How Peptide Drugs Are Made: From Lab Bench to Pharmacy Shelf

Two roads to a peptide drug

There are two fundamentally different ways to manufacture peptides. Chemical synthesis builds the peptide chain one amino acid at a time using organic chemistry — no living cells involved. Recombinant production uses genetically engineered microorganisms (bacteria or yeast) as living factories, exploiting their natural protein-making machinery. The choice between them depends primarily on the peptide’s length: chemical synthesis dominates for peptides under ~50 amino acids, while recombinant methods are preferred for larger peptides and proteins.

Solid-phase peptide synthesis (SPPS)

Solid-phase peptide synthesis (SPPS), invented by Robert Bruce Merrifield in 1963 (Nobel Prize in Chemistry, 1984), is the dominant method for manufacturing peptide drugs. The revolutionary insight was to anchor the growing peptide chain to an insoluble resin bead, allowing all liquid reagents and byproducts to be washed away after each reaction step — dramatically simplifying purification and enabling automation.

The SPPS cycle

SPPS builds peptides from C-terminus to N-terminus (opposite to biological synthesis). Each amino acid addition follows a repeating cycle of four steps:

Step 1: Deprotection. The α-amino protecting group on the resin-bound peptide is removed. In Fmoc chemistry (the modern standard), 20% piperidine in DMF removes the Fmoc group in 5–15 minutes. The released dibenzofulvene absorbs UV at 301 nm, enabling real-time monitoring of deprotection efficiency.

Step 2: Activation. The incoming amino acid’s carboxyl group is activated with a coupling reagent (HATU, HBTU, or DIC/Oxyma) to create a reactive ester that will readily form a peptide bond.

Step 3: Coupling. The activated amino acid is added to the resin-bound peptide, forming a new peptide bond. Coupling times range from 30 minutes to 2 hours depending on the amino acid and sequence position. Coupling efficiency must exceed 99.5% per step for acceptable final yields.

Step 4: Washing. Excess reagents and byproducts are washed away with DMF solvent. The resin-bound peptide remains on the solid support.

This cycle repeats for every amino acid in the sequence. A 30-residue peptide like semaglutide requires 30 complete deprotection-coupling cycles — typically 2–4 days on a modern automated synthesizer.

The yield problem

Even at 99.5% coupling efficiency per step, losses compound exponentially. After 30 steps: 0.995^30 = 86% theoretical crude yield. After 50 steps: 0.995^50 = 78%. At 99.0% efficiency: 30 steps gives only 74%, and 50 steps gives just 60%. This exponential decay is why chemical synthesis becomes impractical beyond about 50–70 amino acids. “Difficult sequences” — regions prone to aggregation, secondary structure formation on resin, or steric hindrance — can reduce individual coupling efficiencies below 99%, compounding the yield problem.

Cleavage and global deprotection

After chain assembly, the completed peptide must be released from the resin and all side-chain protecting groups must be removed simultaneously. In Fmoc chemistry, this is accomplished with a cleavage cocktail, typically 95% trifluoroacetic acid (TFA) with scavengers (water, triisopropylsilane, ethanedithiol) to trap reactive cation intermediates. Cleavage takes 2–4 hours and yields a crude peptide that must then be purified.

Recombinant production

For peptides longer than ~50 amino acids, recombinant production is typically more practical and economical. The gene encoding the peptide is cloned into an expression vector and transformed into a host organism — most commonly E. coli bacteria for simpler peptides, or yeast (Pichia pastoris, Saccharomyces cerevisiae) and Chinese hamster ovary (CHO) cells for peptides requiring complex post-translational modifications like glycosylation.

The host cells are grown in large-scale fermentation bioreactors (hundreds to thousands of liters), producing the target peptide as part of their normal protein synthesis. The peptide is then extracted, purified through multiple chromatography steps, and formulated. Recombinant human insulin (Humulin, produced in E. coli by Genentech/Eli Lilly, approved 1982) was the first recombinant peptide drug and remains one of the highest-volume biopharmaceuticals produced globally.

