Position: CSO Large Scale Biology, Gaithersburg, MD
Prior Experience: Co-head of Molecular Anatomy Program, Argonne National Laboratory
In a biotech business environment swarming with start-ups, Leigh Anderson’s proteomics division of Large Scale Biology stands out as a wizened veteran. Founded in 1987, the Gaithersburg, Md.-based company devoted itself to analyzing proteins en masse long before proteomics became a buzzword, and the company was apparently the first to bet that cataloging the human proteome could translate into a successful business venture.
Nor is Anderson, the company’s chief scientific officer, a recent convert to the proteomics cause. He became interested in applying 2D gel electrophoresis to studying large numbers of proteins while finishing his PhD in 1975 at Cambridge University, UK, and soon after began working with his father, Norm Anderson, to develop protein analysis methods at Argonne National Laboratory in Argonne, Ill. There, he and Anderson senior contributed to the “early phases of the development of high-throughput 2D gels and software,” he said in a recent interview.
Suffice to say that the Andersons are staunch believers in the value of 2D gels. Over its 14 years in existence, the proteomics division of Large Scale Biology has tackled the thorny issues confronting practioners of the technique, namely speed of analysis and reproducibility, by designing in-house robotics and automation systems. Only completed within the last year, these systems handle first dimension gels, transfer them to slab gels, and stain, scan, and cut slab gels for mass spectrometry analysis in a variety of instruments, including Bruker and Applied Biosystems MALDI/TOF, Thermo Finnigan ion trap, and Micromass QTOF mass spectrometers.
The primary advantage of his company’s automated protein analysis system, Anderson said, boils down to statistics. Without reliable, reproducible data, it would be impossible to definitively identify protein markers of disease, or potential therapeutic proteins. “Anybody can run one good 2D gel, or two of them, or four of them. But there is a major class of interesting problems from a biological point of view that require running hundreds or thousands of samples. That’s where you just can’t do it without the automation,” he said.
Anderson doesn’t claim to have solved every problem associated with running large numbers of 2D gels, however. Automation may no longer present the biggest challenge — although he wouldn’t give a specific per-day rate at which the company analyzes gels — but the sensitivity of 2D gel analysis still gives Large Scale Biology scientists headaches. Nor are purely mass-spectrometry-based techniques any better at identifying low abundance proteins, Anderson said.
Instead, Anderson and his colleagues at Large Scale Biology are betting that protein microarrays may provide a solution to the sensitivity problem — at least when trying to detect proteins that have already been discovered. The company has formed an alliance with Biosite, a San Diego-based antibody and chip platform manufacturer, to design 2,000 antibodies to proteins identified by Large Scale Biology. The resulting arrays, when they hit the market in late 2002, should be able to detect proteins at a sensitivity three to four orders of magnitude beyond what a 2D gel or mass spectrometry-based system can currently handle, Anderson said.
Protein microarrays’ ultimate application, Anderson posited, lie in diagnostics. The technology is ideally suited for monitoring the progress of disease or drug effects, because it will enable scientists to measure the quantity and presence of large numbers of proteins simultaneously, and be able to track which type of tissue expressed the protein. “It’s a very useful method for characterizing patients in clinical trials,” he said. “You want to detect any kind of adverse reaction very easily, very early, and you want to understand as much as possible about what exactly has happened to each patient.”
In a broader sense, proteomics must face the challenge of convincing the genomics community, particularly within large pharmaceutical companies, to get more than their feet wet in proteomics, Anderson said. Big pharma’s large investments in genomics, as well as skepticism about the value of methods such as 2D gels, have both contributed to an initial reluctance to fully embrace proteomics. “All of that is true,” Anderson said. “But the issue is that you just have to automate it, and when automated there are a lot of things that can be done with it.”