The disputed article, which would eventually appear under the title “A New Type of Microrespirometer” in the Journal of Biochemistry in January 1939, isn’t a terribly important scientific paper, but it is a gold mine of foreshadowing. First, it revealed Florey’s special talent for preserving the morale, and so the productivity, of all his research assets; when Chain claimed that he deserved credit as an author for the paper, Florey was able to give it to him without simultaneously angering Heatley (at least, not very much). In the final article, as directed by Florey, Heatley is the first-named author, but Chain is the last—in the world of scholarly publishing, the second-best spot. (In between Heatley and Chain was a research fellow at the Dunn named Isaac Berenblum, who would become a world-famous oncologist after emigrating to the new state of Israel in 1949.)
More foreshadowing: On display are not only Florey’s careful management style and Chain’s ambition, but also Heatley’s great talent for building lab equipment out of spare parts and discards. Describing the magnetized iron balls needed to mix the droplets under investigation, he wrote, “Steel-bearing balls, 1/16 in. in diameter, are given several coats of Bakelite varnish . . . the balls are then heated to 100° in paraffin wax for some minutes, the surplus wax being removed by rolling the balls on hot filter paper. . . . They are then rolled in the palm of a warm, but clean and dry hand with some well washed kaolin. . . .”
The article was the first time the peculiar mix of talents of the Dunn team stood revealed. It also marked, or rather caused, a permanent breach in the relationship between Heatley and Chain. From that moment forward, at Heatley’s insistence, and with Florey’s tacit approval, all communication and direction for the young man from Kent would come from the Dunn School’s director, rather than its chief biochemist.
By this time, Chain had plenty of other subjects to keep him busy, most especially the one that had been Florey’s reason for bringing him—and E. A. H. Richards—to the Dunn in the first place: finding the substrate for lysozyme. In 1937, assisted by one of that year’s Rhodes Scholars, a medical student from Missouri named Leslie Epstein, Chain had found his answer: The substrate was determined to be a polysaccharide, which meant that lysozyme was a polysaccharidase, whose (mild) antibacterial action was that it broke down the polysaccharides (some of them, anyway—E. coli alone makes more than two hundred different polysaccharides) that coat the cell walls of bacteria. Epstein had found the subject for his thesis: “The Actions of Certain Bacteriolytic Principles.”
Around the same time, Florey found Alexander Fleming’s 1929 paper on penicillin.
No one knows precisely how the paper came to the attention of the Dunn investigators, or even which of them first read it. To his death, Chain was adamant that Florey never thought about penicillin until Chain suggested it. “Something seemed to click in my [i.e., Chain’s] mind” after reading Fleming’s paper. Florey was equally insistent that he had brought it up to Chain. All that can be known for certain is that, during the preceding eight years, virtually no other researcher had cited Fleming’s work.
Fleming’s discovery was a classic dead end: an interesting compound that was so unstable that even its discoverer couldn’t reliably produce it for future experiment, nor could anyone else. Though Florey’s predecessor at the Dunn, George Dreyer, had been intrigued enough by Fleming’s mold to secure some for the pathology lab, he had done so for dozens of potentially interesting compounds, and no one had been any more fortunate than Fleming himself in understanding its actions. No one, that is, until 1937, when Florey and Chain began planning an ambitious survey of all the antibacterial substances produced by microorganisms. The planned survey would include dozens of different strains of bacteria, but also fungi, particularly the Penicillium molds.
The first experiments of the new survey, though, were still focused on lysozyme and other potential antibacterial substances, since penicillin, in its decidedly impure, “natural” form, was such an unreliable antibacterial agent. In any case, Chain believed penicillin to be a kind of “mould lysozyme,” an enzyme that acted, like Fleming’s egg-white lysozyme, on bacterial cell walls, but also on pathogens like staph and strep, so research on lysozyme was likely to be applicable to penicillin anyway.
