Metabolite Variation between Nematode and Bacterial Seed Galls in Comparison to Healthy Seeds of Ryegrass Using Direct Immersion Solid-Phase Microextraction (DI-SPME) Coupled with GC-MS

Annual ryegrass toxicity (ARGT) is an often-fatal poisoning of livestock that consume annual ryegrass infected by the bacterium Rathayibacter toxicus. This bacterium is carried into the ryegrass by a nematode, Anguina funesta, and produces toxins within seed galls that develop during the flowering to seed maturity stages of the plant. The actual mechanism of biochemical transformation of healthy seeds to nematode and bacterial gall-infected seeds remains unclear and no clear-cut information is available on what type of volatile organic compounds accumulate in the respective galls. Therefore, to fill this research gap, the present study was designed to analyze the chemical differences among nematode galls (A. funesta), bacterial galls (R. toxicus) and healthy seeds of annual ryegrass (Lolium rigidum) by using direct immersion solid-phase microextraction (DI-SPME) coupled with gas chromatography–mass spectrometry (GC-MS). The method was optimized and validated by testing its linearity, sensitivity, and reproducibility. Fifty-seven compounds were identified from all three sources (nematode galls, bacterial galls and healthy seed), and 48 compounds were found to be present at significantly different (p < 0.05) levels in the three groups. Five volatile organic compounds (hexanedioic acid, bis(2-ethylhexyl) ester), (carbonic acid, but-2-yn-1-yl eicosyl ester), (fumaric acid, 2-ethylhexyl tridec-2-yn-1-yl ester), (oct-3-enoylamide, N-methyl-N-undecyl) and hexacosanoic acid are the most frequent indicators of R. toxicus bacterial infection in ryegrass, whereas the presence of 15-methylnonacosane, 13-methylheptacosane, ethyl hexacosyl ether, heptacosyl acetate and heptacosyl trifluoroacetate indicates A. funesta nematode infestation. Metabolites occurring in both bacterial and nematode galls included batilol (stearyl monoglyceride) and 9-octadecenoic acid (Z)-, tetradecyl ester. Among the chemical functional group, esters, fatty acids, and alcohols together contributed more than 70% in healthy seed, whereas this contribution was 61% and 58% in nematode and bacterial galls, respectively. This study demonstrated that DI-SPME is a valid technique to study differentially expressed metabolites in infected and healthy ryegrass seed and may help provide better understanding of the biochemical interactions between plant and pathogen to aid in management of ARGT.


Compounds
RI RT

Multivariate Analysis to Classify Nematode Galls, Bacterial Galls and Healthy Seeds
Distribution of compounds in the three treatment sources was analyzed using principal component analysis (PCA) ( Figure 3) and a hierarchical cluster analysis heat map ( Figure 4). PCA is an unsupervised method that helps to visualize covariance and correlation among the three treatment sources and metabolites expressed in each.
There is clustering of metabolites from the three replicates of each treatment source, and the three treatment ellipses do not overlap ( Figure 3). This separation indicates that the three treatment sources have different chemical profiling, and this profiling can be used as a marker to detect nematode and bacterial infection in annual ryegrass. The first principal component explains 62.2% of the variance, whereas the second principal component explains 33.2%. Similarly, Figure 4 shows the cluster and heat map of all three Molecules 2023, 28, 828 5 of 12 treatment sources used in this study. The heat map shows a clear expression level difference between healthy seed, nematode and bacterial galls.
Further, to know the differential compounds between the individual groups, supervised analysis was conducted. Partial least squares-discriminant analysis (PLS-DA) and orthogonal PLS-DA (OPLS-DA) score plots were drawn between nematode gall versus healthy seed and bacterial gall versus healthy seed ( Figure 5). The compounds 2-cyclohexen-1-ol and 1-hexadecanol, 2-methyl clearly distinguish healthy seeds from nematode galls by showing their abundance in healthy seeds, whereas due to the abundant nature of carbonic acid, but-2-yn-1-yl eicosyl ester and fumaric acid, 2-ethylhexyl tridec-2-yn-1-yl ester in the bacterial gall, they were separated from the healthy seeds.
Chemical functional groups differed among the three. Esters, fatty acids, and alcohols were the dominant groups in healthy seed, representing 34%, 17%, and 13% of the total number of compounds, respectively, followed by amides, ethers, hydrocarbons, and phenols, each at 8%, and heterocyclic compounds at 4% ( Figure 6). In nematode galls, the proportions of esters, hydrocarbons, and acids were 48%, 20%, and 10%, respectively, followed by ethers and phenols, each at 7%, and alcohols and aldehydes, each at 3%. In bacterial galls, the proportions of esters, acids, alcohols, hydrocarbons were 36%, 16%, 12%, and 12%, respectively, followed by amides, ethers, and phenols, with 8% in each group. Interestingly, no amides and only the aldehydes were present in the nematode galls, and heterocyclic compounds were found only in healthy seed. In addition to this, esters were highly expressed in each treatment, and the fraction of hydrocarbons was greater in infected seed than healthy seed.

