Skip to main content
NCBI home page
Plant Physiology logoLink to Plant Physiology
. 2015 Aug 17;169(3):1595–1606. doi: 10.1104/pp.15.00573

Comprehensive Assessment of Transcriptional Regulation Facilitates Metabolic Engineering of Isoprenoid Accumulation in Arabidopsis1,[OPEN]

Iris Lange 1, Brenton C Poirier 1, Blake K Herron 1, Bernd Markus Lange 1,*
PMCID: PMC4634052  PMID: 26282236

A population of Arabidopsis lines with modulated levels of expression of each gene involved in isoprenoid precursor pathways was used to develop engineering strategies for increasing sterols.

Abstract

In plants, two spatially separated pathways provide the precursors for isoprenoid biosynthesis. We generated transgenic Arabidopsis (Arabidopsis thaliana) lines with modulated levels of expression of each individual gene involved in the cytosolic/peroxisomal mevalonate and plastidial methylerythritol phosphate pathways. By assessing the correlation of transgene expression levels with isoprenoid marker metabolites (gene-to-metabolite correlation), we determined the relative importance of transcriptional control at each individual step of isoprenoid precursor biosynthesis. The accumulation patterns of metabolic intermediates (metabolite-to-gene correlation) were then used to infer flux bottlenecks in the sterol pathway. The extent of metabolic cross talk, the exchange of isoprenoid intermediates between compartmentalized pathways, was assessed by a combination of gene-to-metabolite and metabolite-to-metabolite correlation analyses. This strategy allowed the selection of genes to be modulated by metabolic engineering, and we demonstrate that the overexpression of predictable combinations of genes can be used to significantly enhance flux toward specific end products of the sterol pathway. Transgenic plants accumulating increased amounts of sterols are characterized by significantly elevated biomass, which can be a desirable trait in crop and biofuel plants.

Isoprenoids are ubiquitous metabolites with diverse biological functions in plants as membrane components (sterols), photosynthetic pigments (carotenoids and chlorophylls), side chains of quinones involved in electron transport systems, phytohormones (GAs, brassinosteroids, abscisic acid, and cytokinins), and antioxidants (carotenoids and tocopherols; Bouvier et al., 2005). There are also numerous commercial uses for isoprenoids in the flavor and fragrance industry (e.g. essential oils and resins), as nutritional supplements (e.g. vitamins A, D, E, and K), and as pharmaceutical drugs (e.g. the anticancer diterpene taxol or the antimalarial sesquiterpene artemisinin; Holstein and Hohl, 2004). Across all kingdoms of life, isoprenoids are derived from two universal five-carbon building blocks, isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP). In most plant cells, the mevalonate (MVA) pathway, which is localized to the cytosol, endoplasmic reticulum, and peroxisomes, is the predominant source of IPP and DMAPP for the synthesis of sterols (C30), whereas the plastidial methylerythritol 4-phosphate (MEP) pathway provides the bulk of IPP and DMAPP for the synthesis of carotenoids (C40) and the C20 isoprenoid side chain of chlorophylls (Hemmerlin et al., 2012; Fig. 1). The exchange of common intermediates of isoprenoid biosynthesis between different compartments has been inferred from feeding experiments with labeled precursors, but the extent of this metabolic cross talk varies significantly between different experimental systems and is still poorly understood at a mechanistic level (Rodríguez-Concepción, 2006; Hemmerlin et al., 2012).

Figure 1.

Figure 1.

Open in a new tab

Outline and compartmentation of isoprenoid biosynthesis in plants. The enzymes involved in the MVA and MEP pathways are numbered as follows: 1, acetoacetyl-CoA thiolase (AACT); 2, 3-hydroxy-3-methylglutaryl-CoA synthase (HMGS); 3, 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGR); 4, mevalonate kinase (MK); 5, phosphomevalonate kinase (PMK); 6, mevalonate 5-diphosphate decarboxylase (MPDC); 7, 1-deoxy-d-xylulose 5-phosphate synthase (DXS); 8, 1-deoxy-dxylulose 5-phosphate reductoisomerase (DXR); 9, 2C-methyl-d-erythritol 4-phosphate cytidyltransferase (MCT); 10, 4-(cytidine 5′-diphospho)-2C-methyl-d-erythritol kinase (CMK); 11, 2C-methyl-d-erythritol 2,4-cyclodiphosphate synthase (MDS); 12, 1-hydroxy-2-methyl-2-butenyl 4-phosphate synthase (HDS); and 13, (E)-4-hydroxy-3-methyl-but-2-enyl diphosphate reductase (HDR). Multistep reactions are depicted with hollow arrows. The potential exchange of isoprenoid intermediates across the plastidial envelope membrane (metabolic cross talk) is indicated with a gray circle. Evidence for the existence of an alternative pathway involving isopentenyl monophosphate (IP) has been presented, but its role in plants is currently unknown (depicted with broken arrows). ER, Endoplasmic reticulum.

To evaluate isoprenoid pathway regulation and opportunities for metabolic engineering, several groups have generated transgenic plant lines in which the expression levels of individual genes involved in the MVA and MEP pathways are modulated. The focus of these studies has been on genes that encode enzymes assumed to catalyze rate-limiting reactions. In animals, where the MEP pathway does not operate, the enzyme HMGR plays an important role in controlling flux through the MVA pathway (Chappell et al., 1995). In plants, the up-regulation of HMGR transcript levels in transgenic plants has led to increases in the major sterol end products in some cases (Schaller et al., 1995; Harker et al., 2003; Holmberg et al., 2003; Hey et al., 2006; Muñoz-Bertomeu et al., 2007; Sales et al., 2007), but not in others (Re et al., 1995). Increased sterol levels have also been reported for transgenic plants overexpressing the HMGS gene, which codes for the enzyme just upstream of HMGR (Wang et al., 2012).

Some authors suggested a limiting role for DXS to provide precursors of chlorophylls and carotenoids via the MEP pathway (Lois et al., 2000; Estévez et al., 2001; Rodríguez-Concepción et al., 2001; Enfissi et al., 2005), but there are also reports in which no correlation between DXS transcript levels and these plastidial end products was observed (Muñoz-Bertomeu et al., 2006). Several studies demonstrated correlations between the expression levels of the genes encoding DXR (Mahmoud and Croteau, 2001; Carretero-Paulet et al., 2002; Hasunuma et al., 2008) or HDR (Botella-Pavía et al., 2004) with chlorophyll and carotenoid concentrations in transgenic plants, whereas others did not find evidence for a limiting role of DXR in carotenoid biosynthesis (Rodríguez-Concepción et al., 2001).

Although we are only beginning to appreciate the relevance of posttranscriptional control, the currently available evidence suggests that the MVA and MEP pathways are both regulated at multiple feedback levels (Rodríguez-Concepción, 2006; Espenshade and Hughes, 2007; Cordoba et al., 2009). For the development of metabolic engineering strategies, it is thus imperative to understand to what extent carbon flux through a pathway of interest is regulated at the transcriptional level. Here, we report, to the best of our knowledge, the first systematic investigation evaluating the relative importance of the expression levels of each individual MVA and MEP pathway gene for controlling the accumulation of marker isoprenoids in Arabidopsis (Arabidopsis thaliana). This information was then used to eliminate bottlenecks in the sterol pathway and dramatically enhance flux toward sterol end products.

