Reference
What makes a peptide expensive
Prices in this field look arbitrary until you know what the number of residues does to a yield. Then most of them stop looking arbitrary.
A four-residue peptide and a thirty-nine-residue peptide are made by the same method on the same machine by the same people. One of them is cheap. The reason is arithmetic, and it is worth doing once because it explains most of a price list without any reference to demand.
Every residue is a coupling, and every coupling loses something
Solid-phase synthesis builds a chain one residue at a time: anchor the first to a bead, add the next, wash away what did not react, repeat. Each of those additions is a chemical reaction, and no chemical reaction goes to completion. A good coupling might reach 99.5%.
That sounds close enough to perfect to ignore, and over one step it is. Over many steps it is not, because the losses multiply rather than add. At 99.5% per coupling:
| Chain length | Couplings | Full-length chains remaining |
|---|---|---|
| 4 residues | 3 | 98.5% |
| 10 residues | 9 | 95.6% |
| 16 residues | 15 | 92.8% |
| 29 residues | 28 | 86.9% |
| 39 residues | 38 | 82.6% |
| 44 residues | 43 | 80.6% |
Drop the per-step figure to 99% — still a good coupling — and a forty-four-residue chain finishes at 65%. At 98%, which is what a difficult sequence gives you, it is 42%. More than half the material is something other than what you were making before purification has even started.
What the lost material becomes
It does not vanish. A coupling that fails leaves a chain one residue short, and the next coupling then adds to that short chain quite happily. What comes off the resin is the target peptide mixed with a family of deletion sequences, each missing one or more residues somewhere in the middle.
Those are the hardest impurities to remove, because a chain missing one residue out of thirty-nine is chemically almost identical to the one you want. It is very nearly the same size, very nearly the same charge, and it comes off a chromatography column at very nearly the same time. This is the specific reason purification, not synthesis, is where the money goes on a long peptide.
Difficult sequences
Some sequences are harder than their length suggests. A stretch of residues with a tendency to form beta sheets will fold back on itself while still attached to the resin, and a chain that has folded has buried the end the next residue needs to attach to. Couplings that were running at 99.5% start running at 90%, and the effect is not recoverable further down the chain.
Synthetic chemists have a vocabulary for this — difficult couplings, aggregation, the need for pseudoproline dipeptides or a different resin or double couplings at particular positions. All of it means the same thing commercially: the same length, more work.
Residues that are not standard
Several compounds in this range contain amino acids that do not occur in proteins, and those residues are bought rather than made. Alpha-aminoisobutyric acid in the incretin analogues, norleucine in Melanotan I, D-phenylalanine in place of the L-form, 2,6-dimethyltyrosine in SS-31. A protected non-standard residue can cost many times what a protected standard one costs, and if it sits at position 30 of a 39-residue chain, every failed coupling after it wastes material that was already expensive.
D-amino acids are the clearest case. They are the same molecule as their L-partner in a mirror, and making them is a separate manufacturing problem from making the common form.
Rings, bridges and the things done after the chain is built
Not every peptide is finished when the last residue is on. Melanotan II is cyclic and has to be closed. Oxytocin has a disulfide bridge, which means two particular cysteines have to be joined to each other and not to any of the other cysteines in the vessel. Several of the incretin analogues carry a fatty acid chain attached at a specific residue, which is its own reaction with its own yield.
Each of these is a step after the expensive part, performed on material that already represents the whole cost of the synthesis. A 70% yield on a cyclisation is a 30% loss of finished peptide.
The last percent of purity is the most expensive percent
Purification by preparative chromatography works by throwing material away. The column separates the target from everything near it, and you collect only the part of the peak you are confident about. Collect generously and the yield is good and the purity is moderate. Collect narrowly and the purity is high and the yield is not.
Going from 95% to 98% costs some yield. Going from 98% to 99.5% costs a great deal more, because the impurities still present at 98% are the ones that come off the column closest to the target — the deletion sequences — and separating those means discarding good material along with them. A higher purity specification is not a different process; it is the same process run at a worse yield on purpose.
What this adds up to
- A short chain of standard residues with no modifications is cheap, and should be. KPV is three residues; Epithalon is four.
- A long chain is expensive roughly in proportion to how badly the yield arithmetic runs against it, which is faster than in proportion to its length.
- Non-standard residues, cyclisation, bridges and attached fatty acids each add cost on top of length rather than instead of it.
- Two suppliers quoting the same compound at very different prices are usually not quoting the same purity specification, and sometimes not the same net peptide content.
How peptides are made · The families of research peptides · The range
