Herein we examine the origin of the high-temperature(350 K) behavior of a thermophilic mutant enzyme (labeled as 5-3H5; see Zhao and Arnold Prot. Eng. 1999, 12, 47-53) derived from subtilisin E by eight amino acid substitutions. Through the use of molecular dynamics (MD) simulations, we have provided molecular-level insights into how point mutations can affect protein structure and dynamics. From our simulations we observed a reduced rmsd in several key regions, an increased overall flexibility, an increase in the number of hydrogen bonds, and an increase in the number of stabilizing interactions in the thermophilic system. We also show that it is not a necessary requirement that thermophilic enzymes be less flexible than their mesophilic counterparts at low temperatures. However, thermophilic enzymes must retain their three-dimensional structures and flexibility at high temperatures in order to retain activity. Furthermore, we have been able to point out the effects of some of the single substitutions. Even if ii is not possible yet to give general rules for rational protein design, we are able to make some predictions on how a protein should be stabilized in order to be thermophilic. In particular, we suggest that a promising strategy toward speeding up the design of thermally stable proteins would be to identify fluxional regions within a. protein through the use of MD simulations (or suitable experiments). Presumably these regions allow for autocatalytic reactions to occur and are also involved in allowing water to gain access to the interior of the protein and initiate protein unfolding. These fluxional regions could also adversely affect the positioning of the catalytic machinery, thereby decreasing catalytic efficiency. Thus, once these locations have been identified, "focused" directed evolution studies could be designed that stabilize these "fluxional" regions.