Male Sterile2 encodes a plastid-localized fatty acyl carrier protein reductase required for pollen exine development in Arabidopsis.

Male Sterile2 (MS2) is predicted to encode a fatty acid reductase required for pollen wall development in Arabidopsis (Arabidopsis thaliana). Transient expression of MS2 in tobacco (Nicotiana benthamiana) leaves resulted in the accumulation of significant levels of C16 and C18 fatty alcohols. Expression of MS2 fused with green fluorescent protein revealed that an amino-terminal transit peptide targets the MS2 to plastids. The plastidial localization of MS2 is biologically important because genetic complementation of MS2 in ms2 homozygous plants was dependent on the presence of its amino-terminal transit peptide or that of the Rubisco small subunit protein amino-terminal transit peptide. In addition, two domains, NAD(P)H-binding domain and sterile domain, conserved in MS2 and its homologs were also shown to be essential for MS2 function in pollen exine development by genetic complementation testing. Direct biochemical analysis revealed that purified recombinant MS2 enzyme is able to convert palmitoyl-Acyl Carrier Protein to the corresponding C16:0 alcohol with NAD(P)H as the preferred electron donor. Using optimized reaction conditions (i.e. at pH 6.0 and 30°C), MS2 exhibits a Km for 16:0-Acyl Carrier Protein of 23.3 ± 4.0 μm, a Vmax of 38.3 ± 4.5 nmol mg−1 min−1, and a catalytic efficiency/Km of 1,873 m−1 s−1. Based on the high homology of MS2 to other characterized fatty acid reductases, it was surprising that MS2 showed no activity against palmitoyl- or other acyl-coenzyme A; however, this is consistent with its plastidial localization. In summary, genetic and biochemical evidence demonstrate an MS2-mediated conserved plastidial pathway for the production of fatty alcohols that are essential for pollen wall biosynthesis in Arabidopsis.

In flowering plants, the life cycle alternates between diploid sporophyte and haploid gametophyte generations. Pollen grains play a biologically protective role for the haploid male sperm cells surrounded by the outer cell wall lipidic biopolymers called the exine (Blackmore et al., 2007;Ariizumi and Toriyama, 2011). Pollen exine protects the game-tophyte against pathogen attack, dehydration, and UV irradiation as well as facilitates the pollination process, including pollen recognition and adhesion to the stigma. The highly durable exine that occurs throughout flowering plants is thought to play an essential role in land colonization by plants (Chaloner, 1976).
The exine is mainly composed of the biopolymer sporopollenin and contains two sublayers, the sexine and nexine (Zinkl et al., 1999). The biochemical nature of pollen exine remains largely unknown because of the technical difficulties in purifying and obtaining large quantities of materials for analysis. In addition, sporopollenin is highly insoluble, resistant to degradation, and exceptionally stable (Brooks and Shaw, 1968;Bubert et al., 2002). Current evidence suggests that the major components of sporopollenin are derivatives of aliphatics, such as fatty acids and phenolic compounds (Bubert et al., 2002;Blackmore et al., 2007).
Tapetum, the innermost sporophytic anther wall layer, is thought to play a major role in actively synthesizing and secreting sporopollenin precursors onto the microspore surface for pollen exine polymerization and patterning (Bedinger, 1992;Ariizumi and Toriyama, 2011). The model dicot plant Arabidopsis (Arabidopsis thaliana) and many other plants have a secretory-type tapetum with spe-cialized structures such as tapetosomes in the cells, which accumulate lipidic components (Huysmans et al., 1998). The outer surface of Arabidopsis pollen grains displays elegant reticulate cavities with abundant pollen coat (tryphine) deposited inside the pollen exine.
MS2 was assumed to encode a FAR-like protein that converts fatty acids to alcohols. MS2 was shown to be expressed in the tapetum shortly after the microspore was released from the tetrad (Aarts et al., 1997). ms2 mutants display abnormal pollen wall development, which is sensitive to acetolysis treatment, causing reduced pollen fertility (Aarts et al., 1997;Dobritsa et al., 2009). However, detailed biochemical characterization of the MS2 enzyme has not been performed. Recently, recombinant bacteria expressing five Arabidopsis FAR homologs were shown to produce fatty alcohols with carbon lengths of C14, C16, and C18 from endogenous bacterial fatty acids. Bacteria expressing MS2 are able to form C14:0, C16:0, and C18:1 alcohols (Doan et al., 2009). Furthermore, yeast cells expressing FAR1, FAR4, and FAR5 are able to produce alcohols using distinct but overlapping substrates with a chain length ranging from C18:0 to C24:0 (Domergue et al., 2010).
In this study, we report the biochemical characterization of MS2. We show that MS2 encodes a fatty acyl-Acyl Carrier Protein (ACP) reductase, and the purified recombinant MS2 enzyme from Escherichia coli is able to convert the preferred substrate palmitoyl-ACP to C16:0 alcohol in the presence of NAD(P)H. In addition, MS2 possesses an N-terminal transit peptide that is necessary for localization to the plastid. The biological significance of MS2 subcellular localization, and the presence of conserved domains/motifs within MS2, were demonstrated by genetic complementation of ms2 mutants. This work, therefore, demonstrates the involvement of the plastid in primary fatty alcohol synthesis required for pollen wall development in Arabidopsis.

