The Discovery of Insulin

by Michael Bliss

  As he worked, Collip could not have been at all happy about the behaviour of Banting, who seemed to have undercut the group’s arrangements by turning the purification problem into some kind of competition between Banting and Best on the one hand, and Collip on the other. What was the point of this? Especially because there was no true competition, for the trio had already pooled its methods. In making their extract for Thompson, for example, Banting and Best had apparently adopted the improvements Collip had worked out in December, notably the use of a vacuum still and the technique of not evaporating off all the alcohol.46 What were Banting and Best up to in testing that extract on Thompson? Were they hoping to take credit away from the other members of the team, hoping to say they made the extract first used on humans? If so, would they acknowledge that even this process relied on contributions by others?

  Collip might have been very angry at the breach of the experimental plan, perhaps regarding it as a breach of faith or trust. Both Macleod and Collip might well have regretted the scientific blunder and embarrassment of premature testing. Relations between laboratory experimenters and clinicians are seldom without stress. At that time in Toronto, with Duncan Graham’s particularly strong views about clinical experimentation, and in a climate of deep public suspicion (caused by a struggle over appointments) about the university’s relationship with Toronto General Hospital,47 Macleod must have found the Thompson test something of a humiliation.*

  He found it had another disastrous consequence when a day or two later Best came into his office with a reporter from the Toronto Star, the city’s dynamic evening newspaper which was just entering a period of all-out enterprising reporting. In fact it was a sign of the Star’s enterprise that the reporter, Roy Greenaway, had somehow found out about the test on Thompson and was about to scoop the world on this new treatment for diabetes. He had found his way to Best, who thought the best way to handle the situation was to give him to Macleod. Macleod was appalled at the prospect of the impact of premature publicity, especially on diabetics desperate for treatment. He probably urged Greenaway not to publish anything; Greenaway agreed that he would emphasize that the work was preliminary. He did, more or less. The article, appearing on January 14, emphasized Macleod’s cautions. “We’ve really no hope to offer any one at all as yet,” Macleod was quoted as saying. “We don’t know anything yet that would warrant a hope for cure. But we are working intensively at the thing with a hope that some day we may be able to help on a little bit.” Last summer’s experiments had not been new by any means. Hundreds of people all over the world had been working on the problem of sugar and the blood. “At New Haven we were able to report results that were more definite; that was all. We are working very conservatively striving to awaken no false hopes.”48

  To Fred Banting everything about the article, which barely mentioned Best and himself, was a distortion. To understand why, reread the last paragraph from Banting’s point of view. Think about his situation on January 14; think about Macleod’s use of “we.” Banting’s near paranoia about Macleod is surely understandable.

  Macleod probably did not give Banting’s sensitivity any thought, concerned as he was about the trouble the publicity was bound to cause. The Star’s report was picked up by other papers. Within a few days letters started to arrive from diabetics and their relations asking about the extract.49

  Macleod’s innocence of Banting’s suspicions ended a day or two later when Duncan Graham came to see him to report a conversation he had just had with Banting. Banting was accusing him, Macleod, of stealing his work, Graham said. He had been scattering these accusations freely for some time and now wanted to see Macleod.

  Macleod’s first reaction was not to take it seriously. But Graham “impressed me with the serious character of these charges of Banting, and as Banting had been discussing this matter with other people advised me to take steps to put things right.” Macleod immediately went to C.L. Starr, whom he knew Banting trusted. Starr agreed to see Banting. Starr and Banting had a talk; possibly Starr and Macleod had a second talk (Banting’s and Macleod’s recollections of the comings and goings do not agree). Banting was instancing both the New Haven session and the Star interview as evidence of Macleod’s bad faith, and apparently was able to call on others who had been at New Haven to at least corroborate Macleod’s dominance of the meeting.50

