Research ArticleRESEARCH ARTICLES - FOCUS ISSUEF
Open Access

A Feedback-Insensitive Isopropylmalate Synthase Affects Acylsugar Composition in Cultivated and Wild Tomato

Jing Ning, Gaurav D. Moghe, Bryan Leong, Jeongwoon Kim, Itai Ofner, Zhenzhen Wang, Christopher Adams, A. Daniel Jones, Dani Zamir, Robert L. Last

Published November 2015. DOI: https://doi.org/10.1104/pp.15.00474

Abstract

Acylsugars are insecticidal specialized metabolites produced in the glandular trichomes of plants in the Solanaceae family. In the tomato clade of the Solanum genus, acylsugars consist of aliphatic acids of different chain lengths esterified to sucrose, or less frequently to glucose. Through liquid chromatography-mass spectrometry screening of introgression lines, we previously identified a region of chromosome 8 in the Solanum pennellii LA0716 genome (IL8-1/8-1-1) that causes the cultivated tomato Solanum lycopersicum to shift from producing acylsucroses with abundant 3-methylbutanoic acid acyl chains derived from leucine metabolism to 2-methylpropanoic acid acyl chains derived from valine metabolism. We describe multiple lines of evidence implicating a trichome-expressed gene from this region as playing a role in this shift. S. lycopersicum M82 SlIPMS3 (Solyc08g014230) encodes a functional end product inhibition-insensitive version of the committing enzyme of leucine biosynthesis, isopropylmalate synthase, missing the carboxyl-terminal 160 amino acids. In contrast, the S. pennellii LA0716 IPMS3 allele found in IL8-1/8-1-1 encodes a nonfunctional truncated IPMS protein. M82 transformed with an SlIPMS3 RNA interference construct exhibited an acylsugar profile similar to that of IL8-1-1, whereas the expression of SlIPMS3 in IL8-1-1 partially restored the M82 acylsugar phenotype. These IPMS3 alleles are polymorphic in 14 S. pennellii accessions spread throughout the geographical range of occurrence for this species and are associated with acylsugars containing varying amounts of 2-methylpropanoic acid and 3-methylbutanoic acid acyl chains.

Plants make large numbers of metabolites with great structural diversity, differences in cell and tissue type distribution and concentrations, and physiological roles. Beyond evolutionarily ancient and highly conserved central metabolic processes such as photosynthetic carbon fixation and amino acid metabolism, plants produce specialized metabolites (historically referred to as secondary metabolites) that typically are found in a subset of taxa (Milo and Last, 2012). These specialized biosynthetic pathways use building block metabolites derived from more ancient central metabolism (Modolo et al., 2009; de Kraker and Gershenzon, 2011).

Gene duplications and mutations have been documented to modify the activity, expression, or functional regulation of enzymes involved in central metabolism to yield catalysts for specialized metabolism. An especially elegant example is found in grass species, which have evolved a novel form of the Trp synthase α-subunit (indole synthase or indole-3-glycerolphosphate lyase) that produces indole in the first committing step of defensive cyclic hydroxamic acid biosynthesis (Frey et al., 1997; Melanson et al., 1997; Gierl and Frey, 2001). This evolutionary protein engineering created an enzyme that diverts indole away from amino acid production and toward specialized metabolism.

Other examples center upon the key theme of evolutionary modification of the regulation of the expression or activity of a committing enzyme of primary metabolism to produce an enzyme that is involved in specialized metabolism (Moghe and Last, 2015). For example, elicitor-treated cell cultures of Ruta graveolens express a variant anthranilate synthase that is insensitive to allosteric inhibition by the end product Trp (Bohlmann et al., 1995, 1996). These changes in gene expression and allosteric regulation allow the variant enzyme, which is not subject to feedback inhibition by Trp, to be expressed under conditions in which anthranilate-derived alkaloid production is high. Another example of protein engineering is evident in the case of METHYLTHIOALKYLMALATE SYNTHASE (MAM). This evolutionary variant of ISOPROPYLMALATE SYNTHASE (IPMS), the enzyme that catalyzes the committing step in Leu biosynthesis in bacteria, fungi, and plants, catalyzes the first step in chain elongation in Met-derived glucosinolate biosynthesis in Arabidopsis (Arabidopsis thaliana) and other crucifers (de Kraker and Gershenzon, 2011). The MAM enzymes have evolved two novel functions. First, point mutations cause altered substrate specificity, which allows them to use methylthio derivatives of the more typical IPMS substrate 2-oxoisovalerate (OIV). Second, these enzymes lack the C-terminal protein domain that is responsible for Leu end product inhibition in IPMS enzymes. These three examples illustrate how gene duplication and neofunctionalization can lead to the conversion of amino acid biosynthetic activities to specialized metabolic enzymes by changing genetic and biochemical regulation as well as substrate preference (Milo and Last, 2012; Moghe and Last, 2015).

Many species within the Solanaceae (nightshade family) accumulate a diverse group of acylsugar specialized metabolites in their glandular trichomes (Fobes et al., 1985; McDowell et al., 2011; Schilmiller et al., 2012). The acylsugars identified thus far consist of short to medium (C2–C12) chain length aliphatic acids in normal (straight chain), iso (branched at the penultimate carbon), or anteiso (branched at the antepenultimate carbon) configuration esterified to Suc or Glc (Fig. 1A; Severson et al., 1985; Burke et al., 1987; King et al., 1990; Walters and Steffens, 1990). Acylsugars contribute to direct defense against a wide variety of insect herbivores (Rodriguez et al., 1993; Juvik et al., 1994; Chortyk et al., 1997; Puterka et al., 2003), and substitutions in either the sugar or acyl chain component alter their insecticidal activity (Puterka et al., 2003). Acylsugars also function in multitrophic defense: young Manduca sexta (tobacco hornworm) larvae metabolize Nicotiana attenuata acylsucroses and release volatile short-chain organic acids that attract predatory ants (Weinhold and Baldwin, 2011). Their activity in plant protection led to interest in breeding for altered acylsugar quantities and composition in Solanum lycopersicum (cultivated tomato). Because acylsugars are especially abundant and structurally varied in wild tomato species (Shapiro et al., 1994; Schilmiller et al., 2010, 2015; Kim et al., 2012; Ghosh et al., 2014), these relatives provide an attractive source of variation for improvement in S. lycopersicum pest tolerance (Maluf et al., 2010; Leckie et al., 2012, 2014).

Figure 1.
Figure 1.

Characterization of the IL8-1-1 candidate gene and expression of GFP driven by the IPMS3 upstream regulatory region. A, Structure of the abundant S. lycopersicum M82 acylsucrose S4:17 (2,5,5,5) as determined from NMR spectroscopy (Schilmiller et al., 2010). Acylsucrose molecules are defined as S for the Suc moiety followed by the number of acyl chains and the total number of carbons in the acyl chains [e.g. S4:17 (2,5,5,5) is an acylsucrose with four acyl chains of length 2, 5, 5, and 5, for a total of 17 carbons]. B, Proposed pathways to branched acyl chains from BCAA biosynthesis. The enzymes involved in each step are highlighted in blue. Enzyme abbreviations are as follows: TD, Thr deaminase; AHAS, acetohydroxyacid synthase; KARI, ketoacid reductoisomerase; DHAD, dihydroxyacid dehydratase; BCAT, branched-chain aminotransferase; IPMDH, isopropylmalate dehydrogenase. The Solyc names of genes with strong trichome-enriched expression are shown in red, with log2 fold change calculated between isolated trichomes and shaved stems. Detailed expression information is available in Supplemental Data Set S1. C, Genomic locations of the three IL8-1-1 IPMS genes. Black boxes represent IPMS genes, with the arrowheads representing the direction of transcription. The introgressed segments appear as solid bars in which the boundary edge of each segment is indicated by inclusive (+) and exclusive (−) flanking markers. Numbers at the bottom denote the genomic positions according to a previous publication (Chitwood et al., 2013). D, ProSlIPMS3:GFP-GUS plants show GFP expression in the apical cells of type I/IV trichomes. No expression is detected in type VI trichomes (indicated by the arrows).

We have been investigating acylsucrose biosynthesis in S. lycopersicum and wild relatives to understand the biochemical basis of the intraspecific and interspecific acylsugar diversity (Kim et al., 2012; Schilmiller et al., 2015) and to develop foundations for engineering insect tolerance. Two acyltransferases of the BAHD (for benzyl alcohol O-acetyltransferase [BEAT], anthocyanin O-hydroxycinnamoyltransferase [AHCT], N-hydroxycinnamoyl/benzoyltransferase [HCBT], deacetylvindoline 4-O-acetyltransferase [DAT]) superfamily involved in acylsugar biosynthesis were recently described: ACYLSUCROSE ACYLTRANSFERASE3 (SlASAT3) and SlASAT4 (formerly SlAT2) catalyze the third and fourth steps, respectively, of tetraacylsucrose biosynthesis in S. lycopersicum M82 (Schilmiller et al., 2012, 2015). Allelic variation in these enzymes leads to diversity in the types of acylsugars present in accessions of the wild tomato Solanum habrochaites (Kim et al., 2012; Schilmiller et al., 2015).

Less is known about the genes and enzymes responsible for the diversity of acyl-CoA ester donor substrates of acylsugar biosynthesis across the Solanaceae. Feeding studies with 14C- and 2H-labeled amino acids showed that short, branched acyl chains (2-methylpropanoic acid [iC4], 3-methylbutanoic acid [iC5], and 2-methylbutanoic acid) on acylsugars are derived from branched-chain amino acids (BCAAs; Val, Leu, and Ile) metabolism (Fig. 1B; Kandra et al., 1990; Kandra and Wagner, 1990; Walters and Steffens, 1990). The branched acyl chains of C6 to C12 are derived from BCAAs by elongation through a fatty acid synthase-mediated mechanism or α-keto acid elongation pathway (van der Hoeven and Steffens, 2000; Kroumova and Wagner, 2003). In addition, Slocombe et al. (2008) used virus-induced gene silencing to reduce expression of the BCAA catabolic enzyme BRANCHED-CHAIN KETO ACID DEHYDROGENASE (BCKDH) and KETOACYL-ACYL CARRIER PROTEIN SYNTHETASE I genes, associated with BCAA degradation and fatty acid synthase, respectively. Their results are consistent with the model that acyl-CoA production from BCAAs is involved in acylsugar biosynthesis.

We seek to identify enzymes that control the range of acyl-CoA substrates used for acylsugar biosynthesis. Screening of S. lycopersicum M82 × Solanum pennellii LA0716 chromosome introgression lines identified a region of S. pennellii chromosome 8 that caused M82 to shift dramatically from acylsucroses with abundant iC5 chains to iC4 acyl chains (Schilmiller et al., 2010). This change led to the hypothesis that there is an S. pennellii allele in this region that shifts BCAA metabolic flux from Leu (the source of iC5) to Val (the source for iC4; Fig. 1B; Schilmiller et al., 2010). Here, we report the identification of S. lycopersicum Solyc08g014230 (SlIPMS3), which encodes a type I/IV trichome-expressed IPMS. This enzyme differs from other plant and microbial IPMSs that function in Leu biosynthesis in that it has a C-terminal deletion that renders the protein insensitive to feedback inhibition by Leu without causing the loss of enzymatic activity. The S. pennellii LA0716 IPMS3 protein has a larger C-terminal deletion that eliminates in vitro IPMS activity and causes the protein to be defective in the yeast two-hybrid (Y2H) protein-protein interaction assay. This enzymatically inactive allele was found to be homozygous in LA0716 and other S. pennellii accessions that accumulate large amounts of iC4 acyl chains in their acylsugars. In contrast, S. pennellii with higher iC5 acyl chains contain at least one copy of the enzymatically active IPMS3 allele. These results indicate that a relatively simple change in IPMS structure and function can cause dramatic changes in the types of acylsugar acyl chains that accumulate in cultivated and wild tomato species. These findings have implications for the understanding of insect defense in natural populations and approaches to traditional breeding and transgenic manipulation of insect resistance in crop plants.

