The genomes of G. raimondii, G. arboreum, G. hirsutum and G. barbadense were downloaded from CottonGen (https://www.cottongen.org/). Blastp (E-value ≤ 10^-10) was carried out in cotton genome using GhWRKY53 encoded amino acid sequence as a probe to obtain WRKY53 protein sequence in each of the four cotton species. Multiple sequence alignment of GhWRKY53 was performed with DNAMAN (Kumar et al., 2016). Conserved motifs of the WRKY53 protein sequence were predicted with the use of online software, MEME (https://meme-suite.org/). RT-qPCR was used to analyze the gene expression level of GhWRKY53 in the roots, stems, and leaves of resistant and susceptible cotton varieties.

Tobacco rattle virus (TRV) derived vectors, pTRV1 and pTRV2, were used for VIGS (Liu et al., 2002). The VIGS vector construction and experimental procedure were carried out according to the protocol previously described (Lu et al., 2021). The fragment targeting GhWRKY53 was amplified using the forward primer of 5’-GAATTCGGGCAAAAAGACATCCTGGG-3’ and the reverse primer 5’-GGTACCGAAAGAAGTTGCCATCTCGGT-3’ (the underlined nucleotides in the forward and reverse primers represent the restriction sites of Eco R1 and Kpn I, respectively). The PCR cycles and confirmation of the fragment were done according to Lu et al. (2021). The cotton variety used in VIGS was Upland cotton ZZM2 and XLZ7. The method of V. dahliae infection of cotton was done according to Li et al. (2023). All primers used for the VIGS were designed with Primer3 software and are shown in Table S1 .

For the SA and MeJA treatment experiments, two-leaf stage seedlings of cotton were sprayed with 1mmolL⁻¹ SA or 100 µmol L⁻¹ MeJA (Xiong et al., 2019). Patients treated for 0 hours were selected as the control group. Collect root samples of cotton seedlings at 0, 6, 12, 24, and 48 hours after treatment. Roots of TRV::00 and TRV::GhWRKY53 inoculated with V. dahliae for 0 and 48 h were selected to determine the endogenous levels of SA and MEJA. Moreover, the roots from five individual seedlings of cottons at two-leaf stage were collected to measure the expression profiles of GhPAL, GhPR1, GhAOS1, GhPDF1.2 and GhLOX after V.dahliae infection in TRV::00 and TRV::GhWRKY53 plants by qRT-PCR. The differences between groups were compared using Student’s t-test (* P < 0.05; ** P < 0.01).

Total RNA was extracted from leaves of TRV1/TRV2::00 or TRV1/TRV2::GhWRKY53 treated ZZM2 and XLZ7 plants (both control and V.dahliae treated), and then reverse transcribed into cDNA to be used in qRT-PCR to analyze the effect of inoculation with V.dahliae treatment on gene expression changes. qRT-PCR was carried out by the SYBR Green (Roche, Rotkreuz, Switzerland) on a Light Cycler 480II (Roche, Germany) with default parameters. All primers used for the validation experiments were designed with Primer3 software and are shown in Table S1 . The GhUBQ7 gene served as an internal control to normalize differences between samples. The relative expression levels of genes from three biologically independent experiments were calculated using the 2^-ΔΔCTmethod (Livak and Schmittgen, 2001).

SPSS 26.0 (SPSS, Chicago, USA) was used for data processing and analysis of variance. The data were analyzed by a one-way analysis of variance (ANOVA) and the significance of the difference was tested using the Duncan multiple comparison method. Origin 2022 (OriginLab, Northampton, USA) was used to plot the figures.

The 2, 2, 4, and 4 WRKY53 genes were identified in G. arboreum, G. raimondii, G. barbadense, and G. hirsutum, respectively. Further, the conserved domains of GhWRKY53 were demonstrated by multiple protein sequence alignments of GhWRKY53. Almost all proteins contained the WRKY domain and zinc finger structure in the form of CX7C23HXC ( Figure 1A ). Based on the results of a motif analysis of GhWRKY53 proteins using MEME suite, a total of ten conserved motifs were identified. GrWRKY53-2, GbWRKY53-2D and GhWRKY53 contained nine motifs, while the others contained ten motifs ( Figure 1B ). Additionally, the promoter region of WRKY53 in each cotton species contained at least two plant hormone cis-acting elements ( Figure 1C ). Of them, an ETH cis-responsive element was found in the promoter region of each WRKY53 gene. And a MeJA cis-responsive element and an SA were found in four WRKY53 (GaWRKY53-2, GbWRKY53-2A, GbWRKY53-1D and GbWRKY53-2D). The results also showed that the GaWRKY53-1, GhWRKY53-1A and GhWRKY53-1D exclude MeJA and SA cis-responsive element ( Figure 1C ). The above results indicated that cotton WRKY53 was regulated by a variety of hormones and was involved in different types of hormone response.