ms2 Pollen Grains Have Defective Pollen Exine
Defective pollen grains of ms2 were incubated with 4#,6-diamidino-2-phenylindole (DAPI) and iodinepotassium iodide (I 2 -KI). DAPI staining showed that both the wild type and ms2 display normal mature pollen grains with a single vegetative nucleus and two sperm nuclei at the mature pollen developmental stage (Fig. 1, A and B). I 2 -KI staining revealed that ms2 pollen grains had a rough outer surface compared with the wild type ( Fig. 1, C and D), consistent with the defective pollen wall development of ms2. In agreement with previous reports (Aarts et al., 1997;Dobritsa et al., 2009), ms2 pollen grains were sensitive to acetolysis treatment (Fig. 1, E and F), and scanning electron microscopy (SEM) analysis revealed that the ms2 pollen grains collapsed and attained a severely shrunken appearance ( Fig. 1, G and H).

The MS2 Protein Is Localized to Plastids
Previous investigation predicted that MS2 (At3g11980) and a MS2 homolog (At3g56700) were localized in chloroplasts based on the presence of an N-terminal transit peptide (Doan et al., 2009). Our TargetP 1.1, ChloroP 1.1, and MitoProt II version 1.101 sequence analyses consistently suggested that the MS2 protein contained a putative plastid-targeting (or mitochon- drion-targeting) peptide with 46 amino acid residues at the N terminus. In order to determine the subcellular localization of MS2, we made two constructs, pro35S:MS2-GFP and pro35S:MS2MN-GFP, driven by the cauliflower mosaic virus (CaMV) 35S promoter. pro35S:MS2-GFP contained the full-length MS2 cDNA translationally fused to the 5# terminus of GFP, and in pro35S:MS2MN-GFP, the N-terminal transit peptide portion of the MS2 cDNA was deleted and the MS2 polypeptide was translationally fused with GFP ( Fig.  2A). Chloroplasts were identified by the autofluorescence of chlorophyll, and a translational fusion construct of the mitochondrial F1ATPase g-subunit with red fluorescent protein (RFP) was used to mark the mitochondrion . Analysis using Arabidopsis protoplasts from leaf cells showed that the MS2-GFP signal colocalized with the chlorophyll autofluorescence signal in chloroplasts but not obviously with the F1ATPase-g-RFP signal in Arabidopsis protoplasts (Fig. 2, B-E), consistent with MS2 localization in plastids. In contrast, the MS2MN-GFP signals were observed in the cytoplasm and did not colocalize with the RFP signal or the chlorophyll autofluorescence signal (Fig. 2, F-I). Furthermore, we fused the DNA fragment of the MS2 N-terminal transit peptide to GFP and confirmed that the transit peptide of MS2 is able to target the GFP into chloroplasts using pro35S: TP-YFP (for transit peptide-yellow fluorescent protein; Fig. 2, J-L). In addition, the localization of MS2 in plastids was also confirmed in transgenic tobacco (Nicotiana benthamiana) leaves transiently expressing MS2-GFP, MS2MN-GFP, and free GFP (Supplemental Fig. S1).
The expression of MS2 was shown only in tapetal cells at or very shortly after young microspores were released from the tetrad by in situ RNA hybridization and promoter fused with the GUS reporter gene analyses (Aarts et al., 1997). In this study, we observed the GFP signal of proMS2-GFP in the tapetal cells at the free microspore stage in the stable transgenic lines (Fig. 3, A-C). Furthermore, the fluorescent signal in the anthers of transgenic lines harboring proMS2MN-GFP showed higher expression signals in tapetal cells and anther locules compared with those of transgenic lines containing proM2-GFP (Fig. 3, D-F), suggesting that the N-terminal transit peptide of MS2 is responsible for correct MS2 localization, the absence of which possibly affects its turnover.
The biological importance of the MS2 localization to plastids was confirmed from the observation that the construct proM2:MS2MN without the N-terminal transit peptide-encoding sequence of MS2 could not complement the defective pollen wall development in ms2 homozygous mutants, whereas proM2:MS2 rescued the defect of ms2 pollen wall development (Fig. 4, A, B, C, H, and I). The Arabidopsis Rubisco small subunit protein contains an N-terminal transit peptide (RTP) that has been shown to effectively target this protein to chloroplasts (Kim et al., 2010a). To test whether RTP can functionally replace the N-terminal transit peptide of MS2 protein, we made the construct proMS2:RTP-MS2MN#, in which the fragment of RTP translationally replaced the N-terminal transit peptide of MS2 ( Fig.  5A) and transformed it to ms2 mutants. Twenty-one independent transgenic lines containing proM2:RTP-MS2DN# in the ms2 homozygous background were obtained. Acetolysis treatment and SEM analysis showed that 10 transgenic lines displayed normal pollen exine development compared with the wild type (Fig. 5, B and E), suggesting the rescued pollen wall phenotype of ms2 mutants by this construct. The pollen grains of four transgenic lines showed partially resistant to acetolysis treatment and less defective outer surface analyzed by SEM, suggesting a partially rescued phenotype of pollen exine formation (Fig. 5,C and F). Seven transgenic lines showed defects of pollen exine development similar to ms2 (Fig. 5, D and G), which may have resulted from the difference between the RTP and the N-terminal transit peptide of MS2. These results confirmed that the localization of MS2 to plastids is critical for its function in anther development.

