In recent years, the rapid development of novel biopharmaceuticals, including oligonucleotide therapeutics, has increased the demand for robust and intensified manufacturing platforms. In downstream operations, stringent purity specifications and the presence of impurities similar to the product of interest pose major manufacturing challenges. In this framework, chromatography emerges as a key enabling technology to achieve these strict quality requirements while simultaneously maintaining industrial standards. However, traditional discontinuous single-column chromatography is limited by its intrinsic yield-purity trade-off and low productivity. Continuous strategies, such as Multicolumn Countercurrent Solvent Gradient Purification (MCSGP), have been proposed to address these limitations by internally recycling off‑specification fractions between two (or more) columns, operating simultaneously, thereby increasing the throughput while reducing waste generation. This thesis aims to develop a mechanistic model for the MCSGP of a recombinant oligonucleotide. The model is intended to support process design and in silico optimization, reducing the experimental effort, buffer consumption, time and overall costs. The mechanistic model of MCSGP was first developed and tested on a reference 20‑mer oligonucleotide. In the simulations, the column dynamics was described by a convective‑dispersive model, while the mass transfer was accounted for through the lumped rate model. In parallel, ad hoc experiments were conducted to calibrate the model for the description of the chromatographic purification of a synthetic pre‑miR‑29b-1 by anion‑exchange chromatography. Specifically, the axial dispersion coefficients of both pre-miR-29b-1 and the modifier, as well as the column porosities (total, intra-particle and inter-particle) were estimated. Additionally, equilibrium and kinetic parameters were identified through the Yamamoto approach, using the stoichiometric displacement model, yielding estimates of isotherm parameters and the overall mass‑transfer coefficient. The resulting validated parameters provide the basis for subsequent model‑driven optimization of the purification of recombinant pre‑miR‑29b-1 produced via bacterial perfusion cultures, ultimately supporting the design of an end-to-end continuous manufacturing platform.
Negli ultimi anni, il crescente sviluppo di biofarmaci, inclusi quelli a base di oligonucleotidi, ha aumentato la domanda di nuovi processi produttivi intensificati. Nell’ambito della purificazione, le stringenti specifiche in termini di purezza, unitamente alla presenza di impurezze con struttura simile a quella del prodotto di interesse, rappresentano degli aspetti critici. In questo contesto, la cromatografia emerge come una tecnologia all’avanguardia, capace di soddisfare i requisiti di qualità, fondamentali nella produzione farmaceutica. Tuttavia, la cromatografia discontinua tradizionale è limitata dal compromesso tra resa e purezza e da basse produttività. Strategie di cromatografia in continuo, come la Multicolumn Countercurrent Solvent Gradient Purification (MCSGP), sono state proposte per superare tali limitazioni, tramite il riciclo interno, in continuo, delle frazioni fuori specifica, tra due o più colonne operanti in simultanea, aumentando la produttività e riducendo gli scarti produttivi. Questa tesi ha come obiettivo lo sviluppo di un modello meccanicistico di un processo MCSGP per la purificazione di un oligonucleotide ricombinante. Il modello è finalizzato al supporto della progettazione di processo e all’ottimizzazione in silico, con lo scopo di ridurre il numero di esperimenti, il consumo di buffer, i tempi e i costi complessivi. Inizialmente, il modello batch è stato testato su un sistema di riferimento per la purificazione di un oligonucleotide a base di DNA, composto da 20 nucleotidi, ed è stato successivamente esteso alla configurazione operativa in continuo. Nella simulazione del sistema, la dinamica della colonna è stata descritta come un modello convettivo-dispersivo, mentre il trasferimento di massa è stato simulato con un lumped rate model. Parallelamente, sono state condotte prove sperimentali a impulso e sfondamento ad hoc per calibrare il modello per la purificazione del pre-miR-29b-1 sintetico mediante cromatografia a scambio anionico. In particolare, sono stati stimati il coefficiente di dispersione assiale del pre-miR-29b-1 e del sale e le porosità della colonna. Inoltre, i parametri cinetici e di equilibrio sono stati identificati tramite l'approccio di Yamamoto, applicato allo stoichiometric displacement model, ottenendo i parametri dell'isoterma e il coefficiente complessivo di trasferimento di massa. I parametri così validati, insieme al modello MCSGP sviluppato, forniscono le basi per il design e l’ottimizzazione di un processo di purificazione del pre-miR-29b-1 ricombinante prodotto mediante fermentazione batterica in perfusione, a supporto della progettazione di una piattaforma produttiva continua integrata.
Mechanistic modeling for the continuous oligonucleotide chromatographic purification via MCSGP
Oliveti, Beatrice
2024/2025
Abstract
In recent years, the rapid development of novel biopharmaceuticals, including oligonucleotide therapeutics, has increased the demand for robust and intensified manufacturing platforms. In downstream operations, stringent purity specifications and the presence of impurities similar to the product of interest pose major manufacturing challenges. In this framework, chromatography emerges as a key enabling technology to achieve these strict quality requirements while simultaneously maintaining industrial standards. However, traditional discontinuous single-column chromatography is limited by its intrinsic yield-purity trade-off and low productivity. Continuous strategies, such as Multicolumn Countercurrent Solvent Gradient Purification (MCSGP), have been proposed to address these limitations by internally recycling off‑specification fractions between two (or more) columns, operating simultaneously, thereby increasing the throughput while reducing waste generation. This thesis aims to develop a mechanistic model for the MCSGP of a recombinant oligonucleotide. The model is intended to support process design and in silico optimization, reducing the experimental effort, buffer consumption, time and overall costs. The mechanistic model of MCSGP was first developed and tested on a reference 20‑mer oligonucleotide. In the simulations, the column dynamics was described by a convective‑dispersive model, while the mass transfer was accounted for through the lumped rate model. In parallel, ad hoc experiments were conducted to calibrate the model for the description of the chromatographic purification of a synthetic pre‑miR‑29b-1 by anion‑exchange chromatography. Specifically, the axial dispersion coefficients of both pre-miR-29b-1 and the modifier, as well as the column porosities (total, intra-particle and inter-particle) were estimated. Additionally, equilibrium and kinetic parameters were identified through the Yamamoto approach, using the stoichiometric displacement model, yielding estimates of isotherm parameters and the overall mass‑transfer coefficient. The resulting validated parameters provide the basis for subsequent model‑driven optimization of the purification of recombinant pre‑miR‑29b-1 produced via bacterial perfusion cultures, ultimately supporting the design of an end-to-end continuous manufacturing platform.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_03_Oliveti_Tesi.pdf
non accessibile
Descrizione: Testo tesi
Dimensione
11.62 MB
Formato
Adobe PDF
|
11.62 MB | Adobe PDF | Visualizza/Apri |
|
2026_03_Oliveti_ExecutiveSummary.pdf
non accessibile
Descrizione: Testo Executive summary
Dimensione
1.24 MB
Formato
Adobe PDF
|
1.24 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/253609