Targeted proteomics using selected reaction monitoring reveals the induction of specific terpene synthases in a multi-level study of methyl jasmonate-treated Norway spruce (Picea abies)

Authors


*For correspondence (fax +01 604 822 2114; e-mail bohlmann@msl.ubc.ca).

Summary

Induction of terpene synthase (TPS) gene expression and enzyme activity is known to occur in response to various chemical and biological stimuli in several species of spruce (genus Picea). However, high sequence identity between TPS family members has made it difficult to determine the induction patterns of individual TPS at the protein and transcript levels and whether specific TPS enzymes respond differentially to treatment. In the present study we used a multi-level approach to measure the induction and activity of TPS enzymes in protein extracts of Norway spruce (Picea abies) bark tissue following treatment with methyl jasmonate (MeJA). Measurements were made on the transcript, protein, enzyme activity and metabolite levels. Using a relatively new proteomics application, selective reaction monitoring (SRM), it was possible to differentiate and quantitatively measure the abundance of several known TPS proteins and three 1-deoxy-d-xylulose 5-phosphate synthase (DXS) isoforms in Norway spruce. Protein levels of individual TPS and DXS enzymes were differentially induced upon MeJA treatment and good correlation was generally observed between induction of transcripts, proteins, and enzyme activities. Most of the mono- and diterpenoid metabolites accumulated with similar temporal patterns of induction as part of the coordinated multi-compound chemical defense response. Protein and enzyme activity levels of the monoTPS (+)-3-carene synthase and the corresponding accumulation of (+)-3-carene was induced to a higher fold change than any other TPS or metabolite measured, indicating an important role in the induced terpenoid defense response in Norway spruce.

Introduction

Conifers are long-lived organisms that must defend themselves against a myriad of insects, pathogens and herbivores. Both the chemical and the physical defenses of conifers depend in a large part on the production of terpenoid-rich oleoresin in specialized anatomical structures (Bohlmann, 2008). Oleoresin is composed of three structurally diverse classes of terpenoid compounds, monoterpenes (C10), sesquiterpenes (C15) and diterpenoids (C20), which are synthesized by their respective mono-, sesqui- and di-terpene synthase (TPS). In Norway spruce (Picea abies), approximately 95% of oleoresin is composed of monoterpenes and diterpenoids in approximately equal proportions (Martin et al., 2002). Oleoresin is stored in specialized constitutive resin ducts (CRD) found in the cortex tissue and in xylem traumatic resin ducts (TRD), which are formed de novo upon insect attack or treatment with the defense hormone methyl jasmonate (MeJA) (Franceschi et al., 2002; Martin et al., 2002). Treatment of Norway spruce with MeJA has been shown to be a good mimic of insect attack and has led to new knowledge about the biochemical and molecular underpinnings of induced terpenoid defense (Miller et al., 2005). Along with TRD formation, MeJA induces TPS gene expression, enzyme activity and terpenoid accumulation in both bark and xylem tissues of Sitka spruce (Picea sitchensis) and Norway spruce (Martin et al., 2002; Fäldt et al., 2003; Miller et al., 2005).

The prenyl diphosphate precursors to terpenoid biosynthesis are formed by the condensation of isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP), which originate from the methyl erythritol 4-phosphate (MEP) and mevalonate (MVA) pathways (Figure 1), although the MEP pathway is the principal route for defense-related terpenoid biosynthesis (Lange and Ghassemian, 2003). Geranyl diphosphate (GPP), farnesyl diphosphate (FPP) and geranylgeranyl diphosphate (GGPP) are the substrates for mono-, sesqui- and diTPS, respectively (Figure 1). The TPSs of conifers are encoded by large gene families with high nucleotide and amino acid identity within the conifer specific TPS-d family (Martin et al., 2004). However, the specific biochemical functions of individual TPS family members cannot be predicted based on sequence similarity alone, as changes in only a few amino acids can lead to drastic changes in the terpenoid profile of a given TPS enzyme (Keeling et al., 2008). This has made it difficult to determine protein and transcript abundance of individual TPSs using liquid chromatography-tandem mass spectrometry (LC-MS/MS), western blotting or hybridization-based techniques such as northern blotting due to cross-reactivity and cross-hybridization. In addition, many TPSs are multi-product enzymes and can often give rise to mixtures of the same compounds in differing proportions (Martin et al., 2004). The complexity of the TPS-d family of multi-product TPS enzymes has made it difficult to study, in a targeted fashion, the roles played by specific TPS transcripts or proteins as part of the overall oleoresin terpenoid defense response. The TPS proteins are also present in very low abundance compared with many other proteins and are typically not observed using standard protein profiling methods (Lippert et al., 2007, 2009).

Figure 1.

 Biochemical pathways of terpenoid biosynthesis in Norway spruce (Picea abies).
The methyl erythritol 4-phosphate (MEP) and mevalonate (MVA) pathways supply isoprenoid precursors [isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP)] for mono-/diterpene and sesquiterpene biosynthesis, respectively. Prenyltransferases condense IPP and DMAPP to produce geranyl diphosphate (GPP), farnesyl diphosphate (FPP) or geranylgeranyl diphosphate (GGPP), respectively. Mono-, sesqui- and diterpene synthases (TPSs) form the respective terpenoid compounds. Percentages indicate the per cent total of the specific terpenoids that the TPS enzyme produces (Martin et al., 2004). Percentages below 1% are not shown. The olefin products of the diTPS PaTPS-LAS and PaTPS-Iso are rapidly oxidized to the corresponding diterpenoid resin acids (not shown), which accumulate in the oleoresin. DXS, 1-deoxy-d-xylulose 5-phosphate synthase; IPPI, isopentenyl diphosphate isomerase.