Purification: HPLC

Whether produced by SPPS or recombinant methods, crude peptides must be purified to pharmaceutical grade. Reverse-phase high-performance liquid chromatography (RP-HPLC) is the gold standard. The peptide mixture is dissolved in a water/acetonitrile gradient and passed through a C18 or C8 column. Different peptide species (the target peptide plus deletion sequences, truncation products, and other impurities) elute at different times based on their hydrophobicity. The target peptide fraction is collected, and purity is typically assessed by analytical HPLC (target: ≥95% for pharmaceutical grade, often ≥98%).

Quality control: proving it’s right

Before a peptide drug can be administered to patients, it must pass a battery of quality control tests documented on a Certificate of Analysis (COA):

Identity: Mass spectrometry confirms the molecular weight matches the target peptide (typically within 0.1 Da). LC-MS/MS fragmentation can verify the full amino acid sequence.

Purity: Analytical HPLC quantifies the percentage of the target peptide in the sample. Related substances (deletion peptides, oxidized forms, degradation products) are identified and quantified individually.

Potency: Bioassays measure the peptide’s biological activity (e.g., insulin’s ability to lower blood glucose in a standardized assay).

Sterility: For injectable peptides, sterility testing confirms the absence of microbial contamination. Endotoxin levels are measured by the LAL (limulus amebocyte lysate) assay to ensure they are below safe thresholds.

Water content: Karl Fischer titration measures residual water in lyophilized (freeze-dried) peptides, which affects stability and shelf life.

Scale: from milligrams to metric tons

The scale of peptide manufacturing varies enormously. Research peptides are synthesized in milligram quantities on benchtop synthesizers. Clinical trial supply requires grams to kilograms. Commercial drugs like insulin and semaglutide require metric tons per year — Novo Nordisk alone has invested billions of dollars in expanding peptide manufacturing capacity to meet global demand for GLP-1 drugs. The peptide CDMO (contract development and manufacturing organization) market has grown rapidly, with companies like Bachem, PolyPeptide Group, and Lonza operating large-scale GMP peptide manufacturing facilities.

Difficult sequences: when synthesis fights back

Not all peptide sequences synthesize equally well. Certain amino acid combinations cause the growing peptide chain to aggregate on the resin, forming secondary structures (particularly beta-sheets) that block subsequent coupling reactions. These “difficult sequences” are the bane of peptide chemists and can drop individual coupling efficiencies below 95%, catastrophically reducing overall yield.

Common problem patterns include consecutive hydrophobic residues (Val-Val, Ile-Ile, Ala-Ala stretches), sequences prone to beta-sheet formation, and sterically hindered residues following proline or beta-branched amino acids. The industry has developed several strategies to combat aggregation: pseudoproline dipeptides (temporary proline-like structures that disrupt beta-sheets during synthesis, then revert to the native sequence during cleavage), backbone protection (Hmb or Dmb groups on nitrogen atoms to prevent inter-chain hydrogen bonding), microwave-assisted synthesis (elevated temperature disrupts on-resin aggregation), and chaotropic additives in the coupling solvent.

For pharmaceutical manufacturing, difficult sequences add significant cost. A “clean” peptide with no difficult stretches might synthesize at $1–5 per gram at scale, while a difficult sequence requiring specialized chemistry could cost $50–500 per gram. Semaglutide’s synthesis is considered moderately complex due to the unnatural Aib residue at position 8 and the C18 fatty diacid side-chain attachment, which requires specialized linker chemistry and post-cleavage conjugation steps.

Fragment condensation: building big from small

For peptides too long for single-run SPPS (roughly >50 residues), fragment condensation offers a hybrid approach. The target peptide is divided into 2–4 shorter fragments, each synthesized and purified separately by SPPS, then joined together by native chemical ligation (NCL) or enzymatic ligation. NCL, developed by Philip Dawson and Stephen Kent in 1994, exploits a chemoselective reaction between a C-terminal thioester on one fragment and an N-terminal cysteine on the next, forming a native peptide bond at the ligation site without racemization.

Fragment strategies are increasingly important as the industry pushes toward longer therapeutic peptides. Insulin (51 amino acids with 3 disulfide bonds) has historically been produced recombinantly, but chemical synthesis routes using fragment condensation plus directed disulfide bond formation are now competitive at manufacturing scale. This gives manufacturers flexibility to choose the most cost-effective route based on their existing infrastructure and expertise.

