Stability of the Metabolite Signature Resulting from the OsSULTR3;3

Subscriber access provided by University of South Dakota

Omics Technologies Applied to Agriculture and Food

Stability of the Metabolite Signature Resulting from the OsSULTR;3 Mutation in Low Phytic Acid Rice (Oryza sativa L.) Seeds upon Cross-breeding Chenguang Zhou, Yuanyuan Tan, Sophia Gossner, Youfa Li, Qingyao Shu, and Karl-Heinz Engel J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.8b03921 • Publication Date (Web): 15 Aug 2018 Downloaded from http://pubs.acs.org on August 22, 2018

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 34

Journal of Agricultural and Food Chemistry

Stability of the Metabolite Signature Resulting from the OsSULTR;3 Mutation in Low Phytic Acid Rice (Oryza sativa L.) Seeds upon Cross-breeding

Chenguang Zhou1, Yuanyuan Tan2, Sophia Goßner1, Youfa Li3, Qingyao Shu2, and Karl-Heinz Engel1*

1

Chair of General Food Technology, Technical University of Munich,

Maximus-von-Imhof-Forum 2, D-85354 Freising-Weihenstephan, Germany 2

State Key Laboratory of Rice Biology and Zhejiang Provincial Key

Laboratory of Plant Germplasm, Institute of Crop Sciences, Zhejiang University, Hangzhou, China 3

Jiaxing Academy of Agricultural Sciences, Jiaxing, China

* Corresponding Author (telephone: +49 8161 714250; fax +49 8161 714259; e-mail: [email protected])

1 ACS Paragon Plus Environment


1

ABSTRACT

2

The low phytic acid (lpa) rice (Oryza sativa L.) mutant Os-lpa-MH86-1, resulting

3

from the mutation of the putative sulfate transporter gene OsSULTR3;3, was

4

crossed with a commercial rice cultivar. The obtained progenies of generations F4

5

to F7 were subjected to a non-targeted metabolite profiling approach allowing the

6

analyses of a broad spectrum of lipophilic and hydrophilic low molecular weight

7

constituents. The metabolite profiles of the homozygous lpa progenies were

8

characterized not only by a decreased concentration of phytic acid, but also by

9

increased contents of constituents from various classes, such as sugars, sugar

10

alcohols, amino acids, phytosterols and biogenic amines. Statistical assessments

11

of the data via multivariate and univariate approaches demonstrated that this

12

mutation-induced metabolite signature was nearly unaffected by the cross-

13

breeding step and consistently expressed over several generations. The data

14

demonstrate that even for complex metabolic changes resulting from a mutation,

15

cross-breeding can be employed as a tool to generate progeny rice seeds stably

16

exhibiting the mutation induced traits.

17 18

Key words: low phytic acid mutant, metabolite profiling, rice (Oryza sativa L.),

19

cross-breeding, OsSULTR3;3

20


Page 2 of 34

Page 3 of 34


21

INTRODUCTION

22

Phytic acid (myo-inositol-1,2,3,4,5,6-hexakisphosphate) constitutes the major

23

storage form of phosphorus (P) in cereal grains.1 Owing to the chelation of

24

divalent cations it may reduce the bioavailability of nutritionally important

25

minerals such as Zn2+ and Fe2+.2 Additionally, the excretion of undigested

26

phytate from the manure of monogastric animals contributes to P-pollution of

27

the environment.3 Therefore, during the past years several low phytic acid (lpa)

28

crops have been developed using genetic engineering as well as mutation

29

breeding via chemical mutagenesis and γ-irradiation.4

30

Induced mutations in lpa crops have been shown to result not only in the

31

intended decreased levels of phytic acid but also in other metabolic changes.

32

In lpa soybean seeds, for example, decreased levels of myo-inositol, stachyose,

33

raffinose and galactosyl cyclitols and increased levels of sucrose compared to

34

the wild-type cultivar have been detected via targeted analysis5 as well as via

35

metabolite profiling.6 Application of metabolite profiling to lpa rice mutants

36

also revealed consistent changes of the contents of myo-inositol, raffinose,

37

galactose and galactinol compared to the wild-type cultivars.7

38

Recently, it has been demonstrated that the disruption of OsSULTR3;3, an

39

ortholog of the sulfate transporter family group 3 gene, resulted not only in a

40

pronounced reduction of the phytic acid content in the lpa rice mutant Os-lpa-

41

MH86-1, but also in significant changes of the metabolite profile of the lpa

42

rice grain compared with the corresponding wild-type MH86.8 The mutation-

43

induced metabolite signature comprised altered levels of a broad spectrum of

44

constituents, e.g. reduced content of cysteine, increased concentrations of

45

various amino acids (e.g. serine, threonine and isoleucine), organic acids (e.g.



46

citric acid) and other nutritionally relevant compounds, such as γ-aminobutyric

47

acid (GABA).

48

Lpa mutant crops often show inferior agronomic performance, e.g. lower

49

grain yield and seed viability compared with the wild-types, and therefore

50

approaches such as cross and selection breeding are applied to minimize these

51

negative effects.9 Studies with soybean (Glycine max L. Merr.) lines showed

52

that the field emergence of lpa progenies was significantly improved by

53

crossing parental lpa lines with elite cultivars.10, 11

54

However, information on the potential impact of the cross-breeding of lpa

55

mutants with commercial cultivars on the metabolite profiles of the resulting

56

lpa mutant progenies is lacking. Therefore, based on the application of a non-

57

targeted metabolite profiling approach, the objective of this study was to use

58

the lpa rice mutant Os-lpa-MH86-1 as example (i) to investigate the impact of

59

cross-breeding with a commercial cultivar on the metabolic phenotype of the

60

homozygous lpa mutant, and (ii) to assess the stability of the mutation-specific

61

metabolite signature in the lpa progenies over several generations.

62 63

MATERIALS AND METHODS

64

Chemicals. Internal standards tetracosane, 5α-cholestan-3β-ol, phenyl-β-D-

65

glucopyranoside and 4-chloro-L-phenylalanine were purchased from Fluka

66

(Buchs, Switzerland). Reference compounds were supplied by VWR

67

International (Darmstadt, Germany), Fluka (Buchs, Switzerland), Sigma-

68

Aldrich (Steinheim, Germany) and Roth (Karlsruhe, Germany). All other

69

reagents and solvents were obtained from VWR International (Darmstadt,


Page 4 of 34

Page 5 of 34


70

Germany), Sigma-Aldrich (Steinheim, Germany) and Merck (Darmstadt,

71

Germany) as HPLC grade.

