Corynebacterium accolens Releases Antipneumococcal Free Fatty Acids from Human Nostril and Skin Surface Triacylglycerols

Abstract

Bacterial interspecies interactions play clinically important roles in shaping microbial community composition. We observed that Corynebacterium spp. are overrepresented in children free of Streptococcus pneumoniae (pneumococcus), a common pediatric nasal colonizer and an important infectious agent. Corynebacterium accolens, a benign lipid-requiring species, inhibits pneumococcal growth during in vitro cocultivation on medium supplemented with human skin surface triacylglycerols (TAGs) that are likely present in the nostrils. This inhibition depends on LipS1, a TAG lipase necessary for C. accolens growth on TAGs such as triolein. We determined that C. accolens hydrolysis of triolein releases oleic acid, which inhibits pneumococcus, as do other free fatty acids (FFAs) that might be released by LipS1 from human skin surface TAGs. Our results support a model in which C. accolens hydrolyzes skin surface TAGS in vivo releasing antipneumococcal FFAs. These data indicate that C. accolens may play a beneficial role in sculpting the human microbiome.

Importance: Little is known about how harmless Corynebacterium species that colonize the human nose and skin might impact pathogen colonization and proliferation at these sites. We show that Corynebacterium accolens, a common benign nasal bacterium, modifies its local habitat in vitro as it inhibits growth of Streptococcus pneumoniae by releasing antibacterial free fatty acids from host skin surface triacylglycerols. We further identify the primary C. accolens lipase required for this activity. We postulate a model in which higher numbers of C. accolens cells deter/limit S. pneumoniae nostril colonization, which might partly explain why children without S. pneumoniae colonization have higher levels of nasal Corynebacterium. This work narrows the gap between descriptive studies and the needed in-depth understanding of the molecular mechanisms of microbe-microbe interactions that help shape the human microbiome. It also lays the foundation for future in vivo studies to determine whether habitat modification by C. accolens could be promoted to control pathogen colonization.

FIG 1

Nasopharyngeal Corynebacterium spp. are overrepresented in children without S. pneumoniae nasal colonization. Shown is a linear discriminant analysis (LDA) plot of the nasopharyngeal bacterial genera overrepresented when S. pneumoniae was absent (dark gray) or present (light gray) by cultivation. Abbreviations indicate the following genera: Str, Streptococcus; Dol, Dolosigranulum; Cor, Corynebacterium. Data were obtained from 27 nasopharyngeal swabs collected from healthy children aged 6 months to 7 years. Swabs used to inoculate medium for S. pneumoniae cultivation were frozen and later used to harvest nucleic acids, which were analyzed by 16S rRNA V1-3 tag pyrosequencing. Sequences were analyzed as described in Materials and Methods.

FIG 2

Model human skin surface TAGs triolein (triO) and trilinolein (triL) support the growth of the fatty acid-requiring species C. accolens. Black arrows indicate the oleic acid esters in (A) triolein and the linoleic esters in (B) trilinolein. (C) The mean turbidity (OD₆₀₀) after 48 h of shaken aerobic growth at 37°C demonstrated increased C. accolens growth in BHI broth supplemented with one of the TAGs compared to solvent-only controls (n = 9). Data were analyzed using an ANOVA and Tukey’s multiple-comparisons test. Error bars represent standard errors of the means. ****, P ≤ 0.0001.

FIG 3

C. accolens inhibits S. pneumoniae. C. accolens KPL1818 (Cac) inhibited S. pneumoniae 603 (Spn) growth on BHIC agar spread with 50 µl of a 100-mg/ml solution of (A) triolein solubilized in chloroform (triO + CHCl₃) or (C) trilinolein emulsified in ethanol (triL + EtOH), but not on medium spread with 50 µl of solvent alone: (B) chloroform (CHCl₃) or (D) EtOH. Representative images are shown (n = 3). To permit visualization of partial inhibition of the pneumococcal spot, S. pneumoniae was inoculated at various distances from C. accolens, as shown by the varied sizes of ZOIs, which are likely due to differences in diffusion of the antipneumococcal molecule(s).

FIG 4

Oleic acid is enriched in methanolic extracts of C. accolens CFCM. We used mass spectrometric analysis to identify compounds uniquely enriched in C. accolens CFCM. Methanolic extracts were analyzed by direct line infusion into a triple-quadrupole mass spectrometer operating in negative-ion scanning mode from m/z 50 to m/z 500. (A) To identify novel compounds present in the CFCM, the limited mass spectrum of the methanolic extract from unconditioned medium was subtracted from the spectrum for C. accolens CFCM. (B) Limited mass scanning of the novel m/z 281 peak present in CFCM, determined via high-resolution mass spectrometry, which gave a measured mass of m/z 281.2485, consistent with the calculated mass of oleic acid (m/z 281.2486). (C) An FFA standard mixture containing authentic oleic acid and heptadecanoic acid (HdA), as an internal standard, was analyzed by reversed-phase ultraperformance liquid chromatography coupled to tandem mass spectrometry, operating in multiple reaction mode monitoring transition of m/z 281 to m/z 281 at 10 eV for HdA and m/z 269 to m/z 269 at 10 eV for oleic acid, respectively. (D) CFCM with HdA added as an internal standard, analyzed under the same chromatographic conditions. (E) Unconditioned medium with HdA added as an internal standard, analyzed under the same chromatographic conditions. Representative spectra (A and B) or chromatographs (C, D, and E) are shown.

FIG 5

C. accolens LipS1 is necessary for growth on triolein (triO) as the sole source of fatty acids. (A) A C. accolens mutant strain (LipS1⁻) harboring an in-frame deletion in lipS1 did not grow in BHI supplemented with triolein (0.1 mg/ml), whereas the wild-type (WT) did. Growth was measured as the mean OD₆₀₀ following 48 h of aerobic growth with shaking at 37°C. Data were analyzed using a Welch two-sample t test. n = 3. Error bars represent standard deviations. *, P ≤ 0.05. (B) The lipase-deficient mutant carrying the empty vector (LipS1⁻ pCGL⁺) did not grow in BHI supplemented with triolein; however, the complemented mutant (LipS1⁻ cLipS1⁺) did. For comparison, the empty vector and complemented vector controls for the parental strain (WT pCGL⁺ and WT cLipS1⁺, respectively) are included. Data (n = 3) were analyzed using an ANOVA and Tukey’s multiple-comparisons test. Error bars represent standard deviations. **, P ≤ 0.01; ****, P ≤ 0.0001.

FIG 6

One model is that C. accolens modifies the nasal and/or skin habitat, making it inhospitable for S. pneumoniae. C. accolens produces a lipase, likely extracellular, that releases FFAs from host TAGs. Some of these FFAs have antibacterial activity and might impede the colonization and proliferation of S. pneumoniae and possibly other FFA-susceptible colonizers.