Multivariate Analysis to Classify Nematode Galls, Bacterial Galls and Healthy Seeds
Distribution of compounds in the three treatment sources was analyzed using principal component analysis (PCA) ( Figure 3) and a hierarchical cluster analysis heat map ( Figure 4). PCA is an unsupervised method that helps to visualize covariance and

Multivariate Analysis to Classify Nematode Galls, Bacterial Galls and Healthy Seeds
Distribution of compounds in the three treatment sources was analyzed using principal component analysis (PCA) (Figure 3) and a hierarchical cluster analysis heat map (Figure 4). PCA is an unsupervised method that helps to visualize covariance and correlation among the three treatment sources and metabolites expressed in each.       and orthogonal PLS-DA (OPLS-DA) score plots were drawn between nematode ga versus healthy seed and bacterial gall versus healthy seed ( Figure 5). The compounds cyclohexen-1-ol and 1-hexadecanol, 2-methyl clearly distinguish healthy seeds fro nematode galls by showing their abundance in healthy seeds, whereas due to th abundant nature of carbonic acid, but-2-yn-1-yl eicosyl ester and fumaric acid, ethylhexyl tridec-2-yn-1-yl ester in the bacterial gall, they were separated from the health seeds. Figure 5. Partial least squares-discriminant analysis (PLS-DA) and orthogonal PLS-DA (OPLS-DA score plots of ryegrass samples. (a,b) PLS-DA score plots for bacteria gall versus healthy seed an nematode gall versus healthy seed, respectively; (c,d) OPLS-DA score plots for bacteria gall vers healthy seed and nematode gall versus healthy seed, respectively. Red represents bacterial grou blue represents nematode group, and green represents healthy seeds for bacteria gall versus health seed and nematode gall versus healthy seed, respectively. PLS-DA score plots for bacteria gall versus healthy seed and nematode gall versus healthy seed, respectively; (c,d) OPLS-DA score plots for bacteria gall versus healthy seed and nematode gall versus healthy seed, respectively. Red represents bacterial group, blue represents nematode group, and green represents healthy seeds for bacteria gall versus healthy seed and nematode gall versus healthy seed, respectively.
Molecules 2023, 28, 828 8 of 13 Chemical functional groups differed among the three. Esters, fatty acids, and alcohols were the dominant groups in healthy seed, representing 34%, 17%, and 13% of the total number of compounds, respectively, followed by amides, ethers, hydrocarbons, and phenols, each at 8%, and heterocyclic compounds at 4% (Figure 6). In nematode galls, the proportions of esters, hydrocarbons, and acids were 48%, 20%, and 10%, respectively, followed by ethers and phenols, each at 7%, and alcohols and aldehydes, each at 3%. In bacterial galls, the proportions of esters, acids, alcohols, hydrocarbons were 36%, 16%, 12%, and 12%, respectively, followed by amides, ethers, and phenols, with 8% in each group. Interestingly, no amides and only the aldehydes were present in the nematode galls, and heterocyclic compounds were found only in healthy seed. In addition to this, esters were highly expressed in each treatment, and the fraction of hydrocarbons was greater in infected seed than healthy seed.

Discussion
To optimize the DI-SPME extraction protocol, the fiber was exposed at room temperature (25 ± 2 °C) for the analyte extraction for different times (30, 45, 60 and 120 min) and desorption time (3 min, 6 min and 10 min). The optimum extraction time, 60 min,

