A library-based bioinformatics services program*

Support for molecular biology researchers has been limited to traditional library resources and services in most academic health sciences libraries. The University of Washington Health Sciences Libraries have been providing specialized services to this user community since 1995. The library recruited a Ph.D. biologist to assess the molecular biological information needs of researchers and design strategies to enhance library resources and services. A survey of laboratory research groups identified areas of greatest need and led to the development of a three-pronged program: consultation, education, and resource development. Outcomes of this program include bioinformatics consultation services, library-based and graduate level courses, networking of sequence analysis tools, and a biological research Web site. Bioinformatics clients are drawn from diverse departments and include clinical researchers in need of tools that are not readily available outside of basic sciences laboratories. Evaluation and usage statistics indicate that researchers, regardless of departmental affiliation or position, require support to access molecular biology and genetics resources. Centralizing such services in the library is a natural synergy of interests and enhances the provision of traditional library resources. Successful implementation of a library-based bioinformatics program requires both subject-specific and library and information technology expertise.

An antisense approach to phenotype-based gene cloning in Tetrahymena

We report a pioneering approach using Tetrahymena thermophila that permits rapid identification of genes based on their null or hypomorphic phenotypes. This technique involves cell transformation with a library of plasmids that encode 26S ribosomal subunits containing short insertions. The insertions correspond to antisense sequences for a large number of genes. The majority of cells each acquires a single antisense sequence, which silences a single genomic locus. Because the insertion site within the ribosomal sequence is known, the silenced gene is easily amplified. We demonstrate that this approach can be used to identify genes required for dense core granule exocytosis.

Isolation of a superfamily of candidate disease-resistance genes in soybean based on a conserved nucleotide-binding site.

The tobacco N and Arabidopsis RPS2 genes, among several recently cloned disease-resistance genes, share highly conserved structure, a nucleotide-binding site (NBS). Using degenerate oligonucleotide primers for the NBS region of N and RPS2, we have amplified and cloned the NBS sequences from soybean. Each of these PCR-derived NBS clones detected low-or moderate-copy soybean DNA sequences and belongs to 1 of 11 different classes. Sequence analysis showed that all PCR clones encode three motifs (P-loop, kinase-2, and kinase-3a) of NBS nearly identical to those in N and RPS2. The intervening region between P-loop and kinase-3a of the 11 classes has high (26% average) amino acid sequence similarity to the N gene although not as high (19% average) to RPS2. These 11 classes represent a superfamily of NBS-containing soybean genes that are homologous to N and RPS2. Each class or subfamily was assessed for its positional association with known soybean disease-resistance genes through near-isogenic line assays, followed by linkage analysis in F2 populations using restriction fragment length polymorphisms. Five of the 11 subfamilies have thus far been mapped to the vicinity of known soybean genes for resistance to potyviruses (Rsv1 and Rpv), Phytophthora root rot (Rps1, Rps2, and Rps3), and powdery mildew (rmd). The conserved N- or RPS2-homologous NBS sequences and their positional associations with mapped soybean-resistance genes suggest that a number of the soybean disease-resistance genes may belong to this superfamily. The candidate subfamilies of NBS-containing genes identified by genetic mapping should greatly facilitate the molecular cloning of disease-resistance genes.

Identification of a functionally important sequence in the cytoplasmic tail of integrin beta 3 by using cell-permeable peptide analogs.

Integrins are major two-way signaling receptors responsible for the attachment of cells to the extracellular matrix and for cell-cell interactions that underlie immune responses, tumor metastasis, and progression of atherosclerosis and thrombosis. We report the structure-function analysis of the cytoplasmic tail of integrin beta 3 (glycoprotein IIla) based on the cellular import of synthetic peptide analogs of this region. Among the four overlapping cell-permeable peptides, only the peptide carrying residues 747-762 of the carboxyl-terminal segment of integrin beta 3 inhibited adhesion of human erythroleukemia (HEL) cells and of human endothelial cells (ECV) 304 to immobilized fibrinogen mediated by integrin beta 3 heterodimers, alpha IIb beta 3, and alpha v beta 3, respectively. Inhibition of adhesion was integrin-specific because the cell-permeable beta 3 peptide (residues 747-762) did not inhibit adhesion of human fibroblasts mediated by integrin beta 1 heterodimers. Conversely, a cell-permeable peptide representing homologous portion of the integrin beta 1 cytoplasmic tail (residues 788-803) inhibited adhesion of human fibroblasts, whereas it was without effect on adhesion of HEL or ECV 304 cells. The cell-permeable integrin beta 3 peptide (residues 747-762) carrying a known loss-of-function mutation (Ser752Pro) responsible for the genetic disorder Glanzmann thrombasthenia Paris I did not inhibit cell adhesion of HEL or ECV 304 cells, whereas the beta 3 peptide carrying a Ser752Ala mutation was inhibitory. Although Ser752 is not essential, Tyr747 and Tyr759 form a functionally active tandem because conservative mutations Tyr747Phe or Tyr759Phe resulted in a nonfunctional cell permeable integrin beta 3 peptide. We propose that the carboxyl-terminal segment of the integrin beta 3 cytoplasmic tail spanning residues 747-762 constitutes a major intracellular cell adhesion regulatory domain (CARD) that modulates the interaction of integrin beta 3-expressing cells with immobilized fibrinogen. Import of cell-permeable peptides carrying this domain results in inhibition "from within" of the adhesive function of these integrins.

