Towards sustainable agarwood production: integrating microbial interactions, anatomical changes, and metabolite biosynthesis

Abstract

Agarwood is a highly valuable non-timber forest product mainly derived from the Aquilaria genus, widely traded in the perfumery, religious items, and traditional medicine industries. Naturally, agarwood forms within the xylem as part of the tree's defense mechanism against environmental stressors and microbial infection. The escalating demand for agarwood has led to the overexploitation of Aquilaria species, with some now classified as critically endangered. Despite advancements in artificial induction methods for sustainable agarwood supply, the intricate links between physiological and molecular mechanisms governing its formation remain poorly understood. This review addresses these knowledge gaps by examining the interplay between morphological changes in xylem structure during tylose formation and molecular alterations, particularly the biosynthesis of 2-(2-phenylethyl)chromones (PECs), key compounds in agarwood. Additionally, it integrates findings from multi-omics approaches including genomics, transcriptomics, proteomics, and metagenomics to reveal how secondary metabolite biosynthesis, including PECs and terpenes, is regulated across various Aquilaria species, regions, and induction techniques. The role of microbial communities, particularly endophytes such as Fusarium, in regulating agarwood formation is also discussed, emphasizing their involvement in both natural and artificial induction strategies. Furthermore, this review explores the role of reactive oxygen species in mediating morphological and biochemical defense responses, alongside the functions of transcription factors (TFs), protein kinases, and signaling molecules in balancing defense and growth. However, the crosstalk between key genes such as chalcone synthases, MAPK, cytochromes, NADPH oxidases, TFs, and miRNAs require further study to fully understand the complex defense mechanisms in Aquilaria trees. Overall, this review aims to bridge the current knowledge gaps by linking morphological and biochemical changes in agarwood formation, particularly PEC biosynthesis, while proposing metabolite engineering using microbial hosts as a promising tool for sustainable and technology-driven agarwood production.

One-Sentence Summary: This review explores the physiological and molecular processes behind agarwood formation in Aquilaria malaccensis, highlighting the roles of tyloses, microbial interactions, secondary metabolite biosynthesis particularly 2-(2-phenylethyl)chromones and the integration of biotechnology for sustainable production and metabolic engineering.

Graphical Abstract

Graphical Abstract.

Introduction

The global agarwood market is projected to reach USD204.31 million by 2031 (Cognitive Market Research 2022) as the most valuable minor non-timber forest products (NTFP) traded in the form of essential oil as well as chips and powder as raw materials (Gusmailina & Sumadiwangsa 2020). There are two well-known plant genera that are capable of producing agarwood, namely Aquilaria and Gyrinops from the family Thymelaeaceae (Dwianto et al. 2019; Samsudin et al. 2021) where agarwood from Aquilaria is often associated with high quality agarwood due to its distinctive aroma, whereas Gonystylus agarwood is considered as low quality agarwood (Oktavianawati et al. 2023). The Aquilaria genus is listed in the Convention on International Trade in Endangered Species of Wild Fauna and Flora under Appendix II (CITES 2021), with five of its twenty species classified as critically endangered (IUCN 2018). Beyond its role in the fragrance industry, agarwood and Aquilaria leaves are widely used in traditional medicine due to their diverse biological activities (Eissa et al. 2018; Hashim et al. 2016; Li et al. 2021; Wang et al. 2018).

Naturally, agarwood formation is an immune response of the plants towards injury caused by lightning strikes or being gnawed by ants or insects, which initiates the microbial invasion of the white Aquilaria wood (Kristanti et al. 2018). Microbial induction plays a crucial role in this process, as invading endophytes and opportunistic microbes trigger the host's defense mechanisms, leading to the production and accumulation of secondary metabolites such as sesquiterpenes and chromones that form the fragrant resin. This process occurs over a long period of up to 25 years while only around 10% of Aquilaria trees form agarwood (Barden et al. 2000). The high demand for agarwood in multiple supply chains and its historical and cultural importance (López-Sampson & Page 2018) have led to illegal trade and overexploitation of Aquilaria trees through unsustainable harvesting (Shivanand et al. 2022). Hence, to protect Aquilaria trees in the wild, CITES made the trade of wild agarwood illegal. This regulation led to advancements in artificial induction techniques by mimicking natural agarwood formation processes, including physical wounding, microbial inoculation, and chemical induction, which offers a more sustainable approach. These methods are increasingly adopted by both large scale and small scale plantations across Asia to meet the rising global demand for agarwood (Snelder & Lasco 2008). However, different artificial induction techniques produce different qualities of agarwood (Ngadiran et al. 2023; Yan et al. 2019; Zhang et al. 2021a). Moreover, agarwood sourced from different species and locations has also been shown to form different qualities of agarwood. Recent metagenomic studies have highlighted that microbial communities vary significantly across induction techniques and geographical origins, influencing the metabolic profile and quality of the resulting agarwood (Du et al. 2022a; Fu et al. 2024). A previous review by Tan et al. (2019) discussed the biosynthetic pathways involved in agarwood formation from various induction techniques while providing insights into the omics approach in agarwood formation. However, specific gaps still need to be addressed to understand the complete process of agarwood formation.

Firstly, the link between the morphological changes and the molecular alteration during agarwood formation remains a gap influenced by microbial infection. It has been known that agarwood is a resinous fragrance wood that can be formed in any part of the Aquilaria trees either through the sapwood (Jalil et al. 2022; Subasinghe & Hettiarachchi 2013) or the heartwood (Buchanan 2021). Specifically, the healthy white Aquilaria wood undergoes morphological changes by tylose formation (Pang et al. 2024) coupled with the synthesis of secondary metabolites mainly consisting of terpenoids, 2-(2-phenylethyl)chromones (PECs), and other phenolic compounds as a defense mechanism (Chen & Rao 2022). Secondly, despite the vast research in profiling agarwood constituents and advancement in molecular biology techniques to understand its formation, significant gaps remain in linking these pathways and the molecular mechanisms of agarwood formed under different conditions, including in vivo and in vitro. A clearer understanding of these biosynthetic pathways, particularly those activated during microbial infection, may facilitate microbial engineering approaches to optimize the production of specific high-value metabolites through artificial induction. Thirdly, the putative biosynthetic pathway of key agarwood compounds, particularly the PECs, must be unraveled. The PECs are not only known for their pharmaceutical properties but are also associated with the unique aroma of agarwood and their lasting pleasant odor (Yang et al. 2021).

Therefore, this paper discusses the morphological changes in Aquilaria trees during the agarwood formation by providing more details on the unique structure of Aquilaria wood and its interactions with microbial communities that initiate defense-related tissue responses. In addition, it also discusses different multi-omics approaches, which include genomics, transcriptomics, and proteomics studies to unravel the formation of agarwood sourced from different species, locations, and induction techniques. Finally, this paper also provides an overview of PEC biosynthesis and prospects of transforming the agarwood industry into a technology-integrated sustainable agarwood industry and diversification of industry via microbial metabolite engineering to conserve this depleting natural resource as well as to meet the market demands.

Physiological Perspective and Anatomical Changes during Agarwood Formation

Anatomy of Aquilaria

Aquilaria trees exhibit a unique xylem structure that supports their adaptation to stress and defense mechanisms (Fig. 1, Step A). The xylem primarily facilitates water, nutrient transport, and carbon circulation in healthy trees throughout the woody plant elongation (Pfautsch et al. 2015). Its unique xylem structure is mainly composed of four distinct parts, which are the interxylary phloem, xylem rays, vessels, and wood fibers which mostly consist of living parenchyma cells (Fig. 1, yellow boxes), as shown through 4,6-diamidino-2-phenylindole (DAPI), Safranin, and fast green staining (Liu et al. 2019). The secondary phloem embedded within the xylem is known as the interxylary phloem, a distinct feature in Aquilaria wood, unlike other angiosperms in which phloem develops outside the cambium (Chen et al. 2023; Nobuchi & Siripatanadilok 1991). These structures lead to apoptosis (Zhao et al. 2024) and formation of agarwood throughout the whole Aquilaria tree (Liu et al. 2013b; Zhang et al. 2022e), including the interxylary phloem, xylem elements, fibers, as well as axial parenchyma and ray parenchyma (Mohamed et al. 2013; Nobuchi & Siripatanadilok 1991; Sun et al. 2024; Zhang et al. 2022d).

