S  A  C  R  E  D      B  O  V  I  N  E  S
The  Ironies  of  Misplaced Assumptions  in  Biology

Penicillin and Chance*

by Douglas Allchin

Alexander Fleming's discovery of penicillin is one of the most celebrated cases of chance, or accident, in science. In the conventional story, a stray mold spore was borne through an open window and landed on an exposed bacterial culture, Fleming later noticed a clear zone where the bacteria had been killed, he immediately recognized the thera-peutic significance of the event, and it was only a matter of time before penicillin became a miracle drug. Fleming himself often underscored the role of chance in his work. Despite the numerous honors and awards he received, he was fond of reminding others, "I did not invent penicillin. Nature did that. I only discovered it by accident."

There was even more "chance" to the story than is often told, however. In addition, the traditional account obscures a considerable amount of scientific work that identified the efficacy of penicillin as an antibacterial agent. Without several researchers, who aggressively pursued the potential in Fleming's initial observation, penicillin would probably not have become a "discovery" on this occasion. The fuller story suggests a more complex view of science--as guided both by the contingencies of circumstance and by the focused effort of researchers.

Renewed interest in the history of Fleming's work began quite a few years ago when a bacteriologist in London noted that the windows of Fleming's lab at St. Mary's Hospital were so constructed that they could not open. How could a stray mold spore have wandered in, even by chance? Second, he observed, spores of Penicillium will not germinate under the conditions described by Fleming. Someone else then observed that the particular species of Penicillium would not likely have been floating in the air of London. Though common bread mold is a variety of Penicillium, it was the much rarer P. notatum that produced Fleming's penicillin.

The most likely source of the mold, it now appears, was a mycology lab downstairs from Fleming. There were likely spores all over the building. Further, Fleming was never known for neatness in his lab. Open cultures would not have been uncommon. It almost seems inevitable, then, that the mold would contaminate one of his cultures sooner or later.

The conditions of contamination would also have been important. Fleming believed, based on his earlier work on lysozyme, that penicillin acted by lysing bacteria open. This would certainly have accounted for the watery appearance of the area on his culture where the bacteria were absent. In this case, the spore would merely have needed to land on the culture plate--and this is how Fleming reported his own chance observation. But we have since learned that penicillin acts by blocking the synthesis of chemicals used by bacteria to build cell walls. Penicillin does not kill bacteria outright. Rather, it prevents their effective reproduction. A spore landing on an existing culture would thus be unlikely to have any immediate observable effect. The mold would have had to establish itself first if it was to prevent the further growth of bacteria. Temperature conditions while Fleming was away from his lab on vacation may have allowed this, or Fleming may have inoculated a plate that was already moldy. In either case, a stray mold spore alone would not have created what Fleming observed.

The circumstance whereby Fleming noticed the original culture also seems quite improbable. Fleming did not notice the mold's effect while routine-ly examining his cultures, though he did inspect them when he returned from his one-month summer vacation in 1928. In fact, he had discarded the now famous culture and left it to soak in a tray of lysol. A former member of his lab stopped by to visit, however, and Fleming showed him several cultures. Among these he casually selected the critical culture from the top of the discarded stack, where it had escaped the liquid disinfectant. Only then was Fleming struck by the unusual pattern of growth. He was obviously impressed, though, because he showed the culture to numerous colleagues the rest of the day and went on to investigate some of the strange antibacterial properties he saw.

Fleming was certainly not the first scientist to have noticed the antibacterial effects of molds. In 1871, Joseph Lister (noted for introducing antiseptic practice into surgery) had found that a mold in a sample of urine seemed to be inhibiting bacterial growth. In 1875 John Tyndall reported to the Royal Society in London that a species of Penicillium had caused some of his bacteria to burst. In 1877 Louis Pasteur and Jules Joubert observed that airborne microorganisms could inhibit the growth of anthrax bacilli in urine that had been previously sterilized.

Most dramatically, Ernest Duchesne had completed a doctoral dissertation in 1897 on the evolutionary competition among microorganisms, focusing on the interaction between E. coli and Penicillium glaucum . Duchesne reported how the mold had eliminated the bacteria in culture. He had also inoculated animals with both the mold and a lethal dose of typhoid bacilli, showing that the mold prevented the animals from contracting typhoid. He urged more research, but went into the army following his degree and died of tuberculosis before ever returning to research. Chance, here, worked against his discovery (or potential discovery?) bearing fruit.

Several other researchers--almost certainly unknown to Fleming--had noticed the effects of Penicillium molds on bacteria. Fleming was not unique in this regard. But noticing a phenomena does not always mean that it will be followed up. The chance in Fleming's case may have been less the appearance of the moldy culture itself than that Fleming had a habit of pursuing odd phenomena. Fleming pursued his observation.

Still, Fleming did not follow through on his own "discovery" in ways that we might expect, knowing the current role and importance of penicillin. Fleming originally observed the action of penicillin in 1928. Yet he did not initiate clinical trials. Nor did he strongly advocate the use of penicillin in treating humans until 1940. The events during this twelve-year hiatus are perhaps the most telling in the history of penicillin.

