1
1
2 Effect of pulsed light on postharvest disease control-related metabolomic variation in melon
3 (Cucumis melo) artificially inoculated with Fusarium pallidoroseum
4
5 Short title: Pulsed light as tool to disease control in melons and its metabolomic changes
6
7 Francisco Oiram Filho1; Ebenézer de Oliveira Silva2; Mônica Maria de Almeida Lopes3; Paulo
8 Riceli Vasconselos Ribeiro2; Andréia Hansen Oster2; Jhonyson Arruda Carvalho Guedes4; Patrícia
9 do Nascimento Bordallo2; Guilherme Julião Zocolo2*
10
11
12
13 1Department of Chemical Engineering, Science Center, Federal University of Ceará, Fortaleza,
14 Ceará, Brazil.
15 2Multiuser Laboratory of Natural Products Chemistry, Embrapa Agroindústria Tropical, Fortaleza,
16 Ceará, Brazil.
17 3 Department of Biochemistry and Molecular Biology, Science Center, Federal University of Ceará,
18 Fortaleza, Ceará, Brazil.
19 4 Department of Analytical and Physical-Chemical Chemistry, Science Center, Federal University
20 of Ceará, Fortaleza, Ceará, Brazil.
21
22
23 *Corresponding author
24 E-mail: [email protected] (GJZ)
25
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
2
26 Abstract
27 Pulsed light, as a postharvest technology, is an alternative to traditional fungicides, and can be used
28 on a wide variety of fruit and vegetables for sanitization or pathogen control. In addition to these
29 applications, other effects also are detected in vegetal cells, including changes in metabolism and
30 production of secondary metabolites, which directly affect disease control response mechanisms.
31 This study aimed to evaluate the possible applications of pulsed ultraviolet light in controlling
32 postharvest rot, mainly caused by Fusarium pallidoroseum in yellow melon 'Goldex', in natura, and
33 its implications in the disease control as a function of metabolomic expression to effect fungicidal
34 or fungistatic. The dose of pulsed light (PL) that inhibited F. pallidoroseum growth in melons
35 (Cucumis melo var. Spanish) was 9 KJ m-2. Ultra performance liquid chromatography (UPLC)
36 coupled to a quadrupole time-of-flight (QTOF) mass analyzer identified 12 compounds based on
37 the MS/MS fragmentation patterns. Chemometric analysis by Principal Components Analysis
38 (PCA) and Orthogonal Partial Least Squared Discriminant Analysis (OPLS-DA and S-plot) were
39 used to evaluate the changes in fruit metabolism. PL technology provided protection against
40 postharvest disease in melons, directly inhibiting the growth of F. pallidoroseum through
41 upregulation of specific fruit biomarkers such as pipecolic acid (11), saponarin (7), and orientin (3),
42 which acted as major markers for the defense system against pathogens. PL can thus be proposed as
43 a postharvest technology to avoid chemical fungicides and may be applied to reduce the decay of
44 melon quality during its export and storage.
45
46
47
48
49
50
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
3
51 Introduction
52 The melon (Cucumis melo L.) is a tropical fruit and is an economically crucial part of
53 Brazilian exports. However, the major fragilities in the postharvest chain of melon are incidences of
54 postharvest pathologies, particularly the rot caused by Fusarium pallidoroseum, which is
55 responsible for postharvest losses of melons. The melon is a ground plant whose fruit is in contact
56 with soil, thus facilitating its contamination with F. pallidoroseum [1]. F. pallidoroseum is a
57 widespread and common species in tropical, subtropical, and Mediterranean areas, often isolated
58 from plants with complex diseases; it is also known to be toxigenic [2].
59 Fungal disease from genus Fusarium in fruit is traditionally controlled by application of
60 synthetic fungicides that leave chemical residues, which could be harmful to the consumer, and also
61 encourage the development of fungicide-resistant strains of fungal pathogens. To overcome these
62 challenges, several alternatives or integrative approaches, including physical methods, are
63 imperative to develop fruit with increased natural defense responses through the resistance
64 mechanisms induced by abiotic stress [3].
65 Among these strategies, application of technologies using light as an abiotic factor aiming
66 to promote a regulatory and signaling role in the developmental and metabolic processes of plants is
67 encouraged [4]. Ultraviolet light (UV-C) is a continuous radiation that is widely reported in the
68 induction of resistance mechanisms and control of postharvest diseases, thus extending the shelf-life
69 of fruit and vegetables [5, 6]. However, studies involving pulsed light (PL) as radiation from the
70 perspective of inducing resistance mechanisms are still discreet [7].
71 PL is a new, non-thermal technology, where a lamp containing an inert gas such as xenon
72 emits high-frequency radiation pulses between wavelengths of 180 to 1100 nm [8]. This technology
73 is used to sanitize fruit and vegetable surfaces by acting on microorganism cells and breaking and
74 altering their DNA sequences, thus inhibiting the pathogen reproductive capacity [9]. Moreover, the
75 use of PL as a physical method can stimulate the production of phytochemicals in plant tissues to
76 minimize the possible deleterious effects caused by radiation [10].
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
4
77 Metabolomics has been emphasized within the “omics” sciences, and efficiently evaluates
78 a large part of the metabolites of an organism both quantitatively and qualitatively, at a given time
79 and in a specific situation [11]. Metabolomics can identify a change in the concentration of
80 compounds involved in primary metabolism or production of secondary metabolites. These
81 metabolic alterations may arise from cellular lesions, metabolic adjustments to restore cellular
82 homeostasis, or synthesis and/or accumulation of metabolites resulting from cellular pathways [12].
83 The major objective of this study was to investigate the fungicidal or fungistatic effect to
84 Fusarium pallidoroseum in the melon cv. Spanish by PL treatment using ultra-performance liquid
85 chromatography-quadrupole-time-of-flight mass spectrometry (UPLC-QTOF-MSE) and
86 chemometric tools to identify possible biomarkers associated with metabolic changes in the defense
87 mechanism.
88
89 Material and methods
90 Chemical compounds
91 Acetonitrile (PubChem CID: 6432), formic acid (PubChem CID: 284), methanol
92 (PubChem CID: 887), sodium hypochlorite (PubChem CID: 23665760), MilliQ water (PubChem
93 CID: 962), liquid nitrogen (PubChem CID: 947).
94
95 Plant material
96 Melons (C. melo var. Spanish) were obtained at maturity stage (10–12 ºBrix) from a
97 commercial growing field of Norfruit Northeast Fruit, located at Mossoro-RN, Brazil (04”54’9,4”S
98 and 37”21’59,9”W). Mature melons were surface disinfected with 200 μg L-1 sodium hypochlorite
99 solution for 2 min, rinsed, and allowed to drain. The melons were then artificially inoculated.
