Growth-Environment Dependent Modulation of Staphylococcus aureus Branched-Chain to Straight-Chain Fatty Acid Ratio and Incorporation of Unsaturated Fatty Acids

Abstract

The fatty acid composition of membrane glycerolipids is a major determinant of Staphylococcus aureus membrane biophysical properties that impacts key factors in cell physiology including susceptibility to membrane active antimicrobials, pathogenesis, and response to environmental stress. The fatty acids of S. aureus are considered to be a mixture of branched-chain fatty acids (BCFAs), which increase membrane fluidity, and straight-chain fatty acids (SCFAs) that decrease it. The balance of BCFAs and SCFAs in strains USA300 and SH1000 was affected considerably by differences in the conventional laboratory medium in which the strains were grown with media such as Mueller-Hinton broth and Luria broth resulting in high BCFAs and low SCFAs, whereas growth in Tryptic Soy Broth and Brain-Heart Infusion broth led to reduction in BCFAs and an increase in SCFAs. Straight-chain unsaturated fatty acids (SCUFAs) were not detected. However, when the organism was grown ex vivo in serum, the fatty acid composition was radically different with SCUFAs, which increase membrane fluidity, making up a substantial proportion of the total (37%) and BCFAs (>36%) making up the rest. Staphyloxanthin, an additional major membrane lipid component unique to S. aureus, tended to be greater in content in cells with high BCFAs or SCUFAs. Cells with high staphyloxanthin content had a lower membrane fluidity that was attributed to increased production of staphyloxanthin. S. aureus saves energy and carbon by utilizing host fatty acids for part of its total fatty acids when growing in serum. The fatty acid composition of in vitro grown S. aureus is likely to be a poor reflection of the fatty acid composition and biophysical properties of the membrane when the organism is growing in an infection in view of the role of SCUFAs in staphylococcal membrane composition and virulence.\n\nFundingThis work was funded in part by grant 1R15AI099977 to Brian Wilkinson and Craig Gatto and grant 1R15GM61583 to Craig Gatto from the National Institutes of Health