Introduction

Dendrobium, one of the largest genera in the family Orchidaceae with approximately 1800 species worldwide [1], are popular economic plants owing to their beautiful flowers, scientific values and health benefits. Especially, D. catenatum is a highly prized Dendrobium species with a wealth of therapeutic functions in antitumor, anti-angiogenesis, anti-oxidation, anti-inflammation, diabetes alleviation, liver protection, stomach nourishing, body fluids supplementation, and immunity enhancement [2,3,4,5]. The stems of D. catenatum are the principal medicinal part containing large amounts of polysaccharides (> 30% of dry weight) and relatively low levels of ethanol extractives (~ 4.93% of dry weight), such as flavonoids, bibenzyls, phenanthrene and fluorenone [6]. The leaves, comprising approximately half of the total biomass of D. catenatum, are new sources of bioactive molecules, including phenolic compounds [7], flavonoids, polysaccharides, and amino acids [8]. The flowers of D. catenatum are rich with phenolic components (> 30% dry weight), while other substances such as essential and non-essential amino acids, polysaccharides, and volatile components are also found [9].

Many of the secondary metabolites are glucosylated to increase their solubility and stability, and activation of the glucosylated compounds is mediated by enzymes called β-glucosidases (BGLUs) [10]. BGLUs are groups of glycoside hydrolase 1 (GH1) family members found in all domains of living organisms with essential functions in removing nonreducing terminal glucosyl residues from glycosyl esters, oligosaccharides, and glycosides [10]. There are a great many glycosidases in higher plants with extensive redundant functions such as defense, cell wall remodeling, phytohormone activation, scent release, microbe/insect interactions, and secondary metabolism [11, 12]. In the last decades, genome-wide analysis of GH1-BGLUs has been carried out in a few plant species: Arabidopsis thaliana with 47 members (10 subfamilies), Oryza sativa with 40 members (8 subfamilies), Zea mays with 26 members (4 subfamilies), Brassica rapa with 64 members (10 subfamilies), and Medicago truncatula with 51 members (7 subfamilies) [13, 14]. For instance, a specific cytoplasmic BGLU has been demonstrated to hydrolyze the monoterpene alkaloid intermediate strictosidine to produce various monoterpene alkaloids [15]. Although some BGLUs related to metabolite biosynthesis were cloned and characterized from several plant species, limited information is available about BGLU family in orchids. Several BGLU genes expressed in the stems of three Dendrobium species (Dendrobium huoshanese, D. catenatum, and Dendrobium moniliforme) have been identified as hub genes that are possibly involved in polysaccharides biosynthesis [16]. In Tongan vanilla, β-glucosidase has been exogenously applied on green beans to hydrolyze glucovanillin into vanillin [17]. In Cymbidium sinense, BGLU genes are upregulated during ovule development, when β-glucosidase-mediated hydrolysis of cellulose may occur [18]. Fifteen BGLU genes are also significantly downregulated in C. sinense leaves, which may be an important reason for decreased starch content and abnormal sugar metabolism in the chloroplast [19]. In D. catenatum, BGLU genes are significantly upregulated in symbiotically germinated seeds [20], and BGLU genes associated with the glutathione metabolism pathway are upregulated in response to cadmium stress [21]. In Dendrobium crumenatum, a significant increase in β-glucosidase activity is observed during floral bud development [22].

Usually, the accumulation of medicinal metabolites, especially alkaloids, are low across tissues, making them hard to meet the threshold for drug-making [23]. Although genetic engineering and molecular modification are exceedingly helpful in creating improved varieties, the physiological and molecular mechanisms underlying metabolites production remain largely unexplored in Dendrobium plants [24]. Nevertheless, the great achievements in genome sequencing for D. catenatum [25, 26] make it convenient to conduct a genome-wide search for potential genes associated with important traits.

Previously, we sequenced the genome of D. catenatum [25] and reported DcBGLU2 as a key regulator for polysaccharide accumulation in response to phytohormone treatments [27]. However, determining the exact number of DcBGLUs in D. catenatum genome and their corresponding roles in medicinal compounds production deserve more attention. The present study identified 22 candidate GH1 DcBGLUs members through a genome-wide analysis. Then their sequence feature, molecular phylogenetic relationship, conserved motif, gene structure, chromosomal localization, and cis-elements were characterized. Furthermore, the dynamic expression patterns in three more tissues were examined based on our and other published RNA-seq data. These results provide valuable information on the DcBGLU family in D. catenatum and lay a foundation for further exploring its function in plant metabolism.

