Volume IX · IX.552 references

General Peptide Monograph · No. GPM 15 · Research Use Only

Manufacturing Therapeutic Peptides Facilities, equipment, inputs, processes, scale-up, and commercial production

A peptide medicine does not become real when a sequence looks right on a screen. It becomes real when a plant — with water systems, solvent farms, reactors or fermenters, chromatography trains, people, and a quality system that can say no — can repeatedly turn qualified inputs into a released batch. This monograph is about that plant and the engineering logic inside it.

Compiled 5 August 2026
References 52 cited
Sources Peer-reviewed literature and named manufacturing guidance
Constraint No human use, dose, route or schedule is recommended anywhere in this document

How to read this document

Read for industrial and scientific understanding: what facilities and equipment classes do, why scale changes process behaviour, where cost and environmental burden concentrate, and how chemical and recombinant routes reshape the plant. Sibling monographs own adjacent lanes — GPM 09 for critical quality attributes and certificate-of-analysis literacy; GPM 14 for who may produce what under which legal regime; and the production-challenges monograph (prompt 16) for impurities, contamination, sterility failures, producer-category failure comparison, and patient risk. Shared words such as “impurity” or “sterile” appear here only as plant and process consequences, not as a second failure atlas. This document does not recommend human use of any peptide, and it does not provide sterile injectable recipes or unlicensed manufacturing procedures.

Part One

From pancreas to plant: how manufacturing understanding relocated

Section 01Manufacturing is a system property

Every therapeutic peptide that reaches a sealed vial has survived more than chemistry. It has survived a schedule. Someone had to qualify amino-acid derivatives or a host cell bank; someone had to keep water systems within specification; someone had to decide whether a chromatographic fraction was good enough to pool; someone had to hold a batch because an environmental monitor was out of trend. The molecule is the product of a system — facilities, utilities, equipment occupancy, materials, people, and disposition authority — not of a single clever coupling.

Figure 1 Manufacturing value stream. Eight linked stages from design to release. Purification, drying, and fill-finish commonly bind the calendar more than synthesis itself.

The controlling idea of this document is therefore blunt. Sequence correctness is not manufacturing completeness. A beautiful laboratory synthesis can be industrially impossible. A recombinant route can solve length and still drown in host-cell impurities. A factory can own every tank and still fail for lack of solvent recovery, lyophilizer capacity, or a quality culture that stops a bad batch. Figure 1 maps the value stream the rest of the Parts unpack: design and route selection, input qualification, assembly or expression, recovery, purification, isolation, formulation and fill-finish, release, and distribution.

Reader’s contract

This monograph explains engineering principles, industrial history and facility logic for research readers. It does not recommend that any person use any peptide, and it specifies no dose, route or schedule. It does not provide sterile-suite recipes, fill volumes, sterilisation cycles or step-by-step instructions for unlicensed manufacture of injectable products. Where literature reports process conditions, those reports are historical and scientific observation, not operational advice. Illustrative cost and occupancy figures are labelled as such and are not product-specific claims.

Scale is not a single number. Milligrams for an assay, grams for toxicology, kilograms for a regional commercial product, and metric-ton ambitions for a global incretin franchise are different plant problems even when the amino-acid sequence is identical. Figure 2 compares five practical bands — research, process development, pilot and clinical, intermediate commercial, and large commercial — without pretending that rigid thresholds fit every potency and patient population.

Figure 2 Scale taxonomy. Mass bands are illustrative. Potency, dose, population, yield, and campaign frequency move a product between rows.

Section 02Tissue extracts and the first industrial peptide problem

The modern peptide industry begins with insulin, and insulin begins as a supply crisis wrapped in a miracle. In the early 1920s, Banting, Best, Collip and Macleod turned pancreatic extracts into a medicine that could reverse diabetic ketoacidosis. The triumph immediately created a manufacturing question that still structures the field: how do you make enough of a fragile biological active, of reproducible potency, when the starting material is an animal organ and the molecule itself has not yet been sequenced?

Early industrial insulin was an extraction-and-purification plant problem before it was a synthetic chemistry problem. Potency was defined operationally — by physiological response — long before primary structure was known. Batch-to-batch variability, species differences, and impurity-driven immunogenicity taught the industry that “natural source” is not a synonym for “simple factory.” The anatomical and physiological understanding of the endocrine pancreas, the medical urgency of diabetes, and the pharmaceutical need for potency units arrived tangled together. That tangle is the ancestor of every later argument about critical quality attributes: activity, identity, and purity are different questions asked of the same vial.

Section 03Merrifield, automation, and the birth of the synthesis suite

Solid-phase peptide synthesis (SPPS), introduced by R. B. Merrifield in the 1960s, relocated the manufacturing problem again. Instead of isolating a hormone from tissue, one could assemble a chain on a resin bead, wash away excess reagents, and — in principle — automate the cycle. Deprotection, washing, activation, coupling, washing: the factory logic was already visible in the laboratory cycle. What looked like glassware was, in embryonic form, a train of solvent deliveries, filtration steps, and in-process checks.

Figure 3 Manufacturing history timeline. The scarce resource relocated from pancreas supply to synthesis automation, recombinant DSP, then commercial reactor and dryer capacity.

SPPS did not abolish difficulty. Incomplete couplings create deletion sequences; aspartimide formation, racemisation, and aggregation on resin punish certain sequences; cleavage and global deprotection generate a crude mixture that must be purified. But it did invent a platform: resins, protected amino acids, coupling reagents, and later Fmoc chemistry that could be purchased, qualified, and scheduled. Automation of that cycle — Merrifield’s second vision after the resin itself — is what made the synthesis suite a recognisable facility type rather than a craft bench.³³ Solvent distribution, waste handling, and eventually large jacketed reactors followed the same logic. Figure 3 places that shift beside later recombinant and capacity-race eras.

Section 04Recombinant insulin relocates the factory

When recombinant human insulin entered commerce in the 1980s, the plant changed species. Fermenters, media kitchens, cell banks, harvest centrifuges, refolding tanks, and endotoxin control joined — or replaced — the solid-phase narrative for that molecule. Downstream processing of recombinant insulin and its analogues from E. coli inclusion bodies illustrates the industrial grammar: expression is only the opening act; capture, cleavage, oxidative folding, and chromatographic polishing determine whether a fermentation becomes an API.⁴⁴

The medical and physiological story is familiar — identical human sequence, reduced immunogenicity concerns relative to some animal insulins — but the manufacturing story is about relocated impurities. Host-cell proteins, host-cell DNA, endotoxin, and misfolded species replace many of the deletion-peptide themes of SPPS. Facility design follows the impurity grammar: biosafety and cleaning regimes, segregation of live and downstream areas, and viral-safety thinking where mammalian systems are used. A recombinant peptide plant is not a synthesis lab with a fermenter wheeled into the corner; it is a different organism of rooms and utilities.

Section 05Hybrid chemistry, incretin scale, and the capacity race

The twenty-first century peptide plant is often hybrid in two senses. First, chemically: long or heavily modified sequences may combine solid-phase fragments, solution couplings, enzymatic or chemical ligation, and late-stage lipidation or conjugation. Platforms that assemble long peptides through interchain strategies show how synthesis engineering still invents new machines when linear SPPS alone becomes punishing.² Peer-reviewed hybrid syntheses of commercial-class incretin peptides — fragment assembly joined by ligation chemistry and assisted by tangential-flow operations — make the same point at manufacturing-relevant scale.²² Second, commercially: originator companies, API specialists, and contract development and manufacturing organisations (CDMOs) share — and compete for — reactor volume, preparative chromatography capacity, and lyophilization slots.

Public announcements in 2025–2026 describe multi-hundred-million to multi-billion currency investments in SPPS trains, purification, lyophilization, and fill-finish across Europe, North America, and Asia. Industry secondary market notes and manufacturer knowledge-centre essays attribute the surge largely to incretin and related metabolic peptide demand and describe API synthesis services and sterile fill-finish as co-binding constraints. Those figures are capacity and investment claims, not measured yields for any named product in this monograph, but they correctly identify the scarce reagents of the decade: reactor volume, purification solvent systems, dryer shelves, aseptic line time, and skilled people. The plant, once again, has become the limiting starting material.

