Butia spp. (Arecaceae) LC-MS-based metabolomics for species and

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Article

Butia spp. (Arecaceae) LC-MS-based metabolomics for species and geographical origin discrimination Jessica Fernanda Hoffmann, Ivan Ricardo Carvalho, Rosa Lia Barbieri, Cesar Valmor Rombaldi, and Fabio Clasen Chaves J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.6b03203 • Publication Date (Web): 16 Dec 2016 Downloaded from http://pubs.acs.org on December 17, 2016

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 free 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 accessible to all readers and 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.

Journal of Agricultural and Food Chemistry 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 36

Journal of Agricultural and Food Chemistry

PCA for species discrimination LC-MS-based metabolomics of Butiá fruit

ACS Paragon Plus Environment

PLS-DA for location discrimination


Page 2 of 36 1

1

Butia spp. (Arecaceae) LC-MS-based metabolomics for species and

2

geographical origin discrimination

3 4

Jessica Fernanda Hoffmanna, Ivan Ricardo Carvalhoa, Rosa Lia Barbierib, Cesar

5

Valmor Rombaldia, and Fabio Clasen Chavesa*

6 7

a

8

Postal 354, CEP 90010-900, Pelotas, RS, Brazil

9

b

10

Faculdade de Agronomia Eliseu Maciel, Universidade Federal de Pelotas, Caixa

Embrapa Clima Temperado, Caixa Postal 403, CEP 96001-970, Pelotas, RS, Brazil

*Corresponding author: +55 53 32757284; [email protected]

11


Page 3 of 36

Journal of Agricultural and Food Chemistry 2

12

Abstract

13

The metabolic variability of fruit from Butia spp. (Arecaceae) genotypes from different

14

geographical locations was characterized using untargeted metabolomics by liquid

15

chromatography-mass spectrometry (LC-MS) followed by multivariate data analyses.

16

Principal component analysis (PCA) and partial least squares-discriminant analysis

17

(PLS-DA) from LC-MS data sets showed a clear distinction among B. catarinensis, B.

18

odorata, B. paraguayensis, and B. yatay. The major metabolites that contributed to

19

species discrimination were primary metabolites including sugars and organic acids,

20

and specialized metabolites such as tetrahydroxy-trans-stilbene and rutin. B. odorata

21

fruit from Tapes, RS showed a high content of organic acids and flavonoids, while B.

22

odorata fruits from Capão do Leão, RS showed a high sugar content. The results

23

demonstrate

24

chemometric analysis can be used to discriminate among Butia species and between

25

geographical origins of Butia odorata, and to identify primary and specialized

26

metabolites responsible for the discrimination.

that

LC-ESI-qToF-MS-based

metabolic

profiling

coupled

with

27 28

Keywords: metabolic variability; bioactive compounds; native fruit; metabolomics;

29

LC-ESI-qTOF-MS; multivariate analysis.

30



Page 4 of 36 3

31

1.

Introduction

32

Brazil has the planet’s largest biodiversity with many unexplored fruit-bearing

33

plants distributed in different biomes. The Pampa Biome (subtropical and temperate

34

climates) has a number of underutilized native fruit species possessing commercial

35

potential, especially for smallholders and family farmers, for new product

36

development focusing on emerging markets for nutritional and functional foods 1. In

37

southern Brazil, there are eight native species of Butia and the species B. odorata, B.

38

yatay, B. paraguayensis, and B. catarinensis are the most predominant. B. witeckii is

39

the most recently described species and is considered to be most closely related to

40

B. yatay and B. paraguayensis, differing from these species in fruit size and weight,

41

endocarp, and inflorescence among other morphological characteristics 2.

42

Butia palms are used in landscaping and their dried leaves are used in the 3,4

43

making of hats, brooms, baskets, and other crafts

44

acid-sweet taste, intense flavor and aroma, and are rich in phenolic compounds,

45

carotenoids, vitamin C, and potassium

46

pulp is used to produce juices, liqueurs, sweets, jellies, and ice cream 4,8. The uses of

47

these palms have been recorded since pre-historic times and their utilization, like

48

most natives fruit species, has occurred from natural populations without any

49

systematic cultivation. In addition, the areas occupied by naturally occurring

50

‘butiazais’ are subject to protection by the State Forestry Code and therefore,

51

management and commercial exploration of Butia spp. is limited 9.

5-8

. The fruit are fibrous, have an

. They are usually consumed fresh or the

52

In order to conserve and characterize the variability of Butia, research

53

institutions maintain germplasm banks. The accessions belonging to these

54

collections need to be evaluated for agronomic, technological, phytochemical, and

55

biological potential to identify genotypes with superior quality and commercial


Page 5 of 36


56

potential. The knowledge of variability (molecular, phenological, morphological, and

57

chemical) provides support for the establishment of conservation strategies,

58

incorporation in regional production systems, sustainable management, and use 1,10.

59

Several studies have been developed resulting in an increase in scientific

60

knowledge about Butia palm trees. However, these studies have primarily focused on

61

agronomic, morphological, and molecular characterization of Butia spp. as well as

62

some isolated preliminary chemical characterization of the fruit 3. With

63

recent

developments

in

analytical

techniques

such

as

liquid

64

chromatography coupled to mass spectrometry (LC-MS), it is possible to

65

simultaneously detect hundreds of metabolites and thus compare the differences and

66

similarities among samples. This non-targeted approach has been used to obtain

67

information on the metabolic variability in germplasm banks, differentiation of

68

species, and to establish the geographic origin of different fruit

69

metabolomics can be used for chemical classification of plants by chemotaxonomy.

70

Differentiation between different species of Butia based on their metabolomic

71

profiling has not been carried out, and their speciation has been based predominantly

72

on morphological characters and their regional occurrence. Thus, the aim of this

73

study was to evaluate the metabolic variability of Butia spp. genotypes using

74

untargeted metabolomics by LC-ESI-qToF-MS.

11,12

. Moreover,

75 76

2.

Material and methods

77

2.1 Sample collection and preparation

78

This study was divided into two experiments. In the first, the interest was to

79

evaluate differences among Butia catarinensis Noblick and Lorenzi, B. odorata (Barb.

80

Rodr.) Noblick, B. paraguayensis (Barb. Rodr.) L.H. Bailey, and B. yatay (Mart.)



Page 6 of 36 5

81

Becc.. In the second, differences among genotypes of B. odorata, the predominant

82

species in southern Brazil, from different geographical origins were examined. Butia

83

species were identified and classified by Barbieri, R.L. based on morphological

84

characteristics according to Lorenzi et al. 5.

85

B. catarinensis fruit were collect in in Laguna, SC (28º 28' 57" S; 48º 46' 51"

86

W, and altitude of 2m), B. paraguayensis fruit in Rondinha, RS (27º 49' 41" S; 52º 54'

87

35" W, and altitude of 440m), and B. yatay fruit in Giruá, RS (28º 01' 42" S; 54º 20'

88

59" W, and altitude of 429m). For B. catarinensis, B. paraguayensis, and B. yatay, a

89

pool of 50 ripe fruit from three plants of each species was collected. For B. odorata,

90

fruit was collected from 57 genotypes. Fruit of B. odorata were collected in two

91

locations in the state of Rio Grande do Sul (RS), in the municipalities of Capão do

92

Leão (31º 52’ 00" S; 52º 21’ 24" W, and altitude of 13m) and Tapes (30° 31’ 22.34” S;

93

51° 21’ 35.23” W, and altitude of 10m). Fifty ripe fruit from each of 38 genotypes were

94

collected in Capão do Leão, and 50 ripe fruit from each of 19 genotypes were

95

collected in Tapes for a total of 57 genotypes of B. odorata.

