How Was Insulin Made From Pigs Before Modern Methods?

Pig insulin was extracted directly from the pancreases of slaughtered hogs, ground up, dissolved in acid, and painstakingly purified through a series of chemical steps that turned raw organ tissue into a life-saving medicine. For roughly six decades, from the early 1920s until recombinant DNA technology took over in the 1980s, this process supplied virtually all of the world’s insulin. The story of how it worked is part industrial chemistry, part meatpacking logistics, and part a quiet triumph of pharmaceutical science that kept millions of people with diabetes alive.

Why Pig Insulin Worked in Humans

Pig insulin is remarkably close to human insulin. The two molecules differ by just a single amino acid: at position 30 of the B-chain, human insulin has threonine where pig insulin has alanine. That tiny difference meant porcine insulin could bind to human insulin receptors and lower blood sugar almost identically to the body’s own hormone. Research comparing the binding of human, porcine, and bovine insulin to receptors from human brain, muscle, and fat tissue found no meaningful tissue-specific differences in how the three species’ insulins interacted with those receptors.¹ Bovine (cow) insulin, by contrast, differs from human insulin at three amino acid positions, which made it more likely to trigger immune reactions.

This closeness in structure is not a coincidence. Insulin is an ancient hormone, evolutionarily conserved across mammals because it performs an essential metabolic function. The single amino acid swap between pigs and humans was small enough that for most patients, injected porcine insulin worked well. But “most” was not “all,” and the impurities that rode along with early preparations caused real problems, as we’ll see below.

From Slaughterhouse to Syringe

The production of animal-source insulin began at the meatpacking plant. Pancreases were harvested from pigs immediately after slaughter, frozen or chilled, and shipped to pharmaceutical facilities. Timing mattered: insulin-producing cells in the pancreas begin to degrade quickly after an animal dies, so the organs needed to reach the factory while still fresh enough to yield usable hormone. Major insulin manufacturers maintained close relationships with slaughterhouses and meatpacking operations, effectively building a pharmaceutical supply chain on top of the food industry.

Once at the factory, the pancreases were minced or ground and mixed with acidified alcohol, usually ethanol at a low pH. The acid-alcohol mixture served two purposes: it dissolved the insulin out of the tissue and it denatured many of the other proteins in the pancreas, making them easier to separate out later. The resulting slurry was filtered to remove solid tissue debris, leaving a crude extract that contained insulin along with a host of other pancreatic proteins, including glucagon, proinsulin, C-peptide, and various digestive enzymes.

The sheer scale was staggering. It took pancreases from thousands of pigs to produce enough insulin for even a modest number of patients. One commonly cited estimate is that roughly 23,500 pig pancreases were needed to produce about one pound of purified insulin. This enormous demand meant that any disruption in the pork supply chain, whether from disease, drought, or shifting consumer demand for pork, could ripple into insulin shortages.

Purification and the Problem of Contaminants

The crude acid-alcohol extract was only the starting point. Getting from that murky solution to an injectable medicine required multiple rounds of purification. Early methods relied on fractional precipitation: adjusting the pH and adding salts or solvents to coax insulin out of solution in a relatively pure crystalline form while leaving contaminants behind. Repeated recrystallization improved purity, but even “crystalline” insulin from the 1930s through the 1960s still contained significant levels of proinsulin, C-peptide, and other peptide fragments.

Those contaminants mattered clinically. Patients treated with early insulin preparations frequently developed antibodies against the injected material. High levels of these antibodies sometimes caused immune-mediated insulin resistance, where the body’s own immune system neutralized the injected hormone and the patient needed ever-larger doses to control their blood sugar.² Injection-site reactions, including lumps, redness, and fat tissue changes under the skin (called lipodystrophy), were also common.

