Biologics are made by living cells, not by chemical reactions in a flask. A manufacturer engineers a cell line to produce the desired protein, banks frozen stocks of those cells, grows them in carefully controlled bioreactors, harvests the protein they secrete, purifies it through multiple chromatography and filtration steps, confirms it is free of viruses and contaminants, formulates it for stability, and fills it into vials or syringes under sterile conditions. The entire journey from a single engineered cell to a finished injectable product can take months and involves dozens of individual process steps, each of which can subtly alter the final molecule.
Why Living Cells Are Necessary
Conventional drugs are small molecules, typically synthesized through straightforward chemical reactions. Aspirin, for instance, is a simple compound you could draw on a napkin. Biologics are enormously larger and structurally more complex. A typical monoclonal antibody weighs in at roughly 150,000 daltons, while most small-molecule drugs fall between 100 and 1,000 daltons.1ScienceDirect. Biologics vs. small molecules: Drug costs and patient access That size difference matters because proteins of this complexity fold into intricate three-dimensional shapes and carry sugar chains (glycans) that no purely chemical process can reliably reproduce. Only living cells have the molecular machinery to fold, assemble, and decorate these proteins correctly.
This reliance on biology introduces a challenge that defines the entire manufacturing process: biologics are exquisitely sensitive to how they are made. Change the cell line, the growth medium, the temperature, or the timing, and you can end up with a molecule that looks similar on paper but behaves differently in a patient. Small molecules, by contrast, retain their chemical identity no matter which factory makes them.1ScienceDirect. Biologics vs. small molecules: Drug costs and patient access That sensitivity is the reason biologic manufacturing demands such tight control at every stage.
Picking and Engineering the Production Cell Line
Most therapeutic antibodies today are produced in Chinese hamster ovary (CHO) cells. CHO cells became the industry workhorse decades ago because they grow well in suspension culture, tolerate the serum-free media preferred for pharmaceutical manufacturing, and attach human-like sugar patterns to proteins. Other host systems exist, including bacteria like E. coli (used for simpler proteins like insulin) and yeast, but CHO dominates the antibody space.
Getting a CHO cell to produce your antibody starts with inserting the gene encoding that protein into the cell’s genome. The two most widely used industrial systems rely on either dihydrofolate reductase (DHFR) with methotrexate selection or glutamine synthetase (GS) with methionine sulfoximine selection.2PubMed. Improving the efficiency of CHO cell line generation using glutamine synthetase gene knockout cells In the GS system, for example, the gene for glutamine synthetase is linked to the antibody gene. When the cells are grown in medium lacking glutamine and spiked with the inhibitor methionine sulfoximine, only cells that have taken up the inserted genes and are expressing them at high levels survive.3PubMed Central. Attenuated glutamine synthetase as a selection marker in CHO cells to efficiently isolate highly productive stable cells for the production of antibodies and other biologics This biochemical pressure pushes the population toward high-producing clones.
Researchers also experiment with newer gene-insertion technologies. Transposon-based systems, for instance, have shown the ability to boost antibody expression levels two- to five-fold over conventional methods, and in some configurations over a hundred-fold.4PubMed Central. Monoclonal antibodies expression improvement in CHO cells by PiggyBac transposition regarding vectors ratios and design Higher expression per cell means fewer cells and smaller bioreactors to produce the same amount of drug, which has obvious cost implications.
Once a population of high-expressing cells is generated, individual clones are isolated and screened. Scientists look for the clone that combines high productivity with stable growth and consistent product quality. That single winning clone becomes the foundation for everything that follows.
