Miracle Cure

by William Rosen

The bad news was that Chain hadn’t found a complex molecule, but a relatively simple one, though, again in Chain’s words, “it became very interesting to find out which structural features were responsible for the instability. It was clear that we were dealing with a chemically very unusual substance.” The good news was that the mice tolerated it almost completely, which suggested that, unlike almost every other promising antibacterial compound, it seemed safe. Even better: the extract, once excreted in urine, was a) nearly as brown as the compound injected, and b) still strongly bactericidal.

  The Dunn team was on the verge of isolating a substance that killed pathogens without damaging their hosts. Which made it highly interesting medically.

  Howard Florey made penicillin the Dunn School’s number one priority. Two months after Chain’s first test, at 11:00 A.M. on May 25, 1940, Florey infected eight mice with Streptococcus pyogenes, the pathogen responsible for infections like strep throat, impetigo, erysipelas, and even the flesh-eating nightmare, necrotizing fasciitis. At noon, two mice were given 10 milligrams of penicillin, and another two were given 5 milligrams. Follow-up doses were given at 4:15, 6:20, and 10:00. Just before midnight, as Heatley recorded in his lab notebook, “one mouse got up and staggered about for a few seconds, then fell down, twitched once or twice, and was dead.” By 4:00 A.M. on May 26, all four of the controls had died.

  All four of the treated mice survived.

  The following day, as the Battle of France raged, nearly seven hundred British ships and small craft started evacuating what Prime Minister Winston Churchill called the “root and core and brain of the British Army” from Dunkirk. It would be four years before they would return, on June 6, 1944: D-day. When they began the liberation of France, the most important and valuable medicine used by their medics and field hospitals would be the same substance that had saved four of the Dunn School’s mice.

  FOUR

  “The People’s Department”

  The Dunkirk rescue was the best news that Britain would experience for months, as the seemingly unstoppable Wehrmacht conquered Belgium, Norway, the Channel Islands, and—on June 25—France. The Dunn team was acutely aware of the German threat, but even more focused on the progress of their own investigations. The success of the May 25 experiment launched the already highly energetic Howard Florey, and everyone else at Dunn, into another gear. Within days, more successful experiments with mice followed, revealing the potency of even highly diluted concentrations of penicillin.

  It also exposed a pressing need for more staff. Though the mice experiments showed the effectiveness of the penicillin broth, they revealed little about why it worked. Bacteriologists Duncan Gardner and Jean Orr-Ewing were brought on board to investigate the mechanism by which the compound was performing its magic. Florey himself, along with Jim Kent, planned and executed a series of experiments comparing the results of different dosages of both streptococci and penicillin at different stages of infection.

  The chemical investigations were accelerating. The Dunn team needed more penicillin at more powerful concentrations, and that could only happen with a better understanding—any understanding, really—of penicillin’s chemical structure. A twenty-seven-year-old chemist named Edward Penley Abraham was assigned to work with Chain on the daunting problem of purification.

  Abraham and Chain were a well-matched team. They used a variety of purification techniques aimed at improving the compound’s potency, as measured by the ratio of the active ingredient to the rest of the solution: then as little as 0.5 milligrams per liter, less than one part in a million. The most effective technique turned out to be freeze-drying: First, dissolving the penicillin with dry methanol, diluting it with water, then freezing the liquid. Second, exposing it to a high vacuum to sublime the solvent, transforming it directly from a solid to a gas. Third, after all the sublimable solvent has been removed, separating the rest of it by exposure to high heat. The real eye-opener for the two chemists was the discovery that, even when diluted to one part in a million, penicillin still stopped bacteria from growing, which made it at least twenty times more powerful than the strongest sulfa drugs.

  This was promising and frustrating in equal measure. Penicillin might be the most powerful and effective medicine ever discovered, but it was also one of the most difficult to produce. The lack of raw material was the most daunting bottleneck for further research. How, then, to produce more? Despite the grants from the Rockefeller Foundation and the Medical Research Council, Florey was still scrambling for money, and now needed larger-scale production resources.