100
101 Fruit inoculation
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
5
102 F. pallidoroseum spore suspension was produced from pure colonies of the fungus at a
103 concentration of 106 spores mL-1. Melons were inoculated near the peduncles and adjacent regions
104 with 100 µL inoculums of F. pallidoroseum (n = 5 inoculants). After inoculation, the melons were
105 subjected to PL treatment.
106
107 Disease control by PL treatment
108 After 12 h there is already pathogen effect [13] on fruits inoculated with F. pallidoroseum.
109 Then, the fruits were irradiated in a PL chamber (XeMaticA-2LXL model, SteriBeam® GmbH,
110 Germany) equipped with two xenon flash lamps and Teflon® transparent supports that allowed the
111 melons to be uniformly exposed over 360º by both lamps with uniformity at distance of 0.07 m
112 between fruit and lamps. The lamps produced short-time pulses of 0.3 µs, delivering broad-
113 spectrum white light (200-1100 nm) with approximately 15–20 % of UV-C, according to the system
114 built-in photodiode readings. Evaluation of disease control against F. pallidoroseum was carried out
115 by dose screening, as follows: 0 (non-treated with PL), 6, 9, and 12 KJ m-2, according to the
116 standard limits proposed [14]. After PL-screening, the inoculated and PL-treated melons were
117 incubated in a box protected from light for 48 h at 28 ºC ± 1 ºC with relative humidity of 92 % to
118 ensure optimal conditions of mycelium growth, and then stored at room temperature for 21 d. After
119 this time, the occurrence of a severe degree of fungal disease was analyzed by measuring the radial
120 lesion diameter in the inoculum region using a digital pachymeter (Digimess®, São Paulo, Brazil),
121 and expressing it in meters (m) and percentage of disease incidence (%). These analyses were
122 performed using randomized design, each treatment comprised 8 replicates and data were subjected
123 to analysis of variance (ANOVA) followed by Tukey’s test at 5% probability.
124
125 Disease control promoted by suitable PL-dose against F.
126 pallidoroseum
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
6
127 After PL-screening, an additional experiment was conducted using the suitable PL-dose
128 capable of controlling fungal disease in the fruit. To evaluate disease control, melons were
129 inoculated with F. pallidoroseum under the same inoculation conditions described in Fruit
130 inoculation section. Inoculated melons were then irradiated with a PL-dose of 9 KJ m-2 in the same
131 instrumental conditions described in Disease control by PL treatment section. The control group
132 included inoculated and non-inoculated melons. After 96 h, biological triplicates of each treatment
133 were subjected to extraction processes for injection into a UPLC system.
134
135 Extract preparation
136 Extracts were obtained Moore, Farrant (15] with modifications. The pellets were extracted
137 from the different adjacent regions (0.005 m) to the inoculum with thickness approximately of rind
138 melon. Peel melon powder (1.00 g) was resuspended in 4 mL of MeOH/H2O (7:3 v/v). The
139 homogenate was then subjected to sonication (Ultracelaner 1450, Unique®, Brazil) for 30 min,
140 followed by centrifugation at 6,000 × g, for 5 min. The pellet was re-extracted twice using 3 mL of
141 MeOH/H2O (7:3 v/v) using the same conditions of ultrasound and centrifugation. The supernatant
142 was filtered through a 0.22 μm polytetrafluoroethylene membrane (PTFE) (Biotechla®, Bulgaria)
143 and injected directly into a UPLC system.
144
145 Chromatographic analysis by UPLC-QTOF-MSE
146 The analyses were accomplished on an Acquity UPLC (Waters, USA) system coupled to a
147 Xevo Quadrupole and Time-of-Flight mass spectrometer (Q-TOF, Waters). Separations were
148 performed on a C18 column (Waters Acquity® UPLC C18 - 150 mm × 2.1 mm, 1.7 μm). For
149 metabolic fingerprinting, a 2-μL aliquot of phenolic extract was subjected to UPLC using an
150 exploratory gradient with the mobile phase composed of deionized water (A) and acetonitrile (B),
151 both containing formic acid (0.1% v/v). The phenolic extracts from melons were subjected to
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
7
152 exploratory gradient as follows: 2–95% for 15 min, with a flow rate of 500 µL min-1. Ionization was
153 performed with an electrospray ionization source in negative mode ESI, acquired in the range of
154 110–1200 Da. The optimized instrumental parameters were as follows: capillary voltage at -2800 V,
155 cone voltage at -40 V, source temperature at 120ºC, desolvation temperature at 330ºC, flow cone
156 gas at 20 L h-1, desolvation gas flow at 600 L h-1, and the microchannel voltage of the plate (MCP) -
157 detector at -1900 V. The mode of acquisition was MSE and system was controlled using MassLynx
158 4.1 software (Waters Corporation). The extracts were injected in triplicate.
159
160 Statistical analysis
161 The UPLC-MS data were processed using MassLynx® software (Waters Co., Milford,
162 MA, USA), under the following conditions: retention time deviation tolerance ± 0.05 min, mass
163 range 110–1200 Da (accurate mass tolerance of ± 0.05 Da) and noise elimination level 5. For
164 structural identification of metabolites, molecular formulas were considered and the values for m/z
165 were obtained from high-resolution spectra observed in the chromatogram at higher intensity. The
166 relative error is given in ppm for each formula. Margins of error less than 10 ppm are considered for
167 studies in MS/MS.
168 The structural proposals of molecules were performed using MS/MS data through the
169 establishment of rational fragmentation patterns reported in literature [16-19]. A list of identities of
170 peaks was created using the retention time (rt) and error (m/z). For unidentified peaks, all possible
171 molecular formulas were derived (elements C, H, N, and O, with the tolerance of 10 ppm, at least 2
172 carbon atoms) using the elemental composition tools available in MassLynx® software.
173 The UPLC-MS data analyzed by chemometrics were processed using Markerlynx®
174 software for Principal Components Analysis (PCA) and Orthogonal Partial Least Squared
175 Discriminant Analysis (OPLS-DA and S-plot). S-plots were obtained via OPLS-DA analysis to
176 determine potential biomarkers that significantly contributed to the difference among groups [20-
177 23].