  Banting and Macleod finally met. Each claimed later that the other apologized. Macleod told Banting he regretted having taken over the discussion at New Haven, but did it to emphasize the real value of the work. Banting agreed he had misunderstood Macleod’s action, “and assured me that he would do his utmost to undo the harm he had done among his friends.” The two apparently agreed on a modus vivendi, described by Macleod:

  Dr. Banting assured me that he would not misunderstand me in the future and would not conceal any doubts he might have as to whether I was treating him properly. I agreed to continue collaborating with him and I assured him that I had no intention of robbing him of any of the glory that was his due. I agreed further to have the names of those who participated in the researches, then underway in my department, in which the physiological action of pancreatic extracts was being investigated, published with the names in alphabetical order. This placed his name first and Best’s second.

  Banting’s sour 1940 comment on the meeting was that “Macleod thought I had been working overhard, advised a holiday and smoothed everything over with a sticky candy.”51

  V

  While all of this was going on, Collip was working in his lab trying to produce a purified extract. Later in his career J.B. Collip’s skill at extracting hormones made him something of a legend in Canadian medical research. He was part chef, part brewer, part wizard, and, to his critics, part “messer,” as he mixed and filtered, distilled and evaporated, concentrated and diluted, centrifuged and blended. A restless man by temperament, endlessly crisscrossing North America on marathon automobile trips, talking so quickly and disconnectedly people had trouble following him, Collip would finish with one batch of extract and go on to another and another, never making them the same way twice, sometimes working so quickly he had trouble recalling what he had done. It was laboratory research, but the most practical kind of tinkering – a touch of this, a dash of that, what Collip later referred to as “bathtub chemistry.”

  Rough and ready as Collip’s methods were, they were just what the Toronto group needed in January 1922. There are no records of Collip’s trials and progress as he mixed up batch after batch of pancreatic extract, testing each one for potency, perhaps several times at different stages, on his rabbits. We know his starting point was fresh whole beef pancreas ground up in alcohol. Then the permutations and combinations of possible treatments seemed practically endless. How long should it stand before the first filtration? At what temperature? Should acid be added to the alcohol? At what concentration? How should the alcohol be evaporated? How much evaporation? How many more filtrations? What about using other solvents? How do you get the fats and salts out? And so on, and on.

  Actually the chemistry was fairly comprehensible, especially after the fact. The pancreatic tissue consisted of fats and proteins, water, salts, smaller quantities of other organic materials, and the mysterious active principle. Different kinds of proteins were soluble in alcohol at different concentrations and different degrees of acidity. The active principle was soluble in alcohol at the approximately 50 per cent concentration Banting and Best had first hit upon. Would it be possible to find a concentration of alcohol at which the active principle would be still soluble and most of the non-insulin protein contaminants insoluble? Or, for that matter, vice versa – the proteins still dissolved, the active principle precipitated out? The fats could be dealt with fairly easily by known chemical methods, the salts with a little more finesse. While the alcohol was critical to the whole operation, it was a constant problem to get rid of it without also somehow destroying the active principle. Banting and Best’s results seemed t
o show that heat destroyed the active principle in aqueous solution.

  Collip’s method involved gradually increasing the concentration of alcohol in the mixtures, finding that the active principle stayed in solution at higher and higher concentrations, while most of the proteins precipitated out and the lipids and salts could eventually be extracted by centrifuging and washing. It was late on a January night, probably the evening of the 19th,52 when Collip discovered a limit. (He may have been looking for it because of his observations in late December about the potency of the precipitate in one of his early batches.) At a certain concentration of alcohol, somewhere over 90 per cent, the active principle itself was precipitated out. There it was. You could “trap” the active principle (as Collip put it), or isolate it, by first producing the concentration of alcohol in which it was soluble but most of its protein contaminants were not, and then moving to the concentration that would precipitate it. The night he discovered this, Collip wrote in 1949, “1 experienced then and there all alone in the top story of the old Pathology Building perhaps the greatest thrill which has ever been given me to realize.”53 Describing the chemical procedure at a dinner for Collip in 1957, Dr. R.F. Farquharson, Professor of Medicine at Toronto, ended the account of the purification by saying, “As Walter Campbell used to say, Collip then actually saw insulin.”54

  Actually that was an exaggeration, for the powder Collip produced was eventually found to consist of a little active principle in a lot of impurities. But it was far purer than any previous extract. Collip tested its potency on rabbits, waited a few days to check for abscesses, and knew he had an extract that could go back to the clinic. The treatment of Leonard Thompson with injections of pancreatic ex tract, Collip’s extract this time, resumed on January 23.

