Extra‐large G‐proteins influence plant response to Sclerotinia sclerotiorum by regulating glucosinolate metabolism in Brassica juncea

Abstract Heterotrimeric G‐proteins are one of the highly conserved signal transducers across phyla. Despite the obvious importance of G‐proteins in controlling various plant growth and environmental responses, there is no information describing the regulatory complexity of G‐protein networks during pathogen response in a polyploid crop. Here, we investigated the role of extra‐large G‐proteins (XLGs) in the oilseed crop Brassica juncea, which has inherent susceptibility to the necrotrophic fungal pathogen Sclerotinia sclerotiorum. The allotetraploid B. juncea genome contains multiple homologs of three XLG genes (two BjuXLG1, five BjuXLG2, and three BjuXLG3), sharing a high level of sequence identity, gene structure organization, and phylogenetic relationship with the progenitors’ orthologs. Quantitative reverse transcription PCR analysis revealed that BjuXLGs have retained distinct expression patterns across plant developmental stages and on S. sclerotiorum infection. To determine the role of BjuXLG genes in the B. juncea defence response against S. sclerotiorum, RNAi‐based suppression was performed. Disease progression analysis showed more rapid lesion expansion and fungal accumulation in BjuXLG‐RNAi lines compared to the vector control plants, wherein suppression of BjuXLG3 homologs displayed more compromised defence response at the later time point. Knocking down BjuXLGs caused impairment of the host resistance mechanism to S. sclerotiorum, as indicated by reduced expression of defence marker genes PDF1.2 and WRKY33 on pathogen infection. Furthermore, BjuXLG‐RNAi lines showed reduced accumulation of leaf glucosinolates on S. sclerotiorum infection, wherein aliphatic glucosinolates were significantly compromised. Overall, our data suggest that B. juncea XLG genes are important signalling nodes modulating the host defence pathways in response to this necrotrophic pathogen.


