Oligometrics.com	"useful results"

---

What is the best way to deal with sequence space? Our approach is to jump in and make a lot of sequences, using basic chemical ideas, intuition, and just plain curiosity to design them. For instance, what happens when you make a peptide with 20 randomly selected aromatic residues? It turns out the GFP-peptide fusion is insoluble. It can be made soluble again by placing 8 carboxylates in the chain. What happens if you try to make a hydrophobic peptide with alternating Ile and Leu residues? The E. coli colonies have no fluorescence. What happens when a different hydrophobic peptide is made from alternating Pro-Pro-Val trimers? You get bright colonies and a soluble fusion. Apart from the fascination of actually conceiving and then making these peptides, they offer insights regarding the kinds of peptides you can think about making and isolating. So, too many aromatic groups (especially Trp), or ILILIL, make for a fusion that is hard to work with, while proline is well tolerated. The idea was hatched to design a peptide around a ligand to see if will bind to it. This turns out to be very difficult, in part because the peptide is not pre-organized and does not have a stable shape complementary to the ligand. This is in contrast to many proteins, whose tertiary structure defines a cleft or pocket for binding small molecule ligands. But what is the size cutoff? How small can a protein/peptide be and still take on a pre-organized shape and start to act like a protein?		We have a lot of evidence at this point that 20-mers with a host of different sequences behave as if they have shapes. That is, they seem to fold like proteins, even if the evidence for this is roundabout and not definitive. The key experiments showing this is enhanced IMAC affinity in 20-mers with only two histidines which are separated by many amino acid residues: if the peptide binds to the metal in the IMAC resin, how can it be doing this if the histidines are not acting together to chelate the metal? And if they are acting together to chelate the metal, then they must be close together in space, which means the intervening amino acids are folding and holding the histidines in close proximity. This seemed like such a big discovery that many variations of several core pepides were made, many of which helped to build a convincing folding story. Unfortunately, when new IMAC resin was ordered, most of the experiments could no be reproduced, in spite of the fact the replacement product was from the same supplier with the same catalog number. A look at the patent literature has offered some hints as to what might account for the change in IMAC behavior, but patents are written to protect an idea, not to explain in the kind of detail needed to solve a riddle--on top of which a patent might have a dozen specific examples, any one of which could be the material in a patented commercial product. So now the peptide work is being supplemented by work trying to make solid supports which will give the same selectivity for good and bad folding as was observed with the first batch of IMAC resin. __________________________
What choice do we have?

---