Metabolic engineering of terpene metabolism in lavender

Abstract

Background

Several members of the Lamiaceae family of plants produce large amounts of essential oil [EO] that find extensive applications in the food, cosmetics, personal hygiene, and alternative medicine industries. There is interest in enhancing EO metabolism in these plants.

Main body

Lavender produces a valuable EO that is highly enriched in monoterpenes, the C₁₀ class of the isoprenoids or terpenoids. In recent years, substantial effort has been made by researchers to study terpene metabolism and enhance lavender EO through plant biotechnology. This paper reviews recent advances related to the cloning of lavender monoterpene biosynthetic genes and metabolic engineering attempts aimed at improving the production of lavender monoterpenes in plants and microbes.

Conclusion

Metabolic engineering has led to the improvement of EO quality and yield in several plants, including lavender. Furthermore, several biologically active EO constituents have been produced in microorganisms.

1 Background

Lavender (genus Lavandula) belongs to the family Lamiaceae and is known for producing large amounts of essential oil [EO]. The genus Lavandula encompasses over 30 known species, each characterized by a unique EO profile [1, 2]. The presence of specific monoterpenes in lavender EO contributes to its value, making it a significant resource in the production of perfumes, medicinal products, food flavorings, and antiseptics [3, 4]. The most abundant monoterpenes in lavender EO are linalool, linalyl acetate, borneol, camphor, and 1,8-cineole [5, 6]. Other notable monoterpenes include limonene, lavandulol, lavandulyl acetate, and α-terpineol [4, 7, 8]. In addition, lavender produces certain monoterpenes in response to environmental conditions [8, 9].

Among all lavender species, Lavandula angustifolia, Lavandula latifolia, and their natural hybrid, Lavandula x intermedia, hold significant commercial importance [10]. L. angustifolia is highly valued in the perfumery industry due to its high levels of linalool and linalyl acetate. EO obtained from L. latifolia is important in the medical sector because of its high concentrations of linalool, camphor, and 1,8-cineole. L. x intermedia EO contains constituents found in both parents, and is widely used in personal care and hygiene products, [3].

This review highlights recent efforts aimed at biotechnological improvement of lavender EO in engineered plants. It also provides information on the production of lavender EO constituents in microorganisms.

2 Main text

2.1 Methodology

Published articles were sourced from various search engines and databases, including Web of Science, Google Scholar, ScienceDirect, Scopus, and SciFinder. Keywords used in the search included lavender, along with essential oil, glandular trichomes, gene cloning, transformation, metabolism, engineering, terpene synthases, terpene biosynthesis, and genome editing. Additionally, separate searches were conducted for yeast, bacteria, and cyanobacteria in conjunction with key monoterpenes from lavender EO such as linalool, 1,8-cineole, and borneol.

2.2 Terpenes

The terpenoids (terpenes), also known as isoprenoids, represent the most abundant class of plant secondary metabolites. They serve crucial roles in plant growth, development, overall metabolism, and defense against predators, diseases, and competition [11]. Many terpenes have applications in cosmetics, pharmaceuticals, insecticides, and potential biofuels. In plants, the biosynthesis of terpenes is divided into three stages. The first stage involves the production of the common precursors, isopentenyl diphosphate [IPP] and its isomer dimethylallyl diphosphate [DMAPP] via the 2-C-methyl-D-erythritol 4-phosphate [MEP] pathway, also known as the 1-deoxy-D-xylulose-5-phosphate [DXP] pathway, and/or mevalonate [MVA] pathway. During the second stage, intermediates such as geranyl diphosphate [GPP], farnesyl diphosphate [FPP], and geranylgeranyl diphosphate [GGPP] are synthesized from IPP/DMAPP by isoprenyl diphosphate synthases [IDSs], also known as prenyltransferases. The final stage involves the production of various terpenes, catalyzed by terpene synthases [TPSs] such as linalool synthase [LINS] and 1,8-cineole synthase [CINS], along with terpene-modifying enzymes. It has been shown that the enzymes involved in terpene production have distinct subcellular localizations: all MEP pathway enzymes are located in plastids, while MVA pathway enzymes are found in the cytosol or peroxisomes [13, 14]. IDSs and TPSs exhibit more diverse localizations and are frequently associated with the subcellular site of terpene biosynthesis [12].

