A Targeted Mass Spectrometry-based Metabolomics Approach toward

Publication Date (Web): September 16, 2018 ... alternations among oranges, mandarins and grapefruits revealed that host responses to CLas were differe...
0 downloads 0 Views 660KB Size
Subscriber access provided by UNIV OF FINDLAY

Omics Technologies Applied to Agriculture and Food

A Targeted Mass Spectrometry-based Metabolomics Approach toward the Understanding of Host Responses to Huanglongbing Disease Wei-Lun Hung, and Yu Wang J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.8b04033 • Publication Date (Web): 16 Sep 2018 Downloaded from http://pubs.acs.org on September 18, 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 40

Journal of Agricultural and Food Chemistry

A Targeted Mass Spectrometry-based Metabolomics Approach toward the Understanding of Host Responses to Huanglongbing Disease

Wei-Lun Hung,†‡ and Yu Wang*†

†Citrus Research and Education Center, Department of Food Science and Human Nutrition, University of Florida, Lake Alfred, FL 33850, USA ‡School of Food Safety, Taipei Medical University, 250 Wu-Hsing Street, Taipei, 11031, Taiwan

*Please send all correspondence to: Dr. Yu Wang Citrus Research and Education Center Department of Food Science and Human Nutrition University of Florida 700, Experiment Station Rd, Lake Alfred, FL 33850 USA Phone: 863-956-8673 Fax: 863-956-4631 E-mail: [email protected]

1

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

1

ABSTRACT

2

Candidatus Liberibacter asiaticus (CLas) is the major culprit of Huanglongbing (HLB), the most

3

destructive citrus disease worldwide. Polymerase chain reaction (PCR) is the most common

4

method for detecting the presence of CLas in the tree. However, due to uneven distribution of

5

bacteria and a minimum bacterial titer requirement, an infected tree may test false negative. Thus,

6

our current study profiled primary and secondary metabolites of CLas-free leaves harvested from

7

citrus undercover protection system (CUPS) to prevent a misjudgment of CLas infection.

8

Functional enrichment analysis revealed several metabolic pathways significantly affected by

9

CLas infection, mainly biosynthesis of amino acids and secondary metabolites. Comparisons of

10

CLas-infected metabolite alterations among oranges, mandarins and grapefruits revealed that

11

host responses to CLas were different. The metabolite signature highlighted in this study will

12

provide a fuller understanding into how CLas bacteria affects biosynthesis of primary and

13

secondary metabolites in different hosts.

14

15

Keywords: citrus, metabolomics, Huanglongbing, Candidatus Liberibacter, metabolic pathway

16

2

ACS Paragon Plus Environment

Page 2 of 40

Page 3 of 40

Journal of Agricultural and Food Chemistry

17

INTRODUCTION

18

Huanglongbing, also known as citrus greening disease, is a devastating disease affecting all

19

varieties of citrus worldwide. This destructive disease is associated with infection of the gram-

20

negative bacteria Candidatus Liberibacter asiaticus (CLas) transmitted by the Asian citrus

21

psyllid (ACP) Diaphorina citri, the vector of HLB. HLB was first reported in southern China in

22

1919, and subsequently has spread throughout to different countries in Africa, Asia, South and

23

North Americas.1 In the United States, HLB was first confirmed in south Miami-Dade County in

24

2005, and it is present in all citrus-growing areas including Texas and California.2-3 Different

25

symptoms can be found in all parts of the tree including leaves, twigs and fruits. In leaves, the

26

typical symptoms include blotchy mottles, yellow veins and green islands, ultimately followed

27

by the death of the entire plant (Figure S1).4 Furthermore, HLB also profoundly affects flavor

28

and the taste quality of citrus fruits. Several studies have found that HLB-affected oranges had a

29

lower soluble solids content (SSC)/titratable acid (TA), resulting in the HLB-affected juice being

30

perceived as less sweet than the juice from healthy trees.5 Meanwhile, descriptive sensory

31

evaluation from 16 panelists indicated that HLB-affected orange juice was perceived as more

