A Novel Multifunctional C-23 Oxidase, CYP714E19, is Involved in Asiaticoside Biosynthesis

A Novel Multifunctional C-23 Oxidase, CYP714E19, is Involved in Asiaticoside Biosynthesis Abstract Centella asiatica is widely used as a medicinal plant due to accumulation of the ursane-type triterpene saponins asiaticoside and madecassoside. The molecular structure of both compounds suggests that they are biosynthesized from α-amyrin via three hydroxylations, and the respective Cyt P450-dependent monooxygenases (P450 enzymes) oxidizing the C-28 and C-2α positions have been reported. However, a third enzyme hydroxylating C-23 remained elusive. We previously identified 40,064 unique sequences in the transcriptome of C. asiatica elicited by methyl jasmonate, and among them we have now found 149 unigenes encoding putative P450 enzymes. In this set, 23 full-length cDNAs were recognized, 13 of which belonged to P450 subfamilies previously implicated in secondary metabolism. Four of these genes were highly expressed in response to jasmonate treatment, especially in leaves, in accordance with the accumulation patterns of asiaticoside. The functions of these candidate genes were tested using heterologous expression in yeast cells. Gas chromatography–mass spectrometry (GC-MS) analysis revealed that yeast expressing only the oxidosqualene synthase CaDDS produced the asiaticoside precursor α-amyrin (along with its isomer β-amyrin), while yeast co-expressing CaDDS and CYP716A83 also contained ursolic acid along with oleanolic acid. This P450 enzyme thus acts as a multifunctional triterpenoid C-28 oxidase converting amyrins into corresponding triterpenoid acids. Finally, yeast strains co-expressing CaDDS, CYP716A83 and CYP714E19 produced hederagenin and 23-hydroxyursolic acid, showing that CYP714E19 is a multifunctional triterpenoid oxidase catalyzing the C-23 hydroxylation of oleanolic acid and ursolic acid. Overall, our results demonstrate that CaDDS, CYP716A83 and CYP714E19 are C. asiatica enzymes catalyzing consecutive steps in asiaticoside biosynthesis. Introduction Triterpenoids are a large group of secondary metabolites comprising >300 different carbon skeletons, with one or more functional groups that may be glycosylated to form saponins (Xu et al. 2004). The bioactive saponins of many plant species are derived from oleanane-type aglycones, such as glycyrrhizin accumulating in licorice, saikosaponins from Bupleurum falcatum and soyasaponins from Glycine max. Many medicinal plants also contain active saponins with ursane skeletons. Among them, Centella asiatica (L.) Urban leaves are widely used in pharmaceutical applications towards wound healing, for treating vein diseases and memory improvement, and as anti-histamine, anti-ulcer, anti-leprosy, anti-depressant, anti-bacterial, anti-fungal or anti-oxidant agents (Brinkhaus et al. 2000, James and Dubery 2009). The bioactive compounds involved are ursane-type triterpenoid saponins, most importantly madecassoside and asiaticoside. The processes involved in the formation of C. asiatica saponins can be predicted based on triterpenoid biosynthesis in other plant species, typically proceeding in stages of precursor formation, cyclization, functionalization and conjugation. First, common precursors are synthesized by isoprene oligomerization involving squalene synthase, and activation by squalene epoxidase. Then, triterpenoid pathways start to branch, and cyclization by oxidosqualene cyclases (OSCs) yields either phytosterol or saponin intermediates. Some OSCs form multiple products, and many plant species have two or more OSC homologs, giving rise to arrays of isomeric triterpene carbon backbone structures (Phillips et al. 2006). For example, the Arabidopsis thaliana genome encodes 13 OSCs, all with distinct biochemical functions. Similarly, OSCs from many other plant species have been characterized, in many cases as product-specific enzymes forming β-amyrin. The C. asiatica saponin, asiaticoside, has a ursane carbon skeleton and is thus thought to be biosynthesized from α-amyrin (Fig. 1). Relatively few OSC enzymes have been described to date that form α-amyrin in other plant species, and all were characterized as multifunctional enzymes synthesizing multiple triterpenoid products. OSCs producing predominantly α-amyin (α-amyin and β-amyrin in a 5:1 ratio) were described in Catharanthus roseus and Malus×domestica (Brendolise et al. 2011, Yu et al. 2013). Two Centella asiatica OSCs, CaCYS and CaDDS, were initially characterized as cycloartenol and dammarenediol synthases, respectively (Kim et al. 2005, Kim et al. 2009). In contrast, CaDDS was recently reported to form α-amyrin, β-amyrin and dammarenediol-II in a ratio of 88:11:1, respectively (Moses et al. 2014a), suggesting that this enzyme is forming the α-amyrin precursor for asiaticoside biosynthesis. However, due to the conflicting evidence in these two studies, the exact function of CaDDS needs to be re-evaluated. Fig. 1 View largeDownload slide Putative biosynthetic pathways leading to triterpene saponins in C. asiatica. The precursors, α-amyrin and β-amyrin, are transformed by a series of oxidative reactions to oleanane- and ursane-type aglycones, respectively, which are then converted by UGTs to asiaticoside and scheffoleoside A. The oxidosqualene cyclase enzymes are represented as: CaDDS, multifunctional α-amyrin synthase; bAS, β-amyrin synthase. Colored arrows signify reactions catalyzed by P450 enzymes on the triterpene saponin pathway of C. asiatica. Fig. 1 View largeDownload slide Putative biosynthetic pathways leading to triterpene saponins in C. asiatica. The precursors, α-amyrin and β-amyrin, are transformed by a series of oxidative reactions to oleanane- and ursane-type aglycones, respectively, which are then converted by UGTs to asiaticoside and scheffoleoside A. The oxidosqualene cyclase enzymes are represented as: CaDDS, multifunctional α-amyrin synthase; bAS, β-amyrin synthase. Colored arrows signify reactions catalyzed by P450 enzymes on the triterpene saponin pathway of C. asiatica. In the next stage of triterpenoid biosynthesis, the pre-formed carbon skeletons undergo oxidative modifications (Thimmappa et al. 2014, Seki et al. 2015), and Cyt P450 monooxygenases (P450s) are thought to catalyze most of the regiospecific functionalizations. Several P450 enzymes catalyzing triterpenoid oxidations have been described to date, introducing hydroxyl or carboxyl functions at the C-6 to C-30 positions, respectively, of β-amyrin, dammarenediol, thalianol or lupeol (Qi et al. 2006, Field and Osbourn 2008, Seki et al. 2008, Fukushima et al. 2011, Han et al. 2011, Seki et al. 2011, Han et al. 2012, Geisler et al. 2013, Fukushima et al. 2013, Moses et al. 2014a, Moses et al. 2014b, Moses et al. 2015a, Moses et al. 2015b, Yasumoto et al. 2016, Zhang et al. 2016, Miettinen et al. 2017, Yasumoto et al. 2017). In contrast, hydroxylations of α-amyrin and its derivatives leading to the less common ursane-type saponins, including asiaticoside and madecassoside, are relatively poorly understood. En route to asiaticoside, α-amyrin must be hydroxylated at positions C-2α, C-23 and C-28 to yield asiatic acid (while further hydroxylation at C-6β yields madecassic acid). Very recently, a study based on publicly available transcriptome information identified C. asiatica candidate P450 enzymes for asiaticoside biosynthesis, all belonging to the subfamily CYP716 (Miettinen et al. 2017). Two of the enzymes, CYP716A83 and CYP716A86, were found to catalyze the C-28 oxidation of the pentacyclic triterpenoid skeleton, while CYP716C11 oxidized the C-2α position. Although two enzyme activities required for asiaticoside functionalization were thus reported, the third P450 responsible for C-23 oxidation remained unidentified. Candidate P450s for particular metabolic pathways cannot be identified based on sequence similarity alone due to high redundancy within the enzyme family (e.g. the A. thaliana genome contains 273 P450 genes and pseudogenes; Nelson and Werck-Reichhart 2011). Furthermore, even P450 homologs with relatively high sequence identity, within one species, may differ drastically in substrate or product specificity (Augustin et al. 2011). Therefore, approaches to identify the missing asiaticoside biosynthesis P450 must be both broad and specific, covering a range of enzyme homologs and filtering them with pathway-specific information. The formation of many secondary metabolites can be enhanced by environmental stress, frequently mediated by phytohormones such as methyl jasmonate (MeJA) (Zhao et al. 2005) and involving transcriptional up-regulation (Wasternack 2007). Accordingly, MeJA elicitation may be used to enhance transcriptomic analyses of secondary metabolism, including triterpene saponin biosynthesis (Lambert et al. 2011). In particular, concerted transcript and metabolite profiling of stress/elicitor-treated plants or cell cultures has proven to be a powerful approach for determining gene function in secondary metabolism (De Geyter et al. 2012). However, this strategy has not yet been used to investigate asiaticoside biosynthesis. The goal of the present study was to characterize genes potentially involved in asiaticoside biosynthesis, with particular focus on the C-23 oxidation step probably catalyzed by a P450 enzyme. To this end, we analyzed our previous transcriptome data of C. asiatica leaves elicited with MeJA to short-list candidate P450 enzymes and tested their expression patterns. We then characterized select C. asiatica P450 genes using heterologous expression in yeast and gas chromatography–mass spectrometry (GC-MS) analysis of triterpenoid products. Results The present investigation aimed at the isolation and characterization of a P450-dependent enzyme catalyzing the C-23 methyl group oxidation of triterpenoid intermediates en route to asiaticoside in Centella asiatica using a transcriptomic approach. In a previous study, 40,046 unique sequences had been identified in the transcriptome of C. asiatica leaves elicited by MeJA, and 64.8% of them had been annotated based on BLAST similarity searches against four public databases (Kim et al. 2017), including many sequences encoding enzymes likely to be involved in asiaticoside formation. Re-evaluation of the CaDDS enzyme To establish a heterologous expression system that would enable the characterization of candidate C. asiatica P450s, it was necessary first to identify enzymes that could provide relevant substrates. Therefore, we aimed to isolate and characterize at least one OSC enzyme forming the asiaticoside precursor α-amyrin. We identified five OSC candidate genes in the C. asiatica leaf cDNA library, with high sequence homology to previously characterized α-amyrin synthases, β-amyrin synthases and lupeol synthases from other species. One of the OSC sequences in the transcriptome data set encoded CaDDS, a C. asiatica enzyme initially characterized as a dammarenediol synthase (Kim et al. 2009) and more recently as a multifunctional OSC producing mainly α-amyrin, along with β-amyrin, traces of dammarenediol-II and another, unidentified product (Moses et al. 2014a). To re-assess the biochemical function of CaDDS, it was overexpressed in the yeast GIL77 mutant engineered to accumulate the OSC precursor oxidosqualene. In contrast to the previous reports, dammarenediol-II could not be detected in the resulting transgenic yeast, either in total ion chromatograms or in single ion traces of characteristic fragments (m/z 147, 199). However, the transgenic yeast produced six compounds not found in the respective empty vector controls. Among them, δ-amyrin (d1), α-amyrin (u1), β-amyrin (o1), ψ-taraxasterol (p1) and taraxasterol (t1) were identified based on GC retention times and mass spectral fragmentation patterns matching those of identical triterpenoid structures in the cuticular wax of tomato (Bauer et al. 2004, Wang et al. 2011). Another triterpenoid alcohol in the extract of the transgenic yeast could not be identified. The six triterpenoids were present in a ratio of 1:67:26:4:1:1, respectively (Supplementary Fig. S1). Overall, our re-evaluation of the CaDDS product spectrum revealed that this OSC is a multifunctional triterpene synthase forming mainly α-amyrin, the predicted precursor for asiaticoside. Therefore, CaDDS was identified as a suitable OSC for providing α-amyrin as substrate in co-expression experiments to test the activity of P450 enzymes that may be involved in the asiaticoside biosynthesis pathway. Selection of C. asiatica P450 enzymes involved in triterpenoid biosynthesis To identify C. asiatica P450 enzymes that may be involved in asiaticoside biosynthesis, three consecutive experiments were carried out. First, the transcriptome of MeJA-elicited C. asiatica leaves was analyzed, resulting in 149 unigenes annotated as P450s (Supplementary Table S1). Among them, 23 unigenes were present as full-length cDNAs, 13 of them belonging to various subfamilies with members previously implicated in secondary metabolism in other plant species (Table 1). Table 1 Full-length P450 cDNAs identified through 454 sequencing in C. asiatica leaves treated with MeJA Gene name GenBank accession No. EST number RPKM Putative function and source E-value CYP76AF2 KF004516 2,203 644.9 Cyt P450 76C4 [Vitis vinifera] 1E-117 CYP71D409 KF004517 642 273.9 Cyt P450 hydroxylase [Hyoscyamus muticus] 0 CYP74A1 KF004518 695 247.1 Cyt P450 [Panax notoginseng] 0 CYP716A83 KF004519 491 227.1 Cyt P450 CYP716A52v2 [Panax ginseng] 0 CYP714E19 KF004520 470 201.5 Cyt P450 [Theobroma cacao] 0 CYP710A58 KF004521 455 194.9 Cyt P450 710A1-like [Vitis vinifera] 0 CYP736A118 KF004522 489 184.1 Cyt P450 [Panax ginseng] 0 CYP72A309 KF004523 635 182.7 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP82H27 KF004524 462 169.4 Cyt P450 [Ammi majus] 0 CYP82D69 KF004525 365 121.0 Cyt P450 82A3-like [Vitis vinifera] 0 CYP78A112 KF004526 232 71.97 Cyt P450 78A4 isoform 1 [Vitis vinifera] 0 CYP736A119 KF004527 146 60.0 Cyt P450 [Panax ginseng] 0 CYP94B47 KF004528 116 56.1 Cyt P450 94A2 [Theobroma cacao] 0 CYP96A81 KF004529 235 49.8 Cyt P450 86B1 [Vitis vinifera] 0 CYP72A312 KF004530 103 45.6 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A313 KF004531 84 40.1 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A314 KF004532 73 36.3 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP81B59 KF004533 69 31.4 Cyt P450 [Panax notoginseng] 0 CYP716A84 KF004534 57 28.6 Cyt P450 CYP716A41 [Bupleurum chinense] 0 CYP86A87 KF004535 56 20.3 Cyt P450 86A2 [Vitis vinifera] 0 CYP97A39 KF004536 37 12.6 Cyt P450, putative [Ricinus communis] 0 CYP716A85 KF004537 26 10.8 Cyt P450 716B2-like [Solanum lycopersicum] 0 CYP716A86 KF004538 20 10.1 Cyt P450 716B2-like [Solanum lycopersicum] 0 Gene name GenBank accession No. EST number RPKM Putative function and source E-value CYP76AF2 KF004516 2,203 644.9 Cyt P450 76C4 [Vitis vinifera] 1E-117 CYP71D409 KF004517 642 273.9 Cyt P450 hydroxylase [Hyoscyamus muticus] 0 CYP74A1 KF004518 695 247.1 Cyt P450 [Panax notoginseng] 0 CYP716A83 KF004519 491 227.1 Cyt P450 CYP716A52v2 [Panax ginseng] 0 CYP714E19 KF004520 470 201.5 Cyt P450 [Theobroma cacao] 0 CYP710A58 KF004521 455 194.9 Cyt P450 710A1-like [Vitis vinifera] 0 CYP736A118 KF004522 489 184.1 Cyt P450 [Panax ginseng] 0 CYP72A309 KF004523 635 182.7 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP82H27 KF004524 462 169.4 Cyt P450 [Ammi majus] 0 CYP82D69 KF004525 365 121.0 Cyt P450 82A3-like [Vitis vinifera] 0 CYP78A112 KF004526 232 71.97 Cyt P450 78A4 isoform 1 [Vitis vinifera] 0 CYP736A119 KF004527 146 60.0 Cyt P450 [Panax ginseng] 0 CYP94B47 KF004528 116 56.1 Cyt P450 94A2 [Theobroma cacao] 0 CYP96A81 KF004529 235 49.8 Cyt P450 86B1 [Vitis vinifera] 0 CYP72A312 KF004530 103 45.6 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A313 KF004531 84 40.1 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A314 KF004532 73 36.3 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP81B59 KF004533 69 31.4 Cyt P450 [Panax notoginseng] 0 CYP716A84 KF004534 57 28.6 Cyt P450 CYP716A41 [Bupleurum chinense] 0 CYP86A87 KF004535 56 20.3 Cyt P450 86A2 [Vitis vinifera] 0 CYP97A39 KF004536 37 12.6 Cyt P450, putative [Ricinus communis] 0 CYP716A85 KF004537 26 10.8 Cyt P450 716B2-like [Solanum lycopersicum] 0 CYP716A86 KF004538 20 10.1 Cyt P450 716B2-like [Solanum lycopersicum] 0 Table 1 Full-length P450 cDNAs identified through 454 sequencing in C. asiatica leaves treated with MeJA Gene name GenBank accession No. EST number RPKM Putative function and source E-value CYP76AF2 KF004516 2,203 644.9 Cyt P450 76C4 [Vitis vinifera] 1E-117 CYP71D409 KF004517 642 273.9 Cyt P450 hydroxylase [Hyoscyamus muticus] 0 CYP74A1 KF004518 695 247.1 Cyt P450 [Panax notoginseng] 0 CYP716A83 KF004519 491 227.1 Cyt P450 CYP716A52v2 [Panax ginseng] 0 CYP714E19 KF004520 470 201.5 Cyt P450 [Theobroma cacao] 0 CYP710A58 KF004521 455 194.9 Cyt P450 710A1-like [Vitis vinifera] 0 CYP736A118 KF004522 489 184.1 Cyt P450 [Panax ginseng] 0 CYP72A309 KF004523 635 182.7 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP82H27 KF004524 462 169.4 Cyt P450 [Ammi majus] 0 CYP82D69 KF004525 365 121.0 Cyt P450 82A3-like [Vitis vinifera] 0 CYP78A112 KF004526 232 71.97 Cyt P450 78A4 isoform 1 [Vitis vinifera] 0 CYP736A119 KF004527 146 60.0 Cyt P450 [Panax ginseng] 0 CYP94B47 KF004528 116 56.1 Cyt P450 94A2 [Theobroma cacao] 0 CYP96A81 KF004529 235 49.8 Cyt P450 86B1 [Vitis vinifera] 0 CYP72A312 KF004530 103 45.6 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A313 KF004531 84 40.1 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A314 KF004532 73 36.3 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP81B59 KF004533 69 31.4 Cyt P450 [Panax notoginseng] 0 CYP716A84 KF004534 57 28.6 Cyt P450 CYP716A41 [Bupleurum chinense] 0 CYP86A87 KF004535 56 20.3 Cyt P450 86A2 [Vitis vinifera] 0 CYP97A39 KF004536 37 12.6 Cyt P450, putative [Ricinus communis] 0 CYP716A85 KF004537 26 10.8 Cyt P450 716B2-like [Solanum lycopersicum] 0 CYP716A86 KF004538 20 10.1 Cyt P450 716B2-like [Solanum lycopersicum] 0 Gene name GenBank accession No. EST number RPKM Putative function and source E-value CYP76AF2 KF004516 2,203 644.9 Cyt P450 76C4 [Vitis vinifera] 1E-117 CYP71D409 KF004517 642 273.9 Cyt P450 hydroxylase [Hyoscyamus muticus] 0 CYP74A1 KF004518 695 247.1 Cyt P450 [Panax notoginseng] 0 CYP716A83 KF004519 491 227.1 Cyt P450 CYP716A52v2 [Panax ginseng] 0 CYP714E19 