We love history at Retraction Watch, but with few exceptions, such as covering what seems to have been the first-ever English language retraction in 1756, the daily march of retractions doesn’t leave us much time to take steps back. So we’re very glad to be able to present a guest post by our friend Jeff Perkel about a classic paper that scientists have known to be wrong for most of its nearly 60-year-life — and yet remains in the literature.
The date is December 31, 1952. Linus Pauling, the CalTech wunder-chemist who had recently solved the secondary structure of proteins by describing the alpha-helix and the beta-sheet, has just submitted a “Proposed Structure for the Nucleic Acids” to the Proceedings of the National Academy of Sciences. The PNAS description appears in February 1953 and runs for 14 pages, with seven figures and two tables (compare that with Watson and Crick’s one-pager two months later in Nature).
Pauling also submitted a short note to Nature to alert that journal’s readers to the basics of the PNAS paper. The Nature note appeared in the journal’s February 21 issue:
The structure involves three intertwined helical polynucleotide chains. Each chain, which is formed by phosphate di-ester groups and linking b-D-ribofuranose or b-D-deoxyribofuranose residues with 3´, 5´ linkages, has approximately twenty-four nucleotide residues in seven turns of the helix. The helixes have the sense of a right-handed screw. The phosphate groups are closely packed about the axis of the molecule, with the pentose residues surrounding them, and the purine and pyrimidine groups projecting radially, their planes being approximately perpendicular to the molecular axis…
That structure – a triple helix with the phosphates in the middle and the bases radiating outwards – was similar to one Watson and Crick had first advanced a year earlier and then rejected on both chemical and physical grounds. It also in no way accommodates Chargaff’s observation that the abundance of A in DNA approximately equals T, and C equals G. It also fails to explain the biology and replication of DNA.
As detailed in the excellent Oregon State University archive, “Linus Pauling and the Race for DNA,” around Christmas 1952, Pauling wrote of his discovery to a Cambridge chemist.
I have practically no doubt. . . The structure really is a beautiful one.
Yet in the PNAS paper, says the OSU archive, Pauling was “uncharacteristically tentative.”
This was “a promising structure,” Pauling wrote, but “an extraordinarily tight one”; it accounted only “moderately well” for the x-ray data and gave only “reasonably satisfactory agreement” with the theoretical values obtained by the Crick formula; the atomic positions, he wrote, were “probably capable of further refinement.”
Still, sometimes being first is better than being absolutely right. Pauling was aiming for
not the last word on DNA but the first. He wanted the initial publication that would be cited by all following. It did not have to be precise.
Watson described his feelings upon reading the Pauling manuscript on page 102 of The Double Helix:
At once I felt something was not right. I could not pinpoint the mistake, however, until I looked at the illustrations for several minutes. Then I realized that the phosphate groups in Linus’ model were not ionized, but that each group contained a bound hydrogen atom and so had no net charge. Pauling’s nucleic acid in a sense was not an acid at all. Moreover, the uncharged phosphate groups were not incidental features. The hydrogens were part of the hydrogen bonds that held together the three intertwined chains. Without the hydrogen atoms, the chains would immediately fly apart and the structure vanish.
At physiological pH, those hydrogens would disappear, and so would Pauling’s structure. Today, of course, we know that the structure of DNA is a double-stranded, antiparallel helix with the sugar-phosphate backbone on the outside and the bases paired in the center like the rungs of a ladder. It is, undeniably, an acid.
How could Pauling have made such a mistake? Everybody had an opinion, says the Oregon State archive, but they basically came down to two: “hurry and hubris.” Pauling didn’t take “a serious stab” at DNA until late November 1952, and then devoted basically a month to hammering it out. By comparison,Pauling’s work on the protein alpha helix
was the result of more than a decade of off-and-on analysis and thousands of man-hours of meticulous crystallographic work.
More to the point, Pauling wanted to be first to the prize, beating out his UK rivals led by Sir William Lawrence Bragg.
He rushed, and he thought he could get away with it because of his pride in his own ability. He wanted the prize, he gambled, and he lost.
Sixty years later, that gamble remains on the books. “A Proposed Structure for the Nucleic Acids” has never been retracted. It has been cited 109 times, including at least once in 2012.
Pauling died in 1994. His coauthor Corey died in 1971. Still, the question remains: Should a paper, known for over five decades to be dead wrong, be retracted? Does it even matter?
Weigh in on our poll prompted by the lively discussion in our comments.