Miracle Cure

by William Rosen

Keefer and Richards did their best to keep the lid on, but a sampling of newspaper headlines that followed Cocoanut Grove will give the flavor:

  7 HOURS TO LIVE—SCARCEST DRUG RUSHED TO BABY

  GIRL, 20, DEAD AFTER REFUSAL OF PENICILLIN

  PENICILLIN . . . DEW OF MERCY*

  A woman in Oklahoma City wrote to President Roosevelt because “I do not know who else has the authority to help me, or if you can possibly tell me where my son can get the medicine penicillin. . . .” Another mother wrote the president a letter in which she allowed that “I know you are busy with the war, but . . . I am in great need of your help. My husband is in great need of the new drug penicillin. . . .” In the face of such demand, the New York Herald-Tribune actually provided hopelessly optimistic instructions for making penicillin in a home kitchen.

  The result was that demand and supply were seriously out of balance. During the first five months of 1943, American production of penicillin was only 400 million units; and, since more than 2 million units were needed to treat simple staph infections, the total supply was sufficient to treat fewer than a hundred patients. For skin infections like the acute streptococcal skin disease known as erysipelas, a single treatment could require 9 million units or even more: 200,000 to 400,000 units three times daily for ten days.

  Penicillin manufacturing had to be industrialized.

  In June 1943, Richards and Keefer attended a meeting in Washington hosted by Elihu Root of the National Academy of Sciences. In attendance were Robert Coghill from the Northern Lab, and members of the War Production Board, which had unprecedented authority over all allocations of resources—private and public—for the duration of the conflict. Three months later, a slightly larger version of the same group met again, this time as the WPB’s Penicillin Producers Industry Advisory Committee. Two key items were on the agenda. The first was the appointment of a “penicillin czar,” formally the coordinator of the penicillin program: Albert Elder, a chemical engineer from the U.S. Patent Office’s Chemical Division. The second was to recruit a sufficient number of qualified American corporations to ramp up penicillin production.

  The new agenda demanded a wider ambit. At the meetings called by Richards in October and December 1941, only four pharmaceutical companies had been represented: Merck, Pfizer, Squibb, and Lederle . . . and only the first three agreed to commit any resources to the project. By 1943, matters had changed dramatically. The penicillin project was now a national priority, and virtually every company that had anything to do with medicinal compounds, or even fermentation, was invited to apply for consideration, most of them given only the vaguest knowledge of the nature of the project.

  From the 175 companies that applied, Richards, Elder, Keefer, and Coghill selected seventeen. Some were obvious, like the first three: Merck, Squibb, and Pfizer. The others included drug companies like Lederle, Eli Lilly, Sharp & Dohme, Abbott Laboratories, Parke-Davis, Winthrop, Upjohn, Cutter Laboratories, Roche Nutley (the American subsidiary of the Swiss drug company Hoffmann-La Roche, located in Nutley, New Jersey), and Bristol-Myers’s Cheplin Laboratories; but also companies with experience in fermentation for other uses, such as Allied Molasses, Schenley Distillers, the Heyden Chemical Corporation (like Merck, a once-upon-a-time German-owned firm, seized by the Office of the Alien Property Custodian during the First World War), and the Commercial Solvents Corporation.* Each was promised free access to all publicly available information about penicillin fermentation, plus patentable ownership of any techniques they developed while part of the program.

  By any standard, this was a dramatic change. As later recalled by Sir Robert Robinson:

  Richards . . . recognized the simple truth that the commercial interests would not develop penicillin unless they were guaranteed some enjoyment of the fruits of their labors and investment. Somehow the companies that participated in the development of penicillin had to be permitted exclusive rights to their discoveries. Once the research and development of penicillin were completed, no one would be allowed to jump on the penicillin bandwagon for a free ride. The CMR thus found itself in the awkward position of needing to devise a system by which private companies would gain patent rights to processes and products developed, at least in part, with public money.