Materials and methods

Genome-wide identification of GH1 BGLUs in D. catenatum

The genome sequences and protein data of the D. catenatum were downloaded from NCBI (https://www.ncbi.nlm.nih.gov/genome/?term=JSDN00000000) under the accession code JSDN00000000 [25]. To identify D. catenatum BGLU candidates of the GH1 family (DcBGLUs), the hidden Markov model (HMM) profile of Glyco_hydro_1 (PF00232) from the Pfam database (http://pfam.xfam.org/) were downloaded and searched against the BGLU domain in D. catenatum protein sequence data using the HMMER software (version 3.2.1, http://hmmer.org/download.html). After removing repeated or incomplete proteins, the remaining potential candidates were double-checked using the SMART (http://smart.embl-heidelberg.de/) and CDD (https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi) databases. The gene identifiers used in this study are listed in the Table S1. The molecular weight (MW) and isoelectric point (pI) of each BGLU protein were calculated with the online ProtParam (https://web.expasy.org/protparam/) tool. The subcellular localization of each BGLU protein was predicted using the online ProtComp v. 9.0 server (http://www.softberry.com). The signal sequences were predicted using the SignalP (https://sFervices.healthtech.dtu.dk/service.php?SignalP-5.0) tool, and the N-glycosylation sites were detected by the NetNGlyc 1.0 server (https://services.healthtech.dtu.dk/service.php?NetNGlyc-1.0).

Phylogenetic analysis of DcBGLUs

To classify and investigate phylogenetic relationships of BGLUs, the protein sequences from A. thaliana and D. catenatum were aligned using MEGA X [28]. The phylogenetic tree was constructed according to the neighbor-joining method with the bootstrap set at 1000 replicates, and then visualized and modified using EVOLVIEW (https://www.evolgenius.info/evolview/).

Gene structures and conserved motifs

The exon/intron organization of DcBGLU genes was analyzed using the WebScipio server (https://www.webscipio.org/) [29]. The conserved motifs of DcBGLUs were identified by MEME suite (version 5.4.1) [30]. At the same time, the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html) was used to identify potential cis-elements in the 2000-bp gene promotors.

RNA isolation and gene expression analysis

Approximately 300 mg of D. catenatum leaves transiently transformed with BGLU-OE plasmids, were homogenized in liquid nitrogen and then subjected to total RNA isolation using TRIzol reagent (Thermo Fisher Scientific) following the manufacturer’s instructions. RNA integrity was checked on a 1.5% agarose gel. Subsequently, 1.0 mg total RNA was reverse transcribed into first strand cDNA with a Prime Script™ reagent Kit with gDNA Eraser (TaKaRa, Japan), according to the manufacturer’s instructions. The cDNAs were subjected to a quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) using SYBR Green Premix Kit (Toyobo, Japan) in the ABI PRISM 7500 Fluorescent Quantitative PCR System (Thermo Fisher Scientific). All the primers used were listed in Table S2, with Actin as the internal control.

Expression profiling of DcBGLUs from Transcriptomic data

Based on the transcriptome data from D. catenatum (GSE155403 and SRP150489) [31, 32], D. huoshanense (SRP122499) [33], and Dendrobium nobile (PRJNA338366) [34], the expressions of DcBGLUs and the corresponding orthologue genes were screened across four more tissues, mainly leaves, stems, roots, and protocorm-like body (PLBs), and under MF23 treatment. All the raw SRA reads were transformed into fastq format, filtered, aligned, assembled, and estimated for expression levels. The results were visualized using heatmap generated from the TBtools [35].

Cloning of DcBGLUs and transient transformation in D. catenatum leaves

The full-length coding sequences of DcBGLU2, DcBGLU6, DcBGLU8, and DcBGLU13 were amplified using D. catenatum cDNA template with the primer sets listed in Table S3. The amplified fragments were cloned into pNC-Cam1304 binary vectors (NC biotech, Hainan, China) for gene overexpression. The corresponding constructs were then infiltrated into 2-year-old D. catenatum leaves following an Agrobacterium-mediated method [36] (Fig. S1).

