Insulin's 100-Year History Inspires Today's Biotech Industry
By Jeffrey C. Baker, Ph.D.
I’ve given seminars on the history of insulin manufacturing for years. They have grown more special to me recently as we celebrate the 100th anniversary of insulin, the first protein therapeutic, and the 40th anniversary of the first rDNA therapeutic, recombinant human insulin.
The industry also recently marked the 25th anniversary of the first insulin analog, a non-natural sequence protein, which was not a replacement therapy but a protein designed to have properties far better suited to therapeutic use than one secreted from a gland.
The insulin story shows how necessity and curiosity in the past give insight into today's issues of scale-up and continuous manufacturing. The creators of commercial insulins built and rolled out novel ways to harness automation, new delivery methods, and monitoring technology. They did it all decades ago. We stand on the shoulders of giants. Their vision and stories can inform our own.
Insulin’s Leaders Built The Modern Biotech Industry
A seminar on the history of insulin manufacturing is a joyful presentation to give.
I love telling the stories of well-known scientists like George Best and Frederick Banting, still the youngest winner of the Nobel Prize in Medicine. In a time when the thought leaders of the day were searching for a diabetes germ, they boldly relieved diabetic symptoms by injecting a pancreatic extract later found to contain a protein called insulin. These injections enabled a quality of life for diabetics in a time when the standard of care therapy was a draconian, life-prolonging starvation diet in an asylum. I like telling the story of Josiah Lilly and George Clowes’ vision of and commitment to large-scale biomanufacturing. Thirty years before the Kefauver-Harris Amendment, they engaged more than 7,000 patients and doctors in testing their insulin prior to bringing it to market. I like telling the stories of clinicians Marie Krogh and Hans Christian Hagedorn and engineers Thorvald and Harald Pederson, who revolutionized insulin formulations by first collaborating, then forming cooperatively competing companies that remerged in 1989 to form NovoNordisk.
I find my enthusiasm takes off when I tell stories of how science was transformed by insulin manufacturing. There’s big-concept stuff like the idea of a metabolic intervention (hey, we still haven’t found that diabetes germ!) or diagnostic blood tests, or even sterile filtration or prefilled syringes in 1927.
There’s the stuff that gives goosebumps to biochemistry geeks: the first use of isoelectric precipitation, the first protein sequenced, the first protein synthesized, the first protein crystallized, the first identification of a protein precursor or “prohormone” that was post-translationally modified by enzymes to give its active form, the concept of species-specific isoforms. We found insulin works by binding to another protein called a “receptor,” which undergoes a conformational change and activates a kinase that makes yet more post-translational changes. That motif is central to our understanding of metabolic regulation today.
Biomanufacturing technologies were also nurtured by less well-known names. In the 1920s, the insulin process quit working in Toronto, and a chemical engineer named George Walden set it right with a systemic root-cause analysis that brought the process back up and increased purity a hundredfold. His work revolutionized record-keeping with standardized lab notebooks and operational effectiveness using instructional records called “tickets.” Another less-known innovator, Richard “Doc” Jackson, put the first chromatography columns in place to purify insulin by size and ion exchange in the 1970s, driving the manufacture of the families of resins and supports used today. Insulin scientists did things I had been taught simply could not be done. For example, Bruce Frank showed that you can renature denatured protein, and in a 20,000-liter tank. And then there was Jan Markussen, who converted pork insulin to human insulin by using trypsin to first make selective cleavages and then add a new sequence back through transpeptidation. Before we had cell phones and chatbots, William “Wild Bill” Muth would get coded updates on his belt pager from the insulin fermenters informing him of changes to dissolved oxygen, glucose feeds, and other automated process changes seemingly just to keep a human in the loop.
In 1981, diabetics in the United States alone needed the glands of 56 million animals — 10,000 pounds of pancreas — to make enough insulin to meet their diabetic needs. In 1982, three years after the first human insulin crystals were made from genetically engineered E. coli, and one year after the first human dose, biosynthetic human insulin was approved for market in the U.S., U.K., and Germany, the first therapeutic form rDNA technology. This extraordinary accomplishment was achieved by combining process engineering with molecular life sciences, disruptive pharmaceutical manufacturers with open-minded, patient-oriented regulators, and an educated patient population who embraced the science of this unprecedented healthcare solution.
Yesterday’s Lessons Inform Today’s Narratives
As I give these talks, I find the old stories resonate with contemporary issues. Women scientists and engineers respond to the unstoppable Marie Krogh, herself a diabetic, who, after visiting Banting and Best in 1923, founded the Nordisk Insulinlaboratorium. There’s also Dorothy Hodgkin, who received the Nobel Prize for her work in X-ray crystallography and, after 35 years, solved the 3D structure of insulin. And I can’t forget Nancy Jones, a woman with an associate degree who is a co-inventor on the patent for crystallizing the first insulin analog. This work later helped show the role of phenolics in stabilizing insulin hexamer structure, in turn leading to novel insulin formulations.
Analytical biochemists nod when they hear about insulin potency being determined by physicochemical methods rather than animal testing. Sometimes their jaws drop when they hear this precedent was set in the 1980s.
