Back in 2005, Eric Peters and co-workers from the Genomics Institute of the Novartis Research Foundation (GNF) published the seminal paper on the use of fluorous tags in proteomics. There they described their work in using fluorous tags to enrich samples for specific peptide subsets. The proteins that these peptides originated from could then be ID’ed by peptide mass fingerprinting using MS/MS and standard programs such as MASCOT or SEQUEST. Since that time FTI and GNF have been working to refine the labeling and separation protocols and to expand the number of applications using fluorous tags.
For those not familiar with proteomics, here’s a little background. Proteomics is the study of the “product” of the genome, that is the proteins that are expressed by the genome. I’ll make the imperfect analogy of a supermarket. The genome would be all the ingredients in order listed on every product in that supermarket. The proteome would then be all products. As you can see there are a lot more products than there would be ingredients, due to the various sizes, brands, quantities, etc. of each product. The problem when studying the proteome, therefore, is often one of complexity. Estimates seem to bounce around, but once you take into account not only all the proteins the genome codes, but also all the post-translational modifications, you may be talking about 500,000 different compounds. Some sort of sample simplification, therefore, is usually needed to make things more manageable. Luckily there are a number of ways to do this. For example, if you digest all the proteins to peptides, then tag the cysteine containing peptides, and analyze them, you can still identify the proteins they originated from but have reduced the complexity of the sample by about 1.5 orders of magnitude. To carry the supermarket analogy a little further, it would be like clipping and scanning all the UPC codes and associating them with a product instead of trying to note every box in the store.
Methods to conduct this enrichment include affinity tags, with biotin or His tags being favorites. Biotin and it’s affinity partner, streptavidin, are probably still the gold standard, but there are problems, notably non-specific binding to streptavidin and getting the darn tagged molecules off the streptavidin. What Eric Peters’ group did was substitute the affinity tag with a fluorous tag then used a fluorous solid phase extraction to separate the non-tagged peptides from the tagged peptides. The fluorous tag has some nice advantages over other methods including low non-specific binding, ease of separation, and a mass defect. The last one is interesting since fluorine has a mass slightly below 19. Since C, H, N, O, and S all have masses slightly above the whole #, one can easily tell if a peptide is fluorous tagged using this mass defect.
Functionally, the fluorous tag is similar to an affinity tag. So much in fact that the Peters’ paper calls it “fluorous affinity”. That, however, is a misnomer since there is no affinity between fluorous groups. The separation is partitioning based which allows one to do things that you could not do with affinity tags. For example, they demonstrated that you could separate doubly tagged peptides from singly tagged peptides and they have also shown that you can separate peptide subsets with different length fluorous tags.
Sounds great, huh? So why hasn’t it taken off more? Well, there are some issues which need to be sorted out, such as labeling efficiency. FTI has secured an SBIR grant and in collaboration with GNF we are working on these issues and expanding into quantitative proteomics. At this point, we are closing in on a protocol that should be robust, simple, and above all, kit-able. We’ll keep you informed.