1. Introduction
Apocynum, commonly known as kendir or Luobuma since pre-historic times due to its excellent quality fiber, is a genus in the Apocynaceae family used as a medicine and source of fiber. They are considered native to a large number of countries in northern and south Eastern Europe, northern America, and Asia. Although there is much divergence in their taxonomy, there exist mainly two species, Apocynum venetum and Apocynum hendersonii (also called Apocynum pictum) in China distributed widely in the northwest and the coast of the Yellow Sea [1,2]. Both species are eco-economic perennial rhizomatous plants with high stress resistance and economic value, with rhizomes extending up to 5–6 m horizontally underground, and adapted to survive in extreme environments of low rainfall, saline and alkaline deserts, river banks, alluvial plains, and the Gobi desert [3].
A. venetum, in addition to serving as source of bast fiber, is very well known for possessing broad medicinal uses as anti-platelets [4], anti-depressive [5], anti-hypertensive [6], anti-oxidant [7], anti-aging [8], anxiolytic activity [1], stimulant as tea [9] etc. Meanwhile, A. hindersonii is morphologically similar to A. venetum but with a major visible distinction being in the color of stem and leaf, and the shape of their flower, owing likely to their local names “White Hemp” and “Red Hemp”, respectively.
Applications of A. hendersonii are usually limited to fibers for spinning and paper making, rather than medicinal purpose [8], due to the relative low total flavonoid content. Chan et al. tentatively identified ten kinds of flavonoid components by ultra-visible and mass spectra, and hyperoside was identified as a critical parameter or a suitable chemical marker for discrimination of two species [7]. However, as A. venetum and A. hendersonii are distributed widely from the northwest to the coast of the Yellow Sea in China, the tremendous differences of the natural conditions may lead to the great differences in phenotype, structure, and chemical compositions. The whole biological effects of both plants are borne out by their bast fiber and flavonoids [10].
Recently, researches on the two species have attracted widespread attention. Previous researches were mainly focused on the identification of morphological characteristics or extraction of phytochemicals, and combinative technologies of bulked segregate analysis (BSA) coupled with randomly amplified polymorphic DNA (RAPD) were developed to distinguish the two species [2]. Yet very few studies have assessed the genetics of A. venetum and A. hendersonii, and inadequacy of gene sequence resources have limited their in-depth researches. Flavonoids play important roles in a wide range of physiological processes. In China and Japan, the leaves of A. Venetum and A. hendersonii are used as traditional herbal teas due to their higher flavonoid content than Camellia sinensis [9]. Flavonoids are synthesized through a long pathway, which has largely been characterized in Arabidopsis thaliana and Zea mays. Briefly, the upstream pathway consists in the formation of the core, the basic skeleton of all flavonoids (starting from malonyl-CoA and 4-coumaroyl-CoA to naringenin). The oxidation of naringenin yields the dihydrokaempferol that can subsequently be hydroxylated, producing, respectively, dihydroquercetin or dihydromyricetin [11,12,13]. Researchers have tended to focus on specific compounds, rather than thoroughly assessing all of the constituents. Previous studies found that the content of flavonoid in A. hendersonii was lower than in A. Venetum. So, identifying and investigating the regulatory mechanisms behind flavonoid biosynthesis is essential to understanding their roles in Apocynum to improve flavonoid production by metabolic engineering, and a comparative analysis of their chemical composition or pharmacological activities is certainly needed.
In order to analyze crucial genes in the synthesis of flavonoids and fully mine molecular biology data from A. venetum and A. hendersonii, we present here the metabolome profile and transcriptome data to analyze metabolite pathways associated with flavonoid biosynthesis. In addition, we conducted a genomic survey of A. venetum and A. hendersonii. The availability of genomic information may provide an opportunity to elucidate genes involved in the biosynthesis of flavonoids, and also have significance for genomics and evolutionary research of Apocynum genus plants.
2. Results and Discussion
2.1. Genome Survey of A. venetum and A. hendersonii
A total of 116.45 Gb for A. venetum and 120.61 Gb for A. hendersonii raw base were obtained by the Illumina Hiseq platform (Table 1). After filtering low-quality data, the clean read rates of 98.12% and 97.34% raw base were obtained for A. venetum and A. hindersonii, respectively. The sequencing quality evaluation result also shows raw Q30 base rate (92.16%) and clean Q30 base rate (92.75%) for A. venetum. Similarly, raw Q30 base rate (91.92%) and clean Q30 base rate (92.21%) were recorded for A. hendersonii, indicating high quality and accuracy of the high-throughput sequencing.
