Promoter Analysis and Transcriptional Profiling of Ginkgo biloba 3- Hydroxy-3-Methylglutaryl Coenzyme A Reductase (GbHMGR) gene in Abiotic Stress Responses

The terpene trilactones (TTLs) are believed to be important for the pharmacological properties of Ginkgo biloba leaves extract. 3-Hydroxy-3-methylglutaryl coenzyme A reductase (HMGR) is a critical enzyme involved in the biosynthetic pathway of TTLs. In this study, an 1.2-kb fragment of 5’ flanking region of the HMGR gene (GbHMGR), was isolated from G. biloba by genome walking. Extensive sequence analysis revealed the presence of evolutionarily conserved and over-represented putative cis-acting elements in light-regulated transcription, hormone signaling (gibberellic acid, jasmonate and salicylic acid), elicitor and stress responses (cold/dehydration responses), and plant defense signaling (W-box/WRKY) that are common to the promoter region of GbHMGR. EMSA analysis suggested possible functionality of W-box in GbHMGR promoter region. The behavior of gene transcripts in ginkgo callus upon light, low temperature, MeJA and SA treatments further verified the regulatory function of GbHMGR promoter. A significant positive relationship between gene expression level and total TTL contents suggested that GbHMGR might be one of key genes involved in TTL biosynthesis in G. biloba.


Introduction
The maidenhair tree Ginkgo biloba L., known as living fossil, has undergone very little evolutionary changes over 200 million years and is considered to be native to China, Korea, and Japan (Singh et al., 2008). In recent years, standardized extracts of G. biloba leaves have been amongst the top-selling phytomedicines in the world (Gertz and Kiefer, 2004). Active compounds in G. biloba extract improve blood circulation, discourage clot formation, reinforce the walls of the capillaries and protect nerve cells from harm when deprived of oxygen (Mohanta et al., 2014). Flavonoids and terpene trilactones (TTLs) are believed to be associated with most of the pharmacological properties of G. biloba extracts. While flavonoids can be obtained from many other plants, ginkgolides and bilobalide, termed as TTLs, are unique components of G. biloba (Liao et al., 2011). Though TTLs is considered to play a key role in the active ingredients, the content of TTLs is very low, within 0.06% in dry leaves of G. biloba (van Beek and Montoro, 2009). Bioengineering is an ideal way to increase the content of TTLs, but which relies on the overall elucidation of the TTL biosynthetic pathway both at molecular genetics and biochemistry levels in G. biloba.
Terpenoids such as ginkgolides are biosynthesized from a universal 5-carbon building block Isopentenyl diphosphate (IPP) (Schwarz and Arigoni, 1999). IPP can be derived from two pathways: One is the classical cytosolic mevalonic acid (MVA) pathway and the other is the plastidial methylerythritol 4-phosphate (MEP) pathway, which is mevalonate independent. The MVA pathway in the cytosol, starting from 3 acetyl-CoA to finally yield IPP, is responsible for synthesizing sesquiterpenoids and sterols. The MEP pathway producing IPP and dimethylallyl diphosphate (DMAPP) from pyruvate and D-glyceraldehyde 3-phosphate (GAP) is mainly responsible for forming monoterpenoids, diterpenoids constituents. The classical cytosolic MVA pathway and the other is the plastidial MEP pathway, which method described by Zhang et al. (2011). The cultures were incubated in the light (100 μmol m -2 s -1 ) with a 16/8h light/dark photoperiod at 24 ± 1 °C as the control. For dark treatment, fourweek-subcultured callus were covered with a cardboard box to keep in complete darkness and the samples were harvested at 24h from start of the treatment. For low temperature treatment, fourweek-subcultured callus were grown at 15 ± 1°C and samples were harvested at 48 h after the treatment for analysis of gene transcription level and TTLs content. Methyl jasmonate (MeJA) and salicylic acid (SA) with the concentration 2.0 mM were added to two-week callus cultured MS, and control cultures were untreated during cultivation. The samples was harvested and measured at 48h after MeJA and SA treatments.