Fig. 1.

Cross section of xylem structure in Aquilaria and its association with physiological processes during agarwood formation. (Step A) Upon exposure to biotic or abiotic stress, Aquilaria initiates a defense response that activates hormonal signals and transcriptional regulation, triggering tylosis formation in xylem vessels (XV) as a physical barrier. (Step B) Stress signals enrich defense pathways and initiate secondary metabolite biosynthesis in parenchyma-containing tissues including xylem rays (XR), interxylary phloem (IPs), and fibres. (Step C) Tyloses develop within the vessels through parenchyma expansion and occlusion, aiding in embolism and pathogen blockage. (Step D) Simultaneously, parenchyma cells metabolize starch and sugars into resinous oil droplets, particularly in the XR and IP regions. (Step E) These oils, including phenolic and terpene-rich compounds, accumulate in a compartmentalized manner. (Step F) With time, tyloses gradually disintegrate, allowing oil deposition within vessel lumens. (Step G) Fungal invasion further stimulates metabolic activity and oil biosynthesis via mycelial intrusion, leading to host-fungal interactions. (Step H) These events are interconnected with broader physiological changes in other plant organs, involving growth, senescence, and defense modulation. This integrative anatomical—physiological network underpins the characteristic resinous transformation of Aquilaria wood into agarwood.

Formation of Tyloses in Xylem Vessels

Two notable anatomical changes are associated with the defense mechanism in Aquilaria wood. Firstly, it involves the formation of tyloses in the xylem structure (Fig. 1, Step C). Tylosis in plants is formed either as individual (Murmanis 1975; Sun et al. 2006) or grouped structures (Feng et al. 2013) with varying wall thickness, pitting, sclerenchyma (Kim et al. 2024; Rioux et al. 1995), and inclusions such as resins, gels, starch, gums, crystals, or other storage products (Słupianek et al. 2021). It may appear from parenchyma cells and xylem rays, which extend into the lumens of adjacent cells through the pit, allowing vessel occlusion, inhibiting the fluid flow through embolism (Collins et al. 2009; De Micco et al. 2016) and physically limiting or preventing the infection of pathogens (Ingel et al. 2020; Shi et al. 2023). In Aquilaria wood, tylosis has been observed in the xylem vessels (Fig. 1, Step F), which was gradually reduced after resin accumulation in the interxylary phloem and adjacent sapwood tissues (Zhang et al. 2022b). Tyloses occur by forming occlusion and lignification in the vessels to compartmentalize (Fig. 1, Step E), thus creating a horizontal barrier to prevent the spread of oxidative outbursts and infection (Kashyap et al. 2020). Meanwhile, the gradual disappearance of tyloses in the vessels allows the plant to regrow by re-promoting water transport (Fig. 1, Steps F, H) (Castillo-Argaez et al. 2020). This process explains the changes in the healthy Aquilaria white wood, which gradually changes its color to yellowish, brownish, dark brown, or black upon injury or infection, either naturally or artificially induced.

Tylose formation in plants is intricately linked to the accumulation of secondary metabolites, such as lignin and aromatic substances, which are involved in plant defense mechanisms (Fig. 1, Step B). Stress hormones, particularly ethylene and jasmonates, which also regulate defense, growth, and senescence pathways, are also involved in the process. Ethylene is elevated during tylose formation, particularly in response to plant wounding and senescence (Hillis 1975; Leśniewska et al. 2017; McElrone et al. 2010; Pérez-Donoso et al. 2006). The pivotal role of ethylene in tylose formation is confirmed by studies where ethylene inhibition in pruned grapevine stems prevented tylose development (Sun et al. 2007). Its increased concentration was observed in the transition zone of heartwood formation (Taylor et al. 2002) and infection zones (Collins et al. 2009). Additionally, jasmonates, particularly methyl jasmonate (MeJA), increased agarwood formation in Aquilaria (Faizal et al. 2021; Ito et al. 2005), indicating its role in tylose formation. Apart from hormones, Leśniewska et al. (2017) reported the downregulation of PECTIN METHYLESTERASE1 (PtxtPME1) in aspen (Populus tremula × tremuloides), which induces tylose formation and activation of oxidative stress (Allario et al. 2018). This complex interplay between hormones and enzymes further triggers the accumulation of secondary metabolites (Fig. 1, Step E).

Deposition of Secondary Metabolites

While tylosis occurs in the vessels of Aquilaria wood, other tissues exhibit different responses to the stressors, underscoring the multifaceted regulatory network associated with agarwood formation. Microscopic analyses of inoculated and healthy Aquilaria wood revealed distinct layers in the xylem during agarwood formation, including a decayed layer with mycelia, a resin-filled agarwood layer, a transitional layer with vessel occlusions, and healthy tissue (Fig. 1, Step G) (Liu et al. 2019). Many other histochemical analyses of Aquilaria cells also revealed physicochemical changes in the cells due to stress. Typically, resin or brownish oil-like droplets start to accumulate in xylem rays, wood fibers, and essentially in the interxylary phloem (Fig. 1, Steps D–E) (Dwianto et al. 2019; Faizal et al. 2016; Li et al. 2022; Lv et al. 2022; Peng et al. 2014; Zhang et al. 2022a; Zhang et al. 2022b). The starch and soluble sugars in the parenchyma cells slowly disappear as they are transformed into non-starch polysaccharides, phenols, and terpenes (Li et al. 2022; Liu et al. 2019; Nobuchi & Siripatanadilok 1991; Zhang et al. 2022b). According to Li et al. (2024c), parenchyma cells are xylem rays that contain more starch, while parenchyma in interxylary phloem contains more soluble sugars, both of which lead to accumulation of oil droplets, with less amount of resin observed in xylem vessels after years of induction. Hence, this interconnected parenchyma network facilitates cell communication, secondary metabolite biosynthesis, storage, and transportation while preventing infection from spreading (Fig. 1, Steps B–H).

In addition, other parts of the plant are also affected by the stresses, with a positive correlation between agarwood formation and the interxylary phloem area, soluble sugar growth rate, starch utilization, leaf nitrogen content, photosynthetic rate, chlorophyll content, and ray cell wall thickness (Fig. 1, Step H) (Li et al. 2022). On the other hand, the vessel cell wall thickness and ray cell height of the Aquilaria tree were negatively correlated with the agarwood formation. It may be due to the depolymerization of lignin in the cell walls of infected wood, which was observed via phloroglucinol stain accompanied by the absence of starch in interxylary phloem of infected Aquilaria wood compared to the uninfected wood (Adams et al. 2016). These observations further indicate that agarwood formation involves complex structural modifications and enhanced metabolic activities across plant tissues. Furthermore, pathogenic invasion of the microbes and fungal community, either as pathogens or symbionts, might also influence the changes in the anatomy of Aquilaria wood (Ramli et al. 2022). It happens as the fungus extends its mycelia into the Aquilaria cells and utilizes the starch for growth and defense against the plant immune system (Fig. 1, Step G) (Adams et al. 2016). The living parenchyma cells are the most vital elements for the formation of agarwood since they can biosynthesize the resinous substances of agarwood, which can prevent further damage in the infected area (Kuroda 2015; Xu et al. 2023a).