The clean bases were used for preliminary genome assembly, and K-mer value of 35 was selected for constructing contig and scaffold (Table 2). In A. venetum, we obtained a total of 1250,389 contigs with a total length of 322,394,863 bp and the longest sequence assembled of 6632 bp. The length of the N50 Contig was 310 bp. The number of obtained scaffolds was 876,453, having a total length of 377,590,884 bp, and thr longest sequence assembled of 70,393 bp and N50 scaffold of 1225 bp. A total of 741,647 contigs with a total length of 199,859,654 bp was obtained for A. hendersonii, with the longest sequence assembled of 8521 bp and 457-bp-long N50 Contig. A total of 303,264 scaffolds were obtained after further assembly with a total length of 236,656,356 bp, and the longest sequence assembled of 110,436 bp. The length of the N50 scaffold was 4667 bp.
Following exclusion of error effects due to erroneous K-mers (Table 3), the genome sizes were found to be 232.80 Mbp and 233.74 Mbp for A. venetum and A. hendersonii, respectively. A. venetum and A. hendersonii genome sizes are, therefore, relatively small in comparison to published genome sizes of other plants [14,15], suggesting future genome assembly and comparative studies on genome evolution to also be relatively simple. Heterozygosity is an important feature of plant genome, and many crop species are manipulated to increase heterozygosity for heterosis breeding [16,17]. Calculating the proportion of heterozygous sites in the sequences of both species revealed a gene heterozygosity ratio of 0.67% and 0.46% for A. venetum and A. hendersonii, respectively, indicating more heterozygous complexity in A. venetum.
2.2. Transcriptome Sequencing, De Novo Assembly, and Pathway Analysis
RNA libraries were derived from A. venetum and A. hendersonii leaves and were sequenced using Illumina HiSeq2500. A total of 47.9 million clean reads with 44.2% GC content (guanine plus cytosine content) for A. venetum were acquired, and FastQC analysis show 95.3% of the total sequences to possess a quality above Q30. RNA-seq libraries for A. hendersonii yielded 48.1 million clean reads with 44.1% GC content and FastQC analysis show 95.4% of the total sequences with quality above Q30.
Pooled reads of both species were assembled de novo into a total of 32,740 unigenes with N50 of 2081 bp. All of the unigenes were annotated by aligning with public databases and obtained 21,743 (66.41%) significant hits in Nr; 17,328 (52.93%) in Swiss-Prot; 8211 (25.08%) in KEGG (Kyoto encyclopedia of genes and genomes); 12,995 (36.69%) in KOG (Clusters of orthologous groups for eukaryotic complete genomes); 20,918 (63.89%) in eggnog; 15,627 (47.73%) in GO (Gene Ontology); and 37 (0.11%) in Pfam databases. The best-match result of species distribution (Figure 1) gave top hits of 51.09% of unigenes to be from Coffea canephora, followed by Sesamum indicum (8.09%), Vitis vinifera (3.58%), Nicotiana tomentosiformis (3.54%), and Nicotiana sylvestris (3.43%). The 15,627 unigenes reflected in GO were assigned to one or more terms and categorized into 60 functional groups under three main partitions, namely, biological process, molecular function, and cellular component. Of the unigenes, 10,461 (66.94%) in the biological process partition fall under highly encoded proteins that are involved in the cellular process, followed by 9064 unigenes (58.00%) involved in the metabolic process, 4670 unigenes (29.88%) in biological regulation, and 4566 unigenes (29.21%) involved in response to stimulus.
A total of 14538 unigenes were also searched against the KOG database and clustered into 26 classes. The largest category, 2990 unigenes (20.56%), fall into general function prediction, followed by post-translational modification, protein turnover, chaperones (1469 unigenes, 10.10%), and signaltransduction mechanisms (1236 unigenes, 8.50%). Moreover, 577 unigenes were classified as secondary metabolite biosynthesis, transport, and catabolism. These results opined that both A. venetum and A. hendersonii undergo extensive metabolic activity. Additionally, pathway analysis of differentially expressed genes (A. venetum vs. A. hendersonii) identified flavonoid biosynthesis (ko00941) and flavone and flavonol biosynthesis (ko00944) as the top 20 up-regulated KEEG pathways [18], suggesting higher flavonoid content in A. venetum.