RESULTS

Identification of a Novel Trichome-Expressed IPMS Gene

Leaf surface metabolites were profiled previously from S. lycopersicum M82 × S. pennellii LA0716 introgression lines (Eshed and Zamir, 1995), revealing several loci that influence trichome metabolite accumulation (Schilmiller et al., 2009, 2010, 2012, 2015; Kim et al., 2014). The introduction of S. pennellii LA0716 chromosomal DNA from region 8-1-1 at the top of chromosome 8 caused a dramatic increase in the levels of metabolites containing iC4 acyl chains, which are normally of low abundance in S. lycopersicum M82, at the expense of the typically abundant iC5 acyl chains (Schilmiller et al., 2010).

The IL8-1-1 region is predicted to encode approximately 700 genes based on the International Tomato Annotation Group 2.4 annotation (Mueller et al., 2005; Tomato Genome Consortium, 2012). This region was narrowed to 43 genes using selected lines from the 10K SOLCAP SNP-genotyped backcross inbred lines (BILs) that have recombination break points throughout this region (see “Materials and Methods”; Supplemental Table S1; Supplemental Fig. S1). To identify trichome-expressed genes in this interval, RNA sequencing (RNA-Seq) was performed using RNA isolated from stems, trichomes isolated from stems, or stems with the trichomes removed (Supplemental Data Set S1; where TS refers to stems with trichomes, T refers to isolated trichomes, and S refers to shaved stems). Two genes in the region of interest showed more than 2-fold differential expression between isolated trichome and shaved-stem RNA (Supplemental Table S2; significance of P < 0.05; see yellow highlighted rows). These are Solyc08g014190, annotated as a cytochrome P450, and Solyc08g014230, which is one of a cluster of three genes in this region annotated as encoding the first committing enzyme of Leu biosynthesis, IPMS (Fig. 1C). In addition, a fourth IPMS gene is found on chromosome 6 (Tomato Genome Consortium, 2012). We renamed these genes as follows: Solyc06g053400, SlIPMS1; Solyc08g014130, SlIPMS2; Solyc08g014230, SlIPMS3; and Solyc08g014240, SlIPMS4. The RNA-Seq analysis revealed that SlIPMS3 was expressed more than 350-fold higher in isolated stem trichomes compared with shaved stems. In contrast, SlIPMS2 was not differentially expressed, and expression of SlIPMS4 was not detected in any of the tissues we analyzed (Supplemental Table S3) or in complementary DNA (cDNA) sequencing data for other tomato tissues in the Sol Genomics Network (http://solgenomics.net/). These expression results were cross validated by quantitative PCR (qPCR) of cDNA for SlIPMS2 and SlIPMS3 (Supplemental Fig. S2).

Acylsugars accumulate in the long and slender type I/IV long trichome (Fobes et al., 1985; McDowell et al., 2011). Recent work with two BAHD acylsucrose acyltransferases (SlASAT3 and SlASAT4) showed that these genes are specifically expressed in the apical cells of these long trichomes (Schilmiller et al., 2012, 2015), strongly suggesting that the biosynthetic pathway is located in this cell type. To look at the cell type-specific expression of SlIPMS3, transgenic plants that express GFP-GUS under the control of the SlIPMS3 promoter (ProSlIPMS3:GFP-GUS) were generated. Examination of GFP fluorescence in three independent ProSlIPMS3:GFP-GUS lines revealed fluorescence specifically within the apical cells of type I/IV trichomes (Fig. 1D). This expression pattern is indistinguishable from that seen for SlASAT3 and SlASAT4 and corroborates and extends the RNA-Seq and qPCR data.

Sequence Comparison of the S. lycopersicum and S. pennellii LA0716 IPMS3 Genes

In addition to the trichome apical cell expression described above, the S. lycopersicum Heinz 1706 IPMS3 gene model predicts that it encodes an unusual variant missing approximately 160 amino acids at the C terminus of the protein compared with the chromosome 6 SlIPMS1. This would result in deletion of the Leu-binding allosteric regulatory domain described in Arabidopsis (de Kraker and Gershenzon, 2011) and microbes (Soper et al., 1976; Koon et al., 2004; de Carvalho et al., 2005). Resequencing confirmed that the S. lycopersicum M82 SlIPMS3 also contains this truncation, indicating that this is a bona fide change rather than a sequence assembly error. Because introgression of the S. pennellii LA0716 region containing this gene has a major impact on acyl chain composition, we attempted to identify the orthologous gene from this accession using available sequence data (Bombarely et al., 2012; Tomato Genome Consortium, 2012; Bolger et al., 2014a). The SlIPMS3 (Solyc08g014230) and S. pennellii (Sopen08g005140) genes are reciprocal best hits, with 97% identity within a large syntenic region comprising 361 collinear genes (see “Materials and Methods”). In addition, a phylogenetic reconstruction of IPMS sequences in the S. lycopersicum, S. pennellii, and Nicotiana benthamiana genomes suggested that Sopen08g005140 and SlIPMS3 are orthologs (Fig. 2). Thus, we refer to the S. pennellii LA0716 allele Sopen08g005140 as SpIPMS3.

Figure 2.
Figure 2.

Phylogenetic relationships between IPMS and MAM sequences from different species. Sequences were aligned using MUSCLE (Edgar, 2004) and used for phylogenetic reconstruction using the method noted by Benderoth et al. (2009). Specifically, the tree was constructed with Mr. Bayes version 3.2.1 using the wag model (Huelsenbeck and Ronquist, 2001). The Markov Chain Monte Carlo simulation was run with four chains each for 20,000 iterations with a sample frequency of 200 and a burn-in fraction of 0.05. The average potential scale reduction factor was 0.996, and the average sd of the split frequencies was 0.01, indicating convergence of the algorithm. The values on the branches represent the posterior probabilities obtained via the Bayesian simulation. Only the values for the internal branches are shown. The tree was rooted using the Ostreococcus tauri (Ot) IPMS sequence, based on previous knowledge from Benderoth et al. (2009). At, Arabidopsis; Bd, Boechera divaricarpa; Bi, Brassica insularis; Bo, Brassica oleracea; Ca, Capsicum annuum; Nb, N. benthamiana; Os, Oryza sativa; Pt, Populus trichocarpa; Sl, S. lycopersicum; Sopen, S. pennellii; Zm, Zea mays. The sequences that start with PG are for Solanum tuberosum. Numbers represent accessions from the Sol Genomics Network database (http://solgenomics.net/) and GenBank accession numbers.

To determine whether the sequence of SpIPMS3 from the chromosome 8 introgression lines is the same as that in the S. pennellii LA0716 sequence, IPMS3 cDNA and genomic DNA were sequenced from the S. lycopersicum M82 and S. pennellii LA0716 mapping parents as well as IL8-1-1 and IL8-1. cDNA amplification of IPMS3 from M82 yielded a product of identical size and sequence to that predicted from the S. lycopersicum Heinz 1706 annotation. In contrast, sequence analysis of SpIPMS3 cDNA revealed a 109-bp deletion at the 3′ end of the coding region, as was expected based on the S. pennellii LA0716 genome sequence annotation. This deletion is predicted to cause a frame-shift mutation and premature stop codon at amino acid 395, which would remove the entire subdomain II as well as the Leu-binding domain (Fig. 3, A and D). To determine whether the deletion is found in the genome, DNA from S. lycopersicum M82 and S. pennellii LA0716 were amplified by using PCR primers flanking the deletion break points. The result showed that the SpIPMS3 genomic DNA PCR product is 1,029 bp shorter than SlIPMS3 (Fig. 3B). Seventeen single-nucleotide polymorphisms were detected between M82 and LA0716 across the full-length SpIPMS3 cDNA (Supplemental Fig. S3), 10 of which are predicted to affect the amino acid sequence (Supplemental Fig. S4). Thus, SpIPMS3 from LA0716 is predicted to encode an enzyme of 395 amino acids, which would produce a protein missing 51 amino acids from the C terminus relative to the full-length SlIPMS3.

Figure 3.
Figure 3.

An internal deletion in SpIPMS3 genomic DNA results in the loss of subdomain II. A, Short cDNA transcripts of IPMS3 were detected by reverse transcription-PCR from S. pennellii LA0716 trichome cDNA. SlIPMS3, 1,338 bp; SpIPMS3, 1,229 bp. B, Genomic DNA deletion of IPMS3 was detected by PCR from S. pennellii LA0716, which is consistent with the results of reverse transcription-PCR in A. SlIPMS3, 1,354 bp; SpIPMS3, 325 bp. C, Gene structure of SlIPMS3. Exons, introns, and untranslated regions are indicated with black boxes, black lines, and white boxes, respectively. The thin horizontal black line indicates the location of the 1,029-bp deletion in S. pennellii LA0716 IPMS3. Positions of the primers used for PCR in B are indicated by the arrows. D, Predicted protein structures of SlIPMS3 and SpIPMS3. A 109-bp deletion in the SpIPMS3 cDNA sequence, indicated by the dashed line, results in a premature translational termination (indicated with the asterisk) that is predicted to delete the sequence encoding subdomain II. Positions of the primers used for reverse transcription-PCR in A are indicated by the arrows.

SlIPMS3 Encodes a Feedback-Insensitive Enzyme, and SpIPMS3 Is Nonfunctional in Vitro

IPMS from plants and bacteria contains two core domains, an N-terminal catalytic region and a C-terminal allosteric regulatory domain, which is documented to mediate Leu feedback inhibition in microbes (Soper et al., 1976; Koon et al., 2004; de Carvalho et al., 2005) and Arabidopsis (de Kraker and Gershenzon, 2011). These two domains are separated by two smaller subdomains (I and II; Koon et al., 2004). In the IPMS gene cluster on chromosome 8, SlIPMS3 and SlIPMS4 (64% amino acid identity with Arabidopsis IPMS; Supplemental Fig. S4) are predicted to lack approximately 160 amino acids at the C terminus, which include the Leu-binding allosteric regulatory domain. Lack of the C-terminal allosteric domain led us to hypothesize that SlIPMS3 encodes a trichome-expressed IPMS protein that is not subject to Leu feedback inhibition.

To test this hypothesis, SlIPMS1 and SlIPMS3 were expressed in Escherichia coli as fusion proteins containing an N-terminal His tag. These proteins were purified with a nickel-nitrilotriacetic acid agarose column and assayed for IPMS activity using the IPMS substrate OIV. Both proteins used this substrate with kinetic properties similar to published Arabidopsis IPMS enzymes (Table I; de Kraker and Gershenzon, 2011). Two lines of reasoning led us to test the activity of SlIPMS3 with the MAM substrate 4-methylthio-2-oxobutyrate (MTOB). First, the SlIPMS3 enzyme has strong sequence similarity to the Met-derived glucosinolate biosynthetic enzymes AtMAM1 (AT5G23010) and AtMAM3 (AT5G23020; 56% and 58% amino acid identity, respectively; Fig. 2; Supplemental Figs. S4 and S5; Kroymann et al., 2001; Textor et al., 2004, 2007). Second, removal of the C terminus of AtIPMS2 increased its activity with MTOB substrate compared with the full-length AtIPMS2 (de Kraker and Gershenzon, 2011). While SlIPMS3 used MTOB as a substrate, the catalytic efficiency (kcat/Km) of SlIPMS3, which lacks the Leu allosteric domain, was 14-fold higher for the classic IPMS substrate OIV than for the MAM substrate MTOB (Table I). In contrast, the full-length SlIPMS1 did not exhibit detectable activity with MTOB.