  One intriguing possibility would be to enlist Britain’s pharmaceutical industry, whose companies, though generally far smaller than Germany’s huge conglomerates, nonetheless possessed manufacturing expertise, investment capital, and, especially, a powerful interest in any compound with the potential to treat infectious disease. Glaxo, founded in New Zealand in 1880, had expanded into the United Kingdom less as a pharmaceutical company than as a manufacturer and processor of milk fortified with vitamin D;* by the 1930s, they were the country’s largest producer of nutrition products. Beecham’s Pills were Britain’s most popular laxative, and the foundation of Beecham Limited. Imperial Chemical Industries, the result of an I. G. Farben–style merger between competing companies (including Nobel Explosives and the British Dyestuffs Corporation), had been, since 1926, by far Britain’s largest chemical company, but, unlike the Germans, had little or no interest in pharmaceuticals. The small chemical manufacturing firm Kemball, Bishop and Company showed some interest, visiting the Dunn in March 1940 accompanied by Sir Henry Dale, president of the Royal Society, but its resources weren’t up to the task.

  More promising was Burroughs Wellcome, then Britain’s most technologically sophisticated pharmaceutical company. The company had been founded by two American expatriates, Henry S. Wellcome and Silas M. Burroughs, graduates of the Philadelphia College of Pharmacy, who had decamped to London and opened for business in 1880. Sixteen years later, the forward-thinking Americans had built Britain’s first industrial biological research facility, the Physiological Research Laboratories, largely to produce the company’s own version of Robert Koch’s antidiphtheria serum.* When two chemists from Burroughs Wellcome visited the Dunn in July 1940, they politely declined Florey’s suggestion that they take on the challenge of purifying and producing penicillin.

  With no other takers, the Dunn would have to produce the penicillin it needed for research in-house, making up in ingenuity what was lacking in money. What this meant, in practice, was throwing the job to Norman Heatley.

  Heatley’s first challenge in producing sufficient quantities of the precious substance was a shortage of properly sized vessels for growing the mold itself. Faced with extraordinary difficulty in procuring appropriate glassware (or anything; Florey was so penurious that he had the Dunn’s elevator shut down in order to save £25 annually), Heatley turned to larceny. Pie trays and baking dishes mysteriously vanished from the Dunn’s kitchens. Sixteen bedpans likewise disappeared from the Radcliffe Infirmary, to reappear in the pathology department’s labs.

  Over the course of 1940, Heatley continued to improve his technique for extracting penicillin from the broth produced in his bedpan factories. The most effective process he found depended on a truly remarkable machine, able to mix ether with penicillin, acidify it, and separate the concentrated broth.

  Credit: Wellcome Library, London

  The modified ceramic bedpans, used by the Dunn School team to grow penicillin broth

  The homemade apparatus in operation resembled nothing so much as a Rube Goldberg cartoon.

  Three bottles—of broth, ether, and acid—are held upside down in a frame, until

  The glass ball stopper in the bottle containing broth is moved aside; liquid flows

  Into a glass coil surrounded by ice. Once cooled, the acidified liquid combines with acid from bottle number three and is jet-s
prayed in droplets

  That arrive in one of six parallel separation tubes. Meanwhile

  The stopper on bottle number two, containing ether, is moved aside, releasing ether into the bottom of the whole arrangement. The filtrate in the separation tube is sprayed into a tube of ether rising in a four-foot-long tube. As penicillin has a chemical affinity for the ether, it transfers into that tube, leaving the remaining components of the original broth behind, to be drained out.

  Then, the penicillin-plus-ether (later acetate) solution is introduced into another tube, with slightly alkaline water. The penicillin-plus-water mixture—about 20 percent of the volume of the filtered broth that started the whole rigamarole—was drawn off.