Microbial Role in Regulating Aquilaria Tissue Dynamics and Secondary Metabolite Biosynthesis

Microorganisms, particularly endophytic fungi and bacteria, are pivotal in the initiation and regulation of agarwood formation in Aquilaria species, serving as both biotic stressors and metabolic inducers (Li et al. 2024c). Metagenomics and next-generation sequencing (NGS) have unveiled a dynamic microbial succession during different stages of A. malaccensis agarwood development (Bora et al. 2025), with key genera such as Bacillus, Klebsiella, Pantoea, and Saccharomyces dominating various phases of agarwood formation (Bora et al. 2025). Meanwhile, metagenomics studies on different A. sinensis tissues and agarwood found that uninduced white trunks, brown trunks, and bark harbored distinct microbial communities, with dominant genera including Burkholderia, Sphingomonas, Listeria, and Fonsecaea more dominant in agarwood compared to white wood which contained Mortierella and Aspergillus (Fu et al. 2024). Some taxa such as Colletotrichum, Pseudomonas and Acinetobacter were shared across tissues (Fu et al. 2024), suggesting the microbiota role in regulating growth and defense in Aquilaria trees. Other studies involving different Aquilaria species and induction methods, such as drilling, fermentation, and insect attacks, have reported comparable shifts in microbial diversity and structure, reinforcing the association between specific microbes and agarwood formation (Du et al. 2022a; Wang et al. 2022b).

Despite these variations, certain microbial groups particularly endophytic fungi from the Nectriaceae family like Fusarium and bacterial genera like Bacillus are consistently reported as key players in agarwood induction and metabolite release (Du et al. 2022a). For instance, artificial inoculation using specific microbial strains, including Fusarium proliferatum, Lasiodiplodia theobromae (Chen et al. 2017; Tian et al. 2013), and Phaeoacremonium rubrigenum (Liu et al. 2022a), including combination of different fungi (Ramli et al. 2022; Li u et al 2013  a; Zhang et al. 2025a). Notably, P. rubrigenum triggered sesquiterpene biosynthesis via the mevalonate (MVA) pathway, modulating key transcription factors (MYB, bZIP, WRKY) linked to terpene production (Liu et al. 2022a). Furthermore, saprophytic Bacillus strains assist in compound release through cellulose degradation in dead wood (Yang et al. 2022). Understanding these microbial functional groups is critical for optimizing artificial inoculation techniques and developing sustainable, biologically driven agarwood production strategies through a synergistic host-microbe relationship (Li et al. 2023; Li et al. 2024c).

Crosstalk Between ROS and Anatomical Changes

Reactive oxygen species (ROS) are constantly produced in plants as by-products of metabolic processes, primarily within chloroplasts and peroxisomes. Additionally, ROS also originates from the unavoidable leakage of electrons to oxygen during electron transport in chloroplasts, mitochondria, and plasma membranes (Mansoor et al. 2022). Under biotic and abiotic stress conditions, ROS levels increase significantly, leading to oxidative damage within plant cells. The key changes in the anatomy of Aquilaria tissues facilitate the trade-off between growth, defense, and programmed cell death by managing the levels of oxidative stress and pathogenic invasion. This crosstalk is mainly regulated by the ROS, which can damage the tissues and simultaneously program the cells to death (Ye et al. 2021). Usually, plant defense mechanisms stop the spread of oxidative outburst from wounded cells to the adjacent cells via the apoplast (Evans et al. 2016; Shapiguzov et al. 2012) and plasmodesmata (Evans et al. 2016; Yu 1994). This crosstalking of the ROS signals to all neighboring cells triggers the action of antioxidant enzymes, which at a certain level initiates the biosynthesis of secondary metabolites such as alkaloids, flavonoids and phenolics, thus providing direct resistance against the invading pathogen (Zaynab et al. 2018).

Tylosis, a key anatomical change associated with agarwood formation, is also linked to ROS, with studies showing that pruning Aquilaria stems induces varying levels of H₂O₂, correlated with increased vessel occlusion and higher sesquiterpene abundance, such as α-humulene and β-seline (Zhang et al. 2013). Studies also revealed that different H₂O₂ concentrations either initiate the sesquiterpene gene expression at excessive level or impede its biosynthesis at insufficient levels (Gechev & Hille 2005; Zhang et al. 2013). Furthermore, the regulation of respiratory burst oxidase homologs (Rboh), particularly AaRbohA and AaRbohC, which has been linked to ROS accumulation, is observed with higher expression levels in naturally induced stems and calli treated with H₂O₂ (Begum et al. 2024). These findings highlighted the complex role of timing and concentrations of ROS outburst with the anatomical changes associated with tylosis and secondary metabolite accumulation during agarwood formation.

In a nutshell, oxidative stress acts as a switch to the structural changes in the xylem, which directly influence the quality and quantity of agarwood produced. Tylosis formation, resin deposition, and pathogenic invasion lead to the mixture of aromatic compounds in agarwood, which contribute to its therapeutic properties. A comprehensive understanding of these anatomical adaptations enables targeted induction techniques to optimize agarwood yield and quality, as well as the integration of imaging technology in identifying Aquilaria trees with agarwood. The application of omics approaches, such as transcriptomics and proteomics, may help elucidate the molecular pathways underpinning these structural transformations, bridging the gap between physiological observations and molecular mechanisms.

Multi-Omics Approaches in Unravelling the Formation of Agarwood

Transcriptomics and Genomics Studies Reveal Stress Response Under Different Induction Conditions Influenced by Timing and Germplasm

Analyzing transcriptomics data from different studies on Aquilaria species highlights commonalities such as the enrichment of terpenoid and phenylpropanoid biosynthesis and variations in the specifics of regulatory mechanisms and gene expression patterns responding to differences in induction methods, germplasms, and timings as listed in Table 1. Additionally, Aimi-Wahidah et al. (2021) identified pathways linked to structural defense mechanisms, including lignin and flavonoid synthesis, as critical to agarwood formation. The differences in enriched pathways often reflect the unique environmental and experimental conditions used in these studies, such as salt stress (Wang et al. 2016) or light conditions (Kuo et al. 2015).

Table 1.

Transcriptomics Studies in Aquilaria Species

No	Species	Plant part	Pathways	Transcription factors	Signalling molecules	References
1	A. sinensis	Healthy vs wounded stems	Terpenoid and sesquiterpene metabolism (UR), Photosynthesis (DR), MAPK signaling (UR)	WRKY (UR), MYB (UR)	Methyl jasmonate (UR), Ethylene (UR)	Xu et al. 2013
2	A. agallocha	Callus treated with red light vs callus treated with far red light vs callus under normal light	Red and farRed light response pathways (UR), Defense response pathways (UR)	TGA, AP2	NA	Kuo et al. 2015
3	A. sinensis	Different types of trunks treated with formic acid	MVA pathway (UR), Terpene synthases (UR)	WRKY, MYC	Jasmonic acid (JA) (UR), SA (UR)	Ye et al. 2016
4	A. malaccensis	Healthy vs senescing callus	Phenylpropanoid metabolism (UR), Flavonoid biosynthesis (UR)	MYB, AP2	Ethylene (UR), Brassinosteroid (UR)	Siah et al. 2016
5	A. sinensis	Untreated calli vs 150 mM NaCl induced calli at 24 and 120 hr	Stilbenoid, diarylheptanoid, and gingerol biosynthesis (UR), SA signal transduction (UR) MAPK signaling (UR)	NAC (UR), TGA (DR)	Brassinosteroid (UR), Auxin (DR)	Wang et al. 2016
6	A. malaccensis	Healthy stem vs infected white stem vs infected black stem	Plant defense mechanisms (UR), Enzyme regulator activity (UR)	AP2, bHLH	ABA (UR), Cytokinin (DR)	Aimi-Wahidah et al. 2021
7	A. malaccensis	Healthy vs wild vs artificial trunk	NA	NA	NA	Abdul Kadir et al. 2021
8	A. sinensis and ChiNan germplasm	A. sinensis wounded branch vs ChiNan wounded germplasm branch	Starch and sucrose metabolism (UR), Oxygen metabolism (DR)	MYC, bHLH	Cytokinin (UR), Auxin (DR)	Lv et al. 2022
9	A. sinensis and easier induced agarwood A. sinensis (EIAA)	A. sinensis stems vs EIAA stems at various time points after drilling: 2, 4, 6, 8, 12, 24, 48, and 72 hr	Terpenoid backbone biosynthesis (UR), Flavonoid biosynthesis (UR)	bHLH, GRAS	Auxin (DR), JA (UR)	Zhang et al. 2022a
10	A. sinensis	Untreated vs mechanically wounded xylem tissues	Phenylalanine metabolism (UR), JA pathway (UR)	bZIP, WRKY (UR)	JA (UR), SA (UR)	Xu et al. 2023a
11	A. sinensis	Untreated region (UA) vs mechanically wounded xylem area of the same plant (TA) from 3 A. sinensis replicate	Lignin biosynthesis (UR), Plant hormone signal transduction (UR), Lignin biosynthesis (UR)	GRAS, TGA, WRKY	Salicylic acid (UR), JA (UR), Auxin, ABA (UR)	Du et al. 2022b

UR = upregulated, DR = downregulated.