KF004520 470 201.5 Cyt P450 [Theobroma cacao] 0 CYP710A58 KF004521 455 194.9 Cyt P450 710A1-like [Vitis vinifera] 0 CYP736A118 KF004522 489 184.1 Cyt P450 [Panax ginseng] 0 CYP72A309 KF004523 635 182.7 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP82H27 KF004524 462 169.4 Cyt P450 [Ammi majus] 0 CYP82D69 KF004525 365 121.0 Cyt P450 82A3-like [Vitis vinifera] 0 CYP78A112 KF004526 232 71.97 Cyt P450 78A4 isoform 1 [Vitis vinifera] 0 CYP736A119 KF004527 146 60.0 Cyt P450 [Panax ginseng] 0 CYP94B47 KF004528 116 56.1 Cyt P450 94A2 [Theobroma cacao] 0 CYP96A81 KF004529 235 49.8 Cyt P450 86B1 [Vitis vinifera] 0 CYP72A312 KF004530 103 45.6 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A313 KF004531 84 40.1 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A314 KF004532 73 36.3 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP81B59 KF004533 69 31.4 Cyt P450 [Panax notoginseng] 0 CYP716A84 KF004534 57 28.6 Cyt P450 CYP716A41 [Bupleurum chinense] 0 CYP86A87 KF004535 56 20.3 Cyt P450 86A2 [Vitis vinifera] 0 CYP97A39 KF004536 37 12.6 Cyt P450, putative [Ricinus communis] 0 CYP716A85 KF004537 26 10.8 Cyt P450 716B2-like [Solanum lycopersicum] 0 CYP716A86 KF004538 20 10.1 Cyt P450 716B2-like [Solanum lycopersicum] 0 In a second experiment, the candidate P450 enzymes for asiaticoside biosynthesis were further evaluated by studying gene expression in hairy root cultures. As both the asiaticoside products and the OSC CaDDS transcript accumulate rapidly after MeJA induction in C. asiatica hairy roots and whole plants (Mangas et al. 2006, Kim et al. 2007), it seemed plausible that expression of asiaticoside-forming P450 enzymes is also induced by MeJA treatment. Consequently, the expression levels of the 13 genes from P450 subfamilies previously associated with secondary metabolism were monitored by real-time reverse transcription–PCR (RT–PCR) as a function of time after MeJA induction (Supplementary Fig. S2A). Eight of the unigenes showed particularly high expression levels in hairy roots starting 12 h after elicitation. In contrast, other P450s were either expressed at constant levels with and without MeJA induction or not expressed at all. To corroborate the results of the RT–PCR analysis, we performed further quantitative RT–PCR analysis of the eight inducible genes (Fig. 2A). All genes, except CYP72A309, showed dramatically increased expression already 12 h after MeJA induction, and varying transcript levels thereafter. Fig. 2 View largeDownload slide Expression patterns of select P450 genes in C. asiatica. RT–PCR analysis of (A) gene expression in hairy roots as a function of time after induction with MeJA, and (B) expression levels in different organs. The expression levels of genes were normalized to the expression levels of β-actin in the same sample. Densitometric analysis showing the mean ± SE from three independent experiments, each performed in triplicate. Fig. 2 View largeDownload slide Expression patterns of select P450 genes in C. asiatica. RT–PCR analysis of (A) gene expression in hairy roots as a function of time after induction with MeJA, and (B) expression levels in different organs. The expression levels of genes were normalized to the expression levels of β-actin in the same sample. Densitometric analysis showing the mean ± SE from three independent experiments, each performed in triplicate. Finally, to filter the list of potential asiaticoside-forming enzymes further, the organ-specific expression of candidate P450s was assessed. As asiaticoside accumulates mainly in leaves of C. asiatica, those P450s with preferential leaf expression were considered as good candidates for triterpenoid biosynthesis. CYP716A83, CYP716A85 and CYP714E19 were highly expressed in all organs of C. asiatica investigated, most notably in leaves, and CYP736A118 was more highly expressed in leaves than in other organs (Fig. 2B). These findings are in sharp contrast to many other P450s which were expressed at very low levels in leaves, or not at all (Supplementary Fig. S2B). Therefore, taking the organ-specific expression patterns together with MeJA inducibility, CYP716A83, CYP716A85, CYP714E19 and CYP736A118 were the primary candidates for P450 enzymes to be involved in asiaticoside biosynthesis, and in particular for catalyzing hydroxylations of α- and β-amyrin. To evaluate the selected P450 candidates for asiaticoside biosynthesis, their sequences were compared with those of other C. asiatica enzymes belonging to P450 subfamilies also associated with secondary metabolism (Fig. 3). In this context, the C. asiatica CYP716 genes were of primary interest because several members of this P450 subfamily had previously been characterized as triterpenoid oxidases from diverse plant species (Fukushima et al. 2011, Han et al. 2011, Han et al. 2012, Han et al. 2013, Moses et al. 2014a, Moses et al. 2015b). In particular, several CYP716A genes had previously been characterized as specific triterpenoid C-28 oxidases accepting various substrates to form oleanolic acid, ursolic and/or betulinic acids (Thimmappa et al. 2014, Seki et al. 2015). The high sequence similarity between CYP716A genes from distantly related plant species suggests that this biochemical function is an ancestral evolutionary trait that is highly conserved among taxa (Fukushima et al. 2011, Huang et al. 2012, Han et al. 2013, Moses et al. 2015a). One of the C. asiatica candidate enzymes, CYP716A83, grouped with P450s known as triterpenoid C-28 oxidases (along with CYP716A84 and CYP716A86). Our phylogenetic analysis of CYP716A83 thus underlined the findings of a previous report showing that this enzyme catalyzes the C-28 oxidation of α- and β-amyrin to ursolic and oleanolic acid, respectively (Miettinen et al. 2017). Fig. 3 View largeDownload slide Phylogenetic analysis of P450s from C. asiatica and other plant species. The gene names and sequences as well as the full names of species are given in Supplementary Table S2. Enzymes encoded by highly MeJA-inducible C. asiatica genes are highlighted in blue. The C. asiatica P450s CYP716A83 and CYP714E19 were characterized in this study. Fig. 3 View largeDownload slide Phylogenetic analysis of P450s from C. asiatica and other plant species. The gene names and sequences as well as the full names of species are given in Supplementary Table S2. Enzymes encoded by highly MeJA-inducible C. asiatica genes are highlighted in blue. The C. asiatica P450s CYP716A83 and CYP714E19 were characterized in this study. Our phylogenetic analysis further showed that the second C. asiatica P450 of interest, CYP716A85, had sequence similarity to various other P450 enzymes known to oxidize triterpenoids in various positions other than C-28 (Fig. 3). Recently, the sequence initially named CYP716A85 in our study has been re-classified as CYP716E41 due to additional sequence information accrued over time (D. Nelson, personal communication), and under this name has been partially characterized (Miettinen et al. 2017). The information available so far suggested that CYP716A85 (alias CYP716E41) hydroxylates the C-6 position of maslinic acid (i.e. 2α-hydroxyoleanolic acid) or oleanolic acid. Our gene expression data now confirm that this enzyme is likely to be involved in asiaticoside formation, probably oxidizing triterpenoid precursors on a position other than C-28. Beyond the CYP716 subfamily, several other P450 sequences in the C. asiatica leaf transcriptome also had similarities to triterpenoid metabolism enzymes. First, one CYP71 (CYP71D409) sequence was identified in phylogenetic clusters with enzymes known to oxidize triterpenoids on C-23 or C-28 (Krokida et al. 2013, Kranz-Finger et al. 2018). Secondly, a C. asiatica CYP72 subfamily member (CYP72A309) had sequence similarity to triterpenoid C-22, C-23 or C-30 oxidases in other species (Fukushima et al. 2013), including the oleanane C-23 oxidase CYP72A68v2 from Medicago truncatula (Supplementary Fig. S3). In contrast, the amino acid sequence of the C. asiatica P450 CYP714E19 was only 30% identical to that of CYP72A68v2 (Supplementary Fig. S4). Thirdly, several C. asiatica P450 sequences fell into the CYP74, CYP76, CYP82, CYP710, CYP714 and CYP736 subfamilies where no triterpenoid oxidases had been characterized before. Conversely, none of the C. asiatica P450s clustered with the CYP93E subfamily, many members of which oxidize triterpenoid C-24 methyls into carboxyl functions, and which have high levels of mutual identity (80–87%; Moses et al. 2014b). Overall, our C. asiatica sequence analyses and expression data showed that CYP714E19 and CYP736A118 were the primary enzyme candidates for catalyzing the C-23 oxidation of asiaticoside precursors. Therefore, the biochemical characteristics of both enzymes had to be studied in detail. Our results thus far further confirmed previous findings that CYP716A83 and CYP716A85 are also involved in asiaticoside formation, probably as C-28 and C-6 oxidases, respectively. The activities of these two enzymes hence had to be investigated as well, both to complement previous biochemical data and in the new context of the CYP714E19 and CYP736A118 enzymes. Characterization of C. asiatica P450 enzymes For biochemical characterization of the selected P450 enzyme candidates, they were co-expressed in yeast together with CaDDS to provide triterpenoid substrates. The yeast strain WAT21 was used for these in vivo characterization experiments, as it harbors the Arabidopsis reductase ATR to serve as an electron donor for testing of hydroxylase function. Ethyl acetate extracts of the yeast cells were analyzed by GC–MS, using total ion counts (TICs) to capture all compounds and extracted ion chromatograms (EICs) to inspect further those peaks with enhanced signal-to-noise ratios. Two rounds of experiments were carried out, first testing the activity of single candidate P450s and then testing the combined activities of select pairs of P450 enzymes. In the first set of experiments, control yeast harboring only CaDDS produced four triterpenoids not found in the empty vector control (Fig. 4A), confirming our results on CaDDS expressed in GIL77 yeast (compare Supplementary Fig. S1). In contrast, a yeast line co-expressing CaDDS and CYP716A83 produced six additional compounds not found in the controls (Fig. 4B), four of which had GC retention times (Fig. 4C) and MS fragmentation patterns (Fig. 5) matching those of (TMSi-derivatized) authentic standards of uvaol (u2a), erythrodiol (o2a), ursolic acid (u4a) and oleanolic acid (o4a). Two more products in the transgenic yeast were, based on their MS fragmentation patterns combining characteristics of ursane/oleanane skeletons with likely molecular ions m/z 512 and deformylation products [M-30]+. (Fig. 6A), tentatively identified as ursolic aldehyde (u3a) and oleanolic aldehyde (o3a). Therefore, the results of this yeast co-expression experiment show that CYP716A83 converts α-amyrin (u1) via uvaol (u2a) and ursolic aldehyde (u3a) into ursolic acid (u4a), and β-amyrin (o1) via erythrodiol (o2a) and oleanolic aldehyde (o3a) into oleanolic acid (o4a). Overall, this enzyme thus catalyzes the stepwise oxidation of the C-28 methyl groups (-CH3) of various triterpenoid substrates all the way to the corresponding C-28 carboxylic acids (-COOH). Fig. 4 View largeDownload slide Biochemical characterization of C. asiatica CYP716A83. GC-MS analysis of (A) extracts from yeast expressing CaDDS alone, (B) extracts from yeast expressing CaDDS together with CYP716A83, and (C) standards. Total ion counts (TICs) and extracted ion chromatograms (EICs) for fragments m/z 189 and m/z 203 are shown. Compounds formed by CaDDS were identified based on GC and MS behavior (see Fig. 5) comparison with matching authentic standards of δ-amyrin (d1), β-amyrin (o1), α-amyrin (u1), ψ-taraxasterol (p1) and taraxasterol (t1). Compounds o2a, u2a, o4a and u4a, formed by CaDDS and CYP716A83 together, showed retention behavior similar to that of authentic standards of erythrodiol, uvaol, oleanolic acid and ursolic acid, respectively. Further products were tentatively identified based on mass fragmentation patterns (see Fig. 6) as oleanolic aldehyde (o3a), ursolic aldehyde (u3a), urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a). IS, internal standard lupeol. Fig. 4 View largeDownload slide Biochemical characterization of C. asiatica CYP716A83. GC-MS analysis of (A) extracts from yeast expressing CaDDS alone, (B) extracts from yeast expressing CaDDS together with CYP716A83, and (C) standards. Total ion counts (TICs) and extracted ion chromatograms (EICs) for fragments m/z 189 and m/z 203 are shown. Compounds formed by CaDDS were identified based on GC and MS behavior (see Fig. 5) comparison with matching authentic standards of δ-amyrin (d1), β-amyrin (o1), α-amyrin (u1), ψ-taraxasterol (p1) and taraxasterol (t1). Compounds o2a, u2a, o4a and u4a, formed by CaDDS and CYP716A83 together, showed retention behavior similar to that of authentic standards of erythrodiol, uvaol, oleanolic acid and ursolic acid, respectively. Further products were tentatively identified based on mass fragmentation patterns (see Fig. 6) as oleanolic aldehyde (o3a), ursolic aldehyde (u3a), urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a). IS, internal standard lupeol. Fig. 5 View largeDownload slide MS identification of triterpenoids from yeast expressing C. asiatica CYP716A83 by comparison with standards. (A) Spectra of compounds u2a and o2a detected in the yeast extract (top row) matched those of standards (bottom row), identifying u2a and o2a as uvaol and erythrodiol, respectively. (B) Spectra of compounds u4a and o4a detected in the yeast extract (top row) matched those of standards (bottom row), identifying u4a and o4a as ursolic acid and oleanolic acid, respectively. TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. Fig. 5 View largeDownload slide MS identification of triterpenoids from yeast expressing C. asiatica CYP716A83 by comparison with standards. (A) Spectra of compounds u2a and o2a detected in the yeast extract (top row) matched those of standards (bottom row), identifying u2a and o2a as uvaol and erythrodiol, respectively. (B) Spectra of compounds u4a and o4a detected in the yeast extract (top row) matched those of standards (bottom row), identifying u4a and o4a as ursolic acid and oleanolic acid, respectively. TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. Fig. 6 View largeDownload slide Tentative MS identification of triterpenoids from yeast expressing C. asiatica CYP716A83 based on fragmentation patterns. (A) Compounds tentatively identified as ursolic aldehyde (u3a) and oleanolic aldehyde (o3a). (B) Compounds tentatively identified as urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a). TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. Fig. 6 View largeDownload slide Tentative MS identification of triterpenoids from yeast expressing C. asiatica CYP716A83 based on fragmentation patterns. (A) Compounds tentatively identified as ursolic aldehyde (u3a) and oleanolic aldehyde (o3a). (B) Compounds tentatively identified as urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a). TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. Two more compounds (9 and 13) were detected in the yeast co-expressing CaDDS and CYP716A83 (Fig. 4B), with mass spectral characteristics suggesting urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a) structures, respectively (Fig. 6B). Based on these tentative structure assignments, it seems plausible that CYP716A83 catalyzes the oxidation of the C-28 methyl groups of ψ-taraxasterol (p1) as well as of α/β-amyrin (u1/o1). Three further yeast lines co-expressing CaDDS together with CYP716A85, CYP714E19 or CYP736A118 contained the same triterpenoid alcohols as the CaDDS-only control, mostly β-amyrin and α-amyrin, but no further triterpenoids generated by oxidation of the alcohols (Supplementary Fig. S5). Neither of the three candidate enzymes thus showed monooxygenase activity on either the oleanane- or the ursane-type triterpenoid alcohol precursors. In a second round of experiments, we tested the potential activities of the CYP716A85, CYP714E19 or CYP736A118 candidate P450s in combination with CYP716A83. However, yeast expressing CaDDS and CYP716A83 together with CYP716A85 or CYP736A118 produced the same triterpenoids as the CaDDS/CYP716A83 control, without any additional oxidation products (Supplementary Fig. S6). Finally, co-expression of CaDDS, CYP716A83 and CYP714E19 led to the formation of various compounds not present in the respective controls. Among the newly formed triterpenoids, 23-hydroxyursolic acid (u4b), hederagenin (o4b) and gypsogenic acid (o4d) were identified based on retention times (Fig. 7) and mass spectra (Fig. 8) matching those of (TMSi-derivatized) authentic standards. Using GC-MS analyses in total ion and single ion (m/z 189 and 203) modes, five more compounds were recognized as triterpenoids. Among them, 3-hydroxyurs-12-ene-23,28-dioic acid (u4d) was identified based on its MS fragmentation pattern (Fig. 9A) similar to that of gypsogenic acid (o4d) (Fig. 8C) and a GC retention time difference between both compounds (Fig. 7B) matching that between other isomeric ursane- and oleanane-type triterpenoid derivatives (compare, for example, the retention time difference between α-amyrin u1 and β-amyrin o1). Two other compounds were tentatively identified as 3-hydroxy-23-oxours-12-en-28-oic acid (u4c) and gypsogenin (o4c) based on their MS fragmentation patterns (Fig. 9B) and relative GC retention times (Fig. 7B). Finally, two more compounds detected in the transgenic yeast had MS fragmentation patterns consistent with 23-hydroxyuvaol (u2b) and 23-hydroxyursolic aldehyde (u3b), respectively (Fig. 9C); however, their exact structures could not be assigned based on the spectroscopic evidence acquired here. Fig. 7 View largeDownload slide Biochemical characterization of C. asiatica CYP714E19. GC-MS analysis of (A) extracts from yeast expressing CaDDS and CYP714E19, (B) extracts from yeast expressing CaDDS, CYP716A83 and CYP714E19, and (C) standards. Total ion counts (TICs) and extracted ion chromatograms (EICs) for fragments m/z 189 and m/z 203 are shown. All major compounds were identified based on GC and MS behavior (see Figs. 5, 8) matching authentic standards of δ-amyrin (d1), β-amyrin (o1), α-amyrin (u1), ψ-taraxasterol (p1), taraxasterol (t1), erythrodiol (o2a), uvaol (u2a), oleanolic acid (o4a), ursolic acid (u4a), 23-hydroxyursolic acid (u4b), hederagenin (o4b) and gypsogenic acid (o4d). Further products were tentatively identified based on fragmentation patterns (see Figs. 6, 7, 9) as urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a), 23-hydroxyuvaol (u2b), 23-hydroxyursolic aldehyde (u3b), 3-hydroxy-23-oxours-12-en-28-oic acid (u4c), gypsogenin (o4c) and 3-hydroxyurs-12-ene-23,28-dioic acid (u4d). IS, internal standard lupeol. Fig. 7 View largeDownload slide Biochemical characterization of C. asiatica CYP714E19. GC-MS analysis of (A) extracts from yeast expressing CaDDS and CYP714E19, (B) extracts from yeast expressing CaDDS, CYP716A83 and CYP714E19, and (C) standards. Total ion counts (TICs) and extracted ion chromatograms (EICs) for fragments m/z 189 and m/z 203 are shown. All major compounds were identified based on GC and MS behavior (see Figs. 5, 8) matching authentic standards of δ-amyrin (d1), β-amyrin (o1), α-amyrin (u1), ψ-taraxasterol (p1), taraxasterol (t1), erythrodiol (o2a), uvaol (u2a), oleanolic acid (o4a), ursolic acid (u4a), 23-hydroxyursolic acid (u4b), hederagenin (o4b) and gypsogenic acid (o4d). Further products were tentatively identified based on fragmentation patterns (see Figs. 6, 7, 9) as urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a), 23-hydroxyuvaol (u2b), 23-hydroxyursolic aldehyde (u3b), 3-hydroxy-23-oxours-12-en-28-oic acid (u4c), gypsogenin (o4c) and 3-hydroxyurs-12-ene-23,28-dioic acid (u4d). IS, internal standard lupeol. Fig. 8 View largeDownload slide MS identification of triterpenoids from yeast expressing C. asiatica CYP714E19. (A) The spectrum of compound u4b detected in the yeast extract (top) matched that of a standard (bottom), identifying it as 23-hydroxyursolic acid. (B) The spectrum of compound o4b detected in the yeast extract (top) matched that of a standard (bottom), identifying it as hederagenin. (C) The spectrum of compound o4d detected in the yeast extract (top) matched that of a standard (bottom), identifying it as gypsogenic acid. Fig. 8 View largeDownload slide MS identification of triterpenoids from yeast expressing C. asiatica CYP714E19. (A) The spectrum of compound u4b detected in the yeast extract (top) matched that of a standard (bottom), identifying it as 23-hydroxyursolic acid. (B) The spectrum of compound o4b detected in the yeast extract (top) matched that of a standard (bottom), identifying it as hederagenin. (C) The spectrum of compound o4d detected in the yeast extract (top) matched that of a standard (bottom), identifying it as gypsogenic acid. Fig. 9 View largeDownload slide Tentative MS identification of triterpenoids from yeast expressing C. asiatica CYP714E19 based on fragmentation patterns. (A) Compounds tentatively identified as 23-hydroxyuvaol (u2b) and 23-hydroxyursolic aldehyde (u3b). (B) Compounds tentatively identified as 3-hydroxy-23-oxours-12-en-28-oic acid (u4c) and gypsogenin (o4c). (C) Compound tentatively identified as 3-hydroxyurs-12-ene-23,28-dioic acid (u4d). TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. Fig. 9 View largeDownload slide Tentative MS identification of triterpenoids from yeast expressing C. asiatica CYP714E19 based on fragmentation patterns. (A) Compounds tentatively identified as 23-hydroxyuvaol (u2b) and 23-hydroxyursolic aldehyde (u3b). (B) Compounds tentatively identified as 3-hydroxy-23-oxours-12-en-28-oic acid (u4c) and gypsogenin (o4c). (C) Compound tentatively identified as 3-hydroxyurs-12-ene-23,28-dioic acid (u4d). TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. To assess the relative quantities of various triterpenoids in the transgenic yeast harboring CaDDS, CYP716A83 and CYP714E19, the abundances of the common fragment m/z 203 were quantified in all GC-MS peaks. The overall triterpenoid mixture consisted of 17% 23-hydroxyuvaol (u2b), 6% 23-hydroxyursolic aldehyde (u3b), 33% 23-hydroxyursolic acid (u4b), together with 2% 3-hydroxy-23-oxours-12-en-28-oic acid (u4c) and 4% 3-hydroxyurs-12-en-23,28-dioic acid (u4d), and also together with 30% hederagenin (o4b), 4% gypsogenin (o4c) and 5% gypsogenic acid (o4d). These relative amounts imply that the various structures with alcohol, aldehyde or acid functionalities at the C-23 position comprised approximately 86, 6 and 9%, respectively, showing that CYP714E19 plays a major role in oxidizing the C-23 methyl group of various triterpenoid precursors into corresponding 23-hydroxy compounds (i.e. alcohols), and only a lesser role in the two further oxidation steps to the corresponding aldehydes and acids. All yeast expression experiments were repeated three times, always with nearly identical results. Overall, our results show that CYP714E19 is a multifunctional triterpenoid C-23 oxidase, accepting uvaol and erythrodiol, as well as ursolic and oleanolic acids as substrates. Based on this result, the extracts from yeast co-expressing CaDDS and CYP714E19 were re-analyzed using GC-MS in single-ion mode (m/z 189, 203, 216 and 571) to search specifically for 23-hydroxy-α-amyrin and 23-hydroxy-β-amyrin (Supplementary Fig. S7). However, neither of the compounds was detected, indicating that the amyrins do not serve as substrates for the C-23 oxidase. Discussion In the present study, the transcriptome of MeJA-induced C. asiatica leaves was analyzed to identify candidate genes potentially involved in asiaticoside biosynthesis. It is well established that MeJA induction leads to asiaticoside accumulation in C. asiatica leaves (Kim et al. 2004, Mangas et al. 2008), and we therefore expected that genes encoding asiaticoside pathway enzymes should be up-regulated upon MeJA treatment. We hypothesized that at least three P450s with different regiospecificity probably oxidize the C-2α, C-23 and C-28 positions of the ursane skeleton starting from α-amyin as the initial substrate (Fig. 1). The transcriptome of C. asiatica leaves contained several P450 genes with sequence homology to enzymes involved in triterpenoid metabolism. Among them, candidate genes belonging to the CYP716A, CYP714E and CYP736A subfamilies were of particular interest due to their relatively high expression in leaves and MeJA-elicited expression, both matching the accumulation patterns of asiaticoside. We found that one of the candidate P450s, CYP716A83, converts α-amyrin (u1) and β-amyrin (o1) into ursolic acid (u4a) and oleanolic acid (o4a), respectively. Our results thus confirm an earlier report showing that this enzyme has C-28 oxidase activity on ursane and oleanane substrates (Miettinen et al. 2017). In the light of our gene expression data, we conclude that CYP716A83 is a crucial enzyme catalyzing the initial oxidation steps converting the first ursane skeleton precursors into intermediates dedicated to the formation of asiaticoside. Our analyses also corroborated previous evidence on C-28 oxidation intermediates of CYP716A83, on the one hand confirming the formation of erythrodiol (o2a) en route from β-amyrin (o1) to oleanolic acid (o4a) and adding GC-MS information to substantiate the formation of oleanolic aldehyde (o3a), and on the other hand identifying uvaol (u2a) and substantiating ursolic aldehyde (u3a) as intermediates between α-amyrin (u1) and ursolic acid (u4a). Of note, we also detected additional CYP716A83 products heterobetulin (p2a) and heterobetulinic acid (p4a), probably formed by C-28 oxidation of ψ-taraxasterol instead of the amyrins. The enzyme thus accepts various substrates with isomeric backbone structures, while being specific in its activity on only one carbon center in them. Taken together, our results suggest that CYP716A83 catalyzes a three-step oxidation, forming respective C-28 alcohol and C-28 aldehyde products in a first and second round of hydroxylation, respectively, and releasing at least a part of them. It remains to be determined whether the enzyme can also re-bind the same intermediates for further oxidation. The oleanolic acid and ursolic acid formed by CYP716A83 may not only serve as intermediates en route to asiaticoside, but may also be converted into other bioactive products in C. asiatica or play physiological roles themselves. Because oleanolic acid, ursolic acid and betulinic acid are widespread and relatively abundant triterpenoids in plants, Fukushima et al. (2011) speculated that CYP716As could have an important role in plant development or signaling, along with their intermediate role in triterpenoid pathways. Accordingly, Carelli et al. (2011) reported that M. truncatula mutants defective in CYP716A12 expression, with impeded synthesis of oleanolic acid, showed a strong decrease in growth, confirming a possible dual role for this triterpenoid in plant defense and development. However, the detailed physiological functions of CYP716A subfamily enzymes catalyzing the oxidation at C-28 of amyrins in other species are not clear. A second C. asiatica CYP716 subfamily enzyme, CYP716A85, had expression patterns suggesting an involvement in asiaticoside formation, and based on sequence similarities was predicted to oxidize triterpenoid substrates on positions other than C-28. Accordingly, this P450 (re-named as CYP716E41) had recently been found to catalyze the C-6 hydroxylation of ursolic acid (u4a) and oleanolic acid (o4a) (Miettinen et al. 2017), albeit at very low levels varying strongly between parallels. Here, we could not detect the respective products, possibly due to low substrate and/or enzyme amounts or due to other differences in experimental conditions. In our experiments, CYP716A85 also showed no activity on α-amyrin (u1) and β-amyrin (o1), similar to results of Miettinen et al. (2017). As a major result of the current investigations, we found that a third candidate enzyme, CYP714E19, oxidizes ursolic acid into 23-hydroxyursolic acid (u4b). Together with our expression data, this result suggests that CYP714E19 is the enzyme catalyzing the C-23 hydroxylation crucial for asiaticoside formation. Interestingly, the P450 enzyme also formed several minor side products with C-23 oxygen functions, including the 23-oxo (u4c) and 23-carboxylate (u4d) derivatives of ursolic acid probably produced by repeat C-23 oxidation, showing that it may either carry out multistep oxidations or re-bind oxidation products. However, the major activity of the enzyme was for the first round of oxidation, leading to the 23-hydroxy function. The CYP714 subfamily, along with the CYP72, 734, 749, 721 and 749 clades, belongs to the CYP72 clan of P450 enzymes. While many CYP72 members have been characterized, there is relatively little information on the plant CYP714A–D subfamily to date. Most interestingly, some CYP714 enzymes were found to play roles in gibberellin biosynthesis, as C-13 and C-16 oxidases as well as C-16α,17 epoxidases (Zhu et al. 2006, Magome et al. 2013, Wang et al. 2016). However, a phylogenetic analysis showed that the CYP714E19 protein of C. asiatica has relatively little sequence homology with other CYP714 subfamily proteins (Supplementary Fig. S8). Our findings for CYP714E19 are similar to those of Fukushima et al. (2013) for CYP72A68v2 in M. truncatula. Co-expression of the latter gene with the Lotus japonicus bAS, CPR and CYP716A12 genes in yeast demonstrated that CYP72A68v2 also catalyzed specific oxidation of C-23 of oleanolic acid. However, the activity of this C-23 oxidase was not tested on other substrates, such as ursolic acid, impeding further comparisons with our results on ursane-type substrates. Furthermore, CYP72A68v2 also catalyzes a series of oxidation reactions from oleanolic acid via hederagenin to gypsogenic acid, similar to CYP714E19. The yeast co-expressing CYP716A83 and CYP714E19 also produced two novel products tentatively identified as 23-hydroxyuvaol (u2b) and 23-hydroxyursolic aldehyde (u3b). All our GC retention and MS fragmentation data are in good agreement with these structure assignments; however, the identification cannot be confirmed by comparison with authentic standards at this point. Meanwhile, these putative 23-hydroxy ursane products suggest that CYP714E19 may also use substrates with various functional groups on C-28 for C-23 oxidation. This raises the possibility that CYP716A83, co-expressed in our experiments, may have further flexibility on its substrates and convert 23-hydroxyuvaol and 23-hydroxyursolic aldehyde by C-28 oxidation into 23-hydroxyursolic acid. The oxidations catalyzed by CYP716A83 may thus either precede or follow those of CYP714E19, leading to a flexible sequence of oxidations rather than a linear pathway. In order to test respective enzyme specificities and relative C-28 oxidation activities on various substrates, quantitative in vitro activity assays will have to be established. Interestingly, co-expression of CaDDS and CYP714E19 alone did not lead to the formation of 23-hydroxy-α-amyrin or 23-hydroxy-β-amyrin under the conditions tested here (Supplementary Fig. S7). This result suggests that CYP714E19 is specific for oleanolic acid and ursolic acid as substrates, but discriminates against the corresponding α- and β-amyrin precursors lacking a C-28 functionality, even though they have oleanane and ursane skeletons, respectively. It therefore seems very likely that the asiaticoside biosynthesis pathway proceeds via initial oxidation of C-28, yielding ursolic acid, followed by C-23 oxidation to 23-hydroxyursolic acid (and/or C-2 oxidation). Our finding that CYP714E19 accepts ursolic acid as substrate indicates that C-23 oxidation can occur without prior C-2α oxidation, leading us to predict a C-2α oxidase with substrate specificity for 23-hydroxyursolic acid. However, it cannot be ruled out that such a C-2α oxidase may also accept ursolic acid and that CYP714E19 may catalyze C-23 hydroxylation of the resulting corosolic acid. In a previous study, the C. asiatica enzyme CYP716C11 was shown to catalyze the C-2α hydroxylation of oleanolic acid or 6-hydroxyoleanolic acid, implicating this as another P450 involved in asiaticoside formation (Miettinen et al. 2017). However, it was not tested whether CYP716C11 may also accept, or even prefer, 23-hydroxyursolic acid as substrate, and it hence remains to be determined whether C-23 oxidation precedes or follows C-2 hydroxylation. Finally, it should be noted that C. asiatica leaves also contain not only the ursane-type triterpenoid classes of asiaticoside and madecassoside, but also oleanane-type saponins such as scheffoleoside A and asiaticoside B (Matsuda et al. 2001). All these compounds share 2α-, 6β- and 23-hydroxy functionalities in their oleanane- and ursane-type sapogenins and saponins. Since all our results showed similar activities of P450 enzymes on ursane- and oleanane-type substrates, it seems likely that the hydroxylases are involved in both asiaticoside and scheffoleoside A formation. In conclusion, this work led first to the identification of MeJA-responsive transcripts in leaves of C. asiatica, and on to the functional characterization of two P450s that are involved in asiaticoside biosynthesis. In particular, CYP714E19 was identified as the enzyme catalyzing the C-23 oxidation, probably of ursolic acid, and thus the last missing P450 along the asiaticoside pathway. Now that all the oxidases involved in asiaticoside formation are known, further investigations into the sequence of oxidation steps on the pathway can be envisaged. The toolbox of C. asiatica P450 enzymes will also enable the engineering of other, possibly novel, hydroxylated triterpenoids with carbon skeletons other than ursanes and oleananes, by co-expression of the various P450s with diverse OSC enzymes in yeast. Finally, it is now also possible to attempt (ectopic) overexpression of the P450s in C. asiatica or heterologous expression, for example in Nicotiana benthamiana, to confirm their biochemical functions and to engineer the formation of asiaticoside and its precursors. Materials and Methods Plant materials and sample preparation For expressed sequence tag (EST) sequencing, whole-plant cultures of C. asiatica were established from node segments as previously described by Kim et al. (2004). The node segments of four individual plants were cultured on 1/2 Murashige and Skoog (MS) liquid medium supplemented with 3% sucrose at 23 ± 2°C in 16/8 h light/dark cycles at 100 r.p.m. After 2 weeks of cultivation, MeJA (Sigma-Aldrich) was added to the medium to a concentration of 0.1 mM, and leaves were harvested 24 h after elicitation and immediately frozen in liquid nitrogen until use for total RNA extraction. For RT–PCR time-course analysis of gene expression, an established hairy root culture system was used as described by Kim et al. (2007). Total RNAs were extracted from hairy roots at 0, 12, 24, 48 and 72 h after induction with 0.1 mM MeJA. For RT–PCR analysis of gene expression in various organs, whole plants of C. asiatica were cultured on 1/2 MS medium supplemented with 3% sucrose and 0.8% agar adjusted to pH 5.8 at 23 ± 2°C in 16/8 h light/dark cycles. After cultivation for 6 weeks, total RNA was extracted from leaves, stems, nodes and roots. Transcriptome analysis In a previous study, the RNA samples of C. asiatica treated with MeJA were sequenced using the Genome sequencer FLX (454 Life Sciences, Roche) and information of the annotations based on the public databases and expression levels calculated using RPKM (reads per kilobase per mapped million reads) for unigenes was used to select candidates of the P450 gene associated with triterpenoid biosynthesis of asiaticoside (Kim et al. 2017). Phylogenetic analysis For phylogenetic analyses, the deduced amino acid sequences were aligned using the interactive web site CLUSTAL W (http://clustalw.ddbj.nig.ac.jp/top-j.html). A phylogenetic tree was generated using MEGA5.2 based on the Neighbor–Joining method (Tamura et al. 2011). Amino acid distances were calculated using the Dayhoff PAM matrix method. Bootstrap analysis with 1,000 replicates was used to assess the strength of nodes in the tree. Sequence information for 23 of the P450 genes of C. asiatica has been deposited in GenBank with accession numbers as shown in Table 1. RT–PCR analysis Total RNA was extracted from each sample and first-strand cDNA was synthesized using the AMV reverse transcriptase (Promega) and 2 μg of total RNA. SYBR Green Master Mix (Bio-Rad) was used for quantification with the CFX96™ Real-Time System (Bio-Rad) and