- Research article
- Open Access
Genetic variation and metabolic pathway intricacy govern the active compound content and quality of the Chinese medicinal plant Lonicera japonica thunb
BMC Genomicsvolume 13, Article number: 195 (2012)
Traditional Chinese medicine uses various herbs for the treatment of various diseases for thousands of years and it is now time to assess the characteristics and effectiveness of these medicinal plants based on modern genetic and molecular tools. The herb Flos Lonicerae Japonicae (FLJ or Lonicera japonica Thunb.) is used as an anti-inflammatory agent but the chemical quality of FLJ and its medicinal efficacy has not been consistent. Here, we analyzed the transcriptomes and metabolic pathways to evaluate the active medicinal compounds in FLJ and hope that this approach can be used for a variety of medicinal herbs in the future.
We assess transcriptomic differences between FLJ and L. japonica Thunb. var. chinensis (Watts) (rFLJ), which may explain the variable medicinal effects. We acquired transcriptomic data (over 100 million reads) from the two herbs, using RNA-seq method and the Illumina GAII platform. The transcriptomic profiles contain over 6,000 expressed sequence tags (ESTs) for each of the three flower development stages from FLJ, as well as comparable amount of ESTs from the rFLJ flower bud. To elucidate enzymatic divergence on biosynthetic pathways between the two varieties, we correlated genes and their expression profiles to known metabolic activities involving the relevant active compounds, including phenolic acids, flavonoids, terpenoids, and fatty acids. We also analyzed the diversification of genes that process the active compounds to distinguish orthologs and paralogs together with the pathways concerning biosynthesis of phenolic acid and its connections with other related pathways.
Our study provides both an initial description of gene expression profiles in flowers of FLJ and its counterfeit rFLJ and the enzyme pool that can be used to evaluate FLJ quality. Detailed molecular-level analyses allow us to decipher the relationship between metabolic pathways involved in processing active medicinal compounds and gene expressions of their processing enzymes. Our evolutionary analysis revealed specific functional divergence of orthologs and paralogs, which lead to variation in gene functions that govern the profile of active compounds.
Flos Lonicerae Japonicae (FLJ, Lonicera japonica Thunb.) is used as a herbal medicine with anti-inflammatory effect . The first record in the literature on this herb is found in one of the world earliest pharmacopoeias, the Shen-Nong’s Herbals. The commercial value of FLJ in herbal medicine trading markets has increased over 400% in recent years, and over 30% of current traditional Chinese medicine prescriptions contain FLJ. This important herb is used to treat various diseases, such as severe acute respiratory syndromes, H1N1 influenza, and hand-foot-and-mouth disease. FLJ extracts also show other biological and pharmaceutical properties, including, anti-bacterial, anti-inflammatory, anti-viral, liver protection , anti-angiogenic, and antinociceptive activities . However, the quality of FLJ as a medicinal herb is rather inconsistent and largely due to its uncharacterized active compound content.
Chlorogenic acid and luteoloside are biomarkers used by the Chinese Pharmacopoeia (Chinese Pharmacopoeia Commission, 2010) for evaluating the quality of FLJ. Pharmacological experiments show that luteolin has a spectrum of biological activities, particularly antioxidative and anti-inflammatory properties. Luteolin has a direct inhibitory effect on lung fibrosis . Other phenolic compounds, including phenolic acids, have been identified in the methanolic extract of FLJ by liquid chromatography with time-of-flight mass spectrometry . Among these compounds, loganin and sweroside also exhibit anti-inflammatory and analgesic activities, which are candidate active components of the FLJ extract . Rutin is demonstrated to reduce oxidative stress-mediated myocardial damage in both in vitro and in vivo models and may prove beneficial in the treatment of myocardial infarction .
FLJ has other commercial applications, such as scent additive used in foods and cosmetics . The main volatile component is linalool, but other floral volatile compounds, including germacrene D, cis-jasmone, E,E-α-farnesene, nerolidol, cis-3-hexenyl acetate, hexyl acetate, cis-hexenyl tiglate, and indole, have been detected based on headspace analysis and gas chromatography–mass spectrometry (GC-MS) .
The concentrations of active components and volatile compounds are closely correlated with floral developmental stages . The content is higher in the early stages of cell differentiation status and the flower buds often show the highest medical value, whereas the chlorogenic acid content is significantly reduced in blooming flowers. However, the budding period is both short and not synchronized among individual plants, making it problematic for mass harvest.
A limited number of publications have assessed the relationship between the abundance of active compounds and floral development of FLJ based on molecular biological techniques. One of the studied showed that hydroxycinnamoyl-CoA quinate hydroxycinnamoyl transferase (HQT) gene, encoding a protein of 439 amino acids and identified in FLJ, has a tissue distribution that correlated with the pattern of chlorogenic acid abundance . Another study cloned multi-copied allene oxide synthase LjAOS (GenBank accession: DQ303120) from FLJ and demonstrated that LjAOS mRNA is most abundant in flower buds, and its expression correlates with the concentration of chlorogenic acid .
The content of active compounds also differs significantly among the species and varieties  of the Lonicerae family. L. japonica Thunb. var. chinensis (Watts) (rFLJ) is a Chinese local variety and its corolla has purple outer (upper) and white inner (lower) portions and the whole flower has different active compound contents when compared with FLJ . Qin et al. reported different chlorogenic acid contents between FLJ and rFLJ. Changes of the active compound contents often result in different pharmacological activity and medicinal quality . However, a systematic study to compare gene expression and active compound variations in the active parts among varieties and closely related plant species is of essence.
The high-throughput of the next generation RNA sequencing (RNA-seq) technologies offers rapid genome-wide transcriptomic studies and is widely used to define gene structure and expression profiles in model organisms [17–19]. The assembly of de novo transcriptomes based on short reads generated from RNA-seq method allows gene discovery in organisms without reference genomes. In this study, we applied RNA-seq to the study of floral transcriptomes of FLJ and rFLJ. We generated over 100 million reads using the Illumina GAII platform, and analyzed over 6,000 expressed genes from each of the three flowering stages: bud, blooming, and bloomed. We correlated the transcriptomic profiles with metabolic activities of the relevant active compounds, including phenolic acids, flavonoids, terpenoids, and fatty acids, to elucidate effects of enzymatic divergence on biosynthetic pathways.
Flower samples (corollas or all petals) were randomly collected five each from 3-year old FLJ and rFLJ in Doudian plantation (Beijing, China) (Figure 1). The flowering stages are: (1) the bud stage (white, FLJ and red, rFLJ) when the flower bud has not bloomed into a full-size flower yet; (2) the flower1 stage when the white inner petals and white (FLJ) or red (rFLJ) outer petals has just bloomed into full-size flowers; and (3) the flower2 stage when the yellow inner petals and white (FLJ) or red (rFLJ) outer petals bloomed into full-size flowers. We separated the samples into 2 groups; group 1 is used to compare the FLJ flower buds with its flowers from flower1 and flower2 stages, and group 2 is used to compare the flower buds between FLJ and rFLJ. Fresh samples were used for gas chromatography–mass spectrometry (GC-MS), and freeze-dried flowers were used for HPLC. Quick-frozen flowers (in liquid nitrogen) were used for RNA extraction.
RNA isolation and sequence acquisition
Total RNA was extracted from flower samples by using Concert Plant RNA Reagent (Invitrogen, Cat. 12322–012) according to the manufacturer’s protocol. RNA integrity was measured by using gel electrophoresis and spectrophotometer (Nonodrop). An Oligotex-dT30 Super mRNA Purification Kit from TaKaRa (D9086) was used to extract mRNA.
De novo sequence assembly and contig clustering
Prior to assembly and mapping, we removed low quality reads (such as >30% “N” in a read and polyA tails) from the raw data and assembled the processed reads into contigs using ABySS http://www.bcgsc.ca/platform/bioinfo/software/abyss; . We used contigs longer than 100 bp for further annotations. Since the genome sequence of FLJ has not been available, we used BLAST  to align the contigs to the NCBI non-redundant sequence database. Because V. vinifera full-length cDNA sequences provided the most annotations, we clustered the FLJ/rFLJ contigs in reference to the Vitis vinifera cDNA sequences.
