How Bruce Merrifield made the peptide era possible — solid-phase peptide synthesis
Before 1963, synthesising a new peptide was a months-long act of purification, loss, repetition. Bruce Merrifield at the Rockefeller Institute had a disarmingly simple idea: anchor the growing peptide chain to a solid polymer resin so that excess reagents can simply be washed away after each step. Two decades later every commercial peptide synthesiser was an application of this principle — and in 1984 Merrifield received the Nobel Prize in Chemistry.
The problem with classical peptide synthesis
In the 1950s peptide synthesis was done in liquid phase. Every single amino acid coupling required: activation of one amino acid, reaction with the other, purification of the product from reagents and side products, drying, characterisation. For a decapeptide this cycle repeated nine times, and at each step 5-20% of material was lost. Cumulative yield was often below 10% of starting material. Insulin — 51 amino acids in two chains — was synthesised between 1953 and 1963 over multi-year work; that was at the time a spectacular individual feat.
Bruce Merrifield, a young biochemist at the Rockefeller Institute in New York, formulated the problem differently in 1959: the obstacle was not the individual reactions but the purification between steps. What if the peptide were anchored to something one can trivially separate — an insoluble resin?
The idea: a resin as anchor
Merrifield published the first description of the method in JACS (Journal of the American Chemical Society) in 1963: the C-terminus of the growing peptide chain is covalently bound to polystyrene beads sitting in a glass reactor. A protected amino acid is added, coupled, and the reaction solution together with excess reagents is simply washed out of the reactor through a filter. The protecting group on the α-amino group of the appended amino acid is removed, washed, and the next amino acid is added. At the end, the finished peptide is cleaved off the resin by chemistry.
What used to take weeks now took hours. What used to have a cumulative yield of 1-10% now had 60-90%. What previously needed an individual specialised chemist per substance could now be automated — and that is precisely what Merrifield did: in 1965 he published the first semi-automated peptide synthesiser. In 1969 his group synthesised ribonuclease A (124 amino acids) on the machine — a complete enzymatically active protein synthesis that definitively validated the method.
What the method opened for pharmacology
Without SPPS, practically none of today's approved peptide medicines would have come into being. Octreotide (1988), the GnRH agonists (leuprolide 1985, goserelin 1989), calcitonin analogs, later the GLP-1 substances (exenatide 2005, liraglutide 2010, semaglutide 2017), bremelanotide (2019), cagrilintide (2024) and all GHRH/GHRP variants are direct or indirect follow-on applications of the solid-phase method. Recombinant production via bacteria or yeasts (for longer proteins such as recombinant insulin from 1982 or recombinant growth hormone from 1985) complemented SPPS but did not replace it — for mid-length peptides (10-50 amino acids), SPPS remained superior.
Economically, the method has been the basis of its own industry to this day: specialised contract-synthesis houses such as Bachem (Bubendorf, Switzerland), CEM, Bachem North America and PolyPeptide Laboratories produce synthetic peptides at industrial scale — from research quantities (milligrams) to tonnage scale for approved medicines.
„The idea was basically simple: why not anchor one end of the chain instead of having to isolate every intermediate? But if something only looks simple in retrospect, it usually wasn't simple before."
The iteration: Fmoc instead of Boc
Merrifield's original method used tert-butyloxycarbonyl (Boc) as the temporary protecting group and hydrogen fluoride (HF) for cleavage from the resin. HF was highly corrosive and required special apparatus. In the 1970s and 1980s Louis Carpino developed an alternative protecting group — 9-fluorenylmethyloxycarbonyl (Fmoc) — that could be removed under mild basic conditions and allowed a substantially more peaceful solvent system (trifluoroacetic acid instead of HF) for the final cleavage. Fmoc-SPPS is today the standard in most laboratories and industrial facilities.
Further iterations include microwave-assisted coupling, flow-chemistry-based continuous SPPS and Native Chemical Ligation (NCL) — the latter enables synthesis of proteins with over 100 amino acids by chemoselective joining of multiple SPPS-synthesised fragments. All of these methods build on Merrifield's basic principle.
What the SPPS story shows
Three observations are methodologically interesting. First: a technical innovation can make an entire pharmacological class possible in the first place. Without SPPS, modern peptide medicines would be either impossible to synthesise or so expensive that only small research quantities would be conceivable. Second: Merrifield's step from concept (1959) to validation (1963) to Nobel recognition (1984) took 25 years — recognition followed not the invention but the demonstration that it had actually transformed science. Third: SPPS is a classical example that the most important innovation is often not new chemistry but a new architecture of chemistry — the reactions SPPS uses were largely known in the 1950s; what was missing was the idea of chaining them in a heterogeneous format.
Open questions
- What further iteration leaps for SPPS are foreseeable — flow chemistry, in-vivo translation, new protecting-group chemistries?
- How does the increasing availability of Native Chemical Ligation change the line between 'synthetic peptide' and 'recombinant protein'?
- What environmental aspects does industrial peptide synthesis have — solvent consumption, atom economy — and what greener alternatives are emerging?
- Does the economic consolidation of specialised peptide CMOs (Bachem, PolyPeptide) have consequences for supply security of approved medicines?