Section 06Four lessons that structure the Parts ahead

History leaves four lessons that the rest of this document treats as design constraints rather than nostalgia.

First, the bottleneck moves. Tissue supply gave way to synthesis automation, then to fermentation DSP, then to chromatography and lyophilizer capacity. Optimising the wrong unit operation is a classic way to lose a year.

Second, impurity grammar follows the route. SPPS, LPPS, hybrid ligation, and recombinant expression create different characteristic impurities and therefore different analytical, cleaning, and facility demands. GPM 09 develops the quality-attribute consequences; here the point is that the building must match the grammar.

Third, scale changes physics. Heat and mass transfer, resin swelling, reagent distribution, hold times, and chromatographic loading are not laboratory habits multiplied by a constant. Part Five returns to this with a bottleneck map.

Fourth, make-versus-buy is a manufacturing decision. Internalising a plant buys control and learning at the price of capital and utilisation risk; outsourcing buys speed and specialised trains at the price of allocation and technology-transfer friction. Part Six treats that choice as engineering economics, not ideology.

What follows is the plant tour those lessons require: routes, facilities, equipment and inputs, process flow and scale-up, then quality systems, people, money, and environmental burden — always at the level of industrial analysis, never as a license to manufacture sterile injectables outside a lawful pharmaceutical quality system.

Part Two

Routes: chemistry, biology, and the industrial choice between them

Section 07Solid-phase peptide synthesis as an industrial machine

SPPS is easiest to understand as a repeating machine. A resin bead carries a growing chain; each cycle removes a temporary protecting group, washes, activates an incoming protected amino acid, couples, washes again, and — when justified — caps unreacted amines before the next cycle. Monitoring may be as simple as a colour test or as instrumented as inline analytics. At the end, cleavage and side-chain deprotection release a crude peptide into a world of scavengers, precipitation solvents, and chromatographic reality.

Figure 4 SPPS cycle. The industrial unit is a completed cycle. Solvent washes and incomplete coupling — not the amide-bond cartoon alone — set plant cost and risk.

Figure 4 is deliberately a cycle diagram, not a recipe. The industrially important points are structural. Resin loading and swelling set how much product a reactor volume can hold and how reagents access the growing chain. Protecting-group strategy (classically Boc or Fmoc chemistries) sets the reagent and solvent vocabulary of the suite. Incomplete coupling and side reactions — deletion sequences, racemisation at sensitive residues, aspartimide formation — write the impurity catalogue that purification must later read. Aggregation on resin can turn a sequence that looks ordinary on paper into a plant crisis.

Solvent use is not a footnote. An ACS Green Chemistry Institute Pharmaceutical Roundtable cross-company compilation assessed process mass intensity (PMI) across forty synthetic peptide processes and reported average SPPS PMI on the order of thirteen thousand kilograms of input per kilogram of product — far above typical small-molecule medians (roughly one hundred sixty-eight to three hundred eight in the same industry framing) and still above the biopharmaceutical average near eighty-three hundred cited in that benchmarking.²⁴ Synthesis and purification dominate the mass burden; solvent hazards include DMF, DMAc, NMP, TFA, and chlorinated or ethereal work-up solvents that force storage, recovery, and permitting to sit beside reactors in any serious SPPS facility conversation. Roundtable solvent-selection guidance exists precisely because solvent choice is a plant decision, not a notebook preference.¹ Wash-minimisation and alternative-solvent research are attempts to move that industrial constraint, not aesthetic green labelling.⁸¹¹

Automation is what turned Merrifield’s single-vessel vision into a recognisable industry. Reviews of automated SPPS for therapeutic peptides note both the diversity of commercial synthesizer architectures and the existence of multikilogram chemical manufacture for selected approved peptides — while longer hormones such as insulin historically preferred recombinant routes.³³ Automation today scales from benchtop parallel instruments to development-scale machines to large solid-phase reactors with jacketed vessels, metering skids, and solvent distribution. Single-use components appear more often in formulation and bioprocess than in classical organic-solvent SPPS, where reusable, cleanable metallurgy and glass still dominate many commercial trains — until a site standardises otherwise. Scale limitations are less about “SPPS cannot be large” and more about heat and mass transfer, filtration times, resin handling, and the purification train waiting downstream.

Section 08Liquid-phase synthesis and the return of isolation

Liquid-phase peptide synthesis (LPPS) rebuilds the chain in solution, often as fragments that are isolated, crystallised, or otherwise purified before further coupling. The industrial appeal is selective: for some sequences and fragment plans, solution chemistry offers impurity control points that SPPS washes cannot match, lower reagent stoichiometry on short fragments, and — in the same ACS Roundtable PMI compilations — sometimes more favourable mass intensity than linear SPPS.²⁴¹¹ The tradeoff is familiar to process chemists: every isolation is a chance to purge impurities and a chance to lose yield, add cycle time, and multiply solvent exchanges. Industry process-strategy notes for late-stage development therefore treat SPPS, LPPS, and hybrid SPPS/LPPS as a portfolio of facility footprints rather than as a single preferred chemistry.

LPPS plants look more like classical fine-chemical API facilities — jacketed reactors, crystallisers, filters, dryers — than like resin-handling SPPS suites. That resemblance is exactly why LPPS can be attractive when existing small-molecule infrastructure can be adapted, and exactly why it is unattractive when fragment strategy is brittle or when SPPS automation already owns the schedule.

Section 09Hybrid routes, fragments, and ligation

Hybrid manufacturing admits that long or heavily modified peptides may be unwise as a single linear SPPS marathon. Chemically synthesised fragments can be condensed; native chemical ligation and related chemistries can join unprotected or selectively protected pieces; enzymatic ligation appears in specialised niches. Convergent strategies reduce the exponential pain of stepwise yield loss, at the cost of process complexity: more intermediates, more analytical methods, more hold points, and more opportunities for mismatch at the ligation junction. Contemporary case literature on gram-scale convergent synthesis of a dual GIP/GLP-1 receptor agonist peptide illustrates the pattern: fragment assembly, ligation or desulfurization chemistry, and tangential-flow operations sit in one manufacturing story rather than in a single linear SPPS campaign.²² Interchain assembly platforms likewise show synthesis engineering inventing machines when length alone punishes stepwise chemistry.²

From a facility perspective, hybrid routes often mean both a synthesis suite and a solution-chemistry or bioconjugation suite, plus purification capable of resolving ligation scars and incompletely joined species. TFF and preparative chromatography become ligation partners, not afterthoughts. The impurity grammar is a collage. That is not an argument against hybrid routes; it is an argument for honest equipment lists and honest tech-transfer packages.

Section 10Recombinant production as a downstream story

Recombinant peptide production begins with host selection and an expression construct, but the plant earns its keep after harvest. Intracellular expression may produce inclusion bodies that require disruption and refolding; secretion may simplify capture and complicate titres; fusion tags may ease isolation and demand a cleavage step that itself creates impurities. Centrifuges, homogenisers, depth filters, capture chromatography, refolding tanks, and polishing trains define the footprint.

Figure 5 Recombinant workflow. Expression is the opening act. Capture, refolding, and polishing determine whether fermentation becomes API — and set the plant footprint.

⁴⁴

Figure 5 emphasises the train, not a host manual. Endotoxin control, host-cell protein and DNA clearance, and — where applicable — viral safety shape both process design and facility segregation. Post-translational modifications that a cell performs (or fails to perform) can decide whether a recombinant route is even eligible. Recombinant plants consume water, cleaning chemicals, and quality-control microbiology at rates that surprise teams who only budgeted fermenter steel.

Section 11Enzymatic and cell-free approaches

Enzymatic synthesis, chemoenzymatic fragment joining, and cell-free translation systems attract attention because they promise selectivity, milder conditions, or speed in research settings. Their current maturity for broad commercial therapeutic peptide manufacturing remains limited relative to SPPS and established recombinant platforms. Potential advantages — reduced protecting-group burden, orthogonal bond formation, rapid prototyping — must be weighed against enzyme supply, lysate cost, scale of reactors, and regulatory familiarity. In this monograph they appear as an emerging equipment and process class, not as a default commercial answer.

