CU Medical Center Scientist Retracts Flawed Research

Correction appended

When a Columbia scientist published a paper in 2002 challenging a
prevailing theory in developmental biology, he raised a few
eyebrows. But when he revealed earlier this month that one of his
lab members had fabricated much of the data, he attracted a bit
more attention.

Gary Struhl, a professor of genetics and development at Columbia
University Medical Center, co-authored the original paper with
Siu-Kwong Chan, then a post-doctoral fellow in his lab. According
to the retraction, printed in the Feb. 6 issue of the journal Cell,
Chan invented or manipulated much of the experimental data on which
the paper was based.

Struhl's lab studies proteins that direct spatial development in
embryonic fruit flies. The controversial 2002 paper concerned the
details of a kind of molecular communication known as Wnt
signaling, an important process in the early development of the
fruit fly and other animals.

The incident may prove to be only a minor setback for Struhl, a
widely respected researcher who is also an investigator of the
prestigious Howard Hughes Medical Institute, because Chan was
responsible for the fabricated experiments.

Struhl notified the University's Committee for the Conduct of
Science as soon as he discovered the abnormalities in the data,
according to Anne Bayne, director of external relations for CUMC.
Chan had already left Columbia, so "the university felt no further
investigation was needed," Bayne wrote in an e-mail yesterday.

In fact, Chan left Columbia in November. He worked at the Albert
Einstein College of Medicine until last month, according to a
spokeswoman, resigning just days before the publication of the
retraction.

The misconduct came to light when Struhl found he could not
replicate the results reported in the 2002 paper. When he
confronted Chan, he learned that many of the experiments "were
either not performed or gave different results than presented in
the paper," he wrote in the retraction.

"Occurrences like this are very rare," Bayne wrote
yesterday.

The publication of the 2002 paper prompted Xi He, an assistant
professor of neurosciences at Harvard Medical School, to embark on
a set of experiments to see if the results Struhl and Chan were
correct. His findings, published this month in Public Library of
Science Biology, demonstrated that Struhl and Chan were wrong, and
that the prevailing theory of Wnt signaling was correct.

In his retraction, Struhl apologized for any "adverse
consequences" that the false data might have caused.

Struhl and Chan could not be reached for comment.

Correction

"CU Medical Center Scientist Retracts Flawed Research" (Feb. 24)
reported that Harvard neuroscientist Xi He conducted experiments to
verify the work of a Columbia team. The experiments were done by
Princeton's Nicholas Tolwinski and Eric Wieschaus.