All E. grandis genome sequences and annotation were obtained from the NCBI nucleotide sequence repository (https://www.ncbi.nlm.nih.gov/datasets/taxonomy/71139/). The Pfam database (http://pfam.sanger.ac.uk/) was used to obtain an HMM file of predicted WRKY structural domain (PF03106). A basic HMM search using TBtools v2.086 was conducted to identify WRKY family proteins in the E. grandis protein database with an e-value cutoff of 1 × 10⁻⁵ (Chen et al., 2023). The NCBI CDD (http://www.ncbi.nlm.nih.gov/cdd/) and SMART databases (http://smart.embl.de/) were used to confirm the existence of the WRKY domain in all Eucalyptus WRKY proteins.

The Arabidopsis (Arabidopsis thaliana) WRKY (AtWRKY) protein sequences were downloaded from Phytozome (https://phytozome-next.jgi.doe.gov/) (Goodstein et al., 2012). Multiple sequence alignment of AtWRKY and EgWRKY proteins was performed using ClustalW with MEGA11.0 software (Tamura et al., 2021) with default parameters and used to construct a maximum likelihood phylogenetic tree of AtWRKY and EgWRKY protein sequences.

TBtools (Chen et al., 2023) was used to confirm the chromosomal locations and gene structures of the EgWRKYs and to generate a chromosome distribution map and gene structures for all EgWRKYs. To gain a better understanding of the functions of the EgWRKYs, the deduced protein sequences were analyzed using MEME (Multiple Expectation Maximization for Motif Elicitation: http://meme-suite.org/tools/meme/) to identify and examine conserved motifs, with the following parameters: the repeat count was adjusted to zero or one, with a maximum of ten motifs (Bailey et al., 2009). TBtools was also utilized to generate Circos plots. The One Step MCScanX program in TBtools was used to identify and investigate duplication types and collinear blocks and to build Circos plots.

The isoelectric point (pI) and molecular weight (Mw) functions in ExPASy (https://web.expasy.org/protparam/) were used to estimate the pI and Mw of full-length EgWRKY proteins (Gasteiger et al., 2003). TBtools was used to harvest 2,000 base-pair (bp) sequences upstream of the start codon (ATG) of each EgWRKY gene in the Eucalyptus genome (Chen et al., 2023). After analyzing sequences with PlantCARE online (Lescot et al., 2002) and New PLACE (https://www.dna.affrc.go.jp/PLACE/?action=newplace), cis-acting elements were identified and their distribution in each promoter was determined using TBtools.

E. grandis seeds were surface sterilized with 1% NaClO for 15 min and placed on sterilized quartz sand for germination. After 14 days, seedlings with uniform growth status were transferred to individual pots and inoculated with R. irregularis 197198 (Ri) at the roots using 400 spores per plant. At 45 days post inoculation (dpi) with Ri, mycorrhizal colonization was detected in E. grandis roots after WGA488 staining and observed under a fluorescence microscope (Trouvelot et al., 1986). At 90 dpi, the roots were watered with 20 mL of a 1 × 10⁹ CFU/mL suspension of R. solanacearum. The plants were placed in an artificial climatic chamber with 16 h of light, 30°C, 80% relative humidity, light intensity of 100–200 Wm^–2, and 8 h of darkness at 28°C, 80% relative humidity. Mycorrhizal and non-mycorrhizal E. grandis root samples were collected at 0, 24, 48, and 96 hours post-infection (hpi) with Rs, flash-frozen in liquid nitrogen, and stored at –80°C.

Total RNA was extracted from the samples using a Plant RNA Kit (BIO-TEK R6827, OMEGA). The integrity of the total RNA was assessed using 1.5% agarose gel electrophoresis, while its concentration and quality were evaluated using an ultra microspectrophotometer (Nanophotometer N50 Touch, Implem). RNAs from the roots of non-mycorrhizal and mycorrhizal E. grandis seedlings infected with Rs at 0, 24, 48, and 96 hpi were sent to Lianchuan Biotechnology, Hangzhou for cDNA library construction and double-end sequencing of the cDNA library using the Illumina NovaSeq™ 6000 platform (LC-Bio Technologies Co., Ltd., Hangzhou, China). The RNA-Seq (PRJNA893422) data were annotated based on the E. grandis reference genome. The ballgown package was used for segmental per kilobase million (FPKM) quantification. FPKM = all exon sequences/matched sequences × exon length (KB). At the same time, TBtools was used to perform normalization analysis of the FPKM data and to identify significant differences (|log2FC ≥ 1| was considered to be significant). The heatmap diagram was also constructed with log₂ values using TBtools.

