Metabolic Engineering Reveals LUP5 as a Determinant of Saponin Composition and Insect Resistance in Barbarea vulgaris

Metabolic Engineering Reveals LUP5 as a Determinant of Saponin Composition and Insect Resistance in Barbarea vulgaris | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results Metabolic Engineering Reveals LUP5 as a Determinant of Saponin Composition and Insect Resistance in Barbarea vulgaris View ORCID Profile Jincheng Shen , View ORCID Profile Jan Günther , View ORCID Profile Sebastian Kjeldgaard-Nintemann , View ORCID Profile Pablo D. Cárdenas , View ORCID Profile Søren Bak doi: https://doi.org/10.1101/2025.08.04.665341 Jincheng Shen 1 Department of Plant and Environmental Sciences, Faculty of Science, University of Copenhagen , 1871 Frederiksberg C, Denmark Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jincheng Shen Jan Günther 1 Department of Plant and Environmental Sciences, Faculty of Science, University of Copenhagen , 1871 Frederiksberg C, Denmark Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jan Günther Sebastian Kjeldgaard-Nintemann 1 Department of Plant and Environmental Sciences, Faculty of Science, University of Copenhagen , 1871 Frederiksberg C, Denmark Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Sebastian Kjeldgaard-Nintemann Pablo D. Cárdenas 1 Department of Plant and Environmental Sciences, Faculty of Science, University of Copenhagen , 1871 Frederiksberg C, Denmark Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Pablo D. Cárdenas For correspondence: bak{at}plen.ku.dk Søren Bak 1 Department of Plant and Environmental Sciences, Faculty of Science, University of Copenhagen , 1871 Frederiksberg C, Denmark Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Søren Bak For correspondence: pdcardenas{at}plen.ku.dk Abstract Full Text Info/History Metrics Preview PDF Summary Plant–insect coevolution has been a major driver of specialized metabolite diversification, yet the genetic basis of natural variation in defensive chemistry remains poorly understood. The wild crucifer Barbarea vulgaris comprises two ecotypes, an insect-resistant G-type and a susceptible P-type, characterized by distinct triterpenoid saponin profiles. To investigate the causal relationship between saponin composition and insect resistance, we established a stable transformation system for B. vulgaris . Expression of the G-type β-amyrin synthase gene LUP5 in the susceptible P-type conferred up to a 95% reduction in Plutella xylostella feeding, accompanied by increased accumulation of three hederagenin-derived monodesmosidic saponins. Comparison of LUP5 expression driven by its native promoter and by the constitutive 35S promoter revealed that the native promoter leads to increased hederagenin accumulation and is activated later in development, which may prevent early metabolic stress and allow coordinated expression of downstream pathway genes. In contrast, silencing of CYP72A552 by RNAi decreased total hederagenin levels by approximately 40% without affecting resistance, indicating threshold-dependent defense. Our results provide direct in planta evidence that LUP5 is a key determinant of natural variation in insect resistance in B. vulgaris , underscoring the pivotal role of the saponin backbone in herbivore deterrence. By linking promoter activity to metabolite structural diversity, this work provides mechanistic and conceptual insight into how plants coordinate specialized metabolism and defense. Introduction Plants have evolved a wide array of specialized metabolites to deter insect herbivores ( Fürstenberg-Hägg, Zagrobelny and Bak, 2013 ; Yactayo-Chang et al., 2020 ; Beran and Petschenka, 2022 ). Among these, triterpenoid saponins, which consist of a hydrophobic sapogenin backbone and one or more hydrophilic sugar moieties, represent a structurally diverse and ecologically important class of defensive compounds. Due to their structural similarity to sterols and steroid hormones, triterpenoid saponins have been hypothesized to interact with eukaryotic cell membranes ( Cárdenas, Almeida and Bak, 2019 ). In addition, their inherent bitterness may deter herbivores, contributing to their role as effective defense agents ( Augustin et al., 2011 ). Thousands of triterpenoid saponins have been reported to date ( Xu, Fazio and Matsuda, 2004 ; Augustin et al., 2011 ; Moses, Papadopoulou and Osbourn, 2014 ; Huang et al., 2020 ). Importantly, specific saponin structures have been implicated in insect resistance, highlighting the functional significance of their chemical diversity ( Stevenson, Nyirenda and Veitch, 2010 ; Cui et al., 2019 ; Hussain et al., 2019 ; Liu et al., 2019 ; Dolma et al., 2021 ). The structure–activity relationships underlying saponin bioactivity are critical to their defensive function. For example, hederagenin 3- O -monoglucoside has higher toxicity and antifeedant activity than oleanolic acid 3- O -monoglucoside (which lacks C-23 hydroxylation), and gypsogenic acid 3- O -monoglucoside (which is further oxidized at C-23 to the ketone) ( Liu et al., 2019 ). Additionally, monodesmosidic hederagenin and oleanolic acid exhibited higher toxicity and antifeedant activity than their bidesmosidic counterparts ( Tian et al., 2021 ; Dervishi et al., 2025 , in press). These observations suggest that even subtle modifications to the sapogenin backbone, glycosylation pattern, or sugar linkage can significantly alter saponin bioactivity. This structural diversity makes saponins a valuable system for studying molecular evolutionary mechanisms of specialized metabolism, particularly the structure– activity relationships underlying their ecological functions in insect defense. Despite growing knowledge of these structure–activity relationships, the genetic and biochemical mechanisms underlying saponin diversification remain poorly understood in most species. One particularly promising ecological model to explore this question is the wild crucifer Barbarea vulgaris , which naturally segregates in two ecotypes or chemotypes that alter in both saponin profiles and insect resistance. B. vulgaris has emerged as a powerful ecological model for investigating how specialized metabolites contribute to herbivore resistance ( Kuzina et al., 2011 ; Hauser, Toneatto and Nielsen, 2012 ; Byrne et al., 2017 ). The species comprises two ecotypes: the glabrous insect-resistant G-type, and the pubescent insect-susceptible P-type ( Agerbirk, Olsen and Nielsen, 2001 ). This intraspecific divergence likely arose from isolation in different Ice Age refugia ( Hauser, Toneatto and Nielsen, 2012 ) and is associated with marked differences in saponin composition ( Kuzina et al., 2009 ; Khakimov et al., 2012 ; Khakimov et al., 2016 ). Gene duplication events have further contributed to the chemical divergence and insect resistance ( Khakimov et al., 2015 ; Liu et al., 2019 ). Insect resistance in G-type B. vulgaris correlates with four oleanane-type defensive saponins, including oleanolic acid cellobioside, hederagenin cellobioside, gypsogenin cellobioside, and 4-epihederagenin cellobioside ( Agerbirk et al., 2003 ; Kuzina et al., 2011 ). Among these, hederagenin cellobioside is particularly toxic and confers G-type insect resistance ( Agerbirk et al., 2003 ; Nielsen et al., 2010 ; Kuzina et al., 2011 ). The biosynthetic pathway leading to hederagenin cellobioside in G-type B. vulgaris has been elucidated up to the first glycosylation step ( Fig. 1 ). The oxidosqualene cyclase LUP5 is highly expressed in the G-type and primarily produces β-amyrin ( Khakimov et al., 2015 ). Subsequent enzymatic steps involve CYP716A80, CYP72A552, and UGT73C11, which sequentially convert β-amyrin into oleanolic acid, hederagenin, and hederagenin 3- O -glucoside ( Fig. 1 ) ( Augustin et al., 2012 ; Khakimov et al., 2015 ; Liu et al., 2019 ). The enzyme responsible for the final step in forming hederagenin cellobioside remains unidentified ( Erthmann, Agerbirk and Bak, 2018 ). In contrast, LUP5 is expressed at very low levels in B. vulgaris P-type and primarily produces α-amyrin due to