BioJava: an open-source framework for bioinformatics.

SUMMARY: BioJava is a mature open-source project that provides a framework for processing of biological data. BioJava contains powerful analysis and statistical routines, tools for parsing common file formats and packages for manipulating sequences and 3D structures. It enables rapid bioinformatics application development in the Java programming language. AVAILABILITY: BioJava is an open-source project distributed under the Lesser GPL (LGPL). BioJava can be downloaded from the BioJava website (http://www.biojava.org). BioJava requires Java 1.5 or higher. All queries should be directed to the BioJava mailing lists. Details are available at http://biojava.org/wiki/BioJava:MailingLists.

Functional antagonism between E2F family members.

E2F is a heterogenous transcription factor and its role in cell cycle control results from the integrated activities of many different E2F family members. Unlike mammalian cells, that have a large number of E2F-related genes, the Drosophila genome encodes just two E2F genes, de2f1 and de2f2. Here we show that de2f1 and de2f2 provide different elements of E2F regulation and that they have opposing functions during Drosophila development. dE2F1 and dE2F2 both heterodimerize with dDP and bind to the promoters of E2F-regulated genes in vivo. dE2F1 is a potent activator of transcription, and the loss of de2f1 results in the reduced expression of E2F-regulated genes. In contrast, dE2F2 represses the transcription of E2F reporters and the loss of de2f2 function results in increased and expanded patterns of gene expression. The loss of de2f1 function has previously been reported to compromise cell proliferation. de2f1 mutant embryos have reduced expression of E2F-regulated genes, low levels of DNA synthesis, and hatch to give slow-growing larvae. We find that these defects are due in large part to the unchecked activity of dE2F2, since they can be suppressed by mutation of de2f2. Examination of eye discs from de2f1; de2f2 double-mutant animals reveals that relatively normal patterns of DNA synthesis can occur in the absence of both E2F proteins. This study shows how repressor and activator E2Fs are used to pattern transcription and how the net effect of E2F on cell proliferation results from the interplay between two types of E2F complexes that have antagonistic functions.

The Drosophila Ret gene is transcribed in multiple alternatively spliced forms.

We have cloned and sequenced the 5' and 3' ends of the Drosophila homolog of the vertebrate c-ret gene, Ret, and have derived from it the predicted protein sequence of Ret. The extracellular domain of Ret is very widely diverged from that of its vertebrate counterparts but the cadherin motif present in vertebrate c-ret proteins can also be discerned in Ret. As with the vertebrate gene, multiple splice variants were detected at the 5'-end of Ret, one of which inserts an exon with a protein-terminating frameshift into the cDNA. In contrast to human c-ret, which may vary its signalling specificity by using splicing-derived, alternative C-terminal sequences, Ret cDNAs showed no variation at their 3'-ends.

The balance between isoforms of the prickle LIM domain protein is critical for planar polarity in Drosophila imaginal discs.

The tissue polarity mutants in Drosophila include a set of conserved gene products that appear to be involved in the control of cytoskeletal architecture. Here we show that the tissue polarity gene prickle (pk) encodes a protein with a triple LIM domain and a novel domain that is present in human, murine, and Caenorhabditis elegans homologs which we designate PET. Three transcripts have been identified, pk, pkM, and sple, encoding 93-, 100-, and 129-kD conceptual proteins, respectively. The three transcripts span 70 kb and share 6 exons that contain the conserved domains. The pk and sple transcripts are expressed with similar tissue-specific patterns but have qualitatively different activities. The phenotypes of pk mutants, and transgenic flies in which the different isoforms are overexpressed show that the balance between Pk and Sple is critical for the specification of planar polarity. In addition, these phenotypes suggest a tessellation model in which the alignment of wing hairs is dependent on cell shape and need not reflect fine-grained positional information. Lack of both pk and sple transcripts gives a phenotype affecting the whole body surface that is similar to those of dishevelled and frizzled (fz) suggesting a functional relationship between pk and fz signaling.

The Epstein-Barr virus gene BHRF1, a homologue of the cellular oncogene Bcl-2, inhibits apoptosis induced by gamma radiation and chemotherapeutic drugs.

Analysis of apoptosis, active and controllable cell death, has demonstrated that the size of a cell population can be regulated by changes in the cell death rate as well as in the rates of proliferation and differentiation. Factors which alter the rate of cell death, such as expression of the proto-oncogene bcl-2, can therefore directly affect the number of cells within a population. Bcl-2 has been shown to suppress apoptosis in response to a variety of stimuli and to act as a complementary survival signal for the random acquisition of other oncogenic mutations, such as deregulated c-myc. The Epstein Barr virus (EBV) gene BHRF1 was the first of a family of bcl-2 homologues now being identified. BHRF1 and bcl-2 share 25% primary amino acid sequence homology. Here we show that gamma radiation and several cytotoxic anticancer agents induce apoptosis in Burkitt's lymphoma (BL) cell lines, as has been found in several other systems. Using gene transfection studies we have also shown that expression of either BHRF1 or bcl-2 in BL cell lines significantly suppresses apoptosis in response to a variety of anticancer treatment. This has confirmed that BHRF1 is functionally homologous to bcl-2 in B-cells and suggests that BHRF1 may act to prevent apoptosis during EBV infection, maximising virus particle production, as has been suggested for other human and insect viral genes. Suppression of chemotherapeutic drug induced cell death by bcl-2 and BHRF1 as demonstrated in this cell system, results in resistance to a variety of different agents and may represent an alternative mechanism by which multidrug resistance arises during chemotherapy.