In this study we have used selected reaction monitoring (SRM) (Lange et al., 2008) to specifically detect and quantitatively measure several individual members of the TPS gene family in the MeJA-induced defense response of Norway spruce bark tissue. To produce an integrative study of Norway spruce terpenoid biosynthesis in response to MeJA, we have also used quantitative real-time (qRT) PCR for transcript measurement, gas chromatography-tandem mass spectrometry (GC-MS/MS) and gas chromatography-flame ionization detection (GC-FID) for metabolite measurement and ex vivo measurements of TPS enzyme activity in protein extracts. In addition, we have used the same integrated approach to measure, in a transcript- and protein-specific manner, induction of the three isoforms of 1-deoxy-d-xylulose 5-phosphate synthase (DXS), which catalyze the first step in the MEP pathway (Phillips et al., 2007). This study provides a multi-level analysis of the defense-related induction of transcripts, proteins, enzyme activity and metabolite products of TPS and upstream DXS proteins involved in defense-related oleoresin formation in spruce with fine temporal resolution and using highly specific approaches.

Results

Differential induction of TPS proteins and three DXS isoforms in response to MeJA treatment

Protein profiling by SRM was performed to determine the differential induction of individual TPS and MEP pathway enzymes in response to MeJA treatment over time. A total of 19 peptides representing 13 known TPS and three DXS proteins from Norway spruce were designed for SRM analysis of MeJA-treated and control Norway spruce bark tissue (Table S1 in Supporting Information). The 13 TPS and three DXS targets include PaTPS-Lin, PaTPS-Pin, PaTPS-Car, PaTPS-Lim, PaTPS-Myr, PaTPS-Bis, PaTPS-Lon, PaTPS-Far, PaTPS Iso, PaTPS-LAS, two monoTPS-like, one sesquiTPS-like, PaDXS1, PaDXS2a and PaDXS2b (Fäldt et al., 2003; Martin et al., 2004; Phillips et al., 2007). Using these 19 peptides, we detected five TPS (as shown in Figure 1) and all three DXS isoforms by SRM. Two peptides from each DXS protein were measured and the results for both peptides did not substantially differ. Of the TPSs detected, three were monoTPSs (PaTPS-Pin, PaTPS-Car, PaTPS-Lim) and two were diTPSs (PaTPS-LAS, PaTPS-Iso); however, no sesquiTPSs were detected. Levels of all detected TPS and DXS proteins were induced upon MeJA treatment over the 32-day time course (Figure 2), with the exception of PaDXS1 where protein levels in MeJA-treated tissue were similar to controls (Figure 2f). Levels of PaTPS-Car increased rapidly at 2 days, peaked at 8 days in MeJA-treated trees (Figure 2a) and declined thereafter. Notably, the average fold change protein abundance over the time course of PaTPS-Car in MeJA-treated bark compared with control is greater than any of the other proteins measured by SRM (Table 1). PaTPS-Lim displayed a different induction profile in MeJA-treated trees, with an increase from control levels at 4 days and a continuous increase in abundance over the time course of 32 days (Figure 2b). Levels of PaTPS-Pin increased rapidly compared with controls 4 days after treatment with MeJA, peaking at 8 days and decreasing thereafter (Figure 2c). The two diTPSs measured also displayed differential induction patterns (Figure 2d,e). Levels of PaTPS-Iso protein increased 4 days after treatment and peaked at 16 days. PaTPS-LAS increased more rapidly, and peaked 8 days after treatment. The difference in induction patterns of these proteins that are >90% identical highlights the ability of SRM to discriminate among closely related proteins. Of the three DXS proteins measured, levels of PaDXS2a and PaDXS2b both increased over the time course, with a peak abundance between 4 and 8 days (Figure 2g,h). However, induction of PaDXS1 was less pronounced for MeJA-treated in comparison to control bark tissue (Figure 2f).

Figure 2.

 Protein levels of terpene synthase (TPS) and 1-deoxy-d-xylulose 5-phosphate synthase (DXS) enzymes as detected by selective reaction monitoring (SRM) from a time course of methyl jasmonate (MeJA)-treated and control Norway spruce (Picea abies) bark protein extracts.
The results are the peak area ratio of the endogenous peptide (light) to the isotopically labeled internal standard (heavy). Data are presented are the mean of four biological replicates for each time point.

Table 1.   Fold changes of transcripts, proteins and metabolites measured in methyl jasmonate (MeJA)-treated Norway spruce (Picea abies) bark tissue relative to control levels
Days after treatment2 days4 days8 days16 days32 days
  1. SRM, selective reaction monitoring.

Transcript profiling
 PaTPS-Car34.3616.2711.5837.6515.81
 PaTPS-Lim43.9745.14298.3270.58124.82
 PaTPS-Pin10.896.4032.9812.053.54
 PaTPS-LAS0.290.620.440.580.80
 PaTPS-Iso16.508.0717.8813.204.19
 PaDXS19.666.743.642.411.03
 PaDXS2A10.748.8725.865.602.51
 PaDXS2B14.2915.4137.4928.123.64
Protein profiling (SRM)
 PaTPS-Car99.1921.1536.1013.7039.54
 PaTPS-Lim0.002.534.506.9232.51
 PaTPS-Pin2.602.244.784.312.62
 PaTPS-LAS1.392.305.204.161.81
 PaTPS-Iso0.801.803.493.302.05
 PaDXS12.971.601.531.161.65
 PaDXS2A2.312.104.244.633.11
 PaDXS2B0.003.264.603.553.03
Enzyme assays
 (+)-3-Carene21.1561.5651.7769.41100.05
 (−)-Limonene1.782.925.072.003.38
 (−)-β-Phellandrene1.391.031.050.800.86
 (−)-α-Pinene1.131.401.921.371.57
 (−)-β-Pinene0.751.151.631.331.76
 Myrcene3.392.261.541.331.30
 (−)-Sabinene12.668.897.804.784.62
 Terpinolene3.743.945.053.593.56
 Levopimaradiene0.022.0912.561.290.00
 Abietadiene10.625.9118.3022.890.58
 Neoabietadiene0.146.1114.9225.224.62
 Isopimaradiene0.312.172.742.360.82
Metabolite profiling
 (+)-3-Carene0.831.264.575.012.05
 (−)-Limonene1.071.782.873.042.86
 (−)-β-Phellandrene0.991.361.821.961.05
 (−)-α-Pinene0.861.261.721.981.14
 (−)-β-Pinene0.861.381.742.021.14
 Myrcene0.951.491.882.061.18
 (−)-Sabinene0.811.662.032.251.23
 Terpinolene0.851.111.081.091.04
 Levopimaric acid0.771.663.363.711.94
 Palustric acid0.761.822.152.361.51
 Abietic acid0.661.522.572.591.65
 Neoabietic acid0.801.312.192.701.43
 Isopimaric acid0.741.351.762.071.20

Mono- and diTPS enzyme assays

Mono- and diTPS enzyme assays were performed to determine if a MeJA-induced increase in protein abundance results in a corresponding increase in enzyme activity. Protein extracts were assayed with the appropriate substrate (GPP for monoTPS assays and GGPP for diTPS assays), the terpenoid products were analyzed using GC-MS, then identified and quantified using authentic standards. All of the monoTPSs detected by SRM are multi-product enzymes, some of which have partially overlapping product profiles (Martin et al., 2004). Although this makes it difficult to directly correlate the formation of terpene products with the enzyme activity of a single protein, some relevant comparisons can be made.