NAD(P)H-Binding Domain and Sterile Domain Are Essential for MS2 Function
Consistent with a previous investigation (Aarts et al., 1997), analysis of the MS2 protein using the SMART protein structure prediction tool (http://smart.emblheidelberg.de/) revealed that the MS2 protein contains two conserved domains: the NAD(P)H-binding domain (NBD) between residues 135 and 438 and the sterile domain (SD) between residues 534 and 614.
MS2 and other FARs were observed to contain the conserved NAD(P)H-binding motif between residues 133 and 143 [(I/V/F)X(I/L/V)TGXTGFL(G/A)] and the active site motif YX(3)K of NADP-dependent proteins between residues 357 and 368 ( Fig. 4A; Aarts et al., 1997;Doan et al., 2009), implying a functional role of these motifs.
To examine whether these domains/motifs are essential for MS2 function, we made several constructs containing the mutated MS2 fragments driven by the MS2 promoter (Fig. 4A). The construct containing the deleted SD sequence (proMS2:MS2MSD) or deletion of the NBD domain (proMS2:MS2MNBD) could not rescue the defective pollen exine development in the ms2 homozygous background (Fig. 4, A, D, E, J, and K), suggesting an essential role of both SD and NBD domains in MS2 function. Furthermore, the constructs deleted for the NAD(P)H-binding motif or the active site motif YX(3)K of NADP-dependent proteins failed to complement the defective pollen exine phenotype of ms2 in the transgenic plants (Fig. 4, A, F, G, L, and M), suggesting the essential role of these two conserved motifs in the NBD of MS2.