96

After collection, fruit were kept in a freezer (-20 °C) until analysis. For

97

physicochemical analysis, fruit were manually depulped, and approximately 100 g of

98

sample were ground in a ball mill with liquid nitrogen to fine powder. For metabolic

99

profile by LC-ESI-qToF-MS, samples were lyophilized.

100 101

2.2 Physicochemical analysis

102

2.2.1 Color, soluble solids, pH, and acidity

103

Fruit skin color (measured in 9 fruits per genotype) was determined using a

104

colorimeter (Minolta Chronometer, CR 300) in the CIELab color system. The values

105

of a* and b* were used to calculate hue angle (ºHue= tan -1 b*/a*).


Page 7 of 36


106

Soluble solids (SS) content was determined by refractometry, and results were

107

expressed as ºBrix. For analysis of pH and acidity (AC), one gram of sample was

108

diluted with 40 mL of distilled water. The pH was determined with a pH meter (Hanna

109

Instruments HI2221). For analysis of acidity, the mixture was titrated with NaOH (0.1

110

M) until pH 8.1. Results were expressed as mg of citric acid equivalent (CAE) 100 g-1

111

fresh weight (fw). Analyses were performed according to AOAC methods 13.

112 113 114

2.2.2 Vitamin C Vitamin C content was determined using the titrimetric method with 2,613

115

dichlorophenolindophenol reagent

. One gram of ground sample was mixed with 15

116

mL of extracting solution (3% metaphosphoric acid – 8% acetic acid solution). The

117

mixture was homogenized and centrifuged at 6600 x g for 15 min. Four milliliters of

118

the supernatant was diluted with 2 mL of the extracting solution and 50 mL of water.

119

This solution was titrated with a 2,6-dichlorophenolindophenol solution (0.01%).

120

Titration endpoint was defined when the solution turned a persistent (15 sec) pink

121

color. Results were expressed as mg of L-ascorbic acid equivalents 100 g-1 fw.

122 123 124

2.2.3 Carotenoid content Total carotenoid content analysis was carried out according to the AOAC 970.64

13

125

method

.

Fifteen

milliliters

of

126

(hexane:acetone:ethanol:toluene 10:7:6:7) were added to 2.5 g of ground sample in

127

a polyethylene tube (50 mL) protected from light and stirred for one min by vortexing.

128

Then, one mL of potassium hydroxide in methanol (10% w/v) was added and the

129

mixture was stirred for one min and then subjected to hot saponification (20 min in a

130

water bath at 56°C). The mixture stayed at room temperature for one hour, and then


extraction

solution


Page 8 of 36 7

131

15 mL of hexane and 29 mL of sodium sulfate solution (10% w/v) were added. After

132

one hour, the absorbance was measured at 450 nm (Jenway 6705 UV-Vis). A

133

calibration curve (0–20 µg mL-1) was prepared using β-carotene (Sigma-Aldrich) in

134

hexane. Results were expressed as mg of β-carotene equivalents 100 g-1 fw.

135 136

2.2.5 Preparation of extracts

137

Two grams of ground sample was homogenized with 20 mL of methanol for

138

one min in an IKA® Ultra Turrax (digital T18), and subsequently centrifuged for 10 min

139

at 9900 x g (Eppendorf, 5430) at 15 °C. The supernatant was separated and used as

140

a crude extract to determine total phenolic content, flavonoid content, and total

141

antioxidant capacity (DPPH and ABTS).

142 143

2.2.4 Phenolic compounds content

144

Folin–Ciocalteau reagent was used to determined the total phenolic

145

compounds content according to Beskow et al. 7. A calibration curve (0–250 µg mL-1)

146

was prepared using gallic acid (Sigma-Aldrich) in methanol. Absorbance was

147

measured at 725 nm and results were expressed as mg of gallic acid equivalents

148

(GAE) 100 g-1 fw.

149 150 151

2.2.5 Flavonoid content Flavonoid content was determined according to Zhishen et al.

14

. 500 µL of

152

the extract was mixed with two mL of distilled water and 150 µL of sodium nitrite (5%

153

w/v) and allowed to react for five min. 150 µL of aluminum chloride (10% w/v) was

154

added and allowed to react for 6 minutes, followed by the addition of one mL of

155

sodium hydroxide (1 M) and 1.2 mL of distilled water. Absorbance was measured at


Page 9 of 36


156

510 nm. A calibration curve (0-150 µg mL-1) was prepared using (+)-catechin (Sigma-

157

Aldrich) in methanol. Results were expressed as mg of (+)-catechin equivalents 100

158

g-1 fw.

159 160 161

2.2.6 Radical scavenging assay DPPH

(2,2-diphenyl-1-picrylhydrazyl)

and

ABTS

(2,2'-azino-bis(3-

162

ethylbenzothiazoline-6-sulphonic acid) radical scavenging capacity of crude extract

163

was measured according to a method previously reported by Pereira et al.15. The

164

scavenging capacity of the extracts was expressed as percentage (%) DPPH and

165

ABTS radical remaining according to the equation:

166

% inhibition= [(Acontrol−Asample)/Acontrol] x 100

167

where Acontrol is the absorbance of the control (containing all reagent except the

168

sample), Asample is the absorbance of the sample.

169 170

2.3 Preparation of extracts for metabolic profiling by LC-ESI-qTOF-MS

171

The lyophilized fruit pulp sample was divided into three subsamples of 100 mg

172

each and 1 mL of ice-cold sample extraction solution (75% methanol and 0.1%

173

formic acid) and 10 µg of reserpine mL-1 (as internal standard for relative

174

quantification) were added. The samples were vortexed, sonicated in a water bath at

175

room temperature for 15 min, centrifuged (9900 x g, 15 min), and the supernatants

176

were collected. The extraction process was repeated once 16. The supernatants were

177

combined, filtered through a 0.2 µm nylon membrane and stored at −80 °C until

178

analysis.

179 180

2.4 LC-ESI-qToF-MS analysis



Page 10 of 36 9

181

The LC-ESI-qToF-MS analysis was performed on a Prominence UFLC system

182

(Shimadzu, Japan) coupled to a quadrupole time-of-flight mass spectrometer (Impact

183

HD, Bruker Daltonics, Bremen, Germany). Separation of metabolites was performed

184

using a Bidentate C18 column (100 x 2.1 mm, MicroSolv Technology Corporation,

185

Eatontown, NJ, EUA). Mobile phases were 0.1% aqueous formic acid (pH 2.8;

186

solvent A) and acetonitrile (solvent B). The gradient program was set as follows:

187

started at 5% B and increased linearly to 90% B at 15 min and was maintained for 3

188

min at 90% B; returned to 5% B in two minutes and was maintained at 5% B for 6

189

more minutes at a flow rate of 0.2 mL min-1. The injection volume was 10 µL. All

190

samples were injected in duplicate.

191

Parameters for MS analysis were set using negative and positive ionization

192

mode with spectra acquired over a mass range from 50 to 1200 m/z. The parameters

193

were: capillary voltage 3.5 kV; drying gas temperature 180 °C; drying gas flow 8.0 L

194

min-1; nebulizing gas pressure 2 bar; collision RF 300 Vpp; transfer time 120 µs, and

195

pre-pulse storage 8 µs. Collision energy intensity was adjusted for automatic MS/MS

196

experiments according to m/z ratios: m/z 100, 15 eV; m/z 500, 35 eV; m/z 1000, 50

197

eV, using nitrogen as collision gas.

198 199

2.5 Data processing and statistical analysis

200

The MS data were analyzed using Data Analysis 4.0 software (Bruker

201

Daltonics, Bremen, Germany). ProfileAnalysisTM software (version 2.0, Bruker

202

Daltonics, Bermen, Germany) was used for processing (data mining, alignment, and

203

normalization) of LC-MS/MS records. The negative ionization data matrices

204

comprising lists of peaks characterized for each sample by retention time (RT), m/z,


Page 11 of 36


205

and intensity, were obtained and total area sum normalization and Pareto scaling

206

(the square root of the standard deviation) was performed for each sample.