2.3. Metabolic Profiles of A. venetum and A. hendersonii
We performed a comparative metabolomics analysis between the two species based on the LC/MS (liquid chromatograph-mass spectrometer) data obtained from the individual species leaf. The results show that both of the A. venetum and A. hendersonii were rich in secondary metabolites, and more than 7500 kinds of components were detected and identified in the two species. Positive and negative ionization modes were used to evaluate the results, while the negative ionization mode being more selective and sensitive for presenting metabolomic profile with higher number of compounds and intensities [19] was chosen for further analysis. Upon comparison of observed m/z values with data of mass spectra of agrochemicals, pharmaceuticals, and physiologically active compounds (including drugs, steroid hormones, or endocrine disruptors) from mass spectral libraries NIST and WILEY (
The data obtained were initially submitted to a PCA (Principal Component Analysis) approach using the Pareto scale. The secondary metabolites were clearly separated and grouped into two clusters corresponding to A. venetum and A. hindersonii. The differential metabolites were selected on the basis of the combination of a statistically significant threshold of variable influence on projection (VIP) values obtained from the OPLS-DA (Orthogonal partial least squares discriminant analysis) model and p-values from a two-tailed Student’s t-test on the normalized peak areas (Figure S1). We observed that in both the A. venetum and A. hendersonii leaf, the general pattern of metabolites was relatively similar, while the accumulation content of 665 metabolites such as flavonoids and sterols was found to be significantly different. These compounds could be used to explore the potential chemical markers that contribute to the diversity between the two species. The hierarchical clustering and correlation analysis of the top 50 differentially accumulated metabolites between A. venetum and A. hendersonii are shown in Figure 3, which visually shows the intensities of the potential chemical biomarkers between A. venetum and A. hendersonii.
2.4. Major Flavonoids Identified in A. venetum and A. hendersonii
Flavonoids, which are widespread in plants and fruits, have lots of structural variations that can be subdivided into six major groups according to the hydroxylation pattern and conjugation between the aromatic rings; namely, flavones, flavanones, flavonols, flavanols, anthocyanins, and isoflavones. In our analyzed leaf extracts, two flavones (luteolin, apigenin), six flavonols (rutin, hyperin, isoquercitrin, quercetin, kaempferol, tamarixetin), three flavanols (catechin, epicatechin, epigallocatechin), one flavanones (hesperidin), one isoflavanone (trifolin), seven anthocyanidins (cyanidin, procyanidin c1, procyanidin, delphinidin, pelargonidin, malvidin, peonidin), and two chalcones (carthamin, neocarthamin) were found in A. venetum and A. hendersonii leaves. A few other flavonoids (acacetin, chrysoeriol 7-O-glucoside, acacetin-7-O-rutinoside) not detected before were also identified. The chemical characterizations of the flavonoids are described in Table 4. Among these compounds, flavonols are the major flavonoids. This result is consistent with the previous data [20,21], while more anthocyanins were detected.
Quantitative analysis shows that flavonols are the main differentiated flavonoids between the A. venetum and A. hendersonii leaf samples. The content of hyperin, isoquercitrin, quercetin, kaempferol, and rutin in A. venetum was much larger than that in A. hendersonii—evidence of its higher flavonol content. This result conforms to the reports that hyperoside and isoquercetrin could serve as good chemical markers to differentiate the two species [22].
2.5. Analysis of Flavonoids Biosynthesis-Related Transcript Level and Metabolites Content
The flavonoid pathway consists of a number of enzymatic steps, each catalyzing a sequential reaction. Phenylpropanoid is a direct precursor for the synthesis of anthocyanins and other flavonoids. The basic skeleton of all flavonoids, starting from three molecules of malonyl-CoA and one of 4-coumaroyl-CoA, produces a colorless flavanone named naringenin. The related enzymes, including phenylalanine ammonialyase (PAL, Gene ID: DN7770, DN12095), cinnamic acid 4-hydroxylase (C4H, Gene ID: DN15816), chalcone isomerase (CHI, Gene ID: DN13226, DN3483, and DN8427), and chalcone synthase (CHS, Gene ID: DN18628), were identified in both A. venetum and A. hendersonii. Transcriptome analysis revealed there was no significant difference in the expression level of these genes, and the content of metabolites such as naringenin that were detected in the first few biosynthesis pathway of both A. venetum and A. hendersonii, which formatted the core and basic skeleton of flavonoids, also did not show significant difference.