A few studies highlighted that mechanical wounding alone can activate immediate stress responses even in a short duration (Xu et al. 2023b; Zhang et al. 2022b). It includes the upregulation of WRKY and Myeloblastosis (MYB) transcription factors (Xu et al. 2013), driving terpenoid and sesquiterpene biosynthesis even without infection by pathogens. However, at the onset of wounding, certain WRKY and basic helix-loop-helix (bHLH) transcription factors (e.g. WRKY54, WRKY22 and bHLH93) are downregulated (Xu et al. 2023a) while positively correlated to the rate-limiting enzyme, 1-deoxy-D-xylulose-5-phosphate synthase (DXS), in terpene biosynthesis (Xiang et al. 2007). Temporal downregulation may serve to prioritize the synthesis of immediate defense-related compounds while reserving energy-intensive pathways, such as sesquiterpene biosynthesis, for later stages. Hence, Xu et al. (2023b) hypothesized that terpene biosynthesis occurs later after the PEC biosynthesis as early induction stimulates phenylpropanoid biosynthesis and JA signaling. Timing of induction illustrates dynamic regulatory shifts and enrichment of different pathways.

Meanwhile, in prolonged responses, lignin biosynthesis and plant hormone signal transduction dominate, reflecting tissue repair and secondary metabolite accumulation (Du et al. 2022b). In addition, germplasm-specific differences reveal significant variability in secondary metabolite production. For instance, Chi-Nan germplasm and easier-induced agarwood germplasm (EIAA) demonstrated enriched terpenoid pathway with higher sesquiterpene levels and enhanced JA signaling compared to normal A. sinensis (Lv et al. 2022; Zhang et al. 2022b). Prolonged wounding in normal A. sinensis showed a decrease in the enrichment of certain stress-related pathways, suggesting that Chi-Nan germplasm can acclimatize to stress more efficiently than EIAA germplasm, resulting in a lower accumulation of secondary metabolites (Zhang et al. 2022b). These differences in expressions suggest that specific germplasms may possess genetic predispositions forming agarwood with different chemical depositions at different rates.

Genomics studies further increase the number of scaffolds that can be annotated, advancing the exploration of gene expression in Aquilaria trees. Two Aquilaria species genomes have been sequenced and annotated from previous studies (Table 2) (Chen et al. 2014; Ding et al. 2020). The first draft genome (NCBIBioProject: PRJNA240626) was obtained from the callus DNA of A. agallocha (Chen et al. 2014). Meanwhile, the A. sinensis genomics and transcriptomics (NCBI: txid210372) analyses revealed dynamic expansion and contraction of gene families, offering insight into its evolutionary adaptation (Ding et al. 2020). Expanded gene families were enriched in pathways related to plant circadian rhythm, tricarboxylic acid cycle (TCA), propanoate metabolism, ribosome biogenesis, and aminoacyl-tRNA biosynthesis, indicating enhanced capacities in energy metabolism. Conversely, contracted gene families were associated with starch and sucrose metabolism, linoleic acid metabolism, and notably, sesquiterpenoid and triterpenoid biosynthesis key pathways which are involved in agarwood formation. These results reflect long-term genomic adaptation and possibly specialization of A. sinensis towards specific metabolic outputs. Understanding the adaptation of Aquilaria might provide more insights on the gene families responsible for the intricate defense system in Aquilaria while providing knowledge for future genetic improvements.

Table 2.

Genomics Studies in Aquilaria Species

No	Species	Plant part	Method	Reference
1	A. agallocha	In vitro material	Illumina HiSeq 2000	Chen et al. 2014
2	A. sinensis	Fresh leaves	Oxford Nanopore, Illumina HiSeq 2500, Hi-C sequencing	Ding et al. 2020

In addition, the formation of agarwood and the defense response in Aquilaria can be further explored by proteomics studies listed in Table 3. Artificial induction of a 50-year-old A. sinensis tree with Phaeoacremonium rubrigenum caused proteomic changes, with MVA pathway proteins upregulated, MEP pathway proteins downregulated, and three TFs (ARR2, ALR2, NAC17) upregulated (Liu et al. 2022a). Meanwhile, mechanical wounding of A. malaccensis stems identified 564 time-point specific proteins and 21 wound-response proteins linked to defense, enriching pathways like JA biosynthesis and phenylalanine metabolism (Hishamuddin et al. 2019). These studies corroborate with the transcriptomics studies as the terpene synthase pathway is observed to be enriched in the proteomics studies. Multi-omics approaches allow a more comprehensive exploration of the Aquilaria genetics during agarwood formation.

Table 3.

Proteomics Studies in Aquilaria Species

No	Species	Plant part	Method	Pathways enriched	Reference
1	A. sinensis	White wood vs agarwood	iTRAQ LC-MS/MS	Sesquiterpene biosynthesis, glycan biosynthesis, fatty acid biosynthesis, plant defense, peroxide removal, and disease resistance pathways.	Ye et al. 2018
2	A. sinensis	P. rubrigenum treated tree vs non-treated tree	Tandem Mass Tags (TMT)	Sesquiterpene biosynthesis (upregulated MVA pathway; downregulated MEP pathway), stress resistance, secondary metabolism, plant development.	Liu et al. 2022
3	A. malaccensis	Mechanical wounding at different time points (0, 2, 6, 12, and 24 h)	Time-based LC-MS/MS	JA biosynthesis, terpenoid backbone biosynthesis, phenylalanine metabolism, and wound-response biological processes.	Hishamuddin et al. 2019

Role of ROS, MAPK, TFs, and Signaling Molecules in Stress Response Pathways and Terpene Biosynthesis

Aquilaria trees undergo a complex and interconnected pathway during the formation of agarwood comprising the outburst of ROS, leading towards the expression of genes responsible for the deposition of secondary metabolites while mediating between growth and defense stages (Fig. 2, Stages 0–5). Under normal conditions, the plant is primarily engaged in growth-related processes, relying on glycolysis and the pentose phosphate pathway to generate precursors such as phosphoenolpyruvate (PEP) and erythrose-4-phosphate for energy production, storage, and primary metabolism (Fig. 2, Stage 0). Upon exposure to stress, elevated ROS levels trigger signaling cascades, transcription factor activation, and phytohormone responses that lead to the biosynthesis of secondary metabolites such as sesquiterpenes and chromones, or ultimately, programmed cell death (Stages 1–5). ROS serve as critical initiators of plant stress responses, acting both directly by regulating defense-related genes (Huang et al. 2019; Su et al. 2024) and indirectly through signaling cascades such as the mitogen-activated protein kinase (MAPK) pathway (Fig. 2, Stages 1–2) (Takata et al. 2020) and Ca²⁺ signaling (Ravi et al. 2023) (Fig. 2; Wang et al. 2016, Stage 1). This dual role enables plants to adapt quickly to environmental stressors by activating key stress response genes or the TFs like WRKY, bHLH, MYB, and NAM, ATAF and CUC domain protein (NAC) (Fig. 2, Stage 3) (Wang et al. 2016; Xu et al. 2013), which regulate stress-responsive genes including terpene and PECs biosynthesis (Fig. 2, Stage 4) (Liu et al. 2024). Furthermore, the spike in ROS also increases the expression of signaling molecules such as abscisic acid (ABA), JA, salicylic acid (SA), and ethylene (Fig. 2, Stages 1–2).