207

To explore the metabolite multi-dimensional data set, both unsupervised

208

principal component analysis (PCA) and supervised partial least square discriminant

209

analysis (PLS-DA) multivariate statistical methods were used after an analysis of

210

variance (ANOVA, p≤0.05). A variable importance score (VIP) greater than 1.0 was

211

chosen to select the most discriminating variables. Multivariate analyses were

212

performed in MetaboAnalyst 3.0.

213

After statistical analysis, significant peaks were subject to the identification

214

process. The elemental composition of each compound was selected according to

215

accurate masses and isotopic pattern through Smart Formula tools (Bruker Compass

216

DataAnalysis™), which provides a list of possible molecular formulas by combining

217

accurate mass and isotopic distribution reflected in their error (ppm) and mSigma

218

values, respectively. Tentative metabolite identification was performed by matching

219

the accurate m/z values and MSn fragmentation patterns with data from databases

220

(METLIN, KEGG compounds, PubChem, Mass bank, Maven, FooDB, and ReSpect)

221

and reference literature with a mass accuracy window of 5 ppm. The identity of malic

222

acid, sucrose, citric acid, hydroxibenzoic acid, chlorogenic acid, catechin, syringic,

223

epicatechin, caffeic acid, vanilic acid, rutin, quercetin, p-coumaric, ferulic, luteolin,

224

hesperetin, and pinocembrin were confirmed with external standards (Sigma-Aldrich).

225

Relative quantification was performed by comparison to internal standard 17

, except for epicatechin (y=57.8x+545.4 R2=0.9990) and rutin

226

(reserpine)

227

(y=177.4x+731.4 R2=0.9998), which were quantified using an external standard

228

calibration curve (39 – 5000 ng mL-1).

229



Page 12 of 36 11

230

Results

231

3.1

Physicochemical characteristics of Butia spp.

232

Butia spp. fruit can be yellow, green, orange, reddish-orange, bright red, or

233

purple 10. Mature B. paraguayensis and B. catarinensis fruit were more reddish, while

234

B. odorata and B. yatay had yellow fruit (Figure 1). pH values varied from 3.3 (B.

235

yatay and B. odorata) to 4.0 (B. catarinensis). Fruit of B. catarinensis were the least

236

acidic (0.4% of citric acid) while fruit of B. odorata were the most acidic (1.9% of citric

237

acid). Soluble solids content varied from 10.9 (B. yatay) to 13.7 ºBrix (B.

238

catarinensis). Fruit of B. catarinensis showed the highest SS/AC ratio (Table 1).

239

These fruit presented lower acidity and higher soluble solids when compared with the

240

other species in this study. Vitamin C content varied from 20.2 (B. catarinensis) to

241

64.6 mg of L-ascorbic acid 100g-1 (B. yatay). Carotenoid content varied from 10.7 (B.

242

odorata) to 26.7 mg of β-carotene equivalent 100 g-1 fw (B. paraguayensis). Phenolic

243

content varied from 142.4 mg (B. paraguayensis) to 200.1 mg GAE 100 g-1 (B. yatay).

244

B. paraguayensis showed the lowest total flavonoid content (61.4 mg), while B.

245

catarinensis showed the highest content (100.1 mg). The lowest level of antioxidant

246

capacity was observed in B. catarinensis fruit for both DPPH and ABTS radicals (52.1

247

and 33.2 %, respectively), and B. odorata had the highest antioxidant capacity (83.3

248

and 67.8 % for DPPH and ABTS, respectively) (Table 1). It is reported in the

249

literature that antioxidant capacity is strongly correlated with phenolic content. In this

250

study, antioxidant capacity against the DPPH radical showed intermediate correlation

251

with total phenolic content (R2=0.6433) and flavonoid content (R2=0.5712), and

252

antioxidant capacity against the ABTS radical showed low correlation with total

253

phenolic content (R2=0.4257) and intermediate with flavonoid content (R2=0.6116).

254


Page 13 of 36


255

3.2 Non-targeted LC-MS analysis of Butia spp.

256

3.2.1 Multivariate statistical analysis

257

Compared to the positive-ion mode, negative-ion MS spectra revealed higher

258

sensitivity and more observable peaks, especially for organic acids, phenolic acids,

259

and flavonoids (Figure S1 – Supplementary material). Thus all samples and

260

metabolite identifications were carried out in the negative ionization mode (Figure

261

S2). Figure 2A shows the PCA score plot for components 1 and 2. These

262

components explained 77.7% of total variance. Principal component analysis showed

263

clear differentiations among B. odorata, B. yatay, B. paraguayensis, and B.

264

catarinensis species.

265

In order to obtain more distinct separation of the samples and determine the

266

metabolites contributing to the discrimination, PLS-DA was used on the LC-MS

267

dataset. For the PLS-DA model, both Q2 max (0.86) and R2 values (0.91) were higher

268

in the permutation test than in the real model, suggesting good predictability and

269

goodness of fit. Figure 2B shows the separation of species by PLS-DA analysis.

270 271

3.2.2 Key components in Butia spp. differentiation by PLS-DA

272

VIP scores were used (VIP values > 0.96) to determine which compounds

273

contributed to the separation between groupings (Fig. 2C). Table 2 presents

274

metabolites most useful in discriminating butiá species.

275

Peaks with [M-H]- m/z 293.1239 (C12H22O8, peak 5) and [M-H]- m/z 259.0608

276

(C14H12O5, peak 7) were not identified, but were significant markers for differentiating

277

B. catarinensis and B. yatay fruit, respectively.

278

Galabiose had a parent ion of [M+M+Cl-2H]- m/z 719.2018, and fragment ions

279

of m/z 377.0851 [M-Cl]-, 341.1089 [M-H]- and 215.0318 (Fig. 3F). Galabiose content



Page 14 of 36 13

280

varied from 42.8 (B. yatay) to 190.5 (B. odorata) µg g-1 (Fig. 4). This disaccharide

281

influenced the separation of B. odorata fruit from other species.

282

Parent ions observed at [M-H]- m/z 341.1103 were identified as sucrose

283

confirmed by MS/MS with online databases and external standard. The fragment ion

284

detected at m/z 179.0553 may be produced by the neutral loss of the hexose [M–H-

285

162]-. Sucrose content varied from 151.7 to 227.9 µg g-1 (Fig. 4) and was responsible

286

for separation of B. catarinensis from the other species. This species had sweeter

287

fruit than others, as indicated by the soluble solids content (Table 1).

288

Malic acid had a parent ion in negative mode of [M-H]- m/z 133.0147 and a

289

fragment ion of m/z 115.0041 resulting from a loss of water (Fig. 3B). The

290

concentration of malic acid varied from 1975.4 (B. catarinensis) to 3818.0 (B. yatay)

291

µg g-1 (Fig. 4).

292

Citric acid had a parent ion in negative mode of [M-H]- m/z 191.0200 and

293

fragments ion of m/z 129.0195, [M-H-CO2-2H2O]- 111.0089, and 87.0090 (Fig. 3G).

294

Citric acid content varied from 295.8 (B. catarinensis) to 549.0 (B. odorata) µg g-1

295

(Fig. 4). Malic and citric acid influenced the separation of B. yatay fruit from the other

296

studied species.

297

Isopropylmalic acid was characterized by parent ion of [M-H]- m/z 175.0608

298

and fragment ions of [HCO2+CH3]- m/z 115.0389 and m/z 85.0648 (Fig. 3E)

299

influenced the separation of B. catarinensis. The content varied from 19.9 (B.

300

paraguayensis) to 124.0 µg g-1 (B. catarinensis) (Fig. 4). B. catarinensis had malic

301

and citric acid contents approximately 50% less than that of the other species while

302

isopropylmalic acid content in B. catarinensis was 84% greater than other species

303

tested.


