Graham Ellis-Davies says January 25th was one of the worst days of his life. That was when the journal ChemBioChem retracted an article, published barely two weeks earlier, for a mistake Ellis-Davies blames squarely on himself.
The fallout has been nearly two months of painful self-recrimination, a tattered friendship and, perhaps most significant, he adds, the wasted effort of a postdoc who poured months of research into a now worthless publication.
The tale begins in the early 2000s, when Ellis-Davies, a renowned chemical biologist at Mount Sinai School of Medicine, first hooked up with Jeff Magee, then a promising neuroscientist at Louisiana State University. The two spent a week together at Magee’s New Orleans lab, collaborating on what would become a 2003 publication in the Journal of Physiology.
The researchers remained friends, and had a chance to work together again not long ago when Ellis-Davies and a few of his colleagues, including his postdoc Srinivas Kantevari, went to visit Magee at his new digs in Ashburn, Va., at the Janelia Farm campus of the Howard Hughes Medical Institute.
The two groups were interested in exploiting a technique that Ellis-Davies helped discover to get over the serious hurdle of solubility of molecules, such as drugs, in the body.
Solubility is really a fundamental problem, particularly for the central nervous system and peripheral nervous system. About 50% of a drug, by weight, is non-clinical material that promotes solubilization in plasma or other buffer.
For drugs that have to cross the blood-brain barrier — a membrane that acts as a filter to protect the brain from large, potentially toxic molecules — all that extra matter is like a sign that screams “don’t let me in,” making the development of effective neuroactive medications difficult.
A solution to the solubility problem would be to dispense with the chaperoning substance entirely. That, of course, raises a second issue: that the drug would become immediately active in the body, and thus far less useful, if useful at all.
Ellis-Davies has been working on ways of avoiding the solubility problem by keeping potentially therapeutic molecules inert until they can be activated by a signal. Ellis-Davies and his colleagues graft fat-loving molecules onto the end of neurotransmitters, rendering them inert. But by shining a finely tuned laser onto the fatty chain — which he describes as a kind of chemical antenna — they can activate the neurotransmitter.
The laser is so precise that it can stimulate the firing of a single synapse, thus releasing the same amount of a neurotransmitter as would be spit forth from a lone neuron.
The beam becomes like a conductor’s baton, a magic wand.
At least, that’s the theory. In fact, what Ellis-Davies discovered was that the idea is much harder to implement in practice.
The product of the one-day site visit (at which Magee agreed to provide the Mount Sinai team with tracers to conduct their experiments) was the hard-won realization that the creating a reliable “photochemical scaffold” for releasing neurotransmitters is maddeningly difficult, Ellis-Davies says.
Not wanting to squander the leg work, he and Kantevari wrote up a paper, “Development of Anionically Decorated 2-(ortho-Nitrophenyl)-Propyl-Caged Neurotransmitters for Photolysis in vitro and in vivo,” describing the effort. The article was not a boast about success, Ellis-Davies insists, but an effort highlight for other scientists the pitfalls of making the best possible scaffolds. Indeed, the discussion section carries a “cautionary note” for the field about how the problem is “not trivial” to solve.
It’s an account of how we didn’t solve the problem. We wrote it in a way to say we didn’t succeed.
Ellis-Davies says he showed the manuscript to collaborators at the University of Pennsylvania, who okayed the paper, but that he neglected to send a copy to Magee prior to submitting it to ChemBioChem. Although the journal asked Ellis-Davies to complete an electronic form vouching for the data and authorship, it did not require him to collect signatures from every author — a step, he says, that doubtless would have prevented what followed.
The article appeared on January 10. The retraction notice came 15 days later:
The article Srinivas Kantevari, Judit K. Makara, Attila Losonczy,Tommaso Fellin, Philip G. Haydon, Jeffrey C. Magee, and Graham C. R. Ellis-Davies, published online on 10th January 2011 in Wiley Online Library (www.onlinelibrary.wiley.com, 10.1002/cbic.201000254) has been retracted (25th January 2011) by the journal editor-in-chief, Dr. Peter Gölitz, and Wiley-VCH, in accordance with the “EUCheMS Ethical Guidelines for Publication in Journals and Reviews” adhered to by the journal (EUCheMS Ethical Guidelines for Publication in Journals and Reviews, 2006, http://www.euchems.org/Publications/index.asp). J.K.M., A.L. and J.C.M. were not informed about the manuscript prior to its publication and hold that the data presented in Figure 3 and related text are too preliminary for publication.
Magee, who has not responded to a Retraction Watch request for comment on the paper, contacted Ellis-Davies and said he had regarded the research as “anecdotal” and not for publication — at least, not with his name on it. A “very surprised” Ellis-Davies responded that he regarded the work “as proof of concept” and thus worth reporting. Such experiments are very common in his field, he told us.
In the end, Ellis-Davies realized that the group at Janelia would not consent to be part of the paper, and he agreed to retract it.
The paper couldn’t really exist without the work done in his lab. It has no meaning.
A contrite Ellis-Davies says the episode has been the worst of his career — indeed, of his entire life. “I made a huge mistake,” he says, insisting that Magee and his Janelia colleagues were “perfectly correct” to seek the retraction. And he feels especially for Kantevari, who is now back in India.
The first author has put a huge amount of work into it, he’s the one that’s most affected by it.
But he’s not ready to throw away the idea.
I hope to revisit the biology with another friend of mine, when the dust settles. In a way, I actually quite like the paper. I think the core of it is correct.
The next time he has a cross-lab collaboration, though, he’ll be sure to vet whatever manuscript that emerges with all of his co-authors:
I’m 53. Why am I still learning things the hard way?