Construction of genomic library and isolation of promoter region of GbHMGR gene
Genomic DNA was extracted from ginkgo seed using modified CTAB method (Xu et al., 2008a). Ginkgo Genome walker libraries were constructed using the Genome Walker Universal Kit (Clontech, USA). To clone the promoter region of GbHMGR, two round PCR were performed using gene-specific primers (GbHMGRP1 and GbHMGRP2) that were designed according to the sequence of GbHMGR gDNA (AY741133; Pang et al., 2006), and the adapter primers (AP1 and AP2) of the kit. Their sequences were shown in Table 1. After the nested PCR was carried out, amplified fragments were cloned and then sequenced. The sequences that extended upstream of the gDNA of GbHMGR were isolated as the 5'upstream region of GbHMGR gene and used for further analysis. The isolated 5'-upstream sequence was analyzed for the putative cis-acting regulatory elements using the PLACE (http://www.dna.affrc.go.jp/PLACE) and the Signal Scan Program PlantCARE..(http://bioinformatics.psb.ugent.be/webtools/plantcare /html/) database.

Expression of purification of GbWRKYs protein in E. coli
To obtain the WRKY protein, the open reading frame (ORF) of GbWRKY1 and GbWRKY2 of G. biloba (GbWRKY1 and GbWRKY2 Sequences in Supplemental Figure S1 and S2) was cloned into the expression vector pET-28a(+), yielding His-tagged GbWRKY1 and GbWRKY2. After sequence confirmation, the resulting recombinant plasmid was transferred into the E. coli strain BL21 (DE3) cells with heat shock method. A single colony of E. coli strain BL21 (DE3) cells harboring the plasmid pET28a-GbWRKY1/GbWRKY2 was inoculated in LB medium at 37 °C containing kanamycin (50 mg L -1 ), and was grown with 200 rpm shaking at 37 °C until the optical density (OD600) reached about 0.6. Expression of the recombinant protein was induced by adding isopropyl-β-D-1-thiogalactopyranoside (IPTG) and cells were harvested at 3h. The recombinant protein was extract and purified using Nickel-CL agarose affinity chromatography (Bangalore Genei, India) and used for electrophoretic mobility shift assay.
In the present study, we report the isolation and characterization of promoter region of GbHMGR gene to understand molecular regulation mechanism of GbHMGR expression underlying TTL biosynthesis in G. biloba. The transcript level of the GbHMGR and TTL content was examined after environmental stresses and plant hormone treatments. The correlation between transcript levels and TTL content in G. biloba callus further substantiated putative role of GbHMGR gene in regulating TTL biosynthesis.

Plant materials and growth condition
Seeds of ginkgo were harvested from 14-year-old grafts of G. biloba cultivar 'Jiafoshou' grown in Botanical Garden of Yangtze University, China in October, 2012. The seeds were used when the embryos were at the cotyledonary stage.

Callus induction of G. biloba and treatments
The embryo-derived callus of G. biloba was induced using the complementary labeled strands were mixed together in an equimolar ratio and annealed at 25 °C after denaturation at 90 °C. Gel mobility shift assays were performed by incubating 0.5 ng of labeled probe with recombinant GbWRKYs protein and competing oligonuleotides in binding buffer (10 mM Tris-HCl, pH 7.5, 50 mM NaCl, 1 mM dithiothreitol, 1mM EDTA, 5% glycerol, and 1μg/μL poly(dI•dC)) at 25 °C for 20 min. Mixtures were size-fractionated on a non-denaturing 46% polyacrylamide gel followed by drying and transfer to nitrocellulose membranes and detection by streptavidin-HRP/chemiluminescence for biotin-labeled probes.

Quantitative Real-Time PCR analysis of transcript levels
The transcription levels of GbHMGR of G. biloba callus were determined at different stress or hormone treatments. RNA was isolated from the ginkgo callus at different treatments using CTAB methods (Cai et al., 2007). First-strand cDNA synthesis was performed in triplicate for each sample according to the instructions of the manufacturer (PrimeScript TM RT Reagent Kit, Dalian TaKaRa, China). Quantitative Real-Time PCR (qRT-PCR) was carried out using a Applied Biosystems 7500 Real-Time PCR System with SYBR ® Premix Ex Taq™ II Kit (Dalian TaKaRa, China) according to the protocol of the manufacturer. The primers of GbHMGR and G. biloba house-keeping gene 18S (GenBank accession no. D16448) for qRT-PCR are listed in Table 1. Reactions were performed in triplicate using 10 μL of SYBR ® Premix Ex Taq II, 0.8 μL of each primer, 0.4 μL ROX Reference Dye II, 2 μL of diluted cDNA, and nuclease free water to a final volume of 20 μL. The PCR reaction conditions were preincubated at 95 °C for 30 s, followed by 30 cycles of amplification (95°C for 5 s, 60 °C for 34 s), with melt curve stage (95 °C for 15 s, 60 °C for 1 min, and 95 °C for 15 s). Fluorescence was measured at the end of each annealing step. Raw data were analyzed with Applied Biosystems 7500 software, and expression was normalized to 18S gene to minimize the variation in the cDNA template levels. Real-time PCR data were technically replicated with error bars, representing mean ± SE (n =3).

Extraction and determination of TTLs
Ginkgolide A (GA), ginkgolide B (GB), ginkgolide C (GC) and bilobalide (BB) were extracted and determined using gas chromatography with a wide bore capillary column (Liao et al., 2008). The content of TTLs was the sum of the contents of GA, GB, GC, and BB and expressed as μg·g -1 DW. All the tests were carried out in triplicate, and data represent the means ± standard errors (SE).

Statistic analysis
Data were analyzed with one-way ANOVA using SPSS 11.0 (SPSS Inc., Chicago, Illinois) for Windows and means were compared with Duncan's multiple range test at P < 0.05.

Results and discussions
Cloning and analysis of GbHMGR promoter The promoter sequence of GbHMGR was obtained by constructing a Genomic Walker DNA library from G. biloba leaves after 2 rounds of nested PCR using chromosome walking techniques. The promoter sequence of GbHMGR gene was 1,264 bp in length (Fig. 1). The cis-acting elements of GbHMGR promoter was predicted using PlantCARE and PLACE 27 database. Various cis-acting elements, along with their functions and location in the promoter of GbHMGR, are shown in Table  2 and Fig. 2. We found that the GbHMGR promoter sequence contains many important cis-regulatory elements such as TATA boxes, CAAT boxes, ACGTT boxes, W boxes, E boxes, etc. Specifically, in silico analysis showed that one TATA boxes are present within the promoter region of GbHMGR, at position 1233, i.e. 32 bp upstream of the transcription start site. The TATA box is necessary in promote gene transcription by combination to RNA polymerase II, and affects the rate of transcription (Smale and Kadonaga, 2003). Another conserved eukaryotic cis-element, CAAT boxes, were also observed at position 1109, i.e. 124 bp upstream from TATA box. The CAAT box controls the transcription initiation frequency and impacts conversion rates of target genes (Edwards et al., 1998). The changes in these basic elements within the promoter region will greatly affect the level of transcription of target genes. In addition to these essential cis-acting elements, other corresponding cis-element with roles in regulation of gene expression in the promoter region of GbHMGR were also predicted and listed in Table 2. These include cis-elements associated with hormone regulations and found in other plant gene promoters, in detail including one ABRERATCAL motif function as Ca 2+ -responsive element (Kaplan et al., 2006), one CATATGGMSAUR element involved in the induction of gene by Auxin (Xu et al., 1997), two Arabidopsis Response Regulator 1 type B (ARR1AT) transcription factor recognition sequences involved in cytokinin signaling and two sequences critical for cytokinin-enhanced binding (Sakai et al., 2000), one MYBGAHV element and three GT1 motifs known as to be GA and SA-responsive elements (Gubler et al.,1999;Zhou, 1999), respectively. We also found the presence of eight Dof (DNA binding with one finger) transcription factor recognition core sequences, which could be involved in auxin, jasmonate or ethylene responsiveness as previously reporter (Baumann et al., 1999;Yanagisawa and Schmidt, 1999;Nakano et al., 2006). In addition, the GbHMGR promoter contains cis-elements previously associated with low-temperature and light responsiveness, including three GT1 motifs (Zhou, 1999), one CRTDREHVCBF2 element (Xue, 2003) and one T box (Chan et al., 2001). Interestingly, we also found one MYB-box (MYB1AT), three W-box and four E-boxes in promoter region of GbHMGR. The W-box and MYB-box have shown to be WRKY (Eulgem et al., 2000) and MYB (Abe et al., 2003) transcription factor binding sites, respectively, involved in plant defense signaling. The consensus E-box sequence has been shown to be recognized by the basic helix-loop-helix (bHLH) proteins and involved in light regulation (Hartmann et al., 2005). The bioinformatic analysis also revealed the presence of root motif for root specific expression in GbHMGR promoter region (Elmayan and Tepfer, 1995). Previous study has shown that the transcript of GbHMGR was present specifically in roots . Three ROOT-motif elements (positions 184, 407 and 753) were identified as root-responsive ciselements, which were consistent with the expression pattern of GbHMGR gene. Based on the predictive identification of putative cis-acting element, it could be hypothesized that the transcriptional activity of the GbHMGR promoter is regulated by different signals.
Comparison of the known promoter sequences of the genes such as GbDXS and GbGGPPS (Xu et al., 2013), involved in TTL biosynthetic pathway in G. biloba, reveals that cis-elements binding Dof proteins. Dof proteins are plant-specific transcription factors with a highly conserved DNA-binding 28 domain, which presumably induces a single C2-C2 zinc finger. Zinc finger proteins like members of the transcription factor IIIA zinc finger protein family (ZCTs) have been previously isolated in medicinal plant Catharanthus roseus and were shown to act as transcriptional repressors of TDC and STR promoter activity in the regulation of induced terpenoid metabolism (Pauw et al., 2004). Dof transcription factors have been suggested to participate in the regulation of vital processes exclusive to plants such as photosynthetic carbon assimilation, light-regulated gene expression, accumulation of seed-storage proteins, germination, dormancy and response to phytohormones (Yanagisawa 2004). The relative abundance of Dof binding sites in GbHMGR promoter suggest that this specific protein could play a significant role of TTL gene expression in G. biloba.

GbWRKYs bind with the W-box sequence of GbHMGR promoter
Several reports have shown that WRKY proteins regulate the expression of gene involved in terpenoid biosynthesis by combining the W-box, and W-box was found in present in gene promoters related with MEP pathway such as LPS (Kim et al., 2012), IDSs (Kang et al., 2013), GbDXS and GbGGPPS (Xu et al., 2013). The W-box sequences predicted as TTTGAC were  Fig. S3, lans 2 and 3) and GbWRKY2 (Supplemental Fig. S3, lan 4 and 5) was expressed as a major protein product in the total celluar soluble protein. The molecular weight of the expressed recombinant GbWRKY1 and GbWRKY2 proteins was estimated to be about 36.7 and 39.7 kDa with the His-tag, respectively, the size of which were in good agreement with that predicted through bioinformatics. The interaction between GbWRKY1 and GbWRKY2 with GbHMGR promoter sequence was assayed with EMSA. No binding bands were detected with crude proteins of E. coli without or with empty vector pET28a (Fig. 3, lans 1 and 2). GbWRKY1 and GbWRKY2 specifically bind with the W-box sequence, and unlabeled probes inhibit the binding (Fig. 3). These results confirmed that GbWRKY proteins could combine to the W-box sequence of GbHMGR gene which is the target gene of WRKY protein. Some studies have also reported the WRKY proteins participated in the control of sesquiterpene and enzylisoquinoline alkaloid biosynthesis and their transcriptional induction by methyljasmonate (Xu et al., 2004;Kato et al., 2007;Ma et al., 2009). Data on EMSA suggested that GbWRKY1 and GbWRKY2 might regulate TTL accumulation through regulating the transcript level of target gene GbHMGR in G. biloba.

Effect light and low temperature on expression of GbHMGR and TTL content
As shown in Fig. 4A, GbHMGR expression level of ginkgo callus was significantly (P<0.05) higher by 281.3% in light as compared to those under dark after 24h treatment. Also, total TTL content of ginkgo callus increased by 18.0% in light as compared to the callus maintained under dark after 24 treatment (Fig. 4B). The above results implied that dark conditions might starve the plants and it is likely that gene expression and TTL content in dark could be a reflection of the effect of carbon limitation. Light would affect carbon pool through photosynthesis and the role of carbon pool in regulating secondary metabolites has been shown in Pinus sylvestris (Heyworth et al., 1998) and Hypericum perforatum (Mosaleeyanon et al., 2005). Our previous work also demonstrated that chlorocholine chloride induced the biosynthesis of TTL and expression of key genes related with ginkgolide biosynthesis by promoting the photosynthesis and carbon pool . Also, the possibility exists that light modulated gene expression independent of carbon pool (Fey et al., 2005). Moreover, GbHMGR expression was in agreement with the promoter data wherein the motifs (e.g. GT1 motif, T box, and E box) for light were present. The role of light in modulating a range of terpenoids and the corresponding transcripts has been documented, but there is no universal behavior and it varies depending upon the metabolite type and the plant species. The same gene, HMGR for example, it stimulated by light in Triticum aestivum (Aoyagi et al., 1993) and potato (Korth et al., 2000). Effect of light on the activity of HMGR promoter has been reported that explained the light mediated alteration in HMGR transcripts (Learned and Connolly, 1997;Kawoosa et al., 2010).
In the case of low temperature, both expression of GbHMGR and total TTL content were significantly (P < 0.05) higher by 143.1% and 21.0%, respectively at 15 °C as compared to those at 24 °C (Fig. 5 A and B). The up-regulation of GbHMGR by low temperature is expected because one low temperature responsive motif was identified in GbHMGR promoter region. Picrorhiza HMGR gene has also been reported to be up-regulated by low temperature due to the motif for low temperature presence in the promoter region of this gene. Likewise, our previous work also found that low temperature could up-regulated the genes such as GbPAL (Xu et al., 2008a), GbANS (Xu et al., 2008b), and GbFLS (Xu et al., 2012) involved in flavonoid biosynthesis in G. biloba. Taken together, it can be suggested that secondary metabolite production can be induced by low temperature under the temperature condition of satisfy the growth of ginkgo callus. Soitamo et al. (2008) showed that light at low temperature induced expression genes involved in synthesis of phenylpropanoids, carotenoids, and terpenoids. In Arabidopsis, light-dependent flavonoid (Fuglevand et al., 1996) and phenylpropanoid (Hemm et al., 2004) biosynthesis has been attributed to upregulation of relevant genes at transcript level, and the involvement of the enhancement of primary metabolite production. Low temperature mediated increase in secondary metabolites in Arabidopsis has mainly been attributed to the transcriptional upregulation of genes of secondary metabolism, which in turn, has been suggested due to the over-expression of relevant transcription factors at low temperature (Hannah et al., 2005). It also possible that upregulation of GbHMGR by light and low temperature increased carbon partitioning towards terpenoid metabolism resulting in higher TTL content in G. biloba.