Fleming was certainly searching for antibacterial agents in 1928 and he investigated penicillin's potential. But he was not impressed. He found that penicillin was not toxic to animals and that it did not harm white blood cells (leucocytes), yet he also found that penicillin would not be absorbed if taken orally. Penicillin taken by injection, alternatively, was excreted in the urine in a matter of hours--well before it could have its effects. For Fleming, penicillin's therapeutic potential was limited, perhaps to topical antisepsis.

Fleming did continue to use and advocate penicillin in the years following his initial discovery. But he saw the value of penicillin primarily in the context of bacteriology. Penicillin suppressed the growth of certain bacterial species, allowing one to selectively culture certain others (such as those causing influenza, acne and whooping cough). In this role penicillin became a valuable tool in the manufacture of vaccines--a major task Fleming managed at St. Mary's Hospital. Production of penicillin continued on a weekly basis throughout the 1930s, but all for purifying bacterial cultures. The penicillin was crude--good enough for Fleming's purpose, but hardly strong enough to destroy a serious human infection. Meanwhile, Fleming had turned his research to another group of chemical bactericides, the sulphonamides.

The pursuit of penicillin in treating human infections was due ultimately to another lab, led by Howard Florey in Oxford. In 1938 Ernst Chain, an associate of Florey's, began a search for natural antibacterial agents, as part of an effort to under-stand their mechanisms more fully. He chose three to study, penicillin among them. Fleming's 1929 paper offered a thread of information that Chain could pick up, though with a quite different purpose in mind. By early 1939 Chain and Florey began to suspect the medical potential of penicillin. But they could not simply test it: penicillin was difficult to produce and to purify. Florey had difficulty finding funding. By that time, the war effort in Britain meant that extra funds were not available for exploring mere possibilities. Support eventually came in late 1939 from the Rockefeller Foundation in the U.S.

Florey shifted the resources of his department to the penicillin project. Before they could demonstrate the efficacy of penicillin, they had several technical challenges. They needed to improve extraction methods, refine an assay for determining the strength of their extracts, and scale up production. After five months of work--in May, 1940--they had enough of the brown powder to test on mice. The penicillin allowed several mice injected with lethal doses of virulent streptococci to survive. The potential of penicillin for treating infections then seemed demonstrably real. Florey and Chain repeated their tests as a double-check, and then went on to determine appropriate dosages and treatment duration, publishing their results in August.

But the research was hardly done. Would the results transfer to humans? To know, they had to scale up production yet again. Based on relative weight, a human would need roughly 3,000 times the penicillin used by a mouse. And commercial support was still not forthcoming. In the Oxford labs, flasks and biscuit tins used for the mold cultures gave way to hundreds of bedpan-like vessels stored on bookshelves. Purification turned from the laboratory to dairy equipment. Column chromatography allowed the group to isolate the relevant fractions and to concentrate their solutions. All this was in the service of a clinical test. --And after the first test in early 1941, they had to return to their methods to find a way to remove some impurities that had caused side effects. The tests eventually went quite well, but it had required two professors, five graduates and ten assistants working almost every day of the week for several months to produce enough penicillin to treat six patients.

Fleming took notice of the striking results. But he did not disturb his research agenda. He knew that the value of penicillin still lay in research on economical mass production. Thus, the research--and, in a sense, the discovery--was still not complete. Florey took his cause to America once again, where work began on the scale of breweries. One key technical assistant found a new medium for the mold cultures, increasing yields tenfold. Other drug companies in England were by now interested, but the scale of production was at first somewhat limited. After a second set of clinical trials in 1942-43, though, production began in earnest. In another half-year, industry could produce enough for treating 200 persons per month. Two years later, the U.S. was producing enough to treat a quarter-million patients per month.

Many scientists, Fleming among them, were confident that determining the chemical structure of penicillin would enable chemists to produce it synthetically and thus more economically. Once the structure was determined, however, synthesis proved to be at least as costly as extraction. The "failure" seems an exception in this tale otherwise graced by good fortune. But not all research ventures pay off as expected--chance works both ways in science.

Nobel Prize winner Peter Medawar once commented, "I was sorry that the traditional story of Fleming's discovery did not stand up to critical scrutiny because I should have liked to have believed it true; but even if it had been true, it would not have told us very much about the efficacy of luck." Here, Medawar referred to the substantial work that transforms a lucky event into a genuine discovery. There is more to science that what meets the eye. First, one must recognize and be ready to pursue the meaning of one's observations. Fleming had a habit of playing in the lab and of toying with oddities. He pursued a chance phenomenon that even his colleagues found insignificant, even without guessing its ultimate significance.

Further, the import of an observation is not always obvious. Chain and Florey recognized a therapeutic potential where Fleming saw it only vaguely. And they were willing to invest resources to pursue it. Fleming, Chain and Florey all shared in the Nobel Prize in Medicine in 1945. Their joint award reminds us that the discovery of penicillin was more than a mere chance event.

Further Reading

  • Macfarlane, Gwyn. 1985. Alexander Fleming: The Man and the Myth. Oxford University Press.
  • Judson, Horace Freeland. 1981. "Chance." In The Search for Solutions. Holt, Rinehart, Winston.

*This essay first appeared in the SHiPS Network News, Vol.5, No. 4, pp. 1, 4-5 (1995).