Bonnie Watson	Zhentian Lei	Satish Nagaraj

Introduction
The impressive achievements in genomic sequencing over the past several years have yielded a wealth of information for many model organisms, including the model plants Arabodopsis thaliana and Medicago truncatula. At the same time, the knowledge that sequence information alone is insufficient to answer many biological questions concerning gene function, developmental/regulatory biology, and the biochemical kinetics of life have defined the need for more comprehensive approaches. To address these questions, many are turning to quantitative and qualitative analyses of gene expression products at the transcriptome, proteome, and even metabolome levels. Transcriptome approaches using microarray and SAGE technologies are powerful tools, however, mRNA abundances may only represent putative function since there is still a questionable correlation between mRNA and protein levels [xxiii]. In contrast, proteomics provides a more physiologically accurate snapshot of biochemical processes by revealing the actual protein constituents performing the enzymatic, regulatory, and structural functions encoded by the genome and transcriptome at a given point in time.  Recent improvements in high resolution 2-DE [xxiv], increased content of protein and nucleotide databases, and the increased capabilities for protein identification utilizing modern mass spectrometry methods such as MALDI-TOF/MS [xxv] has made the large scale identification of proteins a dynamic new area of research in plant biology.

FIGURE 1.  Illustration of the proteomic process starting with extraction, profiling,
in-gel digestion, peptide mass mapping, database searching and protein identification.

We are utilizing proteomics to study the model legume Medicago truncatula (barrel medic). Legumes are valuable agricultural and commercial crops that serve as important nutrient sources for both humans and animals. For example, over 60 million acres of Glycine max (soybeans) are planted annually and have an estimated annual value in excess of 10 billion dollars. This crop provides more than 50% of the world's supply of oilseed and the residual soy meal is used as a high protein animal feed. Similarly, Medicago sativa (alfalfa) is a high protein forage with over 26 million acres planted annually which have a US estimated annual value approaching 12 billion dollars. Legumes are characterized by their symbiotic relationships with both nitrogen fixing bacteria and arbuscular mycorrhizal fungi. These host-symbiot interactions not only result in the ability to fix atmospheric nitrogen, but also effect mutualistic and defense related biosynthetic pathways. Many of these pathways are unique to legumes and are not found in the current model plant Arabidopsis thaliana, making legumes an important taxa for molecular-genetic studies. These biosynthetic pathways include natural products such as the isoflavones that have been reported to possess antimicrobial properties, anti-carcinogenic and health promoting effects. Other secondary metabolites such as the triterpenes are also of interest as novel pharmaceuticals. The major disadvantages of using many of the agriculturally important legumes such as soybean and alfalfa as a model species are their large genome size and complex ploidy. M. truncatula is diploid and has a smaller genome relative to other legumes. The diploid nature of the M. truncatula genome provides a system with more manageable genetics. These traits, along with its autogamous nature, short generation time and prolific seed production have made M. truncatula a more suitable and successful model legume for genetic and biological studies.

We have chosen two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) as our primary profiling tool.  Since its introduction in 1975 by O'Farrell,[i] 2D-PAGE has been established as the dominant technique for proteomic analysis.[ii]  This technique utilizes isoelectric focusing and polyacrylamide gel electrophoresis for first and second dimension separation, respectively.  Currently, 2D-PAGE technology is capable of resolving some 10,000 proteins with 2,000 proteins being somewhat routine [iii]. A recent review describes the role of 2D-PAGE in proteomic and genetic studies of plant systems, including its use as a tool to investigate genetic diversity, phylogenetic relationships, mutant characterization, and drought tolerance.[iv]  2D-PAGE has also been successfully utilized in studies on defense-associated responses [v] and responses to methyljasmonate.[vi]  We recognize the limitations of a 2D-PAGE approach, e.g. difficulties with membrane proteins, narrow dynamic range, and difficulty in identifying low level proteins,[vii] however we feel it is currently the most proven technique and more immediately applicable to addressing biological question.  We will continue to monitor developing techniques such as capillary LC/MS/MS of protein digests, accurate mass tags, and isotope affinity tag [viii] technologies. 2D-PAGE gels are stained with Coomassie Brilliant Blue, silver,[xv] [xvi] or fluorescent dyes.  Currently we are considering the addition of DIGE capabilities in our research to minimize gel to gel variation  