Section 12Choosing a route without romance

Figure 6 summarises the industrial choice.

Figure 6 Route comparison. Scores are qualitative industrial fit, not a ranking of molecules. Route choice relocates cost, impurity grammar, and facility type.

Chemical synthesis is often preferable for shorter chains, noncanonical amino acids, and sequences that cells handle poorly. Recombinant production becomes compelling as length and structural complexity grow, when native-like folding is required, or when cellular throughput beats stepwise chemical yield. Hybrid strategies appear when neither pure story fits. Volume matters: a potent peptide with tiny annual API mass can live in a flexible multiproduct suite; a high-dose, high-prevalence product pushes toward dedicated trains and solvent recovery whether the chemistry is elegant or not.

Route selection is therefore simultaneously a chemistry decision, a facility decision, and a supply-chain decision. Changing route late is not a change of flask; it is a change of impurity grammar, cleaning validation, analytical methods, and sometimes of CDMO. The next Part asks what buildings and utilities those route choices require.

Part Three

The factory: rooms, flows, utilities, and archetypes

Section 13Functional areas from gate to waste

A peptide manufacturing facility is a map of risks dressed as rooms. Receiving and quarantine decide what enters. Raw-material warehouses and controlled-temperature storage protect what has been released. Solvent and hazardous-material storage isolate flammability and toxicity. Weighing and dispensing convert inventory into labelled charges. Synthesis suites or fermentation suites do the assembly work. Downstream rooms — buffer preparation, chromatography, tangential-flow filtration — turn crude into bulk. Formulation, aseptic fill-finish, lyophilizer loading, component preparation, packaging, and cold storage turn bulk into finished goods. Quality-control and microbiology laboratories, stability chambers, and retained-sample storage make disposition possible. Utilities, waste handling, solvent recovery, maintenance workshops, gowning, and documentation spaces keep the organism alive.

Figure 7 Facility block flow. A campus is risk-segregated functional areas tied by planned flows. This is a block diagram, not an architectural plan.

Figure 7 is a block-flow diagram, not an architectural plan. The point is adjacency and segregation: solvent farms near synthesis but not casually open to aseptic cores; live fermentation separated from purified downstream; QC samples moving without dragging warehouse dust into Grade A fantasies. Expansion capacity is a first-day design question: the site that cannot add a chromatography suite or a second lyophilizer without breaking flows will rediscover that fact at the worst moment in a product’s life.

Section 14Personnel, material, and waste flows

Contamination control is largely a traffic plan. Personnel move through gowning and airlocks along pressure cascades that make air prefer to flow from cleaner toward less clean. Materials move through pass-throughs and status labels — quarantined, released, rejected — so that a drum cannot socially engineer its way onto the floor. Waste moves out without crossing inbound paths. Figure 8 states the obvious that facilities violate when they grow by improvisation.

Figure 8 Flow segregation. Personnel, materials, and waste each need planned paths. Improvised crossings defeat cleanroom investment.

Multiproduct facilities intensify the problem. Dedicated equipment and suites reduce cross-contamination risk at the price of utilisation. Shared equipment demands cleaning validation, changeover procedures, and sometimes campaign policies that keep high-risk molecules apart in time. None of that is visible in a glossy rendering of a lobby; all of it is visible in the batch record when something goes wrong.

Section 15Utilities: the invisible plant

Reactors are celebrities; utilities are the government. Electrical service, backup generators, and uninterruptible power decide whether a lyophilizer cycle or a cleanroom pressure cascade survives a grid event. Purified water and water for injection (where the process requires them), clean steam, process gases, nitrogen, compressed air, vacuum, chilled water, and heating systems are qualified systems with their own alert/action frameworks. HVAC, dust and vapour extraction, solvent handling and recovery, wastewater treatment, building management systems, environmental monitoring systems, data historians, and cybersecurity close the loop. Cold rooms, freezers, and occasionally ultra-low-temperature storage hold materials and product whose failure modes are silent until a logger tells the truth.

Figure 9 Utilities map. Every synthesis or fill suite is a dependent of qualified utilities. A grid event or WFI failure stops the visible plant.

Figure 9 is a checklist of dependence. A synthesis chemist can describe a coupling in a paragraph; a facility engineer needs pages to describe how the nitrogen blanketing, solvent distribution, and waste segregation make that coupling lawful and repeatable inside a building.

Section 16Cleanrooms, pressure, and hazardous atmospheres

Cleanroom zoning is risk-based theatre with real physics. Aseptic fill-finish cores sit at the high-grade end; gowning and support areas step downward; pressure cascades and airlocks enforce directionality. Pass-throughs reduce door traffic. Occupational exposure controls — containment for potent peptides, local exhaust for solvents and corrosives — coexist with product-protection logic; the two are not identical. Flammable solvent inventories bring explosive-atmosphere design, fire protection, and environmental permitting into the same conversation as gowning philosophy.

This monograph describes those control concepts because they define what “a suitable facility” means at clinical or commercial scale. It does not provide executable zoning drawings or sterilisation recipes for unlicensed sterile manufacture.

Section 17Four facility archetypes

Figure 10 offers four non-operational models.

Figure 10 Facility archetypes. Silhouette scale tracks capital and quality-system depth. These are analytical models, not executable facility designs.

Research peptide laboratory. Milligram-to-gram synthesis, analogue libraries, analytical method development, early formulation tests. Capital burden concentrates in synthesizers, fume extraction, and analytical instruments. Quality systems may be good laboratory practice rather than full commercial GMP. Sterile clinical supply is typically outsourced.

Development and pilot facility. Gram-to-kilogram work, route optimisation, impurity characterisation, engineering batches, toxicology and early clinical API. Limited clean zones, preparative HPLC, TFF, and documented tech-transfer packages appear. Staffing mixes process chemists, engineers, and analytical scientists.

Flexible multiproduct commercial API facility. Multiple-kilogram annual demand across products, campaign manufacturing, partial automation, solvent recovery, and CDMO-like changeover discipline. Major capital sits in synthesis trains, purification, utilities, and cleaning validation. Primary risks are cross-contamination, utilisation swings, and schedule congestion.

Integrated large-scale API, formulation, and fill-finish campus. High-volume API, extensive automation, large purification systems, redundant utilities, on-site or tightly coupled lyophilization and aseptic filling, sometimes device assembly. Capital and operating burdens are dominated by buildings, utilities, specialised equipment lead times, and quality systems that can support global supply. Expansion strategy is modular trains and parallel dryers, not heroic overtime.

Section 18Dedicated lines, shared suites, and expansion

Dedication buys simplicity of cleaning and psychological comfort for regulators and clients; sharing buys capital efficiency. The rational choice tracks toxicity, potency, commercial value, and how often changeovers would eat the calendar. Stainless-steel multi-product trains and single-use strategies (more common in bioprocess and formulation than in heavy organic solvent SPPS) are partial answers to the same question: how much permanent plant should a transient product own?

Expansion capacity is strategy disguised as civil engineering. Sites that leave no room for a second lyophilizer, a solvent-recovery upgrade, or a parallel chromatography hall discover that commercial success can be a facility failure mode. The next Part inventories the machines and materials those halls contain.

Part Four

Equipment and inputs: what the work is done with

Section 19Synthesis and expression equipment

Equipment selection is scale selection. Benchtop and automated parallel synthesizers serve discovery and analogue libraries. Development-scale synthesizers and small jacketed vessels support process chemistry. Pilot and large solid-phase reactors, agitated filter reactors, nutsche filters, pressure-rated vessels, metering pumps, reagent-delivery skids, solvent distribution, inline monitoring, resin handling, cleavage and precipitation vessels, centrifuges, and filter or vacuum dryers populate commercial SPPS trains. Recombinant sites add seed and production bioreactors — single-use or stainless — media preparation, feed systems, harvest centrifuges or depth filtration, homogenisers, microfiltration and ultrafiltration, refolding tanks, and capture chromatography.

