The C. briggsae genome was annotated using various gene-finding programs (such as Genefinder 14) and by comparison with C. elegans. As different programs often disagree with each other in predicting genes, Stein et al. 3 adopted a 'hybrid' approach by combining the predictions made by multiple gene-finding programs and selecting the consensus. In cases in which a consensus could not be obtained, the authors chose the predictions with best overall similarity with the C. elegans genome. This analysis identified 19,507 genes in C. briggsae. In general, the 'hybrid' approach was twice as accurate as any single gene-prediction program. Conversely, the C. briggsae genome has been extremely useful in the annotation of C. elegans genes: 1,275 new genes were predicted in the C. elegans genome (6% of the new total of 20,621 genes) on the basis of the C. briggsae-C. elegans comparison. Almost 300 of these are confirmed by the open reading frame sequence tag data of Reboul et al. 15. Gene finding is of supreme importance because essentially all C. elegans genes are being studied, for example by RNAi, and this 6% increase in the number of genes will have enormous impact on the intensive analysis of the organism. Further refinements in C. elegans gene-structure predictions are likely as many more genes are compared carefully in the two species.

Using two different approaches - best reciprocal BLASTP matches and conserved gene order (synteny) - Stein et al. 3 have defined orthologs of C. elegans genes in the C. briggsae genome (62% of predicted C. briggsae genes or 12,155; Figure 3). The identity at the protein level between C. briggsae-C. elegans orthologous pairs (mean of 75%) is comparable to mouse-human orthologs (median 78.5%) 312. Comparison of the orthologs reveals that about 11% of introns are species-specific, with C. elegans having almost twice as many unique introns (C. elegans has about 4,400 and C. briggsae about 2,200; see also 16). On average, half of the genes with orthologs differ in the presence of an intron between the two species. Just 4% (807) of the C. briggsae genes do not have significant BLASTP matches in C. elegans (Figure 3); these are likely to be highly divergent and novel (species-specific) genes.

The C. briggsae genome contains 5,211 genes that have multiple matches in the C. elegans genome (Figure 3) and that correspond to various gene families. Although in most cases such families contain comparable numbers of proteins in the two species (for example, there are 376 protein kinases in C. elegans and 399 in C. briggsae), the olfactory-type chemosensory receptors and the cyclin-like F-box proteins have significantly more members in C. elegans (718 and 243 members, respectively) than in C. briggsae (464 and 98, respectively) 3. For the chemoreceptors, gene expansion is likely to have occurred in the C. elegans-specific families. The functional significance of such divergence might indicate physiological or ecological differences between these species.

In addition to finding coding exons, blocks of alignable sequence were found using the WABA algorithm 16. About 1.3 million such blocks were identified, covering coding exons (32.2%), introns (35.3%) and other regions (untranslated regions, intergenic regions, and so on). These conserved regions are a rich resource for further investigation, but it would be useful to have additional genome sequences in order to help discriminate between regions that are conserved because of selection and those that are conserved because they have not yet faded away during neutral evolution. The WABA algorithm distinguishes different codon positions and thus reveals many of the presumed regions of intergenic conservation to have the signature of coding exons. The C. briggsae genome sequence has been a tool to help analyze cis-regulatory sequences for some time already. In general, the C. briggsae-C. elegans comparison has been helpful in locating functionally critical regions of non-coding DNA 1718 and may be a good filter for eliminating false positives for some types of bioinformatic searches.

By aligning clear orthologs, Stein et al. 3 were able to evaluate the syntenic relationships for much of the two genomes. Alignment of ortholog pairs reveals over 3,000 rearrangements, and these are ten times as common within a chromosome as between chromosomes. A higher-resolution view of the syntenic relationships between the genomes will require additional physical mapping, or genetic mapping, for example using single-nucleotide polymorphisms. Comparison of the genomic rearrangements between C. briggsae-C. elegans and Drosophila species reveals a breakpoint rate roughly five times higher in nematodes.

One striking feature of nematode genome organization is the existence of trans-spliced operons, in which a primary transcript is processed to give multiple protein-coding transcripts by a trans-splicing reaction with the SL2 spliced leader sequence. There are estimated to be about 1,000 operons in C. elegans (WS108 release) 19. Of the 800 well-characterized C. elegans operons (WS77 release), 32 are broken in C. briggsae, either by insertion, transposition or rearrangement.

