An associate professor at the Massachusetts Institute of Technology retracted a 2010 paper that was published in Cell because her lone co-author “falsified or fabricated figures,” according to the retraction notice (1-2).
The method described in the paper, called biotin labeling of intercellular contacts, or BLINC for short, got some attention since it seemed more sensitive than previous methods and allowed visualization of protein interactions in living neurons. The method involved expressing biotin ligase linked to the protein neurexin in the presynaptic cell and a so-called acceptor peptide linked to the protein neuroligin in the opposing neuron. When biotin and ATP were added, the biotin ligase connected biotin to the acceptor peptide on neuroligin, allowing detection of the interaction using a fluorescent probe attracted to biotin.
At the time of publication, nothing seemed fishy: “As I knew of Alice's great technology for protein labeling, I thought that the BLINC was an extension of it,” said Atsushi Miyawaki, laboratory head at the Riken Brain Science Institute in Wako, Japan, in an email. Miyawaki wrote a News and Views article on the paper in 2010 for Nature Chemical Biology (3).
The problem came when Ting’s own laboratory could not reproduce the published results, according to the retraction notice. After she brought her concerns to MIT, the institute conducted an independent investigation that found Thyagarajan had engaged in research misconduct by falsifying or fabricating figures in the report. As a result, Ting retracted the paper without Thyagarajan’s signature.
When asked about the retraction, Ting replied in an email, “Please take note of our new paper in PLoS One, which addresses the technical aspects of the BLINC method.”
The paper that she is referring to was published in Public Library of Science One on February 14, the same day that the Cell retraction was published. In the new paper, Ting and colleagues, minus Thyagarajan, reported their methodical troubleshooting of the BLINC protocol. They found that the constructs for the acceptor protein and biotin ligase were faulty. So the team replaced the promoters and optimized the design, and expressed the genes for a longer period of time.
But even after these modifications, BLINC still did not work in neurons. The images suggested that the biotin ligase-neurexin fusion protein remained stuck in the cell body instead of moving to the synapses. So the group attached the biotin ligase to a different isoform of neurexin, which allowed the proteins to traffic to the distal processes of the neurons.
In the end, the BLINC labeling technique does work, but is not robust in neurons. So, the team developed a new protocol called interaction-dependent probe incorporation mediated by enzymes, or ID-PRIME for short, that “gives stronger and more localized labeling, and is more versatile than BLINC in terms of signal readout,” according to the authors.
Thyagarajan is no longer with the Ting lab at MIT, and although he was listed as a technology specialist at the intellectual property law firm of Clark+Elbing LLP in Boston, his name has been removed from the firm’s website.
1. Thyagarajan, A., and A. Y. Ting. 2013. Retraction notice to: Imaging Activity-Dependent regulation of Neurexin-Neuroligin interactions using trans-Synaptic enzymatic biotinylation. Cell 152(4).
2. Thyagarajan, A., and A. Y. Ting. 2010. Imaging activity-dependent regulation of neurexin-neuroligin interactions using trans-synaptic enzymatic biotinylation. Cell 143(3):456-469.
3. Liu, D. S., K. H. Loh, S. S. Lam, K. A. White, and A. Y. Ting. 2013. Imaging Trans-Cellular Neurexin-Neuroligin interactions by enzymatic probe ligation. PLoS ONE 8(2):e52823+.