Development of analytical methods based on Near Infrared Spectroscopy for monitoring of pharmaceutical and biotechnological processes and control of new psychoactive substances

Abstract

L’espectroscòpia de l’infraroig proper (NIR per les seves sigles en anglès) és una tècnica analítica basada en la interacció de la radiació electromagnètica en el rang 780-2500 nm amb la matèria. L’espectre NIR es pot considerar una empremta única per cada substància o barreja d’aquestes, format per bandes resultants dels sobretons i combinacions de les vibracions fonamentals de la regió de l’infraroig mig. Addicionalment, en mostres sòlides, l’espectre NIR és sensible a la dispersió causada per les característiques físiques del producte, de manera que ofereix simultàniament informació de canvis físics i químics. L’aprofitament de la informació que contenen els espectres NIR implica l’aplicació de mètodes d’anàlisis multivariable. L’objectiu general d’aquesta tesis ha estat el desenvolupament de mètodes analítics basats en NIR pel seguiment de processos farmacèutics i biotecnològics, així com pel control de noves substàncies psicoactives (NPS per les seves sigles en anglès). Per a fer-ho, s’han explorat les següents condicions: i) extensió de rangs de concentració de principi actiu (API per les seves sigles en anglès) en l’aplicació de l’estratègia de l’espectre de procés (PS per les seves sigles en anglès) durant les etapes de granulació i compactació de preparats farmacèutics sòlids; ii) identificació de NPS utilitzant instruments portàtils i de laboratori, i iii) seguiment de la producció de Lipasa B recombinant de Candida antarctica en Pichia pastoris utilitzant glicerol com a font de carboni.

La espectroscopía de infrarrojo cercano (NIR, por sus siglas en inglés) es una técnica analítica basada en la interacción entre la radiación electromagnética en el rango 780-2500 nm y la materia. El espectro NIR puede considerarse una huella única para cada sustancia o mezcla de ellas, compuesta por bandas resultantes de los sobretonos y combinaciones de las vibraciones fundamentales en la región del infrarrojo medio. Adicionalmente, en muestras sólidas, el espectro NIR es sensible a la dispersión causada por las características físicas de las muestras, de modo que ofrece simultáneamente información sobre cambios físicos y químicos. El aprovechamiento de la información contenida en espectros NIR implica la aplicación de métodos de análisis multivariable. El objetivo general de esta tesis ha sido el desarrollo de métodos analíticos basados en NIR para el seguimiento de procesos farmacéuticos y biotecnológicos, así como para el control de nuevas sustancias psicoactivas (NPS, por sus siglas en inglés). Para ello, se han explorado las siguientes condiciones: i) rangos extendidos de concentración de ingrediente activo (API, por sus siglas en inglés) en la aplicación de la estrategia del espectro de proceso (PS, por sus siglas en inglés) durante etapas de granulación y compactación de preparados sólidos farmacéuticos; ii) identificación de NPS utilizando instrumentos portátiles y de laboratorio, y iii) seguimiento de producción recombinante de Lipasa B de Candida antarctica en Pichia pastoris usando glicerol como fuente de carbono.

The Near Infrared Spectroscopy (NIRS) is an analytical technique based on the interaction of electromagnetic radiation in the wavelength range 780-2500 nm and matter. The NIR spectrum can be considered as a “fingerprint” of each chemical compound or mixture of them, which contains absorption bands that are the result of overtones and combinations of the fundamental vibrations observed in the Mid-Infrared region. Additionally, NIRS of solids is sensible to the scattering effect caused by physical characteristics of the samples, therefore provides simultaneous sensitivity to chemical and physical changes of solids. Because of NIR spectra show broad and overlapped bands makes it necessary the use of Chemometrics, which implies the application of statistical and mathematical methods to such spectral data, for achieving the maximal extraction and collection of useful information from it. The general objective of this doctoral thesis is developing analytical methods based on NIRS for monitoring pharmaceutical and biotechnological processes and for the control of illicit drugs. The following conditions have been considered i) extended active pharmaceutical ingredient (API) concentration ranges during granulation and tableting using the process spectrum (PS), ii) on-site identification of new psychoactive substances (NPS) during police seizing procedures with hand-held and bench-top instruments and iii) inline monitoring of the production of recombinant Lipase B from Candida antarctica in Pichia pastoris using Glycerol as carbon source. i) Extended API concentration ranges during granulation and tableting using the PS The PS is a methodology for preparing calibration sets by adding the changes due to the manufacturing process to NIR spectra of samples prepared at the laboratory, using an algebraic procedure. The PS has successfully included the process contributions during modelling in the central point of API concentration values of diverse formulations, however the properties of this methodology at extreme points of API concentration ranges have not been studied yet. For evaluating such properties, in this work the PS was applied to samples in the range of ± 30% of a nominal API value. Results have shown that the PS performance can be affected by API concentration changes in the studied range, and classical pre-treatments are not enough to overcome this condition. 4 ii) Comparison of the performance of bench-top and hand-held NIR instruments concerning the identification of NPS The NPS are 'legal highs' with molecular differences regarding the structures of illicit controlled drugs, whose emergence have expanded the current synthetic drugs market in a very important way. The feasibility of using portable NIRS instruments for the fast identification of NPS have been previously demonstrated, however, their performance has not been faced to the performance of bench-top instruments. Results presented in this thesis expose that, even when models developed using data from NIRS miniaturized instruments are limited in performance regarding those developed using data provided by bench-top instruments, classification models of NPS based on data from hand-held instruments can be useful to make real-time and on-site decisions that can be confirmed later using high performance analytical instrumentation. iii) Inline monitoring of the production of recombinant Lipase B from Candida antarctica in Pichia pastoris using glycerol as carbon source. The use of new constitutive promoters and recycled carbon sources in the recombinant production of industrial proteins, such as lipases, in the cell factory Pichia pastoris is advantageous for improving production yields and minimizing the cost of the culture medium. The capabilities of a NIR spectrometer with fiber optic coupling for immersion of a transflectance probe were employed for the inline monitoring of the cultivation mentioned in the headline. Quantitative models have been developed for Biomass, Total protein, Nitrogen and Activity, which have demonstrated better prediction capability during the feed batch stage than during the batch stage. Predictions of glycerol values has been probably affected by the formation of hydrogen bonds in the aqueos medium.