Biotech and AI help win a production challenge: it is the beginning of a technological revolution
2018 was a turning point for Dompé farmaceutici. "The increase in sales volumes of our first biotech drug, has made possible investments in R&D that aim to achieve results of flexibility and productivity unthinkable until a few years earlier" underlines Domenico Bonanni, New Technologies and Innovation Specialist at Dompé in L'Aquila. The active ingredient of this drug is a recombinant protein produced by genetically modified microorganisms, known as rhNGF (recombinant human Nerve Growth Factor. The growing demand for this drug for the treatment of a rare disease previously without a cure, has raised new problems: the production of NGF needs to remain compliant with very strict quality criteria, it's so easy to increase the volumes.
Improving the productivity and robustness of the process was not a simple challenge. Since 2018, the company has been producing rhNGF (recombinant human NGF, which is a molecule identical to that naturally synthesized by the human body) on an industrial scale extracting it from Escherichia coli, a common bacterium also presented in the human intestine. Using molecular biology techniques, a gene sequence coding for human NGF was inserted into a sample of E. coli, inducing it to produce the NGF protein. Once modified, a small population of microorganisms is induced to replicate through cell division in a fermenter. At the end of this process the number of bacteria will have increased exponentially as well as the amount of recombinant protein produced. At the end of the fermentation, repeated chromatographic cycles allow to extract and purify the protein.
"It is a complex process – says Bonanni - and hard to model as it involves living organisms, making it difficult to scale the volume of the final product". The key to working successfully with living organisms is, in fact, having full control over their growth environment - culture media, pH, oxygen, temperature, growth rate and so on - because an alteration of just few parameters could compromise the amount and quality of the protein produced. It goes without saying that while controlling a single bacterium is quite straightforward, controlling hundreds of millions of cells per cubic milliliter is a different ballgame”.
"Our team could have followed an academic approach by developing an analytical model to describe (and then control) the fermenting process - says the researcher - but pursuing this path would have required a timing incompatible with our industrial research project”.
The solution was a combination an Industry 4.0 approach with biotechnologies. In the Fall of 2020, Bonanni’s team in L’Aquila installed eight micro-fermenters equipped with an artificial intelligence system, creating a smaller plant running in parallel to the production one. Its aim: collect and analyze process data through a dedicated Ai system. "The biological processes taking place inside each fermenter generate a huge amount of data used to train Neural Networks to recognize the parameter values maximizing the yield of the fermentation - explains Bonanni - This “machine learning” approach, has become the cornerstone of many artificial intelligence systems used in industry ".
Artificial Intelligence (Ai) is the field of computer science dedicated to the design of hardware systems and development of software giving machines abilities considered until recently be the exclusive of living beings as, for example, visual recognition, the perception of space-time and many decision-making skills. To learn and interact correctly with the surrounding environment, modern Ai systems continuously acquire and process information, literally “learning” through data. From self-driving vehicles to real-time translations, Ai is changing the way we interact, making every interface intuitive and "intelligent". However, it shouldn’t be forgotten that precisely due to the enormous potential for change, Ai is still a controversial discipline today. "I fear the day when technology will go beyond our humanity: the world will then be populated by a generation of idiots," Einstein once said. For many, this is a well-founded fear that can only be exorcised by putting technology to the service of humanity and aligning ethics and innovation. "This requires an interdisciplinary approach that brings together ethics, politicians and technicians under the same roof - comments Bonanni - Only in this way will we be able to foresee all the possible consequences of the impact of these technologies and impose a direction of sustainable development to the proliferation of possible applications ".
What has been done with micro fermenters is a small step in this direction. Thanks to a cutting-edge project, co-financed by the Italian Ministry for Economic Development, it was possible to combine mathematics, biology, engineering and AI but above all, and without taking it for granted, human intelligence.