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
8
178
179 Results and discussion
180
181 The growth of F. pallidoroseum under PL treatment
182 The lesion diameter of fungal infection (Fig 1) and percentage of disease incidence (data
183 not shown) were analyzed to determine the suitable PL-dose of treatment for the control or
184 inhibition of F. pallidoroseum in melon fruit. The growth of the pathogen on melon fruit inoculated
185 with P. pallidoroseum was strongly inhibited by PL treatment (Fig 1). This is corroborated as
186 shown in Fig 1, by a mean lesion diameter of 0.013 m, with a 100% incidence of fungal disease in
187 melons without PL radiation.
188
189 Fig 1. Screening of PL doses to evaluate the lesion diameter (m) in melon inoculated with F.
190 pallidoroseum. Mean values followed by the same small letter did not differ significantly between
191 PL-treatments, by Tukey’s test at 5% probability.
192
193 Disease progression in the PL-treated group that received 6 KJ m-2 was significantly lower
194 than that in the untreated group, with a significant reduction in lesion diameter (0.007 m) (Fig 1)
195 and 62.5% incidence of fungal disease. Moreover, a dose of 9 KJ m-2 was more effective than all
196 the doses employed and was associated with a low mean lesion diameter of 0.004 m, (Fig 1), and
197 33.3% disease incidence by F. pallidoroseum. The hormesis concept was used to explain the
198 behavior with a dose radiation of 9 KJ m-2.
199 Hormesis is a phenomenon in which low levels of potentially damaging radiation elicit
200 beneficial responses, i.e. physiological stimulation of beneficial responses in plants by low levels of
201 stressors that otherwise elicit harmful responses. Hormetic doses of ultraviolet light (UV-C)
202 radiation are involved in plant susceptibility towards diseases, and are capable of elicit plant
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
9
203 resistance mechanisms including production of anti-fungal compounds such as specific phenolics
204 including phytoalexins that act both as light quenchers that absorb damaging wavelengths of light,
205 and antioxidants that prevent reactive oxygen species (ROS) mediated cellular damage [7, 24, 25].
206 The UV-C radiation might also have fungistatic effect promoted by phenolic compounds which
207 plays a barrier role physically against pathogens attacks and the same barrier reduce water and
208 nutrients diffusion important to the growth pathogen [26].
209 However, a recent study compared the application of low-intensity UV-C and high-
210 intensity PL sources as elicitors of hormesis in tomato fruit (Solanum lycopersicum cv. Mecano)
211 [7]. Curiously, these authors showed that postharvest hormetic treatment with 16 pulses of PL (7.4
212 KJ m-2) with a spectral range (240 – 1050 nm) significantly delayed ripening along with inducing
213 disease resistance to Botrytis cinerea in tomato fruit with 41.7% reduction in disease progression
214 compared to 38.1% reduction with the conventional low-intensity UV-C (254 nm) at 0.37 KJ m-2.
215 Thus, according to the authors, PL treatment, although rich in UV-C (broader spectral output),
216 elicited the same pathways or responses as hormesis induced by conventional low UV sources
217 (narrower spectral range), making the PL-treatment more commercially attractive, as this treatment
218 allows a substantial reduction in treatment time from seconds to microseconds.
219 The last dose applied, 12 KJ m-2, corresponded to a disease incidence statistically equal to
220 the treatment with 9 KJ m-2, but showed a mean lesion diameter twice large (0.008 m) compared to
221 that in treatment with 9 KJ m-2 (Fig 1). These results showed a possible damaging effect on melon,
222 where excess of PL can inhibit the fruit defenses against fungal disease, and that the dose of 12 KJ
223 m-2 stipulated by the [14] was limiting to the conditions assessed.
224 Based on these results, we can hypothesize that the PL-treatment with 9 KJ m-2 applied
225 here acted as a fungistatic factor inhibiting the mycelial growth of F. pallidoroseum and that this
226 behavior is linked to a probable hormetic effect associated with the induction of metabolites
227 synthesis, which possibly activated defense responses against this fungal disease in melon cv.
228 Spanish. For this reason, 9 KJ m-2 was chosen for identification of the metabolites produced using a
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
10
229 UPLC system and chemometric tools to associate the identification of specific metabolites with
230 disease control mechanism in melon infected with F. pallidoroseum.
231
232 Putative metabolite identification by UPLC-QTOF-MS
233 Initially, infected and uninfected melon samples were evaluated in the positive and
234 negative ionization modes. Finally, the negative ionization mode was chosen because it presents a
235 metabolomic profile with a higher number of compounds at higher intensities, and is more selective
236 and sensitive.
237 The chemical profile of melon samples was established by analyzing the negative mode
238 (ESI-) chromatograms, Fig 2, together with mass spectra. The peaks were numbered according to
239 their elution order, the compounds were tentatively identified by interpretation of their MS and
240 MS/MS spectra determined by QTOF-MS, along with data from literature and open-access mass-
241 spectra databases as MassLynx®. Table 1 lists the MS data of tentatively identified compounds,
242 including experimental and calculated m/z for the molecular formula, error, and the fragments
243 obtained by MS/MS, as well as the proposed compound for each peak. In general, 12 metabolites of
244 distinct chemical classes, organic acids, non-protein amino acids, and phenolics, have been
245 tentatively identified.
246
247 Fig 2. Base-peak chromatogram of melon inoculated with F. pallidoroseum and subjected to PL-
248 treatment with 9 KJ m-2.
