Conserved amino acid sequence motifs in different groups of NTP-utilizing enzymes were studied using computer methods of sequence analysis to the end of predicting NTPase activity of unexplored proteins, designing schemes for identification of NTPases in sequence databases and generating a sequence-based classification of this type of enzymes. NTPases are characterized by well-defined conserved motifs that are implicated in substrate binding and hydrolysis. Amino acid sequence databases were searched for the so-called A motif that is involved in phosphate binding and the resulting set of proteins was explored in detail with respect to their similarities with other proteins, and the available data on NTPase activity. Three distinct protein families that bring together NTPases with classical and deviant forms of the A motif were characterized. A modified NTP-binding motif was detected in E.coli DnaC protein and it was shown that this protein is distantly related to DnaA leading to implications for the evolution of the DNA replication machinery. A putative ATPase that is also related to DnaA was identified in an archaebacterial virus. Helicase activity was predicted for several protein whose functions were unknown. Unexpectedly, helicase motifs were found in E.coli protein SecA leading to a hypothesis that this protein may have a second activity, RNA duplex unwinding, that might be important for autoregulation of translation. Murine protein MOV10 that was previously proposed to be a GTPase was shown to be related to yeast helicases. MOV10 appears to be the first helicase from a higher eukaryote that belongs to the so-called superfamily I. A protein (a purported transcription factor) was revealed upon database search that appeared to encompass only two of the seven conserved motifs typical of helicases and may be considered a helicase domain isolated in a separate protein. Work on delineation of sequence patterns that will serve as unique identifiers of different families of helicases is in progress. Another ongoing development is an attempt to generalize the A motif to include other types of NTP-binding sites and also binding sites for other small molecules, e.g. for S-adenosylmethionine in methyltransferases. The significance of the project is in the prediction of NTPase activity for many proteins with unknown functions, characterization of allowed deviations in NTP-binding motifs, derivation of identifying patterns for different groups of NTPases, and development of a sequence-based classification for a vast enzyme class. GRAMT=Z01LM00036 Computer-assisted comparisons of amino acid sequences of proteins of large DNA viruses with each other and with related cellular proteins were performed, with the emphasis on proteins that may specifically influence virus virulence. Vast collections of gene sequences of large DNA viruses are currently available, including one complete poxvirus genome and several herpesvirus genomes. The functions of a large fraction of proteins encoded by these viruses are not understood. Of particular interest are genes that are not essential for virus growth in tissue culture but influence virus virulence. A family of poxvirus proteins was described that contain two distinct domains similar to apparently unrelated cellular proteins. The N-terminal domain is related to similarly located domains of several transcription factors, and the C-terminal domain resembles a murine placental-specific protein MIPP. The latter domain consists of an array of four or five imperfect repeats. Our analysis revealed also truncated members of this protein family in poxviruses, suggesting a complex evolutionary pathway that might have involved duplications and deletions. Functions of several viral NTP-utilizing enzymes were predicted, including two poxvirus helicases and a putative NTPase that may be involved in virus maturation. A protein was identified in fowlpox virus that is related to mammalian cytidylate kinase and is proposed to possess the same enzymatic activity. A tentative phylogeny of viral and cellular thymidine and thymidylate kinases was generated and a hypothesis was developed that genes coding for enzymes of nucleotide metabolism could be acquired independently by different viruses via recombination with the respective cellular genes. A specific version of Zn finger motif was identified in a protein encoded by a gene of mousepox virus that has been sequenced in the Laboratory of Viral Diseases and has been shown to be essential for the virus to be virulent for its natural host. The comparative analysis of the Zn finger proteins of the respective family is in progress.
