Seeds of the upland cotton (G. hirsutum) resistant cultivar Zhongzhimian 2 (original strain no. GK44) were provided by the Cotton Research Institute, Chinese Academy of Agricultural Sciences (Anyang, Henan). The germinated seeds were cultured in nutrient soil and vermiculite (2:1, w/w) under a photoperiod of 16 h/8 h. After vernalization for 3 days at 4°C in Murashige-Skoog (MS) culture medium, both the wild-type (Columbia-0) and transgenic Arabidopsis plants were grown in a mixture of nutrient soil and vermiculite (1:1, w/w) under a controlled environment with 16 h light (25°C)/8-h dark (22°C). The aggressive defoliating isolate Vd991 of V. dahliae was cultured on potato dextrose agar (PDA) at 25°C for 7 days and then inoculated in Czapek liquid media. The experiments were performed using conidial suspensions of 1 × 10⁷ conidia/mL. Briefly, the cotton plants (2 weeks after VIGS) and Arabidopsis plants (4-week-old) were gently uprooted and the roots were dipped in conidial suspensions (1 × 10⁷ conidia/mL) for 5 min. Then they were replanted in a mixture of nutrient soil and vermiculite.

Total RNA was extracted using an RNA extraction kit (Biomed Gene Technology Co., Ltd., Wuhan, China), and the cDNA was synthesized using the manufacturer’s instructions with a FASTQuant cDNA RT Kit (Tiangen Biotech Co., Ltd., Beijing, China). The forward primer 5′-ACGCGTCGACATGGAAAAAAGTACTGGAAAC-3′ (SalI, underlined restriction site) and the reverse primer 5′-ATAGACTAGTTGGGCCAAGCTTGTGCAA-3′ (SpeI, underlined restriction site) were designed according to the GhPLP2 sequence (Gene ID: Gh_D02G0923) from cotton protein databases (G. hirsutum, NAU assembly) in the Cotton Functional Genomics Database¹ to amplify the open reading frame (ORF) using the cDNA template. The PCR product was transferred into the pMD18-T vector according to the manufacturer’s instructions (TaKaRa, Dalian, China), and the positive clone was sequenced.

The theoretical isoelectric point (pI) and molecular mass were calculated with ProtParam.² All the sequences were retrieved from the website of the National Center for Biotechnology Information database,³ and the accession numbers of the patatin proteins are listed in Supplementary Table 1. Multiple amino acid sequence alignments were performed using Clustal Omega,⁴ and the multiple alignment file was shaded with BoxShade.⁵ A phylogenetic tree was constructed with the neighbor-joining method using MEGA 7 with bootstrap values from 1,000 replicates indicated at the nodes, and the motifs were annotated using MEME⁶ (Bailey et al., 2009; Kumar et al., 2016). The homology model of GhPLP2 was generated using SWISS-MODEL, and three-dimensional models were analyzed and visualized using EzMol, Version 1.20 (Bordoli et al., 2009; Reynolds et al., 2018). The crystal structure of SeMet Patatin (1oxw.1.B) was selected as a template to predict the theoretical model.

For abiotic stresses and hormone treatments, 2-week-old cotton seedlings were gently uprooted and the roots were immersed in Hoagland solution that contained 100 μM methyl jasmonate (Sigma-Aldrich, St. Louis, MO, United States), 2 mM ethylene (Sigma-Aldrich), and 2.5% (w/v) PEG 6000, respectively (Champion et al., 2009; Pei et al., 2019). For the treatments with methyl jasmonate and ethylene, the whole plant was collected at 0, 0.5, 3, 6, 9, 12, 24, and 30 h for RNA extraction. For PEG 6000 treatment, the whole plant was collected at 0, 1, 3, 6, 12, 24, and 48 h for RNA extraction. For V. dahliae treatment, seedling roots were inoculated with conidial suspensions for 5 min and then transplanted into sterile soil. Control samples were treated with sterile water. The whole plant was collected at 0, 0.5, 12 h, 1, 3, 5, and 7 days for RNA extraction.

