Bioinformatics for rice resources.

The distinguishing feature of the 'new biology' is that it is information intensive. Not only does it demand access to and assimilation of vast data sets accumulated by engineered laboratory processes, but it also demands a previously unimaginable level of data integration across data types and sources. There are various information resources available for rice. In addition, there are various information resources that are not focused on rice but that contain rice data. The challenge for rice researchers and breeders is to access this wealth of data meaningfully. This challenge will grow significantly as international efforts aimed at sequencing the entire rice genome come into full swing. Only through concerted efforts in bioinformatics will the power of these public data be brought to bear on the needs of rice researchers and breeders worldwide. These efforts will need to focus on two large but distinct areas: (1) development of an effective bioinformatics infrastructure (hardware systems, software systems, and software engineers and support staff) and (2) computational biology research in visualization and analysis of very large, complex data sets, such as those that will be developed using high-throughput expression technologies, large-scale insertional mutagenesis, and biochemical profiling of various types. In the midst of the large flow of high-throughput data that the international rice genome sequencing efforts will produce, it is also imperative that integration of those data with unique germplasm data held in trust by the CGIAR be a part of the informatics infrastructure. This paper will focus on the state of rice information resources, the needs of the rice community, and some proposed bioinformatics activities to support these needs.

Improved statistical inference from DNA microarray data using analysis of variance and a Bayesian statistical framework. Analysis of global gene expression in Escherichia coli K12.

We describe statistical methods based on the t test that can be conveniently used on high density array data to test for statistically significant differences between treatments. These t tests employ either the observed variance among replicates within treatments or a Bayesian estimate of the variance among replicates within treatments based on a prior estimate obtained from a local estimate of the standard deviation. The Bayesian prior allows statistical inference to be made from microarray data even when experiments are only replicated at nominal levels. We apply these new statistical tests to a data set that examined differential gene expression patterns in IHF(+) and IHF(-) Escherichia coli cells (Arfin, S. M., Long, A. D., Ito, E. T., Tolleri, L., Riehle, M. M., Paegle, E. S., and Hatfield, G. W. (2000) J. Biol. Chem. 275, 29672-29684). These analyses identify a more biologically reasonable set of candidate genes than those identified using statistical tests not incorporating a Bayesian prior. We also show that statistical tests based on analysis of variance and a Bayesian prior identify genes that are up- or down-regulated following an experimental manipulation more reliably than approaches based only on a t test or fold change. All the described tests are implemented in a simple-to-use web interface called Cyber-T that is located on the University of California at Irvine genomics web site.

Open Source Software meets gene expression.

Use of the Open Source Software (OSS) development model has been crucial in a number of recent technological areas, including operating systems, applications and bioinformatics. The rationale for why OSS is often a better development model than proprietary development and some of the results of this model in the field of Gene Expression are reviewed. The paper concludes with a discussion of why funding agencies should endorse OSS and require funded software projects to be released Open Source.

Characterization of the Borna disease virus phosphoprotein, p23.

Borna disease virus infection is diagnosed by the presence of serum antibodies reactive with the major viral proteins, p40 and p23. Although p40 and p23 are unrelated in amino acid sequence structure, cross-reactive antibodies are described. Protein fragments and synthetic peptides were analyzed to characterize the specificities of antibodies to p23. Epitope mapping revealed eight continuous epitopes accessible on the surface of a predicted structural model for the monomeric and the disulfide-linked dimeric forms of p23. None of these epitopes was reactive with antibodies to p40. Cross-reactivity with monospecific sera and monoclonal antibodies to p40 was found for one discontinuous epitope located at the amino terminus of p23.

A family of POU-domain and Pit-1 tissue-specific transcription factors in pituitary and neuroendocrine development.

The anterior pituitary gland provides a model for investigating the molecular basis for the appearance of phenotypically distinct cell types, within an organ, a central question in development. The rat prolactin and growth hormone genes are selectively expressed in distinct cell types (lactotrophs and somatotrophs) of the anterior pituitary gland, which reflect differential mechanisms of gene activation or restriction because of interactions of multiple factors binding to these genes. We find that the pituitary-specific 33,000 dalton transcription factor, Pit-1, normally expressed in somatotrophs, lactotrophs, and thyrotrophs, can bind to and activate both growth hormone and prolactin promoters in vitro at levels even tenfold lower than those normally present in pituitary cells. In the case of the prolactin gene, high levels of expression in transgenic animals required two cis-active regions; a distal enhancer (-1.8 to -1.5 kb) and a proximal region (-422 to +33 bp). Each of these regions alone can direct low levels of fusion gene expression to prolactin-producing cell types in transgenic mice, but a synergistic interaction between these regions is necessary for high levels of expression. The initial appearance of the prolactin transgene expression closely follows the appearance of high levels of Pit-1, but later increases in expression coincident with appearance of mature lactotrophs suggest the operation of additional, critical positive factor(s). Unexpectedly, transgenes containing the distal enhancer removed from its normal context are expressed in both the prolactin-producing lactotrophs and the TSH-producing thyrotrophs, thereby suggesting that sequences flanking this enhancer are necessary to restrict expression to the correct cell type within the pituitary. These data indicate that distinct processes of gene activation and restriction are necessary for the fidelity of cell-type specific expression within an organ. Consistent with this model, we find that lactotroph cell lines that cannot express the growth hormone gene contain high levels of functional Pit-1. We suggest a large, highly related POU-domain gene family, potentially exceeding 100 members, has been conserved and expanded in evolution to meet the increasing requirements for more intricate patterns of cell phenotypes. The POU-domain subgroup of the homeodomain gene family, in concert with other homeodomain proteins and with other classes of transcription factors, is likely to contribute to the establishment of the mammalian neuroendocrine system.

