Building the best possible mini-liver (without making it too complex)
How do organs work in the body, and how can we create mini-organs to study diseases and test new medicines? That’s the idea behind organ-on-a-chip technology. During his PhD, Flavio Bonanini worked on developing the best possible mini-livers. ‘Make them as simple as possible, and as complex as needed.’
‘For me, the model was successful when I saw that the cells arranged themselves in a way similar to a real liver.’ During his PhD, Flavio Bonanini worked on creating the most realistic liver model possible. He performed his research at Mimetas, a Leiden-based company specialising in making miniature versions of organs (organ-on-a-chip).
Bonanini did not have to start from scratch; his work built on decades of research. Scientists began with simple, flat layers of cells in petri dishes. Over time, these evolved into 3D clusters that better mimicked real tissue. To improve these models even further, research introduced different cell types, biochemical signals, and supportive structures. Bonanini’s focus was to add blood vessels to his mini-liver. ‘Blood vessels play a crucial role in organ function,’ he explains.
Growing blood vessels from scratch
To introduce blood vessels, Bonanini studied how they form naturally. When tissue needs a new blood supply—due to injury or tumor growth—it releases chemical signals to attract vessels. ‘We grew a tiny blood vessel and guided it toward a chemical source. Once it reached the target, we removed the signal and introduced mini-livers, tricking the vessels into growing within the tissue.’
Another approach mimicked embryonic development, where immature cells self-organise into blood vessels. ‘With the right conditions, vessels formed naturally within the liver tissue. Even more interestingly, liver cells recognized these vessels and positioned themselves correctly, just like in a real liver.’
Blood flowing through mini-livers
Where there are blood vessels, there must be blood—or at least a liquid that mimics it. ‘We use a nutrient-rich fluid to support cell growth,’ Bonanini explains. To create flow, he didn’t use pumps. ‘They’re complex and impractical.’ Instead, he used gravity-driven flow: tilting a platform back and forth to move liquid through the vessels without extra equipment.
Sometimes, a simple model is the better choice
Using real blood could be possible, but it complicates the model unnecessarily, making it harder to interpret results. The key is to control each variable. If you want to test how a molecule interacts with a system, that should be the only factor you change. ‘You can always add more elements—blood, new cell types, even a nervous system—but do you need that complexity to answer your question?’ Sometimes, a simpler model works best.
One of the biggest challenges in drug development is liver toxicity. ‘The liver processes many drugs, and toxicity is a major reason why medications fail. Animal models don’t always predict human liver toxicity well because metabolism differs across species. If our model can detect toxicity that animal models miss, that would be a major breakthrough.’
Industry versus academia
Doing a PhD at a company has both challenges and advantages. ‘In an industry PhD, you engage with academic research but also work on practical applications,’ Bonanini says. Industry moves fast—results matter. At the same time, working in a company means collaborating beyond research, with teams in business development, marketing, and regulation. ‘It was inspiring to see my work contribute to something with real value for pharmaceutical companies and toxicology research.’
PhD defence
Flavio Bonanini defended his thesis Microphysiological liver systems for in vitro modeling and industry implementation on 3 April. His supervisors were Thomas Hankemeier and Dorota Kurek.