View this table:
Table I. Kinetic parameters for SlIPMS1, SlIPMS3, and SpIPMS3

All values are means ± sd with three replicates. ND, No detectable activity.

We also tested the hypothesis that the absence of the C-terminal domain abolishes Leu-mediated inhibition of SlIPMS3 enzyme activity. As shown in Figure 4, while the activity of SlIPMS1 declined to 50% at 200 μm Leu and was reduced to 30% at higher concentrations, SlIPMS3 activity was insensitive to Leu concentrations up to the highest concentration measured (10 mm). These results are consistent with the hypothesis that SlIPMS3 encodes an IPMS enzyme that is not subject to Leu inhibition.

Figure 4.
Figure 4.

SlIPMS3 activity is insensitive to Leu. SlIPMS1 and SlIPMS3 activities are presented as percentages of the enzyme activity in the absence of Leu. Data points represent means ± se (n = 3).

The shift from high iC5- to iC4-containing acylsugars in IL8-1-1 suggested that the S. pennellii LA0716 allele has reduced production of iC5-CoA due to a defect in Leu biosynthesis (Schilmiller et al., 2010). We asked whether the longer predicted C-terminal deletion in SpIPMS3, including the loss of subdomain II (Figs. 3 and 5A), impaired IPMS activity, as recently reported for Leptospira biflexa IPMS (Zhang et al., 2014). Recombinant SpIPMS3 protein was generated with a C-terminal glutathione S-transferase (GST) tag, and its activity was compared with the GST-tagged SlIPMS3, which retains subdomain II. In contrast to the enzymatically active SlIPMS3 C-terminal GST fusion, no IPMS activity was seen with SpIPMS3 fusion protein (Table I). These results are consistent with the hypothesis that SpIPMS3 is compromised in IPMS activity.

Figure 5.
Figure 5.

Y2H analysis of IPMS and MAM3 proteins. A, Structural domains of SlIPMS1, SlIPMS3, SpIPMS3, and AtMAM3. The start of the deletion in SpIPMS3 compared with SlIPMS3 is indicated by the arrow. Numbers denote amino acid positions. B, SpIPMS3 is unable to dimerize in the Y2H assay. The interactions were tested by comparing yeast growth on selective (SD/-Leu-Trp-His with 3-amino-1,2,4-triazole [3AT]) and nonselective (SD/-Leu-Trp) synthetic dropout medium. Empty signifies an empty GAL4 activation domain (AD) or GAL4 DNA-binding domain (BD) vector. Dilutions from the same starting yeast cultures are indicated at the bottom.

Arabidopsis IPMS behaves as a dimer or tetramer both in vivo and in vitro (Textor et al., 2004; de Kraker and Gershenzon, 2011). The lack of detectable in vitro IPMS activity led us to ask whether SpIPMS3 has altered ability to dimerize in the Y2H system. Full-length IPMS3 coding sequences were fused to the GAL4 DNA-binding domain or the GAL4 activation domain and coexpressed in Saccharomyces cerevisiae. As shown in Figure 5B, SlIPMS3 gave a positive result, demonstrating homodimerization. In contrast, no evidence was obtained for SpIPMS3 homodimerization or for interaction between SlIPMS3 and SpIPMS3 (Fig. 5B). The positive control SlIPMS1 and negative control AtMAM3 showed the expected results in the Y2H system. Taken together, the Y2H results suggested that the SpIPMS3 variant was unable to form oligomers.

We tested the hypothesis that the trichome-expressed SlIPMS3 does not function in Leu biosynthesis by measuring free BCAA levels in trichomes. No differences in Ile, Leu, or Val were detected in M82 (with an enzymatically active SlIPMS3) compared with IL8-1-1 (with the short form SpIPMS3). Taken together with the in vitro assay and Y2H data, these results are consistent with the hypothesis that SlIPMS3 functions in acylsugar biosynthesis and does not directly impact BCAAs (Supplemental Table S4).

Genetic Manipulation of IPMS3 Alters Acylsugar Profiles in S. lycopersicum in Unexpected Ways

To test the hypothesis that SlIPMS3 influences acylsugar composition in vivo, an SlIPMS3 gene-specific fragment was used for RNA interference (RNAi) expression in M82 transgenic plants, and the resulting lines were screened for acylsugar profiles. SlIPMS3 RNAi lines were recovered that exhibited substantial reduction in levels of iC5 acyl chain-containing acylsugars and increased iC4 acyl chain-containing acylsugars compared with M82 (Fig. 6A). Notably, transgenic plants accumulated two acylsugars, S3:20 (4,4,12) and S4:22 (2,4,4,12), that were detected in IL8-1-1 but were near the detection limit in M82 (Fig. 6B; Supplemental Fig. S6 shows the mass spectra of peaks labeled in Fig. 6). This change from predominantly iC5-containing acylsugars to those with increased iC4 acyl chains argues that the accumulation of acylsugars with high iC5 requires active SlIPMS3. Taken together with the lack of in vitro SpIPMS3 activity, this result predicts that the SlIPMS3 allele is dominant to SpIPMS3.

Figure 6.
Figure 6.

Liquid chromatography-mass spectrometry (LC-MS) analysis of acylsugars in M82 transformed with an SlIPMS3 RNAi construct. A, Quantification of acylsugar peak areas showing that IL8-1-1 and M82 transformed with an SlIPMS3 RNAi construct have similar acylsugar profiles. Acylsugar abundance is shown as integrated extracted ion chromatogram peak areas normalized to the internal standard (IS) and dry weight (dw) of the extracted leaflet. The four most abundant acylsugars are labeled with triangles. The number and length of acyl chains on each acylsugar were determined based on collision-induced dissociation mass spectrometry in negative ionization mode (Schilmiller et al., 2010; Ghosh et al., 2014). R16 and R22 indicate independent T1 transgenic lines. Each bar represents the fold difference of the acylsugar abundance compared with M82 (mean ± se; n = 4). B, LC-MS extracted ion chromatograms of leaf dip extracts of M82, IL8-1-1, and the SlIPMS3 RNAi R22 T1 transgenic line. Extracted ion chromatograms display specific mass-to-charge ratios (m/z) of 709 [formate adducts of S3:20 (5,5,10) and S3:20 (4,4,12)] and 751 [formate adducts of S4:22 (2,5,5,10) and S4:22 (2,4,4,12)].

A contradictory result was obtained when acylsugars were analyzed from M82 × IL8-1-1 F1 heterozygous plants, which contain one copy of SlIPMS3 and one copy of SpIPMS3: instead of complete reversion to the M82 phenotype of high iC5 and low iC4 expected for a recessive loss-of-function mutation, a partial conversion was observed (Fig. 7A, left). Peak areas of iC4 acyl chain-containing acylsugars were approximately 3-fold higher in the F1 plants than in M82, but no changes in iC5 acyl chain-containing acylsugars were observed. This result suggests that expression of the SpIPMS3 protein, which is inactive in vitro, influences acyl chain composition even in the presence of active SlIPMS3 enzyme.

Figure 7.
Figure 7.

Acylsugar levels of IPMS3 variant plants. A, Quantification of acylsugar extracted ion chromatogram peak areas from LC-MS analysis showing that M82 × IL8-1-1 F1 and M82 transformed with ProSlASAT4:SpIPMS3 T1 lines (3, 9, and 17) have similar acylsugar profiles. Acylsugar abundance is shown as integrated extracted ion chromatogram peak areas normalized to the internal standard (IS) and dry weight (dw) of the extracted leaflet. The four most abundant acylsugars are labeled with triangles. Each bar represents the fold difference of the acylsugar abundance compared with M82 (mean ± se; n = 6 for F1, n = 3 for transgenic plants). Asterisks indicate significant differences relative to M82 (*, P < 0.05; **, P < 0.01; and ***, P < 0.001, Student’s t test). B, Quantification of acylsugar peak areas showing IL8-1-1 transformed with ProSlASAT4:SlIPMS3. Total acylsugar peak areas are shown for each T0 independent transgenic line, M82 and IL8-1-1. Acylsugar abundance is shown as integrated peak areas normalized to the internal standard and dry weight of the extracted leaflet.

To test this hypothesis, IL8-1-1 was transformed with SlIPMS3 driven by the type I/IV apical cell-specific SlASAT4 promoter (referred to as SlAT2 in Schilmiller et al., 2012; ProSlASAT4:SlIPMS3). We predicted that this transgene would increase iC5 acyl chain-containing acylsugars and decrease iC4 acyl chain-containing products. As predicted, 12 of the 16 independent T0 primary transgenic lines had total iC5 acyl chain-containing acylsugar levels modestly higher than measured in IL8-1-1. However, the expected decrease in iC4 acyl chain-containing acylsugars was not observed in these 12 lines (Fig. 7B; Supplemental Fig. S7). Thus, expression of the enzymatically active SlIPMS3 enzyme did not fully complement the SpIMPS3 genotype.

This lack of full complementation could be due to poor expression of the transgene or a negative effect of the SpIPMS3 protein on SlIPMS3 activity. Evidence for the latter hypothesis was obtained from transformation of S. lycopersicum M82 with the SpIPMS3 cDNA driven by the SlASAT4 promoter (ProSlASAT4:SpIPMS3). Transgenic lines were obtained with increased iC4-containing acylsugars compared with the M82 parent (Fig. 7A, right). This phenotype is reminiscent of that of the M82 × IL8-1-1 F1 plants. The combination of phenotypic results from the M82 × IL8-1-1 F1 and transgenic plants is consistent with the hypothesis that the SpIPMS3 allele influences the acylsugar phenotype in the presence of the enzymatically active SlIPMS3 by a currently unknown mechanism.

The Two IPMS3 Alleles Correlate with iC4/iC5 Acyl Chain Variation in S. pennellii Accessions

Acylsugar acyl chain composition varies among different accessions of wild tomato species (Shapiro et al., 1994; Kim et al., 2012). For example, some southern S. pennellii accessions, including LA0716, contained predominantly iC4 acyl chains, while other accessions had higher levels of iC5 (Shapiro et al., 1994). To test whether the two IPMS3 alleles found in IL8-1-1 and M82 are associated with the iC4/iC5 acyl chain variation in these native plants, acylsugar phenotypes and IPMS3 alleles were analyzed for 14 S. pennellii accessions collected across Peru (Fig. 8; Supplemental Fig. S8; Supplemental Table S5). In addition to LA0716, three other accessions (LA2963, LA1941, and LA1674) were found to be homozygous for the SpIPMS3-like short-form allele (labeled S/S in Fig. 8). Each of these accessions exhibited high levels of acylsugars containing iC4 acyl chains. In contrast, four accessions that were homozygous for the SlIPMS3-like long-form genotype (LA1809, LA2657, LA2560, and LA1649; labeled L/L in Fig. 8) accumulated predominantly iC5 acyl chains. The remaining six accessions (LA1356, LA1376, LA1282, LA1367, LA1911, and LA1946) were heterozygous SlIPMS3-like/SpIPMS3-like (labeled L/S in Fig. 8). Five of the heterozygotes showed a high iC5-iC4 ratio similar to that of the SlIPMS3-like homozygous plants. One of the heterozygous accessions (LA1946) had increased iC5 compared with SpIPMS3-like/SpIPMS3-like homozygotes, but the iC5-iC4 ratio was intermediate to that of the homozygotes. These results show that the natural IPMS3 allelic pattern was correlated with iC5/iC4 acyl chain variation and related to a specific geographical distribution.

