Comparative Proteomics of Yeast-Elicited Medicago truncatula Cell

Comparative Proteomics of Yeast-Elicited Medicago truncatula Cell Suspensions Reveals Induction of Isoflavonoid Biosynthesis and Cell Wall Modifications Zhentian Lei, Fang Chen, Bonnie S. Watson, Satish Nagaraj, Aaron M. Elmer, Richard A. Dixon, and Lloyd W. Sumner* Plant Biology Division, The Samuel Roberts Noble Foundation, 2510 Sam Noble Parkway, Ardmore, Oklahoma 73401, United States Received May 12, 2010

The temporal proteome response of Medicago truncatula suspension cell cultures to yeast elicitation (which mimics fungal infection) was investigated using two-dimensional polyacrylamide gel electrophoresis (2-DE) and nanoliquid chromatography coupled to tandem mass spectrometry (nano LC-MS/ MS). Reproducibility of 2-DE was assessed using the number of the visualized protein spots and spot volume. Average coefficient of variation was determined to be less than 6% for the number of spots and around 50% for spot volume. About 4% of the total visualized proteins, that is, 34 out of 861, were differentially accumulated in the suspension cells 24 h after yeast elicitation, including isoflavononid biosynthetic enzymes and a putative laccase. The induction of the putative laccase was highly correlated with the polymerization of phenolics such as 4-hydroxybenzoic acid, 4-hydroxybenzaldehyde, and ferulic acid into cell walls. In contrast, lignin was only induced at the later stages of the temporal study, indicating that this specific laccase is primarily involved in cell wall modifications and/or fortifications rather than in lignification in response to yeast elicitation. Keywords: proteomics • Medicago truncatula • 2-DE • gel reproducibility • LC-MS/MS • lignin • wallbound phenolics • cell suspension • yeast elicitation

1. Introduction Legumes are economically important crops rich in protein and a wide variety of natural products including flavonoids, isoflavonoids, saponins, and alkaloids.1,2 Many of these natural products are associated with plant disease/defense responses and plant-microbe interactions such as nodulation gene inducers/repressors in Rhizobia. The unique and legume specific Rhizobia symbiosis enables legumes to fix atmospheric nitrogen, a trait of great economical and ecological significance and one that cannot be studied using the well-characterized plant model Arabidopsis thaliana. Medicago truncatula (common name barrel medic) has emerged as a model legume due to its small diploid genome size, short generation time, and high transformation efficiency.3,4 It has also been chosen as a model species for studying forage quality traits such as digestibility, nutritional value, and natural product/ secondary metabolite biosynthesis due to its close relationship with the economically important alfalfa (Medicago sativa).2 Several legume natural products, especially those derived from the isoflavonoid pathway such as genistein and daidzein, have been documented for their anticancer properties5,6 and, therefore, have potential in biotechnological applications. Isoflavonoids and their pterocarpan derivatives such as medicarpin are predominantly legume-specific and normally absent or rare in * To whom correspondence should be addressed. Lloyd W. Sumner. Phone: (580) 224-6710. Fax: (580) 224-6692. E-mail: [email protected].

6220 Journal of Proteome Research 2010, 9, 6220–6231 Published on Web 09/21/2010

other plant families including the well-characterized plant model Arabidopsis, making M. truncatula an excellent model for studying the biosynthesis of these natural products.2 Plant cell cultures have been successfully used to study gene expression and associated natural product biosynthesis in response to both abiotic and biotic stress elicitors. Using elicitor-induced plant cell cultures, Suzuki and co-workers7 showed that a large number of genes involved in secondary metabolism including L-phenylalanine ammonia-lyase, cinnamate 4-hydroxylase, caffeic acid 3-O-methyltransferase and chalcone synthase were differentially regulated in response to different elicitors in M. truncatula. This report demonstrated that the M. truncatula cell cultures are an excellent system for understanding the regulation of legume secondary metabolism.8 Historically, investigations of plant responses to stress elicitors have focused on specific enzymes or biochemical pathways. Little data is available concerning the more holistic response to stress but is much needed to complete an integrated functional understanding. Recent advances in proteomics have made large-scale protein profiling possible for specific tissues and in a temporal manner. High-resolution, two-dimensional polyacrylamide gel electrophoresis (2-DE) still provides one of the highest resolution separation techniques for proteins where separation is based upon protein isoelectric point (pI) and molecular weight (MW).9 Mass spectrometry (MS) is used to identify proteins of interest following in-gel digestion. Proteomics has been used 10.1021/pr100439k