Figure 11 Equipment matrix. Selection follows cleanability, heat/mass transfer, and whether the asset will be the campaign bottleneck.

Figure 11 organises classes by stage. Limitations matter as much as capabilities: heat-transfer area, cleanability, pressure drop through resin beds, materials of construction compatible with TFA or caustic CIP, and whether a skid can be multiproduct without becoming a permanent cleaning science project. Automated synthesizer reviews emphasise that commercial instruments differ in solution transfer, mixing, and accessible scale — equipment geometry is therefore part of the process description, not a vendor footnote.³³

Section 20Purification, concentration, isolation, drying

Preparative HPLC and low-pressure chromatography systems, ion exchange, reversed-phase, size-exclusion, and occasional affinity steps form the purification vocabulary. Chromatography skids, column packing, fraction collection, solvent blending, buffer preparation, inline dilution, membrane filtration, and TFF/UF/DF systems move mass between purity states. For synthetic crudes, reversed-phase preparative chromatography is often the workhorse that resolves deletion and closely related impurities; for recombinant insulin-class trains, orthogonal chromatography sequences (size, charge, hydrophobicity) dominate published DSP maps.⁴⁴ Orthogonal ion-exchange ahead of reversed-phase is a recurring industrial pattern when crude complexity or CIP-resistant media can protect expensive RP cycles. Figure 12 and Figure 13 show why purification often owns the schedule: pooling rules, solvent demand, membrane fouling, and hold times are plant clocks.

Figure 12 Preparative chromatography. Pooling rules and solvent demand make prep HPLC a schedule asset. The shaded band marks a conceptual pool window, not a method.

Figure 13 TFF system. TFF concentrates and buffer-exchanges between purification and isolation. Fouling and shear are scale-sensitive.

Isolation and drying — precipitation, centrifugation, agitated nutsche filter dryers, tray or vacuum ovens, occasional spray drying, bulk freeze dryers — convert solutions into stable solids. Continuous chromatography and simulated moving-bed ideas appear where solvent intensity and volume justify the engineering. None of these unit operations is “just work-up” at commercial scale; each is a capital asset with a queue.

Section 21Formulation, fill-finish, and lyophilizers

Formulation equipment includes compounding vessels, mixing and monitoring for pH and conductivity, sterile filtration skids, hold vessels, single-use bags and transfer tubing where appropriate, and controlled-temperature systems. Fill-finish adds vial washing, depyrogenation tunnels, stopper preparation, filling machines (peristaltic or time-pressure among other designs), isolators or restricted-access barrier systems, lyophilizer loading interfaces, stoppering, capping, visual inspection, checkweighing, container-closure integrity testing, labelling, serialization, cartoning, and cold-chain packaging.

Figure 14 Fill-finish line. Primary packing moves through increasing contamination control, then out to inspection and cold chain. Principle-level only.

Figure 15 Lyophilizer anatomy. Chamber, condenser, and vacuum form one capacity-limited asset. Principle-level only — not a cycle recipe.

Figure 15 treats the lyophilizer as chamber, shelves, condenser, vacuum, controls, and cleaning interfaces. Figure 14 orders the line as a contamination-control sequence. Both figures are principle-level. Primary packs — vials, syringes, cartridges, pens — change downstream equipment and device assembly more than they change SPPS chemistry, which is why integrated campuses treat drug-product assets as strategic, not cosmetic. Mid-2020s CDMO expansion journalism and manufacturer capability statements repeatedly pair peptide API growth with aseptic fill-finish and lyophilization investment; secondary market notes even treat formulation and fill-finish as among the faster-growing service segments. Those are industry observations about asset scarcity, not instructions for building an unlicensed sterile suite.

Section 22Analytical and support equipment as infrastructure

Release and characterisation laboratories are part of manufacturing capacity. HPLC/UPLC, LC-MS and high-resolution or tandem MS, amino-acid analysis, capillary electrophoresis, spectroscopic tools, Karl Fischer, osmometry, particle counters, endotoxin systems, microbiological incubators, sterility-testing isolators, environmental-monitoring gear, stability chambers, and container-closure integrity instruments generate the data that batch disposition requires. Support equipment — parts washers, autoclaves, CIP/SIP, balances, material handling, warehouse systems, computerised maintenance, LIMS, MES, electronic batch records — decides whether the plant can prove what it did.

Figure 16 Digital architecture. Wider layers are foundational. Data integrity and cybersecurity are manufacturing controls, not IT afterthoughts.

A site that under-builds analytics creates a hidden bottleneck: reactors idle while methods and investigations catch up. Digital architecture (Figure 16) is therefore drawn as a manufacturing figure, not an IT brochure.

Section 23Material inputs and consumables

Figure 17 divides inputs into chemical synthesis materials (protected amino acids, specialty residues, resins, linkers, coupling reagents, bases, cleavage reagents, scavengers, solvents, chromatographic media, membranes, reference standards), recombinant inputs (hosts, constructs, media, feeds, enzymes, resins, single-use assemblies), formulation inputs (water systems outputs, buffers, bulking and tonicity agents, stabilisers, surfactants, antioxidants, chelators, primary and secondary packaging), and consumables (tubing, filters, PPE, garments, wipes, disinfectants, EM plates, labels).

Figure 17 Material taxonomy. Each leaf carries grade, supplier qualification, storage, testing, shelf life, and supply-risk questions.

For each class the industrial questions repeat: function, quality grade, supplier qualification, storage, testing, shelf life, supply risk, and impact on cost and yield. A resin lot that swells differently, a protected amino acid with a chiral impurity, or a stopper formulation that adsorbs peptide can rewrite a validated process. Supplier qualification is not paperwork theatre; it is how the plant extends its quality system beyond the fence line.

Section 24Multipurpose assets and product-specific traps

Balances, many chromatography skids, autoclaves, and utilities are multipurpose. Dedicated HPLC columns, product-specific device assembly tooling, certain conjugation reactors, and sometimes entire suites become product-specific because cleaning cannot convincingly erase the last campaign or because geometry only fits one presentation. The cost of idle dedicated assets is a strategic input to make-versus-buy decisions. The next Part follows a batch through the value stream and asks where time, mass, and money disappear when scale changes.

Part Five

Process flow, yield loss, and why scale-up changes the chemistry

Section 25The pipeline from sequence to distribution

Figure 18 compresses the commissioned end-to-end list into decision stages: sequence and route selection; raw-material qualification; synthesis or expression; crude recovery; cleavage or harvest; purification; concentration; folding or conjugation when required; isolation and drying; API characterisation; formulation; sterile filtration; filling; lyophilization where applicable; inspection; packaging; QC testing; batch review; release; storage; distribution; stability monitoring.

Figure 18 Process-flow map. A manufacturing system is a chain of operate / measure / decide loops. Missing the decision point is how bad batches become shipped batches.

At each stage the same checklist applies even though the equipment changes: objective; major equipment class; major inputs; critical quality concerns; typical failure modes; scale-up issues; environmental burden; cost drivers; data generated; decision points. A plant that cannot name the decision point — pool or reject, reprocess or scrap, release or investigate — is not yet a manufacturing system.

Section 26Where yield and calendar time go to die

Mass disappears into incomplete couplings, cleavage losses, precipitation mother liquors, chromatographic side cuts, membrane hold-up, adsorption to surfaces, and rejected fractions. Time disappears into long coupling or refolding holds, filtration, column equilibration, lyophilizer cycles, cleaning, environmental monitoring waits, and analytical release. Environmental burden concentrates where solvents and aqueous washes multiply — exactly where PMI studies find synthesis and purification dominating peptide process mass.²⁴

Yield loss is not only a chemist’s sorrow; it is a capacity multiplier. A process that loses half its mass in polishing needs twice the upstream reactor time for the same released grams. That is why purification strategy and route selection are economic documents as much as scientific ones.

Section 27Scale-up is not multiplication

Figure 19 lists the usual offenders: mixing and heat transfer gradients; mass transfer into swollen resin; reagent distribution; reaction kinetics that looked instantaneous in a speck of bead; incomplete coupling that was invisible at milligram scale; pressure drop; filtration time; precipitation behaviour that changes with tip speed and cooling rate; impurity profiles that shift when hold times lengthen; solvent volumes that break waste permits; column loading versus resolution; product adsorption; oxygen exposure; shear in TFF or homogenisation; refolding windows; microbial risk in aqueous holds; batch size versus cycle time; cleaning and campaign planning; equipment occupancy and facility scheduling.

Figure 19 Scale-up bottlenecks. Bars are relative industrial emphasis, not measured cycle times for a named product. Transport phenomena rewrite lab processes.

Illustrative process parameters in development reports are not universal setpoints. A coupling excess that is casual on the bench can be a solvent and cost crisis in a thousand-litre solvent inventory. Comparability after scale change is a scientific argument, not a hope expressed in a slide.

Section 28Process development, CPP/CMA/CQA, and PAT

Process development is how a route becomes a controllable plant story: route scouting; design of experiments; critical process parameters; critical material attributes; critical quality attributes; process analytical technology; impurity fate and purge maps; mass balance; yield optimisation; solvent and resin selection; coupling efficiency studies; purification strategy; robustness; scale-down models; technology transfer; engineering batches; validation batches; continuous process verification. ICH pharmaceutical development, quality risk management, and drug-substance development guidances (Q8, Q9, Q11) supply the shared vocabulary for that work.¹⁹²⁰¹⁸

PAT is not a gadget tax. Inline or at-line information that shortens cycles or prevents failed batches earns its keep; dashboards that nobody trusts do not. Scale-down models are only as good as the question they were built to answer — mixing, chromatography loading, and hold-time effects are different questions.

Section 29Technology transfer and the batches that teach

Engineering batches teach the plant; clinical and validation batches teach the file. Technology transfer between a sponsor and a CDMO, or between two internal sites, multiplies differences in water quality, raw-material grades, equipment geometry, and analyst practice. A transfer package that ships only a notebook synthesis is incomplete. Cleaning validation, analytical method transfer, and materials of construction belong in the same crate as the coupling description.

ICH Q7 frames GMP expectations for APIs across chemical and biological routes and reminds manufacturers that stringency increases as the process approaches the finished API.¹⁷ That gradient matters for where facilities invest in segregation and documentation density. For synthetic peptides that refer to listed drugs of recombinant origin, FDA industry guidance on highly purified synthetic peptide ANDAs sharpens CMC attention on starting-material quality, impurity characterisation, and analytical comparability — documentation expectations that feed directly into what a manufacturing and analytical plant must be able to prove.⁴⁹ Detailed CQA and CoA literacy remain in GPM 09; here the point is that regulatory CMC framing is itself a facility and equipment requirement list.

Section 30Batch time and equipment occupancy

Figure 20 shows illustrative relative occupancy across synthesizer or reactor time, recovery, preparative chromatography, TFF and isolation, lyophilization or fill-finish, and QC release.

Figure 20 Equipment occupancy. Overlapping bars show why dryer, prep HPLC, and QC queues — not only reactors — set campaign length. Illustrative only.

The bars are pedagogical, not a priced schedule for a named product. Their lesson is scheduling: the binding constraint is often the dryer, the prep HPLC, or the QC queue, not the glamorous reactor. Campaign planning is the art of not letting cleaning and changeover steal the year.

Part Six

Quality systems, people, money, and the environment of the plant

Section 31Quality systems as manufacturing infrastructure

Good laboratory practice, good manufacturing practice, data integrity, change control, deviations, corrective and preventive action, out-of-specification and out-of-trend investigations, supplier qualification, process validation, cleaning validation, analytical method validation, computerised-system validation, environmental monitoring, aseptic process simulation, batch records, product-quality review, recalls, complaint handling, and stability programs are not an appendix bolted onto a reactor. They are how the plant remembers and how it refuses. ICH Q7, Q8, Q9, Q10, and Q11 name much of that operating system.¹⁷²¹¹⁹²⁰¹⁸

GPM 09 develops what release analytics can and cannot claim. The production-challenges monograph (prompt 16) develops what goes wrong when those systems fail — impurity and contamination pathways, sterility escapes, producer-category differences, recalls and warning letters, and patient risk. Here the emphasis is organisational infrastructure only: without disposition authority and documented evidence, equipment is only expensive sculpture. Aseptic process simulation and environmental monitoring are named as facility capabilities that define clinical and commercial suitability — not as failure trees, clinical case studies, or instructions for building an unlicensed sterile suite.

Section 32Personnel across scales

Peptide chemists, process chemists, chemical and biochemical engineers, fermentation and purification scientists, formulation scientists, analytical chemists, microbiologists, QC and QA personnel, validation and automation engineers, maintenance technicians, EHS staff, warehouse personnel, production operators, project managers, regulatory-affairs professionals, supply-chain managers, facility engineers, and IT/cybersecurity staff all appear somewhere between research and commercial scale. Research teams are chemist- and analyst-heavy. Pilot plants add engineers and documentation discipline. Commercial sites add shift cultures, maintenance depth, and quality units with real stop-the-line power.

Staffing is itself a capacity constraint. A lyophilizer without trained operators and a QC laboratory without reviewers is a parked asset. Training and knowledge management are therefore manufacturing controls.

Section 33Capital-cost framework

Capital accumulates in land and building shell; cleanrooms; utilities; synthesis or fermentation equipment; purification; lyophilizers; fill-finish lines; laboratories; environmental controls; solvent storage; waste treatment; automation; validation and commissioning; contingency; and expansion capacity left unused on purpose. Public CDMO and API-manufacturer announcements in 2025–2026 illustrate order-of-magnitude investment programs — hundreds of millions to more than a billion in stated currency for multi-year expansions — but those figures mix geography, scope, and marketing. This monograph does not convert press releases into a universal CAPEX calculator. It insists that any serious estimate state year, geography, capacity assumptions, and inclusions.

Section 34Operating costs and cost of goods

Figure 21 illustrates common operating-cost centres: protected amino acids and resins; solvents and recovery burden; chromatography media and filters; labour and quality systems; utilities and waste; fill-finish, lyophilization, and packaging; yield loss, rejects, and idle capacity.

Figure 21 Cost-of-goods waterfall. Relative shares illustrate common cost centres discussed in open PMI and industry analyses — not a bill of materials for any product.

Relative bars are illustrative. Open PMI literature shows why solvents punch above their casual reputation in synthetic peptide COGS discussions.²⁴ Depreciation, contract services, cold chain, and testing join the bill whether or not a slide acknowledges them.

Minimum economic batch size is the quiet tyrant of peptide portfolios. A process that is elegant at five grams may be unaffordable at the batch size a commercial forecast implies — or, conversely, a high-potency product may never need the mega-train a capacity headline celebrates.

Section 35Build versus buy, campaigns, and supply risk

Figure 22 frames internalization against outsourcing.

Figure 22 Build versus buy. Choose internalization when volume and learning justify capital; choose CDMOs when speed and specialised trains dominate.

Build when volume, learning, and supply security justify capital and when the organisation can staff the quality system. Buy when speed, flexibility, or specialised trains dominate — and when quality agreements, capacity reservation, and tech-transfer competence are real. Dedicated versus multiproduct, stainless versus single-use, batch versus continuous, local versus offshore sourcing, and small flexible plants versus large centralised plants are variations on the same theme: utilisation, risk, and learning.

The cost of idle capacity and the cost of supply failure are both real. Peptide markets in the mid-2020s have demonstrated that reactor volume, purification solvent systems, and dryer shelves can become strategic commodities — especially where metabolic-disease peptide demand collides with long equipment lead times. Secondary industry estimates of CDMO market growth, API-service share, and fill-finish bottlenecks should be read as estimates with year stamps and vendor incentives, not as laboratory measurements. Open PMI compilations remain the better quantitative window into why solvents and purification dominate synthetic-peptide mass and cost discussions even when press releases celebrate reactor counts.²⁴

Section 36Environment, occupation, and the standing constraint

Figure 23 gathers solvent consumption, hazardous reagents, worker exposure, fire and explosion risk, corrosive materials, toxic and aqueous waste, emissions, recovery, water and energy use, lyophilizer demand, and greener chemistry opportunities.

Figure 23 Environmental impact map. ACS GCIPR member-company PMI compilations show synthetic peptide processes with very high process mass intensity, largely from solvents.

Fluorinated and other persistent materials appear where process chemistry uses them; they are called out as stewardship topics, not as a complete regulatory inventory. Occupational hygiene and emergency systems are part of facility suitability.

Process intensification, alternative solvents, improved atom economy, and wash reduction are active research and industrial programmes precisely because peptide PMI and hazard profiles are difficult.⁸¹¹⁴⁶¹ Stewardship that damages CQAs is not stewardship; quality that ignores waste and worker risk is not a complete manufacturing ethic.

Standing constraint

No human use, dose, route or schedule is recommended anywhere in this document. No unsafe operational sterile-manufacturing SOP is provided. Research-use-only materials remain research-use-only regardless of how complete their manufacturing story appears. Industrial literacy is not a licence to manufacture.

The controlling idea returns. A therapeutic peptide is manufactured when a system — not a sequence — can repeatedly deliver a released batch. History moved the bottleneck from pancreas supply to automated synthesis to recombinant DSP to chromatography, dryers, and global capacity. Routes relocate impurity grammar and facility type. Scale changes physics. Make-versus-buy allocates capital and risk. The Apparatus that follows gathers glossary, abbreviations, equipment and materials appendices, utilities and quality-system summaries, assumptions, and the references that pin the claims.

Apparatus

Glossary, appendices, assumptions, and references

Section A1References

The list below is numbered and sorted by first-author surname (or issuing body). Journal articles were resolved from source-record metadata where harvested; named ICH guidances and selected open sustainability papers are cited by official designation or DOI. In-text citations are the superscript numbers throughout the document.