Freezing the Starting Point with Cell Banks
A biologic might be manufactured for twenty years or more. The manufacturer needs to guarantee that the cells used in batch number one thousand are genetically identical to those in batch number one. Cell banking solves this problem. The selected clone is expanded and frozen into hundreds of vials at a carefully controlled rate, creating a master cell bank. This master bank is the single source of truth for the product. From it, a smaller working cell bank is derived, and individual vials from the working bank are thawed to start each production campaign.5PubMed Central. Cell bank system, establishment, and application in the virus research, diagnosis, and biopharmaceutical industries
Both banks undergo extensive testing to confirm that the cells are free of bacterial, fungal, and viral contamination, that they carry the correct genetic insert, and that they behave as expected in culture. The frozen vials are remarkably stable. Studies have demonstrated that both master and working cell banks can maintain their defined viability and concentration for well over three years after cryopreservation.6Translational Medicine Communications. Cell Banking of HEK293T cell line for clinical-grade lentiviral particles manufacturing For companies managing multiple related cell lines from a single donor, testing strategies have been developed to pool samples from sibling lines, reducing validation costs without sacrificing safety.7PubMed Central. Cost-effective master cell bank validation of multiple clinical-grade human pluripotent stem cell lines from a single donor
Growing Cells at Scale
A single frozen vial contains only a few million cells. A production bioreactor may hold 10,000 liters or more and needs billions of cells to start. Bridging that gap is the seed train: a stepwise expansion process where cells are thawed, grown in progressively larger flasks or small bioreactors, and eventually used to inoculate the production vessel. Traditionally, this expansion takes weeks. One method developed to speed things up uses high-density frozen bags instead of small vials, allowing cells to be thawed and transferred directly into a controlled bioreactor, cutting the expansion timeline by 25 to 30 days compared to the conventional approach.8PubMed Central. A new seed-train expansion method for recombinant mammalian cell lines A related approach, sometimes called FASTEC, involves growing cells to very high densities in a perfusion bioreactor and cryopreserving them in disposable bags, so each manufacturing run starts from a consistent, high-density thaw.9PubMed. Development of a new bioprocess scheme using frozen seed train intermediates to initiate CHO cell culture manufacturing campaigns
Once the production bioreactor is inoculated, the cells are grown under tightly monitored conditions. Temperature, pH, dissolved oxygen, nutrient levels, and waste metabolite concentrations are tracked continuously. Most antibody manufacturing uses a fed-batch process, where the cells are given a starting volume of growth medium and then fed concentrated nutrient solutions at intervals. The goal is to keep the cells alive and productive as long as possible, often 12 to 15 days. Process intensification, where cells are seeded at higher initial densities, can shorten this timeline while boosting yields. One study comparing standard and high-density inoculation found that the high-density approach roughly doubled the antibody concentration for one molecule while achieving the result three days faster.10PubMed Central. Impact of fed-batch process intensification on the productivity and product quality of two CHO cell lines expressing unique novel molecular format proteins
Harvesting the Protein
When the production run ends, the bioreactor contains a complex soup: the target antibody dissolved in the liquid, plus billions of cells (many dead or dying), cell debris, DNA, host cell proteins, lipids, and residual nutrients. The first job is to separate the protein from everything else. This primary recovery step typically combines centrifugation with depth filtration. Centrifuges spin the broth at high speed to pellet cells and large debris, and the clarified liquid then passes through depth filters, which trap smaller particles and some soluble impurities.11PubMed. Advances in primary recovery: centrifugation and membrane technology Depth filtration has also become popular as a standalone harvest method, especially for smaller-scale operations, because it can remove process-related impurities at the same time.12PubMed Central. Depth filter material process interaction in the harvest of mammalian cells
The choice between centrifugation-based and filtration-based harvest depends on scale, facility design, and economics. At large manufacturing scale, disc-stack centrifuges followed by depth filters remain standard. Smaller or single-use facilities sometimes skip the centrifuge entirely and rely on depth filters alone.13Biochemical Engineering Journal. Technical and economic considerations of cell culture harvest and clarification technologies
Purification Through Chromatography
After harvest, the clarified liquid still contains the target antibody mixed with host cell proteins, DNA, leached Protein A (from the next step), and other contaminants. Purification happens in stages, and the first capture step is usually Protein A affinity chromatography. Protein A is a bacterial surface protein that binds specifically to the Fc region of most antibodies. The harvested fluid is flowed over a column packed with resin to which Protein A is attached. The antibody sticks; everything else washes through. Then a low-pH buffer is used to release the antibody. This single step typically achieves very high purity in one pass. The elution pH is the most critical variable, governing both how much antibody is recovered and whether unwanted aggregates form.14PubMed. Statistically-aided development of protein A affinity chromatography for enhancing recovery and controlling quality of a monoclonal antibody
After Protein A capture, one or two polishing steps remove the remaining trace impurities. Anion-exchange chromatography in flow-through mode is common: the antibody, which carries a positive charge at the operating pH, passes straight through the column, while negatively charged impurities like DNA and many host cell proteins bind and are retained.15PubMed. Membrane ion-exchange chromatography for process-scale antibody purification Membrane adsorbers, a newer alternative to traditional resin columns, can handle this same polishing task at higher flow rates and with less hardware. One study found that a membrane-based polisher reduced host cell protein levels from 8,000 parts per million down to as low as 10 parts per million.16PubMed. Efficient host cell protein clearance: A study of membrane adsorbers and resins in biopharmaceutical processes Multiple commercial membrane products are available for this purpose, each with different binding chemistries suited to different contaminant profiles.17PubMed. Anion exchange membrane adsorbers for flow-through polishing steps: Part II. Virus, host cell protein, DNA clearance, and antibody recovery