Page 15 of 36


304

Rutin was characterized by an [M-H]- with m/z 609.1485, with a fragment ion

305

at m/z 300.0290 derived from a loss of hexose and deoxyhexose (-308 Da) (Fig. 3I).

306

B. odorata fruit had the highest rutin content, which varied from 2.5 (B. yatay) to 10.2

307

µg g-1 (B. odorata) (Fig. 4).

308

Tetrahydroxy-trans-stilbene was characterized by an [M-H]- with m/z 243.0661

309

and fragments ions of m/z 201.0555 and m/z 159.0455 (Fig. 3H), and was

310

approximately 95% higher in B. paraguayensis than other species. It is not possible

311

to distinguish between cis and trans isomers by mass spectra; however, the cis

312

isomer is highly unstable and its presence in plants is highly unlikely. Tetrahydroxy-

313

trans-stilbene content varied from 9.8 (B. catarinensis) to 254.0 µg g-1 (B.

314

paraguayensis) (Fig. 4). To the best of our knowledge, this is the first time that this

315

compound was reported in Butia spp.

316 317

3.3 Non-targeted LC-MS analysis for Butia odorata

318

3.3.1 Multivariate statistical analysis for B. odorata fruits

319

To verify how location influenced the metabolic profile, 57 genotypes of B.

320

odorata collected in Capão do Leão and Tapes, RS, Brazil, were evaluated by LC-

321

ESI-qToF-MS.

322

The first two principal components explained 35.3% of the total variation (Fig.

323

5A). Two separate groups of B. odorata genotypes were formed based on location.

324

For the PLS-DA model, both Q2 max (0.83) and R2 values (0.88) were higher in the

325

permutation test than in the real model, suggesting good predictability and goodness

326

of fit. Figure 5B shows the separation of two groups by PLS-DA analysis.

327



Page 16 of 36 15

328

3.3.2 Key components in Butia odorata differentiation by PLS-DA

329

The compounds contributing to the separation of B. odorata fruit from different

330

locations are shown in Table 3. Metabolites that presented VIP scores greater than

331

one were selected and presented in Figure S3.

332

The ion [M+M+Cl-2H]- m/z 719.2018 (peak 2, Fig. S3) and fragment ions of

333

m/z 377.0851 [M-Cl]-, 341.1089 [M-H]- and 215.0318 were tentatively identified as

334

galabiose, and fruit collected from Capão do Leão were rich in this compound.

335

Fumaric acid (peak 3, Fig. S3) was characterized by an [M-H]- with m/z

336

115.0040 and fragment ions of m/z 84.9859 and m/z 71.0146 resulting from a loss of

337

methyl ester (-31 Da) and carboxylic acid (-44 Da), respectively, contributed to

338

discrimination of fruit from Tapes.

339

Malic acid ([M-H]- m/z 133.0140) and citric acid ([M-H]- m/z 191.0197)

340

influenced separation of the fruit from Tapes. These fruit were more acidic than fruit

341

from Capão do Leão (data not shown). On the other hand, fruit from Capão do Leão

342

were sweeter than fruits from Tapes and had high sucrose content (peak 5, Fig. S3).

343

Furoic acid (peak 7, Fig. S3) with a parent ion of [M-H]- m/z 111.0090 and

344

fragment ions of m/z 67.0192 and m/z 49.0095 resulting from a loss of carboxylic

345

acid and water, 44 and 62 Da, respectively, was higher in fruit from Tapes.

346

Citramalic acid was characterized by an [M-H]- at m/z 147.0300, with a

347

fragment ion at m/z 129.0189 derived from a loss of water (-18 Da) and m/z

348

101.0242 derived from a loss of water and carbon monoxide (-46 Da), characteristic

349

of carboxylic acids. Fruit from Tapes showed higher content for this compound.

350

Isopropylmalic acid (peak 10, Fig. S3) with a parent ion of [M-H]- m/z 175.0620

351

and fragment ion of m/z 115.0389 and m/z 85.0648, resulting from a loss of C2H4O2

352

(60 Da) and C3H6O3 (90 Da), was higher in fruit from Capão do Leão (Table 3).


Page 17 of 36


353

(-)-Epicatechin (peak 12, Fig. S3) had a parent ion of [M-H]- m/z 289.0720 and

354

fragment ion of m/z 245.0821, resulting from a loss of carboxylic acid. The identity of

355

this compound was confirmed by MS/MS with online databases and external

356

standard. (-)-Epicatechin and (+)-catechin isomers were distinguished by retention

357

time matching that of the external standards and not by mass spectrum. Fruit from

358

Tapes showed higher (-)-epicatechin than fruit from Capão do Leão.

359

Peak 13 (Fig. S3) had a parent ion of [M-H]- m/z 623.1620 (C28H32O16), with a

360

main fragment ion at m/z 315.0510 derived from a loss of hexose and deoxyhexose

361

(-308 Da). This compound is an isorhamnetin derivative (m/z 315). Fruit from Tapes

362

showed high content for this compound (Table 3).

363

Peaks with [M-H]- m/z 439.0810 (C10H16N8O12, peak 1), [M-H]- m/z 293.1240

364

(C12H22O8, peak 9), [M-H]- m/z 259.1911 (C14H28O4, peak 14), and [M-H]- m/z

365

417.1555 (C22H26O8, peak 11) were not identified but were significant markers for fruit

366

from Capão do Leão.

367 368

3.4 Qualitative profile

369

In the present work, a total of 39 semi-polar compounds have been tentatively

370

characterized (Table S1 – Supplementary material). A total of nine known organic

371

acids were identified in all Butia species tested. These compounds eluted between

372

1.9 and 5.7 min. The identified compounds were fumaric ([M-H]- m/z 115.0040), malic

373

([M-H]- m/z 133.0140), citric ([M-H]- m/z 191.0200), furoic ([M-H]- m/z 111.0090),

374

aconitic ([M-H]- m/z 173.0092), citraconic ([M-H]- m/z 129.0195), citramalic ([M-H]-

375

m/z 147.0300), pantothenic ([M-H]- m/z 218.1032), and isopropylmalic acid ([M-H]-

376

m/z 175.0620). Two sugars, galabiose ([M-H]- m/z 719.2002) and sucrose ([M-H]- m/z

377

341.1090), were identified.



Page 18 of 36 17

378

Phenolic acid derivatives were found in all four species of butiá. These

379

compounds eluted between 3.1 and 7.7 min and were classified as hydroxybenzoic

380

and hydroxycinnamic derivatives. Among the hydroxybenzoic derivatives, aglycones

381

were detected including hydroxybenzoic acid ([M-H]- m/z 137.0242), syringic acid

382

([M-H]- m/z 197.0445), and vanillic acid ([M-H]- m/z 167.0347), as well as one

383

glycoside ester, hydroxybenzoic acid hexose ([M-H]- m/z 299.0070) characterized by

384

the neutral loss of the glycosidic moiety (162 Da). Hydroxybenzoic hexose was

385

identified in B. catarinensis and B. odorata fruit. Four hydroxycinnamic derivatives,

386

chlorogenic acid ([M-H]- m/z 353.0871), caffeic acid ([M-H]- m/z 179.0348), p-

387

coumaric acid ([M-H]- m/z 163.0403), and ferulic acid ([M-H]- m/z 193.0505) were

388

identified. p-coumaric acid was not detected in B. paraguayensis fruit.

389

Flavonoid compounds are divided into different classes. In butiá fruit tested,

390

flavan-3-ol, flavonol, flavone, flavanone, stilbene and flavonoid glycoside were

391

identified. The flavonoids eluted between 5.7 and 12.2 min. Three flavan-3-ol,

392

catechin

393

catechin/epicatechin dimer ([M-H]- m/z 577.1350) were identified by comparing

394

retention times relative to external standards and MS/MS fragmentation patterns.

