Metabolomics Integrated with Transcriptomics Reveals Redirection of

Subscriber access provided by Macquarie University

Omics Technologies Applied to Agriculture and Food

Metabolomics Integrated with Transcriptomics Reveals Redirection of the Phenylpropanoids Metabolic Flux in Ginkgo biloba Jie Meng, Bo Wang, Guo He, Yu Wang, Xianfeng Tang, Shumin Wang, Yubin Ma, Chunxiang Fu, Guohua Chai, and Gongke Zhou J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.8b06355 • Publication Date (Web): 25 Feb 2019 Downloaded from http://pubs.acs.org on February 26, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 38

Journal of Agricultural and Food Chemistry

Metabolomics integrated with transcriptomics reveals redirection of the phenylpropanoids metabolic flux in Ginkgo biloba Jie Meng1,2, Bo Wang2, Guo He2, Yu Wang2, Xianfeng Tang1,2, Shumin Wang1,2, Yubin Ma2, Chunxiang Fu2, Guohua Chai1,2*, Gongke Zhou1,2* Affiliation: 1, College of Resources and Environment, Qingdao Agricultural University, Qingdao 266109, China. 2, Qingdao Engineering Research Center of Biomass Resources and Environment, Qingdao Institute of Bioenergy and Bioprocess Technology, Chinese Academy of Sciences, Qingdao, China * Corresponding

authors: Guohua Chai, [email protected]; Gongke Zhou [email protected].

Tel: +86-532-80662731. Fax: +86-532-80662778.

Jie Meng: [email protected]; Bo Wang: [email protected]; Guo He：[email protected]; Yu Wang: [email protected]; Xianfeng Tang: [email protected]; Shumin Wang: [email protected]; Yubin Ma: [email protected]; Chunxiang Fu: [email protected]; Guohua Chai: [email protected]; Gongke Zhou: [email protected].

ACS Paragon Plus Environment


1

Abstract

2

Ginkgo biloba is monotypic species native to China with great economic and

3

ecological values. Leaves extract of this tree contains about 24% flavonoids, which

4

are widely used for pharmaceutical industry. However, the flavonoids biosynthesis

5

pathway is poorly understood in Ginkgo. In this study, we comprehensively compared

6

the transcriptome and metabolite profiles of Ginkgo high-flavonoids mutant (ZY1)

7

and Anlu1 (control) leaves. A total of 122 significantly changed metabolites (SCMs)

8

and 1683 differentially expressed genes (DEGs), including 45 transcription factors,

9

were identified in ZY1 compared with Anlu1. An integrated analysis of metabolic and

10

transcriptomic data revealed that the abundances of some major flavonoids (especially

11

flavone and flavonol) were most significantly increased, while other

12

phenylpropanoid-derived products and lipids showed the most largely reduced

13

abundances in ZY1 compared to that in Anlu1. qRT-PCR results confirmed the

14

alterations in the expression levels of genes encoding components of pathways

15

involved in phenylpropanoids and lipids. The redirection of metabolic flux may

16

contribute to increased accumulation of flavonoid levels in ZY1 leaves. Our results

17

provide valuable information for metabolic engineering of Ginkgo flavonoids

18

biosynthesis.

19

KEYWORDS: metabolomics, transcriptomics, phenylpropanoids metabolic flux,

20

flavonoids biosynthesis, Ginkgo biloba


Page 2 of 38

Page 3 of 38


21

Introduction

22

Ginkgo (Ginkgo biloba L.) is one of the oldest living tree species, with distinguishing

23

features such as slow growing and strong resistance to environmental stresses,

24

microbial disease and pests. For thousands of years, Ginkgo is thought to be an

25

important medicinal tree, because its leaves and nuts possess many pharmacological

26

properties including flavonoids and terpene lactones, which have been widely used for

27

treating and protecting from various illnesses 1. However, the mechanism underlying

28

flavonoids biosynthesis in Ginkgo remains unclear.

29

Flavonoids are extensively distributed in the plant kingdom and their

30

biosynthesis pathways have been relatively well elucidated in the Arabidopsis seed 2.

31

Flavonoid scaffolds are formed from the building blocks of a phenylpropanoid primer

32

(p-coumaroyl-CoA) and polyketide condensing unit (malonyl-CoA) by a serial of

33

reactions, including condensation, isomerization, oxidation and reduction. The

34

scaffolds are further modified to various sub-classes of flavonoids by different classes

35

of enzymes. Functional analysis of Arabidopsis mutants identifies a battery of genes,

36

which encode key components of the flavonoids biosynthesis pathway and its

37

constituent enzymes. These genes mainly include CHALCONE SYNTHASE (CHS),

38

CHALCONE ISOMERASE (CHI), FLAVANONE 3-HYDROXYLASE (F3H),

39

FLAVONOID 3’HYDROXYLASE (F3’H), ISOFLAVONE SYNTHASE (IFS),

40

FLAVONOL SYNTHASE (FLS), DIHYDROFLAVONOL 4-REDUCTASE (DFR),

41

LEUCOANTHOCYANIDIN REDUCTASE (LAR), ANTHOCYANIDIN REDUCTASE

42

(ANR), and ANTHOCYANIDIN SYNTHASE (ANS). Compared with the single-copy

43

genes in Arabidopsis, Ginkgo employs multigene families to control each step of the

44

flavonoid biosynthesis pathway, resulting in a more complex network 3. Although the

45

catalyzed activities of several Ginkgo homologs (GbCHS, GbCHI, GbF3H, GbFLS

46

and GbANS) are characterized in Escherichia coli 4-8, it is still unclear how these



47

enzymes synthesize flavonoid compounds in planta due to lack of related mutants.

48

Transcriptome analysis of Ginkgo tissues has successfully identified genes that

49

are involved in ginkgolides and bilobalide biosynthesis 9,10, sex determination 11 and

50

circadian rhythms-mediated regulation of flavonoid contents 12. In particular, Ni et al.

51

showed that different flavonoids varied their contents differently during the day–night

52

cycles 12. Genes involved in the flavonoids biosynthesis were down-regulated at

53

midnight compared with that at midday. However, no systematic analysis of the

54

alterations of flavonoid biosynthesis genes expression is conducted in Ginkgo leaves.

55

Recently, application of metabolomics to medicinal plants has significantly facilitated

56

the identification of the metabolic pathways of the active medicinal compounds in

57

plants 13. Furthermore, the integration of transcriptomics and metabolomics has larger

58

advantages to reveal the biosynthetic mechanisms of key metabolic pathways 14,15.

59

Therefore, it is feasible to analyze the flavonoid biosynthesis pathway in Ginkgo

60

through joint application of the two technologies.

61

We recently cultivated a high-flavonoids Ginkgo variety ZY1, whose leaves

62

extract contains higher (28%) level of flavonols than Anlu1 (CK, a widely planted

63

variety in China) 16. In this study, we further clarified the mechanism of enhanced

64

flavonoids levels in ZY1 leaves by integrating the transcriptome and metabolome

65

data. The levels of some major flavonoid compounds (especially flavone and

66

flavonol) were drastically increased, whereas other phenylpropanoid-derived products

67

and fatty acids showed lower abundances in ZY1, compared with CK. These were

68

correlated with the alterations in the expression levels of genes associated with these

69

metabolic processes. Our current finding may be useful for metabolic engineering of

70

Ginkgo flavonoids biosynthesis in host cells.


Page 4 of 38

Page 5 of 38


71

Materials and Methods

72

Plant Materials

73

The Ginkgo varieties “Anlu1” (control) and “ZY1” (bred from “Anlu1”) grafted on

74

three-year-old rootstock (the variety “Damaling”) were grown in the field (Laoshan

75

district, Qingdao, China; 36°09′13″N, 120°28′53″E). After two years of growth,

76

fifteen leaves in three trees for each variety were sampled at middle part of the main

77

stems in October, 2017. Each leaf was split into two equal parts along the mid vein:

78

one half of the halves was used for metabolomic analysis and the other half for

79

transcriptomic analysis. All the materials were frozen in liquid nitrogen immediately

80

and then stored at -80°C until used.

81

Metabolite Extraction

82

The frozen leaves were crushed using a bead beater (1.5 min, 30 Hz, three repetitions,

83

MM 400, Retsch). 100 mg of the powdered sample was extracted overnight at 4 °C

84

using 1 mL 70% aqueous methanol containing 0.1 mg L-1 lidocaine. After

85

centrifugation at 10,000 g for 10 min, the supernatants were filtrated by using 0.22-

86

μm hydrophilic PTFE syringe filters (SCAA-104, ANPEL, Shanghai, China) before

87

metabolomics analysis. The quality control samples (mix1-3) were injected every two

88

experimental samples throughout the analytical run to provide a set of data from

89

which repeatability could be assessed.

90

Metabolite Profiling Using Liquid Chromatography-Electrospray Ionization -

91

Tandem Mass Spectrometry (LC-ESI-MS/MS)

92

Metabolite profiling was conducted using a LC-ESI-MS/MS system (HPLC, UFLC

93

SHIMADZU CBM30A system; MS, Applied Biosystems 4500 Q TRAP) and an

94

Agilent 6520 accurate-mass time-of-flight mass spectrometry. Chromatographic



95

separation was performed on an ACQUITY UPLC HSS T3 C18 column (2.1

96

mm×100 mm×1.8 μm; Waters) using mobile phase A (0.04% acetic acid in deionized

97

water) and mobile phase B (0.04% acetic acid in acetonitrile). The elution profile was

98

used as follow: 95:5 V(A)/V(B) at 0 min, 5:95 V(A)/V(B) at 11.0 min, 5:95

99

V(A)/V(B) at 12.0 min, 95:5 V(A)/V(B) at 12.1 min, 95:5 V(A)/V(B) at 15.0 min.

100

The flow rate was maintained at 0.4 mL min-1.

101

Mass data acquisition was performed in electrospray ionization-positive/negative

102

mode using the following parameters: ion spray voltage of (+/-)5.5 kV; ion source gas

103

I (GSI) of 55 psi; gas II (GSII) of 60 psi; curtain gas (CUR) of 25 psi; turbo spray

104

temperature of 550 °C. Instrument tuning and mass calibration were performed with

105

10 and 100 μmol L-1 polypropylene glycol solutions in triple quadrupole (QQQ) and

106

linear ion trap (LIT) modes, respectively. Declustering potential (DP) and collision

107

energy (CE) for individual multiple reaction monitoring (MRM) transitions were

108

performed with specific DP and CE optimization. A specific set of MRM transitions

109

were monitored for each period based on the metabolites eluted within this period.

110

Qualitative and Quantitative Analysis of Metabolites

111

To facilitate the identification of metabolites by widely targeted metabolomics

112

approach (MetWare, Wuhan, China), accurate m/z value of each precursor ions (Q1)

113

was obtained. This method has been previously described 14,17. In brief, metabolites

114

were identified by comparing the m/z values, the retention time (RT) and the

115

fragmentation patterns with the standards in a self-compiled database (MetWare).

116

Significantly changed metabolites (SCMs) were filtered according to |Log2 (fold

117

change)| ≥1, p-value < 0.05. Principle component analysis (PCA) of SCMs was

118

performed by R (www.r-project.org/) to study metabolite variety-specific accumulation

119

18.


Page 6 of 38

Page 7 of 38


120

RNA Extraction and Illumine Sequencing

121

Total RNAs were isolated from frozen leaves using the modified CTAB method 19.

122

Quality and concentration of RNAs were determined using a Nanodrop 2000

123

spectrophotometer (Thermo Scientific, Wilmington, USA) and an Agilent 2100

124

Bioanalyzer (Agilent Technologies, Santa Clara, USA). RNA purification, library

125

construction and RNA-seq were performed by Biotree Bio-technology Co., Ltd.

126

(Shanghai, China). Briefly, a library per sample was generated by 3 ug RNA and then

127

sequenced using Illumina HiSeq4000 (6G 150bp paired-end reads). Sequence data

128

with base-pair qualities in the Q ≥ 20 were extracted by custom Perl scripts. The

129

filtered reads were mapped to a Ginkgo reference genome

130

(http://gigadb.org/dataset/100209) using the TopHat2 software with default

131

parameters 20. FPKM was used for gene/transcript level quantification. Based on the

132

raw count data, edgeR package (http://www.bioconductor.org/packages/release/bioc

133

/html/edgeR.-html) was adopted to identify differentially expressed genes (DEGs)

134

using the following criteria: (1) FPKM of CK or ZY1 >1; (2) false discovery rate

135

(FDR) < 0.05; (3) Log2FC is greater than 1 or less than -1. DEGs were subjected to

136

Gene ontology (GO) and pathway analysis. GO analysis was conducted using the

137

GOseq R package based on biological process categories (corrected P-values < 0.05).

138

Pathway analysis was conducted to elucidate significant pathways of DEGs according

139

to the Kyoto Encyclopedia of Gene and Genomes (KEGG)

140

(http://www.genome.jp/kegg) databases. The RNA-Seq data set is available in the

141

Gene Expression Omnibus (accession No.--).

142

Integrative Analysis of Metabolome and Transcriptome

143

Metabolites and DEGs involved in phenylpropanoids biosynthesis and lipids

144

metabolism in KEGG pathways were selected for integrative analysis. Metabolites

145

used for correlation analysis were filtered according to Variable Importance in the



146

Project (VIP) > 1, p-value < 0.05 and |Log2(Fold Change)|≥1. Pearson correlation

147

coefficients and p-values were calculated for metabolome and transcriptome data

148

integration using the Spearman method 21.

149

Quantitative Realtime RT-PCR (qRT-PCR)

150

Total RNAs isolation, DNase I digestion and first-strand cDNA synthesis were

151

performed previously described 19. Reactions were carried out on a Lightcycler 480

152

Real-Time PCR Detection System (Roche, Germany) using transStart® Top/Tip

153

Green qPCR (TransGen Biotech, China). The amplification cycling program was as

154

follows: 94 °C for 30 s, 45 cycles of 94 °C for 5 s, 59 °C for 15 s and 72 °C for 10 s.

155

Primers were listed in Supplemental Table 6. The relative expression was calculated

156

using the 2−ΔΔCT method 22. The 18S gene (Gb_32799) was used as an internal control.

157

Data from three independent biological replicates were analyzed by one-way analysis of

158

variance. P-values of < 0.05 and < 0.01, calculated using Dunnett’s test, were regarded as

159

statistically significant and highly significant, respectively.

160

Measurement of Lignin Content

161

Alcohol insoluble residues (AIR) were prepared from the leaves that are used for

162

metabolomic and transcriptomic analysis following our previous method 23. To detect

163

lignin content, about 3 mg AIRs were solubilized by acetyl bromide solution, and 2 M

164

sodium hydroxide and 0.5 M hydroxylamine hydrochloride were added to stop the

165

reaction. Absorbance at 280 nm was recorded by an UV-visible spectrophotometer of

166

VARIAN Cary 50 (VARIAN, USA). Percentage of acetyl bromide soluble lignin (%

167

ABSL) was calculated by the formula (% ABSL = 0.236 × absorbance at 280 nm /

168

weight of AIRs).


Page 8 of 38

Page 9 of 38


169

Results

170

Metabolic Differences between the Leaves of ZY1 and CK

171

To compare the differences of metabolite compositions in two G. biloba varieties

172

ZY1 and Anlu 1 (CK) (Supplemental Figure 1), the leaves sampled in October were

173

subjected to LC-ESI-MS/MS analysis. In this work, 702 metabolites including 200

174

flavonoids and 81 amino acids, two types of major medical components, were

175

identified and quantified in Ginkgo leaves (Supplemental Figure 2). PCA analysis

176

showed that ZY1 and CK were clearly separated in the PC1×PC2 score plots (Figure

177

1A). Volcano plot of the metabolite contents indicated the significant differences

178

between these two varieties (Figure 1B). Of the 702 metabolites, 122 (17.4%) were

179

markedly changed in ZY1 compared with CK, with 62 metabolites being up-regulated

180

and 60 metabolites being down-regulated (Figure 1C; Supplemental Table 1). The 122

181

significantly changed metabolites (SCMs) were generally divided into 22 categories,

182

including flavonoids, other phenylpropanoid-derived products, lipids, amino acids,

183

and nucleotides and its derivatives. Interestingly, most of flavonoids abundances were

184

markedly increased, while other phenylpropanoid-derived products and lipids were

185

reduced in ZY1, compared to that in CK (Supplemental Table 1a). This was

186

consistent with our previous HPLC analysis of flavonoids contents in ZY1 and CK16.

187

Differential Expression of Genes between ZY1 and CK

188

To identify genes differentially expressed in the leaves of ZY1 compared with CK, a

189

transcriptomic comparison of the materials mentioned above was carried out. A total

190

of 40.4-49.8 (98.3%-98.4% of total clean reads) million clean reads were obtained

191

through sequencing of the cDNA libraries after stringent quality checks and data

192

cleanup. 39.1–48.7 million reads were mapped to the Ginkgo genomic database

193

(http://gigadb.org/dataset/100209), with match ratios in the range of 88.1 -91.5%.



194

Comparing the two types of library with respect to the FPKM calculation, 26104 and

195

26173 genes were detected in the ZY1 and CK leaves. High correlation coefficients

196

(R2 > 0.99) of gene expression between biological replicates indicated the

197

effectiveness of data (Supplemental Figure 3). With the filter criteria of fold change ≥

198

2 and FDR < 0.05, 1683 DEGs were identified in ZY1 compared with CK, of which

199

739 and 944 genes were up-regulated and down-regulated, respectively (Supplemental

200

Table 1b). GO analysis of the 1683 DEGs showed enrichment of five major biological

201

processes, including single-organism process, metabolic process and cellular process.

202

In the molecular function aspect, these DEGs were enriched in catalytic activity and

203

binding. In the cellular component aspect, most of DEGs were enriched in four

204

categories including membrane, organelle, and cell part (Figure 2A). KEGG pathway

205

enrichment analysis (Q-value < 0.05) revealed that these DEGs were mainly enriched

206

in several metabolic processes including phenylpropanoid biosynthesis, lipid

207

metabolism, and amino acid metabolism (Figure 2B; Supplemental Table 2).

208

Alterations of Phenylpropanoid and Lipid Biosynthesis in the Leaves of ZY1

209

Compared with CK

210

To understand the regulatory mechanism of enhanced flavonoids biosynthesis in the

211

leaves of ZY1, 72 metabolites (flavonoid biosynthesis: 43, other phenylpropanoid -

212

derived products: 15, and lipid metabolism: 14) and 44 DEGs (flavonoid biosynthesis:

213

10, other phenylpropanoid-derived products: 16, and lipid metabolism: 18) were

214

selected to carried out correlation tests. PCA results showed strong positive and

215

negative correlation (r > 0.8 or

Metabolomics Integrated with Transcriptomics Reveals Redirection of

Recommend Documents