two amino acid substitutions as compared to the LUP5 in the G-type ( Günther et al., 2021 ). Despite 98% sequence identity at the amino acid level, the G- and P-type LUP5 enzymes differ markedly in both their gene expression and product profiles. The P-type plants preferentially express LUP2, an oxidosqualene cyclase over 80% identical at the amino acid level to LUP5 , but it produces predominantly lupeol (98%) and only 2% β-amyrin-derived saponins ( Khakimov et al., 2015 ). Notably, the genes encoding for the remaining enzymes responsible for the biosynthesis of hederagenin 3- O -glucoside (CYP716A80, CYP72A552, and UGT73C11) are present in both P-type and G-type and exhibit similar substrate and product specificity ( Augustin et al., 2012 ; Khakimov et al., 2015 ; Liu et al., 2019 ). However, the cyclases are differentially regulated suggesting that upstream regulation and oxidosqualene cyclase product specificity may be key drivers of saponin structural and insect resistance difference. Additionally, CYP72A552, which catalyzes the C23 hydroxylation of oleanolic acid to hederagenin, the precursor of the more toxic hederagenin-type saponins, has not yet been validated in planta for its ecological contribution to insect resistance in B. vulgaris . Download figure Open in new tab Fig. 1. Biosynthetic pathways for lupeol and β–amyrin derived triterpenoid saponins in B. vulgaris . Monodesmosidic saponin, a single sugar chain attached at preferentially C-3 (R1) or C-28 (R2); bidesmosidic saponin, sugar chains attached at both C-3 (R1) and C-28 (R2). To investigate the genetic and biochemical basis of saponin-mediated insect resistance, we leveraged the natural variation in B. vulgaris to develop a model system linking in planta metabolic engineering of specialized metabolism with ecological function. We provide functional evidence that engineering saponin biosynthesis through manipulation of LUP5 expression substantially enhances insect resistance in this wild crucifer and explains the molecular basis of the difference in insect resistance between the G- and P-type B. vulgaris . By comparing constructs driven by the native LUP5 and constitutive 35S promoters, we further reveal how promoter activity shapes metabolite structural diversity. These findings offer new mechanistic insight into plant–insect coevolution and established a model system for dissecting genotype– phenotype relationships in plant chemical specialization. Materials and methods Plant and insect materials Seeds of Barbarea vulgaris G-type (glabrous, accession B44, Herbarium-code: CP0057358) and P-type (pubescent, accession B4, herbarium-code: CP0057347) were provided by Associate professor Niels Agerbirk ( Agerbirk et al., 2021 ) and used for stable transformation. Both transformed and wild-type plants were cultivated in soil (Krukvaxtjord med lera och kisel, SW Horto AB, Hammenhog, Sweden) in a greenhouse maintained at 19 °C under a 16 h light/8 h dark photoperiod. Irradiance was supplemented with LED lamps whenever natural light intensity dropped below 250 µmol m -2 s -1 , with full-spectrum sunlight available when curtains were open. Plants were fertigated on Thursdays and Fridays with a nutrient solution (electrical conductivity 2.2 mS cm -1 ; pH 5.8) prepared by dissolving 100 g L -1 of a compound fertilizer (Pioner Basis 13-2-23+3+ME; containing 13% N, 2% P ₂ O ₅ , 23% K ₂ O, 3% MgO, plus micronutrients) together with 0.5 g L -1 of iron chelate (Pioner Fe-EDDHA 6%; 6% Fe) in water. Brassica napus plants were grown in the same greenhouse under identical conditions. Plutella xylostella (diamondback moth) eggs were obtained from Dr. Patrick Hughes (Boyce Thompson Institute, Ithaca, NY, USA). The laboratory colony was originally established at the New York State Agricultural Experiment Station (Geneva, NY, USA) in 1994. The insects were reared on B. napus at 20 °C under a 16 h light/8 h dark photoperiod in cages. Third instar P. xylostella larvae were used for the insect feeding assay. Plasmids construction For stable transformation, the pJCV51 vector was employed to introduce the G-type LUP5 coding sequence and to assess tissue-specific promoter activity (native LUP5 and constitutive 35S), while pK7FWG2 was utilized for CYP72A552 silencing ( Karimi, Inzé and Depicker, 2002 ). The LUP5 coding sequence, its native promoter (1988 bp), and the CYP72A552 coding sequence were PCR-amplified from B. vulgaris G-type cDNA and genomic DNA ( Supplementary Table S1 and S2 ). G-type LUP5 expression was driven either by the 35S promoter or its native promoter, while CYP72A552 was driven by the 35S promoter. View this table: View inline View popup Download powerpoint Tab. S1. Primers used in this study View this table: View inline View popup Download powerpoint Tab. S2. Nucleotide sequence of the G-type LUP5 promoter The pJCV51-p35S::LUP5, pJCV51-p35S::eGFP, and pK7FWG2-CYP72A552 RNAi plasmids were assembled using Gateway technology (Gateway™ BP Clonase™ II Enzyme mix, Thermo Fisher Scientific, 11789020; Gateway™ LR Clonase™ II Enzyme mix, Thermo Fisher Scientific, 11791020) (Supplementary Table S1). pDONR207 was used as the entry clone vector for both constructs (Thermo Fisher Scientific, 117207-021). Subsequently, the pJCV51-pLUP5::LUP5 and pJCV51-pLUP5::eGFP plasmids were generated by replacing the 35S promoter in pJCV51-p35S::LUP5 or pJCV51-p35S::eGFP with the G-type LUP5 promoter using the SalI sites. Finally, the plasmids were introduced into Agrobacterium tumefaciens strain AGL1 via electroporation ( Hayta et al., 2021 ). Stable transformation in wild crucifer B. vulgaris Seed sterilization was achieved by immersing seeds in 70% ethanol for 30 s, followed by a 12.5 min treatment in 3% sodium hypochlorite (Thermo Fisher Scientific) with 0.1% Tween20 (Bio-RAD, 1706531) and thorough rinsing with sterile water. Seeds were sown on germination medium in Magenta boxes and incubated in darkness for 6–8 days ( Supplementary Table S3 ). View this table: View inline View popup Download powerpoint Tab. S3. Meida for tissue culture Hypocotyl explants (∼3 mm) from 6–10-day-old seedlings were incubated for 15 min with A. tumefaciens strains (OD₆₀₀ = 0.5 in 10 mM MgCl₂, 100 µM acetosyringone) with gentle agitation and then co-cultivated in darkness for three days ( Supplementary Table S3 ). The pJCV51-pLUP5::LUP5 and pJCV51-35S::LUP5 constructs were transformed into both B. vulgaris ecotypes, while pJCV51-p35S::eGFP, pJCV51-pLUP5::eGFP, and pK7FWG2-CYP72A552 RNAi were introduced only into the G-type. After rinsing in Milli-Q water containing 1 µL/mL timentin, the explants were transferred sequentially to callus induction (one week), shoot induction (two rounds, two weeks each), and root induction media (two rounds, two weeks each) at 25°C under a 16:8 h light–dark cycle ( Supplementary Table S3 ). Throughout these stages, the explants were maintained in a climate chamber at 25°C under a 16:8 h light-dark cycle. Kanamycin (50 ng/mL) was used for selecting transformed explants, whereas wild-type explants were cultured without kanamycin supplementation in the medium. CYP72A552 RNAi plants were confirmed by NPTII (neomycin phosphotransferase II) ELISA Kit (Agdia, PSP 73000/0288). All the other transgenic plants were confirmed by RFP fluorescence (for LUP5 expression and promoter activity constructs). Fluorescence was analyzed on a Leica M205FA fluorescence dissection microscope (Leica Microsystems). RFP and eGFP were imaged using the dsRed plant (excitation 546/10 nm; emission 600/40 nm) or ET GFP filters (excitation 470/40 nm; emission 525/50 nm), respectively. Subsequently, the confirmed plants were transferred to soil and grown in a greenhouse at 19°C under a 16:8 h light–dark cycle. Choice insect feeding assay Insect resistance was evaluated using at least three individual plants (biological replicates) per transgenic line. For each transgenic plant, five transgenic and five wild-type leaf discs (1.57 cm² each) were alternately arranged in a petri dish (9.4 x 1.6 cm) (Greiner BIO-ONE, 633180) with filter paper moistened