The induction pattern of the formation of (+)-3-carene in monoTPS enzyme assays (Figure 3a) is similar to that of PaTPS-Car protein abundance (Figure 2a), which peaks 8 days after treatment with MeJA. (+)-3-Carene is the major product of PaTPS-Car (Fäldt et al., 2003) and is not known to be produced by any other monoTPS in Norway spruce, with the exception of PaTPS-Lin (linalool synthase) where (+)-3-carene is 0.2% of the total product profile (Martin et al., 2004). PaTPS-Lin was not detected in this study using SRM, which is consistent with previous studies showing that (−)-linalool is induced in needles to a greater degree than in stems of Norway spruce upon MeJA treatment (Martin et al., 2003). (−)-Sabinene is also a product of PaTPS-Car, and the enzyme activity profiles of (−)-sabinene formation and (+)-3-carene formation are similar 8–32 days after treatment. However (−)-sabinene peaks at 2 days after treatment while (+)-3-carene peaks at 8 days (Figure 3a,g). These differences may either reflect additional enzymatic activity responsible for (−)-sabinene production or a differential synthesis of two products by the same enzyme in response to MeJA treatment. Similar to the SRM results, the fold change of (+)-3-carene as measured by enzyme assays was the highest of all terpenoids measured (Table 1).

Figure 3.

 Monoterpene synthase enzyme assay products as detected by gas chromatography-mass spectrometry (GC-MS) from a time course of methyl jasmonate (MeJA)-treated and control Norway spruce (Picea abies) bark protein extracts.
Data are presented as means of four biological replicates at each time point.

Protein abundance of PaTPS-Lim and enzyme assay profiles of (−)-limonene formation are also similar (Figures 2b and 3b); however, (−)-limonene is also synthesized by PaTPS-Pin at >1% of the product profile (Figure 1; Martin et al., 2004). The enzyme activity profiles for the formation of (−)-α-pinene and (−)-β-pinene, which are the two major products of PaTPS-Pin, are very similar; however, the levels of these two monoTPS enzyme assay products peak at 8 days after treatment (Figure 3c,d) and the protein abundance profile of PaTPS-Pin peaks at 8 days (Figure 2c). Myrcene was the only monoTPS enzyme assay product to rapidly decrease in abundance 2–16 days after its initial increase at 2 days after treatment (Figure 3f). PaTPS-Myr produces myrcene as a sole product, and it was not detected using SRM. This is consistent with the low abundance and decrease of myrcene formation as measured in enzyme assays (Figure 3f). However, myrcene is also a minor product of both PaTPS-Car and PaTPS-Lim (Figure 1).

All the enzymatic products of both diTPSs, PaTPS-LAS and PaTPS-Iso (Figure 1), were detected in the metabolite analysis of diTPS enzyme assays, with the exception of palustradiene which was present in the enzyme assay products in very low amounts and was difficult to chromatographically resolve above background. This result is not surprising, since palustradiene is the least abundant of the four products of PaTPS-LAS (Figure 1; Martin et al., 2004), and was shown to be present at low levels relative to the other products of PaTPS-LAS and PaTPS-Iso in untreated Norway spruce bark tissue and decreased in abundance upon MeJA treatment (Martin et al., 2002). The three other products of PaTPS-LAS (Figure 1; levopimaradiene, abietadiene and neoabietadiene) all showed similar induction patterns in tissue extract diTPS enzyme assays, with a peak 8–16 days after MeJA treatment (Figure 4a,b,c). Isopimaradiene, the sole product of PaTPS-Iso (Figure 1), increased sharply after 8 days, peaked at 16 days and declined thereafter (Figure 4d). These patterns match well with protein abundance data for both PaTPS-LAS and PaTPS-Iso (Figure 2d,e).

Figure 4.

 Diterpene synthase enzyme assay products as detected by gas chromatography-mass spectrometry (GC-MS) from a time course of methyl jasmonate (MeJA)-treated and control Norway spruce (Picea abies) bark protein extracts.
Data are presented as means of four biological replicates at each time point.

Metabolite profiling of monoterpenes and diterpene resin acids in MeJA-treated bark

For the next level of a systems analysis of induced terpenoid biosynthesis of the Norway spruce oleoresin defense, we performed metabolite profiling targeting monoterpenes and diterpene resin acids to determine how these metabolite pools change in response to MeJA treatment over time and how these changes compare with protein abundance and enzyme activity measurements. While diterpene olefins (i.e. levopimaradiene, abietadiene, neoabietadiene, palustradiene and isopimaradiene) are the immediate products of the diTPS enzymes PaTPS-LAS and PaTPS-Iso, these metabolites are efficiently converted into the corresponding diterpene resin acids by the activity of cytochrome P450 enzymes (Ro et al., 2005). The diterpene resin acids then accumulate in the stem tissues (Martin et al., 2002). Most monoterpenes and diterpenoid resin acids peaked in abundance 16 days after MeJA treatment (Figures 5 and 6). This result is noteworthy, since protein abundance and enzymatic activity of the corresponding mono- and diTPSs showed a more pronounced pattern of differential response. However, there was some variation in the accumulation profiles for individual terpenoid compounds. The most notable were those for (+)-3-carene and (−)-limonene, where the highest abundances were observed at 8 and 32 days, respectively. Similar to protein abundance and products formed in enzyme assays, (+)-3-carene had the highest fold change among any other terpenoid measured (Table 1). Despite the requirement for additional P450 activity to convert the olefin products of PaTPS-LAS and PaTPS-Iso to the corresponding diterpenoid resin acids, the levels of the diterpenoid resin acids in MeJA-treated trees (Figure 6) correlated well with protein abundance (Figure 2d,e) and enzyme activities (Figure 4) for PaTPS-LAS and PaTPS-Iso.

