Self Expression Magazine

Fantastic Yeasts (and Bacteria) – How Insulin Manufacturing Transformed Biotechnology

Posted on the 28 January 2025 by Jhouser123 @jhouser123

I want to take a brief interlude in the series to talk about how a fantastic drug becomes a fantastic drug. The story of insulin, covered in the previous article, can teach us many lessons about drug discovery and commercialization. The advancements in purification and analytical techniques and the paradigm shift in characterizing and producing animal-derived therapeutics still have relevance a century later.

Today I want to talk about how the advent of recombinant protein production led us to create an entirely new class of tools: genetically modified cells that produce bioactive products. This co-opting of biology for industrial production shows how science transitions to useful technologies, further enabling interesting science. In many ways, insulin production was the proving ground that inspired scientists to engineer life itself to serve our needs—sometimes in ways that the original discoverers could never have imagined.

Extractions and Excipients

The incredible part about the discovery of insulin is that they had pretty much no idea what the substance actually was before creating a commercialized product. Early preparations were simply an acetone extraction from pancreas tissue followed by an alcohol phase separation. The final product was only “pure” in the sense that it didn’t cause harmful reactions in animals or people once injected (as opposed to other preparations which very much did). When insulin production was brought to commercial scale, they further refined the method (using pH-based precipitation which selectively purifies proteins by their isoelectric point) but still did not completely understand what was in the vial.

It took over a decade from the official launch of full-scale insulin production for researchers at Johns Hopkins University to prove that a single molecule was the active ingredient in the products that were being sold. This seemingly simple realization transformed the way we think about drug manufacturing. Now instead of chasing the product down a series of extractions (hoping for good recovery and removal of excipients) we can rationally design purification processes that not only purify our specific product but remove specific problematic impurities along the way. This model now underpins the engineering of downstream purification systems for every drug on the market.

Cleaning up the Feed Stream

The journey from crude product to final purified material can be an arduous one, and the reality is that the “complexity” of the crude product plays a crucial role in how you design your purification strategy. In pharmaceutical bioprocesses, thankfully, the input product is never sewage, but rather a relatively innocuous mix of proteins, lipids, nucleic acids, and salts, the vast majority of which you need to get rid of. When the starting point is the discarded pancreases of cows and pigs there is a lot to remove, and the risks of an impure product are high.

One of the major milestones in this early insulin era was the drive to reduce the immunogenic responses some patients encountered when they received porcine (pig) or bovine (cow) insulin. Although these forms are structurally similar to human insulin, subtle differences in amino acid sequences triggered reactions in certain individuals. The shortage of suitable animal pancreas sources—and the rising demand as diabetes diagnoses climbed—further fueled the quest to find a better way to produce insulin reliably that removed the livestock pancreas from the equation entirely.

The partnership between Genentech and Lilly to solve this problem yielded an incredible innovation: a genetically engineered bacteria strain that stably synthesized human insulin. Combined with recent advances in chromatography, the product could achieve purities even higher than the most advanced extraction-based methods, and by 1982 the first recombinant protein drug, Humulin, was FDA approved. This opened the door to an entirely new way of thinking about drug development. Now instead of being confined to the substances we could find in nature in enough abundance to purify, we could engineer biomolecules and produce them at scales large enough to commercialize nearly any protein one could think of.

In many ways, this shift was a watershed moment not just for diabetes treatment but for the entire pharmaceutical industry. Suddenly, supply constraints tied to animal organs were gone, and the potential for creating recombinant human proteins—ones more likely to be tolerated by patients—was front and center. This transition also accelerated improvements in fermentation, bioreactor design, and quality control processes, all of which laid the groundwork for the modern biotech sector.

The Unsuspecting Single-Cell Factory

All this innovation was made possible by something viruses have been exploiting for probably as long as cells have been around: a cell is just a factory waiting for instructions. Normally it gets them from the genome, but if you can get some recombinant nucleic acid in there you can hijack the process. In fact, we are able to do this so efficiently that the cell will stop almost all other normal functions to make your product under the right conditions.

Bacteria happen to be one of the easier types of organisms to manipulate in this way because horizontal gene transfer by plasmid is relatively common in the kingdom. It also helps that translation and transcription are so closely linked that there are very few checks on protein production (as compared to mammalian cells with long-lived, extensively modified RNA). The E. coli strains used in modern biomanufacturing applications are selected for their growth characteristics and production yields, and can be scaled to tens of thousands of liters of fermentation volume for the production of insulin and other products.

While the use of E. coli revolutionized insulin production by enabling large-scale synthesis of human insulin, it introduced certain limitations. One major drawback is the inability of E. coli to perform post-translational modifications, such as glycosylation, which are essential for the proper folding, stability, and bioactivity of many biologic drugs. This limits its utility for more complex proteins beyond insulin. Additionally, E. coli secretes few, if any, proteins, requiring extensive downstream purification to separate the target product from intracellular debris and endotoxins, which can provoke severe immune reactions in humans. The scalability of bacterial systems, while impressive, also comes with challenges in maintaining consistent yields and preventing contamination during fermentation processes, making alternative hosts worth exploring for biologic manufacturing.

Why Yeast?

Yeast systems, particularly Saccharomyces cerevisiae and Pichia pastoris, address many of the limitations posed by E. coli. Unlike bacteria, yeast cells are eukaryotic, enabling them to perform essential post-translational modifications, which opens the door to producing a wider range of biologics. Moreover, yeast secretes proteins into the growth medium, simplifying purification and reducing the risk of contamination by intracellular components such as endotoxins. Yeast also offers robust growth in scalable fermentation systems, withstanding the high-density cultures required for cost-effective production. The choice between yeast and bacterial systems involves many factors, including the complexity of the protein, the desired post-translational modifications, and the economic considerations for large-scale manufacturing.

The flexibility and capabilities of yeast as a production platform are paving the way for new innovations in biologics. Advances in synthetic biology and genome engineering now allow scientists to tailor yeast strains for specific production needs, including optimizing secretion pathways or introducing rare post-translational modifications. Yeast can also be engineered to produce complex molecular assemblies like virus-like particles and nanobodies. Furthermore, yeast systems are being explored for their potential in producing personalized medicines, where rapid strain engineering could allow the on-demand synthesis of individualized biologics. As techniques for strain optimization and bioprocessing advance, yeast is set to play an increasingly central role in the future of biopharmaceutical innovation.

One additional upside for yeast is its centuries-long history in fermentation (think beer and bread). This familiarity means we already have a foundational understanding of how yeast grows and how to manipulate growth conditions. While we’ve obviously pushed far beyond the conditions of your average bakery or microbrewery, that cultural knowledge has helped guide industrial-scale processes and troubleshooting. It’s a small but fascinating example of how ancient practices gave modern biotechnology a running start.

What About Mammalian Systems?

Woah, woah, woah. Don’t get ahead of me. It makes total sense though: if we use yeast because the post-translational modifications are more human-like, then why not just use human cells to get the rest of the way there? Well, I don’t mean to end on a cliff hanger, but this will be the topic of another interlude. Mammalian cells indeed hold the key to producing some of the most complex and high-value therapeutics on the market today—but they also come with their own set of hurdles, from cost to complexity of growth media to regulatory considerations.

Stay tuned, because this Friday we introduce monoclonal antibodies which are the next step in the evolution of clinical biologics. This major leap forward creates a world of possibilities for harnessing the power of the immune system. This is where we will meet the challenges and fascinating possibilities of mammalian production systems.

Stay curious!


Back to Featured Articles on Logo Paperblog