  VI

  One of the more remarkable personal confrontations in the history of science occurred sometime between January 17 and January 24. There are no contemporary accounts of it, no references whatever by Collip, and only the two following accounts, neither of which should be considered totally reliable. Banting wrote in 1940 as follows:

  The worst blow fell one evening toward the end of January. Collip had become less and less communicative and finally after a week’s absence he came into our little room about five thirty one evening. He stopped inside the door and said “Well fellows I’ve got it.”

  I turned and said, “Fine, congratulations. How did you do it?”

  Collip replied, “I have decided not to tell you.”

  His face was white as a sheet. He made as if to go. I grabbed him with one hand by the overcoat where it met in front and almost lifting him I sat him down hard on the chair. I do not remember all that was said but I remember telling him that it was a good job he was so much smaller – otherwise I would “knock hell out of him.” He told us that he had talked it over with Macleod and that Macleod agreed with him that he should not tell us by what means he had purified the extract.55

  Best, not having read Banting’s account, gave his version of the incident in a letter to Sir Henry Dale, written in 1954 and intended for the historical record:

  One evening in January or February, 1922, while I was working alone in the Medical Building, Dr. J.B. Collip came into the small room where Banting and I had a dog cage and some chemical apparatus. He announced to me that he was leaving our group and that he intended to take out a patent in his own name on the improvement of our pancreatic extract. This seemed an extraordinary move to me, so I requested him to wait until Fred Banting appeared, and to make quite sure that he did I closed the door and sat in a chair which I placed against it. Before very long Banting returned to the Medical Building and came along the corridor to this little room. I explained to him what Collip had told me and Banting appeared to take it very quietly. I could, however, feel his temper rising and I will pass over the subsequent events. Banting was thoroughly angry and Collip was fortunate not to be seriously hurt. I was disturbed for fear Banting would do something which we would both tremendously regret later and I can remember restraining Banting with all the force at my command.56

  Except for a veiled but important reference in Banting’s 1922 account, there are no other useful written records of this incident. Clark Noble once drew a cartoon, unfortunately now lost, of Banting sitting on Collip, choking him; he captioned it “The Discovery of Insulin.”57

  The one surviving artifact of the fight is an agreement signed by Banting, Best, Collip, and Macleod, dated January 25, 1922, and entitled, “Memorandum in Reference to the Co-operation of the Connaught Anti-Toxin Laboratories in the Researches of Dr. Banting, Mr. Best and Dr. Collip – Under the General Direction of Professor J.J.R. Macleod to obtain an Extract of Pancreas Having a Specific Effect on the Blood Sugar Concentration.” The two key conditions of Connaught’s co-operation with the team were:

  1. Dr. Banting, Mr. Best and Dr. Collip each agrees not to take any steps which will result in the process of obtaining an extract or extracts of pancreas, being patented, prepared by any commercial firm with aid of any of the above or otherwise exploited during the period of co-operation with the Connaught Anti-Toxin laboratories. 2. That no step involving any modification in policy concerning these researches be taken without preliminary joint conference between Dr. Banting, Mr. Best and Dr. Collip, and Professor Macleod and Professor Fitzgerald be held.58

  The rest of the document spelled out technical and financial details.

  What had happened? What had Collip said to Banting to cause the attack? Why had he said it? There seems little doubt that Collip said three things that night in the lab to Banting and Best: first, he would not tell them how he had made his breakthrough; second, he had told Macleod, who had agreed that Collip did not have to tell Banting and Best; and third, he might go ahead and take out a patent on his process.