qRT-PCR analysis showed that GhWRKY53 was expressed in roots, stems and leaves of the ZZM2 and XLZ7, especially in the roots and stems. And its expression levels in leaves were relatively low. For XLZ7, a susceptible variety, GhWRKY53 expression levels in the stems were significantly higher than roots. However, GhWRKY53 expression levels were higher in the roots of disease-resistant variety ZZM2 ( Figure 2 ). To further analyze the function of GhWRKY53 in cotton verticillium wilt resistance, the two-leaf stage seeding was inoculated with V. dahliae by the root irrigation method (Xiong et al., 2019). The results showed that GhWRKY53 expression levels were significantly up-regulated in the root and stem tissues of ZZM2 after 12 h of V. dahliae inoculated ( Figures 3A, B ). The GhWRKY53 showed significantly up-regulated after 24 h of V. dahliae inoculated in the root and stem of XLZ7 ( Figures 3A, B ). However, the up-regulated of GhWRKY53 expression levels of XLZ7 in roots and stems was not significant than ZZM2. In the resistant variety ZZM2, GhWRKY53 had higher expression levels in the root and was more sensitive to the stress response of V. dahliae. It also indicated that GhWRKY53 may play an important role in cotton resistance to verticillium wilt.

The upstream promoter sequence of GhWRKY53 contains cis-acting elements of SA and MeJA, suggesting that GhWRKY53 may be involved in disease resistance related to hormone response. Cotton seedlings treated with different hormones were sampled to detect the changes in GhWRKY53 expression levels in the roots of the ZZM2 and XLZ7 after SA and MeJA treatment. The results show that the GhWRKY53 expression levels were significantly up-regulated after 6, 12, and 24 h after SA treatment in ZZM2 ( Figure 3C ). However, the expression levels of GhWRKY53 in sensitive variety XLZ7 showed a slow upward trend from 6 to 24 h after SA treatment ( Figure 3C ). After MeJA treatment, GhWRKY53 expression levels in both XLZ7 and ZZM2 showed a trend of first downregulation and then upregulation, and GhWRKY53 expression levels in both cotton varieties were significantly up-regulated at 48 h ( Figure 3D ). These results indicated that GhWRKY53 was specifically regulated by SA and MeJA in different verticillium wilt resistant varieties.

