Mixed-Charge, pH-Responsive Nanoparticles for Selective Interactions with Cells, Organelles, and Bacteria

CONSPECTUS: Charged interfaces are ubiquitous in biology, from proteins, to membranes enclosing cells and organelles, to bacterial cell walls. For nanomaterials to interact with these biological entities effectively and selectively, their electrostatic properties must be fine-tuned. To date, charged nanomaterials have had their surfaces decorated with a single type of charged ligands (positive, negative, or zwitterionic), sometimes "diluted" with additional uncharged/hydrophobic ligands. Although such materials have shown some promise, they can be nonselectively cytotoxic or membrane-permeable, and can nonspecifically adsorb serum proteins or display limited cellular uptake. In this Account, we review a decade of our work on pH-responsive nanoparticles (NPs) covered with mixtures of oppositely charged ligands. These so-called mixed-charge nanoparticles (MCNPs) comprise d similar to 4-12 nm Au cores protected with mixed monolayers comprising positively charged N,N,N-trimethyl (11-mercaptoundecyl) ammonium chloride (TMA, [+]) and negatively charged mercaptoundecanoic acid (MUA, [-]) ligands at varying and precisely adjustable proportions. The MCNPs are pH-responsive because MUA ligands can be protonated at a lower pH and, instead of interacting electrostatically, can engage in hydrogen bonding. Importantly, by adjusting the proportion of TMA and MUA functional groups on the particles' surfaces, (chi(TMA):chi(MUA))(NP), and/or particle sizes, the relative balance between electrostatic, hydrogen bonding, and van der Waals forces can be tuned, translating into pH-dependent stability versus aggregation of the MCNPs. In this way, MCNPs resist aggregation at the pH of healthy tissues and blood, but they precipitate or even crystallize at a narrow range of acidic pHs, characteristic of extracellular space surrounding tumors as well as intracellular endolysosomal compartments. We harnessed this propensity of MCNPs to aggregate at the desired pH to target low-pH lysosomes in cancer cells. As we showed, MCNPs with compositions around (chi(TMA):chi(MUA))(NP) = 80:20 are noncytotoxic to normal cells and do not present adverse effects in mice but kill multiple different cancer cell types through a pH-dependent assembly of micron-sized crystals inside cancer lysosomes. Due to impaired transport and clearance mechanisms, this crystallization leads to an increase in osmotic pressure and lysosome swelling, which permeabilizes lysosomal membranes triggering lysosomal cell death only in cancer cells. The same NPs display only limited aggregation in higher-pH autolysosomes of noncancerous cells. MCNPs can also selectively interact with and disrupt differentially charged bacterial cell walls, leading to Gram-selective antibiotic activity. The most effective ligand composition against Gram-negative bacteria is (chi(TMA):chi(MUA))(NP) = 80:20, while for Gram-positive bacteria it is 48:52. Mechanistically, cationic ligands help to attach the particles to Gram-negative cell walls, while partly protonated MUA ligands compete for hydrogen bonds with cell surface lipopolysaccharides and peptidoglycans. As for Gram-positive bacteria, in addition to hydrogen bonding, deprotonated MUAs may contribute to the disruption of the cell wall by coordinating metal cations. Although MCNPs can be nowadays prepared in various and precisely controlled compositions, their properties and in vivo fates merit further study. In this spirit, we conclude this Account by outlining several unresolved problems important for the application of mixed-charge nanoconstructs in vivo and in the clinic.