant genome has elements in it that move around. One concern is that the transgenic DNA, or parts of it, may end up in such "transposable elements" and affect the expression of proteins other than that encoded for by the transgene.

Gepts said genetic engineers usually require several attempts before they are successful at inserting transgenic DNA into a GMO. It is the successful inserts that are then selected for. In particular, genetic engineers look for a GMO that will have a stable phenotypic trait.
He suggests that taking a closer look at why transformation events fail may provide useful information in understanding genetic stability.

Former Calgene research head Robert Goodman concurs. Although the idea of transgenes moving around freely in a host genome is not generally accepted, he says a lot is not understood about what happens to transgenes once they are inserted.

"It is not surgically precise," he said. "It's a bit of a fling at the genome. Odd things happen."

"Why some lines don't continue to perform are not understood," he said. "There are so few people who have looked at those kinds of questions."

Goodman says the fact that plant's appear to have a "plastic genome" complicated the question of whether transgenes could be moving around in the host genome,

"He said separating out the behaviour of transgenes against the background of how plant genes behave is an area that is in its infancy," he said.

Some scientists, such as Mae-Wan Ho of the UK's Open University, argue that transgenes can disrupt the plant's own regulation of its "fluid genome". For her, this is a warning sign against the wisdom of genetic engineering.

However, while Goodman says concerns about the possible unpredictable effects of the Cauliflower Mosaic Virus (CaMV) promoter in transgenic constructs may be legitimate, he believes that with more understanding it will be possible to predict and control negative impacts of transgenes.

"I'm not trying to be scary. I'm just saying there is a background of this and we don't really understand it much," he said.

"We don't know all the rules yet and we could do something that could have unexpected consequences. I think we should proceed with appropriate caution."

What exactly is meant by "appropriate caution" is of course a matter of judgment.

For example, while scientists such as Mae Wan Ho argue it's too dangerous to proceed at this point, University of Queensland molecular biologist John Mattick argues such concerns are unwarranted.

"DNA is dynamic, so I would accept that in principle DNA in cells may get rearranged from time to time, although in reality this is a very rare event. In any case the odds that this is going to produce anything that is going to have any ability to compete with wild populations is effectively nil," he said.

Meanwhile, the yet-to-be implemented Cartagena Biosafety Protocol refers to the stability of inserted genes in its section on "potential risks" of GMOs. And a new European directive on the deliberate release of GMOs will require companies to prove the genes in the GMO they wish to release are stable over successive generations.

According to Mae Wan Ho, it is the prospect of having to prove genetic stability that most worries the biotechnology industry.

Due process?

So where does all this leave Nature's retraction?

Papers are usually only retracted when the author acknowledges there are mistakes or serious fraud has been shown to occur. So what line did Quist and Chapela cross to deserve the treatment they got?

Scientists who support Quist and Chapela claim Nature should have let the normal scientific process of contestation proceed, using Quist and Chapela's claims and data to repeat, verify or refute their findings without additional editorial comment.

Quist and Chapela's detractors, on the other hand, claim the journal should not have reprinted the article and expected the public to decide for themselves whether this was good science, but should have retracted the paper properly and apologised for misleading everyone.

Early this month BBC Newsnight accused Nature of ignoring the advice of most of its advisers when it decided to retract the Quist and Chapela research.

Editor Philip Campbell was asked why he retracted the whole paper the focus of concerns was on the "secondary" findings about fragmentation of the transgenes, and when only one referee had recommended retraction of the whole paper.

"What we were doing was giving our judgement based on a variety of pieces of advice we'd received," he told the program. "The paper as a whole shouldn't have been published."

Doubts linger, however, to the extent Nature could have been immune to the powerful forces at work and the influence such forces could have had on the "variety of pieces advice" received by the journal.

Negotiation over acceptable methods and interpretation are a normal part of scientific discourse although but for controversies like this they don't usually happen in public.

Donald Kennedy Editor in Chief of Nature's rival, Science, recently gave some insight into backstage negotiations over a controversial paper on cold fusion. He described the pressure on his journal from distinguished scientists and others not to publish and comments the case is not unique.

The motivations of those who try to stop publication, he said, sometimes rest on "serious legitimate scientific differences of opinion, although sometimes that is no so clear".

Professional pride, policy implications and financial matters are just some of the contexts that can surround such negotiation.

"The peer review process is like democracy, it's the least imperfect system," comments Paul Gepts.

There has been much discussion about improving the system of scientific publication by disclosing the conflicts of interest of authors.

Italian clinical psychologist and medical journal editor Giovanni Fava thinks the system could be improved further if this requirement was extended to referees themselves:

"Disclosure of conflict of interest becomes crucial when data are interpreted," he says.

Certaintly there have been claims that Nature's failure to reveal the interests of players in the drama, including the one referee which demanded a retraction of Quist and Chapela's article, and the journal's own reliance on advertising from biotechnology companies, has done little to engender trust in the peer review system.

Perhaps it's time for a wholesale overhaul of the process of peer review and scientific publication to ensure the transparency that's necessary in a context where there is concern that commercial pressures are threatening the integrity of independent science?