Fig. 2.

Schematic representation of ROS-triggered defense responses and terpene biosynthesis in Aquilaria under stress conditions. In normal conditions (Stage 0), Aquilaria primarily engages in glycolysis and the pentose phosphate pathway (PPP) to sustain growth and maintain basal levels of signaling molecules. Under stress, reactive oxygen species (ROS) are generated (Stage 1) via cellular oxidation and the activity of Respiratory Burst Oxidase Homologs (Rboh), elevating ROS beyond normal physiological levels. This increase initiates a plant defense cascade by activating defense-related genes and signaling pathways (Stage 2), including MAPK phosphorylation and the accumulation of stress phytohormones (e.g. ABA, ethylene, jasmonic acid). These signals converge on transcription factors (TFs) such as WRKY, MYB, bHLH, and NAC (Stage 3), which regulate the biosynthesis of secondary metabolites, including sesquiterpenes (e.g. δ-guaiene) and 2-(2-phenylethyl)chromones (Stage 4). Terpenes are specifically synthesized via the MVA and MEP pathways and are further modified. In extreme stress, ROS accumulation may lead to programmed cell death (PCD) (Stage 5), a defense response that limits pathogen spread. The system balances defense and metabolic adaptation, eventually restoring growth processes.

Signaling molecules further regulate the expression networks of stress response genes by directly activating TFs and indirectly modulating downstream pathways (Lv et al. 2019; Wang et al. 2016). Signaling molecules like jasmonates and ABA have been reported to increase with the expression of some key genes involved in pathogen resistance, such as S-adenosyl-l-methionine-dependent methyltransferases (SAM-Mtases) (Seo et al. 2001), CYP450s (Liao et al. 2017; Liu et al. 2022b; Xu et al. 2018), and transcription factors like vascular plant one-zinc-finger (VOZ) (Prasad et al. 2018; Schwarzenbacher et al. 2020) (Fig. 2, Stage 2). SAM-Mtases including O-methyltransferases (OMTs), are key enzymes involved in the modification, diversification, and mediation of signaling molecules (Dong et al. 2022) and secondary metabolites, particularly the phenylpropanoid pathway (Joshi & Chiang 1998; Zhang et al. 2021b) by catalysing the transfer of methyl groups to different nucleophilic atoms (Kozbial & Mushegian 2005; Lashley et al. 2023). Apart from that, VOZ acts as a regulator of signaling pathways for defense against pathogen infection (Kumar et al. 2018; Prasad et al. 2018; Xu et al. 2024). Moreover, the VOZ cis-element was also identified in the promoter region of 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGR) in A. agallocha (Chen et al. 2014). HMGR is a rate-limiting enzyme in the mevalonic acid (MVA) pathway involved in the biosynthesis of terpenoids, sterols, and isoprenoids (Wei et al. 2019).

In addition, protein kinases, especially MAPK, also play a mediating role in the stress response of Aquilaria trees (Wang et al. 2016) potentially by regulating cell growth, differentiation (Chen et al. 2001; Taj et al. 2010; Xu & Zhang 2015), and programmed cell death (Ren et al. 2002; Vilela et al. 2010) for plant's innate immunity (Yue & López 2020). In direct pathways, MAPK may phosphorylate TFs like WRKY or bHLH, enabling their binding to promoter regions of the bioactive genes producing terpenes and flavonoids (Ishihama et al. 2011; Singh et al. 2024; Yao et al. 2022; Zhang et al. 2023) while working together with other signaling molecules like jasmonates (Cox 2022) (Fig. 2, Stages 2–4). In addition, MAPK also signals oxidative outburst via the phosphorylation of the promoter region of NADPH oxidase (Adachi et al. 2015; Manna et al. 2023) and indirectly regulating the biosynthesis of secondary metabolites (Fig. 2, Stage 2). For instance, Xu et al. (2017) reported that MYC2 is involved in the regulation of sesquiterpene biosynthesis in A. sinensis, which is reported to be activated by MAPK (Sethi et al. 2014), suggesting its complex role during the formation of agarwood. Apart from MAPK, the proline-rich receptor-like kinase (PERK) and cysteine-rich receptor-like protein kinase (CRR-RLK) genes are also upregulated together with terpene synthase (TPS), chalcone synthase (CHS) and O-methyltransferase (OMT) gene expressions (Wang et al. 2016; Xu et al. 2013). PERK and CRR-RLK are associated with cell wall integrity during cell expansion (Borassi et al. 2021), suggesting their possible role during plant exposure to fungal attack and programmed cell death during agarwood formation.

In plant response to stress, TFs serve as a switch to the expression of secondary metabolite biosynthetic genes and genes responsible for anatomical changes in plants, particularly WRKY, bHLH, NAC, Heat Shock Factor (HSF), and MYB TFs (Fig. 2, Stage 3). For example, WRKY TFs regulate terpene synthase genes during defense (Wang et al. 2016), while MYB and AP2 facilitate phenylpropanoid and terpenoid biosynthesis (Li et al. 2021) (Fig. 2; Xu et al. 2013, Stages 3–4). These TFs, influenced by ROS and MAPK signaling, act by bridging hormonal cues and enzymatic activity (Li et al. 2021) (Fig. 2; Xu et al. 2017, Stages 1–3).

Terpene biosynthesis begins with the cytosolic mevalonic acid (MVA) and plastidial methylerythritol phosphate (MEP) pathways, generating the precursors, isopentenyl pyrophosphate (IPP) and dimethylallyl pyrophosphate (DMAPP) (Tholl 2015). These precursors are converted into substrates like geranyl diphosphate, farnesyl diphosphate (FPP), and geranyl geranyl diphosphate (GGDP) by prenyl transferases, which are subsequently cyclized by terpene synthases (TPS) to form monoterpenes, sesquiterpenes, and diterpenes, respectively. Notably, sesquiterpene enzymes can synthesize multiple products since their domain supports multiple substrate specificity, a feature that enhances chemical diversity. For instance, the same sesquiterpene synthase can catalyze the production of δ-guaiene, α-guaiene, and α-humulene, critical components of agarwood (Xu et al. 2017). Terpenoids are also further modified by cytochrome P450 monooxygenases (CYPs), which introduce functional groups, linking terpenoids with phenylpropanoids to form unique chemical complexes like sesquiterpene-chromones (Bathe & Tissier 2019).

As the plant overcomes stress, the regulatory focus shifts from defense to growth, where carbohydrate metabolism is essential in ensuring a constant supply of energy for the cell processes (Fig. 2, Stages 1–0). ROS and MAPK signaling transition from promoting defense mechanisms like programmed cell death (PCD) (Fig. 2, Stage 5) to facilitating tissue remodeling and cell differentiation (Lv et al. 2019). This shift, marked by the reactivation of auxin and cytokinin pathways, underscores the ability of the plant to balance survival and development under fluctuating conditions, maintaining both metabolic resilience and growth potential (Das et al. 2021; Liu et al. 2021). Together, all these elements create a robust regulatory system that adapts dynamically to stress conditions, ensuring efficient biosynthesis of secondary metabolites (Xu et al. 2013; Xu et al. 2023a).