72

Development of sample materials. The lpa rice mutant Os-lpa-MH86-1

73

has been previously generated from the corresponding wild-type MH86 via γ-

74

irradiation.12 The mutation of Os-lpa-MH86-1 was attributed to a 1-bp

75

deletion of the putative sulfate transporter gene (OsSULTR3;3).8 In the present

76

study, the wild-type line Jiahui 99 (JH99), the restorer line of the hybrid

77

cultivar Jiayou 99, was used to cross with Os-lpa-MH86-1 and to generate

78

wild-type and lpa progenies, which were used for evaluation of the effect of

79

genetic background on the metabolite signature of the lpa mutants. F2 seeds

80

were bulk-harvested and grown into plants. Two hundred F2 plants were

81

genotyped for the lpa mutation; consequently, F2 plants were classified into

82

three types, i.e. homozygous wild-type, homozygous lpa type, and

83

heterozygous progenies (Figure 1). Plants were also segregating for other

84

agronomic traits, e.g. flowering time, plant height and fertility; therefore, five

85

F2 plants each of the homozygous and 10 each of the heterozygous genotype

86

were chosen based on similarity of flowering time, and their F3 seeds were

87

harvested. While seeds of homozygous F2 plants were stored for growth until

88

2013, those from heterozygous F2 plants were grown into F2:3 plots and their

89

F3 plants were genotyped for the lpa mutation and classified into three types as

90

in F2 (Figure 1). Similar work was performed in F4-F6, and seeds harvested

91

from homozygous F2 to F5 plants were grown in a common paddy field in

92

2013 in Jiaxing for the production of sample materials (F4 to F7, as well as the

93

two parental lines) for metabolite profiling (Figure 1).



94

For the genotyping, genomic DNA was extracted from leaf tissues

95

following a modified CTAB method, and high-resolution melting curve

96

analysis was utilized to distinguish among homozygous wild-types,

97

homozygous lpa mutants and heterozygous progenies, according to a

98

previously described procedure.13

99

The homozygous F4 to F7 seeds, as well as the original wild-type MH86,

100

the progenitor lpa mutant Os-lpa-MH86-1 and the crossing parent JH99 were

101

harvested at a field trial in Jiaxing, 2013. For each generation of the progenies,

102

seeds from 3 to 4 plants were available for analysis. Owing to the limited

103

amount of the sample materials, for each generation rice seeds from 3 to 4

104

plants of each phenotype were pooled. After dehulling and grinding with a

105

cyclone mill equipped with a 500 µm sieve (Cyclotec, Fodd, Germany), the

106

flour was immediately freeze-dried for 48 h and then stored at -20 °C until

107

analysis.

108

Phytic acid content analysis. Extraction and analysis of phytic acid were

109

performed according to a procedure described for dried distillers grains.14 For

110

each sample, three aliquots (100 mg) of freeze-dried rice flour were weighed

111

into 50 mL polystyrene centrifuge tubes. Twenty milliliters of 0.5 M HCl were

112

added to each aliquot. The extraction of phytic acid was performed under

113

sonication for 20 min after the mixture had been vortexed thoroughly. After

114

centrifugation at 2000 g for 20 min, the supernatants were filtered through a

115

0.22 µm polyether sulfone filter disk, and 100 µL were subjected to ion

116

chromatography.

117

The analysis of phytic acid was performed on a Thermo Scientific Dionex

118

(Dreieich, Germany) ICS-5000 HPIC (high pressure ion chromatography)


Page 6 of 34

Page 7 of 34


119

system equipped with a CarboPac PA100 guard column (4 x 50 mm) and a

120

CarboPac PA100 analytical column (4 x 250 mm) constantly thermostatted to

121

30 °C. The flow rate of the column was 1 mL/min and a gradient elution was

122

performed by two mobile phases, i.e. deionized water and 0.5 M HCl. The

123

initial proportion of gradient HCl of 5% was increased linearly to 100% within

124

4 min, and then held for 10 min. The flow rate of post-column derivatization

125

with ferric nitrate (solution of 0.1% Fe(NO3)3 in 0.33 M perchloric acid) was

126

0.4 mL/min, and peak detection was performed at 290 nm. Quantification of

127

phytic acid was based on an external standard calibration curve (R2 = 0.9997);

128

the recovery rate was 98.8%, the limit of detection and the limit of

129

quantification were 0.7 mg/L and 2.2 mg/L, respectively.15

130

Metabolite Profiling. Extraction and fractionation of the rice flour were

131

performed as previously described.7 Briefly, for each sample, three aliquots

132

(600 mg) of freeze-dried rice flour were weighted into empty cartridges (3 mL

133

volume; Supelco, Munich, Germany). The lipophilic constituents of the flour

134

were eluted with 4 mL of dichloromethane. The polar constituents were eluted

135

with 10 mL of a mixture of methanol and deionized water (80 + 20, v/v). After

136

adding 50 µL and 100 µL of internal standard solution I (1.5 mg/mL of

137

tetracosane in hexane) and II (0.6 mg/mL of 5α-cholestan-3β-ol in

138

dichloromethane) to the lipophilic extract, respectively, the lipids were

139

transesterified with sodium methylate and separated by solid-phase extraction

140

into fraction I containing fatty acid methyl esters (FAME) and hydrocarbons,

141

and fraction II containing minor lipids (free fatty acids, fatty alcohols and

142

sterols). Prior to fractionation of the polar constituents, 50 µL and 100 µL of

143

internal standard solution III (1.6 mg/mL of phenyl-β-D-glucopyranoside in



144

deionized water) and IV (0.8 mg/mL of 4-chloro-L-phenylalanine in deionized

145

water) were added to the polar extract, respectively. Fraction II and IV were

146

silylated with 50 µL of N-methyl-N-(trimethylsilyl)-trifluoroacetamide by

147

heating at 70 °C for 15 min, fraction III with 100 µL trimethylsilylimidazole

148

by heating at 70 °C for 20 min. Selective hydrolysis of silylated derivatives

149

was applied to separate the polar extracts into fraction III (sugars and sugar

150

alcohols) and fraction IV (acids, amino acids and amines). The obtained four

151

fractions were analyzed by capillary gas chromatography (GC) coupled with a

152

flame ionization detector and a mass spectrometer (MS), respectively, under

153

the previously described conditions.7

154

The identification of metabolites was based on the comparison of retention

155

times and mass spectra with those of reference compounds, with data from the

156

NIST08 mass spectra library and the literature.16, 17 The amounts of identified

157

metabolites from fractions I-IV were expressed as relative peak intensities (i.e.

158

metabolite peak intensity / internal standard peak intensity × 100) based on the

159

internal standards.

160

Quality Control. Commercially available rice seeds were analyzed

161

regularly together with actual samples as reference material. The results of the

162

quality control analysis were evaluated based on the reproducibility of selected

163

representative compounds from each fraction: C16:0 FAME, C18:0 FAME

164

and squalene for fraction I, C16:0 FFA, C18:0 FFA, campesterol and β-

165

sitosterol for fraction II, sucrose and raffinose for fraction III and serine,

166

aspartic acid, glutamic acid and citric acid for fraction IV. In accordance with

167

previous studies,18, 19 the data were considered acceptable when the relative

168

standard deviation (RSD) of these selected compounds did not exceed 25%.


