pH-Responsive Bioactive Polymer Plp-NDA: Computational Analysis on Membrane Interactions and Destabilization

Poly(l-lysine iso-phthalamide) grafted with decylamine (PLP-NDA) is a bioactive polymer with potential for pH-responsive intracellular drug delivery. When drugs are internalized by cells through endocytosis, they are compartmentalized inside endosomes, which eventually merge with lysosomes causing degradation of the drug. This can be prevented by releasing the drug to the cytosol through the pH-triggered lysing of late endosomes by PLP-NDA. In past works, we have demonstrated that PLP-NDA exhibits membrane-lysis under acidic endosomal pH, but not at physiological pH, allowing for the lysis of endosomes without the destruction of cell membranes, preserving the cells. Because the strength of lysis depends on the degree of grafting with NDA, the polymer’s activity can be fine-tuned for other biomedical applications.

Despite PLP-NDA’s demonstrated activity, the underlying mechanisms of its interactions with membranes have remained relatively unexplored, especially in relation to its chemical structure, rendering rational design challenging. Past experimental works, observing the formation of ghost cells, have hypothesised a pore formation mechanism facilitated by the anchoring of hydrophobic NDA chains to the membrane. However, it remains unclear how a large polymer could efficiently anchor onto a membrane and only exhibit significant disruption upon protonation.

This study seeks to understand the biomolecular interactions of PLP-NDA and membranes, identifying the mechanisms of its pH-responsiveness through molecular dynamics simulations. Because the procedure is generalisable to other polymers, it allows for the computer-aided design of novel drug-delivery polymers through screening prior to experimentation. In this study, we investigate polymer binding and its effects on membrane deformation under different degrees of protonation and NDA grafting at intermediate polymer lengths.

With complete NDA grafting (100%), PLP-NDA does not bind persistently to the membrane, remaining unbound for 20% of the microsecond simulation time. In contrast, PLP-NDA with partial grafting (18%, experimentally optimal for lysis) or no grafting (0%) binds persistently for more than 90% of the simulation time, forming hydrogen bonds with the membrane. This result suggests the importance of lysine’s hydrophilic carboxyl group in facilitating the initial binding and maintaining it. The bound PLP-NDA 18% and 0% are observed to cause time-averaged and localised membrane-thinning. PLP-NDA 100% did not cause membrane-thinning. Ionised PLP-NDA demonstrated weaker membrane-thinning, consistent with experimental findings of less membrane-lysis at higher pH. Although it is very challenging to observe membrane disruption in time-scales of atomistic simulations, these results may explain the early initiation of pores and demonstrates a quantitative criterion to benchmark novel polymers.