Ester-linked polar lipid fatty acid profiles of soil microbial communities: a comparison of extraction methods and evaluation of interference from humic acids

Analyses of polar lipid fatty acids isolated from soil are frequently used for characterization of microbial communities, and any interference from fatty acids derived from dead organic material is assumed to be negligible. We studied the initial extraction of lipid material from eight different soils and from purified humic acids using four different combinations of solvent (chloroform or dichloromethane), methanol and buffer (potassium phosphate, pH 7.4 or sodium citrate, pH 4). The quantitative yields of polar lipid fatty acids (PLFA) and PLFA composition of soils and humic acids were compared with absorbance spectra (200–850 nm) of lipid extracts for evaluation of extraction efficiency and potential interference. Chloroform + citrate buffer generally gave the highest, and dichloromethane + phosphate buffer the lowest PLFA yields, and it was estimated that <20% of the yield difference between extraction methods could be explained by interference from humic acids. Principal component analyses of PLFA composition suggested an effect of extraction method for several soils, but when all soils were analyzed together the differences between soils were much more important than the choice of extraction method. Co-extraction of lipids from living cells during preparation of humic acids was quantified and, correcting for this, it was estimated that the interference from non-microbial sources in PLFA analyses was probably not more than 5–10% with the extraction methods employed.