RESULTS

Gene-to-Metabolite Correlation Analysis Identifies Genes with Highest Impact on Isoprenoid Accumulation

Complementary DNAs (cDNAs) for all MVA and MEP pathway genes were cloned into suitable plant transformation vectors (three different selectable markers to facilitate subsequent crossing), and following Agrobacterium tumefaciens GV3101-mediated transformation, homozygous transgenic Arabidopsis lines were selected. Wild-type controls and homozygous transgenic lines were grown on soil under tightly controlled conditions in a growth chamber, and rosette leaves were harvested right before bolting. We harvested tissue only from plants with a normal visible phenotype to avoid complications from analyzing pleiotropic effects resulting from secondary outcomes of transgene modulation. In particular, we observed that lines with extremely low sterol amounts were severely dwarfed, and those with extremely low chlorophyll or carotenoid levels had an albino phenotype. Very high overexpression of the DXS gene (and corresponding increases in chlorophylls and carotenoids) has previously been described as causing dwarfism (Estévez et al., 2001), indicating that detrimental effects can occur when isoprenoid levels are increased above a certain level. We determined the relative expression levels of transgenes and the quantities of the major isoprenoid end products (sterols derived from the MVA pathway; chlorophylls and carotenoids as end products of the MEP pathway). Transgenic lines transformed with a vector containing a GUS gene (tested as controls with a gene insert unrelated to isoprenoid biosynthesis) yielded results that introduced more variation, but were statistically indistinguishable from those obtained with untransformed wild-type plants (Supplemental Fig. S1), and we thus used the latter in all further analyses.

The correlation between transgene expression levels and the amounts of marker isoprenoids was evaluated by linear regression analysis. A positive slope of the trend line (m ≥ 0.005), a significantly high Pearson correlation coefficient (r ≥ 0.30), and a low P value for the significance of the found correlation (deemed acceptable at a confidence threshold of 0.05) were obtained for the following gene-to-metabolite correlations: HMGR-sterols (overexpression of the HMG1 gene), DXR-chlorophylls and carotenoids, and MDS-chlorophylls and carotenoids (Table I; Supplemental Figs. S2 and S3). Although no significant correlation between the transcript levels of other transgenes and isoprenoid end products was detectable when all data points were considered, individual transgenic lines in which an MVA pathway gene was overexpressed did accumulate increased amounts of chlorophylls and carotenoids (Fig. 2A). Analogously, in individual transgenic lines, a positive correlation between the expression levels of MEP pathway genes with sterols was observed (Fig. 2A).

Table I. Gene-to-metabolite correlation analysis to identify genes with the highest impact on isoprenoid accumulation.

The expression levels of genes involved in the cytosolic/peroxisomal MVA and plastidial MEP pathways were modulated in transgenic Arabidopsis lines, and major end products derived from these pathways (sterols [MVA] and chlorophylls/carotenoids [MEP]) were quantified. Important parameters of the statistical analysis (fold change versus untransformed controls for both genes and metabolites) are provided (m = slope of correlation trend line; r = Pearson correlation coefficient; P = Student’s t test P value). Gray background is used to indicate positive gene-to-metabolite correlations. Full data are presented in Supplemental Figs. S2 and S3.

Gene Metabolite
Chlorophylls
 Carotenoids
 Sterols
m r P m R P m r P
MVA pathway
 AACT 0.007 0.19 0.293 0.012 0.16 0.395 0.002 0.08 0.690
 HMGS 0.003 0.03 0.423 0.005 0.27 0.244 7 × 10−5 0.004 0.978
 HMGR 0.003 0.03 0.423 0.002 0.14 0.360 0.04 0.49 1.4 × 10−7
 MK 2 × 10−4 0.09 0.612 3 × 10−4 0.18 0.315 7 × 10−5 0.05 0.776
 PMK 0.002 0.26 0.097 0.002 0.16 0.311 0.003 0.31 0.038
 MPDC 0.009 0.19 0.293 0.01 0.20 0.221 0.003 0.05 0.755
MEP pathway
 DXS 0.025 0.11 0.456 0.061 0.29 0.055 0.013 0.12 0.446
 DXR 0.006 0.31 0.05 0.006 0.37 0.016 0.004 0.25 0.104
 MCT 5 × 10−5 0.01 0.929 5 × 10−4 0.11 0.491 5 × 10−4 0.12 0.456
 CMK 3 × 10−4 0.08 0.609 2 × 10−5 0.004 0.979 0.001 0.32 0.032
 MDS 0.011 0.37 0.019 0.012 0.40 0.012 8 × 10−5 4 × 10−4 0.981
 HDS 0.002 0.05 0.738 0.002 0.03 0.834 0.002 0.04 0.819
 HDR 0.008 0.07 0.691 0.004 0.02 0.900 0.011 0.16 0.384
Open in a new tab

Figure 2.

Figure 2.

Open in a new tab

Transgenic lines with elevated sterol levels accumulate pathway intermediates. A, Selected transgenic lines with high isoprenoid concentrations. Asterisks indicate the level of significance in a Student’s t test (*, P ≤ 0.05; and **, P ≤ 0.01). B, Representative chromatogram of HPLC-diode array analysis of chlorophylls and carotenoids (top, untransformed controls; bottom, transgenic lines with increased flux into chlorophylls and carotenoids [MDS overexpressor]). The chromatograms of wild-type and transgenic plants overexpressing the MDS gene are essentially identical, and no accumulation of chlorophyll or carotenoid pathway intermediates is apparent. C, Representative total ion current chromatogram of gas chromatography-mass spectrometry analysis of sterols (top, untransformed controls; bottom, transgenic lines with increased flux into sterols [HMG1 overexpressor]). There is an increased accumulation of the sterol pathway intermediates isofucosterol, cycloartenol, and 24-methylene-cycloartanol in transgenic lines overexpressing the HMG1 gene. D, Correlation analysis for sterol pathway end products against total sterols (left) and sterol pathway intermediates against total sterols (right; all expressed as fold change versus control). Trend lines for linear (blue) and logarithmic (black) correlations are shown for selected end products.