Gene annotation and expression analysis
We used BLASTX  to search against the NCBI non-redundant (nr) database to identify transcripts and annotated the transcripts using KEGG  and COG  with an E value cut-off of 10−5. We applied InterPro  and Blast2GO  to the annotation of protein motifs/domains and Gene Ontology (GO) terms. GO annotation enrichment analyses were conducted based on a Benjamini and Hochberg false discovery rate correction with significance set at p < 0.05 by using the Cytoscape plug-in BiNGO.
We mapped the sequence reads and contigs using SOAP http://soap.genomics.org.cn/soapaligner.html;  and handled isoforms/spliced variants with cautions . We used sequence similarity information and the Vitis vinifera full-length cDNAs for transcriptome mapping and tag counting using LASTZ  after clustering the contigs into ESTs. Only uniquely mapped reads were counted. The expression profiling was done by normalizing the total mapped reads and contig length as RPKM (reads per kilobase of transcript sequence per million mapped reads; 19). The effective size was used to adjust RPKM values in subsequent analyses. DEGseq1.2.2  was used to roughly identify the differentially expressed genes (DEGs) via the p value and the RPKM fold-change value. The DEGs were further studied based on pathway/expression analyses and real-time PCR.
Gas chromatography–mass spectrometry profiling
The concentrations of ethanol, acetate, alkane, and terpene in the flower samples were determined based on gas chromatography–mass spectrometry (GC-MS) ( Additional file 1: Figure S1). Fresh flower samples (1.00 g, n = 3) were washed twice with distilled water, subjected to ultrasonic extraction (90 Hz) with 10 ml ethyl acetate for 40 minutes, and filtered through a microfiltration membrane (0.45 μm). Extracted metabolites were analyzed as follows: 1 μl of sample was injected at a split ratio of 10:1 into a Shimadzu GCMS-QP2010 instrument. A VF-5MS capillary column coated with 5% phenyl and 95% dimethylpolysiloxane (30 m × 0.25 mm i.d. and 0.1 μm film thickness; Varian, USA) was employed for separation. Injection temperature was 230°C and the interface temperature was set to 250°C. The ion source was adjusted to 230°C and the solvent cut-time was set to 3 minutes. Helium was the carrier gas at a flow-rate of 1.05 ml minute−1. The temperature program was: an initial temperature of 50°C, programmed at 5°C minutes−1 to 150°C and held for 10 minutes, then ramped at 10°C minute−1 to 260°C and held for 20 minutes. The mass spectrometric detector was operated in the electron impact ionization mode with an ionizing energy of 70 eV, scanning from 40–400 m/z. Peak identification was performed by employing AMDIS (NIST, Gaithersburg, MD, USA) and WILEY7n (Palisade Corporation, NY, USA) databases with a spectral match quality >90%. An internal standard of pentadecanol was added to correct for differences in derivatization efficiency and changes in sample volume during heating. Peaks were quantified by area integration and concentrations were normalized to the quantity of the internal standard recovered. Two technical replicates were analyzed for three biological samples from each flowering stage.
The dried flowers were separately comminuted with a miller. Each solid sample (40 mesh, 0.50 g) was accurately weighed and extracted with 50 ml of 70% aqueous ethanol by ultrasonication for 30 minutes. The extract was cooled to 25°C and diluted to 50 ml with 70% aqueous ethanol, filtered with a 0.45 μm Millipore filter membrane. Then, 10 μl of the filtrate was injected into the HPLC system for analysis ( Additional file 1: Figure S2).
The HPLC system was an Agilent 1200LC series (Agilent Technologies, Palo Alto CA, USA), consisting of an online vacuum degasser (G1379B), a Bin pump SL (G1312B), an auto-sampler (GB67C), a thermostatic column compartment (G1316B), and a diode-array detector (DAD) (G1315C) coupled with an analytical workstation. The column configuration was an Agilent TC-C18 reserved-phase column (5 μm, 250 mm × 4.6 mm). The sample injection volume was 10 μl. The detection wavelength was set at 242 nm for analysis with the flow rate at 1.0 ml minute−1, and the column temperature remaining at 25°C. The mobile phase contained deionized water, formic acid (A; 99:1, v/v), and methanol (B). The elution conditions are shown in Additional file 2: Table S1. To determine the linearity of the chromatographic techniques, calibration plots of eight standards were constructed on the basis of peak areas (y) using seven different concentration solutions ( x). All plots were linear in the examined ranges, and linear ranges were shown at different concentrations for the standard compounds (μg ml−1). The r value in Additional file 2: Table S2 refers to the correlation coefficient of the equation for calculating the content of compound. All the standard compounds showed good linearity (r > 0.9994) in a relatively wide concentration range. The standard compounds, chlorogenic acid (110753), caffeic acid (110885), ferulic acid (110773), rutin (100080), huteoloside (111720), Hyperoside (111521), quercitrin (111538), and quercetin (100081), were purchased from National Institutes for Food and Drug Control in China.
Principal component analysis and statistical assessment of GC-MS and HPLC data
Data sets containing more than two independent biological replicates per samples were statistically analyzed based on the Student’s t test with a significance cutoff of p < 0.05. To assess the metabolic changes or differences between samples and to identify metabolic changes involved in group discrimination, multivariate analyses (PCA, PLS-DA) were performed by using the SIMCA-P + (22.214.171.124.0) program (Umetrics AB, Tvistevdgen 48 Umea 907 19, Sweden).
Phylogeny and identification of paralogs and orthologs
We used the PFAM database  for validating all the gene families and protein sequences and constructed neighbor-joining trees for all sequences (ClustalW2). To identify orthologs, we performed an all-against-all sequence comparison using BLAST (cutoff <10-20) and determined orthologs from the best reciprocal hits >80% alignment length; .
Experimental validation of transcribed sequences
We used RNA samples extracted from the flower samples of FLJ and rFLJ to perform qRT−PCR and M-MLV reverse transcriptase cDNA Synthesis Kit from Takara. The PCR primers are shown in Additional file 2: Table S5. The amplification condition was set as follows: initiated by 1-minute incubation at 95°C, followed by 35 cycles at 95°C for 15 seconds, 57–60°C for 30 seconds, and 68°C for 30 seconds. PCR results were evaluated by using 2–3% NuSieve agarose gels.
Paired-end sequencing and de novo assembly
We designed a paired-end sequencing strategy and acquired nearly-saturated raw sequence data for all five libraries, including FLJ bud, FLJ flower1, FLJ flower2, rFLJ bud, and rFLJ flower2 in a range of 27–41 million reads per library (Table 1). After quality filtering and assembly, the usable sequence reads per library totaled 13–32 million reads. Given the read lengths of either 76 bp or 81 bp, the net transcriptome coverage is deemed adequate.
We employed ABySS (http://www.bcgsc.ca/platform/bioinfo/software/abyss; 20), an assembler developed specifically for the next-generation short-read sequences, to assemble the processed sequences and obtained a total of 180,220 contigs, ranging from 25,232 to 41,796 for each library. We assembled all reads using the SOAP aligner tool , allowing up to two base mismatches. About half (51,607,107 reads) of the total reads are mapped to the contigs, and 49,821,911 reads remain unmapped. Specifically, 11,434,981 (46.01%) reads are mapped to the contigs in the rFLJ bud; 8,202,791 (56.30%) to rFLJ flower2; 17,927,893 (56.27%) to FLJ bud; 8,943,545 (54.21%) to FLJ flower2; and 4,697,897 (36.30%) to FLJ flower1. The average contig lengths are less than 1,000 bp, but the N50 contig sizes are over 1,000 bp for all libraries.
Gene annotation and expression analysis
We used the available public information of plant genes and genomes for annotation and performed a similarity search against the Genbank non-redundant protein database (Genbank nr) using the BLASTx algorithm [21, 33] with an E-value threshold of 10−5 and a size threshold >100 bp. We have 119,965 contigs (66.64%) shown significant similarity to known proteins based on 45,549 unique proteins. Based on the BLAST search, 86% of the contigs show similarities in the six plant species, including Vitis vinifera Ricinus communis Populus trichocarpa Arabidopsis lyrata Glycine max, and Nicotiana tabacum ( Additional file 1: Figure S3), and the fractions of sequences that match to what in V. vinifera are more than 50% for all five libraries. Due to the absence of genome information for FLJ, the full-length cDNA set of V. vinifera (RefSeq project by NCBI) served as the best reference for clustering and combining FLJ and rFLJ data. Moreover, our results indicate that the proportion of the sequences with matches in the Genbank nr database is greater among the longer contigs. For instance, we observed 98.6% matching efficiency for the sequences longer than 2,000 bp but it decreased to 50.8% when the sequence lengths dropped to 100 to 500 bp. The matching efficiencies for the sequences ranging in 500–1,000 bp, 1,000–1,500 bp, and 1,500–2,000 bp, are 90.5%, 96.6%, and 98.2%, respectively.
We defined the FLJ/rFLJ genes using LASTZ  and V. vinifera full-cDNAs (15,674 V. vinifera genes and E value of 10−10) as the reference. Fragmented genes were also identified and joined as ESTs. The FLJ/rFLJ transcriptomes were defined based on the criterion: at least one contig mapped to a reference gene. Nearly 30% of the total reference genes have matches to the FLJ/rFLJ contigs. Finally, we have 5,480, 5,310, 5,818, and 5,131 unigenes identified in rFLJ bud, rFLJ flower2, FLJ bud, and FLJ flower2, respectively. Only the FLJ flower1library (4,483 genes) has less than 5,000 unigenes identified.
We carried out functional and pathway analyses using the Kyoto Encyclopedia of Genes and Genomes (KEGG; http://www.genome.jp/kaas-bin/kaas_main?mode=est_b;22), and 180,020 sequences with significant matches were assigned to 276 KEGG pathways. Of the total, 21,692 unigenes have enzyme commission (EC) numbers ( Additional file 2: Table S3). We attempted to map major compounds that are involved in the biosynthesis of phenylalanine, terpenoid backbone, and fatty acid to the citric acid cycle, glycolysis, and sucrose metabolic pathways based on sequence homologies to the known plant genes ( Additional file 1: Figure S4). We categorized a total of 1,321 unigenes involved in the biosynthetic pathways. In addition to genes assigned to metabolic pathways, 18,417 unigenes are attributable to functions in genetic information processing, membrane transport, signal transduction, immune system, and environmental adaptation. These results demonstrated the power of high-throughput sequencing in identifying novel genes in non-model organisms, providing a valuable resource for investigating specific processes, functions, and pathways involved in active compound formation and their accumulation in FLJ flowers.
Analysis of differentially-expressed genes (DEGs)
We calculated gene expression level based on unique read counts and RPKM values (reads per kilobase of exon model per million mapped reads) for each contig and ESTs (the length of the ESTs was the exon length for the contigs). Two methods were used to define DEGs: DEGseq software based on p value estimates and 2-fold RPKM differences between the two libraries as a threshold for each gene. The numbers of DEGs between different datasets are: (1) 2,316 between the FLJ bud and flower1, (2) 1,713 between the FLJ bud and flower2, (3) 1,163 between FLJ flower1 and flower2, (4) 666 between rFLJ and FLJ bud, and (5) 692 between rFLJ and FLJ flower2.
We identified 262 (26%) up-regulated and 663 (67%) down-regulated genes in the flower bud in comparison with flower1 and flower2 in FLJ (the group 1 comparison; Additional file 1: Figure S5). These DEGs are concentrated on certain pathways, such as biosynthesis of plant hormones (41 DEGs), biosynthesis of terpenoids and steroids (28 DEGs), ribosomes (25 DEGs), biosynthesis of phenylpropanoids (24 DEGs), and biosynthesis of alkaloids derived from terpenoids and polyketides (21 DEGs). Up-regulated genes are absent in certain pathways, such as fatty acid and unsaturated biosyntheses. The total number of down-regulated genes in the three fatty acid-related pathways is 12, which is higher than that found in other pathways without up-regulated genes ( Additional file 1: Figure S6).
One of the down-regulated DEGs in pathogenesis (GO:0009405) is glyceraldehyde-3-phosphate dehydrogenase (G3PD, EC:126.96.36.199; Additional file 1: Figure S7). G3PD catalyzes the conversion of glyceraldehyde-3-phosphate to 1,3-bisphosphoglycerate in glycolysis and plays a critical role in the control of plant metabolism and development . Munoz-Bertomeu reported  that the expression and catalytic activity of G3PD in anthers are necessary for mature pollen development of Arabidopsis. Exine formation in developing pollen exhibits an intricate pattern, primarily comprised of a polymer of fatty acids and phenolic compounds . We show here that the transcription level of sc_FLJ_007660 is 9.57-fold and 27.54-fold higher in the group 1 comparison.
The up-regulated DEGs are involved in transport (GO:0006810), transmembrane transporter activity (GO:0022857), and substrate-specific transporter activity (GO:0022892). One of the DEGs involved in localization (GO:0051234) is a regulator of Vps4 activity in the MVB protein pathway and related to pollen tube growth  ( Additional file 1: Figure S7). Rapid pollen tube growth requires a high rate of sugar metabolism to meet energetic and biosynthetic demands . The transcription level of sc_FLJ_015256 is 0.35-fold and 0.43-fold in the group 1 comparison and the result suggests a distinct sucrose metabolism.
DEGs and their related metabolic pathways
We further investigated several metabolic pathways and selected several representative pathways for more detailed analyses, including phenolic acids, terpenoid and fatty acid metabolism, glycolysis, and TCA cycles (Figure 2 and Additional file 1: Figure S8).
Phenolic acid pathway
Most phenolic compounds are derived from the phenylpropanoid pathway (Figure 3). L-Phenylalanine is first catalyzed to trans-cinnamic acid by phenylalanine ammonia-lyase (PAL, EC:188.8.131.52), and subsequently to p-coumaric acid by trans-cinnamate 4-hydroxylase (C4H). P-coumaric acid can be converted into caffeic acid by p-coumarate 3-hydroxylase (C3H) or to p-coumaroyl-CoA by 4-coumarate-CoA ligase (4CL, EC:184.108.40.206). The four known products of the p-coumaroyl-CoA-related reactions are as follows: (1) caffeoyl-CoA O-methyltransferase (COMT, EC:220.127.116.11) for feruloyl-CoA, (2) cinnamyl-alcohol dehydrogenase (EC:18.104.22.168) for p-coumaryl alcohol, (3) shikimate O-hydroxycinnamoyltransferase (HCT) for p-coumaroyl quinic acid and chlorogenic acid, and (4) chalcone synthase (CHS, EC:22.214.171.124) and chalcone isomerase (CHI, EC:126.96.36.199) for naringenin. In addition, naringenin can be catalyzed to form dihydrokaempferol by dihydroflavonol −4-reductase (DFR, EC:188.8.131.52), to become apiforol by flavanone 4-reductase (F4R) or a DFR homolog, or to become eriodictyol by flavonoid 3′-hydroxylase (F3′H). Eriodictyol can be further catalyzed to become luteoforol by a homolog of DFR (EC:184.108.40.206) and to be luteolin by flavones synthase (FNS). Dihydrokaempferol can also be further converted into cis-3,4-leucopelargonidin by flavanone 3β-hydroxylase (F3H), and eventually to quercetin and rutin by flavonol synthase (FLS).
In the group 1 comparison, we found that the transcription levels of PAL, EC:220.127.116.11, EC:18.104.22.168, and EC:22.214.171.124 are down-regulated, whereas EC:126.96.36.199 and EC:188.8.131.52 are up-regulated in FLJ buds (Figure 2). In the group 2 comparison, the transcription levels of seven DEGs (EC:184.108.40.206, EC:220.127.116.11, EC:18.104.22.168, EC:22.214.171.124, EC:126.96.36.199, EC:188.8.131.52, and EC:184.108.40.206) involved in phenylpropanoid and flavonoid biosynthesis are down-regulated. Our data strongly support the correlation between metabolic pathways and their related gene expressions (Figure 4).
In the group 1 comparison, we obtained five DEGs in terpenoid backbone biosynthesis. The transcription levels of 3-hydroxy-3-methylglutaryl-CoA reductase (HMGR, [EC:220.127.116.11]) in the mevalonate pathway, and 4-diphosphocytidyl-2-C-methyl-D-erythritol kinase [EC:18.104.22.168], isopentenyl-diphosphate delta-isomerase [EC:22.214.171.124], (E)-4-hydroxy-3-methylbut-2-enyl-diphosphate synthase [EC:126.96.36.199], and 4-hydroxy-3-methylbut-2-enyl diphosphate reductase [EC:188.8.131.52] in the plastid MEP/DXP pathway are all down-regulated. Furthermore, the transcription levels of six DEGs (EC:184.108.40.206, EC:220.127.116.11, EC:18.104.22.168, EC:22.214.171.124, EC:126.96.36.199) are down-regulated but farnesyl diphosphate synthase [EC:188.8.131.52 184.108.40.206] is up-regulated. The combination of the different trends results in an increase in germacrene D content in FLJ buds.
Fatty acid pathway
The key enzymes involved in fatty acid metabolism, aldehyde dehydrogenase [EC:220.127.116.11,1 18.104.22.168], acyl-CoA oxidase [EC:22.214.171.124], and long-chain acyl-CoA synthetase [EC:126.96.36.199], are all down-regulated in the group 1 comparison. Other DEGs, such as 3R-hydroxymyristoyl ACP dehydrase [EC:4.2.1.-], acyl-ACP desaturase [EC:188.8.131.52], fatty acyl-ACP thioesterase B [EC:184.108.40.206 3.1.2.-], acyl-CoA oxidase [EC:220.127.116.11], and omega-3 fatty acid desaturase (delta-15 desaturase) [EC:1.14.19.-], involved in the biosynthesis of fatty acids and unsaturated fatty acids, are also down-regulated. In addition, five DEGs (EC:1.14.19.-, EC:1.3.1.-, EC:18.104.22.168/3.1.2.-) involved in fatty acid and unsaturated fatty acid biosynthesis are also down-regulated in the group 2 comparison. Our expression data again are in complete agreement with the metabolic data (Figure 4).
Our group 1 comparison results lead to the identifications of down-regulated DEGs: eight (EC:22.214.171.124, EC:126.96.36.199, EC:188.8.131.52, EC:184.108.40.206, EC:220.127.116.11, EC:18.104.22.168, EC:22.214.171.124/126.96.36.199) in the citrate cycle, eleven (EC:188.8.131.52, EC:184.108.40.206, EC:220.127.116.11, EC:18.104.22.168, EC:22.214.171.124, EC:126.96.36.199, EC:188.8.131.52, EC:184.108.40.206/220.127.116.11) in glycolysis/gluconeogenesis, and five (EC:18.104.22.168, EC:22.214.171.124, EC:126.96.36.199, EC:188.8.131.52 and EC:184.108.40.206) in starch and sucrose metabolism. In addition, we found five up-regulated genes in the group 1 comparison: succinate dehydrogenase (ubiquinone) iron-sulfur protein [EC:220.127.116.11], pyruvate kinase [EC:18.104.22.168], α, α-trehalose-phosphate synthase (UDP-forming) [EC:22.214.171.124], 4-α-glucanotransferase [EC:126.96.36.199], and beta-amylase [EC:188.8.131.52]. In the group 2 comparison, we defined five down-regulated genes (EC:184.108.40.206, EC:220.127.116.11, EC:18.104.22.168, EC:22.214.171.124, EC:126.96.36.199, and EC:188.8.131.52) and they are involved in the citrate cycle, glycolysis/gluconeogenesis, and starch and sucrose metabolism. In contrast, we have eight up-regulated genes: EC:184.108.40.206, EC:220.127.116.11, EC:18.104.22.168, EC:22.214.171.124, EC:126.96.36.199, EC:188.8.131.52, EC:184.108.40.206, EC:220.127.116.11, and EC:18.104.22.168.
We also validated some less-abundant DEGs as representatives for most of the above-mentioned pathways using qRT-PCRs, and the results are consistent with the RNA-seq data ( Additional file 2: Table S5).
Integrated analysis of active compound abundance and gene expression
According to a previous report, the content of chlorogenic acid in the FLJ flower bud is higher than that of the flowers . Our study shows that the contents of chlorogenic acid and four other active compounds (caffeic acid, ferulic acid, luteoloside, and quercitrin) are actually lower, aside from three others (isopropyl laurate, linalool and germacrene D) that remain unchanged in the three different flowering stages (Figure 5A). We also found that related enzymes PAL, CHS, and CHI are up-regulated in the group 1 comparison (Figure 3). These results indicate that there are reduced expressions of the key genes involved in phenolic compound skeleton biosynthesis, leading to changes in the FLJ active compounds during flowering, which provide additional information for the medicinal value of the flower buds. We further found that the content of isopropyl laurate has a similar trend as what we observed in fatty acid biosynthesis.
The MEP/DXP and MVA pathways are thought to be interdependent during the biosynthesis of isoprenoids . Sesquiterpene is synthesized by the MVA pathway during the isoprenoid biosynthetic system in plants and the HMGR multi-gene family catalyzes the synthesis of mevalonate, which is a precursor in this pathway . The inducible HMGR enzyme activity is required for sesquiterpene accumulation in tobacco cell suspension cultures . The transcription level of HMGR has no effect and farnesyl diphosphate sythase (FPS) is only transiently induced after emission of (−)-germacrene D in poplar . However, our results suggest that the content of sesquiterpene may be negatively controlled by HMGR and the up-regulation of FPS may increase the production of germacrene D.
Analysis of active compounds in FLJ and rFLJ
To gain additional insight into the active compounds and metabolic pathway maps in FLJ and rFLJ, we generated metabolic profiles of the active compounds from different tissues using HPLC (Figure 5) and observed reduced production of chlorogenic acid, caffeic acid, ferulic acid, luteoloside, and quercitrin after flowering (Figure 5A). Our GC-MS analysis also indicates variable fatty acid and terpenoid contents in the flowering stages (Table 2). In addition, we observed the decreased content of isopropyl laurate and the increased contents of linalool and germacrene D in the flowering process (Figure 5B). Isopropyl laurate, linalool, and germacrene D are all known major volatile chemicals released in full-bloomed flowers .
We also carried out PCA analysis on selected organic compounds; phenolic acids, fatty acids, and terpenoids are all detected by using GC-MS and HPLC. These compounds form two independent groups (Figure 6), where the contents of chlorogenic acid, luteoloside, quercitin, and isopropyl laurate are higher overall in the rFLJ flower buds as compared to those of FLJ but the germacrene D content is lower in rFLJ.
Interactions of secondary and primary/central metabolisms
The production of secondary metabolites is tightly associated with pathways of primary/central metabolism, such as glycolysis, the shikimate pathway, the production of aromatic amino acids, and aliphatic amino acids . GC-MS and HPLC analysis detected three compound groups produced from primary metabolisms in FLJ and rFLJ: phenolic acids, terpenoids, and fatty acids ( Additional file 1: Figure S4).
Phosphoenolpyruvate (PEP) is an example of glycolysis intermediates. It is indispensable for energy metabolism in the cytosol and delivers ATP and pyruvate catalyzed by cytosolic pyruvate kinase (PK) . Inside the plastids, PEP acts as a precursor for at least four metabolic pathways: fatty acids, isoprenoids, branched chain amino acids, and the shikimate pathway . However, chloroplasts and most non-green plastids lack the ability to produce PEP via glycolysis, because their enolase is either absent or has low activity . In this study, we found that enolase (sc_FLJ_010870) transcription is 1232.8-fold higher in the FLJ flower buds when compared with flower1, and no detectable expression of the enzyme in flower2. Therefore, the formation of PEP is negatively correlated with FLJ flowering. In contrast, the transcription level of PK (sc_FLJ_000591), a competing enzyme, is 0.11-fold and 0.54-fold in the group 1 comparison. We neither observed significant changes in the flavonoid content nor in ABA and fatty acids contents as what was reported in an eno1 mutant when compared with the wild type of A. thaliana. We did found that the transcription level of enolase is up-regulated in buds as compared with the two flower stages, as well as in the group 2 comparison, but showed no correlation with the phenolic acid content. In contrast, the transcription level of PK is higher in the same comparison. Our combined analysis of biosynthesis of phenolic acid, terpenoids, and fatty acids did indicate that there is a potential correlation between PK expression level and the contents of flavonoids, terpenoids, and fatty acids in FLJ.
ATP synthesis could also have influences on the interaction of primary and secondary metabolisms. We assessed the predicted phosphorylation sites of the relevant enzymes ( Additional file 2: Table S4) and found that all are candidates for phosphorylation-based regulation. Of the defined DEGs, a study in animals showed that HMGR exists in both active (dephosphorylated) and inactive (phosphorylated) forms . PK has also been shown to be more susceptible to inhibition by ATP . We analyzed the differential expression of ATP synthesis-coupled electron transporters and found that this gene family is down-regulated in the group 1 comparison and up-regulated in the group 2 comparison. Prediction of phosphorylation sites from protein sequences indicates that HMGR (sc_FLJ_012288) has nine phosphorylation sites, whereas pyruvate kinase (sc_FLJ_000591) has four phosphorylation sites.
Previous studies on primary and secondary metabolisms have indicated that they are linked via the phenylalanine pool [50, 51]. Therefore, it is possible that a reduced carbon flux in the phenylpropanoid pathway may affect carbohydrate metabolism.
As Matt et al.  showed that an increase in the sugar/amino acid ratio resulted in an elevated production of carbon-rich phenylpropanoids, we found that biosynthesis of phenylpropanoids is down-regulated in both group 1 and group 2 comparisons, and the transcription level of β-glucosidase that generates β-D-glucose and α-D-glucose is down-regulated in the group 2 comparison. Furthermore, sucrose synthase that creates sucrose is down-regulated in the group 1 comparison but β-amylase that generates maltose is up-regulated in the group 1 comparison and down-regulated in the group 2 comparison.
Moriizumi  reported that glucose-regulated transcription of pyruvate kinase is mediated by its glucose response element; the carbohydrate response elements are composed of two E box-like motifs separated by 5 bp and is recognized by two basic helix-loop-helix/leucine zipper (bHLH/LZ) proteins [54, 55]. In the group 2 context, down-regulation of pyruvate kinase may be mediated by down-regulating glucose biosynthesis, and in the group 1 context, however, the transcription level of glucose biosynthetic enzymes is insignificant albeit up-regulated pyruvate kinase. Finally, two bHLH (sc_FLJ_008421 and sc_FLJ_006390) proteins are seen up-regulated and down-regulated in the group 1 and 2 comparisons, respectively.
A putative sequence for a carbohydrate-response-element binding protein (sc_FLJ_004075) is obtained based on sequence homology (a homolog in Norway rat; EMBLCDS:BAB77523 and in chicken GenBank: ABV72703.1). The transcript is up-regulated in the group 1 comparison, suggesting that bHLH transcription factors may be involved in regulating the response of PK to glucose in FLJ.
The complexity of defining orthologs and paralogs for key metabolic pathways
Since gene duplication is very common in plant genomes, we made an effort to differentiate orthologs and paralogs from all homologs. The overall sequence identity between FLJ and rFLJ contigs is 99.0%. In our analysis, we selected 55 DEGs sequences from all pathway-related genes, where six of the selected genes have slightly lower identities, about 97.4% ( Additional file 2: Table S6).
We identified the orthologs and paralogs of PAL, CHS, HMGR, and PK based on the genome sequences of Arabidopsis and grape (http://www.phytozome.net; Additional file 2: Table S7) and built phylogenetic trees (Figure 7 and Additional file 1: Figure S9). First, between FLJ and rFLJ, the PAL family genes clustered into two groups; one contains a pair of orthologs that have no detectable expression in the flower buds and another has paralogs expressed at high levels. The total RPKM of PAL paralogs is 5.9-fold higher in rFLJ when compared to FLJ. Second, we have two pairs of CHS orthologs in both FLJ and rFLJ; the FLJ paralogs expressed at high levels, 17-fold higher than those in rFLJ. Third, in FLJ and rFLJ, HMGR genes are also clustered into two groups and each has two pairs of orthologs. The FLJ paralogs are expressed at low levels to the extent that they may become pseudogenes already. The total RPKM of HMGR paralogs is 1.6-fold higher in rFLJ than in FLJ. Finally, there are four groups of PK genes and the expressed PK paralogs primarily present in two of the four clusters (Cluster 3 and Cluster 4; Figure 7).
Gene expression data provide more comprehensive understanding of FLJ as medicinal plant
Although several studies have reported that the content of chlorogenic acid is higher in the flower buds as compared to that in flowers, it has long been disputed as to which organ has the highest medical value due to the applications of controversial evaluation methods. In this study, we obtained a sufficient amount of transcriptomic data from both young buds and mature flowers, and carried out an integrated analysis on the variations of gene expression and the contents of active compounds. Our data from different flowering stages indicate that the contents of eight major active compounds either decreased (five of them) or remain unchanged (three of them) and that the biosynthesis of the active compounds is overall higher in the buds than in the flowers. In addition, many key enzymes, such as PAL, CHS, and CHI, are up-regulated in the FLJ flower buds (Figure 4). PAL is a key enzyme in the synthesis of phenylpropanoid phytoalexins and other phenolics . Previous data have shown that the PAL expression levels decline during flowering in Nicotiana, and the decrease enzyme activity corresponds to the decreased major phenylpropanoid compounds. Our data add further support for this notion, as we found that the CHS and CHI genes are also significantly down-regulated in the later flowering stages. CHS functions to produce flavonoid precursors, and CHI plays a major part in the cyclization reaction from chalcone to flavanone .
The quality of herbal medicine has been very difficult to control and to evaluate primarily because of the complexity and incomplete knowledge of the active medicinal compounds. The primary methods that have been used for quality evaluation of Chinese herbal medicines are chemical and pharmacological analyses. Chemical evaluations showed that chlorogenic acid and luteoloside are two common active compounds found in FLJ. However, content and fingerprint analysis of one or more of these compounds are not indicative for the medicinal value of the plant, and our genomic approach provides a comprehensive survey. Our study generated gene expression data for terpenoids and fatty acid biosynthesis and increased valuable knowledge on other FLJ compounds.
Both fatty and phenolic acid biosyntheses may depend on the interaction between the two pathways [59, 60]. In FLJ, the transcription level of long-chain acyl-CoA synthetase (sc_FLJ_011431) and 4-coumarate-CoA ligase (sc_FLJ_001317) decreases after flowering. Since the contents of phenolic compounds and fatty acids are down-regulated in both the group 1 and 2 comparisons, some of the biosynthetic genes for phenypropanoid compounds and fatty acids may be regulated in a similar way or function as similar enzymes (such as sharing catalytic domains).
Changes in phenolic acids, fatty acid biosynthesis, and the MEP/DXP pathway show a decreased trend in gene expression from the medicinal organ (buds) to the non-medicinal organs (flower1 and flower2). In addition, we used phylogenetic tools for paralog analysis and revealed that the mRNA levels of these enzymes have higher expressions in the flower buds of both FLJ and rFLJ. We found that a total of eighteen enzymes in these biosynthetic networks are differentially expressed ( Additional file 2: Table S9).
Enzymatic divergence of orthologs and paralogs results in gene function variation and active compound content
Secondary metabolism varies intensively, even between two closely related taxonomic groups, and the underlining functional variations, such as enzyme activities, often lead to the production of unique compounds. In this study, we identified several orthologous enzymes as well as their related paralogs and evaluated their evolutionary relatedness between FLJ and rFLJ. We observed that the RPKM values of PAL, CHS, and HMGR are higher in rFLJ than FLJ, and these variations in expression may lead to alternations in the active compound contents between in the two plants.
In contrast to specific evolutionary changes to individual enzymes during speciation, most of the functional variations appear to be related to gene or genome duplications . After gene duplication events, most paralogs are lost over time, but those have survived often gain new functions (neo-functionalization), partition the original function into different time and tissues (subfunctionalization), or have lost their functions (nonfunctionalization). Our data indicate that gene duplication in FLJ may provide an opportunity for neo-functionalization, whereby the PK gene and its orthologs and paralogs may evolve to have complementary enzyme functions. For instance, an FLJ PK paralog turns out to have a high RPKM value and may undergo neofunctionalization, whereas the expression of another paralog in FLJ is higher than that of rFLJ. There is a third PK paralog whose expression levels are quite similar in both plants. Nevertheless, the divergence of gene expression due to duplicated genes appears to play direct roles in the production of active compounds in FLJ and rFLJ.
We used a comparative approach to address whether transcriptomes can be informative for the analysis of active medicinal compounds in herbal plants. Our study not only provided an initial description of the expression profiles of FLJ flowers, but also identified the enzyme pool that can be used to evaluate FLJ quality in future studies. We also associated metabolic pathways involved in processing active medicinal compounds to the expressions of their catalytic enzymes. We also used sequence evolution as a tool to identify orthologs and paralogs, as well as pathways for the biosynthesis of phenolic acid and its interactions with other pathways (Figure 8), and revealed that functional divergence of orthologs and paralogs may lead to variations in gene functions that control the active compound contents among different tissues and plants.
Lonicera japonica Thunb
Lonicera japonica Thunb. var. chinensis (Wats.)
Quantitative reverse transcriptase PCR
Reverse transcriptase PCR
Hong SZJ, Yang G, Luo X: Earliest report of Flos Lonicerae Japonicae and its medical organ. Zhongyaocai. 1997, 20: 2-
Xiang T, Xiong QB, Ketut AI, Tezuka Y, Nagaoka T, Wu LJ, Kadota S: Studies on the hepatocyte protective activity and the structure-activity relationships of quinic acid and caffeic acid derivatives from the flower buds of Lonicera bournei. Planta Med. 2001, 67: 322-325. 10.1055/s-2001-14337.
Yoo HJ, Kang HJ, Song YS, Park EH, Lim CJ: Anti-angiogenic, antinociceptive and anti-inflammatory activities of Lonicera japonica extract. J Pharm Pharmacol. 2008, 60: 779-786.
Hsu SL, Chen CY, Peng WH, Wu LC, Wu CC: Luteolin Ameliorates Experimental Lung Fibrosis Both in Vivo and in Vitro: Implications for Therapy of Lung Fibrosis. J Agr Food Chem. 2010, 58: 11653-11661. 10.1021/jf1031668.
Li P, Qi LW, Chen CY: Structural characterization and identification of iridoid glycosides, saponins, phenolic acids and flavonoids in Flos Lonicerae Japonicae by a fast liquid chromatography method with diode-array detection and time-of-flight mass spectrometry. Rapid Commun Mass Sp. 2009, 23: 3227-3242. 10.1002/rcm.4245.
Chung JH, Ryu KH, Rhee HI, Kim JH, Yoo H, Lee BY, Um KA, Kim K, Noh JY, Lim KM: Anti-Inflammatory and Analgesic Activities of SKLJI, a Highly Purified and Injectable Herbal Extract of Lonicera japonica. Biosci Biotech Bioch. 2010, 74: 2022-2028. 10.1271/bbb.100279.
Chang KC, Jeong JJ, Ha YM, Jin YC, Lee EJ, Kim JS, Kim HJ, Seo HG, Lee JH, Kang SS: Rutin from Lonicera japonica inhibits myocardial ischemia/reperfusion-induced apoptosis in vivo and protects H9c2 cells against hydrogen peroxide-mediated injury via ERK1/2 and PI3K/Akt signals in vitro. Food Chem Toxicol. 2009, 47: 1569-1576. 10.1016/j.fct.2009.03.044.
Ge B, LX, Yi K, Tian Y: The Active Constituent and Pharmaceutical Action of Flos Lonicerae and Its Application. Chinese Wild Plant Resources. 2004, 23: 5-
El-Sayed AM, Mitchell VJ, McLaren GF, Manning LM, Bunn B, Suckling DM: Attraction of New Zealand Flower Thrips, Thrips obscuratus, to cis-Jasmone, a Volatile Identified from Japanese Honeysuckle Flowers. J Chem Ecol. 2009, 35: 656-663. 10.1007/s10886-009-9619-3.
Geng S, Ning X, Wu H, Lin H, Zhao S, Xu H: The Structure of Flower in Different Developmental Stages in Relation to the Varieties of Chlorogenic Acid Content in Lonicera confusa. Acta Botanica Yunnanica. 2005, 27 (3): 279-287.
Bai GB, Peng XX, Li WD, Wang WQ: Cloning and Characterization of a cDNA Coding a Hydroxycinnamoyl-CoA Quinate Hydroxycinnamoyl Transferase Involved in Chlorogenic Acid Biosynthesis in Lonicera japonica. Planta Med. 2010, 76: 1921-1926. 10.1055/s-0030-1250020.
Jiang K, Pi Y, Huang Z, Hou R, Zhang Z, Lin J, Sun X, Tang K: Molecular cloning and mRNA expression profiling of the first specific jasmonate biosynthetic pathway gene allene oxide synthase from Hyoscyamus niger. Russ J Genet. 2009, 45: 430-439. 10.1134/S1022795409040073.
Yang M, HF, Zhang L: Comparison on the Chlorogenic Acid Contents in Different Kinds and Different Parts of Flos Lonicerae. Journal of Instrumental Analysis. 2006, 25: 122-123.
Lonicerae Japonicae. Edited by: HC. 1998, Science Press, Beijing
Qin S, YY, Hu G, Chen X, Li X: Comparison of active compounds between Lonicera japonica Thunb and their variation. China J of Exper Tran Med Form. 2010, 16: 2-
Lei Z, ZR, Zeng R, He Y: Comparative experiments on antipyretic effect between Lonicera macrathodes Hands Mazz. and the certified Flos Lonicera. Journal of Hunan Traditional Chinese Medicine University of Hunan. 2005, 5: 2-
Bahler J, Wilhelm BT, Marguerat S, Watt S, Schubert F, Wood V, Goodhead I, Penkett CJ, Rogers J: Dynamic repertoire of a eukaryotic transcriptome surveyed at single-nucleotide resolution. Nature. 2008, 453: U1239-U1239. 10.1038/nature07002.
Snyder M, Nagalakshmi U, Wang Z, Waern K, Shou C, Raha D, Gerstein M: The transcriptional landscape of the yeast genome defined by RNA sequencing. Science. 2008, 320: 1344-1349. 10.1126/science.1158441.
Wold B, Mortazavi A, Williams BA, Mccue K, Schaeffer L: Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Methods. 2008, 5: 621-628. 10.1038/nmeth.1226.
Simpson JT, Wong K, Jackman SD, Schein JE, Jones SJ, Birol I: ABySS: a parallel assembler for short read sequence data. Genome Res. 2009, 19: 1117-1123. 10.1101/gr.089532.108.
Altschul SF, Madden TL, Schaffer AA, Zhang JH, Zhang Z, Miller W, Lipman DJ: Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997, 25: 3389-3402. 10.1093/nar/25.17.3389.
Kanehisa M, Goto S: KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 2000, 28: 27-30. 10.1093/nar/28.1.27.
Tatusov RL, Fedorova ND, Jackson JD, Jacobs AR, Kiryutin B, Koonin EV, Krylov DM, Mazumder R, Mekhedov SL, Nikolskaya AN: The COG database: an updated version includes eukaryotes. BMC Bioinforma. 2003, 4: 41-10.1186/1471-2105-4-41.
Hunter S, Apweiler R, Attwood TK, Bairoch A, Bateman A, Binns D, Bork P, Das U, Daugherty L, Duquenne L: InterPro: the integrative protein signature database. Nucleic Acids Res. 2009, 37: D211-D215. 10.1093/nar/gkn785.
Aparicio G, Gotz S, Conesa A, Segrelles D, Blanquer I, Garcia JM, Hernandez V, Robles M, Talon M: Blast2GO goes Grid: Developing a Grid-Enabled Prototype for Functional Genomics Analysis. St Heal T. 2006, 120: 194-204.
Maere S, Heymans K, Kuiper M: BiNGO: a Cytoscape plugin to assess overrepresentation of gene ontology categories in biological networks. Bioinformatics. 2005, 21: 3448-3449. 10.1093/bioinformatics/bti551.
Li R, Li Y, Kristiansen K, Wang J: SOAP: Short oligonucleotide alignment program. Bioinformatics. 2008, 24: 713-714. 10.1093/bioinformatics/btn025.
Dewey CN, Li B, Ruotti V, Stewart RM, Thomson JA: RNA-Seq gene expression estimation with read mapping uncertainty. Bioinformatics. 2010, 26: 493-500. 10.1093/bioinformatics/btp692.
Harris RS: Improved pairwise alignment of genomic DNA. Ph. D. Thesis. 2007, The Pennsylvania State University,
Wang XW, Wang LK, Feng ZX, Wang X, Zhang XG: DEGseq: an R package for identifying differentially expressed genes from RNA-seq data. Bioinformatics. 2010, 26: 136-138. 10.1093/bioinformatics/btp612.
Bateman A, Birney E, Durbin R, Eddy SR, Howe KL, Sonnhammer ELL: The Pfam protein families database. Nucleic Acids Res. 2000, 28: 263-266. 10.1093/nar/28.1.263.
Tatusov RL, Koonin EV, Lipman DJ: A genomic perspective on protein families. Science. 1997, 278: 631-637. 10.1126/science.278.5338.631.
Cameron M, Williams HE, Cannane A: Improved gapped alignment in BLAST. Ieee Acm T Comput Bi. 2004, 1: 116-129.
Munoz-Bertomeu J, Cascales-Minana B, Alaiz M, Segura J, Ros R: A critical role of plastidial glycolytic glyceraldehyde-3-phosphate dehydrogenase in the control of plant metabolism and development. Plant Signal Behav. 2010, 5 (1): 67-69. 10.4161/psb.5.1.10200.
Munoz-Bertomeu J, Cascales-Minana B, Mulet JM, Baroja-Fernandez E, Pozueta-Romero J, Kuhn JM, Segura J, Ros R: Plastidial glyceraldehyde-3-phosphate dehydrogenase deficiency leads to altered root development and affects the sugar and amino acid balance in Arabidopsis. Plant Physiol. 2009, 151: 541-558. 10.1104/pp.109.143701.
Lee Y, Choi H, Jin JY, Choi S, Hwang JU, Kim YY, Suh MC: An ABCG/WBC-type ABC transporter is essential for transport of sporopollenin precursors for exine formation in developing pollen. Plant J. 2011, 65: 181-193. 10.1111/j.1365-313X.2010.04412.x.
Jiang LW, Wang H, Tse YC, Law AHY, Sun SSM, Sun YB, Xu ZF, Hillmer S, Robinson DG: Vacuolar sorting receptors (VSRs) and secretory carrier membrane proteins (SCAMPs) are essential for pollen tube growth. Plant J. 2010, 61: 826-838. 10.1111/j.1365-313X.2009.04111.x.
Mellema S, Eichenberger W, Rawyler A, Suter M, Tadege M, Kuhlemeier C: The ethanolic fermentation pathway supports respiration and lipid biosynthesis in tobacco pollen. Plant J. 2002, 30: 329-336. 10.1046/j.1365-313X.2002.01293.x.
Rodriguez-Concepcion M, Fores O, Martinez-Garcia JF, Gonzalez V, Phillips MA, Ferrer A, Boronat A: Distinct light-mediated pathways regulate the biosynthesis and exchange of isoprenoid precursors during Arabidopsis seedling development. Plant Cell. 2004, 16: 144-156. 10.1105/tpc.016204.
Ha SH, Kim JB, Hwang YS, Lee SW: Molecular characterization of three 3-hydroxy-3-methylglutaryl-CoA reductase genes including pathogen-induced Hmg2 from pepper (Capsicum annuum). Bba-Gene Struct Expr. 2003, 1625: 253-260. 10.1016/S0167-4781(02)00624-3.
Chappell J, Vonlanken C, Vogeli U: Elicitor-Inducible 3-Hydroxy-3-Methylglutaryl Coenzyme-a Reductase-Activity Is Required for Sesquiterpene Accumulation in Tobacco Cell-Suspension Cultures. Plant Physiol. 1991, 97: 693-698. 10.1104/pp.97.2.693.
Arimura G, Huber DP, Bohlmann J: Forest tent caterpillars (Malacosoma disstria) induce local and systemic diurnal emissions of terpenoid volatiles in hybrid poplar (Populus trichocarpa x deltoides): cDNA cloning, functional characterization, and patterns of gene expression of (−)-germacrene D synthase, PtdTPS1. Plant J. 2004, 37: 603-616. 10.1111/j.1365-313X.2003.01987.x.
Zhuang X, Klingeman WE, Hu J, Chen F: Emission of volatile chemicals from flowering dogwood (cornus Florida L.) flowers. J Agric Food Chem. 2008, 56: 9570-9574. 10.1021/jf801651v.
Aharoni A, Galili G: Metabolic engineering of the plant primary-secondary metabolism interface. Curr Opin Biotechnol. 2011, 22: 239-244. 10.1016/j.copbio.2010.11.004.
Givan CV: Evolving concepts in plant glycolysis: two centuries of progress. Biol Rev. 1999, 74: 277-309. 10.1017/S0006323199005344.
Hausler RE, Prabhakar V, Lottgert T, Geimer S, Dormann P, Kruger S, Vijayakumar V, Schreiber L, Gobel C, Feussner K: Phosphoenolpyruvate Provision to Plastids Is Essential for Gametophyte and Sporophyte Development in Arabidopsis thaliana. Plant Cell. 2010, 22: 2594-2617. 10.1105/tpc.109.073171.
Vanderstraeten D, Rodriguespousada RA, Goodman HM, Vanmontagu M: Plant Enolase - Gene Structure, Expression, and Evolution. Plant Cell. 1991, 3: 719-735.
Angelin B, Einarsson K, Liljeqvist L, Nilsell K, Heller RA: 3-Hydroxy-3-Methylglutaryl Coenzyme-a Reductase in Human-Liver Microsomes - Active and Inactive Forms and Cross-Reactivity with Antibody against Rat-Liver Enzyme. J Lipid Res. 1984, 25: 1159-1166.
Morin RJ, Noble NA, Srikantaiah MV: Modulation of 3-Hydroxy-3-Methyl Glutaryl Coa Reductase by 2,3-Diphosphoglyceric Acid. Experientia. 1984, 40: 953-955. 10.1007/BF01946455.
Properties of Rat Pyruvate Kinase Isozymes. Edited by: Ibsen KH, TP, Basabe J. 1975, Academic Press, Inc, New York
Rohde A, Morreel K, Ralph J, Goeminne G, Hostyn V, De Rycke R, Kushnir S, Van Doorsselaere J, Joseleau JP, Vuylsteke M: Molecular phenotyping of the pal1 and pal2 mutants of Arabidopsis thaliana reveals far-reaching consequences on phenylpropanoid, amino acid, and carbohydrate metabolism. Plant Cell. 2004, 16: 2749-2771. 10.1105/tpc.104.023705.
