Modeling across scales of soft materials for nucleic acids delivery

Nucleic acid (NA) therapies represent a transformative frontier in modern medicine, yet their clinical success depends heavily on the development of efficient delivery carriers to overcome biological barriers. This Thesis focuses on two distinct classes of materials: nanoparticles formed by the complexation of NAs with block copolymers (polyplexes) and peptide/NA coacervates. While the former are well-established delivery systems, the latter have been proposed only recently in this field. We center our study on representative systems and, by combining Molecular Dynamics (MD) simulations at different levels of resolution with experimental investigations carried out by collaborating groups, we shed light on the molecular mechanisms that control the stability and the structural organization of the systems. A first objective is to disentangle the individual effects of NA physical properties, such as length, charge density, and flexibility, which strongly depend on the hybridization state, on the stability and internal structure of the NA/peptide coacervates. To this end, we focus on well-defined model systems under controlled conditions: specifically, electroneutral mixtures comprising a moderately charged, disordered peptide and a series of systematically varied oligonucleotides (ONTs), including single-stranded (ss), double-stranded (ds), and partially hybridized structures. We combine experimental characterization with simulations based on a chemically specific coarse-grained (CG) model, wherein each interaction site corresponds to a nucleotide or an amino acid. We find that coacervate stability scales positively with the degree of base pairing, and we quantitatively assess the dominant role of charge distribution stemming from the phosphate backbone. Bending stiffness, which differs markedly between ss- and dsONTs, is key to the emergence of liquid crystal phases within coacervates formed by duplexes. We provide a detailed account of how the organization and mobility of peptides and ONTs vary across the different phases. Extending the CG model to incorporate end-to-end stacking interactions between dsONTs, we explore the mechanisms behind the complex phase behavior reported recently for short DNA duplex/peptide coacervates under the control of salt concentration. We find that this mechanism is determined by a synergy of hierarchical assembly processes: stacking interactions drive the formation of long ONT columns, thereby enhancing LLPS and stabilizing liquid crystal order in dense coacervate environments; these, in turn, further promote stacking and the growth of ordered columns. In this mechanism, salt, by modulating electrostatic peptide/ONT interactions, controls the proximity of ONTs that facilitate stacking. Finally, we focus on the polyplexes formed by a series of block copolymers containing mannosyl, agmatinyl (diblock), and additional butyl (triblock) units, which were recently proposed as vectors for ssDNA oligomer delivery. We combine all-atom (AA) and CG simulations to examine how polymer architecture, topology, and length impact polyplex properties. AA simulations of samples comprising a few ONTs and varying copolymer ratios are crucial for demonstrating the role of specific interactions between the different chemical moieties. In particular, AA simulations provide a molecular rationale for the stabilizing effect of the butyl blocks and the unexpectedly low targeting efficacy of mannosyl units. CG modeling enables the simulation of systems at scales comparable to experiments, making it possible to directly relate copolymer structure to the size, morphology, and, ultimately, the delivery efficacy of polyplexes. Overall, this Thesis demonstrates that computational methods, by providing quantitative insights into how molecular and supramolecular interactions can be fine-tuned, serve as valuable tools for the development of NA delivery materials.