with 2 mL of water. After a 3– 5-hour starvation period, ten third-instar P. xylostella larvae were placed at the center of the plate, following a modified protocol from Liu et al ( Liu et al., 2019 ). Leaf consumption was quantified when half of the wild-type leaves were consumed using ImageJ (approx 7.5 h). Preparation of plant extracts and metabolite analysis The extraction of saponins and sapogenins was adapted from Khakimov et al. (2016) . B. vulgaris leaf samples were ground with liquid nitrogen using a mortar and pestle. For saponin detection, 100 mg of leaf powder was weighed into 1.5 mL Eppendorf tubes and extracted with 300 µL of 85% methanol (v/v) via sonication for 30 minutes, followed by centrifugation at 16,000 × g for 10 minutes. Supernatants (200 µL) were filtered through a 0.22-μm filter before LC-qTOF-ESI-MS/MS analysis. To detect total and free sapogenins, 200 mg of leaf powder was weighed into 2 mL Eppendorf tubes and extracted with 600 µL of 85% methanol (v/v) at 100°C while mixing at 1,400 × rpm in a thermomixer (Eppendorf, Denmark). The samples were cooled on ice and centrifuged at 16,000 × g for 3 min. After filtration through a 0.22- μm filter, 100 µL of supernatant was collected for free sapogenin detection via LC-qTOF-APCI-MS/MS. An additional 300 µL of supernatant was collected and dried under nitrogen flow. To cleave off sugar residues, samples were treated with 500 µL of 2 M hydrochloric acid at 100°C while mixing at 1,400 × rpm for 1.5 hours in a thermomixer (Eppendorf, Denmark). After cooling on ice, a double volume of ethyl acetate was added to the acid-water mixture, followed by vortexing and centrifugation at 3,500 × g for 5 minutes. The samples were extracted three times with ethyl acetate to collect sapogenins. To remove residual acid, an equal volume of Milli-Q water was added, the samples were subsequently vortexed and centrifugated at 3,500 × g for 5 minutes. This washing step was repeated three times. The final extracts were dried under nitrogen flow, re-solubilized in 300 µL of 100% methanol, and filtered through a 0.22- μm filter before total sapogenin detection via LC-qTOF-APCI-MS/MS. LC-qTOF-ESI-MS/MS was performed according to Liu et al. (2019) . LC-qTOF-APCI-MS/MS was performed as described by Patel et al. (2020) . Untargeted metabolite data were analyzed using XCMS ( Tautenhahn et al., 2012 ), and targeted saponin and sapogenin analyses were conducted with DataAnalysis software (v.4.3; Bruker, Bremen, Germany) following Khakimov et al. (2016) . Hederagenin cellobioside and oleanolic acid cellobioside standards were used for saponin identification. Lupeol, β- amyrin, oleanolic acid, and hederagenin standards were employed for sapogenin identification. Gene expression analysis Total RNA was extracted using the Spectrum™ Plant Total RNA Kit (Sigma-Aldrich, STRN50-1KT) and then synthesized to cDNA by iScript™ cDNA Synthesis Kit (Bio-Rad, 1708891). Relative expression of LUP5 was quantified via qPCR using kapa SYBR® fast kit (Merck-Sigma, KK4607) ( Khakimov et al., 2015 ). Tubulin (GenBank accession no. EU555191 ) was used as reference gene, and LUP5 expression levels were normalized to the wild-type P-type control (Liu TongJin et al., 2016). Primer sequences are provided in Supplementary Table S1 ( Khakimov et al., 2015 ). Results Establishment of a transformation and regeneration system in the wild crucifer B. vulgaris To enable functional studies of insect resistance in B. vulgaris as an ecological model system, we developed a stable transformation and regeneration method for both the G-type and P-type ecotypes ( Fig.1 , Supplementary Fig. S1 , Supplementary Table S3 ). To metabolically engineer the saponin compositions, we introduced the G-type LUP5 coding sequence into both ecotypes and attempted to silence CYP72A552 in the G-type. LUP5 expression was driven either by the constitutive 35S promoter or the native LUP5 promoter. Transformation efficiency, calculated as the percentage of positive transgenic plants to all obtained plants from tissue culture, reached up to 94% across experiments ( Table 1 ). The native LUP5 promoter exhibited approximately two-fold higher transformation efficiency than the 35S promoter in both the G- and P-type, indicating a selection pressure against LUP5 overexpression from the 35S promoter ( Table 1 ). Download figure Open in new tab Fig. S1. Agrobacterium -mediated stable plant transformation process. View this table: View inline View popup Download powerpoint Table 1. A stable transformation system in B. vulgaris enabled up to 94% efficiency. Stable expression of G-type LUP5 in the P-type confers insect resistance To evaluate the impact of G-type LUP5 expression on insect resistance, a choice insect feeding assay was conducted by presenting alternating wildtype and transgenic B. vulgaris leaf discs to diamondback moth larvae. Leaf area consumed was quantified when approximately half of the total leaf material of the wild type had been eaten (approx. 7.5 hr). The best performing LUP5 -transformed P-type plants exhibited a significant reduction in larval consumption, ranging from 76 - 95% ( Fig. 2 , A and B). QPCR on the best performing 35S and LUP5 promoter lines, confirmed that the G-type LUP5 was transformed and expressed in the P-type both under the 35S promoter or the native G-type LUP5 promoter ( Fig. 2C ). Interestingly, the 35S promoter expressed the LUP5 at approximately sixteen-fold higher levels than that in the best pLUP5::LUP5 transformed P-type and the wildtype P- and G-type ( Fig. 2C ). Download figure Open in new tab Fig. 2. Expression of the G-type LUP5 in B. vulgaris P-type reduces feeding by diamondback moth larvae. A) Consumption area (%) after 7.5 h (choice test, three individual plants per transgenic line; each replicate represents one Petri dish containing five transgenic and five wildtype leaf discs). B) Example of the insect choice feeding assay, and the resulting leaf condition after 7.5 hr insect exposure. C) Relative LUP5 expression levels (three individual plants per transgenic line). PWT, wildtype P-type B. vulgaris ; GWT, wildtype G-type B. vulgaris ; p35S::LUP5, p35S::LUP5 transformed P-type B. vulgaris ; pLUP5::LUP5, pLUP5::LUP5 transformed P-type B. vulgaris , LUP5, P-type B. vulgaris expressing the G-type LUP5 . Statistical significance was assessed using Student’s t -test: *(p<0.5), ** (p < 0.05) and *** (p < 0.005). Expression of G-type LUP5 induced β-amyrin-derived saponins in P-type B. vulgaris Based on the highest reduction of leaf area consumed, three transgenic lines were selected for each promoter type (35S and native LUP5 ) and subjected to untargeted and targeted LCMS analyses. The untargeted metabolite analysis revealed approximately 9,500 distinct m/z features, each defined as a unique m/z and retention time (RT) pair. Features differing between LUP5 -expressing and wild-type plants (P < 0.5) were ranked by fold change, and the top 100 increased and 100 decreased features were analyzed further. However, extraction of the m/z features using DataAnalysis software revealed that the majority lacked clear MS² fragmentation patterns, preventing structural annotation. Many of these features likely represent in-source fragments, background ions, or uncharacterized metabolites with inconclusive fragmentation. Consequently, a targeted saponin analysis was conducted using DataAnalysis. B. vulgaris saponins were detected by extracted ion chromatograms of our previously tentatively identified aglycones and their structures were inferred from characteristic fragmentation patterns. Based on 455 m/z, 469 m/z, and 471 m/z, the characteristic fragmentation patterns of oleanolic acid, gypsogenin and hederagenin were tentatively identified, respectively. This targeted approach identified ten putative distinct saponins in P-type lines and thirty in G-type lines ( Supplementary Table S4 and S5 ). All detected saponins were classified as monodesmosidic, based on their fragmentation patterns. Saponin abundance was subsequently quantified by calculating peak areas