Figure 5.

 Monoterpene metabolite levels as detected by gas chromatography-mass spectrometry (GC-MS) from a time course of methyl jasmonate (MeJA)-treated and control Norway spruce (Picea abies) bark.
Data are presented as means of four biological replicates at each time point.

Figure 6.

 Diterpenoid resin acid metabolite levels as detected by gas chromatography-flame ionization detection (GC-FID) from a time course of methyl jasmonate (MeJA)-treated and control Norway spruce (Picea abies) bark.
Data are presented as means of four biological replicates at each time point.

Transcript profiling of proteins quantified by SRM

In our analysis upstream of the proteome, transcript levels for all proteins detected by SRM were measured to determine the relationship between transcript and protein abundance in response to MeJA treatment. In general, profiles of MeJA-induced transcript accumulation (Figure 7) show good correlation with protein abundance in MeJA-treated trees (Figure 2). Transcripts corresponding to PaTPS-Car exhibited peak abundance 2 days after treatment (Figure 7a) and declined thereafter, correlating well with the rapid induction of the corresponding protein detected by SRM (Figure 2a). Transcript levels of PaTPS-Lim peaked 8 days after treatment (Figure 7b); however, protein levels peaked 32 days after treatment (Figure 2b). PaTPS-Pin transcript profiles (Figure 7c) agreed well with protein levels (Figure 2c), both showing a peak of abundance 8 days after MeJA treatment. Transcript levels corresponding to PaTPS-LAS and PaTPS-Iso also agreed well with protein levels; however, transcript levels of PaTPS-Iso peak at 16 days and protein levels peak at 8 days after MeJA treatment (Figures 2d,e and 7d,e). Transcript profiling results corresponding to the three DXS isoforms agree well with protein abundance data (Figure 2f–h), showing an increased level of transcript for PaDXS2a and PaDXS2b, which remained up-regulated for at least 16 days after MeJA treatment (Figure 7g,h). In contrast, MeJA treatment only caused an initial spike of PaDXS1 transcript accumulation detected at 2 days, followed by a rapid decline (Figure 7f). Our transcript profiling data for PaDXS1, PaDXS2a and PaDXS2b also agree with previously published work investigating the effects of MeJA treatment on DXS transcript levels in Norway spruce (Phillips et al., 2007).

Figure 7.

 Transcript levels of monoterpene and diterpenoid synthases and 1-deoxy-d-xylulose 5-phosphate synthase (DXS) isoforms using quantitative real-time PCR.
Expression levels are relative to the endogenous control β-tubulin. Data are presented as means of four biological replicates and two technical replicates at each time point.

Discussion

Integrated analysis of terpenoid biosynthesis in spruce

Terpenoid synthases of the TPSd family (Keeling and Bohlmann, 2006) are responsible for much of the chemical diversity found in conifer oleoresin and are regulated in part at the transcriptional level in response to various treatments such as wounding, insect feeding and MeJA treatment; with a corresponding general induction of enzyme activity (Martin et al., 2002, 2004; Byun McKay et al., 2003, 2006; Fäldt et al., 2003). However, measuring changes in abundance of specific TPS proteins by western blotting or LC-MS/MS has not previously been possible due to high amino acid sequence identity within the TPSd family (Martin et al., 2004). Furthermore, TPS products from enzyme assays of tissue protein extracts do not necessarily reflect the activity of individual TPS enzymes, since many TPSs are multi-product enzymes with partly overlapping product profiles. Also, previously reported transcript profiling of TPS in Sitka spruce using hybridization-based methods such as northern blotting suffered from cross-hybridization of highly similar probes (Miller et al., 2005; Byun McKay et al., 2006). In this work we describe a multi-level investigation of MeJA-induced defense-related terpenoid biosynthesis in Norway spruce using SRM in combination with qRT-PCR and GC-MS to overcome previous challenges in measuring protein and transcript levels of highly similar TPS proteins. To our knowledge, SRM has only been reported once before for use in a plant system to measure the abundance of sucrose phosphate synthase proteins in Arabidopsis thaliana (Lehmann et al., 2008). Measuring transcript, and especially protein levels of individual TPSs is important because the fine regulation of TPS enzymes is thought to be responsible for much of the complex and dynamic nature of defense-related terpenoid accumulation in the oleoresin secretion of spruce (Keeling and Bohlmann, 2006; Bohlmann, 2008). Also, it cannot be presumed that an increase in a transcript will result in a similar increase in protein abundance, enzyme activity and ultimately metabolite accumulation, as many regulatory mechanisms are active at each level. By examining the differential induction profiles of mono- and diTPS transcripts, proteins, enzyme activities and monoterpenes and diterpenoid resin acids in MeJA-treated and untreated trees, we can investigate multiple regulatory levels of defense-related terpenoid production.

To monitor the protein abundance of highly similar TPS proteins we used SRM to measure protein-specific peptide ions and fragments, which provide specificity in the quantitative measurement of TPS proteins. Although in the absence of a complete genome sequence for spruce one cannot completely exclude the possibility of the target peptides detecting more than one protein, we have substantially reduced this possibility by selecting peptides based on comparison with over 500 000 spruce expressed sequence tags (ESTs) and full length (FL) cDNAs and all publicly available plant gene databases. Our results show that distinct profiles of protein abundance could be measured for proteins with >90% amino acid identity such as PaTPS-LAS and PaTPS-Iso (Martin et al., 2004). This level of resolution would be difficult to achieve with antibodies in traditional western blot analysis. We also showed that SRM is an effective means of obtaining quantitative measurements of protein abundance in specialized (i.e. secondary) metabolism, where proteins within the TPSd family are of relatively low abundance such that they were previously not detected in spruce tissue or cell culture extracts by other means of proteome analysis such as two-dimensional electrophoresis (2DE)-PAGE or iTRAQ (Lippert et al., 2007, 2009). In fact, SRM is able to detect proteins in the mid to low attomole range against a complex background of more abundant proteins (Kuzyk et al., 2009). We were able to measure detectable levels of three monoTPSs and two diTPSs involved in oleoresin production in Norway spruce (Figure 2). However, no sesquiTPSs were detected, which is supported by previous results showing that MeJA has a weak effect on sesquiterpene accumulation and associated sesquiTPS and FPP synthase activity in Norway spruce (Martin et al., 2002).