32

bitter than healthy orange juice due to high contents of limonoids and flavonoids.5-6 Recently,

33

sensory descriptions of active taste compounds of orange juice isolated from centrifugal partition

34

chromatography and preparative high performance liquid chromatography (HPLC) suggested

35

that hydroxycinnamic acids and other compounds may also contribute to the bitterness of HLB-

36

affected Valencia orange juice in addition to limonoids and flavonoids.7

37

Since HLB is one of the most destructive diseases in citrus, different strategies have been

38

developed to delay the progression of HLB, including antimicrobials, thermotherapy and

39

nonantimicrobial compounds.8 For example, a recent study documented the effects of

3

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

40

oxyteracycline hydrochloride trunk injections on HLB-affected Hamlin orange trees.9 The

41

population density of CLas remained significantly lower in treated trees compared to that from

42

untreated controls for 9 months. In addition, thermotherapy appears to be a potential strategy to

43

suppress phytopathogen titer, at least, in the green house and growth chamber.10-11 Combined

44

with applications of broad-spectrum insecticides and foliar nutrients, fruit yield increased

45

significantly.12 Although these treatments could delay the progression of HLB, there is no cure to

46

eliminate this disease. Therefore, according to the latest statistical report from the Florida

47

Department of Agriculture and Consumer Services, citrus production in Florida decreased from

48

94.2 million boxes in the 2015-2016 season to 78.1 million boxes in the 2016-2017 season.13

49

Over the last decade, metabolomics has emerged as a comprehensive approach to achieve

50

a global view of biological systems through the profiling of metabolites from bodily fluids, cells

51

and tissues. Among different methodologies used for metabolite discovery, mass spectrometry is

52

an excellent analytical platform for metabolomics analysis due to its high sensitivity and

53

versatility.14 In recent years, liquid chromatography or gas chromatography coupled with mass

54

spectrometry has been employed to understand plant’s responses to HLB. For example, a

55

targeted metabolomics approach using high performance liquid chromatography-mass

56

spectrometry (LC-MS) was carried out to profile alterations of secondary metabolites in sweet

57

orange leaves infected with CLas.15 In flavoromics, changes of key aroma compounds in citrus

58

fruits after CLas infection have been characterized using gas chromatography-mass spectrometry

59

(GC-MS).6, 16 Our recent research revealed a positive link between lipid oxidation products and

60

long-chain fatty acids in response to HLB using a targeted metabolomics approach.17

61

Polymerase chain reaction (PCR) technique is the most common method used to detect

62

the presence of CLas in citrus trees.18 However, limitations of PCR such as requirement of a

4

ACS Paragon Plus Environment

Page 4 of 40

Page 5 of 40

Journal of Agricultural and Food Chemistry

63

minimum bacterial titer and uneven distribution of bacteria throughout the tree make PCR prone

64

to false negatives, especially at early stages of infection. A nuclear magnetic resonance (NMR)-

65

based metabolomics study revealed that healthy Hamlin oranges could not be completely

66

separated from asymptomatic fruits in the partial least squares discriminant analysis (PLS-DA)

67

model, suggesting that some of the healthy fruits were possibly false negatives due to uneven

68

distribution of bacteria.19 Our recent metabolomics study has also compared volatile and

69

nonvolatile profiles of healthy Hamlin orange fruits collected from citrus undercover protection

70

systems (CUPS) with HLB asymptomatic and symptomatic fruits.16 CUPS, consisting of a pole

71

and cable frame architecture, physically prevents the trees from coming into contact with ACP.20

72

However, no consistent pattern could be observed in both volatile and nonvolatile metabolites as

73

compared to previous studies, indicating that the metabolite profile of Hamlin orange fruits

74

significantly changed earlier than detecting CLas infection by PCR. To better understand how

75

CLas infection affects metabolite changes in the different hosts and prevent a misjudgment of

76

health status due to uneven distribution of CLas and detection limits of PCR, our current study

77

profiled primary and secondary metabolites of CLas-free leaves from Hamlin oranges, Murcott

78

mandarins and Ray Ruby grapefruits, harvested from CUPS. Comparisons of metabolite profiles

79

of truly CLas-free trees and their corresponding CLas-infected trees will provide a fuller

80

understanding of the host responses to CLas bacteria in different citrus species as well as acquire

81

useful information for developing potential treatments against HLB.

