Experiment Eleven

by Peter Pringle

  Back at work after a week, Schatz began the crucial tests to see how his microbes would deal with the H37Rv strain. He took what precautions were available. The safety equipment had not progressed much beyond the crusted and cracked rubber aprons of the 1930s. By modern standards they were pitiful. He grew the germs in narrow test tubes, trying to limit exposure of the growing TB cultures to the open air when he removed the cotton wool caps. He had to sterilize all his own glassware. This meant putting the test tubes into the autoclave before leaving the lab—often in the early hours of the morning. The tests took much longer than those of the other germs because H37Rv was one of the slowest-growing—multiplying every two days instead of every twenty minutes like most bacteria. He started his first H37Rv test on March 24 and noted it as Experiment 10 in his new 1944 notebook.

  It was a potentially deadly operation, handling germs that had killed millions of people without gloves or face masks or proper ventilation. The TB germ spreads on airborne droplets, but Doris Jones later remembered seeing Schatz wash out his mouth with antiseptic after a day’s work. No other remedies were available. For the entire time he was using the virulent strain of TB, he later recalled, Waksman never came near the laboratory, and he told Schatz never to bring the germs to the upstairs labs. Schatz believed that his professor was afraid. It was not an unreasonable belief: In his retelling of the discovery of streptomycin, Waksman never mentioned being in the basement laboratory during these crucial experiments.

  Rows of test tubes showing how streptomycin destroys the human strain of tuberculosis, H37, in Albert Schatz’s experiments in the basement laboratory. At left, the tubes start with the TB microbes, which are gradually destroyed. (Photograph by Julius Schatz, courtesy Vivian Schatz)

  By the middle of April, Schatz had made enough streptomycin to send ten grams to the Mayo Clinic, and Feldman and Hinshaw started tests with four guinea pigs on April 27, quickly running through the initial supply. Schatz obliged by keeping his stills going night and day.

  THE MAYO GUINEA pig experiment presented a major problem of logistics. The animals had to be injected with streptomycin every five hours, which meant Feldman and Hinshaw driving through the winter snows to the lab, plus dealing with gas shortages and a lack of lab staff because of the war. Feldman, with Waksman’s permission, brought in a third researcher, with “one of the finest labs,” to test animals infected with bacteria other than TB. Initial results were promising on everything but cholera.

  Waksman began to worry that he was somehow losing control of streptomycin. In the fiercely competitive world of patents and scientific discovery, priority matters. The recognition and the rewards go to the scientist who publishes first. The research field was suddenly crowded. Schatz had done the initial isolation. Doris Jones had done the first toxicity tests on chick embryos. The vet at Rutgers was testing streptomycin on mice. Merck researchers were working on ways to extract the drug from the nutrient broth in which A. griseus, the streptomycin-producing microbe, was grown, and the company had also done animal tests. And besides TB, the Mayo Clinic was also testing streptomycin on plague, tularemia, pneumonia, and cholera. Understandably, Waksman wanted his results—or, more correctly, Schatz’s results—on in vitro experiments with H37Rv to be published first. The Mayo researchers had a quick and easy route to publication through their own Proceedings of the Mayo Clinic, which was published every two weeks. The publishing routes for Waksman took a month at least—as Feldman had surmised.

  Waksman told Feldman that he wanted to be sure that the Mayo people would warn him if they were going to publish anything, and would let him “look over critically” any papers reporting results from their experiments. This would avoid “any possible confusion” among the various research groups. Feldman agreed.

  Feldman and Hinshaw ended their initial four-guinea-pig test on June 20 after fifty-four days. Streptomycin worked against TB, and it was not toxic. They concluded, “The new substance was well tolerated at a therapeutic level sufficient to exert a marked suppressive effect on otherwise irreversible tuberculous infection in guinea pigs.” And after five years of trying, this was the most positive result they had ever seen; perhaps they had even found a cure.