([M-H]-

m/z

289.0720),

epicatechin

([M-H]-

m/z

289.0720),

and

395

Flavonol was the predominant group of flavonoids with seven compounds

396

identified. Four quercetin derivatives (main fragments at m/z 300.0271) were

397

identified, including rutin or quercetin-O-rutinoside ([M-H]- m/z 609.1448), quercetin-

398

O-glucoside ([M-H]- m/z 463.0885), quercetin-O-malonylglucoside ([M-H]- m/z

399

549.0881), and the aglycone quercetin ([M-H]- m/z 301.0343). In addition, three

400

kaempferol derivatives (main fragments at [M-H]- m/z 285.0404 and [M-H]- m/z

401

284.0335)

were

identified:

kaempferol-O-rutinoside

([M-H]-


m/z

593.1513),

Page 19 of 36


402

kaempferol-O-glucoside ([M-H]- m/z 447.0936), and kaempferol-O-acetylglucoside

403

([M-H]- m/z 489.1041).

404

Luteolin ([M-H]- m/z 285.0402) was the only flavone identified in butiá fruit, and

405

was found in all species. The identity of this compound was confirmed with an

406

external standard. Additionally, the flavanones pinocembrin and hesperetin were

407

identified. Pinocembrin ([M-H]- m/z 255.0659) was not identified in B. paraguayensis

408

fruit, while hesperetin ([M-H]- m/z 301.0707) was only identified in B. odorata fruit.

409

The identities of these compounds were confirmed by external standards. A flavonoid

410

glycoside tentatively identified as an isorhamnetin derivative ([M-H]- m/z 623.1620)

411

with a main fragment at m/z 315.0510 was identified. Finally, a stilbene was identified

412

as tetrahydroxy-stilbene ([M-H]- m/z 243.0667).

413 414

4 Discussion

415

In relation to physicochemical characteristics, this study demonstrated that B.

416

catarinensis fruit were the sweetest (13.7 º Brix) and B. odorata fruit the most acidic

417

with 1.9% citric acid equivalents (CAE). Other species such as B. capitata have been

418

shown to have a mean acidity of 0.4% CAE, pH 3.0

419

eriospatha pH ranged from 2.4 to 3.1, acidity from 0.4 to 1.9% CAE

420

solids from 6.4 to 9.3 º Brix 18.

18

and 9.3 º Brix

19,20

while for B.

19

and soluble

421

The ascorbic acid content found in butiá fruit (20.2 to 64.6 mg L-ascorbic acid

422

100 g-1) was similar to that reported for green peas (40 mg), cauliflower (49 mg),

423

orange (54 mg), kiwifruit (59 mg), strawberry (77 mg), and broccoli (112 mg)

424

variation in L-ascorbic acid content in this study was greater than previously found for

425

B. odorata (23 to 63 mg), B. eriospatha (21 to 31 mg), and B. capitata (38 to 73 mg)


21

. The


Page 20 of 36 19

426

3

427

recommended daily intake of L-ascorbic acid for adult women (60 mg).

. Fruit of B. yatay, B. odorata, and B. paraguayensis tested contained more than the

428

In this study, B. paraguayensis fruit had the highest carotenoid content (26.7

429

mg β-carotene equivalent 100 g-1) among species tested. Carotenoids are

430

considered health beneficial compounds, as some carotenoids are precursors to

431

vitamin A and have high antioxidant potential. These properties make fruit and

432

vegetables rich in carotenoids interesting food sources since their consumption has

433

been associated with a reduced incidence of many types of cancers, degenerative,

434

and cardiovascular diseases

435

carotenoid content (10.7 to 26.7 mg β-carotene 100 g-1) than other commonly

436

consumed fruit and vegetables such as carrot (18.3 mg), mango (13.1 mg), spinach

437

(5.6 mg), lettuce (1.3 mg), and tomato (1.2 mg) 23.

22–24

. Butia fruit tested in this study had higher

438

Butia odorata has been previously shown7 to have a total phenolic content

439

greater than 280 mg gallic acid equivalent (GAE) 100 g-1. In this study, B. yatay fruit

440

showed the highest phenolic content (200.1 mg GAE 100 g-1), followed by B. odorata

441

fruit (171.1 mg GAE 100 g-1). For total flavonoids, B. catarinensis had the highest

442

content (100.0 mg (+)-catechin equivalents 100 g-1) and B. paraguayensis fruit the

443

lowest content (61.4 mg (+)-catechin equivalent 100 g-1). Fruits and vegetables are

444

the main source of phenolic compounds in the diet, and the regular consumption of

445

these foods is increasingly associated with health benefits such as anti-cancer

446

effects which have been associated with the antioxidant capacity of these

447

compounds 25.

448

Metabolomics is currently considered one of the most popular approaches for

449

the creation of profiles based on primary and specialized metabolites in plants, and

450

has been used for the identification of markers for detection of physiological changes,


Page 21 of 36

Journal of Agricultural and Food Chemistry 20 26,27

451

genotype differences, geographic origin, and quality control

452

compounds responsible for the discrimination among Butia species were both

453

general and specialized metabolites. Primary metabolites include carbohydrates,

454

amino acids, lipids, and organic acids, which are essential for normal growth,

455

development, and reproduction of organisms. In this study, several organic acids and

456

sugars known to be associated with the degree of ripeness of fruit were identified as

457

markers of Butia spp. Sucrose is a source of energy and a signaling molecule in

458

plants

459

typically responsible for their sour taste, and are important at the cellular level with

460

roles in energy production, formation of precursors for amino-acid biosynthesis, and

461

at the whole-plant level in modulating adaptation to the environment 29.

28

. In this study, the

. Organic acids have strong organoleptic influence in fruits and vegetables,

462

Specialized metabolites are associated with plant fitness and their capacity to

463

cope with environmental conditions, and provide defense against herbivores, micro-

464

organisms, and other pathogens. Additionally, specialized metabolites found in fruits

465

and vegetables play relevant roles both in human health because of their biological

466

activities, and in influencing food color, flavor, and nutritional characteristics

467

specialized metabolite profile is unique to individuals within a species or a close

468

taxonomic group; however, it can be altered because biosynthetic pathways may be

469

influenced by environmental conditions such as climate (temperature, light, and

470

water), soil properties, and attack by pathogens and pests

471

phenolic compounds were identified in butiá fruit, and a flavonol (rutin), a flavan-3-ol

472

(epicatechin), a glycosylated flavonoid (isorhamnetin), and a stilbene (tetrahydroxy-

473

trans-stilbene) were significant markers for discrimination of geographic origin and

474

species. Flavonoids protect plants from UV-B radiation and are relevant for human

475

health acting as anti-inflammatory, anti-microbial, anti-tumor, and anti-asthma agents


31

30

. The

. In this study, several


Page 22 of 36 21

476

32

477

(phytoalexins). These compounds may also be involved as chemical signals in

478

allelopathy or in response to oxidative stress generated by UV irradiation, suggesting

479

that these compounds are important components of defense responses and may

480

constitute indicators of disease resistance 33. For human health, several studies have

481

reported that tetrahydroxy-trans-stilbene has potential activity against lung cancer,

482

with antioxidant, anti-aging, and anti-angiogenesic properties 34.

. In plants, stilbenes pre-exist or are synthesized after a pathogen attack

483

This is the first report of luteolin, pinocembrin, and hesperetin in butiá fruit.

484

Luteolin has been previously identified in other fruits in the Arecaceae family, such as

485

açai (Euterpe oleraceae) and date palm (Phoenix dactylifera). It is reported that

486

luteolin has high antioxidant, anti-inflammatory, and antimicrobial activities

487

Hesperetin occurs predominantly in citrus fruits and contributes to the antioxidant

488

capacity of orange juice 36, while pinocembrin has anticancer properties 25.

35

.

