Methods and strategies for gene structure curation in WormBase.

The Caenorhabditis elegans genome sequence was published over a decade ago; this was the first published genome of a multi-cellular organism and now the WormBase project has had a decade of experience in curating this genome's sequence and gene structures. In one of its roles as a central repository for nematode biology, WormBase continues to refine the gene structure annotations using sequence similarity and other computational methods, as well as information from the literature- and community-submitted annotations. We describe the various methods of gene structure curation that have been tried by WormBase and the problems associated with each of them. We also describe the current strategy for gene structure curation, and introduce the WormBase 'curation tool', which integrates different data sources in order to identify new and correct gene structures. Database URL: http://www.wormbase.org/.

Complete genome sequence of Salmonella enterica serovar Typhimurium LT2.

Salmonella enterica subspecies I, serovar Typhimurium (S. typhimurium), is a leading cause of human gastroenteritis, and is used as a mouse model of human typhoid fever. The incidence of non-typhoid salmonellosis is increasing worldwide, causing millions of infections and many deaths in the human population each year. Here we sequenced the 4,857-kilobase (kb) chromosome and 94-kb virulence plasmid of S. typhimurium strain LT2. The distribution of close homologues of S. typhimurium LT2 genes in eight related enterobacteria was determined using previously completed genomes of three related bacteria, sample sequencing of both S. enterica serovar Paratyphi A (S. paratyphi A) and Klebsiella pneumoniae, and hybridization of three unsequenced genomes to a microarray of S. typhimurium LT2 genes. Lateral transfer of genes is frequent, with 11% of the S. typhimurium LT2 genes missing from S. enterica serovar Typhi (S. typhi), and 29% missing from Escherichia coli K12. The 352 gene homologues of S. typhimurium LT2 confined to subspecies I of S. enterica-containing most mammalian and bird pathogens-are useful for studies of epidemiology, host specificity and pathogenesis. Most of these homologues were previously unknown, and 50 may be exported to the periplasm or outer membrane, rendering them accessible as therapeutic or vaccine targets.

Intercistronic region required for polycistronic pre-mRNA processing in Caenorhabditis elegans.

In Caenorhabditis elegans, polycistronic pre-mRNAs are processed by cleavage and polyadenylation at the 3' ends of the upstream genes and trans splicing, generally to the specialized spliced leader SL2, at the 5' ends of the downstream genes. Previous studies have indicated a relationship between these two events in the processing of a heat shock-induced gpd-2-gpd-3 polycistronic pre-mRNA. Here, we report mutational analysis of the intercistronic region of this operon by linker scan analysis. Surprisingly, no sequences downstream of the 3' end were important for 3'-end formation. In contrast, a U-rich (Ur) element located 29 bp downstream of the site of 3'-end formation was shown to be important for downstream mRNA biosynthesis. This approximately 20-bp element is sufficient for SL2 trans splicing and mRNA accumulation when transplanted to a heterologous context. Furthermore, when the downstream gene was replaced by a gene from another organism, no loss of trans-splicing specificity was observed, suggesting that the Ur element may be the primary signal required for downstream mRNA processing.

WormBase: network access to the genome and biology of Caenorhabditis elegans.

WormBase (http://www.wormbase.org) is a web-based resource for the Caenorhabditis elegans genome and its biology. It builds upon the existing ACeDB database of the C.elegans genome by providing data curation services, a significantly expanded range of subject areas and a user-friendly front end.

Sequence and analysis of chromosome 5 of the plant Arabidopsis thaliana.

The genome of the model plant Arabidopsis thaliana has been sequenced by an international collaboration, The Arabidopsis Genome Initiative. Here we report the complete sequence of chromosome 5. This chromosome is 26 megabases long; it is the second largest Arabidopsis chromosome and represents 21% of the sequenced regions of the genome. The sequence of chromosomes 2 and 4 have been reported previously and that of chromosomes 1 and 3, together with an analysis of the complete genome sequence, are reported in this issue. Analysis of the sequence of chromosome 5 yields further insights into centromere structure and the sequence determinants of heterochromatin condensation. The 5,874 genes encoded on chromosome 5 reveal several new functions in plants, and the patterns of gene organization provide insights into the mechanisms and extent of genome evolution in plants.

Conservation and novelty in the evolution of cell adhesion and extracellular matrix genes.