MeJA induction of (+)-3-carene formation

PaTPS-Car exhibited the highest fold change at the protein and enzyme activity levels and was matched by some of the highest fold changes at the metabolite level (Table 1). (+)-3-Carene is a common constituent of conifer oleoresin, and in some conifer–insect interactions has been associated with resistance [i.e. lodgepole pine (Pinus contorta; Rocchini et al., 2000) and Scots pine (Pinus sylvestris; Passquier-Barre et al., 2001)]. Infection of wound sites of lodgepole pine with the bark beetle-associated blue-staining fungus Ceratocystis clavigera resulted in an increase of 3-carene (Croteau et al., 1987), and a similar pattern exists in Scots pine phloem infected with Leptographium wingfieldii (Fäldt et al., 2002). In Sitka spruce, (+)-3-carene synthase was shown to be transcriptionally induced using northern blot analysis as soon as 3 h after MeJA treatment, with a peak between 12 h and 2 days and a subsequent decline until 32 days (Miller et al., 2005). In the present study, PaTPS-Car was very rapidly induced at all levels measured (Figure 8).

Figure 8.

 Schematic summary of the integrated analysis of methyl jasmonate (MeJA)-induced monoterpene (a) and diterpenoid (b) biosynthesis in Norway spruce (Picea abies).
The MeJA-induced changes of abundance of transcripts, proteins, products of enzyme assays and metabolite accumulation are shown with color-coded heat maps. The fold increase at each time point was normalized to the peak fold increase, which was set to 1 in each case. Detailed quantitative results are shown in Table 1 and Figures 2–7. Percentages indicate the per cent total of the terpenoids that the specific terpene synthase (TPS) enzyme produces (Martin et al., 2004). Percentages below 1% are not shown. DXS, 1-deoxy-d-xylulose 5-phosphate synthase; MEP, methyl erythritol 4-phosphate; MVA, mevalonate; IPPI, isopentenyl diphosphate isomerase; GPP, geranyl diphosphate; GGPP, geranylgeranyl diphosphate.

Induction of the formation of other monoterpenes

PaTPS-Pin is of known biological significance in the Norway spruce interaction with the bark beetle Ips typographus, which uses α-pinene for host selection (Renwick et al., 1976; Lanier et al., 1980). α-Pinene is detoxified by the insect in the hind gut to form (+)-trans-verbenol from (+)-α-pinene and (+)-cis-verbenol from (−)-α-pinene, with only the (+)-cis configuration being effective in attracting the female beetle. In previous studies using Sitka spruce, transcripts of (−)-α/β-pinene synthase were induced in a similar manner to (+)-3-carene synthase as detected by northern blot analysis (Miller et al., 2005). In the present study using SRM and qRT-PCR, differences in the induction profiles of protein and corresponding transcript abundance of PaTPS-Pin and PaTPS-Car were resolved (Figure 8). Enzyme activity assays showed a similar peak of accumulation for both (−)-α-pinene and (−)-β-pinene at 8 days; however, a full resolution of the molecular basis of this pattern of metabolite accumulation is complicated by the fact that both monoterpenes, (−)-α-pinene and (−)-β-pinene, are produced by several different monoTPSs (Figures 1 and 8).

Levels of PaTPS-Lim proteins and metabolites increased with a peak at 32 days after MeJA treatment, a pattern which is unique among all TPSs measured (Figure 8) and may indicate a special role for (−)-limonene in a sustained defense response. In previous studies, transcripts of (−)-limonene synthase were found to be induced in drill-wounded apical leaders of a genotype of Sitka spruce known to be resistant to white pine weevil attack (Byun McKay et al., 2006) and accumulation of (−)-limonene was induced four-fold in Norway spruce seedling bark 16 days after MeJA treatment (Martin et al., 2002).

MeJA-induced diterpenoid biosynthesis

The two Norway spruce diTPSs detected in this study, PaTPS-LAS and PaTPS-Iso, produce diterpene olefins (Martin et al., 2004; Keeling et al., 2008) which are subsequently oxidized by cytochrome P450 enzymes (Ro et al., 2005) to the corresponding diterpenoid resin acids that dominate the oleoresin composition (Martin et al., 2002). Northern blot analysis has shown that transcripts corresponding to PaTPS-LAS and PaTPS-Iso in Sitka spruce are induced with peaks at 3–12 h and 2–4 days after MeJA treatment (Miller et al., 2005). However, it is likely that the results generated using northern blot analysis are subject to cross-hybridization due to the high similarity between PaTPS-LAS and PaTPS-Iso (>90% amino acid identity). In the present work, we were able to design gene-specific primers and protein-specific peptides to dissect the unique induction patterns of PaTPS-LAS and PaTPS-Iso transcripts and proteins. The patterns of MeJA-induced transcript abundance of the two diTPSs agree well with protein abundance, enzyme activities and corresponding metabolite profiles (Figure 8).