489

A metabolomic fingerprinting of date palm fruit (Phoenix dactylifera) indicated

490

flavonols (rutin and isoquercetrin) and sugars (glucose, fructose, and sucrose) as

491

biomarkers contributing to the classification of date palm fruit

492

evaluation of the profile of semi-polar compounds in four species of peppers

493

(Capsicum spp.) by LC-MS revealed differentiation among species based on acyclic

494

diterpenoid glycosides, phenolic acids (coumaric acid and ferulic acid), and

495

flavonoids (luteolin, apigenin, quercetin, and kaempferol)

496

geographic origin of Goji berry (Lycium barbarum) by metabolite profiling was

497

influenced by flavonoids and phenolic acids 39. Metabolomics followed by multivariate

498

analyses were also useful in discriminating tea 40 and blueberry 41 samples.

38

37

. Similarly, an

. Discrimination of

499

This study demonstrated that the use of LC-MS followed by multivariate

500

statistical analyses was a useful tool to differentiate Butia spp. and provided


Page 23 of 36


501

biomarkers for discrimination among B. catarinensis, B. odorata, B. paraguayensis,

502

and B. yatay. Organic acids were markers for B. yatay, rutin and galabiose for B.

503

odorata, sucrose and isopropylmalic acid for B. catarinensis, which was sweeter and

504

less acidic, and tetrahydroxy-trans-stilbene for B. paraguayensis. Currently, Butia

505

species are identified by morphological markers, which can be difficult and

506

inconclusive. Thus, metabolomics can be a useful additional tool for the identification

507

of species. However, in order to confirm the validity of the biomarkers identified in this

508

study, future evaluations should be expanded to include more Butia genotypes and

509

species.

510

B. odorata fruit from Tapes showed high content of organic acids and

511

flavonoids, while fruit from Capão do Leão had high sugar content. A previous study

512

using molecular markers observed less variation among populations than within

513

populations of Butia odorata 42. To the best of our knowledge, this was the first time

514

that a metabolomics approach was used to reveal compositional differences among

515

Butia species and between fruit from the same species grown in different locations.

516

Typically, outcrossing species have greater variation within individuals of the same

517

population than among populations

518

included, this study identified metabolic markers that can be employed to assess

519

Butia population diversity and assist in species identification.

43

. Although genetic diversity measures were not

520 521

Acknowledgements

522 523

The authors would like to acknowledge Capes and CNPq (457947/2014-4) for

524

scholarships and financial support for research.

525 526



Page 24 of 36 23

527

References

528

1. Barbieri, R.; Costa Gomes, J.; Alercia, A.; Padulosi, S. Agricultural biodiversity

529

in southern Brazil: integrating efforts for conservation and use of neglected

530

and underutilized species. Sustainability 2014, 6, 741–757.

531 532

2. Soares, K.P.; Longhi, S.J. A new Butia (Becc.) Becc. species (Arecaceae) of Rio Grande Do Sul, Brazil. Cienc Florest. 2011. 21, 203–208.

533

3. Büttow, M. V.; Barbieri, R. L.; Neitzke, R. S.; Heiden, G. Traditional knowledge

534

associated with the use of butia palm (Butia spp., Arecaceae) in the southern

535

of Brazil. Rev. Bras. Frutic. 2009, 31, 1069–1075.

536 537 538 539

4. Hoffmann, J. F.; Barbieri, R. L.; Rombaldi, C. V.; Chaves, F. C. Butia spp. (Arecaceae): An overview. Sci. Hortic. 2014, 179, 122–131. 5. Lorenzi, H.; Noblick, L.; Kahn, F.; Ferreira, E. Flora Brasileira: Arecaceae (Palmeiras). Nova Odessa: Instituto Plantarum, Brazil, 2010.

540

6. Moura, R. C.; Lopes, P. S. N.; Brandão Júnior, D. D. S.; Gomes, J. G.;

541

Pereira, M. B. Fruit and seed biometry of Butia capitata (Mart.) Beccari

542

(Arecaceae), in the natural vegetation of the North of Minas Gerais, Brazil.

543

Biota Neotrop. 2010, 10, 415–419.

544

7. Beskow, G. T.; Hoffmann, J. F.; Teixeira, A. M.; Fachinello, J. C.; Chaves, F.

545

C.; Rombaldi C. V. Bioactive and yield potential of jelly palms (Butia odorata

546

Barb. Rodr.). Food Chem. 2015, 172, 699–704.

547

8. Cruxen, C. E. dos S.; Hoffmann, J. F.; Zandona, G. P.; Fiorentini, A. M.;

548

Rombaldi, C. V.; Chaves, F.C. Probiotic butia (Butia odorata) ice cream:

549

Development, characterization, stability of bioactive compounds, and viability

550

of Bifidobacterium lactis during storage. LWT - Food Sci. Technol. 2017, 75,

551

379–385.

552

9. Mistura, C. C.; Barbieri, R. L.; Castro, C. M.; Padulosi, S.;

Alercia, A.

553

Descriptors for Butiá [Butia odorata (Barb. Rodr.) Noblick]. Bioversity

554

International, Rome (Italy); Brazilian Agriculture Research Corporation,

555

Embrapa Temperate Agriculture, Pelotas, RS (Brazil), 2015, 51 p.

556

10. Mistura, C. C.; Barbieri, R. L.; Castro, C. M.; Padulosi, S.; Alercia, A.

557

Descriptors for on-farm conservation and use of Butia odorata natural

558

populations. Plant Genet. Resour. 2015, 1–6.


Page 25 of 36


559

11. Cevallos-Cevallos, J. M.; Reyes-De-Corcuera, J. I.; Etxeberria, E.; Danyluk, M.

560

D.; Rodrick, G. E. Metabolomic analysis in food science: a review. Trends

561

Food Sci. Technol. 2009, 20, 557–566.

562

12. Oms-Oliu, G.; Odriozola-Serrano, I; Martín-Belloso, O. Metabolomics for

563

assessing safety and quality of plant-derived food. Food Res. Int. 2013, 54,

564

1172–1183.

565 566

13. AOAC (Association of Official Analytical Chemists). Official methods of analysis. Arlington, VA, USA: 2005, 18th ed.

567

14. Zhishen, J.; Mengcheng, T.; Jianming, W. The determination of flavonoid

568

contents in mulberry and their scavenging effects on superoxide radicals.

569

Food Chem. 1999, 64, 555–559.

570

15. Pereira, M. C.; Steffens, R. S.; Jablonski, A.; Hertz, P. F.; Rios, A. O.; Vizzotto,

571

M.; Flôres, S. H. Characterization, bioactive compounds and antioxidant

572

potential of three Brazilian fruits. J. Food Compos. Anal. 2013, 29, 19–24.

573

16. De Vos, R.C.; Moco, S.; Lommen, A.; Keurentjes, J.J.; Bino, R.J.; Hall, R.D.

574

Untargeted large-scale plant metabolomics using liquid chromatography

575

coupled to mass spectrometry. Nat. Protoc. 2007, 2, 778–791.

576

17. Tolstikov, V. V.; Fiehn, O.; Tanaka, N. Application of Liquid Chromatography–

577

Mass Spectrometry Analysis in Metabolomics in Methods in Molecular Biology,

578

vol. 358: Metabolomics: Methods and Protocols, Edited by: Weckwerth, W.

579

Humana Press Inc., Totowa, NJ, 2007.

580

18. Haminiuk, C. W. I.; Sierakowski, M. R.; Maciel, G. M.; Vidal, J. R. M. B.;

581

Branco, I. G.; Masson, M. L. Rheological properties of Butia pulp. Int. J. Food

582

Eng. 2006, 2, 1–8.

583

19. Magro, N. G. D.; Coelho, S. R. M.; Haida, K. S.; Berté, S. D.; Moraes, S. S. de.

584

Physicochemical comparison of frozen fruit of Butia eriospatha (Mart.) Becc.