The breakthrough in purification came with chromatographic techniques, especially gel filtration and ion-exchange chromatography, introduced in the late 1960s and 1970s. These methods could separate proteins by size and charge far more precisely than crystallization alone. The result was what manufacturers called “monocomponent” insulin: porcine insulin purified to the point where essentially only insulin remained, with proinsulin and other contaminants reduced to trace levels. The immunological improvement was dramatic. A study of 137 patients who had never previously used insulin showed that purified pork insulin triggered significantly fewer antibodies than the older mixed beef-pork preparations of lower purity.³ The widespread adoption of highly purified porcine insulin made immune-mediated insulin resistance extremely rare.²

Why Beef Insulin Was More Troublesome

Porcine insulin was not the only animal-derived option. Bovine insulin was also widely used, often blended with pork insulin in a single vial. But bovine insulin differed from human insulin at three positions rather than one, and that greater structural distance made it more provocative to the immune system. In the same study of treatment-naive patients, even trace contamination of beef insulin mixed in with purified pork insulin made the preparation significantly more immunogenic than pure pork insulin alone.³ This finding helped drive the industry toward pork-only preparations and, eventually, toward converting porcine insulin into a molecule identical to human insulin.

Turning Pig Insulin Into Human Insulin

Because pig and human insulin differ by only that single amino acid at the end of the B-chain, chemists realized they could enzymatically swap one for the other. The technique, called semisynthesis, became commercially important in the late 1970s and 1980s as a bridge between purely animal-derived insulin and the fully recombinant human insulin that was just beginning to emerge.

The process worked by using the enzyme trypsin to catalyze a transpeptidation reaction. In essence, trypsin clipped off the terminal alanine on the porcine insulin B-chain and, in the presence of a chemically activated threonine derivative, stitched threonine into its place. The result was a molecule structurally identical to human insulin. The Danish manufacturer Novo pioneered a commercial version of this approach, intertwining the chromatographic purification steps for making monocomponent porcine insulin with the chemical conversion reactions: a trypsin-catalyzed transpeptidation followed by a nonenzymatic cleavage of an ester bond to yield the finished human insulin product.⁴

Later research explored alternative enzymes that could perform the transpeptidation more efficiently. One study showed that a bacterial enzyme called lysyl endopeptidase could catalyze the same amino acid swap more effectively than trypsin, using a single-chain insulin precursor as the starting material.⁵ These enzymatic semisyntheses represented a clever workaround, but they still depended entirely on the supply of porcine insulin as a starting material. That dependency, along with the growing concern about whether pig pancreas supplies could keep pace with rising global diabetes rates, made the process vulnerable.⁶

Quality Testing and the Rabbit Blood Sugar Test

Once purified, every batch of insulin had to be tested to confirm it actually worked. For most of the twentieth century, this meant injecting it into live rabbits and measuring the drop in their blood sugar. The rabbit blood sugar test became the gold standard for bioassay, enshrined in the United States Pharmacopeia. Manufacturers had to show that each batch contained at least 15 units of biological activity per milligram of insulin.⁷

The scale of animal testing was substantial. One major manufacturer reported using many thousands of rabbits to support routine batch-release and stability studies before cell-based laboratory assays began replacing the animals in 2018.⁷ An in vitro cell-based alternative was introduced into the USP standard in 2020 for qualitative batch release, though quantitative stability and comparability testing still required rabbits at that time. The shift away from live-animal bioassays has been slow and methodical, reflecting how deeply the rabbit test was embedded in regulatory frameworks.

Formulation for Longer Action

Pure insulin, whether from pigs or humans, acts fast and fades fast. Injected on its own, it peaks within an hour or two and wears off in a few hours. That meant patients needed multiple injections throughout the day just to keep blood sugar controlled around meals, with no overnight coverage.

To extend the duration, manufacturers developed intermediate- and long-acting formulations. The most widely used was NPH insulin (neutral protamine Hagedorn), which mixed insulin with a protein called protamine and a small amount of zinc. The protamine-zinc complex formed a suspension of tiny crystals that dissolved slowly after injection, releasing insulin over a longer period. Clinical studies comparing porcine NPH and human NPH insulins at a moderate dose found similar blood-sugar-lowering effects both early and late after injection.⁸ At lower doses, though, the semi-synthetic human version showed a slightly more prolonged glucose-lowering effect.⁸ Other long-acting approaches included lente and ultralente formulations, which used different zinc concentrations and crystal sizes to fine-tune how slowly the insulin entered the bloodstream.