As we have seen, despite the morphological similarity, there are significant molecular differences between the C. briggsae and C. elegans genomes. The C. briggsae genome contains about 800 genes with no apparent match in C. elegans 3. Together with the divergent genes and gene families (Figure 3), nearly one third of the genome is arguably different from C. elegans. Are these changes reflected in significant biological differences? Careful examination has revealed a number of subtle differences between the two species. For example, the excretory system in C. elegans plays a critical role in osmoregulation 20, and comprises three cells including a single duct cell 21. Although the excretory system looks morphologically identical in the two species, C. briggsae animals have a more anterior duct opening than C. elegans 22. This difference is the result of the altered expression of lin-48, which encodes a member of the Ovo family of transcription factors. Expression of lin-48 is observed in the excretory duct cell in C. elegans but not in C. briggsae 2223. This is the first example of a morphological difference between the two nematode species with a known molecular basis.

Studies of vulval development in C. briggsae have revealed that, although the overall vulval morphology is similar to that in C. elegans (Figure 1b,c), there are some differences in the underlying mechanisms. In C. elegans, six vulval precursor cells are competent to respond to a gonad-derived inductive signal and to produce the vulval progeny cells 24. For example, ablation of all vulval precursor cells but the anterior-most one, known as P3.p, allows the latter to generate vulval cells, even though it does not do so in an intact animal. By contrast, P3.p in C. briggsae is not competent in this assay 25. On the other hand, in certain multivulva mutants of C. briggsae, P3.p is induced and makes vulval progeny cells (B.P.G., unpublished observations). Thus, there is a subtle difference in the competence of the P3.p cell between the two species. Another example of the differences in vulval development comes from the study of the glp-1 gene function in three Caenorhabditis species: C. briggsae, C. elegans and C. remanei. In C. elegans, GLP-1, a receptor of the LIN-12/Notch family, mediates cell-cell communication during development 26. Loss-of-function mutations in glp-1 do not affect C. elegans vulval development, but in C. briggsae, injection of glp-1 double-stranded RNA causes a multivulva phenotype, suggesting that glp-1 inhibits vulval development in C. briggsae 27.

Nematodes are morphologically diverse and have two different modes of reproduction (hermaphroditic, such as in C. briggsae and C. elegans, and male-female, such as in C. remanei and Caenorhabditis species CB5161); they thus provide an excellent opportunity to study the molecular basis of sexual differentiation and evolution. Several of the known sex-determination genes are remarkably diverged between C. briggsae and C. elegans 28; for example, FEM-3 and TRA-2, which physically interact with each other, have coevolved and function strictly in a species-specific manner 29. RNAi targeting the fem-1 and fem-2 genes in C. briggsae have revealed functional differences from C. elegans (28 and references therein).

An immediate reward of the C. briggsae genome sequence has been to increase our understanding of C. elegans gene structure. The C. briggsae-C. elegans comparison will inspire many new projects to study gene function and evolutionary changes. Although sequence comparison has revealed a significant number of gene sequences conserved between C. briggsae and C. elegans (62% orthologs; see Figure 3), it is not clear whether all these genes are functionally conserved. Furthermore, conservation in the coding sequences might be misleading if the regulatory elements have evolved differentially (see, for example, 17). In addition, there are significant differences in some gene families (see above). An understanding of the evolutionary conservation and divergence between the two genomes will require systematic study of the function of C. briggsae genes. The recent success of large-scale RNAi screens in C. elegans raises the promise of a similar approach in C. briggsae. If there is a difference in the RNAi phenotype of a particular gene, however, it will be unclear whether it is a quantitative difference in the use of that gene or whether the gene differs in its susceptibility to RNAi. Thus, classical genetics or targeted gene knockouts are likely to be valuable. To facilitate such experiments, a classical genetic linkage map (B.P.G., R. Johnsen, T. Inoue, A. Mah, G. Jo, D. Baillie and P.W.S., unpublished results) and a single-nucleotide polymorphism map (R. Miller, S. Baird, L. Fulton and R. Waterston, personal communication) are being developed for C. briggsae. These maps will help order contigs in the genome and allow researchers to carry out genetic analysis, as well as helping clone C. briggsae genes with novel mutant phenotypes and study their biological functions.

The C. briggsae genome sequence will probably also help in understanding parasitic nematode genomes such as that of the filarial parasite Brugia malayi. Additional genomes could bridge the gap from C. briggsae and C. elegans to other species (Figure 2). Intensive analysis of transcriptional regulatory networks, in particular cis-regulatory elements, has clearly been helped by the availability of C. briggsae sequence. There is significant value for computational analysis in having additional close nematode genomes (E. Schwarz, J. DeModena, E. Moon, H. Shizuya, B. Wold and P.W.S., unpublished observations). The biological differences between the other nematodes might make the sequence comparisons slightly less informative, but it is possible to test the function of sequences by reciprocal transformation experiments 30, providing a way to know whether there is indeed conservation in gene function. We look forward to the new insights into nematode biology that more genome sequences and further analysis of the C. briggsae genome will bring.