The subsequent reactions in the flavonoid biosynthesis pathway are from naringenin to dihydroflavonol, anthocyanins, flavonols, or other flavonoids with the different enzymes. Among these key enzymes, two flavonol 3-hydroxylase (F3H)-encoding genes (ID: DN13255 and DN9170), three anthocyanidin reductase (ANR)-encoding genes (ID: DN11675, DN11945, and DN20760), five flavonoid 3-O-glucosyltransferase-encoding genes (ID: DN11850, DN14334, DN16440, DN19250, and DN6934) were identified, and their expression levels were significantly up-regulated in A. venetum rather than in A. hendersonii. Other differentially expressed genes, such as kaempferol 3-O-beta-D-galactosyltransferase (ID: DN10026) and leucoanthocyanidin reductase (ID: DN11675) encoding genes, were also identified with a log2 fold change value of 5.08 and 1.65 in A. venetum vs. A. hendersonii samples, respectively. The content of corresponding products (kacmpferol, quercetin, isoqucrcitrin, epicatechin, and epiafzelechin) in A. venetum were also higher than in A. hendersonii. Further confirmation by quantitative PCR (qPCR) analysis shows that all of the selected genes had the same expression pattern, revealed by the differential analysis results from high throughput sequencing (Table S1). The positive correlation between the transcript level and total flavonoid content (Figure 4) obtained in this study indicates that accumulation of flavonoids was accompanied by the induction of flavanone 3-hydroxylase, anthocyanidin reductase, and flavonol 3-O-glucosyltransferase.
The flavonoid biosynthesis pathway is one of the important pathways influencing Apocynum quality, with the biosynthesis starting from phenylalanine, then channeled to different branches of the pathway to yield many different compounds. Anthocyanins, one of the products of the pathway, are a class of water-soluble pigments providing certain color for various plant parts [13,22,23]. Transcription factors regulating anthocyanin synthesis via different mechanisms were previously identified in plants with dark- or light-colored phenotype of fruit or other organs, such as red-skinned and red-fleshed apple, purple and green asparagus or eggplants. The coloring of red-skinned apples is primarily determined by anthocyanin [24,25,26,27,28]. Purple and green asparagus cultivars were analyzed for anthocyanins biosynthetic and regulatory genes, and the purple asparagus was reported to have much higher anthocyanin than all other purple vegetables [29]. Our transcriptome and metabolome content correlation analyses show higher transcriptomic and metabolomic levels related to anthocyanins in A. venetum than in A. hendersonii, especially the anthocyanidin reductase genes which play roles in the early unbranched segment of the flavonoid biosynthetic pathway were found to be significantly up-regulated in A. venetum. We conclude that the complexity of the regulation mechanism of flavonoid biosynthesis is the embodiment of the diversity between A. venetum and A. hendersonii, and the red coloring of A. venetum (Figure 5) may be primarily determined by anthocyanins.
The HPLC (High performance liquid chromatography) method was adopted for quantitative analysis of hypersoside, isoquercitrin, quercetin, and rutin. The results show the amount of flavonoids compounds selected in the two species to have varied tremendously in accordance with the LC/MC profile (Table 5). The highest level of isoquercitrin (5.254 mg/g of DW, dry weight) was obtained in A. venetum, which was significantly higher than that obtained in A. hendersonii (1.718 mg/g of DW), suggesting that, in addition to hyperoside [7], isoquercitrin could also be a good chemical marker to differentiate A. venetum and A. hendersonii. The amount of total anthocyanins was determined using the previously described pH differential method. The average content of total anthocyanin in A. venetum was also much higher than in A. hendersonii. The amount of anthocyanin in A. venetum leaves (0.9017 mg/g) was comparable to the amount reported in Rosa rugose (1.1790 mg/g), Rosa chinensis (1.1010 mg/g), Hibiscus sabdariffa (0.8610 mg/g), Camellia japonica (0.8280 mg/g), and Dianthus caryophyllus (0.8180 mg/g). Despite the low anthocyanins in A. hindersonii, the amount obtained (0.4133 mg/g) was higher than in Myosotis sylvatica (0.2770 mg/g) [30]. Thus, both A. venetum and A. hendersonii are good choices as an anthocyanin supplement for consumers.
3. Materials and Methods
3.1. Plant Materials and Experimental Conditions
The asexual tissue cultured A. venetum and A. hendersonii were transfer to 15 L pots and grown in an air-conditioned greenhouse under long day condition (16 h, 22 °C day; 8 h, 18 °C night). In all experiments, additional fertilizer and/or pesticides were applied to the plants when necessary. Plants with similar canopy (diameter = 1 m) were randomly selected, and the 4–6 cm long leaves which were 20 cm away from the apex on every branch from the 2-year-old plants were collected for genomic survey, LC/MS, and RNA-sequence analyses.
