Let’s start with a joke. “What are three Germans doing that you have put into one room? – Founding an association!” In Germany we have associations for everything in the smallest village. Associations of hen breeders, associations of stamp collectors, associations of local singers, associations of hobby gardeners, associations of wine drinkers, associations of The Kelly Family concert visitors, and so on. Since late 2001 we additionally have the German Society for Proteome Research (DGPF), whose very first founding charter was wrote down on a beer mat (well, we are in Germany, aren’t we).
The foundation of the DGPF by scientists and industry representatives was a reaction on latest market and application movements towards protein research. Germany already has had strong Proteomics (& protein) research when others were still chasing the holy grail Genomics. But – to my impression – it was never really well communicated. So, one major aim of the DGPF will be to improve the international knowledge about the high level of German Proteomics.
But why are researchers and the industry more and more focussed on Proteomics? One of the major disadvantages of Genomics approaches is the missing connection between a gene and its cellular function. The fact that a gene has been sequenced does not give us the cellular function of the gene product. That makes genomic results so difficult to interpret. Even the sequence analysis with bioinformatics tools does not yield the full picture. Additional problems arise through the organisation of the genetic information as well as the fact that only a subset of genes is active in a specific cell at a specific stage.
So, scientists are moving to the functional level, to the gene products, to the proteins. And they developed the new term, “Proteomics”, for the complete set of proteins (functions) of a cell in a specific stage, in analogy to “Genomics” that addresses the complete set of genes (information) of a cell.
Similar to other attempts with a large-scale option in industrial applications (drug discovery e.g.), it will depend on the technology developing and supplying industry if Proteomics will get its chance. When I was doing the research for this article I had the impression that some companies just stuck the Proteomics label onto their existing products. This is neither a solution nor does it really fit the researchers needs. But where are the bottlenecks and what has to be done?
There is a dramatic increase of complexity while switching from the genetic to the functional level. A gene is a gene is a gene. There is slight variation caused by introns and foreign elements as well as expression control. But our scientific thinking is dominated by the “one gene – one protein” paradigm, even since the knowledge about posttranscriptional modifications has shown that it is not just that simple.
With proteins one has to view every single candidate in the context of multifunctionality and networking. In many cases one protein is not just one function. It is part of a high-complex cellular network of interacting and cascading activities. The function of most regulatory proteins for example depends on environment (regarding ‘cellular clock’ and location), posttranslational modifications and interacting partners. As a result one protein might have a couple of functions depending on where, when and with whom it is. This puts Proteomics to trouble.
At first, there are still no powerful technologies for many aspects in large-scale protein research available. Friedrich Lottspeich, head of the protein analysis group at the Max-Planck-Institute for Biochemistry in Munich and DGPF-chairman, said that recent methods exhibit great potential but are not yet ready for the industrial job, in drug discovery for example. There are only few suitable solutions for automation and high-throughput. Early stage MALDI-TOF applications work pretty well, in Structural Proteomics e.g.. But problems with high-throughput sample preparation, low abundant and hydrophobic proteins are unsolved. In Functional Proteomics automated interaction-screens based on the 2-Hybrid, SPR (surface plasmon resonance) or TAP (tandem affinity purification) technologies – that are essential to discover the networking aspect of proteins – are at its infancy. Antibody-based biochips already show the direction.
At second, Proteome research results in huge amounts of data. Corresponding to the higher complexity, Proteomics causes exponentially more data than Genomics does. But drug discovery (and scientific research in common) is not just collecting data, even if one might suspect some scientists to think so. No, the scientific progress depends on results derived by the analysis and interpretation of collected data. And this is getting more and more difficult with increasing complexity.
Finally, the complexity of protein functionality has to be taken into account while moving forward. An attempt to this is the field of Integrated Proteomics that considers various views by the combination of data coming from different approaches and sources. But . this again increases not only the total amount of data to be analysed but also the level of complexity. According to Thomas Franz, head of the Proteomics core facility at EMBL Heidelberg, existing bioinformatics solutions are not able to quantitatively and qualitatively analyse the produced data. This opinion is shared by a couple of colleagues working in the field. Scientific teams are analysing the data manually again because this is more effective and still yields the most meaningful results.
The conclusion is an answer to my question what has to be done. There is a deep need for at least a) large-scale protein research technologies, b) suitable bioinformatics solutions and c) Proteomics-optimised devices.
I am curious about the future development of Proteomics. It might be overrun by other “-omics” in public attention. But I am convinced that Proteomics will contribute important findings to our understanding of how a cell works. And for sure it is and will be a major market for technology suppliers and bioinformatics companies.
Originally published in April 2002 by Inside-Lifescience, ISSN 1610-0255.