The authors of a Nature paper outlining a mechanism for multisensory memories in Drosophila melongaster have retracted the work after they were unable to replicate a set of imaging experiments.
The paper’s overall conclusions remain unaffected, says study investigator Scott Waddell, professor of neurobiology at the University of Oxford. The team plans to resubmit the work without the imaging experiments to a new journal. “This is the right thing to do, despite how painful it is,” he says.
The paper, which was published in 2023 and has been cited 63 times, according to Nature, investigated how neuronal activity changes to support a multisensory memory. Fruit flies learned to associate an odor, a color or both with a sugar reward. The memories formed through multisensory training were stronger than the memories that relied on a single sense, even when memory recall was tested using only one of the cues.
When Waddell and his team blocked neurons in the mushroom body—a sensory integration brain region—that respond to visual cues, the flies couldn’t recall the multisensory memory when presented with just the odor; the blockage had no effect on flies that were trained on odor cues alone.
Based on this finding as well as other behavioral and connectome-based experiments, Waddell and his team concluded that multisensory learning reshapes neural coding so neurons that originally responded to visual cues become necessary for remembering olfactory ones.
Subsequent voltage imaging experiments supported this takeaway: After multisensory training, an odor cue did not inhibit the visually responsive neurons as expected but instead induced excitation.
The paper was “exciting” when it came out, and many labs discussed it in journal clubs, says Annika Barber, assistant professor of molecular biology and biochemistry at Rutgers University, who was not involved in the work. The field knew that flies were capable of multisensory learning, “but this was the first paper that really combined colors with odors, showed that that could improve memory, and then sought to identify the cellular basis through which this is actually occurring.”
After the paper was published, Gaby Maimon, professor and head of the Laboratory of Integrative Brain Function at Rockefeller University, reached out to Waddell and asked to view his data because he was curious about the voltage imaging results, Waddell says. (Maimon declined a request to be interviewed for this story.)
Maimon’s team found an imaging processing error in the voltage data that meant the recorded signals were actually an artifact. “They were kind enough to point out the error rather than going public with it. I can’t praise them enough for that,” Waddell says. Waddell and his team reexamined the imaging data and identified the same error.
The team replicated the paper’s behavioral and connectomics experiments but could not replicate the imaging results after fixing the processing error. “I didn’t want to be in a position where we are essentially chasing a result,” Waddell says, and he also did not want to leave data in the literature that are “just wrong,” so he chose to retract the paper. “I want our published work to be as truthful as possible.”
Nature did not have any comment beyond what is stated in the retraction note.
T
Waddell stands behind the broader conclusion that multisensory learning reroutes neural circuitry so visually responsive neurons become required for olfactory behavior. “The conclusions were drawn before we had any imaging data anyway,” Waddell says. His group is still studying the potential mechanism for this process using different methods.
Barber and Cervantes-Sandoval agree that the behavioral data are robust and say they are interested to see what the neuronal mechanism turns out to be.
The amount of work required to redo the experiments and determine that the voltage imaging data don’t hold up is “not trivial,” Barber says. “I think the willingness to put in that level of effort for a retraction speaks highly of Scott’s group.”
Barber says she is also pleased to see a research group choose to retract a paper from such a high-profile journal. In some cases, the reluctance to retract papers of this status “has resulted in some very sketchy but splashy findings, that maybe aren’t as well validated as they could be, standing in the literature.”