In one well-remembered discussion over afternoon tea, Florey reminded his listeners that penicillin was also notoriously difficult to work with; not only had Fleming been unable to stabilize his own compound, Harold Raistrick, a skilled and experienced biochemist, had no better luck. Chain reacted by saying Raistrick couldn’t be a very good chemist, since it “must” be possible to produce it in a stable form. Whether deliberately or not, Florey had challenged Chain.
Chain, who was as competitive as a pit bull, responded, though it is worth a reminder that both he and Florey saw the research as an interesting scientific challenge far more than as a way to add to medicine’s therapeutic arsenal. Prontosil and the other sulfanilamides were rightly regarded as revolutionary therapies, and there seemed little need or desire to supplant them.
Chain was already investigating the substance produced by Pseudomonas pyocyanea (the bacterium responsible for, among other things, septic shock and a number of skin infections; since the 1880s, extracts of P. pyocyanea had been shown to destroy other bacteria . . . and to be highly toxic to mammals), and the somewhat more promising Bacillus subtilis, a hardy microbe with some demonstrated ability to stimulate the immune system. To them, he added the remaining frozen samples of Penicillium notatum that Dreyer had left behind, but his first results were unimpressive. He could study the mold if he had enough, but it was “impracticable to grow the [Penicillium] mould and carry out chemical studies simultaneously.”
Heatley stepped into the breach. Despite his self-effacing modesty, he was, in the words of Gwyn Macfarlane, a hematologist at the Radcliffe Infirmary and later one of Florey’s biographers, “a most versatile, ingenious, and skilled laboratory engineer on any scale, large or minute. To his training in biology and biochemistry, he could add the technical skills of optics, glass and metalworking, plumbing, carpentry, and as much electrical work as was needed.” Most important of all: “He could improvise—making use of the most unlikely bits of laboratory or household equipment to do the job with the least possible waste of time.”
When he was drafted to increase the yield of the antibacterial substance produced by the Penicillium mold, he knew relatively little about it, except that the fungus grew adequately on agar, but did best in shallow vessels, no more than 1.5 centimeters deep. At that depth, the branchlike mycelia of the mold could grow above the surface of the agar, and then dry out. Once dry, yellow drops of Fleming’s “juice” formed on the dried-out mycelia and could be collected using a glass pipette. Even more valuably, other penicillin droplets settled into the agar itself and turned it yellow. The most productive time for penicillin “farming,” therefore, was just after the broth turned rich enough to be harvested, but before it became so saturated that the agar couldn’t grow another batch. By careful observation, Heatley learned to identify the agar’s phase of maximum productivity.
Agar offered a good base for growing Penicillium, but continued to produce frustratingly small amounts of Fleming’s broth. The stakes were high; only with significant quantities of broth could any investigation proceed, and Heatley knew it. He fertilized it with everything he could find on the Dunn’s shelves: nitrates, salts, sugars, glycerol, meat extracts. He dosed the media with oxygen and CO2. In December 1939, he tried adding brewer’s yeast, which improved the yield only slightly, but did cut the time it took for the mold to produce the broth from three weeks to ten days.
It took months before Heatley determined the best recipe. First, he incubated the fungus on a nutrient solution known as Czapek-Dox medium: a stew of inorganic salts, sugar, and agar. Once the mold bloomed, brewer’s yeast was added. In days, a film formed on top of the medium, and soon thereafter, green spores of the Penicillium would appea
r. Over the course of ten days, the fungus would grow, after which Heatley drew off the penicillin-laced broth and replaced the growth medium, twelve times if he was lucky, two or three times if not.
This provided sufficient amounts of broth, but did little to gauge its strength. Though Fleming and those who followed had been able to demonstrate the antibacterial activity of the liquid extracted from mold, what they had wasn’t penicillin, but a broth of which penicillin was a component. How much of the liquid was biologically active? No one knew. Heatley needed a yardstick for measuring antibacterial activity, and once again he found an ingenious solution: He cut disks out of the bottoms of Petri dishes and replaced them with glass tubes, creating concave indentations in the center of the vessels, into which colonies of bacteria were introduced. He then added measured amounts of the cultivated yellow broth, and noted the size of the bacteria-free halo around each cylinder. The bigger the halo, the more potent the compound.
