Models don't always have to be correct to be helpful. In 2004, Paul Talalay and his team at Johns Hopkins University published evidence that the sensory protein Keap1 dissociates from the transcription factor Nrf2, allowing Nrf2 to induce detoxifying enzymes that protect the cell from carcinogens and oxidants.
The team demonstrated that external chemical cues, such as sulforaphane found in broccoli, directly modified two highly reactive Keap1 cysteines. When they systematically mutated each cysteine in mice, they found evidence that these inducers cause Keap1 to form dimers via intermolecular disulfide bridges.1
The resulting model - that dissociation of Keap1 from Nrf2 frees it to activate downstream genes - appears incorrect, however. Findings about other important cysteines in Keap1 and evidence that electrophilic agents alone do not actually dissociate Keap1 from Nrf2 has led to a new model in which Keap1 regulates the ubiquitin-mediated degradation of Nrf2, says Andrew Mesecar at the University of Illinois, Chicago. "We're now trying to deconvolute ubiquitination, de-ubiquitination, ... and factors in the cell that are responsible for doing that."
Nevertheless, says Mark Hannink at the University of Missouri-Columbia, Talalay's work was critical in showing that Keap1 uses cysteine residues as a sensing mechanism in the first place.