82

83

MATERIALS AND METHODS

84

Chemical and reagents

5

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

Page 6 of 40

85

All solvents used in this study were LC-MS grade and purchased from Fisher Scientific

86

(Waltham, MA, U.S.A). Authentic standards of organic acids, amino acids, sugars and sugar

87

alcohols were purchased from Sigma Co. (St. Louis, MO, U.S.A). Authentic standards of

88

eriocitrin, narirutin, hesperidin, didymin, hesperetin, isosinensetin, sinensetin, 5,6,7,3’,4’,5’-

89

hexamethoxyflavone, nobiletin, isosakuranetin and neoeriocitrin were purchased from Indofine

90

Chemical Co. (Hillsborough, NJ, U.S.A). Rutin, apigenin, diosmetin, limonin, luteolin, naringin,

91

neohesperidin and poncirin were sourced from Sigma Co. Meranzin, isomeranzin, tetramethoxy-

92

o-isoscutellarein, tetramethoxy-o-scutellarein and 3,5,6,7,8,3’,4’-heptamethoxyflavone were

93

purchased from Yuanye Biotech (Shanghai, China). Scopoletin was purchased from ChromaDex

94

(Irvine,

95

hydrochloride were purchased from Sigma Co.

96

Plant materials

97

Leaf samples from Hamlin orange (Citrus sinensis (L.) Osbeck), Murcott mandarin (Citrus

98

reticulata Blanco) and Ray Ruby grapefruit (Citrus paradisi Macf.) used in this study were

99

grown at the University of Florida’s Citrus Research Education Center (CREC, Lake Alfred,

100

Florida). All trees were planted in August 2014 and leaf samples were harvested in June 2016.

101

The healthy samples were collected from CUPS to guarantee the tree was truly CLas-free,20

102

while the CLas-infected samples were collected from the field. CUPS and the field are located at

103

CREC. Three plants (n=3) from each species were sampled from CUPS (healthy, CLas-free) and

104

the field (CLas-infected). The CLas-free and Clas-infected leaves were from the same rootstock.

105

The rootstocks of Hamlin oranges, Murcott mandarins and Ray Ruby grapefruits were Swingle,

106

Kuharske and Sour orange, respectively. Several leaves were collected from different locations

107

(top, middle and bottom) of each tree. Typical symptoms of HLB could be found in CLas-

CA,

U.S.A).

N-methyl-N-(trimethylsily)trifluoacetamide

6

ACS Paragon Plus Environment

and

methoxyamine

Page 7 of 40

Journal of Agricultural and Food Chemistry

108

infected leaves, including blotchy mottles, green islands and yellow veins (Figure S1). Leaf

109

samples were immediately frozen in liquid nitrogen and then stored at -80

110

Leaves from different locations of each tree were pooled and ground before metabolite extraction.

111

The presence of CLas in the infected trees and its absence in healthy trees were confirmed by

112

qPCR.18 The qPCR data of CLa-infected leaves revealed average cycle threshold values of

113

26.91±0.49 for Hamlin oranges, 27.51±1.02 for Murcott mandarins and 26.81±2.74 for Ray

114

Ruby grapefruits. Generally, Ct values under 30 are considered confidently positive, while

115

higher numbers are considered ambiguous.21

116

Sugar, sugar alcohol and organic acid analysis

117

Thirty three milligrams of ground leaves were extracted with 470 µL 80% methanol containing

118

20 µL adonitol as internal standard (1 mg/mL in methanol) by agitation for 10 min. After

119

centrifugation at 1000g for 10 min at 4 o C, the supernatant was collected and the solvent and

120

moisture were removed by a SpeedVac evaporator (Thermo Scientific, Waltham, MA). The

121

dried residue was mixed with 30 µL of methoxyamine hydrochloride (20 mg/mL in pyridine)

122

and

123

(trimethylsily)trifluoacetamide was added to the mixture and then shaken for an additional 30

124

min prior to GC-MS analysis. A 7890 gas chromatograph coupled with an Agilent 5975C mass

125

spectrometer (Santa Clara, CA, U.S.A) was employed to analyze sugar, sugar alcohols and

126

organic acids. The carrier gas was helium and a split ratio was 1:20 with a 1.1 mL/min flow rate.