Page 8 of 34

Page 9 of 34


169

Data Processing and Statistical Analysis. Chromatographic data were

170

acquired and integrated by Chrom-Card 2.3 (Thermo Electron). Peaks below

171

noise level were discarded on the basis of a threshold of 1 % peak height

172

relative to the internal standards. Standardization of peak heights and retention

173

time

174

(http://www.chrompare.com).20 For data pretreatment, log transformation and

175

Pareto scaling were carried out.21 Principal component analysis (PCA) and

176

heat map-analysis were conducted using XLSTAT (version 19.5, France).

177

Orthogonal Partial Least Squares - Discriminant Analysis (OPLS-DA) was

178

performed

179

http://www.metaboanalyst.ca/faces/home.xhtml). The OPLS-DA model was

180

validated by a 10-fold cross validation and a permutation test (1000 times).22

181

In accordance with a guideline on the univariate statistical analysis of

182

metabolomics-derived data,23 Student’s t-test (p < 0.05) and ANOVA with

183

Tukey’s post hoc test (p < 0.05) were performed for those metabolites

184

exhibiting a normal distribution and homogeneity of variance. Only for

185

phosphoric acid, which did not fulfill these requirements, the non-parametric

186

Mann-Whitney test (p < 0.05) and the Kruskal-Wallis test with Dunn’s post

187

hoc test (p < 0.05) were performed by XLSTAT.

alignment

using

were

the

performed

web-based

tool

by

Chrompare

MetaboAnalyst

(version

1.1

4.0;

188 189

RESULTS

190

Multivariate analysis of rice seed metabolite profiles

191

Homozygous lpa mutant, homozygous wild-type and heterozygous progenies

192

of F2-F6 plants resulting from the cross of the lpa mutant Os-lpa-MH86-1 with

193

the commercial rice cultivar JH99 were genotyped by high-resolution melting



194

curve analysis.13 Seeds of homozygous wild-type and homozygous lpa mutant

195

progenies of generations F4 - F7 were obtained from the same field trial

196

(Figure 1) and subjected to metabolite profiling. The previously established

197

GC/MS-based approach7,

198

hydrophilic low molecular weight rice constituents and resulted in four

199

fractions containing fatty acid methyl esters (FAMEs) and hydrocarbons

200

(fraction I), free fatty acids (FFAs), fatty alcohols and sterols (fraction II),

201

sugars and sugar alcohols (fraction III), and amino acids, acids and amines

202

(fraction IV). A total of 279 peaks were detected; 118 metabolites were

203

identified by comparison of retention times and mass spectral data to those of

204

reference compounds and/or data from mass spectra libraries (Tables S1 and

205

S2, Figure S2).

24

covered a broad spectrum of lipophilic and

206

A PCA score plot based on the combined fractions I – IV revealed a clear

207

separation not only of the lpa progenitor Os-lpa-MH86-1 and the

208

corresponding wild-type MH86, but also a comparable and consistent

209

separation of the homozygous lpa mutant and the homozygous wild-type

210

progenies from generations F4 to F7 (Figure 2A). The homozygous lpa mutant

211

progenies of generation F4 clustered closely to the progenitor lpa mutant and

212

the wild-type progenies closely to the original wild-type. In addition, the

213

crossing parent JH99 clustered together with the wild-type MH86 and the F4

214

wild-type progenies, and exhibited a clear separation from Os-lpa-MH86-1

215

and F4 lpa progenies. These results indicated that the mutation-specific

216

metabolite signature in lpa mutants was nearly unaffected by the cross-

217

breeding step.


Page 10 of 34

Page 11 of 34


218

For the progenies of generations F5 to F7, there were shifts along PC1 and

219

PC2 compared to generation F4; however, the separation between homozygous

220

lpa

221

demonstrated the stability of the mutation-induced metabolite signature in lpa

222

progenies over generations.

mutant and homozygous wild-type progenies remained, which

223

PCA score plots of the single fractions are shown in Figure 2B-E. The score

224

plots of fractions III (Figure 2D) and IV (Figure 2E) also showed clear

225

separations of the homozygous wild-type and the homozygous lpa progenies.

226

This demonstrated that the constituents contained in these polar fractions are

227

the main contributors to the differentiation of these two phenotypes seen in the

228

score plot of the combined fractions (Figure 2A). In contrast, the lipophilic

229

fractions I (Figure 2B) and II (Figure 2C) showed separations between

230

progenies from generation F4 and those from generations F5-F7. This indicated

231

that these fractions are the main contributors to the shifts between generations

232

observed in the score plot of the combined fractions (Figure 2A) for both wild-

233

type and lpa progenies.

234 235

Metabolic differences between homozygous wild-type and homozygous lpa

236

mutant progenies.

237

In order to determine individual metabolites responsible for the observed

238

clustering (Figure 2A), the respective PCA loading plot (Figure S3) was used.

239

The metabolites were quantitated (relative peak intensities based on the

240

internal standards in each fraction), and mapped in simplified biosynthetic

241

pathways adapted from the KEGG pathway database.25 Figures 3 and 4 show

242

comparisons of the contents of metabolites in (i) the original wild-type MH86



243

and the lpa mutant Os-lpa-MH86-1, (ii) the crossing parent JH99, and (iii) the

244

homozygous wild-type and homozygous lpa mutant progenies of generations

245

F4 to F7, resulting from the cross-breeding of Os-lpa-MH86-1with JH99.

246

As shown in Figure 3, the phytic acid content in the homozygous F4 lpa

247

mutant (4.54 ± 0.05 mg/g), determined separately via targeted analysis using

248

HPIC, was significantly lower than that in the corresponding wild-type

249

progeny (7.13 ± 0.03 mg/g). For the following generations F5 to F7, despite the

250

variations of phytic acid contents, the homozygous lpa mutant progenies from

251

each generation always exhibited significantly lower levels of phytic acid than

252

the corresponding wild-type progenies. In addition, the phytic acid contents of

253

the homozygous lpa mutant progenies of generations F4 to F7 were

254

consistently lower than that of the original wild-type MH86.

255

The metabolite signatures of the homozygous lpa mutant progenies

256

resulting from the cross-breeding were characterized not only by this intended

257

reduction of the phytic acid content but also by changes in the levels of a

258

broad spectrum of metabolites from different classes. Statistically significant

259

differences in the levels of metabolites previously reported to discriminate

260

between MH86 and its lpa mutant Os-lpa-MH86-18 were similarly observed

261

for the homozygous wild-type and the homozygous lpa mutant progenies of

262

generation F4 resulting from the crossing of Os-lpa-MH86-1 with JH99. These

263

included sugars related to the biosynthesis of phytic acid (myo-inositol,

264

glucose, fructose and sucrose), a broad spectrum of amino acids, organic acids

265

(citric acid, succinic acid), sugar alcohols (sorbitol, mannitol) and biogenic

266

amines (GABA) (Figures 3 and 4).


Page 12 of 34

Page 13 of 34


267

In addition, the mutation-specific signature was consistently expressed in

268

lpa progenies from the generations F4 to F7. Although the absolute

269

concentrations of metabolites varied depending on the generations, in nearly

270

all cases, the differences between the homozygous wild-type and the

271

homozygous lpa mutant progenies were statistically significant for each

272

generation.

273

Univariate comparisons of the metabolite levels were also carried out

274

between the original wild-type MH86 and the homozygous lpa mutant

275

progenies of generations F4 - F7. There were a few sporadic examples where

276

increases observed for metabolites of the lpa progenies compared with the

277

original wild-type were not statistically significant for one of the generations.

278

However, in most cases the differences between MH86 and the homozygous

279

lpa mutant progenies were consistently expressed over all investigated

280

generations (Table 1).

281 282

Metabolic differences between progenies depending on the generation.

283

Apart from the clear differentiation of the metabolite profiles of the

284

homozygous lpa and the homozygous wild-type progenies, there was also a

285

metabolic shift among crossbred progenies of generations from F4 to F7 for

286

both wild-types and lpa mutants on the PCA score plot along PC 1 and PC 2,

287

respectively (Figure 2A). The extent of the shift was most pronounced from F4

288

to F5, then decreased with ascending generations from F5 to F7. The dynamics

289

of the changes of these metabolites depending on the generations are

290

illustrated by the heat map shown in Figure 5. The metabolic changes over the

291

generations were mainly attributable to lipophilic compounds. Apart from a



292

few exceptions, most of the FAMEs, FFAs and fatty alcohols showed

293

increasing levels with ascending generations both for homozygous wild-type

294

and lpa mutant progenies. This is in agreement with the PCA results which

295

showed similar metabolic shifts along ascending generations of progenies both

296

in fractions I and II (Figures 2B and 2C). On the contrary, polar constituents

297

contributed to a lesser extent to the differentiation according to the generation;

298

in fractions III and IV galactose, malic acid, GABA and serine exhibited

299

reduced levels, and only the tryptophan levels were increased from the early to

300

the late generations.

301 302

Separation of homozygous wild-type and lpa mutant progenies via

303

supervised statistical assessment

304

The results of the multivariate analysis via PCA gave a first indication that it is

305

possible to consistently differentiate between wild-type MH86 and

306

homozygous wild-type progenies resulting from the cross-breeding of the Os-

307

lpa-MH86-1 mutant with the commercial cultivar JH99, on the one hand, and

308

Os-lpa-MH86-1 mutants and the homozygous lpa mutant progenies resulting

309

from the cross-breeding step, on the other hand. To confirm this observation, a

310

supervised OPLS-DA was performed, taking into account all presently

311

available metabolite profiling data on MH86 and Os-lpa-MH86-1 mutants.

312

Figure 6A shows the OPLS-DA score plot of metabolite profiling data of

313

MH86 and the Os-lpa-MH86-1 mutant generated in this study and of MH86

314

and lpa mutants from three other field trials previously performed at other

315

locations and in different years, applying the same standard agronomic

316

practice (fertilization and water management).8 There were similar separation


Page 14 of 34

Page 15 of 34


317

patterns within the wild-type and the lpa mutant groups pre-defined in the

318

supervised model on the orthogonal component. Despite this variability, a

319

pronounced separation between wild-types and lpa mutants could be observed

320

along the predictive component. This separation remained after incorporation

321

of the metabolite profiling data from homozygous wild-type and homozygous

322

lpa mutant progenies from different generations obtained after cross-breeding

323

of the Os-lpa-MH86-1 mutant with JH99 performed in this study (Figure 6B).

324

For both plots, the high values of R2 and Q2 demonstrated the good fitness and

325

the predictability of the employed OPLS-DA models.26 In addition, the results

326

of permutation tests (n = 1000) demonstrated the models to be robust without

327

overfitting (Figure S5).

328

The OPLS-DA loadings S-plot22 was employed to reveal the metabolites

329

contributing to the differentiation between wild-type and lpa mutants. For the

330

main metabolites responsible for the differentiation seen in Figure 6B,

331

boxplots depicting the statistically significant differences in the mean

332

concentrations (expressed as relative peak intensities) of samples in the wild-

333

type and in the lpa mutant OPLS-DA groups are shown in Figure 7.

334 335

DISCUSSION

336

The phytic acid contents of homozygous lpa mutant progenies resulting from

337

the cross-breeding of the Os-lpa-MH86-1 mutant with JH99 were lower than

338

those of the corresponding homozygous wild-type progenies and of the

339

original wild-type MH86 (Figure 3, Table 1). This demonstrated that the lpa

340

trait, i.e. the significantly reduced content of phytic acid, remained nearly

341

unaffected by the crossing step and was consistently expressed in lpa



342

progenies over generations (F4 to F7). This result fulfilled a first major

343

prerequisite for the implementation of cross and selection breeding with

344

commercial cultivars as part of the procedure to generate lpa rice seeds.

345

The metabolic differences between the lpa mutant Os-lpa-MH86-1 and the

346

respective wild-type MH86 have been shown to result from the mutation of

347

the putative sulfate transporter gene OsSULTR3;3.8 The mutation affects the

348

expression of a number of genes involved in grain phosphorous and sulfur

349

homeostasis and metabolism. The reduced concentration of cysteine and the

350

increased concentration of its precursor serine suggest a deficiency in sulfide

351

supply. Overall, the lpa mutant exhibited a metabolite profile similar to that

352

reported in plants grown under phosphorous deficiency and/or sulfate

353

starvation, e.g. increased concentrations of sugars, sugar alcohols, free amino

354

acids and GABA.27, 28 The complex metabolite signature has been discussed to

355

be the result of a cross-talk between phosphate and sulfate homeostasis and

356

signaling8.

357

This study demonstrated that these mutation-induced metabolic differences

358

could also be observed between homozygous wild-type and homozygous lpa

359

progenies resulting from the cross-breeding of the Os-lpa-MH86-1 mutant

360

with JH99. The consistent metabolic differences observed for individual

361

metabolites linked via the biosynthetic pathways shown in Figures 3 and 4

362

form the molecular basis for the separations seen in the PCA plots (Figure 2).

363

The metabolite signature was shown to be superimposed by metabolic changes

364

depending on the generation. The shift in metabolite profiles between F4 seeds,

365

on the one hand, and F5-F7 seeds, on the other hand, observed in the PCA

366

score plot (Figure 2A) can be explained by the differences in genetic diversity


Page 16 of 34

Page 17 of 34


367

to be expected from the employed breeding scheme (Figure 1). All

368

investigated rice seeds originated from F2 plants. Homozygous wild-type and

369

homozygous lpa mutant F4 seeds were directly obtained by growing the F3

370

seeds from the homozygous wild-type and the homozygous lpa mutant F2

371

plants. In contrast, F5 to F7 seeds were obtained by growing the segregating F3

372

seeds of the heterozygous F2 plant. The subsequent process then included two

373

further stages of growing the segregating seeds of F4 and F5 plants,

374

respectively. The resulting metabolic differences between progenies of

375

generations F4 and F5-F7, respectively, were mainly attributable to changes in

376

the concentrations of lipophilic constituents. They occurred in both the

377

homozygous wild-type and the homozygous lpa progenies; they were

378

therefore reflected in a parallel translation of the respective clusters in the PCA

379

score plot (Figure 2A) and thus did not hamper their discrimination. The

380

mechanism underlying this effect on the concentrations of lipophilic

381

metabolites still needs further investigation.

382

The

application

of

a supervised

statistical assessment

approach

383

demonstrated that it was possible to assign the progenies resulting from the

384

employed cross-breeding step to the respective wild-type and lpa mutant

385

groups. The OPLS-DA score plots reflected variability within these groups

386

regarding environmental impact. For example, the metabolite profiles of the

387

rice samples obtained from the field trial performed in this study (Jiaxing,

388

2013) clustered together with those from a field trial performed at another

389

location in the temperate zone (Hangzhou, 2011) but were separated from

390

those of two field trials in the tropical zone (Hainan, 2012 and 2013). The

391

OPLS-DA score plots also showed differences between the homozygous wild-



392

type and homozygous lpa mutant progenies depending on the generations.

393

However, none of these factors finally interfered with the clear separation of

394

the wild-type and lpa mutant groups in the supervised statistical assessment.

395

The specific mutation-induced metabolite signature expressed in the lpa

396

mutant Os-lpa-MH86-1 as well as in the homozygous lpa progenies

397

encompasses the discriminating metabolites shown in Figure 7. Except for

398

cysteine, all other metabolites exhibited significantly increased levels in lpa

399

mutant rice seeds compared with wild-type seeds. It is noteworthy that these

400

metabolites also include some nutritionally relevant compounds, e.g. essential

401

amino acids, phytosterols and the biogenic amine GABA, a well-known

402

neurotransmitter.

403

In conclusion, the results demonstrate that both the lpa trait and the

404

concomitant metabolic changes resulting from the OsSULTR3;3 mutation

405

were nearly unaffected by cross-breeding of the lpa mutants with a

406

commercial cultivar. Despite of variations of the metabolite levels, the

407

mutation-induced metabolite signature in homozygous lpa mutants was

408

consistently expressed over several generations.

409

For the lpa mutant Os-lpa-MH86-1, the actual mechanism underlying the

410

reduction of the concentration of phytic acid as a result of the disruption of the

411

sulfate transporter gene has not been fully elucidated.8 Considering the broad

412

spectrum of metabolic alterations observed in the lpa mutant compared to the

413

wild-type MH86, it is noteworthy that the resulting complex mutation-induced

414

metabolite signature remained stable upon the employed cross-breeding step.

415

Of course, the results elaborated in this study for the metabolite profiles of

416

progenies obtained upon cross-breeding of the lpa mutant Os-lpa-MH86-1


Page 18 of 34

Page 19 of 34


417

with the commercial cultivar JH99 should not be unconditionally extrapolated

418

to any other lpa mutant. Depending on the induced genetic disruptions, for

419

each type of mutation specific features of changes in the metabolic network

420

and their susceptibility to a cross-breeding step will have to be taken into

421

account. However, the stability of a rather complex metabolite signature

422

shown in this study is encouraging and indicates that cross-breeding might be

423

a suitable tool to obtain lpa crops retaining the desired metabolic phenotype

424

induced by the mutation.

425

As shown in Figures 3 and 4, the metabolite profile of the crossing parent

426

JH99 employed in this study was similar to that of the wild-type MH86.

427

Therefore, it might also be of importance to follow the impact of cross-

428

breeding on the mutation-induced metabolite signature with a commercial

429

cultivar showing more pronounced own metabolic characteristics compared

430

with the MH86.

431

Owing to the limited amount of plant materials available from the field

432

trials, only preliminary assessments of the agronomic parameters could be

433

performed (Table S3). They indicated improved field emergence of the lpa

434

progenies; however, larger field trials yielding more plant materials would be

435

required to elaborate robust data and to draw valid conclusions regarding the

436

agronomic performance.

437

Nevertheless, from a breeder´s point of view, the achieved metabolite

438

profiling data are promising. The lpa mutant Os-lpa-MH86-1 exhibited its

439

favorable potential as useful germplasm resource to be exploited for

440

generating valuable lpa rice cultivars. The data indicate that even for complex

441

metabolic changes resulting from a mutation, cross-breeding can be employed



442

to obtain progeny rice seeds stably exhibiting the mutation-induced traits.

443

Further studies with different crossing parents should be performed to

444

substantiate the usefulness of this approach.

445 446

Funding

447

This project was financially supported by the Sino-German Center for

448

Research Promotion (project number GZ 932) and the China Scholarship

449

Council (CSC).

450

Notes

451

The authors declare no competing financial interest.

452

Supporting Information

453

Compounds identified in fractions I -IV of rice seeds. Agronomic parameters

454

of selected rice materials. Genotyping results via high-resolution melting

455

(HRM) analysis. GC chromatograms of fractions I-IV of MH86. PCA score

456

and loading plots of metabolite profiling data. Permutation tests of OPLS-DA

457

models.

458

459


Page 20 of 34

Page 21 of 34


460

REFERENCES

461

1.

462

2003, 64, 1033-1043.

463

2.

464

phytate and phytase in human nutrition: A review. Food Chem. 2010, 120, 945-

465

959.

466

3.

467

Broilers Waste through Supplementation of Wheat Based Broilers Feed with

468

Phytase. J Chem-Ny 2015, 3, 1-3.

469

4.

470

Inositol phosphates: Linking agriculture and the environment, 1st ed.; Turner, B.

471

L.; Richardson, A. E.; Mullaney, E. J., Eds. CAB international: Wallingford, UK,

472

2007; pp 111-132.

473

5.

474

molecular characterization of a mutation that confers a decreased

475

raffinosaccharide and phytic acid phenotype on soybean seeds. Plant Physiol.

476

2002, 128, 650-660.

477

6.

478

Low Phytic Acid (lpa) Soybean Mutants. J. Agric. Food Chem. 2009, 57, 6408-

479

6416.

480

7.

481

profiling of two low phytic acid (lpa) rice mutants. J. Agric. Food Chem. 2007, 55,

482

11011-11019.

483

8.

484

H.; Huang, J.; He, Z.; Poirier, Y.; Engel, K.-H.; Shu, Q., Disruption of

485

OsSULTR3;3 reduces phytate and phosphorus concentrations and alters the

486

metabolite profile in rice grains. New Phytol. 2016, 211, 926-939.

487

9.

488

Effect of non-lethal low phytic acid mutations on grain yield and seed viability in

489

rice. Field Crops Res. 2008, 108, 206-211.

Raboy, V., myo-Inositol-1,2,3,4,5,6-hexakisphosphate. Phytochemistry

Kumar, V.; Sinha, A. K.; Makkar, H. P. S.; Becker, K., Dietary roles of

Abdel-Megeed, A.; Tahir, A., Reduction of Phosphorus Pollution from

Raboy, V., Seed phosphorus and the development of low-phytate crops. In

Hitz, W. D.; Carlson, T. J.; Kerr, P. S.; Sebastian, S. A., Biochemical and

Frank, T.; Nörenberg, S.; Engel, K. H., Metabolite Profiling of Two Novel

Frank, T.; Meuleye, B. S.; Miller, A.; Shu, Q. Y.; Engel, K. H., Metabolite

Zhao, H.; Frank, T.; Tan, Y.; Zhou, C.; Jabnoune, M.; Arpat, A. B.; Cui,

Zhao, H. J.; Liu, Q. L.; Fu, H. W.; Xu, X. H.; Wu, D. X.; Shu, Q. Y.,



490

10.

Boehm, J. D.; Walker, F. R.; Bhandari, H. S.; Kopsell, D.; Pantalone, V.

491

R., Seed Inorganic Phosphorus Stability and Agronomic Performance of Two

492

Low-Phytate Soybean Lines Evaluated across Six Southeastern US Environments.

493

Crop Sci. 2017, 57, 2555-2563.

494

11.

495

soybean lines with low phytate. Crop Sci. 2007, 47, 1354-1360.

496

12.

497

Generation and characterization of low phytic acid germplasm in rice (Oryza

498

sativa L.). Theor. Appl. Genet. 2007, 114, 803-814.

499

13.

500

Shu, Q. Y., Functional molecular markers and high-resolution melting curve

501

analysis of low phytic acid mutations for marker-assisted selection in rice. Mol.

502

Breed. 2013, 31, 517-528.

503

14.

504

in Dried Distillers Grains with Solubles. Thermo Fisher Scientific, Sunnyvale, CA,

505

USA. 2014, 1-10.

506

15.

507

methodology Q2 (R1), International Conference on Harmonization, Geneva,

508

Switzerland, 2005; pp 11-12.

509

16.

510

lipids of Sesamum indicum and related wild species in Sudan. The sterols. J. Sci.

511

Food Agric. 1992, 59, 327-334.

512

17.

513

gamma-Oryzanol in rice bran oil. J. Agric. Food Chem. 1999, 47, 2724-2728.

514

18.

515

of induced drought stress on the metabolite profiles of barley grain. Metabolomics

516

2015, 11, 454-467.

517

19.

518

From sample treatment to biomarker discovery: A tutorial for untargeted

519

metabolomics based on GC-(EI)-Q-MS. Anal. Chim. Acta 2015, 900, 21-35.

Spear, J. D.; Fehr, W. R., Genetic improvement of seedling emergence of

Liu, Q. L.; Xu, X. H.; Ren, X. L.; Fu, H. W.; Wu, D. X.; Shu, Q. Y.,

Tan, Y. Y.; Fu, H. W.; Zhao, H. J.; Lu, S.; Fu, J. J.; Li, Y. F.; Cui, H. R.;

Oates, K.; De Borba, B.; Rohrer, J., Determination of Inositol Phosphates

Guideline, I. H. T. In Validation of analytical procedures: text and

Kamal‐Eldin, A.; Appelqvist, L. Å.; Yousif, G.; Iskander, G. M., Seed

Xu, Z. M.; Godber, J. S., Purification and identification of components of

Wenzel, A.; Frank, T.; Reichenberger, G.; Herz, M.; Engel, K.-H., Impact

Mastrangelo, A.; Ferrarini, A.; Rey-Stolle, F.; Garcia, A.; Barbas, C.,


Page 22 of 34

Page 23 of 34


520

20.

Frenzel, T.; Miller, A.; Engel, K. H., A methodology for automated

521

comparative analysis of metabolite profiling data. Eur. Food Res. Technol. 2003,

522

216, 335-342.

523

21.

524

van der Werf, M. J., Centering, scaling, and transformations: improving the

525

biological information content of metabolomics data. BMC Genomics 2006, 7,

526

142.

527

22.

528

S.; Xia, J., MetaboAnalyst 4.0: towards more transparent and integrative

529

metabolomics analysis. NAR 2018, 46, W486-W494.

530

23.

531

Guideline to Univariate Statistical Analysis for LC/MS-Based Untargeted

532

Metabolomics-Derived Data. Metabolites 2012, 2, 775.

533

24.

534

Fractionation method for analysis of major and minor compounds in rice grains.

535

Cereal Chem. 2002, 79, 215-221.

536

25.

537

Genomes. Nucleic Acids Res 2000, 28.

538

26.

539

Metabolomics 2013, 1, 92-107.

540

27.

541

root responses to phosphorus starvation. JExB 2008, 59, 93-109.

542

28.

543

Kempa, S.; Morcuende, R.; Scheible, W.-R.; Hesse, H.; Hoefgen, R., Effect of

544

sulfur availability on the integrity of amino acid biosynthesis in plants. Amino

545

Acids 2006, 30, 173-183.

Van den Berg, R. A.; Hoefsloot, H. C.; Westerhuis, J. A.; Smilde, A. K.;

Chong, J.; Soufan, O.; Li, C.; Caraus, I.; Li, S.; Bourque, G.; Wishart, D.

Vinaixa, M.; Samino, S.; Saez, I.; Duran, J.; Guinovart, J. J.; Yanes, O., A

Frenzel, T.; Miller, A.; Engel, K. H., Metabolite profiling - A

Kanehisa, M.; Goto, S., KEGG: Kyoto Encyclopedia of Genes and

Bradley, W.; Robert, P., Multivariate Analysis in Metabolomics. Current

Hammond, J. P.; White, P. J., Sucrose transport in the phloem: integrating

Nikiforova, J. V.; Bielecka, M.; Gakière, B.; Krueger, S.; Rinder, J.;