We identified 117 WRKY sequences in the E. grandis genome ( Figure 1 , Supplementary Table S1 ). The EgWRKYs encode proteins ranging from 139 (EgWRKY7) to 1,831 (EgWRKY63) amino acids long. The molecular weights of the EgWRKYs range from 15.796 kDa (EgWRKY72) to 204.607 kDa (EgWRKY63), with the majority falling between 20 and 85 kDa. The calculated isoelectric points (pI) range from 4.75 (EgWRKY85) to 11 (EgWRKY80), with an average of 7.11. The pI values of 48 EgWRKYs are greater than 7, while the pI values of 69 EgWRKYs are less than 7.

Sequence analysis revealed a highly conserved middle WRKY domain (WD) in the EgWRKY proteins ( Figure 2 ), covering approximately 60 amino acid residues. The WRKY proteins were classified into four types based on the specific zinc-finger motifs and the number of WDs. Group I members possess two WRKY motifs, while group II and III members possess a single WRKY motif ( Figure 1 ). Group I and II members harbor a C₂H₂ zinc-finger motif following the WRKY structural domain, whereas group III WRKY proteins contain a C₂HC zinc-finger motif ( Figure 2 ). By combining the results of phylogenetic analysis and analysis of zinc-finger structures, we further classified the EgWRKYs into subgroups. Cluster II was classified into six subgroups (II-a–II-f): II-a (C-X₅-C-X₂₃-H-X₁-H), II-b (C-X₅-C-X₂₃-H-X₁-H), II-c (C-X₄-C-X₂₃-H-X₁-H), II-d (C-X_4/5-C-X₂₃-H-X₁-H), II-e (C-X₅-C-X₂₃-H-X₁-H), and II-f (C-X₄-C-X₂₂-H-X₁-H). Group III was subdivided into two subgroups: III-a (C-X₇-C-X₂₃-H-X₁-C) and IIIb (C-X_5/7-C-X₂₅-H-X₁-C) ( Figures 1 , 2 ). By contrast, the presence of incomplete WRKY domains (KQVQ) or zinc fingers in group IV ( Figure 2 ) suggests that EgWRKY65, EgWRKY72, EgWRKY78, EgWRKY51, and EgWRKY71 may have lost WRKY TF function. In addition to variations in the zinc-finger motif, mutations in the WRKYGQK sequence were also observed in E. grandis, including WKKHGQK (II-a), WRKYGKK (II-c), WRKYGQR (II-f), WRKYAQE (III-a), WRKYGSK (III-a), WRKYDQK (III-b), and WTMYRQR(III-b). Since EgWRKY3 and EgWRKY7 each contain a zinc-finger motif, they were classified into group III-a rather than group IV, although their zinc fingers exhibit minor differences with those of other group III-a members.

The 117 EgWRKYs are unevenly distributed across all 11 chromosomes of E. grandis ( Figure 3 ). Most of these EgWRKY genes are located on chromosomes 1 and 7, accounting for 15.38% (18 genes) and 23.08% (21 genes) of the total number of genes, respectively. Six EgWRKY genes are dispersed on chromosomes 2 and 4, each accounting for 5.12%. Chromosome localization mapping also revealed that EgWRKYs exist in multiple gene clusters on different chromosomes, such as EgWRKY87–EgWRKY90 on chromosome 8 and EgWRKY91–EgWRKY95 on chromosome 9, with most members of the same cluster categorized in the same subfamilies.

Collinearity analysis revealed 21 pairs of segmental duplicates and 4 tandem duplicates among the 117 EgWRKYs. The four tandem duplicates include EgWRKY4 with EgWRKY5, EgWRKY2 with EgSCL8 (encoding scarecrow-like protein 8), EgWRKY1 with EgWRKY7, and EgWRKY74 with E. grandis “Probable Disease Resistance Protein” (XM_039316797.1) ( Figure 4A , Supplementary Table S2 ). Additionally, we identified 68 pairs of segmental duplicates between E. grandis and Arabidopsis WRKYs ( Figure 4B , Supplementary Table S3 ).

The quantity of introns and exons differ among the EgWRKYs. Specifically, EgWRKY family genes contain one to twelve introns and two to nine exons. In terms of exon length and intron number, members of the same subfamily exhibit comparable exon/intron distribution patterns. For example, most members of subfamilies II-c, II-d, and II-e have only one to two introns, while most members of subfamilies I, II-a, and IV possess three or more introns. Notably, subfamily IV members EgWRKY51 and EgWRKY6 contain an intron of more than 18,000 bp, and subfamily IIIa members EgWRKY5 and EgWRKY9 possess an even larger intron of ~38,000 bp ( Figure 5B , Supplementary Table S1 ).