2.3 Glandular trichomes [GTs]

GTs are specialized plant structures dedicated to the synthesis and accumulation of EO. Lavender possesses two types of GTs: peltate GTs and capitate GTs. Peltate GTs produce and store a significant amount of monoterpene-rich EO in lavender [6, 15]. Each peltate GT comprises a basal cell, anchoring it to the epidermis, a stalk cell, up to eight secretory cells, and a storage cavity.[16]. A plasma membrane separates the storage cavity from the secretory cells [17, 18]. Within the secretory cells, EO synthesis occurs in two compartments: the cytosol and the leucoplast, where the MVA and MEP pathways operate, respectively [17, 19, 20]. EO constituents synthesized in the plastids via the MEP pathway are transported to the cytosol and subsequently secreted into the storage cavity, either directly or after processing [18].

Trichome formation, which is widely studied in Arabidopsis, is regulated in part by transcription factors [TFs] that can have either positive or negative effects. Additionally, other TFs function upstream or downstream of these regulators [21]. Recently, Zhang et al. [22] studied GT formation in lavender using a genomics approach, identifying several TFs belonging to R2R3-MYB subfamily associated with GT development.

2.4 Monoterpene biosynthesis

Monoterpenes are synthesized within the plastids of photosynthetic organisms, utilizing primary metabolites [12, 23]. In plants, chloroplasts serve as the major sites for monoterpene biosynthesis via the MEP pathway (Fig. 1) [24]. The pathway initiates with the condensation and decarboxylation of pyruvate [PYR] and glyceraldehyde 3-phosphate [G3P] by DXP synthase [DXS], yielding DXP. Subsequently, DXP reductoisomerase [DXR] catalyzes the reduction of DXP to MEP via NADPH-dependent isomerization [25]. MEP undergoes a series of enzymatic reactions involving MEP cytidyltransferase [MCT], 4-(cytidine 5ˈ-diphospho)- 2-C-methyl-D-erythritol-kinase [CMK], and 2-C-methyl-D-erythritol-2,4-cyclodiphosphate [MEcPP] synthase [MDS], leading to the formation of MEcPP [26, 27]. MEcPP is further reduced to 4-hydroxy-3-methylbut-2-enyl diphosphate [HMBPP] by HMBPP synthase [HDS]. IPP and DMAPP, the final products of MEP pathway, are generated in a ratio of approximately 5:1through the reduction of HMBPP by HMBPP reductase [HDR] [28,29,30]. Prenyltransferases, including geranyl diphosphate synthase [GPPS] and neryl diphosphate synthase [NPPS], catalyze the condensation of IPP and DMAPP to synthesize GPP and neryl diphosphate [NPP], which are precursors of monoterpenes [31]. In addition to monoterpenes, IPP/DMAPP derived from the MEP pathway is used to produce numerous other terpenes including photosynthetic pigments such as phytol and carotenoid precursors [32, 33].

Fig. 1

Biosynthesis of isopentenyl diphosphate [IPP] and dimethylallyl diphosphate [DMAPP] through 2-C-methyl-D-erythritol 4-phosphate [MEP] pathway. PYR [pyruvate], G3P [glyceraldehyde 3–phosphate], DXP [1-deoxy-D-xylulose-5-phosphate], CPP-ME [cytidyl diphosphate-methyl-D-erythritol], CPP-MEP [cytidyl diphosphate-methyl-D-erythritol 4-phosphate], MEcPP [2-C-methyl-D-erythritol -2,4-cyclodiphosphate], HMBPP [1-hydroxy-2-methyl-2-(E)-butenyl-4-diphosphate], GPP [geranyl diphosphate], NPP [neryl diphosphate], and GGPP [geranylgeranyl diphosphate]

