NSF Project: Results

Optimization of cell culture conditions (Dixon)

We found that root-derived cell suspension cultures of M. truncatula responded to both yeast elicitor (for induction of phenylpropanoid natural products) and to methyl jasmonate (MeJA) (for production of triterpene saponins). Concentrations of these elicitors were optimized for maximum response. Importantly, yeast elicitor treatment did not significantly induce triterpenes in the cultures, and, conversely, methyl jasmonate did not appear to induce the phenylpropanoid pathway.

By mining existing M. truncatula EST datasets, we were able to identify, and subsequently functionally characterize, the first three enzymes at the entry point to the triterpene pathway (squalene synthase, squalene epoxidase and ?-amyrin synthase, ?-AS, a triterpene cyclase). These will be important markers for triterpene pathway induction. Following exposure of test cultures to 50 µM MeJA, there was a 30-fold induction of ?-AS transcripts, but a decrease in PAL and CHS transcripts from the phenylpropanoid pathway. A manuscript on this work has been submitted (Achnine et al., 2002), and a second paper on the optimization of the system with different elicitors is in preparation. A patent (PCT provisional) has also been filed on the use of elicited M. truncatula cell cultures for triterpene pathway gene discovery.

Preliminary microarray conditions (Dixon, May)

Using the 70 mer oligo arrays, we have generated preliminary data on the induction of known natural product pathway genes in response to a range of elicitors in M. truncatula cell suspension cultures. The results indicate that the microarrays are working and reveal up-regulation of specific members of the multigene families encoding several key biosynthetic enzymes such as L-phenylalanine ammonia-lyase. We are now ready to begin the detailed kinetic analysis of transcriptional changes in the large-scale elicitor time course experiments.

Time course expression profiling (May and Dixon)

Expression profiling of the MeJA and yeast elicitor time courses have begun. Control vs. Reference RNA sample and Elicited vs. Reference RNA sample hybridizations are being performed serially and in triplicate. Comparisons are being made of gene expression profiles over the course of the elicitation experiments. Methods for expression analyses data transfer and data upload between Noble and VBI are being evaluated. Typical results from microarray hybridizations are represented in Transcriptomics.

The use of polysomal RNA vs. total RNA in expression profiling experiments - DNA microarray analyses are routinely used to evaluate steady-state mRNA levels, but RNA abundance does not always correlate with protein levels. We have compared polysome-bound mRNA versus total RNA profiling in yeast-elicited cell cultures. Approximately 10% of the genes demonstrated two-fold or greater differences that are the exact opposite in the profiles between polysome-bound and total mRNAs. For example, transcript abundance for dihydroflavonol-4-reductase appears to increase in yeast-elicited cell cultures when total RNA is incorporated in the experiment while relative abundance of this same transcript appears to decrease under these same conditions when polysomal RNAs are used. Quantitative RT-PCR has been used to confirm and validate these results.

Optimization of protocols for proteomics (Sumner)

Currently over 300 proteins have been identified from Medicago truncatula tissues during the studies for optimization of proteomics protocols. A representative 2-DE gel reference map obtained for cell cultures is provided in Figure 1. To date approximately 50 proteins have been identified from cell culture gels and are also listed in Figure 1. We were encouraged by the observation and identification of several flavonoid pathway enzymes such as isoflavone reductase and chalcone-flavone isomerase. Our work will continue to increase the number of proteins identified from the cell culture reference map and will pursue comparative proteome expression experiments in response to elicitation.

Results of experiments carried out to quantify the level of analytical and biological variance associated with 2-DE of Medicago truncatula are shown in Figure 2. From these experiments, differences of approximately 100% are necessary to be real at the 95% confidence level.

Figure 1 - Two-dimensional polyacrylamide gel electrophoresis (2-DE) reference map for Medicago truncatula suspension cell cultures with a tabular listing of proteins identified to date (Note multiple isoforms observed for some proteins such as chitinase).

Figure 2 . Summary of analytical and biological variances associated with quantitative profiling of protein expression. Data suggests that differences of approximately 100% are necessary for the observation to be statistically supported as real.

Optimization of protocols for proteomics (Sumner)

Currently over 300 proteins have been identified from Medicago truncatula tissues during the studies for optimization of proteomics protocols. A representative 2-DE gel reference map obtained for cell cultures is provided in Figure 1. To date approximately 50 proteins have been identified from cell culture gels and are also listed in Figure 1. We were encouraged by the observation and identification of several flavonoid pathway enzymes such as isoflavone reductase and chalcone-flavone isomerase. Our work will continue to increase the number of proteins identified from the cell culture reference map and will pursue comparative proteome expression experiments in response to elicitation.