Metabolite-to-Gene Correlation Analysis Identifies Flux Bottlenecks

The results presented above are indicative of shared control, with certain genes exerting a higher level of control compared with others. At the enzymatic level, this notion is consistent with the summation theorem of metabolic control analysis, which implies that metabolic fluxes are systems-level properties and that regulation is shared by all reactions (Kacser and Burns, 1973; Heinrich and Rapoport, 1974). Due to the substantial experimental challenges associated with determining specific activities of all enzymes of the MVA and MEP pathways (Hemmerlin et al., 2012), the calculation of flux control coefficients for metabolic control analysis was impractical, and we thus investigated the utility of metabolite-to-gene correlations to identify flux bottlenecks. The major end products of the plastidial chlorophyll and carotenoid pathways (chlorophyll a, chlorophyll b, trans-lutein, trans-β-carotene, trans-violaxanthin, and cis-neoxanthin) accumulated in all Arabidopsis plants, whereas pathway intermediates accessible with our analytical methods were present only in small amounts (α-carotene) or were undetectable (e.g. phytoene, lycopene, δ-carotene, and [nonprenylated] chlorophyllides). Transgenic plants with increased total chlorophyll and carotenoid levels, when compared with controls, did not accumulate higher levels of these plastidial isoprenoid pathway intermediates (Fig. 2B). In contrast, transgenic lines with significantly increased sterol levels appeared to accumulate disproportionately high amounts of certain biosynthetic intermediates (Fig. 2C). When data sets from all transgenic lines were considered, there was a correlation between total sterol amounts and sterol pathway end products (β-sitosterol and campesterol) that was best described (highest R2 value) by a logarithmic function (strong increase at lower fold change, reaching a plateau for larger fold change values; Fig. 2D). In contrast, for sterol pathway intermediates (isofucosterol, cycloartenol, and 24-methylene cycloartanol), there was a strong linear correlation with total sterol levels (fold change versus controls; Fig. 2D). Based on the position of the accumulating intermediates in the sterol pathway (Fig. 3), we hypothesized that two enzymatic activities, those acting on cycloartenol/24-methylene-cycloartanol or isofucosterol as substrates (sterol methyl oxidase1 [SMO1] and sterol C24 reductase [DWF1], respectively), exerted an increased level of control in transgenic lines engineered for high flux through the sterol pathway.

Figure 3.

Figure 3.

Open in a new tab

Outline of sterol biosynthesis with an emphasis on the reactions catalyzed by SMO1 and DWF1. Triple arrows indicate multiple enzymatic steps.

Metabolic Engineering of Sterol Biosynthesis Using Predictable Combinations of Genes Results in Transgenic Plants with Increased Biomass

To test the hypothesis that SMO1 catalyzes, in addition to sterol methyltransferase 1 as demonstrated previously (Sitbon and Jonsson, 2001), a rate-limiting reaction of the sterol pathway in leaves, we generated a population of SMO1 overexpressors and selected a line with dramatically increased SMO1 transcript levels (740-fold increase compared with appropriate wild-type controls; Fig. 4A; Supplemental Fig. S4) and 2.4-fold higher sterol amounts (Fig. 4B). Importantly, the amounts of the SMO1 substrates, cycloartenol and 24-methylene-cycloartanol, in the highest SMO1 overexpressor were comparable with those in untransformed control plants (Fig. 4C). These results demonstrate that, in SMO1 transgenics, there were higher sterol end product pools but no differences in intermediate pools (compared with untransformed controls) at a defined harvest time point, which is an indication of increased flux through the sterol pathway. The developmental time course was the same in transgenic lines and controls. To investigate if additive effects of transgene overexpression would further increase flux through the sterol pathway, we performed crosses to generate transgenic lines overexpressing both HMG1 and SMO1. A transgenic HMGR/SMO1 double overexpressor accumulated sterol levels (3.8-fold higher than wild-type controls) comparable with those observed with HMG1 overexpressor (3.4-fold higher than wild-type controls; Fig. 4, A and B). However, the HMG1/SMO1 double overexpressor contained significantly lower amounts of the SMO1 substrate cycloartenol (1.5% of total sterols) compared with HMG1 overexpressor (7.1% of total sterols), with an overall composition of sterols similar to that of wild-type controls (Fig. 4C). A separate set of transgenic lines overexpressing DWF1 was also evaluated to decrease the accumulation of the DWF1 substrate isofucosterol (Supplemental Fig. S4). Interestingly, an HMG1/DWF1 double overexpressor indeed contained significantly reduced amounts of isofucosterol when compared with the HMG1 overexpressor, while maintaining the high sterol amounts (Fig. 4, A–C). This is particularly noteworthy because DWF1 expression levels in the HMG1/DWF1 double overexpressor were not as high as in the DWF1 (single) overexpressor. Furthermore, transgenic lines in which the flux into the sterol pathway is increased (HMG1, SMO1, and HMG1/SMO1 overexpressors) were characterized by a significantly higher above-ground biomass (up to 54% increase in HMG1/SMO1 double overexpressor), with less significant effects on stem length (Fig. 4, D and E).

Figure 4.

Figure 4.

Open in a new tab

Effects of overexpressing combinations of genes involved in precursor biosynthesis and the core sterol pathway. A, Expression levels of transgenes in selected single and double overexpressors. WT, Wild type; Ox, overexpression construct. B, Total sterol amounts (as fold change versus untransformed controls) in selected single and double overexpressors. C, Individual sterol profiles (as percentage of total sterols) in selected single and double overexpressors. D, Above-ground biomass in selected single and double overexpressors. E, Stem height enhancements in selected single and double overexpressors. Asterisks indicate the level of significance in a Student’s t test (*, P ≤ 0.05; and **, P ≤ 0.01).

Gene-to-Metabolite and Metabolite-to-Metabolite Correlation Analyses Indicate that Metabolic Cross Talk Is Negligible

Most studies investigating metabolic cross talk have used feeding experiments with isotopically labeled precursors to investigate the distribution of label in isoprenoid end products, and then inferred the extent of exchange between plastidial and cytosolic intermediate pools (Hemmerlin et al., 2012). However, such an approach is essentially impossible to implement with the large number of transgenic lines generated as part of the efforts described here. As an alternative, we used a series of statistical correlation analyses to assess the extent of metabolic cross talk between compartmentalized isoprenoid pathways.

To evaluate the likelihood of metabolic cross talk, we tested whether the expression levels of any gene coding for an extraplastidial enzyme correlated with plastidial end products, or whether there was evidence for a correlation of transcript abundance for a plastidial enzyme and extraplastidial isoprenoids. No statistically significant correlation (neither positive nor negative) was detectable for the expression levels of MVA pathway genes (which encode extraplastidial enzymes) and plastidial isoprenoid end products (chlorophylls and carotenoids; Table I). There was also a lack of correlation between the expression levels of MEP pathway genes (which encode plastidial enzymes) and extraplastidial isoprenoid end products (sterols; Table I). To further evaluate the possibility of metabolic cross talk, we evaluated metabolite-to-metabolite correlations. The rationale was that, if isoprenoid intermediates were exchanged between compartments, plastidial and nonplastidial isoprenoid end products should both be increased in elite transgenic lines. The amounts of the two major classes of end products from the MEP pathway, carotenoids and chlorophylls, were positively correlated across all transgenic and control lines (m = 0.76, r = 0.67, P = 3.8 × 10−64; Fig. 5). In contrast, no correlation, neither positive nor negative, was detected for chlorophylls/carotenoids and MVA pathway-derived sterols (Fig. 5). Taken together, these results provide evidence that, in this study, metabolic cross talk across the chloroplast envelope membrane was negligible.

Figure 5.

Figure 5.

Open in a new tab

Metabolite-to-metabolite correlation analysis of transgenic lines: chlorophylls versus carotenoids (A), carotenoids versus sterols (B), and chlorophylls versus sterols (C; all expressed as fold change versus untransformed controls). Important parameters of the statistical analysis (fold change versus untransformed controls for both genes and metabolites) are provided (m = slope of correlation trend line; r = Pearson correlation coefficient; P = Student’s t test P value). WT, Wild type.

DISCUSSION

Are There Rate-Limiting Steps in Arabidopsis Sterol Biosynthesis?

