From: Dave Groves []
Sent: Monday, February 18, 2002 2:15 PM
Subject: Insulin Information
– How is “Human” insulin made?

There are many ways to produce "human" insulin and I will outline my understanding of the various means and who is reported to be using which method(s).

1. Natural human insulin can be extracted, crystallized, and purified from cadaverous human pancreas tissue as it is from beef pancreas and pig pancreas. No one does this, as the few human pancreases available are far to needed for pancreas transplantation, islet cell transplantation, and other research for curing diabetes. Such insulin would likely cost well over $5,000 per vial (it would require well over 3 cadavers per vial) and there would be almost no supply, perhaps enough for 10 diabetics in the world. It would be no better than any other "human" insulin and it would be subject to most, if not all, of the problems that we know exist for many of us with the semi-synthetic, synthetic, and natural animal insulins.

2. Semi-synthetic human insulin can be made, biochemically, from natural pork insulin, by removing the B30 amino acid (Alanine) from pork insulin and substituting a Theronine amino acid in its place. For all vertebrate animals, each insulin molecule consists of precisely 2 peptide chains (A and B) bound together by sulfa bonds at the A7-B7 Cysteine site and at the A20-B19 Cysteine site and there is an additional Cysteine sulfa bond at the A6-A11. All insulin molecules consist of this two chain structure, with an A chain of 21 amino acids and a B chain of 30 amino acids, for a total of 51 amino acid molecules bound by 3 sulfa bonds, any molecule with more or fewer amino acids or different locations for the 3 sulfa bonds cannot be insulin, though it may be (like Lantus/Glargine from Aventis) an insulin-like molecule. There are, further, about 16 mutant forms of human insulin (up to 4 amino acids may be changed to one or more specific other amino acids) produced by human beings' pancreases, though most of the mutant forms seem to cause hypoglycemia or Type 2 diabetes. When Lilly brought rDNA "human" insulin to market in 1981-4, after acquiring the Genentech patents in 1981, Novo and Aventis began producing "human" insulin from pork insulin to compete with the massive advertising program that Lilly has run for the past 20 years. International patent and copyright and trademark law allows about a 20 year monopoly for drug companies which create new drugs, in theory, to allow them to recover their research and development costs. Lilly, allegedly, stole the Genentech patents for producing insulin from rDNA modified e. coli bacteria, and it took about 6 years for Aventis (then Hoechst) to develop a way to brew insulin from yeast cells.

3. Synthetic human and "analog" insulins can all be made in at least 3 different ways, from at least two different species of cells (e. coli bacteria and brewer's yeast cells) by almost identical processes.

a. Pro-insulin can be made by infecting e. coli bacteria or yeast cells with a "virus" that alters their DNA causing them to form cells which excrete pro-insulin. Pro-insulin consists of both the A and B chains of insulin, without the distinct sulfa bonds of insulin, plus the C-peptide chain the last molecule of which binds back to the A chain in a circular formation, like a snake biting its tail. In all of the patents and research papers that I have read, the first generation of "infected" (genetically modified) bacteria or yeasts (germs) must be cultured (grown) and infected in an agar nutrient medium made from beef brain, spinal chord and heart tissues of unknown origin. These germs are then cleaned up and re-infected and then raised to brew pro-insulin in large vats similar to the brewing vats for beer or wine. For the e. coli, the brewing medium is an agar made from the beef and heart tissues of beef, for the yeast brewing, an agar made from beef liver is used, according to the patents, but this may have been changed due to irrational fears of BSE contamination. The resulting pro-insulin is then hydrolyzed to from insulin by removing the C-peptide chain and forming the 3 Cysteine sulfa bonds and the 21 amino acid A chain and the 30 amino acid B chain which is then crystallized and purified. Fortunately, the crystallization of the insulin, followed by a 14 stage high pressure liquid chromatography process, eliminates virtually all possibility of BSE prion contamination for all insulins, regardless of species. I think that Lilly used this procedure initially but that they have gone on to use b or c below though I think that some may still use the procedure, but as you have pointed out, without subpoena power, it is "company confidential" at each of the insulin makers' labs.

