Streptococcus pneumoniae rapidly kills Staphylococcus aureus by producing membrane-permeable hydrogen peroxide (H2O2). The mechanism by which S. pneumoniae-produced H2O2 mediates S. aureus killing was investigated. An in vitro model that mimicked S. pneumoniae-S. aureus contact during colonization of the nasopharynx demonstrated that S. aureus killing required outcompeting densities of S. pneumoniae. Compared to the wild-type strain, isogenic S. pneumoniae ΔlctO and S. pneumoniae ΔspxB, both deficient in production of H2O2, required increased density to kill S. aureus. While residual H2O2 activity produced by single mutants was sufficient to eradicate S. aureus, an S. pneumoniae ΔspxB ΔlctO double mutant was unable to kill S. aureus. A collection of 20 diverse methicillin-resistant S. aureus (MRSA) and methicillin-susceptible S. aureus (MSSA) strains showed linear sensitivity (R2 = 0.95) for S. pneumoniae killing, but the same strains had different susceptibilities when challenged with pure H2O2 (5 mM). There was no association between the S. aureus clonal complex and sensitivity to either S. pneumoniae or H2O2. To kill S. aureus, S. pneumoniae produced ∼180 μM H2O2 within 4 h of incubation, while the killing-defective S. pneumoniae ΔspxB and S. pneumoniae ΔspxB ΔlctO mutants produced undetectable levels. Remarkably, a sublethal dose (1 mM) of pure H2O2 incubated with S. pneumoniae ΔspxB eradicated diverse S. aureus strains, suggesting that S. pneumoniae bacteria may facilitate conversion of H2O2 to a hydroxyl radical (·OH). Accordingly, S. aureus killing was completely blocked by incubation with scavengers of ·OH radicals, dimethyl sulfoxide (Me2SO), thiourea, or sodium salicylate. The ·OH was detected in S. pneumoniae cells by spin trapping and electron paramagnetic resonance. Therefore, S. pneumoniae produces H2O2, which is rapidly converted to a more potent oxidant, hydroxyl radicals, to rapidly intoxicate S. aureus strains. IMPORTANCE Streptococcus pneumoniae strains produce hydrogen peroxide (H2O2) to kill bacteria in the upper airways, including pathogenic Staphylococcus aureus strains. The targets of S. pneumoniae-produced H2O2 have not been discovered, in part because of a lack of knowledge about the underlying molecular mechanism. We demonstrated that an increased density of S. pneumoniae kills S. aureus by means of H2O2 produced by two enzymes, SpxB and LctO. We discovered that SpxB/LctO-produced H2O2 is converted into a hydroxyl radical (·OH) that rapidly intoxicates and kills S. aureus. We successfully inhibited the toxicity of ·OH with three different scavengers and detected ·OH in the supernatant. The target(s) of the hydroxyl radicals represents a new alternative for the development of antimicrobials against S. aureus infections.

Streptococcus pneumoniae colonizes the upper airways of children and the elderly. Colonization progresses to persistent carriage when S. pneumoniae forms biofilms, a feature required for the development of pneumococcal disease. Nasopharyngeal biofilms are structured with a matrix that includes extracellular DNA (eDNA), which is sourced from the same pneumococci and other bacteria. This eDNA also allows pneumococci to acquire new traits, including antibiotic resistance genes. In this study, we investigated the efficacy of lactoferrin (LF), at physiological concentrations found in secretions with bactericidal activity [i.e., colostrum (100 μM), tears (25 μM)], in eradicating pneumococcal biofilms from human respiratory cells. The efficacy of synthetic LF-derived peptides was also assessed. We first demonstrated that LF inhibited colonization of S. pneumoniae on human respiratory cells without affecting the viability of planktonic bacteria. LF-derived peptides were, however, bactericidal for planktonic pneumococci but they did not affect viability of pre-formed biofilms. In contrast, LF (40 and 80 μM) eradicated pneumococcal biofilms that had been pre-formed on abiotic surfaces (i.e., polystyrene) and on human pharyngeal cells, as investigated by viable counts and confocal microscopy. LF also eradicated biofilms formed by S. pneumoniae strains with resistance to multiple antibiotics. We investigated whether treatment with LF would affect the biofilm structure by analyzing eDNA. Surprisingly, in pneumococcal biofilms treated with LF, the eDNA was absent in comparison to the untreated control (∼10 μg/ml) or those treated with LF-derived peptides. EMSA assays showed that LF binds S. pneumoniae DNA and a time-course study of DNA decay demonstrated that the DNA is degraded when bound by LF. This LF-associated DNase activity inhibited acquisition of antibiotic resistance genes in both in vitro transformation assays and in a life-like bioreactor system. In conclusion, we demonstrated that LF eradicates pneumococcal-colonizing biofilms at a concentration safe for humans and identified a LF-associated DNAse activity that inhibited the acquisition of resistance.

The World Health Organization has designated Streptococcus pneumoniae as a priority pathogen for research and development of new drug treatments due to extensive multidrug resistance. Multiple strains of S. pneumoniae colonize and form mixed biofilms in the human nasopharynx, which could enable exchange of antibiotic resistance determinants.