GhPLP2 was cloned into the 6 × His-Tagged protein expression vector pET-22b (+) (Novagen, Madison, WI, United States). The primers are listed in Supplementary Table 2. The recombinant vector was expressed in Escherichia coli BL21 (DE3). Single colonies were cultured at 37°C in LB broth that contained 100 μg/mL ampicillin. When the OD₆₀₀ reached 0.6, the culture was induced with 0.4 mM IPTG for another 12 h at 22°C at 200 rpm. The protein was purified using a 6 × His-Tagged Protein Purification Kit (CWbio, Beijing, China) following the manufacturer’s instructions and detected using SDS-PAGE. The concentration of purified protein was determined using a Bradford Protein Assay Kit (TaKaRa, Dalian, China).

The LAH activity of the GhPLP2 protein was assayed using the artificial substrate p-nitrophenyl palmitate (p-NPP) and phospholipids (Hong et al., 2008). Briefly, the reaction mixture was composed of 200 μL of Tris-HCl (pH 8.0), 10 mM CaCl₂, 0.05% (v/v) Triton X-100, and 2 mM p-nitrophenyl palmitate to which 25 μg of purified protein was added and incubated at 37°C for 30 min. A volume of 700 μL of absolute ethanol was added and centrifuged for 2 min at 7,000 rpm. The absorbance of p-nitrophenol (NP) liberated from p-NPP was measured at 405 nm.

The substrates for phospholipase assay included 1,2-distearoyl-sn-glycerol-3-phosphoethanolamine (PE), 1,2-distearoyl-sn-glycerol-3-phosphocholine (PC), or L-α-phosphatidylglycerol (PG) (Aladdin, Shanghai, China). The reaction mixture was composed of 50 mM Tris-HCl (pH 8.0), 10 mM CaCl₂, 0.05% (v/v) Triton X-100, 500 μg substrate, and 10 μg purified protein in a final volume of 600 μL. The reactions were performed at 37°C for 1 h and terminated by the addition of 500 μL chloroform: methanol (2:1, v/v). The fatty acids products released from the substrates were separated and analyzed as previously described (La Camera et al., 2005). Individual fatty acid methyl esters (FAME) were quantified using gas chromatography (GC) equipped with an Agilent column (Agilent Technologies, Santa Clara, CA, United States) (30 m by 0.25 mm, 0.25 μm film) and a flame ionization detector (FID).

The PCR product with a SalI/SpeI restriction site was inserted into a modified pCAMBIA 1300 vector that harbored hygromycin phosphor-transferase (hptII) and green fluorescent protein (GFP) under the control of the CaMV 35S promoter (Supplementary Figure 1; Wang et al., 2011). By particle bombardment (PDS-1000; Bio-Rad, United States), the CaMV 35S:GhPLP2-GFP and 35S:GFP empty vector were transformed into onion epidermal cells. The transformed cells were incubated on the MS medium (with 20 g/L sucrose) in the dark for 18–24 h at 22°C.

For the generation of GhPLP2-transgenic plants, the construct was introduced into Agrobacterium tumefaciens strain GV3101 by the freeze-thaw method (Jyothishwaran et al., 2007), and Arabidopsis was transformed by the floral dip method (Clough and Bent, 1998). Transgenic Arabidopsis seeds were screened on MS plates (with 20 g/L sucrose) that contained 25 mg/mL hygromycin B and T3 homozygous lines that displayed 100% hygromycin resistance were used for additional experiments.

The virus-induced gene silencing (VIGS) transient expression methods were previously described by Gao et al. (2011). The silenced fragments of GhCLA1 (Cloroplastos alterados 1) and GhPLP2 were amplified from cotton cDNA and inserted into the TRV:00 vector to generate the TRV:GhCLA1 and TRV:GhPLP2 vectors. Then plasmids of TRV:GhCLA1 and TRV:GhPLP2 were transformed into A. tumefaciens strain GV3101 by heat shock (Dong et al., 2007). TRV:GhCLA1 plants were used as positive controls. Two weeks after infiltration, when the GhCLA1-silenced plants showed clear signs of albinism in the leaves, the efficiency of GhPLP2 was evaluated by quantitative real-time reverse transcriptase–PCR (qRT-PCR). GhUBQ7 (DQ116441) was amplified as an internal control with the primers qUBQ7-F/R. All the primers are listed in Supplementary Table 2.