Parallel human genome analysis: microarray-based expression monitoring of 1000 genes.

Microarrays containing 1046 human cDNAs of unknown sequence were printed on glass with high-speed robotics. These 1.0-cm2 DNA "chips" were used to quantitatively monitor differential expression of the cognate human genes using a highly sensitive two-color hybridization assay. Array elements that displayed differential expression patterns under given experimental conditions were characterized by sequencing. The identification of known and novel heat shock and phorbol ester-regulated genes in human T cells demonstrates the sensitivity of the assay. Parallel gene analysis with microarrays provides a rapid and efficient method for large-scale human gene discovery.

Mouse mast cell gp49B1 contains two immunoreceptor tyrosine-based inhibition motifs and suppresses mast cell activation when coligated with the high-affinity Fc receptor for IgE.

Mouse mast cells express gp49B1, a cell-surface member of the Ig superfamily encoded by the gp49B gene. We now report that by ALIGN comparison of the amino acid sequence of gp49B1 with numerous receptors of the Ig superfamily, a newly recognized family has been established that includes gp49B1, the human myeloid cell Fc receptor for IgA, the bovine myeloid cell Fc receptor for IgG2, and the human killer cell inhibitory receptors expressed on natural killer cells and T lymphocyte subsets. Furthermore, the cytoplasmic domain of gp49B1 contains two immunoreceptor tyrosine-based inhibition motifs that are also present in killer cell inhibitory receptors; these motifs downregulate natural killer cell and T-cell activation signals that lead to cytotoxic activity. As assessed by flow cytometry with transfectants that express either gp49B1 or gp49A, which are 89% identical in the amino acid sequences of their extracellular domains, mAb B23.1 was shown to recognize only gp49B1. Coligation of mAb B23.1 bound to gp49B1 and IgE fixed to the high-affinity Fc receptor for IgE on the surface of mouse bone marrow-derived mast cells inhibited exocytosis in a dose-related manner, as defined by the release of the secretory granule constituent beta-hexosaminidase, as well as the generation of the membrane-derived lipid mediator, leukotriene C4. Thus, gp49B1 is an immunoreceptor tyrosine-based inhibition motif-containing integral cell-surface protein that downregulates the high-affinity Fc receptor for IgE-mediated release of proinflammatory mediators from mast cells. Our findings establish a novel counterregulatory transmembrane pathway by which mast cell activation can be inhibited.

Gene recognition via spliced sequence alignment.

Gene recognition is one of the most important problems in computational molecular biology. Previous attempts to solve this problem were based on statistics, and applications of combinatorial methods for gene recognition were almost unexplored. Recent advances in large-scale cDNA sequencing open a way toward a new approach to gene recognition that uses previously sequenced genes as a clue for recognition of newly sequenced genes. This paper describes a spliced alignment algorithm and software tool that explores all possible exon assemblies in polynomial time and finds the multiexon structure with the best fit to a related protein. Unlike other existing methods, the algorithm successfully recognizes genes even in the case of short exons or exons with unusual codon usage; we also report correct assemblies for genes with more than 10 exons. On a test sample of human genes with known mammalian relatives, the average correlation between the predicted and actual proteins was 99%. The algorithm correctly reconstructed 87% of genes and the rare discrepancies between the predicted and real exon-intron structures were caused either by short (less than 5 amino acids) initial/terminal exons or by alternative splicing. Moreover, the algorithm predicts human genes reasonably well when the homologous protein is nonvertebrate or even prokaryotic. The surprisingly good performance of the method was confirmed by extensive simulations: in particular, with target proteins at 160 accepted point mutations (PAM) (25% similarity), the correlation between the predicted and actual genes was still as high as 95%.

A computer-based systematic survey reveals the predominance of small inverted-repeat elements in wild-type rice genes.