 2010 American Chemical Society

Isoflavonoid Biosynthesis and Cell Wall Modifications 10,11

to study the proteomes of Arabidopsis chloroplast, vacuoles,12 root microsomal fractions,13 and root-microbe interactions.14 Proteomic approaches have also been successfully employed to investigate M. truncatula proteins associated with symbiosis,15 seed development,16 and in six differentiated tissues17 as well as embryogenic cell cultures.18 Previously, we used 2-DE to generate a high-resolution proteome reference map for M. truncatula cell cultures that included over 1600 resolved proteins of which over 1300 were identified.19 To add to our understanding of M. truncatula stress responses, the changes in the global protein populations in M. truncatula cell cultures elicited with yeast cell walls were studied at a high temporal resolution and integrated with parallel transcriptomics and metabolomics data sets. The yeast cell wall elicitor mimics a fungal pathogen and induces defense responses against fungi. The specific observed temporal proteome changes and their relationship to plant defense are reported and discussed here.

2. Materials and Methods 2.1. Materials. HPLC grade acetonitrile (ACN) and water were purchased from Burdick & Jackson. Formic acid, iodoacetamide (IAA), dithiothreitol (DTT), BF3 etherate, dioxane and ethanethiol were purchased from Sigma-Aldrich. 3-[(3-Cholamidopropyl)dimethyl ammonio] propanesulfonic acid (CHAPS) was purchased from USB Corporation. 2.2. Suspension Cell Cultures and Yeast Elicitation. Suspension cell cultures were initiated from M. truncatula cv Jemalong A17 root calli grown on modified Schenk and Hilderbrandt (SH)20 agar plates in the dark at 25 °C. Callus cells (5.0 g) were suspended in modified SH medium (40 mL) in a 125 mL Erlenmeyer flask in the dark at 22 °C with shaking at 130 rpm. Cultures were transferred to 250 mL flasks and subcultured approximately every 2 weeks until elicited. The cells were elicited during the 11th subculture passage by adding 2.5 mL of a 5 mg/mL aqueous solution of a yeast cell wall preparation for a final concentration of 50 µg glucose equivalents per mL.21 Cells (both control and elicited) were harvested in triplicate from separate culture flasks at 21 time points, that is, 0, 5, 15, 30, 45 min and 1, 2, 3, 4, 6, 8, 10, 12, 15, 18, 21, 24, 30, 36, 42, 48 h, washed with 50 mL of a 25% dilution of the original modified SH culture medium described above, divided into four 50 mL tubes, frozen in liquid nitrogen and stored at -80 °C until analyzed. 2.3. Protein Extraction, 2-DE Separation, and Image Analysis. Extraction and 2-DE separation of proteins were performed as previously described.19 Frozen cells were ground in liquid nitrogen, extracted with 40 mM Tris buffer (pH 9.5) containing 1 mM phenylmethylsulfonyl fluoride and 120 units/mL endonuclease. After centrifugation (5000× g at 4 °C for 10 min), the supernatant was brought to a final concentration of 12.5% (w/v) trichloroacetic acid. The precipitated proteins were recovered by centrifugation (15 000× g at 4 °C for 20 min), washed extensively, and resuspended in 2-DE solubilization buffer consisting of 9 M urea, 3% CHAPS, 2% Triton X-100, 20 mM DTT and 0.5% ampholytes. The protein concentration was determined using the Bradford method and a commercial dye reagent (Bio-Rad) with bovine serum albumin as a standard.22 First dimensional separation of proteins (1500 µg of protein in 450 µL) was performed on Immobilized pH gradient strips (Immobiline Dry Strips 24 cm, pH 3-10 nonlinear, Amersham) for a total of 75 000 V-hours. The second dimensional separation was performed on a 10% acrylamide gel following DTT