Figure 8.
Figure 8.

Relationship of geography, acylsugar acyl chain profiles, and IPMS3 genotype of S. pennellii accessions. The acyl chains on acylsugars extracted from leaf dip samples were transesterified to the corresponding ethyl esters and analyzed by gas chromatography-mass spectrometry (GC-MS) as described in “Materials and Methods.” The mean fractional acyl chain composition of each accession is indicated by the segments within each circle. The accessions that are homozygous for the SlIPMS3-like alleles are depicted as L/L (L refers to long), the accessions that are homozygous for the SpIPMS3-like alleles are represented as S/S (S refers to short), and SlIPMS3-like/SpIPMS3-like heterozygotes are shown as L/S.

DISCUSSION

BCAA metabolism has a central role in producing a wide variety of alkyl-acylated plant specialized metabolites (Hirosawa et al., 1995; Pérez et al., 2002; Karppinen et al., 2007; Wang et al., 2008; Clark et al., 2013; Zabalza et al., 2013), making it of interest to understand the various regulatory mechanisms that control flux from amino acid metabolism into specialized metabolite production. We describe the characterization of SlIPMS3, a gene encoding a variant IPMS enzyme that is expressed in the apical cells of type I/IV trichomes in the cultivated tomato S. lycopersicum. The gene produces an IPMS enzyme that is active in vitro and exhibits insensitivity to Leu-mediated allosteric inhibition due to an approximately 160-amino acid deletion at the C terminus. In contrast, the S. pennellii LA0716 gene SpIPMS3 encodes a protein that has a 51-amino acid longer C-terminal truncation, which renders the protein unable to use OIV and acetyl-CoA as substrates when assayed in vitro. The SpIPMS3 allele in the homozygous state (IL8-1-1 and several BILs) changes the pattern of acylsucroses from that typical of S. lycopersicum M82, with abundant iC5 acyl chains derived from Leu metabolism, to iC4 acyl chains derived from Val metabolism (Schilmiller et al., 2010). Of particular note is that the IPMS3 genotypes found in S. pennellii accessions from throughout the species range in Peru show strong association with the relative abundance of iC5 and iC4 acylsugar acyl chains (Fig. 8). These results indicate that variation in IPMS3 structure and function plays an important role in the observed acyl chain diversity in S. pennellii. This study also provides indirect evidence that CoA ester availability influences the types of acyl chains found on acylsucrose products, presumably because both iC4 and iC5 CoA esters can be used by acylsugar acyltransferases (see Supplemental Figures 7C and 7D in Schilmiller et al., 2015).

SlIPMS3 Has Altered Feedback Regulation of Enzyme Activity and Gene Expression

IPMS3 was identified by screening S. lycopersicum M82 × S. pennellii LA0716 chromosome introgression lines; this is the same approach that led to the identification of SlASAT3 and SlASAT4 enzymes, which catalyze the third and fourth steps of acylsucrose biosynthesis in S. lycopersicum (Schilmiller et al., 2012, 2015). Our results revealed that the type I/IV trichome apical cell-expressed SlIPMS3 enzyme catalyzes the first committing step of Leu biosynthesis using OIV and acetyl-CoA as substrates (Table I). In this regard, SlIPMS3 resembles previously identified IPMS enzymes from Arabidopsis, fungi, and bacteria (Roeder and Kohlhaw, 1980; Wiegel, 1981; de Kraker et al., 2007).

However, the SlIPMS3 protein has characteristics reminiscent of the MAM enzymes of Arabidopsis (Textor et al., 2004, 2007), which catalyze the committing step of Met-derived glucosinolate biosynthesis. First, both are missing the C-terminal allosteric regulatory domain and as a result are insensitive to Leu inhibition (Fig. 4). Second, as reported previously for C-terminal deletion alleles of both full-length Arabidopsis IPMS enzymes (de Kraker and Gershenzon, 2011), the naturally occurring SlIPMS3 truncated enzyme exhibits a change in substrate specificity, having acquired detectable MAM activity not seen in the full-length SlIPMS1 enzyme (Table I). Finally, the SpIPMS3 variant enzyme, which is inactive in vitro and fails to give a positive result in the Y2H assay (Fig. 5B), leads to a dramatic shift in acylsugar acyl chain composition in S. pennellii accessions (Supplemental Table S5). This is analogous to the situation with MAM alleles, including loss-of-function variants, which influence the types of Met-derived glucosinolates found in Arabidopsis accessions (Kroymann et al., 2003; Benderoth et al., 2009).

The evolutionary recruitment from a Leu biosynthetic enzyme to one that diverts OIV toward the production of iC5-CoAs for iC5 acylsugar synthesis also involves changes in the developmental regulation of mRNA accumulation. The most compelling evidence for cell type-specific gene expression comes from the analysis of transgenic plants containing the ProSlIPMS3:GFP-GUS transgene (Fig. 1D). These plants exhibit GFP expression in the apical cells of type I/IV trichomes, the long trichomes on stems, petioles, and leaves that are visible to the naked eye (Fig. 1D). That reporter gene expression faithfully represents the expression of the gene in trichomes is corroborated by RNA-Seq (Supplemental Table S2) and qPCR data (Supplemental Fig. S2), which show strong enrichment of the SlIPMS3 transcript in trichome RNA compared with shaved stems. This is in contrast to the other IPMS sequences in M82 cultivated tomato, whose gene expression is not enriched in stem and petiole trichome preparations (Supplemental Table S3). The observed selectivity of expression in type I/IV apical cells matches that seen for the S. lycopersicum ASAT3 and ASAT4. These results strongly suggest that the apical cells are metabolic factories that contain the enzymatic machinery needed to produce acylsucroses from simple products of central metabolism, including acetyl-CoA, Suc, and amino acids.

Evidence That BCAAs Are Precursors to Acylsugar Acyl Chains

The involvement of a variant Leu biosynthetic enzyme in producing acylsugar iC5 side chains provides strong evidence for the proposed role of acyl-CoA in vivo fluxes as determinants of acylsugar composition (Slocombe et al., 2008; Schilmiller et al., 2012, 2015). RNA-Seq data with S. lycopersicum total trichome RNA (Supplemental Table S3) provide additional evidence for this hypothesis (Fig. 1B). Trichome-enriched expression was detected for the entire known pathway for the production of iC5-CoAs from the Leu biosynthetic pathway: IPMI (catalyzing the step after IPMS) and three subunits of the BCKDH complex, which produces iC5-CoA from the product of IPMI (Fig. 1B). These results are consistent with previously published findings indicating the involvement of BCAA metabolism in the production of branched acyl chains in acylsugar biosynthesis (Kandra et al., 1990; Kandra and Wagner, 1990; Walters and Steffens, 1990). For example, virus-induced gene silencing of the E1 β-subunit of the BCKDH complex caused the reduction of acylsugar accumulation in N. benthamiana (Slocombe et al., 2008). Finally, we recently demonstrated the involvement of BAHD acyltransferases and acyl-CoA esters, including iC5-CoA, in acylsucrose biosynthesis (Schilmiller et al., 2012, 2015).

The Influence of IPMS on Acyl Chain Composition in Cultivated and Wild Tomato

One or more copies of the IPMS3 short form originally identified in IL8-1-1 and S. pennellii LA0716 was found in the majority of S. pennellii accessions that we analyzed (labeled S in Fig. 8). The short allele was found exclusively in a total of four accessions, including LA0716, and these were collected from the southern part of the species distribution. In contrast, only the longer SlIPMS3-like allele was identified in four other accessions (labeled L in Fig. 8), with three from the northern part of the distribution and one from the south. Finally, both long and short IPMS3-like sequences were isolated from six accessions collected from the large central range of the species. This pattern of contrasting homozygous genotypes at the geographical extremes of the species, with polymorphic genotypes in the center, is reminiscent of isozyme diversity data published more than 30 years ago by Charles Rick and colleagues for S. pennellii (Rick and Tanksley, 1981) as well as for S. pimpinellifolium (formerly Lycopersicon pimpinellifolium) and S. habrochaites (formerly Lycopersicon hirsutum; Rick et al., 1979). In those studies, polymorphic isozyme markers were identified that showed evidence for contrasting genotypes in the north and south of the geographic range and were more highly polymorphic in the central regions.

The IPMS3 genotypes are well correlated with iC5 and iC4 acylsugar acyl chain phenotypic patterns in wild and cultivated tomato. We found high iC5 and low iC4 acyl chains in trichome extracts of S. pennellii accessions for which only SlIPMS3-like long alleles were recovered (presumably homozygotes) and all but one accession, LA1946, that contains both long and short alleles (putative heterozygotes; Fig. 8). In contrast, iC4 acyl chains were abundant in acylsugars of the four accessions from which only short alleles were recovered (presumptive homozygotes). Taken together, these results support the hypothesis that this locus plays a major role in determining the types of acylsugars found in S. pennellii. Our phenotypic results show good agreement with the work of Shapiro et al. (1994), who reported that the high iC4 phenotype was found for more southerly accessions, including LA0716, LA1674, LA1946, and LA1941. They also found high iC5 acylsugars for the LA1376, LA2560, and LA1809, which we identified as having iC5-dominant acylsugar acyl chains associated with the presence of at least one SlIPMS3-like long-form allele in the genome. The influence of this region of S. pennellii chromosome 8 on S. lycopersicum iC4 versus iC5 composition is also consistent with published quantitative trait locus mapping and breeding studies from Martha Mutschler and colleagues (Blauth et al., 1999; Leckie et al., 2014).

While our results support the hypothesis that IPMS3 influences iC4 and iC5 acylsugar acyl chain accumulation, SpIPMS3 does not behave as a simple loss-of-function mutation in S. lycopersicum. The SlIPMS3-like/SpIPMS3-like heterozygote plants created by crossing IL8-1-1 with M82 have peak areas of iC4 acyl chain-containing acylsugars approximately 3-fold higher than in M82 (Fig. 7A, left). In addition, transgenic introduction of the IL8-1-1 SpIPMS3 (from S. pennellii LA0716) under the control of the type I/IV apical cell-specific SlASAT4 promoter (ProSlASAT4:SpIPMS3) resulted in increased iC4 acylsugar acyl chains (Fig. 7A, right). These genetic results are consistent with the hypothesis that the SpIPMS3 protein can exert a dominant inhibitory effect on SlIPMS3 enzyme activity in vivo. None of the short-form IPMS3 alleles from S. pennellii have accumulated other mutations that would cause the synthesis of more highly truncated forms of the enzyme (Supplemental Fig. S8), suggesting that the short-form SpIPMS3-like protein(s) may have some function. However, we have no data indicating what, if any, biochemical role is played by SpIPMS3 protein. For example, the Y2H assay revealed no evidence for SlIPMS3-SpIPMS3 interaction in the nucleus of yeast (Fig. 5B), and mixing E. coli-expressed extracts of His-SlIPMS3 and SpIPMS3-GST proteins did not result in the inhibition of IPMS enzyme activity. Another possible mechanism to explain the influence of SpIPMS3 is that the enzyme utilizes OIV to make an alternative product, indirectly reducing the production of iC5-CoA esters.