New proteins and modules have been invented throughout evolution. Gene "birth dates" in Caenorhabditis elegans range from the origins of cellular life through adaptation to a soil habitat. Possibly half are "metazoan" genes, having arisen sometime between the yeast-metazoan and nematode-chordate separations. These include basement membrane and cell adhesion molecules implicated in tissue organization. By contrast, epithelial surfaces facing the environment have specialized components invented within the nematode lineage. Moreover, interstitial matrices were likely elaborated within the vertebrate lineage. A strategy for concerted evolution of new gene families, as well as conservation of adaptive genes, may underlie the differences between heterochromatin and euchromatin.

Relationship between 3' end formation and SL2-specific trans-splicing in polycistronic Caenorhabditis elegans pre-mRNA processing.

About 25% of the genes in the nematode Caenorhabditis elegans are in operons, polycistronic transcription units in which the genes are only 100-400 bp apart. The operon pre-mRNAs are processed into monocistronic mRNAs by a combination of cleavage and polyadenylation at the 3' end of the upstream mRNA and SL2 trans-splicing at the 5' end of the downstream mRNA. To determine whether 3' end formation and SL2 trans-splicing are coupled mechanistically, we tested a gpd-2/gpd-3 operon construct driven by a C. elegans heat shock promoter, and measured the effects of inhibition of 3' end formation and/or trans-splicing on the processing of the polycistronic RNA in vivo. The results indicate that proper 3' end formation of the upstream mRNA in an operon is required for SL2-specificity of downstream mRNA trans-splicing. In contrast, trans-splicing of the downstream mRNA is not necessary for correct 3' end formation of the upstream mRNA. In addition, shortening the distance between the 5' cap and the AAUAAA of gpd-2 (the upstream gene) decreases the efficiency of 3' end formation and is accompanied by a replacement of SL2 with SL1 at the trans-splice site of gpd-3, the downstream gene. These results indicate that SL2 trans-splicing, in C. elegans, is coupled mechanistically to 3' end formation in the processing of polycistronic pre-mRNAs.

Gene structure and organization in Caenorhabditis elegans.

The sequencing of the 100 Mb Caenorhabditis elegans genome-containing approximately 14,000 genes-is approximately 50% complete. One of its most interesting features is its compactness; introns and intergenic distances are unusually small and, surprisingly, approximately 25% of genes are contained in polycistronic transcription units (operons) with only approximately 100 bp between genes.

Operons as a common form of chromosomal organization in C. elegans.

Although eukaryotic genes are usually transcribed individually, at least a few Caenorhabditis elegans genes appear to be transcribed polycistronically in clusters resembling bacterial operons. The spliced leader SL2 (ref. 2) is specific for trans-splicing to downstream genes in these operons. In addition, many C. elegans pre-mRNAs are trans-spliced to SL1 (ref. 3) near the 5' ends of pre-mRNAs. Because operons have not previously been found in higher eukaryotes, we have investigated how widespread they are in the C. elegans genome. We identified gene clusters using the extensive data generated by the genome project and tested seven for trans-splicing specificity. All were found to fit expectations for polycistronic transcription. In addition, we surveyed reported C. elegans genes for trans-splicing specificity. Both methods indicate that the pre-mRNAs of about 70% of C. elegans genes are trans-spliced and as many as a quarter are transcribed in operons.

Analysis of the VPE sequences in the Caenorhabditis elegans vit-2 promoter with extrachromosomal tandem array-containing transgenic strains.

The Caenorhabditis elegans vit genes, encoding vitellogenins, are abundantly expressed in the adult hermaphrodite intestine. Two repeated elements, vit promoter element 1 (VPE1 [TGTCAAT]) and VPE2 (CTGATAA), have been identified in the 5' flanking DNA of each of the vit genes of C. elegans and Caenorhabditis briggsae. These elements have previously been shown to be needed for correctly regulated expression of a vit-2/vit-6 fusion gene in low-copy-number, integrated transgenes. Here we extend the analysis of the function of VPE1 and VPE2 by using transgenic lines carrying large, extrachromosomal arrays of the test genes. The results validate the use of such arrays for transgenic analysis of gene regulation in C. elegans, by confirming previous findings showing that the VPE1 at -45 and both VPE2s are sites of activation. Additional experiments now indicate that when the -45 VPE1 is inverted or replaced by a VPE2, nearly total loss of promoter function results, suggesting that the highly conserved -45 VPE1 plays a unique role in vit-2 promoter function. In contrast, single mutations eliminating the three upstream VPE1s are without effect. However, in combination in double and triple mutants, these upstream VPE1 mutations cause drastic reductions in expression levels. The -150 VPE2 can be replaced by a XhoI site (CTCGAG), and the -90 VPE2 can be eliminated, as long as the overlapping VPE1 is left intact, but when these two replacements are combined, activity is lost. Thus, the promoter must have at least one VPE2 and it must have at least two VPE1s, one at -45 and one additional upstream element.