| INTRODUC TI ON
Signal transduction is one of the important biological processes through which organisms perceive extracellular stimuli and transmit it intracellularly. Heterotrimeric G-proteins (G-proteins), consisting of G-alpha (Gα), G-beta (Gβ), and G-gamma (Gγ) subunits, are a class of signal transducers conserved across eukaryotes. The model plant species Arabidopsis thaliana has a limited set of core G-protein subunits, with a canonical Gα and Gβ, together with three Gγ subunits (Urano et al., 2013). The Arabidopsis genome additionally encodes three extra-large Gα-like proteins (XLGs), namely, XLG1, XLG2, and XLG3 (Ding et al., 2008;Lee & Assmann, 1999). XLGs are twice as large as the canonical Gα protein, which led to their terminology as extra-large GTP binding proteins. The C-terminal region of XLGs shows structural homology to canonical Gα proteins; however, their N-terminal region is quite distinct (Ding et al., 2008). Despite the differences, XLGs show physical interaction with all three possible Gβγ dimers at the plasma membrane (Chakravorty et al., 2015;Heo et al., 2012;Maruta et al., 2015). XLGs thus represent additional nodes in plant G-protein signalling, having multiple roles during plant development and stress responses (Ding et al., 2008;Maruta et al., 2015;Pandey, 2019;Urano et al., 2016).
Various studies using inhibitors, agonists, and loss-of-function mutants suggest the role of G-protein signalling in plant defence (Trusov et al., 2009;Urano et al., 2016;Zhong et al., 2019). In general, a higher susceptibility response to bacterial and fungal pathogens was observed in the Arabidopsis Gβ-null mutant (agb1), xlg123-triple mutant, and Gγ triple mutant (agg123), but not in the null mutant of the canonical Gα (gpa1) and Col-0 wild type, indicating that biotic stress responses are biased towards XLGs rather than canonical Gα (Urano et al., 2016). The expression of XLG2 and XLG3 genes was found to be rapidly induced by the bacterial pathogen Pseudomonas syringae, whereas XLG1 transcription was not affected (Ding et al., 2008). Studies in Arabidopsis demonstrated that XLGs are positive regulators of resistance to the biotrophic pathogen P. syringae, which largely triggers the salicylic acid (SA)-responsive pathway; however, no significant difference was observed for the necrotrophic fungi Alternaria brassicicola and Botrytis cinerea, which trigger the jasmonic acid (JA)-responsive pathway (Maruta et al., 2015;Zhu et al., 2009).
The xlg2 mutation leads to compromised induction of pathogenresponsive (PR1 and PR2) genes during infected conditions. In addition, XLG-mediated resistance against hemibiotrophic fungi and bacteria is associated with receptor-like kinases (RLKs) and the defence mechanism is based upon the activation of programmed cell death (PCD). The Arabidopsis XLGs thus have distinct functions in disease resistance through their interaction with the Gβγ dimer (Zhong et al., 2019).
The genus Brassica is economically the most important genus of the Cruciferae family, crops of which are used for human nutrition as vegetables, oilseeds, and condiments. The inherent susceptibility of Brassica species to biotic stresses is a major factor that has limited their productivity in recent decades. Sclerotinia sclerotiorum, the causal agent of sclerotinia stem rot in Brassica crops, causes extensive (40%-80%) yield losses worldwide (Sharma et al., 2018) and has proved hard to control, with host resistance being insufficient. The persistence of this fungus is due to its sclerotia, which remain viable under adverse conditions and can be retained in soil for many years (Brustolin et al., 2016). This aggressive fungus is known to hijack plant defence by modulating a wide range of signalling cascades, defence phytohormones, and stress-associated metabolites (Liang & Rollins, 2018;Novakova et al., 2014). Recent reports in A. thaliana and Brassica species have established the key role of glucosinolates and their hydrolysis products against the necrotrophic pathogen S. sclerotiorum (Abuyusuf et al., 2018;Augustine & Bisht, 2015a;Chen et al., 2020Chen et al., , 2021Hopkins et al., 2009;Sotelo et al., 2015).
Brassica species have polyploid genomes, serving as an excellent crop model to study genome evolution and trait domestication (Augustine et al., 2014). We earlier reported that the mesohexaploid genomes of Brassica rapa and Brassica nigra contain multiple homologs of G-protein genes, which show distinct expression patterns in response to defence phytohormones and conditions mimicking biotic stress, suggesting putative involvement of G-protein signalling in plant defence Kumar et al., 2014). There is a growing consensus that the signalling molecules, particularly JA, SA, and ethylene (ET), interact in complex ways to fine-tune plant defence metabolites, including glucosinolate accumulation (Augustine & Bisht, 2015b;Mewis et al., 2005). At the core of our study is the hypothesis that XLGs could be the key signalling determinant in the plant's response to S. sclerotiorum, possibly by modulating the defence pathways and stress-associated metabolites. Here, we report identification of multiple homologs of XLG-encoding genes from the allotetraploid B. juncea and its progenitors, which are shaped by whole-genome duplication and triplication events (Augustine et al., 2014). Using a series of molecular genetic and metabolite analyses, we demonstrated that specific silencing of B. juncea XLG homologs led to an altered susceptibility against S. sclerotiorum infection, which was further correlated with the altered expression of defence pathway genes and glucosinolate accumulation during pathogen infection. Our study is a novel endeavour to delineate the XLG-mediated defence strategy in plants against S. sclerotiorum.

| Identification and evolution of B. juncea XLGencoding genes
We first investigated the inventory and molecular evolution of XLG genes in the polyploid Brassica species. Using AtXLG1-3 protein sequences as reference queries on the publicly available databases, multiple XLG-like sequences were retrieved from the allotetraploid B. juncea (AB genome) and its progenitors, namely, B. rapa (A genome) and B. nigra (B genome). The diploid B. rapa and B. nigra genomes contain one XLG1, three XLG2, and two XLG3 sequences each (Table 1). Subsequently, a total of two XLG1, five XLG2, and three XLG3 sequences could be identified in B. juncea; the A and B subgenome-specific homologs could be ascertained using the B. rapa and B. nigra gene sequences. The XLG genes identified in this study were named following the nomenclature adopted for Brassica genes (Østergaard & King, 2008 TA B L E 1 Summary of extra-large Gprotein (XLG)-encoding genes identified from Brassica juncea and its progenitor genomes five characteristic domains (G1-G5) required for guanine nucleotide binding ( Figure S1; Ding et al., 2008).

