While his colleagues fear reproducibility problems and sample prep, this veteran has the nerve to use 2D — and like it
by John S. MacNeil
In today’s environment of ultra-high-throughput proteomics, Julio Celis might seem a bit conservative. While other researchers scramble to find alternatives to running 2D gels as a method of separating proteins — or at the very least try to automate the tedious process — Celis, 60, continues to plod away in his laboratory in the Institute of Cancer Biology at the Danish Cancer Society in Copenhagen, manually running 2D gels in his quest to identify markers for various types of cancer in tissues and fluids.
Not that this bothers him. Celis, who is the institute’s scientific director, readily admits that 2D gels are not the fastest or most user-friendly approach to separating proteins, but the shortfalls do nothing to take away from the biological insight the technique provides. In fact, Celis views other researchers’ trepidation with 2D gels more as a function of their inability to master the technique rather than an indictment of the method itself. “A lot of people are a bit nervous about it, because I’m not so sure many people can run them in a reproducible way,” he says. “Most people don’t realize the problems that are involved with sample preparation and interpretation of data derived from complex samples.”
Celis’ perspective comes from years of experience; arguably he may be one of the first in the field — along with Leigh Anderson, former CSO of Large Scale Biology — to see the potential for 2D gels as a method for identifying proteins associated with disease. His first exposure to gel electrophoresis came while a postdoc in the lab of Sydney Brenner and John Smith at Cambridge University in the early 1970s. At the time, Celis was studying the genetics of suppressor transfer RNAs, in particular protein fragments derived from nonsense mutations, and needed a way to separate them. Enter 1D gels.
After finishing his postdoc, Celis took a faculty position at the University of Aarhus, in Denmark, where he became interested in studying the cell cycle because of its relevance to cancer. “There were very few people doing cell biology in Denmark, so it was a good opportunity and they have a good quality of life.”
Celis and his late wife Ariana, who were both born in Chile, set up their new lab in Aarhus to accommodate tissue culture and immunofluorescence experiments, but Celis didn’t stray from gel electrophoresis for long. By 1977 he had discovered 2D gels and begun using the technique to study the proteins involved in regulating the cell cycle. One of his first notable successes was to independently discover the PCNA protein, a marker for cell proliferation, in 1981.
Around this time, Celis also began to consider building a database of proteins using the thousands of 2D gels he had collected. To identify the proteins’ function, he wrote off to vendors and other researchers for antibodies that might bind with them in blotting experiments. The first version of Celis’ database, completed around 1980, consisted primarily of stacks of manually annotated 2D gels. Toward the end of the decade the databases gradually became computerized, and today several comprehensive databases are accessible at http://biobase.dk/cgi-bin/celis.
New technologies have also helped Celis in his efforts to identify the proteins separated by his gels. In the late 1980s, Celis and his collaborators were among the first to use Edman sequencing for protein identification. Mass spectrometry stepped onto the scene several years later, and today Celis’ lab uses Bruker Daltonics MALDI-TOFs to help identify proteins isolated from gels.
But Celis firmly believes that research tools should be used to answer a biological question, rather than to just collect data unrelated to a hypothesis. “Let the biology lead you where you want to go, to the questions you want to ask,” he says.
With all the tools falling under the banner of proteomics, Celis also believes that it is imperative for researchers with expertise in different techniques to work together. Through his role as chair of the HUPO meetings committee, he hopes to bring researchers together and has led the planning for the first HUPO Congress to be held in November in Versailles, France. “I think this is one of the reasons why the field has moved so fast,” he said. “There’s been this integration between the people who come from the technology and those who come from the biology — in the end they need to collaborate.”
This article was originally published in ProteoMonitor, the global newsweekly of proteomics technology published by GenomeWeb, LLC.