Operons in C. elegans: polycistronic mRNA precursors are processed by trans-splicing of SL2 to downstream coding regions.

The mRNAs of six C. elegans genes are known to be trans-spliced to SL2. We report here that a similarly oriented gene is located 100-300 bp upstream of each. We present evidence that the genes in these clusters are cotranscribed and downstream mRNAs are formed by cleavage at the polyadenylation site and trans-splicing. From one three-gene cluster we isolated cDNA clones representing both polycistronic RNAs and mRNAs polyadenylated at the free 3' end created by trans-splicing, suggesting that polycistronic RNAs can be processed by trans-splicing. Several experiments indicate that SL2 trans-splicing is a consequence of a gene's downstream location in an operon. In particular, when an SL1-accepting gene was moved to a downstream location, its mRNA was trans-spliced largely to SL2. The possible regulatory significance of cotranscription of C. elegans genes is discussed.

Regulation of vitellogenin gene expression in transgenic Caenorhabditis elegans: short sequences required for activation of the vit-2 promoter.

The Caenorhabditis elegans vitellogenin genes are subject to sex-, stage-, and tissue-specific regulation: they are expressed solely in the adult hermaphrodite intestine. Comparative sequence analysis of the DNA immediately upstream of these genes revealed the presence of two repeated heptameric elements, vit promoter element 1 (VPE1) and VPE2. VPE1 has the consensus sequence TGTCAAT, while VPE2, CTGATAA, shares the recognition sequence of the GATA family of transcription factors. We report here a functional analysis of the VPEs within the 5'-flanking region of the vit-2 gene using stable transgenic lines. The 247 upstream bp containing the VPEs was sufficient for high-level, regulated expression. Furthermore, none of the four deletion mutations or eight point mutations tested resulted in expression of the reporter gene in larvae, males, or inappropriate hermaphrodite tissues. Mutation of the VPE1 closest to the TATA box inactivated the promoter, in spite of the fact that four additional close matches to the VPE1 consensus sequence are present within the 5'-flanking 200 bp. Each of these upstream VPE1-like sequences could be mutated without loss of high-level transgene expression, suggesting that if these VPE1 sequences play a role in regulating vit-2, their effects are more subtle. A site-directed mutation in the overlapping VPE1 and VPE2 at -98 was sufficient to inactivate the promoter, indicating that one or both of these VPEs must be present for activation of vit-2 transcription. Similarly, a small perturbation of the VPE2 at -150 resulted in reduction of fp155 expression, while a more extensive mutation in this element eliminated expression. On the other hand, deletion of this VPE2 and all upstream DNA still permitted correctly regulated expression, although at a very low level, suggesting that this VPE2 performs an important role in activation of vit-2 expression but may not be absolutely required. The results, taken together, demonstrate that both VPE1 and VPE2 are sites for activation of the vit-2 promoter.

elt-1, an embryonically expressed Caenorhabditis elegans gene homologous to the GATA transcription factor family.

The short, asymmetrical DNA sequence to which the vertebrate GATA family of transcription factors binds is present in some Caenorhabditis elegans gene regulatory regions: it is required for activation of the vitellogenin genes and is also found just 5' of the TATA boxes of tra-2 and the msp genes. In vertebrates GATA-1 is specific to erythroid lineages, whereas GATA-2 and GATA-3 are present in multiple tissues. In an effort to identify the trans-acting factors that may recognize this sequence element in C. elegans, we used a degenerate oligonucleotide to clone a C. elegans homolog to this gene. We call this gene elt-1 (erythrocytelike transcription factor). It is single copy and specifies a 1.75-kb mRNA that is present predominantly, if not exclusively, in embryos. The region of elt-1 encoding two zinc fingers is remarkably similar to the DNA-binding domain of the vertebrate GATA-binding proteins. However, outside of the DNA-binding domains the amino acid sequences are quite divergent. Nevertheless, introns are located at identical or nearly identical positions in elt-1 and the mouse GATA-1 gene. In addition, elt-1 mRNA is trans-spliced to the 22-base untranslated leader, SL1. The DNA upstream of the elt-1 TATA box contains eight copies of the GATA recognition sequence within the first 300 bp, suggesting that elt-1 may be autogenously regulated. Our results suggest that the specialized role of GATA-1 in erythroid gene expression was derived after separation of the nematodes and the line that led to the vertebrates, since C. elegans lacks an erythroid lineage.

