| Nat Biotechnol 1996 Oct;14(10):1219-20 |
Institut für Physikalische und Theoretische Chemie, Technische Universität München,
D-85748 Garching, Germany.
KAIROS Scientific Inc., 3350 Scott Blvd., Bldg. 62, Santa Clara, CA 95054, USA
Although fluoroescent proteins are widespread in nature, GFP has turned out to be unique. Rather than binding a fluorophore formed through a complex biosynthetic pathway (e.g. tetrapyrrole as in phycobiliproteins), GFP's fluorophore forms autocatalytically on the nascent apoproteins's backbone. A structure for the fluorophore was purposed in a landmark 1993 publication. The new findings substantiate all of this earlier work.
A fairly simple model for the mechanism of GFP fluorescence, supported by recent mutagenesis data, has been suggested on the basis of the proposed fluorophore excitation spectrum as a function of pH. In this model, at high pH or after certain mutations are introduced near the GFP fluorophore, Tyr66 is postulated to be in the phenolate form, whereas at low pH, it is in the hydroxyl form. Given a small Stokes shift and a large zero point crossing (i.e., significant spectral overlap), one would envisage that the 477-nm excitation band and 510-nm emission band of GFP correspond to transitions involving the first excited singlet state of the phenolate form of Tyr66. Similarly, one would model the hydroxyl form of the fluorophore to be responsible for 395-nm excitation band, albeit with a remarkably large 105-nm Stokes shift. In the point mutant Ser65 -> Thr or the combinatorial mutant RSGFP4, the phenolate form of Tyr66 with the smaller Stokes shift would be favored. In this case, one might be tempted to assign the 395-nm band to a higher-lying S2 state. As more recent data indicate, this last interpretion is not correct.
| HOME |