Measurement of the content of medicinal components

Contents of reducing-polysaccharides were determined following Tonukari et al.’s methods [37]. Contents of soluble-polysaccharides were measured using a plant soluble-polysaccharide assay kit (BC0035, Solarbio, Beijing, China) according to the manufacturer’s instructions. Contents of flavonoids were measured using a flavonoid assay kit (BC1335, Solarbio, Beijing, China) according to the manufacturer’s instructions. Contents of alkaloids were estimated following Wang et al.’s method [38]. Transiently transformed D. catenatum leaves (0.5 g) were harvested, grounded in liquid nitrogen, and used for extraction of compounds. The isolated compounds were determined by using a spectrophotometer (BeckmanCoulter DU730).

Statistical analysis

For qRT-PCR gene expression analysis, Actin from D. catenatum was used as the internal control. The 2-ΔΔCt method was used to calculate relative gene expression. Statistical significance is defined as follows: ** p < 0.01, *** p < 0.001, and **** p < 0.0001 (Student’s t-test).

Results

Genome-wide identification and characterization of BGLU genes in D. catenatum

The release and continuous update of the complete D. catenatum genome make it easier to conduct genome-wide identification of genes. A total of 22 candidate DcBGLU genes of the GH1 family were identified in the D. catenatum genome after strict HMMER screening and domain confirmation, and they were designated as DcBGLU1 to DcBGLU22 following the LOC accession numbers in the NCBI gene bank (Table S1). Among these 22 listed DcBGLUs, few were functionally characterized. For clarity, information on our previously published DcBGLUs was provided in Table S1 as well.

Physicochemical characteristics of the predicted DcBGLU proteins, including amino acid number, molecular weight, signal peptide, isoelectric point, GRAVY, N-gly site, and possible subcellular localization, are listed in Table 1. Approximately half of the predicted DcBGLU proteins (12/22) were predicted to have signal peptides ranging from 17 to 38 amino acids, targeting them to the secretory pathway. The length of the predicted precursor proteins varied between 245 aa (DcBGLU11) and 1050 aa (DcBGLU18), which correspond to protein molecular weight (MW) varied from 27.87 to 119.28 kDa. Most DcBGLU proteins contain one to ten N-linked glycosylation sites, except four DcBGLUs (DcBGLU11, 12, 16, and 21).

Table 1 Properties and locations of DcBGLUs in D. catenatum
Full size table

The theoretical isoelectric points (pI) of the predicted proteins varied widely from 5.18 (DcBGLU19) to 8.30 (DcBGLU13). DcBGLU5, 13, and 17 were basic proteins, DcBGLU1 and DcBGLU21 were neutral proteins, and the rest were acidic proteins (Table 1). The GRAVY ranged from − 0.576 to 0.022, suggesting that these DcBGLUs are all hydrophilic proteins. Additionally, most of the DcBGLUs were predicted to be in the vacuole, chloroplast and cytosol (10/22), chloroplast and/or vacuole (7/22). These results showed significant differences among the DcBGLU proteins, reflecting their diversified functions in D. catenatum.

Phylogenetic analysis and classification of DcBGLU proteins

Multiple sequence alignment was conducted to further classify and characterize DcBGLU proteins, and results showed high conservation among the members (Fig. S2). Due to the high sequence similarity, the evolutionary relationship of DcBGLU proteins was investigated. Specifically, a sequence-based neighbor-joining phylogenetic tree was constructed for the proteins, including BGLU proteins both from D. catenatum and A. thaliana, using the MEGA X software (Fig. 1). The results from the phylogenetic tree showed that DcBGLUs could be divided into seven distinct clusters. Among the 22 DcBGLUs, 6, 8, and 2 belong to clusters I, II, and, III, respectively. While cluster IV contains six members only from D. catenatum, clusters AtI (16 members), AtII (6 members), and AtIII (11 members) contain members only from A. thaliana, suggesting that gene deletion could occur during the evolution of D. catenatum.

