Scaling peptide manufacture
A peptide that is straightforward at fifty milligrams can be a difficult manufacturing problem at fifty grams. The chemistry does not change, but the economics invert, and decisions that were irrelevant at research scale become the entire route selection.
What actually changes with scale
At research scale the dominant cost is labor and turnaround. Reagents are cheap relative to a chemist's week, so nobody optimizes excess. Four or five equivalents of protected amino acid per coupling is normal, double couplings are routine insurance, and the crude is purified on whatever column is free.
At kilogram scale the amino acid derivatives and the solvent become the cost. Arginine, tryptophan and histidine derivatives are expensive, and at five equivalents you are discarding four of every five. Solvent volume drives waste disposal, and DMF in particular carries a regulatory burden in the European Union that has pushed real substitution work toward greener alternatives.
Purification inverts too. A preparative run that takes an afternoon at research scale becomes a multi-day campaign with a large solvent inventory, and the yield at the pooling step is now worth arguing about.
Route selection
Straight solid-phase synthesis remains the right answer for most peptides up to roughly thirty residues, including at commercial scale. It tolerates a wide range of sequences, and it is what most contract manufacturers are already set up to run.
Hybrid synthesis builds protected fragments on resin, cleaves them with side chains still protected, and joins them in solution. This is the standard approach above about fifty residues. Each fragment is short enough to synthesize cleanly, purification happens at the fragment stage where the impurities differ more from the product, and the convergent assembly means a failure costs one fragment rather than the whole chain.
Native chemical ligation joins unprotected fragments through a thioester and an N-terminal cysteine, in water. It makes genuinely long chains accessible and is the route to synthetic proteins of a hundred residues or more. It constrains the sequence, because you need a cysteine at the junction or a desulfurization step afterward.
Recombinant expression deserves consideration above about forty residues, and is usually cheaper if the peptide tolerates it. The tradeoffs are the ones that always apply: no non-natural residues without engineered machinery, a fusion and cleavage strategy to handle, and host-derived impurities to clear.
The cost drivers worth knowing before you ask for a quote
Sequence length sets the number of cycles and therefore the raw material consumption, and yield compounds across cycles, so length hurts twice.
Difficult residues raise the price disproportionately. Arginine is expensive and couples slowly. Cysteine needs a milder activation to avoid racemization and then needs a controlled oxidation step. Multiple cysteines needing defined connectivity turn folding into its own development project with its own yield.
Modifications change the route. A C-terminal amide is routine and costs almost nothing, being a resin choice. PEGylation, lipidation, cyclization and non-natural residues each add steps, each with a yield, and each with an analytical burden to demonstrate they happened where intended.
Purity specification is the quiet driver. Moving a specification from ninety-five to ninety-eight percent can halve the recovery at the pooling step, which means buying twice the crude for the same delivered mass.
Practical sequencing of the work
Commit to the final sequence and salt form before scale-up development starts. A single residue change invalidates the route work, and a counterion change invalidates the stability data.
Develop and lock the analytical methods first. Scale-up decisions are made by comparing lots, and comparing lots across two different gradients tells you nothing.
Run one intermediate scale between research and full production. The problems that appear at ten grams, aggregation on resin, an impurity that only resolves on a wider column, a filtration that takes a day, are cheaper to find there than in the production campaign.