Building peptides from the wrong end

UK chemists have cracked a long-standing problem in peptide synthesis that has prevented amino acid chains being grown from both ends. The insight could open up efficient ways to make peptide-based drugs.

Lawrence Harwood and colleagues at the University of Reading say they have discovered a way to extend peptides from the carboxylic acid end without scrambling the stereochemistry - solving a problem that has stumped chemists for decades. The Reading team has now won a ?63,000 grant for commercialisation of the technique, to license to companies working in the rapidly growing market of peptide-based drugs. 

All peptides have a carbon terminus at one end, and a nitrogen terminus at the other. But because of issues with stereochemistry, chemists are currently limited to extending peptides from the nitrogen end.

’Conventional peptide synthesis is always carried out in a very linear manner, building the peptide chain from the nitrogen terminus, as attempts to grow peptide chains from the acid terminus result in scrambling of the stereochemistry,’ explains Harwood. ’This is such a profound problem that the impossibility of synthesising peptides from the acid terminus has become dogma in the field of peptide synthesis.’

However, Harwood is new to the peptide field. ’We weren’t steeped in this dogma, and didn’t realise this reaction shouldn’t be able to happen, so we just went away and did it,’ he adds. Harwood says the new technique could be used to make peptides more easily, because it means short peptides chains could be clipped together - freeing chemists from having to add amino acids one by one to the nitrogen end of the chain.

Stereochemical scrambling

Extending a peptide from the carboxylic acid end of the chain requires that the acid group is pre-activated, increasing its reactivity towards the incoming amino acid so that the new peptide bond will form. However, this activated intermediate can transiently react with itself, forming a cyclic oxazolone. This structure is prone to temporarily losing a hydrogen ion, mixing up the stereochemistry at that point.

Harwood says he has now overcome this issue, activating the C-terminus by converting it into a morpholine ring. ’We’d been studying these compounds for well over a decade, and for most of that time we’d been trying to suppress the labile nature of the morpholine ring to be opened by nucleophiles. Only two or three years ago, I sat down and thought, instead of making this a problem, could we use it to our advantage?’ Harwood realised the morpholine could be a surrogate for a dipeptide which could be coupled at the C-terminus.

Traditional stepwise peptide construction is invariably done using a peptide synthesiser, with the C-terminus attached to a resin bead. Harwood plans to use his grant, awarded by the CommercialiSE fund, to develop his reaction to work while attached to a bead. ’Once we get it onto solid phase, we believe that, while the chemistry is totally novel, the resins we want to make will just go straight into the standard peptide synthesising kit - but now you’ll be able to do divergent peptide synthesis, so you’ll be able to build up from both ends.’

Big advantages

Alethea Tabor researches peptide chemistry at University College London, UK, and agrees there would be several big advantages in being able to extend peptides from the carbon terminus. ’The most obvious advantage is being able to do peptide ligations [to join two peptide chains together], because there’s a limit to building up peptides in a linear fashion which is much shorter than the length of the average protein. In theory, a general method that would allow you to make peptides in the other direction would enable you to ligate more or less anything.’

Making cyclic peptides would be another key advantage, Tabor adds. ’If the cyclisation is difficult or slow you stand a good possibility of scrambling that chiral centre. So again, if you had a reliable method of C-terminal activation, that could be used for cyclisation to make cyclic peptides. And of course a lot of peptide antibiotics, receptor-binding ligands and so on are in fact cyclic peptides.’

James Mitchell Crow