Vitellogenin motifs conserved in nematodes and vertebrates.

Caenorhabditis elegans vitellogenins are encoded by a family of six genes, one of which, vit-5, has been previously sequenced and shown to be surprisingly closely related to the vertebrate vitellogenin genes. Here we report an alignment of the amino acid sequences of vitellogenins from frog and chicken with those from three C. elegans genes: vit-5 and two newly sequenced genes, vit-2 and vit-6. The four introns of vit-6 are all in different places from the four introns of vit-5, but three of these eight positions are identical or close to intron locations in the vertebrate vitellogenin genes. The encoded polypeptides have diverged from one another sufficiently to allow us to draw some conclusions about conserved positions. Many cysteine residues have been conserved, suggesting that vitellogenin structure has been maintained over a long evolutionary distance and is dependent upon disulfide bonds. In addition, a 20-residue segment shows conservation between the vertebrate and the nematode vitellogenins. This sequence may play a highly conserved role in vitellogenesis, such as specific recognition by oocytes. On the whole, however, selection may be acting more strongly on amino acid composition and codon usage than on amino acid sequence, as might be expected for abundant storage proteins: The amino acid compositions of vit-2, vit-5, and vit-6 products are remarkably similar, despite the fact that the sequence of the vit-2 protein is only 22% and 50% identical to the sequences of vit-6 and vit-5 proteins, respectively.

Insertion of part of an intron into the 5' untranslated region of a Caenorhabditis elegans gene converts it into a trans-spliced gene.

In nematodes, the RNA products of some genes are trans-spliced to a 22-nucleotide spliced leader (SL), while the RNA products of other genes are not. In Caenorhabditis elegans, there are two SLs, SL1 and SL2, donated by two distinct small nuclear ribonucleoprotein particles in a process functionally quite similar to nuclear intron removal. We demonstrate here that it is possible to convert a non-trans-spliced gene into a trans-spliced gene by placement of an intron missing only the 5' splice site into the 5' untranslated region. Stable transgenic strains were isolated expressing a gene in which 69 nucleotides of a vit-5 intron, including the 3' splice site, were inserted into the 5' untranslated region of a vit-2/vit-6 fusion gene. The RNA product of this gene was examined by primer extension and PCR amplification. Although the vit-2/vit-6 transgene product is not normally trans-spliced, the majority of transcripts from this altered gene were trans-spliced to SL1. We termed the region of a trans-spliced mRNA precursor between the 5' end and the first 3' splice site an "outron." Our results suggest that if a transcript begins with intronlike sequence followed by a 3' splice site, this alone may constitute an outron and be sufficient to demarcate a transcript as a trans-splice acceptor. These findings leave open the possibility that specific sequences are required to increase the efficiency of trans-splicing.

Regulated expression of a vitellogenin fusion gene in transgenic nematodes.

In Caenorhabditis elegans the vitellogenin genes are expressed abundantly in the adult hermaphrodite intestine, but are otherwise silent. In order to begin to understand the mechanisms by which this developmental regulation occurs, we used the transformation procedure developed for C. elegans by A. Fire (EMBO. J., 1986, 5, 2673-2680) to obtain regulated expression of an introduced vitellogenin fusion gene. A plasmid with vit-2 upstream and coding sequences fused to coding and downstream sequences of vit-6 was injected into oocytes and stable transgenic strains were selected. We obtained seven independent strains, in which the plasmid DNA is integrated at a low copy number. All strains synthesize substantial amounts of a novel vitellogenin-like polypeptide of 155 kDa that accumulates in the intestine and pseudocoelom, but is not transported efficiently into oocytes. In two strains examined in detail the fusion gene is expressed with correct sex, tissue, and stage specificity. Thus we have demonstrated that the nematode transgenic system can give proper developmental expression of introduced genes and so can be used to identify DNA regulatory regions.