For LC/MS analysis, leaves from each of the eight plants of the two species were separately collected as biological replicates. All of the samples were oven-dried to a constant weight at 60 °C. The dried leaves were ground to powder using a mixer mill with zirconia bead for 1.5 min at 30 Hz, and then passed through a 1.2-mm-size mesh. One hundred milligrams of the powdered samples was separately dissolved in 1.0 mL 70% methanol at 4 °C overnight for extraction. Then, after centrifuging at 10,000× g for 10 min, the extracts were passed through the solid-phase extraction cartridge (ANPEL, Shanghai, China) at 3 mL/min, filtrated through a 0.22 μm pore size membrane (ANPEL, Shanghai, China), and transferred to LC vials. The vials were stored at −80 °C before LC/MS analysis. QC samples (quality control samples) were prepared by mixing aliquots of the all samples to be a pooled sample
As for genomic and transcriptome analysis, total DNA and RNA were isolated from the same plants used for LC/MS analysis. The genomic DNA and total RNA were extracted from A. venetum and A. hendersonii leaf material using the Qiagen BioSprint and Qiagen RNeasy Plant Mini Kits (Qiagen, Hilden, Germany), according to the manufacturer’s protocol, respectively, and stored at −80 °C before use. The quantity and quality of genomic DNA were measured by an ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, CA, USA) and the RNA 6000 Nano Labchip Kit (Agilent Technologies, Santa Clara, USA) was used for total RNA measuring.
3.2. Genomic Survey, De Novo Transcriptome Assembly, and Analysis
Genomic paired-end libraries with 170 bp and 500 bp insertions were constructed according to the manufacturer’s standard protocol for Illumina Genome Analysis (Illumina Hiseq2000, San Diego, CA, USA). The SOAPdenovo2 software was used to carry out preliminary genome assembly [24]. The estimation of genome size, repetitive sequences, and heterozygosity were based on the information on peak depth, and the results of 25 mers Kmer analysis. The heterozygous rate estimation in the early stage of the genome project was according to the method modified and addressed by Liu et al. [31].
RNA-seq libraries were constructed using an RNA Library Prep Kit according to the manufacturer’s instruction (NEB, Ipswich, MA, USA), and the quality was measured using the Agilent Bioanalyzer 2100 system. The RNA-seq sequencing and assembly were performed by Biomarker Technology Co. (Biomarker, Beijing, China) with the Illumina HiSeq™ 2500 platform (BGI, Shenzhen, China). Prior to assembly, the raw data were filtered by trimming adaptor sequences and removing low quality sequences (Q < 20) with more than 10% uncertain (N) bases, the clean reads were then de novo assembled into unigenes according to the Trinity program [32,33].
3.3. Gene Function Annotation, Classification, and Digital Gene Expression Profile
Gene function was annotated according to the Nr, Nt, Pfam, Swiss-Prot, KO, and GO biological databases. For each sequence, top hits were extracted and used for further process. Differential expression analysis of two samples was performed using the DEGseq R package [34]. Based on the read count of each transcript, the reads per kilo base per million (RPKM) value was calculated and same taken to Gene Spring 12.5 GX to get fold changes (FC) through libraries with default parameter. Fold changes were calculated for category A. venetum vs A. hindersonii. q-value < 0.005 and |log2(foldchange)| ≥1 were set as thresholds for significant differential expression. Finally, sequences were filtered with RPKM ≥ 0.3, FDR > 0.001 and FC ≥ 2. The KOBAS software was used to test the statistical enrichment of the DEGs in KEGG [35].
Annotation references to flavonoid biosynthesis obtained from the database, such as gene descriptions, GO terms, and sequence descriptions, were manually classified according to the known keywords (referred by published articles and KEGG pathways) for the selection of candidate flavonoid biosynthesis-related genes. Gene expression was quantified using the Maxima SYBR Green qPCR Master Mix (Invitrogen Corp, Carlsbad, CA, USA) and the CFX-96 Thermal Cycler (Bio-Rad, Hercules, CA, USA) using primers appropriate for each gene.
3.4. LC/MS, Data Preprocessing, and Statistical Analysis
The ACQUITY UHPLC system (Waters Corpo MA, Milford, USA) and Triple TOF 5600 System (AB SCIEX, Framingham, MA, USA) were used to analyze metabolic profiling, and the 1.7 μm, 2.1 × 100 mm BEH C18 column was employed in both the positive and negative modes. The injection volume was 50 μL, and the binary gradient elution system consisted of water (0.1% formic acid, v/v) and acetonitrile (0.1% formic acid, v/v). The separation was achieved using the gradient (0 min, 5% acetonitrile; 2 min, 20% acetonitrile; 4 min, 25% acetonitrile; 9 min, 60% acetonitrile; 14 min, 100% acetonitrile; 18 min, 100% acetonitrile; 18.1 min, 5% acetonitrile and 19.5 min, 5% acetonitrile) with the flow rate 0.4 mL/min and column temperature 45 °C.
Data acquisition was performed in full scan mode (m/z ranges from 70 to 1000) combined with IDA mode. Parameters of mass spectrometry were: 550 °C (+) and 550 °C (−) ion source temperature; 5500 V (+) and 4500 V (−) ion spray voltage; 35 PSI curtain gas; 100 V (+) and −100 V (−) declustering potential; 10 eV (+) and −10 eV (−) collision energy; 550 °C (+) and 600 °C (−) interface heater temperature. The range of m/z was set as 25–1000, and the collision energy was set as 30 eV for IDA analysis. The QCs were injected at regular intervals (every 10 samples) throughout the analytical run to provide a set of data from which repeatability can be assessed.
The acquired LC/MS raw data were analyzed by the progqenesis QI software (Waters Corporation, Milford, MA, USA). The 0.3 mg/ml 2-Chloro-L-phenylalanine (CAS: 103616-89-3) in methanol solution was used as reference substance, and the data analysis was based on peak area normalization. The detection parameters of reference substance were deselected for peak retention time alignment with minimum intensity at 15% of base peak intensity, noise elimination level at 10.00, and isotopic peaks were excluded. The internal standard was used for data QC (RSD < 40). The metabolic alterations among different experimental groups were visualized by principle component analysis (PCA) and (orthogonal) partial least-squares-discriminant analysis (O)PLS-DA (Orthogonal-projection to latent structures discriminant analysis). After mean centering (Ctr) and Pareto variance (Par) scaling, the differential metabolites were selected on the basis of the combination of a statistically significant threshold of variable influence on projection (VIP) values. The VIP values were obtained from the OPLS-DA model and p-values from a two-tailed Student’s t-test on the normalized peak areas. The metabolites with VIP values larger than 1.0 and p-values less than 0.05 were considered as differential metabolites. The default seven-round cross-validation was applied with 1/seventh of the samples being excluded from the mathematical model in each round, in order to guard against overfitting.
3.5. Flavonoids and Total Anthocyanin Quantifiation
Quantitative analysis of flavonoids hyperoside, isoquercetrin, quercetrin, and rutin was carried out using HPLC–UV method [36]. Briefly, 0.2 g sample from A. venetum and A. hendersonii leaf was refluxed with 40 mL 70% methanol for 3 h, and then centrifuged at 3000× g for 5 min. After washing with 70% methanol and centrifuging at 3000× g for 5 min again, the extracts were marked to 50 mL volumetric flask. Subsequently, the prepared extract solution was filtered through a 0.45 μm PTFE membrane for HPLC–UV analysis. The analysis was carried out on an Agilent Zorbax column (250 × 4.6 mm, 5 μm) at 25 °C, and the Agilent 1100 liquid chromatography system equipped with a quaternary solvent delivery system.
Total anthocyanin contents (TAC) were extracted according to the pH differential method, with minor modifications [37,38,39,40]. Briefly, the main affecting factors, such as ethanol content in the solution (60–80%, v/v), liquid–solid ratio [40:1–60:1, w/v (g/mL)], leaching time (2–5 h), and temperature (30–60 °C), were taken into account. After centrifuging, the combined and filtered extraction solvent were measured at 520 and 700 nm in buffers at pH 1.0 and 4.5 using a UV spectrophotometer (DU730UV VIS, Beckman Coulter, Brea, CA, USA). Anthocyanins ((mg)/(l)) = (A × Mw × DF × 103)/(ε × L), where A is absorbance given by A = (A_520 nm–A_700 nm ) pH1.0 − (A_520 nm–A_700 nm ) pH 4.5; Mw, molecular weight of anthocyanins (433.2 g/mol); DF, dilution factor; ε, extinction coefficient (29,600 L·mol–1·cm–1); and L, path length (1 cm).
3.6. Transcript–Metabolite Correlation Analysis and qRT-PCR Validation of the Selected Key Genes
The entire raw transcriptome dataset and the entire metabolome dataset, formatted in log2 abundance (peak area), were uploaded to the Metabolights with repository:
Metabolites with q-values less than 0.1 were selected for transcript–metabolite correlation analysis. For each selected metabolite, their related genes were identified according to a Pearson correlation coefficient of metabolite–transcript >0.9. The identified genes or transcripts were then used to determine over-represented pathways through MetNetOnline (
Gene expression was conducted using cDNA from three biological replicates, with three technical replicates each. The expression of mRNA PAL (Gene ID: DN7770, DN12095), C4H (Gene ID: DN15816), CHI (Gene ID: DN13226, DN3483 and DN8427), CHS (Gene ID: DN18628), F3H (Gene ID: DN13255 and DN9170), ANR (Gene ID: DN11675, DN11945 and DN20760), and UGT (Gene ID: DN11850, DN14334, DN16440, DN19250 and DN6934) were quantified with RT-qPCR using the SYBR Green qPCR Master Mix (Invitrogen Corp) and the CFX-96 Thermal Cycler (Bio-Rad). The relative expression levels of these selected genes between A. venetum and A. hendersonii leaf were calculated using the comparative threshold cycle method [44], and the expression levels were normalized against housekeeping gene actin. The specific primers for these selected genes and actin gene are shown on Table S2.
4. Conclusions
In this study, the genome survey of two Apocynum plants, A. venetum and A. hendersonii, were conducted to provide a foundation for whole genome sequencing. Results showed both species to have small genome sizes of 232.80 Mb (A. venetum) and 233.74 Mb (A. hendersonii), suggesting future genome assembly and comparative studies on genome evolution to also be relatively simple. Transcriptome and metabolome co-analysis demonstrated that flavonols were the main differentiated flavonoids between the two species, and positive correlations of gene expression levels of flavonone-3 hydroxylase (F3H), anthocyanin reductase (ANR), flavonoid 3-O-glucosyltransferase (UGT), and total flavonoid content were observed. The content of isoquercitrin, hyperoside, and total anthocyanin in A. venetum was found to be much higher than in A. hendersonii, especially the difference of total anthocyanin content between the two species was thought to be the reason for the phenotype diversity in stem and leaf color of A. venetum (red) and A. hendersonii (white).
Supplementary Materials
The following are available online at
Author Contributions
Conceived this study, G.G., A.Z., and C.Y.; genomic survey, P.C., J.C., and K.C.; transcriptome and metabolome analysis, P.C., G.G., N.L., and A.S.A.; metabolite quantification, G.G., X.W., and J.C. All authors analyzed and discussed the data and contributed to writing the manuscript.
Funding
This work was supported by the Agricultural Science and Technology Innovation Project of the Chinese Academy of Agricultural Sciences (CAAS-ASTIP-2019), the China Agricultural Research System (CARS-16), and the Central Public-Interest Scientific Institution Basal Research Fund (Y2019XK15-03).
Conflicts of Interest
The authors declare no conflicts of interest.
Figures and Tables
Enlarge this image.Figure 1. Venngraph (A) of A. venetum and A. hendersonii unigenes annotated against seven databases and the species distribution on the top BLASTX hits (B). (KOG, Clusters of orthologous groups for eukaryotic complete genomes; KEGG, Kyoto encyclopedia of genes and genomes; GO, Gene ontology; NR, Non-redundant protein sequence databas; eggNGO, Evolutionary genealogy of genes: Non-supervised orthologous groups).
Enlarge this image.Figure 2. Typical base peak intensity (BPI) chromatograms showing metabolic profiles of A. venetum and A. hendersonii leaf in the negative ionization mode.
Enlarge this image.Figure 3. Heatmap (A) and correlation (B) among the top 50 differentially accumulated metabolites between A. venetum and A. hendersonii. The content of each metabolite was normalized to complete linkage hierarchical clustering. Each example in (A) is visualized in a single column and each metabolite is represented by a single row. Red indicates high abundance, whereas low relative metabolites are shown in green. The red and blue color in (B) indicate the positive and negative correlations, respectively.
Enlarge this image.Figure 4. Simplified representation of the enrichment of different expressed flavonoid biosynthetic pathway genes of A. venetum vs. A. hendersonii. (4CL, Coumarate Coenzyme A Ligase; PAL, Phenylalanine ammonialyase; 4CH, Cinnamic acid 4-hydroxylase; CHI, Chalcone isomerase; CHS, Chalcone synthase; F3H, Flavonol 3-hydroxylase; ANR, Anthocyanidin reductase; DFR, Dihydroflavonol 4-reductase; CYP75A, Subfamily of cytocrome P450; UGT, Flavonoid 3-O-glucosyltransferase; ANS, Anthocyanidin synthase).
Enlarge this image.Figure 5. Phenotypes and morphology of A. venetum and A. hendersonii leaf, stem, and flower. The stem and leaf of A. venetum show more of a dark red phenotype than A. hendersonii.
