Olfactory Receptor Database: a Sensory Chemoreceptor Resource Emmanouil Skoufos1,2,*, Luis Marenco1, Prakash M

Olfactory Receptor Database: a sensory chemoreceptor resource Emmanouil Skoufos1,2,*, Luis Marenco1, Prakash M. Nadkarni1, Perry L. Miller1 and Gordon M. Shepherd2

1Center for Medical Informatics and 2Section of Neurobiology, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06511, USA

Received September 29, 1999; Revised and Accepted October 8, 1999

ABSTRACT and tissue, share high sequence homology and have common sequence motifs that may be involved in similar signal trans- The Olfactory Receptor Database (ORDB) is a WWW- duction pathways (7). Putative olfactory receptors, which like accessible database that has been expanded from an the ORLs are GPCRs but have no sequence homology to the olfactory receptor resource to a chemoreceptor vertebrate ORLs, have been identified in Caenorhabditis resource. It stores data on six classes of G-protein- elegans (8,9) and Drosophila (10,11). In addition, G-protein coupled sensory chemoreceptors: (i) olfactory coupled pheromone receptors have been cloned from a variety receptor-like proteins, (ii) vomeronasal receptors, of fungi (12), and vomeronasal G-protein coupled putative (iii) insect olfactory receptors, (iv) worm chemo- pheromone receptors have been identified in vertebrates (13,14). receptors, (v) taste papilla receptors and (vi) fungal When ORDB was created in 1995, the original aim was to pheromone receptors. A complementary database of aid the cloning, sequencing and classification of the ORLs. the ligands of these receptors (OdorDB) has been The identification of the additional olfactory-related chemo- constructed and is publicly available in a pilot mode. receptor gene families and the increasing interest in identifying specific ligand–receptor interactions resulted in the broadening The database schema of ORDB has been changed of the scope of ORDB to include information about all GPCR from traditional relational to EAV/CR (Entity- sensory chemoreceptor proteins. ORDB is a member of the Attribute-Value with Classes and Relationships), consortium of the GPCR databases that includes the GPCRDB which allows the interoperability of ORDB with other (15), the GRAP mutant database (16), the mutation analysis of related databases as well as the creation of intra- GPCRs database (17) and the GCRDb (18). ORDB contains entry- database associations among objects. This inter- level links with the other databases of the consortium when feasible. operability facilitates users to follow information from odor molecule binding to its putative receptor, OVERVIEW OF THE DATABASE to the properties of the neuron expressing the receptor, to a computational model of activity of Version 4.0 of ORDB contains 812 sensory chemoreceptor olfactory bulb neurons. In addition, tools and entries from 27 species that represent the sequencing efforts of 77 laboratories worldwide. The different classes of GPCR resources have been added allowing users to access sensory chemoreceptors in the database are: (i) olfactory interactive phylogenetic trees and alignments of receptor-like proteins (ORLs), (ii) C.elegans chemoreceptors sensory chemoreceptors. ORDB is available via the (CCRs), (iii) vomeronasal receptors (VNRs), (iv) insect olfactory WWW at http://ycmi.med.yale.edu/senselab/ordb/ receptors (IORs), (v) fungal pheromone receptors (FPRs) and (vi) taste papilla receptors (TPRs). ORDB contains information about the tissue from which a receptor is cloned, the size INTRODUCTION (partial or full-length) of clones, chromosomal information, The Olfactory Receptor Database (ORDB) (1) is a database and direct links to the PubMed and/or GenBank records of the containing properties and sequences of the olfactory receptor- ORs. A recently introduced resource area contains original like proteins (ORLs), vertebrate G-protein coupled receptors alignments and phylogenetic trees of all the receptor classes (GPCRs) that are thought to be the largest eukaryotic gene and certain tools that are coupled to the database, such as family, including ~1000 different genes in the mouse (2). In the BLAST searching allowing all users to search for similarities olfactory epithelium, ORLs are thought to bind ~10 000 odor between any sequence of their interest and all sequences in the data- molecules. Until recently, however, there has been very little base. OdorDB, a companion database that contains information evidence about OR-ligand specificity (3–6). ORLs are expressed about odor molecules (the ORL ligands) and their effects in various in 25 tissues in addition to olfactory epithelium, suggesting experimental settings has been released as a pilot beta version. that members of this family of proteins may have functions ORDB is implemented using a database structure called beyond odor recognition (7). All ORLs, regardless of species EAV/CR (Entity-Attribute-Value with Classes and Relationships),

