Ensemble-Based Network Aggregation Improves the Accuracy of Gene Network Reconstruction

Reverse engineering approaches to constructing gene regulatory networks (GRNs) based on genome-wide mRNA expression data have led to significant biological findings, such as the discovery of novel drug targets. However, the reliability of the reconstructed GRNs needs to be improved. Here, we propose an ensemble-based network aggregation approach to improving the accuracy of network topologies constructed from mRNA expression data. To evaluate the performances of different approaches, we created dozens of simulated networks from combinations of gene-set sizes and sample sizes and also tested our methods on three Escherichia coli datasets. We demonstrate that the ensemble-based network aggregation approach can be used to effectively integrate GRNs constructed from different studies – producing more accurate networks. We also apply this approach to building a network from epithelial mesenchymal transition (EMT) signature microarray data and identify hub genes that might be potential drug targets. The R code used to perform all of the analyses is available in an R package entitled “ENA”, accessible on CRAN (http://cran.r-project.org/web/packages/ENA/).     

Unusual Regulation of Splicing of the Cholinergic Locus in Caenorhabditis elegans

The essential neurotransmitter acetylcholine functions throughout the animal kingdom. In Caenorhabditis elegans, the acetylcholine biosynthetic enzyme [choline acetyltransferase (ChAT)] and vesicular transporter [vesicular acetylcholine transporter (VAChT)] are encoded by the cha-1 and unc-17 genes, respectively. These two genes compose a single complex locus in which the unc-17 gene is nested within the first intron of cha-1, and the two gene products arise from a common pre-messenger RNA (pre-mRNA) by alternative splicing. This genomic organization, known as the cholinergic gene locus (CGL), is conserved throughout the animal kingdom, suggesting that the structure is important for the regulation and function of these genes. However, very little is known about CGL regulation in any species. We now report the identification of an unusual type of splicing regulation in the CGL of C. elegans, mediated by two pairs of complementary sequence elements within the locus. We show that both pairs of elements are required for efficient splicing to the distal acceptor, and we also demonstrate that proper distal splicing depends more on sequence complementarity within each pair of elements than on the sequences themselves. We propose that these sequence elements are able to form stem-loop structures in the pre-mRNA; such structures would favor specific splicing alternatives and thus regulate CGL splicing. We have identified complementary elements at comparable locations in the genomes of representative species of other animal phyla; we suggest that this unusual regulatory mechanism may be a general feature of CGLs.     

Genetic Analysis of Resistance and Sensitivity to 2-Deoxyglucose in Saccharomyces cerevisiae

Aerobic glycolysis is a metabolic pathway utilized by human cancer cells and also by yeast cells when they ferment glucose to ethanol. Both cancer cells and yeast cells are inhibited by the presence of low concentrations of 2-deoxyglucose (2DG). Genetic screens in yeast used resistance to 2-deoxyglucose to identify a small set of genes that function in regulating glucose metabolism. A recent high throughput screen for 2-deoxyglucose resistance identified a much larger set of seemingly unrelated genes. Here, we demonstrate that these newly identified genes do not in fact confer significant resistance to 2-deoxyglucose. Further, we show that the relative toxicity of 2-deoxyglucose is carbon source dependent, as is the resistance conferred by gene deletions. Snf1 kinase, the AMP-activated protein kinase of yeast, is required for 2-deoxyglucose resistance in cells growing on glucose. Mutations in the SNF1 gene that reduce kinase activity render cells hypersensitive to 2-deoxyglucose, while an activating mutation in SNF1 confers 2-deoxyglucose resistance. Snf1 kinase activated by 2-deoxyglucose does not phosphorylate the Mig1 protein, a known Snf1 substrate during glucose limitation. Thus, different stimuli elicit distinct responses from the Snf1 kinase.     

A Caenorhabditis elegans RNA-Directed RNA Polymerase in Sperm Development and Endogenous RNA Interference

