The Jatropha leaf curl Gujarat virus on infection regulates the sugar and tricarboxylic acid cycle metabolic pathways in Jatropha

Jatropha, a popular biodiesel crop, suffers severe losses due to Jatropha leaf curl Gujarat virus (JLCuGV) infection in Gujarat (India). Gas chromatography-mass spectrometry (GC-MS) analysis of healthy and infected tissue revealed that the metabolites like sugars, polyols, carboxylic acids, fatty acids, polyphenols, and amino acids were regulated on JLCuGV infection. The sugars (glucose, sucrose, and fructose) increased, while carboxylic acids (malic acid, citric acid and quinic acid) and polyols (galactinol, butanetriol, triethylene glycol, myo-inositol, erythritol) decreased remarkably in infected Jatropha tissue. The increased accumulation of polyphenols in JLCuGV infected tissue suggests activation and involvement of polyphenol biosynthesis pathways as a defence response to geminivirus. All these metabolic variations indicated that sugar metabolism and tricarboxylic acid (TCA) cycle pathways are regulated as a defence response as well as a disease development response to geminivirus infection in Jatropha.


Introduction
Jatropha (Jatropha curcas L.) is an economically important biodiesel crop which grows on wastelands. The Jatropha biodiesel is successfully used to run the vehicles without any modi cations in engines [1,2]. Jatropha, being a non-edible crop, serves as a better and viable option compared to other food crops like rapeseed, sun ower, palm, soybean, corn, sugarcane, etc. for biofuel production. Recently, Jatropha biofuel was for the rst time used in aircraft (Spicejet's Bombardier Q400) [3], indicating its high e ciency as fuel. The major problem in Jatropha biodiesel production is its lower yield because of various diseases like collar rot, leaf spots, powdery mildew, seedling blight, seedling stem rot and viral infestation [4]. The leaf curl mosaic disease caused by geminivirus belonging to Begomovirus genus and is a severe threat to Jatropha, causing up to 80% yield losses [5].
The metabolic changes in response to fungal and bacterial disease infection are well studied in plants as compared to the viral infection [6]. The interaction between the host plant and geminivirus changes the plant's physiology leading to the development of disease symptoms [7]. The metabolite changes are also among the typical responses of host plant after virus infestation [6]. The qualitative and quantitative variation in the plant's metabolite variation can help understand the plant physiological response to virus infection. Due to systemic infection, entire plant physiology is affected in geminivirus infected plant involving disruption in plant's replication machinery, metabolic resources, gene regulation which negatively affects the plant's growth and development [7]. Still, the symptoms are more prominent on leaves of the plant. The altered cellular metabolism helps the virus to multiply in the host [8]. Plants also tend to produce some metabolites which act as a defence against virus infection such as amino acids, secondary metabolites etc. [9,10]. The secondary metabolites like terpenes/terpenoids, avonoids, alkaloids, and phenol derivatives are mainly produced from primary metabolites like sugars, amino acids, carboxylic acids, and their concentration gets affected on various virus infection in plants [10]. The concentration of phenylpropanoids and the avonoids also gets increased as a defence response against geminivirus infection [9]. The terpenoid biosynthesis suppressed after geminivirus infection in tomato [11], suggesting that depletion of terpenoids might be necessary for geminiviruses to establish a successful infection. The virus infection in plants causes an alteration in the metabolism of carbohydrates, starch, sugars and amino acids [12], indicating that the virus infection disrupts the metabolic pathways and damages the plant physiology. The carboxylic acids, which are the by-products of tricarboxylic acid (TCA) cycle, also get affected in geminivirus infection [9].
The studies on metabolite variations on geminivirus infection in Jatropha are limited. An earlier study by Sood and Chauhan [13] documented the upregulation of terpenoid biosynthesis and fatty acid metabolism genes, but the actual metabolite variation not studied. The metabolic changes were studied in J. curcas with Jatropha mosaic virus (JMV) infection by Sidhu et al. [14] using high-resolution magic angle spinning (HR-MAS) NMR spectra, and only 11 metabolites were detected. The comparative metabolic pro les of an infected and healthy plant can help to unravel the changes in biochemical pathways on geminivirus infection. Earlier, we had identi ed the Jatropha leaf curl Gujarat virus (JLCuGV) infecting the Jatropha from Gujarat [15]. In the present study, the modi cations in metabolite pro ling were analysed to understand the regulation of the metabolic pathways on natural JLCuGV infection in Jatropha. The collected leaf samples were stored at -80°C until use.