Several recent reports indicate that mobile elements are frequently found in and flanking many wild-type plant genes. To determine the extent of this association, we performed computer-based systematic searches to identify mobile elements in the genes of two "model" plants, Oryza sativa (domesticated rice) and Arabidopsis thaliana. Whereas 32 common sequences belonging to nine putative mobile element families were found in the noncoding regions of rice genes, none were found in Arabidopsis genes. Five of the nine families (Gaijin, Castaway, Ditto, Wanderer, and Explorer) are first described in this report, while the other four were described previously (Tourist, Stowaway, p-SINE1, and Amy/LTP). Sequence similarity, structural similarity, and documentation of past mobility strongly suggests that many of the rice common sequences are bona fide mobile elements. Members of four of the new rice mobile element families are similar in some respects to members of the previously identified inverted-repeat element families, Tourist and Stowaway. Together these elements are the most prevalent type of transposons found in the rice genes surveyed and form a unique collection of inverted-repeat transposons we refer to as miniature inverted-repeat transposable elements or MITEs. The sequence and structure of MITEs are clearly distinct from short or long interspersed nuclear elements (SINEs or LINEs), the most common transposable elements associated with mammalian nuclear genes. Mobile elements, therefore, are associated with both animal and plant genes, but the identity of these elements is strikingly different.

Evolutionary relationships of the coelacanth, lungfishes, and tetrapods based on the 28S ribosomal RNA gene.

The origin of land vertebrates was one of the major transitions in the history of vertebrates. Yet, despite many studies that are based on either morphology or molecules, the phylogenetic relationships among tetrapods and the other two living groups of lobe-finned fishes, the coelacanth and the lungfishes, are still unresolved and debated. Knowledge of the relationships among these lineages, which originated back in the Devonian, has profound implications for the reconstruction of the evolutionary scenario of the conquest of land. We collected the largest molecular data set on this issue so far, about 3,500 base pairs from seven species of the large 28S nuclear ribosomal gene. All phylogenetic analyses (maximum parsimony, neighbor-joining, and maximum likelihood) point toward the hypothesis that lungfishes and coelacanths form a monophyletic group and are equally closely related to land vertebrates. This evolutionary hypothesis complicates the identification of morphological or physiological preadaptations that might have permitted the common ancestor of tetrapods to colonize land. This is because the reconstruction of its ancestral conditions would be hindered by the difficulty to separate uniquely derived characters from shared derived characters in the coelacanth/lungfish and tetrapod lineages. This molecular phylogeny aids in the reconstruction of morphological evolutionary steps by providing a framework; however, only paleontological evidence can determine the sequence of morphological acquisitions that allowed lobe-finned fishes to colonize land.

A yeast artificial chromosome-based map of the region of chromosome 20 containing the diabetes-susceptibility gene, MODY1, and a myeloid leukemia related gene.

We have generated a physical map of human chromosome bands 20q11.2-20q13.1, a region containing a gene involved in the development of one form of early-onset, non-insulin-dependent diabetes mellitus, MODY1, as well as a putative myeloid tumor suppressor gene. The yeast artificial chromosome contig consists of 71 clones onto which 71 markers, including 20 genes, 5 expressed sequence tags, 32 simple tandem repeat DNA polymorphisms, and 14 sequence-tagged sites have been ordered. This region spans about 18 Mb, which represents about 40% of the physical length of 20q. Using this physical map, we have refined the location of MODY1 to a 13-centimorgan interval (approximately equal to 7 Mb) between D20S169 and D20S176. The myeloid tumor suppressor gene was localized to an 18-centimorgan interval (approximately equal to 13 Mb) between RPN2 and D20S17. This physical map will facilitate the isolation of MODY1 and the myeloid tumor suppressor gene.

Purification of a ligand for the EPH-like receptor HEK using a biosensor-based affinity detection approach.

Advances in screening technologies allowing the identification of growth factor receptors solely by virtue of DNA or protein sequence comparison call for novel methods to isolate corresponding ligand growth factors. The EPH-like receptor tyrosine kinase (RTK) HEK (human EPH-like kinase) was identified previously as a membrane antigen on the LK63 human pre-B-cell line and overexpression in leukemic specimens and cell lines suggested a role in oncogenesis. We developed a biosensor-based approach using the immobilized HEK receptor exodomain to detect and monitor purification of the HEK ligand. A protein purification protocol, which included HEK affinity chromatography, achieved a 1.8 X 10(6)-fold purification of an approximately 23-kDa protein from human placental conditioned medium. Analysis of specific sHEK (soluble extracellular domain of HEK) ligand interactions in the first and final purification steps suggested a ligand concentration of 40 pM in the source material and a Kd of 2-3 nM. Since the purified ligand was N-terminally blocked, we generated tryptic peptides and N-terminal amino acid sequence analysis of 7 tryptic fragments of the S-pyridylethylated protein unequivocally matched the sequence for AL-1, a recently reported ligand for the related EPH-like RTK REK7 (Winslow, J.W., Moran, P., Valverde, J., Shih, A., Yuan, J.Q., Wong, S.C., Tsai, S.P., Goddard, A., Henzel, W.J., Hefti, F., Beck, K.D., & Caras, I.W. (1995) Neuron 14, 973-981). Our findings demonstrate the application of biosensor technology in ligand purification and show that AL-1, as has been found for other ligands of the EPH-like RTK family, binds more than one receptor.

