A common structural feature in promoter sequences of E. coli

Structural prediction of A- and B-DNA duplexes based on coordinates of the phosphorus atoms

The sequence-dependent structure of DNA double helices was studied extensively during the past 10 years. How the backbone structure correlates with the base structure in a duplex conformation is still an important yet open question. Using a set of reduced coordinates and a least-squares fitting procedure, we have developed a method to predict structures for B-DNA duplexes based on coordinates of the phosphorus atoms. This method can be used to predict all-atom structures for both bent and straight molecules. We estimated the accuracies of the predictions by studying a set of 10 oligonucleotides with their structures available from the Protein Data Bank. We used this method to construct a modeled structure for the bacteriophage lambda cro operator for which the phosphorus coordinates were known from 3.5-angstrum resolution crystal data (4CRO).

Prediction of function in DNA sequence analysis

Recognition of function of newly sequenced DNA fragments is an important area of computational molecular biology. Here we present an extensive review of methods for prediction of functional sites, tRNA, and protein-coding genes and discuss possible further directions of research in this area.

Cloning and characterization of an alpha-amylase gene from Streptomyces lividans

The alpha-amylase gene (amy) of Streptomyces lividans TK24 was cloned in an amylase deficient mutant strain S. lividans M2. The cloned gene contained an open reading frame (ORF) of 2757 nucleotides (919 amino acids) coding for a protein of 100 kDa. Sequencing of the amino terminus of the extracellular alpha-amylase protein revealed the presence of a signal peptide of 33 amino acid residues. The transcriptional initiation site was mapped by the primer extension method with T4 DNA polymerase and was found to be transcribed from an unique promoter. The alpha-amylase protein produced by S. lividans was larger than those derived from other origins. It also contained the four common conserved regions characteristic of other alpha-amylase proteins.

Analysis of patterns of twist angles in DNA double helix

We have analysed theoretically the patterns of twist angles of B-DNA by the Tung-Harvey model mainly. It is shown that for a sequence of twist angles a smaller twist angle tends to follow a larger one and vice versa. Therefore the sequence of twist angles always takes a gentle zig-zag form. For simplicity we convert the sequence of twist angles to a symbolic sequence consisting of L and S, where L or S represents a large or a small angle, respectively. The -10 and -35 regions of 112 well-defined promoters for E. coli RNA polymerase, which were compiled by Hawley and McClure, have been analysed in terms of LS sequences in detail. The results shows that the number of LS sequences for promoters is considerably limited and the promoter mutations do not change the patterns of LS sequences in most cases. Several new ideas, which are believed to be useful in the further study, have been presented.

The YYRR box: a conserved dipyrimidine-dipurine sequence element in Drosophila and other eukaryotes

We have discovered a novel DNA sequence element in Drosophila which is based upon a CTGA tandem repeat. This element has been named the YYRR box to emphasize its dipyrimidine-dipurine nature which is predicted to have unusual structural features. Southern hybridization analysis of genomic DNA indicates the presence of 25-30 copies of the YYRR box in each of three Drosophila species (melanogaster, pseudoobscura, and virilis) and conservation of genomic location within species. Similar analysis of human and rat DNA indicates the presence of YYRR related sequences in mammals as well. YYRR boxes have been localized to two genetic loci in Drosophila: Gld and a gene tentative identified as ted. These two genes exhibit correlated patterns of developmental expression and an identical mutant phenotype. Sequence analysis of the Gld YYRR box in three Drosophila species revealed a high degree of conservation despite its intronic location.

Computer modelling of DNA structures involved in chromosome maintenance

Sequence-dependent DNA bending of synthetic and natural molecules was studied by computer analysis. Modelling of synthetic oligonucleotides and of 107 kb of natural sequences gave results which closely resembled published electrophoretic data, demonstrating the powerful predictive capacity of the procedure. The analysis was extended to the study of DNA structures involved in chromosome maintenance. Centromeric DNAs from yeast were found to have sequences in their functional elements which cause them to be unusually straight. Autonomous replicating sequences were found to have two structural domains, one consisting of unusually straight sequences surrounding the consensus and the other of bending elements in flanking DNA. In addition to a structural homology, centromeric and autonomous replicating sequences share common sequence elements. These observations show that computer modelling of natural sequences is a viable approach to the study of the biological implications of alternative DNA structures.