Helix forming peptides as spacers and receptors for fluorescent proteins

Albert Griffin

Appointment Period: 1994-1997, Grant Years: [10,11,12]

Our goal is to develop a new means of in vivo protein labeling, based on the high affinity of arsenoxides for pairs of cysteines. A cysteine-rich peptide will be genetically fused to the protein of interest and an arsenoxide containing fluorophore will be targeted to that peptide. During the past year several advances have been made toward this goal.

Our first in vivo experiments indicate that toxicity may be circumvented by protecting arsenoxides with propanedithiol(PDT). This result implies that endogenous binding sites for arsenoxides cannot compete effectively with the six-membered ring formed by PDT chelation. For our in vivo protein labeling to work, we must have an arsenoxide-peptide couple of higher affinity than the arsenoxide-PDT pair. To do this we are employing organic compounds with two arsenoxides and a short peptide with two pairs of cysteines. This year five additional bisarsenoxide compounds have been prepared, bringing the total to ten. A model peptide with four cysteines is able to displace in vitro the PDT protecting groups from three of these. Based on the structure of the tightest binder, a fluorescent bisarsenoxide has been designed and is being prepared. With this in hand, more extensive in vivo tests will be performed. We continue to explore other spacings of the two arsenoxides which may afford even higher affinity. We also plan to screen the best bisarsenoxides with phage display peptide libraries to optimize the peptide side of the pair.

In addition, we continue to explore thermodynamics and kinetics of arsenic-dithiol interactions. Accurate binding constants of arsenoxides with high-affinity dithiols such as PDT and 2,3-dimercaptopropanol (BAL) have not been reported in the literature, presumably due to the difficulty in measuring them. These affinities can be determined, however, by measuring perturbations of the pH profile of cadmium-dithiol adducts when arsenoxides are allowed to compete with cadmium for the dithiol. Using this method we determined the Kd of p-sulfonatephenylarsenoxide chelated with 2,3-dimercaptopropanesulfonate (a water soluble BAL analog) to be 100 pM. Successful completion of this project will provide a general method for protein labeling, greatly facilitating the study of protein-protein interactions in living cells.

PUBLICATIONS (resulting from this training)

Griffin BA, Adams SR, Tsien RY. (1998) Specific covalent labeling of recombinant protein molecules inside live cells. Science. 281:269-72.

Griffin BA, Adams SR, Jones J, Tsien RY. (2000) Fluorescent labeling of recombinant proteins in living cells with FlAsH. Methods Enzymol. 327:565-78.