Paper About Reactome Array Stirs Controversy

View Enlarged Image Adapted from Science

Revisions Much of the chemistry originally depicted in Figure 1 of the reactome paper has been revised (red). Still unclear is the nature of the cobalt-linker complex, depicted by the green blobs. Poly(A) is a tract of multiple adenosine monophosphates.

A biochemistry paper published in the Oct. 9, 2009, issue of Science is generating controversy on blogs, Twitter, and other networking forums. The paper describes a reactome array, a sensitive metabolite array for obtaining detailed quantitative profiles of a cell's metabolic networks (Science 2009, 326, 252). It has, at worst, been viewed as fraudulent and, at best, as a glaring example of the pitfalls of refereeing interdisciplinary research.

The outcry has been so intense that Science Editor-in-Chief Bruce Alberts has issued an "Editorial Expression of Concern" to alert the journal's readers "to the fact that serious questions have been raised about the methods and data presented" in the article. Alberts writes that the journal has "requested evaluation of the original data and records by officials at the authors' institutions: These officials have agreed to undertake this task."

One of the first people to raise concerns about the paper was Laura L. Kiessling, professor of chemistry and biochemistry at the University of Wisconsin, Madison, and editor of ACS Chemical Biology. "The chemistry just doesn't make any sense," Kiessling says. Like many other experts, she is puzzled by the highly unstable array linkages depicted in Figure 1 of the paper, characterizing them as unlikely at best.

"The more I looked at Figure 1, the less I understood," says Timothy Mitchison, professor of systems biology at Harvard University. The figure depicts the reactome strategy, in which a small-molecule substrate is joined to both a dye and a cobalt-containing linker. When a metabolic enzyme reacts with the substrate, the dye is purportedly released, giving off a glowing signal, and the cobalt linker captures the enzyme. The array is said to be made of over 1,000 such constructs.

In particular, Figure 1 leaves unclear what sort of cobalt-linker complex is proposed, says Ben G. Davis, a chemical biologist at Oxford University who called for more scrutiny of the work on the "Faculty of 1000" website.

The synthesis of the array components described in the supporting information raised serious doubts, Mitchison says. One intermediate contained an improbable anhydride, and the team claimed to form tough-to-make bonds such as aryl sulfonamides without a necessary activation step.

The study's authors stand behind their array. Figure 1, "as published, contains some mistakes that have been corrected and sent to Science," says co-corresponding author Manuel Ferrer of the Spanish National Research Council's (CSIC) Institute of Catalysis, in Madrid. A corrected figure and supporting information are available on a website hosted by CSIC, he says.

Davis notes that rigorous structural verifications of the team's intermediates are still largely absent from the revised supporting information. Ferrer says that all the team's structural information will be provided to the CSIC commission charged with evaluating the data, "and after that, they will be incorporated to the Web page."

"It's great that biological people are trying to use chemistry," Kiessling says, but she worries that the chemistry in some instances doesn't get enough scrutiny. "There should be appropriate reviewers," she says, noting, as have others, that no chemists reviewed the reactome paper. "There should be high standards," she adds.

Science will continue to monitor its review processes, in particular the review of supporting information, Alberts tells C&EN. "Reviewers are often overwhelmed by the amount of information contained there," he adds, "especially when an immense amount of information is added in a revision, as happened in this case."

Chemical & Engineering News

ISSN 0009-2347

Copyright © 2011 American Chemical Society