TreeGeneBrowser: phylogenetic data mining of gene sequences from public databases.

MOTIVATION: Sequence databases represent an enormous resource of phylogenetic information, but there is a lack of tools for accessing that information in order to assess the amount of evolutionary information in these databases that may be suitable for phylogenetic reconstruction and for identifying areas of the taxonomy that are under-represented for specific gene sequences. RESULTS: We have developed TreeGeneBrowser which allows inspection and evaluation of gene sequence data for phylogenetic reconstruction. This program improves the efficiency of identification of genes that may be useful for particular phylogenetic studies and identifies taxa and taxonomic branches that are under-represented in sequence databases.

Analysis of sequence contexts flanking T.G mismatches leads to predictions about reactivity of the mismatched T to osmium tetroxide.

Osmium tetroxide and hydroxylamine are used to detect mutations in DNA and RNA after hybridization of mutant and wild-type DNA. Mismatched T and C bases, respectively, are modified by these reagents and the DNA strand cleaved at the mismatched bases by subsequent treatment with piperidine. This allows detection and location of the mutation. Although most T.G mismatches have been reported to be reactive to osmium tetroxide, some have been reported to be unreactive. The aim of this study was to collect and analyze the reactive and unreactive T.G mismatches. We have collected sequence contexts of all reactive and unreactive T.G mismatches for analysis. This involves 10 unreactive T.G mismatches (plus one T.C) and 19 reactive T.G mismatches. Sequence effects of bases surrounding these mismatches must influence this reactivity. There must be many types of such sequence effects. We postulate that because of the dominance of 5' G bases near the T of unreactive T.G mismatches and the absence of 5' G bases in reactive T.G mismatches that the stacking of the 5' G on the mismatched T is the reason for this lack of reactivity in the majority of the cases studied here.

Two new mutations in the dihydropteridine reductase gene in patients with tetrahydrobiopterin deficiency.

Two new mutations have been identified within the dihydropteridine reductase (DHPR) gene in two patients with DHPR deficiency. The total coding sequence of the cDNA has been screened by chemical cleavage of mismatch in both patients and selected portions of the cDNA have been sequenced. The first mutation identified causes a glycine to aspartic acid substitution at codon 23 and seems particularly frequent in Mediterranean patients. Its occurrence within a glycine string common to the amino-terminal region in NADH dependent enzymes suggests a possible causal mechanism for the defect. The second change involves a tryptophan to glycine substitution at codon 108 and is carried by both alleles in the second patient. It occurs in a motif which shows similarities with a region of dihydrofolate reductase (DHFR) and is highly conserved within different animal species.

Comparison of genotype and intellectual phenotype in untreated PKU patients.

We have screened 55 untreated phenylketonuria patients from 42 families for common mutations of the phenylalanine hydroxylase gene and determined both causative alleles in 12 families. The correlation between genotype and intellectual phenotype of patients in these families was examined. Our results were compared to a study which predicted phenylalanine hydroxylase activity based on genotype and examined its correlation with the biochemical phenotype of treated patients. Some of the intellectual phenotypes of patients in our study correlated well with the predicted activities. However, we found one family with a genotype expected to have no activity of phenylalanine hydroxylase where the patients were not severely retarded. Major differences in intellectual phenotype were found in patients with the same genotype both between unrelated subjects and within families, suggesting that there is not a simple correlation between genotype and intellectual phenotype.

Characterization of the amdA-regulated aciA gene of Aspergillus nidulans.

The structure, function and regulation of the acetate inducible aciA gene of Aspergillus nidulans was analysed. The aciA locus was mapped to chromosome 1 at a position where no acetate inducible gene has been previously located. The nucleotide sequence of aciA was determined, the structures of two transcripts were determined and the sequences of the polypeptide products of the gene were deduced. Construction of an aciA loss-of-function mutant was achieved via insertional inactivation, but it did not reveal a phenotype for an aciA- strain. The larger polypeptide, AciA, was found to have a putative dinucleotide cofactor binding site. Acetate inducibility of aciA was found to be dependent on the amdA regulatory gene. Use of a 5' deletion series of an aciA--lacZ fusion and an in vivo regulatory protein binding (titration) assay allowed the region required for amdA activity to be localized to a 124 bp segment 5' to aciA. A 13 bp region of sequence similarity was observed between aciA and the coregulated amdS gene in the regions required for amdA-mediated regulation of these genes. This sequence may have a role in amdA regulation of the two genes.

CpG hotspot causes second mutation in codon 408 of the phenylalanine hydroxylase gene.

A new mutation has been identified in exon 12 of the gene encoding phenylalanine hydroxylase at codon 408. The single base change from guanine to adenine changes the amino acid arginine to glutamine; thus, the mutation is defined as R408Q. This codon is the site of a mutation known to causes phenylketonuria. Both these mutations are located at the same CpG site.

Complete mutation detection using unlabeled chemical cleavage.

We have developed a strategy for the complete detection of point mutations, small insertions and deletions by chemical cleavage based on the methodology of Cotton et al. (1988). The technique was extended by the development of a nonisotopic cleavage product detection system using silver staining after gel electrophoresis. The complete mutation detection was achieved by use of mutant and wild-type DNAs in equimolar quantities in duplex formation, thus any mismatches that are resistant to chemical cleavage (e.g., some T.G mismatches) are easily detected by cleavage of the complementary heteroduplex (e.g., A.C mismatch). With such a strategy mutant DNAs can be screened for mutations and polymorphisms. The advantages of complete unlabeled mutation detection are considerable.

A mutation affecting amdS expression in Aspergillus nidulans contains a triplication of a cis-acting regulatory sequence.

In Aspergillus nidulans expression of the acetamidase structural gene, amdS, is under the control of at least four regulatory genes including the trans-acting amdA regulatory gene. A cis-acting mutation (amdI66) consisting of an 18 bp duplication in the 5' region of the amdS gene results in very high levels of acetamidase activity but only in strains carrying semi-dominant mutations in the amdA gene. In selecting for increased amdS expression in an amdI66 and A+ strain, an A. nidulans strain with a mutation in the 5' region of the amdS gene was isolated. The nucleotide sequence was determined of the region containing the mutation, designated amdI666. The mutant strain carries three tandem copies of the 18 bp sequence that is duplicated in the amdI66 mutation. Thus, from a strain carrying a duplication of an apparent regulatory protein binding site with little effect on gene expression, a strain has been derived that carries a triplication of the site with consequent major effects on regulation. The multiple copies of regulatory sites present in many genes may have been generated by a similar mechanism.