Journal refuses to publish scientist’s defence of his paper warning of unexpected outcomes from GM mosquitoes release. Report: Claire Robinson
As the citizens of Florida and Texas prepare to act as subjects in an experiment in which millions of Oxitec's genetically modified mosquitoes would be released in their states, a scandal has emerged around a journal's treatment of a scientific article that drew attention to unanticipated outcomes and risks of the project.
The article, published on 10 September 2019 in the Nature-owned journal Scientific Reports, was authored by Jeffrey R. Powell of Yale University and colleagues. The scientists reported the results of experimental trials held in Brazil, in which genetically engineered mosquitoes produced by the biotech company Oxitec (owned by Intrexon) were released. The yellow fever mosquitoes (Aedes aegypti) were genetically engineered to make it impossible for their offspring to survive. Oxitec claimed its mosquitoes had a lethal gene that made them "self-limiting". After release the GM mosquitoes were supposed to mate with female mosquitoes of the species, which transmit infectious diseases, to reduce the natural population.
But there was a surprise. Dr Powell and his colleagues found that some of the male mosquitoes that Oxitec released in Brazil had successfully interbred with the local mosquito population, and that their hybrid offspring were now spreading out of control and beyond the release area. This was despite Oxitec claiming its mosquitoes had a lethal gene that made them "self-limiting" (a claim Oxitec subsequently removed from its website but which is still visible in the archived page).
The paper came under immediate attack from Oxitec. That’s not surprising, as many papers exposing risks and problems with GMOs are attacked by the industry. But what was different in this case is that joining the attack was one of the paper's co-authors, Margareth Capurro of the University of São Paulo in São Paulo, Brazil. Capurro and five Brazilian co-authors even called for the paper's retraction.
Capurro's objections are laid out in articles for the Brazilian science publication Questão de Ciência (A Question of Science) and Science magazine. They boil down to:
* she didn't see the final version of the paper prior to submission
* the hybrid mosquitoes didn't carry the transgene and thus "there are no GM mosquitoes flying over Brazilian Northeastern skies", but this is not stated in the text
* there was nothing "unexpected" about the finding of hybridization, and
* she objects to the statement in the paper that the hybrid population of mosquitoes may be "more robust" than the pre-release population "due to hybrid vigor".
Capurro largely blames the lead author, Jeffrey Powell, for these alleged shortcomings.
In response to the criticisms, on 17 September, just 7 days after the paper's publication, the editors of Scientific Reports published an editor's note at the foot of Powell et al's article, stating, "readers are alerted that the conclusions of this paper are subject to criticisms that are being considered by editors". Then in May this year, the journal published an "expression of concern" about the paper, which closely reflects Capurro's points. While this is not the same as a retraction, the move could have the effect of decreasing confidence in the paper amongst the public and regulators.
The journal's editors state that they "received a response to the concerns from the corresponding author" – Jeffrey Powell – and "sought further advice from expert peer reviewers regarding both the issues raised and the response received. The reviewers confirmed that the scientific concerns are valid and should be addressed."
The editors add that they "have offered the authors the opportunity to submit a Correction which will be peer reviewed. However, the authors have not notified the Journal that they have been able to reach agreement on the content of a Correction that would fully address the issues raised."
They note that while six authors agree with the expression of concern, the four remaining authors disagree with it.
What the editors do not say is that they refused to publish Dr Powell's response. Dr Powell has sent us his response and we have published it in full, with his permission, on our website.
Below we publish the journal editors' concerns along with Dr Powell's responses to them.
Editors: "The title ['Transgenic Aedes aegypti Mosquitoes Transfer Genes into a Natural Population'] does not make it clear that the authors only examined genomes of specimens that lacked the transgenes and sampled during the release period;"
Dr Powell: "The title is accurate and informative; no editor or reviewer found fault with it during review process. It is clearly stated in the text, and data in the figures indicate, that we genotyped fluorescent larvae carrying the transgene separately from wildtype larvae not carrying the transgene. It is the wildtype larvae that give evidence of introgression which is clearly what is stated in the paper and therefore it is abundantly clear the transgene itself is not introgressing."
Editors: "The Abstract and Introduction use language which is not justified given the evidence present in the peer reviewed literature and the data presented in this Article. No sampling for this study was conducted more than a few weeks after the release program, and as such there is no evidence in the Article to establish whether the non-transgenic, introgressed sequences from the released strain remained in the population over time. Furthermore, previous work from some of the authors (Reference 6 in the Article) showed that over time, the transgene is lost from the population, but the Article does not disclose this information;"
Dr Powell: "Prior to publication neither reviewers nor editors found the Abstract or Introduction inaccurate or misleading. The issue of whether the introgressed sequences remain in the population over time is clearly addressed in the paper. We explicitly point out that the proportion of introgressed individuals decreased over time (Table 1 and supplementary table E2). We state in the Discussion 'This observation also implies that introgressed individuals may be at a selective disadvantage causing their apparent decrease after release ceased...' As to not stating the transgene is lost from the population, this is moot as there is no evidence (nor do we claim) it was ever in the population."
Editors: "In the Discussion, the authors claim that because of the distinct genetic backgrounds of different mosquito populations (two used to create OX513A mosquitoes, and one local population), the existing population in Jakobina is more robust than the original wild population due to hybrid vigour. There are no data in the Article to support this point; furthermore, data included in the Article indicate that a number of hybrid individuals rapidly declined post-release.”