As early as the mid-1990s, Chappell et al. (1995) posed the question of whether HMGR, an enzyme widely regarded as being critical for controlling flux through the MVA pathway in animals (Espenshade and Hughes, 2007), played a similar role in plants. The authors expressed a hamster HMGR gene in transgenic tobacco plants (Nicotiana tabacum) and observed a 3-fold increase (compared with controls) in leaf sterol amounts. However, it was also reported that cycloartenol, an intermediate of the sterol pathway, accumulated to very high levels (70-fold increase) in HMGR overexpressors (Chappell et al., 1995). Similar results were obtained in several independent studies in which HMGR genes from various species were overexpressed in tobacco (Schaller et al., 1995; Harker et al., 2003; Hey et al., 2006). Only Re et al. (1995) found that HMGR overexpression did not result in measurable changes in sterol amounts, but these authors only studied two transgenic lines, which may not have been sufficient to obtain interpretable data. It was then hypothesized that sterol C24 methyltransferase1 (SMT1), which acts on cycloartenol (accumulated intermediate in transgenic plants) as a substrate, was a sterol pathway-specific rate-limiting enzyme. Indeed, when SMT1 was overexpressed in transgenic tobacco, sterol levels were increased (1.4-fold increase), without a significant accumulation of cycloartenol (Holmberg et al., 2003). The combined overexpression of HMGR and SMT1 resulted in elevated sterol concentrations (2.5-fold increase), with low levels of pathway intermediates (Holmberg et al., 2003). Does this mean that there are only two rate-limiting enzymes in the plant sterol pathway? To investigate this issue systematically, we generated populations of transgenic lines with varying levels of expression for each gene involved in the MVA pathway. Interestingly, almost any MVA pathway gene could be overexpressed to achieve an enhanced accumulation of sterols (Fig. 2A). For some genes, these increases were very small or moderate (MK, PMK, and MPDC), but for both HMGS and HMGR (HMG1 gene), highly significant sterol increases were detected in selected lines. The most dramatic sterol increases (4-fold) were observed with an HMG1 overexpressor. HMG1 transgenics were also the only population in which a positive correlation between transgene expression levels and sterol amounts was determined when all lines were considered (Table I). From these data sets, one can deduce that the HMG1 gene exerts the highest level of control over flux into the sterol pathway, despite the known complexities with the multilevel regulation of HMGR activity (Hemmerlin et al., 2012). However, as mentioned above, HMGS expression also has a considerable effect on sterol accumulation (Fig. 2A), and the term rate limiting would thus be an exaggeration for HMGR.

All Arabidopsis transgenic lines with elevated sterol amounts accumulated disproportionately high levels of cycloartenol (Fig. 2D), which is consistent with transgenic tobacco overexpressing HMGR (Chappell et al., 1995). In addition, we detected dramatically increased concentrations of the intermediates 24-methylene-cycloartanol and isofucosterol. These findings provided evidence that, in addition to SMT1 (Holmberg et al., 2003), SMO1 (which uses cycloartenol and 24-methylene-cycloartanol as substrates [Darnet and Rahier, 2004]) and DWF1 (which acts on isofucosterol as a substrate [Choe et al., 1999]) may contribute significantly to the regulation of the sterol pathway. Indeed, transgenic lines overexpressing combinations of genes, namely, HMG1/SMO1 or HMG1/DWF1, were characterized by increased sterol amounts, with low accumulation levels of intermediates (Fig. 4, A–C). Taken together, the results presented here indicate that control over flux through the sterol pathway is shared by several enzymes.

Are There Rate-Limiting Steps in Arabidopsis Carotenoid and Chlorophyll (Side Chain) Biosynthesis?

In our hands, almost any MEP pathway gene could be overexpressed to achieve an enhanced accumulation of carotenoids and chlorophylls (Fig. 2A). Transcriptional regulation therefore plays an important role in determining the accumulation of isoprenoid end products. However, since we did not observe a positive correlation between expression levels and end products for all MEP pathway genes (exceptions are DXR and MDS as discussed below), the transcript level may have to be titrated just right to facilitate an increased flux through the pathway. Alternatively, posttranscriptional control, which is still poorly understood (Rodríguez-Concepción, 2006), may become a flux-limiting factor when the expression levels of MEP pathway genes are modulated above or below a certain threshold. The only transgenes for which we did not identify a single line with consistently increased marker isoprenoid levels were those encoding DXS and HDS. HDS had previously been described as being nonlimiting for the biosynthesis of Arabidopsis carotenoids (Flores-Pérez et al., 2008). Yet, a nonlimiting role of DXS for chlorophyll and carotenoid biosynthesis in Arabidopsis contrasts with the data of Estévez et al. (2001). However, these authors reported that DXS overexpressors with elevated plastidial isoprenoid levels were significantly affected by growth retardation, whereas we did not include transgenic lines with severe phenotypes. Therefore, a direct comparison with the study by Estévez et al. (2001) is not possible. Phenotypically normal transgenic plants had only a fairly narrow range of DXS expression levels (1.1- to 3.9-fold higher than wild-type controls) in our population, which might be explained by the fact that DXS is the most highly expressed gene of the MEP pathway in rosette leaves of wild-type plants (Schmid et al., 2005).

When carotenoids and chlorophylls were assayed as metabolic end products, positive gene-to-metabolite correlations across entire populations of transgenics were only detected in lines with altered DXR or MDS expression levels (Table I). Furthermore, significant enhancements (>50%) in the amounts of isoprenoid end products derived from the MEP pathway were detected solely when the transcript levels of MDS were appreciably increased (Table I). Interestingly, the product of the MDS-catalyzed reaction, 2C-methyl-d-erythritol-2,4-cyclodiphosphate, was recently demonstrated to be a plastid-to-nucleus signaling metabolite that elicits stress-responsive genes (Xiao et al., 2012). It is intriguing to speculate that MEP pathway regulation, and thus the accumulation of plastidial isoprenoids, involves retrograde signaling, and that MDS might play an important role in this process.

HPLC chromatograms of extracts from transgenic lines with significantly elevated carotenoid and chlorophyll amounts had the same metabolite pattern as those of untransformed controls (with higher signal intensities; Fig. 2, B and C). The fact that we did not find evidence for the accumulation of carotenoid pathway intermediates is consistent with previous studies demonstrating that phytoene synthase, the entry enzyme into the carotenoid pathway, plays a regulatory role in Arabidopsis vegetative tissues (von Lintig et al., 1997; Rodríguez-Villalón et al., 2009). One way to interpret this might be that regulation of the carotenoid pathway, under favorable conditions, is less complex when compared with the sterol pathway. However, in this context, it is important to note that the highest carotenoid and chlorophyll amounts, without negatively affecting plant growth, were in a fairly narrow range (0.6- to 1.7-fold change compared with untransformed controls), whereas significantly increased sterol levels (up to 6-fold change compared with untransformed controls) did not have negative effects on growth. Regulatory complexity arises under stress conditions, when a higher xanthophyll accumulation is desirable and additional levels of control need to be considered (Davison et al., 2002).