546



Figure Captions Figure 1. Flowchart of the cross-breeding steps applied to produce the rice materials. Figure 2. PCA score plots of metabolite profiling data of combined fractions I – IV (A) and single fractions I (B), II (C), III (D) and IV (E) from wild-type (black) and lpa mutant (blue) rice seeds: original wild-type MH86 (open circles) and Oslpa-MH86-1 mutant (open circles); crossing parent JH99 (black solid circles); homozygous progenies of generations F4 (inverted triangles), F5 (diamonds), F6 (squares) and F7 (triangles). Figure 3. Simplified biosynthetic pathways of selected rice seed metabolites involved in phytic acid, sugar and lipid metabolism. The bars are displayed in the following order: (A) original wild-type MH86 (black) and progenitor Os-lpaMH86-1 mutant (blue); (B) crossing parent JH99 (white); homozygous wild-type progenies (black) and homozygous lpa mutant progenies (blue) of generations F4, F5, F6 and F7. The phytic acid contents are expressed in mg/g of dry matter. All other metabolites are expressed as relative peak intensities, i.e. metabolite peak intensity / internal standard peak intensity × 100. Values represent means ± standard deviations resulting from the analysis of three aliquots of freeze-dried flour. Asterisks represent statistically significant differences (Student’s t-test or Mann-Whitney test, p < 0.05) between the wild-type and the corresponding lpa mutant. MIPS, 1D-myo-inositol 3-phosphate synthase; IMP, myo-inositol monophosphatase; MIK, myo-inositol kinase; PtdIns, phosphatidyl inositol; PtdInsK, phosphatidyl inositol phosphate kinase; PtdInsP, phosphatidyl inositol phosphate; InsPK, inositol phosphate kinase; UDP, uridine diphosphate. Figure 4. Simplified biosynthetic pathways of selected rice seed metabolites involved in amino acid metabolism, phytosterol metabolism and tricarboxylic acid cycle. The bars are displayed in the following order: (A) original wild-type MH86 (black) and progenitor Os-lpa-MH86-1 mutant (blue); (B) crossing parent JH99 (white); homozygous wild-type progenies (black) and homozygous lpa mutant progenies (blue) of generations F4, F5, F6 and F7. The metabolites are expressed as 24 ACS Paragon Plus Environment

Page 24 of 34

Page 25 of 34


relative peak intensities, i.e. metabolite peak intensity / internal standard peak intensity × 100. Values represent means ± standard deviations resulting from the analysis of three aliquots of freeze-dried flour. Asterisks represent statistically significant differences (Student’s t-test, p < 0.05) between the wild-type and the corresponding lpa mutant. Figure 5. Heat map of metabolites contributing to the metabolic shift among the wild-type and lpa mutant rice seeds of different generations. Figure 6. OPLS-DA score plots of metabolite profiling data. (A) Original wildtype MH86 (black) and Os-lpa-MH86-1 mutants (blue) from the four field trials Hangzhou, 2011 (open diamonds); Hainan, 2012 (open triangles); Hainan, 2013 (open squares) and Jiaxing, 2013 (open circles). (B) Same sample set as in (A) plus homozygous wild-type progenies (solid black) and homozygous lpa mutant progenies (solid blue) of generations F4 (inverted triangles), F5 (diamonds), F6 (squares) and F7 (triangles). The boundaries of the clusters of wild-type and lpa mutant seeds correspond to the 95% Hotelling's T2 ellipses. Figure 7. Metabolites contributing to the OPLS-DA separation of wild-type and lpa mutant rice seeds. All metabolites exhibited significantly different levels between wild-type and lpa mutant rice seeds (Benjamini–Hochberg adjusted-p < 0.05). For each metabolite, the relative peak intensity of wild-type was the mean value of MH86 from four field trials (three aliquots per field trial) and homozygous wild-type progenies of generations F4 to F7 (three aliquots per generation); the relative peak intensity of lpa mutants was the mean value of Oslpa-MH86-1 from four field trials (three aliquots per field trial) and homozygous lpa progenies of generations F4 to F7 (three aliquots per generation).



Page 26 of 34

Table 1. Contents of phytic acid and relative peak intensities of selected metabolites in the original wild-type MH86 and the lpa progenies of generations F4, F5, F6 and F7 obtained from field trial Jiaxing, 2013.a,b,c original wild-type homozygous lpa mutant progenies metabolite MH86 F4 F5 F6 F7 phytic acid phosphoric acid

9.69 ± 0.2 a 67.6 ± 30.0 b

4.54 ± 0.05 d 609.3 ± 21.1 a

6.12 ± 0.01 b 504.4 ± 23.3 a

5.38 ± 0.11 c 488.2 ± 21.1 a

6.08 ± 0.06 b 569.8 ± 19.5 a

glucose sucrose mannitol

19.3 ± 0.3 e 1319 ± 25.9 c 3.9 ± 0.1 e

32.9 ± 0.4 b 1782 ± 15.9 b 5.2 ± 0.3 d

55.5 ± 0.4 a 2158 ± 37.2 a 12.4 ± 0.1 a

24.2 ± 0.3 d 1859 ± 30.3 b 7.2 ± 0.02 c

31.5 ± 0.4 c 2219 ± 47.5 a 10.5 ± 0.1 b

glycine leucine isoleucine threonine cysteine aspartic acid glutamic acid asparagine ornithine citrulline GABA

4.5 ± 0.1 2.6 ± 0.1 5.0 ± 0.3 3.9 ± 0.1 0.56 ± 0.03 27.7 ± 1.2 33.8 ± 1.3 74.7 ± 0.4 0.49 ± 0.02 0.43 ± 0.02 2.9 ± 0.1

c c c c a d b d b d c

8.6 ± 0.9 4.8 ± 0.5 7.0 ± 0.6 6.6 ± 0.6 0.26 ± 0.01 53.2 ± 0.4 48.9 ± 3.3 177.8 ± 8.8 1.01 ± 0.20 0.56 ± 0.07 5.4 ± 0.7

b a b b bc a a a a c a

8.8 ± 0.6 4.6 ± 0.2 8.8 ± 0.4 7.8 ± 0.3 0.31 ± 0.05 43.9 ± 1.6 47.6 ± 1.3 116.8 ± 2.8 0.99 ± 0.22 1.09 ± 0.04 4.7 ± 0.3

ab a a a b b a c a a a

10.4 ± 0.3 3.6 ± 0.2 7.7 ± 0.3 6.2 ± 0.2 0.24 ± 0.02 40.1 ± 1.1 48.2 ± 1.5 108.2 ± 3.2 1.16 ± 0.05 0.66 ± 0.04 3.6 ± 0.1

a b ab b bc c a c a bc b

8.0 ± 0.7 3.5 ± 0.4 7.5 ± 0.3 5.9 ± 0.4 0.22 ± 0.03 42.0 ± 1.2 47.9 ± 1.5 137.6 ± 2.4 0.81 ± 0.05 0.70 ± 0.02 2.3 ± 0.2

b b b b c bc a b a b d

citric acid

34.2 ± 1.4

d

247.0 ± 6.6

a

160.6 ± 7.8

c

191.2 ± 5.5

b

159.5 ± 5.8

c

stigmasterol

47.0 ± 3.7

b

57.4 ± 3.9

a

63.1 ± 4.2

a

57.3 ± 1.5

a

61.4 ± 1.3

a

a

The phytic acid contents are expressed in mg/g of dry matter. All other metabolites are expressed as relative peak intensities, i.e. metabolite peak intensity / internal standard peak intensity × 100. b Values represent means ± standard deviations resulting from the analysis of three aliquots of freeze-dried flour. c For each metabolite, different letters indicate significant differences (ANOVA with Tukey’s post hoc test or Kruskal-Wallis test with Dunn’s post hoc test, Benjamini-Hochberg corrected p < 0.05) among the original wild-type MH86 and the homozygous lpa mutant progenies of generations F4, F5, F6 and F7.