Discussion
To optimize the DI-SPME extraction protocol, the fiber was exposed at room temperature (25 ± 2 • C) for the analyte extraction for different times (30,45, 60 and 120 min) and desorption time (3 min, 6 min and 10 min). The optimum extraction time, 60 min, and desorption time, 6 min, were optimized based on the number of peaks and peak areas. Similarly conditioned optimization parameters for SPME extraction were reported for VOC detection in barley [31].
Nontargeted metabolite analysis revealed relative expression levels of compounds present in the sample. This enables identification of the prominent compounds in each treatment, information that can be used to develop diagnostic assays and inform breeding efforts. Bacterial gall samples expressed 2,4-dihydroxyheptadecyl acetate derivatives, and this has been described previously [32,33]. Fatty acids play an important role in bacterial lipid membranes, and it has previously been reported that the presence of C-15 saturated branched chain acid, pentadecanoic acid, C-17 saturated branched chain acids and normal saturated fatty acids, such as lauric, myristic, stearic and arachidic, are major components of cell membranes in Rathayibacter species. [34][35][36]. Here, the closely related fatty acids tetradecanoic acid (myristic), hexadecenoic acid, and hexacosanoic acid were identified in bacterial galls. The presence of oct-3-enoyl amide, N-methyl-N-undecyl-, a structural derivative of an alpha-amino acid, and may be responsible in biological pathways for producing 2,4-diaminobutyric acid, which is a characteristic of R. toxicus [37]. Maulidia et al. [38] reported steroidal compounds as the major component in methanolic extracts of Gram-positive Bacillus thuringiensis associated with root knot nematodes, and similarly, this study shows the presence of stigmastane-3,6-dione (5α), which is classified as a steroid compound, in Gram-positive R. toxicus bacterial galls. Fumaric acid is an important intermediate of the tricarboxylic acid (TCA) cycle in plant systems [39]. Expression of fumaric acid, 2-ethylhexyl tridec-2-yn-1-yl ester, in the bacterial galls suggests that this bacterium could be interfering with the host TCA cycle.
There is no published information available on volatile organic compounds generated in nematode-infested cereal seed galls. However, there are reports that showed the protein content is reduced during nematode infestation in ear cockle disease of wheat [40]. Similarly, compounds containing amide groups, such as pent-4-enoyl amide, 2-methyl-N-dodecyl-, pent-4-enoyl amide, 2-methyl-N-tetradecyl-and oct-3-enoylamide, and N-methyl-Nundecyl-were downregulated in nematode gall. Compounds expressed by soybean cyst nematode and analyzed with GC-MS were methyl ester derivatives, a component of nematode pheromones [41]. In our study, the presence of the derivatives of methyl, ethyl and vinyl esters in the nematode galls could indicate the presence of pheromones associated with A. funesta. Nematodes are also known to contain significant quantities of lipids in their body cuticle [42], and fatty acids and fatty aldehydes are major components of the nematode cuticle [43]. Phospholipids play an important role in the formation of nematode cellular membranes and in signal transduction pathways [44]. In our results, the presence of octadecanal, which is a phospholipid derivative [45], could be a potential marker to identify A. funesta. Wei et al. [46] mentioned the role of 17-octadecynoic acid during synthesis of phospholipids in Caenorhabditis elegans, suggesting that 17-octadecynoic acid, methyl ester and (E)-9-octadecenoic acid ethyl ester may play a role in the development of the nematode in the nematode gall. Among the fatty acids, the presence of saturated fatty palmitic (16:0) and stearic (18:0) acids were reported from the parasitic nematode Ascaridia galli [43], whereas we found the unsaturated fatty octadecenoic acid (18:1) and nonadecanoic acid (19:1), which may indicate different requirements for fatty acids in A. funesta.

Sample Preparation and Extraction Using DI-SPME
Twenty-five milligrams of each treatment source of plant sample (10-15 seeds or galls without removal of lamella or the gall walls) was transferred into 2 mL microtube containing 1 mL of HPLC grade acetonitrile. Three separate samples were prepared for each source. The microtube was sealed with a screw cap and shaken in a bedbug homogenizer (Benchmark BeadBug-Microtube Homogenizer, 115V, China), using two milling balls for 2 min at 4000 RCF, then another 1mL of acetonitrile was added, shaken for 2 min, and then centrifuged at 4000 RCF for 3 min using a minicentrifuge (Dynamica Velocity 13µ, Techcomp Europe). After that, the supernatant (~1.5 mL) was collected in a 2 mL GC vial with septum and stored at 4 °C until further use.
For conducting DI-SPME, a three-phase fiber (50/30 µm DVB/CAR/PDMS, Stableflex 2 cm, Supelco®, Bellefonte, PA, USA) was used. Three replicate sets of each sample were exposed to three different fibers. All new fibers were preconditioned for 60 min at 270 °C as per the manufacturer's instructions. The SPME fiber was directly immersed by inserting into the GC vial containing the sample for 60 min at room temperature (25 ± 2 °C), and thereafter the fiber was taken out and injected directly into the GC-MS instrument with desorption for 6 min at inlet temperature 290 °C to analyze the sample.