249
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
11
250 Table 1. Secondary metabolites tentatively identified by UPLC-QTOF-ESI-MSE in melons treated with PL.Positive ion mode Negative ion mode
Peaktr
(min) [M+H]+
Observed
[M+H]+
Calculated
Error
(ppm)
[M-H]-
Observed
[M-H]-
CalculatedMS/MS
Error
(ppm)
Molecular
formulaCompounds
1 3.16 295.1405 295.1393 4.1 293.1232 293.1236 131.0708 -1.4 C12H22O8Hydroxybutanoic acid ethyl
ester-hexoside
2 3.58 611.1619 611.1612 1.1 609.1454 609.1456
489.1016
429.0793
309.0857-0.3 C27H30O16 Isoorientin-2”- O-glucoside
3 3.78 449.1066 449.1084 -4.0 447.0936 447.0927
357.0532
327.0522
285.01972.0 C21H20O11 Orientin
4 3.96 595.1645 595.1663 -3.0 593.1486 593.1506413.0811
293.0408 -3.4 C27H30O15 4′′ -O-glucosylvitexin
5 4.11 625.1752 625.1769 -2.7 623.1591 623.1612443.1021
323.0592 3.4 C28H32O16 Isoscoparin 2″-O-glucoside
6 4.23 433.1118 433.1135 -3.9 431.0968 431.0978341.0652
311.0542 -2.3 C21H20O10 Vitexin
7 4.28 595.1641 595.1663 -3.7 593.1498 593.1506
503.1073
473.9605
341.0679-1.3 C27H30O15 Isovitexin-7”- O-glucoside (Saponarin)
8 4.40 463.1236 463.1240 -0.9 461.1052 461.1084
371.0816
341.0660
299.0436-2.8 C22H22O11 Diosmetin-6-C-glucoside
9 4.49 801.2205 801.2242 -4.6 799.2114 799.2086461.1159
341.0681 3.5 C38H40O19 Isoscoparin 7-O-[6′′-feruloyl]-glucoside
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
12
10 4.59 771.2106 771.2136 -3.9 769.1970 769.1980
623.1699
443.0993
413.0831-1.3 C37H38O18
Isoscoparin-2”-O-(6’’’-ρ-coumaroyl)-
glucoside
11 4.76 130.0866 130.0868 -1.5 128.0710 128.0712 84.0796 -1.6 C6H11NO2 Pipecolic acid*
12 6.62 273.0753 273.0763 -3.7 271.0605 271.0606
177.0180
151.0026
119.0489-0.4 C15H12O5 Naringerin
251 * Compound tentatively identified in negative and positive ionization mode, with the MS/MS ion in positive mode.252
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
13
253 The peak 1 (tr = 3.16 min) was tentatively identified as hydroxybutanoic acid
254 ethyl ester-hexoside. This compound showed the ion m/z 293.1232 [M-H]- in MS and
255 the fragment ion m/z 131.0708 [M-H-162]- in MS/MS, indicating a pattern loss of the
256 hexoside moiety [27].
257 The mass spectrum of peak 2 (tr = 3.58 min) showed the precursor ion m/z
258 609.1454 [M-H]-. The MS/MS spectrum showed fragment ions m/z 489.1016 [M-H-
259 120]-, 429.0793 [M-H-180]-, and 309.0857 [M-H-180-120]-. The losses of 180 and 120
260 u are significant for diglucosides like sophoroside (1-2 linkages of two glucose
261 molecules). Through the correlation of ions observed in MS and MS/MS, the compound
262 was tentatively identified as luteolin-6-C-glucosyl-2"-O-glucoside, also known as
263 isoorientin-2"-O-glucoside [28].
264 The mass spectrum of peak 3 (tr = 3.78 min) presents the precursor ion m/z
265 447.0936 [M-H]- that exhibited fragments ions m/z 357.0532 [M-H-90]-, 327.0522 [M-
266 H-120]-, and 285.0197 [M-H-162]-. The fragment ion m/z 285.0197 [M-H-162]- is the
267 aglycone formed from the loss of the glycosidic group (Table 1 and Fig. 2.) [29]. Thus,
268 the compound was tentatively identified as 8-C-glucosyl luteolin, also known as
269 orientin [28].
270 The compound 4 (tr = 3.96 min) showed in the mass spectrum as the precursor
271 ion m/z 593.1486 [M-H]-, also exhibiting the fragment ions m/z 413.0811 [M-H-180]-
272 and 293.0408 [(aglycone+41)-18)]-, which are characteristic of flavone O-glucosyl-C-
273 glucoside, indicating the presence of sophoroside, and apigenin as aglycone. This
274 compound was characterized as apigenin-6-C-glucosyl-2”-O-glucoside, also known as
275 4’’-O-glucosylvitexin or isovitexin-2”-O-glucoside [28, 30].
276 The peak 5 (tr = 4.11 min) showed in the mass spectrum a precursor ion m/z
277 623.1591 [M-H]- and its respective fragment ion m/z 443.1021 [M-H-162+18]-, which
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
14
278 suggested loss of a glucose moiety and one unit of water and the fragment ion m/z
279 323.0592 [M-H-120]-. These patterns of fragmentation indicate the a presence of a
280 diglucoside linkage and thus, compound 5 was tentatively identified as isoscoparin 2”-
281 O-glucoside [31].
282 The peak 6 (tr = 4.23 min), presented in the mass spectrum the precursor ion
283 m/z 431.0968 [M-H]-. In the MS/MS spectrum, it showed the fragments ions m/z
284 341.0652 [M-H-90]- and 311.0542 [M-H-120]-, indicating the presence of hexose as the
285 monosaccharide and apigenin as aglycone. Therefore, the compound was tentatively
286 identified as 8-C-glucosyl apigenin, also known as vitexin [28].
287 The mass spectrum of peak 7 (tr = 4.28 min) showed the ion m/z 593.1498 [M-
288 H]- and the fragment ions m/z 503.1073 [M-H-90]- and 473.9605 [M-H-120]-. This
289 pattern of loss indicates the presence of a diglucoside similar to that in compound 5 and
290 m/z 341.0679 [M-H-120-132]-, the loss of 132 Da represents a pentoside. Thus, the
291 compound was tentatively identified as isovitexin-7”-O-glucoside, also known as
292 saponarin. [32].
293 Peak 8 (tr = 4.40 min) in the mass spectrum showed the precursor ion m/z
294 461.1052 [M-H]- and its fragments m/z 371.0816 [M-H-90]-, 341.0660 [M-H-120]-, and
295 m/z 299.0436 [M-H-162]-. The fragment ion 299.0436 [M-H-162]- represents the
296 aglycone, formed by the loss of a glucoside moiety. Thus, the compound was
297 characterized as diosmetin-6-C-glucoside [33]
298 The peak 9 (tr = 4.49 min) represents compound 9 with a mass spectrum
299 showing the ion m/z 799.2114 [M-H]- and its fragment m/z 461.1159 [M-H-338]-
300 representing a loss of feruloyl plus a glucoside moiety and m/z 341.0681 [M-H-feruloyl-
301 Glucoside-120]-.Thus, based on fragmentation, the metabolite was tentatively identified
302 as isoscoparin 7-O-[6′′-feruloyl]-glucoside [31]
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
15
303 The peak 10 (tr = 4.59 min) presents in the mass spectrum, the ion m/z
304 769.1970 [M-H]- and its fragments m/z 623.1699 [M-H-146]- showing a loss of the
305 coumaroyl moiety; the fragment ion m/z 443.0993 [M-H-coumaroyl-162-18]- is the
306 result of successive losses of coumaroyl, glucoside, and a unit of water. The correlation
307 of the ions observed indicates that the metabolite in question is isoscoparin-2”-O-(6′′′-p-
308 coumaroyl)-glucoside [4]
309 Negative and positive ionization modes were used to identify pipecolic acid, a
310 non-protein amino acid (homolog of proline). Therefore, the peak 11 (tr = 4.76 min)
311 showed the ions m/z 128.0710 [M-H]- and m/z 130.0866 [M+H]+ in the MS in the
312 negative and positive ionization modes, respectively. Corroborating with chemical
313 identification, the fragment ion m/z 84.0796 [M+H-COOH]+ was observed in the
314 positive mode, referring to an aromatic core obtained from the cleavage of the carboxyl
315 group (Table 1 and Fig 2) [34].
