A University of Pennsylvania postdoc extensively manipulated data in three published papers according to an Office of Research Integrity
(ORI) announcement released last month.
"This is a fairly big case for us. There is a great number of falsified images," John Dahlberg, ORI's director of the Division of Investigative Oversight, told The Scientist
. "I would say this is up in the top third."
The postdoc, Kristin Roovers, was working on cell cycle dynamics in the lab of Morris Birnbaum
at the School of Medicine at the University of Pennsylvania in Philadelphia when the fraud was uncovered. The ORI investigation, published in the Federal Register on July 16, concluded that Roovers had faked 19 panels of Western blot data, which appeared on 11 figures in 3 studies published between 2001 and 2003, Dahlberg said.
Two of the studies, published in Molecular and Cellular Biology
and Developmental Cell
, were cited 53 and 31 times respectively before they were retracted in 2006. Corrections were recommended for a third study, published in Nature Cell Biology
, which has been cited more than 130 times.
The case was flagged in the spring of 2005 by editors at the Journal of Clinical Investigation (JCI
), who noticed irregularities in the figures of a paper that had been accepted for publication and was undergoing final revisions, said Birnbaum, who has served as a deputy editor at JCI
since March, 2007. That study, a collaboration between Roovers and Birnbaum and Anthony Muslin
of Washington University in St. Louis, examined the role of a protein kinase called Akt
in cardiac hypertrophy in mice.
Muslin induced hypertrophy either physiologically, by forcing the mice to swim, or pathologically, by placing a band around the aorta, Birnbaum said. Roovers used a more artificial model, in which the mice were infused with drugs or hormones. After sacrificing the mice, she sent the hearts to Muslin, who conducted assays to compare expression of Akt isoforms in the different cases.
"Much of [Roovers'] data couldn't be manipulated because the analysis was done in Washington University," Birnbaum told The Scientist
. However, he said, Roovers ran most of the controls for the study herself. The editors "found various pieces of Western blots cut and pasted either within a figure or between figures," said Birnbaum, who said. Muslin later confirmed the fraudulent JCI
data in his own lab and published the findings
After the manipulated data were traced to Roovers, the university opened an investigation in July, 2005, and sent its findings to the ORI in December, according to Glen Gaulton
, Executive Vice Dean of the School of Medicine at UPenn.
Roovers first joined Birnbaum's lab in 2003, a year after completing her doctorate with Richard Assoian
in the same department. She had moved with Assoian from the University of Miami, Florida, to UPenn when he took a position there, Gaulton said. She was "a perfect postdoc" and "an incredibly popular person in the lab," said Birnbaum. "Ironically, she was one of the smartest people there. People would bounce ideas off her before they came to talk to me."
When Birnbaum learned that the figures in the submitted paper were manipulated, he said, he and Assoian closely examined Roovers' past work in both their labs. "If you look at the papers in chronological order, they have increasing numbers of manipulations, till we came to mine where basically all of [the images] were manipulated," Birnbaum said.
The retracted papers examined the relationship between actin stress fibers and the protein integrin in Rho kinase signaling. "There's this long-standing question: What does tension do to promote cell cycle progression?" said Martin Schwartz
of the University of Virginia, who was a co-author on the Nature Cell Biology
study. "I thought [the Molecular Cell Biology
study] answered the question."
The findings, he said, suggested that tension triggers integrin clustering and integrin signals mediated cell-cycle progression. "It put integrin signaling in a primary role," he said. "I don't think any other studies have been directly based on that paper, but it has certainly affected my thinking on that question," he said.
of Harvard University, who cited the Molecular Cell Biology
study in his own work
describing a novel intermediary protein in Rho signaling, said that the retracted findings were unlikely to affect the landscape of the field. "[The study] linked the Rho pathway to the ERK pathway," said Stossel. "It seems to me that someone else came to a similar conclusion."
Assoian declined several times to comment on the case. "It is inappropriate for me to make any comment on the validity of the conclusions in the retracted papers," he wrote in an Email to The Scientist
. "As formally retracted papers, they should no longer be considered when reviewing the scientific literature. As the [Nature Cell Biology
] paper was not retracted, its conclusions are intact."
"[Assoain] is the last person this should have happened to," said Schwartz. "He's a really careful guy. When we were working on this paper, we talked endlessly about every little glitch in the data and what everything meant."
According to the ORI notice, Roovers has been banned for five years from working with any agency of the US government. Birnbaum and Gaulton declined to disclose contact information for Roovers, and other former colleagues did not return calls for comment. Birnbaum said Roovers had returned to Canada and had left research.
Links within this article:
The Office of Research Integrity
K. Roovers and R.K. Assoian, "Effects of rho kinase and actin stress fibers on sustained extracellular signal-regulated kinase activity and activation of G(1) phase cyclin-dependent kinases," Mol. Cell Biol.
, June, 2003.
K. Roovers et. al., "Nuclear translocation of LIM kinase mediates Rho-Rho kinase regulation of cyclin D1 expression," Dev. Cell
, August, 2003.
C.F. Welsh et. al., "Timing of cyclin D1 expression within G1 phase is controlled by Rho," Nat. Call Biol.
, November, 2001.
D. Secko, "Mending broken hearts," The Scientist
, August 11, 2003.
B. DeBosch et al., "Akt1 is required for physiological cardiac growth," Circulation
, May 2, 2006.
Y. Ohta et al., "FilGAP, a Rho- and ROCK-regulated GAP for Rac binds filamin A to control actin remodeling," Nat. Cell Biol,