Two weeks after VIGS, when the true leaves of GhCLA1-silenced cotton plants showed clear signs of albinism, the plants were inoculated with V. dahliae (1 × 10⁷ conidia/mL) as previously described (Liu et al., 2018). The plant disease index (DI) was monitored using the following formula as described by Wang et al. (2020): DI = [Σ (n × the number of seedlings at level n)/(4 × the number of total seedlings)] × 100, n denotes disease level, cotton seedlings were divided into five levels based on their disease severity after inoculation with V. dahliae (level 0, 1, 2, 3, 4). For Arabidopsis plants, 4-week-old GhPLP2-transgenic and wild-type (WT) Arabidopsis plants were inoculated with V. dahliae spores as previously described (Gao et al., 2013). The controls were dipped in sterilized water. The disease index and classification of symptoms were determined as previously reported (Pei et al., 2019). Data were collected from three independent technical replicates (n ≥ 30).

The V. dahliae recovery assay was conducted and modified as described by Fradin et al. (2009), and the antibiotic in the PDA is kanamycin in this study. The stem section above the cotyledons was taken from cotton plants 3 weeks after inoculation with V. dahliae. Six slices were sterilized with 75% ethanol and 10% sodium hypochlorite solution and transferred onto PDA supplemented with kanamycin (50 mg/L) after the slices were washed three times with sterile water. The V. dahliae biomass was quantified as previously described (Ellendorff et al., 2009). Inoculated cotton (first internode) and Arabidopsis plants (whole plants) were harvested at 21 days post-inoculation (dpi) and 14 dpi, respectively. DNA was extracted from 100 mg of the fine powder. V. dahliae biomass was determined by qRT-PCR using fungus-specific ITS1-F (Liguori et al., 2007) and V. dahliae-specific ST-Ve1-R primers (Fradin et al., 2011). GhUBQ7-F/R (DQ116441) was used as a reference gene in the cotton plants, and AtEF1α-F/R (AT5G60390) was used as a reference gene in the Arabidopsis plants. All the primers are listed in Supplementary Table 2.

The primers were designed to obtain the PevD1 fragment from V. dahliae (Supplementary Table 2), which was then cloned into the pET-28a (+) vector. The recombinant vector was expressed in E. coli BL21 (DE3), and the protein was induced with 0.1 mM IPTG for 10 h at 22°C while shaken at 200 rpm. The protein was purified using a 6 × His-Tagged Protein Purification Kit (CWbio, Beijing, China) following the manufacturer’s instructions and detected using SDS-PAGE. The concentration of purified protein was determined using a Bradford Protein Assay Kit (TaKaRa, Dalian, China). The treatment with PevD1 elicitor was performed as described by Li et al. (2019). The expression of HR marker genes GhHSR203 and GhHIN1 triggered by PevD1were detected in cotton leaves at 24 h after infection. Callose deposition was stained using 0.1% aniline blue as previously described (Wang et al., 2018). Images were observed using FLUOVIEW FV1000 on a confocal laser scanning microscope (Olympus, Tokyo, Japan). The level of H₂O₂ was measured by freshly made FOX reagent (Fe[NH₄]₂∙[SO₄]₂) (250 mM), xylenol orange (100 μM), sorbitol (100 μM), and H₂SO₄ (25 mM), and 1% ethanol as described by Pei et al. (2019).