  By the end of 1943, the CMR was spending public money like water. The scope of the effort was really stunning in retrospect: The office had recruited thirty-six universities and hospitals, twenty-two separate companies, and four federal, state, local, and national organizations. Some were multimillion-dollar corporations, like Squibb, but not all; the first company to make a real contribution to the effort was the relatively tiny Chester County Mushroom Laboratories of West Chester, Pennsylvania, which was processing forty-two thousand surface cultures a day. In 1943 alone, the CMR approved fifty-four contracts totaling more than $2.7 million for research on penicillin, and agreed to pay penicillin producers $200 for each million units (Chester Keefer was allocated $1.9 million to buy the stuff needed for his clinical trials). In addition, the War Production Board approved sixteen new penicillin-manufacturing plants, on which the pharmaceutical companies spent nearly $23 million. They also sweetened the deal by offering, as a wartime priority, so-called certificates of necessity, along with tax breaks that allowed companies like Merck and Pfizer to depreciate their investments in only five years. The WPB also spent nearly $8 million of federal money on six penicillin-manufacturing plants, all of which were sold to private companies after the war ended, as designated “scrambled facilities,” a term of art for assets in which the private and public investments were almost impossible to disentangle—a metaphor for the entire penicillin project.*

  It was, to free-market purists, either the greatest of heresies or—more likely—the source of much cognitive dissonance. At the end of the 1920s, pharmaceutical development and manufacturing was the sixteenth most profitable industry in America. By 1944, it was, by far, the most profitable. It would remain so for nearly twenty years.

  Moreover, the industry, which had been made up of hundreds of firms, none possessing more than 3 percent of the national market, had consolidated into twenty or so companies that held, in the aggregate, 80 percent of the market for all drugs, and that market had grown tenfold. What separated the twenty winners from everyone else was possession of a penicillin contract: Each firm that got an OSRD manufacturing contract quickly outstripped its peers; one contract, in economic terms, was the equivalent of finding three hundred additional researchers or $10 million in profit (the entire profit for a company the size of Squibb in a good year prior to the project). It is almost impossible to overstate the importance of this. A mediocre company—measured by profitability or growth—without an OSRD contract was transformed into one of the most profitable simply by winning the CMR sweepstakes. It was equivalent to giving the winners a two-decade head start on the rest of an entire industry. The only comparable events in American economic history were the deals that built the transcontinental railroad and allocated the radio broadcast spectrum.

  The penicillin project would prove a game changer for companies like Merck and especially Pfizer, which staked its future on penicillin. When Jasper Kane, who had represented Pfizer at the October 1941 meeting of the OSRD, brought his plan for fermenting penicillin in the same sort of deep tanks the company used for producing citric acid to his boss, Pfizer president John L. Smith, he was asked, “Is it worth it?” Smith explained, “The mould is as temperamental as an opera singer, the yields are low, the isolation is difficult, the extraction is murder, the purification invites disaster and the assay is unsatisfactory. Think of the risks and then think of the expensive investment in big tanks—think of what it means if you lose a 2,000-gallon tank against what you lose if a flask goes bad.”

  Kane’s reaction is unrecorded, but it must have been persuasive. Smith had a friend whose daughter had been cured of erysipelas by a series of penicillin injections, and Smith hadn’t forgotten. In
early 1943, he charged his chief engineer, John McKeen, with building a pilot penicillin plant in an old ice factory and, once the plant had been fitted with fourteen 7,500-gallon fermenting tanks, “followed every tank, every potency and everything on a day to day basis.”

  He had to. Separating high-grade penicillin from all the other components in the fermentation soup was hard enough in the Northern Lab’s washing machine–sized tanks. To scale up production by several orders of magnitude, McKeen hired a process-engineering firm that was experienced with another separation challenge: turning crude oil into petroleum, kerosene, and aviation fuel. Within months, Pfizer’s subcontractor, the chemical engineering company E. B. Badger & Sons (“Process Engineers and Constructors for the Petroleum, Chemical, and Petro-Chemical Industries”; their lead engineer, Margaret Hutchinson Rousseau, had been the first woman to receive a doctorate in chemical engineering from MIT) had solved the problem of distributing sterile air throughout the fermenting liquid evenly, by introducing pressurized sterile air at the tank’s bottom, while agitators mixed it evenly into the broth. By June 1944, penicillin production exceeded 100 billion units a month, and the dramatic increase in supply had a predictable impact on price. The United States government had, the year before, guaranteed the members of the penicillin consortium that the price for the drug would be set at $200 per million units. When the price supports were removed, the market was able to find a price where supply met demand: $20 for a million units of penicillin . . . on its way to $6.*