Purified Recombinant MS2 Enzyme Has Fatty Acyl-ACP Reductase Activity
Although MS2 is predicted as a fatty acid reductase involved in fatty alcohol formation and recombinant bacteria expressing MS2 can produce C14:0, C16:0, and C18:1 alcohols (Aarts et al., 1997;Doan et al., 2009), a comprehensive biochemical characterization of this enzyme has not been performed. To this end, we cloned the full-length MS2 cDNA into the expression vector pET30a (Novagen) and introduced it into E. coli BL21 (DE3). The recombinant MS2 protein was purified by affinity chromatography and subjected to SDS-PAGE. Compared with the predicted size of about 73 kD, the molecular mass of the affinity-purified protein ran somewhat faster on SDS-PAGE (Fig. 6). Because the data presented herein show that MS2 is localized in plastids in which long-chain fatty acids are de novo synthesized and while esterified to ACP (Li-Beisson et al., 2010), we analyzed MS2 activity with acyl-ACPs. Using 14 C-labeled fatty acyl-ACPs, we detected strong activity only with [1-14 C]16:0-ACP, trace activity with [1-14 C]18:0-ACP, and no activity with [1-14 C]14:0-ACP, [1-14 C]16:1-ACP, and [1-14 C]18:1-ACP (Supplemental Table S1), indicating that MS2 prefers palmitoyl-ACP as a substrate. Furthermore, no activity was detected with the equivalent [1-14 C]acyl-CoA substrates (data not show). To further investigate the substrate specificity of recombinant MS2, unlabeled myristoyl-ACP, palmitoyl-ACP, palmitoleoyl-ACP, and stearoyl-ACP were tested with the purified MS2 enzyme in the presence of NAD(P)H, and only palmitoyl-ACP was reduced to C16:0 alcohol (Fig. 7). The specificity of MS2 against palmitoyl-ACP was confirmed by using deuterium-labeled 7,7,8,8-d 4palmitoyl-ACP as substrate, which yielded a product at the same retention time as hexadecanol but with a mass ion +4 relative to the non-mass-labeled standard (Supplemental Fig. S2).
To more comprehensively characterize the MS2 enzyme, we compared the palmitoyl-ACP-reducing activity in the presence of NAD(P)H versus NADH. Under the same experimental conditions, chromato-graphic analysis revealed that the recombinant MS2 has higher activity in the presence of NAD(P)H relative to that using NADH, and no C16:0 alcohol was detectable if NAD(P)H or NADH was omitted (Fig. 8), suggesting that MS2 activity is dependent on the presence of either NAD(P)H or NADH. Furthermore, we show that the optimal pH value for the MS2 activity is 6.0 (Fig. 9). Additionally, assays at various temperatures showed that the optimal working temperature for the MS2 converting palmitoyl-ACP to the C16:0 alcohol is 30°C (Fig. 9). Under the condition of optimum pH and temperature, the recombinant MS2 has a K m of 23.3 6 4.0 mM for palmitoyl-ACP and a V max of 38.3 6 4.5 nmol mg 21 min 21 (Table I).

Fatty Alcohol Production in Tobacco Leaves Transiently Expressing MS2
To demonstrate the physiological role of MS2 activity in producing fatty acyl alcohols, we transiently expressed the MS2 protein in tobacco leaves using a cowpea mosaic virus RNA-2-based expression system (Sainsbury and Lomonossoff, 2008). The constructed  vector pEAQ:MS2 was transformed into tobacco leaves, and overexpression of MS2 protein was confirmed by western-blot analysis using an MS2-specific antibody (see "Materials and Methods") of the tobacco leaf extracts 3 d after Agrobacterium tumefaciens infiltration (Supplemental Figs. S3 and S4). The lipidic compositions of chloroform-extractable cuticular waxes, cutin monomers, and total soluble lipids were analyzed in the transgenic lines using gas chromatography-mass spectrometry (GC-MS) and gas chromatography-flame ionization detection (GC-FID; Bonaventure et al., 2004;Franke et al., 2005) and compared with lines carrying the empty vector pEAQ as the control.
From GC-FID analysis, we observed that the amounts of total wax and total soluble lipids per cm 2 of tobacco leaves in the pEAQ:MS2 lines were very close to those of the control lines. In contrast, the total cutin amount per cm 2 of tobacco leaves was significantly higher in the pEAQ:MS2 lines compared with the control lines (Fig. 10A). Interestingly, compared with the control lines, significant increases of the contents of hexacosanol, octadecanol, and C29 alkane were observed in the wax of tobacco leaves harboring pEAQ:MS2 ( Fig. 10B; Supplemental Table S2) as well as increased levels of hexacosanol and C20 fatty acid in internal lipids ( Fig. 10C; Supplemental Table S3). In addition, significant accumulations of hexacosanol, octadecanol, octadecanoic acid, oleic acid, eicosanoic acid, docosanoic acid, and unknown cutin monomer were observed in the cutin monomers of the transgenic tobacco leaves (Fig. 10D; Supplemental Table S4). These results suggested that MS2 can synthesize fatty alcohols in plants.