  What was going on in Collip’s mind and what Banting and Best said to him in the course of the conversation can only be speculated upon. He was probably tired – they were all probably tired – after days of hard work and extreme pressure. I presume that Collip and Macleod had little use for Banting’s conduct in the past several weeks, particularly Banting’s breaking of the spirit of the collaboration by himself and Best making the extract for the first clinical test. And, it appeared, Banting had appropriated some of Collip’s improvements in making that extract. Banting had shown his distrust of them; now they had no reason to trust him. It was Collip’s job to purify the extract, not Banting and Best’s. Collip and Macleod may have decided that Banting was trying to take credit away from Collip – that if he knew the process for making the extract he would claim it as his own. They may have believed, after the misadventure of January 11, that Banting could not be trusted not to try to forestall the rest of the team by applying for a patent. Paranoia begat paranoia. So Collip and Macleod decided not to tell Banting and Best the secret of making an effective anti-diabetic extract.

  Speculating further, the kind of things Collip likely said that night are these: “Why should I tell you?…It’s my job, not yours, to get it ready… What do you want to know for? So you can run your own test again?… Stick to your job, I’ll do mine...You’ll know in good time when we see how it works…Don’t worry, you’ll get your share of credit for the work you’ve done...I’m not going to let you take credit for my work…you’ve already tried to do it once… I don’t have to put up with your kind of nonsense… maybe I’ll just go back to Alberta and patent my method….” Collip could not have known Banting very well, could not have known how much of his life revolved around the work, how terribly insecure Banting was at the best of times, how desperately unhappy, suspicious, and frightened he had become as the awful pattern of recent weeks had unfolded, and how the only final outlet this blunt, unsophisticated veteran had for all his frustration and rage was to fight back. Literally.

  The Connaught agreement of January 25 was probably the result of meetings in the day or so after the fight involving the principals, Velyien Henderson as the profess
or Banting trusted, and J.G. Fitzgerald, the director of the Connaught Laboratories. Andrew Hunter, the professor of pathological chemistry, may also have been involved. Again, except for the written agreement, there is no record of these discussions. Banting’s 1922 account suggests Hunter and Henderson supported his view that Collip wanted to patent his process. Henderson’s behaviour and motives in all his dealings with Banting are obscure. A number of fairly detached observers, also Best, thought Henderson deliberately fanned the flames of Banting’s suspicion, perhaps because he intensely disliked Macleod. On the other hand, there may have been some concern on the part of some of the Toronto people, including Fitzgerald of the Connaught,59 that Collip did have a purification process which might be patentable separately from anything anyone else at Toronto had done. Collip was a visitor to the university, free to go back to Alberta at any time, scheduled to leave when the term ended.60 It would be a disaster if he left town taking his knowledge with him, as Banting was insistently and angrily claiming he intended to do. So it was time to tie Collip down, tie Banting and Best down too, and try to settle the whole mess once and for all, by putting the principles of the collaboration down on paper and getting them all to sign it. Then there would be no more need to refer to the unfortunate incident in the lab.61

  VII

  At 11 o’clock on the morning of Monday, January 23, Walter Campbell gave Leonard Thompson five cc. of the new extract made by Collip. At 5:00 that afternoon the boy was given twenty cc. The next day there were two injections of ten cc. each. Thompson’s glycosuria almost disappeared. His ketonuria did disappear. His blood sugar early on the 23rd had been .520. On the 24th it dropped to .120. No extract was given on the 25th and 26th, perhaps while Collip made a new batch. It seems to have been more concentrated, with two four cc. injections becoming the normal daily dose. The urine tests continued to be favourable, “the boy became brighter, more active, looked better and said he felt stronger.”62 This was the first unambiguously successful clinical test of the internal secretion of the pancreas on a human diabetic. Collip’s process worked. Not being a medical doctor, Collip was probably not present at these tests.