Improved method for converting an unmodified peptide to an energy-transfer substrate for a proteinase

This paper describes refinements to a method for generating energy-transfer-based fluorogenic substrates for proteinases [Geoghegan, K. F., et al. (1993) Bioconjugate Chem. 4, 537-544]. An unprotected peptide is taken as starting material and coupled by three steps of chemistry to two dyes that form an energy donor-energy acceptor pair. The donor group is Lucifer Yellow CH (lambda(ex) 420-430 nm, lambda(em) 530 nm), and the acceptor is 5-carboxytetramethylrhodamine, a strong chromophore that effectively quenches Lucifer Yellow fluorescence in intact substrates. Periodate oxidation of N-terminal Ser gives the peptide an alpha-N-glyoxylyl moiety, OHC-CO-peptide, which is allowed to react with Lucifer Yellow CH (a carbohydrazide derivative) to attach the dye through an adequately stable hydrazone bond. Derivatization is completed by allowing the succinimidyl ester of 5-carboxytetramethylrhodamine to react with the epsilon-amino group of a single Lys placed near the C terminus. In the revised method, the first two steps of chemistry are completed sequentially in a single vessel to eliminate the labor and losses associated with isolating the alpha-N-glyoxylyl intermediate. In addition, peptides taken as starting material for this method are now designed according to the scheme (Ser)-(cleavable sequence)-(lys-Arg). The addition of Arg at the C terminus promotes aqueous solubility of the final substrate without complicating the chemistry; multiple Arg might also be used, though this has not been tried here. It has also been found that relative reversed-phase HPLC retention of Lucifer Yellow/5-carboxytetramethylrhodamine substrates is predictable on the basis of the hydrophobicity of the original peptide. Substrates prepared from peptides with Bull and Breese indices of > 200 cal/mol are readily separated from residual dye in the second and final chromatographic step of the synthesis, simplifying the only moderately difficult step in the preparation. As with all fluorogenic substrates, assays using these substrates are subject to internal filtering effects that can lead to serious error. This problem becomes significant at a range of substrate concentration that is predictable when the size and geometry of the fluorescence cell are taken into account. The reasoning applied in determining this range is applicable to substrates constructed with any donor-acceptor pair.