MVA and MEP Transgenic Lines Are a Valuable Resource for Assessing the Functional Roles of Isoprenoids

We demonstrate that transgenic plant lines with significantly increased sterol levels are characterized by increased above-ground biomass (Fig. 4D). Biomass yield is a critical trait for bioenergy crops, particularly because the chemical composition of feedstocks is only a minor concern for emerging technologies such as hydroprocessing combined with zeolite catalysis (Vispute et al., 2010). On a different token, sterols have been shown to contribute to meristem organization and vascular patterning independent of brassinosteroid hormones (Souter et al., 2002; Men et al., 2008), and our transgenic lines are an invaluable resource to shed light on important but poorly understood developmental functions of sterols. Furthermore, isoprenoid volatiles have recently been described as determinants of Arabidopsis resistance to bacterial infection and insect herbivory, and their biosynthesis appears to be constrained mostly by precursor availability (Tholl and Lee, 2011). Our transgenic lines with significantly increased flux through the MVA or MEP pathways may therefore release altered levels of isoprenoid volatiles (has not been tested yet) and could contribute to a better understanding of the roles of isoprenoids in plant defense.

MATERIALS AND METHODS

Generation of Transgenic Arabidopsis Lines

Genes involved in the MVA and MEP pathways of isoprenoid biosynthesis were PCR amplified from several pooled Arabidopsis (Arabidopsis thaliana) cDNA libraries (Supplemental Table S1). Purified PCR products were sequence verified and recombined with the pDONR201 Gateway vector (Invitrogen), thus yielding an entry clone. Cassettes containing the genes of interest were then transferred from the entry clone to the p*7WG2 transfer DNA destination vector suite (Karimi et al., 2002). These vectors are engineered to contain one of the plant selectable marker genes encoding neomycin phosphotransferase II, hygromycin phosphotransferase, or bialaphos acetyltransferase, which confer resistance against kanamycin, hygromycin, or glufosinate ammonium (Basta), respectively. Target genes (13 genes with three different selection markers; 39 constructs total) were placed between the promoter and terminator of the Cauliflower mosaic virus 35S gene (Odell et al., 1985). The destination vector was transformed into competent Agrobacterium tumefaciens GV3101 cells by electroporation (Koncz and Schell, 1986), and transformed into Arabidopsis plants (Columbia ecotype accession) using the floral dip method (Clough and Bent, 1998). Seedlings from the T2 seed generation were grown to select for a 3:1 survivor to nonsurvivor rate on the appropriate antibiotic/herbicide (as expected for single copy gene insertions following Mendelian genetics) using standard laboratory protocols (Weigel and Glazebrook, 2002), and seed from surviving plants was collected (T3 seed). Homozygous T3 lines were identified by a 100% survival rate of their progeny seedlings under selection pressure. In addition, homozygosity was tested using routine PCR methods (Weigel and Glazebrook, 2002). To generate double transgenics, line HMG1-9.1.7 (kanamycin resistant) was used as a pollen donor in a cross with SMO1-1.1 (Basta resistant), and line HMG1-6.1.7 (Basta resistant) was the pollen donor for a cross with DWF1-7.1 (hygromycin resistant). F1 seedlings derived from cross pollination were selected by 10 d of growth on appropriate selective medium. The resulting F1 plants were self-pollinated and grown to maturity for seed collection, providing a segregating F2 population for each cross. To identify double-homozygous plants from the F2 generation, seed was collected from each F2 plant and plated on medium containing two selection agents (one for each overexpression cassette). Plants with a progeny survival rate greater than 98% were considered homozygous for both transgenes.

Plant Growth and Harvest

Wild-type and homozygous transgenic lines were grown on soil (in 6- × 6-cm pots) and maintained in a growth chamber (16-h-day/8-h-night photoperiod; 100 μmol m−2 s−1 light intensity at soil level; constant temperature at 23°C; 70% relative humidity). Pots were arranged in a random grid and rotated once a day. Three independent pools of plants were grown at different times to generate biological replicates. Rosette leaves were collected at growth stage 5.10 (just as the inflorescence starts to appear; Boyes et al., 2001), pooled, weighed, and shock frozen in liquid nitrogen. Harvested plant material from each pool was homogenized (Ball Mill MM301, Retsch) in the presence of liquid nitrogen, weighed, and aliquots were stored in 2-mL Eppendorf tubes. Representative plants of each line were grown to maturity (growth stage 8.0; prior to onset of senescence; Boyes et al., 2001) and phenotypic measurements taken (stem length; number of flowers; number and length of siliques; diameter, length, and number of rosette leaves; and visual phenotype scoring). For dry weight biomass measurements, tissue was placed in an oven at 50°C for 8 d.

Quantitative Real-Time PCR

RNA was extracted using the Trizol reagent (Life Technologies) according to the manufacturer’s instructions. Isolated RNA (1,000 ng) was treated with RNase-free DNase (Thermo Scientific) and first strand cDNA synthesized using Superscript III reverse transcriptase (Life Technologies). In a 10-μL quantitative PCR reaction, concentrations were adjusted to 150 nm (primers), 1× Power SYBR Green PCR Master Mix (Life Technologies), and 10× diluted first strand cDNA as template (see Supplemental Table S2 for primer sequences). Reactions were performed in a 96-well optical plate at 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 10 min in a 7500 Real-Time PCR system (Life Technologies). Fluorescence intensities of three independent measurements (technical replicates) were normalized against the ROX reference dye (ThermoFisher Scientific). Relative transcript levels were calculated based on the comparative CT method as specified in the manufacturer’s instructions (Life Technologies) using β-actin (At3g18780) as the constitutively expressed endogenous control and the expression levels of the corresponding wild-type allele as the calibrator.

Sterol Analyses

Tissue homogenate (60–70 mg) was extracted at 75°C for 60 min with 4 mL of CHCl3/MeOH (2:1, v:v; containing 1.25 mg L−1 epi-cholesterol as an internal standard). Samples were evaporated to dryness (EZ2-Bio, GeneVac), and the remaining residue was saponified at 90°C for 60 min in 2 mL of 6% (w/v) KOH in MeOH. Upon cooling to room temperature, 1 mL of hexane and 1 mL of H2O were added, and the mixture was shaken vigorously for 20 s. Following centrifugation (3,000g for 2 min) to separate the phases, the hexane phase was transferred to a 2-mL glass vial, the aqueous phase reextracted with 1 mL of hexane as above, centrifuged as above, and the hexane phase added to the 2-mL glass vial containing the hexane phase from the first extraction. The combined organic phases were evaporated to dryness as above, 50 µL of N-methyl-N-trimethylsilyltrifluoroacetamide was added to the residue, the sample was shaken vigorously for 20 s, and the mixture was transferred to a 2-mL autosampler glass vial with a 100-µL conical glass insert. After capping the vial, the reaction mixture was incubated at room temperature for 5 min. Gas chromatography-mass spectrometry analyses were performed on an Agilent 6890N gas chromatograph coupled to an Agilent 5973 inert mass selective detector (MSD) detector. Samples were loaded (injection volume 1 µL) with a LEAP CombiPAL autosampler onto an Agilent HP-5MS fused silica column (30 m × 250 µm; 0.25-µm film thickness). The temperatures of the injector and MSD interface were both set to 280°C. Analytes were separated at a flow rate of 1 mL min−1, with He as the carrier gas, using a thermal gradient starting at 170°C (hold for 1.5 min), a first ramp from 170°C to 280°C at 37°C/min, a second ramp from 280°C to 300°C at 1.5°C/min, and a final hold at 300°C for 5.0 min. Analytes were fragmented in electron impact mode with an ionization voltage of 70 eV and data acquired using MSD ChemStation software (revision D.01.02.SP1, Agilent Technologies). Background was subtracted and peaks were deconvoluted using Automatic Mass Spectral Deconvolution and Identification Software. Analytes were identified based on their mass fragmentation patterns by comparison with those of authentic standards using the National Institute of Standards and Technology Mass Spectral Search Program (version D.05.00). Peak areas were obtained from the total ion chromatogram for all detectable peaks with a phytosterol mass fragmentation signature. Raw data were exported to Microsoft Excel and peak areas normalized to tissue mass and internal standard. A blank injection was performed after each sample run, and the background signal from the blank subtracted from the sample values for the entire run. Prior to sample analyses, and then after every 20 samples, a standard mix was run to evaluate the reproducibility of the analyses. The adjusted peak areas for sterols were added up and compared with the corresponding sum of sterol peak areas for untransformed controls (expressed as fold change versus control). Linear regression analyses for gene-to-metabolite, metabolite-to-gene, and metabolite-to-metabolite correlations were performed in Microsoft Excel.