Implications of Multiple Instances of IPMS Recruitment for Plant Specialized Metabolism

Protein sequence comparisons strongly suggest that independent duplication and diversification events led to the evolution of the lineage-specific IPMS variants in the Brassicaceae (MAMs) and Solanaceae (the IPMS3 orthologous group). Our phylogenetic analysis of plant IPMS homologs (Figure 2; Supplemental Fig. S5) supports the hypothesis that the Brassicaceae IPMS gave rise to MAM, as proposed by de Kraker and Gershenzon (2011). Inclusion of the full-length and shorter variants from the Solanaceae in the tree reveals that the C-terminal truncated Solanaceae IPMS3 genes were derived from an IPMS after the last common ancestor of these two families (Supplemental Fig. S2). Multiple instances of IPMS recruitment within eudicots are not surprising, given that loss of the C terminus renders the enzyme free from Leu inhibition and allows it to function independent of amino acid biosynthesis and accumulation. This simple mechanism, combined with the large number of plant specialized metabolites derived from BCAA metabolism, leads us to predict that more examples of independent IPMS recruitment to specialized metabolic pathways will be found.

The rapid diversification of trichome-specialized metabolic enzymes within the genus Solanum (Sallaud et al., 2009; Schilmiller et al., 2009, 2015; Gonzales-Vigil et al., 2012; Kim et al., 2012, 2014; Fig. 8) creates challenges for studies of metabolic pathways using traditional biochemical approaches, such as isotopic labeling studies. For example, surprising differences in trichome acyl chain labeling patterns were reported for different Solanaceae species (van der Hoeven and Steffens, 2000; Kroumova and Wagner, 2003). Kroumova and Wagner (2003) reported evidence that N. benthamiana trichomes use the one-carbon-chain elongation pathway for the elongation of acyl chains. In contrast, they and others reported that acyl chain elongation in S. pennellii LA0716 proceeds by a two-carbon-chain elongation pathway (van der Hoeven and Steffens, 2000; Kroumova and Wagner, 2009). Sequence analysis (Supplemental Fig. S5) revealed a correlation between their labeling patterns and types of trichome-expressed IPMS3. N. benthamiana does one-carbon elongation and has a long-form SlIPMS3-like (which resembles the enzymatically active SlIPMS3), whereas S. pennellii LA0716 does two-carbon elongation and contains only the truncated short form. We hypothesize that the absence of an enzymatically active trichome-expressed IPMS (i.e. SpIPMS3-like short form) disables the normal one-carbon pathway, which presumably operates by the combined activities of IPMS and IPMI.

MATERIALS AND METHODS

BIL Generation and Genotyping

The BIL population was constructed from a cross between the cultivated tomato Solanum lycopersicum M82 and the wild species Solanum pennellii LA0716. Seedlings were grown in Jiffy peat pots (Hummert International) in a controlled growth chamber maintained for 16 h in the light (300 µE m−2 s−1; mixed cool-white and incandescent bulbs) at 28°C and 8 h in the dark at 20°C. To reduce the wild parent genomic representation, all the lines were backcrossed to the cultivated parent for two to three generations. The BILs were then self-pollinated for eight to nine generations, resulting in around 550 lines. The four BILs that include IL8-1-1 were screened for acylsugars. DNA was isolated from leaf tissue, and the population was genotyped with 10K SOLCAP SNP Chip analysis (Sim et al., 2012). Polymorphic markers between the two parents were used for the mapping analysis.

RNA Extraction and Sequencing

Total RNA was extracted from the stems with trichomes, shaved stems, and isolated trichomes of 3-week-old S. lycopersicum M82 plants using the RNAeasy Plant Mini kit (Qiagen). Six samples (three tissues × two replicates) of total RNA were submitted for sequencing in a single lane. Sequencing libraries were prepared using the Illumina TruSeq Stranded mRNA Library Preparation Kit LT. After quality control and quantitation, libraries were pooled and sequenced on the Illumina HiSeq 2500 Rapid Run flow cell (version 1) using Rapid sequencing by synthesis reagents according to the manufacturer’s specifications. Sequencing was performed in a 2- × 100-bp paired-end (PE100) format. Base calling was performed using Illumina Real Time Analysis version 1.17.21.3. The output of Illumina Real Time Analysis was demultiplexed and converted to FastQ files with Illumina Bcl2Fastq version 1.8.4. The RNA-Seq reads were deposited in the National Center for Biotechnology Information Sequence Read Archive under the Bioproject PRJNA271503.

Differential Expression Analyses

Approximately 22 to 30 million 100-bp paired-end read clusters were obtained for each submitted RNA sample. The paired-end reads were filtered using Trimmomatic version 0.32 (Bolger et al., 2014b) with the following settings: ILLUMINACLIP, TruSeq3-PE-2.fa:2:30:10; LEADING, 20; TRAILING, 20; SLIDINGWINDOW, 4:20; MINLEN, 50. Approximately 88% of the reads in each data set passed the quality filter. The trimmed and filtered reads were then mapped to the S. lycopersicum genome version 2.3 with TopHat version 1.4.1 using the following parameters: –library-type, fr-firststrand; -I, 2000; -g, 1; segment-length, 25; –mate-inner-dist, 200; –mate-std-dev, 50. Seventy-four percent to 88% of the filtered reads were found to map to the genome successfully. For differential expression analyses, the TopHat-mapped reads were counted for each gene using HTSeq version 0.6.1 (Anders et al., 2015), and the counts were used to calculate the differential expression for each tissue pair using edgeR (Robinson et al., 2010). Genes were called differentially expressed if log2 fold change was greater than 2 and P, corrected for multiple testing, was less than 0.05 (Storey, 2002).

Determination of Orthology

We determined orthology between the SlIPMS3 and SpIPMS3 copies using reciprocal best match, synteny, and phylogenetic analysis. For reciprocal best match, we performed an all-versus-all BLAST between all annotated S. lycopersicum and S. pennellii proteins (Bolger et al., 2014a) and selected pairs that were reciprocal best hits of each other at E < 1e-10. We determined genome-wide synteny between reciprocal best hits using MCScanX (Wang et al., 2013) using the default parameters. For phylogenetic reconstruction of IPMS and MAM sequences, we combined information from two sources. First, we obtained the MAM and IPMS sequences used in a previous study (Benderoth et al., 2009). Second, we obtained homologous sequences in Nicotiana benthamiana, Capsicum annuum, and Solanum tuberosum using a BLAST search with E < 1e-10. All amino acid sequences were aligned using MUSCLE (Edgar, 2004) and used for phylogenetic reconstruction using a Bayesian approach as implemented in Mr. Bayes version 3.2.1 (Huelsenbeck and Ronquist, 2001) using the procedure described by Benderoth et al. (2009). To determine the robustness of the tree inference, we also generated a maximum likelihood tree using RaxML version 8.0.6 (Stamatakis, 2014) using the approach noted in Supplemental Figure S5.

Analysis of IPMS Gene Expression in Trichomes

Total RNA was isolated from the stems with trichomes, shaved stems, and isolated trichomes as described above. Reverse transcription was performed using SuperScript II reverse transcriptase (Invitrogen) following the manufacturer’s recommendation. qPCR was performed on the 7500 Fast real-time PCR system (Applied Biosystems) using Fast SYBR Green PCR master mix (Applied Biosystems) according to the manufacturer’s protocol. The cultivated tomato EF1α gene was used as the endogenous control. The comparative threshold cycle method was used to evaluate the quantitative variation of replicates examined. The gene-specific primers are listed in Supplemental Table S6.

IPMS Sequence Analysis

Total RNAs from the leaves of S. lycopersicum M82 and S. pennellii LA0716, IL8-1-1, and IL8-1 were extracted with the RNeasy Plant Mini Kit (Qiagen). cDNA was synthesized as described above. The mRNA coding sequence of IPMS3 was amplified from these four plants using primers IPMS3 FL F and IPMS3 FL R (Supplemental Table S6). Genomic DNAs from the leaves of S. lycopersicum M82 and S. pennellii LA0716, IL8-1-1, and IL8-1 were obtained using the DNeasy Plant Mini Kit (Qiagen). The SpIPMS3 deletion was confirmed through sequencing the PCR products amplified from S. pennellii LA0716, IL8-1-1, and IL8-1 genomic DNA using primer pair IPMS3 DEL F and IPMS3 DEL R (Supplemental Table S6). The same fragment from S. lycopersicum M82 was amplified and sequenced as well. All these PCRs were performed using high-fidelity Phusion polymerase (New England Biolabs), and samples were separated by agarose gel electrophoresis, stained with ethidium bromide, and photographed under UV light. The DNA sequence was determined using the ABI PRISM 3730 Genetic Analyzer (Applied Biosystems).

Y2H Analysis

All GAL4 activating domain and GAL4 binding domain fusion-encoding plasmids that correspond to the desired coding sequences were generated by PCR and cloned into the pGADT7 and pGBKT7 vectors using SfiI and BamHI. The primers used for cloning are listed in Supplemental Table S6. The yeast strain Saccharomyces cerevisiae AH109 (Clontech) was cultured and transformed as recommended by the manufacturer using standard synthetic dropout medium (Clontech) as indicated. HIS3 reporter assays were performed according to the manufacturer’s instructions (Clontech). Interactions were selected on synthetic dropout/-Leu-Trp-His medium supplemented with 1.5 mm 3-amino-1,2,4-triazole (Sigma-Aldrich) after 48 h of growth at 30°C.

Protein Expression and Enzyme Assay

To generate His-tagged recombinant proteins, the full-length open reading frames (ORFs) for SlIPMS3, SpIPMS3, and SlIPMS1 were cloned into BamHI and NotI sites of pET28b (Novagen). Because soluble His-tagged SpIPMS3 was not obtained from Escherichia coli, recombinant SpIPMS3 proteins were generated with N-terminal or C-terminal GST tags and compared with the activity of the GST-tagged SlIPMS3. A summary of the protein expression vectors is presented in Supplemental Table S7. To generate N-terminal GST-tagged recombinant proteins, the full-length ORFs for SlIPMS3 and SpIPMS3 were cloned into pGEX-4T-1 (GE) using BamHI and NotI sites. The forward primer used for this construct contains a sequence encoding a 6× His tag. To generate C-terminal GST-tagged recombinant proteins, GST was amplified from pGEX-4T-1 using the primers GST NOT1 F and GST NOT1 R, subsequently digested with NotI, and cloned into the His-tagged SlIPMS3 and His-tagged SpIPMS3 mentioned above. The primers used for cloning are listed in Supplemental Table S6. E. coli strain BL21 Rosetta cells (EMD Millipore) containing the correct construct were grown overnight at 37°C in small Luria-Bertani liquid cultures. Overnight cultures were diluted 50 times into 1 L of Luria-Bertani medium and incubated at 37°C to an optical density at 600 nm of 0.6. Then, the culture was induced with 1 mm isopropyl-β-d-thiogalactopyranoside and incubated at 16°C overnight. The protein was partially purified by nickel-nitrilotriacetic acid agarose (Qiagen) as described previously (de Kraker and Gershenzon, 2011). The purified protein was immediately desalted and concentrated into a 50 mm Tris buffer, pH 8, containing 10 mm MgCl2 and 10% (v/v) glycerol using a 10-kD Centrifugal Filter Unit (Millipore). Kinetic analysis was performed and the effect of Leu on the enzyme activity measurement was determined following the methods described (de Kraker and Gershenzon, 2011). The enzyme assay under these conditions was linear, as determined by assessing the activity at different protein concentrations and time intervals. Three replicates were tested at a saturating concentration of 500 µm acetyl-CoA. Reaction without acetyl-CoA was used as a blank.