gene-specific primers (Supplementary Table S3). The PCR conditions were as follows: initial denaturation at 95°C for 3 min, followed by 40 cycles of amplification for 15 s at 95°C, for 15 s at 58°C and for 30 s at 72°C. After completing the reactions, the threshold cycle (Ct) value for each reaction was obtained, and the differences were calculated using the delta-delta-Ct method and β-actin as internal control. The fold change in transcript levels of each gene (considered for quantitaive RT–PCR) is presented as the mean and SE of three independent experimental analyses. Generation of plasmid vectors The Invitrogen Gateway system was used to facilitate subcloning. All PCR primers are listed in Supplementary Table S3. The open reading frames (ORFs) were amplified by PCR from the original cDNA with attB-modified custom primers for P450s, and the PCR products were inserted into the pDONR/Zeo entry vector to make an entry clone, following the manufacturer’s protocol (Invitrogen). Plasmid DNAs were prepared from several transformants and sequenced. The yeast expression vectors were produced using the LR reaction (Invitrogen) with each of the entry vectors and the destination vector pYES-DEST52 (–Ura) for CaDDS, pAG423GAL-ccdB (–His) for CYP716A83, and pAG425GAL-ccdB (–Leu) for CYP714E19. After sequence confirmation, the pYES-CaDDS, pAG423-CYP716A83 and pAG425-CYP714E19 plasmids were used for yeast transformation. Co-expression of CaDDS and CYP716A83 or CYP714E19 in yeast In one experiment, CaDDS was expressed alone in Saccharomyces cerevisiae strain GIL77 as described before (Wang et al. 2011). In a second experiment, the yeast strain WAT21, which carries A. thaliana NADPH-CYP reductase (Pompon et al. 1996), was transformed with pYES-CaDDS or empty pYES2 as controls. After confirming the presence of amyrins in yeast harboring pYES-CaDDS and the absence of pentacyclic triterpenes in yeast with empty pYES2, both strains were further transformed with pAG423-CYP716A83 or pAG425-CYP714E19. Recombinant yeast cells were grown for 2 d in SC-Ura-His or SC-Ura-Leu selection media, including 2% glucose. Then cells were collected and used to inoculate 15 ml cultures; 2% galactose was added immediately to each culture to induce gene expression, and cells were grown for a further 3 d. Finally, cells were extracted twice with 10 ml of ethyl acetate at 90°C for 30 min with intermittent vortexing. After centrifuging at 4,000 r.p.m. for 5 min, the organic phase was transferred to a new vial and then dried under a stream of nitrogen gas. The extracts were re-suspended in 400 μl of chloroform–methanol (1:1), and 100 μl of the resulting solution was transferred to a vial and dried down under nitrogen gas. Finally, the hydroxyl-containing compounds in the mixture were converted into trimethylsilyl derivatives by reaction with 50 μl of bis-N,O-(trimethylsilyl)trifluoroacetamide for 30 min at 70°C. After confirming the formation of oleanolic acid and ursolic acid in yeast cells transformed with pYES-CaDDS/pAG423-CYP716A83, they were further transformed with pAG425-CYP714E19. The resulting yeast cells were selected in SC medium lacking uracil, histidine and leucine, induced with 2% galactose, grown for a further 3 d, extracted and analyzed as described above. GC-MS analysis GC-MS analysis was performed as described by Wang et al. (2011). A 1 μl aliquot of the yeast extract was analyzed using a 6890 N gas chromatograph (Agilent) equipped with a cool on-column injector, mass spectrometric detector (5973 N, Agilent) and an HP-1 capillary column (Agilent; length 30 m, i.d. 320 μm, 1 μm film thickness). The oven temperature was programmed to hold at 50°C for 2 min after injection, a 40°C min–1 ramp to 200°C, a plateau at 200°C for 2 min, a rise to 320°C at a rate of 3 min–1 and holding at 320°C for 30 min. Triterpenoids were identified by comparison with authentic compounds (erythrodiol, uvaol, oleanolic acid, ursolic acid, hederagenin, 23-hydroxyursolic acid and gypsogenic acid) and with literature data. All sapogenin standards were of analytical grade, all purchased from Sigma-Aldrich, except that 23-hydroxyursolic acid was kindly provided by Dr. Jinsook Kim (Korea Institute of Oriental Medicine). Supplementary Data Supplementary data are available at PCP online. Funding This work was supported by the Cooperative Research Program for Agricultural Science & Technology Development [Project No. PJ00849903]; the Rural Development Administation, Republic of Korea; the Natural Sciences and Engineering Research Council of Canada; and the Canada Foundation for Innovation. Acknowledgments We thank Dr. David Nelson for help with the nomenclature of P450 genes, Dr. Y.E. Choi for providing yeast strain WAT21, and Dr. J.S. Kim for providing the 23-hydroxyursolic acid standard. Disclosures The authors have no conflicts of interest to declare. References Augustin J.M. , Kuzina V. , Andersen S.B. , Bak S. ( 2011 ) Molecular activities, biosynthesis and evolution of triterpenoid saponins . Phytochemistry 72 : 435 – 457 . Google Scholar CrossRef Search ADS PubMed Bauer S. , Schulte E. , Thier H.-P. ( 2004 ) Composition of the surface wax from tomatoes. I. Identification of the components by GC/MS . Eur. Food Res. Technol. 219 : 223 – 228 . Google Scholar CrossRef Search ADS Brendolise C. , Yauk Y.-K. , Eberhard E.D. , Wang M. , Chagne D. , Andre C. , et al. . ( 2011 ) An unusual plant triterpene synthase with predominant α-amyrin-producing activity identified by characterizing oxidosqualene cyclases from Malus × domestica . FEBS J . 278 : 2485 – 2499 . Google Scholar CrossRef Search ADS PubMed Brinkhaus B. , Lindner M. , Schuppan D. , Hahn E.G. ( 2000 ) Chemical, pharmacological and clinical profile of the East Asian medical plant Centella asiatica . Phytomedicine 7 : 427 – 448 . Google Scholar CrossRef Search ADS PubMed Carelli M. , Biazzi E. , Panara F. , Tava A. , Scaramelli L. , Porceddu A. , et al. . ( 2011 ) Medicago truncatula CYP716A12 is a multifunctional oxidase involved in the biosynthesis of hemolytic saponins . Plant Cell 23 : 3070 – 3081 . Google Scholar CrossRef Search ADS PubMed De Geyter N. , Gholami A. , Goormachtig S. , Goossens A. ( 2012 ) Transcriptional machineries in jasmonate-elicited plant secondary metabolism . Trends Plant Sci . 17 : 349 – 359 . Google Scholar CrossRef Search ADS PubMed Field B. , Osbourn A.E. ( 2008 ) Metabolic diversification-independent assembly of operon-like gene clusters in different plants . Science 320 : 543 – 547 . Google Scholar CrossRef Search ADS PubMed Fukushima E.O. , Seki H. , Ohyama K. , Ono E. , Umemoto N. , Mizutani M. , et al. . ( 2011 ) CYP716A subfamily members are multifunctional oxidases in triterpenoid biosynthesis . Plant Cell Physiol . 52 : 2050 – 2061 . Google Scholar CrossRef Search ADS PubMed Fukushima E.O. , Seki H. , Sawai S. , Suzuki M. , Ohyama K. , Saito K. , et al. . ( 2013 ) Combinatorial biosynthesis of legume natural and rare triterpenoids in engineered yeast . Plant Cell Physiol . 54 : 740 – 749 . Google Scholar CrossRef Search ADS PubMed Geisler K. , Hughes R.K. , Sainsbury F. , Lomonossoff G.P. , Rejzek M. , Fairhurst S. , et al. . ( 2013 ) Biochemical analysis of a multifunctional cytochrome P450 (CYP51) enzyme required for synthesis of antimicrobial triterpenes in plants . Proc. Natl Acad. Sci. USA 110 : E3360 – E3367 . Google Scholar CrossRef Search ADS Han J.Y. , Hwang H.S. , Choi S.W. , Kim H.J. , Choi Y.E. ( 2012 ) Cytochrome P450 CYP716A53v2 catalyzes the formation of protopanaxatriol from protopanaxadiol during ginsenoside biosynthesis in Panax ginseng . Plant Cell Physiol . 53 : 1535 – 1545 . Google Scholar CrossRef Search ADS PubMed Han J.Y. , Kim H.J. , Kwon Y.S. , Choi Y.E. ( 2011 ) The Cyt P450 enzyme CYP716A47 catalyzes the formation of protopanaxadiol from dammarenediol-II during ginsenoside biosynthesis in Panax ginseng . Plant Cell Physiol . 52 : 2062 – 2073 . Google Scholar CrossRef Search ADS PubMed Han J.Y. , Kim M.J. , Ban Y.W. , Hwang H.S. , Choi Y.E. ( 2013 ) The involvement of β-amyrin 28-oxidase (CYP716A52v2) in oleanane-type ginsenoside biosynthesis in Panax ginseng . Plant Cell Physiol . 54 : 2034 – 2046 . Google Scholar CrossRef Search ADS PubMed Huang L. , Li J. , Ye H. , Li C. , Wang H. , Liu B. , et al. . ( 2012 ) Molecular characterization of the pentacyclic triterpenoid biosynthetic pathway in Catharanthus roseus . Planta 236 : 1571 – 1581 . Google Scholar CrossRef Search ADS PubMed James J.T. , Dubery I.A. ( 2009 ) Pentacyclic triterpenoids from the medicinal herb, Centella asiatica (L.) Urban . Molecules 14 : 3922 – 3941 . Google Scholar CrossRef Search ADS PubMed Kim O.T. , Bang K.H. , Shin Y.S. , Lee M.J. , Jung S.J. , Hyun D.Y. , et al. . ( 2007 ) Enhanced production of asiaticoside from hairy root cultures of Centella asiatica (L.) Urban elicited by methyl jasmonate . Plant Cell Rep. 26 : 1941 – 1949 . Google Scholar CrossRef Search ADS PubMed Kim O.T. , Jin M.L. , Lee D.Y. , Jetter R. ( 2017 ) Characterization of the asiatic acid glucosyltransferase, UGT73AH1, involved in asiaticoside biosynthesis in Centella asiatica (L.) Urban . Int. J. Mol. Sci . 18 : 2630 . Google Scholar CrossRef Search ADS Kim O.T. , Kim M.Y. , Ahn J.C. , Hwang S.J. , Hwang B. ( 2005 ) Cloning and molecular analysis of cDNA encoding cycloartenol synthase from Centella asiatica (L.) Urban . Biotechnol. Bioproc. Eng. 10 : 16 – 22 . Google Scholar CrossRef Search ADS Kim O.T. , Kim M.Y. , Hong M.H. , Ahn J.C. , Hwang B. ( 2004 ) Stimulation of asiaticoside production from Centella asiatica whole plant cultures by elicitors . Plant Cell Rep. 23 : 339 – 344 . Google Scholar CrossRef Search ADS PubMed Kim O.T. , Lee J.W. , Bang K.H. , Kim Y.C. , Hyun D.Y. , Cha S.W. , et al. . ( 2009 ) Characterization of a dammarenediol synthase in Centella asiatica (L.) Urban . Plant Physiol. Biochem . 47 : 998 – 1002 . Google Scholar CrossRef Search ADS PubMed Kranz-Finger S. , Mahmoud O. , Ricklefs E. , Ditz N. , Bakkes P.J. , Urlacher V.B. ( 2018 ) Insights into the functional properties of the marneral oxidase CYP71A16 from Arabidopsis thaliana . Biochim. Biophys. Acta 1866 : 2 – 10 . Google Scholar CrossRef Search ADS PubMed Krokida A. , Delis C. , Geisler K. , Garagounis C. , Tsikou D. , Pena-Rodriguez L.M. , et al. . ( 2013 ) A metabolic gene cluster in Lotus japonicus discloses novel enzyme functions and products in triterpene biosynthesis . New Phytol. 200 : 675 – 690 . Google Scholar CrossRef Search ADS PubMed Lambert E. , Faizal A. , Geelen D. ( 2011 ) Modulation of triterpene saponin production: in vitro cultures, elicitation, and metabolic engineering . Appl. Biochem. Biotechnol. 164 : 220 – 237 . Google Scholar CrossRef Search ADS PubMed Magome H. , Nomura T. , Hanada A. , Takeda-Kamiya N. , Ohnishi T. , Shinma Y. , et al. . ( 2013 ) CYP714B1 and CYP714B2 encode gibberellin 13-oxidases that reduce gibberellin activity in rice . Proc. Natl. Acad. Sci. USA 110 : 1947 – 1952 . Google Scholar CrossRef Search ADS Mangas S. , Bonfill M. , Osuna L. , Moyano E. , Tortoriello J. , Cusido R.M. , et al. . ( 2006 ) The effect of methyl jasmonate on triterpene and sterol metabolisms of Centella asiatica, Ruscus aculeatus and Galphimia glauca cultured plants . Phytochemistry 67 : 2041 – 2049 . Google Scholar CrossRef Search ADS PubMed Mangas S. , Moyano E. , Osuna L. , Cusido R.M. , Bonfill M. , Palazón J. ( 2008 ) Triterpenoid saponin content and the expression level of some related genes in calli of Centella asiatica . Biotechnol. Lett . 32 : 315 – 319 . Matsuda H. , Morikawa T. , Ueda H. , Yoshikawa M. ( 2001 ) Medicinal foodstuffs. XXVII. Saponin constituents Gotu Kola (2): structures of new ursane- and oleanane-type triterpene oligoglycosides, centellasaponin B, C, and D, from Centella asiatica cultivated in Sri Lanka . Chem. Pharm. Bull . 49 : 1368 – 1371 . Google Scholar CrossRef Search ADS PubMed Miettinen K. , Pollier J. , Buyst D. , Arendt P. , Csuk R. , Sommerwerk S. , et al. . ( 2017 ) The ancient CYP716 family is a major contributor to the diversification of eudicot triterpenoid biosynthesis . Nat. Commun. 8 : 14153. Google Scholar CrossRef Search ADS PubMed Moses T. , Pollier J. , Almagro L. , Buyst D. , Montagu M.V. , Pedreño M.A. , et al. . ( 2014a ) Combinatorial biosynthesis of sapogenins and saponins in Saccharomyces cerevisiae using a C-16α hydroxylase from Bupleurum falcatum . Proc. Natl. Acad. Sci. USA 111 : 1634 – 1639 . Google Scholar CrossRef Search ADS Moses T. , Pollier J. , Faizal A. , Apers S. , Pieters L. , Thevelein J.M. , et al. . ( 2015a ) Unraveling the triterpenoid saponin biosynthesis of the African shrub Maesa lanceolata . Mol. Plant 8 : 122 – 135 . Google Scholar CrossRef Search ADS Moses T. , Pollier J. , Shen Q. , Soetaert S. , Reed J. , Erffelinck M.L. , et al. . ( 2015 ) OSC2 and CYP716A14v2 catalyze the biosynthesis of triterpenoids for the cuticle of aerial organs of Artemisia annua . Plant Cell 27 : 286 – 301 . Google Scholar CrossRef Search ADS PubMed Moses T. , Thevelein J.M. , Goossens A. , Pollier J. ( 2014b ) Comparative analysis of CYP93E proteins for improved microbial synthesis of plant triterpenoids . Phytochemistry 108 : 47 – 56 . Google Scholar CrossRef Search ADS Nelson D. , Werck-Reichhart D. ( 2011 ) A P450-centric view of plant evolution . Plant J. 66 : 194 – 211 . Google Scholar CrossRef Search ADS PubMed Phillips D.R. , Rasbery J.M. , Bartel B. , Matsuda S.P.T. ( 2006 ) Biosynthetic diversity in plant triterpene cyclization . Curr. Opin. Plant Biol . 9 : 305 – 314 . Google Scholar CrossRef Search ADS PubMed Pompon D. , Louerat B. , Bronine A. , Urban P. ( 1996 ) Yeast expression of animal and plant P450s in optimized redox environments . Methods Enzymol . 272 : 51 – 64 . Google Scholar CrossRef Search ADS PubMed Qi X. , Bakht S. , Qin B. , Leggett M. , Hemmings A. , Mellon F. , et al. . ( 2006 ) A different function for a member of an ancient and highly conserved cytochrome P450 family: from essential sterols to plant defense . Proc. Natl. Acad. Sci. USA 103 : 18848 – 18853 . Google Scholar CrossRef Search ADS Seki H. , Ohyama K. , Sawai S. , Mizutani M. , Ohnishi T. , Sudo H. , et al. . ( 2008 ) Licorice β-amyrin 11-oxidase, a cytochrome P450 with a key role in the biosynthesis of the triterpene sweetener glycyrrhizin . Proc. Natl. Acad. Sci. USA 105 : 14204 – 14209 . Google Scholar CrossRef Search ADS Seki H. , Sawai S. , Ohyama K. , Mizutani M. , Ohnishi T. , Sudo H. , et al. . ( 2011 ) Triterpene functional genomics in licorice for identification of CYP72A154 involved in the biosynthesis of glycyrrhizin . Plant Cell 23 : 4112 – 4123 . Google Scholar CrossRef Search ADS PubMed Seki H. , Tamura K. , Muranaka T. ( 2015 ) P450s and UGTs: key players in the structural diversity of triterpenoid saponins . Plant Cell Physiol. 56 : 1463 – 1471 . Google Scholar CrossRef Search ADS PubMed Tamura K. , Peterson D. , Peterson N. , Stecher G. , Nei M. , Kumar S. ( 2011 ) MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods . Mol. Biol. Evol . 28 : 2731 – 2739 . Google Scholar CrossRef Search ADS PubMed Thimmappa R. , Geisler K. , Louveau T. , O’Maille P. , Osbourn A. ( 2014 ) Triterpene biosynthesis in plants . Annu. Rev. Plant Biol. 65 : 225 – 257 . Google Scholar CrossRef Search ADS PubMed Wang C. , Yang Y. , Wang H. , Ran X. , Li B. , Zhang J. , et al. . ( 2016 ) Ectopic expression of a cytochrome P450 monooxygenase gene PtCYP714A3 from Populus trichocarpa reduces shoot growth and improves tolerance to salt stress in transgenic rice . Plant Biotechnol. J . 14 : 1838 – 1851 . Google Scholar CrossRef Search ADS PubMed Wang Z. , Guhling O. , Yao R. , Li F. , Yeats T.H. , Rose J.K. , et al. . ( 2011 ) Two oxidosqualene cyclases responsible for biosynthesis of tomato fruit cuticular triterpenoids . Plant Physiol . 155 : 540 – 552 . Google Scholar CrossRef Search ADS PubMed Wasternack C. ( 2007 ) Jasmonates: an update on biosynthesis, signal transduction and action in plant stress response, growth and development . Ann. Bot . 100 : 681 – 697 . Google Scholar CrossRef Search ADS PubMed Xu R. , Fazio G.C. , Matsuda S.P.T. ( 2004 ) On the origins of triterpenoid skeletal diversity . Phytochemistry 65 : 261 – 291 . Google Scholar CrossRef Search ADS PubMed Yasumoto S. , Fukushima E.O. , Seki H. , Muranaka T. ( 2016 ) Novel triterpene oxidizing activity of Arabidopsis thaliana CYP716A subfamily enzymes . FEBS Lett. 590 : 533 – 540 . Google Scholar CrossRef Search ADS PubMed Yasumoto S. , Seki H. , Shimizu Y. , Fukushima E.O. , Muranaka T. ( 2017 ) Functional characterization of CYP716 family P450 enzymes in triterpenoid biosynthesis in tomato . Front. Plant Sci. 8 : 21 . Google Scholar CrossRef Search ADS PubMed Yu F. , Thamm A.M. , Reed D. , Villa-Ruano N. , Quesada A.L. , Gloria E.L. , et al. . ( 2013 ) Functional characterization of amyrin synthase involved in ursolic acid biosynthesis in Catharanthus roseus leaf epidermis . Phytochemistry 91 : 122 – 127 . Google Scholar CrossRef Search ADS PubMed Zhang J. , Dai L. , Yang J. , Liu C. , Men Y. , Zeng Y. , et al. . ( 2016 ) Oxidation of cucurbitadienol catalyzed by CYP87D18 in the biosynthesis of bogrosides from Siraitia grosvenorii . Plant Cell Physiol. 57 : 1000 – 1007 . Google Scholar CrossRef Search ADS PubMed Zhao J. , Davis L.C. , Verpoorte R. ( 2005 ) Elicitor signal transduction leading to production of plant secondary metabolites . Biotechnol. Adv . 23 : 283 – 333 . Google Scholar CrossRef Search ADS PubMed Zhu Y. , Nomura T. , Xu Y. , Zhang Y. , Peng Y. , Mao B. , et al. . ( 2006 ) ELONGATED UPPERMOST INTERNODE encodes a cytochrome P450 monooxygenase that epoxidizes gibberellins in a novel deactivation reaction in rice . Plant Cell 18 : 442 – 456 . Google Scholar CrossRef Search ADS PubMed Abbreviations Abbreviations EIC extracted ion chromatogram GC-MS gas chromatography–mass spectrometry MeJA methyl jasmonate MS Murashige and Skoog OSC oxidosqualene cyclase P450 Cyt P450 RT–PCR real-time reverse transcription–PCR TIC total ion count © The Author(s) 2018. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Plant and Cell Physiology Oxford University Press

A Novel Multifunctional C-23 Oxidase, CYP714E19, is Involved in Asiaticoside Biosynthesis

Loading next page...
 
/lp/ou_press/a-novel-multifunctional-c-23-oxidase-cyp714e19-is-involved-in-3n7JpdGHqA
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
0032-0781
eISSN
1471-9053
D.O.I.
