| dc.description.abstract | Thus, the simulations of 100 ps each of RNase A and RNase S at temperatures of 300 K and 330 K indicate similar dynamic behaviour about their timeaveraged structures. An exhaustive analysis of data from the simulations revealed only minor differences in the dynamics between RNase A and RNase S. These differences were largely confined to loop regions and to the chain termini. RNase S contains two additional chain termini relative to RNase A. Of these two, only the Cterminus of the Speptide helix exhibits significantly greater dynamic flexibility.
Although some other residues also exhibit altered behaviour in the RNase S simulations, it is unclear whether these differences arise from minor differences in the starting structures used for the two simulations. The simulations strongly suggest that apart from these minor differences, both RNase A and RNase S in solution exhibit similar fluctuations about their respective timeaveraged structures. This appears to contradict previous proteolysis (Allende & Richards, 1962) and hydrogenexchange data (Haris et al., 1985; Rosa & Richards, 1981). This apparent discrepancy may simply reflect the fact that simulation times were not long enough to capture the reported experimental differences.
However, it is noteworthy that the simulation results for RNase A showed reasonable correlation with experimental hydrogenexchange measurements, despite the differences in timescales. Secondly, even at an elevated temperature of 330 K, no substantial dynamic differences between the two proteins were observed. Therefore, we chose to investigate an alternative explanation for the apparent disagreement between simulation and experiment.
Because MD simulations showed similar dynamic behaviour for RNase A and RNase S, we reexamined experimental data suggesting that RNase S should exhibit greater dynamic flexibility. Two experimental probes for dynamics are used-hydrogen exchange and trypsin digestion. RNase S can be digested by trypsin at room temperature (Allende & Richards, 1960), whereas RNase A requires extreme conditions, such as 60°C (Ooi et al., 1963). A fundamental distinction between the two proteins is that RNase A is unimolecular, while RNase S is bimolecular. Thus, the fraction of RNase A that is folded is independent of total protein concentration. In contrast, folded RNase S is always in equilibrium with free Speptide and Sprotein; therefore, the fraction of folded RNase S depends on the total protein concentration.
Because the dissociation constant of RNase S is around 10 nM, it has been generally assumed that the small amount of free Sprotein present in most experiments may be neglected. However, more detailed analysis shows that this assumption is incorrect. Using both experimental probes of dynamics, the concentration dependence in RNase S has been demonstrated. For trypsin cleavage, concentration dependence of RNase S dynamics has been shown (Nadig et al., 1996).
A true test of these models would be hydrogenexchange studies of RNase A and RNase S as a function of total protein concentration. Figure 5.9 shows hydrogenexchange rates of RNase A and RNase S as a function of total protein concentration (unpublished results, P.K. Madhu, G.S. Ratnaparkhi, Anil Kumar and Raghavan Varadarajan).
These results clearly demonstrate that RNase A exchanges in a concentrationindependent manner, while exchange rates of RNase S are concentrationdependent, indicating that the exchange occurs via the second pathway. Thus, we conclude that apparent differences in dynamics between RNase A and RNase S are due to the small amount of free Sprotein in solution, and not due to intrinsically greater dynamic fluctuations in RNase S. | |