An Overview of 2-(2-Phenylethyl)Chromone Biosynthesis

Occurrence and Types of 2-(2-Phenylethyl)chromone

Chromone is a compound with an aromatic ring and pyran ring known as benzopyran, associated with many medicinal properties (Khadem & Marles 2011; Reis et al. 2017; Semwal et al. 2020). There are many different types of chromones, such as benzochromones, furanochromones, pyranochromones, hydroxypyranochromones, and PECs. In Aquilaria trees, there are about 180 sesquiterpenoids and 200 PECs, and 10 sesquiterpenoid-chromone derivatives have been identified in agarwood. However, the exact pathway involved in the formation of PEC is still not comprehensively understood despite it being the main component of agarwood (Li et al. 2021; Yu et al. 2022). Transcriptomics analysis revealed the enrichment of the phenylpropanoid pathway, with chalcone synthases (CHS) and O-methyltransferases (OMT) upregulated and hypothesized to produce key precursors for PEC biosynthesis (Wang et al. 2016; Xu et al. 2023a). 2-(2-Phenylethyl)chromone or 2-(2-phenylethyl)chromen-4-one is a rare type of chromone and has only been found in Aquilaria, Gyrinops (Aqmarina Nasution et al. 2019; Shao et al. 2016), Gonystulus, Imperata cyclindrica (Liu et al. 2013a; Yoon et al. 2006), Cucumis melo (Ibrahim 2010; Ibrahim & Mohamed 2015), Bothriochloa ischaemum (Wang et al. 2001) and Botryosphaeria rhodian (Zhang et al. 2017), a pathogenic fungus from A. sinensis. These findings suggested that while PECs are synthesized in plants via a distinct biosynthetic pathway, they can also be found in fungi (Zhang et al. 2017).

Different PEC derivatives are classified into oxido-type, agaro-type and the most abundant which is flinder type (Sun et al. 2024) or flindersiachromone (Fig. 3A). According to Yan et al. (2019), PECs with the prefix tetrahydro- and agarotetrol- are classified as agaro-types. Meanwhile, PECs with prefixes diepoxy- and epoxy- are known as the oxido-types. Lastly, the flinder-type PECs, the most common PECs, have phenylethyl at carbon 2 of the chromone backbone with an aromatic ring and pyran ring, known as benzopyran. PECs are known as chemical markers to assess the quality of agarwood (Gao et al. 2014; Hung et al. 2014; Ismail et al. 2017) and its pharmacological properties including antioxidant properties (Fan et al. 2022; Li et al. 2020) and anti-cancer (Li et al. 2020; Suzuki et al. 2017), neuro-protective, acetylcholinesterase inhibitory, antibacterial and anti-inflammatory (Chen et al. 2012; Zhang et al. 2022a; Zhang et al. 2022c). Furthermore, PECs were found to act as partial agonists of peroxisome proliferator-activated receptor γ (PPARγ), increasing the production of adiponectin, which in low levels is associated with diverse human diseases, including diabetes, obesity, atherosclerosis, and cancer (Ahn et al. 2019). This finding explains why agarwood can ameliorate various diseases as traditional medicine.

Fig. 3.

Biosynthesis and classification of 2-(2-phenylethyl)chromones (PECs) in Aquilaria. (A) PECs are structurally categorized into three major types: flinder-type (flindersiachromones), agaro-tetrol-type, and oxido-type. Flinder-type PECs are the most abundant and are characterized by a chromone backbone substituted with a phenylethyl group at the C-2 position and a fused aromatic-pyran ring system. Agaro-type PECs include tetrahydro- and agarotetrol derivatives, while oxido-types feature epoxy or diepoxy substituents. (B) Polyketide synthase (PECPS) catalyzes the formation of C6-C5-C6 diarylpentanoid scaffolds, a key precursor for PEC biosynthesis. PECPS employs a unique two-stage catalytic mechanism, first catalyzing the decarboxylative condensation of 4-hydroxyphenylpropionyl-CoA and malonyl-CoA to form a β-diketide acid intermediate. This intermediate is then hydrolyzed and used in a second, separate condensation with benzoyl-CoA, which involves another decarboxylation to generate the final diarylpentanoid product.(C) Flinder-like compound with pyran ring can also be synthesized by type III polyketide synthases such as AsPKS3, AsPKS4, and AsPKS5. These enzymes catalyze the condensation of various CoA-activated precursors with one or two malonyl-CoA units to form intermediates such as p-hydroxybenzalacetone and 4-hydroxy-6-phenethyl-2H-pyran-2-one. These reactions contribute to the chemical diversity of PECs observed in agarwood.

Unravelling the PEC Biosynthesis: Integrated Insights from Metabolomics, Transcriptomics, and Enzymatic Studies

The biosynthesis of PECs is influenced by agarwood induction methods and the associated defense-related metabolic pathways. Metabolomics studies revealed that the induction types affect the composition of different PECs. Takamatsu & Ito (2021) showed that mechanical wounding using stainless steel pins produced an even distribution of oxido-, agaro-, and flinder-type PECs, while chemical treatments with MeJA, SA, and ethylene-generating materials favor flinder-type PECs. This finding suggests flinder-type PECs are produced after the signaling hormones are enriched. This notion is supported by Wei et al. (2014), who observed flinder-type detected weeks later after the detection of oxido-types and agaro-types. The diepoxy-tetrahydro-2-(2-phenylethyl)chromones are suggested as the precursor which undergoes rearrangement to form the oxido-types, which are, in turn, the intermediates forming the flinder-types (Wei et al. 2014). Takamatsu & Ito (2021) corroborated this, showing a temporal decrease in oxido-type PECs as flinder-type PECs increased, indicating sequential biosynthetic conversion. This rearrangement may be catalyzed by O-methyltransferases, flavone 3-O-methyltransferase, and caffeoyl-CoA O-methyltransferase (Wang et al. 2016).

Meanwhile, it is known from transcriptomics studies that the phenylpropanoid pathway provides precursor molecules to catalyse the rate-limiting step for PEC biosynthesis. This pathway begins with phosphoenolpyruvate (PEP) and erythrose-4-phosphate, producing shikimate, which is converted into phenylalanine by phenylalanine ammonia-lyase (PAL). Phenylalanine is a precursor for p-coumaroyl CoA, a substrate for type III polyketide synthases (PKS) like chalcone synthase (Vogt 2010). A few polyketide synthases (PKSs) including chalcone synthases (AsCHSs) were upregulated in the transcriptomics data of the salt stress callus which produced high abundance of PECs (Wang et al. 2016). Then, all putative PKSs including PECPS were expressed in Escherichia coli where their enzymatic reaction products lead to the identification and characterization of 2-(2-phenylethyl)chromone precursor synthase (PECPS). PECPS is a diarylpentanoid-producing PKS, which catalyzes the formation of diarylpentanoid scaffolds (C6-C5-C6) using benzoyl-CoA, 4-hydroxyphenylpropionyl-CoA and malonyl-CoA (Wang et al. 2022a) (Fig. 3B). The expression of these genes is regulated by the spike of ROS during early stages of stress where salt stress has been observed to increase H₂O₂ production, which is decreased with the synthesis of PECs, particularly 6,7-dimethoxy-2-[2-(4'-methoxyphenyl)ethyl]chromone. The presence of PECs during the early stage of salt stress was also correlated with elevated expression of AsRbohsA, AsRbohsB, and AsRbohsC (Wang et al. 2018). Moreover, H₂O₂ scavenger and NA DPH oxidase inhibitor increased PEC accumulation, suggesting that H₂O₂ regulates PEC biosynthesis during early salt stress.

