Antitumor effect of Melaleuca alternifolia essential oil and its main component terpinen-4-ol in combination with target therapy in melanoma models

Abstract

Essential oils (EOs) have been recently emerging for their promising biological activities in preventing tumorigenesis or progression of different tumor histotypes, including melanoma. In this study, we investigated the antitumor activity of a panel of EOs in different tumor models. The ability of Melaleuca alternifolia (tea tree oil) and its main component, terpinen-4-ol, to sensitize the target therapy currently used for melanoma treatment was also assessed. Our results demonstrated that EOs differently affect the viability of human cancer cells and led us to select six EOs effective in melanoma and lung cancer cells, without toxic effects in human fibroblasts. When combined with dabrafenib and/or trametinib, Melaleuca alternifolia synergistically reduced the viability of melanoma cells by activating apoptosis. Through machine learning classification modeling, α-terpineol, tepinolene, and terpinen-4-ol, three components of Melaleuca alternifolia, were identified as the most likely relevant components responsible for the EO’s antitumor effect. Among them, terpinen-4-ol was recognized as the Melaleuca alternifolia component responsible for its antitumor and proapoptotic activity. Overall, our study holds promise for further analysis of EOs as new anticancer agents and supports the rationale for their use to improve target therapy response in melanoma.

Machine Learning Data Augmentation as a Tool to Enhance Quantitative Composition–Activity Relationships of Complex Mixtures. A New Application to Dissect the Role of Main Chemical Components in Bioactive Essential Oils

“…Using the DA technique, it is possible to reshape unbalanced datasets by augmenting records proportionally to the class occurrences. Given the continuous interest in the dermo-cosmetic field, the proposed method has been applied to 61 EOs whose chemical compositions were already known [3,6] and here investigated for an anti-dermatophyte bioactivity as a model approach. The obtained ML models have been then used to dissect the chemical component's role in the inhibition of the dermatophytes using perturbation-based techniques, impurity-based techniques, and coefficients analysis.…”

“…5 F1 Score is the harmonic mean of the precision and recall. 6 The cross-validated Accuracy. 7 The cross-validated MCC.…”

“…The 61 essential oil have been already chemically characterized by Papa et al [3] and Di Martile et al [6] therefore the reader is referred to their published articles.…”

“…The EOs' market value worldwide is expected to rise from around 17 billion U.S. dollars in 2017 to about 27 billion dollars by 2022 (https: //www.statista.com/topics/5174/essential-oils/ 22 July 2021). Although thousands of scientific articles are published each year on EOs (scopus accessed 15 September 2021) only a few reported the application of machine learning (ML) algorithms with the purpose to dissect the chemical components which are mainly responsible for the associated biological profile to derive quantitative composition-activity relationships (QCAR) [1][2][3][4][5][6], while more commonly the use of ML to EOs was just confined in chromatography, to relate the chemical composition to the retention time of each compound [7][8][9].…”

“…Matthew correlation coefficient. 5 F1 Score is the harmonic mean of the precision and recall 6. The cross-validated Accuracy 7.…”