A pituitary POU domain protein, Pit-1, activates both growth hormone and prolactin promoters transcriptionally.

The anterior pituitary gland provides a model for investigating the molecular basis for the appearance of phenotypically distinct cell types within an organ, a central question in development. The rat prolactin and growth hormone genes are expressed selectively in distinct cell types (lactotrophs and somatotrophs, respectively) of the anterior pituitary gland, reflecting differential mechanisms of gene activation or restriction, as a result of the interactions of multiple factors binding to these genes. We find that when the pituitary-specific 33-kD transcription factor Pit-1, expressed normally in both lactotrophs and somatotrophs, is expressed in either the heterologous HeLa cell line or in bacteria, it binds to and activates transcription from both growth hormone and prolactin promoters in vitro at levels even 10-fold lower than those normally present in pituitary cells. This suggests that a single factor, Pit-1, may be capable of activating the expression of two genes that define different anterior pituitary cell phenotypes. Because a putative lactotroph cell line (235-1) that does not express the growth hormone gene, but only the prolactin gene, appears to contain high levels of functional Pit-1, a mechanism selectively preventing growth hormone gene expression may, in part, account for the lactotroph phenotype.

A tissue-specific transcription factor containing a homeodomain specifies a pituitary phenotype.

Multiple related cis-active elements required for cell-specific activation of the rat prolactin gene appear to bind a pituitary-specific positive transcription factor(s), referred to as Pit-1. DNA complementary to Pit-1 mRNA, cloned on the basis of specific binding to AT-rich cell-specific elements in the rat prolactin and growth hormone genes, encodes a 33 kd protein with significant similarity at its carboxyl terminus to the homeodomains encoded by Drosophila developmental genes. Pit-1 mRNA is expressed exclusively in the anterior pituitary gland in both somatotroph and lactotroph cell types, which produce growth hormone and prolactin, respectively. Pit-1 expression in heterologous cells (HeLa) selectively activates prolactin and growth hormone fusion gene expression, suggesting that Pit-1 is sufficient to confer a characteristic pituitary phenotype. The structure of Pit-1 and its recognition elements suggests that metazoan tissue phenotype is controlled by a family of transcription factors that bind to related cis-active elements and contain several highly conserved domains.

Adrenal gland denervation and diving in ducks.

The extreme elevation in plasma levels of free norepinephrine (NE) and free epinephrine (EP), which occurs during forced diving of ducks (Anas platyrhynchos), was studied before and after denervation of the adrenal glands. In intact animals both NE and EP concentration increased by up to two orders of magnitude in a 4-min dive but by a significantly lesser amount if the duck breathed O2 before the dive. Denervating the adrenal glands reduced the amounts of both catecholamines (CA) released during dives, plasma EP decreased to 10%, and NE to 50% of values obtained before denervation. Breathing O2 before a dive virtually eliminated CA release in denervates, indicating that hypoxia was the important non-neural releasing agent. Hypoxia was also the most important neural releasing agent compared with hypercapnia, acidosis, or hypoglycemia. Adrenal denervation did not cause significant changes in heart rate, blood pressure, arterial blood gas tensions, pH, or plasma glucose during dives, although denervation caused increased variation in some of these variables. In ducks CA release in dives is largely due to decreasing arterial O2 partial pressure, and full expression of the response is dependent on intact innervation of the adrenal gland.

Two different cis-active elements transfer the transcriptional effects of both EGF and phorbol esters.

Short cis-active sequences of the rat prolactin or Moloney murine leukemia virus genes transfer transcriptional regulation by both epidermal growth factor and phorbol esters to fusion genes. These sequences act in a position- and orientation-independent manner. Competitive binding analyses with nuclear extracts from stimulated and unstimulated cells suggest that different trans-acting factors associate with the regulatory sequence of each gene. A model is proposed suggesting that both epidermal growth factor and phorbol esters stimulate the transcription of responsive genes via discrete classes of hormone-dependent, enhancer-like elements that bind different trans-acting factors, even in the absence of hormone stimulation.