Detection and Identi cation of begomoviruses
The DNA from collected Jatropha leaf samples was isolated using a method speci cally developed for Jatropha [18]. The isolated DNA (100 ng) then used as template for detecting the presence of begomovirus by PCR using primers VirLCF (5'-GACCGCATAAATAATTGTGTGGGCCG-3') and VirLCR (5'-GCTGTAGAAATATATGCGTATTTTC-3'), which were designed speci cally for detecting the begomoviruses infecting Jatropha. The complete DNA-A of JLCuGV cloned in pBluescriptII KS(+) vector from our earlier study [15] was taken as a positive control.
Gas chromatography-mass spectrometry (GC-MS) analysis After con rming the presence and absence of JLCuGV in collected infected and healthy leaf samples, respectively, we used these samples for GC-MS analysis. The entire leaves were ground to a ne powder using liquid nitrogen and 100 mg of crushed powder tissue taken for sample preparation. The sample preparation and derivatisation by N-methyl-N-trimethylsilyl tri uoroacetamide (MSFTA) was carried out as given in Lisec et al. [20]. The derivatised samples were then transferred into glass vials suitable for GC-MS analysis. The GC-MS analysis was carried out as given in Paidi et al. [21]. Three replicates of each sample were taken for analysis.

Statistical analysis
The mean values and standard deviations were calculated of three independent biological replicates. Two paired sample t-test was employed as the statistical method to determine the signi cance of the difference in concentration of all the detected metabolites in healthy (control) and virus-infected leaf tissue of Jatropha and p<0.05 was considered to be statistically signi cant.

Geminivirus infection in Jatropha
The Jatropha plants of different accessions like CP9, PCM, Ranpur, Shyamalaji, and Zanjmer were infected by geminivirus at Nesvad, Chorvadla and Bhadraval eld sites of CSIR-CSMCRI with >70% geminivirus disease incidence [15]. The symptoms were similar in all accessions with severe leaf curling. In the present study, we used only CP9 accession from Nesvad location as it is economically more important due to its high seed and oil yield [20]. The healthy plants showed no leaf curl mosaic symptoms (Fig. 1A). In the case of younger plants, all the leaves showed the prominent yellow mosaic curl symptoms (Fig. 1B), while in old plants, only the new young leaves showed the prominent yellow mosaic curl symptoms (Fig. 1C). The virus responsible for causing the leaf curl disease was identi ed as JLCuGV in our earlier study [15]. The collected Jatropha leaf samples (three replicates each of healthy and infected leaf tissue) were checked for geminivirus infection. Only those samples showing leaf curl symptoms were identi ed as positive by PCR (Fig. 1D) and were used for the metabolite analysis.