  There was one important qualification. Like the sulfa drugs, streptomycin had not destroyed all the TB germs, and therefore the action of the drug was bacteriostatic—it stopped the infection from spreading—but not bactericidal, meaning it did not kill the germs. How much streptomycin was an advance on the sulfa drugs was not yet clear. There could also have been errors, but the two researchers were confident of at least one part of the trial: Their tests had shown no adverse toxicity.

  Feldman and Hinshaw celebrated their findings in the shade of an elderly apple tree in Feldman’s garden. They sipped one of Feldman’s “mysterious and delicious concoctions made with laboratory alcohol [200 proof].” But they kept reminding themselves that this was but “one step forward” in their search. The important next step was to try streptomycin on a larger number of guinea pigs over a longer period of time.

  William Feldman (left) and Corwin Hinshaw (right) of the Mayo Clinic in Rochester, Minnesota. They confirmed streptomycin’s effect on tuberculosis in guinea pigs and later carried out the first clinical tests on humans. (Feldman by permission of the Mayo Clinic Historical Unit, Rochester, Minnesota. Hinshaw courtesy of Horton Hinshaw and Dorothy Hinshaw Patent.)

  Before the sun had set in Feldman’s garden on that first evening, they had decided on a plan. To make a more certain appraisal, they needed to give the slow-growing TB germ time to take a proper hold in the guinea pigs’ organs before starting the streptomycin treatment—and this meant infecting the animals six to seven weeks before the start of therapy.

  They also saw a problem of priority. Waksman had indicated that other labs were involved, but he had not named them. Hinshaw, always impulsive, suggested that they start to infect guinea pigs immediately. Although this was strictly against the Mayo Clinic rules—the clinic first had to give its approval—Feldman agreed. He was as impatient as his colleague.

  At the same time, Feldman arranged a meeting with Waksman to collect more streptomycin, but did not mention that they had already started the second experiment. It was a gamble, of course. Feldman thought his request for more streptomycin would carry more weight if Hinshaw came to the meeting, as he was eager to do, and he did not give Waksman a chance to say no. He made reservations for two at the Roger Smith Hotel in New Brunswick for the night of July 9. “We will contact you sometime Monday morning,” he wrote to Waksman.

  Hinshaw was shocked at the meager facilities of Waksman’s lab—and even more surprised that Waksman had never worked with animals himself. He was also curious to find that Waksman had no interest in pathology and no medical expertise. He was a soil scientist, and his labs were not set up to produce the quantities of streptomycin needed. “It was obvious that he would not be able to produce the streptomycin we wanted for an adequate experiment,” Hinshaw concluded. Schatz had exhausted his supplies, and in any case his primitive production line was simply not up to the task.

  Waksman arranged for Feldman and Hinshaw to meet Merck officials to see if some of the company’s own sparse supplies could be released. Waksman asked them to keep this contact with Merck secret; he didn’t want other researchers, who were also interested in samples of the new drug, to know that Merck could release supplies—if, indeed, Merck decided to do so.

  When they arrived at Rahway, the Mayo team were surprised to find more than a dozen Merck researchers gathered for the “secret” meeting and assumed they must be in the wrong room. They quickly retreated and continued down the corridor. But they were called back. They presented their “meager evidence,” and the Merck chemists at first said no, they could not increase their production. Feldman and Hinshaw were about to give up when George Merck himself walked into the meeting. Hinshaw happened to be describing the prevalence of TB among World War One veterans, and Merck nodded his he
ad. He could justify providing the Mayo Clinic with scarce streptomycin as part of the war effort.

  In the end, they got their streptomycin, but as they left, Waksman warned them again against any “premature publicity.” Other companies also researching the new drug would be upset if they knew, he stressed. The Merck supply should be “privileged information.” In fact, Feldman and Hinshaw found his warning unnecessary. In their labs, they had always been careful never to use the real names of drugs that they were testing, referring to them only by code names.

  With new supplies now assured, Feldman and Hinshaw began two crucial experiments. The first was to run over thirty days—“to determine if the initial experiment could be confirmed.” It was important to know if the previous experiments could be repeated. The second, a longer-term experiment, would last into January 1945.