Statistics of high-throughput sequencing result.
| Species | Raw Base |
Clean Base |
Clean Read Number | Clean Read Rate (%) | Raw Q30 Base Rate | Clean Q30 Base Rate |
|---|---|---|---|---|---|---|
| A. venetum | 116,453,678,200 | 114,264,349,800 | 763,459,276 | 98.12 | 92.16 | 92.75 |
| A. hendersonii | 120,605,812,500 | 117,392,457,300 | 782,616,382 | 97.34 | 91.92 | 92.11 |
Contig and scaffold assembled sequences in A. venetum and A. hendersonii.
| Genome Characters | A. venetum | A. hendersonii | ||
|---|---|---|---|---|
| Contig | Scaffold | Contig | Scaffold | |
| Total length (bp) | 322,394,863 | 377,590,884 | 199,859,654 | 236,656,356 |
| Total number | 1250,389 | 876,453 | 741,647 | 303,264 |
| Max length (bp) | 6632 | 70,393 | 8521 | 110,436 |
| N50 length (bp) | 310 | 1225 | 457 | 4667 |
| N90 length (bp) | 132 | 150 | 114 | 217 |
Genome characteristics of A. venetum and A. hendersonii.
| Genome Characters | A. venetum | A. hendersonii |
|---|---|---|
| K-mer number | 93,121,402,015 | 93,494,221,128 |
| K-mer depth | 400 | 400 |
| Genome size (Mbp) | 232.80 | 233.74 |
| Revised genome size (Mbp) | 224.36 | 228.21 |
| Heterozygosity ratio (%) | 0.67 | 0.46 |
Flavonoids identified in the leaf extract of A. venetum and A. hendersonii.
| Metabolites | Retention Time (min) | m/z | Formula | A. venetum vs. A. hendersonii log2 (FC, Fold Change) |
|---|---|---|---|---|
| 1. Flavones | ||||
| Luteolin | 10.377 | 267.031 | C15H10O6 | - |
| Apigenin | 3.338 | 271.060 | C15H10O5 | - |
| 2. Flavonols | ||||
| Rutin | 3.430 | 591.138 | C27H30O16 | 1.788 |
| Hyperin | 3.313 | 465.103 | C21H20O12 | 3.452 |
| Isoquercitrin | 3.314 | 463.087 | C21H20O12 | 2.955 |
| Quercetin | 3.460 | 303.050 | C15H10O7 | 7.967 |
| Kaempferol | 4.066 | 285.039 | C15H10O6 | 4.030 |
| Tamarixetin | 3.550 | 363.07 | C16H14O7 | - |
| 3. Flavanones | ||||
| Hesperidin | 3.632 | 591.172 | C28H34O15 | −2.302 |
| 4. Isoflavanone | ||||
| Trifolin | 3.479 | 429.083 | C21H20O11 | - |
| 5. Flavanols | ||||
| Epigallocatechin | 2.463 | 307.081 | C15H14O7 | −0.144 |
| Catechin | 2.977 | 883.210 | C45H38O19 | −1.001 |
| Epicatechin | 3.763 | 271.062 | C15H14O6 | - |
| 6. Anthocyanidins | ||||
| Cyanidin | 3.216 | 287.055 | C43H42O22 | 1.952 |
| Procyanidin C1 | 3.097 | 867.215 | C15H10O6 | 0.959 |
| Procyanidin | 3.017 | 579.150 | C45H38O18 | - |
| Delphinidin | 3.313 | 303.050 | C15H10O7 | 0.118 |
| Pelargonidin | 11.853 | 611.049 | C15H11ClO5 | - |
| Malvidin | 11.268 | 347.033 | C17H15ClO7 | - |
| Peonidin | 10.826 | 381.038 | C16H13ClO6 | - |
| 7. Others | ||||
| Eriodictyol | 3.001 | 577.135 | C15H12O6 | −1.042 |
| Acacetin-7-O-rutinoside | 4.261 | 677.247 | C25H24O14 | 2.051 |
| Chrysoeriol 7-O-glucoside | 3.583 | 547.109 | C25H24O14 | 4.863 |
| Carthamin | 2.463 | 909.210 | C43H42O22 | - |
| Neocarthamin | 3.518 | 449.108 | C21H22O11 | −0.740 |
Quantitative analysis result of flavonoids and total anthocyanins in A. venetum and A. hendersonii leaves. Values are reported as mean ± standard (SD) of two independent experiments performed in triplicate.
| Species | Hypersoside |
Isoquercitrin |
Quercetin |
Rutin |
Total Anthocyanins |
|---|---|---|---|---|---|
| A. venetum | 3.896 ± 0.043 | 5.254 ± 0.057 | 0.543 ± 0.036 | 3.653 ± 0.012 | 0.9017 ± 0.002 |
| A. hendersonii | 1.262 ± 0.025 | 1.718 ± 0.029 | 0.384 ± 0.015 | 2.275 ± 0.015 | 0.4133 ± 0.002 |
© 2019 by the authors.
Details
; Chen, Ping; Chen, Jikang
; Chen, Kunmei; Wang, Xiaofei; Abubakar, Aminu Shehu; Liu, Ning