Ensuring Viral Safety
Because biologics are made in animal-derived cells, there is an inherent risk that viruses could be present in the product. Regulators require manufacturers to demonstrate robust viral clearance through multiple independent steps. The low-pH hold after Protein A elution serves double duty: it elutes the antibody and inactivates enveloped viruses like retroviruses. Studies have confirmed that even brief low-pH exposure achieves significant virus inactivation, and continuous-flow versions of this step perform equivalently to the traditional batch approach.18PubMed Central. Truly continuous low pH viral inactivation for biopharmaceutical process integration
Low pH alone is not enough. Nanofiltration through virus-retentive filters physically removes smaller, non-enveloped viruses like parvoviruses that can survive acidic conditions. Research has confirmed robust retrovirus inactivation by low pH and parvovirus removal by second-generation virus filters, regardless of which specific antibody is being purified.19PubMed. Viral clearance capability of monoclonal antibody purification The combination of these orthogonal methods, meaning they work by different mechanisms, provides multiple layers of safety. Regulators expect manufacturers to demonstrate that the overall process clears viruses by many orders of magnitude.
Formulation, Fill, and Finish
After purification, the antibody is in a buffer solution that is great for chromatography but not suitable for injection into a patient. The final processing steps exchange this buffer for a formulation buffer containing stabilizers like sugars, surfactants, and amino acids that protect the protein during storage. This buffer exchange is typically done by ultrafiltration and diafiltration, where the protein solution is circulated across a membrane that retains the antibody while allowing the old buffer components to wash through and be replaced by the new formulation.20PubMed. Theoretical analysis of excipient concentrations during the final ultrafiltration/diafiltration step of therapeutic antibody This same step can concentrate the antibody to its target dosing concentration, sometimes over 100 milligrams per milliliter for subcutaneous products.21PubMed. Protein-solute interactions affect the outcome of ultrafiltration/diafiltration operations
The formulated bulk drug substance is then sterile-filtered and filled into vials, pre-filled syringes, or cartridges in a cleanroom environment. Some products are lyophilized (freeze-dried) to extend shelf life, particularly if the protein is not stable as a liquid over the required storage period. Every stage of fill-finish, from the sterile filtration to the filling needles to the stoppering of vials, introduces potential stresses like shear, air-liquid interfaces, and surface contact that can damage proteins. Manufacturers carefully optimize these steps to minimize aggregation and particle formation.
Why Aggregation and Glycosylation Matter So Much
Two quality attributes receive outsized attention in biologic manufacturing: aggregation and glycosylation. Protein aggregates are clumps of antibody molecules that have stuck together, and they are more than just a cosmetic defect. Smaller aggregates with high particle counts tend to be more immunogenic than larger clumps, meaning they are more likely to trigger the patient’s immune system to produce anti-drug antibodies that can neutralize the therapy or cause adverse reactions.22PubMed Central. Aggregation of protein therapeutics enhances their immunogenicity: causes and mitigation strategies The level and type of aggregation varies significantly between different proteins, making it an unpredictable risk that demands constant monitoring.23PubMed Central. Highly variable aggregation and glycosylation profiles and their roles in immunogenicity to protein-based therapeutics
Glycosylation, the pattern of sugar chains attached to the protein, affects how the antibody functions in the body, how long it circulates, and whether it triggers unwanted immune responses. Because glycan patterns depend on the cell line, the culture conditions, and even the specific clone, manufacturers invest heavily in analytical methods to characterize them. Techniques like capillary electrophoresis coupled with mass spectrometry can profile the glycan species present on an antibody at the peptide level, providing a detailed fingerprint of the product.24PubMed. Monoclonal antibody N-glycosylation profiling using capillary electrophoresis – Mass spectrometry: Assessment and method validation Newer methods allow direct glycan analysis from the intact antibody without needing to first clip the sugars off enzymatically.25Communications Chemistry. Direct glycosylation analysis of intact monoclonal antibodies combining ESI MS of glycoforms and MALDI-in source decay MS of glycan fragments The suite of analytical tools used to characterize therapeutic glycoproteins is wide, spanning liquid chromatography, electrophoresis, mass spectrometry, and lectin-based methods.26PubMed Central. Glycan analysis of therapeutic glycoproteins
When the Process Changes
Biologic manufacturing is never truly frozen in place. Over a product’s commercial lifetime, a manufacturer might change suppliers of raw materials, move production to a new facility, switch from stainless-steel to single-use bioreactors, or scale up to larger volumes. Each change raises a question: is the product still the same? Because biologics are so sensitive to their manufacturing process, regulatory agencies require formal comparability studies whenever a significant change occurs. These studies involve running pre-change and post-change batches through an extensive battery of physicochemical, biophysical, and biological tests. When done rigorously, this side-by-side comparison can demonstrate that the manufacturing change did not impact product quality.27PubMed Central. Analytical comparability to evaluate impact of manufacturing changes of ARX788, an Anti-HER2 ADC in late-stage clinical development
This comparability framework is one of the most consequential aspects of biologic manufacturing. It is the reason biosimilars (follow-on versions of branded biologics) face a much higher regulatory bar than generic small-molecule drugs. You cannot prove a biosimilar is “the same” in the way you prove two batches of ibuprofen are the same. You can only show that it is highly similar, with no clinically meaningful differences.