316 In peak 12 (tr = 6.62 min) the ion m/z 271.0605 [M-H]- was observed with a
317 fragmentation pattern in MS/MS showing the loss of ring B at m/z 177.0180 and by
318 Retro Diels-Alder at m/z 151.0026 and m/z 119.0489. Thus, it was tentatively identified
319 as flavanone naringenin [35, 36].
320
321 Chemometric analysis
322 Principal component analysis (PCA) is a multivariate data analysis method that
323 can synthesize data from an original matrix with many variables in the set of smaller
324 orthogonal variables [37, 38]. We confirm that the chemometric analyses were centered
325 in PL-treatment with a dose of 9 KJ m-2, considered here as treatment for disease control
326 in melon inoculated with F. pallidoserum (Fig 1).
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
16
327 Therefore, PCA-2D was performed to discriminate different treatments groups
328 according to their metabolic profiles represented by retention time and the mass-to-
329 charge ratio (rt-m/z) from the UPLC-QTOF-MSE analysis (Fig 3). The PCA-2D showed
330 perfect separation of all groups evaluated with 89% of the total cumulative variance in
331 the diaxial axes PC1 and PC2 (R2X[1] = 0.7401 and R2X[2] = 0.1571), with a data noise
332 level of 6%, showing a robust model regarding data certainty. The formation of groups
333 was related to the similarity between biological triplicates.
334
335 Fig. 3. Discrimination of treatment groups by PCA-2D.
336
337 The PCA-2D in the PC1 showed the inoculated group (negative scores) and
338 non-inoculated group (positive scores); whereas in the PC2 (positive scores),
339 discrimination of the inoculated control and non-PL-treated group was observed. PC2
340 (negative scores) also showed the separation of inoculated PL-treated and non-
341 inoculated groups (Fig 3). The separation of negative and positive groups in PC1 and
342 PC2 is clearly linked to the differences of each metabolomic profile [39]. For this
343 reason, OPLS-DA chemometric analysis was applied to the UPLC-QTOF-MSE data to
344 compare samples according to metabolites that influenced on disease control against F.
345 pallidoroseum in melons treated with PL. OPLS-DA is an analysis method used to
346 study ions that contribute to experimental sample classification, and the classification
347 between two groups in the OPLS-DA model can be visualized in the form of a score
348 chart and scatter plot (S-plot).
349 Fig 4A summarizes the separation between non-inoculated and non-treated
350 fruit (control) and non-inoculated fruit treated with PL (9 KJ m-2), with respect to
351 metabolic responses in the function of PL as abiotic stress, through OPLS-DA
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
17
352 (R2X(cum) = 0.9232). The S-Plot generated from OPLS-DA, with VIP > 1.0 and p <
353 0.05 showed potential biomarkers between the treatments evaluated.
354
355 Fig 4 OPLS-DA, S-Plot graphs, and intensity of biomarkers between non-inoculated
356 control and non-inoculated and PL-treated melons (9 KJ m-2) (A). OPLS-DA, S-Plot
357 graphs, and intensity of markers between inoculated control and inoculated and PL-
358 treated melons (9 KJ m-2) (B).
359
360 The control group showed synthesis of principal secondary metabolites such
361 as naringenin (peak 12), hydroxybutanoic acid ethyl ester-hexoside (1), isoorientin-2”-
362 O-glucoside (2), orientin (3), and 4”-O-glucosylvitexin (4). However, an upregulation
363 of naringenin (12) and hydroxybutanoic acid ethyl ester-hexoside (1) was highlighted
364 (Fig 4A). Naringenin is a flavonoid that serves as a primer for more advanced flavonoid
365 structures and as a substrate for glycosylation reactions. Hydroxybutanoic acid ethyl
366 ester-hexoside is a compound has been previously described in other varieties of melon
367 such as Piel de Sapo, Galia, and Cantaloupe [27]. Interestingly, naringenin (12) and
368 hydroxybutanoic acid ethyl ester-hexoside (1) were downregulated in melons treated
369 with hormetic PL-radiation, whereas specific flavonoid compounds such as orientin (3),
370 4”-O-glucosylvitexin (4), and isoorientin-2”-O-glucoside (2) were up-regulated (Fig
371 4A), thus these compound which were up-regulated might be considered as biomarkers
372 caused by PL-treatment. Accumulation of flavonoid precursors such as phenylalanine
373 ammonia lyase into vacuoles present in the epidermal and subepidermal mesophyll
374 tissues of fruit stimulate plant defense mechanisms under specific PL-radiation
375 conditions [24]. Many phenylpropanoids have been associated with induced disease
376 resistance and disease control. According to Jung et al. (2013) the flavonoids orientin,
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
18
377 isoorientin-2”-O-glucoside, and 4”-O-glucosylvitexin are secondary metabolites
378 involved in antioxidant activities against abiotic stress in rice leaves (Oryza sativa cv.
379 Ilmi) exposed to different conditions of LED light radiation.
380 The differences in inoculated (control) and inoculated plus PL-treated (9 KJ m-
381 2) melons are shown in Fig 4B. Separation of the two treatments groups by the OPLS-
382 DA graph (R2X(cum) = 0.8882) indicated the differences among groups according to
383 their chemical profiles. The S-Plot obtained from the OPLS-DA graph, demonstrated
384 different potential biomarkers between groups for VIP > 1.0. Inoculation of F.
385 pallidoroseum mainly induced the synthesis of glycosylated flavonoids such as
386 diosmetin-6-C-glucoside (8), isoorientin-2"-O-glucoside (2), and 4’’-O-glucosylvitexin
387 (4) (Fig 4B), and this behavior is due to the presence of flavonoids at the infection site
388 that are responsible for defense mechanisms against pathogens [40]. However, these
389 metabolites which are a natural response of fruit against de pathogen were not sufficient
390 to control the growth pathogen as shown by control treatment at Figure 1.
