Microorganisms can be classified into psychrophiles, mesophiles, and hyperthermophiles based on the differences in their optimum growth temperatures. Between mesophiles and hyperthermophiles, there are thermophiles and extreme thermophiles. The functions of the macrobiomolecules from these bacteria, such as a protein, must show different temperature dependencies. For example, the enzymes from hyperthermophiles must exhibit higher thermostabilities than those of the enzymes from mesophiles. In contrast, the enzymes from psychrophiles must exhibit higher activities at a low temperature than those of the enzymes from mesophiles. Therefore, the thermostable enzymes and cold-adapted enzymes produced from hyperthermophiles and psychrophiles, respectively, are usually industrially valuable.

        It has been reported that thermophilic enzymes show poor activity at the temperatures which are optimum for the activities of the mesophilic or psychrophilic enzymes. This is probably because the enzymes acquire the thermostability at the sacrifice of their flexibility. However, our studies with site-directed mutagenesis have revealed that the activity and stability are not always inversely correlated, and we can modify only the activity or the stability by changing a part of an enzyme. In order to establish a method to improve the activity or stability of an enzyme, it is necessary to identify and extract the factors which contribute to the difference in the activity or the stability of an enzyme, and to elucidate their mechanisms. For this purpose, it is thought to be effective to select a pair of the enzymes which are functionally and structurally similar with each other, but are considerably different from each other in the optimum temperatures for activity and thermostability, and to identify and extract the factors which cause the differences.

We have isolated a hyperthermophilic archaeon, Pyrococcus kodakaraensis KOD1, whose optimum growth temperature is 95 ?C, and a psychrophile, Shewanella sp. SIBI, whose optimum growth temperature is around 20 ?C and grows relatively well even at 4 ?C. We are trying to clone the genes encoding various enzymes from these organisms and characterize the recombinant proteins. They include the genes encoding subtilisin, glycerol kinase, and ribonuclease HII (RNase HII) from P. kodakaraensis KOD1, and subtilisin, alkaline phosphatase, chitinase, RNase HI, and RNase HII from Shewanella sp. SIBI. Comparison of the amino acid sequences of subtilisin and glycerol kinase from P. kodakaraensis KOD1 with those from mesophiles suggests that insertion of peptide chains and reinforcement of ion-pair networks are responsible for the high stability of thermophilic subtilisin and thermophilic glycerol kinase, respectively. We are performing site-directed mutagenesis experiments to prove it.

        A hyperthermophile P. kodakaraensis KOD1 is a member of archaea. According to phylogenetic analysis, archaea belong to a different group from bacteria and eukaryotes. However, it is rather closely related to eukaryotes than to bacteria. Thus, a common ancestor of archaea and eukaryotes may have diverged from bacteria, and archaea and eukaryotes may have diverged after that. As the optimum conditions for the growth of archaea are thought to resemble those in the primitive earth, archaea may retain traces of early life forms more strongly than bacteria and eukaryotes. Because archaea, especially hyperthermophiles, are expected to share the properties with both early life forms and eukaryotes, these organisms are valuable sources to trace the evolution of life as well.

        Thus, to get a clue for the understanding of evolution of life is one of the purposes in this subject. We have so far identified some unusual properties that seem to be characteristic to proteins from hyperthermophiles. For example, glycerol kinase from P. kodakaraensis KOD1 seems to work as a dimer, instead of a tetramer. The substrate specificities and metal ion requirements of the enzymes from hyperthermophiles seem to be not so strict. Especially, the RecA/Rad51 homologue from P. kodakaraensis KOD1 is structurally and functionally unique. It is composed of a polypeptide chain which is only the equivalent to the core domain of RecA from E. coli and Rad51 from yeast, and exhibits endo/exo-nuclease activity in addition to ATPase activity. We are now analyzing this protein structurally and functionally in more detail. In addition, we found that TBP-interacting protein ( TIP ) inhibits the binding of TBP to DNA by interacting with TBP. We continue working on this protein to examine whether or not TIP acts as a negative transcriptional regulator.