Related Documents
Banting & Best, The Internal Secretion of the Pancreas, Journal of Laboratory and Clinical Medicine 7:251 (1922)
In a previous essay I wrote about the Edmonton Protocol paper as a diabetes classic. I realized that the second classic I needed to discuss had to be Banting & Best’s discovery of insulin. I had never read it. Then I discovered how hard it was to get.
It is usually easy to find a landmark scientific paper. A google of “watson & crick paper” yields the 1953 Nature announcement of DNA’s structure as the first hit. But “banting & best paper” yields a picture of the scientists but not their paper. I ended up going out the UCSF library archives and making a copy myself, available in PDF form above. I also looked up the first report of purified insulin, which was available from the publisher, the National Academy of Sciences, also available above.
As a man working on discovery research related to insulin, I find much in these papers striking. Perhaps the most amazing thing is how crude the insulin preparations are. The standard today for claiming you have isolated a new compound is ‘purification to homogeneity’. That simply means you have a solution that contains essentially only one thing, your new compound. And by tradition even a compound purified to homogeneity is suspect: if you have a pure compound why haven’t you been able to make a crystal? Scientists have a deep belief that crystals are pure and simple. If you make a crystal, you are certain you have highly purified compound, otherwise a crystal will not form. Finally you can easily measure the melting point of the crystal. This is important because it clearly establishes your priority. Melting point can be estimated, but not very accurately. When others repeat your work and find the same melting point, you are the proven winner in the race.
Banting & Best don’t claim purification to homogeneity, only biological activity in dogs. Abel reported crystallizing insulin four year later. “Professor Swartz of the Department of Geology of the Johns Hopkins University has kindly examined samples of the insulin crystals…and reports them as being very uniform in character suggesting a high degree of purity, as double refracting and as belonging to the rhombohedral division of he hexagonal system.” Below is the first published photograph of an insulin crystal, proving it to be a single and new compound.
I find it interesting that Abel does not cite the work of Banting & Best. He probably seethed when they got the Nobel Prize without crystalizing insulin.
Another surprise is the narrative informality of the report. They describe what they did, why they made certain choices, what went wrong, things that they did not do that might subsequently investigated, and useful tips. Science papers today are written from a Godlike perspective, and almost always in the passive voice. We all go to science meetings to talk to investigators to find out what they really did. Like reading the Banting and Best paper.
Finally the biggest shock is their method of preparing ‘insulin’ — really pancreatic extract. Banting’s new idea was, rather than repeat using extracts from whole dog pancreas, to ligate the pancreatic duct and allow the exocrine pancreas — the part making digestive enzyme — to degenerate. After a few weeks he knew from prior work that the degenerated pancreas still had functioning islets.
So his extract was made as follows:
- Ligate the pancreatic duct of a dog. Wait a few weeks.
- Kill the dog with chloroform and remove the degenerate pancreas.
- Slice it into physiological saline (Ringer’s solution). [Acid extracts work better than neutral or basic.]
- Chill and partly freeze; macerate the pancreatic slices.
- Filter through paper.
- Raise the filtrate to room temperature and inject into the vein of a dog. Measure blood sugars.
Crude indeed.
This first paper reports injections of 75 doses into ten animals. “Since the extract has always produced a reduction of the percentage sugar of the blood and of the sugar excreted in the urine, we feel justified in stating that this extract contains the internal secretion of the pancreas.”
After the Toronto work, many laboratories began to purify these crude extracts to make insulin. The Abel paper is the important crystalline insulin milestone. Abel reports that he started with “highly impure and complex, though therapeutically serviceable insulin extracts of commerce…” revealing that in 1926 the first insulins were, by today’s standards, fantastically impure. Today making such a crude material available for human therapy would be absolutely impossible.
Interestingly, at the time I wrote my first Wall Street report on the diabetes industry in 1970 the insulin purification quest was coming to a close. The final impurity was proinsulin, a byproduct of insulin synthesis in the beta cells that is very difficult to remove from insulin preparations. The standard Lilly insulins had about 100 parts per million (ppm) proinsulin. Novo Industri of Denmark introduced insulin purified so that proinsulin was undetectable (less than 1 ppm). Lilly followed and soon everyone was taking ultra-pure insulin.
It took 50 years.
Dear Scott-
Thank you for continuing this important work. My mother, grandmother and aunt all had Type II diabetes and i have just been diagnosed with pre-diabetes. I realize this is a seriously different animal than Type I diabetes- still, I am now personally interested in all diabetes research.