Four-week-old Arabidopsis and 2-week GhPLP2-silenced cotton plants after VIGS were used to analyze the composition of fatty acids and quantify the hormones. The hormones were quantified as previously described (Li D. et al., 2020). Fresh samples were ground into fine powder in liquid nitrogen. A total of 100 mg powder was extracted with 1 mL of 80% aqueous methanol that contained 1% formic acid using an ultrasonic treatment at 4°C for 10 min and centrifuged at 12,000 rpm at 4°C for 10 min. The supernatant was transferred into a 2 mL centrifuge tube that contained 50 mg of primary secondary amine (CNW Technologies GmbH, Dusseldorf, Germany), dried with nitrogen, and added to 100 μL of 80% aqueous methanol that contained 1% formic acid. The samples were analyzed by an Agilent 6410B Triple Quadrupole HPLC-MS/MS (Agilent Technologies), which is equipped with an HPLC reverse phase C18 column (Athena C18-WP 2.1 × 50 mm, 3 μm). The flow rate was 0.3 mL/min. Methanol and ultrapure water were used as the mobile phases A and B, respectively. Isocratic elution was conducted for 5 min at a 65(A):35(B) ratio. MS was performed using the multiple reaction monitoring modes and negative electrospray ionization. All the parameters were performed as described by Li D. et al. (2020).

The fatty acid composition was analyzed by gas chromatography (GC) on Agilent 7890A (Agilent Technologies, Santa Clara, CA, United States) with a flame ionization detector (FID). Fatty acid methyl esterification was performed as described by Liu et al. (2019) with slight modification. The plant samples were dried at 90°C to a constant weight, and 50 mg of dry powder was weighed. Free fatty acids were methylated in 1 mL 0.5 mol/L KOH-methanol at 60°C for 2 h. Next, 1 mL of hexane that contained 0.01% butylated hydroxytoluene (BHT) and 0.1 mg methyl non-adecanoate was added. After shaking and standing for stratification, the supernatant that contained the FAMEs was separated. The GC parameters were as follows: 180°C for 10 min followed by a ramp to 190°C at 1°C/min, a hold at 190°C for 3 min, and then heating up to 220°C with a gradient at 4°C/min. The final temperature was maintained for 3 min.

Arabidopsis drought treatment was carried out as previously described (Wang et al., 2018). Briefly, water was withheld from 3-week-old Arabidopsis plants for 16 days, followed by watering for 2 days from the bottom. All the Arabidopsis plants were grown in a controlled environment with 16 h light (25°C)/8-h dark (22°C) and a relative humidity at 80%.

Total RNA was obtained from the plant samples with an RNA extraction kit with DNase I (Biomed Gene Technology, Co., Ltd., Beijing, China). The cDNA was synthesized using a PrimeScript^TM RT reagent Kit with gDNA Eraser (Perfect Real Time) (TaKaRa Bio, Dalian, China). qRT-PCR was performed using SYBR Premix Ex Taq (Tli RNaseH Plus) (TaKaRa, Dalian, China) on an ABI 7500 thermocycler (Applied Biosystems, Foster City, CA, United States). AtEF1a (AT5G60390) and GhUBQ7 (DQ116441) were utilized as internal standard genes in Arabidopsis and cotton, respectively. The relative gene expression was calculated using the 2^−ΔΔCt method (Livak and Schmittgen, 2001). The melting curves for each gene studied (with a no template control) are provided in Supplementary Figures 2, 3. The primer sequences are listed in Supplementary Table 2.

The data were obtained from three independent technical or biological replicates per treatment and presented as the mean ± standard error. The statistical analyses were performed by Student’s t-test. Asterisks indicates a significant difference between the groups (^∗P < 0.05, ^∗∗P < 0.01).

From the transcriptome analysis of the cotton cultivar Zhongzhimian 2 that is resistant to infection by V. dahliae Vd991, the PLP gene GhPLP2 was highly induced when the plants were infected with V. dahliae. We isolated the full-length clone of GhPLP2 that consists of a 68 bp 5′untranslated region (5′ UTR), 138 bp 3′ UTR, and 1,218 bp open reading frame (ORF) that encoded a 405 amino acid protein with a theoretical pI of 8.21 and a molecular weight of 44.44 kDa. The GhPLP2 protein contains the conserved serine (S) hydrolase motif GXSXG at residues 63–67 and a conserved aspartic acid (D) residue at 214 within the patatin domain (residues 21–221) (Supplementary Figure 4). The catalytic dyad of S and D residues is critical for the LAH activity of patatin (Rydel et al., 2003; Rietz et al., 2010). Using SeMet patatin, an isozyme of patatin 17 from Solanum cardiophyllum as the template, the homology modeling results indicate that the Ser-Asp catalytic dyad is essential for the LAH activity of GhPLP2 (Supplementary Figure 5). GhPLP2 lacks the transmembrane domain or signal peptide. To study the possible role of GhPLP2 protein in cotton plants, a phylogenetic analysis was generated with homologous patatin proteins with precise functions that had been previously reported from other species. The phylogenetic analysis showed that plant patatins were divided into three groups, and GhPLP2 is located in group II that contains NtPat1 (Cacas et al., 2005), GhPat1 (Cacas et al., 2009), CaPLP1 (Kim et al., 2014), and AtPLP2 (La Camera et al., 2009; Figure 1A). The patatin homologs were divided into three groups based on the conserved motif alignments, which are consistent with the phylogenetic analysis result (Figure 1B). Group II members have the canonical S-D dyad esterase motif constituted by GxSxG and DGG/A in patatin catalytic centers, a conserved anion binding element DGGGxxG and the proline motif APP. GhPLP2 is a patatin group II member with the conserved catalytic dyad residues.

The LAH activity of patatin proteins can hydrolyze membrane lipids into free fatty acids and lysophosphatidic acid, which lead to a series of signaling pathways involved in growth development and defense responses (Ryu, 2004). To understand the role of GhPLP2 in more detail, the LAH activity of the recombinant GhPLP2 protein was studied. The recombinant protein GhPLP2 was produced and purified from E. coli BL21 (DE3). The GhPLP2 protein was induced with 0.4 mM IPTG at 22°C and as early as 1 h after induction (Figure 2A). The GhPLP2 fusion protein was purified using Ni columns, and the elution fractions were confirmed by SDS-PAGE (Figure 2B). When p-NPP was used as the substrate, the His-GhPLP2 protein exhibited high LAH activity (97 nmol min^–1 mg^–1), and the vector control was totally deficient in enzyme activity (Figure 2C). In addition, when the phospholipids PC, PE, and PG were used as substrates, the release of free fatty acids was quantified. The purified GhPLP2 protein released fatty acids from these phospholipids and exhibited variant enzyme activities in different phospholipid substrates (Figure 2D). The results indicate that GhPLP2 may function as a LAH to release free fatty acids in plants.

To determine the subcellular localization of GhPLP2 protein, the GFP signals of GhPLP2-GFP fusion protein and 35S:GFP empty vector were monitored by Nikon A1R SI Confocal (Nikon, Japan). The GFP signal of 35S:GFP empty vector was diffused throughout the cell (Figures 3A–C), whereas the GhPLP2-GFP fusion protein was found to accumulate exclusively in the cell wall or plasma membrane (Figures 3D–F). To further clarify the location of GhPLP2, the onion epidermal cells were treated with 0.8 M mannitol for 10 min. After plasmolysis, the GFP signals of GhPLP2-GFP fusion protein was found in the plasma membrane (Figures 3G–I). These results revealed that GhPLP2 was localized to the plasma membrane.

Expression of the GhPLP2 gene in cotton was examined by qRT-PCR. GhPLP2 was the most abundant in cotton roots (Figure 4A), which are the first physical barrier in cotton plants against infection by V. dahliae. The change in expression of GhPLP2 in response to various stresses was investigated. The expression of GhPLP2 was significantly increased by 9.4- and 7.1-fold at 0.5 h and 7 days after inoculation with V. dahliae, respectively (Figure 4B). We investigated the expression of GhPLP2 after treatment with the defense-related signaling molecules methyl jasmonate and ethylene in more detail. The expression of GhPLP2 was significantly higher at 3 h by 6.2-fold, and another 6.0-fold peak appeared at 24 h after treatment with JA (Figure 4C). In contrast, ethylene increased the expression of GhPLP2 at 0.5 h, with a maximum level of 4.7-fold observed at 3 h (Figure 4C). The treatment with 2.5% PEG 6000 (w/v) markedly increased the gene expression at 1 and 3 h with 5.6- and 3.7-fold, respectively (Figure 4D).