Meanwhile, enzymatic studies further elucidate the roles of PKS and CYPs in PEC diversification. PECPS is a homodimeric enzyme with a conserved catalytic triad (Cys-His-Asn), which allows it to utilize a unique mechanism distinct from other PKS enzymes. The substitutions in key residues, such as Phe340 and Ser134, enable PECPS to accept various substrates and catalyze complex reactions in a single catalytic pocket (Wang et al. 2022a). This promiscuity in PKS substrate specificity adds another layer of complexity. For example, polyketide synthases type III, such as AsPKS3, AsPKS4, and AsPKS5, which are highly expressed in A. sinensis agarwood, are shown to catalyze the formation of flinder-like compound with pyran backbones and other novel compounds by utilizing diverse substrates (Fig. 3C) (Xiao et al. 2022). This enzymatic flexibility underpins the chemical diversity observed in agarwood and suggests an intricate interplay of reductases, oxidoreductases (Wang et al. 2022a), oxygenases (CYPs, 2-oxoglutarate-dependent oxygenase [Zhang et al. 2025b]), and transferases (O-methyltransferase, caffeic acid 3-O-methyltransferase [Wang et al. 2016; Xu et al. 2023b]) in PEC biosynthesis. In addition, CYPs, such as AaCYP73A1 and AaCYP73A2, also contribute to PEC diversification by catalyzing hydroxylation, oxidation, and ring-opening reactions. For instance, agarotetrolchromones are hypothesized to form via the enzymatic opening of epoxy rings in oxido-type PECs, highlighting the role of CYPs in structural modifications (Das et al. 2021).

While significant progress has been made, gaps remain in unraveling the biosynthesis of PEC and other valuable agarwood compounds. This includes characterising various type III PKS genes, substrates and intermediates, as well as exploring alternative pathways and regulators (TFs, kinases) of these type III PKSs and other biosynthetic genes. Advancing these via biological interventions like metabolite engineering will deepen understanding of not only PEC biosynthesis but also other valuable agarwood compound biosynthesis. However, expressing key plant biosynthetic enzymes such as type III PKS in microbial systems remains challenging due to its network complexity, regulatory constraints, and the metabolic burden to the host cell (Gao et al. 2021). Although type III PKSs are structurally simpler homodimeric ketosynthases (Yu et al. 2012), their catalytic promiscuity depends on their ability to condense various CoA-activated starter units with extender units, which is catalysed by enzymes like methyltransferases, oxidoreductases, prenyl transferases, and glycosyltransferases (Bisht et al. 2021; Noel et al. 2005) in a series of downstream reactions. Previous studies have shown that type III PKSs have been reported to catalyze reactions using a range of different substrates including phenylpropionyl-CoA, cinnamoyl-CoA, m-coumaroyl-CoA, dihydro-m-coumaroyl-CoA, p-coumaroyl-CoA, and dihydro-p-coumaroyl-CoA while producing different products (Abe 2020; Yu et al. 2012) leading to inconsistent expression of targeted metabolite. Furthermore, the native metabolic pathways of these microbial systems such as fatty acid synthesis compete with the engineered type III PKS for essential precursors like malonyl-CoA (Palmer & Alper 2018) which limits the production of desired polyketides by engineered PKS. This interconnectedness between the cellular metabolism may lead to unstable phenotypes in engineered strains resulting in cell suicide or uncertainty in targeted gene expression. This competition between host cell and engineered cascades can be mitigated by exploring other malonyl-CoA producing pathways (Milke & Marienhagen 2020; Xu 2020), chemical inhibitors like cerulenin (Kallscheuer et al. 2016; Santos et al. 2011) or employing CRISPR interference (CRISPRi) (Wu et al. 2015). Another common challenge in metabolite engineering is the differences in codon usage which influences the protein folding of not only type III PKS but also other cascades of enzymes, impairing protein solubility and function. For instance, heterologous expression of eukaryotic CYPs, which often function alongside type III PKSs in PEC biosynthesis, is further complicated by their membrane-bound nature and the need for eukaryotic-specific folding and redox interactions (Durairaj & Li 2022). To alleviate this, co-culture engineering has emerged as a promising approach, where complex pathways are divided into modules and distributed among different microbial strains (Cui et al. 2019) allowing cross-feeding of intermediates and reducing host cell burden. Overall, these challenges can be addressed through systems biology, synthetic biology, and chassis engineering to unravel the PEC biosynthesis thus allowing the metabolic engineering of agarwood compounds.

Microbial Metabolite Engineering and In Vitro Induction

Different biotechnological techniques are an addition to the existing agarwood industry offering more sustainable approaches and diversification of the typical agarwood industry life cycle, further protecting this endangered agarwood producing genus. Firstly, the integration of metabolite engineering with different approaches and strategies allows the expression of key plant compounds including flavonoids, phenylpropanoids, and terpenoids and 2-(2-phenylethyl)chromones, in different microbial hosts (Table 4). These studies demonstrate achievements such as the complete reconstruction of biosynthetic pathways, CRISPRi-based flux control, novel enzyme discovery, and record-high production titres of compounds like naringenin, patchoulol, and pinostrobin. Meanwhile, in Aquilaria, similar engineering efforts have enabled microbial hosts to replicate key enzymatic activities involved in agarwood biosynthesis, validating terpene synthases and chromone-producing enzymes. For example, Saccharomyces cerevisiae was engineered to overexpress the entire MVA pathway, boosting the farnesyl pyrophosphate (FPP) pool and enabling the production of major sesquiterpenes like δ-guaiene, α-humulene, and β-eudesmol (Promdonkoy et al. 2022). Similarly, E. coli was utilized to characterize terpene synthase genes such as TPS9 and TPS12, converting FPP into diverse sesquiterpenes including cedrol, nerolidol, and hinesol (Kurosaki et al. 2016; Sundaraj et al. 2023; Yu et al. 2023). Meanwhile, unicellular algae (Chlamydomonas reinhardtii) served as a photosynthetic chassis to produce nine sesquiterpenoid skeletons through a green bioprocess involving in situ extraction and oxygenated functionalisation, mimicking natural agarwood chemistry with minimal environmental impact (Gutiérrez et al. 2024). Besides metabolic engineering, in vitro cultivation of Aquilaria callus for agarwood induction is also a sustainable source of agarwood. Aquilaria shoot and callus cultures that have been elicited with methyl jasmonate and fungal extracts (Fusarium solani) (Faizal et al. 2021; Kumeta & Ito 2010), Trichoderma (Jayaraman & Mohamed 2015) or even with different metal-organic frameworks (MOFs) (Overmans et al. 2025) have previously demonstrated the ability to trigger the accumulation of sesquiterpenes, chromones, and defense-related metabolites (Table 4). These findings suggest a multi-step pipeline comprising target compound identification, pathway elucidation, host selection, gene integration, elicitor screening, and scale-up optimization (Gutiérrez et al. 2024; Promdonkoy et al. 2022; Zhang et al. 2025a) providing alternative options to the current agarwood industry.

Table 4.