Metabolite Extraction and Analyses

Leaflets from the next-to-youngest leaf of 3-week-old plants were used for chemical analysis of acylsugar. One leaflet was dipped in 1 mL of isopropanol:acetonitrile:water (3:3:2, v/v/v) containing 0.1% formic acid by volume and 10 μm propyl-4-hydroxybenzoate that served as an internal standard, with gentle rocking for 2 min. Dry weights of the extracted leaf samples were measured following drying at 50°C. The profiling of acylsugar was performed using liquid chromatography-time of flight-mass spectrometry on a Waters Xevo G2-S QToF instrument as described before (Schilmiller et al., 2015). Nonselective collision-induced dissociation mass spectrometry in negative ionization mode provided information about the number and length of acyl chains present on each acylsugar based on the masses of carboxylate fragment ions (Schilmiller et al., 2010; Ghosh et al., 2014).

To analyze acyl chains on acylsugars, a leaflet was dipped in 1 mL of acetonitrile for 2 min. Each supernatant was transferred to a vial and evaporated to dryness using a SpeedVac. To convert acyl groups to ethyl esters, 300 µL of 21% (v/v) sodium ethoxide in ethanol was added, and the reaction was allowed to proceed for 30 min in a fume hood with gentle vortexing every 5 min. A total of 400 μL of hexane was added to the reaction mixture followed by vortexing. The hexane layer was removed and extracted using saturated aqueous sodium chloride (500 µL, three times). Fifty microliters of the hexane solution was transferred to a gas chromatography autosampler vial and analyzed by capillary GC-MS on a 10-m, 0.1-mm (i.d.) fused silica column with a 0.34-µm-thick stationary phase (DB-5; Agilent). Injection of 1 µL of each hexane extract was performed using splitless mode. The gas chromatography program was as follows: injector temperature, 280°C; initial column temperature, 36°C held for 4 min; ramped at 10°C min−1 to 150°C, at 20°C min−1 to 220°C, and finally at 40°C min−1 to 300°C; and then held for 2 min. The helium carrier gas flow was set to 0.4 mL min−1. All compounds were analyzed using an Agilent 6890N gas chromatograph/Agilent 5975B quadrupole mass spectrometer using 70 eV electron ionization. The following extracted ion chromatograms were integrated to measure relative abundances of ethyl esters of the corresponding fatty acids: m/z 71 for iC4, m/z 57 for 2-methylbutanoic acid, and m/z 88 for iC5, C6, iC10, nC10, C11, and nC12.

Generation of Transgenic Plants and Phenotypic Analyses

Agrobacterium tumefaciens-mediated transformation of cotyledon explants was performed using a published method (McCormick, 1991). To generate RNAi lines, an SlIPMS3-specific fragment of 660 bp, located in the 5′ region of SlIPMS3, was amplified from M82 cDNA using the primers SLIPMS3 RNAi F and SLIPMS3 RNAi R. The purified PCR product was introduced into the entry vector pENTR/D-TOPO (Invitrogen) followed by recombination into pHELLSGATE12 (Helliwell and Waterhouse, 2003). The construct was transformed into M82 plants.

To express SpIPMS3 under the control of the SlASAT4 promoter (ProSlASAT4:SpIPMS3) in M82 plants, the full-length SpIPMS3 ORF was amplified from IL8-1-1 cDNA using the primers IPMS3 FL F and IPMS3 FL R. The purified PCR product was introduced into the entry vector pENTR/D-TOPO (Invitrogen) followed by recombination into modified pK7WG2 (Kim et al., 2014). Acylsugars from T1 transgenic plants were analyzed using the single leaf dip metabolite extraction and negative ionization mode LC-MS analysis as described above. We confirmed the presence of the transgene in the T1 generation with PCR, and the acylsugar phenotypes were consistent with the PCR result. To express SlIPMS3 under the control of the SlASAT4 promoter (ProSlASAT4:SlIPMS3) in IL8-1-1 plants, the full-length ORF of SlIPMS3 was amplified from M82 cDNA using the primers IPMS3 FL F and IPMS3 FL R. The purified PCR product was introduced into the entry vector pENTR/D-TOPO (Invitrogen) followed by recombination into modified pK7WG2 (Kim et al., 2014). Acylsugars from T0 transgenic plants were analyzed using the single leaf dip metabolite extraction and negative ionization mode LC-MS analysis as described above. For in planta reporter gene analysis of SlIPMS3 expression, a 1,436-bp SlIPMS3 promoter was amplified from M82 genomic DNA using the primers SLIPMS3 PRO F and SLIPMS3 PRO R and cloned into pCR8/GW/TOPO followed by recombination into pKGWFS7 (Karimi et al., 2007). The resulting construct was used to transform M82 plants and drive the expression of a GFP-GUS fusion protein under the control of the SlIPMS3 promoter (ProSlIPMS3:GFP-GUS). GFP fluorescence was analyzed in multiple primary transformants as described previously (Schilmiller et al., 2012). The primers used for cloning are listed in Supplemental Table S6.

S. pennellii Accession Acyl Chain Profiling and IPMS3 Sequencing

Accessions of S. pennellii available from the Tomato Genetic Resource Center (http://tgrc.ucdavis.edu/) were selected that covered the geographical range of the species. Latitudinal and longitudinal information was obtained from the Tomato Genetic Resource Center. Locations were mapped using Google Maps. Leaf dip extracts were taken from 4-week-old plants, and the acyl chains on acylsugars were analyzed by GC-MS as described above. The SpIPMS3-like allele deletion was confirmed through sequencing the PCR products amplified from S. pennellii genomic DNA using primers IPMS3 DEL F and IPMS3 DEL R. Full-length SpIPMS3-like alleles were amplified from S. pennellii cDNA using the IPMS3 FL F and IPMS3 SR primers. Full-length SlIPMS3-like alleles were amplified from S. pennellii cDNA using the IPMS3 FL F and IPMS3 LR primers. The primers used for PCR are listed in Supplemental Table S6. The DNA sequence was determined using the ABI PRISM 3730 Genetic Analyzer (Applied Biosystems).

Sequence data from this article can be found in the GenBank database under the following accession numbers: SlIPMS3 (KP863470), SpIPMS3 (KP863471), SlIPMS2 (KP863472), SlIPMS1 (KP885645), SpIPMS3-LA1282-L (KP885646), SpIPMS3-LA1282-S (KP885656), SpIPMS3-LA1356-L (KP885647), SpIPMS3-LA1367-L (KP885648), SpIPMS3-LA1367-S (KP885657), SpIPMS3-LA1376-L (KP885649), SpIPMS3-LA1376-S (KP885658), SpIPMS3-LA1649-L (KP885650), SpIPMS3-LA1674-S (KP885659), SpIPMS3-LA1809-L (KP885651), SpIPMS3-LA1911-L (KP885652), SpIPMS3-LA1941-S (KP885660), SpIPMS3-LA1946-L (KP885653), SpIPMS3-LA1946-S (KP885661), SpIPMS3-LA2560-L (KP885654), SpIPMS3-LA2657-L (KP885655), and SpIPMS3-LA2963-S (KP885662).

Supplemental Data

The following supplemental materials are available.

Acknowledgments

We thank trichome project members for their contributions, especially Anthony Schilmiller, Xiaoxiao Liu, and Pengxiang Fan for helpful comments on the article; Yuki Matsuba and Matthew Bedewitz for suggestions on enzyme assays; Cheng Peng for analyzing trichome BCAA LC-MS data; Kathleen Imre for help with cultivated tomato transformation; the Michigan State University Center for Advanced Microscopy and Research Technology Support Facility Mass Spectrometry and Metabolomics Core Facilities; and Alisdair Fernie and Bjoern Usadel for early access to the S. pennellii LA0716 genomic sequence. IL8-1-1 and S. pennellii seeds were obtained from the C.M. Rick Tomato Genetics Resource Center (University of California, Davis).

Footnotes

  • www.plantphysiol.org/cgi/doi/10.1104/pp.15.00474

  • The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Robert L. Last (lastr{at}msu.edu).

  • J.N., G.D.M., B.L., J.K., and R.L.L. designed the research; J.N., G.D.M., B.L., Z.W., and C.A. performed the research; J.N., G.D.M., B.L., Z.W., I.O., A.D.J., D.Z., and R.L.L. analyzed the data; J.N., G.D.M., and R.L.L. wrote the article.

  • ↵1 This work was supported by the National Science Foundation (grant no. IOS–1025636 to R.L.L. and A.D.J. and PlantGenomics at Michigan State University Summer REU Program grant no. DBI–1004425 to C.A.), the National Institute of General Medical Sciences of the National Institutes of Health (grant no. T32–GM110523 to B.L.), the Michigan AgBioResearch Project (grant no. MICL02143 to A.D.J.), and the European Research Council (YIELD grant to D.Z.).

  • ↵[OPEN] Articles can be viewed without a subscription.

Glossary

OIV
2-oxoisovalerate
iC4
2-methylpropanoic acid
iC5
3-methylbutanoic acid
BCAA
branched-chain amino acid
BIL
backcross inbred line
RNA-Seq
RNA sequencing
cDNA
complementary DNA
qPCR
quantitative PCR
MTOB
4-methylthio-2-oxobutyrate
GST
glutathione S-transferase
Y2H
yeast two-hybrid
RNAi
RNA interference
ORF
open reading frame
GC-MS
gas chromatography-mass spectrometry
m/z
mass-to-charge ratio
  • Received March 27, 2015.
  • Accepted May 15, 2015.
  • Published May 18, 2015.

