Mixture modeling provides an effective approach to the differential expression problem in microarray data analysis. Methods based on fully parametric mixture models are available, but lack of fit in some examples indicates that more flexible models may be beneficial. Existing, more flexible, mixture models work at the level of one-dimensional gene-specific summary statistics, and so when there are relatively few measurements per gene these methods may not provide sensitive detectors of differential expression. We propose a hierarchical mixture model to provide methodology that is both sensitive in detecting differential expression and sufficiently flexible to account for the complex variability of normalized microarray data. EM-based algorithms are used to fit both parametric and semiparametric versions of the model. We restrict attention to the two-sample comparison problem; an experiment involving Affymetrix microarrays and yeast translation provides the motivating case study. Gene-specific posterior probabilities of differential expression form the basis of statistical inference; they define short gene lists and false discovery rates. Compared to several competing methodologies, the proposed methodology exhibits good operating characteristics in a simulation study, on the analysis of spike-in data, and in a cross-validation calculation.

No abstract

No abstract

Brome mosaic virus (BMV), a member of the alphavirus-like superfamily of human-, animal-, and plant-infecting (ϩ)RNA viruses, has been studied as a model for viral RNA replication, encapsidation, recombination, and other processes (3). BMV has three genomic RNAs. RNAs 1 and 2 encode the interacting, multifunctional 1a helicase-like and 2a polymerase RNA replication factors (4, 5), which form endoplasmic reticulum (ER) membrane-associated RNA replication complexes with functional similarities to the replicative cores of retrovirus and double-strand (ds)RNA virus virions (6). RNA3 encodes protein 3a that enables infection spread between cells in natural hosts. The negative-strand [(Ϫ) RNA]3 replication intermediate also serves as a template for synthesis of a subgenomic (sg) mRNA, RNA4, which encodes the viral coat protein (Fig. 1A).The yeast Saccharomyces cerevisiae has proven a valuable model for normal and disease processes in human and other cells. The unusual ability of BMV to direct its genomic RNA replication, gene expression, encapsidation, and other processes in this yeast (7,8) has allowed traditional yeast mutagenic analyses that have identified host genes involved in multiple steps of BMV RNA replication and gene expression. Such host genes encode a wide variety of functions and contribute to diverse replication steps, including supporting and regulating viral translation, selecting and recruiting viral RNAs as replication templates, activating the RNA replication complex through chaperones, and providing a lipid profile compatible with membrane-associated viral RNA replication (9-14; reviewed in refs. 2 and 15).Here, we sought to develop a more rapid, global method to systematically identify yeast host factors with effects on BMV RNA replication by using an ordered array of yeast deletion strains (16) to assay virus replication in the absence of each of Ϸ4,500 yeast factors, which is Ϸ80% of the yeast genome. We describe screening this deletion array by using a whole-cell assay based on BMV-directed Renilla luciferase (Rluc) expression by pathways dependent on viral RNA replication and viral RNAdirected sg mRNA synthesis. The assay identified nearly 100 host genes whose absence repressed or enhanced BMV-directed Rluc expression by 3-to 25-fold. The results provide a significantly expanded view of virus-host interactions and should advance understanding of virus and cell pathways. Materials and MethodsYeast. YMI04 and ded1i yeast were described (11). Strains BY4743 (WT; ref. 17) and the homozygous diploid deletion series (BY4743 strain background; ref. 16) were from Research Genetics (Huntsville, AL). Standard yeast techniques were used (18), except for 96-well transformations, which were based on a one-step procedure (19). Briefly, yeast were grown to saturation overnight at 30°C in 96-well plates (1.2 ml per well), pelleted, suspended in 100 l of transformation mix (0.18 M LiAc, pH 5.5, 36% polyethylene glycol-3350, 90 mM DTT, 0.5 mg͞ml sheared salmon sperm DNA, and 20 g͞ml of each plasmid), incubate...

Inevitably, viruses depend on host factors for their multiplication. Here, we show that hepatitis C virus (HCV) RNA translation and replication depends on Rck/p54, LSm1, and PatL1, which regulate the fate of cellular mRNAs from translation to degradation in the 5 -3 -deadenylation-dependent mRNA decay pathway. The requirement of these proteins for efficient HCV RNA translation was linked to the 5 and 3 untranslated regions (UTRs) of the viral genome. Furthermore, LSm1-7 complexes specifically interacted with essential cis-acting HCV RNA elements located in the UTRs. These results bridge HCV life cycle requirements and highly conserved host proteins of cellular mRNA decay. The previously described role of these proteins in the replication of 2 other positive-strand RNA viruses, the plant brome mosaic virus and the bacteriophage Qß, pinpoint a weak spot that may be exploited to generate broad-spectrum antiviral drugs.deadenylation-dependent mRNA decay ͉ HCV ͉ host factors ͉ LSm1-7 ͉ Rck/p54

