Developing novel descriptors for chromatographic protein separation

Background.

Chromatography is a key technology for the purification of pharmaceutical proteins (Tripathi and Shrivastava 2019) but is currently expensive and laborious to develop (Stamatis et al. 2019). Modeling the separation of proteins a priori has therefore gained substantial interest in academia and industry in the last years (Leweke and Lieres 2018; Briskot et al. 2021; Bernau and Buyel 2023). However, such approaches require a solid understanding of protein properties, that is, these proteins need to be characterized in detail, which requires large quantities of pure protein that are typically not available especially during early process development (Bernau et al. 2021). One potential way around this limitation is the use of hybrid models that derive such protein properties from a limited set of proteins contained in a training data set, similar to a typical quantitative structure activity relationship model approach (Buyel et al. 2013). In this project we will design powerful descriptors that can capture relevant protein properties, such as surface charge distribution, to facilitate the a priori prediction of protein separation, which will significantly reduce modeling efforts and ultimately accelerate downstream process development. To do so, we will use different shape analyses (e.g. spectral shape analysis) to ideate innovative protein descriptors. These descriptors will be applied to both unmodified and glycosylated proteins for which we will collect experimental data for model calibration.

Aims and methods.

In this PhD project, the candidate will first develop new protein descriptors that capture size-related and surface properties, such as charge, in dependence of protein modifications. Specifically, the descriptors will integrate shape and surface properties and thus capture protein features in a holistic manner. For example, the Laplace-Beltrami operator is an interesting starting point in this respect (Reuter et al. 2009; Emonts and Buyel 2023). The student will also collect experimental chromatography data for isotherm and mass transport data determination as well as model calibration. To do so, (s)he will define a representative set of proteins covering diverse features, purify the respective molecules and subject them to a detailed characterization. The student will build and curate a database from the collected experimental data and then establish predictive models linking experimental data and descriptor values. The quality of resulting correlations will be assessed and suitable correlations will be used to predict isotherm parameters for proteins not included in the training data set.

Collaborations.

Developing the descriptors discussed above has substantial collaboration potentials that we plan to exploit. On one hand, an intensive exchange with Chris Oostenbrink and coworkers on the fundamentals of protein structure features will help to assess the plausibility and usefulness of novel parameters. On the other hand, the results of the thesis can be immediately applied. For example, the novel descriptors can help to establish purification processes for protein bodies, the topic of another PhD project within BioTop. Furthermore, there is a broad spectrum of additional collaboration potentials outside of BioToP, e.g. in the context of ongoing FFG-funded projects focusing on downstream process development.

Bernau, C. R.; Buyel, J. F. (2023): The use of antifoam agents to eliminate bubbles during biotechnological sample analysis. In Discov Chem Eng 3 (1). DOI: 10.1007/s43938-023-00021-w.
Bernau, C. R.; Jäpel, R. C.; Hübbers, J. W.; Nölting, S.; Opdensteinen, P.; Buyel, J. F. (2021): Precision analysis for the determination of steric mass action parameters using eight tobacco host cell proteins. In J Chromatogr A 1652, p. 462379. DOI: 10.1016/j.chroma.2021.462379.
Briskot, Till; Hahn, Tobias; Huuk, Thiemo; Hubbuch, Jürgen (2021): Protein adsorption on ion exchange adsorbers: A comparison of a stoichiometric and non-stoichiometric modeling approach. In Journal of chromatography. A 1653, p. 462397. DOI: 10.1016/j.chroma.2021.462397.
Buyel, J. F.; Woo, J. A.; Cramer, S. M.; Fischer, R. (2013): The use of quantitative structure-activity relationship models to develop optimized processes for the removal of tobacco host cell proteins during biopharmaceutical production. In J Chromatogr A 1322, pp. 18–28. DOI: 10.1016/j.chroma.2013.10.076.
Emonts, J.; Buyel, J. F. (2023): An Overview of Descriptors to Capture Protein Properties – Tools and Perspectives in the Context of QSAR Modeling. In Computational and Structural Biotechnology Journal. DOI: 10.1016/j.csbj.2023.05.022.
Leweke, S.; Lieres, E. von (2018): Chromatography Analysis and Design Toolkit (CADET). In Comput Chem Eng 113, pp. 274–294. DOI: 10.1016/j.compchemeng.2018.02.025.
Reuter, Martin; Biasotti, Silvia; Giorgi, Daniela; Patanè, Giuseppe; Spagnuolo, Michela (2009): Discrete Laplace–Beltrami operators for shape analysis and segmentation. In Computers & Graphics 33 (3), pp. 381–390. DOI: 10.1016/j.cag.2009.03.005.
Stamatis, Christos; Goldrick, Stephen; Gruber, David; Turner, Richard; Titchener-Hooker, Nigel J.; Farid, Suzanne S. (2019): High throughput process development workflow with advanced decision-support for antibody purification. In J Chromatogr A 1596, pp. 104–116. DOI: 10.1016/j.chroma.2019.03.005.
Tripathi, N. K.; Shrivastava, A. (2019): Recent Developments in Bioprocessing of Recombinant Proteins. Expression Hosts and Process Development. In Frontiers in Bioengineering and Biotechnology 7, p. 420. DOI: 10.3389/fbioe.2019.00420.