Results of experiments carried out to quantify the level of analytical and biological variance associated with 2-DE of Medicago truncatula are shown in Figure 2. From these experiments, differences of approximately 100% are necessary to be real at the 95% confidence level.

Figure 1 - Two-dimensional polyacrylamide gel electrophoresis (2-DE) reference map for Medicago truncatula suspension cell cultures with a tabular listing of proteins identified to date (Note multiple isoforms observed for some proteins such as chitinase).

Figure 2 . Summary of analytical and biological variances associated with quantitative profiling of protein expression. Data suggests that differences of approximately 100% are necessary for the observation to be statistically supported as real.

Optimization of protocols for metabolite profiling (Sumner, Smith)

Example GC/MS chromatograms of a polar and lipophilic extracts are provided below in Figure 3. Each chromatogram yields approximately 300 components of which about half have been identified. These examples also show that complementary information is obtained from the multiple chromatograms.

Figure 3- Example GC/MS chromatogams of a polar and lipophilic extracts of Medicago truncatula.

Figure 4 . Capillary electrophoresis (CE) of ionic constituents and CE/MS analysis of amino acids from Medicago truncatula.

Representative CE and CE-MS data are outlined in Figure 4. Current efforts are focused on yielding a combination method that will profile all the above mention analytes by CE/MS.

We have recently used LC/MS to show that saponin synthesis is induced through methyl jasmonate elicitation (Figure 5).

Figure 5 - Negative-ion LC/ESI/MS analyses of Medicago truncatula before and 24hr post elicitation with methyl jasmonate showing a dramatic increase in many of the saponin conjugates.

Preliminary data analysis of metabolomic data has been carried out at the Sumner lab. PERL and Java scripts are being developed for data reformatting allowing data transfer to the commercial software Pirouette. Pirouette is then used for hierarchical cluster analysis (HCA) and principal component analyses (PCA) of data sets. An example of 2D PCA of a development study of Medicago truncatula is illustrated in Figure 6. These multivariate statistical analyses will be implemented on the back-end informatics system, these preliminary results demonstrate their usefulness for metabolomic data.

2D PCA of M. truncatula Development, POLAR Extract
(300 samples, aligned, cleaned, 160 peaks, mean center log transformed)

Figure 6 . Two-dimensional PCA analysis of Medicago truncatula development study using GC/MS of the polar fraction. Data shows clear differentiation of early development stages from the later ones. The profiles suggest that multiple components of the TCA cycle are increasing over time.

Optimization of protocols for metabolic profiling (Smith) 

Protocols for the profiling of carbohydrates have been developed in Dr. Smith?s laboratory using capillary electrophoresis (CE). Initially we had developed our methods using M. truncatula plant tissue, which included tagging the extracted carbohydrates with a chromophore, 2-amino benzonitrile, followed by UV detection in CE. When cell cultures were analyzed, a significant variation in the carbohydrate composition was observed. In addition, the media contained several components that interfered with the direct analysis of several carbohydrates. A new CE protocol has been developed that tags the carbohydrates with 2-aminoacridone (2-AC). 2-AC is not only a strong chromophore but an excellent fluorophore. Figure y shows an electropherogram of a control cell culture, the media, and a blank with UV detection. Several carbohydrates are easily detectable but many are present at levels near the limit of detection. Laser induced fluorescence (LIF) was used to enhance the detection sensitivity. Figure 8 shows an electropherogram of the control cell culture using LIF detection, which clearly indicates a number of carbohydrates not detected in the UV mode. 2-AC derivatized carbohydrates are also advantageous because the separation method can be readily adapted to mass selective (MS) detection for positive identification of carbohydrates.

Figure 7. CE-UV of Medicago truncatula cell culture 132. Conditions: capillary 40/48.5 cm x 50 mm; voltage, +15 kV; electrolyte, 100 mM borate, pH 10.0; detection, 260 nm.

Amino acids will be profiled using CE with MS detection. CE-MS allows for amino acids to be rapidly quantified without the need for extensive sample cleanup or derivatization of the amino acids. This methodology was recently developed in Dr. Smith?s laboratory and allows rapid amino acid profiling. A dedicated CE instrument for the MS is currently being acquired from other sources of funding in Dr. Smith?s laboratory that will allow for completely automated instrumentation for the amino acid profiling.

Figure 8. CE-LIF of Medicago truncatula cell culture 132. Conditions: capillary 40/50 cm x 50 mm; voltage, +15 kV; electrolyte, 100 mM borate, pH 10.0; detection, excitation He-Cd laser (442 nm), emission 520 nm.

Top