The GhPLP2 gene was silenced by VIGS in cotton plants to clarify its function by Gao et al. (2013). The cotton gene GhCLA1, which is involved in chloroplast development, was used as a positive control (Gao et al., 2011). To assess its gene silencing efficiency, the expression of GhPLP2 was monitored by semi-quantitative PCR and qRT-PCR in TRV:00 and TRV:GhPLP2 cotton true leaves after 2 weeks of VIGS (Figure 5G and Supplementary Figure 6). The results indicated that the expression of GhPLP2 was effectively reduced in GhPLP2-silenced cotton plants.

To analyze the involvement of GhPLP2 in the interaction between cotton and V. dahliae, cotton plants were inoculated with this pathogen. After inoculation, typical disease symptoms appeared at 10 days in GhPLP2-silenced cotton. At 14 days, leaf chlorosis and necrosis were more severe in the GhPLP2-silenced plants (Figures 5A,B). Dark and necrotic vascular bundles of the dissected stems were more apparent in GhPLP2-silenced plants at 21 days (Figures 5C,D). The fungal renewal cultivation showed that more fungi were recovered from GhPLP2-silenced plants compared with the control (Figures 5E,F). The plant disease index of the silenced plants was higher than that of the control plants (Figure 5H). In addition, the fungal biomass of the stems from the GhPLP2-silenced plants were higher than that in the control plants determined by qRT-PCR analysis (Figure 5I). These results suggest that the silencing of GhPLP2 attenuates the resistance of cotton to V. dahliae.

The HR is an efficient and immediate resistance reaction during the attack of pathogens, which rapidly causes cell death around the attempted infection site that restricts the spread of pathogens (Vailleau et al., 2002). During the HR, plant immune responses, including the accumulation of H₂O₂, deposition of callose, and the expression of HR-related marker genes were rapidly induced (Hamada et al., 2005; Simon et al., 2010; Rossi et al., 2011). PLPs have been shown to participate in HR in response to infection by the Gram-negative bacterial pathogen Xanthomonas campestris pv. vesicatoria in pepper plants and the response to Tobacco mosaic virus (TMV) in tobacco plants (Dhondt et al., 2000; Choi and Hwang, 2011; Kim et al., 2014). Therefore, GhPLP2 may regulate the resistance to V. dahliae by mediating an HR induced by V. dahliae elicitors. To test our hypothesis, we expressed and purified PevD1 from V. dahliae (Supplementary Figure 7), which triggered HR and resistance responses in cotton, Arabidopsis, and tobacco (Wang et al., 2012; Bu et al., 2014; Liu et al., 2016). The HR-like cell death was initiated at 24 h after injection with PevD1, while it was significantly reduced in GhPLP2-silenced leaves (Figure 6A). Moreover, PevD1 triggered the deposition of callose, and the expression of HR marker genes GhHIN1 and GhHSR203 were compromised in the GhPLP2-silenced leaves at 24 h after the infiltration (Figures 6B–D). In addition, the silencing of GhPLP2 significantly compromised the production of H₂O₂ induced by PevD1 (Figure 6E).

To further evaluate the role of GhPLP2 in response to V. dahliae, we examined the resistance of wild-type (WT) and transgenic Arabidopsis seedlings to V. dahliae infection. Three homozygous transgenic (T3 generation) lines with the highest levels of expression of GhPLP2 (Lines 1, 4, and 8) were selected for additional experiments (Supplementary Figure 8).

