How a Peptide Is Actually Made — Building a Molecule One Link at a Time

The Problem Merrifield Was Solving

Before 1963, making a peptide in a laboratory was miserable work. You joined two amino acids in solution, then had to fish the product out of the mixture, purify it, and start again for the third. Every round lost material. A chain of ten amino acids might take a year, and chains much longer than that were effectively out of reach.

Picture threading beads onto a string in a swimming pool. Every time you add a bead you have to find your string again among everything else floating about, dry it off, check nothing fell off, and get back in. It is not that any single step is hard. It is that the retrieval between steps is where all the time and all the losses go.

Robert Bruce Merrifield, working at Rockefeller University, had an idea that now looks obvious and at the time was not: what if you nailed one end of the string to the side of the pool? Then you never have to find it again. You can add a bead, flush the whole pool clean, add the next bead, flush again. The chain stays put throughout, and everything you do not want simply washes away.

That is solid-phase peptide synthesis. The first amino acid is anchored to a small plastic bead, and the chain is grown while attached to it. It collapsed a year of work into days, made almost every peptide in this library possible, and won Merrifield the Nobel Prize in Chemistry in 1984.

The Cycle, Step by Step

The chain is built in a repeating four-step loop. It is worth following once, because everything that can go wrong with a peptide goes wrong somewhere in this loop.

The caps are the clever part, and they are the reason the method works at all. Amino acids are sociable molecules — most have more than one place they could bond. Left to themselves in a flask they would join up in every direction and produce a useless tangle. Capping every reactive site except the one you want is the chemical equivalent of taping over every socket on an appliance except the one you intend to plug into. The chain can only grow in one direction, in the order you choose.

One detail surprises people: the chain is built backwards. Peptide sequences are written from what chemists call the N-terminus, but synthesis starts from the other end and works towards it. So a sequence written as A-B-C is assembled C first, then B, then A. It makes no difference to the finished molecule — it is simply which end is easier to nail down.

When the last amino acid is on, the finished chain is cut free from the bead, the remaining caps are stripped, and the raw peptide is purified — usually by chromatography, which separates the intended chain from everything else the build produced. Then it is freeze-dried into the white powder that ends up in a vial.

Why Longer Peptides Are Harder and Dearer

Each coupling step is very efficient, but not perfect. Suppose 99.5% of chains successfully add the next amino acid. That sounds close enough to flawless to ignore. It is not, because the loop repeats.

This is the compounding problem, and it is the single most useful thing to understand about peptide manufacturing. At 99.5% success per step, a five-amino-acid peptide comes out around 98% correct chains. A thirty-amino-acid peptide, same efficiency, same chemistry, same operator, comes out around 86%. Nothing got worse — the failure rate per step never changed. There were simply six times as many chances to slip.

Two things follow. First, price: a longer peptide costs more not because the raw materials are dearer but because you throw more of the batch away. Second, and more importantly, the near-miss chains — missing a link, doubled a link, stopped early — are exactly the population that shows up as the other 1% on a certificate. They are not contamination that got in. They are the batch's own failures that did not get filtered out.

This also explains something that puzzles people: why two suppliers selling the same compound can genuinely differ in quality. The sequence is identical and the chemistry is public. What differs is coupling efficiency, the quality of the starting amino acids, and how aggressively the crude product is purified afterwards. Same recipe, different kitchens.

The Other Route: Grown, Not Built

Not every peptide is assembled chemically. Longer ones — and anything approaching protein size — are often made biologically instead, by inserting the genetic instructions into bacteria or yeast and letting the cells manufacture it.

The difference is roughly bricklaying versus farming. Solid-phase synthesis is bricklaying: you place every unit deliberately, you control the design exactly, and your errors are placement errors. Recombinant production is farming: you plant the instructions and the organism does the assembly with machinery evolved over billions of years to get it right. The cell makes far fewer sequence mistakes than a chemist can — but you inherit a new problem, which is separating your peptide from everything else the cell was doing at the time.

That is why endotoxin — bacterial debris — is a live concern for recombinant material specifically. The peptide may be sequence-perfect and still arrive with residue from the organism that made it. It is also why the two routes need different testing: chemical synthesis is checked hardest for related-peptide impurities, biological production for host-cell residue.

A third category exists that is neither: peptide preparations extracted from animal tissue, which are mixtures rather than single defined molecules. That distinction has its own page — see extracts vs synthetic peptides.

Frequently Asked Questions

Published References

1. Merrifield RB. “Solid Phase Peptide Synthesis. I. The Synthesis of a Tetrapeptide.” Journal of the American Chemical Society, 1963. Source
2. The Nobel Foundation. “The Nobel Prize in Chemistry 1984 — Robert Bruce Merrifield.” NobelPrize.org, 1984. Source
3. Behrendt R, White P, Offer J. “Advances in Fmoc solid-phase peptide synthesis.” Journal of peptide science : an official publication of the European Peptide Society, 2016. PubMed