The presence of multiple XLG-like sequences in the extant
Brassica species led us to investigate the evolution of the XLG gene family by analysing their genomic structure and phylogenetic relationship. Analysis of genomic structure (http://gsds.cbi.pku.edu.cn/) showed that the XLG genes contain six to eight exons ( Figure 1a).
In comparison, the introns in XLG genes were highly divergent in their sizes and sequences, suggesting independent evolution and expansion of XLG1, XLG2, and XLG3 genes in the Brassica lineage.
To get a better insight into the expansion of the XLG gene family, a (1-month-old plant), unopened flower bud, and root (1-month-old plant), and (b) in response to S. sclerotium isolate Delhi-1 (SSD1) infection from 3 to 48 hr postinfection (hpi). The relative expression value of BjuXLG genes was obtained by quantitative reverse transcription PCR analysis using gene-specific primers (Table S5) and normalizing their expression against the endogenous controls BjuActin (set at 100) during plant developmental stages and BjuTIP41 (set at 1) for SSD1 infection experiment (Chandna et al., 2012). Three independent experiments were performed, each with two technical replicates. Error bars represent ± SE of the mean. Different letters indicate significant differences among the homologs of BjuXLG genes for each developmental stage or time point separately, calculated using one-way analysis of variance (p < 0.05) following Tukey's post hoc test

| Development and molecular analysis of knockdown lines of B. juncea XLG genes
To study the roles of BjuXLG genes during pathogenesis of S.  (Table 2).
To minimize any variation caused due to copy-number and cosuppression of the inserted transgene, only T 0 events showing single-copy integration were selected. Basta (active ingredient glufosinate) segregation analysis in the germinated T 1 progeny (3 resistant:1 susceptible) identified single-copy events for each construct, which were propagated until the T 2 generation (Table S4).
The downregulation of BjuXLG1, BjuXLG2, and BjuXLG3 homologs in single-copy transgenic lines was analysed in 5-day-old T 2 seedlings through RT-qPCR using homolog-specific primers (Table S5).
Expression analysis showed variable levels of downregulation of different BjuXLG homologs in the selected single-copy transgenic

| Expression of defence marker genes in B. juncea XLG-RNAi lines during infection of S. sclerotiorum
To investigate whether the suppression of B. juncea XLG genes leads to expression changes in defence-marker genes, we examined the transcription abundance of the commonly used PDF1.2 (JA pathway) and PR-1 (SA pathway) genes using RT-qPCR (Table S5). During the noninfected (mock) condition, the abundance of the selected defence-marker genes in the XLG-RNAi lines was found to be comparable to the vector control plants ( Figure S4). The expression of these defence-marker genes was analysed in SSD1-infected leaf  in mock condition, which increased to 108.29 ± 6.43 µmol/g DW at an early time point (3 hpi) of pathogen infection (Table S6 and Figure 7a).
We observed that the accumulation of aliphatic glucosinolates in  (Table S6).
To investigate the effect of plant type (vector control and BjuXLG-RNAi lines) and treatments (mock and SSD1 infection) on the alteration of aliphatic, indolic, and total glucosinolates, two-way analysis of variance (ANOVA) was performed. The analysis indicated that plant type had significant impact on the levels of all glucosinolates, whereas the effect of treatment was significant for aliphatic and total glucosinolates ( Table 3). The interaction between plant type and treatment further suggested that aliphatic glucosinolates had a predominant effect during the B. juncea-SSD1 interaction.
We examined the correlation between the disease severity (lesion XLG genes leads to differential alteration of leaf glucosinolate accumulation vis-à-vis altered susceptibility against the necrotrophic pathogen S. sclerotiorum.

B. juncea are shaped by differential gene retention and distinct expression patterns
Early findings in Arabidopsis and rice favoured the opinion that the core components of heterotrimeric G-proteins signalling in plants are far less diverse than present in animals (Temple & Jones, 2007). However, the discovery of the noncanonical XLGs and type III Gγ proteins have expanded the diversity and plasticity of plant G-protein signalling to regulate multiple biological processes and environmental signals (Ding et al., 2008;Lee & Assmann, 1999;Trusov & Botella, 2016). This opinion was further changed when multiple members of G-protein subunit genes were reported in the palaeopolyploid genome of soybean Choudhary et al., 2011). Because >50% of plant species are known to be polyploids, the multiplicity of plant G-protein components is now a norm.

| Suppression of BjuXLG genes results in enhanced susceptibility to S. sclerotiorum
In Arabidopsis among the three XLG genes, XLG2 is known to play a key role in resistance against the hemibiotrophic bacterium P. sy-

| BjuXLG suppression leads to altered expression of defence marker genes
Previous studies in A. thaliana have shown that XLGs are positive regulators of resistance to biotrophic pathogens that largely trigger *p < 0.05; **p < 0.01; ***p < 0.001.