Fig. 1
figure 1

Phylogenetic relationship of the BGLUs from D. catenatum and A. thaliana. Phylogenetic tree of BGLUs using neighbor-joining (NJ) methods was constructed by MEGA X with 47 AtBGLU and 22 DcBGLU proteins. The subfamilies were marked in different colors. The identified DcBGLUs were highlighted by red circles

Full size image

Gene structure and conserved motif of DcBGLU proteins

To identify the conserved functional motifs in DcBGLU proteins, we searched using the MEME online tool and identified ten motifs. These conserved motifs possessed 14 to 82 amino acids, and the number of motifs varied from two to ten. The results showed that more than half of the members (13/22) possessed all these ten motifs and that motif two and nine were the two most conserved ones widely present in DcBGLU proteins (Fig. 2a). However, nine DcBGLU proteins lacked the complete combination of the conserved motifs, including five (DcBGLU3, 11, 13, 16 and 19) with less than four motifs, and the other four (DcBGLU17, 20, 21 and 22) have eight motifs. Nevertheless, members within the same cluster tend to share similar compositions of motifs, indicating the highly conservation between these DcBGLU proteins and the validity of cluster classification.

Fig. 2
figure 2

Architecture of conserved motifs and gene structures of DcBGLUs. a The motif composition and distribution of DcBGLU proteins. The colored boxes represent conserved motif and the grey lines indicate non-conserved lines. b Gene structures of DcBGLU genes. Exons and introns were indicated by blue rectangles and orange lines, respectively

Full size image

In order to better understand the evolutionary relationships of the GH1 family members in D. catenatum, exon/intron structures of all the identified DcBGLU genes were analyzed. The results showed that the exon numbers have varied considerably among the DcBGLU family members, from some having only one exon (DcBGLU11, 13, and 19) to some having up to 27 exons (DcBGLU18), including five with ten or less exons and thirteen with 11-13 exons (Fig. 2b). The exon/intron organization and intron numbers of the most closely related members in the same clusters were very similar.

Promoter analysis and chromosomal distribution of DcBGLU genes

The members of different gene families could display diverse expression patterns due to functional divergence. The regulatory promoter often located upstream of the transcription initiation site of a gene has been recognized as one of the key factors in transcriptional regulation. In order to further investigate the potential regulatory mechanisms of DcBGLU genes in secondary metabolites production in response to environmental stimuli, about 2-kb upstream promoter regions of DcBGLU genes were submitted to the PlantCARE database for scanning the presence of key cis-acting elements. Three types of cis-regulatory elements, i.e., phytohormone-responsive, stress-responsive, and secondary metabolites biosynthesis (Flavonoid) elements, were detected (Fig. 3a). Firstly, five hormone-responsive elements, including MeJA, salicylic acid, abscisic acid, gibberellin, and auxin responses, were commonly presented in the promoter regions of the DcBGLU genes. Secondly, anaerobic, drought, low temperature, and defense and stress responsive elements were detected in these regions. Additionally, many light-responsive regulatory elements and MYB binding sites were widely present in these promoter regions (Supplementary File 1). Among these elements, MeJA responsive elements were the most common type of cis-regulatory elements (16/22) found in the promoters of DcBGLU genes (Fig. S3). This result supported the idea that accumulation of MeJA-induced secondary metabolites might be partially mediated by DcBGLUs. Moreover, flavonoid biosynthesis-associated cis-elements were detected in the promoters of DcBGLU2, 21, and 22, indicating the possible role of DcBGLUs in flavonoid production. The presence of cis-acting elements in the promoters of DcBGLU genes suggested that they might be responsible for adaptation to various stresses and hormone treatments by modulating the production of secondary metabolites.

Fig. 3
figure 3

Cis-elements analysis and chromosomal localization of DcBGLUs. a Cis-elements in promoters of DcBGLU genes. Different colored wedges represented different cis-elements. b Chromosomal localization of DcBGLU genes

Full size image

The genomic distribution of DcBGLU genes was analyzed to provide an overview of the location on chromosomes. The 16 out of 22 DcBGLU genes were unevenly distributed on 12 of the 19 chromosomes (Fig. 3b). Most of the chromosomes have only one DcBGLU gene on each of them, except chromosome 9 with two members (DcBGLU9 and 12) and chromosome 13 with three members (DcBGLU5, 6, and 22). Interestingly, DcBGLU5, 6 and 22, which belonging to the same subgroup I on the same chromosome 13, tended to cluster together, whereas the other family members were clustered separately.

Organ and stress-specific expression patterns of the DcBGLU genes

Plant BGLUs are well documented to play essential roles in response to developmental and environmental stimuli. Nevertheless, the functions of BGLUs in D. catenatum are far from clear. In the present study, four RNA-seq datasets of D. catenatum (GSE155403 and SRP150489) [31, 32], D. huoshanense (SRP122499) [33], and D. nobile (PRJNA338366) [34] were retrieved from NCBI Web Server, including samples across multiple tissues (root, stem, leaf, and PLB) and samples infected with MF23 (Fig. 4a). Overall, 14 differentially expressed DcBGLUs were identified in D. catenatum, among which DcBGLU8, 16, and 18 were more highly expressed in leaves than in roots. However, DcBGLU2 and 14 were more highly expressed in stems than in roots. The expression analysis indicated that four BGLU orthologue genes corresponding to DcBGLU1, 7, 9, and 20, respectively, in D. nobile were induced by MF23 infection and that the expression patterns of corresponding genes expressed in leaves or stems vs roots in D. catenatum and D. huoshanense are very similar.

Fig. 4
figure 4

Expression profiles showing members of DcBGLUs in varied organs and different treatments. a Expression of DcBGLU genes in stem, leaf, PLB, and under MF23 treatment. The log2(TPM values) of genes were shown by different color dots. Red and blue indicate high and low levels of expression, respectively. Each column indicates a treatment, and each row indicates a DcBGLU gene. b Expression patterns of DcBGLU genes in eight tissues. The expression levels of 17 DcBGLU genes were from the RNA-seq data. The eight samples included the column, sepal, stem, leaf, lip, flower bud, white root and green root tip. The color scale represents the values of log2(TPM value)

Full size image

Because of the close relevance between gene expression and function, the expression profiles of DcBGLUs in eight tissues (column, sepal, white root, green root tip, stem, leaf, lip, and flower bud) from RNA-seq datasets (PRJNA348403) [39] were analyzed. DcBGLUs showed different transcriptional patterns across various tissues (Fig. 4b). Some DcBGLUs are specifically expressed in floral organs (e.g., DcBGLU9 in the column, DcBGLU1 and 20 in the lip, DcBGLU2, 7, and 14 in the flower buds), suggesting their potential role in pigmentation and scent release. For DcBGLU10, 11, 12, 14, 17, and 21, they tended to be more highly expressed in white roots and green root tips than in other tissues. Additionally, the DcBGLUs classified in the same cluster did not always have the same expression pattern. For example, compared with the members of DcBGLU2, 9, 10 and 14 in cluster IV, DcBGLU8 and 18 from the same clade were more highly expressed in leaves. These results indicated that the expression pattern of DcBGLU genes was diverse and tissue-specific.

Functional validation of DcBGLU genes in medicinal compounds accumulation

As our previous study suggested that the expression of DcBGLU2 was closely related to reducing-polysaccharide production [27], we thus verified the function by transient overexpression of this gene in D. catenatum leaves. Upregulated expression was detected in the infiltrated leaves 6 hours later (Fig. 5a), with accumulated levels of reducing-polysaccharides, flavonoids, and alkaloids 5 days later. We also overexpressed DcBGLU8, another member in the same cluster as DcBGLU2. Results showed that overexpression of DcBGLU8 had similar promoting roles as DcBGLU2 in flavonoids and alkaloids accumulation but somehow suppressed reducing-polysaccharide production (Fig. 5b). Besides, we also tested the possible functions of the members from the other two clusters, DcBGLU6 from cluster I and DcBGLU13 from cluster II, in medicinal metabolites production. Overexpression of DcBGLU6 greatly induced accumulation of reducing-polysaccharides and flavonoids, and slightly but significantly increased accumulation of alkaloids. Meanwhile, overexpression of DcBGLU13 enhanced the production of soluble-polysaccharides (Fig. 5c) and flavonoids, but slightly reduced alkaloids (Fig. 5d, e). These results reflected the diversified functions of DcBGLUs in regulating medicinal metabolites accumulation in D. catenatum.