MeJA-induced response of DXS proteins and transcripts

Relatively little is known about upstream enzymes in the MEP pathway for terpenoid biosynthesis in conifer defense. The MEP pathway is instrumental in providing the central C5 intermediates, IPP and DMAPP, involved in the biosynthesis of mono- and diterpenes (Lange and Ghassemian, 2003; Phillips et al., 2008). The first step in the MEP pathway is DXS, which condenses the (hydroxyethyl)thiamine derived from pyruvate with the C1 aldehyde group of glyceraldehyde 3-phosphate to produce 1-deoxy-D-xylulose 5-phosphate (DXP) (Sprenger et al., 1997; Lange et al., 1998; Lois et al., 1998). It has also been suggested that DXS plays an important regulatory role in the MEP pathway as a rate-limiting step, and higher DXS levels may be required for increased terpenoid production (Rodríguez-Concepción, 2006). There are two types of DXS enzymes in plants, type I is constitutively expressed in photosynthetic tissue and is probably involved in the biosynthesis of isoprenoids such as carotenoids and phytol in primary metabolism, while type II seems to be involved in the biosynthesis of isoprenoids for specialized (i.e. secondary) metabolism (Walter et al., 2002). Induction patterns of the DXS proteins measured in this study also suggest that PaDXS2a and PaDXS2b play an important role in induced terpenoid biosynthesis of the Norway spruce oleoresin defense, as protein levels of PaDXS2a and PaDXS2b are increased upon MeJA treatment compared with controls (Figure 2g,h). Unlike PaDXS2a and PaDXS2b, PaDXS1 protein levels in MeJA-treated tissue are more similar to controls (Figure 2f), consistent with a role for PaDXS1 in primary metabolism. It has been previously shown that transcripts corresponding to primary metabolic enzymes may be down-regulated in response to biotic stress (Ralph et al., 2006; Zulak et al., 2007). The profiles of differentially induced protein abundance for the three DXS isoforms agree with the corresponding transcript profiles of induction for PaDXS2a and PaDXS2b, as detected in this study and in previous work (Phillips et al., 2007).

Conclusions

In this study of the proteome, transcriptome, enzyme activities and metabolites of MeJA-induced terpenoid oleoresin biosynthesis in Norway spruce bark tissue, there was generally good agreement between transcript and protein abundance in specialized terpenoid metabolism (Figure 8). However, there were some discrepancies between protein abundance and enzyme activity, especially in the case of PaTPS-Lim (Figure 8). This could indicate post-translational regulatory mechanisms affecting enzymatic activity, differences in protein stability or simply overlapping metabolite profiles due to the multi-product function of most monoTPSs. Interestingly, temporal profiles of induced metabolite accumulation were nearly identical for all monoterpenes and diterpenoid resin acids studied (Figure 8), suggesting that additional levels of regulation, such as substrate availability for TPS or mechanisms of terpenoid transport (which are as yet unknown) into the lumen of resin ducts, may also contribute to the coordinated accumulation of many terpenoids with the same temporal profile. In summary, we have established a substantially refined picture of the multi-level induced response of terpenoid oleoresin biosynthesis in Norway spruce, a complex biological system characterized by an enormous diversity of defense-related terpenoid specialized metabolites.

Experimental procedures

Plant materials and treatment

Norway spruce seedlings of a clonal line (Picea abies L. Karst; clone 518473-07) provided by J. D. Irving, Ltd (http://www.jdirving.com/) and grown as in Martin et al. (2002). Each treated tree was sprayed with 20 ml of 0.1% solution (v/v) of MeJA [95% (w/w) pure, Sigma Aldrich (http://www.sigmaaldrich.com/)] dissolved in 0.1% Tween-20 (v/v). Control saplings were sprayed with a 0.1% (v/v) Tween-20 solution to account for solvent effects as reported (Miller et al., 2005). Control and MeJA-treated saplings were kept in separate growth chambers. Four independent biological replicates of control and treated trees were harvested at 2, 4, 8, 16 and 32 days following MeJA treatment and used for all experiments. Needles were stripped and bark was separated from xylem using a razor blade. All tissues were flash-frozen in liquid nitrogen and stored at −80°C until use.

Protein extraction for SRM analysis

Bark tissue (150 mg) was ground to a fine powder under liquid nitrogen with 100 mg sea sand and 50 mg polyvinylpolypyrrolidone (PVPP). Tissue was extracted twice using 2 ml of 10% tricarboxylic acid and 1% dithiothreitol (DTT) in acetone, twice using 2 ml of 0.1 m ammonium acetate and 1% DTT in 80% methanol, and twice using 2 ml of 1% DTT in 80% acetone. Tissue pellets were dried to remove residual acetone and extracted using 600 μl 2-amino-2-(hydroxymethyl)-1,3-propanediol (TRIS)-buffered phenol and 600 μl 30% sucrose, 2% SDS, 0.1 m TRIS (pH 8.0) and 2% DTT. The phenol phase was transferred to a fresh tube, 0.1 m ammonium acetate in methanol added and samples were incubated at −20°C overnight. Samples were then centrifuged for 5 min at 14 000 g at 4°C and washed with 100% methanol and then 80% acetone. Pellets were dried and subjected to SDS–PAGE analysis.

SDS–PAGE analysis

Protein pellets were resuspended in 0.3% (w/v) SDS (100 μl) and boiled for 5 min. Protein concentrations were determined using a Quick Start Bicinchonic Acid protein assay kit (Sigma). In preparation for SDS–PAGE, 20 μg of each sample was denatured by boiling for 5 min in 1.5 × lauryl dodecyl sulfate sample buffer (Invitrogen, http://www.invitrogen.com/) containing 50 mm DTT (Invitrogen). Samples were analyzed by SDS–PAGE using pre-cast, 1.5 mm, 4–12% polyacrylamide gradient Bis-TRIS NuPage gels (Invitrogen) with MES SDS running buffer according to the manufacturer’s protocol. Proteins were visualized by staining gels with Gel Code® Blue Safestain (Thermo Scientific, http://www.thermo.com/).

Selection, synthesis and purification of isotopically labeled tryptic peptides

Cloned Norway spruce TPS enzymes were expressed in Escherichia coli and analysed by tryptic digestion and tandem mass spectrometry [digestion and MS/MS were performed as described in Lippert et al.(2007)]. The peptide sequences selected for SRM analysis and used for the construction of internal peptide standards were chosen empirically from these data. Stable isotope-labeled (13C and 15N) standard (SIS) peptides were synthesized at a 5-μmol scale using Fmoc chemistry with a Protein Technologies Prelude peptide synthesizer (http://www.peptideinstruments.com/) as described previously (Bordeerat et al., 2009). Peptides were synthesized incorporating stable isotope-labeled amino acids, either [13C6]Lys, [13C6]Arg, [13C6][15N1]Leu, [13C6][15N1] Ile, [13C3][15N1]Ala, or [13C2][15N1]Gly, all purchased from Cambridge Isotope Laboratories (http://www.isotope.com/) with a 98% isotope enrichment. [13C6]Lys and [13C6]Arg were conjugated to TentaGel R resin by Rapp Polymere (http://www.rapp-polymere.com/). Subsequent amino acid residues (100 mm) were double-coupled using 20% piperidine as the deprotector and 1H-benzotriazolium 1-[bis(dimethylamino)methylene]-5-chloro-,hexafluorophosphate (1),3-oxide (HCTU) as the activator. The cleavage was performed with 95:2.5:2.5 trifluoroacetic acid (TFA):water:triisopropylsilane.

