Tomatoes and potatoes more resistant to bacteria: how Ai revolutionises plant defence

However, the bacteria are constantly changing, modifying their flagellin to evade recognition and thus bypass the plant's defences. This is where artificial intelligence comes into play: the researchers used AlphaFold 3 to generate three-dimensional models of FLS2 receptors present in plant species such as wild grapevine, Chinese oak, star jasmine and soybean, naturally also capable of recognising flagellin variants. The AI model made it possible to identify 13 amino acids on the receptor surface, which are essential for the recognition of the most elusive bacterial forms.

By specifically modifying these sites using synthetic biology techniques, the team reprogrammed the FLS2 receptor in plants that were previously unable to detect certain pathogens, making them capable of activating an effective immune response even in the presence of previously 'invisible' bacterial variants.

The results, validated both in the laboratory and in the greenhouse, showed that the modified plants develop an enhanced immune response, capable of acting even in the presence of infections already in progress, without side effects on plant growth.

"We were able to resurrect a defeated receptor, one where the bacterium had won, giving the plant a chance to resist the infection in a much more targeted and precise manner," said Gitta Coaker, professor of plant pathology and lead author of the study.

According to the researchers, targeted modification of immune receptors could be a new strategy to strengthen crops without resorting to invasive transgenic techniques. At present, the main objective is to make plants more resistant to bacteria such as Ralstonia solanacearum, which can infect more than 200 plant species and cause huge losses for agriculture.

The UC Davis group is already working on developing machine learning tools - a branch of artificial intelligence in which algorithms learn from data - to identify other key receptors and further reduce the number of genetic modifications required.

In a rapidly changing agrifood sector, AI is emerging as a technology that can redefine crop protection strategies, improve plant resilience and open up new perspectives on sustainability and competitiveness.