The Epstein-Barr virus latent membrane protein-1 (LMP1) mediates activation of NF-kappa B and cell surface phenotype via two effector regions in its carboxy-terminal cytoplasmic domain.

The Epstein-Barr virus (EBV) encoded latent membrane protein, LMP1, is oncogenic in rodent fibroblasts and is an essential effector protein in EBV-induced growth-transformation of human B lymphocytes. Previous structure-function studies with LMP1 have relied largely on rodent fibroblast transformation as a functional readout, with apparently conflicting results. We have now analysed several LMP1 mutants in various human cell types, including B cells, T cells and epithelial cells, using two independent functional assays; (i) activation of NF-kappa B, and (ii) induction of two cell surface activation markers, CD54 and CD40. The results suggest that the cytosolic N-terminus is not essential for LMP1 function in any cell type studied. The third and fourth transmembrane helices and the intracytosolic loops are dispensable for activation of NF-kappa B, but they do influence the induction of CD54 and CD40. The major effector domain appears to be the cytosolic C-terminus in which were identified two 'C-terminal activating regions', CTAR-1 (residues 194-232) and CTAR-2 (residues 351-386). Whilst the exact results depended upon the host cell line, CTAR-2 was generally more important for activation of NF-kappa B, and both CTAR-1 and CTAR-2 were required for optimal induction of CD54 and CD40. Analysis of NF-kappa B activation by LMP1 in Rat-1 fibroblasts indicated that many mutations that were functional in human cells were poorly tolerated in the rodent cells; a result that is in broad agreement with published Rat-1 transformation data.

Upregulation of bcl-2 by the Epstein-Barr virus latent membrane protein LMP1: a B-cell-specific response that is delayed relative to NF-kappa B activation and to induction of cell surface markers.

An ability of the Epstein-Barr virus latent membrane protein LMP1 to enhance the survival of infected B cells through upregulation of the bcl-2 oncogene was first suggested by experiments involving gene transfection and the selection of stable LMP1+ clones (S. Henderson, M. Rowe, C. Gregory, F. Wang, E. Kieff, and A. Rickinson, Cell 65:1107-1115, 1991). However, it was not possible to ascertain whether Bcl-2 upregulation was a specific consequence of LMP1 expression or an artifact of the selection procedure whereby rare Bcl-2+ cells already present in the starting population might best be able to tolerate the potentially toxic effects of LMP1. We therefore reexamined this issue by using two different experimental approaches that allowed LMP1-induced effects to be monitored immediately following expression of the viral protein and in the absence of selective pressures; activation of the NF-kappa B transcription factor and upregulation of the cell adhesion molecule ICAM-1 were used as early indices of LMP1 function. In the first approach, stable clones of two B-cell lines carrying an LMP1 gene under the control of an inducible metallothionein promoter were induced to express LMP1 in all cells. Activation of NK-kappa B and upregulation of ICAM-1 occurred within 24 h and were followed at 48 to 72 h by upregulation of Bcl-2. In the second approach, we tested the generality of this phenomenon by transiently expressing LMP1 from a strong constitutively active promoter in a range of different cell types. All six B-cell lines tested showed NF-kappa B activation in response to LMP1 expression, and this was followed in five of six lines by expression of ICAM-1 and Bcl-2. In the same experiments, all three non-B-cell lines showed NF-kappa B activation and ICAM-1 upregulation but never any effect upon Bcl-2. We therefore conclude that Bcl-2 upregulation is part of the panoply of cellular changes induced by LMP1 but that the effect is cell type specific. Our data also suggest that whilst NF-kappa B may be an essential component of LMP1 signal transduction, other cell-specific factors may be required to effect some functions of the viral protein.

Evolutionary conservation of function among mammalian, avian, and viral homologs of the Bcl-2 oncoprotein.