from extracted ion chromatograms ( Supplementary Table S4 and S5 ). View this table: View inline View popup Download powerpoint Tab. S4. Saponins detected in P-type B. vulgaris . EIC, extract ion chromatogram View this table: View inline View popup Download powerpoint Tab. S5. Saponins detected in G-type B. vulgaris . EIC, extract ion chromatogram The LCMS analysis revealed three hederagenin-derived monodesmosidic saponins (two hexoses; RT 10.7, 11.5, and 12.3 min) and one oleanolic acid monoglucoside (one hexose; RT 9.5 min) that accumulated at significantly higher levels in LUP5 -transformed P-type compared to wild-type P-type ( Fig. 3 , A and B). The most abundant of the three hederagenin-derived saponins eluted at 11.5 min and was increased ∼four-fold in the P-type expressing LUP5 under the native LUP5 promoter compared to the wildtype and was ∼two-fold more abundant than in P-type plants transformed with LUP5 under the 35S promoter ( Fig. 3A ). Hederagenin cellobioside (eluted at 12.3 min) was also increased ∼four-fold but remained at lower levels in pLUP5::LUP5 transformed P-type ( Fig. 3A ). These three hederagenin-derived saponins were also present in the control G-type plants, with hederagenin cellobioside (eluted at 12.3 min) being the most abundant hederagenin-derived saponin ( Supplementary Table S5 and S8 ). The increased oleanolic acid monoglucoside (eluted at 9.5 min) was not detected in G-type ( Supplementary Table S5 ). The other detected saponins remained unchanged ( Supplementary Table S6 ). View this table: View inline View popup Download powerpoint Tab. S6. Non significantly peak area changed saponins in LUP5 transformed P-type B. vulgaris compared to wildtype P-type. EIC, extract ion chromatogram; PWT, wildtype P-type B. vulgaris ; p35S::LUP5, p35S::LUP5 transformed P-type B. vulgaris . Download figure Open in new tab Fig. 3. Expression of the G-type LUP5 increased glycosylated oleanolic acid and hederagenin levels in the transgenic P-type B. vulgaris . A) Monodesmosidic hederagenin (with 2 hexoses) peak area (EIC: m/z = 841.4591 ± 0.2). B) Oleanolic acid monoglucoside peak area (EIC: m/z = 617.4059 ± 0.2). C) Peak area of sapogenins, including total hederagenin (EIC: m/z = 437.3425 ± 0.2) and total oleanolic acid (EIC: m/z = 439.3593± 0.2). PWT, wildtype P-type B. vulgaris ; p35S::LUP5, p35S::LUP5 transformed P-type B. vulgaris ; pLUP5::LUP5, pLUP5::LUP5 transformed P-type B. vulgaris . Error bars represent the standard deviation of the mean from three individual plants. Statistical significance was assessed using Student’s t -test: *(p<0.5), ** (p < 0.05) and *** (p < 0.005). To evaluate the effect of G-type LUP5 expression on P-type B. vulgaris saponin aglycone biosynthesis and accumulation, levels of free and total sapogenins were analyzed. Free sapogenins were measured from untreated samples, representing aglycones that occur naturally in a non-glycosylated form. Total sapogenins were defined as the sum of free and sugar-bound sapogenins, the latter released through hydrochloric acid mediated hydrolysis of glycosylated saponins. A transgenic line (with three vegetative clones) from both p35S:: LUP5 and pLUP5:: LUP5 -transformed P-type B. vulgaris was selected based on reduced leaf area consumption and increased accumulation of four saponins ( Fig. 2A , Fig. 3 , A and B). This analysis revealed a significant increase in total hederagenin and oleanolic acid levels in LUP5 -transformed P-type plants ( Fig. 3C ), while total lupeol, β-amyrin, and free sapogenins remained undetectable ( Supplementary Table S7 ). Higher accumulation of total hederagenin was observed in plants expressing LUP5 under the native promoter compared to those under the 35S promoter ( Fig. 3C ). View this table: View inline View popup Download powerpoint Tab. S7. Identification and relative quantification of total and free sapogenins in B. vulgaris . EIC, extract ion chromatogram To investigate how promoter choice affects gene expression and metabolic composition, we compared the previously described transgenic P-type lines expressing LUP5 under either the constitutive 35S or the native LUP5 promoter. Interestingly, although the 35S promoter drove much higher LUP5 expression levels compared to the native promoter ( Fig. 2C ), this did not result in proportionally increased accumulation of β-amyrin–derived saponins ( Fig. 3A–C ). To investigate this, we made eGFP lines with the LUP5 and 35S promoter respectively. Analysis of these lines the 35S promoter was already active at the shoot induction stage and consistently drove stronger fluorescence signals, whereas the LUP5 promoter only became active only about four weeks after transfer to soil ( Fig. 4A-F ). These observations indicate that the two promoters differ mainly in temporal activation and relative expression strength, which likely contributes to the observed variation in transcript levels, metabolite accumulation, and transformation efficiency. Download figure Open in new tab Fig. 4. Native LUP5 promoter supports higher transformation efficiency via moderate expression in B. vulgaris . During the shoot induction stage, no detectable eGFP fluorescence was observed in the pLUP5::eGFP lines (B). Four weeks after root induction, eGFP fluorescence could be detected in the pLUP5::GFP lines (D, E). However, GFP expression driven by the 35S promoter was still much stronger, as evident by the saturated image F, acquired with the same exposure time. RFP, driven by the 35S promoter, was used for screening transgenic plants by microscopy. A, D: Wild-type plants under white light, eGFP, and RFP filters; B, E: pLUP5::eGFP transformed G-type plants under white light, eGFP, and RFP filters; C, F: p35S::eGFP transformed G-type plants under white light, eGFP, and RFP filters; Scale bars: A–C, 10 mm, D–F, 0.5 mm. Introduction of G-type LUP5 in G-type increased a monodesmosidic hederagenin (with three hexoses) and a oleanolic acid monoglucoside (with two hexoses) in both p35S::LUP5 and pLUP5::LUP5 transformed G-type plants, retention times 8.5, and 11.0 min, respectively ( Fig. 5 , A to C). Monodesmosidic oleanolic acid (with 3 hexose, retention time at 10.7 min) increased in all pLUP5::LUP5 transformed G-type lines ( Fig. 5C ). Twenty-six additional saponins remained unchanged ( Supplementary Table S8 ), and the three slightly elevated saponins remained at low abundance relative to other saponins ( Fig. 5A -C, Supplementary Table S8 ). View this table: View inline View popup Download powerpoint Tab. S8. Non significantly peak area changed saponins in LUP5 transformed G-type B. vulgaris compared to wildtype G-type. EIC, extract ion chromatogram; GWT, wildtype G-type B. vulgaris ; p35S::LUP5, p35S::LUP5 transformed G-type B. vulgaris . Hederagenin cellbioside (EIC: m/z 841.4591± 0.2, retention time 12.3 min) was quantified by after diluting sample 60 times Download figure Open in new tab Fig. 5. Expression of G-type LUP5 increased glycosylated oleanolic acid and hederagenin levels in G-type B. vulgaris . A) Monodesmosidic hederagenin (with 3 hexose) peak area (EIC: m/z = 1003.5115 ± 0.2). B) Monodesmosidic oleanolic acid (with two hexoses) peak area (EIC: m/z = 617.4059 ± 0.2). C) Monodesmosidic oleanolic acid (with 3 hexose) peak area (EIC: m/z = 987.517 ± 0.2). GWT, wildtype G-type B. vulgaris ; p35S::LUP5, p35S::LUP5 transformed G-type B. vulgaris ; pLUP5::LUP5, pLUP5::LUP5 transformed G-type B. vulgaris . Error bars represent the standard deviation of the mean from three individual plants. Statistical significance was assessed using Student’s t -test: *(p<0.5), ** (p < 0.05) and *** (p < 0.005). In conclusion, stable transformation of G-type LUP5 in P-type significantly increased one oleanolic acid monoglucoside (with one hexose) and three hederagenin-derived monodesmosidic saponins (each with two hexoses) by approximately 50% to fourfold, depending on the promoter and compound ( Fig. 3 , A and B). Notably, the oleanolic acid monoglucoside, eluting at 9.5 minutes, was detected exclusively in the transformed P-type ( Fig. 3B , Supplementary Table S5 ). The most abundant hederagenin-derived saponin in P-type eluted at 11.5 min, whereas hederagenin cellobioside (eluted at 12.3 min) remained the dominant saponin in G-type ( Fig. 3A , Supplementary Table S6 and S8 ). Introduction of G-type LUP5 in G-type slightly increased saponin levels, but their concentrations remained low compared to other oleanolic acid- and hederagenin-derived compounds ( Fig. 5 , A to C, Supplementary Table S8 ). Expression of G-type LUP5 by its native promoter resulted in higher transformation efficiency and induced greater accumulation of β-amyrin-derived sapogenins and their glycosylated saponins, despite lower LUP5 transcript levels compared to the 35S promoter ( Fig. 3 , A to C, Fig. 5 , A to C, Table 1 , Supplementary Table S7 ). CYP72A552 silencing decreased to 40% hederagenin derived saponin accumulation in G-type B. vulgaris To evaluate the impact of G-type CYP72A552 silencing on insect resistance, a choice insect feeding assay was conducted as described above. However, neither the transgenic G-type nor the wild-type G-type plants were consumed by diamondback moth larvae as measure after 7.5 hours ( Fig. 6G ). To assess the impact of CYP72A552 downregulation on saponin and sapogenin levels in G-type B. vulgaris , a metabolomic analysis was performed as described above. Silencing of CYP72A552 reduced the accumulation of six saponins in the G-type, including hederagenin monoglucoside and three other hederagenin-derived monodesmosidic saponins (each with two hexoses), monodesmosidic gypsogenin (with two hexoses), and one unknown saponin ( Fig. 6 , A to E). MS/MS analysis indicated that the unknown saponin contains a sapogenin of 457 m/z with one hexose and two methylpentoses ( Fig. 6E ). Twenty-three other saponins, including seven glycosylated oleanolic acids, remained unchanged ( Supplementary Table S9 ). When the sapogenin levels of CYP72A552 -silenced line #13 and wild-type G-type B. vulgaris were compared, the analysis revealed a 40% reduction in total hederagenin, while total and free oleanolic acid levels remained unchanged, consistent with the saponin data ( Fig. 6A–F ).. Lupeol and β-amyrin remained undetectable in G-type B. vulgaris ( Supplementary Table S7 ). In conclusion, silencing CYP72A552 significantly reduced hederagenin biosynthesis and the accumulation of its glycosylated derivatives in G-type B. vulgaris . View this table: View inline View popup Download powerpoint Tab. S9. Non significantly peak area changed saponins in CYP72A552 RNAi G-type B. vulgaris compared to wildtype G-type. EIC, extract ion chromatogram; GWT, wildtype P-type B. vulgaris ; CYP72A552::RNAi, CYP72A552 silenced G-type B. vulgaris . Download figure Open in new tab Fig. 6. Silencing CYP72A552 decreased hederagenin and its derived saponins in G-type B. vulgaris . A) Hederagenin monoglucoside peak area (EIC: m/z = 679.4063 ± 0.2). B) Hederagenin cellobioside peak area (EIC: m/z = 841.4591 ± 0.2). C) Monodesmosidic hederagenin (with two hexoses) peak area (EIC: m/z = 841.4591 ± 0.2). D) Monodesmosidic gypsogenin (with two hexoses) peak area (EIC: m/z = 839.4409 ± 0.2). E) unknown compound-1 peak area (EIC: m/z = 901.2429 ± 0.2). F) Peak area of sapogenins, including total hederagenin (EIC: m/z = 437.3425 ± 0.2), free and oleanolic acid (EIC: m/z = 439.3593± 0.2). G) Example of the insect choice feeding assay, and the resulting leaf condition after 7.5 hr insect exposure. GWT, wildtype P-type B. vulgaris ; CYP72A552::RNAi, CYP72A552 silenced G-type B. vulgaris . Error bars represent the standard deviation of the mean from three individual plants. Statistical significance was assessed using Student’s t -test: *(p<0.5), ** (p < 0.05) and *** (p < 0.005). Discussion Deciphering the genetic and biochemical bases of natural insect resistance is essential for our basic understanding of chemical ecology and for addressing global challenges in crop protection and sustainability. Our study contributes to this effort by elucidating how structural diversification in plant specialized metabolites mediates ecological defense and by identifying regulatory elements that can be harnessed for future resistance breeding. Triterpenoid saponins are key mediators of plant–insect interactions and have attracted attention due to their high chemical diversity and structure-dependent bioactivity and unique modes of action. Key factors influencing insect resistance include the structure of the saponin, the composition of the sapogenin backbone, the types of sugar moieties, the length and number of glycoside chains, and the sites of glycosylation ( Gao et al., 2011 ; Liu et al., 2019 ; Tian et al., 2021 ). Prior metabolic engineering efforts have focused on modifying saponin composition by disrupting biosynthetic genes in legumes ( Confalonieri et al., 2021 ; Hodgins et al., 2024 ). For example, Confalonieri et al. (2021) knocked out CYP93E2 , which hydroxylates β- amyrin at the C-24 position, in Medicago truncatula , resulting in transgenic lines that no longer produced non-hemolytic soyasapogenols saponins but instead redirecting metabolic flux toward the synthesis of hemolytic saponins. Hodgins et al. (2024) knocked out β-amyrin synthase BAS1 in pea ( Pisum sativum ), leading to a 99.8% reduction in saponin content in the seeds. However, these studies did not address how targeted saponin engineering influences ecological traits such as herbivore resistance or consumption. We have previously transiently expressed saponin biosynthetic genes in Nicotiana benthamiana ( Khakimov et al., 2015 ; Liu et al., 2019 ). An unexpected drawback of this transient tobacco system was that hederagenin was further metabolized by endogenous tobacco enzymes, which led to the accumulation of hederagenin-3- O -monoglucoside to levels far below those naturally occurring in Barbarea vulgaris ( Liu et al., 2019 ). When oleanolic acid monoglucoside was transiently produced in N. benthamiana , mainly non-toxic bidesmosidic saponins accumulated, suggesting that the toxic monoglucosides were further glycosylated and thereby detoxified by the plant ( Khakimov et al., 2015 ). In addition, the diamondback moth is a crucifer specialist and will not feed on N. benthamiana . To overcome these major obstacles, we metabolically engineered the saponin biosynthetic pathway directly in the wild crucifer B. vulgaris to test whether specific triterpenoid profiles confer resistance to insect herbivory. We hypothesized that differences in LUP5 expression between G- and P-type B. vulgaris underlie their distinct saponin profiles and insect resistance, based on quantitative trait locus (QTL) analysis and insect feeding assays with saponin-treated leaf disks ( Augustin et al., 2012 ; Khakimov et al., 2015 ; Nielsen et al., 2010 ; Kuzina et al., 2011 ). Our metabolic engineering strategy demonstrates that the susceptibility of the P-type to diamondback moth feeding is primarily due to its lack of LUP5 expression. LUP5 encodes an oxidosqualene cyclase that produces β-amyrin, the precursor of oleanolic acid and hederagenin-derived saponins, both implicated in insect deterrence. Interestingly, our results further suggest that glycoside linkage exerts less influence on feeding behavior than the sapogenin backbone itself. Variation in oxidosqualene cyclase product specificity may be a key driver of saponin-profile differences between the two B. vulgaris ecotypes. Insect-susceptible P-type plants predominantly produce lupeol-derived saponins, reflecting their higher expression of LUP2 , while G-type plants express LUP5 and accumulate β-amyrin-derived saponins, including oleanolic acid and hederagenin cellobioside, both contributing to insect resistance ( Augustin et al., 2012 ; Khakimov et al., 2015 ; Liu et al., 2019 ; Günther et al., 2021 ). By combining gene transfer with ecological assays, our study provides a framework for linking specialized metabolism with ecological function in a non-model plant system. To test this hypothesis in an in planta system, the β-amyrin synthase LUP5 was expressed in