It might have been SARS- or war-related travel fear, an over-abundance of proteomics conferences, the general economic climate, or a combination of all the above: Cambridge Healthtech Institute’s “Proteins to Profits” conference in Munich last month was a far cry from the previous year’s meeting, judged by the sparse attendance and by the rather downbeat mood.
The meeting — one day shorter than last year’s — covered most aspects of proteomics: protein structure, interactions, profiling, arrays, and proteomics applications in drug discovery. However, a few talks seemed only of marginal relevance to the “protein” theme. A head count of the audience revealed no more than 50 participants at any time — including speakers.
But while many of last year’s companies — notably protein microarray firms — were absent this time, several newcomers presented on novel technologies, while others had some progress to report. A summary of selected presentations follows.
Ulf Skoglund, a professor at the Karolinska Institute in Stockholm, presented electron tomography, a cryo-electron microscopy-based technology that allows users to visualize proteins and protein complexes in three dimensions. Low doses of electrons are used to generate images of a protein sample at different angles, and a 3-D picture of individual proteins or protein complexes is reconstructed using a proprietary algorithm called Comet, which was co-developed by Skoglund. The images can reach a resolution of two nanometers and can be overlayed with X-ray crystallography structures. Sidec Technologies, a company founded in March 2000 by Skoglund, offers electron tomography as a commercial service. While the majority of projects deal with proteins in solution, the technique can also image protein complexes from tissue sections. The company is backed by 30 million Krona ($3.7 million) in total VC funding — this year alone, it received 8 million Krona ($1 million) from the Karolinska Investment Fund and Industry Fund. Sidec’s customers include BioInvent, Biovitrum, Amersham Biosciences, Procognia, and the Cleveland Clinic Foundations.
Instead of electron microscopy, Sireen — which stands for Signal Regulation Engineering — of Martinsried, Germany, bets on X-ray crystallography. According to CSO Ismail Moarefi, the company tests selectivity of compound binding to protein kinases in vitro and in cellular assays. By analyzing kinase crystal structures and mutating key residues, the scientists can create kinase mutants to which highly selective inhibitors no longer bind. This, Moarefi explained, reduces the amount of work required in selectivity screens. Sireen, founded in January of 2002, has a high-throughput facility for crystallizing and co-crystallizing proteins. Last spring, it started a long-term strategic alliance with Evotec and received €5 million in venture capital during a first financing round last fall, which was led by Deutsche Venture Capital.
Last year, Mathias Uhlén from the Royal Institute of Technology in Stockholm presented his ideas for a protein-analysis and polyclonal antibody project. This year, it has turned real. With $28 million from the Swedish Wallenberg Foundation, Uhlén plans to study 14,000 gene products over the next four years, creating biological data for half of them. This will involve expressing recombinant proteins in E. coli, raising IgY antibodies against them in chicken, and using these to characterize protein expression in tissues. Images and protein clones will be made publicly available, while the antibodies will be shared through selective collaborations. Goals for 2003 include tackling chromosomes X, Y, and 22.
While GeneProt’s CSO Keith Rose kept mum about the recent downsizing of his company — in which 45 of the 95 employees were let go last month — he stressed its scientific achievements in his talk. In March, GeneProt finished the analysis of plasma samples from patients with coronary artery disease and healthy controls, one of three projects it is performing for Novartis. Using its separation process for small proteins, the company found more than 600 proteins, “quite a few” of which were differentially expressed and have been delivered to Novartis for testing in biological assays. The 2D-gel based approach to separate larger proteins revealed less than 200 proteins, Rose told ProteoMonitor, “some” of which were differentially expressed.
GeneProt is currently analyzing plasma samples from a dementia study and serum samples from a pregnancy study, and expects first results from both projects soon.
— JK