Process engineers’ raise their eyebrows when I talk about 24/7 asynchronous, semi-continuous batch manufacturing. In 1995, insulin manufacturers were harvesting banks of 40,000-liter fermenters every 14 hours into a five-column downstream process. They had five chemistry steps, from enzymatic cleavage to protein folding, managed with online/at-line process analytics. Again, that was in 1995.
Stories for the regulatory scientists and quality professionals show how, for decades, manufacturing led statute and guidance rather than the other way around. They hear how “the Insulin Amendment” of 1941 laid the groundwork for cGMPs and introduced the new concept of FDA user fees. They hear about the subtleties of how statute deemed insulin a biologic in 2020. The FDA published a Final Rule the same year that defined a protein as, “Any alpha amino acid polymer with a specific defined sequence that is greater than 40 amino acids in size,” a decision pivoting around insulin and insulin analogs. These stories and others in the insulin saga inform today’s regulatory constructs and quality management systems and, I hope, catalyze discussion on how we can together drive innovation in regulation and quality oversight as much as in operational practice.
It has been suggested that you can’t draw a perfect line between yesterday’s challenges and what biotechnology faces today. I am reminded that today’s molecules are more complex, and contemporary market factors drive aggressive timelines and reward the shedding of headcount and capital assets. All these things together have had a chilling effect on bold innovation.
Companies will tell me the revolutionary application of 21st-century technologies in commercial biomanufacturing is “not realistic” today because of global regulatory constraints. Sometimes I put my hands in my pockets and grimace, but other times I observe that some things don’t change — anyone who aspires to manipulate the molecular basis of life to save lives has always had to be at ease making big, risk-based decisions in informed ignorance. Constraint doesn’t crush creativity; constraint drives it.
We Need To Know Our Biotech Stories And Make Ourselves A Part Of Them
As I give my insulin presentation at conferences and seminars, audience members act surprised as they remember that biotech was more than mAbs from CHO. The insulin story shows them how the linkage of unit operations and throughput was being managed by automation and algorithms before most of them were born. They hear how insulin innovators developed personalized medicine, real-time response with wearable monitors, and pumps that employed dozens of engineered hormones to address lifestyle needs. In their parents’ era, diabetes was a death sentence. Now, the biotech solutions are so common that insulin pricing can be weaponized in political arenas, and the press can shrug off global recalls of Asian-made insulin biosimilars because of mislabeling just as they would any other generic drug quality issue.
Their surprise has a profound impact on me. How can these stories surprise anyone? What does this tell us about how we learn? What does this say about what we teach? I think about how the Library at Alexandria, contrary to popular myth, was not burned in one barbaric cataclysm but declined through decades of decreased funding, a victim of empire and theocracies and, even then, messaging. We speak to the value of tearing down functional silos and interdisciplinary innovation yet, in manufacturing, profoundly reward divestment and specialization and build temporal silos around the next milestone in the name of focus, fully segregated from historical lessons and cross-sector learning of the past as well as end-in-mind, outcomes-based thinking framing the future.
Lessons From The Past Inform Challenges Of The Present
Lessons for today’s priorities are waiting. In a time when sustainability is a manufacturing priority, we should not forget that for decades, insulin manufacturing was based on process intensifying solid/liquid separations, on continued elimination of solvents, and reusing or eliminating nitrogen emissions. Insulin was manufactured in tanks that were cleaned and reused rather than buried or incinerated. Shipping a kilogram of insulin crystals had a fraction of the expense or carbon footprint of shipping liquid biologic presentations that are mostly water in weight and volume.
In a time when patient centricity and outcomes-based therapies are topics of focus, dozens of insulin analogs are now available in dozens of presentations and remain sufficiently available. Public outrage now is about the pricing rather than the deaths due to shortages. If we are to have a data-driven and long-overdue discussion about the cost of diabetes therapy, I think it is marvelous we can do so while this must-have protein is commoditized rather than a subject of crisis-driven production and distribution. In our time of digital medicine development, we can learn from the parents who monitor their children’s health in real time with smartphones linked to wearable insulin pumps and glucose monitors.
In this day of rethinking critical workforce needs in biomanufacturing, we need to reframe the difference between being qualified for a job on paper and being suited for a job in fact. I was fortunate to learn from people like Nate Utley, a bioprocess development scientist at Eli Lilly, who crystallized the first biosynthetic human insulin 40 years ago because “we only had one lot, and he had the best hands of all of us.” He was the right guy in the right job and was not, thank goodness, vetted by a keyword search of his resume. You see, Nate didn’t have a college degree, but he had experience, diligence, and commitment to delivering quality in whatever he did.
Understanding our past, and how we got to where we are, is critical to mapping our future. The history of insulin manufacturing offers lessons of innovation in biopharmaceutical manufacturing and its place in our society. We need to listen.
About the Author:
Jeffrey Baker, Ph.D., worked in the biopharmaceutical industry for over 20 years in process development and manufacturing and served for 10 years as Deputy Director in the FDA’s Office of Biotechnology Products, CDER. He has retired from FDA but remains active in the biotech community working with NIIMBL and NASEM and several universities.