I found this to be a remarkable paper. Making synthetic peptides into drugs has been something that people have been trying for decades now, but it’s a really hard way to make a living. Proteins get degraded. They get degraded in the gut, in the blood, and in every tissue you can name. That may sound a little odd if you’re just picking up the subject of biochemistry, because proteins themselves are such a key part of the physical and chemical structure of any living creature – how (and why?) would they be so likely to be broken up? There are several answers. The proteins being used by living systems are often tailored to their environments, and they generally have an expiration date on them, anyway – they last as long as they need to last in the particular compartment they hang out in. Some have a half-life of minutes, some last for weeks, and a few (like the lens proteins in the eye) are there for the duration, but in general, proteins are there to be turned over. Their sequences, versus the proteases and other degradation pathways around them, are balanced out evolutionarily.
So when you throw a new one into the soup, it’s on its own, and it’s facing an environment that’s biased towards tearing unknown visitors into pieces. That’s not to say that there are no protein/peptidic drugs out there, far from it, but it does present a challenge. There have been many, many tricks developed over the years to toughen peptides sequences up and produced peptidomimetics that can simultaneously hit their targets and survive degradation, but it’s not a straightforward process. It’s more like “Oh, that didn’t work? Well, try this”. Cyclic peptides, to pick one example from the bag of ideas, can be unusually stable, but making them can present some significant synthetic challenges, and you’re also faced with a huge number of potential cyclic possibilities to explore. So a title like “Accurate de novo design of hyperstable constrained peptides” does call attention to itself.
The authors, a large multicenter team from the University of Washington and several other institutions, have a lot to prove after an opening like that, but I have to say that it’s an impressive paper. It shows a dozen examples of cyclized peptides designed through these techniques, which admittedly is not a very long list compared to the universe of possibilities. But they hit a pretty good variety, though – natural and unnatural amino acids and stereochemistries, several types of linkers, and a number of completely different geometries. It’s computationally fairly intensive work, but that’s nothing compared to the alternative:
Large numbers of peptide backbones were stochastically generated as described in the following sections, combinatorial sequence design calculations were carried out to identify sequences (including disulfide crosslinks) stabilizing each backbone conformation, and the designed sequence–structure pairs were assessed by determining the energy gap between the designed structure and alternative structures found in large-scale structure prediction calculations for the designed sequence. A subset of the designs in deep energy minima were then produced in the laboratory, and their stabilities and structures were determined experimentally.
The disulfide-stabilized natural-amino-acid proteins that they designed could be made in cell culture, and they obtained a crystal structure of one of these, and NMR structures for several others. In every case, it seems like the designed structure matched quite well with the real one, and the proteins had greatly enhanced thermal and chemical stability. The unnatural ones, of course, had to be synthesized by hard-working chemists, so the group added an extra layer of simulation to the calculations to make sure that this wasn’t going to be time wasted. One of the four design classes still managed to go off on its own a bit, on one end of the structure, but the others matched up well. Further tests were backbone-cyclized peptides and one class that has no analogs in nature at all.
This could all be very interesting – it’s sort of the second coming of the “stapled peptide” idea, but that one was best suited to presenting a helical peptide of a certain size, whereas these designs seem to be able to deliver various loops, beta-sheets, and other surfaces. There are a lot of protein-protein targets out there that one would like to apply these to – I’d be very interested in seeing a test case (like the Bcl system) tried out with several of these new constrained geometries to see what the partner protein makes of them. It looks like one should be able to design plausible binding partners to such targets, but we’ll see what happens when someone actually tries that – and after this paper, I think we can assume that someone will. Say the authors:
The hyperstable molecules presented in this study provide robust starting scaffolds for generating peptides that bind targets of interest using computational interface design or experimental selection methods. . .it should be possible to re-engineer the peptide surfaces, incorporating target-binding residues to construct binders, agonists or inhibitors. There has been considerable effort in both academia and industry to use small, naturally occurring proteins as alternatives to antibody scaffolds for library selection-based affinity reagent generation. Our genetically encoded designs offer considerable advantages as starting points for such approaches. . .
Going beyond the re-engineering of our hyperstable designs to bind targets of interest, the methods developed in this Article can be used to design new backbones to fit specifically into target binding pockets. Such ‘on-demand’ target-specific scaffold generation is likely to yield scaffolds with considerably greater shape-complementarity than that of scaffolds generated without knowledge of the target. More generally, our computational methods open up previously inaccessible regions of shape space, and, in combination with computational interface design, should help unlock the pharmacological potential of peptide-based therapeutics.
I hope that they’re right about this, of course. Time to go find out!