Identification of a minimal sequence of the mouse pro-alpha 1(I) collagen promoter that confers high-level osteoblast expression in transgenic mice and that binds a protein selectively present in osteoblasts.

Based on our previous transgenic mice results, which strongly suggested that separate cell-specific cis-acting elements of the mouse pro-alpha 1(I) collagen promoter control the activity of the gene in different type I collagen-producing cells, we attempted to delineate a short segment in this promoter that could direct high-level expression selectively in osteoblasts. By generating transgenic mice harboring various fragments of the promoter, we identified a 117-bp segment (-1656 to -1540) that is a minimal sequence able to confer high-level expression of a lacZ reporter gene selectively in osteoblasts when cloned upstream of the proximal 220-bp pro-alpha 1(I) promoter. This 220-bp promoter by itself was inactive in transgenic mice and unable to direct osteoblast-specific expression. The 117-bp enhancer segment contained two sequences that appeared to have different functions. The A sequence (-1656 to -1628) was required to obtain expression of the lacZ gene in osteoblasts, whereas the C sequence (-1575 to -1540) was essential to obtain consistent and high-level expression of the lacZ gene in osteoblasts. Gel shift assays showed that the A sequence bound a nuclear protein present only in osteoblastic cells. A mutation in the A segment that abolished the binding of this osteoblast-specific protein also abolished lacZ expression in osteoblasts of transgenic mice.

Phylogenetic resolution within the Elephantidae using fossil DNA sequence from the American mastodon (Mammut americanum) as an outgroup.

DNA was extracted from the extinct American mastodon, the extinct woolly mammoth, and the modern Asian and African elephants to test the traditional morphologically based phylogeny within Elephantidae. Phylogenetic analyses of the aligned sequences of the mitochondrial gene cytochrome b support a monophyletic Asian elephant-woolly mammoth clade when the American mastodon is used as an outgroup. Previous molecular studies were unable to resolve the relationships of the woolly mammoth, Asian elephant, and African elephant because the sequences appear to have evolved at heterogeneous rates and inappropriate outgroups were used for analysis. The results demonstrate the usefulness of fossil molecular data from appropriate sister taxa for resolving phylogenies of highly derived or early radiating lineages.

Analysis of homeodomain function by structure-based design of a transcription factor.

The homeodomain is a 60-amino acid module which mediates critical protein-DNA and protein-protein interactions for a large family of regulatory proteins. We have used structure-based design to analyze the ability of the Oct-1 homeodomain to nucleate an enhancer complex. The Oct-1 protein regulates herpes simplex virus (HSV) gene expression by participating in the formation of a multiprotein complex (C1 complex) which regulates alpha (immediate early) genes. We recently described the design of ZFHD1, a chimeric transcription factor containing zinc fingers 1 and 2 of Zif268, a four-residue linker, and the Oct-1 homeodomain. In the presence of alpha-transinduction factor and C1 factor, ZFHD1 efficiently nucleates formation of the C1 complex in vitro and specifically activates gene expression in vivo. The sequence specificity of ZFHD1 recruits C1 complex formation to an enhancer element which is not efficiently recognized by Oct-1. ZFHD1 function depends on the recognition of the Oct-1 homeodomain surface. These results prove that the Oct-1 homeodomain mediates all the protein-protein interactions that are required to efficiently recruit alpha-transinduction factor and C1 factor into a C1 complex. The structure-based design of transcription factors should provide valuable tools for dissecting the interactions of DNA-bound domains in other regulatory circuits.

Prediction of protein folding class using global description of amino acid sequence.

We present a method for predicting protein folding class based on global protein chain description and a voting process. Selection of the best descriptors was achieved by a computer-simulated neural network trained on a data base consisting of 83 folding classes. Protein-chain descriptors include overall composition, transition, and distribution of amino acid attributes, such as relative hydrophobicity, predicted secondary structure, and predicted solvent exposure. Cross-validation testing was performed on 15 of the largest classes. The test shows that proteins were assigned to the correct class (correct positive prediction) with an average accuracy of 71.7%, whereas the inverse prediction of proteins as not belonging to a particular class (correct negative prediction) was 90-95% accurate. When tested on 254 structures used in this study, the top two predictions contained the correct class in 91% of the cases.