LASER TEMPERATURE-JUMP STUDY OF THE HELIX-REVERSIBLE-ARROW-COIL KINETICS OF AN ALANINE PEPTIDE INTERPRETED WITH A KINETIC ZIPPER MODEL

The kinetics of the helix reversible arrow coil transition of an alani ne-based peptide following a laser-induced temperature jump were monit ored by the fluorescence of an N-terminal probe, 4-(methylamino)benzoi c acid (MABA). This probe forms a peptide hydrogen bond to the helix b ackbone, which changes its fluorescence quantum yield. The MABA fluore scence intensity decreases in a single exponential relaxation, with re laxation times that are weakly temperature dependent, exhibiting a max imum value of similar to 20 ns near the midpoint of the melting transi tion. We have developed a new model, the kinetic version of the equili brium 'zipper' model for helix reversible arrow coil transitions to ex plain these results. In this 'kinetic zipper' model, an enormous reduc tion in the number of possible species results from the assumption tha t each molecule contains either no helical residues or a single contig uous region of helix (the single-sequence approximation). The decay of the fraction of N-terminal residues that are helical, calculated from numerical solutions of the kinetic equations which describe the model , can be approximately described by two exponential relaxations having comparable amplitudes. The shorter relaxation time results from rapid unzipping (and zipping) of the helix ends in response to the temperat ure jump, while the longer relaxation time results from equilibration of helix-containing and non-helix-containing structures by passage ove r the nucleation free energy barrier. The decay of the average helix c ontent is dominated by the slower process. The model therefore explain s the experimental observation that relaxation for the N-terminal fluo rescent probe is similar to 8-fold faster than that for the infrared p robe of Williams et al. [(1996) Biochemistry 35, 691-697], which measu res the average helix content, but does not account far the absence of observable amplitude for the slow relaxation in the fluorescence expe riments (<10% slow phase). If we assume that the activation barrier fo r the coil-->helix rate is purely entropic, the model can also explain the maximum in the temperature dependence of the relaxation time for the fluorescent probe. Parameters that best reproduce the melting curv es and the ratio of relaxation times predict a value of the cooperativ ity parameter sigma which is similar to 3-fold larger than previously reported values obtained from fitting equilibrium data only. The helix growth rate of similar to 10(8) s(-1) that reproduces the experimenta l relaxation times is similar to 100-fold slower than those observed i n molecular dynamics simulations. These parameters can be used to simu late the kinetically cooperative formation of a helix from the all-coi l state.