MS2 Encodes a Fatty Acyl-ACP Reductase Essential for Normal Pollen Exine Development
Male reproductive development in higher plants is indispensable for the synthesis of fatty acid derivatives . Our results show that purified recombinant MS2 is able to produce C16:0 alcohol, which is required for pollen exine development. Although the biochemical nature of pollen exine remains exclusive, we previously showed that pollen exine may share a common lipidic biosynthetic pathway with that of cutin, which is an insoluble polymer matrix containing the major constituents of hydroxylated and epoxy hydroxylated C16 and C18 fatty acids as well as minor components such as fatty alcohols Zhang et al., 2010). Fatty alcohols with C16 and C18 occupy 0% to 8% of cutin composition (Pollard et al., 2008). CYP704B2 belongs to a conserved and ancient subfamily of fatty acid hydroxylases and is specifically expressed in the tapetum and the microspore. Mutation of CYP704B2 causes aborted pollen  grains without detectable exine and undeveloped anther cuticle with reduced levels of cutin monomers. Recombinant CYP704B2 catalyzes the production of v-hydroxylated fatty acids with 16 and 18 carbon chains, suggesting that CYP704B2 is involved in the biosynthesis of the two biopolymers sporopollenin and cutin . Accordingly, we propose that the MS2-derived fatty alcohol may account for pollen exine development. Consistently, we recently genetically characterized a close homolog of MS2 in rice called Defective Pollen Wall (DPW). dpw displays defective anther development and aborted pollen grains with an abnormal exine, causing complete male sterility. Additionally, dpw anthers have significantly reduced levels of cutin monomers and cuticular wax composition as well as soluble fatty acids and alcohols (Shi et al., 2011). dpw, like MS2, is able to produce C16 primary alcohols.
In this study, we show that the purified recombinant MS2 has a strong preference for converting palmitoyl-ACP to hexadecanol that is dependent on the electron donor NAD(P)H, but it can also use NADH with lower efficiency. Compared with previous investigations, this study provides biochemical insight into plastidially localized enzymes with sequence homology to the FARs from the green alga Euglena gracilis (Kolattukudy, 1970), jojoba (Metz et al., 2000), oilseed rape (Brassica napus; Hu et al., 2006), and Arabidopsis (Rowland et al., 2006;Doan et al., 2009;Domergue et al., 2010). In the Arabidopsis genome, eight FAR-like sequences were identified, and five FARs have been biochemically investigated (Doan et al., 2009). The Arabidopsis FARs share sequence similarity with the jojoba FAR (30% to 54% identity), and CER4 has the highest sequence identity to that of jojoba (54%; Doan et al., 2009). Yeast cells expressing Arabidopsis CER4 generated C24:0 and C26:0 primary alcohols, and CER4 is required for the development of epidermal cells of aerial organs and roots (Rowland et al., 2006). Bacteria expressing MS2 are capable of producing C14:0, C16:0, and C18:1 alcohols (Doan et al., 2009). We are currently unable to reconcile this observation with our biochemical assays, in which we failed to detect activity with [1-14 C]C14:0 substrates. It is possible that bacteria expressing MS2 have more complex metabolic activities mediated by MS2 and bacterial enzymes that are not revealed in purified enzyme assays. Unlike MS2, the FAR enzyme from jojoba is able to convert acyl-CoAs to C16:0 and C18:0 fatty alcohols, which are required for wax production in jojoba seeds (Metz et al., 2000). Silk moth FAR-expressing yeast cells could convert C14-20 fatty acids to alcohols and preferentially to C15 and C16 fatty acids (Moto et al., 2003). Two mouse enzymes, FAR1 and FAR2, were shown to reduce fatty acyl-CoA ester substrates into fatty alcohols (Cheng and Russell, 2004). Unlike the recombinant MS2, the recombinant DPW can convert both palmitoyl-ACP and palmitoyl-CoA into hexadecanol despite its higher activity to palmitoyl-ACP (Shi et al., 2011). In addition, the K m value of the recombinant MS2 toward palmitoyl-ACP was determined to be about 6.5-fold higher than that of DPW, suggesting that DPW has higher affinity for palmitoyl-ACP than MS2. Moreover, the recombinant MS2 has trace activity against [1-14 C]18:0-ACP that is not detectable for DPW. Furthermore, the DPW activity depends exclusively on NAD(P)H (Shi et al., 2011), whereas MS2 can use either NAD(P)H or NADH. These results suggest that MS2 homologs have undergone substantial diversification during the course of the evolution of monocots and dicots.
Alignment analysis revealed that MS2, DPW, and FARs from Arabidopsis (CER4), jojoba, wheat (TAA1a), silk moth, and mouse share conserved NAD(P)Hbinding and active site motifs (Doan et al., 2009;Shi et al., 2011). In this study, we demonstrate the biological importance of these motifs in anther development by genetic complementation (Fig. 4), implying that these motifs are essential for enzyme function. In addition, the critical role of the SD conserved in MS2 and other homologs (Aarts et al., 1997) was also empirically demonstrated in this investigation (Fig. 4).

