Understanding the effects of system differences for parameter estimation and scale-up of high throughput chromatographic data

Simultaneous prediction of 16 quality attributes during protein A chromatography using machine learning based Raman spectroscopy models

Several key technologies for advancing biopharmaceutical manufacturing depend on the successful implementation of process analytical technologies that can monitor multiple product quality attributes in a continuous in-line setting. Raman spectroscopy is an emerging technology in the biopharma industry that promises to fit this strategic need, yet its application is not widespread due to limited success for predicting a meaningful number of quality attributes. In this study, we addressed this very problem by demonstrating new capabilities for preprocessing Raman spectra using a series of Butterworth filters. The resulting increase in the number of spectral features is paired with a machine learning algorithm and laboratory automation hardware to drive the automated collection and training of a calibration model that allows for the prediction of 16 different product quality attributes in an in-line mode. The demonstrated ability to generate these Raman-based models for in-process product quality monitoring is the breakthrough to increase process understanding by delivering product quality data in a continuous manner. The implementation of this multiattribute in-line technology will create new workflows within process development, characterization, validation, and control.

Mechanistic modeling of cation exchange chromatography scale-up considering packing inhomogeneities

In process development and characterization, the scale-up of chromatographic steps is a crucial part and brings a number of challenges. Usually, scale-down models are used to represent the process step, and constant column properties are assumed. The scaling is then typically based on the concept of linear scale-up. In this work, a mechanistic model describing an anti-Langmuirian to Langmuirian elution behavior of a polypeptide, calibrated with a pre-packed 1 ml column, is applied to demonstrate the scalability to larger column volumes up to 28.2 ml. Using individual column parameters for each column size, scaling to similar eluting salt concentrations, peak heights, and shapes is experimentally demonstrated by considering the model's relationship between the normalized gradient slope and the eluting salt concentration. Further scale-up simulations show improved model predictions when radial inhomogeneities in packing quality are considered.

Demonstration of continuous gradient elution functionality with automated liquid handling systems for high-throughput purification process development

Various high-throughput systems and strategies are employed by the biopharmaceutical industry for early to late-stage process development for biologics manufacturing. The associated increases to experiment productivity and reduction in material consumption makes high throughput tools integral for bioprocess development. While these high-throughput systems have been successfully leveraged to generate high quality data representative of manufacturing scale processes, their data interpretation often requires complex data transformation and time-intensive system characterization. With respect to high throughput purification development, RoboColumns by Repligen operated on Tecan automated liquid handling systems offer superior performance scalability, but lack an optimized liquid delivery system that is representative of preparative chromatography. Particularly, stock Tecan liquid handling systems lack the capability to provide high-capacity continuous liquid flow and ideal linear gradient chromatography conditions. These limitations impact protein chromatography performance and hinder the application of high-throughput gradient elution experiments. In this work, we describe a Tecan Freedom EVO high-throughput purification tool that provides more continuous liquid delivery enabling continuous gradient elution capability for RoboColumn experiments as demonstrated by generation of highly linear conductivity gradients. Results demonstrate that the tool can provide RoboColumn performance and product quality data that is in agreement with larger, bench scale chromatography formats for two model purification methods. The described gradient purification method also provides more consistent performance between RoboColumns and larger column formats compared to step elution methods using the same optimized Tecan system. Lastly, new insights into the impact of discontinuous flow on RoboColumn elution performance are introduced, which may help further improve application of these data towards bioprocess development.

Parameter-by-parameter method for steric mass action model of ion exchange chromatography: Simplified estimation for steric shielding factor

Mechanistic models play a crucial role in the process development and optimization of ion-exchange chromatography (IEC). Recent researches in steric mass action (SMA) model have heightened the need for better estimation of nonlinear parameter, steric shielding factor σ. In this work, a straightforward approach combination of simplified linear approximation (SLA) and inverse method (IM) was proposed to initialize and further determine σ, respectively. An existed, unique, and positive σ can be derived from SLA. Compared with linear approximation (LA) developed in our previous study, σ of the multi-component system can be calculated easily without solving the complex system of linear equations, leading to a time complexity reduction from O(n³) to O(n). The proposed method was verified first in numerical experiments about the separation of three charge variants. The calculated σ was more reasonable than that of LA, and the error of elution profiles with the parameters estimated by SLA+IM was only one-sixth of that by LA in numerical experiments. Moreover, the error accumulation effect could also be reduced. The proposed method was further confirmed in real-world experiments about the separation of monomer-dimer mixtures of monoclonal antibody. The results gave a lower error and better physical understanding compared to LA. In conclusion, SLA+IM developed in the present work provides a novel and straightforward way to determine σ. This simplification would help to save the effort of calibration experiments and accelerate the process development for the multi-component IEC separation.

Parameter-by-parameter method for steric mass action model of ion exchange chromatography: Theoretical considerations and experimental verification

Ion exchange chromatography (IEC) is one of the most widely-used techniques for protein separation and has been characterized by mechanistic models. However, the time-consuming and cumbersome model calibration hinders the application of mechanistic models for process development. A new methodology called "parameter-by-parameter method (PbP)" was proposed with mechanistic derivations of the steric mass action (SMA) model of IEC. The protocol includes four steps: (1) first linear regression (LR1) for characteristic charge; (2) second linear regression (LR2) for equilibrium coefficient; (3) linear approximation (LA) for shielding factor; (4) inverse method (IM) for kinetic coefficient. Four SMA parameters could be one-by-one determined in sequence, reducing the number of unknown parameters per species from four to one, and predicting almost consistent retention. Numerical single-component experiments were investigated firstly, and the PbP method showed excellent agreement between experiments and simulations. The effects of loadings on the PbP and Yamamoto methods were compared. It was found that the PbP method had higher accuracy and robustness than the Yamamoto method. Moreover, a five-experiment strategy was suggested to implement the PbP method, which is straightforward to reduce the cost of calibration experiments. Finally, a real-world multi-component separation was challenged and further confirmed the feasibility of the PbP method. In general, the proposed method can not only reliably estimate the SMA parameters with comprehensive physical understanding but also accurately predict retention over a wide range of loading conditions.