Short interfering RNAs (siRNAs) are a class of regulatory effectors that enforce gene silencing through formation of RNA duplexes. Although progress has been made in identifying the capabilities of siRNAs in silencing foreign RNA and transposable elements, siRNA functions in endogenous gene regulation have remained mysterious. In certain organisms, siRNA biosynthesis involves novel enzymes that act as RNA-directed RNA polymerases (RdRPs). Here we analyze the function of a Caenorhabditis elegans RdRP, RRF-3, during spermatogenesis. We found that loss of RRF-3 function resulted in pleiotropic defects in sperm development and that sperm defects led to embryonic lethality. Notably, sperm nuclei in mutants of either rrf-3 or another component of the siRNA pathway, eri-1, were frequently surrounded by ectopic microtubule structures, with spindle abnormalities in a subset of the resulting embryos. Through high-throughput small RNA sequencing, we identified a population of cellular mRNAs from spermatogenic cells that appear to serve as templates for antisense siRNA synthesis. This set of genes includes the majority of genes known to have enriched expression during spermatogenesis, as well as many genes not previously known to be expressed during spermatogenesis. In a subset of these genes, we found that RRF-3 was required for effective siRNA accumulation. These and other data suggest a working model in which a major role of the RRF-3/ERI pathway is to generate siRNAs that set patterns of gene expression through feedback repression of a set of critical targets during spermatogenesis.REPRESSION of gene expression by small RNAs of ∼20–30 nt in length is important for many aspects of multicellular eukaryotic development. A variety of classes of small RNA with distinct structural features, modes of biogenesis, and biological functions have been identified (reviewed in Hutvagner and Simard 2008). We are particularly interested in a class of small RNAs, called endogenous short interfering RNAs (siRNAs), that are similar to intermediates in exogenously triggered RNA interference (RNAi) in their perfect complementarity to mRNA targets. High-throughput sequencing technology has provided a valuable tool for characterization of endogenous siRNA populations from many diverse sources, including mouse embryonic stem cells (Babiarz et al. 2008), Drosophila heads (Ghildiyal et al. 2008), and Arabidopsis pollen (Slotkin et al. 2009). These siRNAs have been proposed to function in the regulation of both cellular processes and genome defense through downregulation of gene expression. Caenorhabditis elegans, like plants and fungi, utilizes RNA-copying enzymes called RNA-directed RNA polymerases (RdRPs) as part of the RNAi machinery (Smardon et al. 2000; Sijen et al. 2001). While two of the C. elegans RdRPs are nonessential (RRF-1 and RRF-2), mutations in either of the remaining two (EGO-1 or RRF-3) lead to fertility defects (Smardon et al. 2000; Simmer et al. 2002). RRF-3 is functionally distinct from EGO-1 in that the RRF-3 requirement in fertility is temperature dependent. In addition, RRF-3 activity has an inhibitory effect on exogenously triggered RNAi (resulting in an ERI, or enhanced RNAi, mutant phenotype in rrf-3 mutants). Mutants lacking either RRF-3 or another ERI factor, ERI-1, have been used as experimental tools because of their enhanced sensitivity in RNAi-based screens. One proposed mechanism for the enhancement in RNAi in rrf-3 and eri mutants has been a competition for cofactors between the exogenously triggered RNAi pathway and an endogenous RNAi pathway. Consistent with this hypothesis, siRNAs corresponding to several genes have been shown by Northern analysis to depend upon RRF-3 and other ERI factors for their accumulation (Duchaine et al. 2006; Lee et al. 2006; Yigit et al. 2006). Global microarray analyses have also been undertaken to identify messenger RNAs whose expression is affected by RRF-3 and ERI-1 (Lee et al. 2006; Asikainen et al. 2007).A functional significance of the RRF-3/ERI pathway has been inferred by the inability of rrf-3, eri-1, eri-3, and eri-5 mutant strains to propagate at a high growth temperature (Simmer et al. 2002; Duchaine et al. 2006). Rather than producing temperature-sensitive mutant protein effects, RRF-3 and other ERI proteins are thought to act in a temperature-sensitive process, as evidenced by the predicted truncated and presumed nonfunctional protein fragments that would result from the available deletion alleles and by their shared temperature-sensitive phenotypes. rrf-3 mutant animals have been observed to exhibit X-chromosome missegregation (Simmer et al. 2002) and an unusual persistence of a chromatin mark on the X chromosome during male spermatogenesis (Maine et al. 2005). X-chromosome missegregation and defective spermatogenesis have been referred to in previous studies of eri-1 (Kennedy et al. 2004) and eri-3 and eri-5 (Duchaine et al. 2006). Furthermore, eri-3 mutant sterility can be rescued by insemination with wild-type sperm (Duchaine et al. 2006).Here we investigated the role of RRF-3 during spermatogenesis. We found defects evident at multiple stages, including after fertilization, where defects in rrf-3 mutant sperm can produce subsequent nonviable embryos. By using high-throughput sequencing, we characterized a large population of siRNAs present in spermatogenic cells and found a strong enrichment for antisense siRNAs from genes with known mRNA expression during spermatogenesis. While the majority of siRNA production during spermatogenesis does not require RRF-3, we found a set of genes for which siRNA production was dependent upon RRF-3. Existing data indicate increased expression for these genes in rrf-3 and/or eri-1 mutants. Taken together, our analyses suggest a working model in which the RRF-3/ERI pathway generates siRNAs that downregulate specific genes during spermatogenesis, with this regulation playing a key role in generating functional sperm.     