The cleaved peptides were removed from the synthesizer and the TFA was evaporated under a stream of nitrogen. Ether was added to precipitate the peptides and, after centrifugation at 3000 g for 5 min, the ether layer was decanted. Peptides were resolubilized in 0.1% TFA and purified by reversed-phase HPLC (Ultimate 3000, Dionex, http://www.dionex.com/) while monitoring the peptide elution at 230 nm. The crude peptides were separated using a Vydac C18 column (10 × 250 mm, 10 μm resin; Grace Davison Discovery Sciences, Deerfield, IL) with a linear gradient of 0.1% TFA in water (v/v) and 0.085% TFA in 50% acetonitrile (v/v) at a flow rate of 4 ml min−1 over 60 min. Fractions of interest were spotted onto stainless steel matrix assisted laser desorption/ionization (MALDI) plates and measured by MALDI-time-of-flight (TOF) mass spectrometry (Applied Biosystems/MDS SCIEX, http://www.appliedbiosystems.com and http://www.sciex.com/). Fractions containing >80% of the target peptide by MALDI-TOF analysis were pooled and lyophilized.

In-gel trypsin digestion with internal standard peptides

The SDS–PAGE gels were destained with deionized water (18 MΩ). Gel lanes representing each sample were manually excised and divided into 16 equivalent slices. Gel lane slices spanning the 30–220 kDa relative molecular weight range (slices 9–14, Figure S1) were digested separately in a microtiter plate format in a single automated sample preparation run. All digests were analyzed separately by LC-SRM/MS analysis using a Genomic Solutions ProGest (http://www.digilabglobal.com/) as previously described (Parker et al., 2005). This mass range was chosen because the target peptides had been shown empirically to be restricted to only these gel fragments (data not shown). Briefly, gel pieces were destained [50/45/5 (v/v) methanol/water/acetic acid] prior to reduction (10 mm dithiothreitol) and alkylation (100 mm iodoacetamide). The absolute quantification strategy of Kirkpatrick et al. (2005) was adapted to permit multiplexed quantification of the 16 target peptides in each gel slice (Kirkpatrick et al., 2005). The modified sequencing grade porcine trypsin solution (20 ng μl−1, Promega, http://www.promega.com/) contained a mixture of 19 SIS peptides representing 13 different TPS and 3 DXS proteins (Table S1) to permit accurate delivery of all SIS peptides to each gel slice (ranging in concentration from 0.25 to 5.0 pmol per gel slice) and enable relative quantification of TPS expression levels between samples. Proteins were then digested for 8 h at 37°C prior to collection of tryptic digests and acid extraction of the gel slices [50/40/10 (v/v) acetonitrile/water/formic acid]. Samples were then frozen at −80°C and lyophilized.

SRM Q1/Q3 ion pair selection by nano-infusion

Isotopically labeled peptides were diluted to 1 pmol μl−1 (1 mm) in 30% acetonitrile, 0.1% formic acid for infusion at a flow rate of 300 nl min−1 using a Harvard PicoPlus 11 syringe pump (Harvard Apparatus, http://www.harvardapparatus.com/). Infused peptide solutions were analyzed by nano-electrospray using a 4000 QTRAP hybrid triple quadrupole/linear ion trap MS (Applied Biosystems/MDS SCIEX) equipped with a nanospray ionization source. The MS analysis was conducted in the positive ion mode with ion spray voltages in the 1800–2000 V range. The declustering potential was ramped (0–120 V in 2-V increments) during Q1 scans centered on 10-Da wide mass ranges. The SRM scans for optimization of SRM Q1/Q3 ion pairs were conducted with both Q1 and Q3 set to unit resolution [0.6–0.8 Da full width at half-height (FWHH)] while the collision energy was ramped (5–120 V in 2-V increments). An MS operating pressure of 3.5 × 10−5 Torr was used during all SRM scans. Three SRM ion pairs (Table S1) were used in the final acquisition method to quantify all peptides and detect the presence of interferences from co-eluting, non-specific signals.

The LC-SRM/MS analysis of tryptic digests

Immediately prior to LC/MS analysis, in-gel tryptic digests were reconstituted in 10 μl of 0.1% (v/v) formic acid. An Eksigent NanoLC-1Dplus HPLC (http://www.eksigent.com/) was used to inject 2 μl of each sample onto reversed-phase capillary columns (75 m × 15 cm), packed in-house using Magic C18AQ (5-m internal diameter particles, 100-Å pore size, from Michrom, http://www.michrom.com/). A flow rate of 300 nl min−1 of solvent A (2% acetonitrile, 0.1% formic acid) was used for 6 min. Separations were performed using a flow rate of 300 nl min−1 with a 32-min linear gradient from 0 to 23% solvent B (98% acetonitrile, 0.1% formic acid), followed by a 9-min linear gradient from 23 to 43% solvent B. An Applied Biosystems/MDS SCIEX 4000 QTRAP with a nano-electrospray ionization source controlled by Analyst 1.5 software (Applied Biosystems) was used for all LC-SRM/MS analyses. All acquisition methods used the following parameters: 1900–2000 V ion spray voltage, a curtain gas setting of 25 pounds per square inch (p.s.i.), a 200°C interface heater temperature, a collision activated dissociation (CAD) pressure at 3.5 × 10−5 Torr, and Q1 and Q3 set to unit resolution (0.6–0.8 Da FWHH). Spray stability was improved and the lifespan of the uncoated fused silica emitter tips (20 μm inner diameter, 10 μm tip, New Objective, http://www.newobjective.com/) was improved by use of 3–5 p.s.i. sheath gas and post-column, pre-spray addition of make-up solvent [80% (v/v) isopropanol, 10% (v/v) acetonitrile] at a flow rate of 50 nl min−1 using a PicoPlus 11 syringe pump (Harvard Apparatus). The SRM acquisition methods were constructed using 114 SRM Q1/Q3 ion pairs with peptide-specific tuned declustering potential (DP) and collision energy (CE) voltages and retention time constraints. A default collision cell exit potential of 23 V was used for all SRM ion pairs, and the scheduled SRM option was used for all data acquisition with a target scan time of 2 sec and an 8-min SRM detection window.