*To whom correspondence should be addressed. Tel: +1 203 785 3730; Fax: +1 203 785 6990; Email: [email protected] Present address: Emmanouil Skoufos, 3rd Millennium Inc., 125 Cambridge Park Drive, Cambridge, MA 02140, USA 342 Nucleic Acids Research, 2000, Vol. 28, No. 1

Figure 1. ORDB receptor entry screen. A typical ORDB receptor screen is presented. Information includes ORDB sequential name (ORL464), OR type (ORL), organism (mouse), source tissue, chromosome, GenBank accession number, trivial name, source lab, sequencing lab, length of the clone, ligand information, nucleotide and amino acid sequence, type of clone and links to GenBank, PubMed and GPCRDB. Users can retrieve information about like receptors by using the ‘show others’ links or can retrieve information about particular values (e.g., mouse) in all linked databases through ‘bridge screens’ (Fig. 2) accessed by following the underlined link of the value of interest. Users can access any information in the database using the navigation frame on the left.

which has been developed to facilitate the implementation and NeuronDB about neuronal tissue properties. From the receptor integration of databases containing heterogeneous data (19). entry screen (Fig. 1) users have the option of retrieving all receptor The basic EAV model has been used in a number of databases, entries in ORDB with the same values in the field by following the including clinical databases, and involves the computer ‘show other’ link. For example, in Figure 1, one can see all mouse science principle of row modeling. EAV/CR extends the basic receptors by following the ‘show other’ link at the ‘Organism’ level. paradigm to allow complex data values (classes) and the The values themselves are links that direct the users to explicit representation of relationships in the database (19). ‘bridge screens’ (Fig. 2) that contain information about the value in the field from the other related databases. For example, following the ‘olfactory receptor neuron’ link in ORDB INTERFACE AND NAVIGATION Figure 1, the user will arrive at the bridge screen presented in ORDB implements frames to present an intuitive interface to Figure 2. This screen contains information about the olfactory the WWW user (Fig. 1). Users have several navigation choices receptor neuron in all three databases: the ‘Neuron’ link will direct from every page: they can search the database using keywords, the user to the NeuronDB entry of the olfactory receptor neuron search the sequences in the database using BLAST, browse the properties, the ‘Experiments’ link will direct the user to a list of all database records or go to the ORDB tools and resources page experimental data in OdorDB that have olfactory receptor neuron as (Fig. 1). The flexible EAV/CR architecture of ORDB that is tissue and the ‘Olfactory Receptors’ link will direct the user to all shared with OdorDB and NeuronDB (20), allows for inter- receptors in ORDB that have been cloned from olfactory receptor operability of these related databases and for retrieval of neurons. Thus, users receive a multi-dimensional wealth of related related data from any of the databases. information regarding a particular object. A receptor entry screen is presented in Figure 1. Users can access information on receptor parameters such as sequencing ORDB RESOURCES AND TOOLS AND FUTURE laboratory, data source, type of sequence, organism, tissue, CONSIDERATIONS size of OR clone, and be redirected to PubMed, GenBank and GPCRDB entries for the receptor, as in the previous version With the fourth version of ORDB we are introducing a (Fig. 1). Furthermore, in this version users can retrieve information ‘Resources and Tools’ area to organize the available tools and from OdorDB about odor molecule properties and from to provide several resources in response to many user requests. Nucleic Acids Research, 2000, Vol. 28, No. 1 343

Figure 2. ORDB bridge screen. A typical ORDB bridge screen is presented. Through this screen users can retrieve related information about the olfactory receptor neuron in the linked databases: the ‘Neuron’ link will direct the user to the NeuronDB entry of the olfactory receptor neuron properties, the ‘Experiments’ link will direct the user to a list of all experimental data in OdorDB that have olfactory receptor neuron as tissue and the ‘Olfactory Receptors’ link will direct the user to a list of all receptors in ORDB that have olfactory receptor neuron as their source tissue.

