Institute of Biological Sciences, The University of Wales, Aberystwyth, Ceredigion, SY23 3DD, UK
This paper highlights the potential of metabolomics to advance allelopathic research. Metabolomics allows simultaneous analysis of the total biochemical complement of any given sample. This is predominantly achieved from technologies such as nuclear magnetic resonance (NMR), mass spectrometry and infra-red spectroscopy. Metabolomic methods are discussed with reference to variation, time and output. An example spectrum from Fourier Transform Infrared Spectroscopy (FT-IR) of a typical soil leachate is provided along with details of how we can start to unravel the multitude of information and communication belowground to further our understanding of allelopathy.
Modern technologies such as mass spectrometry and infrared spectroscopy will allow the total chemical composition of soil bio-chemicals to be analysed for the first time.
Metabolomics, allelopathy, FTIR, data mining
Soil biodiversity and its related biochemistry represents one of the least known 'black boxes' in terrestrial ecology. This organic substrate is regularly enriched via plant organic deposits from leaf litter decomposition and root exudation. Many of the organic carbon compounds released include phenolic based compounds that can have positive benefits to the plant by attracting nitrogen fixing bacteria or mycorrhizas (Schultze and Kondorosi 1998; Siqueira et al. 1991). Alternatively, these compounds can have negative effects on competing species via allelopathic inhibition of their growth (Sanchez-Moreiras et al., 2003; Singh et al., 2003). This multiplicity of below-ground chemical information has been referred to as the “underground information superhighway” (Bais et al., 2004). However, mining and understanding this information superhighway has always been fraught with difficulties.
Methodological difficulties and debate into ecological relevance of laboratory experiments already hamper allelopathic research (Inderjit and Weston 2000). Furthermore, up until now it has not been possible to holistically analyse the complete network of soil borne biochemical interactions. We advocate metabolomics as offering new possibilities by allowing the analysis of the total chemical complement of a given sample. The metabolome (of which metabolomics is concerned) can be defined as the total biochemical complement of a cell or particular organ (Fiehn et al., 2000a). Advanced multivariate statistical analyses and data mining of the metabolome can provide new potential and opportunity for the understanding of sample biochemistry, offering new insight into the sub-terranean biology of allelopathy.
Materials and Method
Aliquots of 5μl of solution are applied onto 400 wells drilled onto an aluminium plate (10cm by 10cm) which oven-dried at 50║C for one hour prior to analysis. The plate is loaded into an IFS28 FT-IR Spectrometer (Bruker Spectrospin Ltd., Banner Lane, Coventry, UK) supplied with a liquid nitrogen-cooled MCT (mercury-cadmium-telluride) detector. The IFS28 is controlled by an IBM-compatible personal computer using OPUS 2.1 software running under the IBM OS/2 Warp operating system (provided by manufacturers). Spectra are collected over the wavenumber range 4000 cm-1 to 600 cm-1 at a resolution of approximately 3.85 cm-1 and acquired at a rate of 20 s-1. 256 spectra were co-added and averaged. Spectra are displayed in terms of absorbance as calculated from the reflectance-absorbance spectra.
ASCII data are imported from the Opus software used to control the IFS28 and imported into Matlab version 6.0 (The MathWorks, Inc., Natick, MA, USA), which runs under Microsoft Windows 98 on an IBM-compatible personal computer. PC-DFA analysis can then be carried out on the data matrix (Ellis et al., 2002; Gidman et al., 2003; Johnson et al., 2003, 2004).
Results and Discussion
The field of metabolomics has already begun to revolutionise many aspects of biology although its potential for allelopathy research has yet to be realised. Technological advances in equipment such as mass spectrometers and infrared spectroscopes have allowed the metabolome to be analysed for the first time. Metabolomic methods can now provide the rapid and reproducible classification of samples according to their biochemistry by producing a metabolic fingerprint of data (Sumner et al., 2003). Such fingerprints are gained from obtaining spectra from the metabolome and consequently comparing it to others by multivariate statistics (Fiehn 2001). By employing metabolomic approaches to interrogate chemicals from root exudates / soil leachates and leaf litter, it will be in the future be possible to further understanding of below ground allelopathic interactions.
There are numerous technologies currently used in metabolomics and the choice of instrument and method are largely dependent on certain trade-offs. These include the accuracy of the instrument, the range of chemicals that can be identified /analysed and the degree of throughput (Figure 1). Sensitivity is also a major factor with degree of sensitivity increasing from the top of the triangle downwards.
