Retraction Watch

Tracking retractions as a window into the scientific process

Harvard biologist retracts diabetes breakthrough in Cell

with 2 comments

2013 probably felt like it was going to be a great year for stem cell biologist Douglas Melton at Harvard. He had published a buzz-worthy paper in Cell about a new way to potentially boost insulin in diabetics, attracting significant media attention, and eventually gathering nearly 200 citations.

But 2016 is closing out on a less positive tone for Melton — today, he and his colleagues are retracting the paper, after multiple labs (including his own) couldn’t reproduce the findings.

Although the lab has itself already published two articles casting doubt on the original findings, Melton told Retraction Watch he chose to retract the paper to ensure there was no confusion about the original paper’s validity:

I wanted to make sure anyone doing a PubMed search would see this is our present view.


For two years, the researcher community has raised questions about the main findings of the 2013 paper — namely, that a hormone dubbed betatrophin could increase the production of beta cells, which produce insulin.

In 2014, a group of authors led by researchers at Regeneron Pharmaceuticals published a paper in Cell questioning the findings, casting doubt on the link between the hormone and beta cells. In that same issue of the journal, Melton and his co-authors published a correspondence, concurring with these findings:

…we agree with the main conclusion of Gusarova et al., namely that deletion of Angptl8/betatrophin itself does not support the idea that betatrophin alone is a capable of inducing pancreatic beta cell proliferation. The mechanisms controlling beta cell replication are clearly more complicated than we put forth…

That Correspondence by Melton et al generated a heated discussion on PubPeer, with some suggesting the authors should have issued a retraction notice, rather than a Correspondence that notes their results are problematic (but not all agreeing). Meanwhile, in 2015, another group replicated the original findings in rats.

To get to the bottom of the science, Melton asked two other labs — including Jake Kushner at Baylor College of Medicine — to collaborate:

Rather than just letting the controversy slowly die away, I contacted Jake and asked would he be willing to do a blinded experiment to find out how we could be getting such different results.

In July 2016, the researchers published a paper in PLOS ONE that concluded the original results were not valid:

Overall, the hypothesis that [betatrophin] induces dramatic and specific β-cell proliferation can no longer be supported…As an added note, we would like to commend collaborative group efforts, with repetition of results and procedures in multiple laboratories, as an effective method to resolve discrepancies in the literature.

The paper concludes:

One of the two main conclusions of the original paper describing the betatrophin hypothesis needs to be withdrawn.

It’s painful to retract a paper, but the experience showed him how science works best, Melton told us:

It’s an example of how scientists can work together when they disagree, and come together to move the field forward…The history of science shows it is not a linear path.

In retrospect, he said he wished he’d performed the original experiment with more mice (“more attention to the statistical strength is a lesson that I’ve learned”), but the experiments with Kushner and his colleagues taught him the primary mistake was that the lab had miscounted the number of beta cells, which are relatively rare — so should have been tracked under higher magnification.

We’ve learned how to do the experiment better.

Although Melton’s lab had now produced two papers casting significant doubt on the 2013 Cell paper, he told Retraction Watch he wanted to issue a retraction so there was zero confusion about the status of the paper:

I thought it would be most unfortunate if a lab missed the PLOS ONE paper, then wasted time and effort trying to replicate our results.

He added that one aspect of the original paper remains valid — namely, that a molecule known as S961 can block the insulin receptor, which causes a robust replication of beta cells. That was an “unprecedented” finding, Melton said:

I hope that won’t cause any confusion – there are several labs that have reproduced that finding, and many labs use that S961 molecule.

The next step for the field, he said, is to figure out why adding an insulin antagonist boosts beta cells, now that it’s clear the reason is not betatrophin:

We don’t think it’s betatrophin, but we don’t know what it is…I’ll be glad when I’ve moved on to figure out what the real cause of beta cell replication is.

Here’s the retraction notice, posted today by Cell:

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

In this article, we claimed that Angptl-8, which we termed betatrophin and is also referred to in the literature as TD26, RIFL, C19orf80, and Lipasin, was the agent by which the insulin antagonist S961 induces robust β-cell replication in mice. Gusarova et al., in a Matters Arising article, subsequently showed that targeted Angptl-8/betatrophin overexpression in mice increases blood triglycerides but does not induce β-cell growth (Cell, 2014, 159, 691–696). When we repeated our original experiments with a larger number of mice, we also failed to observe β-cell expansion upon Angptl-8/betatrophin overexpression and reported these results in a Correspondence (Cell, 2014, 159, 467–468). We have subsequently repeated a series of blinded experiments with the Kushner lab and have now determined conclusively that our conclusion that Angptl-8/betatrophin causes specific β-cell replication is wrong and cannot be supported (PLoS One, 2016, 11, e0159276). Therefore, the most appropriate course of action is to retract the paper. We regret and apologize for this mistake.

Betatrophin: A Hormone that Controls Pancreatic b Cell Proliferation” has been cited 194 times, according to Clarivate Analytics’ Web of Science, formerly part of Thomson Reuters — more than 100 times since 2014, when the Cell paper led by authors at Regeneron was published.

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.

Written by Alison McCook

December 27th, 2016 at 10:00 am

  • herr doktor bimler December 27, 2016 at 1:23 pm

    Melton’s lab had now produced two papers casting significant doubt on the 2013 Cell paper
    Three papers out of one laboratory / statistical error is not bad going.

  • Deepa Bhartiya January 4, 2017 at 4:01 am

    Melton’s group should also re-look at their existing concept of endo-reduplication of pre-exisisting islets to regenerate pancreas rather than a role for pancreas stem cells. They may refer to two publications from our group (;

  • Post a comment

    Threaded commenting powered by interconnect/it code.