Metabolomics is used to determine how an abiotic stress impacts that actual chemical composition of a plant. The number of chemical metabolites in an organism is far smaller than the number of proteins or genes, and many metabolites are characteristic of multiple biochemical pathways. Metabolomics studies of plants in a non-targeted approach allows scientists to observe how a plant responds to an environmentally-induced stress. Two different analytical instrument platforms have emerged to probe the chemical fingerprints found within the metabolome: nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry (MS). NMR is a nondestructive technique, so the sample can be analyzed by other methods once NMR is conducted. Furthermore, NMR does not require separation of the components of a biological sample, and can measure multiple chemical species simultaneously. While the high reproducibility and simplicity of NMR makes it a suitable platform when levels of a metabolite are sufficiently high enough in the sample of interest, the majority of metabolites are below detectable levels by NMR; furthermore, because separation of the chemical components is not performed, overlapping signals in NMR can greatly complicate interpretation of the data obtained. In contrast, MS has much lower detection limits than NMR, and with tandem mass spectrometry (MS/MS), provides for high selectivity for chemical identification; this is often further enhanced when hyphenating a mass spectrometer with gas chromatography (GC) or liquid chromatography (LC). In addition, the emergence of numerous metabolomics databases based on MS or MS/MS data enable identification of many chemical constituents that are metabolites.
A limitation in the hyphenated chromatography mass spectrometry methods is the low throughput, as chromatography experiments can take 30-60 minutes each. However, it was recognized that high resolution mass spectrometry (HR-MS), because of its high mass accuracy capabilities, could still provide high chemical specificity and identification without the need for hyphenation to chromatography using direct infusion of the samples. This greatly decreases the time for analysis per sample, enabling higher throughput and more replicates to be acquired in shorter time frames. This is of particular importance in plant metabolomics, where numerous types of stresses can be applied to plants (e.g., drought, flooding, high salinity soil, etc.) and collectively examined for their impact. Of the HR-MS instruments, the Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer has the highest mass resolution and greatest mass accuracy, within 1 ppm error. Recognizing the power of direct infusion FT-ICR, this technique was first applied to plant phenotyping, demonstrating the ability to detect differences during the ripening process in strawberries and distinguishing transgenic tobacco from wild-type.
Our group was the first to apply direct infusion FT-ICR to the metabolomics of soybean leaves, in which the metabolic profiles of solvent extracts of soybean leaves harvested from in-field plants was compared between mid-summer and harvest day (autumn). Substantial and easily distinguished profiles of the extracts harvested from the different seasons were obtained. From five technical replicates of each extract procedure, we were able to construct a metabolic “heatmap” of fifty m/z ratios most altered in their expression levels as a function of the growing season.
Subsequently, our group started a collaboration with Professor Henry Nguyen, Curators' Distinguished Professor of Plant Sciences and Director of the Molecular Genetics and Soybean Genomics Laboratory at the University of Missouri. The Nguyen group investigates soybean genomics, and has been working toward the development of soybean cultivars tailored to produce soybean crop even under conditions of abiotic stress. In particular, water stress due to drought and flooding are estimated to cause 40-60% decreases in crop yields; drought itself has a negative impact on all developmental stages of soybean plants. To enhance seed quality and reduce yield losses, the development of drought tolerant cultivars has been deemed imperative. Mutava et al. examined different soybean cultivars in an attempt to gain insight into mechanisms of tolerance to drought and flooding stress. Prince et al. investigated soybean root traits that might help plants adapt to drought, and identified several genotypes that could be used to enhance drought tolerance of elite soybean cultivars through the development of root traits designed for specific soil types.