Figure 1. Simplified diagram showing three trade-offs (bold type) when choosing a metabolomic technology (caps). Italics show best use of instrument. Acronyms denote: ESI-MS (Electrospray Mass Spectrometry), NMR (Nuclear Magnetic Resonance), FT-IR (Fourier Transform Infrared Spectroscopy), GC-MS and LC-MS (Gas and Liquid Chromatography coupled with Mass Spectrometry).
Fourier Transform Infrared Spectroscopy (FT-IR) is particularly suitable for metabolic fingerprinting, low in cost and rapid as a screening tool (Goodacre et al., 1996). The premise of FT-IR is that chemical bonds will absorb particular wavenumbers and consequently vibrate differently when interrogated with light (Bauer & Richter, 1996). Thus the absorbances can be correlated to functional groups on molecules. The mid-infrared region (4000-500 cm-1) is the most informative biologically and consequently used in metabolic fingerprinting.
Figure 2 illustrates a typical FT-IR spectrum from a soil leachate. A typical spectrum from FTIR consists of 882 variables, obtained by measuring the absorbance of the sample at 882 points over the mid infrared region. Given that 400 samples can be analysed at one time, the resultant data is effectively a matrix of 400 cases by 882 variables. This matrix can thus be analysed by multivariate statistics such as Principal Component Analysis (PCA) and Discriminant Function Analysis (DFA) to assess comparative similarities between samples thus forming the basis of a metabolic fingerprint. PCA works by reducing multivariate data into a reduced number of principal components which retain the majority of the variance. The first two principal components are typically plotted together to see if any clustering due to similar metabolic profiles occurs. DFA uses data from PCA to remove any within-group similarities and maximises between group differences. Such DFA plots are usually cross-validated with a test set from PCA to ensure lack of bias.
Figure 2. Example spectrum from a soil leachate of a 12 plant Lolium perenne monoculture.
FT-IR has been successfully used to classify bacterial samples (Winson et al., 1997) and is increasingly being applied to plant biology (Johnson et al., 2003, 2004) and to address ecological questions (Scullion et al., 2003; Gwynn-Jones et al., 2004) including competition (Gidman et al., 2003).
Following FT-IR screening of leachates, exudates and / or litter it is possible to do more targeted analysis of a certain group of compounds (e.g. phenolics) Electrospray mass spectroscopy (ESI-MS) has been used to accurately identify chemical compounds, including phenolics, in plant samples (Goodacre et al., 2002) and root exudates (Narasimhan et al., 2003). Phenolics have often been implemented in allelopathy (Blum et al., 1999; Inderjit, 1996; Inderjit et al., 2002). Alternatively mass spectrometry coupled to liquid or gas chromatography (LC-MS and GC-MS respectively) can be used to comprehensively profile the metabolome (Fiehn et al., 2000a, 2000b) although it is relatively expensive and the degree of throughput is compromised.
Although metabolic fingerprinting remains predominantly used as a classification and screening tool, there have been advances to more effectively mine data via pattern recognition analysis. The reoccurrence of certain peaks has led to the discovery of biomarkers; especially when data are analysed with appropriate chemometric techniques such as machine learning and genetic algorithms (Johnson et al., 2003; Ellis et al., 2002). This further increases the protractibility of the data and could lead to hypothesis development. Furthermore, the current increase in new technologies, such as FT-MS means that the future scope for metabolomic research is being continually expanded.
Future research will simultaneously investigate plant and leachate chemistry under different competitive scenarios in trying to identify allelopathic interactions. FT-IR screening, data mining and more targeted metabolomic approaches now offer enormous potential to uncover the chemical basis of allelopathy and further our understanding of plant-plant interactions.
Bais HP, Park SW, Weir TL, Callaway RM and Vivanco JM (2004). How plants communicate using the underground information superhighway. Trends in Plant Science 9, 26-32.
Bauer G and Richter W (1996). Optical characterization of epitaxial semiconductor layers, pp 225-232. Springer-Verlag, Berlin.
Blum U., Schafer S.R. & Lehman M.E. (1999). Evidence for inhibitory allelopathic interactions involving phenolic acids in field soils: Concepts vs. an experimental model. Critical Reviews in Plant Sciences 18, 673-693.
Ellis D.I., Broadhurst D., Kell D.B., Rowland J.J. & Goodacre R. (2002). Rapid and quantitative detection of the microbial spoilage of meat by Fourier transform infrared spectroscopy and machine learning. Applied and Environmental Microbiology 68, 2822-2828.
Ellis D.I., Harrigan G.G. & Goodacre R. (2003). Metabolomic fingerprinting with Fourier-transform infrared spectroscopy. In Metabolic Profiling: its role in biomarker discovery and gene function analysis, pp 111-124. Edited by Harrigan G.G & Goodacre R.G. Kluwer Academic Publishers.
Fiehn O. (2001). Combining genomics, metabolome analysis, and biochemical modelling to understand metabolic networks. Comparative and Functional Genomics 2, 155-171.
Fiehn O., Kopka J., D÷rmann P., Altmann T., Trethewey R.M. & Willmitzer L. (2000a). Metabolite profiling for plant functional genomics. Nature Biotechnology 18, 1157-1161.
Fiehn O., Kopka J., Trethewey R.N. & Willmitzer L. (2000b). Identification of uncommon plant metabolites based on calculation of elemental compositions using gas chromatography and quadropole mass spectroscopy. Analytical Chemistry 72, 3573-3580.
Gidman E., Goodacre R., Emmett B., Smith A.R. & Gwynn-Jones D. (2003). Investigating plant-plant interference by metabolomic fingerprinting. Phytochemistry 63, 705-710.
Goodacre R., Timmins ╔.M., Rooney P.J., Rowland J.J. & Kell D.B. (1996). Rapid identification of Streptococcus and Enterococcus species using diffuse reflectance-absorbance Fourier transform infrared spectroscopy and artificial neural networks. FEMS Microbiology Letters 140, 233-239.
Goodacre R., York E.V., Heald J. & Scott I.M. (2002). Chemometric discrimination of unfractionated plant extracts analyzed by electrospray mass spectroscopy. Phytochemistry 62, 859-863.
Gwynn-Jones D., Huang W., Easton G., Goodacre R. & Scullion J. (2004). UV-B radiation induced changes in litter quality affects earthworm growth and cast characteristics as determined by metabolic fingerprinting. Pedobiologia 47, 784-787.
Inderjit (1996). Plant phenolics in allelopathy. Botanical Review 62, 182-202.
Inderjit, Streibig J.C. & Olafsdotter M. (2002). Joint action of phenolic acid mixtures and its significance in allelopathy research. Physiologia Plantarum 114, 422-428.
Inderjit & Weston L.A. (2000). Are laboratory bioassays suitable for prediction of field responses? Journal of Chemical Ecology 26, 2111-2118.
Johnson H.E., Broadhurst D., Goodacre R. & Smith A.R. (2003). Metabolomic fingerprinting of salt-stressed tomatoes. Phytochemistry 62, 919-928.
Johnson H.E., Broadhurst D., Kell D.B., Theodorou M.K., Merry R.J. & Griffith G.W. (2004). High-throughput metabolic fingerprinting of legume silage fermentations via Fourier transform infrared spectroscopy and chemometrics. Applied and Environmental Microbiology 70, 1583-1592.
Klironomos J.N. & Allen M.F. (1995). UV-B mediated change on below-ground communities associated with the roots of Acer saccharum. Functional Ecology 9, 923-930.
Narasimhan K., Basheer C., Bajic V.B. & Swarup S. (2003). Enhancement of plant-microbe interactions using a rhizosphere metabolomics-driven approach and its application in the removal of polychlorinated biphenyls. Plant Physiology 132, 146-153.
Sanchez-Moreiras A.M., Weiss O.A. & Reigosa-Roger M.J. (2003). Allelopathic evidence in the Poaceae. Botanical Review 69, 300-319.
Schultze M. & Kondorosi A. (1998). Regulation of symbiotic root nodule development. Annual Review of Genetics 32, 33-57.
Scullion J., Elliott G.N., Huang W.E., Goodacre R., Worgan H., Darby R., Bailey M.J., Gwynn-Jones D., Griffith G.W., Winson M.K., Williams P.A., Clegg C. & Draper J. (2003). Use of earthworm casts to validate FT-IR spectroscopy as a 'sentinel' technology for high-throughput monitoring of global changes in microbial ecology. Pedobiologia 47, 440-446.
Singh H.P., Batish D.R. & Kohli R.K. (2003). Allelopathic interactions and allelochemicals: New possibilities for sustainable weed management. Critical Reviews in Plant Sciences 22, 239-311.
Siqueira J.O., Safir G.R. & Nair, M.G. (1991). Stimulation of vesicular-arbuscular mycorrhiza formation and growth of white clover by flavonoid compounds. New Phytologist 118, 87-93.
Sumner L.W., Mendes P. & Dixon R.A. (2003). Plant metabolomics: large-scale phytochemistry in the functional genomics era. Phytochemistry 62, 817-836.
Winson M.K., Goodacre R., Woodward A.M., Timmins ╔.M., Jones A., Alsberg B.K., Rowland J.J. & Kell D.B. (1997). Diffuse reflectance absorbance spectroscopy taking in chemometrics (DRASTIC). A hyperspectral FT-IR-based approach to rapid screening for metabolite overproduction. Analytica Chimica Acta 348, 273-282.