Introduction

Tyramine, the decarboxylation product of l-tyrosine (Tyr) catalyzed by tyrosine decarboxylase (TyDC) (Fig. 1), is an important biogenic amine. Tyramine plays an essential role in a variety of primary physiological processes (Negrel and Javelle 2001; Kawalleck et al. 1993; Negrel et al. 1993; Trezzini et al. 1993; Borgolivier et al. 1987). Tyramine is a biosynthetic precursor of a large number of pharmaceutically important specialized metabolites such as the benzylisoquinoline alkaloids (Facchini et al. 2000, 2002) and the Amaryllidaceae alkaloids (Jin and Yao 2019; Jin 2009, 2013, 2016).

Fig. 1
figure 1

Biosynthetic pathway of the Amaryllidaceae alkaloids exemplified by galanthamine (GAL). PAL, phenylalanine ammonia-lyase; C4H, cinnamate 4-hydroxylase; OMT, O-methyltransferase; TyDC, tyrosine decarboxylase

Full size image

The Amaryllidaceae alkaloids (> 600 members), natural benzylphenethylamine derivatives isolated from the plants of the unique Amaryllidaceae family only, were identified with huge chemical complexities and various biological activities (Jin and Yao 2019; Jin 2009, 2013, 2016). The plants of the Amaryllidaceae family have been widely applied and ranked top 20 of the most widely applied plant families because several Amaryllidaceae alkaloids have been used in clinic (Jin 2013). Galanthamine (GAL, Fig. 1), a characteristic complex tetracyclic Amaryllidaceae alkaloid, was reported to increase acetylcholine levels via the inhibition of acetylcholinesterase and/or the allosteric modulation of nicotinic cholinergic receptor. Thus, GAL has been used to reduce the symptoms of the Alzheimer’s disease patients. (Marco-Contelles et al. 2006; Marco and Carreiras 2006).

Featuring with a seven-membered azepane ring and an all-carbon quaternary stereogenic center (Fig. 1), GAL had been isolated from the plants snowdrops (Galanthus species), daffodils (Narcissus pseudonarcissus), snowflakes (Leucojum species), and ‘red-tubed lily’ and ‘golden spider lily’ (Lycoris species) (Berkov et al. 2009) of the Amaryllidaceae family. The bulbs of the aforementioned GAL-producing plants are the main tissue in which GAL is present. The contents of GAL ranged from 0.05 to 0.2% (Berkov et al. 2009). The plants of the Lycoris genus (Amaryllidaceae family) are popular as ornamental plants (Zhou et al. 2007). Their bulbs have been used in traditional Chinese medicine to improve memory and cognitive functions (Son et al. 2010; Jin 2009). Distributed in Eastern Asian regions, L. radiata had been grown in China since 2002 to extract GAL (Berkov et al. 2009).

l-Phenylalanine (Phe) and l-Tyr are the two specific aromatic amino acid precursors which were integrated into the biosynthetic pathway of the Amaryllidaceae alkaloids (Fig. 1) (Barton et al. 1963). l-Phe was converted into p-coumaric acid via phenylalanine ammonia-lyase-catalyzed deamination of l-Phe to give trans-cinnamic acid and the following hydroxylation of trans-cinnamic acid by trans-cinnamate 4-hydroxylase (Fig. 1) (Desgagné-Penix 2020; Li et al. 2018; Kilgore and Kutchan 2016; Takos and Rook 2013). p-Coumaric acid will be oxygenated and cleaved a two-carbon unit to generates 3,4-dihydroxybenzaldehyde and its analogues that will be O-methylated by O-methyltransferase to afford the catechol moiety of the Amaryllidaceae alkaloids (Desgagné-Penix 2020; Li et al. 2019, 2018; Sun et al. 2018; Kilgore and Kutchan 2016; Kilgore et al. 2014; Takos and Rook 2013). Tyramine, the non-catechol portion of the Amaryllidaceae alkaloids, originates from l-Tyr decarboxylation reaction catalyzed by TyDC (Desgagné-Penix 2020; Sandmeier et al. 1994; Barton et al. 1963). The Mannich-like condensation reaction between the aforementioned catechol and non-catechol fragments will afford norbelladine and its analogues (Desgagné-Penix 2020; Li et al. 2019; Singh et al. 2018). The condensation products will be converted into diverse Amaryllidaceae alkaloids through oxidative phenolic coupling, rearrangement, oxidation, reduction, and O/N-methylation reactions (Fig. 1).

Herein, two LrTyDCs were cloned from GAL-producing L. radiata and heterologously expressed in E. coli. Recombinant LrTyDCs were shown to decarboxylate l-Tyr to form tyramine only. The transcripts of LrTyDCs and the content of GAL in each plant tissue were evaluated by quantitative real-time PCR and HPLC–DAD, respectively.

Materials and methods

Plant materials, total RNA isolation, and cDNA preparation

The L. radiata plant collection, growth, total RNA extraction and isolation, and cDNA library construction were performed, according to the reported experimental procedures (Li et al. 2018, 2019).

L. radiata transcriptome data analysis and cloning of LrTyDCs

Three transcriptome datasets of L. radiata are available in NCBI Sequence Read Archive (SRA) under the accession number SRX1597502, SRX845524, and SRX5588289, respectively. To obtain the nucleotide sequence for candidate LrTyDC, we mined L. radiata transcriptome datasets via the BLASTN analysis using the nucleotide sequences of NpTyDC1 (GenBank accession number: KT378599) as a query for homologous alignment. Two highly homologous gene fragments, SRR1765339.26865166.1 and SRR1765339.21667464.1, were retrieved from the transcriptome data of L. radiata (SRX845524). The two fragments contain an overlapped region according to the multiple nucleotide sequences alignment results (Fig. S1, Supplementary Materials).

To amplify the partial fragment of candidate LrTyDC, a pair of gene-specific primers, LrTyDC-con-F and -R (Table S1, Supplementary Materials), were designed on the basis of the alignment results of the two nucleotide sequences (Fig. S1, Supplementary Materials). PCR amplification reaction was performed by using the previously prepared cDNA mixture as template and following the PCR cycling conditions: 1 cycle of 95 °C for 2 min, 35 cycles of 95 °C for 20 s, 69 °C for 20 s, and 72 °C for 4 s followed by a final extension at 72 °C for 5 min in a thermal cycler (Eppendorf AG, Hamburg, Germany). The amplified products were gel-purified and ligated into the pGM-T vector (Tiangen Biotech (Beijing) Co., Ltd., China). The constructs were chemically transformed into E. coli DH5α competent cells and sequenced in Sangon Biotech (Shanghai) Co., Ltd. The nucleotide sequence was analyzed by using the similarity search BLAST program.

Based on the nucleotide sequence of the above confirmed amplicon of LrTyDC, the gene-specific primers 3′-RACE-LrTyDC-GSP1 and -GSP2 (Table S1, Supplementary Materials) were designed to perform 3′-rapid-amplification of cDNA ends (RACE) experiments. The 3′-ends of LrTyDC were obtained by using 3′-RACE methodology, according to the manual instructions of SMARTer™ RACE cDNA amplification kit from Clontech Laboratories. Briefly, using the previously prepared 3′-RACE-Ready cDNA as template, 3′-RACE-LrTyDC-GSP1 and Universal Primer A mix (UPM) as primers, the 3′-ends of LrTyDC were obtained by following the PCR cycling conditions: 1 cycle of 94 °C for 3 min; 5 cycles of 94 °C for 30 s, 72 °C for 2 min; 5 cycles of 94 °C for 30 s, 70 °C for 30 s, 72 °C for 2 min; 25 cycles of 94 °C for 30 s, 68 °C for 30 s, 72 °C for 2 min followed by a final extension at 72 °C for 10 min in a thermal cycler. The PCR products were gel-purified and used as template for the second-round PCR amplification reaction using 3′-RACE-LrTyDC-GSP2 and the Nested Universal Primer A (NUP) as primers, following the cycling conditions: 1 cycle of 94 °C for 3 min; 20 cycles of 94 °C for 30 s, 68 °C for 30 s, 72 °C for 2 min followed by a final extension at 72 °C for 10 min. The PCR products were gel-purified and ligated into the pGM-T vector. The E. coli DH5α competent cells were chemically transformed with the constructs following the procedure mentioned above. The nucleotide sequences of the 3′-ends of LrTyDC were sequenced in Sangon Biotech Co., Ltd. Based on the nucleotide sequence of the confirmed core amplicon fragments and 3′-RACE products, the gene-specific primers 5′-RACE-LrTyDC-GSP1 and -GSP2 (Table S1, Supplementary Materials) were designed to perform 5′-RACE experiments. The 5′-ends of LrTyDC were obtained by using 5′-RACE methodology according to the manual instructions of SMARTer™ RACE cDNA amplification kit from Clontech Laboratories and following the identical PCR cycling conditions as described in 3′-RACE experiments. The PCR products were gel-purified and ligated into the pGM-T vector. The E. coli DH5α competent cells were chemically transformed with the respective construct following the procedure mentioned above. The nucleotide sequences of the 5′-ends of LrTyDCs were sequenced in Sangon Biotech Co., Ltd.

LrTyDC-Full-F1, -F2, and -R (Table S1, Supplementary Materials), three gene-specific primers, were designed and synthesized to clone the full-length cDNA of LrTyDCs, according to the above mentioned nucleotide sequences of the core amplicon fragments, 5′- and 3′-RACE products. Using the previously prepared cDNA mixture as template and LrTyDC-Full-F1, -F2, and -R as primers, the open reading frames (ORFs) of LrTyDCs were obtained by following the PCR cycling conditions: 1 cycle of 94 °C for 3 min, 35 cycles of 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 2 min followed by a final extension at 72 °C for 10 min. The PCR amplified products were gel-purified, ligated into the pGM-T vector, and the resulting construct was transformed into E.coli DH5α competent cells following the procedures mentioned above. The nucleotide sequences of the full length of LrTyDCs were sequenced and analyzed by using the similarity search BLAST program.

Bioinformatics analyses and 3-dimensional structure prediction of LrTyDCs

According to the reported procedures (Li et al. 2018, 2019), the basic bioinformatics characteristics of LrTyDCs were analyzed. The 3-dimensional structures of LrTyDCs were predicted by using the crystal of Papaver somniferum TyDC2 (PDB ID: 6LIU) as template. The phylogenetic tree of plant TyDCs was constructed by using the neighbour-joining method of MEGA 7.0 and the bootstrap is 1000.

Heterologous expression of LrTyDCs in E. coli

LrTyDC1-BH-F and -R, LrTyDC2-BH-F and -R (Table S1, Supplementary Materials) were designed and synthesized to amplify the ORFs of LrTyDCs from the full-length cDNA of LrTyDCs by using a HiFi Taq DNA polymerase (TransGen Biotech) with the following cycling conditions: 1 cycle of 94 °C for 3 min, 35 cycles of 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 2 min followed by a final extension at 72 °C for 10 min. The PCR amplification products were gel-purified, digested with BamH I and Hind III endonucleases, and ligated into the expression vector pET28a digested with the same restriction enzymes, respectively. The resulting recombinant plasmids, pET28a-LrTyDC1 and pET28a-LrTyDC2, were transformed into E. coli transetta (DE3) competent cells to generate the recombinant strains, respectively.

The expression of LrTyDCs in E. coli was performed according to the reported procedures (Li et al. 2019) with minor modification. Briefly, 50 μg mL−1 of kanamycin and 34 μg mL−1 of chloromycetin were supplemented into E. coli culture. When the optical density (OD600nm) of the culture reached 0.6–0.8, 0.5 mM of isopropyl β-d-1-thiogalactopyranoside (Sangon Biotech) were added into the culture to induce the expression of the target protein. The culture was shaken at 16 °C, 160 rpm for another 19 h. The cells were harvested to separate the target enzyme by following the reported procedures (Li et al. 2019). Fractions containing His6-tagged LrTyDCs were pooled, desalted, and concentrated by ultrafiltration using an Amicon Ultra 30 K MWCO centrifugal filter (Merck Millipore Ltd., USA) with dialysis buffer (Na2HPO4, 100 mM NaCl, 10% glycerol, pH 7.4, 4 °C) and stored at − 80 °C. The concentrations of LrTyDCs were estimated by using the ε280nm 27,750 M−1 cm−1 (LrTyDC1) and 24,000 M−1 cm−1 (LrTyDC2) calculated from ExPASy ProtParam.

Enzymatic activity assay and characterization of LrTyDCs

The general enzymatic activity assays of His6-tagged LrTyDCs were performed in 100 μL of Tris–HCl buffer (50 mM Tris, pH 8.5) containing 1 mM of PLP (purity ≥ 99%, Aladdin, China), 2 mM of l-Tyr (purity ≥ 99%, Aladdin, China), and 10 μM of enzyme. The enzyme and PLP were mixed in the buffer and incubated at 50 °C for 3 min prior to assay initiation by the addition of substrate. The reaction mixture was incubated at 50 °C with shaking at 250 rpm. An aliquot (20 μL) of the reaction mixture was sampled at 1, 3, and 48 h and quenched by adding 20 μL of chill methanol. The formation of tyramine from the enzymatic reactions was determined by HPLC–DAD equipped with an Altima C18 analytic column (250 mm × 4.6 mm, 5 μm). The mobile phase consisting of solvent A (MeOH) and solvent B (H2O containing 0.3% trifluoroacetic acid) followed a gradient elution program (0 min, 1% solvent A; 25 min, 20% solvent A; 30 min, 99% solvent A; 31 min, 1% solvent A) at a flow rate of 1 mL min−1 at 35 °C, monitored at 276 nm by a diode array detector.

l-Tryptophan (Trp), l-Phe, or 3,4-dihydroxy-l-phenylalanine (DOPA) (purity ≥ 99%, Aladdin, China) was introduced to the enzymatic activity assay system to investigate the substrate specificity of His6-tagged LrTyDCs.

To quantitate the enzymatic tyramine from LrTyDC-catalyzed decarboxylation reactions, the reaction mixtures were directly added 1 mL of NaHCO3-NaOH buffer (pH 10.0) and then extracted twice with ethyl acetate (2 mL × 2). After centrifugation at 10,000 rpm for 2 min, the upper layer was evaporated to dryness. The samples were re-suspended into 200 μL of NaHCO3-NaOH buffer (pH 9.5). To the aforementioned solution the derivatization reagent dansyl chloride (DNS-Cl, dissolved in acetone, 1 eq.) was added and the resulting mixtures were incubated at 60 °C for 20 min. And then the mixtures were cooled to room temperature and the reaction was quenched by addition of 100 μL of NH3·H2O. The resulting mixtures were evaporated to remove acetone. The remaining mixtures were extracted twice with ether (2 mL × 2). The organic layer was evaporated to dryness after centrifugation at 10,000 rpm for 2 min. Finally, the residue was dissolved in 200 μL of methanol and the supernatant was analyzed by HPLC–DAD followed the gradient elution of solvent A (MeOH) and solvent B (H2O) at a flow rate of 1 mL min−1 and 35 °C. The gradient elution program was as follows: 0 min, 50% solvent A; 15 min, 90% solvent A; 20 min, 100% solvent A; 26 min, 50% solvent A; monitored at 264 nm by a diode array detector.

To quantitate the formation of corresponding enzymatic products by HPLC–DAD analyses, the calibration curves of standard DNS-tyramine were established. Briefly, the same derivatization procedure was applied to prepare DNS-tyramine standard. A linear fit for the relationship of specific concentration of DNS-tyramine versus the corresponding peak area integral at 264 nm was used to plot the calibration curve.

Optimization of LrTyDC-catalyzed decarboxylation

To determine the optimal reaction temperature of LrTyDC-catalyzed decarboxylation, the reaction mixtures were incubated at 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, and 70 °C. Different buffer systems were prepared as reported previously (Li et al. 2019) to determine the effects of pH value of the reaction buffer on LrTyDCs catalytic activities. The metal ions Na+, Fe3+, Fe2+, Mg2+, Zn2+, Mn2+, Ca2+, Cu2+, Li+, and Co2+ were added to the reaction solution, respectively, at a final concentration of 1 mM in the activity assay system to investigate their effects on LrTyDCs activity. All the above assays were performed in triplicate and the reaction mixtures were incubated at 50 °C, 250 rpm for 3 h.

To determine the kinetic parameters of LrTyDCs toward l-tyrosine, the reactions were performed in 100 μL of Tris–HCl buffer (50 mM Tris, pH 8.5) containing 10 μM of LrTyDCs, 1.0 mM of PLP, and l-Tyr with different concentrations (0.05–2.0 mM), respectively. All the above assays were performed in triplicate and the reaction mixtures were incubated at 50 °C, 250 rpm for 1 h. The reaction mixtures were derived with DNS-Cl as described above and the formation of DNS-tyramine was analyzed and quantitated by HPLC–DAD analyses. The steady-state kinetic constants were calculated with nonlinear regression analysis using Origin 8.0 software.

Characterization of enzymatic reaction product

To characterize the enzymatic products, a UPLC-HRESIMS was performed on a Waters Acquity I Class system equipped with an Altima C18 analytic column (250 mm × 4.6 mm, 5 μm) and a Vion IMS QToF mass spectrometer. The mobile phase consisting of solvent A (MeOH) and solvent B (H2O containing 0.1% formic acid) followed an isocratic 1% solvent A elution program for 15 min. The flow rate is 1.0 mL min−1, monitored at 276 nm by a diode array detector. The injection volume of the sample was 10 μL. The electrospray ionization mass spectrometer was operated in positive ion mode and the spectra were collected in the enhanced full mass scan mode from m/z 50–1500.

Quantitative real-time PCR analysis of LrTyDCs

To quantitate the relative expression levels of LrTyDCs in different tissues of L. radiata, quantitative real-time PCR (qRT-PCR) experiments were conducted by using a CFX Connect™ real-time PCR System (Bio-Rad, USA), according to the manufacturer's instructions and previously reported procedures (Li et al. 2019). Briefly, two pairs of specific primers, rtLrTyDC1-F and -R, rtLrTyDC2-F and -R (Table S1, Supplementary Materials), were designed on the basis of the full-length cDNA sequences for real-time PCR amplification of LrTyDCs, according to the instructions of real-time PCR system. The relative transcriptional levels of LrTyDCs in different tissues were calculated by using the corresponding transcript levels of the reference gene encoding actin in the same tissue as control. The relationship between relative transcriptional levels of LrTyDCs and accumulation of GAL was analyzed to further determine whether LrTyDCs are involved in the Amaryllidaceae alkaloids biosynthesis in L. radiata (Li et al. 2019).

Results and discussion

Molecular cloning and bioinformatics properties of LrTyDCs

TyDCs had been cloned from plants such as Oryza sativa (Park et al. 2012), Rhodiola sachalinensis (Zhang et al. 2011), Arabidopsis thaliana (Lehmann and Pollmann 2009), Rhodiola rosea (György et al. 2009), Aristolochia contorta (Yang and Zeng 2005), and Papaver somniferum (Facchini and De Luca 1994). In view of the role of NpTyDC1 in the biosynthesis of the Amaryllidaceae alkaloids (Kilgore and Kutchan 2016), the nucleotide sequence of NpTyDC1 (GenBank accession number: KT378599) was used as a query to mine the public available transcriptome datasets of L. radiata from NCBI SRA. A candidate LrTyDC fragment was assembled on the basis of the multiple nucleotide sequences alignment results (Fig. S1, Supplementary Materials). Based on a homology cloning strategy, a 171-bp gene fragment of LrTyDC was obtained from a PCR amplification reaction. Subsequent 3′- and 5′-RACE experiments afforded the 3′- and 5′-end fragments, respectively. The full-length cDNA sequence of LrTyDCs was generated by overlapping the aforementioned DNA fragments, which was confirmed by the following full-length PCR amplification reactions and subsequent DNA sequencing. Two DNA sequences were obtained. One is 1536-bp which encodes a protein with 511 amino acid residues and it was designated as LrTyDC1 (GenBank accession number: MT468182). LrTyDC2 (GenBank accession number: MT468183) was shown to have 510 amino acid residues that is encoded by a 1533-bp ORF.

It should be noted that a TyDC from L. radiata was deposited in NCBI on October 22, 2019 with GenBank accession number: MK690193. Multiple amino acid residue sequence alignment analyses showed that a serine highlighted by the blue arrow is present in all functionally characterized plant TyDCs (Fig. 2). Previous studies revealed that mutation of Ser372 of PsTyDC into Gly enables the mutant PsTyDC to use 5-hydroxytryptophan as a substrate and reduces the enzyme activity toward l-DOPA (Torrens-Spence et al. 2014). However, the reverse mutation in a Catharanthus roseus tryptophan decarboxylase (CrTDC) enabled the mutant enzyme to utilize l-Tyr and l-DOPA substrates with a reduced affinity toward l-Trp (Torrens-Spence et al. 2014). Both TyDC and TDC are plant aromatic amino acid decarboxylase but they exhibit different substrate specificity (Facchini and De Luca 1995). It is a Gly in the corresponding site in the partial CDS of the previously deposited LrTyDC, we propose that it might be a TDC rather than a TyDC in L. radiata.

Fig. 2
figure 2

Multiple amino acid residue sequences alignment of LrTyDCs and the functionally characterized plant TyDCs by using online Clustal Omega multiple alignment tool. The results were graphed by using ESPript 3.0 software. The crystal structure of PsTyDC2 (PDB ID: 6LIU, P54769.1) was used as model. AmTyDC, ACJ76782.1; LaTyDC1, AYH64864.1; PcTyDC, AAA33861.1; RsTyDC1, ABF06560.1. The highly conserved amino acid residues were highlighted in red and the secondary highly conserved amino acid residues were highlighted in yellow. The catalysis lysine (K) was indicated by a vertical red arrow and the key serine (S) was indicated by a vertical blue arrow

Full size image

By using the ExPASy online tools, the theoretical isoelectric point and the molecular weight of LrTyDC1 were predicted to be 4.93 and 62.84 kDa, respectively, while for LrTyDC2, they are 4.93 and 62.93 kDa. LrTyDCs shares 47.9–97.3% amino acid residue identities and similarities with the functionally characterized plant TyDCs (Fig. 2 and Fig. S2, Supplementary Materials). LrTyDCs were shown to contain the following characteristic features of plant TyDCs (Fig. 2) (Wang et al. 2020). The highly conserved catalytic amino acid K321 in LrTyDC1 (K319 in LrTyDC2) covalently bounds to PLP to form an internal aldimine. The key active sites (D289, N318, and S372 in LrTyDC1; and D287, N316, and S370 in LrTyDC2) coordinate with PLP via hydrogen bonds. The highly conserved amino acid residues F126, H205, and T264 in LrTyDC1 (F124, H203, and T263 in LrTyDC2) involve in binding and stabilizing cofactor PLP. The conserved loop (WQIALSRRFR) involves in interactions with phosphate group of PLP to anchor PLP. The phylogenetic tree analysis revealed that LrTyDC1 and LrTyDC2 are phylogenetically close to the functionally characterized NpTyDC1 (Singh and Desgagné-Penix 2017; Kilgore and Kutchan 2016) and LaTyDC1 (Wang et al. 2019). The afore-mentioned information implied that LrTyDCs might function in GAL’s biosynthetic pathway (Fig. 3). The 3-dimensional structure of LrTyDCs (Fig. 4) was predicted by using the SWISS-MODEL online software (http://www.swissmodel.expasy.org) in which the crystal structure of PsTyDC2 (PDB ID: 6LIU, Wang et al. 2020) from Papaver somniferum that shares 73.56% and 74.21% amino acid residue identities with LrTyDC1 and LrTyDC2, respectively, was identified to be a template. Predicted models showed that LrTyDC1 and LrTyDC2 are homodimeric structures (Fig. 4a and c) and each monomer consists of catalytic amino acid residues and conserved domains (Fig. 4b and d).

Fig. 3
figure 3

Phylogenetic tree of plant TyDCs. The amino acid residue sequences were retrieved from the National Center for Biotechnology Information database (http://www.ncbi.nim.nih.gov/). Phylogenetic analysis was performed by using the MEGA 7.0 software package and neighbor-joining program (http://www.megasoftware.net). The computation was performed with 1000 bootstrap replicate and a Poisson correction. The bootstrap values higher than 50% are given at the nodes. The scale bar indicates the phylogenetic distance calculated from the number of differences. LaTyDC1 from Lycoris aurea; NpTyDC1 from Narcissus pseudonarcissus; NaTyDC Narcissus aff. pseudonarcissus MK-2014; NpaTyDC1 from Narcissus papyraceus; EgTyDC1 from Erythranthe guttata; NnTyDC2 from Nelumbo nucifera; DcTyDC2 from Dendrobium catenatum; QsTyDC1 from Quercus suber; AcTyDC from Aristolochia contorta; TfTyDC from Thalictrum flavum subsp. Glaucum; OsTyDC1 from Oryza sativa Japonica Group; RsTyDC1 from Rhodiola sachalinensis; AcTyDC2 from Ananas comosus; AsTyDC2 from Apostasia shenzhenica; RcTyDC1 from Rosa chinensis; JrTyDC1 from Juglans regia; PbTyDC1 from Pyrusx bretschneideri; PcTyDC from Petroselinum crispum; PaTyDC from Prunus avium; PsTyDC2 and PsTyDC7 from Papaver somniferum; AmTyDC from Argemone mexicana; AtTyDC from Arabidopsis thaliana; TcTyDC from Theobroma cacao

Full size image
Fig. 4
figure 4

The predicted 3-D structure of LrTyDCs. Ribbon diagram of quaternary structure of LrTyDC1 (a) and LrTyDC2 (c) homodimer. Schematic key active regions of the monomer of LrTyDC1 (b) and LrTyDC2 (d). The crystal structure of Papaver somniferum TyDC2 (PDB ID: 6LIU) was used as model template to construct the 3-dimensional structure of LrTyDCs. The PLP-binding, -stabilizing, -anchoring, and –coordinating regions were indicated. The catalytic lysine was illustrated. The graphic was generated by PyMOL molecular visualization system

Full size image

Heterologous expression and catalytic function of LrTyDCs

The two LrTyDCs were heterologously overexpressed in E. coli transetta (DE3) cells via pET28a expression vector. The homogeneous N-terminal His6-tagged recombinant LrTyDCs were obtained via affinity column chromatography (Fig. S3, Supplementary Materials). The LrTyDC-catalyzed decarboxylation activity assays were carried out by incubation of l-Tyr, PLP, and LrTyDCs or boiled LrTyDCs. A new peak was observed in LrTyDC-catalyzed reaction mixture that was analyzed by HPLC–DAD (panels II and IV, Fig. 5a), compared with the control reactions in which the boiled LrTyDCs were used as catalysts (panels III and V, Fig. 5a). The aforementioned peaks have identical retention time to that of the standard tyramine (panel I, Fig. 5a). The two products showed identical protonated ion at m/z 138.05496 [M + H]+ in the HRESIMS analyses (Fig. S4, Supplementary Materials), which is in perfect conformity with that of authentic tyramine. Thus both LrTyDCs exhibited decarboxylation activity and LrTyDC1 showed higher activity than LrTyDC2 (panels II and IV, Fig. 5a).

Fig. 5
figure 5

Catalytic function characterization of LrTyDCs. a HPLC–DAD analyses of standard tyramine and the enzymatic reaction mixture of LrTyDCs. b HPLC–DAD analyses of the DNS derivatives of standard tyramine and the enzymatic reaction mixture of LrTyDCs. Panels I, the authentic tyramine and its DNS derivative; Panels II, LrTyDC1-catalyzed reaction mixture and its derivatives; Panels III, boiled LrTyDC1 was used as catalyst; Panels IV, LrTyDC2-catalyzed reaction mixture and its derivatives; Panels V, boiled LrTyDC2 was used as catalyst. The vertical lines indicate that these peaks have identical retention time

Full size image

However, the decarboxylated product of the LrTyDC1-catalyzed decarboxylation reaction (48 h) showed ~ 10 mAU absorption at 276 nm, the characteristic maximum wavelength in the HPLC–DAD condition used in the present work (panel II, Fig. 5a). To enhance the sensitivity of the HPLC–DAD analyses and quantitate the products, the enzymatic products and standard tyramine were separately derived with DNS-Cl (Wu et al. 2018; Zhang et al. 2018). DNS-tyramine showed a sharp peak (panel I, Fig. 5b) in the HPLC–DAD condition used in the present work. DNS-tyramine with ~ 30 mAU absorption at the characteristic maximum wavelength of 264 nm was observed when the LrTyDC1-catalyzed decarboxylation reaction was performed for 3 h (panel II, Fig. 5b).

The optimal enzymatic reaction conditions for LrTyDC-catalzyed decarboxylation are 50 °C (Fig. S5A and 5D, Supplementary Materials) and pH 8.5 (Fig. S5B and 5E, Supplementary Materials) under which LrTyDCs exhibit the maximum catalytic activity. The decarboxylation activity of LrTyDCs was displayed with a slight change when the enzymatic reaction was performed with the addition of different metal ions (Fig. S5C and 5F, Supplementary Materials). The enzymatic activity of LrTyDCs was inhibited by the addition of Fe2+, Fe3+, and Li+ (Fig. S5C and 5F, Supplementary Materials). The metal ions Na+, Zn2+, Mg2+, Mn2+, Ca2+, and Co2+ were shown to slightly increase the decarboxylation activity of LrTyDCs (Fig. S5C and 5F, Supplementary Materials). The apparent kinetic parameters Km and Vmax for LrTyDC1-catalyzed decarboxylation reaction are 0.39 ± 0.14 mM and 3.15 ± 0.84 μM min−1, respectively (Fig. 6). LrTyDC1 showed similar Km and Vmax to LaTyDC1 (Wang et al. 2019), which is consistent with the phylogenic tree analyses (Fig. 3). LrTyDC2 showed higher affinity and lower reaction rate and the Km and Vmax values for LrTyDC2-catalyzed decarboxylation reaction are 0.06 ± 0.007 mM and 2.64 ± 0.04 μM min−1, respectively (Fig. 6).

Fig. 6
figure 6

The steady-state kinetic parameters for LrTyDC1- (a) and LrTyDC2- (b) catalyzed decarboxylation of l-tyrosine. The apparent kinetic parameters Km and Vmax for LrTyDC1-catalyzed decarboxylation reaction were 0.39 ± 0.14 mM and 3.15 ± 0.84 μM min−1. The Km and Vmax values for LrTyDC2-catalyzed decarboxylation reaction were 0.06 ± 0.007 mM and 2.64 ± 0.04 μM min−1, respectively. Values represent the means of three replicates ± SD

Full size image

Animal TyDCs can recognize and catalyze the decarboxylation of L-Tyr and its analogues, including l-DOPA, l-Trp, l-Phe, and l-His, to generate the corresponding decarboxylated products (Christenson et al. 1972). Nevertheless, plant TyDCs usually show high substrate specificity (Marques and Brodelius 1988; Noé et al. 1984; Christenson et al. 1972). L-Tyr was the native substrate of the TyDC from Rhodiola sachalinensis involved in the biosynthesis of salidroside (Zhang et al. 2011). l-DOPA and l-Phe could be decarboxylated by Rhodiola sachalinensis TyDC but l-Trp could not be accepted by the aforementioned TyDC (Zhang et al. 2011). Arabidopsis thaliana TyDC recognized l-Tyr only (Lehmann and Pollmann 2009). Papaver somniferum TyDC5 showed decarboxylation activity toward l-Tyr, l-DOPA, and l-Phe (Maldonado-Mendoza et al. 1996). No activity was observed when l-Trp was used as substrate of PsTyDC5 (Maldonado-Mendoza et al. 1996). PsTyDC1 and PsTyDC2 could accept l-Tyr and l-DOPA (Facchini and De Luca 1994). l-Trp and l-Phe were not recognized as substrate of PsTyDC1 and PsTyDC2 (Facchini and De Luca 1994). TyDCs from Parsley (Petroselinum crispum) (Kawalleck et al. 1993) and Eschscholtzia californica (Marques and Brodelius 1988) could accept l-Tyr and occasionally l-DOPA, but not l-Trp or l-Phe. LaTyDC1, a TyDC from L. aurea, catalyzed the decarboxylation of l-Tyr and l-DOPA (Wang et al. 2019). For l-DOPA, LaTyDC1 exhibited a slight high Km and a low Vmax than those for l-Tyr (Wang et al. 2019). In this study, l-Tyr, l-Trp, l-Phe, and l-DOPA were used as potential substrates for LrTyDC-catalyzed decarboxylation reactions. The results showed that LrTyDCs catalyze decarboxylation of l-Tyr only (Fig. 5).

Transcriptional levels of LrTyDCs and GAL contents

Plant TyDCs have tissue expression specificity and associate with the accumulation of secondary metabolites in which l-Tyr is a precursor. For instance, PsTyDC1 was highly expressed in poppy roots, while PsTyDC2 was highly expressed in poppy roots and stems. Both of them were weakly expressed in leaves (Facchini and De Luca 1994). Herein both LrTyDCs were found to be widely expressed in L. radiata on the basis of qRT-PCR analyses (Fig. 7). The relative transcriptional levels of LrTyDCs in the roots are lower than those in other plant tissues (Fig. 7). Thus, all relative transcriptional levels were correlated with the lowest transcription level of LrTyDC1 in roots (Fig. 7). LrTyDC1 was shown to exhibit the highest transcript in bulbs, while it was the flowers in which LrTyDC2 transcript was highest. LrTyDC2 exhibited a slight lower transcriptional level in bulbs than that in flowers (Fig. 7). The relative transcripts of LrTyDCs in each plant tissue are directly proportionally correlated with the accumulation amount of GAL in the corresponding tissue of L. radiata according to the results reported previously (Fig. 7) (Li et al. 2019). Therefore, it was further confirmed that both LrTyDCs are involved in the decarboxylation of l-Tyr, the fist biosynthetic step of GAL in L. radiata.

Fig. 7
figure 7

The relationship of the transcript levels of LrTyDCs with the accumulation amount of GAL in L. radiata. The relative transcript levels of LrTyDC1 (black column filled with black) and LrTyDC2 (black column filled with red) in stems, leaves, flowers, and bulbs were normalized to that of LrTyDC1 in roots. The contents of GAL were in blue column. Values represent the means of three biological replicates ± SD

Full size image

As an important biogenic amine, tyramine participates in a variety of physiological processes. For instance, tyramine is the essential component of the xylem cell wall (Borgolivier et al. 1987). It is the decisive factor in the overall defense system of plants such as parsley (Kawalleck et al. 1993) and Arabidopsis species (Trezzini et al. 1993). When the cells were wounded, free and bound tyramine in the cell wall protected cells against viral infections (Negrel et al. 1993). Tyramine is a precursor of tyramine-derived natural products such as benzylisoquinoline alkaloids (Facchini et al. 2000, 2002) and the Amaryllidaceae alkaloids (Jin and Yao 2019; Jin 2009, 2013, 2016). The decarboxylation derivation of l-Tyr to generate tyramine is catalyzed by TyDCs, a group of enzymes between the primary and the secondary metabolisms. TyDCs are the pivotal and rate-limiting enzymes that connect primary metabolism and secondary metabolism, which determines the carbon flux of metabolic pathways. Plant TyDCs, a group II PLP-dependent decarboxylases, catalyze the decarboxylation of l-Tyr to generate tyramine, which is biologically important for diverting essential primary metabolites into secondary metabolic pathways (Wang et al. 2020; Sandmeier et al. 1994). TyDCs are the committed enzymes and play an important role in regulating the first step of biosynthesis of secondary metabolites in plants (Guillet and De Luca 2005; Facchini and De Luca 1994). TyDCs are present in most, if not all, biological species (Kawalleck et al. 1993; Zahnow et al. 1991; Eveleth et al. 1986; Morgan et al. 1986; Albert et al. 1987). In this work, two TyDCs were obtained from L. radiata, a GAL-producing plant, via transcriptome data analysis and molecular cloning. Heterologous expression of the aforementioned LrTyDCs in E. coli and following purification enable the in vitro biochemical characterization of recombinant LrTyDCs. Only l-Tyr was recognized by LrTyDCs to be decarboxylated to generate tyramine (Fig. 5). The transcript levels of LrTyDCs are correlated with the GAL contents in each specific tissue of L. radiata (Fig. 7), which implied their involvement in the biosynthesis of the Amaryllidaceae alkaloids. The reported works here will be helpful for biosynthetic investigations on the Amaryllidaceae alkaloids and for biological investigations on physiological processes.

Author contribution statement

YL conceived and designed the experiments. JH, WL, and ZL performed the experiments. YL, JH, WL, and GZ analyzed the data and wrote the manuscript. All authors have read and approved the manuscript in its final form.