Alex Philippidis Senior News Editor Genetic Engineering & Biotechnology News
As Part of Its 35th Anniversary, GEN Recalls the Breakthrough Introduction of Humulin
When Eli Lilly launched Iletin®––the world's first commercially available insulin product for diabetes control in humans—the company relied on material harvested from cows and pigs. That was in 1923. And the company continued to source material from the slaughterhouse for decades. By the early 1970s, however, a study group at Eli Lilly determined that the insulin resources needed to meet patient demands would surpass sources of animal insulin in about 20 years.
“We were all becoming aware that the curve for the number of diabetics, and therefore the insulin demand from that group, would be crossing over the curve of pancreatic availability somewhere toward the end of the 1990s,” recalled Bruce H. Frank, Ph.D., who retired from Eli Lilly in 2000. During his 34 years of service, he not only held senior scientific positions in insulin discovery and development, he also influenced manufacturing and regulatory affairs.
Dr. Frank also noted that at a 1977 forum of the National Academy of Sciences, Irving S. Johnson, then vice president of research with Lilly Research Laboratories, posited several possible alternatives to animal insulin—including drugs mimicking insulin, surgically transplanted pancreases, transplanted islets, chemically synthesized insulin, tissue culture, and insulin production by genetically manipulated cells.
“The group that Lilly management put together [was charged with evaluating] increased production options,” elaborated Dr. Frank, who is now a consultant as well as vice president of development at Thermalin Diabetes. “A report went to management, after which there was more discussion along the lines of, ‘Well, we’ll start work here on the possibilities and keep current on developments.’”
That work grew into the effort to develop a human source of insulin safe and effective enough to have supplanted animal-derived versions. In October 1982, Humulin® won FDA approval as the first marketed human healthcare product derived from recombinant DNA (rDNA) technology.
“Humulin was the first therapeutic recombinant protein which established the feasibility of recombinant proteins,” J. Leslie Glick, Ph.D., an independent corporate management advisor, told GEN. “Humulin established that recombinant human proteins could have therapeutic efficacy, which meant that, essentially, as the science or the technology progressed, you could start thinking about larger and larger proteins.”
Extending a Legacy
Humulin extended an Eli Lilly legacy in diabetes that stretched back to 1921, when Eli Lilly research director George Henry Alexander Clowes, Ph.D., D.Sc., attended an American Physiological Society meeting. At that meeting, University of Toronto researcher Frederick Banting, M.D., and his assistant Charles Best described how they extracted insulin from the pancreases of dogs, then injected it into a severely diabetic dog named Marjorie. The injections, the scientists reported, significantly prolonged Marjorie’s life.
In 1922, Dr. Banting and Best (who later earned his Ph.D.) administered the extract to a 14-year-old boy who had type I diabetes. The insulin saved the boy’s life and proved the hormone’s efficacy in treating human diabetes. The following year, Eli Lilly introduced to market Iletin, which was discontinued in 2005.
“What we had seen in the ’70s and early ’80s was a fair percentage of patients who generated antibodies against insulin,” Robert E. Ratner, M.D., chief scientific and medical officer for the American Diabetes Association, told GEN. One manifestation, Dr. Ratner said, was insulin allergy resulting in hives, a condition almost unheard of with human insulin. The other was neutralizing antibodies.
“If you make antibodies to the insulin, and they bind up the insulin, then the insulin doesn’t work, and so you have individuals who require hundreds, if not thousands of units of insulin, in order to overwhelm the antibodies and have adequate biological response,” Dr. Ratner explained. “Human insulin gets around all of that. Since it’s not a foreign protein, the body generates very few if any antibodies to it. So that was the primary advantage.”
Another advantage to human insulin, he added, is its greater purity compared with animal-sourced counterparts, and its ability to minimize other potential proteins that come out of the pancreatic islets.
To begin producing human-sourced insulin, researchers had to to overcome several hurdles. One was technological, synthesizing the human insulin gene via rDNA. Another was winning regulatory approval for an rDNA product at a time when the fledgling technology sparked debate among researchers over the possibility of creating pathogens dangerous to humans and the environment. A third hurdle was logistical, producing sufficient quantities free of contamination to meet demand from millions of people with diabetes.
The first rDNA human insulin was synthesized by Genentech on August 24, 1978, just two years after the biotech’s founding, after combining A- and B-chains that were individually expressed in E. coli using chemically synthesized genes. The following day, Genentech and Eli Lilly signed an agreement to commercialize rDNA human insulin. Lilly agreed to use Genentech’s plasmids containing the synthetic A- and B-chain genes to develop and commercialize rDNA human insulin, dubbed “biosynthetic human insulin” (BHI).
“Since Lilly had been a premier supplier of insulin throughout its history, Lilly management concluded that Lilly really needed to move into this technology,” said Dr. Frank. This decision, added Dr. Frank, was more than the recognition of a business opportunity. It also reflected the company’s awareness that the supplies had to be available for future diabetes treatments.