ACS Green Chemistry Institute Pharmaceutical Roundtable. ACS GCI Pharmaceutical Roundtable Solvent Selection Guide. ACS GCIPR. 2011.
AlShaer D, Al Musaimi O, Williams DR. Application of the OBIMAP (One-Bead Interchain Multipeptide Assembly Platform) to Long Peptide Synthesis: Liraglutide as a Case Study. ACS Omega. 2026;11(13):21369.
PMID 41970854 · doi:10.1021/acsomega.6c00648 · PMC13063175
Bagherinejad MR, Sadeghi HM, Abedi D, Chou CP, Moazen F, Rabbani M. Twin arginine translocation system in secretory expression of recombinant human growth hormone. Research in Pharmaceutical Sciences. 2016;11(6):461.
PMID 28003839 · doi:10.4103/1735-5362.194871 · PMC5168882
Bai M, Wang N, Zhang M, Zhao Z, Yu J, She W, et al.. Deep eutectic solvent-mediated extraction of collagen peptides from sturgeon skin: Molecular-weight control and mechanistic insights. Food Chemistry: X. 2026;37:104078.
PMID 42317277 · doi:10.1016/j.fochx.2026.104078 · PMC13273213
Baretić D, Missoury S, Patel K, Martinez M, Coste F, Zhu K, et al.. Filament formation and NAD processing by noncanonical human FAM118 sirtuins. Nature Structural & Molecular Biology. 2025;32(12):2526.
PMID 41249820 · doi:10.1038/s41594-025-01715-1 · PMC12700824
Boyd JD, Wang S, Lin H, Hsieh Y, Sun YS, Thibodeaux BA, et al.. Preclinical characterization of an active immunotherapy targeting calcitonin gene-related peptide. Communications Medicine. 2025;5:145.
PMID 40301574 · doi:10.1038/s43856-025-00870-2 · PMC12041250
Chrudinová M, Kirk NS, Chuard A, Venugopal H, Zhang F, Lubos M, et al.. A viral insulin-like peptide inhibits IGF-1 receptor phosphorylation and regulates IGF1R gene expression. Molecular Metabolism. 2024;80:101863.
PMID 38182007 · doi:10.1016/j.molmet.2023.101863 · PMC10831276
Collins JM, Singh SK, White TA, Cesta DJ, Simpson CL, Tubb LJ, et al.. Total wash elimination for solid phase peptide synthesis. Nature Communications. 2023;15:8168.
PMID 38071224 · doi:10.1038/s41467-023-44074-5 · PMC10710472
Eladl O. Biophysical and transcriptomic characterization of LL-37-derived antimicrobial peptide targeting multidrug-resistant Escherichia coli and ESKAPE pathogens. Scientific Reports. 2025;15:36126.
PMID 41102328 · doi:10.1038/s41598-025-22890-7 · PMC12533254
Fan Z, Zhang X, Ge X, Zhu L, Zhu M, Yang L. Biosynthesis of polyamine-polyphosphate granules for colitis alleviation. iScience. 2026;29(2):114679.
PMID 41660257 · doi:10.1016/j.isci.2026.114679 · PMC12874136
Ferrazzano L, et al.. Sustainability in peptide chemistry: current synthesis and purification technologies and future challenges. Green Chemistry. 2022;24.
doi:10.1039/d1gc04387k
Ghade NS, Thappa DK, Lona J, Krishnan AR, Sonar SM. Comparative physicochemical and structural characterisation studies establish high biosimilarity between BGL-ASP and reference insulin aspart. Scientific Reports. 2024;14:4224.
PMID 38378730 · doi:10.1038/s41598-024-54819-x · PMC10879530
Giles MB, Hong JKY, Liu Y, Tang J, Li T, Beig A, et al.. Efficient aqueous remote loading of peptides in poly(lactic-co-glycolic acid). Nature Communications. 2022;13:3282.
PMID 35676271 · doi:10.1038/s41467-022-30813-7 · PMC9177552
Han J, Meade J, Devine D, Sadeghpour A, Rappolt M, Goycoolea FM. Chitosan-coated liposomal systems for delivery of antibacterial peptide LL17-32 to Porphyromonas gingivalis. Heliyon. 2024;10(14):e34554.
PMID 39149035 · doi:10.1016/j.heliyon.2024.e34554 · PMC11325287
He L, Feng D, Guo H, Zhou Y, Li Z, Zhang K, et al.. Pharmacokinetics, distribution, metabolism, and excretion of body-protective compound 157, a potential drug for treating various wounds, in rats and dogs. Frontiers in Pharmacology. 2022;13:1026182.
PMID 36588717 · doi:10.3389/fphar.2022.1026182 · PMC9794587
Hu X, Guo L, Ding Y, Nie S, Fang J, Li J, et al.. Self-assembling peptide inspired by insulin and type 1 insulin-like growth factor for the treatment of androgenetic alopecia. Bioactive Materials. 2025;53:819.
PMID 40822304 · doi:10.1016/j.bioactmat.2025.08.004 · PMC12354804
International Council for Harmonisation. ICH Q7: Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients. ICH Harmonised Tripartite Guideline. 2000.
International Council for Harmonisation. ICH Q11: Development and Manufacture of Drug Substances (Chemical Entities and Biotechnological/Biological Entities). ICH Harmonised Guideline. 2012.
International Council for Harmonisation. ICH Q8(R2): Pharmaceutical Development. ICH Harmonised Guideline. 2009.
International Council for Harmonisation. ICH Q9(R1): Quality Risk Management. ICH Harmonised Guideline. 2023.
International Council for Harmonisation. ICH Q10: Pharmaceutical Quality System. ICH Harmonised Guideline. 2008.
Jalan A, Murzinski ES, Jansen PJ, Miller RD, Embry MC, Scherer RB, et al.. A Convergent Hybrid Gram-Scale Synthesis of Tirzepatide: Tangential Flow Filtration Assisted Native Chemical Ligation-Desulfurization Approach. Angewandte Chemie International Edition. 2026:e20060.
PMID 41437654 · doi:10.1002/anie.202520060
Jarmusch SA, Muhammad T, Göransson U, Strömstedt AA. α‑Helical Peptides Encoded in Collagen Exhibit Antimicrobial Activity with Low Cytotoxicity. Journal of Natural Products. 2026;89(1):242-250.
PMID 41528266 · doi:10.1021/acs.jnatprod.5c01318 · PMC12836314
Kekessie I, Wegner K, Martinez I, Kopach ME, White TD, Tom JK, et al.. Process Mass Intensity (PMI): A Holistic Analysis of Current Peptide Manufacturing Processes Informs Sustainability in Peptide Synthesis. The Journal of Organic Chemistry. 2024;89(7):4261-4282.
PMID 38508870 · doi:10.1021/acs.joc.3c01494 · PMC11002941
Kirk NS, Chen Q, Wu YG, Asante AL, Hu H, Espinosa JF, et al.. Activation of the human insulin receptor by non-insulin-related peptides. Nature Communications. 2022;13:5695.
PMID 36171189 · doi:10.1038/s41467-022-33315-8 · PMC9519552
Knadler MP, Nguyen T, Campanale K, De Veer MJ, Beals JM, Li S, et al.. Addition of 20-kDa PEG to Insulin Lispro Alters Absorption and Decreases Clearance in Animals. Pharmaceutical Research. 2016;33(12):2920-2929.
PMID 27528391 · doi:10.1007/s11095-016-2014-1 · PMC5093203
Kong L, Li X, Liu T, Yao Q, Qin J. Harnessing lactic acid bacteria for nicotinamide mononucleotide biosynthesis: a review of strategies and future directions. Frontiers in Microbiology. 2024;15:1492179.
PMID 39735184 · doi:10.3389/fmicb.2024.1492179 · PMC11681623
Lee S, Han J, Lee E, Kim YK, Hwang S, Lee S, et al.. Structural and functional characterization of recombinant human growth hormone isolated from transgenic pig milk. PLoS ONE. 2020;15(7):e0236788.
PMID 32735629 · doi:10.1371/journal.pone.0236788 · PMC7394428
Li Q, Zhang Q, Kim Y, Gaddam RR, Jacobs JS, Bachschmid MM, et al.. RETRACTED ARTICLE: Deficiency of endothelial sirtuin1 in mice stimulates skeletal muscle insulin sensitivity by modifying the secretome. Nature Communications. 2023;14:5595.
PMID 37696839 · doi:10.1038/s41467-023-41351-1 · PMC10495425
Li Y, Liu W, Liu D, Wang R, Zhang Y, Li X, et al.. Deciphering the significant impact of natural glycosylation on human insulin. Acta Pharmaceutica Sinica. B. 2025;15(11):5880-5890.
PMID 41311403 · doi:10.1016/j.apsb.2025.08.005 · PMC12648096
Lin N, Zheng N, Purushottam L, Zhang YW, Chou DH. Synthesis and Characterization of Phenylboronic Acid-Modified Insulin With Glucose-Dependent Solubility. Frontiers in Chemistry. 2022;10:859133.
PMID 35372263 · doi:10.3389/fchem.2022.859133 · PMC8965884
Ma T, Tian X, Zhang B, Li M, Wang Y, Yang C, et al.. Low-dose metformin targets the lysosomal AMPK pathway through PEN2. Nature. 2022;603(7899):159.
PMID 35197629 · doi:10.1038/s41586-022-04431-8 · PMC8891018
Made V, Els-Heindl S, Beck-Sickinger AG. Automated solid-phase peptide synthesis to obtain therapeutic peptides. Beilstein Journal of Organic Chemistry. 2014;10:1197-1212.
PMID 24991269 · doi:10.3762/bjoc.10.118 · PMC4077397
Matinfar M, Heidari B, Valipour A, Rahdari T, Ramezanpour S, Hadavi M, et al.. Superior growth performance in carp fry achieved with chitosan-alginate encapsulated A-ghrelin versus free peptide: Evidence from physiological, molecular, and morphological analyses. PLOS One. 2025;20(6):e0327235.
PMID 40587525 · doi:10.1371/journal.pone.0327235 · PMC12208430
Mendel HC, Hopping G, Undheim EAB, Zuegg J, Lewis RJ, Forbes BE, et al.. PeptideMiner—neuropeptide discovery across the animal kingdom. GigaScience. 2025;14:giaf078.
PMID 40796375 · doi:10.1093/gigascience/giaf078 · PMC12343078
Moreau F, Kirk NS, Zhang F, Gelfanov V, List EO, Chrudinová M, et al.. Interaction of a viral insulin-like peptide with the IGF-1 receptor produces a natural antagonist. Nature Communications. 2022;13:6700.
PMID 36335114 · doi:10.1038/s41467-022-34391-6 · PMC9637144
Qian X, Dai Y, Li C, Pan J, Xu J, Mu B. Enzymatic synthesis of high-titer nicotinamide mononucleotide with a new nicotinamide riboside kinase and an efficient ATP regeneration system. Bioresources and Bioprocessing. 2022;9(1):26.
PMID 38647612 · doi:10.1186/s40643-022-00514-6 · PMC10992250
Ren Y, Yang X, Ding L, Liu D, Tao Y, Huang J, et al.. Adaptive evolutionary strategy coupled with an optimized biosynthesis process for the efficient production of pyrroloquinoline quinone from methanol. Biotechnology for Biofuels and Bioproducts. 2023;16:11.
PMID 36658601 · doi:10.1186/s13068-023-02261-y · PMC9851590
Ryakhovsky VV, Khachiyan GA, Kosovova NF, Isamiddinova EF, Ivanov AS. The first preparative solution phase synthesis of melanotan II. Beilstein Journal of Organic Chemistry. 2008;4:39.
PMID 19043625 · doi:10.3762/bjoc.4.39 · PMC2587946
Schuster S, Biri-Kovács B, Szeder B, Farkas V, Buday L, Szabó Z, et al.. Synthesis and in vitro biochemical evaluation of oxime bond-linked daunorubicin–GnRH-III conjugates developed for targeted drug delivery. Beilstein Journal of Organic Chemistry. 2018;14:756-771.
PMID 29719573 · doi:10.3762/bjoc.14.64 · PMC5905287
Selis F, Schrepfer R, Sanna R, Scaramuzza S, Tonon G, Dedoni S, et al.. Enzymatic mono-pegylation of glucagon-like peptide 1 towards long lasting treatment of type 2 diabetes. Results in Pharma Sciences. 2012;2:58-65.
PMID 25755995 · doi:10.1016/j.rinphs.2012.09.001 · PMC4167179
Sen S, Ali R, Onkar A, Verma S, Ahmad QT, Bhadauriya P, et al.. Synthesis of a highly thermostable insulin by phenylalanine conjugation at B29 Lysine. Communications Chemistry. 2024;7:161.
PMID 39043846 · doi:10.1038/s42004-024-01241-z · PMC11266353
Shin HJ, Bak SU, La HN, Kang JS, Lee HH, Eom HJ, et al.. Efficient transdermal delivery of functional protein cargoes by a hydrophobic peptide MTD 1067. Scientific Reports. 2022;12:10853.
PMID 35760980 · doi:10.1038/s41598-022-14463-9 · PMC9237094
Siew YY, Zhang W. Downstream processing of recombinant human insulin and its analogues production from E. coli inclusion bodies. Bioresources and Bioprocessing. 2021;8(1):65.
PMID 34336550 · doi:10.1186/s40643-021-00419-w · PMC8313369
Singh RV, Kalia VC, Sambyal K, Singh B, Kumar A, Lee J. Enzymatic approaches to nicotinic acid synthesis: recent advances and future prospects. Frontiers in Bioengineering and Biotechnology. 2025;13:1585736.
PMID 40534613 · doi:10.3389/fbioe.2025.1585736 · PMC12174410
Tatsumi T, Sasamoto K, Matsumoto T, Hirano R, Oikawa K, Nakano M, et al.. Practical N-to-C peptide synthesis with minimal protecting groups. Communications Chemistry. 2023;6:231.
PMID 37884638 · doi:10.1038/s42004-023-01030-0 · PMC10603086
Teli B, Somani S, Arya V, Rajput PJ, Kahale P, Bharmal V, et al.. Semaglutide and Follow-On Peptide Therapeutics: Balancing Innovation, Regulation, and Clinical Outcomes. Cureus. 2026;18(5):e108999.
PMID 42311743 · doi:10.7759/cureus.108999 · PMC13269946
Trier S, Linderoth L, Bjerregaard S, Andresen TL, Rahbek UL, Castanho MARB. Acylation of Glucagon-Like Peptide-2: Interaction with Lipid Membranes and In Vitro Intestinal Permeability. PLoS ONE. 2014;9(10):e109939.
PMID 25295731 · doi:10.1371/journal.pone.0109939 · PMC4190408
U.S. Food and Drug Administration. ANDAs for Certain Highly Purified Synthetic Peptide Drug Products That Refer to Listed Drugs of rDNA Origin: Guidance for Industry. FDA Guidance for Industry. 2021.
Van Gelder TG, Vergoossen D, Zonnenberg IA, Lalmohamed A, Koole-Oostveen MI, van Reij EML, et al.. Physical compatibility of insulin aspart, lidocaine, alprostadil and vancomycin with individualised two-in-one parenteral nutrition used in the neonatal intensive care unit. European Journal of Hospital Pharmacy. 2025;33(1):e004300.
PMID 39863397 · doi:10.1136/ejhpharm-2024-004300 · PMC12772616
Winstel R, Wieland J, Gertz B, Mueller A, Allgaier H. Manufacturing of Recombinant Human Follicle-Stimulating Hormone Ovaleap® (XM17), Comparability with Gonal-f®, and Performance/Consistency. Drugs in R&D. 2017;17(2):305-312.
PMID 28386738 · doi:10.1007/s40268-017-0182-z · PMC5427053
Zhao W, Zhang R, Zhou L, Zhang Z, Du F, Wu R, et al.. Construction and optimization of a genetic transformation system for efficient expression of human insulin-GFP fusion gene in flax. Bioresources and Bioprocessing. 2024;11(1):83.
PMID 39190215 · doi:10.1186/s40643-024-00799-9 · PMC11349960