2.5 Monoterpene metabolism in lavender

In line with other plant species, TPSs are responsible for catalyzing the conversion of the GPP substrate into cyclic and acyclic monoterpenes within the secretory cells of lavender GTs [34, 35]. Numerous researchers have undertaken efforts to clone TPS genes and characterize their functionality in vitro, aiming to elucidate their roles in EO production (Table 1). For instance, Landmann et al. [1] employed a homology-based PCR approach to clone two monoterpene synthases [monoTPSs]: limonene synthase [LaLIMS] and LINS [LaLINS], and one sesquiterpene synthase [sesquiTPSs]: bergamotene synthase [LaBERS] from L. angustifolia leaves and flowers. Subsequently, the cloned cDNAs were expressed in Escherichia coli. LaLIMS produced limonene, terpinolene, camphene, α-pinene, β-myrcene, and traces of β-phellandrene, while LaLINS solely produced (R)-(−)-linalool, and LaBERS converted FPP into bergamotene. In addition, Demissie et al. [36] reported the cloning and functional characterization of β-phellandrene synthase [LaβPHLS] in E. coli. Their results showed that the recombinant LabPHLS converted GPP and NPP into β-phellandrene. In 2012, Demissie et al. [4] identified the CINS [LiCINS] gene from L. x intermedia and cloned it into E. coli. The resulting bacterially generated recombinant protein, approximately 63 kDa in size, converted GPP mainly into 1,8-cineole. As reported again by Demissie et al. [37], L. x intermedia lavandulyl diphosphate synthase [LiLPPS] was cloned using a homology-based cloning method and expressed in E. coli. The resulting protein, with a molecular weight of about 34.5 kDa, catalyzed the fusion of two DMAPP units to form lavandulyl diphosphate [LPP] in vitro. The works of Demissie et al. [4, 36, 37] were continued by the research of Sarker et al. [2, 7, 38], who cloned and expressed borneol dehydrogenase [LiBDH], caryophyllene synthase [LiCPS], and two alcohol acetyltransferases [LiAAT-3 and LiAAT-4] genes in E. coli. The recombinant LiBDH protein converted borneol into camphor, the recombinant LiCPS protein converted FPP into 9-epicaryophyllene, and both recombinant LiAAT-3 and LiAAT-4 proteins converted lavandulol to lavandulyl acetate. In 2015, Benabdelkader et al. [9] cloned two monoTPSs and one sesquiTPS from L. pedunculata and functionally identified them as fenchol synthase [LpFENS], α-pinene synthase [LpPINS], and germacrene A synthase [LpGEAS]. Interestingly, while the expression patterns of FENS and PINS genes aligned with the enzyme product accumulation profile, this correlation was not observed for GEAS. Next, Adal et al. [8] identified and characterized a monoTPS gene, 3-carene synthase [Li3CARS], from L. x intermedia. The results showed that the recombinant Li3CARS transformed GPP into 3-carene as the major product (Fig. 2). In 2019, the same researcher reported the cloning of S-LINS [LiS-LINS] gene from L. x intermedia in bacteria. The cloned LiS-LINS catalyzed the conversion of GPP into S-linalool as the sole product [39]. In 2023, Adal et al. [40] cloned lavender (+)- bornyl diphosphate synthase [LiBPPS] in bacteria, and the recombinant LiBPPS promoted the conversion of GPP to (+)- bornyl diphosphate [BPP] as the main product, accompanied by the formation of several minor monoterpene compounds. In a recent study, Ling et al. [41] identified terpene synthase 7 [LaTPS7] and terpene synthase 8 [LaTPS8] genes from L. angustifolia during the budding phases. Subsequently, they cloned these TPSs into E. coli and Nicotiana benthamiana. The recombinant LaTPS7 generated nine different products in vitro, including camphene, myrcene, and limonene, while LaTPS8 produced eight volatiles using GPP and NPP as substrate. Moreover, the overexpression of LaTPS7 in N. benthamiana resulted in the synthesis of limonene, whereas LaTPS8 yielded α-pinene and sylvestrene.