MS2 Represents a Conserved Plastid-Localized Lipid Metabolism Pathway That Operates during Anther Development
Plant anther cells contain various types of diversified plastids, and plastids within tapetal cells are thought to be associated with lipid synthesis and secretion for pollen wall synthesis (Clement and Pacini, 2001). Plant de novo biosynthesis of fatty acids with short carbon chain lengths of up to C18 is mainly in plastids (Ohlrogge et al., 1979;Li-Beisson et al., 2010). We revealed that the MS2/DPW protein is localized to chloroplasts, and a genetic complementation assay shows that the  Shi et al., 2011). A similar N-terminal transit peptide is also observed in other MS2 homologs (Doan et al., 2009;Shi et al., 2011). Therefore, our data show that a plastid-localized and conserved pathway of converting fatty acids to fatty alcohols is required for male reproductive development in monocots and dicots (Shi et al., 2011).
In conclusion, we have characterized an Arabidopsis plastid-localized fatty acyl-ACP reductase, MS2, that is able to reduce palmitoyl-ACP to C16:0 alcohol. The plastidial MS2 fatty alcohol production pathway is essential for pollen exine synthesis. Moreover, the subcellular localization of MS2 in plastids, mediated by its N-terminal transit peptide, NBD, and SD and the active site motif YX(3)K, is required for MS2 function in anther development. This work reveals a key and conserved step of primary fatty alcohol synthesis for pollen wall biosynthesis in Arabidopsis.

Plant Materials and Growth Conditions
Arabidopsis (Arabidopsis thaliana Columbia ecotype) plants were grown in a greenhouse at 22°C and 70% relative humidity under continuous illumination. The ms2 mutant, a T-DNA insertion line (SAIL_92_C07), was obtained from the Arabidopsis Biological Resource Center. Tobacco (Nicotiana benthamiana) plants were grown in a greenhouse at 25°C with 16 h of supplemental light per day.

Staining, SEM, and Acetolysis Treatment
For DAPI staining, pollen grains were fixed in formaldehyde acetic acid, immersed in 0.2% DAPI, and viewed by fluorescence microscopy under UV light (Leica DM2500). For I 2 -KI staining, pollen grains were immersed in I 2 -KI solution and photographed with a microscope (Leica DM2500). For SEM, pollen grains were air dried in silica, coated with palladium-gold in a sputter coater (Hummer), and then observed in an autoscan scanning electron microscope (JEOL) with an acceleration voltage of 30 kV. Acetolysis treatment was performed as described previously (Aarts et al., 1997).

Laser Scanning Confocal Microscopy
The Arabidopsis protoplasts were prepared and transformed as described (Yoo et al., 2007). Tobacco leaf cells were observed 2 d after argoinfiltration in 50% glycerol. Fluorescence was detected by a Leica TCS SP5 confocal laser scanning microscope with an argon laser at 488 nm for excitation and images for GFP and chlorophyll signals and at 543 nm for mitochondria marker F1ATPase-g-RFP. The signals were collected from 505 to 525 nm for GFP, 650 to 750 nm for chlorophyll, and 560 to 620 nm for F1ATPase-g-RFP.