TA B L E 3
Summary of two-way analysis of variance of the impact of plant type (vector control, BjuXLG1-RNAi lines, BjuXLG2-RNAi lines, and BjuXLG3-RNAi lines) and treatment (noninoculated mock and Sclerotinia sclerotiorum-inoculated) on the levels of aliphatic, indolic, and total glucosinolates SA-responsive genes, but no such information is available about necrotrophic pathogens (Maruta et al., 2015;Zhu et al., 2009). The role of JA/ET phytohormones in providing defence against S. sclerotiorum is also documented (Chen et al., 2021;Guo & Stotz, 2007).
In B. napus, constitutive overexpression of WRKY33 was found to provide resistance against S. sclerotiorum by modulating the expression of the JA/ET marker PDF1.2 (Wang et al., 2014). To strengthen our understanding of XLG-mediated defence responses in B. juncea, the screening of defence marker genes was analysed. The highly compromised expression of PDF1.2 and WRKY33 in BjuXLG-RNAi lines during S. sclerotiorum infection ( Figure 6) in B. juncea suggests that the BjuXLGs are important G-protein signalling nodes upstream of these defence marker genes. Although the SA-mediated defence response against S. sclerotiorum has been documented earlier (Guo & Stotz, 2007), the expression pattern of the SA pathway gene PR-1 was not altered during pathogen infection in our study, thereby indicating that B. juncea predominantly uses the JA-dependent defence response against S. sclerotiorum.
WRKY33 is a pathogen-inducible transcription factor whose expression is regulated by the pathogen-responsive mitogen-activated protein kinase (MPK3/MPK6) cascade in A. thaliana (Mao et al., 2011). It is known that in Arabidopsis WRKY33 impacts the biosynthesis of the indolic phytoalexin camalexin, the major phytoalexin, which probably contributes to resistance towards S. sclerotiorum (Stotz et al., 2011). Thus, the observed reduction of WRKY33 expression in BjuXLG-RNAi lines might lead to compromised accumulation of Brassica-specific phytoalexins, which in turn lead to the observed enhanced susceptibility. Alternatively, WRKY33 can also affect other resistance mechanisms that are not dependent on specialized metabolites.

| Susceptibility to S. sclerotiorum in XLG-RNAi lines is correlated with aliphatic glucosinolate levels
The species belonging to Brassicaceae family contain glucosinolates that, along with their hydrolysis products, play important roles in plant protection against pests and pathogens (Chen et al., 2020;Hopkins et al., 2009;Kumar, Augustine, et al., 2017;Sotelo et al., 2015). Several studies suggest that both glucosinolate content and composition are altered on fungal infection (Abuyusuf et al., 2018;Robin et al., 2017). The necrotrophic fungus S. sclerotiorum causes more severe tissue damage than biotrophs, thereby making it more exposed to glucosinolates and their hydrolysis products (Kliebenstein, 2004). It has been noted that Arabidopsis mutants deficient in aliphatic or indolic glucosinolate biosynthesis are hypersusceptible to S. sclerotiorum (Chen et al., 2020;Stotz et al., 2011).
Glucosinolate levels are known to be positively correlated with oilseed rape (B. napus) resistance to S. sclerotiorum (Abuyusuf et al., 2018;Sotelo et al., 2015). An earlier study in B. juncea also highlighted that biotic elicitors and mechanical damage modulate glucosinolate accumulation (Augustine & Bisht, 2015b). In the current study, we observed that tissue damage by S. sclerotiorum infection induced glucosinolates in B. juncea, at least during the early time point of infection (Table S6). However, glucosinolate status was compromised in XLG knockdown plants (Table S6) as queries in the BRAD v. 2.0 (http://brass icadb.org/brad) and Phytozome v. 10.1 (https://phyto zome.jgi. doe.gov/pz/portal.html) databases using stringent cut-off values (E-value = 0). To study evolutionary relationships, the identified XLG protein sequences were aligned by ClustalW and the phylogenetic tree constructed using the neighbour-joining method, adopting the complete depletion option of the gaps with 500 replicated bootstrap value using MEGA 6.0 (Tamura et al., 2013). Genomic structures showing exon and intron boundaries of XLG genes were identified by gene structure display server (http://gsds.cbi.pku.edu.cn/).