SRM data analysis

All SRM data were processed using MultiQuant 1.0 (Applied Biosystems) with the MQL algorithm for peak integration. A 2-min retention time window, with ‘report largest peak’ enabled and a three-point smooth with a peak-splitting factor of 2 was used. The default MultiQuant values for noise percentage and a baseline subtraction window were used. All data were manually inspected to ensure correct peak detection and accurate integration.

TPS enzyme assays

Activity of mono- and diTPSs were determined as described in Martin et al. (2002) with minor modifications. Namely, for diTPS enzyme assays the protein extract was concentrated by centrifugation in 30K Microsep centrifugal device (PALL, http://www.pall.com/). Concentrated protein extract (100 μg) was incubated with the diTPS assay buffer.

MonoTPS enzyme assay products were analyzed on an Agilent 6890N GC equipped with a 5975 Inert XL MS Detector at 70 eV and fitted with a DB-WAX (0.25 mm × 0.25 μm × 30 m) column (Agilent Technolgies Inc., http://www.home.agilent.com/) with a flow rate of 1 ml min−1 He as in Martin et al. (2002). The diTPS enzyme assay products were analyzed on an Agilent 6890A GC equipped with a flame ionization detector (FID) fitted on AT-1000 column (30 m × 0.25 mm, 0.25 μm thickness, Alltech Associates Inc., http://www.discoverysciences.com) at 1.2 ml min−1 H2 on a 7683 series autosampler using a 0.1-μl injection as in Martin et al. (2002).

Extraction and metabolite profiling of terpenoids

Monoterpenes were analyzed by GC-MS/MS on an Agilent 6890N GC equipped with a SolGel-WAX column (30 m × 0.25 mm × 0.25 μm; SGE Ltd, http://www.sge.com/) at 1 ml min−1 He as in Martin et al. (2002). Chiral analysis of monoterpenes was done on the same GC instrument fitted with a cyclodex B column (0.25 mm × 0.25 μm × 30 m, J&W Scientific, http://www.agilent.com). Diterpene resin acids were methylated and analyzed on an Agilent 6890A GC equipped with a FID as in Martin et al. (2002).

RNA extraction for transcript profiling

Bark tissue (150 mg) was ground to a fine powder under liquid nitrogen with 20 mg PVPP. One milliliter of extraction buffer [400mm TRIS–HCL pH 8.5, 51 mm lauryl sulfate lithium salt, 10 mm lithium chloride, 10 mm disodium salt EDTA, 24 mm deoxycholic sodium salt, 1% (v/v) tergitol NP-40, 1 mm aurintricarboxylic acid, 10 mm dithiotheritol, 5 mm thiourea, 2% PVPP (w/v)] was added, samples were centrifuged at 18 000 g for 10 min at 4°C, and supernatant transferred to a fresh tube. One volume of chilled isopropanol and 1/10 volume of 3.3 m sodium acetate were added, samples were incubated at −80°C overnight and centrifuged the next day for 30 min at 10 000 g at 4°C. The pellet was resuspended in 400 μl TRIS-EDTA (TE) buffer and 400 μl of 5 m sodium chloride, then 200 μl of 10% cetyl trimethylammonium bromide (CTAB) was added and the samples incubated at 65°C for 5 min. Samples were extracted twice with 0.5 ml of chloroform:isoamyl alcohol (24:1) and the aqueous phase transferred to a new tube. Then 50 μl of lithium chloride was added and samples were incubated at −80°C overnight. Samples were then centrifuged for 30 min at 18 000 g at 4°C and the pellet resuspended in 100 μl TE buffer. One volume of chilled isopropanol and 1/10 volume 3.3 m sodium acetate were added and samples were incubated overnight at −80°C. Samples were centrifuged for 30 min at 18 000 g at 4°C and the pellet was washed with 0.5 ml 70% ethanol, dried and resuspended in 15 μl nuclease-free water.

Quantitative real time PCR

Prior to cDNA synthesis 3 μg of RNA was treated with DNase I, amplification grade (Invitrogen), and 1 μg was reverse transcribed using Superscript III reverse transcriptase (Invitrogen) and 50 pmol anchored poly dT primer, then cDNA was diluted to a concentration of 1.67 ng μl−1. Quantitative RT-PCR was performed using the DyNAmo® HS SYBR® Green qPCR kit (Finnzymes, http://www.finnzymes.com/) in a 20-μl reaction containing 6 μl of diluted cDNA 6 pmol forward and reverse primer, and 10 μl of DyNAmo master mix. All primers used are shown in Table S2. Two amplicons from each primer pair were sequenced to confirm primer specificity. In addition to no-template (water) controls, no-reverse transcriptase controls were performed for each primer pair to ensure the absence of genomic DNA contamination. Relative expression and PCR efficiencies were calculated using the Real-Time PCR Miner software application (Stanford University, CA). The reference gene chosen for relative quantification was a Norway spruce β-tubulin (Phillips et al., 2007).

Generation of heatmaps

Data points for each protein/metabolite were normalized to the highest value, which was set to 1 in order to equalize the scale of each level of analysis. Normalized time course data were then inputted into the microarray visualization software MeV hosted on the TIGR website (http://www.tm4.org/mev.html).

Acknowledgements

The Norway spruce clonal line used in this study (518473-07) was generously provided by J. D. Irving Ltd (St John, NB, Canada). We are grateful to Lina L. Madilao for excellent technical assistance, to Karen Reid for excellent laboratory management support and to David Kaplan, Tristan Gillan and Alfonso Lara for greenhouse support. This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC; grant to JB), Genome British Columbia and Genome Canada (grant to JB). Salary support for JB was provided in part by an NSERC Steacie Memorial Fellowship and the University of British Columbia Distinguished University Scholar program.

Ancillary