Ties that don’t bind: Group retracts parathyroid hormone crystallography paper

The authors of a 2008 paper alleging to have described how a particular protein binds to the parathyroid hormone have retracted it. The paper, “Structure of the Parathyroid Hormone Receptor C Terminus Bound to the G-Protein Dimer Gβ12,” has been cited 12 times, according to Thomson Scientific’s Web of Knowledge. According to the notice:

This article has been retracted: please see Elsevier Policy on Article Withdrawal(http://www.elsevier.com/locate/withdrawalpolicy).

This article has been retracted at the request of the Authors.

In this paper, a co-crystal structure was described of the G-protein heterodimer Gβ1γ2 bound to a C-terminal peptide from the parathyroid hormone receptor-1 (PTH1R) at 3.0 Å resolution (PDB id 2QNS). A subsequent refinement was later deposited in the Protein Data Bank (PDB id 3KJ5). While this structure represents a new crystal form of the Gβ1γ2 heterodimer, because of the lack of clear and continuous electron density for the receptor peptide in the complex structure, the paper is being retracted. We apologize for any confusion this may have caused.

Here’s the protein’s entry in the Protein Data Bank, if you want to check it out yourself.

We asked for some help interpreting the notice from someone who tipped us off to this retraction anonymously:

Basically, they fit to noise.  Most of this structure is previously characterized; the novel feature reported in the paper is the binding of a small peptide.  Unfortunately, there isn’t any reason to believe the peptide is present in the crystals – the electron density maps calculated from the diffraction data simply don’t show it.  The structure is not only wrong, it is exceptionally poorly optimized – curiously, the statistics reported in the paper do not match reality.

Our source described the image below, of a key finding in the original paper, in this caption:

The blue mesh surrounds all regions of the crystal where the electron density is one standard deviation above the mean - anything heavier than hydrogen should usually fall inside this. The peptide runs from left to right; the pink dashes are large overlaps between atoms. The density that surrounds some of the atoms in the peptide is probably either an artifact, or possibly some crystallization buffer molecules that tagged along.

So how did the authors come to realize that they’d made an error? We tried contacting corresponding author David Siderovski, of the University of North Carolina-Chapel Hill, but haven’t heard back. In the meantime, our interpreter has a hypothesis:

Since reviewers rarely receive the model or the diffraction data along with the manuscript, it probably would not have been obvious that something was wrong.  However, journals have for some time required deposition of these data in the Protein Data Bank as a condition of publication, which guarantees some form of post-publication peer-review.

It turns out the Structure paper wasn’t the only crystallography withdrawal this week. Nature issued a correction on Thursday of another paper, “Migrastatin analogues target fascin to block tumour metastasis,” that included the withdrawal of two structures, although the paper remains in the literature:

In this Letter, we reported the crystal structure of macroketone bound to fascin (Protein Data Bank number PDB 3LNA) (Fig. 2d). The chemical structure of macroketone was incorrectly shown and has been corrected (PDB 308K). We have been advised that the crystallographic data for this complex are not technically robust, and do not justify the conclusion that macroketone is bound as shown in Fig. 2d. We therefore regretfully withdraw the X-ray structure models PDB 3LNA and PDB 308K from this study. However, we believe that the rest of the Letter, including the observations made using the mutants, is not directly affected.

The study has been cited 20 times since it was published in April 2010.

Hat tips: Bob O’Hara, @biochembelle

6 thoughts on “Ties that don’t bind: Group retracts parathyroid hormone crystallography paper”

  1. Worrying.

    Of all your retractions this one should be cause for very serious concern. Crystal structures of this sort are relied upon at the inception of drug discovery work: if this is wrong, a huge wedge of money can be quickly thrown away.

  2. At what point can I write a paper detailing all the bogus crystallography I’ve found? In the end, I haven’t found anything as egregious as totally fitting the ligand to noise, but I have lots of examples of things that are not right.

      1. The post is from 2009, and still the paper is not retracted or cited as invalid. What happened to the request of deleting this structure from the PDB?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.