Metabolite pro ling in healthy and infected Jatropha leaf
The metabolite analysis of geminivirus infected and healthy Jatropha leaf tissue revealed an accumulation of 98 total metabolites. ( Table 1). Out of them, 49 metabolites were detected in both tissues, 28 and 21 unique metabolites in healthy and infected tissue, respectively ( Fig. 2A, Table 1). All the detected metabolites were categorised in eight different groups like alkanes, amino acids, carboxylic acids, fatty acids, fatty alcohols, polyols, polyphenols and sugars. The compounds like Dgalactopyranoside and 2-ketoglutaric acid were detected in higher amount in healthy tissue only, while hexonic acid and 4-Aminobutyric acid were two compounds which were detected in higher concentration in only infected tissue (Table 1).
Sugars were the largest group of compounds, and the amount of sugar was more in the infected tissue (48%) than healthy tissue (35%; Fig. 2B). Twelve different sugars were identi ed in both healthy and infected tissue. In the infected tissue, sugars like fucopyranose, trehalose, glyceryl-glycoside, and mannobiose decreased, while sucrose, glucose, galactoside, glucopyranose, lactose and fructose increased. L-fucopyranose decreased by 2-fold, while lactose and fructose increased signi cantly by 2.14and 3.23-fold, respectively ( Table 2, Fig. 3).
The carboxylic acid was the second largest group identi ed with 26% and 16% concentration in the healthy and geminivirus infected tissue, respectively (Fig. 2B). Fourteen carboxylic acids were detected in both healthy and infected tissue. Carboxylic acids like glutaric acid, malic acid, quinic acid, pentadeca uorooctanoic acid, succinic acid, citric acid and threonic acid were signi cantly decreased, while glyceric acid was signi cantly increased in infected tissue (Table 2, Fig. 3). In the infected tissue maximum decrease of 24.5-fold was observed in glutaric acid, >2-fold decrease in malic acid, quinic acid, galactaric acid, methoxyacetic acid, pentadeca uorooctanoic acid, succinic acid as compared to healthy tissue. The glyceric acid, gluconic acid, fumaric acid, and shikimic acid increased by <1.2-fold in infected tissue ( Table 2).
In addition to sugars and carboxylic acids, the next larger group was polyols. The polyols accumulation was more in healthy tissue (23%) as compared to infected tissue (19%; Fig. 2B). Total nine polyols were detected in both healthy and infected tissue. Amongst them, galactinol, triethylene and myo-inositol decreased signi cantly, while deoxypentitol increased signi cantly in the infected tissue (Fig.3). Galactinol decreased by 2.04-fold, while deoxypentitol increased by 2.77-fold in infected tissue (Table 2).
Similarly, the concentration of fatty acids was also lower in infected tissue (6%) as compared to healthy tissue (8%). Total ve fatty acids were detected, and among them, docosanoic acid, palmitic acid, and stearic acid were signi cantly decreased by >1.5-fold, while butanoic acid octatriacontyl penta uoropropionate were unaffected (Table 2, Fig. 3). The polyphenols were more in infected tissue (10%) than healthy tissue (6%). Total three polyphenols detected in both tissue and among them, quercetin was increased signi cantly by 3.7-fold in infected tissue (Table 2, Fig. 3).
The other compounds like fatty alcohols, alkanes and amino acids were detected in less quantity (< 2%) in both the healthy and infected tissue, furthermore their concentration decreased on the infection (Fig.   2B). Amongst them, L-serine is the most affected compound with a 19.67-fold decrease in infected tissue ( Table 2). The identi ed compounds mostly participate in sugar metabolism and TCA cycle. Their variation in healthy and infected tissue is shown in the interlinked pathway of sugar metabolism and TCA cycle (Fig. 4). Sugars (glucose, fructose, sucrose, galactoside, glucopyranose and lactose) were mostly increased in infected tissue, while the compounds generated from the metabolism of these sugars like sugar alcohols (butanetriol, erythritol, myo-inositol, galactinol) and carboxylic acids (galactaric acid threonic acid and ribonic acid) were higher in healthy tissue. Glutaric acid, a by-product of amino acid metabolism, was also higher in healthy tissue. Similarly, compounds of TCA cycle (citric acid, succinic acid and malic acid) and derivatives of fatty acids (docosanoic acid, dihydroxybutanoic acid, palmitic acid and stearic acid) were more in healthy tissue (Fig. 4).