b. Insulin can be made by infecting e. coli or yeast cells to produce straight insulin without the C-peptide impurity (which has been related to cardiovascular disease if memory serves me). I do not believe that anyone uses this technique (apparently it is terribly inefficient), but as you have pointed out, without subpoena power, it is "company confidential" at each of the insulin makers' labs.

c. Separate A and B chains can also be made, as described above, then bound back together into insulin molecules and purified. We believe that this is how all of the "human" insulins are currently made and it is probable that the analogs are made in similar fashion. Part of the problem with the analogs is that all 3 (Humalog, Novolog, and Lantus), aside from the known "human insulin problems" that they may share with the "human" insulins, have all caused cancer in the lab animals on which they were initially tested. Additionally, with the so-called "analogs" and specifically with Lantus/Glargine, it is possible that we are creating protein-based prions (the proposed transmission agent for BSE [bovine spongiform encephalopathy] either in the insulin molecules themselves or in the impurities that make it through the purification processes. None of the 3 available analogs has any counterpart in any living vertebrate animal. Most of the impurities are, essentially, unknown proteins of unknown impact on animals and humans other than their capacity to lower blood sugar. Lantus is doubly problematic since it is simply a totally foreign molecule with 52 amino acids, including a change at the terminal A20 site and the addition of 2 molecules of Arginine hooked to the terminal B30 molecule of the insulin molecule, which makes it improper to even call it insulin.

4. All of the factors listed in item 3 are clearly evident from Dr. Brent Hoadley's tremendous patent research work and our continuing medical journal research, and can be well documented.

5. The discovery of the BSE prion raises serious concern over the use of and rDNA synthetic insulins or other medications and may even upset "The Human Genome" project in major ways, since it now appears that DNA is not needed to alter peptide structures as science has theorized since the discovery orRNA and DNA in all living cells and virii.

6. If the concerns in 3-5 can be overcome, the patents reveal that it is just as easy to make recombinant beef or pork insulins as it is to make "human" and the other "analogs" of insulins.

7. There is absolutely no published scientific proof of any clinical benefit to diabetics from any of the semi-synthetic, synthetic, or "analog" insulins.

8. There are mountains of evidence that the new insulins cause major problems for many diabetics.

9. There is not now, nor has there ever been, the alleged shortage of animal pancreatic tissue to produce insulin for humans and animals.

10. Only the "human" R (regular, neutral, or clear) is anything at all like true human insulins, all of the other types contain protamine (made from trout sperm), and/or zinc, and toxic preservatives (even Human R contains these), manufacturing impurities, chemical compounds to buffer or stabilize the insulins and impurities from the rubber stoppers, plastic syringes, coating agents, and the by-products of all of these as they denature and/or oxidize over time. Many of the different insulin molecules form very different crystal structures as a result of the precise order of each of the 51 amino acids in each of the 2 chains. Humalog has exactly the same 51 amino acids as does human insulin. The only difference is that the Lysine, at B28, and the Proline, at B28, are swapped to form an insulin found nowhere in nature (an insulin that allegedly causes cancer in lab animals due to its high affinity for IGF1 receptor sites).

11. If a single molecule of human DNA were unwound and laid out straight, it would be over 6 feet (2 meters) long. While it would be impossible to see because it would be so thin, it is a huge molecule. An insulin molecule would be sub-microscopic, requiring an electron microscope to even measure its length. What we don’t know about the rDNA synthetics is whether or not we are producing infectious proteins (prions) in or even as the synthetic insulin proteins, or whether we are producing by-product, impurity prions or DNA fragments that are being passed along as impurities in the insulin.

12. We do know that the purification process for any insulin would remove any BSE prions in the insulin, and that the pancreas cannot carry BSE prions, while the nutrient medium for the rDNA insulins contain the primary vectors for BSE, Brain, Heart, Spinal Chord, and Liver tissues from cows.

I am passing this along to all on my mailing lists and will publish it to my web site in hopes that it will clarify what is known at this point. I wish to underscore that we do not wish to have any diabetic deprived of any tool that he or she may find useful in controlling this vicious disease.


David Groves, President
Diabetics International Foundation
Please sign our Petition to restore beef insulin to the US at: