Aug 24, 2015 - ABSTRACT: Powdery mildew (PM) disease causes significant loss in watermelon. Due to the unavailability of a commercial watermelon variety ...
Article pubs.acs.org/JAFC
NMR Spectroscopy Identifies Metabolites Translocated from Powdery Mildew Resistant Rootstocks to Susceptible Watermelon Scions Iqbal Mahmud,†,‡ Chandrasekar Kousik,§ Richard Hassell,# Kamal Chowdhury,† and Arezue F. Boroujerdi*,‡ †
Department of Biology, School of Natural Science and Mathematics and ‡Department of Chemistry/Molecular Science Research Center, Claflin University, Orangeburg, South Carolina 29115, United States § U.S. Vegetable Laboratory, Agricultural Research Service, U.S. Department of Agriculture, Charleston South Carolina 29414, United States # Department of Agricultural, Forest, and Environmental Sciences, Coastal Research and Education Center, Clemson University, Clemson, South Carolina 29634, United States S Supporting Information *
ABSTRACT: Powdery mildew (PM) disease causes significant loss in watermelon. Due to the unavailability of a commercial watermelon variety that is resistant to PM, grafting susceptible cultivars on wild resistant rootstocks is being explored as a shortterm management strategy to combat this disease. Nuclear magnetic resonance-based metabolic profiles of susceptible and resistant rootstocks of watermelon and their corresponding susceptible scions (Mickey Lee) were compared to screen for potential metabolites related to PM resistance using multivariate principal component analysis. Significant score plot differences between the susceptible and resistant groups were revealed through Mahalanobis distance analysis. Significantly different spectral buckets and their corresponding metabolites (including choline, fumarate, 5-hydroxyindole-3-acetate, and melatonin) have been identified quantitatively using multivariate loading plots and verified by volcano plot analyses. The data suggest that these metabolites were translocated from the powdery mildew resistant rootstocks to their corresponding powdery mildew susceptible scions and can be related to PM disease resistance. KEYWORDS: grafted watermelon, powdery mildew disease, nuclear magnetic resonance, multivariate principal component analysis, biochemical pathway analysis
■
Table 1. Resistant and Susceptible Watermelon Plant Materialsa
INTRODUCTION Watermelon (Citrullus lanatus) fruits are high in several antioxidants and important vitamins and are one of the leading fruits in terms of lycopene content. Watermelon is grown in 44 states throughout the United States on approximately 141,000 acres valued at $492 million (USDA-NASS, 2012). In the state of South Carolina alone, over 8000 acres of watermelon are grown and valued at $36 million (USDA-NASS, 2012). Diseases and pests are significant factors limiting watermelon production, especially in the southeastern United States, where a major portion of U.S. fruits are produced. Various plant pathogens including fungi, bacteria, and viruses cause disease in watermelon that can significantly reduce yield and in some situations cause complete crop loss.1 Powdery mildew (PM) is a foliar disease caused by Podospharea xanthii that significantly reduces watermelon yield in the United States.2 PM could possibly result in up to ∼$91 million loss in the United States alone.3 This disease can also result in reduced vigor and, in some instances, death of the seedlings. The pathogen is known to infect cotyledons, stem, leaves, and fruit.4−6 Additionally, it has been reported to infect most cucurbits.1,7 Moreover, the PM pathogen can develop rapidly on the underside of leaves, and most currently available fungicides, which are applied to the top part of plants, do not have systemic activity to help manage this situation.2 Host resistance will be highly effective in such situations; however, commercial edible watermelon varieties with PM resistance are not yet available. © 2015 American Chemical Society
rootstock ID
rootstock type
plant type
powdery mildew reaction of rootstock
USVL677-PMS USVL531-PMR USVL516-PMR USVL225-PMR USVL331-PMR ML-UG ML-SG
CLL CLL CLL CLC CLC CL CL
watermelon watermelon watermelon watermelon watermelon watermelon watermelon
susceptible resistant resistant resistant resistant susceptible susceptible
a
Grafted watermelon has two parts: a resistant rootstock (bottom portion) and a scion (top portion), which is from the commercially available Mickey Lee watermelon but susceptible to powdery mildew. Two independent experiments were conducted for each variety of rootstock listed. Similar results were observed for repetitive experiments as well as for both varieties of PMR-CLL; therefore, only the results from one of each experiment were used. PMS, powdery mildew susceptible; PMR, powdery mildew resistant; CLL, Citrullus lanatus var. lanatus; CLC, Citrullus lanatus var. citroides; ML, Mickey Lee; UG, ungrafted; SG, self-grafted.
Grafting watermelon, which is new to the United States, has been practiced since the 1920s in Asia and more recently in Received: Revised: Accepted: Published: 8083
June 8, 2015 August 19, 2015 August 24, 2015 August 24, 2015 DOI: 10.1021/acs.jafc.5b02108 J. Agric. Food Chem. 2015, 63, 8083−8091
Article
Journal of Agricultural and Food Chemistry Table 2. Metabolites Identified in Watermelon Rootstocks and Scionsa metabolite
1 H chemical shifts (ppm) (functional group or specific H, multiplicity of peak)
13
C chemical shifts (ppm) (functional group or specific C)
1
2-hydroxybutyrate
0.86 (CH3, t); 1.6 (CH2, m); 1.7 (CH2, m); 3.9 (CH, m)
10.0 (CH3); 33 (CH2); 29 (CH2); 72 (CH)
2
leucine
0.96, 1.0 (Hδ, d); 1.7 (Hγ, m); 1.7 (Hβ, m); 3.7 (Hα, m)
24.0, 25 (Cδ); 27 (Cγ); 43 (Cβ); 56 (Cα)
3
valine
1.00, 1.05 (Hγ, d); 2.3 (Hβ, m); 3.6 (Hα, d)
19.0, 21.0 (Cγ); 32 (Cβ); 63 (Cα)
4
isoleucine
0.92 (Hδ, t), 0.99 (Hγ, d); 1.2, 1.5 (Hγ, m); 2.0 (Hβ, m); 3.7 (Hα, d)
14.0 (Cδ); 18, 27.0 (Cγ); 39 (Cβ); 62 (Cα)
5
methylmalonate
1.22 (CH3, d); 3.1 (CH, q)
18.0 (CH3); 56 (CH, q)
6
alanine
1.47 (Hβ, d); 3.78 (Hα, q)
19.2 (Cβ); 53.3 (Cα)
7
melatonin
7.4 (4CH, d); 7.2 (1CH, d); 1.91(8CH3, s)
133 (4CH); 120 (1CH); 22.0 (8CH3)
8
methionine
2.12 (Hδ, q)
31.0 (Cδ)
9
levulinate
2.20 (CH3, s); 2.4 (CH2, t)
32.0(CH3); 33 (CH2)
10
4-aminobutyrate
2.28 (CH2, t); 1.9 (CH2, m)
43.0 (CH2); 37 (CH2)
11
malate
2.36, 2.66 (CH2, dd); 4.29 (CH, dd)
46.0 (CH2); 73 (CH)
12
succinate
2.40 (CH2, s)
36.9 (CH2)
13
aspartate
2.80, 2.66 (Hβ, dd); 3.9 (Hα, q)
40.0 (Cβ); 55 (Cα)
14
asparagine
2.83, 2.92 (Hβ, dd); 3.9 (Hα, q)
38.0 (Cβ); 54 (Cα)
15
malonate
3.10 (CH2, s)
50.0 (CH2)
16
choline
3.19 (CH3, s); 3.5, 4.0 (CH2, m)
57.0 (CH3); 70 (CH2)
17
glucose
5.22 (1βCH, d); 4.64 (1αCH, d); 3.23 (2βCH, dd); 3.52 (2αCH, dd); 3.48 95.0 (1βCH); 99.0 (1αCH); 77.1 (2βCH); 74.4 (2αCH); 78.6 (3βCH, t); 3.70 (3αCH, t); 3.41 (4CH, dd); 3.47 (5βCH, m); 3.84 (3βCH); 75.6 (3αCH); 72.5 (4CH); 78.8 (5βCH); 74.4 (5αCH, m); 3.71 (6βCH2, dd); 3.89 (6αCH2, dd) (5αCH); 63.6 (6CH2)
18
maltose
5.2 (1CH, d); 3.60 (2CH, m); 3.65 (3CH, t); 3.41 (4CH, t); 3.7 (5CH, m); 95 (1CH); 74.6 (2CH); 75.4 (3CH); 78.8 (4CH); 75 (5CH); 3.8 (6CH2, m); 5.41 (1′CH, dd); 3.27 (2′CH, t); 3.97 (3′CH, m); 3.7 63 (6CH2); 102.5 (1′CH); 77.0 (2′CH); 76.2 (3′CH); 72 (4′CH, t); 3.68, 3.90 (5αβ′CH, m); 3.8 (6′CH2, m) (4′CH); 79.7, 72.4 (5αβ′CH); 63 (6′CH2)