Effect MeJA and SA on expression of GbHMGR and TTL content
Various signaling molecules that interact with their cognate receptors in the plant plasma memberane activate specific genes, which are responsible for the synthesis of alkaloids among secondary plant metabolites (Menke et al., 1999). Among the various signaling molecules, the elicitors MeJA and SA are thought to activate signal transduction pathways that stimulate expression of the enzymes, which form defense compounds such as terpenoids (Martin et al., 2003;Pu et al., 2009). Thus, this paper study the effects of MeJA and SA on expression of GbHMGR and accumulation of TTLs in ginkgo callus. As shown in Fig. 6, MeJA significantly stimulated GbHMGR expression and total TTL content by 86.5% and 23.9%, respectively, after 48 h treatment. Similarly, SA treatment at 48h caused increasing expression level of GbHMGR expression and total TTL content by 208.6% and 16.5% (Fig. 7), respectively. Response of GbHMGR expression to MeJA and SA could provide clue to the function of the enzymes. MeJA-and SAresponsive sites are present in the promoter sequence of GbHMGR ( Fig. 2 and Table 2). Production of MeJA and SA is involved in plant defense against biotic stresses such as herbivore and pathogen attacks (Robert-Seilaniantz et al., 2011). MeJA and SA interact antagonistically against each other to induce transcription of defense-related genes (Koornneef et al., 2008).  The HMGR genes were also induced by MeJA and SA in Solanum tuberosum (Choi et al., 1994), Brassica juncea (Alex et al., 2000), Artemisia annua , and S. miltiorrhiza (Liao et al., 2009;Dai et al., 2011). C. arachnoidea (Wang et al., 2014). The up-regulation of HMGR expression by MeJA and SA is of particular interest because of the possible limiting role of this enzyme in terpenoid synthesis. However, MeJA and SA are not specific inducer of the GbHMGR gene. They also induces other genes contributing to terpenoid biosynthesis. In G. biloba, genes IDS, DXS, CMK have also been reported to be positively responsive toward MeJA and SA treatment (Gong et al., 2006;Kim et al., 2008a, b). They suggested that the positive response implied the involvement of the gene in ginkgolide biosynthesis. Therefore, it is considered that the increased content of TTLs by MeJA or SA in present study is due to an integrated effect on a cluster of genes related to TTL biosynthesis. Further work will be required to study the effect of MeJA and SA on expression of more related genes involved in TTL biosynthesis.

Relationship between GbHMGR expression and TTL accumulation
To further study relationship between of GbHMGR expression and TTL accumulation. We performed linear regression analysis of gene expression level and total TTL content data in ginkgo callus subjected to the light, low temperature, MeJA and SA treatments. The results showed that the relationship between total TTL content (y) and GbHMGR expression level (x) was significantly positively linearly correlated (Fig. 8), with a correlation coefficent is r 2 = 0.583, and a linear curve equation represented by y = (0.0073 ± 0.0014 ) x + (2.098 ± 0.1112). These results indicated that the GbHMGR gene is responsible for the TTL accumulation and might played a crucial role in TTL biosynthesis.

Conclusions
In summary, the present study isolated and characterized the MVA pathway GbHMGR gene promoter from G. biloba. Functional and bioinformatic analyses revealed that the GbHMGR promoter contains a number of cis-motifs that could bind transcription factors involved in regulation of TTLs. EMSA analysis suggested possible functionality of W-  box in GbHMGR promoter region. The behavior of gene transcripts in ginkgo callus upon light, low temperature, MeJA and SA treatments further verified the regulatory function of GbHMGR promoter. A positive relationship between gene expression level and total TTL contents implied that GbHMGR might be one of key genes involved in TTL biosynthesis in G. biloba. Studies on gene promoters in terpenoid biosynthetic pathway constitute a valuable approach to identify new regulatory factors and/or families that could controlled the TTL biosynthesis in G. biloba cell culture.