Table 1 Cloning and metabolic engineering of lavender genes

Fig. 2

Metabolism of acyclic a and cyclic b monoterpenes in lavender. LIMS [limonene synthase], (−)-α-TS [(−)-α-terpineol synthase], 1,8 CS [1,8-cineole synthase], BPPS [bornyl diphosphate synthase], BDH [borneol dehydrogenase], LINS [linalool synthase], AT [acetyltransferase], and LPPS [lavandulyl diphosphate synthase]

2.6 Cloning other lavender genes

As shown in Table 1, other aspects of lavender have also been studied. For instance, Guitton et al. [42] conducted a study on the concentration of volatile organic compound [VOC] in L. angustifolia throughout inflorescence growth. They found that calyces were the primary sources of VOC accumulation, with three major VOC groups dominating the global fragrance bouquet of inflorescences. The transition of VOCs occurred between the opening of the inflorescence's first flower and the beginning of seed set. There is a need to develop more knowledge on the molecular features of bloom initiation and development in lavender. Wells et al. [43] studied the short vegetative phase [LaSVP] gene of L. angustifolia and transformed it into A. thaliana. Their results showed that expression of LaSVP in A. thaliana delayed flowering, resulted in the production of sepals instead of petals, and prevented the formation of seed pods. Additionally, Adal et al. [44] used RNA-Seq and transcript profiling to identify several TFs potentially regulating floral development in lavender. Their study focused on the roles of two TFs, LaAGAMOUS-like [LaAG-like] and LaSEPALLATA3-like [LaSEP3-like], in flower development. LaAG-like and LaSEP3-like cDNAs were overexpressed in Arabidopsis plants. The results revealed that all transgenic plants exhibited earlier flowering compared to wild-type controls. Furthermore, mildly overexpressed plants grew normally, but those that excessively expressed the transgene had curling leaves. Another researcher, Dong et al. [45] focused on a bHLH TF, LaMYC4, an important regulator for plant terpenoid biosynthesis. This gene was isolated from L. angustifolia following methyl jasmonate [MeJA] treatment and overexpressed in Arabidopsis and tobacco. Results showed that overexpression of LaMYC4 enhanced sesquiterpenoids, including caryophyllene.

2.7 Metabolic engineering to produce lavender monoterpenes

2.7.1 Engineering lavender plants

The initial lavender metabolic engineering endeavor to improve the plastidial MEP pathway for the synthesis of the precursors IPP and DMAPP, was conducted by Munoz-Bertomeu et al. [46]. In their study, Munoz-Bertomeu et al. up-regulated DXS, an enzyme catalyzing the initial step in the MEP pathway. They expressed a cDNA encoding the A. thaliana DXS into spike lavender. Gas chromatography/mass spectrometry [GC–MS] analyses indicated that transgenic plants produced significantly higher EOs compared to control plants, with the accumulated EOs maintained in the T₁ generation. In another effort, Tsuro and Ikedo [47] infected calli derived from lavandin (L. × intermedia) leaves with Agrobacterium rhizogenes, and regenerated plants. However, the regenerated plantlets displayed dwarfism due to short internodes, and low EO content. Moreover, Adal et al. [40] expressed the (+)-LiBPPS gene in both sense and antisense orientation. They observed that when (+)-LiBPPS was orientated in the antisense orientation, there was a reduction in the synthesis of (+)-borneol and camphor, while plant growth and development remained unaffected. Conversely, plants with the sense-orientated (+)-LiBPPS produced higher levels of borneol and camphor, but their growth and development were adversely affected.

2.7.2 Engineering microbes to produce lavender EO constituents

Researchers are also considering engineering microorganisms (Table 2) for the (eventual) large-scale production of lavender EO monoterpene constituents. The following are prominent examples of such studies:

Table 2 Microbial production of major monoterpenes present in lavender

2.7.2.1 Production of lavender monoterpenes in yeast

Deng et al. [48] produced linalool in Saccharomyces cerevisiae by introducing a fusion protein composed of LINS from Actinidia arguta and farnesyl diphosphate synthase [ERG20] from S. cerevisiae. The fusion protein, connected by a proper polypeptide linker between enzymes, exhibited a 69.7% increase in efficiency in linalool production compared to the application of individual free enzymes. Furthermore, Zhang et al. [49] demonstrated an enhancement in linalool production in S. cerevisiae through a series of experiments: They initially integrated the Isopentenol Utilization Pathway [IUP] into S. cerevisiae by incorporating choline kinase and isopentenyl phosphate kinase. Subsequently, LINS from Actinidia arguta was truncated from the N-terminal and introduced into S. cerevisiae. A double mutation was applied to ERG20 to enhance its efficiency, followed by its introduction into S. cerevisiae. linalool production was further enhanced by optimizing isoprenol, prenol, carbon source, and including Mg²⁺ in the medium. Moreover, Zhou et al. [50] applied a combinatorial strategy to enhance linalool content in S. cerevisiae. This strategy involved the overexpression of the entire MVA pathway, as well as a LINS from Mentha citrata, resulting in a significant increase in linalool concentration. Later, they further enhanced linalool production by employing a double mutation in LINS and ERG20. Zhang et al. [51] conducted dual metabolic engineering of the MVA pathway to upgrade linalool content in both the mitochondria and cytoplasm of S. cerevisiae. This was achieved by introducing MVA genes into both cellular compartments, with mitochondrial localization signal [MLS] fused to the genes for transfer into mitochondria. Thus, they constructed a strain of S. cerevisiae in which the expression of LINS from Cinnamomum osmophloeum and ERG20^F96W−N127W occurred in the cytoplasm and mitochondria. This recombinant S. cerevisiae exhibited increased linalool production (7.61 mg L⁻¹). Moreover, they cultured the recombinant S. cerevisiae in media containing varying amounts of PYR and mevalonolactone as carbon sources. Notably, the medium supplemented with 70 mg L⁻¹ mevalonolactone yielded the highest linalool concentration. Park et al. [52] enhanced linalool production by integrating an inducible sensor array into the genomic DNA of S. cerevisiae. This sensor array was comprised of sequences encoding repressor proteins controlled by constitutive promoters and strong terminators, allowing for individual or simultaneous expression upon exposure to sensor molecules in the medium. The sensor molecules employed included xylose, anhydrotetracycline, vanillic acid, and IPTG which induced the expression of LINS from M. citrata.

Kirby et al. [53] significantly enhanced 1,8-cineole production in Rhodotorula toruloides by employing an N-terminal truncated GPPS from Abies grandis and a CINS from Hypoxilon sp. These synthases were introduced into R. toruloides under the control of promoters sourced from the R. toruloides genome. Subsequently, the titer increased further through medium optimization. Ma et al. [54] overexpressed (+)-bornyl diphosphate synthase [BPPS] from Cinnamomum burmanni along with all genes involved in the MVA pathway in S. cerevisiae to produce (+)-borneol. (+)-Borneol production was further significantly increased by N-terminal truncation of BPPS and incorporating a Kozak sequence. In another study by the same researcher, Ma et al. [55], (−)-BPPS from Blumea balsamifera was identified and functionally characterized and then introduced into S. cerevisiae. Similar to the previous experiment, N-terminal truncation of BPPS and the addition of a Kozak sequence were utilized to enhance (−)-borneol production. Finally, the fusion of (−)-BPPS with ERG20^F96W−N127W resulted in a further increase in the production of (−)-borneol.

2.7.2.2 Production of lavender monoterpenes in bacteria

Wu et al. [56] employed a scaffolding strategy to produce linalool in E. coli. They constructed scaffolds consisting of three domains for IPPS, GPPS and LINS enzymes, with different domain repeats for GPPS and LINS. Ligands were attached to the enzymes via linkers. The scaffold featuring one domain for IPPS, one for GPPS, and four for LINS exhibited the highest linalool production. Additionally, they optimized the concentrations of IPTG (0.5 mM), L-arabinose (0.3%), and glycerol (4%) in the medium, with an identified optimal temperature of 20º C for linalool production. Kong et al. [57] produced linalool in E. coli by designing and introducing a heterologous MVA pathway, which included genes involved in IPP and DMAPP accumulation, GPP formation, and linalool production. They showed that GPPS2 from A. grandis had a more significant effect on linalool production compared to ERG20. The lower efficiency of ERG20 in linalool production was attributed to its bifunctional activity, resulting in the production of both GPP and FPP. The recombinant E. coli strain harboring the new MVA pathway produced 15 ± 1.4 mg L⁻¹ linalool. The linalool concentration further increased to 63 ± 5.6 mg L⁻¹ with the overexpression of isopentenyl diphosphate isomerases. In a study by Wang et al. [58], linalool production in E. coli was enhanced through LINS modification. Initially, the most efficient LINS (4.8 mg L⁻¹) was obtained from Streptomyces clavuligerus [bLIS]. Then, bLIS variants with different ribosomal binding sites [RBS] and translation initiation rate [TIR] were constructed. The results demonstrated a positive correlation between bLIS expression and TIR. Additionally, a fusion tag was added to increase bLIS solubility, resulting in enhanced linalool production to 33.4 mg L⁻¹. Further optimization strategy included the addition of GPSS from A. grandis to ensure sufficient GPP availability, leading to linalool production reaching 100.1 mg L⁻¹. Finally, culturing the recombinant E. coli in a bioreactor with fed-batch fermentation achieved a significant increase in linalool production to 1027.3 mg L⁻¹.