Discussion
The young Jatropha plants infected with JLCuGV showed the mosaic curl symptoms all over the plant, suggesting that early infection in the young plant could spread to the entire plant. In contrast, in mature Jatropha plant, only new young leaves show mosaic curl symptoms, indicating that the Jatropha plants infected in early-stage were highly damaged. The extent of damage reduced with delay in infection [22].
The JLCuGV successfully replicated in young leaf cells undergoing endoreduplication [23], but the mature leaves of matured Jatropha plant remained unaffected, indicating the restriction of virus spread in mature cells, leading to less damage.
The compatible plant-virus interactions result in disease development and lead to changes in gene expression and metabolite content. In Jatropha, the regulation of the hormonal signalling of salicylic acid (SA), jasmonic acid (JA), abscisic acid (ABA), auxin and ethylene was reported during the geminivirus infection [13,24]. As the previous study by Sidhu et al. [35] showed limited metabolite analysis, we performed the total metabolite analysis using GC-MS pro ling of Jatropha leaf to reveal the changes in different metabolite pathways during geminivirus infection. In the present study, 98 metabolites were identi ed based on GC-MS analysis of infected and healthy Jatropha plant, and the analysis revealed differential metabolite pro ling.
The sugars were increased in infected tissue of Jatropha because the sugar metabolism was negatively affected and resulted in decreased production of polyols and carboxylic acids, which are the by-products of sugar metabolism (Fig. 4). The similar negative regulation of sugar metabolism in tobacco plants due to JLCuGV infection was also reported in our earlier study [15]. Impaired sugar transport may be the reason for the increased sugar content during virus infection [25]. Among polyols, galactinol, which was earlier reported as a signalling molecule to trigger the plant defence on bacterial pathogen attack [26], was decreased by > 2 fold in infected tissue, indicating that galactinol triggered plant defence might not be activated on JLCuGV infection. Myo-inositol, which plays a role in the membrane and cell wall synthesis [27], decreased in infected tissue, indicating the impaired cell wall biosynthesis in infected Jatropha tissue. The carboxylic acids decreased in infected Jatropha tissue, and the similar result reported in the Papaver somniferum on Ageratum enation virus (AEV) infection [28] and in our earlier study in tobacco on JLCuGV infection [15]. The increase in sugars and decrease in polyols, carboxylic acids in infected Jatropha tissue indicates that both sugar metabolism and TCA cycle are affected in infected Jatropha tissue (Fig. 4). The interruption in the TCA cycle and low accumulation of glucose and sucrose was also reported earlier in Jatropha infected tissue by Sidhu et al. [14].
Fatty acids, which are the precursors of the JA biosynthetic pathway [29], decreased in infected Jatropha tissue, indicating the JA mediated defence response impaired in infected tissue. The similar decrease in fatty acid composition was observed on TRV infection in Arabidopsis [6]. The polyphenols were more in infected tissue, indicating that these polyphenol biosynthesis pathways are activated in response to geminivirus attack. The polyphenols were reported to be produced in the plant to protect them from against pathogens and phytophagous insects [30]. The phenolic compounds also accumulated in response to geminivirus infection in capsicum [31] and pumpkin [32], and the phenolic compounds facilitate the formation of physical barriers to restrict the systemic spread of the virus [33]. Out of the common compounds, which were identi ed in both healthy and infected Jatropha leaf tissue, the concentration of L-serine, a polar amino acid playing a fundamental role in metabolism and cell signalling [34], showed maximum decrease in infected tissue, indicating disturbed metabolism and cell signalling in Jatropha after geminivirus infection.
The compound like hexonic acid was detected only in infected Jatropha tissue in a higher amount, indicating that its synthesis gets activated with geminivirus infection. It was earlier reported that hexonic acid is stress signalling molecule providing hexanoic acid-induced resistance (Hx-IR) against several pathogens like Botrytis cinerea and Pseudomonas syringae in tomato and Alternaria alternata in citrus [35,36]. The hexonic acid also blocked the systemic spread of Melon necrotic spot virus (MNSV; family Tombusviridae) by activating SA pathway and increasing the callose deposition in melon [37]. In Jatropha, both the JA and SA pathway get activated on geminivirus infection [13], suggesting that hexonic acid might be involved in activation of these pathways in Jatropha with geminivirus attack. After hexonic acid, the second-highest compound identi ed in JLCuGV infected leaf tissue of Jatropha was 4aminobutyric acid (gamma-aminobutyric acid (GABA)). A similar increase in GABA concentration in infected Jatropha leaf was observed by Sidhu et al. [14]. GABA found to accumulate in both abiotic and biotic stress [38]. Although its role in geminivirus infection is still not reported, its accumulation increases with herbivores insect feeding [39], so the white y feeding might have stimulated GABA accumulation in Jatropha infected leaves. The 2-ketoglutaric acid (α-ketoglutarate), an intermediate of the TCA cycle which determines the overall rate of a citric acid cycle [40], was detected only in the healthy tissue of Jatropha, indicating that the TCA cycle was e ciently functioning in healthy tissue.
In conclusion, the metabolite analysis in healthy and infected Jatropha leaves revealed that the sugar metabolism and TCA cycle pathway were negatively affected in infected leaves, leading to increased sugar content and decreased polyols and carboxylic acid content. The metabolite ux of Jatropha plant showed JLCuGV infection impairs plant physiology and development of leaf curl mosaic symptoms. The metabolites were regulated as defence response and also as a consequence of the disease incidence in Jatropha after JLCuGV infection. The authors state that there is no con ict of interest regarding the publication of this article. Table 1 The concentration of all the detected metabolites in healthy (control) and virus-infected Jatropha leaf tissue. The compounds that showed a signi cant decrease and increase (as determined by two paired sample t-test at p<0.05) of concentration in the infected tissue than healthy tissue are marked by "*" and "**", respectively. 3-Hydroxyisovaleric acid --0.14 0.00 Table 2 The fold change in the common metabolites in infected tissue as compared to healthy tissue of Jatropha.
Class of Compound   Heat map showing the concentrations of different metabolites detected in both healthy and infected tissue. The metabolites which showed a signi cant decrease and increase (as determined by two paired sample t-test at p<0.05) in infected leaf tissue are marked with "*" and "**", respectively.