When independent replication isn’t really independent – backtracking monitoring

Matt Warman

My lab at Boston Children’s Hospital and Harvard Medical School studies genetic diseases that affect the skeletal system. We became interested in the osteocalcin protein after Gerard Karsenti of Columbia University reported in several research papers using out-of-strength mice – mice that lacked the genes that produce osteocalcin – that osteocalcin is a bone-derived hormone that affects glucose metabolism, insulin production, and male fertility. , and muscle mass. , and realization. If osteocalcin acts similarly in humans, osteocalcin becomes an exciting and clinically important protein.

To independently confirm these results, we generated our own osteocalcin knockout mouse strain. We examined glucose metabolism and male fertility in our rats and found none of the effects reported by Karsenti and colleagues; we We reported our findings in May 2020. A group in Japan created a third osteocalcin-knockout mouse strain that also failed to confirm by Karsenty and colleagues. the findings.

In previous years my lab was also unable to independently confirm other results reported by Karsenty’s group: Paper I co-authored In 2011, no evidence was found that the LRP5 Wnt receptor affects serum serotonin levels, contrary to what Karsenty Lab Publishing.

While I was puzzled by our inability to independently confirm osteocalcin Karsenty’s findings, I contacted Thomas Clemens at the Johns Hopkins Medical Institute, whose lab appeared to have had more success validating Karsenty’s research. In fact, Clemens was the corresponding author in 2010 Paper published in the cell Which appears to confirm one of Karsenty’s main findings: the hormonal effect of osteocalcin on blood glucose. When I wrote to Clemens in June 2020 asking him for his detailed experimental methods, I made an intriguing discovery that only heightened my concern about Karsenti’s published findings.

I specifically asked about the experiment mentioned in Figure 5 of Clemens’ paper, which showed that osteocalcin infusion reduced hyperglycemia in insulin receptor (IR) knocked-out mice. Clemens’ initial email response indicated that the experiments were conducted about a decade ago. He wrote that the detailed methods might be hard to find in “old notebooks” and that he would “talk to the postdoc who was working on this.”

Then, he wrote to me to point out that the experiment in Figure 5 was not actually performed by him, any of his co-authors, or anyone acknowledged in his 2010 paper. Instead, the data I queried was provided by Karsenty’s lab:

not shiny,

Sorry for the delay in getting back to you.

Our study reviewed osteocalcin infusion in IR KO mice (Fig. 5, cell 142, 309–319). This study was conducted in collaboration with Girard’s group who synthesized and validated the recombinant peptide. I suggest you talk to him about the details if you need them.

It is considered,


And the

not shiny,

I apologize for the lack of clarity in my last email. This study was performed in Girard’s laboratory by Mathieu Veron (transcribed above). I’m sorry I can’t provide more details.

It is considered,


Therefore, Clemens’ paper did not independently confirm Karsenty’s findings, but instead reported on the data obtained from Karsenty’s laboratory. This collaboration is not mentioned in Clemens’ paper or in Karsenty special article Posted in the same issue of hive. As a reader who did not see any intervening individuals listed as co-authors or included in the acknowledgments, I have concluded that the work was done independently, not collaboratively.

The accompanying Cell editorial reinforced my impression by describing the two papers as coming from “independent labs,” with Clemens’ group leaking osteocalcin. a Article review The following year, Clemens and Karsenty co-authored it, also stating that their work was done independently, and in 2016, Clemens published Article review He stated that the findings of osteocalcin were independent of Karsenti.

In my nearly 40 years as a scientist, I have never before encountered a situation in which an investigator published successive experiments while failing to acknowledge the collaborator responsible for a portion of the work and without the collaborator apparently raising any objection.

Since both scientists and funding agencies take claims of independent validation of research into account when designing trials and providing funding, I shared my concern that the authors were not sufficiently transparent about their collaboration with the US Office of Research Integrity (ORI), Jones. Office of Policy Compliance at the Hopkins Medical Center, Columbia University Vice President for Research Compliance, Training and Policy, and Cell Editor.

The email response I received from the ORI—which oversees research integrity in work funded by the National Institutes of Health, and which supported Clemens and Karsenty—indicated that it does not consider an investigator publishing another group’s data without attribution to be a problem if the group that provided the data does not object. The email I received from Johns Hopkins Medicine seems to regard it as a dispute over authorship rather than a scientific error. I never received a response from Columbia University.

My correspondence with Cell seems to have prompted revision by Clemens earlier this year. But this correction does not indicate the absence of independent verification. In his correction, Clemens wrote that Mathieu Veron “helped perform the osteocalcin infusion studies shown in Figure 5.” However, Clemens did not disclose that these studies were conducted entirely in Karsenty’s laboratory.

Nor does Clemence’s correction reveal that Karsenty was at the time of publication the founder and member of the scientific advisory board of Escoublac, Inc. The company whose business was based on Karsenty’s claim that osteocalcin plays a role in energy metabolism and related disorders.

More complete correction of Clemens cell paper: should include Veron and Karsenty as collaborators and co-authors; clarification that the Clemens laboratory has not independently shown that osteocalcin affects blood glucose levels; Disclosure of conflict of interest collaborator Karsenty.

The Clemens correction should also indicate, consistent with standard disclosures, whether the Columbia University Institutional Animal Care and Use Committee has approved the import of insulin receptor-knockout animals from the Clemens laboratory, if this has already occurred, and the infusion of osteocalcin into those animals. the animals. If the import of Clemens mice did not occur, the Clemens paper would appear to skew the actual mouse stress obtained from the data in Fig. 5. This would undermine scientific confidence in the robustness of Karsenty’s results across strains of mice.

The larger issue that this issue raises for science is whether, to avoid any doubt, the authorship/attribution guidelines themselves should require disclosure of all physical sources of data in papers. This would avoid the reviewers, readers and funders of erroneous conclusion of independently verified results when they were not validated.

Matt Warman is director of the Orthopedic Research Laboratories at Boston Children’s Hospital and professor of genetics and orthopedics at Harvard Medical School. He would like to thank Eugenie Reich, former science journalist and author of Fantastic plasticdescribing the fraud case of Jan Hendrik Schön, an attorney at Greene LLP in Boston, for helpful input.

Like undo watch? You can make a file A tax deductible contribution to support our workFollow us on Twitterlike us in Facebookadd us to a profile RSS ReaderOr subscribe to our website daily digestion. If you find undo it Not in our databaseCould you Let us know here. For comments or feedback, write to us at team@retractionwatch.com.

Leave a Comment