research articles reduction and IAA alkylation. Gels were stained with Coomassie Brilliant Blue R-250 and images were acquired on a UMax Astra 2400S scanner at 300 dpi and saved as a gray scale TIFF file. Experimental MW and pI were calculated from the digitized image using MW marker proteins and the predicted nonlinear pH gradient provided by Amersham Biosciences. The 2-DE images were analyzed using Phoretix 2D Expression software (Nonlinear Dynamics, U.K.). Briefly, all the 2D gel images were imported into the 2D Expression program. Spot detection was performed using mode of non-spot margin 45. Only spots with an area >200 and peak height >4000 were reported. Background subtraction was performed using mode of non-spot margin 45 and spot volume was normalized to total spot volume. 2.4. In-Gel, Trypsin Digestion and LC-MS/MS. Differentially accumulated proteins were excised from the gels manually and transferred to polypropylene, 96-well plates and sealed. The plates were then stored at -80 °C until further analyses as described previously.19 Briefly, the gel plugs were first destained with 50% ACN in 50 mM ammonium bicarbonate and then dehydrated with 100% ACN. After removal of ACN, the gel spots were dried under vacuum. The dried gel plugs were rehydrated with a bovine trypsin solution (10 ng/µL in 25 mM ammonium bicarbonate) and digested overnight at 37 °C. Digestion was stopped by adding 15 µL of 10% formic acid. The peptides were extracted, combined and concentrated under vacuum to a final volume of about 25 µL. Analyses of peptides were performed on an ABI QSTAR Pulsar i hybrid Q-TOF mass spectrometer coupled to a nanoscale LC system (LC Packings, San Fransico). Samples (5 µL) were first loaded onto a C18 precolumn (0.3 mm i.d. × 1.0 mm, LC Packings) for desalting and concentration. Peptides were then eluted from the precolumn and separated on a nanoanalytical C18 column (75 µm i.d. × 15 cm, LC Packings) at a flow rate of 200 nL/min. The eluted peptides were electrosprayed and ionized using a PicoTip needle (10 µm i.d., New Objectives) at a voltage of 2400 V. TOF-MS and tandem mass spectral data were acquired using an IDA (information dependent acquisition) feature in the Analyst QS software. The IDA experiments selected multiply charged precursors (charge state from 2 to 5) for fragmentation with an intensity threshold of 10 counts per second. Precursor ions were excluded for 90 s using a window of 6 atomic mass units and collision energy settings were automatically determined by the IDA based on the ion m/z values. 2.5. Database Queries and Protein Identifications. The acquired tandem mass spectral data were queried against a custom legume protein database using the MASCOT search engine (version 1.8, Matrix Science Ltd., U.K.) with a mass tolerance of 150 ppm and one trypsin miscleavage. Oxidation of methionine and carbamidomethylation of cysteine were set as variable amino acid modifications. The custom legume database was generated at the Noble Foundation (http:// bioinfo.noble.org/gateway/). Briefly, tentative consensus sequences or gene indices were downloaded from The Institute for Genomic Research (TIGR) (http://www.tigr.org/tdb/tgi. shtml). These sequences included: MtGI.053002 (M. truncatula, 22 652 records), GmGI.052802 (Glycine max, 32 081 records), and LjGI.053102 (Lotus japonicus, 7686 records). The nucleotide sequences were then translated into amino acid sequences (62 621 records) and annotated using an in-house EST analyzer developed at the Noble Foundation. For a given sequence, the EST analyzer searched NCBInr protein database to identify a homologous protein (the best hit of BLASTX search), then used the homologous protein to annotate the query sequence. Based Journal of Proteome Research • Vol. 9, No. 12, 2010 6221

research articles

Lei et al.

Figure 1. 2-DE image of the proteome of M. truncatula suspension cells (control, 0 h time point, replicate 1) using a pH 3-10 nonlinear IPG strip and 10% acrylamide gel. Spots showing differential displays in the time-course experiment are annotated with arrows and number. Inserts are expanded regions of several control and elicited gels and illustrate differential protein accumulation at indicated time points. Spot number and time points are indicated in the inserts, for example, 2, spot number 2; 0H_C, zero hour time point control gel; 0H_E, zero time point elicited gel.

on the alignment between the query sequence and template, frameshift errors were detected and corrected if any. All possible protein sequences were annotated, given pseudo GI numbers, formatted similar to NCBInr to allow queries by MASCOT, and compiled as a plant protein database. The database also included M. truncatula chloroplast sequences (MtChI v.1, 156 records) and Arabidopsis mitochondrial proteins (AtMit v.1, 46 records). Proteins with at least 2 matched peptides and MOWSE scores greater than 2 times the generally accepted significant threshold (determined at 95% confidence level as calculated by MASCOT) were reported in this study and exceed the minimum information about a proteomics experiment (MIAPE).23 2.6. Determination of Lignin Content and Composition. Cell wall lignin content was determined using an acetyl bromide method24 and its composition by thioacidolysis.25,26 Thioacidolysis was performed by reacting dry samples (10 mg) with 3 mL of 0.2 M BF3 etherate in an 8.75:1 dioxane/ ethanethiol mixture. Lignin-derived monomers were identified and quantified by gas chromatography mass spectrometry 6222

Journal of Proteome Research • Vol. 9, No. 12, 2010

(GC-MS) analysis of their trimethylsilyl derivatives.27 GC-MS was performed using a Hewlett-Packard 5890 series II gas chromatograph with a 5971 series mass selective detector (column: HP-1; 60 m × 0.25 mm; 0.25-µm film thickness), and mass spectra were recorded in electron ionization mode (70 eV) with a 60-650 m/z scan range. 2.7. Determination of Wall-Bound Phenolics. The quantification of wall-bound phenolics was performed as previously described.28 Briefly, cells were sequentially extracted with acetone, 80% methanol, methanol, and chloroform to remove cell contents and yield a cell wall preparation that was free of extractable solutes. The extract-free cell wall samples were then dried, weighed and hydrolyzed overnight at room temperature in the dark using 1 N NaOH (1.6 mL). The aqueous supernatant was collected by centrifugation and the cell wall samples hydrolyzed again with 1.6 mL of 1N NaOH for an additional 6 h and then washed with water (1.6 mL). The combined aqueous extracts were acidified to pH 1.2 with 6 N HCl and extracted three times with an equal volume of ethyl acetate. The combined organic phases were dried under nitrogen and

Isoflavonoid Biosynthesis and Cell Wall Modifications

research articles

resuspended in 70% methanol for HPLC analysis. HPLC was carried out on a Beckman System Gold HPLC system consisting of a programmable solvent module 126, a System Gold 508 autosampler and a System Gold 168 diode array detector. A Waters Spherisorb ODS-2 reverse phase column (5 µm particle, 250 × 4.6 mm) was used. Solvent: A, 0.1% phosphoric acid in water; B, acetonitrile. Gradient: 5% B to 25% B in 50 min. The wall-bound phenolics were identified by comparing their UV spectra and retention times with those of the authentic compounds and quantified by using an external standard method.

3. Results and Discussion Purified yeast elicitor (YE), consisting of yeast cell wall polysaccharides, is widely used to study the responses of plant cell cultures to external pathogen attacks.29-31 YE is an inducer of multiple natural product pathways including benzophenanthridine alkaloids,21,32 indole alkaloids,33 isoflavonoids,34 and prenylated isoflavonoids32 in various species. Previously, we reported that YE strongly induces M. truncatula isoflavonoid biosynthesis and related primary metabolism.30,35 More specifically, YE induced phenylalanine ammonia lyase and chalcone synthase expression while also increasing the accumulation of cinnamic acid, benzoic acid and isoflavone-derived compounds in M. truncatula cell cultures treated with yeast elicitor.8 The purpose of this study was to evaluate the temporal proteome response of M. truncatula liquid suspension cell cultures to advance our understanding of the proteome responses of legumes to external stress and to integrate these findings with parallel transcriptomics and metabolomics data. 3.1. Quantitative Analysis of Reproducibility. Comparative proteomics was performed using M. truncatula suspension cell cultures to identify proteins differentially accumulated in response to yeast elicitation. The suspension cells were generated from callus originally derived from M. truncatula root as described in the Materials and Methods. The proteome was evaluated in triplicate at 0, 12, 24 and 48 h after exposure to YE using a 2-DE/MS-based proteomics approach. The total 2-DE protein loading was 1500 µg and the 2-DE protein profiles typically revealed about 900 protein spots as illustrated in Figure 1. The majority of protein spots were observed in the pI range from 4 to 7 with molecular weights from 14 to 110 kD. Reproducibility of 2-DE gels is often a concern due to inherent gel-to-gel variations which affect the protein spot positions and volume (i.e., area × intensity). Thus, statistical analyses of both the number of spots visualized and the normalized spot volumes were performed using triplicate gels generated from both the control and elicited cells collected at the 0 h time point to assess the 2-DE reproducibility in this work. For the triplicate control gels, the average spot number per gel was determined to be 875 with a coefficient of variance (CV and defined as a ratio of standard deviation to the mean) of 6.25%. This CV represents the total variance from biological and analytical sources as the replicates were also independent biological replicates. For the triplicate elicited gels acquired at the 0 h time point, the average spot number was found to be 923 and the CV 4.14%. The data illustrate good reproducibility and low CV for the number of spots visualized. ANOVA (analysis of variance) revealed no significant difference between the number of spots visualized on the control and the elicited gels acquired at the 0 h time point (p value: 0.23).