Although 2D-PAGE analysis has been successfully utilized for the last 25 years in protein profiling, it provides little information concerning protein identification.  Recent advances in mass spectrometry and the establishment of genomic and protein databases have substantially increased the ease and speed with which proteins can be identified.  The union of these technologies is the foundation for modern proteomic studies.  A review, "Mass Spectrometry in the Age of the Proteome" of the different MS techniques and their role  was recently presented. [ix]

We are using 2D-PAGE in conjunction with matrix-assisted laser desorption ionization[xi] (MALDI) and electrospray ionization[xii],[xiii],[xiv] (Yamashita, 1984; Fenn 1990; Smith 1991) MS to investigate differences in protein expression in M. truncatula. First round protein identification is performed using a MALDI-TOF-MS peptide mass mapping approach. Proteins spots are excised, digested in-gel using a protease, and the resultant peptide masses measured. The observed peptide masses can be searched against a theoretical list of proteolytic peptide maps predicted by a given gene or protein database. Increased peptide mass accuracy has increased the success and selectivity of such searches.[x] If the database query is unsuccessful, the protein can be sequenced using tandem mass spectrometry (MS/MS). During the MS/MS experiment, only the peptide mass of interest is isolated or transmitted, thus discriminating against all other components of the mixture with different mass values. After isolation, the peptide is fragmented using a unimolecular or bimolecuar (collision gas) strategy. Fragments observed in the isolated peptide can then be rationalized to a sequence or used to search the databases using a "sequence tag" approach. LC/MS/MS and MALDI/MS/MS will be used as needed for more difficult identifications, sequencing, and post translational modification determinations.

Instrumentation                                                                                                                                                                                   Back to top

Our proteomics efforts involve an array of instrumentation.  These include an ABI Q-Star Pulsar, a PE Biosystems coupled to a LC Packings nano HPLC, a PerSeptive DE-STR MALDI-TOF-MS and a Bruker Esquire Ion-trap MS. We also have multiple 2-DE options. We use both Pharmacia and BioRad isoelectric focusing (IEF) units for the first dimension of the 2-DE experiment. IEF can be performed using various immobilized pH gradient (IPG) strips designed to fit mini (Novex or BioRad 10 x 10cm), midi (BioRad Criterion 10 x 14cm), large (Hoeffer SE-600, 16 x 16cm), or very large (BioRad Protean III, 22 x 22cm) ( Pharmacia Ettan DALTsix and DALTtwelve, 24 x 24cm) 2-DE gel formats.

TIC of cell culture protein gel spot digest

TOF MS/MS spectrum of cell culture protein gel spot digest

Projects                                                                                                                                                                                                               Back to top
Our group has completed an overview of the Medicago truncatula proteome that include leaves, stems, roots, flowers, seed pods, and cell cultures. These data have been released in our 2D-PAGE database following publication (Ref below).  An example of a Medicago truncatula cell suspension proteome is provided below.  The functions of identified proteins have been summarized to provide information on tissue specific proteome expression.  A series of pie charts are also provided below to summarize the functional aspects of proteins identified in specific tissues.

2D-PAGE analysis of M. truncatula cell suspension cultures

See: A Two-dimensional Electrophoresis Proteomic Reference Map and Systematic Identification of 1367 Proteins from a Cell Suspension Culture of the Model Legume Medicago truncatula. Molecular & Cellular Proteomics, 2005, vol 4. p 1812-1825.

Summary of the functions of various proteins identified in specific tissues of M. truncatula.