If Heatley was legendarily resourceful in making equipment for pennies, Florey was no less so in collecting the pennies themselves, wherever they could be found. The use of the word “pennies” is not accidental; to call the Dunn experimental program impoverished is to flatter. At the 1938 Physiological Congress in Zurich, Florey buttonholed Edward Mellanby to beg the Medical Research Council for what seems, in retrospect, a ridiculously tiny sum: £600. Even when he got it, it was scarcely enough. At one point, Florey told Chain that the lab had completely exhausted its funds, and that he must stop ordering everything, up to and including glassware. Though the Medical Research Council finally agreed to renew a portion of the lab’s grant for 1939—for Chain, £300 a year for each of the next four years, plus an additional grant for expenses of £250 annually through 1940—it kept it just north of starvation. When Florey learned that Oxford was proposing to cut the operating budget they supplied to the Dunn because a new university heating plant would lower the laboratory’s utility costs, Florey wrote back, “I have struggled to keep the place warm on money I ought to have devoted to research.”
Matters nearly came to a head in the summer of 1939, when the grant from the Medical Research Council that paid for Chain’s research on the penicillin project was about to expire. Florey might not have loved Chain’s company, but he recognized his value and was determined to find the money needed to keep his biochemist fed, housed, and not too surly for work. Their grant application not only identified fungi as a promising source for antibacterial compounds, but announced the status of their experiments on Alexander Fleming’s nearly forgotten eleven-year-old discovery.* For the first time, penicillin was an explicit part of their research program.
Florey was taking no chances. He sent a virtually identical grant proposal off to the newly established Nuffield Provincial Hospitals Trust (another charity built on a wealthy man’s will, this time a bequest from William Morris, who had made his fortune selling Morris Garages sports cars—MGs), and another to the usual suspects at the Rockefeller Foundation, though by a circuitous route. The original request was sent to the Rockefeller offices in France on November 20, and was then forwarded to the New York office ten days later, where the director of scientific research, Warren Weaver, wrote back, “The application of Florey appeals to me, but I seriously question whether a three-year grant is justified under present circumstances. . . .”
The “present circumstances,” were, of course, the Second World War, which had commenced with Germany’s invasion of Poland on September 1, and within days, declarations of war from France and Britain. By the time Weaver received the Dunn School’s grant application, Poland had been divided between Germany and the Soviet Union, the United States had passed the Neutrality Acts (which allowed France and the United Kingdom to buy arms), and the estuaries of the Thames had been mined by U-boats. The threat of war had already had an impact at the Dunn; Chain, as a refugee perhaps more fearful than most of a German invasion, and more grateful to Britain for offering sanctuary, had volunteered for a Red Cross Certificate in First Aid, and, after becoming a British subject in April 1939, joined the Oxford City Council Air Raid Precautions Department. Heatley was unable to leave England for a fellowship in Copenhagen, and—at Florey’s behest—stayed in Oxford.
Initially, the transition to open hostilities would shave budgets for research, particularly of the “interesting scientific challenge” sort. After some back-and-forth, including reassurances from Weaver that the grant would be renewed as long as the Dunn team showed progress, the money started flowing again, though, as always, through a very narrow straw. Though Florey’s original request to the Medical Research Council for studying penicillin as a therapeutic agent in vivo was for a mere £100, the MRC actually came through with only £25. Luckily, on February 19, 1940, the Dunn team learned that the Rockefeller grant had been approved, with the first payment scheduled to arrive March 1.
Tiny budgets notwithstanding, the Dunn research was progressing, and progressing rapidly. By March 1940, Heatley’s methods had improved yields so much that instead of providing Chain with a milligram of broth at a time, he could make a hundred times as much.