19
fructose
3.56, 3.70 (1CH2, dd); 3.80 (3CH, d); 4.12 (4CH, d); 3.81 (5CH, d); 3.56, 3.69 (6CH2, dd)
20
galactose
3.48 (2CH, q); 3.64 (3CH, dd); 3.74 (3CH, m); 3.9, 3.98 (4CH, q); 4.58, 75.0 (2CH); 71.0 (3CH); 65.0 (3CH); 70 (4CH); 99.0 5.25 (1CH, d) (1CH)
21
IAA
7.38 (4CH, d); 3.58 (1CH2,S); 6.8 (6CH, d)
133.5 (4CH); 62.0(1CH2); (6CH, d); 118(6CH)
22
myo-inositol
3.2 (1CH, t); 3.52, (3CH, dd); 3.6 (2CH, t); 4.0(4CH, t)
77 (1CH); 74.0 (3CH); 75 (2CH, t); 75 (4CH)
23
guanosine
5.93 (1CH, d); 7.9 (2CH, s)
92.0 (1CH); 140 (2CH)
24
adenosine
6.06 (1CH, d); 8.2, 8.3 (2CH, s)
91.0 (1CH); 143 (2CH)
25
fumarate
6.50 (CH, s)
138.2 (CH)
26
glycerate
3.6 (2CH2, q); 3.8 (2CH2, dd); 4.07 (1CH, q)
66.9 (2CH2); 76.0 (1CH)
27
glycine
3.57 (Hα, s)
44.3 (Cα)
66.8 (1CH2); 70.4 (3CH); 77.3 (4CH); 70.5 (5CH); 65.3 (6CH2)
a1
H chemical shifts (ppm) and their corresponding 13C chemical shifts (ppm) are listed. Bold chemical shifts indicate an identified peak in the spectra. Bold metabolite names indicate metabolites that were found to be statistically significant (in concentration change) in the comparison between resistant and susceptible watermelon tissues. 8084
DOI: 10.1021/acs.jafc.5b02108 J. Agric. Food Chem. 2015, 63, 8083−8091
Article
Journal of Agricultural and Food Chemistry
Figure 1. Proton spectral comparisons. Higher intensity of 1H signals were observed for sugar compounds in powdery mildew susceptible rootstocks (PMS RS) compared to powdery mildew resistant rootstocks (PMR RS). The spectra were scaled with the internal reference and standard TMSP (1 mM).
Figure 2. PCA score plots. PC1 versus PC2 score plots are shown for each. The total explained variance is >90%. The ovals represent 95% confidence intervals. Each oval represents a sample group, and each point in the oval represents a single sample/spectrum. R1 represents all of the rootstock samples: UG-ML, ungrafted Mickey Lee (a); SG-ML, self-grafted Mickey Lee (b); powdery mildew susceptible (PMS) rootstock PMS-CLL; (c) powdery mildew resistant (PMR) rootstocks PMR-CLL (d); PMR rootstocks PMR-CLC (e). S1 represents all of the corresponding grafted scions (PMS Mickey Lee) on all rootstocks.