  At the same time, the results from the third Mayo researcher, Dr. Fordyce Heilman, were also positive. All five of the harmful bacteria, including those causing plague and bovine TB in infected mice, were “completely inhibited” by streptomycin. The results were “considerably better” than those demonstrated in experiments using sulfa drugs and the pressure was building on Waksman to let the Mayo team publish.

  IN THE MIDDLE of September 1944, the thirty-day guinea pig test at the Mayo Clinic proved to be a “definite” confirmation of the earlier success. Feldman wanted Waksman to be the first to know and telephoned him with the news, arranging for his secretary to record the historic call in shorthand. This experiment had used two virulent strains of TB, H37Rv and a sputum strain, which he and Hinshaw had isolated a year earlier. Feldman, a cautious man, used the word “definite” to describe the results, but it was underplaying a dramatic breakthrough in the long fight against TB. Oddly, Waksman was unmoved. He was preoccupied, distant.

  “Good,” he said, according to the transcript.

  Feldman pressed on with the astonishing news. To a man of few exclamations himself, this seemed a minimal enthusiasm, at best.

  “There can be no reasonable doubt that streptomycin has a considerable efficacy in combating experimental TB.”

  “Fine,” Waksman responded.

  “I will write you a more detailed account of the results, probably tomorrow,” Feldman said.

  Waksman: “Good. I shall look forward to hearing of the results.”

  Waksman’s seeming lack of interest was not out of character. He was an intensely practical man, not known for eloquent conversation, especially about work. He had a fiercely logical mind, moved swiftly step-by-step, and did not like to waste time. But what Feldman spoke of next suddenly got his attention: publication of the Mayo results.

  “We would like to know your wishes regarding the publication within the next month or so of a preliminary report of this work.”

  Waksman hesitated.

  “Well,” he said finally, “I think it would be all right. I’ll tell you what I will do. We have prepared for publication our in vitro study [Schatz’s streptomycin tests on H37Rv], for the Proceedings of the Society for Experimental Biology and Medicine. I don’t see why you couldn’t do that for your in vivo [guinea pig] study.”

  This was not what Feldman had in mind. The society’s journal would take a couple of months at least to publish Schatz’s results; the Mayo Clinic’s own publication could print their results in two weeks. So he ignored Waksman’s suggestion, continuing, “We would of course submit to you and the Merck people a copy of the manuscript for suggestions before submitting the paper for publication.”

  Waksman summoned an ally. He said that he personally saw “no objection” to the Mayo publication, but he wasn’t sure “how the Merck people will feel.” Merck was also doing animal tests. But Waksman’s aim was to stall Feldman until he could publish Schatz’s crucial in vitro tests. That paper would also have Waksman’s name on it, of course, even though Schatz had done the experiments alone in the basement lab. Waksman wanted to leave Feldman with the understanding that there might be a doubt about Merck’s continuing to supply streptomycin if he didn’t comply with Waksman’s wishes and delay publication.

  Feldman understood the warning. He could not afford to offend Merck. Even though ten companies were now working on the drug at various stages of production, they were about six months behind Merck. There was no alternative supply.

  Even so, as soon as he put the phone down, Feldman wrote Waksman, pressing him again to let the Mayo Clinic publish its results. Feldman felt “it would be proper” to prepare a brief report of his observations for publication. This would be only a preliminary notice to establish the Mayo team’s priority in the animal tests, and would be followed by a much more detailed account at a later time.

  “Since you indicated that you had no objections to the publication of a brief report, I will determine that the people at Merck’s do not object ... We will of course send for your information a copy of the manuscript.”

  Waksman did not reply directly; he passed his problem of priority on to Randolph Major, the director of research at Merck. In a gracious but firm letter, Major told Feldman to delay publication. He “quite understood” Feldman’s desire “to publish these very gratifying results,” but he had talked it over with Waksman, who “felt that you might care to wait until the publication of his initial in vitro results.”

  Over Hinshaw’s objections, Feldman agreed to a delay. He had “discussed very frankly the entire situation” with Hinshaw, he told Major. “We are in entire agreement with Dr. Waksman’s wishes that we withhold publication until such time as his report on the in vitro studies is published.” But he hoped this would delay would probably not exceed four to six weeks.