10.1093/pcp/pcy055
Publisher site
See Article on Publisher Site

Abstract

Abstract Centella asiatica is widely used as a medicinal plant due to accumulation of the ursane-type triterpene saponins asiaticoside and madecassoside. The molecular structure of both compounds suggests that they are biosynthesized from α-amyrin via three hydroxylations, and the respective Cyt P450-dependent monooxygenases (P450 enzymes) oxidizing the C-28 and C-2α positions have been reported. However, a third enzyme hydroxylating C-23 remained elusive. We previously identified 40,064 unique sequences in the transcriptome of C. asiatica elicited by methyl jasmonate, and among them we have now found 149 unigenes encoding putative P450 enzymes. In this set, 23 full-length cDNAs were recognized, 13 of which belonged to P450 subfamilies previously implicated in secondary metabolism. Four of these genes were highly expressed in response to jasmonate treatment, especially in leaves, in accordance with the accumulation patterns of asiaticoside. The functions of these candidate genes were tested using heterologous expression in yeast cells. Gas chromatography–mass spectrometry (GC-MS) analysis revealed that yeast expressing only the oxidosqualene synthase CaDDS produced the asiaticoside precursor α-amyrin (along with its isomer β-amyrin), while yeast co-expressing CaDDS and CYP716A83 also contained ursolic acid along with oleanolic acid. This P450 enzyme thus acts as a multifunctional triterpenoid C-28 oxidase converting amyrins into corresponding triterpenoid acids. Finally, yeast strains co-expressing CaDDS, CYP716A83 and CYP714E19 produced hederagenin and 23-hydroxyursolic acid, showing that CYP714E19 is a multifunctional triterpenoid oxidase catalyzing the C-23 hydroxylation of oleanolic acid and ursolic acid. Overall, our results demonstrate that CaDDS, CYP716A83 and CYP714E19 are C. asiatica enzymes catalyzing consecutive steps in asiaticoside biosynthesis. Introduction Triterpenoids are a large group of secondary metabolites comprising >300 different carbon skeletons, with one or more functional groups that may be glycosylated to form saponins (Xu et al. 2004). The bioactive saponins of many plant species are derived from oleanane-type aglycones, such as glycyrrhizin accumulating in licorice, saikosaponins from Bupleurum falcatum and soyasaponins from Glycine max. Many medicinal plants also contain active saponins with ursane skeletons. Among them, Centella asiatica (L.) Urban leaves are widely used in pharmaceutical applications towards wound healing, for treating vein diseases and memory improvement, and as anti-histamine, anti-ulcer, anti-leprosy, anti-depressant, anti-bacterial, anti-fungal or anti-oxidant agents (Brinkhaus et al. 2000, James and Dubery 2009). The bioactive compounds involved are ursane-type triterpenoid saponins, most importantly madecassoside and asiaticoside. The processes involved in the formation of C. asiatica saponins can be predicted based on triterpenoid biosynthesis in other plant species, typically proceeding in stages of precursor formation, cyclization, functionalization and conjugation. First, common precursors are synthesized by isoprene oligomerization involving squalene synthase, and activation by squalene epoxidase. Then, triterpenoid pathways start to branch, and cyclization by oxidosqualene cyclases (OSCs) yields either phytosterol or saponin intermediates. Some OSCs form multiple products, and many plant species have two or more OSC homologs, giving rise to arrays of isomeric triterpene carbon backbone structures (Phillips et al. 2006). For example, the Arabidopsis thaliana genome encodes 13 OSCs, all with distinct biochemical functions. Similarly, OSCs from many other plant species have been characterized, in many cases as product-specific enzymes forming β-amyrin. The C. asiatica saponin, asiaticoside, has a ursane carbon skeleton and is thus thought to be biosynthesized from α-amyrin (Fig. 1). Relatively few OSC enzymes have been described to date that form α-amyrin in other plant species, and all were characterized as multifunctional enzymes synthesizing multiple triterpenoid products. OSCs producing predominantly α-amyin (α-amyin and β-amyrin in a 5:1 ratio) were described in Catharanthus roseus and Malus×domestica (Brendolise et al. 2011, Yu et al. 2013). Two Centella asiatica OSCs, CaCYS and CaDDS, were initially characterized as cycloartenol and dammarenediol synthases, respectively (Kim et al. 2005, Kim et al. 2009). In contrast, CaDDS was recently reported to form α-amyrin, β-amyrin and dammarenediol-II in a ratio of 88:11:1, respectively (Moses et al. 2014a), suggesting that this enzyme is forming the α-amyrin precursor for asiaticoside biosynthesis. However, due to the conflicting evidence in these two studies, the exact function of CaDDS needs to be re-evaluated. Fig. 1 View largeDownload slide Putative biosynthetic pathways leading to triterpene saponins in C. asiatica. The precursors, α-amyrin and β-amyrin, are transformed by a series of oxidative reactions to oleanane- and ursane-type aglycones, respectively, which are then converted by UGTs to asiaticoside and scheffoleoside A. The oxidosqualene cyclase enzymes are represented as: CaDDS, multifunctional α-amyrin synthase; bAS, β-amyrin synthase. Colored arrows signify reactions catalyzed by P450 enzymes on the triterpene saponin pathway of C. asiatica. Fig. 1 View largeDownload slide Putative biosynthetic pathways leading to triterpene saponins in C. asiatica. The precursors, α-amyrin and β-amyrin, are transformed by a series of oxidative reactions to oleanane- and ursane-type aglycones, respectively, which are then converted by UGTs to asiaticoside and scheffoleoside A. The oxidosqualene cyclase enzymes are represented as: CaDDS, multifunctional α-amyrin synthase; bAS, β-amyrin synthase. Colored arrows signify reactions catalyzed by P450 enzymes on the triterpene saponin pathway of C. asiatica. In the next stage of triterpenoid biosynthesis, the pre-formed carbon skeletons undergo oxidative modifications (Thimmappa et al. 2014, Seki et al. 2015), and Cyt P450 monooxygenases (P450s) are thought to catalyze most of the regiospecific functionalizations. Several P450 enzymes catalyzing triterpenoid oxidations have been described to date, introducing hydroxyl or carboxyl functions at the C-6 to C-30 positions, respectively, of β-amyrin, dammarenediol, thalianol or lupeol (Qi et al. 2006, Field and Osbourn 2008, Seki et al. 2008, Fukushima et al. 2011, Han et al. 2011, Seki et al. 2011, Han et al. 2012, Geisler et al. 2013, Fukushima et al. 2013, Moses et al. 2014a, Moses et al. 2014b, Moses et al. 2015a, Moses et al. 2015b, Yasumoto et al. 2016, Zhang et al. 2016, Miettinen et al. 2017, Yasumoto et al. 2017). In contrast, hydroxylations of α-amyrin and its derivatives leading to the less common ursane-type saponins, including asiaticoside and madecassoside, are relatively poorly understood. En route to asiaticoside, α-amyrin must be hydroxylated at positions C-2α, C-23 and C-28 to yield asiatic acid (while further hydroxylation at C-6β yields madecassic acid). Very recently, a study based on publicly available transcriptome information identified C. asiatica candidate P450 enzymes for asiaticoside biosynthesis, all belonging to the subfamily CYP716 (Miettinen et al. 2017). Two of the enzymes, CYP716A83 and CYP716A86, were found to catalyze the C-28 oxidation of the pentacyclic triterpenoid skeleton, while CYP716C11 oxidized the C-2α position. Although two enzyme activities required for asiaticoside functionalization were thus reported, the third P450 responsible for C-23 oxidation remained unidentified. Candidate P450s for particular metabolic pathways cannot be identified based on sequence similarity alone due to high redundancy within the enzyme family (e.g. the A. thaliana genome contains 273 P450 genes and pseudogenes; Nelson and Werck-Reichhart 2011). Furthermore, even P450 homologs with relatively high sequence identity, within one species, may differ drastically in substrate or product specificity (Augustin et al. 2011). Therefore, approaches to identify the missing asiaticoside biosynthesis P450 must be both broad and specific, covering a range of enzyme homologs and filtering them with pathway-specific information. The formation of many secondary metabolites can be enhanced by environmental stress, frequently mediated by phytohormones such as methyl jasmonate (MeJA) (Zhao et al. 2005) and involving transcriptional up-regulation (Wasternack 2007). Accordingly, MeJA elicitation may be used to enhance transcriptomic analyses of secondary metabolism, including triterpene saponin biosynthesis (Lambert et al. 2011). In particular, concerted transcript and metabolite profiling of stress/elicitor-treated plants or cell cultures has proven to be a powerful approach for determining gene function in secondary metabolism (De Geyter et al. 2012). However, this strategy has not yet been used to investigate asiaticoside biosynthesis. The goal of the present study was to characterize genes potentially involved in asiaticoside biosynthesis, with particular focus on the C-23 oxidation step probably catalyzed by a P450 enzyme. To this end, we analyzed our previous transcriptome data of C. asiatica leaves elicited with MeJA to short-list candidate P450 enzymes and tested their expression patterns. We then characterized select C. asiatica P450 genes using heterologous expression in yeast and gas chromatography–mass spectrometry (GC-MS) analysis of triterpenoid products. Results The present investigation aimed at the isolation and characterization of a P450-dependent enzyme catalyzing the C-23 methyl group oxidation of triterpenoid intermediates en route to asiaticoside in Centella asiatica using a transcriptomic approach. In a previous study, 40,046 unique sequences had been identified in the transcriptome of C. asiatica leaves elicited by MeJA, and 64.8% of them had been annotated based on BLAST similarity searches against four public databases (Kim et al. 2017), including many sequences encoding enzymes likely to be involved in asiaticoside formation. Re-evaluation of the CaDDS enzyme To establish a heterologous expression system that would enable the characterization of candidate C. asiatica P450s, it was necessary first to identify enzymes that could provide relevant substrates. Therefore, we aimed to isolate and characterize at least one OSC enzyme forming the asiaticoside precursor α-amyrin. We identified five OSC candidate genes in the C. asiatica leaf cDNA library, with high sequence homology to previously characterized α-amyrin synthases, β-amyrin synthases and lupeol synthases from other species. One of the OSC sequences in the transcriptome data set encoded CaDDS, a C. asiatica enzyme initially characterized as a dammarenediol synthase (Kim et al. 2009) and more recently as a multifunctional OSC producing mainly α-amyrin, along with β-amyrin, traces of dammarenediol-II and another, unidentified product (Moses et al. 2014a). To re-assess the biochemical function of CaDDS, it was overexpressed in the yeast GIL77 mutant engineered to accumulate the OSC precursor oxidosqualene. In contrast to the previous reports, dammarenediol-II could not be detected in the resulting transgenic yeast, either in total ion chromatograms or in single ion traces of characteristic fragments (m/z 147, 199). However, the transgenic yeast produced six compounds not found in the respective empty vector controls. Among them, δ-amyrin (d1), α-amyrin (u1), β-amyrin (o1), ψ-taraxasterol (p1) and taraxasterol (t1) were identified based on GC retention times and mass spectral fragmentation patterns matching those of identical triterpenoid structures in the cuticular wax of tomato (Bauer et al. 2004, Wang et al. 2011). Another triterpenoid alcohol in the extract of the transgenic yeast could not be identified. The six triterpenoids were present in a ratio of 1:67:26:4:1:1, respectively (Supplementary Fig. S1). Overall, our re-evaluation of the CaDDS product spectrum revealed that this OSC is a multifunctional triterpene synthase forming mainly α-amyrin, the predicted precursor for asiaticoside. Therefore, CaDDS was identified as a suitable OSC for providing α-amyrin as substrate in co-expression experiments to test the activity of P450 enzymes that may be involved in the asiaticoside biosynthesis pathway. Selection of C. asiatica P450 enzymes involved in triterpenoid biosynthesis To identify C. asiatica P450 enzymes that may be involved in asiaticoside biosynthesis, three consecutive experiments were carried out. First, the transcriptome of MeJA-elicited C. asiatica leaves was analyzed, resulting in 149 unigenes annotated as P450s (Supplementary Table S1). Among them, 23 unigenes were present as full-length cDNAs, 13 of them belonging to various subfamilies with members previously implicated in secondary metabolism in other plant species (Table 1). Table 1 Full-length P450 cDNAs identified through 454 sequencing in C. asiatica leaves treated with MeJA Gene name GenBank accession No. EST number RPKM Putative function and source E-value CYP76AF2 KF004516 2,203 644.9 Cyt P450 76C4 [Vitis vinifera] 1E-117 CYP71D409 KF004517 642 273.9 Cyt P450 hydroxylase [Hyoscyamus muticus] 0 CYP74A1 KF004518 695 247.1 Cyt P450 [Panax notoginseng] 0 CYP716A83 KF004519 491 227.1 Cyt P450 CYP716A52v2 [Panax ginseng] 0 CYP714E19 KF004520 470 201.5 Cyt P450 [Theobroma cacao] 0 CYP710A58 KF004521 455 194.9 Cyt P450 710A1-like [Vitis vinifera] 0 CYP736A118 KF004522 489 184.1 Cyt P450 [Panax ginseng] 0 CYP72A309 KF004523 635 182.7 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP82H27 KF004524 462 169.4 Cyt P450 [Ammi majus] 0 CYP82D69 KF004525 365 121.0 Cyt P450 82A3-like [Vitis vinifera] 0 CYP78A112 KF004526 232 71.97 Cyt P450 78A4 isoform 1 [Vitis vinifera] 0 CYP736A119 KF004527 146 60.0 Cyt P450 [Panax ginseng] 0 CYP94B47 KF004528 116 56.1 Cyt P450 94A2 [Theobroma cacao] 0 CYP96A81 KF004529 235 49.8 Cyt P450 86B1 [Vitis vinifera] 0 CYP72A312 KF004530 103 45.6 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A313 KF004531 84 40.1 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A314 KF004532 73 36.3 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP81B59 KF004533 69 31.4 Cyt P450 [Panax notoginseng] 0 CYP716A84 KF004534 57 28.6 Cyt P450 CYP716A41 [Bupleurum chinense] 0 CYP86A87 KF004535 56 20.3 Cyt P450 86A2 [Vitis vinifera] 0 CYP97A39 KF004536 37 12.6 Cyt P450, putative [Ricinus communis] 0 CYP716A85 KF004537 26 10.8 Cyt P450 716B2-like [Solanum lycopersicum] 0 CYP716A86 KF004538 20 10.1 Cyt P450 716B2-like [Solanum lycopersicum] 0 Gene name GenBank accession No. EST number RPKM Putative function and source E-value CYP76AF2 KF004516 2,203 644.9 Cyt P450 76C4 [Vitis vinifera] 1E-117 CYP71D409 KF004517 642 273.9 Cyt P450 hydroxylase [Hyoscyamus muticus] 0 CYP74A1 KF004518 695 247.1 Cyt P450 [Panax notoginseng] 0 CYP716A83 KF004519 491 227.1 Cyt P450 CYP716A52v2 [Panax ginseng] 0 CYP714E19 KF004520 470 201.5 Cyt P450 [Theobroma cacao] 0 CYP710A58 KF004521 455 194.9 Cyt P450 710A1-like [Vitis vinifera] 0 CYP736A118 KF004522 489 184.1 Cyt P450 [Panax ginseng] 0 CYP72A309 KF004523 635 182.7 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP82H27 KF004524 462 169.4 Cyt P450 [Ammi majus] 0 CYP82D69 KF004525 365 121.0 Cyt P450 82A3-like [Vitis vinifera] 0 CYP78A112 KF004526 232 71.97 Cyt P450 78A4 isoform 1 [Vitis vinifera] 0 CYP736A119 KF004527 146 60.0 Cyt P450 [Panax ginseng] 0 CYP94B47 KF004528 116 56.1 Cyt P450 94A2 [Theobroma cacao] 0 CYP96A81 KF004529 235 49.8 Cyt P450 86B1 [Vitis vinifera] 0 CYP72A312 KF004530 103 45.6 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A313 KF004531 84 40.1 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A314 KF004532 73 36.3 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP81B59 KF004533 69 31.4 Cyt P450 [Panax notoginseng] 0 CYP716A84 KF004534 57 28.6 Cyt P450 CYP716A41 [Bupleurum chinense] 0 CYP86A87 KF004535 56 20.3 Cyt P450 86A2 [Vitis vinifera] 0 CYP97A39 KF004536 37 12.6 Cyt P450, putative [Ricinus communis] 0 CYP716A85 KF004537 26 10.8 Cyt P450 716B2-like [Solanum lycopersicum] 0 CYP716A86 KF004538 20 10.1 Cyt P450 716B2-like [Solanum lycopersicum] 0 Table 1 Full-length P450 cDNAs identified through 454 sequencing in C. asiatica leaves treated with MeJA Gene name GenBank accession No. EST number RPKM Putative function and source E-value CYP76AF2 KF004516 2,203 644.9 Cyt P450 76C4 [Vitis vinifera] 1E-117 CYP71D409 KF004517 642 273.9 Cyt P450 hydroxylase [Hyoscyamus muticus] 0 CYP74A1 KF004518 695 247.1 Cyt P450 [Panax notoginseng] 0 CYP716A83 KF004519 491 227.1 Cyt P450 CYP716A52v2 [Panax ginseng] 0 CYP714E19 KF004520 470 201.5 Cyt P450 [Theobroma cacao] 0 CYP710A58 KF004521 455 194.9 Cyt P450 710A1-like [Vitis vinifera] 0 CYP736A118 KF004522 489 184.1 Cyt P450 [Panax ginseng] 0 CYP72A309 KF004523 635 182.7 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP82H27 KF004524 462 169.4 Cyt P450 [Ammi majus] 0 CYP82D69 KF004525 365 121.0 Cyt P450 82A3-like [Vitis vinifera] 0 CYP78A112 KF004526 232 71.97 Cyt P450 78A4 isoform 1 [Vitis vinifera] 0 CYP736A119 KF004527 146 60.0 Cyt P450 [Panax ginseng] 0 CYP94B47 KF004528 116 56.1 Cyt P450 94A2 [Theobroma cacao] 0 CYP96A81 KF004529 235 49.8 Cyt P450 86B1 [Vitis vinifera] 0 CYP72A312 KF004530 103 45.6 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A313 KF004531 84 40.1 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A314 KF004532 73 36.3 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP81B59 KF004533 69 31.4 Cyt P450 [Panax notoginseng] 0 CYP716A84 KF004534 57 28.6 Cyt P450 CYP716A41 [Bupleurum chinense] 0 CYP86A87 KF004535 56 20.3 Cyt P450 86A2 [Vitis vinifera] 0 CYP97A39 KF004536 37 12.6 Cyt P450, putative [Ricinus communis] 0 CYP716A85 KF004537 26 10.8 Cyt P450 716B2-like [Solanum lycopersicum] 0 CYP716A86 KF004538 20 10.1 Cyt P450 716B2-like [Solanum lycopersicum] 0 Gene name GenBank accession No. EST number RPKM Putative function and source E-value CYP76AF2 KF004516 2,203 644.9 Cyt P450 76C4 [Vitis vinifera] 1E-117 CYP71D409 KF004517 642 273.9 Cyt P450 hydroxylase [Hyoscyamus muticus] 0 CYP74A1 KF004518 695 247.1 Cyt P450 [Panax notoginseng] 0 CYP716A83 KF004519 491 227.1 Cyt P450 CYP716A52v2 [Panax ginseng] 0 CYP714E19 KF004520 470 201.5 Cyt P450 [Theobroma