According to Dr. Fran, Eli Lilly focused on scaling up and developing efficient processes for manufacturing, drawing on decades of commercial production experience. The company committed to developing a process of manufacturing A-chains and B-chains separately, then learning how to chemically combine them to produce recombinant human insulin.
“Because Lilly had been a major antibiotic manufacturer for years, fermentation was a real strength of the company,” Dr. Frank noted. “If you combine that with what at that time was probably 50 years of experience with purifying and isolating and characterizing insulin, we were in the position to take the initial workup and take it to the scale that we needed for commercialization purposes.”
Initial fermentations were restricted to 10 L or less, requiring multiple E. coli K12 fermentations to obtain just a few milligrams of each insulin chain. By 1979, Eli Lilly won approval from the Recombinant Advisory Committee for 150 L fermentations which, combined with an improved chain combination method, yielded small but significant amounts of BHI. The rDNA human insulin was identical in all respects to standard pancreatic human insulin, enabling rapid scale-up and advancement into clinical evaluation.
During those early years of rDNA development, only inactive protein could be expressed in E. coli—an outgrowth of the guidelines for rDNA research hammered out by some 150 researchers at the 1975 Asilomar Conference on Recombinant DNA. Conferees agreed that E. coli K12 should be engineered to prevent the colonization of the human digestive tract were the engineered organism to escape into the outside world.
The first rDNA human insulin was administered to volunteers at Guy's Hospital, London, in July 1980. Nearly two years later in May 1982, Eli Lilly submitted NDA 18-790 for BHI to the FDA for approval—which came less than six months later.
The Right Protein
“The FDA got its act together,” Dr. Glick said. “But Genentech knew that human insulin would have an advantage over the pig insulin for that small percentage of people who could not tolerate the pig insulin. Genentech picked the right protein, which made FDA comfortable with dealing with recombinant proteins. Subsequently, a whole bunch of approvals followed.”
During Humulin’s first two to three years on the market, Dr. Ratner noted, some patients developed hypoglycemia because of the lack of antibody binding that occurred.
“There were those who suggested that the antibodies might be a benefit in prolonging the effect of insulin, and diminishing hypoglycemia,” Dr. Ratner said. “The reality was, you simply had to adjust the dose. By adjusting the dose downward, you achieved all of the benefits you wanted without any of the antibody side effects.”
Humulin’s approval laid the groundwork for human insulins launched in the decades since.
In 1987, Novo (now Novo Nordisk) launched a replacement for its pig-derived insulin, a human insulin produced by the yeast Saccharomyces cerevisiae. Nine years later, Eli Lilly won FDA approval for the first insulin analog, Humalog® (insulin lispro injection [rDNA origin]), whose inventors included Dr. Frank.
In 2014, Eli Lilly and Boehringer Ingelheim won tentative FDA approval for Basaglar™ (insulin glargine injection), a basal insulin intended to provide long-lasting blood sugar control in between meals and overnight. The tentative approval is subject to an automatic stay of up to 30 months because Sanofi has sued Eli Lilly, accusing it of infringing four of its patents for Lantus®; Lilly has denied wrongdoing.
“The most amazing thing about this recombinant technology is the doors that this opened,” Dara Schuster, M.D., senior medical director for the U.S., Lilly Diabetes, told GEN: “Once we had the pure, durable, and consistent product developed, we could ask, ‘If we were to make this insulin better, what would that look like?’”
“We learned that one of the first things we needed to do with insulin was make it work faster so that it would be closer to the body’s own insulin action,” Dr. Schuster added.
Humalog allowed users to have more flexibility in the timing of their insulin injections with meals rather than waiting 30 minutes to eat with Humulin. Another fast-acting insulin, Novo Nordisk’s NovoLog®/NovoRapid® (insulin aspart [rDNA origin] injection), was approved in 2000. It is indicated for use within 5 to 10 minutes of the start of a meal.
Among long-acting basal insulins, Sanofi’s once-daily Lantus (insulin glargine [rDNA origin] injection 100 units/mL) was first to win FDA approval in 2000. Last year, Lantus lost the U.S. patent protection that accorded exclusivity to the drug’s active ingredient; Sanofi has since launched Lantus’ successor, the more concentrated once-daily Toujeo® (insulin glargine injection 300 units/mL). In between came Novo Nordisk’s Levemir (insulin detemir [rDNA] injection), approved in 2005 and allowing once-daily or twice-daily in divided doses.
In May 2015, the FDA approved Eli Lilly’s pre-filled KwikPen® containing a more concentrated 200 units/mL formulation of Humalog, the agency’s first approval of a concentrated mealtime insulin analog.
“There’s so much room to continue to explore, investigate, and improve that I think treating diabetes is going to look a lot different even three to five years from now,” Dr. Schuster concluded.