Brome mosaic virus (BMV), a positive-strand RNA virus in the alphavirus-like superfamily, encodes two RNA replication factors. Membrane-associated 1a protein contains a helicase-like domain and RNA capping functions. 2a, which is targeted to membranes by 1a, contains a central polymerase-like domain. In the absence of 2a and RNA replication, 1a acts through an intergenic replication signal in BMV genomic RNA3 to stabilize RNA3 and induce RNA3 to associate with cellular membrane. Multiple results imply that 1a-induced RNA3 stabilization reflects interactions involved in recruiting RNA3 templates into replication. To determine if 1a had similar effects on another BMV RNA replication template, we constructed a plasmid expressing BMV genomic RNA2 in vivo. In vivo-expressed RNA2 templates were replicated upon expression of 1a and 2a. In the absence of 2a, 1a stabilized RNA2 and induced RNA2 to associate with membrane. Deletion analysis demonstrated that 1a-induced membrane association of RNA2 was mediated by sequences in the 5-proximal third of RNA2. The RNA2 5 untranslated region was sufficient to confer 1a-induced membrane association on a nonviral RNA. However, sequences in the N-terminal region of the 2a open reading frame enhanced 1a responsiveness of RNA2 and a chimeric RNA. A 5-terminal RNA2 stem-loop important for RNA2 replication was essential for 1a-induced membrane association of RNA2 and, like the 1a-responsive RNA3 intergenic region, contained a required box B motif corresponding to the T⌿C stem-loop of host tRNAs. The level of 1a-induced membrane association of various RNA2 mutants correlated well with their abilities to serve as replication templates. These results support and expand the conclusion that 1a-induced BMV RNA stabilization and membrane association reflect early, 1a-mediated steps in viral RNA replication.During infection, the genomic RNAs of positive-strand RNA viruses first must be translated to generate RNA replication factor(s) and other proteins and then must serve as templates for negative-strand RNA synthesis. One complication of these dual template functions is the ability of 5Ј-to-3Ј processive ribosomes to block negative-strand RNA synthesis by 3Ј-to-5Ј processive polymerase (11,20). Therefore, positivestrand RNA viruses must have evolved mechanisms to regulate the alternate template functions of genomic RNA, including mechanisms to clear ribosomes from the RNA and allow transfer of the RNA template from translation to RNA replication (20,25).Brome mosaic virus (BMV), a member of the alphavirus-like superfamily of positive-strand RNA viruses, has three genomic RNAs with 5Ј caps and tRNA-like 3Ј ends (1, 51). RNA1 and RNA2 encode nonstructural proteins 1a and 2a, respectively, which direct RNA replication and contain domains conserved with other superfamily members (4, 19, 25). 1a contains an N-terminal domain with m 7 G methyltransferase and covalent GTP binding activities implicated in viral RNA capping (3, 30) and a C-terminal domain with all motifs of DEAD box RNA helicases (22...

Recent results suggest that cytoplasmic mRNAs can form translationally repressed messenger ribonucleoprotein particles (mRNPs) capable of decapping and degradation, or accumulation into cytoplasmic processing bodies (P-bodies), which can function as sites of mRNA storage. The proteins that function in transitions between the translationally repressed mRNPs that accumulate in P-bodies and mRNPs engaged in translation are largely unknown. Herein, we demonstrate that the yeast translation initiation factor Ded1p can localize to P-bodies. Moreover, depletion of Ded1p leads to defects in P-body formation. Overexpression of Ded1p results in increased size and number of P-bodies and inhibition of growth in a manner partially suppressed by loss of Pat1p, Dhh1p, or Lsm1p. Mutations that inactivate the ATPase activity of Ded1p increase the overexpression growth inhibition of Ded1p and prevent Ded1p from localizing in P-bodies. Combined with earlier work showing Ded1p can have a positive effect on translation, these results suggest that Ded1p is a bifunctional protein that can affect both translation initiation and P-body formation.

Plant viral movement proteins mediate the cell-to-cell movement of nucleic acids. This involves either a direct interaction between the viral movement protein and the nucleic acid or an indirect interaction involving host factors. The bipartite geminiviruses possess two movement proteins, BV1 and BC1, that coordinate movement of viral DNA across nuclear and plasmodesmal boundaries, respectively. Here, we demonstrate that both BV1 and BC1 interact directly with DNA and, in addition, that they have the unique property to recognize DNA on the basis of form and size rather than sequence. This is a novel feature for plant virus movement proteins and raises the possibility that BV1 and BC1 may be determinants of genome size in the bipartite geminiviruses.