Vector Construction
The MS2 promoter fragment (1,095-bp upstream fragment of the start codon ATG) was amplified from wild-type Arabidopsis genomic DNA with the primers MS2pro-F and MS2pro-R and cloned into the binary vector pCAMBIA1301 to generate vector p1301proMS2. The full-length MS2 cDNA (1,851 bp) was amplified from wild-type Arabidopsis inflorescence cDNA with the primers MS2-1F/1R and cloned into pMD 18-T vector (Novagen). For deletion analysis of the MS2 gene, serial primers were used to amplify MS2, MS2DN, MS2DSD, MS2DNADP, MS2DNBM, and MS2DASM fragments, which were then digested with NcoI and BstEII and subcloned into the vector under the control of the MS2 promoter, named proMS2:MS2, proMS2:MS2DN, proMS2:MS2DSD, proMS2:MS2DNBD, proMS2:MS2DNBM, and proMS2: MS2DASM, respectively.
For the analysis of MS2 subcellular localization, two MS2 fragments (a 1,848-bp full-length cDNA without the stop codon TAA [MS2] and a 1,701-bp truncated cDNA lacking the 147-bp fragment encoding the N-terminal transit peptide and the stop codon TAA [MS2DN]) were amplified with primers MS2-1F/2R and MS2DN-F/MS2-2R, respectively. The GFP coding region was amplified with primers GFP-F/R. MS2 or MS2DN (digested with NcoI and BglII) and GFP (digested with BglII and BstEII) were subcloned into pMS2 or pCAMBIA 1301 (previously digested with NcoI and BstEII) to generate proMS2:MS2-GFP, proMS2:MS2DN-GFP, pro35S:MS2-GFP, and pro35S: MS2DN-GFP. The N-terminal transit peptide-encoding sequence of MS2 was amplified using primers TP-F/R, digested with SalI and SpeI, and cloned into pA7-YFP (kindly provided by Prof. Hongquan Yang) to generate pro35S:TP-YFP. The DNA fragment encoding the Arabidopsis RTP (240 bp; GenBank accession no. AT1G67090) was amplified from wild-type Arabidopsis inflorescence cDNA with the primers RTP-F/R. The amplified RTP was fused to the N terminus of the truncated MS2 fragment (1,761 bp, encoding MS2 residues 31-616) using overlapping PCR with the primers MS2DN2-F and MS2-1R; the resulting fragment was digested with NcoI and BstEII and then cloned into proMS2, generating proMS2:RTP-MS2DN#.
For overexpressing MS2, the MS2 fragment was amplified with the primers MS2-2F/3R and cloned into the AgeI-and XhoI-digested pEAQ (Sainsbury and Lomonossoff, 2008) to generate pEAQ:MS2. For preparation of the MS2 antibody, the specific MS2 fragment with less similarity to other Arabidopsis genes was amplified with primers MS2f-1F/1R and MS2f-2F/2R. The two PCR products were cloned into the BamHI and XhoI sites of pET30a, forming pET30a:MS2S. For expression of the MS2 enzyme in Escherichia coli, the MS2 fragment was amplified with primers MS2-3F/3R and cloned into pET30a (Novagen) digested with BamHI and SalI, resulting in pET30a:MS2.
All constructs were confirmed by sequencing. Sequences of the primers used are listed in Supplemental Table S5.

Polyclonal Antibody Preparation and Western Blot
pET30a:MS2S was transformed into BL21 (DE3) E. coli cells, and the recombinant cells were grown at 37°C in Luria-Bertani medium containing 50 mg L 21 kanamycin to an optical density at 600 nm of 0.8 and induced using 1 mM isopropyl b-D-1-thiogalactopyranoside. After 48 h of induction at 20°C, the cell pellet was collected by centrifugation at 4°C. The fusion protein purification and antibody preparation were performed as described by Huang et al. (2003). The antiserum was collected after four injections. The antibody was purified according to the method described by Ritter (1991).
Total soluble proteins were extracted from the wild-type buds and pEAQand pEAQ:MS2-transformed tobacco leaves using the protocol described by Huang et al. (2003) and separated by 10% SDS-PAGE. After electrophoresis, the proteins were electronically transferred onto nitrocellulose membranes and incubated with the purified antibodies (1:100 dilution) as described by Huang et al. (2003).
Expression, Purification, and Activity Assay of Recombinant MS2 BL21 (DE3) E. coli cells containing pET30a:MS2 were grown at 30°C to an optical density at 600 nm of 0.6, and MS2 expression was induced by the addition of 0.5 mM isopropyl b-D-1-thiogalactopyranoside and further incubation at 20°C for 48 h. The cell pellet was collected by centrifugation at 4°C and resuspended in phosphate-buffered saline buffer (137 mM NaCl, 2.7 mM KCl, 8.1 mM Na 2 HPO 4 , and 1.76 mM KH 2 PO 4 , pH 7.4) containing 1 mM phenylmethylsulfonyl fluoride and 5 mM dithiothreitol and then lysed by French press. The lysis solution was clarified by centrifugation (45,000g) and loaded onto a nickel-nitrilotriacetic acid agarose (Ni 2+ -NTA) affinity column (Merck) following the manufacturer's instructions. The bound proteins were eluted with phosphate-buffered saline containing 70 mM imidazole and detected by 10% SDS-PAGE. Nonspecific proteins purified from pET30atransformed bacteria were used as a negative control.  (Rock and Garwin, 1979) using recombinant spinach (Spinacia oleracea) ACP-I (Broadwater and Fox, 1999). For the substrate preference analysis using unlabeled fatty acyl-ACP, the reductase activity of MS2 was analyzed as described by Metz et al. (2000). The reaction mixture of 200 mL for the MS2 assay contained 15 mM NAD(P)H, 25 mM HEPES-NaOH, pH 7.5, 1 mM dithiothreitol, 1 mM EDTA, 10% (w/v) glycerol, and 0.3% (w/v) CHAPS plus the desired substrate. NAD(P)H was replaced by NADH or water for the catalytic activity analysis of MS2. The reaction was initiated by addition of the enzyme, and for negative control experiments, an equivalent volume of eluted buffer from the Ni 2+ -NTA column that had been incubated with protein from cells expressing the empty pET30a vector was added. Incubations were performed at 30°C for 60 min and terminated by the addition of an equal volume of cyclohexane. The products were extracted into 1.2 volumes of hexane, and 2 mL of this extract was subjected to GC-MS analysis.
The Agilent GC apparatus was coupled with an Agilent 5975C inert XL MSD with triple axis detector. The detector and injector temperature were maintained at 300°C. The oven temperature of 120°C was increased to 240°C at a rate of 15°C min 21 and held at 240°C for 2 min.
For substrate preference with 14 C-labeled fatty acyl-ACPs, the 150-mL reaction was carried out as described above and incubated at 30°C for 30 min. For pH optimization, 50 mM MES/50 mM Bis-Tris propane buffer with pH range from 5.0 to 9.0 was used, and the reaction was incubated for 15 min. For temperature optimization, the optimum MES buffer (pH 6.0) was used and incubated at 10°C, 16°C, 25°C, 30°C, 35°C, and 42°C. For kinetics analysis of MS2 using [1-14 C]16:0-ACP, the reaction was carried out at optimum pH (50 mM MES, pH 6.0) and temperature (30°C) for 10 min. The reactions were extracted twice with 2 volumes of hexane. The radioactivity of the hexane fraction was measured by liquid scintillation counting. As described above, extracts from bacteria expressing pET30a were used as the negative control.