Matt P, Krapp A, Haake V, Mock HP, Stitt M: Decreased Rubisco activity leads to dramatic changes of nitrate metabolism, amino acid metabolism and the levels of phenylpropanoids and nicotine in tobacco antisense RBCS transformants. Plant J. 2002, 30: 663-677. 10.1046/j.1365-313X.2002.01323.x.
Moriizumi S, Gourdon L, Lefrancois-Martinez AM, Kahn A, Raymondjean M: Effect of different basic helix-loop-helix leucine zipper factors on the glucose response unit of the L-type pyruvate kinase gene. Gene Expr. 1998, 7: 103-113.
Ma L, Sham YY, Walters KJ, Towle HC: A critical role for the loop region of the basic helix-loop-helix/leucine zipper protein Mlx in DNA binding and glucose-regulated transcription. Nucleic Acids Res. 2007, 35: 35-44. 10.1093/nar/gkl730.
Li MV, Chang B, Imamura M, Poungvarin N, Chan L: Glucose-dependent transcriptional regulation by an evolutionarily conserved glucose-sensing module. Diabetes. 2006, 55: 1179-1189. 10.2337/db05-0822.
Maffi D, Iriti M, Pigni M, Vannini C, Faoro F: Uromyces appendiculatus infection in BTH-treated bean plants: ultrastructural details of a lost fight. Mycopathologia. 2011, 171: 209-221. 10.1007/s11046-010-9350-1.
Fukasawa-Akada T, Kung SD, Watson JC: Phenylalanine ammonia-lyase gene structure, expression, and evolution in Nicotiana. Plant Mol Biol. 1996, 30: 711-722. 10.1007/BF00019006.
Nishihara M, Nakatsuka T, Yamamura S: Flavonoid components and flower color change in transgenic tobacco plants by suppression of chalcone isomerase gene. FEBS Lett. 2005, 579: 6074-6078. 10.1016/j.febslet.2005.09.073.
Dobritsa AA, Lei Z, Nishikawa S, Urbanczyk-Wochniak E, Huhman DV, Preuss D, Sumner LW: LAP5 and LAP6 encode anther-specific proteins with similarity to chalcone synthase essential for pollen exine development in Arabidopsis. Plant Physiol. 2010, 153: 937-955. 10.1104/pp.110.157446.
Douglas CJ, Souza CD, Kim SS, Koch S, Kienow L, Schneider K, McKim SM, Haughn G, Kombrink E: A Novel Fatty Acyl-CoA Synthetase Is Required for Pollen Development and Sporopollenin Biosynthesis in Arabidopsis. Plant Cell. 2009, 21: 507-525. 10.1105/tpc.108.062513.
Komatsuda T, Sakuma S, Pourkheirandish M, Matsumoto T, Koba T: Duplication of a well-conserved homeodomain-leucine zipper transcription factor gene in barley generates a copy with more specific functions. Funct Integr Genomic. 2010, 10: 123-133. 10.1007/s10142-009-0134-y.
Flowers were provided by Chinese Shandong YATE Ecological Co. and auth by Dr. Qingjun Yuan. We thank Dr. Haifang Li for helping the GC-MS analysis. The work received financial support from the National Natural Science Foundation of China (81001605).
The authors declare that they have no competing interests.
JY and LQH designed the study. YY and SSQ collected samples and did GC-MS analysis. YY, MHL, and XW did HPLC analysis. LPS, GML, and YY carried out data analysis. XMW contributed to RNA-seq library construction and sequencing. JY and YY contributed writing of the manuscript. All authors read and approved the final manuscript.
Yuan Yuan, Lipu Song contributed equally to this work.
Electronic supplementary material
Additional file 1: Figure S1. Comparing volatile compound of FLJ and rFLJ using Gas chromatography-mass spectrometry. A, Bud, B, flower1, C, flower2. Black line represent FLJ and Red line represent rFLJ. FLJ, Lonicera japonica Thumb; rFLJ, Lonicera japonica Thunb. var. chinensis (Wats.). Figure S2. HPLC analysis of active compounds in FLJ and rFLJ. A, Mix of standard compounds. Chlorogenic acid (RT, 13.20; MW, 354.31); Caffeic acid (RT, 17.39; MW, 180.15); ferulic acid (RT, 24.24; MW, 194.18); Rutin (RT, 24.24; MW, 610.52); Luteoloside(RT, 25.27; MW,448.4); Hyperoside (RT, 25.35; MW, 464.37); Quercitrin (RT, 28.65; MW, 448.38); Quercetin (RT, 38.04; MW, 302.24). B, buds of rFLJ; C, buds of FLJ. FLJ, Lonicera japonica Thumb; rFLJ, Lonicera japonica Thunb. var. chinensis (Wats.). Figure S3. BlastX analysis result of contigs in FLJ and rFLJ with all non-redundant (NR) database in Genebank. Six species,Vitis vinifera, Ricinus communis, Populus trichocarpa, Glycine max, Arabidopsis lyrata, Nicotiana tabacum has highest identity with FLJ and rFLJ bud. E-value cut-off was lower than 1e-5. FLJ, Lonicera japonica Thumb; rFLJ, Lonicera japonica Thunb. var. chinensis (Wats.). Figure S4. Pathways of major chemical compounds in Lonicera japonica Thumb (FLJ). All of contigs from three FLJ libraries were annotated with KEGG database. The six pathways, phenylalanine metabolism, terpenoid backbone, fatty acid biosynthesis, citric acid cycle, glycolysis and sucrose metabolism were analysis. Green rectangles repress enzymes finding in FLJ transcriptome. Figure S5. Analysis of gene differential express using MA-plot-based method. M is the Y -axis and represents the intensity ratio, and A is the X-axis and represents the average intensity for each transcript. The red points are the genes identity as differentially expressed. FLJ, Lonicera japonica Thumb; rFLJ, Lonicera japonica Thunb. var. chinensis (Wats.). Figure S6 Pathway assignment based on KEGG analysis of differential express genes between buds and other two flower developmental periods of Lonicera japonica Thumb (Group1). A, Number of contig with down-regulated and up-regulated differential express genes. B, Number of contig with only up-regulated differential express genes. Figure S7. Gene Ontology classification of differential express genes. The results are summarized in three main categories: Biological process, Cellular component and Molecular function. FLJ, Lonicera japonica Thumb; rFLJ, Lonicera japonica Thunb. var. chinensis (Wats.). A, Gene Ontology classification of differential express genes between buds and two other flower developmental period of FLJ. Bud and flower1, differential express genes between bud and flower1; Bud and flower2, differential express genes between bud and flower2; up-regulated, up-regulated express genes in both in between bud and flower1 and in between bud and flower2; down-regulated, down-regulated express genes in both in between bud and flower1 and in between bud and flower2. B, Gene Ontology classification of differential express genes between buds of FLJ and rFLJ.Group2, differential express genes between buds of FLJ and rFLJ. up-regulated, up-regulated express genes; down-regulated, down-regulated express genes. Figure S8. Gene express level in bud, flower1 and flower2 of Lonicera japonica Thumb. Square represents gene express level and nine kinds of color indicate rpkm of scaffolds. B, bud;F1, flower1; F2, flower2. rpkm of scaffolds was calculated according to Grape full-length cDNA sequence. Figure S9. Phylogenetic analysis of the predicted amino acids sequences and expression level of the Phenylalanine ammonia-lyase (PAL), 3-hydroxy-3-methylglutaryl-CoA reductase (HMGR) and chalcone synthase (CHS) homologues among the buds of FLJ and rFLJ and PAL,HMGR and CHS family genes in Arabidopsis and Grape. The phylogenetic tree was constructed by the neighborjoining method using ClustalW2. Identification of PAL,HMGR and CHS homologues was by searching the domain(PF00221, PF00368 and PF00195, respectively) in PFAM database. FLJ, Lonicera japonica Thumb; rFLJ, Lonicera japonica Thunb. var. chinensis (Wats). (PPT 2 MB)
Authors’ original submitted files for images
Below are the links to the authors’ original submitted files for images.