The nucleotide sequence of a nematode vitellogenin gene.

The nematode, Caenorhabditis elegans, contains a family of six genes that code for vitellogenins. Here we report the complete nucleotide sequence of one of these genes, vit-5. The gene specifies a mRNA of 4869 nucleotides, including untranslated regions of 9 bases at the 5' end and 51 bases at the 3' end. Vit-5 contains four short introns totalling 218 bp. The predicted vitellogenin, yp170A, has a molecular weight of 186,430. At its N terminus it is clearly related to the vitellogenins of vertebrates. However, the vit-5-encoded protein does not contain a serine-rich sequence related to the vertebrate vitellin, phosvitin. In fact, the amino acid composition of the nematode protein is very similar to that of the vertebrate protein without phosvitin. Vit-5 has a highly asymmetric codon choice dictionary. The favored codons are different from those favored in other organisms, but are characteristic of highly expressed C. elegans genes. The strong selection against rare codons is not as great near the 5' end of the gene; rare codons are 15 times more frequent within the first 54 bp than in the next 4.8 kb.

The Caenorhabditis elegans vitellogenin gene family includes a gene encoding a distantly related protein.

While the nematode Caenorhabditis elegans is more primitive than most egg-laying organisms, it's vitellogenins, or yolk protein precursors, appear to be more complex. C. elegans oocytes accumulate two major classes of yolk proteins. The first consists of two polypeptides with an Mr of about 170,000 (yp170A and yp170B) encoded by a family of five closely related genes called vit-1 through vit-5. The second class consists of two smaller proteins with Mr values of 115,000 (yp115) and 88,000 (yp88) which are cut from a single precursor. Here we report the cloning and analysis of a single-copy gene (vit-6) that encodes this precursor. The lengths of the gene and its mRNA are about 5 X 10(3) base pairs. Like vit-1 through vit-5, vit-6 is expressed exclusively in adult hermaphrodites. Comparison of portions of the coding sequence indicates that vit-6 is distantly related to the vit-1 through vit-5 gene family. Thus, even though the two classes of yolk proteins are antigenically and physically distinct, they are encoded by a single highly diverged gene family.

The C. elegans vitellogenin genes: short sequence repeats in the promoter regions and homology to the vertebrate genes.

The nematode Caenorhabditis elegans contains a small family of vitellogenin genes which is expressed abundantly, but only in the intestine of the adult hermaphrodite worm. In order to identify possible regulatory elements, we have sequenced the DNA surrounding the 5' ends of five of the six genes. Contained within regions which have largely diverged from one another, two different heptameric sequences are found repeated within the first 200 bp upstream of each of the genes. The first sequence, TGTCAAT, is present as a perfect heptamer at least once upstream of each gene. It is repeated in both orientations four to six times in each 5' flanking region, allowing a one-base mismatch. The second sequence, CTGATAA, is also present as a perfect heptamer in a restricted region upstream of each gene. These two sequence elements may be involved in regulation of the vitellogenin genes. Remarkably, the CTGATAA sequence is present in a similar location in the promoter regions of vertebrate vitellogenin genes. In fact, our data reveal a surprising degree of similarity between the nematode and vertebrate vitellogenins.

Cloning of a yolk protein gene family from Caenorhabditis elegans.

We have cloned a family of five genes which encode the 170,000 Mr yolk proteins in the nematode Caenorhabditis elegans. The genes and their messenger RNAs are about 5 X 10(3) base-pairs in length. Thus most of the length of each gene is exon, although a few small introns have been discovered. Based on hybridization and restriction mapping experiments, the genes can be subdivided into two subfamilies: YP1-YP2 and YP3-YP4-YP5. Within a subfamily the genes are nearly identical. While most of the genes are not clustered, YP3 and YP4 are tandemly linked. Hybrid-arrest translation experiments demonstrate that the YP3-YP4-YP5 subfamily encodes the yp170A yolk protein, while the YP1-YP2 subfamily encodes the yp170B yolk protein. RNAs homologous to these genes are abundant in the adult hermaphrodite, but missing from larvae and males. Furthermore, RNA isolated from dissected intestines is highly enriched for sequences that hybridize to the genes, whereas RNA from gonad or body wall is nearly devoid of these sequences. Thus, this gene family is apparently expressed only in the intestine of the adult hermaphrodite.