However, the trick of extracting penicillin from the broth in a stable form, which had eluded Raistrick and others, was still undiscovered. Fleming had tried to separate the active ingredient in his mold juice using a simple chemical technique: Dissolve the stuff in ether, which would evaporate quickly, leaving behind concentrated penicillin. It had failed almost completely. Harold Raistrick, a far more skilled chemist, used a method of separation known as liquid-liquid extraction: Pour the mixture into a “separatory funnel,” a piece of apparatus that looks like an inverted teardrop, with a funnel at the top, and a stopcock at the bottom, with a flask underneath. In the top chamber, liquids separate into different layers based on densities, with the heavier “aqueous” layer, containing ions, or charged particles, at the bottom, and a neutral, uncharged “organic” layer at the top. Shake vigorously, open the stopcock, let the bottom layer flow out, and you have your extract.
The key, then, was to add a charge to the molecules that composed the desired portion of the broth.* There are a couple of ways of charging a neutral substance, but one of the most effective is to acidify it, since what makes acids acidic is a freely given positively charged hydrogen atom: a proton. Give a proton to a neutral compound, and it becomes charged. The greater the charge, the greater the amount of the previously neutral substance that heads to the bottom of the separatory funnel. This was Raistrick’s strategy. By adding acidified ether to the solution, a little at a time, he was able to separate the mold juice from superfluous fluid, leaving behind a concentrate of about one-fifth the volume with which he began.
However, when he evaporated the ether (easy enough; it’s a very volatile substance) expecting to find a concentrated form of penicillin in the residue, it had completely vanished: None of the antibacterial activity that Fleming had identified in the mold juice remained.
In the intervening years, no one had been able to improve on Raistrick’s method. This meant that, despite Heatley’s resourcefulness in cultivating Penicillium’s precious broth, no one knew how to convert his harvest of broth into a stable source of penicillin.
Enter Heatley, again. He first tried to stabilize the compound at different temperatures and pH levels, slowly adding alkali, or base, salts to the mixture to get it back to neutral, but it wasn’t a very practical method, the equivalent of baking a soufflé over a campfire. His second idea, what he later called “laughably simple,” reextracted the compound from the acidic ether solution into a neutral medium—water—by taking another trip through the separatory funnel, and then gave it another charge, this time by exposure to a base. On March 19, 1940, he did just that, filtering the mold broth through parachute silk (in order to remove solid particles), then mixing it with ether, which caused it to layer and separate. The ether-plus-penicillin mixture was then mixed with alkaline salts, dissolved in water, and separated a
gain; only this time, the penicillin was in the aqueous layer. And, unlike the ether-plus-penicillin mixture that had lost all its potency for Raistrick, the new mix was stable, even after eleven days at room temperature. A source for experimental quantities of far purer penicillin—though the term is used loosely; while the concentration of penicillin in Fleming’s broth was only one part per million, Heatley’s first batch was still no better than .02 percent pure—had been found.
And Chain was ready to experiment with it. Almost as soon as he received the purer extract of penicillin, he directed John M. Barnes, one of the few researchers at the Dunn licensed by the Home Office to perform animal tests, to inject the entire stock of penicillin extract, 80 milligrams, into the abdomens of two mice.
It isn’t clear what he expected in the way of a reaction. Over the preceding year, Chain’s supposition about the nature of penicillin—that it was a complex protein molecule, probably an enzyme, like lysozyme—had been fading. Though the explicit aim of the 1939 grant was the “preparation from certain bacteria and fungi of powerful bactericidal enzymes, effective against staphylococci, pneumococci, and streptococci,” by the beginning of 1940, Chain had completed a number of experiments that proved penicillin couldn’t be a “bactericidal enzyme,” or indeed a protein of any sort. In one experiment, penicillin dialyzed—broke into component parts—when forced through cellophane tubes—which is something that proteins, because of their size, don’t do; in Chain’s words, his “beautiful working hypothesis dissolved into thin air.” If penicillin had been a protein, the mice would have exhibited an immune response: swelling, perhaps, or inflammation. But they didn’t. The mice might as well have been given saline solution; there was no impact. Penicillin definitely wasn’t any sort of protein.
Miracle Cure Page 12