391 Biotic and abiotic stress might regulate systemically the defense mechanism by
392 induce systemic resistance (ISR) or systemic acquired resistance (SAR) [41, 42]. PL-
393 treatment and inoculation with F. pallidoroseum led by SAR to the upregulation of two
394 major biomarkers in melon, pipecolic acid (11) and orientin (3) (Fig 4B), indicating a
395 change in the metabolic pathway, where the fruit preferentially used these specific
396 compounds as phytochemical markers in response to the treatment [43]. The presence of
397 pipecolic acid was verified in both groups that were inoculated with F. pallidoroseum.
398 The presence of pipecolic acid in pathogen inoculation sites is associated with plant
399 defense responses and acts as a regulator of inducible plant immunity [44]. Thus, when
400 the PL-treatment is applied the pipecolic acid together with orientin is up-regulated
401 significantly achieving disease control (fungistatic effect) which also seen at Figure 1 on
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
19
402 treatment 9 KJ.m2. Curiously, pipecolic acid is the supposed precursor of betaines that
403 are accumulated in the cytoplasm as a result of physiological plant responses to stress
404 phenomena induced in a large part by adverse environmental conditions. These
405 compounds that are biochemically inert in the cell are synthesized from some specific
406 amino acids such as serine, alanine, methionine, the non-protein amino acid γ-
407 aminobutyric acid, and some cyclic amino acids such as proline and pipecolic acid.
408 Biosynthesis of betaines in the cytoplasm under abiotic stress conditions is mainly due
409 to the action of methyltransferases, which utilize S-adenosylmethionine as a methyl
410 group donor [43].
411 Fig 5A shows the results of the OPLS-DA graph (R2X(cum) = 0.9684) between
412 the inoculated and non-inoculated groups in the function of F. pallidoroseum as a biotic
413 stress. The S-Plot originating from OPLS-DA data demonstrated potential biomarkers in
414 each group, with values for VIP > 1.0 and p < 0.05.
415
416 Fig 5. OPLS-DA, S-Plot graphs, and intensity of biomarkers between non-inoculated
417 and inoculated melons (A). OPLS-DA, S-Plot graphs, and intensity of markers between
418 treated non-inoculated (9 KJ m-2) and PL-treated inoculated melons (9 KJ m-2) (B).
419
420 The number of compounds up-regulated in non-inoculated melons is much
421 higher when that compared to inoculated melons, highlighting naringenin (12),
422 hydroxybutanoic acid ethyl ester-hexoside (1) and orientin (3) (Fig 5A). However,
423 saponarin (7) was a biomarker present in group inoculated control melons and
424 isoorientin-2"-O-glucoside (2) shows a little up-regulation on the same group (Fig 5A).
425 Hydroxybutanoic acid ethyl ester-hexoside is a compound linked to amino acid groups
426 and findings have indicated significant interconnections between different branches of
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
20
427 amino acid metabolism and plant resistance to pathogens [45, 46]. Naringenin is a
428 flavonoid showing metabolic activity associated with barley resistance to Fusarium
429 graminearum [47].
430 The results observed in Fig. 5B show the separation of distinct groups, where
431 the OPLS-DA plot (R2X(cum) = 0.9798) presents differences between the non-
432 inoculated fruit and inoculated plus PL-treatment. The S-Plot obtained by the OPLS-DA
433 graph points out the potential markers for each group. Upregulation of flavonoids such
434 as naringenin (12), orientin (3) and 4”-O-glucosylvitexin (4) in non-inoculated PL-
435 treated fruit (9 KJ m-2) is much higher when compared to the inoculated and PL-treated
436 (9 KJ m-2) group (Fig. 5B). The synthesis of flavonoids in detriment to abiotic stress
437 observed in PL treatment occurs due to a change in the metabolic pathway of fruit in
438 response to the medium [48]. On the other hand, inoculated and PL-treated melons
439 accumulated a glycosylated flavonoid identified as saponarin (7), and pipecolic acid
440 (11) as biomarkers [49]. Nevertheless, the presence of pipecolic acid on non-inoculated
441 and PL-treated (9 KJ m-2) group (Fig 5B) shows that this metabolite might be
442 synthetized by abiotic stress also. Saponarin is an antioxidant belonging to the flavones
443 and is to inhibiting malonaldehyde formation in barley. In a normal reaction,
444 malonaldehyde is formed from oxidized lipids on the skin surface in barley leaves by
445 UV irradiation [50]. The effect of PL treatment in melons inoculated with fungi resulted
446 in a response mediated by the synthesis of pipecolic acid in the cellular medium, which
447 resulted in increased levels of betaines in the cell, that induce the fruit immunity system.
448
449 Conclusions450
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
21
451 In the present study, a PL-dose of 9 KJ m-2 in melon inoculated with Fusarium
452 pallidoroseum controlled the disease promoted by this pathogen (fungistatic effect), and
453 induced metabolic variation in the fruit defense system. Pipecolic acid (11) and orientin
454 (3) were the two major biomarkers associated with postharvest disease control against
455 F. pallidoroseum in infected melons treated with a pulsed radiation. This study also
456 showed that fruit subjected to biotic and abiotic stresses, separately, demonstrated
457 different metabolic responses, with a chemical profile in response to each stress. In
458 addiction the compounds orientin (3), 4”-O-glucosylvitexin (4), and isoorientin-2”-O-
459 glucoside (2) are the metabolites in response only PL-treatment. Our findings highlight
460 that the application of PL technology provided control against the postharvest disease of
461 melon cv. Spanish by directly controlling the growth of F. pallidoroseum through the
462 synthesis/up-regulation of specific compounds that acted as principal biomarkers of the
463 defense system against pathogen. Thus, PL can be readily proposed as a new
464 postharvest technology alternative to chemical fungicides and may be applied to reduce
465 the delay caused by F. pallidoroseum in melons var. Spanish during their export and
466 storage.
467
468 Acknowledgments
469 The authors are gratefully acknowledged for support from the Brazilian
470 Company of Agricultural Research (EMBRAPA), National Council for Scientific and
471 Technological Development (CNPq), National Institute of Science and Technology
472 (INCT BioNat, Brazil) and commercial growing field of Norfruit Northeast Fruit.
