Most insulin used by diabetics today is made by genetically engineered microorganisms, primarily bacteria and yeast, that have been programmed with the human insulin gene. This process, called recombinant DNA technology, replaced the older method of extracting insulin from pig and cow pancreases and produces a molecule identical to what the human body makes naturally.
From Animal Pancreases to Engineered Cells
For decades after insulin’s discovery in the 1920s, every vial came from slaughterhouse animal organs. Pharmaceutical companies would collect pancreases from pigs and cattle, grind them up, and extract the insulin through a series of chemical steps. This worked, but animal insulin wasn’t a perfect match for human insulin, and some patients developed allergic reactions or immune responses over time. Supply was also a constant concern: it took roughly two tons of pig pancreases to produce eight ounces of purified insulin.
The breakthrough came in the late 1970s when scientists at Genentech figured out how to insert synthetic human insulin genes into bacteria. Eli Lilly partnered with Genentech to commercialize the process, and on October 28, 1982, the FDA approved Humulin. It was the first biosynthetic human insulin product and, notably, the first approved medical product of any kind made using recombinant DNA technology. The entire FDA review took only five months.
How Bacteria Produce Insulin
Human insulin is a small protein made of two chains, called the A chain and the B chain, that link together. To manufacture it in bacteria, scientists first design synthetic genes coding for each chain. These genes are built in the lab using chemical synthesis methods, with the DNA sequence optimized so that E. coli bacteria can read it efficiently.
Each synthetic gene is inserted into a circular piece of DNA called a plasmid, which acts as a delivery vehicle. The gene gets spliced into the middle of an existing bacterial gene so that when the bacteria read their own DNA and produce proteins, they also produce the insulin chain as part of a larger fused protein. Two separate strains of E. coli are created: one carrying the A chain gene and one carrying the B chain gene.
The bacteria are then grown in large fermentation tanks, where they multiply rapidly and churn out the fused proteins. After harvesting, a chemical treatment clips the insulin chain free from the rest of the fused protein. The A and B chains are purified separately, then combined in a controlled process where oxygen helps them bond together in exactly the right configuration. The result is a complete human insulin molecule, chemically identical to what your pancreas would produce.
Why Yeast Makes the Other Half
Roughly half of the world’s pharmaceutical insulin supply comes not from bacteria but from baker’s yeast, the same species (Saccharomyces cerevisiae) used in bread and beer. Yeast offers a key advantage: as a more complex cell, it can fold and process proteins in ways that bacteria cannot, which simplifies some downstream manufacturing steps.
Instead of making the A and B chains separately, yeast can produce a shortened version of proinsulin, the precursor molecule your body normally makes before trimming it into active insulin. Scientists replace the long connecting segment of natural proinsulin with a much shorter linker, just three amino acids long. This modification eliminates a structural instability that would otherwise reduce how much insulin the yeast can produce. Once the yeast secretes this precursor into the surrounding liquid, enzymes cut away the short linker to release the finished insulin molecule.
This yeast-based approach is the foundation for several major insulin brands and made it possible to engineer insulin analogs, the fast-acting and long-acting versions that give diabetics more flexibility in managing blood sugar throughout the day. By swapping specific amino acids in the insulin sequence before inserting it into yeast, manufacturers can change how quickly insulin absorbs or how long it lasts in the body.
Purification and Quality Testing
Raw insulin harvested from fermentation tanks is far from ready for injection. It contains leftover bacterial or yeast proteins, cell debris, and incomplete insulin molecules that all need to be removed. The purification process uses multiple rounds of chromatography, a technique that separates molecules based on their size, charge, or chemical properties as they pass through specialized columns.
One critical step is reversed-phase high-performance liquid chromatography, which separates insulin from chemically similar impurities that other methods might miss. This technique has been scaled up to purify kilogram quantities of insulin at a time using industrial-sized compression columns. The final product has high chemical purity and retains full biological activity.
The U.S. Pharmacopeia maintains 16 separate monographs covering insulin standards, along with seven reference standards that manufacturers must test against. Quality checks measure potency (how many units of activity per milligram), check for protein clumps called aggregates that would reduce effectiveness, and screen for chemical degradation products that can form over time with changes in temperature or pH. Every batch must pass these tests before it can be packaged and shipped.
Scale of Production
Modern insulin manufacturing happens in fermentation vessels ranging from 5,000 to 25,000 liters. The full cycle from fermentation through final purification typically takes days to weeks per batch. Given that hundreds of millions of people worldwide depend on insulin, production capacity is enormous and still expanding.
Three companies dominate global production. Novo Nordisk announced an investment exceeding $6 billion in 2023 to expand its manufacturing facilities in Denmark. Eli Lilly committed $3 billion in 2024 to expand its Wisconsin facility. Sanofi announced roughly $1 billion to build a new insulin production facility in Beijing. These investments reflect both growing demand and the complexity of maintaining a reliable supply chain for a medicine that people need every single day without interruption.
How Insulin Analogs Are Engineered
The same recombinant technology that produces standard human insulin also enables the creation of insulin analogs, modified versions designed to work faster or last longer than natural insulin. Engineers alter the amino acid sequence slightly before inserting the gene into bacteria or yeast. A single amino acid swap can change how insulin molecules stick together, which directly controls how quickly they enter your bloodstream after injection.
Rapid-acting analogs have changes that prevent insulin molecules from clumping into clusters, so they disperse and absorb within minutes of injection. Long-acting analogs have modifications that encourage slow, steady release, sometimes by making insulin bind to a protein in the blood or form crystals under the skin that dissolve gradually over 24 hours or more. These aren’t different drugs so much as precision-tuned versions of the same molecule, all manufactured through the same fundamental fermentation and purification process.