  Remarkably, Heatley’s collection of discarded junk and baling wire—the hole in the glass needed to produce the right-sized droplets was made by Heatley pushing the point of a sewing needle through hot glass; a cast-off doorbell rang when each bottle was filled—could turn about twelve liters of broth into two liters of decidedly impure penicillin in an hour. A quantity of penicillin sufficient for experimentation (on mice, at any rate) had been guaranteed.

  Credit: Science & Society Picture Library

  The filtration machine built by Norman Heatley to purify penicillin

  Florey decided it was time to go public. The Lancet of August 24, 1940, contained half a dozen articles. Two of them—topically enough, given the Blitz—were on treating blast injuries to the lung. Another article was on meningitis; one described the orthopedic trauma known as “locking wrist.” Right in the middle, though, was the world changer: “Penicillin as a Chemotherapeutic Agent,” by E. Chain, H. W. Florey, A. D. Gardner, N. G. Heatley, M. A. Jennings, J. Orr-Ewing, and A. G. Sanders. (Florey, having already suffered through the sniping between Chain and Heatley over authorial credit, deferred to the alphabet.) The first line of the article reads, “In recent years interest in chemotherapeutic effects has been almost exclusively focused on the sulphonamides and their derivatives. There are, however, other possibilities. . . .”

  The possibilities were detailed in the following two pages, including results from the five key studies performed at the Dunn since March, in which as many as seventy-five mice had been exposed to pathogens as varied as staphylococcus, streptococcus, and clostridium: “During the last year methods have been devised here for obtaining a considerable yield of penicillin, and for rapid assay of its inhibitory power. From the culture medium a brown powder has been obtained which is freely soluble in water. It and its solution are stable for a considerable time and though it is not a pure substance, its antibacterial activity is very great. . . .”

  Eleven years and five months after Alexander Fleming had published “On the Antibacterial Actions of Cultures of a Penicillium,” his discovery had finally been revealed as more than just a Petri dish curiosity: “The results are clear cut, and show that penicillin is active in vivo against at least three of the organisms inhibited in vitro.”

  —

  Even before the Lancet article appeared in August, work at the Dunn had been proceeding on three intersecting tracks. Norman Heatley and the technical staff continued to improve the processes by which penicillin-rich broth was grown, and from which the active ingredient could be extracted. The biochemical team, primarily Chain and E. P. Abraham, were subjecting the compound to a series of experiments intended to establish its structure. And the bacteriologists and pharmacologists were designing new ways to test the antibacterial properties of the compound on laboratory animals.

  It is a testimony to Florey’s great gifts as a scientific administrator that the projects—production, analysis, and effectiveness—all succeeded brilliantly, in spite of the enormous amount of friction produced by the Dunn’s collection of strong personalities. The achievement is notable even though it might be said that at least some of Florey’s challenges in managing the lab were self-inflicted. By 1940, he had begun an affair with Margaret Jennings, a physician and histologist who had joined the Dunn in 1936 and became indispensable to the Australian both as a lab assistant and as the editor of the entire lab’s scientific publications. Which meant that Ethel Florey, who had been assigned to supervise the penicillin team’s clinical trials, was having her papers and reports checked for clarity by her husband’s mistress.*

  By comparison, soothing the sensitivities of Ernst Chain was a challenge scarcely worth mentioning. He and E. P. Abraham knew there was no chance of deciphering the structure of penicillin until it could be crystallized—precipitated into something stable enough to be analyzed. Abraham and Chain were still struggling with the structural chemistry when the Lancet article was published. Aware of this, the journal’s editors appended a note reading, “What [penicillin’s] chemical nature is, and whether it can be prepared on a commercial scale, are problems to which the Oxford pathologists are doubtless addressing themselves,” which reads as a bit of an understated scold, as if the Dunn team were keeping at least some of their work secret for now, possibly in order to ensure that they would receive full credit for the discovery.

  They were on to something. On September 2, Alexander Fleming paid the Dunn a surprise visit—more a surprise for some of the team than others; Chain evidently thought Fleming was dead—to find out what had been done “with my old penicillin.” The maneuvering for credit was well under way.