Figure 2. Distribution of CVs for normalized volumes of spots visualized in (A) triplicate control gels and (B) triplicate elicited gels at the 0 h time point.

Variance in the normalized spot volumes generated by 2D Expression (Nonlinear Dynamics) was also used to quantify gel reproducibility. For the 0 h control gels, the CV for the normalized volumes of spots matched across the triplicate control gels (810 spots) ranged from 0 to 164% and the distributions of CVs for spot normalized volume are shown in Figure 2A. The triplicate elicited gels acquired at 0 h time point were found to have similar CV distribution pattern for spot normalized volume as shown in Figure 2B. The average CVs for normalized spot volume were determined as 48% for the triplicate control gels and 51% for the triplicate elicited gels (0 h time point). The CVs represent cumulative biological and analytical variances and were found to be lower than that (58.6%) reported by Jorge et al.36 and slightly higher than observed by Asirvatham et al.37 at 24.2%. 3.2. Comparative Proteomics Reveals Substantial Alterations in Phenylpropanoid Metabolism. Comparative 2-DE protein analyses were performed using 2D Expression software (Nonlinear Dynamics). Although 2-DE analysis software has improved greatly over the past years, the complexity of the proteome as well as gel-to-gel variations still mandate manual interventions in 2-DE analysis to obtain confident results.38,39 For this reason, manual inspection and validation of differential protein accumulations were employed in this work to correct and minimize any mismatch, possible false-positive and/or false-negative results. Normalized volume of each individual protein spot was used in ANOVA analyses and fold calculations. Journal of Proteome Research • Vol. 9, No. 12, 2010 6223

research articles

Lei et al.

Table 1. List of Differentially Accumulated Proteins at 0, 12, 24, and 48 h after Yeast Treatment spot

fold

p-value

ID

MW

pI

1 2

2.2 1.9

0.03 0.005

Differentially accumulated proteins at 0 h eukaryotic protein synthesis initiation factor 5A-2 (eIF-5A 2) 496 catalase 276

TC76568 TC78044

17285 40052

3 4 5 5

2.1 1.5 n/a n/a

0.0041 0.05 n/a n/a

Differentially accumulated proteins at 12 h unknown protein 219 chalcone-flavonone isomerase 1 661 putative laccase 384 proline-rich protein precursor-kidney bean 135

TC88705 TC85633 TC89065 TC85413

6 5 5 7 7 8 9 10 11 12 12 13 13 14 14 15 16 4 17 18 18 19 20 21 22 23 24 24 24 25 25 26 26 26

4.0 3.0 3.0 3.7 3.7 2.6 2.3 2.2 2.1 2.0 2.0 1.8 1.8 1.7 1.7 1.6 1.5 1.5 0.25 0.27 0.27 0.33 0.41 0.5 0.5 0.54 0.52 0.58 0.58 0.64 0.64 0.65 0.65 0.65

0.01 0.03 0.03 0.05 0.05 0.01 0.02 0.02 0.001 0.03 0.03 0.02 0.01 0.05 0.05 0.01 0.03 0.04 0.03 0.02 0.02 0.0008 0.002 0.03 0.03 0.01 0.03 0.03 0.03 0.04 0.04 0.04 0.04 0.04