REFERENCES

  1. ↵
    1. Anders S,
    2. Pyl PT,
    3. Huber W
    (2015) HTSeq: a Python framework to work with high-throughput sequencing data. Bioinformatics 31: 166–169
    OpenUrlAbstract/FREE Full Text
  2. ↵
    1. Benderoth M,
    2. Pfalz M,
    3. Kroymann J
    (2009) Methylthioalkylmalate synthases: genetics, ecology and evolution. Phytochem Rev 8: 255–268
    OpenUrlCrossRef
  3. ↵
    1. Blauth SL,
    2. Steffens JC,
    3. Churchill GA,
    4. Mutschler MA
    (1999) Identification of QTLs controlling acylsugar fatty acid composition in an intraspecific population of Lycopersicon pennellii (Corr.) D’Arcy. Theor Appl Genet 99: 373–381
    OpenUrlCrossRef
  4. ↵
    1. Bohlmann J,
    2. DeLuca V,
    3. Eilert U,
    4. Martin W
    (1995) Purification and cDNA cloning of anthranilate synthase from Ruta graveolens: modes of expression and properties of native and recombinant enzymes. Plant J 7: 491–501
    OpenUrlCrossRefPubMed
  5. ↵
    1. Bohlmann J,
    2. Lins T,
    3. Martin W,
    4. Eilert U
    (1996) Anthranilate synthase from Ruta graveolens: duplicated AS alpha genes encode tryptophan-sensitive and tryptophan-insensitive isoenzymes specific to amino acid and alkaloid biosynthesis. Plant Physiol 111: 507–514
    OpenUrlAbstract
  6. ↵
    1. Bolger A,
    2. Scossa F,
    3. Bolger ME,
    4. Lanz C,
    5. Maumus F,
    6. Tohge T,
    7. Quesneville H,
    8. Alseekh S,
    9. Sørensen I,
    10. Lichtenstein G, et al.
    (2014a) The genome of the stress-tolerant wild tomato species Solanum pennellii. Nat Genet 46: 1034–1038
    OpenUrlCrossRefPubMed
  7. ↵
    1. Bolger AM,
    2. Lohse M,
    3. Usadel B
    (2014b) Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30: 2114–2120
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Bombarely A,
    2. Rosli HG,
    3. Vrebalov J,
    4. Moffett P,
    5. Mueller LA,
    6. Martin GB
    (2012) A draft genome sequence of Nicotiana benthamiana to enhance molecular plant-microbe biology research. Mol Plant Microbe Interact 25: 1523–1530
    OpenUrlCrossRefPubMed
  9. ↵
    1. Burke BA,
    2. Goldsby G,
    3. Mudd JB
    (1987) Polar epicuticular lipids of Lycopersicon pennellii. Phytochemistry 26: 2567–2571
    OpenUrlCrossRef
  10. ↵
    1. Chitwood DH,
    2. Kumar R,
    3. Headland LR,
    4. Ranjan A,
    5. Covington MF,
    6. Ichihashi Y,
    7. Fulop D,
    8. Jiménez-Gómez JM,
    9. Peng J,
    10. Maloof JN, et al.
    (2013) A quantitative genetic basis for leaf morphology in a set of precisely defined tomato introgression lines. Plant Cell 25: 2465–2481
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Chortyk OT,
    2. Kays SJ,
    3. Teng Q
    (1997) Characterization of insecticidal sugar esters of Petunia. J Agric Food Chem 45: 270–275
    OpenUrlCrossRef
  12. ↵
    1. Clark SM,
    2. Vaitheeswaran V,
    3. Ambrose SJ,
    4. Purves RW,
    5. Page JE
    (2013) Transcriptome analysis of bitter acid biosynthesis and precursor pathways in hop (Humulus lupulus). BMC Plant Biol 13: 12
    OpenUrlCrossRefPubMed
  13. ↵
    1. de Carvalho LPS,
    2. Argyrou A,
    3. Blanchard JS
    (2005) Slow-onset feedback inhibition: inhibition of Mycobacterium tuberculosis alpha-isopropylmalate synthase by L-leucine. J Am Chem Soc 127: 10004–10005
    OpenUrlCrossRefPubMed
  14. ↵
    1. de Kraker JW,
    2. Gershenzon J
    (2011) From amino acid to glucosinolate biosynthesis: protein sequence changes in the evolution of methylthioalkylmalate synthase in Arabidopsis. Plant Cell 23: 38–53
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. de Kraker JW,
    2. Luck K,
    3. Textor S,
    4. Tokuhisa JG,
    5. Gershenzon J
    (2007) Two Arabidopsis genes (IPMS1 and IPMS2) encode isopropylmalate synthase, the branchpoint step in the biosynthesis of leucine. Plant Physiol 143: 970–986
    OpenUrlAbstract/FREE Full Text
  16. ↵
    1. Edgar RC
    (2004) MUSCLE: a multiple sequence alignment method with reduced time and space complexity. BMC Bioinformatics 5: 113
    OpenUrlCrossRefPubMed
  17. ↵
    1. Eshed Y,
    2. Zamir D
    (1995) An introgression line population of Lycopersicon pennellii in the cultivated tomato enables the identification and fine mapping of yield-associated QTL. Genetics 141: 1147–1162
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Fobes JF,
    2. Mudd JB,
    3. Marsden MP
    (1985) Epicuticular lipid accumulation on the leaves of Lycopersicon pennellii (Corr.) D’Arcy and Lycopersicon esculentum Mill. Plant Physiol 77: 567–570
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Frey M,
    2. Chomet P,
    3. Glawischnig E,
    4. Stettner C,
    5. Grün S,
    6. Winklmair A,
    7. Eisenreich W,
    8. Bacher A,
    9. Meeley RB,
    10. Briggs SP, et al.
    (1997) Analysis of a chemical plant defense mechanism in grasses. Science 277: 696–699
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Ghosh B,
    2. Westbrook TC,
    3. Jones AD
    (2014) Comparative structural profiling of trichome specialized metabolites in tomato (Solanum lycopersicum) and S. habrochaites: acylsugar profiles revealed by UHPLC/MS and NMR. Metabolomics 10: 496–507
    OpenUrlCrossRefPubMed
  21. ↵
    1. Gierl A,
    2. Frey M
    (2001) Evolution of benzoxazinone biosynthesis and indole production in maize. Planta 213: 493–498
    OpenUrlCrossRefPubMed
  22. ↵
    1. Gonzales-Vigil E,
    2. Hufnagel DE,
    3. Kim J,
    4. Last RL,
    5. Barry CS
    (2012) Evolution of TPS20-related terpene synthases influences chemical diversity in the glandular trichomes of the wild tomato relative Solanum habrochaites. Plant J 71: 921–935
    OpenUrlCrossRefPubMed
  23. ↵
    1. Helliwell C,
    2. Waterhouse P
    (2003) Constructs and methods for high-throughput gene silencing in plants. Methods 30: 289–295
    OpenUrlCrossRefPubMed
  24. ↵
    1. Hirosawa T,
    2. Saito T,
    3. Tanaka T,
    4. Matsushima H
    (1995) SEM observation and HPLC analysis of the accumulation of alpha-acids and beta-acids in the fresh developing hop (Humulus lupulus L) peltate glandular trichomes. J Electron Microsc (Tokyo) 44: 145–147
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Huelsenbeck JP,
    2. Ronquist F
    (2001) MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17: 754–755
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Juvik JA,
    2. Shapiro JA,
    3. Young TE,
    4. Mutschler MA
    (1994) Acylglucoses from wild tomatoes alter behavior and reduce growth and survival of Helicoverpa zea and Spodoptera exigua (Lepidoptera, Noctuidae). J Econ Entomol 87: 482–492
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Kandra G,
    2. Severson R,
    3. Wagner GJ
    (1990) Modified branched-chain amino acid pathways give rise to acyl acids of sucrose esters exuded from tobacco leaf trichomes. Eur J Biochem 188: 385–391
    OpenUrlPubMed
  28. ↵
    1. Kandra L,
    2. Wagner GJ
    (1990) Chlorsulfuron modifies biosynthesis of acyl acid substituents of sucrose esters secreted by tobacco trichomes. Plant Physiol 94: 906–912
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Karimi M,
    2. Depicker A,
    3. Hilson P
    (2007) Recombinational cloning with plant Gateway vectors. Plant Physiol 145: 1144–1154
    OpenUrlFREE Full Text
  30. ↵
    1. Karppinen K,
    2. Hokkanen J,
    3. Tolonen A,
    4. Mattila S,
    5. Hohtola A
    (2007) Biosynthesis of hyperforin and adhyperforin from amino acid precursors in shoot cultures of Hypericum perforatum. Phytochemistry 68: 1038–1045
    OpenUrlCrossRefPubMed
  31. ↵
    1. Kim J,
    2. Kang K,
    3. Gonzales-Vigil E,
    4. Shi F,
    5. Jones AD,
    6. Barry CS,
    7. Last RL
    (2012) Striking natural diversity in glandular trichome acylsugar composition is shaped by variation at the Acyltransferase2 locus in the wild tomato Solanum habrochaites. Plant Physiol 160: 1854–1870
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Kim J,
    2. Matsuba Y,
    3. Ning J,
    4. Schilmiller AL,
    5. Hammar D,
    6. Jones AD,
    7. Pichersky E,
    8. Last RL
    (2014) Analysis of natural and induced variation in tomato glandular trichome flavonoids identifies a gene not present in the reference genome. Plant Cell 26: 3272–3285
    OpenUrlAbstract/FREE Full Text
  33. ↵
    1. King RR,
    2. Calhoun LA,
    3. Singh RP,
    4. Boucher A
    (1990) Sucrose esters associated with glandular trichomes of wild Lycopersicon species. Phytochemistry 29: 2115–2118
    OpenUrlCrossRef
  34. ↵
    1. Koon N,
    2. Squire CJ,
    3. Baker EN
    (2004) Crystal structure of LeuA from Mycobacterium tuberculosis, a key enzyme in leucine biosynthesis. Proc Natl Acad Sci USA 101: 8295–8300
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Kroumova AB,
    2. Wagner GJ
    (2003) Different elongation pathways in the biosynthesis of acyl groups of trichome exudate sugar esters from various solanaceous plants. Planta 216: 1013–1021
    OpenUrlPubMed
  36. ↵
    1. Kroumova AB,
    2. Wagner GJ
    (2009) Pathways for synthesis, and possibilities for genetic modification of sugar ester acyl groups produced by trichomes of solanaceous species. Gen Appl Plant Physiol 35: 95–110
    OpenUrl
  37. ↵
    1. Kroymann J,
    2. Donnerhacke S,
    3. Schnabelrauch D,
    4. Mitchell-Olds T
    (2003) Evolutionary dynamics of an Arabidopsis insect resistance quantitative trait locus. Proc Natl Acad Sci USA (Suppl 2) 100: 14587–14592
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Kroymann J,
    2. Textor S,
    3. Tokuhisa JG,
    4. Falk KL,
    5. Bartram S,
    6. Gershenzon J,
    7. Mitchell-Olds T
    (2001) A gene controlling variation in Arabidopsis glucosinolate composition is part of the methionine chain elongation pathway. Plant Physiol 127: 1077–1088
    OpenUrlAbstract/FREE Full Text
  39. ↵
    1. Leckie BM,
    2. De Jong DM,
    3. Mutschler MA
    (2012) Quantitative trait loci increasing acylsugars in tomato breeding lines and their impacts on silverleaf whiteflies. Mol Breed 30: 1621–1634
    OpenUrlCrossRef
  40. ↵
    1. Leckie BM,
    2. Halitschke R,
    3. De Jong DM,
    4. Smeda JR,
    5. Kessler A,
    6. Mutschler MA
    (2014) Quantitative trait loci regulating the fatty acid profile of acylsugars in tomato. Mol Breed 34: 1201–1213
    OpenUrlCrossRef
  41. ↵
    1. Maluf WR,
    2. Maciel GM,
    3. Gomes LAA,
    4. Cardoso MD,
    5. Goncalves LD,