Formulation: turning a peptide into a medicine

A purified peptide is not yet a drug. Formulation — the science of turning an active pharmaceutical ingredient (API) into a stable, deliverable, and effective medicine — is a critical step that determines how a peptide drug is stored, administered, and experienced by patients.

Most peptide drugs are formulated as sterile aqueous solutions for injection, typically in prefilled pens or vials. Key formulation challenges include maintaining peptide stability (preventing aggregation, oxidation, and deamidation), achieving the target concentration and viscosity for subcutaneous injection through thin needles, controlling pH (most peptides are formulated at pH 4–7 for optimal stability), and including appropriate excipients (buffers, tonicity agents, preservatives for multi-dose formulations, and surfactants to prevent surface adsorption).

Extended-release formulations represent a major innovation in peptide drug delivery. Bydureon (exenatide) uses biodegradable poly(lactic-co-glycolic acid) (PLGA) microspheres that slowly release the peptide over one week after subcutaneous injection. Lupron Depot (leuprolide) uses similar microsphere technology for monthly or even 6-month dosing. These technologies transform short-acting peptides into long-acting depot formulations, dramatically improving patient convenience and adherence.

Lyophilization (freeze-drying) is the standard preservation method for peptides that are unstable in solution. The peptide solution is frozen, then the ice is removed by sublimation under vacuum, leaving a dry, porous cake that can be stored at room temperature or refrigerated for months to years. Before administration, the lyophilized powder is reconstituted with sterile water or saline. Many research peptides and some clinical peptides (e.g., some GnRH analogs, some GH secretagogues) are supplied in lyophilized form.

The economics of peptide manufacturing

Peptide manufacturing is significantly more expensive than small-molecule drug production, primarily due to the cost of amino acid building blocks, coupling reagents, HPLC-grade solvents, and the need for GMP-compliant cleanroom facilities. A rough cost breakdown for SPPS manufacturing of a typical 30-residue peptide at commercial scale (multi-kilogram):

Raw materials (Fmoc-amino acids, coupling reagents, solvents): ~30–40% of total cost. Fmoc-amino acids range from $50–500 per kilogram depending on the residue. Unnatural amino acids like Aib can cost 10–100x more than standard residues.

Labor and overhead (GMP-trained operators, quality personnel, facility maintenance): ~25–35%. Peptide manufacturing requires highly specialized expertise — the number of skilled peptide chemists globally is limited.

Purification (preparative HPLC columns, solvents, analytical testing): ~20–30%. Purification is often the bottleneck — columns are expensive, runs are slow at scale, and solvent consumption is enormous.

Quality control and documentation (analytical testing, stability studies, regulatory documentation): ~10–15%.

The dramatic increase in GLP-1 drug demand has triggered a global peptide manufacturing capacity crunch. Major CDMOs have announced billions of dollars in expansion: Bachem is building new facilities in Switzerland and the US, PolyPeptide Group has doubled capacity at its Malmö site, and Novo Nordisk has invested over $6 billion in its own peptide API manufacturing to ensure semaglutide supply. This infrastructure build-out is one of the defining stories of the pharmaceutical industry in the mid-2020s.

Green chemistry: making synthesis sustainable

Traditional SPPS is environmentally challenging. A single kilogram of purified peptide can require 5,000–10,000 liters of organic solvent (DMF, DCM, NMP, acetonitrile), most of which becomes hazardous waste. The industry is increasingly focused on green chemistry alternatives: solvent recycling and recovery systems (distillation can recover 70–90% of DMF), replacement of DMF with less toxic alternatives (green solvents like Cyrene, GVL, or DMSO blends), water-based synthesis protocols (still early-stage but promising for certain peptide classes), flow chemistry (continuous-flow SPPS reduces solvent volumes and improves heat transfer), and enzymatic synthesis (using enzymes like peptiligases to form peptide bonds under aqueous conditions at ambient temperature).

These sustainability efforts are driven both by regulatory pressure (DMF is classified as a substance of very high concern under EU REACH regulations) and by economic incentives — solvent costs and waste disposal are major expense items in large-scale peptide manufacturing. A 2025 Nature Communications paper demonstrated a water-based protocol for synthesizing short peptides with comparable yields to DMF-based methods, suggesting that greener manufacturing may become practical for commercial peptides within the next decade.

Key terminology

← Browse the Full Dictionary

Part of the PeptideBond.com education network