Two different cultivars of soybean plants in the field were prepared (latitude 38.895305, longitude -92.205917): a drought-susceptible cultivar, Pana (PI 597387), and one which is drought-tolerant, (PI 567731). These cultivars were prepared in two physiological age groupings in petri dishes: young at 1 week old and old at 2 to 3 weeks old. The cultivars were also grown under two different water-access conditions: drought treatment consisted of no irrigation and rainfall in the field for 3 weeks, and controls, which were within irrigated plots and were irrigated two days prior to collection. These the leaves were flash frozen at -80 °C and transported on dry ice, then stored at -20 °C until solvent extracts were prepared. For the extracts, tissue was weighed and macerated for 1 minute with liquid nitrogen in a mortar and pestle, and internal standard (quinapril HCl) was added alongside an aliquot of 10mL of solvent; maceration of tissue continued for 5 minutes and an additional aliquot of ethanol was used to wash the pestle prior to vacuum filtration. Solid was filtered using Whatman filter papers (Cat. No. 1001-055) and the remainder of the 10mL of solvent was used to wash the mortar and filtration flask. The solvent was subsequently dried in a vacuum oven at ambient temperature (20 in Hg below atmospheric pressure). The dried solutions were reconstituted in 2mL of solvent and diluted to instrumental levels 1:50 in ethanol. The extracts were then processed using direct infusion (DI) ESI FT-ICR MS on a 12T Bruker SolariX FT-ICR MS, methods were developed from previous work on soybean leaves.7 Triplicate instrumental and technical replicates were performed, and peak lists were processed using Bruker Data Analysis 4.2. Principle component analysis (PCA) was used to group metabolites to detect correlations amongst metabolites. PCA was able to show distinct grouping of metabolites, whether young or old leaves of drought-susceptible or drought-tolerant leaves were investigated. Perhaps most notably, both the drought-susceptible and drought-tolerant leaves express similar profiles under drought abiotic stress (circled purple regions). Young leaves show the greatest variance within drought susceptible and treated plants. Meanwhile, old drought and control populations both are tightly grouped within each of the cultivars tested, showing that as the life cycle of the plants progresses the metabolome stabilizes.
Most importantly, direct infusion ESI FT-ICR was used to construct a library of molecular components for both drought-tolerant (PI 567731) and drought-susceptible Pana (PI 597387) soybean cultivars; a manuscript on this is in preparation.
In this proposal, we will mine the existing data to identify possible metabolites in the grouping using on-line databases; when promising candidates are indicated, these species will be subjected to MS/MS in order to identify the molecular species when it is not known. As an example, Table 1 shows several different m/z values that are observed in both young and old leaves of the drought-tolerant cultivar PI 567731. The chlorophyll-related metabolite at m/z 1073.72 is a metabolic product of chlorophyll with an additional C12H20O- group. Abundances of these metabolites are altered depending on age of the leaves and the cultivar used.
With many of these species already identified, as indicated in Table 1, along with their abundances, we can compile lists of metabolites that are overexpressed or underexpressed due to the drought abiotic stress. It is likely that the drought-susceptible and drought-tolerant cultivars will possess different molecular changes when faced with drought. Specifically, we are most interested in deducing, from the molecules identified, which pathways are most impacted by drought stress. Moreover, we wish to investigate what biochemical pathways drought-tolerant cultivars use to adapt to drought stress. Metabolic pathway analysis will be conducted by using the Pathway Analysis Module of Metaboanalyst (http://www.metaboanalyst.com). Using the abundance analysis of metabolites in the catalog for both cultivars under the control and drought-stress conditions, and initiate pathway enrichment analysis. Once this is generated, we manually construct the pathways using Kyoto Encylopedia of Genes and Genomes (KEGG) database (https://www.genome.jp/kegg/) in an effort to identify potential drought “phenotypes” that indicate how soybean plants adapt to drought stress. Moreover, this analysis will also reveal distinguishing features to establish a “young” leaf phenotype from an “old” leaf phenotype.
Table 1. Different Molecular Species Detected in
m/z Formula Adduct Putative Identity
455.116 C18H24O12 Na licoagroside b
631.470 C37H68O5 K Diacylglycerol 34:2
633.143 C27H30O16 Na diglycoside
649.116 C27H30O16 K diglycoside
803.574 C44H83O10P H Phosphoglycerol 34:2
805.663 C55H90O K undecaprenol
871.573 C55H74N4O5 H pheophytin a
909.529 C55H74N4O5 K pheophytin a
911.652 C57H92O6 K Triacylglycerol 54:9
913.669 C57H94O6 K Triacylglycerol 54:8
1051.722 C67H94N4O6 H Chlorophyll related metabolite
1057.732 C67H94N4O6 Li Chlorophyll related metabolite
1073.708 C67H94N4O6Mg H Chlorophyll related metabolite