Fig. 5
figure 5

Overexpression of DcBGLUs altered major medicinal metabolites accumulation. a qRT-PCR verification of the overexpression status of four selected DcBGLU genes in transiently transformed D. catenatum leaves. Samples transformed with empty vectors were used as the controls (Ctrl). b-e Determination of major medicinal metabolite contents in BGLU-OE leaves. b Reducing-polysaccharide content; c Soluble-polysaccharide content; d Flavonoid content; e Alkaloid content. Leaves transiently transformed with Agrobacterium EHA105 carrying empty vector were used as control. Bars marked with stars indicate significant differences (Student’s t-test, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001)

Full size image

Discussion

Dendrobium orchids are highly prized medicinal herbs that have been widely used for many years. The plant contains various bioactive components, including polysaccharides, flavonoids, and alkaloids [40]. Several transcriptome profiles have been conducted in Dendrobium species to reveal the putative genes and pathways involved in active metabolites biosynthesis [25, 41, 42]. Moreover, genome-wide characterization of gene families has also begun to shed light on the biosynthesis pathways in several plant species [43]. Glucosidases are multifunctional enzymes that play crucial roles in plant development, secondary metabolites biosynthesis, and responses to biotic and abiotic stresses [11, 12]. Recently, the genome-wide characterization of BGLU members in different plant species has been performed with the rapid advancements of whole-genome sequencing technologies. Compared with the other plant species, the number of CH1-BGLU genes in Orchidaceae (C. sinense, Apostasia shenzhenica, Vanilla shenzhenica, Phalaenopsis equestris, Phalaenopsis aphrodite and D. catenatum) is much smaller, suggesting that there may be gene loss or pseudogenization in the process of evolution [19]. Nevertheless, in D. catenatum, a systematic analysis of BGLU family members has not been available despite being one of the most important medicinal orchids and the importance of BGLU genes in plant secondary metabolism. The recent availability of high-quality annotated reference genome [25] provides valuable resources for studying the BGLU family in D. catenatum. This study, to our knowledge, provides the first report concerning the systematic analysis and functional role of BGLU family genes from D. catenatum on secondary metabolism. This study identified 22 full-length BGLU genes from the D. catenatum genome, 16 of which were unevenly distributed on 12 chromosomes. We also analyzed the phylogenetic relations of DcBGLUs and evaluated gene expression patterns in different tissues and under biotic stress. In addition, correlation analysis between gene expression and metabolites content verified four candidate genes involved in active metabolites biosynthesis. Overall, the current study comprehensively investigated and presented the BGLU genes in D. catenatum. This endeavor will be beneficial for the in-depth exploration of biological functions of the BGLU gene family and provide potential targets for molecular breeding.

The current study identified 22 DcBGLU genes throughout the D. catenatum genome and characterized them using multi-sequence alignment and phylogenetic analysis. The results showed that the DcBGLUs shared high sequence similarity and conserved domain during evolution. DcBGLUs were mainly targeted into cytosol, chloroplast, and vacuole, where they can access the physiological substrates for catalysis. Besides, most DcBGLUs contain at least one predicted N-glycosylation site, corresponding to the estimation that many of DcBGLUs might hydrolyze their substrates through secretory pathways.

Jasmonate (JA) is a plant-specific signaling molecule broadly associated with the biosynthesis of various secondary metabolites. The exogenous application of methyl jasmonate (MeJA) has been frequently used to manipulate the production of polysaccharides [44] and alkaloids in D. catenatum [45]. In the present study (Fig. 3a), we prove that MeJA-responsive cis-elements are widely distributed in most of DcBGLU promoters (16/22), suggesting that DcBGLUs might partially mediate MeJA-induced metabolites accumulation. Salicylic acid (SA) is another stress signaling molecule that can be used as an elicitor to promote the biosynthesis of plant secondary metabolites [46]. Both MeJA and SA have commonly been used in various plant cultures, including PLB, callus, shoot, and root culture systems [47, 48]. The promoters of DcBGLU members, including DcBGLU1, 5, 6, 11, 13, and 18, have SA-responsive elements, implying their possible roles in elicitor-induced active metabolites accumulation.

