We performed threefold shotgun sequencing of the silkworm ( Bombyx mori ) genome to obtain a draft sequence and establish a basic resource for comprehensive genome analysis. By using the newly developed RAMEN assembler, the sequence data derived from whole-genome shotgun (WGS) sequencing were assembled into 49,345 scaffolds that span a total length of 514 Mb including gaps and 387 Mb without gaps. Because the genome size of the silkworm is estimated to be 530 Mb, almost 97% of the genome has been organized in scaffolds, of which 75% has been sequenced. By carrying out a BLAST search for 50 characteristic Bombyx genes and 11,202 non-redundant expressed sequence tags (ESTs) in a Bombyx EST database against the WGS sequence data, we evaluated the validity of the sequence for elucidating the majority of silkworm genes. Analysis of the WGS data revealed that the silkworm genome contains many repetitive sequences with an average length of <500 bp. These repetitive sequences appear to have been derived from truncated transposons, which are interspersed at 2.5- to 3-kb intervals throughout the genome. This pattern suggests that silkworm may have an active mechanism that promotes removal of transposons from the genome. We also found evidence for insertions of mitochondrial DNA fragments at 9 sites. A search for Bombyx orthologs to Drosophila genes controlling sex determination in the WGS data revealed 11 Bombyx genes and suggested that the sex-determining systems differ profoundly between the two species.

▪ Abstract We review progress in applying molecular genetic and genomic technologies to studies in the domesticated silkworm, Bombyx mori, highlighting its use as a model for Lepidoptera, and in sericulture and biotechnology. Dense molecular linkage maps are being integrated with classical linkage maps for positional cloning and marker-assisted selection. Classical mutations have been identified by a candidate gene approach. Cytogenetic and sequence analyses show that the W chromosome is composed largely of nested full-length long terminal repeat retrotransposons. Z-chromosome-linked sequences show a lack of dosage compensation. The downstream sex differentiation mechanism has been studied via the silkworm homolog of doublesex. Expressed sequence tagged databases have been used to discover Lepidoptera-specific genes, provide evidence for horizontal gene transfer, and construct microarrays. Physical maps using large-fragment bacterial artificial chromosome libraries have been constructed, and whole-genome shotgun sequencing is underway. Germline transformation and transient expression systems are well established and available for functional studies, high-level protein expression, and gene silencing via RNA interference.

In 2008, the genome assembly and gene models for the domestic silkworm, Bombyx mori, were published by a Japanese and Chinese collaboration group. However, the genome assembly contains a non-negligible number of misassembled and gap regions due to the presence of many repetitive sequences within the silkworm genome. The erroneous genome assembly occasionally causes incorrect gene prediction. Here we performed hybrid assembly based on 140 × deep sequencing of long (PacBio) and short (Illumina) reads. The remaining gaps in the initial genome assembly were closed using BAC and Fosmid sequences, giving a new total length of 460.3 Mb, with 30 gap regions and an N50 comprising 16.8 Mb in scaffolds and 12.2 Mb in contigs. More RNA-seq and piRNA-seq reads were mapped on the new genome assembly compared with the previous version, indicating that the new genome assembly covers more transcribed regions, including repetitive elements. We performed gene prediction based on the new genome assembly using available mRNA and protein sequence data. The number of gene models was 16,880 with an N50 of 2154 bp. The new gene models reflected more accurate coding sequences and gene sets than old ones. The proportion of repetitive elements was also reestimated using the new genome assembly, and was calculated to be 46.8% in the silkworm genome. The new genome assembly and gene models are provided in SilkBase (http://silkbase.ab.a.u-tokyo.ac.jp).

To build a foundation for the complete genome analysis of Bombyx mori , we have constructed an EST database. Because gene expression patterns deeply depend on tissues as well as developmental stages, we analyzed many cDNA libraries prepared from various tissues and different developmental stages to cover the entire set of Bombyx genes. So far, the Bombyx EST database contains 35,000 ESTs from 36 cDNA libraries, which are grouped into ≈11,000 nonredundant ESTs with the average length of 1.25 kb. The comparison with FlyBase suggests that the present EST database, SilkBase, covers >55% of all genes of Bombyx . The fraction of library-specific ESTs in each cDNA library indicates that we have not yet reached saturation, showing the validity of our strategy for constructing an EST database to cover all genes. To tackle the coming saturation problem, we have checked two methods, subtraction and normalization, to increase coverage and decrease the number of housekeeping genes, resulting in a 5–11% increase of library-specific ESTs. The identification of a number of genes and comprehensive cloning of gene families have already emerged from the SilkBase search. Direct links of SilkBase with FlyBase and WormBase provide ready identification of candidate Lepidoptera-specific genes.

In insects, the precise timing of molting and metamorphosis is strictly guided by a principal steroid hormone, ecdysone. Among the multiple conversion steps for synthesizing ecdysone from dietary cholesterol, the conversion of 7-dehydrocholesterol to 5β-ketodiol, the so-called ‘Black Box’, is thought to be the important rate-limiting step. Although a number of genes essential for ecdysone synthesis have recently been revealed, much less is known about the genes that are crucial for functioning in the Black Box. Here we report on a novel ecdysteroidgenic gene, non-molting glossy (nm-g)/shroud (sro), which encodes a short-chain dehydrogenase/reductase. This gene was first isolated by positional cloning of the nm-g mutant of the silkworm Bombyx mori, which exhibits a low ecdysteroid titer and consequently causes a larval arrest phenotype. In the fruit fly, Drosophila melanogaster, the closest gene to nm-g is encoded by the sro locus, one of the Halloween mutant members that are characterized by embryonic ecdysone deficiency. The lethality of the sro mutant is rescued by the overexpression of either sro or nm-g genes, indicating that these two genes are orthologous. Both the nm-g and the sro genes are predominantly expressed in tissues producing ecdysone, such as the prothoracic glands and the ovaries. Furthermore, the phenotypes caused by the loss of function of these genes are restored by the application of ecdysteroids and their precursor 5β-ketodiol, but not by cholesterol or 7-dehydrocholesterol. Altogether, we conclude that the Nm-g/Sro family protein is an essential enzyme for ecdysteroidogenesis working in the Black Box.