Predicting zeta potential of liposomes from their structure: A nano-QSPR model for DOPE, DC-Chol, DOTAP, and EPC formulations

Kamila Jarzynska; Agnieszka Gajewicz-Skretna; Krzesimir Ciura; Tomasz Puzyn

doi:10.1016/j.csbj.2024.01.012

Predicting zeta potential of liposomes from their structure: A nano-QSPR model for DOPE, DC-Chol, DOTAP, and EPC formulations

Comput Struct Biotechnol J. 2024 Jan 24:25:3-8. doi: 10.1016/j.csbj.2024.01.012. eCollection 2024 Dec.

Authors

Kamila Jarzynska¹, Agnieszka Gajewicz-Skretna¹, Krzesimir Ciura^{1

2}, Tomasz Puzyn¹

Affiliations

¹ Laboratory of Environmental Chemoinformatics, Faculty of Chemistry, University of Gdansk, Wita Stwosza 63, 80-308 Gdansk, Poland.
² Department of Physical Chemistry, Faculty of Pharmacy, Medical University of Gdansk, Gen. Hallera 107, 80-416 Gdańsk, Poland.

Abstract

Liposomes, nanoscale spherical structures composed of amphiphilic lipids, hold great promise for various pharmaceutical applications, especially as nanocarriers in targeted drug delivery, due to their biocompatibility, biodegradability, and low immunogenicity. Understanding the factors influencing their physicochemical properties is crucial for designing and optimizing liposomes. In this study, we have presented the kernel-weighted local polynomial regression (KwLPR) nano-quantitative structure-property relationships (nano-QSPR) model to predict the zeta potential (ZP) based on the structure of 12 liposome formulations, including 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE), 3ß-[N-(N',N'-dimethylaminoethane)-carbamoyl]cholesterol (DC-Chol), 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP), and L-α-phosphatidylcholine (EPC). The developed model is well-fitted ( $R^{2}$ = 0.96, ${RMSE}_{C}$ = 5.76), flexible ( $Q_{CVloo}^{2}$ = 0.83, ${RMSE}_{CVloo}$ = 10.77), and reliable ( $Q_{Ext}^{2}$ = 0.89 ${RMSE}_{Ext}$ = 5.17). Furthermore, we have established the formula for computing molecular nanodescriptors for liposomes, based on constituent lipids' molar fractions. Through the correlation matrix and principal component analysis (PCA), we have identified two key structural features affecting liposomes' zeta potential: hydrophilic-lipophilic balance (HLB) and enthalpy of formation. Lower HLB values, indicating a more lipophilic nature, are associated with a higher zeta potential, and thus stability. Higher enthalpy of formation reflects reduced zeta potential and decreased stability of liposomes. We have demonstrated that the nano-QSPR approach allows for a better understanding of how the composition and molecular structure of liposomes affect their zeta potential, filling a gap in ZP nano-QSPR modeling methodologies for nanomaterials (NMs). The proposed proof-of-concept study is the first step in developing a comprehensive and computationally based system for predicting the physicochemical properties of liposomes as one of the most important drug nano-vehicles.

Keywords: KwLPR; Liposomes; Nano-QSPR; Nanocarriers; Nanodescriptors; PCA biplot; Zeta potential.