To study the function of GhWRKY53 on the cotton resistance to V. dahliae, VIGS was used to silence GhWRKY53. Seedlings infiltrated with A. tumefaciens carrying pTRV1/pTRV2::GhCHLI (plants-TRV:: GhCHLI) that targeted a gene encoding a ChlI subunit of magnesium chelatase were used as a positive control. After about ten days of injection, plants-TRV::GhCHLI showed a yellowing phenotype, indicating that VIGS system was functioning properly ( Figures 4A , 5A ). Compared to the control plant (plants-TRV2::00), GhWRKY53 expression levels in ZZM2 and XLZ7 respectively infiltrated with A. tumefaciens carrying pTRV1/pTRV2::GhWRKY53, decreased significantly, indicating that GhWRKY53 was effectively silenced in XLZ7 and ZZM2 ( Figures 4B , 5B ). And then cotton seedlings were inoculated with V. dahliae to verify plants resistance to V. dahliae after silencing GhWRKY53. At 14 days after inoculation with V.dahliae, GhWRKY53-silenced plants were more sensitive to V. dahliae infection than control plants in XLZ7. XLZ7-pTRV2::GhWRKY53 also showed severe leaf yellowing and wilting, with more shedding of leaves ( Figure 4C ). The longitudinal section of stems after 21 d pathogen infection showed that the browning degree of the stems in the XLZ7-pTRV2::GhWRKY53 was more significant ( Figure 4D ). Additionally, the pathogen colonies isolated from the stems of XLZ7-pTRV2::GhWRKY53 were larger and darker than that of XLZ7 ( Figure 4E ). At 14 d and 21 d, the disease index of XLZ7-pTRV2::GhWRKY53 plants were 47 ± 2.04 and 85.67 ± 3.70, respectively, significantly higher than that of TRV::00 plants at each time point ( Figure 4F ). Meanwhile, the relative contents of pathogen in the stem tissues and the recovery of pathogen were detected. The results showed that the biomass of pathogen in the stems of XLZ7-pTRV2::GhWRKY53 was significantly higher than that of control plants, and its biomass was about three times that of XLZ7( Figure 4G ). These demonstrated that silencing GhWRKY53 further reduces the resistance of susceptible cotton variety XLZ7 to V. dahliae. In ZZM2, after 14 d inoculation, GhWRKY53-silenced plants (ZZM2-pTRV2::GhWRKY53) exhibited poor growth with severe disease and yellowing of leaves compared with control plants ZZM2. Compared with the control plant, leaves of ZZM2-pTRV2::GhWRKY53 had significant withering ( Figure 5C ). The longitudinal section of stem showed that the accumulation of pathogen in ZZM2-pTRV2::GhWRKY53 with severe browning was more than that in control plants, and the stem browning was severe ( Figure 5D ). The results of the pathogen recovery experiment showed that the colonies isolated from the stems of ZZM2-pTRV2::GhWRKY53 were larger and had a denser villous shape than that of control plants ( Figure 5E ). The disease index of the ZZM2-pTRV2::GhWRKY53 on the 14 d and 21d was 13.17 ± 2.02 and 47.17 ± 3.01 respectively, and far higher than the control plant( Figure 5F ). Compared to the control plants, the stems of ZZM2-pTRV2::GhWRKY53 accumulated more pathogen, and its biomass was about five times that of XLZ7 ( Figure 5G ). These were similar to the results observed in XLZ7, indicating that the silencing of GhWRKY53 also significantly reduced the resistance of cotton to V. dahliae.

To explore whether the deterioration of V. dahliae resistance in GhWRKY53-silenced plants is mediated by JA and SA signaling pathways, we measured JA and SA concentrations in the roots of the GhWRKY53-silenced plants and control plants, and the expression of signaling pathway-related genes also were analyzed. The results showed that SA content was up-regulated in both the control and silenced plants after inoculation with V. dahliae for 48 h in ZZM2 and XLZ7. For XLZ7, at 0 h after V. dahliae infected, the content of SA in GhWRKY53-silenced plants and control plants were 52.16 and 113 15 ng/g, respectively. And at 48 h, the content of SA increased to 122.51 and 130.67 ng/g, respectively ( Figure 6A ). Simultaneously, the content of JA in TRV::GhWRKY53 and TRV::00 were 8.05 and 3.46 ng/g at o h, respectively. And increased to 11.75 and 9.59 ng/g at 48 h, respectively ( Figure 6B ). However, it was noted that the SA content in the GhWRKY53-silenced plants of XLZ7 was lower compared to the control plants at 0 h after V. dahliae inoculation ( Figure 6A ). Consistent with this result, SA biosynthesis gene (GhPAL) and response gene (GhPRI) were significantly up-regulated at 48 h ( Figures 6C, D ). And JA biosynthesis gene (GhLOX) and JA response gene (GhPDF1.2) also followed a similar trend ( Figures 6E, F ). For ZZM2, the SA content of TRV::GhWRKY53 (68.92 ng/g) was significantly lower than TRV::00 (120.32 ng/g) at 48 h post-inoculation (hpi) ( Figure 7A ), with a similar trend being seen in the expression levels of SA biosynthesis gene (GhPAL) and response gene (GhPRI) ( Figures 7C, D ). The content of JA in TRV::GhWRKY53 plants and TRV::00 plants at 48 hpi increased 9.43 and 2.67 ng/g than 0 hpi after V. dahliae inoculation ( Figure 7B ), while the expression levels of JA biosynthesis genes (GhAOS and GhLOX) were also significantly induced at 48 hpi ( Figures 7E, F ). These results were support the notion that GhWRKY53 plays a positive role in V. dahliae resistance by activating JA signaling pathway and repressing SA signaling pathway.