Gas chromatographic metabolic profiling: A sensitive tool for functional microbial ecology Elsa Coucheney a, ⁎, Tim J. Daniell b , Claire Chenu a , Naoise Nunan a a Biogéochimie et Ecologie des Milieux Continentaux, UMR 7618 (UPMC, CNRS, AgroParisTech), Bât. EGER, 78850 Thiverval Grignon, France b Environment Plant Interactions, Scottish Crop Research Institute, Invergowrie, Dundee DD2 5DA, UK abstract article info Article history: Received 13 June 2008 Received in revised form 30 July 2008 Accepted 30 July 2008 Available online 3 August 2008 Keywords: Bacteria Functional ecology Gas chromatography Metabolome Metabolic profiling Temperature Microbial metabolomics, which consists of a non-targeted analysis of the metabolites released from (‘exometabolome’) or existing in (‘endometabolome’) a cell has mostly been used to study the metabolism of particular microbes. Metabolomes also represent a picture of microbial activity and we suggest that the exometabolome may also contain pertinent information for studying microbial interaction networks. Gas chromatography coupled to mass spectrometry is the most commonly used technique in metabolomics studies. It allows a wide range of metabolites to be detected but requires the derivatisation of compounds prior to detection. This type of non-targeted analysis can introduce biases to the detection and quantification of the different metabolites, particularly at the extraction and derivatisation steps. The aims of this study, therefore, were to quantify the sources of variability and to test the sensitivity of the GC metabolic profiling approach to small environmental changes such as shifts in temperature. The temperature sensitivity of metabolic profiles was compared with that of catabolic profiles obtained using Biolog® microplates. Analytical variability was compared with biological variability by incubating bacterial strains isolated from soil with fructose at 20 °C and by replicating each step of the protocol (incubation, extraction and deri- vatisation). For both the endo- and the exometabolome, more than 70% of the total variability was of biological origin and principal components analysis clearly separated the strains along the first ordination axis. The endometabolome distinguished bacterial strains at the species level only, whereas separation was evident at the species and group level with the exometabolome. Temperature had a significant but differ- ential effect on the metabolite production of the bacterial strains whilst their catabolic profiles remained relatively unaffected. The exometabolome was more sensitive to temperature shifts than the endometabo- lome, suggesting that this pool may be of interest for studies in environmental functional ecology. © 2008 Elsevier B.V. All rights reserved. 1. Introduction The aim of functional ecology is to study the role and functions of a particular species, group or whole community in a given ecosystem. This is a complex exercise as interactions and effects can occur at many levels. If abiotic factors like climate or physico-chemical parameters are major determinants of ecosystem function, biological feedbacks to the abiotic world are important in the response to environmental variations (Andren et al., 1999). For example, changes in temperature can affect the functioning of individual microorganisms within a community or indeed the structure of the microbial community and the resultant physiological changes can modify biological reactions that drive biogeochemical processes (Schimel and Gulledge, 1998). Understanding the biogeochemistry of the carbon (C) cycle and how it is affected by environmental variations is of particular interest due to its connection with global change. Microbial respiration is the result of many distinct metabolic processes each of which may have a specific response to changes in environment (Schimel and Gulledge, 1998). Thus, to understand the role of microbial communities in the environment, it is essential to take into account functional diversity (Calbrix et al., 2005). Recently, there has been an explosion of “omics” techniques that have been applied to measure various types of fingerprints (chemical or genetic for example) of microbes or microbial communities. The different techniques have been classified into ‘microbial potentiality’ (genomics, transcriptomics and proteomics), ‘microbial functionality’ (transcriptomics and proteomics) and ‘microbial activity’ (metabolo- mics); and the limitation of “nucleic based methods” for extracting information on microbial function has been highlighted (Maron et al., 2007). Metabolomics is an approach that can describe an organism's phenotype, at cellular, tissue or whole organism level (Lin et al., 2006). It consists of a non-targeted analysis of the low molecular weight (LMW) metabolites produced by or existing in a cell. Metabolites excreted or released outside the cell (the exometabolome) and the metabolites Journal of Microbiological Methods 75 (2008) 491–500 ⁎ Corresponding author. E-mail address: ecoucheney@grignon.inra.fr (E. Coucheney). 0167-7012/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.mimet.2008.07.029 Contents lists available at ScienceDirect Journal of Microbiological Methods journal homepage: www.elsevier.com/locate/jmicmeth