Building therapies where it once seemed impossible

The work of Andrea Cavalli, Group Leader at the Institute for Research in Biomedicine (IRB) in Bellinzona (affiliated with Università della Svizzera italiana), is defined by the balance between understanding how things work and then trying to build them. "I see myself first and foremost as a scientist, driven by curiosity," he says. "That is why I studied physics: I wanted to understand how the world works." But alongside curiosity sits another dimension: that of construction. "I also like to design and build things that work," he explains. "In this sense, I feel a bit like a molecular engineer."

Below is the in-depth report produced by IRB, in collaboration with laRegione. (Italian only)

Pushing the limit

At the IRB, his group works at the meeting point of artificial intelligence, structural biology, and biomedicine. The goal is to tackle problems that, for years, were considered unsolvable, such as the possibility of intervening in certain key proteins involved in tumour development. Among these is Erg, a protein involved in about half of all prostate cancers. "It is a transcription factor, which is to say a protein that regulates the production of other proteins," Cavalli explains. "In cancer, these regulatory mechanisms are altered: Erg begins to activate the wrong genes and cells start to proliferate uncontrollably. Reducing Erg activity, therefore, means stripping the tumour of one of its main engines, but doing so has proven to be anything but simple." For a long time, proteins like Erg were considered difficult to target with traditional drugs because they lack structures that therapeutic molecules can easily "latch onto" and are "hidden" inside the cells. The work of Cavalli's group stems precisely from the attempt to overcome this limit.

Instructions for use

The breakthrough came from a very small modification with significant effects: methylation. This involves adding a chemical group to a specific site on the protein. "In the case of Erg, this modification increases its activity and makes it more aggressive," Cavalli explains. A discovery made together with colleagues from the Institute of Oncology Research (IOR), Giuseppina Carbone, and Carlo Catapano: Erg methylation is a specific characteristic of cancer cells, absent in healthy ones. It is precisely this difference that paves the way for targeted therapies, capable of selectively hitting the tumour without affecting normal cells. Having identified this modified form of Erg as a specific target of cancer cells, the group developed an antibody that recognises and binds to it. In this way, the antibody blocks the transcription factor's function, halting tumour growth.

Another innovative element that distinguishes this potential drug is its method of administration. Instead of administering the therapeutic molecule directly, researchers introduce the genetic instructions that encode it into cells. "We don't administer the antibody directly," Cavalli explains. "We provide the cells with instructions, in the form of DNA or RNA, to produce it internally, somewhat like what happens with RNA vaccines." This approach makes it possible to overcome one of the main limitations of antibody-based therapies, which normally act outside the cells. In this case, however, the action takes place inside, where many of the most difficult targets to reach, such as Erg itself, are located. If confirmed in clinical trials, this strategy could open a new therapeutic modality, particularly for tumour-relevant proteins that have until now been considered "untreatable". "It could be the first example of such an approach against transcription factors," Cavalli observes.

Towards uncertainty and beyond

The group's research is built upon the deep integration of computational methods and laboratory trials. Artificial intelligence allows for the design of novel molecules and the rapid characterisation of various solutions. However, these hypotheses require empirical verification prior to clinical application. This process is governed not just by technical progress, but by a paradigm that embraces risk as a fundamental driver of innovation. "If we only stick to what we know for certain, discovery remains out of reach," Cavalli notes, recalling how many technologies that are established today—from RNA therapies to peptide drugs—were considered unfeasible until only a few years ago. In this sense, research becomes a continuous exploration, in which the boundaries of what is possible shift over time and must be continually tested.

Beyond basic research lies the challenge of translating discoveries into concrete applications. Following the initial laboratory development phase, the path to the patient remains arduous: safety must be verified, optimal dosing defined, efficacy evaluated, and results confirmed through rigorous clinical trials. It is a complex, strictly regulated path that requires significant time and resources, as every new treatment must not only demonstrate its potential but also, crucially, its safety profile. Embracing uncertainty is therefore essential—not just for discovery, but for clinical translation. From the development of new molecules to their clinical application, every stage carries inherent risks that must be recognised and supported, including in terms of funding and research policies. This development can be carried out by large pharmaceutical companies or by biotech startups backed by investors. In our case, this second path was chosen, leading to the founding of MethylX. A project that already received its first major recognition in 2025, when MethylX participated in the Boldbrain Competition run by Fondazione Agire in collaboration with the USI Startup Centre, ranking second and also winning the public prize. Outside the laboratory, Cavalli enjoys simple activities in contact with nature. "I run several times a week and, when I can, I go to the mountains," he says. It is a way to maintain a necessary distance from a job that deals with complex problems with a strong impact on people's lives.