  —

  For a time, Britain was able to keep pace with American developments, though the spigot through which research money flowed remained astonishingly narrow. Research funds continued to depend on patrons with titles. In March 1943, Florey persuaded Lord Nuffield (the former William Morris) to donate £35,000, but only over seven years. Government funds matched this, once Prime Minister Churchill concluded that the Allies might be short of the drug on D-day (or worse, that the entire supply would be reserved for American and Canadian troops), but it is scarcely surprising that British production was severely challenged by the need to keep pace with the Americans.

  Two consortia had been selected by the British government to lead the way in penicillin manufacture. The first was the giant Imperial Chemical Industries, which had been able to produce only a few dozen doses of penicillin a week as late as 1942, but had agreed to invest £300,000 in a new state-of-the-art facility (apparently after touring Pennsylvania’s Chester County Mushroom Laboratories). The other was the Therapeutic Research Corporation, which had been formed in 1941 as a joint venture composed of the Boots Pure Drug Company,* British Drug Houses, Glaxo Laboratories, British Drug Company, May & Baker, and the Wellcome Foundation.

  Because of the support of the prime minister, and heroic efforts on the part of British pharmaceutical firms, British production of penicillin managed to match American output through 1943, though, as noted, the total quantity produced by each country was barely enough for clinical trials.

  Britain’s contribution to basic research, however, remained critical—Dorothy Crowfoot Hodgkin’s in particular. Somehow, Hodgkin had gotten access to the most powerful computing machinery then available, the punch-card calculators that were used by the Royal Navy to assemble the most efficient convoys for transatlantic duty, and by the RAF for bombing tables. The computational work wasn’t easy or cheap—the Medical Research Council questioned her bill for computing, convinced it was a mistake; she assured them it was not. But by May 1943, she reported, “Our analysis reached a stage at which we felt reasonably confident that we had found the atomic positions within the crystal structure of . . . penicillin.” The dispassionate prose of scientific writing underplays both the scale and importance of Hodgkin’s achievement. Without a clear picture of the molecule, attempts to synthesize it—that is, build it from a simpler set of component parts rather than cultivate it in fermentation tanks—were doomed. Improving the antibacterial properties of the compound, however it was produced, was hostage to a clear picture of its structure in three dimensions; if penicillin fought Gram-positive bacteria by disrupting cell walls (it did), some of its molecular components had to be able to latch onto the surface of a pathogen, which required an understanding of where exactly those components were located.

  Even more impressive, Hodgkin had used a remarkable technique for elucidating the structure of penicillin; not by taking a picture of it, but by calculating its atomic positions from empirical knowledge of its activity viewed through the lens of very sophisticated mathematics . . . the biological equivalent of finding an otherwise invisible planet by measuring the effects of its mass on other, visible, objects.

  However, even with the results of the Fourier analysis of the X-ray crystallography, the molecular structure of penicillin remained controversial.* Robert Robinson, Hodgkin’s onetime tutor and now at the Dyson Perrins lab, proposed a structure based on the chemical compound oxazolone. Hodgkin, knowing how difficult it had been for Chain and Abraham to work with the compound, thought oxazolone too stable, and proposed that it was more something else, one that Abraham and Chain had already suspected, and that they therefore “immediately accepted.”

  The “something else” was a beta-lactam ring.