The resources available include interactive originally produced REFERENCES alignments and phylogenetic trees of receptors, based on 1. Skoufos,E., Healy,M.D., Singer,M.S., Nadkarni,P.M., Miller,P.L. and receptor class, which are free to users and which they are Shepherd,G.S. (1999) Nucleic Acids Res., 27, 343–345. encouraged to use in their own research. In the tools and 2. Buck,L. and Axel,R. (1991) Cell, 65, 175–187. resources area we have included a gateway to BLAST 3. Zhao,H., Ivic,L., Otaki,J.M., Hashimoto,M., Mikoshiba,K. and searching (an additional gateway is available through the Firestein,S. (1998) Science, 279, 237–242. 4. Krautwurst,D., Yau,K.W. and Reed,R.R. (1998) Cell, 95, 917–926. navigation frame, Fig. 1). Planned resources include: the 5. Malnic,B., Hirono,J., Sato,T. and Buck,L.B. (1999) Cell, 96, 713–723. classification of the ORLs according to the particular motifs 6. Touhara,K., Sengoku,S., Inaki,K., Tsuboi,A., Hirono,J., Sato,T., present in their sequence (7); the automatic generation of a Sakano,H. and Haga,T. (1999) Proc. Natl Acad. Sci. USA, 96, 4040–4045. multi-level descriptor that includes species, ORDB ID 7. Skoufos,E. (1999) Receptors Channels, in press. (sequential name), laboratory (trivial) name, phylogenetic 8. Troemel,E.R., Chou,J.H., Dwyer,N.D., Colbert,H.A. and Bargmann,C.I. (1995) Cell, 83, 207–218. classification, chromosome information, tissue expressed, full- 9. Robertson,H.M. (1998) Genome Res., 8, 449–463. length or partial clone and ligand information, which can act as 10. Clyne,P.J., Warr,C.G., Freeman,M.R., Lessing,D., Kim,J. and a multi-level, database driven nomenclature of the ORLs; and Carlson,J.R. (1999) Neuron, 22, 327–338. the automatic generation of a sequential name upon receptor 11. Vosshall,L.B., Amrein,H., Morozov,P.S., Rzhetsky,A. and Axel,R. (1999) Cell, 96, 725–736. entry by a source laboratory. 12. Fields,S. (1990) Trends Biochem. Sci., 15, 270–273. In the near future OdorDB will be populated with the experi- 13. Dulac,C. and Axel,R. (1995) Cell, 83, 195–206. mental results of the >10 000 odor molecules. The internal 14. Matsunami,H. and Buck,L.B. (1997) Cell, 90, 775–784. flexibility of the EAV/CR schema, as well as the interoperability 15. Horn,F., Weare,J., Beukers,M.W., Horsch,S., Bairoch,A., Chen,W., introduced between the databases will allow users to make virtual Edvardsen,O., Campagne,F. and Vriend,G. (1998) Nucleic Acids Res., 26, 275–279. links between specific ORLs and their odor molecule ligands. 16. Kristiansen,K., Dahl,S.G. and Edvardsen,O. (1996) Proteins: Struct. Funct. Genet., 26, 81–94. 17. van Rhee,A.M. and Jacobson,K.A. (1996) Drug Dev. Res., 37, 1–38. ACKNOWLEDGEMENTS 18. Kolakowski,L.F.,Jr (1994) Receptors Channels, 2, 1–7. This work has been supported in part by NIH grant R01 19. Nadkarni,P., Marenco,L., Chen,R., Skoufos,E., Shepherd,G. and Miller,P. (1999) JAMIA, in press. DC02307 (Human Brain Project) and NIH grants G08 LM05583 20. Mirsky,J.S., Nadkarni,P.M., Healy,M.D., Miller,P.L. and Shepherd,G.M. and TI5 LM07056 from the National Library of Medicine. (1998) J. Neurosci. Methods, 82, 105–121.