GC-MS Conditions
The compounds were analyzed using a GC-MS 7890B coupled with a 5977B MSD mass spectrometer based on Agilent Technologies, Santa Clara, CA, USA with an Agilent HP-5MS column (30 m length, 0.25 mm internal diameter, 0.25 µm film thickness with 5% phenyl, and 95% dimethylpolysiloxane stationary phase). The following analytical conditions were used: splitless mode with helium as a carrier gas with a flow of 1 mL/min. The GC conditions were adopted from a previously used method [49], with slight modifications in temperature programming, starting at 50 °C with 5 min hold, then ramped at 6 °C/min to 90 °C, 8 °C/min to 140 °C, 6 °C/min to 190 °C, 4 °C/min to 240 °C, finally to 50 °C/min to 300 °C with hold at 300 °C for 12 min. The detector was operated in electron impact (EI) ionization mode at 70 eV and the spectra were acquired from 3 scan/s in a range from 50 to 550 atomic mass units (amu). The transfer line temperature of the MSD was 280 °C, and the ion source temperature was 230 °C with total run time of 51.95 min.

Sample Preparation and Extraction Using DI-SPME
Twenty-five milligrams of each treatment source of plant sample (10-15 seeds or galls without removal of lamella or the gall walls) was transferred into 2 mL microtube containing 1 mL of HPLC grade acetonitrile. Three separate samples were prepared for each source. The microtube was sealed with a screw cap and shaken in a bedbug homogenizer (Benchmark Scientific, Shanghai, China), using two milling balls for 2 min at 4000 RCF, then another 1mL of acetonitrile was added, shaken for 2 min, and then centrifuged at 4000 RCF for 3 min using a minicentrifuge (Dynamica Velocity 13µ, Techcomp Europe, Livingston, UK). After that, the supernatant (~1.5 mL) was collected in a 2 mL GC vial with septum and stored at 4 • C until further use.
For conducting DI-SPME, a three-phase fiber (50/30 µm DVB/CAR/PDMS, Stableflex 2 cm, Supelco®, Bellefonte, PA, USA) was used. Three replicate sets of each sample were exposed to three different fibers. All new fibers were preconditioned for 60 min at 270 • C as per the manufacturer's instructions. The SPME fiber was directly immersed by inserting into the GC vial containing the sample for 60 min at room temperature (25 ± 2 • C), and thereafter the fiber was taken out and injected directly into the GC-MS instrument with desorption for 6 min at inlet temperature 290 • C to analyze the sample.

GC-MS Conditions
The compounds were analyzed using a GC-MS 7890B coupled with a 5977B MSD mass spectrometer based on Agilent Technologies, Santa Clara, CA, USA with an Agilent HP-5MS column (30 m length, 0.25 mm internal diameter, 0.25 µm film thickness with 5% phenyl, and 95% dimethylpolysiloxane stationary phase). The following analytical conditions were used: splitless mode with helium as a carrier gas with a flow of 1 mL/min. The GC conditions were adopted from a previously used method [49], with slight modifications in temperature programming, starting at 50 • C with 5 min hold, then ramped at 6 • C/min to 90 • C, 8 • C/min to 140 • C, 6 • C/min to 190 • C, 4 • C/min to 240 • C, finally to 50 • C/min to 300 • C with hold at 300 • C for 12 min. The detector was operated in electron impact (EI) ionization mode at 70 eV and the spectra were acquired from 3 scan/s in a range from 50 to 550 atomic mass units (amu). The transfer line temperature of the MSD was 280 • C, and the ion source temperature was 230 • C with total run time of 51.95 min.

Data Processing and Statistical Analysis
For the identification of compounds, the National Institute of Standards and Technology Mass Spectrometry (NIST MS) was used with Kovat's retention index (RI) of published data. The calculation for Kovat's index is provided in the supplementary file (Table S1). The data were processed with Agilent Mass Hunter Qualitative Analysis software (Version B.08.00, Build 8.0.8208.0, Santa Clara, CA, USA) [50]. The n-alkane (C 7 −C 40 , catalogue number 49451-U; Castle Hill, NSW, Australia) with concentration 100 µg/mL was used as the external standard. Microsoft Excel 2021 was used for data arrangement and other data processing. To perform multivariate analysis and principal component analysis (PCA) with one-way analysis of variance (ANOVA), MetaboAnalyst 5.0 (2022) (https://www.metaboanalyst.ca/MetaboAnalyst/upload/StatUploadView.xhtml; accessed on 20 June 2022) was used. R was used to draw the diagrams. Tukey's post hoc test (HSD) was performed for statistical differences with p < 0.05).

Conclusions
Direct immersion solid-phase microextraction using a three-phase fiber was employed to collect volatile organic compounds from acetonitrile extract from nematode galls, bacterial galls, and healthy ryegrass seed, followed by analysis in GC-MS. The GC-MS analytical data showed a wide spectrum of metabolites in the three treatment sources. For the first time, this study has successfully identified chemical compounds that differ between nematode galls, bacterial galls, and healthy ryegrass seed. These three treatment sources could be distinguished based on their chemical signatures and could form the basis for development of new analytical tools for determining the risk of ARGT occurrence from feedstuff samples.