| Generation of XLG knockdown transformation constructs
For developing RNAi constructs, 388, 309, and 371 bp fragments targeting the most conserved exon region of BjuXLG1, BjuXLG2, and BjuXLG3 homologs, respectively, were amplified using gene-specific primers from cDNA and cloned into the pENTR/D-TOPO vector (Table S5, Figure S3). These fragments were sequenced and mobilized into the pPZP200GW:lox-bar binary vector (Augustine et al., 2013) in both sense and antisense orientations under the control of a constitutive CaMV 35S promoter, using Gateway-based cloning. All constructs were transformed into Agrobacterium tumefaciens GV3101 using the freeze-thaw method (Nishiguchi et al., 1987)

| Brassica plant inoculation with S. sclerotiorum
Five-day-old mycelia of 4 °C preserved SSD1 fungal strain were cultured on fresh potato dextrose agar and grown for 3 days. Agar plugs (5 mm diameter) were excised from edges of growing colonies and upended onto adaxial surface of the three-to four-leaf stage B. juncea plant. High humidity (>90%) and dark conditions were maintained during the infection period. The lesion size was calculated at 24, 36, and 48 hpi after the appearance of a clear measurable lesion using the formula: where a and b represent the long and short diameters of the lesion, respectively. Every experiment was randomized and repeated at least three times with multiple technical replicates. The SSD1-infected samples were harvested from 3 to 48 hpi for expression analysis of BjuXLG homologs and defence marker genes.

| Histochemical staining and assessment fungal infection level
The growth of S. sclerotiorum in leaves of transgenic and wildtype plants was examined by trypan blue staining. The leaves were first cleared in acetic acid:ethanol solution (1:3) for 12 hr, followed by acetic acid:ethanol:glycerol (1:3:1) for 2 hr, and then stained in 0.01% trypan blue solution. Quantification of S. sclerotiorum mRNA in inoculated leaves was according to Chen et al. (2020). Total RNA was isolated from the infected leaf using the Spectrum Plant Total RNA Kit (Sigma Aldrich). For relative quantification of fungal colonization, the amount of normalized fungal mRNA was calculated by the ratio of fungal Histone gene expression to B. juncea TIP41 gene expression (Table S5).

| Glucosinolate estimation using HPLC
Leaf glucosinolates in BjuXLG-RNAi lines under mock and 3 hpi of SSD1 were determined as desulpho-glucosinolates following an established protocol (Augustine et al, 2013). Briefly, glucosinolates were extracted from 10 mg of lyophilized leaf in 1 ml of 70% methanol with sinalbin added as an internal standard (sinalbin was a kind gift from Dr Michael Rachielt, Max Plank Institute for Chemical Ecology, Jena, Germany). Desulfation of glucosinolates was performed overnight using purified sulfatase (25 mg/ml; Sigma-Aldrich) on a DEAE Sephadex-A25 column, and desulfo-glucosinolates were eluted with 1 ml of HPLC-grade water and 10 µl of eluent was analysed in a Shimadzu Nexera X2 UHPLC device (Shimadzu Corporation).
The programme was set at a solvent B (acetonitrile) gradient of 1%-19%, with respect to solvent A (water) through a 30 min cycle with flow rate of 1 ml/min using a Luna C18 reverse-phase column (150 × 4.6 mm, 0.5 mm i.d.) and glucosinolate compounds were detected at 229 nm. Concentration of glucosinolate fractions, expressed as µmoles per gram dry weight of tissue (µmol/g DW), was calculated relative to the internal standard peak and applying their relative response factors reported earlier (Brown et al., 2003). Data given are the mean of four independent measurements ± SE.

| Statistical analyses
Means of groups were compared by ANOVA Tukey's multiple comparison test, performed using SPSS statistic v. 23.0 (IBM). Jitterplot drawing and correlation analysis was performed using RStudio v.

ACK N OWLED G EM ENTS
The work was funded by the Science and Engineering Board, India (grant no. EMR/2016/006433) and a partial grant from NIPGR to N.C.B., and a fellowship to R.T. from DBT, India. The NIPGR-Plant growth facility and technical help of Vinod Kumar are acknowledged.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

AUTH O R CO NTR I B UTI O N S
R.T., J.K. and N.C.B. planned and designed the research. R.T. performed experiments, conducted field work, analysed and interpreted data. R.T. and N.C.B. wrote the manuscript and all authors edited and approved the article.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author.