Chlorophyll and Carotenoid Analyses

Tissue homogenate (approximately 110–120 mg) was extracted with three aliquots of 2 mL of CHCl3/MeOH (2:1; v/v), containing 2 mg L−1 ubiquinone-10 as an internal standard, in a dark room with supplemental lighting from a 40-W green bulb. Each extraction included a 15-min incubation at room temperature, and the three organic extracts were combined in a glass vial wrapped with aluminum foil to avoid light exposure. The pooled organic phases were extracted against 1.5 mL of 1 m NaCl/50 mm Tris-HCl (pH 7.5), and the vials were centrifuged (3,000g for 2 min) to achieve phase separation. The (lower) organic phase was transferred to a new tube (kept in darkness) and the solvent evaporated to dryness (EZ2-Bio, GeneVac). Analytes were dissolved in 200 µL of ethyl acetate, and the mixture was filtered through a 0.45-µm nylon syringe filter and transferred to a brown 2-mL autosampler glass vial with a 100-µL conical glass insert. HPLC/diode array detector/mass spectrometry analyses were performed using an Agilent Series 1100 HPLC system (including a G1315B diode array detector) coupled to an Agilent G2445D LC/MSD Trap SL mass spectrometer. Samples were loaded (injection volume, 20 µL) onto a Prontosil C30 column (250 × 4.6 mm; 5-µm particle size; Bischoff Chromatography distributed via MAC-MOD Analytical) equipped with a guard column of the same stationary phase. The mobile phase consisted of methanol (A), water/methanol (20:80; v:v) containing 0.2% (w/v) ammonium formate (B), and tert-butyl methyl ether (C). A gradient flow rate of 1 mL min−1 was used to separate analytes (Fraser et al., 2000): 11 min of isocratic gradient of 95% A/5% B, a step change to 90% A/5% B/5% C, a linear gradient (15 min) to 30% A/5% B/65% C, a hold at this condition for 5 min, and a conditioning phase to return to the initial conditions. Ultraviolet/visible absorbance was monitored at 275, 287, 460, and 655 nm, and spectra were recorded from 250 to 750 nm in 2-nm increments. Analytes were identified and quantified by comparison of retention times and ultraviolet/visible spectra with those of corresponding reference standards. To further ensure proper peak assignments, eluting analytes from random samples were also ionized using atmospheric pressure chemical ionization, and mass traces were acquired by single-ion monitoring in positive ion mode to detect the following mass ion transitions: mass-to-charge ratio (m/z) 864 → 680 (ubiquinone-10; internal standard), m/z 538 → 444 (α- and β-carotene), m/z 569 → 551 (lutein), m/z 894 → 615 (chlorophyll a), m/z 907 → 629 (chlorophyll b), and m/z 872 → 593 (pheophytin a). The probe voltage was set to 4.0 kV, the capillary voltage was at 2,200 V, the gas temperature was 350°C, and the nebulizer gas flow was 9.0 L min−1. Raw data were exported to Microsoft Excel and peak areas normalized to tissue mass and internal standard using Microsoft Access. To ensure low background signals, a blank injection was performed after every six samples. Prior to sample analyses, and then after every 20 samples, a standard mix was run to evaluate the reproducibility of the analyses. The adjusted peak areas for carotenoids and chlorophylls were added up separately and compared with the corresponding sum of peak areas for untransformed controls (expressed as fold change versus control). Linear regression analyses for gene-to-metabolite, metabolite-to-gene, and metabolite-to-metabolite correlations were performed in Microsoft Excel.

Three copies of the gene encoding DXS are present in the Arabidopsis genome. The gene encoding the most broadly expressed isoform (At4g15560) has been characterized biochemically and was chosen for the current study. The subsequent steps are catalyzed by DXR, 2C-methyl-d-erythritol 4-phosphate cytidyltransferase, CMK, MECPS, HDS, and HDR, all of which are encoded by single genes in Arabidopsis. Biochemical characterizations have been performed with the Arabidopsis genes encoding DXR (At5g62790), 2C-methyl-d-erythritol 4-phosphate cytidyltransferase (At2g02500), HDS (At5g60600), and HDR (At4g34350). Functional data have also been obtained for a gene encoding CMK from tomato (Solanum lycopersicum), and a gene coding for MECPS has been characterized in Ginkgo biloba, but the putative Arabidopsis orthologs have not yet been characterized. However, when the putative plastidial targeting sequences are excluded, the sequence identity/homology of functionally characterized CMK and MECPS proteins and the translated peptides of At1g63970 and At2g26930 in Arabidopsis are above 85% and 94%, respectively, and the corresponding cDNAs were cloned for the current study. Acetoacetyl-CoA thiolase is encoded by two genes in the Arabidopsis genome, but it was shown that At5g48230 coded for the isozyme involved primarily in sterol biosynthesis. HMGS (At4g11820) and MK (At5g27450) are both encoded by single-copy genes in Arabidopsis, and both of these enzymes have been biochemically characterized. Two HMGR gene copies are present in Arabidopsis, and we cloned a cDNA representing the functionally characterized and constitutively expressed HMG1 isoform (At1g76490). Only a distant homolog of the well-characterized yeast (Saccharomyces cerevisiae) PMK gene is detectable in the Arabidopsis genome (At1g31910), but this enzyme is generally poorly conserved across species. Two MPDC copies are detectable in the Arabidopsis genome, one of which (At2g38700) encodes a protein that has been functionally characterized. The corresponding cDNA was cloned for the activities described here. Three copies of sterol C4 methyl oxidase are present in Arabidopsis, and we selected the most widely expressed isoform (At4g22756) for overexpression. The DWF1 gene (encoding sterol C24 reductase) occurs as a single copy (At3g19820) in the Arabidopsis genome.

Supplemental Data

The following supplemental materials are available.

Supplementary Material

Supplemental Data
supp_169_3_1595__index.html (1.5KB, html)

Acknowledgments

We thank Maximilian J. Feldman, Sean R. Johnson, Jared Mell, Amber Parrish, and Lyuba S. Yurgel for technical assistance.