Single-Use Equipment and Sustainability
Traditional biopharmaceutical facilities are built around stainless-steel bioreactors and piping that must be cleaned and steam-sterilized between batches. Over the past two decades, single-use systems, essentially pre-sterilized plastic bags and tubing that are discarded after one run, have become common, particularly for clinical-scale and small commercial manufacturing. A comparison across 72 clinical manufacturing fermentation runs showed that single-use and stainless-steel bioreactors produce monoclonal antibodies with comparable cell culture performance and product quality.28Engineering Reports. Comparative study for the production of monoclonal antibodies in single‐use vs stainless steel bioreactors based on product quality and stress factor
Single-use systems also carry a surprisingly favorable environmental profile at laboratory and small production scales. A life-cycle assessment comparing the two approaches found that producing one gram of protein in a stainless-steel bioreactor emitted about 14.9 kilograms of CO₂ equivalent, while the single-use system emitted roughly 8.7 kilograms, largely because the stainless-steel system requires energy-intensive cleaning and sterilization cycles.29Journal of Cleaner Production. Comparative life cycle assessment of stainless steel and single-use bioreactor units: A laboratory scale case study The plastic waste from single-use systems is a legitimate concern, but the overall carbon footprint comparison is not as one-sided as many people assume.
Keeping the Product Intact After Manufacturing
Even a perfectly manufactured biologic can be ruined between the factory and the patient’s arm. Most biologics must be stored and transported at controlled temperatures, typically refrigerated between 2 and 8 degrees Celsius. Some require frozen storage. The cold chain, the unbroken sequence of temperature-controlled environments from warehouse to clinic, is a genuine vulnerability. Modeling studies have shown that even when average transport conditions look fine, random variability in real-world operations can produce brief temperature spikes that accelerate protein degradation, potentially pushing a product outside its quality specifications before it reaches the patient.30Processes. Physics-Informed Stochastic Modeling of Temperature Dynamics and Product Degradation in Cold Chains This is why biologics packaging includes temperature monitors and why distribution logistics are far more complex and expensive than for a bottle of pills.
How Recombinant DNA Changed Everything
The manufacturing infrastructure described above did not emerge overnight. Before recombinant DNA technology, the only way to produce biological medicines was to extract them from natural sources. Insulin was harvested from pig and cow pancreases for decades, and those early preparations contained impurities that caused allergic reactions and limited the doses that could be given safely.31PubMed Central. Insulin: evolution of insulin formulations and their application in clinical practice over 100 years The introduction of recombinant DNA technology made it possible to insert the human insulin gene into bacteria and later into mammalian cells, ending reliance on animal tissue entirely.32Drug Development Research. Insulin: moments in history That shift, which began in the early 1980s, set the template for modern biomanufacturing. Every monoclonal antibody, fusion protein, and engineered cytokine on the market today traces its production lineage back to the same core idea: put a human gene into a cell, grow the cell at scale, and purify what it produces.
What has changed since then is the sophistication of each step. Cell lines express antibodies at grams per liter rather than milligrams. Purification platforms are standardized enough that a new antibody can often be plugged into an existing process template. Analytical tools can detect impurities at parts-per-million levels. But the fundamental logic of the process, engineer, bank, grow, harvest, purify, formulate, fill, has remained remarkably stable for four decades.