both the insect resistant G-type and insect susceptible P-type B. vulgaris . In the otherwise insect susceptible P-type, this resulted in up to a 95% reduction in leaf consumption and increased levels of three hederagenin-derived monodesmosidic saponins (each with two hexoses) and one oleanolic acid monoglucoside ( Fig. 2A , Fig. 3 , A and B). In wildtype B. vulgaris G-type, hederagenin cellobioside is the predominant saponin, while only small amounts of the other two hederagenin-derived saponins were detected at retention times of 10.7 and 11.5 min ( Fig. 5 , A and B, Supplementary Table S8 ). In contrast, in the LUP5 -transformed P-type, the monodesmosidic hederagenin eluting at 11.5 min is more abundant than the other two ( Fig. 3A ). This variation highlights differences in saponin glycosylation patterns between P- and G-types, suggesting that these three hederagenin-derived monodesmosidic saponins, particularly the one detected at 11.5 min, may contribute to insect resistance in the transformed P-type plants. Hederagenin 3- O -glucoside exhibits seven-fold stronger feeding reduction to diamondback moth and tobacco hornworm compared to oleanolic acid 3- O -glucoside ( Liu et al., 2019 ). This enhanced efficacy has been linked to C-23 hydroxylation in hederagenin, which causes the C-3 glucose moiety to adopt a different orientation relative to the sapogenin backbone compared to oleanolic acid ( Cárdenas, Almeida and Bak, 2019 ; Liu et al., 2019 ). Considering the bioactivity and the relatively low increase of oleanolic acid monoglucoside compared to hederagenin-derived monodesmosidic saponins in LUP5 - transformed P-type suggest that there is a CYP72A552 ortholog in the P-type that effectively transforms the oleanolic acid to hederagenin. This hederagenin is then glycosylated by different UGTs than in the G-type B. vulgaris to other linkage types than the 1→4 linkage cellobiosides that predominantly accumulate in the G-type. Based on our findings, LUP5 is identified as the key determinant for insect resistance differences between G- and P-type B. vulgaris . Furthermore, three hederagenin-derived monodesmosidic saponins (each with two hexoses) may play a role in insect resistance in LUP5 -transformed P-type, particularly the most abundant saponin at retention time 11.5 min. These findings offer new mechanistic insight into plant–insect coevolution and establish a model for dissecting genotype–phenotype relationships in plant chemical defense. Future work comparing the insect resistance differences between the elucidated structures of these three hederagenin-derived monodesmosidic saponins could provide further insights into how glycosylation patterns affect saponin bioactivity. Glycoside linkage impacts cytotoxicity of saponins in in vivo assays ( Chwalek et al., 2006 ). However, how glycoside linkage affects insect resistance remains unknown. Unfortunately, due to their low concentration, isolating and purifying these compounds for bioactivity testing remains challenging. Future research should focus on identifying candidate glycosyltransferases that produces a variety of glycoside linkage with the aim to produce defense-related hederagenin-derived monodesmosidic saponins in planta . Interestingly, when LUP5 expression was driven by the constitutive 35S promoter, LUP5 was expressed at approximately sixteen-fold higher levels than in the best pLUP5::LUP5 transgenic line and also exceeded expression levels in both wild-type P- and G-type plants ( Fig. 2C ). However, this high expression did not translate into significantly increased saponin production ( Fig. 2 , A and B, Fig. 5 , A to C). In contrast, using the native LUP5 promoter not only resulted in approximately two-fold higher transformation efficiency ( Table 1 ) but also led to greater accumulation of hederagenin and its glycosides in the P-type background despite the overall much lower expression profile ( Fig. 3 , A and C). These findings indicate that expression strength alone does not determine metabolic output. We propose that high and premature expression of LUP5 under the 35S promoter leads to early toxic β-amyrin accumulation, thereby reducing transformation efficiency ( Fig.2 and Fig. 3 ). In B. vulgaris , LUP5 is expressed most strongly in leaves compared with roots and petioles ( Khakimov et al., 2015 ). To better understand the observed differences in transformation efficiency, saponin and sapogenin accumulation, and LUP5 expression levels between the native and constitutive promoters, we compared promoter activity and eGFP fluorescence intensity during plant regeneration and early growth. Microscopy observations support this interpretation: the 35S promoter was already active during shoot induction, whereas the native LUP5 promoter became detectable only four weeks after transfer to soil following root induction ( Fig. 2 ). Such premature activation likely causes β-amyrin to accumulate before downstream pathway genes are expressed, disturbing metabolic balance through substrate depletion or feedback inhibition. By contrast, delayed activation under the native LUP5 promoter allows coordinated expression with downstream enzymes, promoting efficient accumulation of of saponins that are non-toxic to the plant. Thus, temporal coordination of gene expression emerges as a key determinant of successful metabolic engineering. Although constitutive promoters such as 35S are widely used to achieve strong expression, their untimed activity can impose metabolic burdens and reduce transformation efficiency. Similar effects have been reported in other plant systems: in cotton ( Gossypium hirsutum ), continuous 35S activity throughout development caused unintended metabolic load ( Sunilkumar et al., 2002 ), while in maize ( Zea mays ), the ZmUbi1 promoter achieved higher transformation efficiency and better tissue-culture survival than 35S despite lower transcript levels ( Beringer et al., 2017 ). These findings reinforce that expression timing, rather than transcriptional magnitude, is crucial for maintaining metabolic homeostasis and efficient pathway flux. Taken together, although the 35S promoter can drive high levels of transgene expression, its continuous activity often disrupts metabolic balance, decreases transformation efficiency, and induces pleiotropic effects on plant development. In contrast, the native LUP5 promoter confines expression to appropriate developmental stages, providing precise metabolic control and coordinated biosynthetic regulation. These results highlight that promoter choice governs metabolic outcomes not merely through expression strength but through temporal alignment with downstream biosynthetic processes, thereby ensuring efficient in planta production of defense-related hederagenin-derived monodesmosidic saponins. We had anticipated that silencing of CYP72A552 in G-type would reduce insect consumption as hederagenin derived saponins are more bioactive than those derived from oleanolic acid ( Liu et al., 2019 ) . However, our results show that while silencing of CYP72A552 decreased hederagenin biosynthesis by up to 40%, leading to a significant reduction in hederagenin- and gypsogenin (which is hederagening that has been further oxidized at C-23 to the ketone)-derived monodesmosidic saponins. Specifically, the bioactive hederagenin monoglucoside, hederagenin cellobioside and monodesmosidic gypsogenin (with two hexoses) were decreased ( Fig. 6 , A to D). The decrease in monodesmosidic gypsogenin and unknown compound 1 may have been caused by the reduced hederagenin biosynthesis or off-target effects ( Fig. 6 , D and E, Supplementary Fig. S2 ), possibly resulting from the high nucleotide sequence identity of eight tandemly repeated CYP72A homologs ( Liu et al., 2019 ). Previous studies indicate that cellobiosides of oleanolic acid, hederagenin, and gypsogenin both correlate with insect resistance in B. vulgaris in a concentration-dependent manner ( Shinoda et al., 2002 ; Agerbirk et al., 2003 ; Kuzina et al., 2009 ; Liu et al., 2019 ). The hederagenin cellobioside concentration