Differentially accumulated proteins at 24 peroxidase precursor putative laccase proline-rich protein precursor-kidney bean putative laccase beta-cyanoalanine synthase isoflavone reductase-like NAD(P)H-dependent oxidoreductase unnamed protein fructokinase malate dehydrogenase ferredoxin-NADP+ reductase isoflavone reductase-like NAD(P)H-dependent oxidoreductase cytosolic phosphoglycerate kinase alcohol dehydrogenase 1 12-oxophytodienoic acid 10 11-reductase probable dTDP-glucose 4-6-dehydratase dnaK-type molecular chaperone PHSP1 precursor vestitone reductase-alfalfa chalcone-flavonone isomerase 1 probable DNA-binding protein GBP16 auxin-induced protein reversibly glycosylated polypeptide probable 26S proteasome non-ATPase chain S5a probable chaperonin-containing TCP-1 complex gamma chain mitochondrial processing peptidase beta subunit superoxide dismutase (Mn) precursor thioredoxin peroxidase methylenetetrahydrofolate dehydrogenase 1,3-beta-glucanase probable cinnamyl-alcohol dehydrogenase proteasome subunit alpha type 7 60S ribosomal protein L6 inorganic pyrophosphatase expressed protein protein serine carboxypeptidase T18K17.3

5 5 27 28 29

6 6 1.6 1.7 1.6

0.01 0.01 0.04 0.03 0.03

L-ascorbate

score

a

TC

pep

cov

5.41 7.01

9 5

34% 24%

33572 23839 63633 40690

4.45 5.37 6.49 9.54

5 12 5 2

17% 35% 12% 9%

TC85274 TC89065 TC85413 TC89065 TC86072 TC86142 TC87386 TC76829 TC76700 TC86307 TC86142 TC85736 TC85844 TC85810 TC76481 TC85716 TC77308 TC85633 TC86829 TC85705 TC180426 TC93216 TC89333 TC76887 TC77316 TC85924 TC77362 TC85874 TC85435 TC86302 TC77079 TC77951 TC78009 TC77067

24985 63633 40690 63633 47833 34341 33830 33060 38627 44970 34321 42453 41097 27429 43688 74806 36001 23839 37450 37664 37402 44760 60423 68063 26284 20305 31491 40013 32651 28160 38532 32528 28304 28062

9.27 6.49 9.54 6.49 8.99 6.20 4.74 4.97 6.73 8.68 6.20 5.73 5.98 5.63 5.64 6.12 5.42 5.37 5.98 5.60 5.77 4.74 5.92 5.06 7.90 5.51 7.06 6.59 7.78 7.74 9.80 5.83 5.09 6.71

11 5 2 4 4 13 12 13 10 10 9 14 11 10 9 19 20 12 7 14 7 4 11 25 10 14 17 11 8 10 5 14 9 5

25% 12% 9% 11% 10% 44% 33% 37% 32% 26% 40% 40% 30% 48% 27% 36% 49% 35% 39% 25% 28% 20% 19% 55% 38% 54% 53% 39% 22% 25% 15% 36% 42% 12%

TC89065 TC85413 TC85681 TC76900 TC76900

63633 40690 58605 50894 50894

6.49 9.54 7.63 8.26 8.26

5 2 11 18 19

12% 9% 33% 47% 49%

h 723 384 135 192 149 766 550 752 611 437 430 682 552 566 510 1340 1149 820 396 980 329 290 590 1454 598 813 1036 591 403 656 338 645 512 196

Differentially accumulated proteins at 48 h putative laccase 384 proline-rich protein precursor-kidney bean 135 aldehyde dehydrogenase (EC 1.2.1.3) 2A precursor 600 acetyl-CoA acyltransferase 1133 acetyl-CoA acyltransferase 1148

a Fold: fold change expressed as the ratio of the average of normalized volumes of three elicited replicates to that of the control. ID: identification; Score: Mascot MOWSE score; TC: tentative consensus; MW: calculated molecular weight; pI: calculated isoelectrical point; “Pep” denotes the number of matched peptides and “Cov” denotes the sequence coverage of the matched peptides. n/a (non-applicable) denotes that the protein was absent in the control gels and therefore the fold change cannot be calculated.

Using a 1.5 fold change with p value