Effects of growth phase on the membrane lipid composition of the thaumarchaeon Nitrosopumilus maritimus and their implications for archaeal lipid distributions in the marine environment

The characteristic glycerol dibiphytanyl glycerol tetraether membrane lipids (GDGTs) of marine ammonia-oxidizing archaea (AOA) are widely used as biomarkers for studying their occurrence and distribution in marine environments and for reconstructing past sea surface temperatures using the TEX86 index. Despite an increasing use of GDGT biomarkers in microbial ecology and paleoceanography, the physiological and environmental factors influencing lipid composition in AOA, in particular the cyclization of GDGTs, remain unconstrained. We investigated the effect of metabolic state on the composition of intact polar and core lipids and the resulting TEX86 paleothermometer in pure cultures of the marine AOA Nitrosopumilus maritimus as a function of growth phase. The cellular lipid content ranged from 0.9 to 1.9 fg cell(-1) and increased during growth but was lower in the stationary phases, indicating changes in average cell size in response to metabolic status. The relative abundances of monoglycosidic GDGTs increased from 27% in early growth phase to 60% in late stationary phase, while monohydroxylated GDGTs increased only slightly. The proportions of characteristic hexose-phosphohexose GDGTs were up to 7-fold higher during growth than in stationary phase, suggesting that they are valuable biomarkers for the metabolically active fraction of AOA assemblages in the environment. Methoxy archaeol was identified as novel, genuine archaeal lipid of yet unknown function; it is one of the most abundant single compounds in the lipidome of N. maritimus. TEX86 values of individual intact GDGTs and total GDGTs differed substantially, were generally lower during early and late growth phases than in stationary phase, and did not reflect growth temperature. Consequently, our results strongly suggest that biosynthesis is at least partially responsible for the systematic offsets in TEX86 values between different intact polar GDGT classes observed previously in environmental samples. Nevertheless, differences in degradation rates of intact polar GDGTs may influence the TEX86 index because the intact polar lipid precursors differ for individual core GDGTs and moreover their relative abundances change with growth stage, which may result in distinct release rates of core GDGTs from their polar precursors. Overall, our findings stress the need to accurately describe the factors influencing GDGT cyclization in thaumarchaea and thus paleotemperature reconstructions. (C) 2014 Elsevier Ltd. All rights reserved.