Analysis of Wax, Internal Lipids, and Cutin in Tobacco Leaves
Tobacco leaf discs used for the analysis of wax, internal lipids, and cutin were prepared as described (Cameron et al., 2006) with modifications. Leaf discs (1.5 cm diameter) were excised using a cork borer from the leaves 3 d after argoinfiltration of pEAQ or pEAQ:MS2, frozen immediately in liquid N 2 , freeze dried, and then stored in silica gel at room temperature. Each sample contained six leaf discs derived from three individual plants.
Cuticular waxes were extracted from the leaf discs described above by immersing tissues for 30 s in 1 mL of chloroform containing 10 mg of tetracosane (Fluka) as an internal standard. The extracts were transferred to reactive vials and dried under a gentle stream of nitrogen gas. The dried wax residues were derivatized by adding 20 mL of N,N-bis-trimethylsilyltrifluoroacetamide (Macherey-Nagel) and 20 mL of pyridine and incubated for 40 min at 70°C. These derivatized samples were then analyzed by GC-FID (Agilent Technologies) and GC-MS (Agilent gas chromatograph coupled to an Agilent 5973N quadrupole mass selective detector).
For total internal lipid extraction, leaf discs that had been used in the wax extraction were reextracted with 1 mL of chloroform:methanol (1:1, v/v). They were first incubated at 50°C for 30 min and then overnight with constant shaking at room temperature. This extraction was repeated three times to ensure that no internal lipids were left in the samples. The combined lipid extracts were then dried under a gentle stream of nitrogen gas to about 100 mL. The remaining delipidated leaf discs were dried over silica gels and used to analyze the monomer composition of cutin polyester as described by Franke et al. (2005). All internal lipid samples and cutin samples were trans-esterified in 1 mL of 1 N methanolic HCl for 2 h at 80°C. After the addition of 2 mL of saturated NaCl and 20 mg of dotriacontane (Fluka) as an internal standard, the hydrophobic monomers were subsequently extracted three times with 1 mL of hexane. The organic phases were combined and dried under a stream of nitrogen gas, and the remaining samples were derivatized as described above. GC-MS and GC-FID analyses were performed as for the wax analysis. Results of leaf wax, internal lipids, and cutin analyses were expressed relative to unit surface area of tobacco leaves.

Supplemental Data
The following materials are available in the online version of this article.
Supplemental Figure S1. Subcellular localization of MS2 in tobacco leaves.
Supplemental Figure S4. Western-blot analysis of MS2 expressed in tobacco leaves.
Supplemental Table S2. Detailed wax compositions in transient tobacco leaves.
Supplemental Table S3. Detailed internal lipid compositions in transient tobacco leaves.
Supplemental Table S4. Detailed cutin compositions in transient tobacco leaves.
Supplemental Table S5. List of the primers used for vector construction.