Section A2Scope relative to sibling monographs

This monograph focuses on manufacturing systems, facilities, equipment, scale-up and commercial production logic. Sibling monographs cover adjacent lanes that are not duplicated here: GPM 09 (quality verification and CoA literacy); GPM 14 (producer-category law and supply channels); and production prompt 16, When Peptide Production Goes Wrong (impurities, contamination, sterility failures, enforcement lessons, and patient risk). Brief mentions of impurity grammar or aseptic capability in this text are plant-design consequences, not a second failure monograph. See notes/OVERLAP_GPM16.md.

Section A3Glossary

API. Active pharmaceutical ingredient; the bulk drug substance before finished dosage form manufacture.

Campaign manufacturing. Running one product for a defined period on shared equipment, then cleaning and changing over.

CDMO / CMO. Contract development and manufacturing organisation / contract manufacturing organisation.

CIP / SIP. Clean-in-place / steam-in-place.

CPP / CMA / CQA. Critical process parameter / critical material attribute / critical quality attribute.

LPPS. Liquid-phase peptide synthesis.

PAT. Process analytical technology.

PMI. Process mass intensity; total mass of materials used to produce a unit mass of product.

RABS. Restricted-access barrier system.

SPPS. Solid-phase peptide synthesis.

TFF / UF / DF. Tangential-flow filtration / ultrafiltration / diafiltration.

WFI. Water for injection.

Section A4Abbreviations

API, CAPA, CAPEX, CCI, CDMO, CIP, CMO, COGS, CPP, CQA, DSP, EMA, EMS, FDA, GMP, HCP, HPLC, HVAC, ICH, LC-MS, LIMS, LPPS, MES, OOS, OOT, OPEX, PAT, PMI, QA, QC, RABS, SIP, SPPS, TFF, UF/DF, UPS, WFI.

Section A5Equipment appendix (summary)

Stage	Equipment classes	Selection notes
Synthesis / expression	Benchtop to large SPPS reactors; jacketed vessels; bioreactors	Heat/mass transfer, cleanability, solvent compatibility
Purification	Prep HPLC, IEX/SEC, chromatography skids, TFF	Often schedule-binding; solvent intensity
Isolation / drying	Nutsche filter dryers, vacuum/tray dryers, bulk lyo	Lead times; energy demand
Fill-finish	Washers, tunnels, fillers, isolators/RABS, cappers, inspection	Contamination control; presentation-specific
Support / digital	CIP/SIP, autoclaves, MES, LIMS, historians	Data integrity is a control

Section A6Raw materials and consumables appendix

Chemical synthesis inputs: protected amino acids, specialty residues, resins, linkers, coupling reagents, activators, bases, deprotection and cleavage reagents, scavengers, solvents, chromatographic media, membranes, reference standards. Recombinant inputs: hosts, constructs, media, feeds, enzymes, resins, single-use assemblies. Formulation inputs: water-system outputs, buffers, bulking/tonicity agents, stabilisers, surfactants, packaging components and devices. Consumables: tubing, filters, PPE, garments, disinfectants, EM plates, labels. Each class requires grade, supplier qualification, storage, testing, shelf-life, and supply-risk controls.

Section A7Utilities appendix

Electrical service with backup generation and UPS; purified water and WFI where required; clean steam; process gases including nitrogen; compressed air; vacuum; chilled water and heating; HVAC with pressure cascades and extraction; solvent storage and recovery; wastewater and emissions controls; BMS/EMS; data historians and cybersecurity; cold rooms and freezers. Utility qualification and monitoring are part of facility suitability at clinical and commercial scale.

Section A8Quality-system appendix

Core elements referenced in the body: GMP/GLP; data integrity; change control; deviations and CAPA; OOS/OOT; supplier qualification; process, cleaning, analytical, and computerised-system validation; environmental monitoring; aseptic process simulation; batch records; product-quality review; stability; recalls and complaints. Binding instruments cited by name include ICH Q7, Q8(R2), Q9(R1), Q10, and Q11. Detailed CQA and CoA literacy remain in GPM 09.

Section A9Assumptions and estimation register

A1. Printed-page equivalents use 500 words of extracted body text per page (series convention).

A2. PMI averages cited from ACS GCIPR member-company compilations (Kekessie et al., forty synthetic peptide processes; SPPS average near thirteen thousand kg/kg) are portfolio metrics across peptides and phases, not a constant for every process.²⁴

A2b. 2025–2026 API research refresh used Tavily, Exa, Firecrawl, and Europe PMC against C:\Apps\_env keys; harvest notes live under notes/api_research/ and notes/API_EVIDENCE_PACKET.md.

A3. Cost-of-goods and occupancy figures in graphics are illustrative relative patterns, not priced bills of materials.

A4. CDMO capacity and investment announcements are secondary industry claims with year and scope uncertainty; they are used only to identify binding constraints (reactors, purification, lyo, fill-finish), not as proprietary yield tables.

A5. Facility archetypes are analytical models, not construction specifications.

Section A10Unresolved engineering questions

How quickly wash-minimised and alternative-solvent SPPS will displace classical solvent PMI at true commercial tonnage remains partly open. Continuous and hybrid chromatography adoption rates vary by firm and molecule. Public, comparable CAPEX benchmarks for peptide fill-finish lines by geography and presentation are thin. The right long-term balance between dedicated incretin-scale trains and flexible multiproduct suites will be answered by utilisation data that are mostly private. Those gaps are labelled rather than filled with invented precision.

Continue exploring

Related chapters

Browse Companion: Peptide-Adjacent Compounds, Formulation & Laboratory Practice →

Volume IX · IX.1Manufacturing and Verifying Therapeutic PeptidesSynthesis, purity, potency, and product quality Volume IX · IX.4The Therapeutic Peptide Production EcosystemPharmaceutical Companies, Compounders, Contract Manufacturers, and Alternative Supply Channels Volume IX · IX.7Testing Therapeutic PeptidesAnalytical methods, certificates of analysis, independent verification, and laboratory reliability

← Volume IX contents Full series index