  Anyone familiar with the adage about success having a thousand fathers might have predicted what would ensue. The Lancet article had ended, courteously enough, with a footnote from the authors thanking the Nuffield Trust, the Medical Research Council, and the Rockefeller Foundation for their support. It did not have the intended effect. In a foreshadowing of the looming disputes about credit and recognition that would attach to the penicillin discovery, Edward Mellanby scolded Florey about showing more gratitude to the Rockefeller Foundation than his own MRC: “I shall be surprised if the Rockefeller Foundation are supporting the work to anything like [the] extent [of the MRC support; he was unaware of the 1939 Rockefeller grant] . . . if you have a good thing in your own country, you might as well give it proper credit and not follow those people who, in cases of research, find it more convenient to give foreigners boosts than their own colleagues. . . .”

  Sometime shortly after the article’s publication, Florey received a report from Ernest Gäumann of the Swiss Federal Institute of Technology informing him that he had been approached by the Basel-based concern known as Chemische Industrie Basel, or CIBA, to help purify and manufacture penicillin for the company. More alarming, at least to Florey, was the other news from Switzerland: German researchers were eager to examine any available samples of penicillin.

  It was a real dilemma. On the one hand, penicillin promised to be a spectacularly important scientific and medical advance, and one of the canons of twentieth-century science was that such discoveries should be shared as widely as possible. Moreover, there was no legal ground for refusing knowledge about therapies, even when the therapeutic information about penicillin was still the very definition of provisional. On the other hand, Britain was literally fighting for its life. Sharing information about a drug that might accelerate the recovery of wounded German soldiers seemed a lot like giving aid and comfort to the enemy. Florey seems not to have tortured himself very long about competing loyalties. He immediately wrote to Edward Mellanby, saying it was “very undesirable that the Swiss and hence the Germans should get penicillin, and I think it would be well worth while to issue instructions to the National Type Collections not to issue cultures of Penicillium notatum to anyone with [a] possible enemy connection, and to send a letter to [Alexander] Fleming to the same effect.”

  Mellanby’s reply: “I sympathize with your position, but I do not see how the Medical Research Council can ask their National Type Collections to restrict their dispatch of special cultures . . . to a neutral country like Switzerland.” He went on: “If the sulphonamide compounds had not proved to be so
efficacious, I think you might have had a strong case [but] although I do not doubt that penicillin may prove to be superior to the sulphonamide compounds, I have difficulty in believing that this superiority is so great that national interests dictate the withholding of publication.”

  By January 1941, Heatley’s penicillin factory had produced enough penicillin to move from testing the stuff on mice weighing 20 grams or so to 150-pound humans. The purpose was as much to discover whether the compound was dangerous as to test whether it was efficacious. Biologists and pathologists had, after all, discovered dozens of antibacterial compounds that were therapeutically useless because they attacked healthy cells as aggressively as they did pathogens. Florey and the clinical physician he had recruited to administer treatment to human patients, L. J. Witts of the Radcliffe Infirmary, considered who would be the subjects for such a test. Volunteers from Oxford? Someone already at death’s door? The Dunn team decided on the latter. On January 17, Elva Akers, a patient at Radcliffe whose cancer was so far advanced that she was given only a month or so to live, volunteered to receive an injection of a tiny amount of penicillin: 100 milligrams.

  Given the dozens of mice who had received weight-comparable doses of the drug with no ill effect, the injection ought to have been safe for a human. It was not. Mrs. Akers almost immediately experienced a high fever accompanied by seizures. The reason, however, wasn’t the penicillin, but the contaminants that the biochemical team had been unable to segregate from the penicillin itself. This was a side effect of Chain and Abraham’s success in separating the penicillin filtrate into different layers; by making one of them relatively pure—up to 80 percent pure—they had also enriched the other layers with a high percentage of impurities, and at least some of them were pyrogens: fever-causing compounds.