Europe for managing soilborne diseases.8,9 Watermelon scions are grafted on various rootstocks including bottle gourds, intraspecific squash hybrids, and watermelons.10,11 Grafting is generally done using various methods such as approach graft, hole insertion graft, or single-cotyledon graft.10 The U.S. Vegetable Laboratory (USVL), USDA-ARS, in Charleston, SC, USA, has developed a watermelon with resistance to PM as a part of their prebreeding program using grafting techniques. In preliminary experiments, moderate but significant levels of resistance to PM and fruit rot were reported in grafted watermelon (commercial PM susceptible scion grafted on wild PM resistant rootstock). To understand what specific resistance components translocate from resistant rootstocks to the susceptible scion Mickey Lee through grafting,
we utilized NMR-based metabolomics, a novel analytical chemical approach for biomarker discovery. This is rapidly being applied to the development of diagnostic tools for a variety of environmental conditions as well as diseases in animals and humans.12 In food science, various analytical techniques including NMR, UPLC-MS, and GC-MS are widely used in targeted as well as untargeted metabolomics studies.13 Among these, NMR-based metabolomics is a burgeoning analytical tool to investigate metabolism due to its nondestructive and unbiased nature, robustness, reproducibility, and simultaneous wide-ranging detection patterns.12,14 Recently in the plant biotechnology field, the NMRbased metabolomics approach has been extensively applied to develop protocols for plant metabolomics research,14−20 plant 8085
DOI: 10.1021/acs.jafc.5b02108 J. Agric. Food Chem. 2015, 63, 8083−8091
Article
Journal of Agricultural and Food Chemistry disease research,21,22 and plant metabolic profiling23,24 and also in plant tissue culture research.25−27 In this study, we compared metabolic profiles of powdery mildew susceptible (PMS) and resistant (PMR) watermelon rootstocks to identify differences in their metabolic profiles using NMR spectroscopy. We also verified translocation of metabolites from resistant rootstocks to susceptible scions in grafted watermelon plants.
■
MATERIALS AND METHODS
Experimental Design. Mickey Lee (PM susceptible watermelon cultivar) scions individually were grafted using the one cotyledon grafting method10 to two varieties of watermelon rootstocks, C. lanatus var. lanatus (CLL) and C. lanatus var. citroides (CLC), and two lines for each variety (Table 1) for a total of four PMR lines. Each combination of grafting was performed six times (six replications) and planted in trays in a greenhouse. Two independent experiments were conducted for each variety (CLL or CLC) of grafted plant. Additionally, six replicates of ungrafted Mickey Lee (ML-UG) were used as negative controls for both the PMS comparisons and self-grafted Mickey Lee (ML-SG) comparisons. All plants were of the same age and were grown in the same experimental conditions. Sample Harvesting and Metabolism Quenching. To harvest each sample from 8-week-old grafted plants, 1 cm of rootstock (stem) directly below the grafting site and 1 cm of scion (stem) directly above the grafting site were collected. At the same time, 1 cm leaf tissues as well as cotyledon tissue were collected separately from rootstocks and their corresponding scion. The samples were immediately submerged into liquid nitrogen to quench all metabolic processes and to prevent any stress responses from the cutting process. Frozen samples were then lyophilized for 48 h, at −40 °C and 0.05 mPa, to remove all water and permanently quench metabolism. Once the samples were dried, they were manually homogenized prior to extraction. Metabolite Extraction. Methanol−chloroform−water extractions of the dried rootstocks were performed as described in ref 15. Briefly, 20 mg of dried and homogenized plant material was used for each sample/replicate, and the solvent volumes were calculated using the dry mass to achieve a constant ratio of 2:2:1.8 of chloroform/methanol/ water according to the Bligh and Dyer method.28,29 The dried plant material was rehydrated with the appropriate ice-cold methanol/water solution. After vortex-mixing, the mixture was transferred into the ice-cold chloroform/water solution in glass tubes. The sample was incubated on ice for 10 min and centrifuged for 10 min at 2000g at 4 °C. The top layer, that is, the hydrophilic extract, was transferred into new microcentrifuge tubes. The hydrophilic extracts were dried using a Centrivap centrifuge for ∼15 h and stored at −20 °C until further preparation for NMR analysis. NMR Sample Preparation and Spectroscopy. The dried hydrophilic extracts from each sample were resuspended in 620 μL of NMR buffer (100 mM sodium phosphate buffer (pH 7.3), 1 mM 3-(trimethylsilyl)propionic-2,2,3,3-d4 acid (internal standard, TMSP; CAS Registry No. 24493-21-8), and 0.1% sodium azide, in 99.9 atom % D2O) and then vortex-mixed and centrifuged to remove any remaining undissolved material. The experiments were conducted at 298 K using 600 μL of sample in 5 mm NMR tubes (Norell). All NMR spectra were recorded on a Bruker Avance III spectrometer operating at 700 MHz. The first increment of a presat-noesy experiment was acquired using the standard Bruker noesypr1d pulse sequence. This resulted in a onedimensional 1H spectrum (metabolic profile) for each sample. All data were collected using a spectral width of 16.0 ppm and 64K points resulting in an acquisition time of 2.9 s. The spectra were collected with 120 scans, 4 dummy scans, 3 s relaxation delay, and presaturation at the residual water frequency. The 90° pulse widths were measured for each sample using the automatic pulse calculation experiment (pulsecal) in TopSpin 2.1.1 (BrukerBioSpin, Billerica, MA, USA). Two-dimensional 1 H−13C HSQC data were collected for a representative rootstock and scion sample using the Bruker hsqcedetgpsisp2.2 pulse sequence. The 1 H was observed in the F2 channel with a spectral width of 11 ppm,
Figure 3. Pairwise PCA score plots. PC1 versus PC2 score plots are shown for each pairwise comparison. In each, the total explained variance is >90%. The ovals represent 95% confidence intervals. R2, R3, R4, R5, and R6 are pairwise comparisons of rootstocks: R2, PMRCLL (a) versus PMS-CLL (b); R3, PMR-CLC (a) versus PMS-CLL (b); R4, PMR-CLC (a) versus PMR-CLL (b); R5, UG-ML (a) versus PMS-CLL (b); R6, UG-ML(a) versus SG-ML (b). S2, S3, S4, S5, and S6 are pairwise comparisons of the corresponding scions: S2, ML grafted on PMR-CLL (a) versus ML grafted on PMS-CLL (b); S3, ML grafted on PMR-CLC (a) versus ML grafted on PMS-CLL (b); S4, ML grafted on PMR-CLC (a) versus ML grafted on PMR-CLL (b); S5, UG-ML (a) versus ML grafted on PMS-CLL (b); S6, UG-ML (a) versus SG-ML (b). The Mahalanobis distances (DM) are listed for each. 8086
DOI: 10.1021/acs.jafc.5b02108 J. Agric. Food Chem. 2015, 63, 8083−8091
Article
Journal of Agricultural and Food Chemistry Table 3. Statistical Significance Analysisa statistical parameter
R2
S2
R3
S3
R4
S4
R5
S5
R6
S6
DM T2 Fc Ft significance status
4.6 70 4.96 31.9 yes
2.90 25.33 4.96 11.40 yes
9.4 265.4 4.96 119.4 yes
4.21 52.91 4.96 23.81 yes
4.92 72.72 4.96 32.73 yes
1.44 6.21 4.96 2.84 no
1.43 6.09 4.96 2.74 no
1.20 4.56 4.96 2.10 no
1.62 7.16 4.96 3.25 no
1.35 5.72 4.96 2.45 no
a
Mahalanobis distance (DM), two-sample Hotellings’ T2 test, and F test values shown, where F-true > F-critical indicates a statistically significant separation between the groups. R2, R3, R4, S2, and S3 have statistically significant separations, whereas R5, R6, S4, S5, and S6 do not. whereas the 13C was observed in the F1 channel with a spectral width of 180 ppm. NMR Data Analysis. Multivariate Statistical Analysis. Prior to principal component analyses (PCA), bucket tables were generated by using AMIX software (version 3.9.7, BrukerBioSpin). Once data were normalized to total intensities, spectra were binned into 0.01 ppm wide buckets over the region of 0.5−10.0 ppm using advanced bucketing in AMIX. Other bucket widths (either larger or smaller) resulted in similar results. The water region (4.75−4.90 ppm) of the spectra was eliminated for all analyses. No scaling method was found necessary. PCA was performed on the bucket tables generated from AMIX using MetaboAnalyst 2.0 (MetaboAnalyst 2.0, a comprehensive server for metabolomic data analysis).30 For each score plot generated during the analysis, the Mahalanobis distance (DM), two-sample Hotelling’s T2 statistic (T2), F values (Ft) and critical F values (Fc) were calculated using MatLabR2010b.31 Spectral Analysis and Metabolite Identification. To identify the metabolites that are different in concentration when resistant and susceptible rootstocks and scions were compared, fold changes and related p values were determined for each bucket. The calculations of fold changes and t tests were performed with MetaboAnalyst 2.0 to determine whether or not the changes in bucket intensities were statistically significant. Buckets with corresponding p values 1 indicates metabolites are up-regulated for the sample in the numerator. A fold change 1.00 indicate a net increase and fold changes