Functional Genomic Identification of Genes Required for Male Gonadal Differentiation in Caenorhabditis elegans

The Caenorhabditis elegans somatic gonad develops from a four-cell primordium into a mature organ that differs dramatically between the sexes in overall morphology (two arms in hermaphrodites and one in males) and in the cell types comprising it. Gonadal development in C. elegans is well studied, but regulation of sexual differentiation, especially later in gonadal development, remains poorly elucidated. To identify genes involved in this process, we performed a genome-wide RNAi screen using sex-specifically expressed gonadal GFP reporters. This screen identified several phenotypic classes, including ∼70 genes whose depletion feminized male gonadal cells. Among the genes required for male cell fate specification are Wnt/β-catenin pathway members, cell cycle regulators, and genes required for mitotic spindle function and cytokinesis. We find that a Wnt/β-catenin pathway independent of extracellular Wnt ligand is essential for asymmetric cell divisions and male differentiation during gonadal development in larvae. We also find that the cell cycle regulators cdk-1 and cyb-3 and the spindle/cytokinesis regulator zen-4 are required for Wnt/β-catenin pathway activity in the developing gonad. After sex is determined in the gonadal primordium the global sex determination pathway is dispensable for gonadal sexual fate, suggesting that male cell fates are promoted and maintained independently of the global pathway during this period.THE Caenorhabditis elegans gonad derives from a simple primordium of four cells that coalesces during embryogenesis and contains two somatic gonad precursors (SGPs), Z1 and Z4, flanking two germline precursors, Z2 and Z3 (Kimble and Hirsh 1979). The SGPs undergo very different developmental programs in each sex, involving sexually dimorphic cell lineages and migrations and sex-specific cellular differentiation. The result is a two-armed bilaterally symmetrical gonad in the adult hermaphrodite or a single-armed asymmetric gonad in the adult male. The high degree of sexual dimorphism of the mature organ and variety of cellular events that occur sex specifically during its development make the C. elegans gonad an outstanding model for sex-specific organogenesis.Development of the somatic gonad occurs in two phases. The early phase defines the gonadal axes and establishes the precursors of the major gonadal cell types. This takes place during the first larval stage (L1), beginning shortly after hatching with the first division of the SGPs. In both sexes SGP division is asymmetric in terms of both the sizes and the fates of the daughter cells, and establishes the proximal/distal axis of the gonad (Hirsh et al. 1976; Kimble and Hirsh 1979). The global sex determination pathway establishes the future sex of the gonad around the time of hatching (Klass et al. 1976; Nelson et al. 1978), and sexual dimorphism is already apparent when the SGPs divide: the size asymmetry of the SGP daughters is much more pronounced in males than hermaphrodites. In both sexes the asymmetry of the first SGP division requires a Wnt/β-catenin pathway. Mutations compromising this pathway cause a “symmetrical sisters” phenotype in which both daughters adopt the same fate (Miskowski et al. 2001; Siegfried and Kimble 2002; Phillips and Kimble 2009). Sex specificity is imposed on the SGPs by the global sex determining gene tra-1 (Hodgkin 1987) and the gonad-specific sex determining gene fkh-6 (Chang et al. 2004). These genes play opposing roles in SGP sex determination, with tra-1 feminizing and fkh-6 masculinizing the somatic gonad, and they also act redundantly to promote mitotic proliferation of the SGP lineage (Chang et al. 2004). SGP sex determination is linked to cell cycle progression by cyclin D, which is required to overcome repression of fkh-6 expression in the SGPs by E2F (Tilmann and Kimble 2005).The later phase of gonadal development involves the elongation of the gonad, together with cellular proliferation and differentiation, and lasts from L2 to adulthood. During L2 the somatic cells enlarge and leader cells (distal tip cells in the hermaphrodite, linker cell in the male) begin long-range migrations that extend the gonad. During L3, somatic gonad cell division resumes in both sexes, leading to the formation of differentiated somatic cell types by the end of L3 or beginning of L4. Gonadal morphogenesis is completed and gametogenesis begins during L4 (Kimble and Hirsh 1979).Although SGP division and much of hermaphrodite gonadal development have been well studied (Hubbard and Greenstein 2000), sexual cell fate specification in the somatic gonad is more poorly understood, particularly after the L1 stage. Despite the importance of fkh-6 in promoting male differentiation, it is expressed in males only during early L1 and null mutants have incomplete gonadal sex reversal. We have therefore performed a genome-wide RNAi screen to identify additional genes required after hatching for gonadal development in each sex. Among the advantages of this approach is the ability to identify gonadal regulators that also are essential for embryonic development. To our knowledge this is the first functional genomic study of gonadal sex differentiation.The screen identified many genes whose depletion disrupts gonadogenesis in each sex and nearly 70 genes whose depletion causes gonadal feminization in males. Prominent among this latter class were components of a Wnt/β-catenin pathway, cell cycle regulators, and genes involved in mitotic spindle function and cytokinesis. We find that Wnt/β-catenin activity continues in both sexes after SGP division and is required for male cell fate commitment in the gonad. We also find that the cyclin-dependent kinase cdk-1 and its cognate cyclin cyb-3 as well as the mitotic spindle regulator zen-4 are required for gonadal Wnt/β-catenin pathway activity, providing a potential new link between the cell cycle, asymmetric division, and sexual differentiation. The feminization caused by depletion of Wnt/β-catenin pathway components or cdk-1 is independent of the global sex determination pathway, suggesting that sexual fates in the male gonad remain plastic after the primary sex determination decision.     

