Determinación de boldina en cultivos celulares de peumus boldus mediante ATR-FTIR apoyado con métodos de extracción de alcaloides.
Loading...
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad de Concepción
Abstract
Los cultivos celulares ofrecen una alternativa prometedora para la producción de boldina y otros alcaloides a partir de Peumus boldus (boldo), pero se requieren herramientas complementarias para aumentar el rendimiento y estandarizar la producción. La espectrometría es una herramienta poderosa para el análisis de compuestos bioactivos pero no ha sido explorada en boldo. El objetivo de este estudio fue diseñar un modelo de predicción de la concentración de boldina en una matriz compleja de compuestos extraídos por un solvente a partir de celulares. Para ello, se analizaron utilizando HPLC para obtener las concentraciones reales y la variable dependiente, datos fundamentales para diseñar y entrenar el modelo. Los mejores rendimientos en la extracción de la boldina se obtuvieron con metanol. Se empleó ATR-FTIR para obtener los espectros, sobre los cuales se aplicó una regresión de mínimos cuadrados parciales (PLS). De esta manera, el modelo alcanzó una precisión en la predicción entre el 95.1% y 94.8% en una data externa. Los espectros infrarrojos revelaron que el grupo amino y éter presentes en la boldina son responsables de las características espectrales relevantes, no obstante, la huella dactilar de la boldina (1700 – 700 cm-1) entrego los mejores resultados. En conclusión, los espectros infrarrojos obtenidos de extractos metanólicos de cultivos celulares de Peumus boldus, tratados con métodos quimiométricos, permiten predecir la concentración de boldina con un alto nivel de precisión. Estos modelos serán útiles para acelerar los procesos de optimización de producción de boldina mediante cultivos celulares a nivel industrial.
Accurate prediction of boldine alkaloid concentration is essential to improve production and optimisation in cell cultures of Peumus boldus. In Chile, extraction of boldine from leaves threatens the existence of this vulnerable native species. Cell cultures offer a promising alternative, showing potential for alkaloid production, but require optimisation for proper selection of yield strategies and standardisation of production. The aim of this study was to design an optimal model predicting boldine concentration in a complex matrix of methanol-extracted compounds in cell callus. For this purpose, Peumus boldus cell cultures were freeze-dried and extracted in methanol. Subsequently, they were analysed using HPLC to obtain real concentrations and the dependent variable, fundamental data to design and train the model. ATR-FTIR was used to obtain the spectra, on which a partial least squares (PLS) regression was applied. In this way, the model achieved a prediction accuracy between 95.1% and 94.8% on external data. The infrared spectra revealed that the amino and ether groups present in boldine are responsible for the relevant spectral features, however, the boldine fingerprint (1700 - 700 cm-1) gave the best results. In conclusion, infrared spectra obtained from methanolic extracts of Peumus boldus cell cultures, treated with chemometric methods, allow predicting boldine concentration with a high level of accuracy. These models could be useful to standardise optimisation processes and boldine production at industrial level.
Accurate prediction of boldine alkaloid concentration is essential to improve production and optimisation in cell cultures of Peumus boldus. In Chile, extraction of boldine from leaves threatens the existence of this vulnerable native species. Cell cultures offer a promising alternative, showing potential for alkaloid production, but require optimisation for proper selection of yield strategies and standardisation of production. The aim of this study was to design an optimal model predicting boldine concentration in a complex matrix of methanol-extracted compounds in cell callus. For this purpose, Peumus boldus cell cultures were freeze-dried and extracted in methanol. Subsequently, they were analysed using HPLC to obtain real concentrations and the dependent variable, fundamental data to design and train the model. ATR-FTIR was used to obtain the spectra, on which a partial least squares (PLS) regression was applied. In this way, the model achieved a prediction accuracy between 95.1% and 94.8% on external data. The infrared spectra revealed that the amino and ether groups present in boldine are responsible for the relevant spectral features, however, the boldine fingerprint (1700 - 700 cm-1) gave the best results. In conclusion, infrared spectra obtained from methanolic extracts of Peumus boldus cell cultures, treated with chemometric methods, allow predicting boldine concentration with a high level of accuracy. These models could be useful to standardise optimisation processes and boldine production at industrial level.
Description
Tesis presentada para optar al grado de Magister en Ciencias Forestales
Keywords
Boldo, Espectroscopía infrarroja, Compuestos bioactivos