Mendez-Perez et al. [59] achieved significant 1,8-cineole production (228 mg L⁻¹) in E. coli by constructing and introducing a plasmid harboring genes related to the MVA pathway along with the CINS gene from Streptomyces clavuligerus. To further enhance 1,8-cineole production, they inserted an additional plasmid containing another copy of CINS into E. coli. This two-plasmid system led to a 33% increase in 1,8-cineole production, reaching up to 305 mg L⁻¹. The higher level of 1,8-cineole (505 mg L⁻¹) was achieved when CINS and GPPS were harbored in one plasmid and other genes were placed in another plasmid. Karuppiah et al. [60] inserted LINS and CINS genes from Streptomyces clavuligerus into an engineered E. coli strain, where the MVA pathway was regulated by an IPTG-inducible promoter, and an N-terminal truncated GPPS was controlled by a tetracycline-inducible promoter. This approach resulted in the production of a remarkable amount of linalool and 1,8-cineole, with the latter exhibiting a higher purity (96%) compared to those obtained from Salvia fruticose, Arabidopsis thaliana and Citrus unshiu which had purities of 67%, 42% and 63%, respectively.

Lei et al. [61] engineered the de novo production of borneol in E. coli. They co-expressed a mutant BPPS enzyme from Lippia dulcis, where a single-point mutation enhanced enzymatic activity, along with an endogenous phosphatase from E. coli to facilitate the dephosphorylation of precursors to borneol. This strategy led to a notable enhancement in borneol content under optimized fermentation condition.

2.7.2.3 Production of lavender monoterpenes in Cyanobacteria

Matsudaira et al. [62] engineered a cyanobacterium strain capable of producing S-linalool. In this strain, the LINS coding sequence from Actinidia arguta was codon-optimized for the cyanobacterium and expressed under the control of tac promoter. This strain produced 11.4 mg L⁻¹ of S-linalool in shake flask culture. The S-linalool concentration further increased to 11.6 mg L⁻¹ with the expression of a mutated farnesyl diphosphate synthase derived from E. coli.

Sakamaki et al. [63] reported the photosynthetic production of 1,8-cineole in cyanobacteria. They designed and constructed a codon-optimized CINS gene from Streptomyces clavuligerus for producing 1,8-cineole in Synechococcus elongatus. They placed this CINS under the control of an IPTG-dependent promoter since it could produce 1,8-cineole directly from GPP, unlike other CINS that convert terpineol into 1,8-cineole. Although the amount of 1,8-cineole produced in their attempt was not remarkable and further attempts are needed, their study showed the feasibility of producing 1,8-cineole without the need for carbon sources like sucrose.

3 Conclusion

By now, most terpene synthase genes responsible for the production of lavender EO monoterpenes have been cloned and functionally characterized. Further, many of these genes have been used to produce the corresponding monoterpenes in bacteria, yeast, and model plants such as Arabidopsis and tobacco. Additionally, the metabolic engineering of lavender has been investigated. In the latter case, success has been limited as the constitutive overexpression of terpene synthase genes is often detrimental to the host plant, as (presumably) non-GT cells cannot tolerate large amounts of the monoterpenes they produce. This problem may be resolved if GT-specific promoters that can direct the expression of transgenes specifically in GTs are used. In this context, ongoing studies currently focus on the cloning of GT-specific promoters. Such promoters could not only help enhance EO quality and yield in lavender, but also assist researchers in using this plant as a bioreactor for the large-scale production of valuable phytochemicals.