RIDDLE: reflective diffusion and local extension reveal functional associations for unannotated gene sets via proximity in a gene network

The growing availability of large-scale functional networks has promoted the development of many successful techniques for predicting functions of genes. Here we extend these network-based principles and techniques to functionally characterize whole sets of genes. We present RIDDLE (Reflective Diffusion and Local Extension), which uses well developed guilt-by-association principles upon a human gene network to identify associations of gene sets. RIDDLE is particularly adept at characterizing sets with no annotations, a major challenge where most traditional set analyses fail. Notably, RIDDLE found microRNA-450a to be strongly implicated in ocular diseases and development. A web application is available at http://www.functionalnet.org/RIDDLE.     

A Genome-Wide RNAi Screen for Enhancers of par Mutants Reveals New Contributors to Early Embryonic Polarity in Caenorhabditis elegans

The par genes of Caenorhabditis elegans are essential for establishment and maintenance of early embryo polarity and their homologs in other organisms are crucial polarity regulators in diverse cell types. Forward genetic screens and simple RNAi depletion screens have identified additional conserved regulators of polarity in C. elegans; genes with redundant functions, however, will be missed by these approaches. To identify such genes, we have performed a genome-wide RNAi screen for enhancers of lethality in conditional par-1 and par-4 mutants. We have identified 18 genes for which depletion is synthetically lethal with par-1 or par-4, or both, but produces little embryo lethality in wild type. Fifteen of the 18 genes identified in our screen are not previously known to function in C. elegans embryo polarity and 11 of them also increase lethality in a par-2 mutant. Among the strongest synthetic lethal genes, polarity defects are more apparent in par-2 early embryos than in par-1 or par-4, except for strd-1(RNAi), which enhances early polarity phenotypes in all three mutants. One strong enhancer of par-1 and par-2 lethality, F25B5.2, corresponds to nop-1, a regulator of actomyosin contractility for which the molecular identity was previously unknown. Other putative polarity enhancers identified in our screen encode cytoskeletal and membrane proteins, kinases, chaperones, and sumoylation and deubiquitylation proteins. Further studies of these genes should give mechanistic insight into pathways regulating establishment and maintenance of cell polarity.     

Targeted Chromosomal Translocations and Essential Gene Knockout Using CRISPR/Cas9 Technology in Caenorhabditis elegans

Many genes play essential roles in development and fertility; their disruption leads to growth arrest or sterility. Genetic balancers have been widely used to study essential genes in many organisms. However, it is technically challenging and laborious to generate and maintain the loss-of-function mutations of essential genes. The CRISPR/Cas9 technology has been successfully applied for gene editing and chromosome engineering. Here, we have developed a method to induce chromosomal translocations and produce genetic balancers using the CRISPR/Cas9 technology and have applied this approach to edit essential genes in Caenorhabditis elegans. The co-injection of dual small guide RNA targeting genes on different chromosomes resulted in reciprocal translocation between nonhomologous chromosomes. These animals with chromosomal translocations were subsequently crossed with animals that contain normal sets of chromosomes. The F1 progeny were subjected to a second round of Cas9-mediated gene editing. Through this method, we successfully produced nematode strains with specified chromosomal translocations and generated a number of loss-of-function alleles of two essential genes (csr-1 and mes-6). Therefore, our method provides an easy and efficient approach to generate and maintain loss-of-function alleles of essential genes with detailed genetic background information.