Abstract

In this research, in-silico and in-vitro approaches were adopted with the aim to investigate the relationship between the tobacco leaf structures (trichomes) and the production of secondary metabolites with antimicrobial activity. Machine learning techniques were used to know the correlation between phenotypic traits and the production of secondary metabolites in Nicotiana tabacum plants. Then, an in-vitro experimental study was carried out to corroborate the proposed model. The relationship between the morphology and distribution of the different types of trichomes in the upper and lower leaves with the contrasting profiles of the chemical composition (diterpenes and sugar esters) of the leaf exudates between different lines of tobacco were found. We determined the influence of each trichome type with secondary metabolites production and the necessary concentration to achieve antimicrobial and antioxidant activity.