in G-type B. vulgaris leaves is ∼eleven-fold higher than the ED 50 (50% effective dose) for P. xylostella ( Shinoda et al., 2002 ; Liu et al., 2019 ). Thus, even with a 40% reduction, the remaining saponin levels still exceed the toxicity threshold required to deter feeding. Likewise, LUP5 -transformed P-type plants did not reach the full resistance potential of G-type plants, likely due to incomplete accumulation of the most potent hederagenin saponins ( Fig. 2 , A and B, Fig. 3 , A to C, Fig. 6 , A to G). Overexpression of LUP5 in G-type plants produced only marginal increases in minor oleanolic acid and hederagenin-derived saponins, suggesting that G-type plants may already accumulate saponin at or close the acceptable threshold for accumulation ( Fig. 6 , A to C, Supplementary Table S8 ). Download figure Open in new tab Fig. S2. Nucleotide sequence alignment between the RNAi fragment and eight tandemly repeated CYP72A genes in B. vulgaris . Accession numbers of the eight CYP72A genes range from MH252567 to MH252574 . Conclusion In summary, this study demonstrates that targeted metabolic engineering of triterpenoid saponins in the wild crucifer Barbarea vulgaris is a robust approach to dissect the ecological roles of structurally diverse saponins and study insect resistance mediated by specialized metabolism. Expression of the β-amyrin synthase LUP5 under its native promoter was more effective than constitutive expression, leading to higher transformation efficiency and increased accumulation of insect-deterring hederagenin-derived monodesmosidic saponins. This underscores the importance of promoter choice and precise transcriptional regulation in optimizing specialized metabolite production. Our findings establish stable metabolic engineering of saponins not only as a tool for biotechnological enhancement of resistance traits, but also as a powerful approach to understanding the role of plant defense compounds in chemical ecology. In contrast to transient expression systems, which often result in unstable metabolite accumulation and further modification by host enzymes, stable engineering enables consistent expression of pathway genes and reliable assessment of metabolite function. We identified LUP5 as a crucial determinant of insect resistance in Barbarea vulgaris and demonstrated that three specific hederagenin-derived monodesmosidic saponins, each containing two hexose units, are associated with this defense. Taken together, this work introduces a novel application of metabolic engineering to unravel the ecological functions of plant metabolites and provides promising strategies for crop protection through fine-tuned manipulation of biosynthetic pathways. Conflicts of interest Authors declare that there are no conflicts of interest. Author contributions JS, PC, and SB designed the research; JS, PC, JG, SKN and SB analysed the data and wrote the paper; JS, PC and SKN performed the research; JG provided the data analysis strategy and standards. Acknowledgements We thank Thure Hauser for providing Plutella xylostella , Niels Agerbirk for providing wildtype Barbarea vulgaris G- and P-type seeds, Barry Cohen and Hanne Volpin for their help in setting up the transformation system, and Jack Olsen and Mariela Alejandra González Ramírez for their assistance in LCMS. This work was financially supported by Novo Nordisk Foundation (EcoSap NNF20OC0060298 to SB), the China Scholarship Council (CSC, 202108330041 to JS) and Marie Sklodowska-Curie Individual Fellowship (MSCA-IF 752437 to PC). Funder Information Declared Novo Nordisk Foundation , NNF20OC0060298 China Scholarship Council , 202108330041 European Union’s Horizon 2020 research and innovation programme , 752437 Footnotes The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors ( https://academic.oup.com/plphys/pages/General-Instructions ) is Søren Bak. This version of the manuscript includes new experimental data and substantial textual revisions. Specifically, we have added new results on promoter activity to address the previously unresolved question of how promoter choice and transcriptional regulation influence specialized metabolite composition in Barbarea vulgaris. References 1. ↵ Agerbirk N , Hansen CC , Olsen CE , Kiefer C , Hauser TP , Christensen S , Jensen KR , Orgaard M , Pattison DI , Lange CBA , Cipollini D , Koch MA ( 2021 ) Glucosinolate profiles and phylogeny in Barbarea compared to other tribe Cardamineae (Brassicaceae) and Reseda (Resedaceae), based on a library of ion trap HPLC-MS/MS data of reference desulfoglucosinolates . Phytochemistry 185 : 112658 OpenUrl PubMed 2. ↵ Agerbirk N , Olsen CE , Bibby BM , Frandsen HO , Brown LD , Nielsen JK , Renwick JAA ( 2003 ) A saponin correlated with variable resistance of Barbarea vulgaris to the diamondback moth Plutella xylostella . Journal of Chemical Ecology 29 : 1417 – 1433 OpenUrl CrossRef PubMed Web of Science 3. ↵ Agerbirk N , Olsen CE , Nielsen JK ( 2001 ) Seasonal variation in leaf glucosinolates and insect resistance in two types of Barbarea vulgaris ssp. arcuata . Phytochemistry 58 : 91 – 100 OpenUrl PubMed 4. ↵ Augustin JM , Drok S , Shinoda T , Sanmiya K , Nielsen JK , Khakimov B , Olsen CE , Hansen EH , Kuzina V , Ekstrøm CT , Hauser T , Bak S ( 2012 ) UDP-glycosyltransferases from the UGT73C subfamily in Barbarea vulgaris catalyze sapogenin 3- O -glucosylation in saponin-mediated insect resistance . Plant Physiology 160 : 1881 – 1895 OpenUrl Abstract / FREE Full Text 5. ↵ Augustin JM , Kuzina V , Andersen SB , Bak S ( 2011 ) Molecular activities, biosynthesis and evolution of triterpenoid saponins . Phytochemistry 72 : 435 – 457 OpenUrl CrossRef PubMed Web of Science 6. ↵ Beran F , Petschenka G ( 2022 ) Sequestration of plant defense compounds by insects: from mechanisms to insect–plant coevolution . Annual Review of Entomology 67 : 163 – 180 OpenUrl CrossRef PubMed 7. ↵ Beringer J , Chen W , Garton R , Sardesai N , Wang PH , Zhou N , Gupta MJ , Wu HX ( 2017 ) Comparison of the impact of viral and plant-derived promoters regulating selectable marker gene on maize transformation and transgene expression . Plant Cell Reports 36 : 519 – 528 OpenUrl PubMed 8. ↵ Byrne SL , Erthmann PO , Agerbirk N , Bak S , Hauser TP , Nagy I , Paina C , Asp T ( 2017 ) The genome sequence of Barbarea vulgaris facilitates the study of ecological biochemistry . Scientific Reports 7 : 40728 OpenUrl PubMed 9. ↵ Cárdenas PD , Almeida A , Bak S ( 2019 ) Evolution of structural diversity of triterpenoids . Frontiers in Plant Science 10 : 1523 OpenUrl PubMed 10. ↵ Chwalek M , Lalun N , Bobichon H , Plé K , Voutquenne-Nazabadioko L ( 2006 ) Structure-activity relationships of some hederagenin diglycosides:: Haemolysis, cytotoxicity and apoptosis induction . Biochimica Et Biophysica Acta-General Subjects 1760 : 1418 – 1427 OpenUrl 11. ↵ Confalonieri M , Carelli M , Gianoglio S , Moglia A , Biazzi E , Tava A ( 2021 ) CRISPR/Cas9-mediated targeted mutagenesis of CYP93E2 modulates the triterpene saponin biosynthesis in Medicago truncatula . Frontiers in Plant Science 12 : 690231 OpenUrl PubMed 12. ↵ Cui CJ , Yang YQ , Zhao TY , Zou KK , Peng CY , Cai HM , Wan XC , Hou RY ( 2019 ) Insecticidal activity and insecticidal mechanism of total saponins from Camellia oleifera . Molecules 24 : 4518 OpenUrl CrossRef PubMed 13. ↵ Dervishi M , Schmitt FM , Günther J , Cedergreen N , Bak S ( 2025 ) Structure–activity relationship of triterpenoid saponins based on toxicity towards organisms of different phylogenetic origin . Journal of Chemical Ecology . In press. 