127

The temperature of the ion transfer line and injection port was set at 230 oC. An Rxi-5 MS

128

column (30 m Χ 0.25 mm; 0.25 µm film thickness, Restek, Bellefonte, PA, U.S.A) was carried

129

out to separate analytes. The oven temperature was initially set at 70 oC for 5 min, and then

130

ramped up by 4 oC /min to 270 oC, followed by 20 oC /min to 320 oC and held for 5 min. A

then

shaken

for

2

h

at

room

temperature.

7

ACS Paragon Plus Environment

Finally,

80

o

C until analysis.

µL

N-methyl-N-

Journal of Agricultural and Food Chemistry

131

solvent delay of 5 min was applied. Scan of mass spectrometry was ranged from m/z 60 to 650.

132

Compound identification was confirmed by the retention time of authentic standards and the

133

NIST library. Relative concentrations of sugars, sugar alcohols and organic acids were semi-

134

quantitated based on the concentration of the internal standard.

135

Amino acid analysis

136

Fifty milligrams of ground leaves were extracted with 1mL 80% methanol containing 10 µL

137

theanine (1 mg/mL, internal standard) by sonication for 30 min. After centrifugation at 5,000g

138

for 5 min at 4 o C, the supernatant was filtered through a 0.22 µm nylon filter prior to LC-MS

139

analysis. The analytical method was performed as described by our previous work.16 In brief, a

140

Thermo Ultimate 3000 HPLC equipped with a Thermo TSQ Quantiva triple quadrupole

141

electrospray ionization tandem mass spectrometer (Thermo Scientific) was carried out for

142

analyzing amino acids. Chromatographic separations were achieved using a Tosoh TSKgel

143

Amide-80 column. The mobile phase consisted of 2.5 mM ammonium formate and 2.5 mM

144

ammonium acetate in 90% acetonitrile aqueous solution containing 0.15 % formic acid (A) and

145

2.5 mM ammonium formate and 2.5 mM ammonium acetate in 90% acetonitrile containing

146

0.15% formic acid (B). The gradient program was set as follows: 0-10 min, 100-90% B; 10-16

147

min, 90-80% B; 16-20 min, 80-50% B; 20-24 min, 50% B. The injection volume was 5 µL.

148

Authentic standards were directly infused into the mass spectrometer at a flow rate of 0.2

149

mL/min. The product ions, collision energy and RF lens of each analyte were optimized using

150

TSQ Quantiva Tune software (Thermo Scientific).The SRM transitions, collision energy, RF

151

lens and retention time of amino acids are given in Table S1. Relative concentrations of amino

152

acids were semi-quantitated based on the concentration of the internal standard

153

Flavonoid, coumarin and limonoid analysis 8

ACS Paragon Plus Environment

Page 8 of 40

Page 9 of 40

Journal of Agricultural and Food Chemistry

154

One hundred milligrams of ground leaves were extracted with 1 mL methanol containing 10 µL

155

catechin (1 mg/mL, as internal standard) by sonication for 15 min. After centrifugation at 5000g

156

for 10 min, the supernatant was collected and solvent was evaporated by a speedVac evaporator.