Dr Powell: "Twice, in the Abstract ('It is unclear how this may affect disease transmission or affect other efforts to control these dangerous vectors.') and Discussion ('It is not known what impacts introgression from a transgenic strain of Ae. aegypti has on traits of importance to disease control and transmission.'), we explicitly state that we do not know what impact the observed introgression has with regard to disease transmission or control. We do point out that, because the three populations now making up the Jacobina population are genetically distinct (supplementary figure E2), the genetic diversity of the population has increased which 'likely' leads to a more robust genetically diverse population. This phenomenon is very well known and established in evolutionary genetics."
Editors: "The conclusion of the Article highlighting 'the importance of having in place a genetic monitoring program during such releases' could be misunderstood to mean that such program was not in place. The Mosquito release program in Jakobina is monitored by the Brazilian regulator, the National Technical Commission of Biosafety (CTNBio)."
Dr Powell: "To our knowledge, this is the first detailed follow-up genetic monitoring of the effect of transgenic mosquito releases on a target population. It is simply not true, in any meaningful sense, that adherence to Brazilian regulations involved any genetic monitoring follow-up of the effect of the releases. Our point is that such regulators should considering requiring the kind of study we conducted."
Editors: "When contacted about these issues, some of the authors indicated that they had not approved the final version that was submitted for publication."
Dr Powell: "All authors saw versions of this manuscript prior to publication. It was a misunderstanding of journal policy that inadvertently led to not all authors being sent the final version by the corresponding author."
Giving critics the last word
Why did the journal not publish Dr Powell's response? In the interests of fairness to all of the authors as well as scientific transparency, this is what should have happened. Dr Powell told GMWatch that he asked the editors the same question, but they only cited "policy" – whatever that means.
In our view, Dr Powell successfully addresses all the criticisms raised in the editors' expression of concern. Could it be that the journal editors wanted the critics of the paper to have the last word and to deny readers – including the scientific community, regulators, and the residents of areas where GM mosquito releases are planned – the chance to judge for themselves the strength of their arguments?
Dr Powell and his team's findings were at odds with what people living in the areas where GM mosquito field releases have been taking place have been told: that Oxitec’s mosquitoes are incapable of successfully reproducing. This is confirmed by a paper published in the journal Emerging Infectious Diseases, which reports on the findings of a survey into local residents’ "awareness and support" for the field release of Oxitec’s GM mosquitoes in Florida. The authors wrote that because these male mosquitoes’ mating "results in death of offspring in the larval or pupal stage of gestation… outreach activities in the area preceding the survey referred to the mosquitoes as ‘sterile.’"
So local residents were given the impression Oxitec’s GM mosquitoes could NOT produce viable offspring.
In Brazil, too, local people were kept in the dark about the risks, according to Dr José Maria Gusman Ferraz, a member of of Brazil’s biosafety commission (CTNBio) which approved the GM mosquito project. Dr Ferraz noted Oxitec's failure to make clear the survival rate of the GM mosquitoes and the lack of studies on the possible consequences of their mating with local mosquito populations. He said, “The commission was warned about the possibility of this mosquito establishing itself permanently in the environment, as well as about crossbreeding between the GM and wild mosquitoes, but most members of the commission completely dismissed these concerns” – and voted in favour of the release.
It could be argued that the Scientific Reports editors have aided these attempts to pull the wool over the public's eyes in refusing to publish Dr Powell's defence of the paper.
Conflicts of interest
Capurro's attempts to censor a paper that she co-authored must be viewed in the light of the fact that she is not independent from Oxitec. She had an agreement with Oxitec to monitor the effects of the Brazil release. She seems to have been collaborating with Oxitec for years. She is the coordinator for the evaluation of all field releases of Oxitec GM mosquitoes in Brazil. And she has co-authored papers with multiple Oxitec affiliates.
Capurro is also the "principal investigator" for a separate project, not involving Oxitec, to produce transgenic mosquitoes. A press release notes that the project Capurro is leading "has already inspired the creation of several biofactories to produce transgenic insects around the world”. So Capurro’s research centres almost entirely on projects that are completely dependent on the public acceptance of GM mosquito releases.
Astonishingly, however, in the "competing interests" of the Scientific Reports article, all of the authors declared that they had none. While the journal editors note in their "expression of concern" that "concerns were also raised about potentially undisclosed competing interests", they add that they "received confirmation from all of the authors that they have no potential competing interests".
This is clearly unsatisfactory and counter to the interests of scientific objectivity. The editors should have required that all authors declare their conflicts of interest.
Scientists criticise conflicts of interest in Florida/Texas experiment
Regarding the plan to release Oxitec's GM mosquitoes in Florida and Texas, five eminent scientists have criticised the untransparent nature of the US EPA's favourable risk assessment, as well as the agency’s dismissal of public opinion: 31,174 public comments were received opposing the release and only 56 in support. Yet the EPA still gave its approval.
The scientists also draw attention to the conflicts of interest arising from Oxitec's involvement in the trials. They say, "There is a potential bias and conflict of interest when experimental trials and assessments of ecological risk lack political accountability and are performed by, or in close collaboration with, the technology developers. This scenario becomes more troubling with a for-profit technology company when cost- and risk-benefit analyses comparing GM mosquitoes to other approaches aren’t being conducted."
To this conflicted situation we now have to add a journal’s participation in obscuring the risks of the project. The editors of Scientific Reports should, in the interests of giving the public the full picture, publish Dr Powell’s response and apologise to the authors who did not agree with the “expression of concern”.