Even though plant-derived alkaloids are beneficial to human health, for plants, alkaloids are essential for defensive responses to environmental stresses. For example, binary stress can increase indole alkaloid levels in Catharanthus roseus [49], drought stresses can increase the accumulation of alkaloid in roots of motherwort (Leonurus japonicas) [50], Ceratocystis fimbriata infection can enhance alkaloids production in mango (cultivar Ubá) [51], and MF23 infection can increase alkaloids accumulation in D. nobile [34]. Accordingly, many defense and stress responsive elements presented in the promoter regions of DcBGLUs. MF23 induced BGLU1, 7, 19, and 20 expressions (Fig. 4a) with increased dendrobine alkaloid levels [34]. Anthocyanin flavonoids, normally present as glycosides, are the main determinants of flowers colors [52]. Tea made from D. catenatum flowers has been consumed for many years, in which flavonoids might be one of the major beneficial ingredients. DcBGLU1, 2, 7, 9, 14, and 20 were highly expressed in D. catenatum floral organs (Fig. 4b), indicating their possible functions in flavonoid biosynthesis.

Stem-specific expressions of AtBGLU45 and AtBGLU46 in A. thaliana are responsible for hydrolysis of lignin precursors in response to various stresses [53]. Grouped in the same cluster (i.e., cluster I), however, DcBGLU16 was mainly expressed in leaves of D. catenatum (Fig. 4b). Root-specific AtBGLU42 has been known to modulate rhizobacteria-induced systematic resistance in A. thaliana [54]. Even though DcBGLU12 in the same clade (i.e., clade II) expressed in roots, it failed to respond to MF23 infection as revealed by the RNA-seq analysis (Fig. 4a), indicating the diversified functions between AtBGLUs and DcBGLUs.

Plant BGLUs are groups of GH1 enzymes that remove the nonreducing terminal β-D-glucosyl residue from glucoconjugates. BGLU enzymes contribute to various biological functions in plants, including cell wall remodeling, defense and stress response, scent release, phytohormone activation, and secondary metabolism [55]. BGLUs can hydrolyze metabolic intermediates to release glucosyl blocking groups and allow further metabolization to various natural products, many of which are medicinally relevant compounds. For example, the monoterpene alkaloid strictosidine is hydrolyzed by a specific cytoplasmic β-glucosidase to produce various monoterpene alkaloids [15, 56]. In addition to hydrolysis activity, many BGLUs also showed transglucosidase activities for synthesizing glucoconjugates, such as anthocyanin biosynthesis in Agapanthus africanus and A. thaliana [57, 58]. This glycosylation and deglycosylation process perfectly maintained the homeostasis of plant metabolites. In a previous study, we found a close association between the expression of DcBGLU2 and polysaccharide content in D. catenatum [27], suggesting that the DcBGLU2 might work as a transglucosidase in polysaccharide biosynthesis. Consistent with these findings, overexpression of DcBGLU2 increased the reducing-polysaccharides, flavonoids, and alkaloids in D. catenatum (Fig. 5). Moreover, SENSITIVE TO FREEZING 2 (SFR2) encoded enzymes from carnation (Dianthus caryophyllus), delphinium (Delphinium grandiflorum), and A. taliana can transglycosylate monogalactosyl diacylglyceride to di-, tri-, and tetra-galactosyl diacylglycerides [59, 60]. The present study also identified DcBGLU3 and DcBGLU16 as BGLU-like SFR2 genes in cluster I (Table S1), indicating similar functions between them. Overexpression of exogenous β-glucosidase in tobacco chloroplasts increased phytohormone levels and thus the resistance to white flies and aphids [61]. Likewise, the transformation of the same BGLU into Artemisia annua vacuole increased trichome numbers and artemisinin production [62]. These works showed the potential application of manipulating β-glucosidase in plants to prompt medicinal metabolites production.

Conclusion

The present study has investigated the classification and expression profile of 22 GH1 DcBGLUs in different tissues and under stress conditions. Although some family members of AtBGLUs and DcBGLUs shared high sequence similarities, the tissue and stress responsiveness were diversified. Therefore, the present study paves the way for further dissection of the distinct role of DcBGLUs in secondary metabolism and other functions.