157

The dried residue was reconstituted in 500 µL methanol and then subjected to a C18 solid phase

158

extraction cartridge (100 mg, Restek, Bellefonte, PA, U.S.A) and then eluted with 500 µL

159

methanol. The eluents were combined and 10 µL was injected into LC-MS (TSQ Quantiva,

160

Thermo Scientific). The analytical method was modified from our previous work.16 In brief,

161

chromatographic separations were carried out using a Phenomenex Gemini C18 column (3 µm, 3

162

Χ 150 mm, Phenomenex, Torrance, CA. U.S.A) with a mobile phase consisting of 0.1% formic

163

acid in water (A) and 0.1% formic acid in acetonitrile (B). The gradient program was set as

164

follows: 0-20 min, 5-75% B, 25-26 min, 75-95% B, 26-33 min, 95% B. The flow rate and

165

injection volume were set at 0.2 mL/min. The SRM transitions, collision energy, RF lens and

166

retention time of the analytes are given in Table S2. Relative concentrations of the analytes were

167

semi-quantitated based on the concentration of the internal standard.

168

Statistical Analysis

169

Data are expressed as means of relative concentrations of the metabolites compared to that of

170

healthy leaves. Significant differences were statistically detected by Student’s t test using MS

171

Excel (Version 2016, Microsoft, Redmond, Washington, U.S.A). A significant difference was

172

considered when p < 0.05. PLS-DA was performed using SIMCA-P software (Umea, Sweden)

173

with mean centering and unit variance scaling. The confidence of tolerance ellipse in the PLS-

174

DA model was 95% based on Hotelling’s T-squared distribution. The quality of the PLS-DA

175

model was described in terms of three parameters. R2X and R2Y were used to quantitate the

176

good-of-fit and Q2 was used to assess the predictability of the model. To rule out the non-

9

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

177

randomness of separation between groups, a 100-iteration random permutation test was also

178

carried out to validate the PLS-DA model. After a log transformation, the heat map was

179

performed using MetaboAnalyst 4.0 (http://www.metaboanalyst.ca/). Metabolic pathways were

180

constructed using Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database

181

(https://www.genome.jp/kegg/) as a reference.

182 183

RESULTS

184

Alterations of organic acids, sugar, and sugar alcohols in citrus leaves by CLas-infection

185

GC-MS was employed to analyze organic acids, sugars and sugar alcohols in citrus leaves after

186

metabolite derivatization. A total of 10 organic acids, 6 sugars and 3 sugar alcohols were

187

analyzed. In CLas-free Hamlin oranges and Murcott mandarin, no significant changes could be

188

found in most of organic acids as compared to the CLas-infected leaves (Table 1). The levels of

189

fumaric acid and malic acid in CLas-free Hamlin oranges were significantly higher than that

190

from CLas-infected oranges, while the level of citric acids in Murcott mandarins and Ray Ruby

191

grapefruits significantly increased after CLas infection. In Ray Ruby grapefruit, levels of maleic

192

acid, succinic acid, glyceric acid and quinic acid significantly decreased after CLas infection. In

193

the sugar profile, the concentration of melibiose significantly increased in both CLas-infected

194

Hamlin oranges and Murcott mandarins. In grapefruits, CLas-infected leaves had lower levels of

195

glucose, fructose, galactose and myo-inositol as compared to CLas-free leaves (Table 1).

196

Alterations of amino acids in citrus leaves by CLas-infection

197

Amino acids are crucial primary metabolites directly involved in plant growth and metabolism.

198

Here, we used hydrophilic interaction chromatography coupled with MS to determine amino

199

acids in leaves. The results of amino acids in citrus leaves are given in Table 2. The levels of

10

ACS Paragon Plus Environment

Page 10 of 40

Page 11 of 40

Journal of Agricultural and Food Chemistry

200

most of amino acids generally increased after CLas infection regardless of citrus species. In

201

Hamlin orange leaves, phenylalanine, leucine, isoleucine, methionine, valine, alanine, threonine

202

and glycine in CLas-free leaves were significantly higher than that from CLas-infected leaves.

203

Similarly, CLas infection significantly elevated production of several amino acids in Ray Ruby

204

grapefruit leaves, including leucine, methionine, proline, glutamine, asparagine, arginine and

205

lysine. Only isoleucine and valine of Murcott mandarin leaves significantly increased after CLas

206

infection (Table 2).

