View comments(10) | Comment on this blog   

The conversation is not over regarding two recent retractions of papers on enzyme engineering. Two letters published this month in Science say that the explanation of retraction issued by Homme Hellinga's group at Duke University does not account for many of the errors in the original publications. The Grantsmanship blog has also collected long running commentary on the circumstances surrounding the retractions.

On February 1, Hellinga's group issued a statement of retraction for their 2004 Science paper on redesigning ribose-binding protein (RBP) to catalyze triose phosphate isomerase (TIM) activity -- a reaction crucial to glycolysis in almost all types of cells. They explained that John Richard, in the department of biochemistry at the State University of New York, Buffalo, found that the TIM activity of the new enzyme was due to wild-type contamination during his own experiments with the redesigned enzyme. A month later the group retracted a second paper on the same redesigned enzyme from the Journal of Molecular Biology.

But the two letters, one written by Richard himself, published electronically on March 10, both noted that key measurements of enzymatic activity were wrong even if there had been wild-type contamination. In particular, the Michaelis Constant (Km) -- a measurement of enzymatic activity -- in the papers was an order of magnitude lower than the wild-type. Had wild-type contamination truly been the error, then the Km values would have been the same as wild-type Km values. "There's no way you can account for that, no explanation," Jack Kirsch, professor at the University of California, Berkeley, and author of one of the letters, told The Scientist.

The 2004 Science paper "was a very important paper even though it was wrong in some sense," David Baker, from the University of Washington, told The Scientist. In 2006, Baker wrote a feature for The Scientist in which he cited Hellinga's groundbreaking work. Baker's group recently published papers in Science and Nature on enzyme design, citing Hellinga's paper in both. "I felt strange removing the reference because I felt the citation should be in there." He added that Hellinga's papers had formed the groundwork for much of what his group did.

Richard told The Scientist that he discovered the enzyme's inactivity during a routine purification. Given the notoriety that Hellinga has gained from this work, he added, people want to know how he made such a mistake. The Science paper has been cited more than 160 times.

In addition, Hellinga's group had reported that their redesigned enzyme could restore the activity of a TIM knockout E coli. If the in vitro enzyme showed no real activity, Kirsch asked, how could it have restored the knockout? Shortly after the original paper appeared in Science in 2004, Hellinga went to give a seminar in Berkeley to present his new findings. Kirsch said he brought up the issue with the Km and asked to see Hellinga's data but never received it.

Hellinga did not respond to several calls for comment. In a February 13 article, Nature reported that Hellinga told them that his lab had been using the wrong purification method during the experiments reported in Science in 2004. "A mistake was made, and nobody caught it -- including myself," Hellinga told Nature. "We were concerned it might not have been innocent, but it was." Duke University began an investigation of misconduct of the paper's first author, Mary Dwyer, in September of 2007 but cleared her on February 4.

Steve Mayo, a prominent protein designer at Caltech, declined to comment for this article. Several other researchers who have cited Hellinga's papers did not return calls for comment.