The Forgotten History of Penicillin: How a Mold Saved Hundreds of Millions of Lives

Alexander Fleming returned from a Scottish vacation in September 1928 to find his cluttered lab bench covered in petri dishes he had forgotten to clean. One of them — a culture of Staphylococcus — had been contaminated with a mold, and around the mold was a clear ring where the bacteria had been killed. Fleming noticed, was interested enough to identify the mold as Penicillium, published a paper in 1929, and then mostly stopped. His efforts to extract the active substance failed. He couldn't get pure penicillin out of the mold broth, the small amounts he could obtain seemed to lose activity quickly, and he turned his attention to other things. The mold sat in the British scientific literature for a decade, a curiosity nobody knew what to do with.

This is where the textbook story usually pauses to credit Fleming with the discovery and skip to the wartime production miracle. The actual decade between 1929 and 1939 is more interesting than that — partly because of who was working on related problems, and partly because of how close the antibiotic era came to being delayed by another twenty years.

The pre-Fleming context

The idea that one microorganism could kill another wasn't original to Fleming. Pasteur and Joubert had demonstrated bacterial antagonism in 1877. Ernest Duchesne, a French military doctor, had published a doctoral thesis in 1897 describing how Penicillium glaucum killed bacteria in lab cultures and in guinea pigs. Duchesne died of tuberculosis at 30 before he could pursue the work, and his thesis was forgotten in the French archives. Joseph Lister had used Penicillium mold filtrate to treat a wound in 1871, with apparent success, but didn't follow up.

The chemistry of identifying and purifying a small organic molecule from a complex biological broth was beyond the analytical techniques available before the late 1930s. Without chromatography, freeze-drying, and the apparatus for handling small quantities of unstable material, the active compound couldn't be separated from the rest of the broth in usable form. Fleming's failure to extract penicillin in pure form wasn't a personal limitation; it was a limitation of the discipline.

The Oxford team

The breakthrough came from a research group at the William Dunn School of Pathology in Oxford, led by Howard Florey, who had been hired as professor of pathology in 1935. Florey was an Australian who wanted to work on natural antimicrobials systematically. He recruited Ernst Chain, a German Jewish biochemist who had escaped Nazi Germany in 1933, and Norman Heatley, an Oxford graduate student with an unusual gift for improvised laboratory engineering.

The Oxford team started looking at penicillin in 1939, partly because Chain was reading widely in the literature on antibacterial substances and noticed Fleming's old paper. Chain assumed penicillin would be a protein (most known antibacterial substances were enzymes); the discovery that it was a small molecule was the first surprise. The second surprise was that it was extremely unstable in aqueous solution at room temperature, which explained why Fleming hadn't been able to purify it.

Heatley designed a freeze-drying apparatus from scrap materials — milk-bottle cooling baths, repurposed surgical bedpans for culture vessels, a homemade vacuum system — that could extract small amounts of stable penicillin from the mold broth. By May 1940 the team had enough purified penicillin to test on animals. They infected eight mice with lethal doses of Streptococcus, treated four with penicillin, and watched all four survive while all four untreated mice died. Florey reportedly remarked that "it looks like a miracle." He meant it as observation, not poetry.

The first human patient

The first human treatment was a 43-year-old Oxford policeman, Albert Alexander, who had developed septicemia after scratching his face on a rose bush. By February 1941 he was near death — multiple abscesses, fever, sepsis. The Oxford team had collected enough penicillin from months of mold cultivation to attempt treatment. Alexander improved dramatically within 24 hours of receiving penicillin. The team extended his treatment by recovering penicillin from his urine and reinjecting it. They ran out of drug after five days. Alexander relapsed and died.

The lesson the team took from Alexander's case wasn't that penicillin didn't work — it had clearly worked while it was being administered. The lesson was that production needed to scale by orders of magnitude before the drug could be used clinically, and that the war made British production impossible. By summer 1941 Florey and Heatley had taken samples of the Penicillium strain to the United States to seek industrial production capacity.

The American scale-up

The Northern Regional Research Laboratory in Peoria, Illinois, became the unlikely center of penicillin production research. Andrew Moyer at the lab figured out that adding corn steep liquor — a waste product of corn milling that the lab had been studying for unrelated reasons — to the mold broth dramatically increased yield. A laboratory technician in Peoria, Mary Hunt, found a moldy cantaloupe in a market that produced 200 times more penicillin than Fleming's original strain. The cantaloupe strain became the parent of nearly all subsequent industrial production.

Pfizer's deep-tank fermentation, developed for citric acid production in the 1930s, turned out to be exactly the right technology for industrial penicillin production. The company committed to the war effort and built the largest fermentation facility in the world in Brooklyn. By D-Day in June 1944 there was enough penicillin for all Allied casualties. By the end of the war, US production was 650 billion units per month and falling rapidly in price. The drug was released for civilian use in March 1945 in the US and gradually thereafter elsewhere.

What the textbook version misses

Three things the standard story usually skips:

First, that Fleming's contribution was the noticing, not the developing. The 1945 Nobel Prize was shared three ways — Fleming, Florey, and Chain — for exactly this reason. Florey and the Oxford team did the work that turned the curiosity into a drug. The popular framing of Fleming as the lone genius is a media simplification that Florey, characteristically, didn't fight against and that Chain bitterly resented.

Second, that the fifty-year gap between Pasteur's 1877 demonstration of bacterial antagonism and Fleming's 1928 observation wasn't because nobody noticed the phenomenon. People noticed repeatedly, including Duchesne in 1897 with results essentially identical to Fleming's. The bottleneck was chemical — until the 1930s, nobody could separate the active molecule from the soup. The breakthrough required chromatography, freeze-drying, and the analytical techniques of small-molecule chemistry to mature first. Penicillin was waiting for the rest of biochemistry to catch up.

Third, that resistance was visible from the beginning. Fleming explicitly warned in his 1945 Nobel lecture that bacteria could become resistant if penicillin was used carelessly. Resistant Staphylococcus strains appeared within three years of widespread civilian use. The arms race between antibiotics and bacterial evolution started immediately and is now the limiting factor on the entire class of drugs. Penicillin gave humanity an enormous and finite gift; we are partway through spending it.

The deeper observation

The penicillin story is the canonical example of a transformative drug, but the structure of the story is unusual. Most major drugs were produced by deliberate searches — the chemists at Bayer methodically screening dyes for medicinal activity, the pharmaceutical companies of the 1950s synthesizing thousands of analogues of promising leads. Penicillin came from an accident, lay neglected for a decade, and then required a cross-disciplinary team to convert it into a useful drug. The accident was necessary because nobody was seriously looking for non-toxic systemic antibacterials in 1928 — the medical consensus was that the toxicity of antimicrobials would always make them unsuitable for internal use. The mold contradicted the consensus, and the contradiction took eleven years to register because contradicting a consensus requires building the apparatus to demonstrate it.

The hundreds of millions of lives saved by penicillin and its descendants represent a debt to a chain of contingencies — Fleming's untidiness, Chain's reading habits, Heatley's improvisational engineering, Mary Hunt's market trip, the corn-milling industry's waste stream, and the wartime urgency that compressed twenty years of industrial development into three. Any one of these falling differently and the antibiotic era would have arrived later, in a war that needed it less, in a world that had developed the substitutes it would have had to develop. We got lucky. We are not always going to be lucky in the same way again.