  A beta-lactam ring is a fairly simple chemical feature: a square formed by three atoms of carbon, one of them connected to a doubly bound atom of oxygen; and one of nitrogen, connected directly to the oxygenated carbon atom. Because two of the carbon atoms and the oxygen are bonded together at one angle, and the third carbon is attached to the square at a different angle, the square it forms is constantly under tension—imagine trying to build a square out of struts that are bending away from one another. This gives the ring both its instability—which had frustrated everyone from Fleming to Chain—but also its effectiveness. As early as 1940, researchers from Florey’s team at the Dunn had been observing penicillin’s activity against pathogenic bacteria, and reported that it didn’t kill them or dissolve them immediately; rather that the microbes exposed to penicillin went through the same first stage of mitotic division as other bacteria—elongation—but instead of dividing, they just kept elongating (sometimes ten or twenty times their normal length) until they exploded.

  Now, they knew why. All that strain from the different attachment angles made the beta-lactam ring vulnerable to breakage, and the bond that typically broke first—between the oxygenated carbon and the nitrogen atom—ended up adhering the oxygenated carbon to the enzyme needed to create the substance used to make the cell walls of Gram-positive bacteria, the ones that are unprotected by the lipopolysaccharide outer membrane. When the enzyme was locked up by the now-open beta-lactam ring, it couldn’t produce a sufficient quantity of the key component of cell walls, so when the cells divided, the new walls were, metaphorically, missing a lot of bricks, and even more mortar.

  Unsurprisingly, such walls eventually collapse. Though the debate over penicillin’s structure would continue until 1945, when both Hodgkin and the American chemist Robert Burns Woodward were able to produce incontrovertible X-ray crystallographic proof of the beta-lactam ring, the puzzle had, for all intents and purposes, been solved. As a commemorative gift, Hodgkin later presented Chain with a model decorated with pushpins stuck in place to represent the molecule’s structure.

  Hodgkin’s discoveries were exemplary science, just as she was an exemplary scientist. In addition to the Nobel Prize, she was awarded the Order of Merit, the Copley Medal, the annual medal of the Royal Society, and has appeared not once, but twice, on British stamps.* She derived the structure of some of the most medically significant compounds in the history of medicine, including insulin, vitamin B12, and not merely penicillin, but the entire family of antibiotics. At her memorial service in 1994, the molecular biologist Max Perutz, a colleague at both Oxford and Cambridge, said, “She radiated love: for chemistry, her family, her friends, her students, her crystals, and her college. . . . There wa
s magic about her person. She had no enemies, not even amongst those whose scientific theories she demolished or whose political views she opposed.”

  —

  Applying Fourier analysis to X-ray crystallography—to help to synthesize penicillin—was to prove both elusive and an expensive lure for the American and British penicillin efforts. For academics like Dorothy Hodgkin and corporations like ICI and Pfizer, the manufacture of penicillin by fermenting mold exudate seemed slightly disreputable; a temporary stopgap at best, something like treating malaria by boiling cinchona bark to get quinine rather than taking Atabrine tablets. It wasn’t merely that growing medicines, at that moment in history, seemed a bit medieval. It was also far more difficult to standardize dosages, or even, as researchers at the Northern Lab had learned, to find a reliable “pure” strain of Penicillium. Out of a combination of practicality and pride, Merck alone invested nearly $800,000 in synthesis experiments through 1944, and even promised Vannevar Bush at the OSRD—which had provided American universities grants amounting to an additional $350,000 to investigate penicillin synthesis—a bottle of synthetic penicillin by the beginning of 1945. Chemical synthesis was clearly seen as a superior, modern, way to make medicine.

  There was another, more urgent, reason to master the technique. The Allies were in a shooting war with the world’s best chemical synthesists: the Germans, who had first synthesized mepacrine/Atabrine as an antimalarial in 1931. And yet, despite the undeniable fact that the academic and industrial resources of Germany, at least as regarded chemical innovation, were superior to those in the United States or the United Kingdom before the war, they never developed a wartime antibiotics program. The question is, why not?

  At first glance, the answer might appear to be the enormous resources commanded by Vannevar Bush, Alfred Newton Richards, and the OSRD. Or, for that matter, the industrial strength that gave the United States, by the end of the Second World War, nearly half of the entire world’s gross domestic product. It was probably inevitable that American infrastructure would eventually dominate the new business of drug production, as indeed it did in every other measurable economic activity.