585

from Parana and Santa Catarina - Brazil. Rev. Varia Sci. 2006, 6, 33–42.

586

20. Rigo, M.; Bezzera, J. R. M. V.; Córdova, K. R. V. Study of temperature effect

587

on rheological parameters of the butia pulp (Butia eriospatha). Ambiência

588

Guarapuava 2010, 6, 25–36.

589

21. Rickman, J. C.; Barrett, D. M.; Bruhn, C. M. Nutritional comparison of fresh,

590

frozen and canned fruits and vegetables . Part 1. Vitamins C and B and

591

phenolic compounds. J. Sci. Food Agric. 2007, 87, 930–944.



Page 26 of 36 25

592 593 594 595

22. Amorim-Carrilho, K. T.; Cepeda, A.; Fente, C.; Regal, P. Review of methods for analysis of carotenoids. TrAC Trends Anal. Chem. 2014, 56, 49–73. 23. Krinsky, N. I; Johnson, E. J. Carotenoid actions and their relation to health and disease. Mol. Aspects Med. 2005, 26, 459–516.

596

24. Reif, C.; Arrigoni, E.; Schärer, H.; Nyström, L.; Hurrell, R. F. Carotenoid

597

database of commonly eaten Swiss vegetables and their estimated

598

contribution to carotenoid intake. J. Food Compos. Anal. 2013, 29, 64–72.

599

25. Roleira, F. M. F.; Tavares-da-Silva, E. J.; Varela, C. L.; Costa, S. C.; Silva, T.;

600

Garrido, J.; Borges, F. Plant derived and dietary phenolic antioxidants:

601

Anticancer properties. Food Chem. 2015, 183, 235–258.

602

26. Ibáñez, C.; Valdés, A.; García-Cañas, V.; Simó, C.; Celebier, M.; Rocamora-

603

Reverte, L.; Gómez-Martínez, A.; Herrero, M.; Castro-Puyana, M.; Segura-

604

Carretero, A.; Ibáñez, E.; Ferragut, J. A.; Cifuentes, A. Global Foodomics

605

strategy to investigate the health benefits of dietary constituents. J.

606

Chromatogr. A 2012, 1248, 139–153.

607

27. Puiggròs, F.; Solà, R.; Bladé, C.; Salvadó, M.J.; Arola, L. Nutritional

608

biomarkers and foodomic methodologies for qualitative and quantitative

609

analysis of bioactive ingredients in dietary intervention studies. J. Chromatogr.

610

A 2011, 1218, 7399–7414.

611 612

28. Tognetti, J. A.; Pontis, H. G.; Martínez-Noel, G. M. A. Sucrose signaling in plants: A world yet to be explored. Plant Signaling & Behavior. 2013, 8, 1-10.

613

29. López-Bucio, J.; Nieto-Jacobo, M. F.; Ramírez-Rodríguez, V.; Herrera-

614

Estrella, L. Organic acid metabolism in plants: from adaptive physiology to

615

transgenic varieties for cultivation in extreme soils. Plant Sci. 2000, 60, 1–13.

616 617

30. Hounsome, N.; Hounsome, B.; Tomos, D.; Edwards-Jones G. Plant metabolites and nutritional quality of vegetables. J Food Sci. 2008, 73, 48–65.

618

31. Sheshadri, S. A.; Nishanth, M. J.; Simon B. Stress-Mediated cis-Element

619

Transcription Factor Interactions Interconnecting Primary and Specialized

620

Metabolism in planta. Front. Plant Sci. 2016, 7:1725.

621 622 623 624

32. Jiang, N.; Doseff, A. I.; Grotewold, E. Flavones: from biosynthesis to health benefits. Plants 2016, 5(2), 27. 33. Chong, J.; Poutaraud, A.; Hugueney, P. Metabolism and roles of stilbenes in plants. Plant Sci. 2009, 177, 143–155.


Page 27 of 36


625

34. Kasiotis, K. M.; Pratsinis, H.; Kletsas, D.; Haroutounian, S. A. Resveratrol and

626

related stilbenes: Their anti-aging and anti-angiogenic properties. Food Chem.

627

Toxicol. 2013, 61, 112–120.

628 629

35. López-Lázaro, M. Distribution and biological activities of the flavonoid luteolin. Mini-Rev. Med. Chem. 2009, 9, 31-59.

630

36. Erlund, I. Review of the flavonoids quercetin, hesperetin, and naringenin.

631

Dietary sources, bioactivities, bioavailability, and epidemiology. Nutr. Res.

632

2004, 24, 851–874.

633

37. Farag, M. A.; Mohsen, M.; Heinke, R.; Wessjohann, L. A. Metabolomic

634

fingerprints of 21 date palm fruit varieties from Egypt using UPLC/PDA/ESI-

635

qTOF-MS and GC-MS analyzed by chemometrics. Food Res. Int. 2014, 64,

636

218–226.

637

38. Wahyuni, Y.; Ballester, A. R.; Tikunov, Y.; de Vos, R. C. H.; Pelgrom, K. T. B.;

638

Maharijaya, A.; Sudarmonowati, E.; Bino, R. J.; Bovy, A. G. Metabolomics and

639

molecular marker analysis to explore pepper (Capsicum sp.) biodiversity.

640

Metabolomics 2013, 9, 130–144.

641

39. Bondia-Pons, I.; Savolainen, O.; Törrönen, R.; Martinez, J. A.; Poutanen, K.;

642

Hanhineva, K. Metabolic profiling of Goji berry extracts for discrimination of

643

geographical origin by non-targeted liquid chromatography coupled to

644

quadrupole time-of-flight mass spectrometry. Food Res. Int. 2014, 63, 1–7.

645

40. Ku, K. M.; Choi, J. N.; Kim, J.; Kim, J. K.; Yoo, L. G.; Lee, S. J.; Hong, Y. S.;

646

Lee, C. H. Metabolomics analysis reveals the compositional differences of

647

shade grown tea (Camellia sinensis L.). J. Agric. Food Chem. 2010, 58, 418–

648

426.

649

41. Ma, C.; Dastmalchi, K.; Flores, G.; Wu, S. B.; Pedraza-Peñalosa, P.; Long, C.;

650

Kennelly, E. J. Antioxidant and metabolite profiling of North American and

651

neotropical blueberries using LC-TOF-MS and multivariate analyses. J. Agric.

652

Food Chem. 2013, 61, 3548−3559.

653

42. Buttow, M. V.; Castro, C. M.; Schwartz, E.; Tonietto, A.; Barbieri, R. L.

654

Molecular characterization of Butia capitata populations (Arecaceae) in the

655

Southern Brazil estimated by AFLP analysis. Rev. Bras. Frutic. 2010, 32, 230–

656

239.



Page 28 of 36 27

657

43. Cortese, L. M.; Honig J.; Miller, C.; Bonos, S.A. Genetic diversity of twelve

658

switchgrass

populations

using molecular and morphological markers.

659

Bioenerg. Res. 2010, 3, 262–271.