These formulation tricks were independent of the insulin’s species of origin. The same NPH or lente approach worked for porcine, bovine, or semisynthetic human insulin. What mattered was the physical chemistry of the crystal suspension, not whether the insulin molecule had an alanine or a threonine at position B30.

Religious and Cultural Concerns

Porcine-derived insulin raised uncomfortable questions for patients whose religious traditions prohibit contact with pig products. In Islam, pork and pork-derived substances are generally considered haram (forbidden), and the use of porcine insulin by Muslim patients has been a subject of scholarly and ethical debate for decades.⁹ Jewish dietary law (kashrut) also prohibits pork, and some observant Jewish patients had similar reservations. Many religious authorities ultimately issued rulings permitting the use of porcine insulin on the grounds that preserving life takes precedence over dietary prohibitions, especially when no alternative existed. But the discomfort was real, and for some patients it influenced adherence to their treatment.

The arrival of recombinant human insulin, produced in bacteria or yeast without any animal tissue, largely resolved these concerns. It was one of several reasons, alongside supply security and immunogenicity, that the pharmaceutical industry moved away from animal-sourced insulin.

Why the Industry Moved On

By the early 1980s, three converging pressures were making animal-derived insulin increasingly untenable as a long-term solution. The first was supply. Global diabetes prevalence was rising steadily, and the number of pig pancreases available from slaughterhouses was finite. Expanding the pork supply to match insulin demand was not realistic, and the enzymatic semisynthesis route that converted porcine insulin to human insulin only compounded the dependency on the same limited raw material.⁶

The second was immunogenicity. Even highly purified porcine insulin occasionally triggered antibody formation, and bovine insulin was worse. The dream of a truly “human” insulin that matched the patient’s own molecule exactly was achievable through semisynthesis, but the process was costly and still dependent on pig tissue. Recombinant DNA technology promised to produce unlimited quantities of the identical human insulin molecule in fermentation tanks, without any animal tissue at all.

The third was the desire for engineered analogs. Once insulin could be produced by inserting a gene into bacteria or yeast, researchers could modify that gene to create insulin molecules that didn’t exist in nature: rapid-acting analogs that peaked faster, long-acting analogs that provided a flat baseline for 24 hours, and other variations tailored to specific clinical needs. None of that was possible starting from pig pancreases.

Eli Lilly’s Humulin, the first recombinant human insulin, reached the market in 1982. Novo Nordisk continued producing semisynthetic human insulin from porcine starting material for some years before transitioning fully to recombinant production. By the mid-1990s, the animal-derived era was effectively over in most industrialized countries, though porcine insulin remained available in some markets for patients who preferred it or had difficulty tolerating recombinant formulations.

Porcine Insulin Is Not Entirely Gone

Despite the dominance of recombinant insulin, animal-derived insulin has not completely vanished. A small number of patients report that they experience better hypoglycemia awareness or fewer side effects on porcine insulin than on recombinant human insulin or modern analogs. The reasons for this are debated. One study found that porcine insulin had a higher receptor affinity in pig brain tissue compared to human insulin.¹⁰ Whether species-matching of insulin to brain receptors explains differences in hypoglycemia awareness remains unresolved, and a separate study of human brain tissue found no such receptor-binding differences between human and porcine insulin.¹

In the United Kingdom, a porcine insulin product called Hypurin remained available for years specifically for patients who could not tolerate recombinant alternatives. Advocacy groups have periodically campaigned to keep animal-derived insulin on the market. The population using it is tiny compared to the recombinant market, but for those patients, the old-fashioned pig-derived product remains the one that works best for them.

The legacy of porcine insulin also persists in the regulatory infrastructure. The rabbit blood sugar bioassay, originally designed to test batches of animal-derived insulin, shaped how all insulin is tested and standardized. The unit system used to dose insulin today traces back to the biological potency assays developed for those early animal extracts. Even in a world of genetically engineered analogs, the basic framework of how we measure and standardize insulin still carries the fingerprints of six decades of grinding up pig pancreases and figuring out, one purification step at a time, how to turn them into medicine.