cacao] 0 CYP710A58 KF004521 455 194.9 Cyt P450 710A1-like [Vitis vinifera] 0 CYP736A118 KF004522 489 184.1 Cyt P450 [Panax ginseng] 0 CYP72A309 KF004523 635 182.7 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP82H27 KF004524 462 169.4 Cyt P450 [Ammi majus] 0 CYP82D69 KF004525 365 121.0 Cyt P450 82A3-like [Vitis vinifera] 0 CYP78A112 KF004526 232 71.97 Cyt P450 78A4 isoform 1 [Vitis vinifera] 0 CYP736A119 KF004527 146 60.0 Cyt P450 [Panax ginseng] 0 CYP94B47 KF004528 116 56.1 Cyt P450 94A2 [Theobroma cacao] 0 CYP96A81 KF004529 235 49.8 Cyt P450 86B1 [Vitis vinifera] 0 CYP72A312 KF004530 103 45.6 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A313 KF004531 84 40.1 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP72A314 KF004532 73 36.3 Cyt P450 CYP72A129 [Panax ginseng] 0 CYP81B59 KF004533 69 31.4 Cyt P450 [Panax notoginseng] 0 CYP716A84 KF004534 57 28.6 Cyt P450 CYP716A41 [Bupleurum chinense] 0 CYP86A87 KF004535 56 20.3 Cyt P450 86A2 [Vitis vinifera] 0 CYP97A39 KF004536 37 12.6 Cyt P450, putative [Ricinus communis] 0 CYP716A85 KF004537 26 10.8 Cyt P450 716B2-like [Solanum lycopersicum] 0 CYP716A86 KF004538 20 10.1 Cyt P450 716B2-like [Solanum lycopersicum] 0 In a second experiment, the candidate P450 enzymes for asiaticoside biosynthesis were further evaluated by studying gene expression in hairy root cultures. As both the asiaticoside products and the OSC CaDDS transcript accumulate rapidly after MeJA induction in C. asiatica hairy roots and whole plants (Mangas et al. 2006, Kim et al. 2007), it seemed plausible that expression of asiaticoside-forming P450 enzymes is also induced by MeJA treatment. Consequently, the expression levels of the 13 genes from P450 subfamilies previously associated with secondary metabolism were monitored by real-time reverse transcription–PCR (RT–PCR) as a function of time after MeJA induction (Supplementary Fig. S2A). Eight of the unigenes showed particularly high expression levels in hairy roots starting 12 h after elicitation. In contrast, other P450s were either expressed at constant levels with and without MeJA induction or not expressed at all. To corroborate the results of the RT–PCR analysis, we performed further quantitative RT–PCR analysis of the eight inducible genes (Fig. 2A). All genes, except CYP72A309, showed dramatically increased expression already 12 h after MeJA induction, and varying transcript levels thereafter. Fig. 2 View largeDownload slide Expression patterns of select P450 genes in C. asiatica. RT–PCR analysis of (A) gene expression in hairy roots as a function of time after induction with MeJA, and (B) expression levels in different organs. The expression levels of genes were normalized to the expression levels of β-actin in the same sample. Densitometric analysis showing the mean ± SE from three independent experiments, each performed in triplicate. Fig. 2 View largeDownload slide Expression patterns of select P450 genes in C. asiatica. RT–PCR analysis of (A) gene expression in hairy roots as a function of time after induction with MeJA, and (B) expression levels in different organs. The expression levels of genes were normalized to the expression levels of β-actin in the same sample. Densitometric analysis showing the mean ± SE from three independent experiments, each performed in triplicate. Finally, to filter the list of potential asiaticoside-forming enzymes further, the organ-specific expression of candidate P450s was assessed. As asiaticoside accumulates mainly in leaves of C. asiatica, those P450s with preferential leaf expression were considered as good candidates for triterpenoid biosynthesis. CYP716A83, CYP716A85 and CYP714E19 were highly expressed in all organs of C. asiatica investigated, most notably in leaves, and CYP736A118 was more highly expressed in leaves than in other organs (Fig. 2B). These findings are in sharp contrast to many other P450s which were expressed at very low levels in leaves, or not at all (Supplementary Fig. S2B). Therefore, taking the organ-specific expression patterns together with MeJA inducibility, CYP716A83, CYP716A85, CYP714E19 and CYP736A118 were the primary candidates for P450 enzymes to be involved in asiaticoside biosynthesis, and in particular for catalyzing hydroxylations of α- and β-amyrin. To evaluate the selected P450 candidates for asiaticoside biosynthesis, their sequences were compared with those of other C. asiatica enzymes belonging to P450 subfamilies also associated with secondary metabolism (Fig. 3). In this context, the C. asiatica CYP716 genes were of primary interest because several members of this P450 subfamily had previously been characterized as triterpenoid oxidases from diverse plant species (Fukushima et al. 2011, Han et al. 2011, Han et al. 2012, Han et al. 2013, Moses et al. 2014a, Moses et al. 2015b). In particular, several CYP716A genes had previously been characterized as specific triterpenoid C-28 oxidases accepting various substrates to form oleanolic acid, ursolic and/or betulinic acids (Thimmappa et al. 2014, Seki et al. 2015). The high sequence similarity between CYP716A genes from distantly related plant species suggests that this biochemical function is an ancestral evolutionary trait that is highly conserved among taxa (Fukushima et al. 2011, Huang et al. 2012, Han et al. 2013, Moses et al. 2015a). One of the C. asiatica candidate enzymes, CYP716A83, grouped with P450s known as triterpenoid C-28 oxidases (along with CYP716A84 and CYP716A86). Our phylogenetic analysis of CYP716A83 thus underlined the findings of a previous report showing that this enzyme catalyzes the C-28 oxidation of α- and β-amyrin to ursolic and oleanolic acid, respectively (Miettinen et al. 2017). Fig. 3 View largeDownload slide Phylogenetic analysis of P450s from C. asiatica and other plant species. The gene names and sequences as well as the full names of species are given in Supplementary Table S2. Enzymes encoded by highly MeJA-inducible C. asiatica genes are highlighted in blue. The C. asiatica P450s CYP716A83 and CYP714E19 were characterized in this study. Fig. 3 View largeDownload slide Phylogenetic analysis of P450s from C. asiatica and other plant species. The gene names and sequences as well as the full names of species are given in Supplementary Table S2. Enzymes encoded by highly MeJA-inducible C. asiatica genes are highlighted in blue. The C. asiatica P450s CYP716A83 and CYP714E19 were characterized in this study. Our phylogenetic analysis further showed that the second C. asiatica P450 of interest, CYP716A85, had sequence similarity to various other P450 enzymes known to oxidize triterpenoids in various positions other than C-28 (Fig. 3). Recently, the sequence initially named CYP716A85 in our study has been re-classified as CYP716E41 due to additional sequence information accrued over time (D. Nelson, personal communication), and under this name has been partially characterized (Miettinen et al. 2017). The information available so far suggested that CYP716A85 (alias CYP716E41) hydroxylates the C-6 position of maslinic acid (i.e. 2α-hydroxyoleanolic acid) or oleanolic acid. Our gene expression data now confirm that this enzyme is likely to be involved in asiaticoside formation, probably oxidizing triterpenoid precursors on a position other than C-28. Beyond the CYP716 subfamily, several other P450 sequences in the C. asiatica leaf transcriptome also had similarities to triterpenoid metabolism enzymes. First, one CYP71 (CYP71D409) sequence was identified in phylogenetic clusters with enzymes known to oxidize triterpenoids on C-23 or C-28 (Krokida et al. 2013, Kranz-Finger et al. 2018). Secondly, a C. asiatica CYP72 subfamily member (CYP72A309) had sequence similarity to triterpenoid C-22, C-23 or C-30 oxidases in other species (Fukushima et al. 2013), including the oleanane C-23 oxidase CYP72A68v2 from Medicago truncatula (Supplementary Fig. S3). In contrast, the amino acid sequence of the C. asiatica P450 CYP714E19 was only 30% identical to that of CYP72A68v2 (Supplementary Fig. S4). Thirdly, several C. asiatica P450 sequences fell into the CYP74, CYP76, CYP82, CYP710, CYP714 and CYP736 subfamilies where no triterpenoid oxidases had been characterized before. Conversely, none of the C. asiatica P450s clustered with the CYP93E subfamily, many members of which oxidize triterpenoid C-24 methyls into carboxyl functions, and which have high levels of mutual identity (80–87%; Moses et al. 2014b). Overall, our C. asiatica sequence analyses and expression data showed that CYP714E19 and CYP736A118 were the primary enzyme candidates for catalyzing the C-23 oxidation of asiaticoside precursors. Therefore, the biochemical characteristics of both enzymes had to be studied in detail. Our results thus far further confirmed previous findings that CYP716A83 and CYP716A85 are also involved in asiaticoside formation, probably as C-28 and C-6 oxidases, respectively. The activities of these two enzymes hence had to be investigated as well, both to complement previous biochemical data and in the new context of the CYP714E19 and CYP736A118 enzymes. Characterization of C. asiatica P450 enzymes For biochemical characterization of the selected P450 enzyme candidates, they were co-expressed in yeast together with CaDDS to provide triterpenoid substrates. The yeast strain WAT21 was used for these in vivo characterization experiments, as it harbors the Arabidopsis reductase ATR to serve as an electron donor for testing of hydroxylase function. Ethyl acetate extracts of the yeast cells were analyzed by GC–MS, using total ion counts (TICs) to capture all compounds and extracted ion chromatograms (EICs) to inspect further those peaks with enhanced signal-to-noise ratios. Two rounds of experiments were carried out, first testing the activity of single candidate P450s and then testing the combined activities of select pairs of P450 enzymes. In the first set of experiments, control yeast harboring only CaDDS produced four triterpenoids not found in the empty vector control (Fig. 4A), confirming our results on CaDDS expressed in GIL77 yeast (compare Supplementary Fig. S1). In contrast, a yeast line co-expressing CaDDS and CYP716A83 produced six additional compounds not found in the controls (Fig. 4B), four of which had GC retention times (Fig. 4C) and MS fragmentation patterns (Fig. 5) matching those of (TMSi-derivatized) authentic standards of uvaol (u2a), erythrodiol (o2a), ursolic acid (u4a) and oleanolic acid (o4a). Two more products in the transgenic yeast were, based on their MS fragmentation patterns combining characteristics of ursane/oleanane skeletons with likely molecular ions m/z 512 and deformylation products [M-30]+. (Fig. 6A), tentatively identified as ursolic aldehyde (u3a) and oleanolic aldehyde (o3a). Therefore, the results of this yeast co-expression experiment show that CYP716A83 converts α-amyrin (u1) via uvaol (u2a) and ursolic aldehyde (u3a) into ursolic acid (u4a), and β-amyrin (o1) via erythrodiol (o2a) and oleanolic aldehyde (o3a) into oleanolic acid (o4a). Overall, this enzyme thus catalyzes the stepwise oxidation of the C-28 methyl groups (-CH3) of various triterpenoid substrates all the way to the corresponding C-28 carboxylic acids (-COOH). Fig. 4 View largeDownload slide Biochemical characterization of C. asiatica CYP716A83. GC-MS analysis of (A) extracts from yeast expressing CaDDS alone, (B) extracts from yeast expressing CaDDS together with CYP716A83, and (C) standards. Total ion counts (TICs) and extracted ion chromatograms (EICs) for fragments m/z 189 and m/z 203 are shown. Compounds formed by CaDDS were identified based on GC and MS behavior (see Fig. 5) comparison with matching authentic standards of δ-amyrin (d1), β-amyrin (o1), α-amyrin (u1), ψ-taraxasterol (p1) and taraxasterol (t1). Compounds o2a, u2a, o4a and u4a, formed by CaDDS and CYP716A83 together, showed retention behavior similar to that of authentic standards of erythrodiol, uvaol, oleanolic acid and ursolic acid, respectively. Further products were tentatively identified based on mass fragmentation patterns (see Fig. 6) as oleanolic aldehyde (o3a), ursolic aldehyde (u3a), urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a). IS, internal standard lupeol. Fig. 4 View largeDownload slide Biochemical characterization of C. asiatica CYP716A83. GC-MS analysis of (A) extracts from yeast expressing CaDDS alone, (B) extracts from yeast expressing CaDDS together with CYP716A83, and (C) standards. Total ion counts (TICs) and extracted ion chromatograms (EICs) for fragments m/z 189 and m/z 203 are shown. Compounds formed by CaDDS were identified based on GC and MS behavior (see Fig. 5) comparison with matching authentic standards of δ-amyrin (d1), β-amyrin (o1), α-amyrin (u1), ψ-taraxasterol (p1) and taraxasterol (t1). Compounds o2a, u2a, o4a and u4a, formed by CaDDS and CYP716A83 together, showed retention behavior similar to that of authentic standards of erythrodiol, uvaol, oleanolic acid and ursolic acid, respectively. Further products were tentatively identified based on mass fragmentation patterns (see Fig. 6) as oleanolic aldehyde (o3a), ursolic aldehyde (u3a), urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a). IS, internal standard lupeol. Fig. 5 View largeDownload slide MS identification of triterpenoids from yeast expressing C. asiatica CYP716A83 by comparison with standards. (A) Spectra of compounds u2a and o2a detected in the yeast extract (top row) matched those of standards (bottom row), identifying u2a and o2a as uvaol and erythrodiol, respectively. (B) Spectra of compounds u4a and o4a detected in the yeast extract (top row) matched those of standards (bottom row), identifying u4a and o4a as ursolic acid and oleanolic acid, respectively. TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. Fig. 5 View largeDownload slide MS identification of triterpenoids from yeast expressing C. asiatica CYP716A83 by comparison with standards. (A) Spectra of compounds u2a and o2a detected in the yeast extract (top row) matched those of standards (bottom row), identifying u2a and o2a as uvaol and erythrodiol, respectively. (B) Spectra of compounds u4a and o4a detected in the yeast extract (top row) matched those of standards (bottom row), identifying u4a and o4a as ursolic acid and oleanolic acid, respectively. TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. Fig. 6 View largeDownload slide Tentative MS identification of triterpenoids from yeast expressing C. asiatica CYP716A83 based on fragmentation patterns. (A) Compounds tentatively identified as ursolic aldehyde (u3a) and oleanolic aldehyde (o3a). (B) Compounds tentatively identified as urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a). TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. Fig. 6 View largeDownload slide Tentative MS identification of triterpenoids from yeast expressing C. asiatica CYP716A83 based on fragmentation patterns. (A) Compounds tentatively identified as ursolic aldehyde (u3a) and oleanolic aldehyde (o3a). (B) Compounds tentatively identified as urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a). TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. Two more compounds (9 and 13) were detected in the yeast co-expressing CaDDS and CYP716A83 (Fig. 4B), with mass spectral characteristics suggesting urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a) structures, respectively (Fig. 6B). Based on these tentative structure assignments, it seems plausible that CYP716A83 catalyzes the oxidation of the C-28 methyl groups of ψ-taraxasterol (p1) as well as of α/β-amyrin (u1/o1). Three further yeast lines co-expressing CaDDS together with CYP716A85, CYP714E19 or CYP736A118 contained the same triterpenoid alcohols as the CaDDS-only control, mostly β-amyrin and α-amyrin, but no further triterpenoids generated by oxidation of the alcohols (Supplementary Fig. S5). Neither of the three candidate enzymes thus showed monooxygenase activity on either the oleanane- or the ursane-type triterpenoid alcohol precursors. In a second round of experiments, we tested the potential activities of the CYP716A85, CYP714E19 or CYP736A118 candidate P450s in combination with CYP716A83. However, yeast expressing CaDDS and CYP716A83 together with CYP716A85 or CYP736A118 produced the same triterpenoids as the CaDDS/CYP716A83 control, without any additional oxidation products (Supplementary Fig. S6). Finally, co-expression of CaDDS, CYP716A83 and CYP714E19 led to the formation of various compounds not present in the respective controls. Among the newly formed triterpenoids, 23-hydroxyursolic acid (u4b), hederagenin (o4b) and gypsogenic acid (o4d) were identified based on retention times (Fig. 7) and mass spectra (Fig. 8) matching those of (TMSi-derivatized) authentic standards. Using GC-MS analyses in total ion and single ion (m/z 189 and 203) modes, five more compounds were recognized as triterpenoids. Among them, 3-hydroxyurs-12-ene-23,28-dioic acid (u4d) was identified based on its MS fragmentation pattern (Fig. 9A) similar to that of gypsogenic acid (o4d) (Fig. 8C) and a GC retention time difference between both compounds (Fig. 7B) matching that between other isomeric ursane- and oleanane-type triterpenoid derivatives (compare, for example, the retention time difference between α-amyrin u1 and β-amyrin o1). Two other compounds were tentatively identified as 3-hydroxy-23-oxours-12-en-28-oic acid (u4c) and gypsogenin (o4c) based on their MS fragmentation patterns (Fig. 9B) and relative GC retention times (Fig. 7B). Finally, two more compounds detected in the transgenic yeast had MS fragmentation patterns consistent with 23-hydroxyuvaol (u2b) and 23-hydroxyursolic aldehyde (u3b), respectively (Fig. 9C); however, their exact structures could not be assigned based on the spectroscopic evidence acquired here. Fig. 7 View largeDownload slide Biochemical characterization of C. asiatica CYP714E19. GC-MS analysis of (A) extracts from yeast expressing CaDDS and CYP714E19, (B) extracts from yeast expressing CaDDS, CYP716A83 and CYP714E19, and (C) standards. Total ion counts (TICs) and extracted ion chromatograms (EICs) for fragments m/z 189 and m/z 203 are shown. All major compounds were identified based on GC and MS behavior (see Figs. 5, 8) matching authentic standards of δ-amyrin (d1), β-amyrin (o1), α-amyrin (u1), ψ-taraxasterol (p1), taraxasterol (t1), erythrodiol (o2a), uvaol (u2a), oleanolic acid (o4a), ursolic acid (u4a), 23-hydroxyursolic acid (u4b), hederagenin (o4b) and gypsogenic acid (o4d). Further products were tentatively identified based on fragmentation patterns (see Figs. 6, 7, 9) as urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a), 23-hydroxyuvaol (u2b), 23-hydroxyursolic aldehyde (u3b), 3-hydroxy-23-oxours-12-en-28-oic acid (u4c), gypsogenin (o4c) and 3-hydroxyurs-12-ene-23,28-dioic acid (u4d). IS, internal standard lupeol. Fig. 7 View largeDownload slide Biochemical characterization of C. asiatica CYP714E19. GC-MS analysis of (A) extracts from yeast expressing CaDDS and CYP714E19, (B) extracts from yeast expressing CaDDS, CYP716A83 and CYP714E19, and (C) standards. Total ion counts (TICs) and extracted ion chromatograms (EICs) for fragments m/z 189 and m/z 203 are shown. All major compounds were identified based on GC and MS behavior (see Figs. 5, 8) matching authentic standards of δ-amyrin (d1), β-amyrin (o1), α-amyrin (u1), ψ-taraxasterol (p1), taraxasterol (t1), erythrodiol (o2a), uvaol (u2a), oleanolic acid (o4a), ursolic acid (u4a), 23-hydroxyursolic acid (u4b), hederagenin (o4b) and gypsogenic acid (o4d). Further products were tentatively identified based on fragmentation patterns (see Figs. 6, 7, 9) as urs-20-ene-3,28-diol (alias heterobetulin, p2a) and 3-hydroxyurs-20-en-28-oic acid (alias heterobetulinic acid, p4a), 23-hydroxyuvaol (u2b), 23-hydroxyursolic aldehyde (u3b), 3-hydroxy-23-oxours-12-en-28-oic acid (u4c), gypsogenin (o4c) and 3-hydroxyurs-12-ene-23,28-dioic acid (u4d). IS, internal standard lupeol. Fig. 8 View largeDownload slide MS identification of triterpenoids from yeast expressing C. asiatica CYP714E19. (A) The spectrum of compound u4b detected in the yeast extract (top) matched that of a standard (bottom), identifying it as 23-hydroxyursolic acid. (B) The spectrum of compound o4b detected in the yeast extract (top) matched that of a standard (bottom), identifying it as hederagenin. (C) The spectrum of compound o4d detected in the yeast extract (top) matched that of a standard (bottom), identifying it as gypsogenic acid. Fig. 8 View largeDownload slide MS identification of triterpenoids from yeast expressing C. asiatica CYP714E19. (A) The spectrum of compound u4b detected in the yeast extract (top) matched that of a standard (bottom), identifying it as 23-hydroxyursolic acid. (B) The spectrum of compound o4b detected in the yeast extract (top) matched that of a standard (bottom), identifying it as hederagenin. (C) The spectrum of compound o4d detected in the yeast extract (top) matched that of a standard (bottom), identifying it as gypsogenic acid. Fig. 9 View largeDownload slide Tentative MS identification of triterpenoids from yeast expressing C. asiatica CYP714E19 based on fragmentation patterns. (A) Compounds tentatively identified as 23-hydroxyuvaol (u2b) and 23-hydroxyursolic aldehyde (u3b). (B) Compounds tentatively identified as 3-hydroxy-23-oxours-12-en-28-oic acid (u4c) and gypsogenin (o4c). (C) Compound tentatively identified as 3-hydroxyurs-12-ene-23,28-dioic acid (u4d). TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. Fig. 9 View largeDownload slide Tentative MS identification of triterpenoids from yeast expressing C. asiatica CYP714E19 based on fragmentation patterns. (A) Compounds tentatively identified as 23-hydroxyuvaol (u2b) and 23-hydroxyursolic aldehyde (u3b). (B) Compounds tentatively identified as 3-hydroxy-23-oxours-12-en-28-oic acid (u4c) and gypsogenin (o4c). (C) Compound tentatively identified as 3-hydroxyurs-12-ene-23,28-dioic acid (u4d). TMSi on the chemical structures indicate a trimethylsilyl derivative of the natural product. To assess the relative quantities of various triterpenoids in the transgenic yeast harboring CaDDS, CYP716A83 and CYP714E19, the abundances of the common fragment m/z 203 were quantified in all GC-MS peaks. The overall triterpenoid mixture consisted of 17% 23-hydroxyuvaol (u2b), 6% 23-hydroxyursolic aldehyde (u3b), 33% 23-hydroxyursolic acid (u4b), together with 2% 3-hydroxy-23-oxours-12-en-28-oic acid (u4c) and 4% 3-hydroxyurs-12-en-23,28-dioic acid (u4d), and also together with 30% hederagenin (o4b), 4% gypsogenin (o4c) and 5% gypsogenic acid (o4d). These relative amounts imply that the various structures with alcohol, aldehyde or acid functionalities at the C-23 position comprised approximately 86, 6 and 9%, respectively, showing that CYP714E19 plays a major role in oxidizing the C-23 methyl group of various triterpenoid precursors into corresponding 23-hydroxy compounds (i.e. alcohols), and only a lesser role in the two further oxidation steps to the corresponding aldehydes and acids. All yeast expression experiments were repeated three times, always with nearly identical results. Overall, our results show that CYP714E19 is a multifunctional triterpenoid C-23 oxidase, accepting uvaol and erythrodiol, as well as ursolic and oleanolic acids as substrates. Based on this result, the extracts from yeast co-expressing CaDDS and CYP714E19 were re-analyzed using GC-MS in single-ion mode (m/z 189, 203, 216 and 571) to search specifically for 23-hydroxy-α-amyrin and 23-hydroxy-β-amyrin (Supplementary Fig. S7). However, neither of the compounds was detected, indicating that the amyrins do not serve as substrates for the C-23 oxidase. Discussion In the present study, the transcriptome of MeJA-induced C. asiatica leaves was analyzed to identify candidate genes potentially involved in asiaticoside biosynthesis. It is well established that MeJA induction leads to asiaticoside accumulation in C. asiatica leaves (Kim et al. 2004, Mangas et al. 2008), and we therefore expected that genes encoding asiaticoside pathway enzymes should be up-regulated upon MeJA treatment. We hypothesized that at least three P450s with different regiospecificity probably oxidize the C-2α, C-23 and C-28 positions of the ursane skeleton starting from α-amyin as the initial substrate (Fig. 1). The transcriptome of C. asiatica leaves contained several P450 genes with sequence homology to enzymes involved in triterpenoid metabolism. Among them, candidate genes belonging to the CYP716A, CYP714E and CYP736A subfamilies were of particular interest due to their relatively high expression in leaves and MeJA-elicited expression, both matching the accumulation patterns of asiaticoside. We found that one of the candidate P450s, CYP716A83, converts α-amyrin (u1) and β-amyrin (o1) into ursolic acid (u4a) and oleanolic acid (o4a), respectively. Our results thus confirm an earlier report showing that this enzyme has C-28 oxidase activity on ursane and oleanane substrates (Miettinen et al. 2017). In the light of our gene expression data, we conclude that CYP716A83 is a crucial enzyme catalyzing the initial oxidation steps converting the first ursane skeleton precursors into intermediates dedicated to the formation of asiaticoside. Our analyses also corroborated previous evidence on C-28 oxidation intermediates of CYP716A83, on the one hand confirming the formation of erythrodiol (o2a) en route from β-amyrin (o1) to oleanolic acid (o4a) and adding GC-MS information to substantiate the formation of oleanolic aldehyde (o3a), and on the other hand identifying uvaol (u2a) and substantiating ursolic aldehyde (u3a) as intermediates between α-amyrin (u1) and ursolic acid (u4a). Of note, we also detected additional CYP716A83 products heterobetulin (p2a) and heterobetulinic acid (p4a), probably formed by C-28 oxidation of ψ-taraxasterol instead of the amyrins. The enzyme thus accepts various substrates with isomeric backbone structures, while being specific in its activity on only one carbon center in them. Taken together, our results suggest that CYP716A83 catalyzes a three-step oxidation, forming respective C-28 alcohol and C-28 aldehyde products in a first and second round of hydroxylation, respectively, and releasing at least a part of them. It remains to be determined whether the enzyme can also re-bind the same intermediates for further oxidation. The oleanolic acid and ursolic acid formed by CYP716A83 may not only serve as intermediates en route to asiaticoside, but may also be converted into other bioactive products in C. asiatica or play physiological roles themselves. Because oleanolic acid, ursolic acid and betulinic acid are widespread and relatively abundant triterpenoids in plants, Fukushima et al. (2011) speculated that CYP716As could have an important role in plant development or signaling, along with their intermediate role in triterpenoid pathways. Accordingly, Carelli et al. (2011) reported that M. truncatula mutants defective in CYP716A12 expression, with impeded synthesis of oleanolic acid, showed a strong decrease in growth, confirming a possible dual role for this triterpenoid in plant defense and development. However, the detailed physiological functions of CYP716A subfamily enzymes catalyzing the oxidation at C-28 of amyrins in other species are not clear. A second C. asiatica CYP716 subfamily enzyme, CYP716A85, had expression patterns suggesting an involvement in asiaticoside formation, and based on sequence similarities was predicted to oxidize triterpenoid substrates on positions other than C-28. Accordingly, this P450 (re-named as CYP716E41) had recently been found to catalyze the C-6 hydroxylation of ursolic acid (u4a) and oleanolic acid (o4a) (Miettinen et al. 2017), albeit at very low levels varying strongly between parallels. Here, we could not detect the respective products, possibly due to low substrate and/or enzyme amounts or due to other differences in experimental conditions. In our experiments, CYP716A85 also showed no activity on α-amyrin (u1) and β-amyrin (o1), similar to results of Miettinen et al. (2017). As a major result of the current investigations, we found that a third candidate enzyme, CYP714E19, oxidizes ursolic acid into 23-hydroxyursolic acid (u4b). Together with our expression data, this result suggests that CYP714E19 is the enzyme catalyzing the C-23 hydroxylation crucial for asiaticoside formation. Interestingly, the P450 enzyme also formed several minor side products with C-23 oxygen functions, including the 23-oxo (u4c) and 23-carboxylate (u4d) derivatives of ursolic acid probably produced by repeat C-23 oxidation, showing that it may either carry out multistep oxidations or re-bind oxidation products. However, the major activity of the enzyme was for the first round of oxidation, leading to the 23-hydroxy function. The CYP714 subfamily, along with the CYP72, 734, 749, 721 and 749 clades, belongs to the CYP72 clan of P450 enzymes. While many CYP72 members have been characterized, there is relatively little information on the plant CYP714A–D subfamily to date. Most interestingly, some CYP714 enzymes were found to play roles in gibberellin biosynthesis, as C-13 and C-16 oxidases as well as C-16α,17 epoxidases (Zhu et al. 2006, Magome et al. 2013, Wang et al. 2016). However, a phylogenetic analysis showed that the CYP714E19 protein of C. asiatica has relatively little sequence homology with other CYP714 subfamily proteins (Supplementary Fig. S8). Our findings for CYP714E19 are similar to those of Fukushima et al. (2013) for CYP72A68v2 in M. truncatula. Co-expression of the latter gene with the Lotus japonicus bAS, CPR and CYP716A12 genes in yeast demonstrated that CYP72A68v2 also catalyzed specific oxidation of C-23 of oleanolic acid. However, the activity of this C-23 oxidase was not tested on other substrates, such as ursolic acid, impeding further comparisons with our results on ursane-type substrates. Furthermore, CYP72A68v2 also catalyzes a series of oxidation reactions from oleanolic acid via hederagenin to gypsogenic acid, similar to CYP714E19. The yeast co-expressing CYP716A83 and CYP714E19 also produced two novel products tentatively identified as 23-hydroxyuvaol (u2b) and 23-hydroxyursolic aldehyde (u3b). All our GC retention and MS fragmentation data are in good agreement with these structure assignments; however, the identification cannot be confirmed by comparison with authentic standards at this point. Meanwhile, these putative 23-hydroxy ursane products suggest that CYP714E19 may also use substrates with various functional groups on C-28 for C-23 oxidation. This raises the possibility that CYP716A83, co-expressed in our experiments, may have further flexibility on its substrates and convert 23-hydroxyuvaol and 23-hydroxyursolic aldehyde by C-28 oxidation into 23-hydroxyursolic acid. The oxidations catalyzed by CYP716A83 may thus either precede or follow those of CYP714E19, leading to a flexible sequence of oxidations rather than a linear pathway. In order to test respective enzyme specificities and relative C-28 oxidation activities on various substrates, quantitative in vitro activity assays will have to be established. Interestingly, co-expression of CaDDS and CYP714E19 alone did not lead to the formation of 23-hydroxy-α-amyrin or 23-hydroxy-β-amyrin under the conditions tested here (Supplementary Fig. S7). This result suggests that CYP714E19 is specific for oleanolic acid and ursolic acid as substrates, but discriminates against the corresponding α- and β-amyrin precursors lacking a C-28 functionality, even though they have oleanane and ursane skeletons, respectively. It therefore seems very likely that the asiaticoside biosynthesis pathway proceeds via initial oxidation of C-28, yielding ursolic acid, followed by C-23 oxidation to 23-hydroxyursolic acid (and/or C-2 oxidation). Our finding that CYP714E19 accepts ursolic acid as substrate indicates that C-23 oxidation can occur without prior C-2α oxidation, leading us to predict a C-2α oxidase with substrate specificity for 23-hydroxyursolic acid. However, it cannot be ruled out that such a C-2α oxidase may also accept ursolic acid and that CYP714E19 may catalyze C-23 hydroxylation of the resulting corosolic acid. In a previous study, the C. asiatica enzyme CYP716C11 was shown to catalyze the C-2α hydroxylation of oleanolic acid or 6-hydroxyoleanolic acid, implicating this as another P450 involved in asiaticoside formation (Miettinen et al. 2017). However, it was not tested whether CYP716C11 may also accept, or even prefer, 23-hydroxyursolic acid as substrate, and it hence remains to be determined whether C-23 oxidation precedes or follows C-2 hydroxylation. Finally, it should be noted that C. asiatica leaves also contain not only the ursane-type triterpenoid classes of asiaticoside and madecassoside, but also oleanane-type saponins such as scheffoleoside A and asiaticoside B (Matsuda et al. 2001). All these compounds share 2α-, 6β- and 23-hydroxy functionalities in their oleanane- and ursane-type sapogenins and saponins. Since all our results showed similar activities of P450 enzymes on ursane- and oleanane-type substrates, it seems likely that the hydroxylases are involved in both asiaticoside and scheffoleoside A formation. In conclusion, this work led first to the identification of MeJA-responsive transcripts in leaves of C. asiatica, and on to the functional characterization of two P450s that are involved in asiaticoside biosynthesis. In particular, CYP714E19 was identified as the enzyme catalyzing the C-23 oxidation, probably of ursolic acid, and thus the last missing P450 along the asiaticoside pathway. Now that all the oxidases involved in asiaticoside formation are known, further investigations into the sequence of oxidation steps on the pathway can be envisaged. The toolbox of C. asiatica P450 enzymes will also enable the engineering of other, possibly novel, hydroxylated triterpenoids with carbon skeletons other than ursanes and oleananes, by co-expression of the various P450s with diverse OSC enzymes in yeast. Finally, it is now also possible to attempt (ectopic) overexpression of the P450s in C. asiatica or heterologous expression, for example in Nicotiana benthamiana, to confirm their biochemical functions and to engineer the formation of asiaticoside and its precursors. Materials and Methods Plant materials and sample preparation For expressed sequence tag (EST) sequencing, whole-plant cultures of C. asiatica were established from node segments as previously described by Kim et al. (2004). The node segments of four individual plants were cultured on 1/2 Murashige and Skoog (MS) liquid medium supplemented with 3% sucrose at 23 ± 2°C in 16/8 h light/dark cycles at 100 r.p.m. After 2 weeks of cultivation, MeJA (Sigma-Aldrich) was added to the medium to a concentration of 0.1 mM, and leaves were harvested 24 h after elicitation and immediately frozen in liquid nitrogen until use for total RNA extraction. For RT–PCR time-course analysis of gene expression, an established hairy root culture system was used as described by Kim et al. (2007). Total RNAs were extracted from hairy roots at 0, 12, 24, 48 and 72 h after induction with 0.1 mM MeJA. For RT–PCR analysis of gene expression in various organs, whole plants of C. asiatica were cultured on 1/2 MS medium supplemented with 3% sucrose and 0.8% agar adjusted to pH 5.8 at 23 ± 2°C in 16/8 h light/dark cycles. After cultivation for 6 weeks, total RNA was extracted from leaves, stems, nodes and roots. Transcriptome analysis In a previous study, the RNA samples of C. asiatica treated with MeJA were sequenced using the Genome sequencer FLX (454 Life Sciences, Roche) and information of the annotations based on the public databases and expression levels calculated using RPKM (reads per kilobase per mapped million reads) for unigenes was used to select candidates of the P450 gene associated with triterpenoid biosynthesis of asiaticoside (Kim et al. 2017). Phylogenetic analysis For phylogenetic analyses, the deduced amino acid sequences were aligned using the interactive web site CLUSTAL W (http://clustalw.ddbj.nig.ac.jp/top-j.html). A phylogenetic tree was generated using MEGA5.2 based on the Neighbor–Joining method (Tamura et al. 2011). Amino acid distances were calculated using the Dayhoff PAM matrix method. Bootstrap analysis with 1,000 replicates was used to assess the strength of nodes in the tree. Sequence information for 23 of the P450 genes of C. asiatica has been deposited in GenBank with accession numbers as shown in Table 1. RT–PCR analysis Total RNA was extracted from each sample and first-strand cDNA was synthesized using the AMV