207

208

Alterations of flavonoids, coumarins, furanocoumarins and limonoids in citrus leaves by CLas-

209

infection

210

LC-MS was used to identify different classes of flavonoids, including flavanones, flavones and

211

flavonols, in which both flavonoid aglycones and their corresponding glycosides were analyzed

212

in this study. CLas infection increased accumulation of most of flavonoids in Hamlin oranges

213

and Murcott mandarins (Table 3). Levels of naringenin, apigenin, diosmetin, 3,5,6,7,8,3’,4’-

214

heptamethoxyflavone and rutin in Hamlin oranges significantly increased after CLas infection.

215

Similarly, eriocitrin, hesperidin, hesperetin, isosakuranetin, diosmin, diosmetin, nobiletin,

216

tangeretin, rutin and quercetin in CLas-infected Murcott mandarins were significantly higher

217

than that from CLas-free leaves. Conversely, CLas infection decreased accumulation of many

218

flavonoids in Ray Ruby grapefruits, especially flavanones. Levels of neoeriocitrin, narirutin,

219

naringin, didymin, poncirin and isosakuranetin in Ray Ruby grapefruits significantly decreased

220

after CLas-infection. It should be noted that neohesperidin, naringin, poncirin, neodiosmin could

221

be only found in Ray Ruby grapefruits. In coumarins, CLas infection significantly increased

222

scopoletin accumulation in Hamlin oranges. Furanocoumarins are specific groups of secondary 11

ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry

223

metabolites commonly present in grapefruits.22 Our results showed that both 6’,7’-

224

epoxybergamottin and bergamottin had higher levels in CLas-free Ray Ruby grapefruits as

225

compared to CLas-infected leaves. Limonin is one of the major limonoids that contribute to the

226

bitterness of citrus fruits.6 Limonin accumulation was significantly elevated in CLas-infected

227

Hamlin oranges (Table 3).

228

229

Multivariate analysis of metabolites

230

First, Venn diagrams were constructed to illustrate the overlapping metabolites significantly

231

altered after CLas infection (Figure 1). In total, CLas infection significantly altered 21, 15 and 26

232

metabolites in Hamlin oranges, Murcott mandarins and Ray Ruby grapefruits, respectively. A

233

total of 6 metabolites, including melibiose, valine, isoleucine, diosmetin, tangeretin and rutin,

234

were significantly altered in both Hamlin oranges and Murcott mandarins after Clas infection,

235

while only two metabolites were significantly changed in both Hamlin oranges and Ray Ruby

236

grapefruits. However, once CLas infected, no metabolites were simultaneously altered in Hamlin

237

oranges, Murcott mandarins and Ray Ruby grapefruits. Similarly, 16 and 14 metabolites

238

significantly accumulated in Hamlin oranges and Murcott mandarins, respectively, while only 9

239

metabolites significantly increased in Ruby Ray grapefruits. It is noteworthy that CLas infection

240

significantly decreased accumulation of 17 metabolites in Ray Ruby grapefruits, while only 5

241

and 1 metabolites in Hamlin oranges and Murcott mandarins significantly decreased (Figure 1).

242

Meanwhile, PLS-DA was employed to determine whether metabolite profiles significantly

243

changed by CLas infection. First, the clusters of Hamlin oranges, Murcott mandarins, and Ray

244

Ruby grapefruits were separated from each other (Figure 2), indicating that metabolite profiles

245

were significantly different depending on the citrus species. R2X, R2Y and Q2 values were 0.907, 12

ACS Paragon Plus Environment

Page 12 of 40

Page 13 of 40

Journal of Agricultural and Food Chemistry

246

0.982 and 0.800, respectively. Y-intercepts of R2 and Q2 in the permutation test (n=100) were

247

0.719, -0.567, respectively, suggesting a valid model. Importantly, the healthy leaves were

248

clustered together and separated from the CLas-infected leaves in Hamlin oranges, Murcott

249

mandarins and Ray Ruby grapefruits (Figure 2). These results suggest that metabolite profiles of

250

citrus leaves significantly altered after CLas infection regardless of citrus species. To identify the

251

most relevant metabolic pathways involved in metabolic reprogramming after CLas infection,

252

metabolites significantly changed by CLas infection (p