473
474 References
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
22
475
476 1. Sebastiani MS, Bagnaresi P, Sestili S, Biselli C, Zechini A, Orrù L, et al. 477 Transcriptome analysis of the melon-Fusarium oxysporum f. sp. melonis race 1.2 478 pathosystem in susceptible and resistant plants. Frontiers in Plant Science. 479 2017;8:362.480 2. Zaccardelli M, Balmas V, Altomare C, Corazza L, Scotti C. Characterization of 481 Italian isolates of Fusarium semitectum from alfalfa (Medicago sativa L.) by AFLP 482 analysis, morphology, pathogenicity and toxin production. Journal of 483 Phytopathology. 2006;154(7‐8):454-60.484 3. Zhang H, Mahunu GK, Castoria R, Apaliya MT, Yang Q. Augmentation of 485 biocontrol agents with physical methods against postharvest diseases of fruits and 486 vegetables. Trends in Food Science & Technology. 2017;69:36-45.487 4. Jung ES, Lee S, Lim S, Ha S, Liu K, Lee CH. Metabolite profiling of the short-term 488 responses of rice leaves (Oryza sativa cv. Ilmi) cultivated under different LED 489 lights and its correlations with antioxidant activities. Plant Science. 2013;210:61-9.490 5. Jin P, Wang H, Zhang Y, Huang Y, Wang L, Zheng Y. UV-C enhances resistance 491 against gray mold decay caused by Botrytis cinerea in strawberry fruit. Scientia 492 Horticulturae. 2017;225:106-11.493 6. Mohamed NTS, Ding P, Kadir J, Ghazali HM. Potential of UVC germicidal 494 irradiation in suppressing crown rot disease, retaining postharvest quality and 495 antioxidant capacity of Musa AAA “Berangan” during fruit ripening. Food Science 496 & Nutrition. 2017;5(5):967-80.497 7. Scott G, Rupar M, Fletcher AGD, Dickinson M, Shama G. A comparison of low 498 intensity UV-C and high intensity pulsed polychromatic sources as elicitors of 499 hormesis in tomato fruit. Postharvest Biology and Technology. 2017;125:52-8.500 8. Cheigh C, Park M, Chung M, Shin J, Park Y. Comparison of intense pulsed light-501 and ultraviolet (UVC)-induced cell damage in Listeria monocytogenes and 502 Escherichia coli O157: H7. Food Control. 2012;25(2):654-9.503 9. Garvey M, Stocca A, Rowan N. Development of a combined in vitro cell culture–504 Quantitative PCR assay for evaluating the disinfection performance of pulsed light 505 for treating the waterborne enteroparasite Giardia lamblia. Experimental 506 Parasitology. 2014;144:6-13.507 10. Rodov V, Vinokur Y, Horev B. Brief postharvest exposure to pulsed light 508 stimulates coloration and anthocyanin accumulation in fig fruit (Ficus carica L.). 509 Postharvest Biology and Technology. 2012;68:43-6.510 11. Roessner U, Dias DA. Metabolomics tools for natural product discovery: Humana 511 Press; 2013.512 12. Genga A, Mattana M, Coraggio I, Locatelli F, Piffanelli P, Consonni R. Plant 513 Metabolomics: A characterisation of plant responses to abiotic stresses. Abiotic 514 Stress in Plants-Mechanisms and Adaptations Rijeka: InTech. 2011:309-50.515 13. Ingle RA, Stoker C, Stone W, Adams N, Smith R, Grant M, et al. Jasmonate 516 signalling drives time‐of‐day differences in susceptibility of Arabidopsis to the 517 fungal pathogen Botrytis cinerea. The Plant Journal. 2015;84(5):937-48.518 14. FDA. Food for human consumption part 179 irradiation in the production, 519 processing and handling of food. Code of Federal Regulations Title 21. 2017:Sec. 520 179.41.521 15. Moore JP, Farrant JM, Lindsey GG, Brandt WF. The South African and Namibian 522 populations of the resurrection plant Myrothamnus flabellifolius are genetically 523 distinct and display variation in their galloylquinic acid composition. Journal of 524 Chemical Ecology. 2005;31(12):2823-34.
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
23
525 16. Moco S, Vervoort J, Bino RJ, De Vos RCH, Bino R. Metabolomics technologies 526 and metabolite identification. Trends in Analytical Chemistry. 2007;26(9):855-66.527 17. Kind T, Fiehn O. Advances in structure elucidation of small molecules using mass 528 spectrometry. Bioanalytical Reviews. 2010;2(1-4):23-60.529 18. Bouslimani A, Sanchez LM, Garg N, Dorrestein PC. Mass spectrometry of natural 530 products: current, emerging and future technologies. Natural Product Reports. 531 2014;31(6):718-29.532 19. Yuliana ND, Jahangir M, Verpoorte R, Choi YH. Metabolomics for the rapid 533 dereplication of bioactive compounds from natural sources. Phytochemistry 534 Reviews. 2013;12(2):293-304.535 20. Farag MA, Gad HA, Heiss AG, Wessjohann LA. Metabolomics driven analysis of 536 six Nigella species seeds via UPLC-qTOF-MS and GC–MS coupled to 537 chemometrics. Food Chemistry. 2014;151:333-42.538 21. Mncwangi NP, Viljoen AM, Zhao J, Vermaak I, Chen W, Khan I. What the devil is 539 in your phytomedicine? Exploring species substitution in Harpagophytum through 540 chemometric modeling of 1 H-NMR and UHPLC-MS datasets. Phytochemistry. 541 2014;106:104-15.542 22. Sprenger RF, Cass QB. Characterization of four Phyllanthus species using liquid 543 chromatography coupled to tandem mass spectrometry. Journal of Chromatography 544 A. 2013;1291:97-103.545 23. Uarrota VG, Moresco R, Coelho B, Nunes EC, Peruch LAM, Neubert EO, et al. 546 Metabolomics combined with chemometric tools (PCA, HCA, PLS-DA and SVM) 547 for screening cassava (Manihot esculenta Crantz) roots during postharvest 548 physiological deterioration. Food Chemistry. 2014;161:67-78.549 24. Lopes MMA, Silva EO, Canuto KM, Silva LMA, Gallão MI, Urban L, et al. Low 550 fluence pulsed light enhanced phytochemical content and antioxidant potential of 551 ‘Tommy Atkins’ mango peel and pulp. Innovative Food Science & Emerging 552 Technologies. 2016;33:216-24.553 25. Shama G, Alderson P. UV hormesis in fruits: a concept ripe for commercialisation. 554 Trends in Food Science & Technology. 2005;16(4):128-36.555 26. Ribeiro C, Alvarenga B. Prospects of UV radiation for application in postharvest 556 technology. Emirates Journal of Food and Agriculture. 2012:586-97.557 27. Rodríguez-Pérez C, Quirantes-Piné R, Fernández-Gutiérrez A, Segura-Carretero A. 558 Comparative characterization of phenolic and other polar compounds in Spanish 559 melon cultivars by using high-performance liquid chromatography coupled to 560 electrospray ionization quadrupole-time of flight mass spectrometry. Food 561 Research International. 