reverse transcriptase (Promega) and 2 μg of total RNA. SYBR Green Master Mix (Bio-Rad) was used for quantification with the CFX96™ Real-Time System (Bio-Rad) and gene-specific primers (Supplementary Table S3). The PCR conditions were as follows: initial denaturation at 95°C for 3 min, followed by 40 cycles of amplification for 15 s at 95°C, for 15 s at 58°C and for 30 s at 72°C. After completing the reactions, the threshold cycle (Ct) value for each reaction was obtained, and the differences were calculated using the delta-delta-Ct method and β-actin as internal control. The fold change in transcript levels of each gene (considered for quantitaive RT–PCR) is presented as the mean and SE of three independent experimental analyses. Generation of plasmid vectors The Invitrogen Gateway system was used to facilitate subcloning. All PCR primers are listed in Supplementary Table S3. The open reading frames (ORFs) were amplified by PCR from the original cDNA with attB-modified custom primers for P450s, and the PCR products were inserted into the pDONR/Zeo entry vector to make an entry clone, following the manufacturer’s protocol (Invitrogen). Plasmid DNAs were prepared from several transformants and sequenced. The yeast expression vectors were produced using the LR reaction (Invitrogen) with each of the entry vectors and the destination vector pYES-DEST52 (–Ura) for CaDDS, pAG423GAL-ccdB (–His) for CYP716A83, and pAG425GAL-ccdB (–Leu) for CYP714E19. After sequence confirmation, the pYES-CaDDS, pAG423-CYP716A83 and pAG425-CYP714E19 plasmids were used for yeast transformation. Co-expression of CaDDS and CYP716A83 or CYP714E19 in yeast In one experiment, CaDDS was expressed alone in Saccharomyces cerevisiae strain GIL77 as described before (Wang et al. 2011). In a second experiment, the yeast strain WAT21, which carries A. thaliana NADPH-CYP reductase (Pompon et al. 1996), was transformed with pYES-CaDDS or empty pYES2 as controls. After confirming the presence of amyrins in yeast harboring pYES-CaDDS and the absence of pentacyclic triterpenes in yeast with empty pYES2, both strains were further transformed with pAG423-CYP716A83 or pAG425-CYP714E19. Recombinant yeast cells were grown for 2 d in SC-Ura-His or SC-Ura-Leu selection media, including 2% glucose. Then cells were collected and used to inoculate 15 ml cultures; 2% galactose was added immediately to each culture to induce gene expression, and cells were grown for a further 3 d. Finally, cells were extracted twice with 10 ml of ethyl acetate at 90°C for 30 min with intermittent vortexing. After centrifuging at 4,000 r.p.m. for 5 min, the organic phase was transferred to a new vial and then dried under a stream of nitrogen gas. The extracts were re-suspended in 400 μl of chloroform–methanol (1:1), and 100 μl of the resulting solution was transferred to a vial and dried down under nitrogen gas. Finally, the hydroxyl-containing compounds in the mixture were converted into trimethylsilyl derivatives by reaction with 50 μl of bis-N,O-(trimethylsilyl)trifluoroacetamide for 30 min at 70°C. After confirming the formation of oleanolic acid and ursolic acid in yeast cells transformed with pYES-CaDDS/pAG423-CYP716A83, they were further transformed with pAG425-CYP714E19. The resulting yeast cells were selected in SC medium lacking uracil, histidine and leucine, induced with 2% galactose, grown for a further 3 d, extracted and analyzed as described above. GC-MS analysis GC-MS analysis was performed as described by Wang et al. (2011). A 1 μl aliquot of the yeast extract was analyzed using a 6890 N gas chromatograph (Agilent) equipped with a cool on-column injector, mass spectrometric detector (5973 N, Agilent) and an HP-1 capillary column (Agilent; length 30 m, i.d. 320 μm, 1 μm film thickness). The oven temperature was programmed to hold at 50°C for 2 min after injection, a 40°C min–1 ramp to 200°C, a plateau at 200°C for 2 min, a rise to 320°C at a rate of 3 min–1 and holding at 320°C for 30 min. Triterpenoids were identified by comparison with authentic compounds (erythrodiol, uvaol, oleanolic acid, ursolic acid, hederagenin, 23-hydroxyursolic acid and gypsogenic acid) and with literature data. All sapogenin standards were of analytical grade, all purchased from Sigma-Aldrich, except that 23-hydroxyursolic acid was kindly provided by Dr. Jinsook Kim (Korea Institute of Oriental Medicine). Supplementary Data Supplementary data are available at PCP online. Funding This work was supported by the Cooperative Research Program for Agricultural Science & Technology Development [Project No. PJ00849903]; the Rural Development Administation, Republic of Korea; the Natural Sciences and Engineering Research Council of Canada; and the Canada Foundation for Innovation. Acknowledgments We thank Dr. David Nelson for help with the nomenclature of P450 genes, Dr. Y.E. Choi for providing yeast strain WAT21, and Dr. J.S. Kim for providing the 23-hydroxyursolic acid standard. Disclosures The authors have no conflicts of interest to declare. References Augustin J.M. , Kuzina V. , Andersen S.B. , Bak S. ( 2011 ) Molecular activities, biosynthesis and evolution of triterpenoid saponins . Phytochemistry 72 : 435 – 457 . Google Scholar CrossRef Search ADS PubMed Bauer S. , Schulte E. , Thier H.-P. ( 2004 ) Composition of the surface wax from tomatoes. I. Identification of the components by GC/MS . Eur. Food Res. Technol. 219 : 223 – 228 . Google Scholar CrossRef Search ADS Brendolise C. , Yauk Y.-K. , Eberhard E.D. , Wang M. , Chagne D. , Andre C. , et al. . ( 2011 ) An unusual plant triterpene synthase with predominant α-amyrin-producing activity identified by characterizing oxidosqualene cyclases from Malus × domestica . FEBS J . 278 : 2485 – 2499 . Google Scholar CrossRef Search ADS PubMed Brinkhaus B. , Lindner M. , Schuppan D. , Hahn E.G. ( 2000 ) Chemical, pharmacological and clinical profile of the East Asian medical plant Centella asiatica . Phytomedicine 7 : 427 – 448 . Google Scholar CrossRef Search ADS PubMed Carelli M. , Biazzi E. , Panara F. , Tava A. , Scaramelli L. , Porceddu A. , et al. . ( 2011 ) Medicago truncatula CYP716A12 is a multifunctional oxidase involved in the biosynthesis of hemolytic saponins . Plant Cell 23 : 3070 – 3081 . Google Scholar CrossRef Search ADS PubMed De Geyter N. , Gholami A. , Goormachtig S. , Goossens A. ( 2012 ) Transcriptional machineries in jasmonate-elicited plant secondary metabolism . Trends Plant Sci . 17 : 349 – 359 . Google Scholar CrossRef Search ADS PubMed Field B. , Osbourn A.E. ( 2008 ) Metabolic diversification-independent assembly of operon-like gene clusters in different plants . Science 320 : 543 – 547 . Google Scholar CrossRef Search ADS PubMed Fukushima E.O. , Seki H. , Ohyama K. , Ono E. , Umemoto N. , Mizutani M. , et al. . ( 2011 ) CYP716A subfamily members are multifunctional oxidases in triterpenoid biosynthesis . Plant Cell Physiol . 52 : 2050 – 2061 . Google Scholar CrossRef Search ADS PubMed Fukushima E.O. , Seki H. , Sawai S. , Suzuki M. , Ohyama K. , Saito K. , et al. . ( 2013 ) Combinatorial biosynthesis of legume natural and rare triterpenoids in engineered yeast . Plant Cell Physiol . 54 : 740 – 749 . Google Scholar CrossRef Search ADS PubMed Geisler K. , Hughes R.K. , Sainsbury F. , Lomonossoff G.P. , Rejzek M. , Fairhurst S. , et al. . ( 2013 ) Biochemical analysis of a multifunctional cytochrome P450 (CYP51) enzyme required for synthesis of antimicrobial triterpenes in plants . Proc. Natl Acad. Sci. USA 110 : E3360 – E3367 . Google Scholar CrossRef Search ADS Han J.Y. , Hwang H.S. , Choi S.W. , Kim H.J. , Choi Y.E. ( 2012 ) Cytochrome P450 CYP716A53v2 catalyzes the formation of protopanaxatriol from protopanaxadiol during ginsenoside biosynthesis in Panax ginseng . Plant Cell Physiol . 53 : 1535 – 1545 . Google Scholar CrossRef Search ADS PubMed Han J.Y. , Kim H.J. , Kwon Y.S. , Choi Y.E. ( 2011 ) The Cyt P450 enzyme CYP716A47 catalyzes the formation of protopanaxadiol from dammarenediol-II during ginsenoside biosynthesis in Panax ginseng . Plant Cell Physiol . 52 : 2062 – 2073 . Google Scholar CrossRef Search ADS PubMed Han J.Y. , Kim M.J. , Ban Y.W. , Hwang H.S. , Choi Y.E. ( 2013 ) The involvement of β-amyrin 28-oxidase (CYP716A52v2) in oleanane-type ginsenoside biosynthesis in Panax ginseng . Plant Cell Physiol . 54 : 2034 – 2046 . Google Scholar CrossRef Search ADS PubMed Huang L. , Li J. , Ye H. , Li C. , Wang H. , Liu B. , et al. . ( 2012 ) Molecular characterization of the pentacyclic triterpenoid biosynthetic pathway in Catharanthus roseus . Planta 236 : 1571 – 1581 . Google Scholar CrossRef Search ADS PubMed James J.T. , Dubery I.A. ( 2009 ) Pentacyclic triterpenoids from the medicinal herb, Centella asiatica (L.) Urban . Molecules 14 : 3922 – 3941 . Google Scholar CrossRef Search ADS PubMed Kim O.T. , Bang K.H. , Shin Y.S. , Lee M.J. , Jung S.J. , Hyun D.Y. , et al. . ( 2007 ) Enhanced production of asiaticoside from hairy root cultures of Centella asiatica (L.) Urban elicited by methyl jasmonate . Plant Cell Rep. 26 : 1941 – 1949 . Google Scholar CrossRef Search ADS PubMed Kim O.T. , Jin M.L. , Lee D.Y. , Jetter R. ( 2017 ) Characterization of the asiatic acid glucosyltransferase, UGT73AH1, involved in asiaticoside biosynthesis in Centella asiatica (L.) Urban . Int. J. Mol. Sci . 18 : 2630 . Google Scholar CrossRef Search ADS Kim O.T. , Kim M.Y. , Ahn J.C. , Hwang S.J. , Hwang B. ( 2005 ) Cloning and molecular analysis of cDNA encoding cycloartenol synthase from Centella asiatica (L.) Urban . Biotechnol. Bioproc. Eng. 10 : 16 – 22 . Google Scholar CrossRef Search ADS Kim O.T. , Kim M.Y. , Hong M.H. , Ahn J.C. , Hwang B. ( 2004 ) Stimulation of asiaticoside production from Centella asiatica whole plant cultures by elicitors . Plant Cell Rep. 23 : 339 – 344 . Google Scholar CrossRef Search ADS PubMed Kim O.T. , Lee J.W. , Bang K.H. , Kim Y.C. , Hyun D.Y. , Cha S.W. , et al. . ( 2009 ) Characterization of a dammarenediol synthase in Centella asiatica (L.) Urban . Plant Physiol. Biochem . 47 : 998 – 1002 . Google Scholar CrossRef Search ADS PubMed Kranz-Finger S. , Mahmoud O. , Ricklefs E. , Ditz N. , Bakkes P.J. , Urlacher V.B. ( 2018 ) Insights into the functional properties of the marneral oxidase CYP71A16 from Arabidopsis thaliana . Biochim. Biophys. Acta 1866 : 2 – 10 . Google Scholar CrossRef Search ADS PubMed Krokida A. , Delis C. , Geisler K. , Garagounis C. , Tsikou D. , Pena-Rodriguez L.M. , et al. . ( 2013 ) A metabolic gene cluster in Lotus japonicus discloses novel enzyme functions and products in triterpene biosynthesis . New Phytol. 200 : 675 – 690 . Google Scholar CrossRef Search ADS PubMed Lambert E. , Faizal A. , Geelen D. ( 2011 ) Modulation of triterpene saponin production: in vitro cultures, elicitation, and metabolic engineering . Appl. Biochem. Biotechnol. 164 : 220 – 237 . Google Scholar CrossRef Search ADS PubMed Magome H. , Nomura T. , Hanada A. , Takeda-Kamiya N. , Ohnishi T. , Shinma Y. , et al. . ( 2013 ) CYP714B1 and CYP714B2 encode gibberellin 13-oxidases that reduce gibberellin activity in rice . Proc. Natl. Acad. Sci. USA 110 : 1947 – 1952 . Google Scholar CrossRef Search ADS Mangas S. , Bonfill M. , Osuna L. , Moyano E. , Tortoriello J. , Cusido R.M. , et al. . ( 2006 ) The effect of methyl jasmonate on triterpene and sterol metabolisms of Centella asiatica, Ruscus aculeatus and Galphimia glauca cultured plants . Phytochemistry 67 : 2041 – 2049 . Google Scholar CrossRef Search ADS PubMed Mangas S. , Moyano E. , Osuna L. , Cusido R.M. , Bonfill M. , Palazón J. ( 2008 ) Triterpenoid saponin content and the expression level of some related genes in calli of Centella asiatica . Biotechnol. Lett . 32 : 315 – 319 . Matsuda H. , Morikawa T. , Ueda H. , Yoshikawa M. ( 2001 ) Medicinal foodstuffs. XXVII. Saponin constituents Gotu Kola (2): structures of new ursane- and oleanane-type triterpene oligoglycosides, centellasaponin B, C, and D, from Centella asiatica cultivated in Sri Lanka . Chem. Pharm. Bull . 49 : 1368 – 1371 . Google Scholar CrossRef Search ADS PubMed Miettinen K. , Pollier J. , Buyst D. , Arendt P. , Csuk R. , Sommerwerk S. , et al. . ( 2017 ) The ancient CYP716 family is a major contributor to the diversification of eudicot triterpenoid biosynthesis . Nat. Commun. 8 : 14153. Google Scholar CrossRef Search ADS PubMed Moses T. , Pollier J. , Almagro L. , Buyst D. , Montagu M.V. , Pedreño M.A. , et al. . ( 2014a ) Combinatorial biosynthesis of sapogenins and saponins in Saccharomyces cerevisiae using a C-16α hydroxylase from Bupleurum falcatum . Proc. Natl. Acad. Sci. USA 111 : 1634 – 1639 . Google Scholar CrossRef Search ADS Moses T. , Pollier J. , Faizal A. , Apers S. , Pieters L. , Thevelein J.M. , et al. . ( 2015a ) Unraveling the triterpenoid saponin biosynthesis of the African shrub Maesa lanceolata . Mol. Plant 8 : 122 – 135 . Google Scholar CrossRef Search ADS Moses T. , Pollier J. , Shen Q. , Soetaert S. , Reed J. , Erffelinck M.L. , et al. . ( 2015 ) OSC2 and CYP716A14v2 catalyze the biosynthesis of triterpenoids for the cuticle of aerial organs of Artemisia annua . Plant Cell 27 : 286 – 301 . Google Scholar CrossRef Search ADS PubMed Moses T. , Thevelein J.M. , Goossens A. , Pollier J. ( 2014b ) Comparative analysis of CYP93E proteins for improved microbial synthesis of plant triterpenoids . Phytochemistry 108 : 47 – 56 . Google Scholar CrossRef Search ADS Nelson D. , Werck-Reichhart D. ( 2011 ) A P450-centric view of plant evolution . Plant J. 66 : 194 – 211 . Google Scholar CrossRef Search ADS PubMed Phillips D.R. , Rasbery J.M. , Bartel B. , Matsuda S.P.T. ( 2006 ) Biosynthetic diversity in plant triterpene cyclization . Curr. Opin. Plant Biol . 9 : 305 – 314 . Google Scholar CrossRef Search ADS PubMed Pompon D. , Louerat B. , Bronine A. , Urban P. ( 1996 ) Yeast expression of animal and plant P450s in optimized redox environments . Methods Enzymol . 272 : 51 – 64 . Google Scholar CrossRef Search ADS PubMed Qi X. , Bakht S. , Qin B. , Leggett M. , Hemmings A. , Mellon F. , et al. . ( 2006 ) A different function for a member of an ancient and highly conserved cytochrome P450 family: from essential sterols to plant defense . Proc. Natl. Acad. Sci. USA 103 : 18848 – 18853 . Google Scholar CrossRef Search ADS Seki H. , Ohyama K. , Sawai S. , Mizutani M. , Ohnishi T. , Sudo H. , et al. . ( 2008 ) Licorice β-amyrin 11-oxidase, a cytochrome P450 with a key role in the biosynthesis of the triterpene sweetener glycyrrhizin . Proc. Natl. Acad. Sci. USA 105 : 14204 – 14209 . Google Scholar CrossRef Search ADS Seki H. , Sawai S. , Ohyama K. , Mizutani M. , Ohnishi T. , Sudo H. , et al. . ( 2011 ) Triterpene functional genomics in licorice for identification of CYP72A154 involved in the biosynthesis of glycyrrhizin . Plant Cell 23 : 4112 – 4123 . Google Scholar CrossRef Search ADS PubMed Seki H. , Tamura K. , Muranaka T. ( 2015 ) P450s and UGTs: key players in the structural diversity of triterpenoid saponins . Plant Cell Physiol. 56 : 1463 – 1471 . Google Scholar CrossRef Search ADS PubMed Tamura K. , Peterson D. , Peterson N. , Stecher G. , Nei M. , Kumar S. ( 2011 ) MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods . Mol. Biol. Evol . 28 : 2731 – 2739 . Google Scholar CrossRef Search ADS PubMed Thimmappa R. , Geisler K. , Louveau T. , O’Maille P. , Osbourn A. ( 2014 ) Triterpene biosynthesis in plants . Annu. Rev. Plant Biol. 65 : 225 – 257 . Google Scholar CrossRef Search ADS PubMed Wang C. , Yang Y. , Wang H. , Ran X. , Li B. , Zhang J. , et al. . ( 2016 ) Ectopic expression of a cytochrome P450 monooxygenase gene PtCYP714A3 from Populus trichocarpa reduces shoot growth and improves tolerance to salt stress in transgenic rice . Plant Biotechnol. J . 14 : 1838 – 1851 . Google Scholar CrossRef Search ADS PubMed Wang Z. , Guhling O. , Yao R. , Li F. , Yeats T.H. , Rose J.K. , et al. . ( 2011 ) Two oxidosqualene cyclases responsible for biosynthesis of tomato fruit cuticular triterpenoids . Plant Physiol . 155 : 540 – 552 . Google Scholar CrossRef Search ADS PubMed Wasternack C. ( 2007 ) Jasmonates: an update on biosynthesis, signal transduction and action in plant stress response, growth and development . Ann. Bot . 100 : 681 – 697 . Google Scholar CrossRef Search ADS PubMed Xu R. , Fazio G.C. , Matsuda S.P.T. ( 2004 ) On the origins of triterpenoid skeletal diversity . Phytochemistry 65 : 261 – 291 . Google Scholar CrossRef Search ADS PubMed Yasumoto S. , Fukushima E.O. , Seki H. , Muranaka T. ( 2016 ) Novel triterpene oxidizing activity of Arabidopsis thaliana CYP716A subfamily enzymes . FEBS Lett. 590 : 533 – 540 . Google Scholar CrossRef Search ADS PubMed Yasumoto S. , Seki H. , Shimizu Y. , Fukushima E.O. , Muranaka T. ( 2017 ) Functional characterization of CYP716 family P450 enzymes in triterpenoid biosynthesis in tomato . Front. Plant Sci. 8 : 21 . Google Scholar CrossRef Search ADS PubMed Yu F. , Thamm A.M. , Reed D. , Villa-Ruano N. , Quesada A.L. , Gloria E.L. , et al. . ( 2013 ) Functional characterization of amyrin synthase involved in ursolic acid biosynthesis in Catharanthus roseus leaf epidermis . Phytochemistry 91 : 122 – 127 . Google Scholar CrossRef Search ADS PubMed Zhang J. , Dai L. , Yang J. , Liu C. , Men Y. , Zeng Y. , et al. . ( 2016 ) Oxidation of cucurbitadienol catalyzed by CYP87D18 in the biosynthesis of bogrosides from Siraitia grosvenorii . Plant Cell Physiol. 57 : 1000 – 1007 . Google Scholar CrossRef Search ADS PubMed Zhao J. , Davis L.C. , Verpoorte R. ( 2005 ) Elicitor signal transduction leading to production of plant secondary metabolites . Biotechnol. Adv . 23 : 283 – 333 . Google Scholar CrossRef Search ADS PubMed Zhu Y. , Nomura T. , Xu Y. , Zhang Y. , Peng Y. , Mao B. , et al. . ( 2006 ) ELONGATED UPPERMOST INTERNODE encodes a cytochrome P450 monooxygenase that epoxidizes gibberellins in a novel deactivation reaction in rice . Plant Cell 18 : 442 – 456 . Google Scholar CrossRef Search ADS PubMed Abbreviations Abbreviations EIC extracted ion chromatogram GC-MS gas chromatography–mass spectrometry MeJA methyl jasmonate MS Murashige and Skoog OSC oxidosqualene cyclase P450 Cyt P450 RT–PCR real-time reverse transcription–PCR TIC total ion count © The Author(s) 2018. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Plant and Cell PhysiologyOxford University Press

Published: Mar 22, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off