Glossary

IPP

isopentenyl diphosphate

DMAPP

dimethylallyl diphosphate

MVA

mevalonate

MEP

methylerythritol 4-phosphate

HDR

(E)-4-hydroxy-3-methyl-but-2-enyl diphosphate reductase

cDNA

complementary DNA

HDS

1-hydroxy-2-methyl-2-butenyl 4-phosphate synthase

CMK

4-(cytidine 5′-diphospho)-2C-methyl-d-erythritol kinase

MECPS

2C-methyl-d-erythritol 2,4-cyclodiphosphate synthase

m/z

mass-to-charge ratio

Footnotes

1

This work was supported by the National Science Foundation (grant no. NSF–MCB–0920758 to B.M.L.).

[OPEN]

Articles can be viewed without a subscription.

References

  1. Botella-Pavía P, Besumbes O, Phillips MA, Carretero-Paulet L, Boronat A, Rodríguez-Concepción M (2004) Regulation of carotenoid biosynthesis in plants: evidence for a key role of hydroxymethylbutenyl diphosphate reductase in controlling the supply of plastidial isoprenoid precursors. Plant J 40: 188–199 [DOI] [PubMed] [Google Scholar]
  2. Bouvier F, Rahier A, Camara B (2005) Biogenesis, molecular regulation and function of plant isoprenoids. Prog Lipid Res 44: 357–429 [DOI] [PubMed] [Google Scholar]
  3. Boyes DC, Zayed AM, Ascenzi R, McCaskill AJ, Hoffman NE, Davis KR, Görlach J (2001) Growth stage-based phenotypic analysis of Arabidopsis: a model for high throughput functional genomics in plants. Plant Cell 13: 1499–1510 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Carretero-Paulet L, Ahumada I, Cunillera N, Rodríguez-Concepción M, Ferrer A, Boronat A, Campos N (2002) Expression and molecular analysis of the Arabidopsis DXR gene encoding 1-deoxy-d-xylulose 5-phosphate reductoisomerase, the first committed enzyme of the 2-C-methyl-d-erythritol 4-phosphate pathway. Plant Physiol 129: 1581–1591 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chappell J, Wolf F, Proulx J, Cuellar R, Saunders C (1995) Is the reaction catalyzed by 3-hydroxy-3-methylglutaryl coenzyme A reductase a rate-limiting step for isoprenoid biosynthesis in plants? Plant Physiol 109: 1337–1343 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Choe S, Dilkes BP, Gregory BD, Ross AS, Yuan H, Noguchi T, Fujioka S, Takatsuto S, Tanaka A, Yoshida S, et al. (1999) The Arabidopsis dwarf1 mutant is defective in the conversion of 24-methylenecholesterol to campesterol in brassinosteroid biosynthesis. Plant Physiol 119: 897–907 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Clough SJ, Bent AF (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J 16: 735–743 [DOI] [PubMed] [Google Scholar]
  8. Darnet S, Rahier A (2004) Plant sterol biosynthesis: identification of two distinct families of sterol 4alpha-methyl oxidases. Biochem J 378: 889–898 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Davison PA, Hunter CN, Horton P (2002) Overexpression of bold beta-carotene hydroxylase enhances stress tolerance in Arabidopsis. Nature 418: 203–206 [DOI] [PubMed] [Google Scholar]
  10. Enfissi EM, Fraser PD, Lois LM, Boronat A, Schuch W, Bramley PM (2005) Metabolic engineering of the mevalonate and non-mevalonate isopentenyl diphosphate-forming pathways for the production of health-promoting isoprenoids in tomato. Plant Biotechnol J 3: 17–27 [DOI] [PubMed] [Google Scholar]
  11. Espenshade PJ, Hughes AL (2007) Regulation of sterol synthesis in eukaryotes. Annu Rev Genet 41: 401–427 [DOI] [PubMed] [Google Scholar]
  12. Estévez JM, Cantero A, Reindl A, Reichler S, León P (2001) 1-Deoxy-D-xylulose-5-phosphate synthase, a limiting enzyme for plastidic isoprenoid biosynthesis in plants. J Biol Chem 276: 22901–22909 [DOI] [PubMed] [Google Scholar]
  13. Flores-Pérez U, Pérez-Gil J, Rodríguez-Villalón A, Gil MJ, Vera P, Rodríguez-Concepción M (2008) Contribution of hydroxymethylbutenyl diphosphate synthase to carotenoid biosynthesis in bacteria and plants. Biochem Biophys Res Commun 371: 510–514 [DOI] [PubMed] [Google Scholar]
  14. Fraser PD, Pinto MES, Holloway DE, Bramley PM (2000) Application of high-performance liquid chromatography with photodiode array detection to the metabolic profiling of plant isoprenoids. Plant J 24: 551–558 [DOI] [PubMed] [Google Scholar]
  15. Harker M, Holmberg N, Clayton JC, Gibbard CL, Wallace AD, Rawlins S, Hellyer SA, Lanot A, Safford R (2003) Enhancement of seed phytosterol levels by expression of an N-terminal truncated Hevea brasiliensis (rubber tree) 3-hydroxy-3-methylglutaryl-CoA reductase. Plant Biotechnol J 1: 113–121 [DOI] [PubMed] [Google Scholar]
  16. Hasunuma T, Takeno S, Hayashi S, Sendai M, Bamba T, Yoshimura S, Tomizawa K, Fukusaki E, Miyake C (2008) Overexpression of 1-Deoxy-D-xylulose-5-phosphate reductoisomerase gene in chloroplast contributes to increment of isoprenoid production. J Biosci Bioeng 105: 518–526 [DOI] [PubMed] [Google Scholar]
  17. Heinrich R, Rapoport TA (1974) A linear steady-state treatment of enzymatic chains. General properties, control and effector strength. Eur J Biochem 42: 89–95 [DOI] [PubMed] [Google Scholar]
  18. Hemmerlin A, Harwood JL, Bach TJ (2012) A raison d’être for two distinct pathways in the early steps of plant isoprenoid biosynthesis? Prog Lipid Res 51: 95–148 [DOI] [PubMed] [Google Scholar]
  19. Hey SJ, Powers SJ, Beale MH, Hawkins ND, Ward JL, Halford NG (2006) Enhanced seed phytosterol accumulation through expression of a modified HMG-CoA reductase. Plant Biotechnol J 4: 219–229 [DOI] [PubMed] [Google Scholar]
  20. Holmberg N, Harker M, Wallace AD, Clayton JC, Gibbard CL, Safford R (2003) Co-expression of N-terminal truncated 3-hydroxy-3-methylglutaryl CoA reductase and C24-sterol methyltransferase type 1 in transgenic tobacco enhances carbon flux towards end-product sterols. Plant J 36: 12–20 [DOI] [PubMed] [Google Scholar]
  21. Holstein SA, Hohl RJ (2004) Isoprenoids: remarkable diversity of form and function. Lipids 39: 293–309 [DOI] [PubMed] [Google Scholar]
  22. Kacser H, Burns JA (1973) The control of flux. Symp Soc Exp Biol 27: 65–104 [PubMed] [Google Scholar]
  23. Karimi M, Inzé D, Depicker A (2002) GATEWAY vectors for Agrobacterium-mediated plant transformation. Trends Plant Sci 7: 193–195 [DOI] [PubMed] [Google Scholar]