2013;54(2):1519-27.562 28. Negri G, Santi D, Tabach R. Chemical composition of hydroethanolic extracts from 563 Siparuna guianensis, medicinal plant used as anxiolytics in Amazon region. 564 Revista Brasileira de Farmacognosia. 2012;22(5):1024-34.565 29. Abad-García B, Berrueta LA, Garmón-Lobato S, Gallo B, Vicente F. A general 566 analytical strategy for the characterization of phenolic compounds in fruit juices by 567 high-performance liquid chromatography with diode array detection coupled to 568 electrospray ionization and triple quadrupole mass spectrometry. Journal of 569 Chromatography A. 2009;1216(28):5398-415.570 30. Dou J, Lee VSY, Tzen JTC, Lee M. Identification and comparison of phenolic 571 compounds in the preparation of oolong tea manufactured by semifermentation and 572 drying processes. Journal of Agricultural and Food Chemistry. 2007;55(18):7462-8.573 31. Piasecka A, Sawikowska A, Krajewski P, Kachlicki P. Combined mass 574 spectrometric and chromatographic methods for in‐depth analysis of phenolic
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
24
575 secondary metabolites in barley leaves. Journal of Mass Spectrometry. 576 2015;50(3):513-32.577 32. Ibrahim RM, El-Halawany AM, Saleh DO, El Naggar EMB, El-Shabrawy AEO, 578 El-Hawary SS. HPLC-DAD-MS/MS profiling of phenolics from Securigera 579 securidaca flowers and its anti-hyperglycemic and anti-hyperlipidemic activities. 580 Revista Brasileira de Farmacognosia. 2015;25(2):134-41.581 33. Bresciani L, Calani L, Cossu M, Mena P, Sayegh M, Ray S, et al. (Poly) phenolic 582 characterization of three food supplements containing 36 different fruits, vegetables 583 and berries. PharmaNutrition. 2015;3(2):11-9.584 34. Piraud M, Vianey‐Saban C, Petritis K, Elfakir C, Steghens J, Morla A, et al. 585 ESI‐MS/MS analysis of underivatised amino acids: a new tool for the diagnosis of 586 inherited disorders of amino acid metabolism. Fragmentation study of 79 molecules 587 of biological interest in positive and negative ionisation mode. Rapid 588 Communications in Mass Spectrometry. 2003;17(12):1297-311.589 35. Sánchez‐Rabaneda F, Jáuregui O, Casals I, Andrés‐Lacueva C, Izquierdo‐Pulido M, 590 Lamuela‐Raventós RM. Liquid chromatographic/electrospray ionization tandem 591 mass spectrometric study of the phenolic composition of cocoa (Theobroma cacao). 592 Journal of Mass Spectrometry. 2003;38(1):35-42.593 36. Akpanika GA, Winters A, Wilson T, Ayoola GA, Adepoju-Bello AA, Hauck B. 594 Polyphenols from Allanblackia floribunda seeds: Identification, quantification and 595 antioxidant activity. Food Chemistry. 2017;222:35-42.596 37. Sainani KL. Introduction to principal components analysis. Physical Medicine and 597 Rehabilitation Journal. 2014;6(3):275-8.598 38. Abdi H, Williams LJ. Principal component analysis. Wiley Interdisciplinary 599 Reviews: Computational Statistics. 2010;2(4):433-59.600 39. Mao Q, Bai M, Xu J, Kong M, Zhu L, Zhu H, et al. Discrimination of leaves of 601 Panax ginseng and P. quinquefolius by ultra high performance liquid 602 chromatography quadrupole/time-of-flight mass spectrometry based metabolomics 603 approach. Journal of Pharmaceutical and Biomedical Analysis. 2014;97:129-40.604 40. Mierziak J, Kostyn K, Kulma A. Flavonoids as important molecules of plant 605 interactions with the environment. Molecules. 2014;19(10):16240-65.606 41. Conrath U, Beckers GJM, Langenbach CJG, Jaskiewicz MR. Priming for enhanced 607 defense. Annual Review of Phytopathology. 2015;53:97-119.608 42. Scott G, Dickinson M, Shama G, Rupar M. A comparison of the molecular 609 mechanisms underpinning high-intensity, pulsed polychromatic light and low-610 intensity UV-C hormesis in tomato fruit. Postharvest Biology and Technology. 611 2018;137:46-55.612 43. Servillo L, Giovane A, Balestrieri ML, Ferrari G, Cautela D, Castaldo D. 613 Occurrence of pipecolic acid and pipecolic acid betaine (homostachydrine) in 614 Citrus genus plants. Journal of Agricultural and Food Chemistry. 2011;60(1):315-615 21.616 44. Návarová H, Bernsdorff F, Döring A, Zeier J. Pipecolic acid, an endogenous 617 mediator of defense amplification and priming, is a critical regulator of inducible 618 plant immunity. The Plant Cell. 2012;24(12):5123-41.619 45. Cecchini NM, Monteoliva MI, Alvarez ME. Proline dehydrogenase contributes to 620 pathogen defense in Arabidopsis. Plant Physiology. 2011;155(4):1947-59.621 46. Stuttmann J, Hubberten H, Rietz S, Kaur J, Muskett P, Guerois R, et al. 622 Perturbation of Arabidopsis amino acid metabolism causes incompatibility with the 623 adapted biotrophic pathogen Hyaloperonospora arabidopsidis. The Plant Cell. 624 2011;23(7):2788-803.
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
25
625 47. Bollina V, Kushalappa AC, Choo TM, Dion Y, Rioux S. Identification of 626 metabolites related to mechanisms of resistance in barley against Fusarium 627 graminearum, based on mass spectrometry. Plant Molecular Biology. 2011;77(4-628 5):355.629 48. Aguiló-Aguayo I, Charles F, Renard CMGC, Page D, Carlin F. Pulsed light effects 630 on surface decontamination, physical qualities and nutritional composition of 631 tomato fruit. Postharvest Biology and Technology. 2013;86:29-36.632 49. Bernsdorff F, Döring A, Gruner K, Schuck S, Bräutigam A, Zeier J. Pipecolic acid 633 orchestrates plant systemic acquired resistance and defense priming via salicylic 634 acid-dependent and-independent pathways. The Plant Cell. 2016;28(1):102-29.635 50. Kamiyama M, Shibamoto T. Flavonoids with potent antioxidant activity found in 636 young green barley leaves. Journal of Agricultural and Food Chemistry. 637 2012;60(25):6260-7.
638
639
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint
.CC-BY 4.0 International licensenot certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which wasthis version posted July 10, 2019. . https://doi.org/10.1101/698407doi: bioRxiv preprint