Four-week-old transgenic and WT plants were infected with V. dahliae spores by the root dipping method (Pei et al., 2019). After inoculation, the WT plants showed more serious wilting, yellowish coloring and necrosis compared with the transgenic plants at 14 dpi (Figure 7A); this was consistent with the disease index investigation (Figure 7B). The fungal biomass of the WT plants was remarkably high compared with that on the transgenic plants as determined by qRT-PCR analysis (Figure 7C). These results indicate that the overexpression of GhPLP2 in Arabidopsis plants confers enhanced resistance to V. dahliae.

Free fatty acids (FFAs) are among the hydrolyzates of membrane lipids owing to PLPs with LAH activity (Rivas and Heitz, 2014). The endogenous LAH activity of crude proteins from different genotypes plants was assayed using p-NPP as the substrate (Lin et al., 2011). The LAH activity of GhPLP2-overexpressed Arabidopsis lines was higher than that of the WT plants, whereas the LAH activity decreased in TRV:GhPLP2 cotton plants compared with the TRV:00 plants (Supplementary Figure 9). To further assess the metabolic differences between different genotypes of plants, the composition of fatty acids was examined in Arabidopsis and cotton plants 3 days after the inoculation of V. dahliae and water (mock) (Figure 8). As shown in Figure 8A, the analyses of the composition of fatty acids indicated that the levels of linoleic acid (LA, 18:2) and α-linolenic acid (ALA, 18:3) were higher in GhPLP2-overexpressed Arabidopsis plants than in the WT plants under treatment with water. After inoculation with V. dahliae, the content of different types of fatty acids increased in both genotypes of Arabidopsis plants, and the increase in GhPLP2-overexpressed Arabidopsis plants was more apparent compared with the WT (Figure 8A). Moreover, GhPLP2-silenced cotton plants exhibited a significant reduction in their contents of LA and ALA regardless of the treatment with V. dahliae (Figure 8B). In addition, the content of palmitic acid (16:0) and stearic acid (18:0) also increased significantly after the inoculation of cotton plants with V. dahliae.

It has been reported that the oxidation and further conversion of C18 polyunsaturated FAs with lipoxygenases (LOXs) may generate the production of JA (Walley et al., 2013; Sun et al., 2014; Fu et al., 2015). We then investigated the JA contents in GhPLP2-overexpressed Arabidopsis and GhPLP2-silenced cotton plants. The contents of JA in GhPLP2-overexpressed Arabidopsis lines 4 and 8 were 84.97 and 101.99% higher than that of the WT, respectively, when they had not been inoculated with V. dahliae. After inoculation with V. dahliae, the JA contents of GhPLP2-overexpressed Arabidopsis lines 4 and 8 were 51.63 and 74.42% higher than that of the WT, respectively (Figure 8C). The levels of JA in the TRV:GhPLP2 cotton plants were 41.21 and 33.83% lower than that of the TRV:00 cotton plants under mock and V. dahliae inoculations, respectively (Figure 8D).

We also conducted qRT-PCR to detect the expression of a series of the JA-synthesis and -signaling genes from Arabidopsis and cotton plants, namely AOS, LOX2, OPR3, and PDF1.2. All of these genes were highly induced by infection with V. dahliae in both Arabidopsis and cotton at 3 dpi (Figures 8E,F). Consistent with the contents of JA, we found that the expression of these genes was significantly upregulated in GhPLP2-overexpressed Arabidopsis plants compared with the WT and downregulated in TRV:GhPLP2 cotton plants relative to the TRV:00 plants in both treatments (Figures 8E,F). Taken together, the results showed that GhPLP2 contributes to the accumulation of fatty acids and is involved in the JA signaling pathway in both Arabidopsis and cotton plants.

Based on our observation that PEG 6000 induced the expression of GhPLP2, we investigated the drought tolerance of GhPLP2-overexpressed Arabidopsis plants. However, the GhPLP2-overexpressed Arabidopsis did not show the distinct phenotype of drought sensitivity compared with the WT plants (Supplementary Figure 10).