  24. Koncz C, Schell J (1986) The promoter of TL-DNA gene 5 controls the tissue-specific expression of chimaeric genes carried by a novel type of Agrobacterium binary vector. Mol Gen Genet 204: 383–396 [Google Scholar]
  25. Lois LM, Rodríguez-Concepción M, Gallego F, Campos N, Boronat A (2000) Carotenoid biosynthesis during tomato fruit development: regulatory role of 1-deoxy-D-xylulose 5-phosphate synthase. Plant J 22: 503–513 [DOI] [PubMed] [Google Scholar]
  26. Mahmoud SS, Croteau RB (2001) Metabolic engineering of essential oil yield and composition in mint by altering expression of deoxyxylulose phosphate reductoisomerase and menthofuran synthase. Proc Natl Acad Sci USA 98: 8915–8920 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Men S, Boutté Y, Ikeda Y, Li X, Palme K, Stierhof YD, Hartmann MA, Moritz T, Grebe M (2008) Sterol-dependent endocytosis mediates post-cytokinetic acquisition of PIN2 auxin efflux carrier polarity. Nat Cell Biol 10: 237–244 [DOI] [PubMed] [Google Scholar]
  28. Muñoz-Bertomeu J, Arrillaga I, Ros R, Segura J (2006) Up-regulation of 1-deoxy-d-xylulose-5-phosphate synthase enhances production of essential oils in transgenic spike lavender. Plant Physiol 142: 890–900 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Muñoz-Bertomeu J, Sales E, Ros R, Arrillaga I, Segura J (2007) Up-regulation of an N-terminal truncated 3-hydroxy-3-methylglutaryl CoA reductase enhances production of essential oils and sterols in transgenic Lavandula latifolia. Plant Biotechnol J 5: 746–758 [DOI] [PubMed] [Google Scholar]
  30. Odell JT, Nagy F, Chua NH (1985) Identification of DNA sequences required for activity of the cauliflower mosaic virus 35S promoter. Nature 313: 810–812 [DOI] [PubMed] [Google Scholar]
  31. Re EB, Jones D, Learned RM (1995) Co-expression of native and introduced genes reveals cryptic regulation of HMG CoA reductase expression in Arabidopsis. Plant J 7: 771–784 [DOI] [PubMed] [Google Scholar]
  32. Rodríguez-Concepción M. (2006) Early steps in isoprenoid biosynthesis: multilevel regulation of the supply of common precursors in plant cells. Phytochem Rev 5: 1–15 [Google Scholar]
  33. Rodríguez-Concepción M, Ahumada I, Diez-Juez E, Sauret-Güeto S, Lois LM, Gallego F, Carretero-Paulet L, Campos N, Boronat A (2001) 1-Deoxy-D-xylulose 5-phosphate reductoisomerase and plastid isoprenoid biosynthesis during tomato fruit ripening. Plant J 27: 213–222 [DOI] [PubMed] [Google Scholar]
  34. Rodríguez-Villalón A, Gas E, Rodríguez-Concepción M (2009) Phytoene synthase activity controls the biosynthesis of carotenoids and the supply of their metabolic precursors in dark-grown Arabidopsis seedlings. Plant J 60: 424–435 [DOI] [PubMed] [Google Scholar]
  35. Sales E, Muñoz-Bertomeu J, Arrillaga I, Segura J (2007) Enhancement of cardenolide and phytosterol levels by expression of an N-terminally truncated 3-hydroxy-3-methylglutaryl CoA reductase in Transgenic digitalis minor. Planta Med 73: 605–610 [DOI] [PubMed] [Google Scholar]
  36. Schaller H, Grausem B, Benveniste P, Chye ML, Tan CT, Song YH, Chua NH (1995) Expression of the Hevea brasiliensis (H.B.K.) Müll. Arg. 3-hydroxy-3-methylglutaryl-coenzyme A reductase 1 in tobacco results in sterol overproduction. Plant Physiol 109: 761–770 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Schmid M, Davison TS, Henz SR, Pape UJ, Demar M, Vingron M, Schölkopf B, Weigel D, Lohmann JU (2005) A gene expression map of Arabidopsis thaliana development. Nat Genet 37: 501–506 [DOI] [PubMed] [Google Scholar]
  38. Sitbon F, Jonsson L (2001) Sterol composition and growth of transgenic tobacco plants expressing type-1 and type-2 sterol methyltransferases. Planta 212: 568–572 [DOI] [PubMed] [Google Scholar]
  39. Souter M, Topping J, Pullen M, Friml J, Palme K, Hackett R, Grierson D, Lindsey K (2002) hydra mutants of Arabidopsis are defective in sterol profiles and auxin and ethylene signaling. Plant Cell 14: 1017–1031 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Tholl D, Lee S (2011) Terpene specialized metabolism in Arabidopsis thaliana. The Arabidopsis Book 9: e0143, [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Vispute TP, Zhang H, Sanna A, Xiao R, Huber GW (2010) Renewable chemical commodity feedstocks from integrated catalytic processing of pyrolysis oils. Science 330: 1222–1227 [DOI] [PubMed] [Google Scholar]
  42. von Lintig J, Welsch R, Bonk M, Giuliano G, Batschauer A, Kleinig H (1997) Light-dependent regulation of carotenoid biosynthesis occurs at the level of phytoene synthase expression and is mediated by phytochrome in Sinapis alba and Arabidopsis thaliana seedlings. Plant J 12: 625–634 [DOI] [PubMed] [Google Scholar]
  43. Wang H, Nagegowda DA, Rawat R, Bouvier-Navé P, Guo D, Bach TJ, Chye ML (2012) Overexpression of Brassica juncea wild-type and mutant HMG-CoA synthase 1 in Arabidopsis up-regulates genes in sterol biosynthesis and enhances sterol production and stress tolerance. Plant Biotechnol J 10: 31–42 [DOI] [PubMed] [Google Scholar]
  44. Weigel D, Glazebrook J (2002) Arabidopsis: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York [Google Scholar]
  45. Xiao Y, Savchenko T, Baidoo EE, Chehab WE, Hayden DM, Tolstikov V, Corwin JA, Kliebenstein DJ, Keasling JD, Dehesh K (2012) Retrograde signaling by the plastidial metabolite MEcPP regulates expression of nuclear stress-response genes. Cell 149: 1525–1535 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Data
supp_169_3_1595__index.html (1.5KB, html)
supp_pp.15.00573_00573SupplFigS3_Final.pptx (157.6KB, pptx)
supp_pp.15.00573_00573SupplFigS1_Final.pptx (45.6KB, pptx)
supp_pp.15.00573_00573SupplTableS2_Final.xls (36KB, xls)
supp_pp.15.00573_00573SupplFigS4_Final.pptx (221.5KB, pptx)
supp_pp.15.00573_00573SupplTableS1_Final.xls (37.5KB, xls)
supp_pp.15.00573_00573SupplFigS2_Final.pptx (133.3KB, pptx)

Articles from Plant Physiology are provided here courtesy of Oxford University Press

RESOURCES