The bcl-2 gene was originally cloned because of its involvement in B-cell lymphomas and encodes a 25-kD integral membrane protein that has been shown to inhibit programmed cell death (also termed apoptosis) in a wide variety of circumstances. The Epstein-Barr Virus (EBV) also has been implicated in B-cell malignancies and interestingly contains an open reading frame (BHRF-1) predicting a 19-kD protein with 22% homology to Bcl-2. To compare the functions of p26-Bcl-2 and p19-BHRF-1, we stably introduced expression plasmids encoding these proteins into a murine interleukin-3 (IL-3)-dependent hemopoietic cell line, 32D. Removal of IL-3 from cultures of control-transfected 32D cells resulted in internucleosomal DNA cleavage (a hallmark of programmed cell death) and loss of cell survival. In contrast, 32D cells containing high levels of p26-Bcl-2 or p19-BHRF-2 proteins exhibited prolonged survival and markedly delayed DNA degradation under the same conditions of IL-3 deprivation. As a first attempt to determine the functional importance of amino acid sequences that are conserved between the Bcl-2 and BHRF-1 proteins, we used site-specific mutagenesis to replace two conserved cysteine residues with alanines (positions 158 and 219) in the human Bcl-2 protein. Comparisons of the wild-type and cysteine-minus human Bcl-2 proteins in S49 lymphoma cells revealed equivalent ability to block glucocorticoid-induced cell death and DNA fragmentation, indicating that these two conserved cysteines are not critical for Bcl-2 oncoprotein function. Investigations in 32D cells of an avian homolog of Bcl-2 cloned from the chicken also revealed conservation of function with the human Bcl-2 protein, despite the presence of a 48-amino-acid region of divergent sequence. Taken together, these data demonstrate that despite marked differences in their predicted amino-acid sequences, the human, chicken, and EBV versions of Bcl-2 have retained the structural characteristics necessary to interface with pathways involved in the regulation of programmed cell death in murine cells. The findings thus contribute to the mapping of functional domains in Bcl-2 proteins, and raise the possibility that the EBV-encoded p19-BHRF-1 protein may be able to substitute for p26-Bcl-2 in the development of some types of cancer.

Epstein-Barr virus-coded BHRF1 protein, a viral homologue of Bcl-2, protects human B cells from programmed cell death.

Epstein-Barr virus, a human herpesvirus that persists within the B-lymphoid system, can enhance the survival potential of latently infected B cells in vitro through up-regulation of the cellular survival protein Bcl-2. The possibility that an analogous effect is operative in lytically infected cells was suggested by the observation of distant sequence homology between an Epstein-Barr virus-coded early lytic cycle protein, BHRF1, and Bcl-2. Here we show by gene transfer that BHRF1 resembles Bcl-2 both in its subcellular localization and in its capacity to enhance B-cell survival. Thus confocal microscopic analysis of cells acutely cotransfected with BHRF1 and Bcl-2 expression vectors revealed substantial colocalization of the two proteins in the cytoplasm. In subsequent experiments, stable BHRF1 gene transfectants of Burkitt lymphoma cells paralleled Bcl-2 transfectants in their enhanced survival under conditions that induce cell death by apoptosis. Despite their limited sequence conservation, therefore, the two proteins appear to be functionally homologous. We suggest that BHRF1 provides an alternative, Bcl-2-independent, means of enhancing B-cell survival that may operate during the virus lytic cycle.

Dilated heart failure in transgenic mice expressing the Epstein-Barr virus nuclear antigen-leader protein.

The Epstein-Barr virus nuclear antigen-leader protein (EBNA-LP) is required for high efficiency B lymphocyte growth transformation by the virus. To test the potential contribution of EBNA-LP to tumorigenesis in vivo, we produced transgenic mice carrying an EBNA-LP cDNA construct, using the widely expressed metallothionein promoter. Expression of EBNA-LP was detected in liver, kidney, heart, lung and spleen. There were no apparent oncogenic consequences of EBNA-LP expression. Unexpectedly however, at ages ranging from about 4 months to over a year, transgenic mice developed symptoms of congestive heart failure, including left ventricular dilatation, right ventricular hypertrophy, left atrial thrombosis, pulmonary oedema and hydrothorax. Fibrillation was not apparent in the electrocardiograph; however a reduction in T-wave amplitude suggested that the development of an abnormality of ventricular repolarization may precede the manifestation of overt symptoms. The highly predictable development of dilated heart failure in these transgenic mice suggests they may be a useful model for the pathophysiological changes associated with human dilated cardiomyopathy.

The expression of the retinoblastoma gene product Rb1 in primary and adenovirus-transformed human cells.

Polyclonal antibodies to the human retinoblastoma gene product (Rb1) have been produced in rats by immunisation with a fusion protein comprising part of Rb1 together with the E. coli beta-Gal sequence. We have used these antibodies in Western blotting studies to screen a number of human foetal tissues and organs and found approximately similar levels of expression of Rb1 in all of them. The protein seems to be somewhat more abundant in some cell lines produced by transfection of human embryo retinal (HER) cells with adenovirus 12 early region 1 (Ad 12 E1), Ad 5 E1, Ad 2 E1A + mutant N-ras or SV40 DNA. Using co-immunoprecipitation followed by Western blotting we have shown that the Rb1 protein binds to Ad 12 E1A 266 and 235 amino acid proteins. This interaction is ionic strength dependent but is unaffected by non-ionic detergent up to a concentration of at least 1%. In Ad 12 infected human cells it appears that less E1A is bound to Rb1 than in the transformants. These results are discussed in view of the known similarities and differences between the amino acid sequences of Ad 12 and Ad 5 E1A proteins.