  Hinshaw was upset: He wanted his and Feldman’s work to be acknowledged as soon as possible. This was a breakthrough for them—after five long years of trials. They had completed their initial experiment with four guinea pigs before Schatz had completed his in vitro tests. He didn’t see why they had to wait.

  In the end, Waksman won the day, overwhelmingly. He established priority not only over Feldman and Hinshaw but also over Albert Schatz. The Mayo Clinic invited Waksman to give a lecture at its Rochester, Minnesota, headquarters on October 5, 1944. It was written up in the Mayo’s Proceedings, and Schatz’s results beat Feldman and Hinshaw’s into print.

  On the day of the lecture, the Mayo Clinic director, Dr. D. C. Balfour, introduced Waksman. This is how Waksman later wrote up Dr. Balfour’s remarks.

  “We have with us today a representative from an agricultural institution, one who has no medical degree and has never received any medical training. He is going to address us tonight on a subject that is at the moment of great importance to medical science and clinical practice, dealing with the discovery and application of new chemotherapeutic agents in the treatment of infectious diseases. The fact that you, Dr. Waksman, have been invited to deliver an address before a great medical organization, such as ours, suggests that you are bringing to us a very important message.”

  Then Waksman told the assembled physicians of his background on the Rutgers college farm, of his longtime interest in the actinomycetes, and of his transition from compost and humus to antibiotics. Two events in 1939, he said, “provided the necessary stimulus.” One was a former student’s work on gramicidin—he did not mention René Dubos by name. The other was World War Two and the need for new agents.

  He recalled finding the first antibiotic, actinomycin, but determining that it was very toxic. The second, streptothricin, was “a great step forward,” but it was also toxic. Finally, “in September, 1943, my assistants and I [emphasis added] succeeded in isolating in our laboratory an antibiotic which possessed properties similar to those of streptothricin but was less toxic.” This was streptomycin.

  Albert Schatz was not at the lecture, and Waksman did not mention his name. Feldman and Hinshaw were in the audience, but in Waksman’s retelling of the events, they were not mentioned either. Waksman gave equal importance to streptothricin and strepto
mycin, even though he had already been informed by Merck that, in its opinion, streptothricin was too toxic for use as an injected drug.

  His paper was published on November 15 by the Mayo Clinic, and this time his name came first and Schatz’s last, with Betty Bugie’s in the middle. Schatz’s prospective paper on his in vitro work on H37Rv, where his name would be first and Waksman’s second, was mentioned in a footnote as being “in press.” Feldman and Hinshaw’s work was not mentioned. In addition, Waksman created a myth about how Schatz had isolated his streptomycin-producing strain of A. griseus. He told the Mayo physicians that antibiotics in his lab were isolated according to a “six-step” method that, he implied, he had devised. He wanted to leave the impression that he should be given credit for the way Schatz had isolated his microbe. The first two of the six steps were ways of finding which bacteria were likely to produce antibiotics, including soil enrichment. (The other four steps, including the streak test, involved testing the microbes, once found, to see which germs they were capable of destroying and then how to produce the antibiotic.)

  In Waksman’s Mayo Clinic paper, he said that “both streptothricin and streptomycin were isolated and studied by the use of these procedures.” This was not true, and as the supervisor of Schatz’s thesis, and one of his examiners, he should have known that. The thesis shows that Albert Schatz did not use the first step in isolating his microbe. And when he used the second step, he got no positive results. If he had used only Waksman’s “six steps,” the chances are that he “would not have isolated” A. griseus because those first two steps did not produce a clear zone. He isolated the streptomycin-producing strain through random selection, using the same kind of intuition that a gardener might use to pick a good, healthy plant for breeding. However, Waksman’s myth would create the impression among many observers of his antibiotic program that all his graduate students, including Schatz, followed a strict method, laid out before Schatz started his Ph.D. That Schatz found his microbes through the serendipitous method of random selection and that “perhaps some intuition was needed” would not be acknowledged by Schatz’s peers for another half century.