660


Page 29 of 36


661

662 663 664 665 666 667 668 669 670 671 672

Table 1. Physicochemical characteristics of Butia spp. B. odorata+ B. paraguayensis++ B. catarinensis++ B. yatay++ ºHUE 74.5±0.6 b 50.1±1.5 c 66.9±3.0 bc 88.5±3.8 a pH 3.3±0.0 b 3.4±0.0 b 4.0±0.0 a 3.3±0.0 b 1 12.5±0.1 ns 11.7±0.2 13.7±0.3 10.9±0.2 SS AC2 1.9±0.0 a 1.0±0.0 b 0.4±0.0 c 1.3±0.0 ab SS/AC 7.1±0.2 c 11.5±0.3 b 32.6±1.6 a 8.5±0.2 c 58.3±2.4 a 56.7±0.1 a 20.2±0.1 b 64.6±0.1 a ASC3 4 CAR 10.7±0.5 b 26.7±0.5 a 11.8±0.5 b 10.9±0.2 b PHE5 171.1±2.6 ns 142.4±2.8 160.8±3.2 200.1±1.9 99.4±4.1 ns 61.4±0.6 100.0±1.7 90.7±0.6 FLA6 DPPH7 83.3±1.4 a 57.4±0.4 ab 52.1±0.7 b 82.6±1.0 ab ABTS7 67.8±1.5 a 36.5±0.4 b 33.2±2.1 b 64.5±0.4 a Results expressed as mean ± standard error of the mean (+ n=171; ++ n=3). Different letters within the same row indicate significant differences by Tukey’s test (p≤0.05). ns= not significant at p≤0.05. ºHUE= tonality, pH= Hydrogen potential; SS=soluble solids; AC= acidity; SS/ AC = soluble solids/acidity ratio; FIB= dietary fiber; ASC= ascorbic acid; CAR= total carotenoid; PHE= total phenolic; FLA= total flavonoid; DPPH and ABTS = antioxidant capacity. 1 ºBrix; 2 g citric acid 100g-1 fresh weight (fw); 3 mg of L-ascorbic acid equivalents 100 -1 g fw; 4 mg β-carotene equivalent 100 g-1 fw; 5 mg gallic acid equivalent 100g-1 fw; 6 mg catechin equivalent 100g1 fw; 7 % inhibition.



Page 30 of 36

29

673

674 675

Table 2. Compounds differentiating Butia spp. # RT Measured Theoretical Error Molecular mSigma Compound VIP (min) m/z m/z (ppm) Formula score 1 1.63 719.2018 719.2005 -0.3 53.0 1.167 C24H44ClO22 Galabiose 2 1.95 133.0147 133.0142 -2.3 C4H6O5 2.9 Malic acid 3.387 3 2.07 341.1103 341.1089 0.8 C12H22O11 14.9 Sucrose 1.834 4 2.52 191.0200 191.0197 -1.3 C6H8O7 1.2 Citric acid 1.154 5 4.85 293.1239 293.1242 1.6 C12H22O8 3.7 Unknown 3.710 6 5.74 175.0608 175.0615 1.7 C7H12O5 8.4 Isopropylmalic acid 1.658 7 6.56 259.0608 259.0612 -0.4 C14H12O5 3.7 Unknown 2.599 8 6.87 609.1485 609.1461 -0.9 C27H30O16 55.3 Rutin 0.950 9 7.73 243.0661 243.0663 -1.6 C14H12O4 3.0 Tetrahydroxy-trans-stilbene 1.049 RT = retention time; VIP= variable importance in the projection. mSigma = fit between measured and theoretical isotopic pattern. The smaller the mSigma value the better the isotopic fit.


Page 31 of 36


30

676

677 678 679

Table 3. Compounds differentiating Butia odorata fruit collected from two locations in RS, Brazil. Molecular formula

mSigma

Compound

VIP score

32.0

C10H16N8O12

10.7

Unknown

2.812

Quantification (µg g-1) CAP TAP 89.7* 44.5

41.6

C24H44ClO22

31.1

Galabiose

1.123

94.4*

15.8

C4H4O4

5.7

Fumaric acid

1.478

2140.0* 2363.5

16.7

C4H6O5

5.5

Malic acid

2.088

4015.6* 4257.4

27.1

C12H22O11

2.2

Sucrose

1.218

241.4*

206.0

19.6

C6H8O7

0.4

Citric acid

1.529

536.0*

593.0

15.6

C5H4O3

11.1

Furoic acid

1.425

174.6*

201.1

17.4

C5H8O5

0

Citramalic acid

1.520

202.4*

292.9

24.7

C12H22O8

4.5

Unknown

1.476

359.7*

249.7

18.8

C7H12O5

17.6

Isopropylmalic acid

1.074

36.4*

23.0

30.9

C22H26O8

23.9

Unknown

1.314

31.0*

10.8

24.5

C15H14O6

4.0

(-)-Epicatechin

1.392

85.8*

120.0

38.7

C28H32O16

11.7

Isorhamnetin derivative

1.598

37.7*

71.4

-

C14H28O4

28.2

Unknown

1.409

21.5*

3.5

# RT (min)

Measured Theoretical Error Fragment ion Energy of m/z m/z (ppm) m/z collision (eV)

1

1.56

439.0855

439.0855

0.1

2

1.63

719.2002

719.2045

-0.3

3

1.92

115.0040

115.0037

3.7

4

1.95

133.0140

133.0142

-0.2

5

2.07

341.1090

341.1089

-0.6

6

2.52

191.0200

191.0197

-1.3

7

2.53

111.0090

111.0088

-1.4

8

2.69

147.0300

147.0299

-3.6

9

4.85

293.1240

293.1242

1.1

10

5.74

175.0620

175.0625

1.7

11

5.85

417.1542

417.1555

3.1

12

6.15

289.0720

289.0718

-1.2

13

7.37

623.1620

623.1649

-4.4

14

11.63

259.1911

259.1911

1.4

259.0241 377.0843 341.1077 179.0562 84.9859 71.0146 115.0038 89.0249 71.0152 179.0521 119.0359 111.0089 87.0094 67.0192 49.0095 129.0189 101.0242 85.0296 101.0248 89.0253 115.0389 85.0648 341.0715 177.0163 154.0275 245.0821 109.0298 315.0510 101.0238 -

58.0

RT = retention time; VIP= variable importance in the projection; CAP = fruit of B. odorata from Capão do Leão; TAP = fruit of B. odorata from Tapes. * Significant difference by LSD test at p≤0.05



Page 32 of 36

31

Figure 1. Fruit of Butia spp. A) B. catarinensis (photo by M.P Eslabão). B) B. paraguayensis (photo by G. Heiden). C) B. yatay (photo by G. Heiden). D) B. odorata (photo by G.T. Beskow).


Page 33 of 36


32

Figure 2. Principal component analysis score plot (A), partial least square (PLS-DA) (B), and VIP scores by PLS-DA (C) derived from LC–MS data using negative electrospray ionization of Butia spp. fruit.



Page 34 of 36

33

Figure 3. Mass spectra (MS/MS) in the negative ion mode of discriminating compounds in Butia spp. fruit. A = unknown (m/z 293.1239); B= malic acid (m/z 133.0147); C= unknown (m/z 259.0608); D= sucrose (m/z 341.1103); E= isopropylmalic acid (m/z 175.0608); F= galabiose (m/z 719.2018); G= citric acid (m/z 191.0200); H= tetrahydroxy-trans-stilbene (m/z 243.0661); I= rutin (m/z 609.1485).


Page 35 of 36


34

600 450

a B. catarinensis B. odorata B. paraguayensis B. yatay a

a b

c

cc

b

2000

b c

b

b

1000

c

c

-1

c

a a

B. catarinensis B. odorata B. paraguayensis B. yatay

15

µg g

µg g-1

bc

3000

a

b

300 150

a

4000

a

bc

ab

a a

a 10

100

5

b

b

50

c

b

b b b

0 Galabiose

Rutin

Sucrose

m/z 293.1237 m/z 259.0613

b

b

0 Citric acid 2-Isopropylmalic acid Malic acid

Figure 4. Quantitative analysis (µg g-1) of discriminating compounds in Butia spp. fruit.


Stilbene


Page 36 of 36

35

Figure 5. Principal component analysis score plot (A), partial least square (PLS-DA) (B) and VIP scores by PLS-DA (C) derived from LC–MS data using negative electrospray ionization of 57 genotypes of Butia odorata fruit collected from two locations in RS, Brazil.


Butia spp. (Arecaceae) LC-MS-based metabolomics for species and

Recommend Documents