    6. da Silva EC,
    7. Knapp M
    (2010) Broad-spectrum arthropod resistance in hybrids between high- and low-acylsugar tomato lines. Crop Sci 50: 439–450
    OpenUrlCrossRef
  42. ↵
    1. McCormick S
    (1991) Transformation of tomato with Agrobacterium tumefaciens. Plant Tissue Culture Manual B6: 1–9
    OpenUrl
  43. ↵
    1. McDowell ET,
    2. Kapteyn J,
    3. Schmidt A,
    4. Li C,
    5. Kang JH,
    6. Descour A,
    7. Shi F,
    8. Larson M,
    9. Schilmiller A,
    10. An L, et al.
    (2011) Comparative functional genomic analysis of Solanum glandular trichome types. Plant Physiol 155: 524–539
    OpenUrlAbstract/FREE Full Text
  44. ↵
    1. Melanson D,
    2. Chilton MD,
    3. Masters-Moore D,
    4. Chilton WS
    (1997) A deletion in an indole synthase gene is responsible for the DIMBOA-deficient phenotype of bxbx maize. Proc Natl Acad Sci USA 94: 13345–13350
    OpenUrlAbstract/FREE Full Text
  45. ↵
    1. Milo R,
    2. Last RL
    (2012) Achieving diversity in the face of constraints: lessons from metabolism. Science 336: 1663–1667
    OpenUrlAbstract/FREE Full Text
  46. ↵
    1. AE Osbourn,
    2. V Lanzotti
    1. Modolo LV,
    2. Reichert AI,
    3. Dixon RA
    (2009) Introduction to the different classes of biosynthetic enzymes. In AE Osbourn, V Lanzotti, eds, Plant-Derived Natural Products: Synthesis, Function, and Application. Springer, New York, pp 143–163
  47. ↵
    1. Moghe GD,
    2. Last RL
    (2015) Something old, something new: conserved enzymes and the evolution of novelty in plant specialized metabolism. Plant Physiol 169: 1512–1523
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. Mueller LA,
    2. Solow TH,
    3. Taylor N,
    4. Skwarecki B,
    5. Buels R,
    6. Binns J,
    7. Lin C,
    8. Wright MH,
    9. Ahrens R,
    10. Wang Y, et al.
    (2005) The SOL Genomics Network: a comparative resource for Solanaceae biology and beyond. Plant Physiol 138: 1310–1317
    OpenUrlAbstract/FREE Full Text
  49. ↵
    1. Pérez AG,
    2. Olías R,
    3. Luaces P,
    4. Sanz C
    (2002) Biosynthesis of strawberry aroma compounds through amino acid metabolism. J Agric Food Chem 50: 4037–4042
    OpenUrlCrossRefPubMed
  50. ↵
    1. Puterka GJ,
    2. Farone W,
    3. Palmer T,
    4. Barrington A
    (2003) Structure-function relationships affecting the insecticidal and miticidal activity of sugar esters. J Econ Entomol 96: 636–644
    OpenUrlAbstract/FREE Full Text
  51. ↵
    1. Rick CM,
    2. Fobes JF,
    3. Tanksley SD
    (1979) Evolution of mating systems in Lycopersicon hirsutum as deduced from genetic variation in electrophoretic and morphological characters. Plant Syst Evol 132: 279–298
    OpenUrlCrossRef
  52. ↵
    1. Rick CM,
    2. Tanksley SD
    (1981) Genetic variation in Solanum pennellii: comparisons with two other sympatric tomato species. Plant Syst Evol 139: 11–45
    OpenUrlCrossRef
  53. ↵
    1. Robinson MD,
    2. McCarthy DJ,
    3. Smyth GK
    (2010) edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26: 139–140
    OpenUrlAbstract/FREE Full Text
  54. ↵
    1. Rodriguez AE,
    2. Tingey WM,
    3. Mutschler MA
    (1993) Acylsugars of Lycopersicon pennellii deter settling and feeding of the green peach aphid (Homoptera, Aphididae). J Econ Entomol 86: 34–39
    OpenUrlAbstract/FREE Full Text
  55. ↵
    1. Roeder PR,
    2. Kohlhaw GB
    (1980) α-Isopropylmalate synthase from yeast: a zinc metalloenzyme. Biochim Biophys Acta 613: 482–487
    OpenUrlCrossRefPubMed
  56. ↵
    1. Sallaud C,
    2. Rontein D,
    3. Onillon S,
    4. Jabès F,
    5. Duffé P,
    6. Giacalone C,
    7. Thoraval S,
    8. Escoffier C,
    9. Herbette G,
    10. Leonhardt N, et al.
    (2009) A novel pathway for sesquiterpene biosynthesis from Z,Z-farnesyl pyrophosphate in the wild tomato Solanum habrochaites. Plant Cell 21: 301–317
    OpenUrlAbstract/FREE Full Text
  57. ↵
    1. Schilmiller A,
    2. Shi F,
    3. Kim J,
    4. Charbonneau AL,
    5. Holmes D,
    6. Jones AD,
    7. Last RL
    (2010) Mass spectrometry screening reveals widespread diversity in trichome specialized metabolites of tomato chromosomal substitution lines. Plant J 62: 391–403
    OpenUrlCrossRefPubMed
  58. ↵
    1. Schilmiller AL,
    2. Charbonneau AL,
    3. Last RL
    (2012) Identification of a BAHD acetyltransferase that produces protective acyl sugars in tomato trichomes. Proc Natl Acad Sci USA 109: 16377–16382
    OpenUrlAbstract/FREE Full Text
  59. ↵
    1. Schilmiller AL,
    2. Moghe GD,
    3. Fan P,
    4. Ghosh B,
    5. Ning J,
    6. Jones AD,
    7. Last RL
    (2015) Functionally divergent alleles and duplicated loci encoding an acyltransferase contribute to acylsugar metabolite diversity in Solanum trichomes. Plant Cell 27: 1002–1017
    OpenUrlAbstract/FREE Full Text
  60. ↵
    1. Schilmiller AL,
    2. Schauvinhold I,
    3. Larson M,
    4. Xu R,
    5. Charbonneau AL,
    6. Schmidt A,
    7. Wilkerson C,
    8. Last RL,
    9. Pichersky E
    (2009) Monoterpenes in the glandular trichomes of tomato are synthesized from a neryl diphosphate precursor rather than geranyl diphosphate. Proc Natl Acad Sci USA 106: 10865–10870
    OpenUrlAbstract/FREE Full Text
  61. ↵
    1. Severson RF,
    2. Arrendale RF,
    3. Chortyk OT,
    4. Green CR,
    5. Thome FA,
    6. Stewart JL,
    7. Johnson AW
    (1985) Isolation and characterization of the sucrose esters of the cuticular waxes of green tobacco leaf. J Agric Food Chem 33: 870–875
    OpenUrlCrossRef
  62. ↵
    1. Shapiro JA,
    2. Steffens JC,
    3. Mutschler MA
    (1994) Acylsugars of the wild tomato Lycopersicon pennellii in relation to geographic-distribution of the species. Biochem Syst Ecol 22: 545–561
    OpenUrlCrossRef
  63. ↵
    1. Sim SC,
    2. Durstewitz G,
    3. Plieske J,
    4. Wieseke R,
    5. Ganal MW,
    6. Van Deynze A,
    7. Hamilton JP,
    8. Buell CR,
    9. Causse M,
    10. Wijeratne S, et al.
    (2012) Development of a large SNP genotyping array and generation of high-density genetic maps in tomato. PLoS ONE 7: e40563
    OpenUrlCrossRefPubMed
  64. ↵
    1. Slocombe SP,
    2. Schauvinhold I,
    3. McQuinn RP,
    4. Besser K,
    5. Welsby NA,
    6. Harper A,
    7. Aziz N,
    8. Li Y,
    9. Larson TR,
    10. Giovannoni J, et al.
    (2008) Transcriptomic and reverse genetic analyses of branched-chain fatty acid and acyl sugar production in Solanum pennellii and Nicotiana benthamiana. Plant Physiol 148: 1830–1846
    OpenUrlAbstract/FREE Full Text
  65. ↵
    1. Soper TS,
    2. Doellgast J,
    3. Kohlhaw GB
    (1976) Mechanism of feedback inhibition by leucine: purification and properties of a feedback-resistant alpha-isopropylmalate synthase. Arch Biochem Biophys 173: 362–374
    OpenUrlCrossRefPubMed
  66. ↵
    1. Stamatakis A
    (2014) RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 30: 1312–1313
    OpenUrlAbstract/FREE Full Text
  67. ↵
    1. Storey JD
    (2002) A direct approach to false discovery rates. J R Stat Soc B 64: 479–498
    OpenUrlCrossRef
  68. ↵
    1. Textor S,
    2. Bartram S,
    3. Kroymann J,
    4. Falk KL,
    5. Hick A,
    6. Pickett JA,
    7. Gershenzon J
    (2004) Biosynthesis of methionine-derived glucosinolates in Arabidopsis thaliana: recombinant expression and characterization of methylthioalkylmalate synthase, the condensing enzyme of the chain-elongation cycle. Planta 218: 1026–1035
    OpenUrlCrossRefPubMed
  69. ↵
    1. Textor S,
    2. de Kraker JW,
    3. Hause B,
    4. Gershenzon J,
    5. Tokuhisa JG
    (2007) MAM3 catalyzes the formation of all aliphatic glucosinolate chain lengths in Arabidopsis. Plant Physiol 144: 60–71
    OpenUrlAbstract/FREE Full Text
  70. ↵
    1. Tomato Genome Consortium
    (2012) The tomato genome sequence provides insights into fleshy fruit evolution. Nature 485: 635–641
    OpenUrlCrossRefPubMed
  71. ↵
    1. van der Hoeven RS,
    2. Steffens JC
    (2000) Biosynthesis and elongation of short- and medium-chain-length fatty acids. Plant Physiol 122: 275–282
    OpenUrlAbstract/FREE Full Text
  72. ↵
    1. Walters DS,
    2. Steffens JC
    (1990) Branched chain amino acid metabolism in the biosynthesis of Lycopersicon pennellii glucose esters. Plant Physiol 93: 1544–1551
    OpenUrlAbstract/FREE Full Text
  73. ↵
    1. Wang G,
    2. Tian L,
    3. Aziz N,
    4. Broun P,
    5. Dai X,
    6. He J,
    7. King A,
    8. Zhao PX,
    9. Dixon RA
    (2008) Terpene biosynthesis in glandular trichomes of hop. Plant Physiol 148: 1254–1266
    OpenUrlAbstract/FREE Full Text
  74. ↵
    1. Wang Y,
    2. Li J,
    3. Paterson AH
    (2013) MCScanX-transposed: detecting transposed gene duplications based on multiple colinearity scans. Bioinformatics 29: 1458–1460
    OpenUrlAbstract/FREE Full Text
  75. ↵
    1. Weinhold A,
    2. Baldwin IT
    (2011) Trichome-derived O-acyl sugars are a first meal for caterpillars that tags them for predation. Proc Natl Acad Sci USA 108: 7855–7859
    OpenUrlAbstract/FREE Full Text
  76. ↵
    1. Wiegel J
    (1981) Alpha-isopropylmalate synthase as a marker for the leucine biosynthetic pathway in several Clostridia and in Bacteroides fragilis. Arch Microbiol 130: 385–390
    OpenUrlCrossRefPubMed
  77. ↵
    1. Zabalza A,
    2. Zulet A,
    3. Gil-Monreal M,
    4. Igal M,
    5. Royuela M
    (2013) Branched-chain amino acid biosynthesis inhibitors: herbicide efficacy is associated with an induced carbon-nitrogen imbalance. J Plant Physiol 170: 814–821
    OpenUrlCrossRefPubMed
  78. ↵
    1. Zhang Z,
    2. Wu J,
    3. Lin W,
    4. Wang J,
    5. Yan H,
    6. Zhao W,
    7. Ma J,
    8. Ding J,
    9. Zhang P,
    10. Zhao GP
    (2014) Subdomain II of α-isopropylmalate synthase is essential for activity: inferring a mechanism of feedback inhibition. J Biol Chem 289: 27966–27978
    OpenUrlAbstract/FREE Full Text
View Abstract
Back to top

Table of Contents

Print
Download PDF
Email Article
Citation Tools
Request Permissions
A Feedback-Insensitive Isopropylmalate Synthase Affects Acylsugar Composition in Cultivated and Wild Tomato
Jing Ning, Gaurav D. Moghe, Bryan Leong, Jeongwoon Kim, Itai Ofner, Zhenzhen Wang, Christopher Adams, A. Daniel Jones, Dani Zamir, Robert L. Last
Plant Physiology Nov 2015, 169 (3) 1821-1835; DOI: 10.1104/pp.15.00474
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

In this issue

View this article with LENS

Subjects