Review Identi®cation and analysis of the myosin superfamily in Drosophila: a database approach R. A. YAMASHITA, J. R. SELLERS* and J. B. ANDERSON Laboratory of Molecular Cardiology, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, MD, USA Received 27 July 2000; accepted in revised form 16 August 2000 Abstract The recent sequencing of the genome of Drosophila melanogaster has provided a valuable resource for mining the database for genes of interest. We took advantage of this opportunity in an attempt to identify novel myosins in Drosophila and con®rm the presence of the previously identi®ed myosins from classes I, II, III, V, VI, and VII. The Drosophila database annotators predicted the structure of three additional proteins which we identi®ed as novel unconventional myosins, two of which fell into classes XV and XVIII, respectively. Our own eorts predicted the presence of four additional partial sequences that appear to be myosin proteins which did not fall into any speci®c class. In the future comparative genomics will hopefully lead to the placement of these myosins into new classes. Introduction Myosins constitute a diverse superfamily of actin-based motors which are typically composed of three major domains (Mermall et al., 1998; Sellers, 1999). The motor domain which is usually found at the amino terminus is the most conserved region of myosins. It contains the actin binding sites, the nucleotide binding site and many highly conserved amino acids within its core. Surface loops are responsible for most of the diversity in motor domains among the many dierent myosins. The neck region, following the motor domain, is a helical segment of the myosin heavy chain which binds calmodulin or calmodulin-like light chains. The target sequence for light chain binding is termed the IQ motif based on its consensus sequence of IQXXXRGXXXR (Bement and Mooseker, 1995). Myosins may have zero to six IQ motifs in the neck region. It is likely that the neck region acts as a lever arm to amplify small movements in the motor domain to create the power stroke (Howard, 1997). In some myosins, the neck region is also involved in regulation of the enzymatic activity of the molecule (Sellers, 1999). Together, the motor and neck domains are sometimes referred to as the head. The third domain of myosin is the tail which likely serves to target the myosin to dierent cellular localization, to anchor the myosin so that it may move actin ®laments within the cell or to bind cargo that is destined to be moved on actin ®lament tracks (Mermall et al., 1998; Sellers, 1999). The tail domains are the most diverse portions of myosin molecules and may contain several dierent functional motifs. Among these are coiled-coil forming sequences that allow for homodimerization to form two headed molecules, SH3 domains, FERM domains, PH domains, GAP domains, and MyTH4 domains. The most recent reviews of the myosin superfamily describe 15 classes along with several outliers which do not fall into any of the classes (Mermall et al., 1998; Soldati et al., 1999; Sellers, 2000). The Myosin Home- Page Web Site (http://www.mrc-lmb.cam.ac.uk/myosin/ myosin.html) has expanded this number to 17 with the inclusion of several very recently published myosins. Phylogenetic analyses of the motor domain usually give the same clustering of molecules into classes as does analyses of the whole molecule or of just the tail domain (Soldati et al., 1999). This is due to the fact that the tail domains within a class often have the same functional motifs arrayed in a similar order, suggesting conserva- tion of function. Dierent organisms contain only a subset of the myosin classes. For example, the completed genomic sequence of Saccharomyces cerevisiae reveals only ®ve myosin genes, two of class I, one of class II and two of class V (Brown, 1997). In contrast, more complex genomes show more diversity. The Caenorabditis elegans genome contains 17 myosin genes including three of class I, nine of class II, and one each of classes V, VI, VII, IX and XII. To date, more than 20 myosin genes have been uncovered in humans. Prior to the publication of the Drosophila genome, nine Drosophila myosins were described representing the following classes: two myosin Is (IA and IB), two myosin IIs (mhc and zip), one myosin III (ninaC), one myosin V (didum), one myosin VI (jar or 95F) and two myosin VIIs (ck and VIIb) (Table 1). The genome project annotators described an *To whom correspondence should be addressed: Fax: +1-301-402- 1542; E-mail: jsellers@helix.nih.gov Journal of Muscle Research and Cell Motility 21: 491±505, 2000. 491 Ó 2000 Kluwer Academic Publishers. Printed in the Netherlands.