Olivia
It is interesting that after von Mehring and Minkowski conclusively demonstrated in Berlin in 1889 that diabetes was caused by a lack of something that the pancreas was producing, it took another 32 years to isolate a clinically effective sample of what this was. Zuelzer and Paulescu probably isolated insulin before Banting and Best did, but Zuelzer stopped using it on patients after he found it caused what he thought was a massive anaphylactic shock, which was in fact probably severe hypoglycemia, and Paulescu later claimed he forgot about developing his discovery because “he was distracted by World War I.” Admittedly, World War I was interesting, but one could have hoped that that a scientist would have found improving the lot of diabetics equally interesting.
I am not sure that the ultimate purification of insulin was entirely a good thing, however. The early, impure samples of insulin used contained about 10% c-peptide, which came to be regarded as a useless metabolic by-product of insulin production. Recently, however, researchers have found that it may well hold the key to preventing diabetic complications, independently of the blood sugar control achieved by insulin treatment. See: Y. Ide, et al, “Prevention of Vascular and Neuronal Destruction in Diabetic Rats by C-Peptide,” Science, 277(5325) 563 (1997).
I must politely disagree on the purity of insulin. When I was first diagnosed I started on then-standard Lilly pig insulin. Highly purified insulin’s did not become standard until the 1980’s. My first insulin had enough lipase to damage the fat cells at the injection site, so so I still have divots in my skin. But if there were zero lipase and equimolar c-peptide that would be OK.
I am of two minds about the c-peptide claims. On the one hand it is naturally secreted with insulin in equimolar amounts, suggesting it has some function. On the other hand c-peptides between species rapidly diverge, whereas insulin is highly conserved, suggesting that the c-peptide function is not specific or maybe non-existent. The Ido paper is intriguing, but it was twelve years ago, in rats. Can you cite a c-peptide clinical study you find persuasive? I have not seen one.
Lots of people were looking for the anti-diabetes factor in the pancreas in the early 20th century. It think it shows how primitive protein chemistry was. Nowadays it would be solved in months, if not a week.
C-Peptide supplementation seems to harm type 2 diabetics by promoting atherosclerosis, since they already have too much insulin and too much c-peptide in the circulation. (D. Walcher and N. Marx, “Advanced Glycation End Products and C-Peptide Modulators,” Seminars in Immunopathology, 31(1) 103 (2009); L. Nordquist and M. Johansson,”Proinsulin C-Peptide: Friend or Foe in the Development of Diabetic Complications?” Vascular Health Risk Management, 4(6) 1288 (2008))
But in type 1 diabetics, c-peptide supplementation seems to improve nerve function, limit the progression of incipient nephropathy, and promote microcirculation. (B. Johansson, et al, “Beneficial Effects of C-Peptide in Incipient Nephropathy in Patients with Type 1 Diabetes Mellitus,” Diabetes Medicine, 17(3) 181 (2001); J. Wahren, et al, “Role of C-Peptide in Human Physiology,” American Journal of Physiology, Endocrinology, and Metabolism, 278(5) E759-68 (2008); J. Wahren, “C-Peptide: New Findings and Therapeutic Implications in Diabetes,” Clincial Physiology and Functional Imaging, 24(4) 180 (2004); K. Ekberg, et al, “C-Peptide Replacement Therapy and Sensory Nerve Function in TYpe 1 Diabetic Neuropathy,” Diabetes Care, 30(1) 71 (2007); T. Forst, et al, “Role of C-Peptide in the Regulation of Microvascular Blood Flow,” Experimental Diabetes Research, 176245 (2008) Article ID 384219)
One company has been working for years to commercialize an injectable c-peptide product for diabetics, but progress has been unaccountably slow. Human c-peptide of injectable quality can already be obtained from chemical supply companies, and the appropriate replacement dose for type 1 diabetics has already been narrowed down to 1.5 mg to 4.5 mg/day in four divided doses. However, since the latest prices I have seen for it amount to about $200 per mg, and I don’t happen to have a spare $320,000 a year to spend on it at the moment, it will be some time before I can report the results of any experimental trial on myself.
It speaks volumes about the negative effects of the tunnel vision of most researchers, who have been obsessed with blood glucose control to the exclusion of all other possible methods to interrupt the long chain of events causing diabetic complications, that early studies of c-peptide in the 1980s dismissed it as useless because it did not improve blood sugar control! Preventing complications by lowering blood sugar levels is difficult to achieve, costs a huge amount of time and energy, and can cause lethal or permanently damaging episodes of severe hypoglycemia, but few seem prepared to address these problems by finding ways other than blood sugar control to reduce diabetic complications.
You might consider adding “Invisible Frontiers: The Race to Synthesize a Human Gene” by Stephen S. Hall to your list of classics regarding type 1 diabetes. Its a story of how biosynthetic “human” insulin (its not really human, no more than artificial vanilla is the same as real vanilla to a chef) was created and the various challenges and lengths that were gone through to create it, and the story of big egos and whatnot that were involved in doing this. A classic that should be added to the list; now out-of-print, but should be available via your local library without too much trouble.