14. ↵ Dolma SK , Suresh PS , Singh PP , Sharma U , Reddy SGE ( 2021 ) Insecticidal activity of the extract, fractions, and pure steroidal saponins of Trillium govanianum Wall. ex D. Don for the control of diamondback moth ( Plutella xylostella L.) and aphid ( Aphis craccivora Koch) . Pest Management Science 77 : 956 – 962 OpenUrl PubMed 15. ↵ Erthmann PO , Agerbirk N , Bak S ( 2018 ) A tandem array of UDP-glycosyltransferases from the UGT73C subfamily glycosylate sapogenins, forming a spectrum of mono- and bisdesmosidic saponins . Plant Molecular Biology 97 : 37 – 55 OpenUrl CrossRef PubMed 16. ↵ Fürstenberg-Hägg J , Zagrobelny M , Bak S ( 2013 ) Plant defense against insect herbivores . International Journal of Molecular Sciences 14 : 10242 – 10297 OpenUrl PubMed 17. ↵ Gao GC , Lu ZX , Tao SH , Zhang S , Wang FZ ( 2011 ) Triterpenoid saponins with antifeedant activities from stem bark of Catunaregam spinosa (Rubiaceae) against Plutella xylostella (Plutellidae) . Carbohydrate Research 346 : 2200 – 2205 OpenUrl PubMed 18. ↵ Günther J , Erthmann PØ , Khakimov B , Bak S ( 2021 ) Reciprocal mutations of two multifunctional β-amyrin synthases from Barbarea vulgaris shift α/β-amyrin ratios . Plant Physiology 188 : 1483 – 1495 OpenUrl 19. ↵ Hauser TP , Toneatto F , Nielsen JK ( 2012 ) Genetic and geographic structure of an insect resistant and a susceptible type of Barbarea vulgaris in western Europe . Evolutionary Ecology 26 : 611 – 624 OpenUrl 20. ↵ Hayta S , Smedley MA , Clarke M , Forner M , Harwood WA ( 2021 ) An efficient Agrobacterium -mediated transformation protocol for hexaploid and tetraploid wheat . Current Protocols 1 : e58 OpenUrl 21. ↵ Hodgins CL , Salama EM , Kumar R , Zhao Y , Roth SA , Cheung IZ , Chen JY , Arganosa GC , Warkentin TD , Bhowmik P , Ham BK , Ro DK ( 2024 ) Creating saponin-free yellow pea seeds by CRISPR/Cas9-enabled mutagenesis on β-amyrin synthase . Plant Direct 8 : e563 OpenUrl 22. ↵ Huang FQ , Dong XS , Yin XJ , Fan Y , Fan YM , Mao CC , Zhou W ( 2020 ) A mass spectrometry database for identification of saponins in plants . Journal of Chromatography A 1625 : 461296 OpenUrl PubMed 23. ↵ Hussain M , Debnath B , Qasim M , Bamisile BS , Islam W , Hameed MS , Wang L , Qiu D ( 2019 ) Role of saponins in plant defense against specialist herbivores . Molecules 24 : 2067 OpenUrl PubMed 24. ↵ Karimi M , Inzé D , Depicker A ( 2002 ) GATEWAY™ vectors for Agrobacterium -mediated plant transformation . Trends in Plant Science 7 : 193 – 195 OpenUrl CrossRef PubMed Web of Science 25. ↵ Khakimov B , Amigo JM , Bak S , Engelsen SB ( 2012 ) Plant metabolomics: resolution and quantification of elusive peaks in liquid chromatography–mass spectrometry profiles of complex plant extracts using multi-way decomposition methods . Journal of Chromatography A 1266 : 84 – 94 OpenUrl CrossRef PubMed Web of Science 26. ↵ Khakimov B , Kuzina V , Erthmann PØ , Fukushima EO , Augustin JM , Olsen CE , Scholtalbers J , Volpin H , Andersen SB , Hauser TP , Muranaka T , Bak S ( 2015 ) Identification and genome organization of saponin pathway genes from a wild crucifer, and their use for transient production of saponins in Nicotiana benthamiana . The Plant Journal 84 : 478 – 490 OpenUrl CrossRef PubMed 27. ↵ Khakimov B , Tseng LH , Godejohann M , Bak S , Engelsen SB ( 2016 ) Screening for triterpenoid saponins in plants using hyphenated analytical platforms . Molecules 21 : 1614 OpenUrl CrossRef PubMed 28. ↵ Kuzina V , Ekstrøm CT , Andersen SB , Nielsen JK , Olsen CE , Bak S ( 2009 ) Identification of defense compounds in Barbarea vulgaris against the herbivore Phyllotreta nemorum by an ecometabolomic approach . Plant Physiology 151 : 1977 – 1990 OpenUrl Abstract / FREE Full Text 29. ↵ Kuzina V , Nielsen JK , Augustin JM , Torp AM , Bak S , Andersen SB ( 2011 ) Barbarea vulgaris linkage map and quantitative trait loci for saponins, glucosinolates, hairiness and resistance to the herbivore Phyllotreta nemorum . Phytochemistry 72 : 188 – 198 OpenUrl CrossRef PubMed 30. ↵ Liu Q , Khakimov B , Cárdenas PD , Cozzi F , Olsen CE , Jensen KR , Hauser TP , Bak S ( 2019 ) The cytochrome P450 CYP72A552 is key to production of hederagenin-based saponins that mediate plant defense against herbivores . New Phytologist 222 : 1599 – 1609 OpenUrl CrossRef PubMed 31. Liu TongJin LT , Zhang XiaoHui ZX , Yang HaoHui YH , Agerbirk N , Qiu Yang QY , Wang HaiPing WH , Shen Di SD , Song JiangPing SJ , Li XiXiang LX ( 2016 ) Aromatic glucosinolate biosynthesis pathway in Barbarea vulgaris and its response to Plutella xylostella infestation . Frontiers in Plant Science 7 : 83 OpenUrl PubMed 32. ↵ Moses T , Papadopoulou KK , Osbourn A ( 2014 ) Metabolic and functional diversity of saponins, biosynthetic intermediates and semi-synthetic derivatives . Critical Reviews in Biochemistry and Molecular Biology 49 : 439 – 462 OpenUrl CrossRef PubMed 33. ↵ Nielsen JK , Nagao T , Okabe H , Shinoda T ( 2010 ) Resistance in the plant, Barbarea vulgaris , and counter-adaptations in flea beetles mediated by saponins . Journal of Chemical Ecology 36 : 277 – 285 OpenUrl CrossRef PubMed 34. ↵ Patel VS , Chhalotiya UK , Patel SB , Nuruddin J ( 2020 ) Simultaneous quantification of betulinic acid, lupeol, and β-sitosterol in Madhuca longifolia methanolic extract of bark by liquid chromatography-tandem mass spectrometric method . Journal of AOAC International 104 : 498 – 505 OpenUrl 35. ↵ Shinoda T , Nagao T , Nakayama M , Serizawa H , Koshioka M , Okabe H , Kawai A ( 2002 ) Identification of a triterpenoid saponin from a crucifer, Barbarea vulgaris , as a feeding deterrent to the diamondback moth, Plutella xylostella . Journal of Chemical Ecology 28 : 587 – 599 OpenUrl CrossRef PubMed Web of Science 36. ↵ Stevenson PC , Nyirenda SP , Veitch NC ( 2010 ) Highly glycosylated flavonoids from the pods of Bobgunnia madagascariensis . Tetrahedron Letters 51 : 4727 – 4730 OpenUrl 37. ↵ Sunilkumar G , Mohr L , Lopata-Finch E , Emani C , Rathore KS ( 2002 ) Developmental and tissue-specific expression of CaMV 35S promoter in cotton as revealed by GFP . Plant Molecular Biology 50 : 463 – 479 OpenUrl CrossRef PubMed Web of Science 38. ↵ Tautenhahn R , Patti GJ , Rinehart D , Siuzdak G ( 2012 ) XCMS online: a web-based platform to process untargeted metabolomic data . Analytical Chemistry 84 : 5035 – 5039 OpenUrl CrossRef PubMed 39. ↵ Tian X , Li Y , Hao N , Su X , Du J , Hu J , Tian X ( 2021 ) The antifeedant, insecticidal and insect growth inhibitory activities of triterpenoid saponins from Clematis aethusifolia Turcz against Plutella xylostella (L .). Pest Management Science 77 : 455 – 463 OpenUrl CrossRef PubMed 40. ↵ Xu R , Fazio GC , Matsuda SPT ( 2004 ) On the origins of triterpenoid skeletal diversity . Phytochemistry 65 : 261 – 291 OpenUrl CrossRef PubMed Web of Science 41. ↵ Yactayo-Chang JP , Tang HV , Mendoza J , Christensen SA , Block AK ( 2020 ) Plant defense chemicals against insect pests . Agronomy 10 : 1156 OpenUrl View the discussion thread. Back to top Previous Next Posted October 17, 2025. Download PDF Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Metabolic Engineering Reveals LUP5 as a Determinant of Saponin Composition and Insect Resistance in Barbarea vulgaris Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Metabolic Engineering Reveals LUP5 as a Determinant of Saponin Composition and Insect Resistance in Barbarea vulgaris Jincheng Shen , Jan Günther , Sebastian Kjeldgaard-Nintemann , Pablo D. Cárdenas , Søren Bak bioRxiv 2025.08.04.665341; doi: https://doi.org/10.1101/2025.08.04.665341 Share This Article: Copy Citation Tools Metabolic Engineering Reveals LUP5 as a Determinant of Saponin Composition and Insect Resistance in Barbarea vulgaris Jincheng Shen , Jan Günther , Sebastian Kjeldgaard-Nintemann , Pablo D. Cárdenas , Søren Bak bioRxiv 2025.08.04.665341; doi: https://doi.org/10.1101/2025.08.04.665341 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Plant Biology Subject Areas All Articles Animal Behavior and Cognition (7629) Biochemistry (17660) Bioengineering (13881) Bioinformatics (41909) Biophysics (21435) Cancer Biology (18576) Cell Biology (25479) Clinical Trials (138) Developmental Biology (13366) Ecology (19887) Epidemiology (2067) Evolutionary Biology (24301) Genetics (15598) Genomics (22482) Immunology (17726) Microbiology (40359) Molecular Biology (17162) Neuroscience (88529) Paleontology (666) Pathology (2830) Pharmacology and Toxicology (4820) Physiology (7636) Plant Biology (15125) Scientific Communication and Education (2044) Synthetic Biology (4290) Systems Biology (9817) Zoology (2269)