Our group is also involved in the proteomics aspect of a Medicago truncatula functional genomics project that is underway at the Noble Foundation. Analytical and biological variance studies have been carried out to quantify inherent irreproducibility in 2D PAGE experiments due to isoelectric focusing, SDS electrophoresis, staining, etc. Analytical variance was quantified by obtaining ten replicate 2D gels (protein samples aliquoted from the same pool) and measuring differing protein spot volumes (intensities) by image analysis and pattern matching. The biological variance was quantified from ten 2D gels of leaf protein extracts from ten different plants grown under identical conditions. Protein expression levels were quantified from gel images using a computer imaging system (BioRad Fluor-S) and software package (Phoretix, NonLinear Dynamics) for spot detection, background subtraction, spot volume quantification, and spot matching across gels.  We have used two methods of spot detection: automated and manual/user guided spot detection. In the automated spot detection mode, a "spot detection wizard" was used set the basic parameters for spot detection through a reiterative visual process. These parameters were then applied to all gels. A single gel was arbitrarily chosen as a reference gel and the spots in the reference gel were then matched across the nine other gels after adding user "seeds" or reference points. Unfortunately, only 50 of the 500 spots could be matched across different gels in a totally automated mode; therefore, only spots that matched in six or more gels were chosen for the statistical analysis. Fifty spots in the biological gel and fifty spots in the analytical gel were found to be common in six or more gels. A semi-automated spot detection method requiring user guided spot detection was also performed that enhanced the software's ability to match spots more accurately. User directed spot detection was performed through visual examination of the gel images and employing circular perimeters. This method was used to determine the spot volumes for the same fifty protein spots described above for the same ten gels. Using this approach all but two spots (i.e. 2 out of 500) in the reference gel were successfully matched across all other gels. The average coefficients of variation (CV) of the 50 normalized spot volumes quantified through automated detection were determined to be 39% for the analytical variance and 55% for the biological variance. The coefficients of variation for semi-automated analytical and biological variations were calculated to be 16% and 22%, respectively. Variances in molecular weight and pI were also calculated from fifty semiautomatic detected spots and analytical and biological CVs calculated. The results are tabulated in Table 1.  We propose from our variance results that for a protein expression difference to be statistically significant at the 95% confidence level, it has to differ by 3.92s (correlates to 94% difference in CV values).

Average Analytical and Biological Coefficients of Variance Determined for Protein

Expression Levels from Ten Replicate 2-DE Analysis of Medicago truncatula Leaf Extracts

Table I.

2-D Database                                                                                                                                                                                                          Back to top
The database is meant to provide a repository for 2D-PAGE and proteomics information for plant research. The majority of the data is focused on the model legume, Medicago truncatula; however, data from other species are included and welcomed.

We welcome your comments and encourage your participation through data submission.

References                                                                                                                                                                                                                  

Back to top

 

	Cell_R250: Protein was extracted from 9th passage Medicago trancatula suspension cell culture. The first dimension was run with 1800 mg of protein loaded and a total of 67,500 Volt-Hours. The second dimension was run on 12.5% Acrylamide-PDA for 15 hours at 3 watts per gel. The gel was stained with 0.1% coomassie brilliant blue R-250.
	Cell_R350: Protein was extracted from 6th passage Medicago trancatula suspension cell culture. The first dimension was run with 1500 mg of protein loaded and a total of 67,500 Volt-Hours. The second dimension was run on 12.5% Acrylamide-PDA for 15 hours at 3 watts per gel. The gel was stained with PlusOne? Coomassie? Blue PhastGel? R-350 (Amersham Biosciences).
	Cell _G250: Protein was extracted from 6th passage Medicago trancatula suspension cell culture. The first dimension was run with 1500 mg of protein loaded and a total of 67,500 Volt-Hours. The second dimension was run on 12.5% Acrylamide-PDA for 15 hours at 3 watts per gel. The gel was stained with 0.1% coomassie brilliant blue G-250.
	Cell_Sypro: Protein was extracted from 2th passage Medicago trancatula suspension cell culture. The first dimension was run with 1200 mg of protein loaded and a total of 67,500 Volt-Hours. The second dimension was run on 12.5% Acrylamide-PDA for 15 hours at 3 watts per gel. The gel was stained with Sypro Ruby.

Gel Images of Stem Proteins from 6 Different Medicago truncatula Ecotypes

- click to enlarge -

Proteins were extracted by grinding in liquid nitrogen and precipitating with 10%TCA in acetone. Only the top 2 internodes of 5 week-old chamber-grown plants were used. Resolubilization buffer consisted of 8 M urea, 4% CHAPS, 20 mM DTT, and 0.5% biolytes. One mg of each sample was focused on 24 cm pH 3-10 strips to 90000 Vhrs. and separated on 12% acrylamide gels. The gels were stained overnight with Coomassie R-250.