Dual-Peptide Nanoplatform: Mesoporous Silica Nanoparticles Functionalized With a Cell-Penetrating Peptide and Loaded With Rationally Designed Antimicrobial Peptides for Tuberculosis Therapy

Abstract

Tuberculosis remains a major infectious disease worldwide and demands new therapeutic approaches capable of targeting intracellular Mycobacterium tuberculosis. Using a machine learning (ML)–guided design strategy based on the plectasin scaffold, a peptide library was generated and DC05 was identified as the most active analog. As a free peptide, DC05 inhibited the H37Rv strain with an MIC90 of 12.46 ± 0.06 µm, displayed low cytotoxicity in MRC-5 fibroblasts (IC50 >250 µm) and J774A.1 macrophages (115.18 ± 8.49 µm), and increased ethidium bromide accumulation to levels comparable to verapamil, with minimal hemolysis up to 256 µm. DC05 was then encapsulated into mesoporous silica nanoparticles (MSN), yielding (MSN@DC05) and further functionalized with tuftsin, a cell-penetrating peptide (CPP) (MSN-CPP@DC05). In simulated pulmonary fluid at 0.5 mg mL−1, all formulations remained colloidally stable, with hydrodynamic diameters between 122 and 164 nm and low polydispersity. Adsorption onto MSN followed a Langmuir-type profile (qmax 39.5 ± 4.9 mg g−1). Release was strongly dependent on the surrounding medium, showing rapid diffusion in simulated pulmonary fluid, slower release in simulated intestinal fluid, and almost no detectable peptide in simulated gastric fluid, where acid-driven degradation is likely. Nanoencapsulation enhanced antimycobacterial potency, reducing the MIC90 to 5.74 ± 0.15 µm for MSN@DC05 and 4.13 ± 0.64 µm for MSN-CPP@DC05, while maintaining high viability in mammalian cells up to 500 µg mL−1 in macrophages and 2000 µg mL−1 in fibroblasts. Ames testing showed mutagenicity indices below 2, and Galleria mellonella assays at 2000 mg kg−1 confirmed high survival. Scanning electron microscopy combined with molecular dynamics supported a membrane-disruptive mechanism for MSN-CPP@DC05. These results indicate that integrating ML-guided peptide design with a functionalized mesoporous carrier improves intracellular delivery and antimycobacterial activity while maintaining favorable safety profiles.