Page 27 of 34


Os-lpa-MH86-1

x

JH99

F1 seeds HM: homozygous lpa mutant HW: homozygous wild-type (WT) HET: heterozygous

F1 plants

samples harvested in Jiaxing, 2013 and subjected to metabolite profiling

F2 seeds F2 plants genotyping

HM plant

HW plant

HM F3 seeds

HW F3 seeds

HET plant segregating F3 seeds

F3 plants genotyping

HM plant

HW plant

HM F4 seeds

HW F4 seeds

HET F3 plant segregating F4 seeds F4 plants genotyping

HM plant

HW plant

HET F4 plant

HM F5 seeds

HW F5 seeds

segregating F5 seeds F5 plants genotyping

HM plant

HW plant

HM F6 seeds

HW F6 seeds

Os-lpa-MH86-1

Grown in Jiaxing 2013

HM F4 seeds

HM F5 seeds

HM F6 seeds

HM F7 seeds

MH86

HW F4 seeds

HW F5 seeds

HW F6 seeds

HW F7 seeds

Os-lpa-MH86-1

Figure 1. 27 ACS Paragon Plus Environment

MH86



Page 28 of 34

Page 29 of 34


phosporic aicd phosphoric acid

800

*

600

phytic acid phytic acid

15

InsPK

*

*

*

*

400

*

10

*

*

InsP3

*

*

Phospholipase C

5

200 0

A B

F4

F5

F6

0

F7

A B

F4

F5

F6

PtdInsP intermediates

F7

MIK

pyruvic acid

PtdIns

lower InsP (InsP2 - InsP5) 75

glucose glucose

50

* *

25 0

*

A B

F5

glucose-6phosphate

*

F6

F7

acetylCoA sorbitol sorbitol

*

2.4

*

*

10

1.2

0

0.0

A B

F4

F5

F6

* *

glycerol glycerol

20

*

A B

*

0

A B

F4

F5

0

A B

F6

F6

F7

12

glycerol-3phosphate

galactinol

sucrose sucrose

F7 1800

mannitol mannitol

*

*

6

*

A B

F4

F5

F6

0

F7

*

A B

F4

*

*

*

900

*

*

*

F5

F6

F7

raffinose C16:0 C16:0 FAME FAME

6000

F5

F5

*

MIK

2700

F6

18

0

*

10

F4

*

6

*

*

F4

F7

*

*

UDP-glucose

*

fructose-6phosphate 30

*

UDP-galactose 20

*

IMP

*

fructose fructose

30

3.6

12

myoinositol-1phosphate

MIPS

myo-inositol myo-inositol

18

InsPK

*

F4

PtdInsK

InsPK

4000

F7

2000

triglycerides FAME

free fatty acids (FFA)

460

*

*

*

0

F4

F5

F6

F7

F4

F5

F6

F7

C18:1 FAME FAME C18:1

6000

*

300

A B

A B

9000

*

600

230 0

C18:1 FFA C18:1 FFA

900

C16:0 FFA C16:0 FFA

690

0

3000

A B

F4

F5

F6

F7

Figure 3.


0

A B

F4

F5

F6

F7


leucine leucine

7.5 5.0

*

*

18

*

*

*

2.5 0.0

A B

F4

F5

alanine alanine

27

F6

*

*

*

F7

A B

F4

F5

F6

* 0.0

F4

F5

serine serine

45

F6

*

pyruvic acid 0

A B

F4

*

*

*

F5

F6

F7

malic acid

*

*

0

*

200

*

*

succinic acid succinic acid

6

citric acid citric acid

*

*

*

F4

0

A B

F4

F5

F6

A B

F7

F6

*

ketoglutaric acid squalene

F4

F5

F6

A B

F4

F5

F6

60

aspartic acid aspartic acid

40

*

* *

0

A B

44

F7

*

6

0

F7

*

*

*

A B

7-avenasterol ∆7-avenasterol

18 12

*

*

*

*

*

200

*

*

F4

F5

*

*

*

A B

F4

F5

F6

F7

0

60

F5

F6

*

*

*

*

F4

F5

F6

F7

0

*

F6

F4

F5

F6

*

*

*

F4

F5

F6

F7

asparagine asparagine

*

A B

F4

*

*

*

F5

F6

F7

F7

* *

*

0

A B

GABA GABA

12

F4

8

F5

*

*

F6

F7

*

*

*

4 0

ornithine ornithine

1.5

*

1.0

*

A B

F4

*

F5

*

F6

F7

*

F7

0.0

*


F4

F5

F6

F7

*

1.0 0.5

A B

citrulline citrulline

1.5

*

0.5

A B

F7

F7

30

A B

F6

F5

glutamine glutamine

4

*

stigmasterol stigmasterol

90

100

6 0

sitosterol β-sitosterol

300

F4

F4

100

proline A B

*

200

22 0

A B

300

*

8

*

*

threonine threonine

0

glutamic acid glutamic acid

66

0

F7 9

*

12

0

F5

homoserine

*

*

2

100

A B

*

*

4

20

*

4

*

*

*

*

3

fumaric acid

300

0.0

*

*

8

*

4

tryptophane

oxaloacetic acid

0.5

glycine glycine

8

15

F7

12

*

30

acetyl-CoA

*

A B

*

*

isoleucine isoleucine

12

*

1.0

0.4

F7

methionine methionine

1.5

0.8

O-acetylserine

9 0

cysteine cysteine

1.2

*

Page 30 of 34

0.0

* A B

* F4

F5

*

*

F6

F7

Page 31 of 34


wild-type MH86

F4

F5

lpa mutant F6

F7

MH86

C14:1 FAME C14:0 FAME C15:1 FAME C15:0 FAME C16:1 FAME C18:0 FAME C20:0 FAME C22:0 FAME C24:0 FAME C26:0 FAME C28:0 FAME C14:0 FFA C16:1 FFA C18:1 (9Z) FFA C18:1 (9E) FFA C20:1 FFA C20:0 FFA C22:0 FFA C24:0-OH C26:0-OH campesterol 24-MCA galactose serine GABA tryptophan malic acid

Figure 5.


F4

F5

F6

F7


Figure 6.


Page 32 of 34

Page 33 of 34


relative peak intensity



900

300

15

citric acid

mannitol

9

600

200

10

6

300

100

5

3

0

0

0

0

2400

sucrose

75

glucose

1.2

cysteine

12

2000

50

0.8

8

1600

25

0.4

4

1200

0

0.0

0

30

1.5

serine

ornithine

7.5

tyrosine

12

20

1.0

5.0

8

10

0.5

2.5

4

0

0.0

0.0

0

10


phosphoric acid

threonine

15

GABA

270

sitosterol

90

7

10

230

70

4

5

190

50

1

0

wildtype

lpa mutant

wildtype

glycine

isoleucine

stigmasterol

30

150

lpa mutant

sorbitol

wildtype

Figure 7.


lpa mutant

wildtype

lpa mutant


Table of Contents Graphics


Page 34 of 34

Stability of the Metabolite Signature Resulting from the OsSULTR3;3

Recommend Documents