By Mary Welch

Structural GenomiX raised $7.5 million in its initial round of financing, enabling the young company to proceed with its mission of determining the 3-dimensional structures of large numbers of proteins in order to facilitate the design of drugs and agricultural compounds.

¿The $7.5 million was a perfectly reasonable amount of money and we¿re quite happy with it,¿ said Tim Harris, the company¿s president and CEO. ¿We¿ll have a Series B round of financing next year in which we¿ll go after funds considerably north of this amount. We really needed this money to get going and build some value before going after a larger sum of money.¿

Participating in the financing was Atlas Venture, of Boston; Prospect Venture Partners, of Palo Alto, Calif.; and AppleTree Partners, of Westport, Conn.

Structural GenomiX (SGX) was founded in April by Harris, former senior vice president of research and development for Axys Pharmaceuticals Inc., of South San Francisco, and Barry Honig and Wayne Hendrickson, both professors of biochemistry and molecular biophysics at Columbia University in New York. The fourth founder is Linda Grais, a physician and attorney who has represented several biotechnology companies.

¿The other three founders had been talking for several months about starting a company that used high-throughput X-ray crystallography to determine the 3-dimensional structures of proteins,¿ Harris said. ¿I was looking for something else to do and we met. I found their ideas quite attractive and not difficult to follow. They appreciated my business experience as well as my enthusiasm and understanding of their ideas. And I thought their ideas would be very successful.¿

The San Diego-based company was formed to use the flood of genomics information available to apply a genomics approach to structure determination. Instead of targeting a single gene from one organism for structure determination, SGX will express a set of homologues of the target. By solving the structure of any one member of a set, information on the whole group is obtained. SGX scientists choose targets from a diverse group of protein targets, including protein families that include proven drug targets, proteins implicated in diseases and proteins involved in biological systems of particular interest to industry.

¿We intend to take 3-dimensional protein structures and integrate them into pharmaceutical and agricultural compound design,¿ Harris said. ¿These structures will enable companies to select targets and optimize leads faster and more precisely. We want to become the premier source for protein structures and intend to solve up to 200 structures per year. We fully expect that we¿ll be able to do that by 2001.¿

By understanding a protein¿s 3-dimensional structure, insight is gained as to how that protein works in the cell, what biological pathway it may be associated with and how it may perform its biochemical action.

With structural information as a guide, detailed descriptions of surface properties can be used to identify new sequence-structure-function relationships for target validation and lead generation. The company intends to first establish the technology through strategic alliances with pharmaceutical and agricultural companies. It then plans to license it through database subscriptions with structure-based drug design an eventual goal, Harris said.

The SGX database of targets will allow researchers to access hundreds of novel proteins in the context of known structures. The database organizes protein sequence information into families and superfamilies and characterizes each sequence family in structural terms. It also systematizes the construction of homology models and contains extensive links to external databases including relevant information on biological pathways, functional annotation and interacting small molecules.

¿While sequence information has grown exponentially in recent years, there has been no comparable increase in the number of proteins for which there is structural information,¿ Harris said. ¿Of the 150,000 or so proteins encoded, only about 2,500 have had their structures solved, and only about 500 currently serve as drug targets. This leaves a large number of human proteins ¿ potential drug targets ¿ with no structural information available. Several companies are working on 2-dimensional genomic sequence information, but we will be the company creating the standard for 3-dimensional protein information.¿

SGX has 24 employees and aims to have 60 by the end of next year. It intends to be able to solve up to 200 structures a year. One technique the company said helps determine the structures so rapidly is the multiwavelength anomalous dispersion phasing technique for X-ray crystal structure determination, which was pioneered by Hendrickson, Harris said.

¿You want to be realistic and bullish at the same time,¿ Harris said. ¿If we said we could do 500 structures in a year, that would be unrealistic. But 200 is certainly within our capabilities by 2001.¿