Engineered Microbial Hosts, In Vitro Systems and Metabolites Produced

Host/System	Engineering/elicitation strategy	Key metabolites produced	Highlights	Reference
E. coli	Overexpression of (−)-patchoulol synthase (PTS) and exogenous mevalonate (MVA) pathway; fermentation optimisation; solid − liquid phase partitioning	(−)-Patchoulol, α-Bulnesene, trans-β-Caryophyllene	First (−)-patchoulol production in E. coli with SLPPC; 5-fold titre increase; 99.65% recovery using Diaion HP20	Aguilar et al. 2020
E. coli	Expression of flavone/flavonol biosynthesis genes; RBS tuning; tyrosine pathway enhancement; cerulenin addition	Apigenin, Kaempferol, Chrysin, Luteolin	Highest apigenin (128 mg/L) and kaempferol (151 mg/L) titres in E. coli; enabled flavonoid biosynthesis without exogenous tyrosine	Yiakoumetti et al. 2023
E. coli and S. venezuelae cocultivation	E. coli expressing phenylpropanoid backbones + S. venezuelae expressing SaOMT2 for O-methylation	Tri-methylated stilbenes, Di-/Tri-methylated flavanones/flavones	First high-yield microbial production of trimethoxystilbene and derivatives; S. venezuelae-expressed SaOMT2 showed high regiospecificity	Cui et al. 2019
E. coli	Genome engineering to overproduce pinocembrin and derivatives; enzyme screening; glycerol as carbon source	Pinocembrin, Chrysin, Pinostrobin, Pinobanksin, Galangin	Produced 353 mg/L pinocembrin; 153 mg/L pinostrobin; highest reported titres for several flavonoids	Hanko et al. 2024
E. coli	Directed evolution of CHS; thioesterase knockout; co-expression of CHIL	Naringenin, CTAL, Naringenin chalcone	Highest naringenin titre (1082 mg/L); minimized byproduct CTAL; improved enzyme specificity	Xiang et al. 2024
E. coli	CRISPRi-based fine-tuning of central carbon metabolism; malonyl-CoA redirection; flavonoid pathway construction	(2S)-Naringenin, Malonyl-CoA	421.6 mg/L naringenin titre (7.4 × increase); malonyl-CoA increased by 433%; first CRISPRi system for flux tuning in flavonoid biosynthesis	Wu et al. 2015
E. coli	Codon-optimized longifolene synthase expression; mevalonate pathway integration; two-phase fed-batch fermentation	Longifolene, FPP, IPP, Mevalonate	382 mg/L longifolene; first microbial system for longifolene production; enhanced FPP flux using E. coli IspA	Cao et al. 2019
E. coli	Transient expression of 7 enzymes; heterologous enzyme expression in E. coli and yeast; comparative omics in Apiaceae	Prim-O-glucosylcimifugin (POG), 5-O-methylvisamminoside (5-O-MVG)	First complete furochromone pathway; 5-O-MVG biosynthesis (17.48 μg/g DW); discovery of lineage-specific enzymes (SdPCS, SdPC)	Zou et al. 2025
S. cerevisiae	Integration of isopentenol utilisation pathway (IUP); knockout of ERG13; enzyme engineering; pathway flux redirection	IPP, DMAPP, Limonene, β-Carotene, Amorphadiene, Taxadiene, Lycopene	IUP simplified terpenoid synthesis; 695-fold increase in squalene; 20-fold increase in limonene; universal yeast platform for terpenoids	Li et al. 2024a
S. cerevisiae	Diauxie-inducible expression; ERG20M & GGPP synthase integration; IUP and MVA pathway coordination	Limonene, Amorphadiene, Lycopene, β-Amyrin, Taxadiene	IUP boosted GGPP by 374-fold; up to 4.3-fold improvement in terpene titres; versatile and controllable yeast platform	Ma et al. 2022
S. cerevisiae	Overexpression of 8 mevalonate pathway genes; terpene synthase screening	δ-Guaiene, α-Humulene, β-Eudesmol	ERG9 repression, enzyme fusion enhanced yield; first β-eudesmol biosynthesis	Promdonkoy et al. 2022
E. coli	Expression of terpene synthase genes (TPS9, TPS12, AmDG2)	α-Humulene, β-Caryophyllene, Cedrol, Nerolidol, γ-Eudesmol	Functional validation of cloned enzymes from Aquilaria spp.	Kurosaki et al. 2016; Sundaraj et al. 2023; Yu et al. 2023
Chlamydomonas reinhardtii	Genetic engineering: CO₂-fed photobioreactor; in situ extraction and oxidation	9 Sesquiterpenoid skeletons, Oxygenated STPs	Green bioprocess, solvent recycling, CO₂ fixation, continuous production	Gutiérrez et al. 2024
Aquilaria (callus or shoot culture)	Methyl jasmonate and F. solani crude extract elicitation	Sesquiterpenes, Chromone derivative, Aromatic compounds	MeJA induced agarwood-like metabolites; fungal elicitor modulated profile	Faizal et al. 2021; Kumeta & Ito 2010
A. crassna (callus culture)	Exposure to various Metal-Organic Frameworks (MOFs), e.g. ZIF-67	Secondary metabolites with fragrance potential	ZIF-67 elicited highest metabolite yield; metal-ligand specificity observed	Overmans et al. 2025

Conclusions and Future Prospects

Sustainable farming of Aquilaria trees is essential for maintaining biodiversity and ensuring the production of high-quality agarwood. Traditional methods of mechanical induction, such as chisel wounds and nail insertion, have mainly been replaced by modern chemical and biological techniques, including the use of MeJA, SA (Mohi-Ud-Din et al. 2021; Vǎn Thành et al. 2015), ethylene gas, and fungal inoculation with pathogens like Fusarium (Faizal et al. 2020). These methods leverage an understanding of biochemical pathways to mimic natural defense responses, producing more consistent and high-quality resin (Fig. 4). Innovations in tissue culture and micropropagation (Li et al. 2024b) need to be further explored to advance cultivation practices for a steady supply of healthy saplings with desirable traits or generation of phytochemicals in vitro (Espinosa-Leal et al. 2018; Mohaddab et al. 2022). Furthermore, the integration of advanced technologies in sustainable farming practices can transform the pre-harvesting, harvesting, and post-harvesting stages of agarwood production (Fig. 4). Genetic markers can detect and verify Aquilaria species, ensuring proper species identification and selection for optimal resin production. Islam & Banu (2021) discovered biomarkers that can differentiate wild or artificially induced agarwood-containing trees to avoid overexploitation of this critically endangered species. Imaging techniques, such as fluorescence spectral imaging (Huang et al. 2015) or microwave tomography (Rahiman et al. 2019), are also helpful to enable the non-invasive detection of agarwood resin in standing trees, improving efficiency and reducing unnecessary tree damage. Post-harvest technologies further enhance quality control, with chemical markers, electronic noses (e-noses) (Aditama et al. 2020; Mohamad et al. 2022), and biological or chemical markers (Jayachandran et al. 2014) employed to differentiate agarwood quality and authenticity. Robust processing technologies, such as supercritical fluid extraction (Pan et al. 2023) and advanced distillation techniques, ensure efficient extraction of high-value compounds while minimizing waste (Fig. 4).

Fig. 4.

The whole life cycle of sustainable agarwood industry. This figure illustrates the four key stages of the agarwood industry: (1) Cultivation, involving sustainable plantation practices, agroforestry integration, and precision agriculture; (2) Induction, where resin production is stimulated through mechanical, chemical, or biological methods; (3) Harvesting, which emphasizes sustainable harvesting, quality assessment, and advanced extraction techniques; and (4) Waste Management, promoting circular economy practices by recycling by-products and creating value-added products from waste. The cycle underscores the integration of sustainable practices across the agarwood value chain.

Overharvesting wild Aquilaria trees and reliance on non-standardized methods have led to biodiversity loss, reduced genetic diversity, and habitat destruction, placing many species at risk of extinction. Sustainable solutions, including precision agriculture technologies, such as Internet of things (IoT)-based monitoring and remote sensing, which optimize soil health and pest management (Dhanaraju et al. 2022), are needed to develop a sustainable agarwood industry. Marker-assisted breeding programs targeting resin production traits (Ahmar et al. 2021), coupled with sustainable harvesting techniques like partial harvesting and staggered induction, can enhance tree productivity and reduce environmental impacts. Integrating agarwood farming into circular economy models offers additional opportunities by repurposing by-products into biochar, essential oils, and fertilizers, minimizing waste while diversifying income streams (Toplicean & Datcu 2024). Agroforestry systems with Aquilaria intercropped alongside pepper, turmeric, or medicinal plants (Rao et al. 2004) provide further benefits, improving soil health, conserving biodiversity, and enhancing farmer livelihoods, demonstrating a pathway to sustainable agarwood production that balances economics and environmental goals (Fig. 4).

Authors Contributions

Y.S., K.K.R. and R.O. contributed to the paper's structure. Y.S. prepared the original draft, while R.O., and K.A.A. provided supervision and ideas. K.K.R. and Y.K.A. contributed ideas too. Y.S., K.K.R., and N.S.M. collected data and developed the figures and tables. R.O. and Y.S. reviewed and revised the manuscript.

Funding

This work was supported by the grant from the Ministry of Higher Education Malaysia under the grant number FRGS/1/2021/WAB04/UKM/02/2 received by R.O. and K.A.A..

Conflict of Interest

None declared.

Data Availability

No new data were generated or analysed in support of this research.