Ten motifs were predicted in the EgWRKYs, with two to six motifs present in each EgWRKY except for some members of group IV ( Figures 5A–C ). Some motifs are shared by most members of groups I–III, such as motif 1 and motif 2, while others are specific to the subfamily. For example, motif 3 and motif 5 are present only in groups I and II-f. However, members of group I contain motif 4, motif 1, and motif 2, whereas group II-f members lack these motifs. This unique structure may cause members of subfamily group II-f to be outliers in the evolutionary tree ( Figures 1 , 5A ). Most group II-a and group II-b members possess motif 7 and motif 6, while group II-a members contain an additional motif 4 linked to motif 1 than group II-b; groups II-d, II-e, and III-b only contain motif 1 and motif 2. In addition, groups III-a, IV-a, and IV-c contain similar motifs in terms of number and arrangement, which could explain why they were grouped in the same branch ( Figure 1 ), whereas EgWRKY4, EgWRKY6, EgWRKY51, and EgWRKY78 were classified in subfamily IV based on the structures of their zinc fingers ( Figure 2 ).

To investigate the putative regulatory mechanisms of EgWRKY genes, we identified the cis-elements in the 2-kb sequences upstream of their translation start sites (ATGs) and visualized them using TBtools ( Figure 6A ). The cis-elements ABREs (abscisic acid responsive), TGACG motifs (MeJA responsive), and CGTCA motifs (MeJA responsive), which are associated with defense hormones, accounted for more than 50% of all EgWRKY subfamilies members ( Figure 6B ). Abscisic acid and jasmonic acid are important phytohormones in the spatio-temporal dynamics of AMF-induced resistance (Cameron et al., 2013). These findings suggest that these EgWRKYs might be involved in the mycorrhiza-induced resistance of Eucalyptus against Rs. The presence of auxin-responsive elements, such as TGA-box and AuxRR-core elements (Lescot et al., 2002), suggests that some EgWRKYs may play a role in regulating plant growth. Notably, 60% of subgroup II-b members contain GC-motif elements, whereas 80% and 100% of subgroup II-b and IV-b members contain MBS elements, 73% and 100% of subgroup II-b and II-e members contain LTR elements, and 14% and 5% subgroup I and II-c members contain SARE elements, respectively. Finally, whereas subgroup II-f members lack ARE elements, all subgroup III-b, IV-a, IV-b, and IV-c members contain this element.

In addition, we identified cis-acting elements associated with mycorrhizal symbiosis in the EgWRKYs, such as P1BS (GNATATNC), CTTC-motif (NTTCTTGTTN), and AW-box (CG(N)7CNANG) (Shi et al., 2023; Duan et al., 2024). Members of all subfamilies except groups IV-b and IV-c contain AW-boxes. In addition more than 20% of members of groups II-a and II-c contain P1BS, CTTC-motif, and AW-boxes, suggesting that these two subfamilies may have a role in the AMF symbiosis. The W-box has been implicated in plant resistance to Rs (Dang et al., 2014; Phukan et al., 2016; Mukhi et al., 2021; Yang et al., 2021). A relatively high number of family members contain W-box elements (71%, 83%, 90%, 85%, 50%, and 67% of group I, II-a, II-b, II-c, II-e, and IV-a members contain W-box elements, respectively, and in other subfamilies, 100% of members contain W-boxes), suggesting that EgWRKYs play important roles in the E. grandis–AMF–R. solanacearum system ( Figure 6B ). In addition, 7 EgWRKY genes that exhibited significant differences in expression at different time points between mycorrhizal and non-mycorrhizal E. grandis contain W-boxes ( Figures 6C , 7 ).

During the Rs infection process, EgWRKY genes showed different expression patterns in the roots of Ri mycorrhizal and non-mycorrhizal E. grandis plants ( Figure 7 ). Among the 58 EgWRKYs, 21 genes responded to R. irregularis in the absence of Rs infection. In detail, eight genes were downregulated in response to this treatment, including three from group I (EgWRKY10, EgWRKY15, and EgWRKY57) and five from group II (EgWRKY90, EgWRKY17, EgWRKY37, EgWRKY21, and EgWRKY85). The remaining 13 EgWRKYs were upregulated by this treatment, including two from group I (EgWRKY26 and EgWRKY116), eight from group II (EgWRKY18, EgWRKY35, EgWRKY38, EgWRKY62, EgWRKY69, EgWRKY107, EgWRKY109, and EgWRKY111), and three from group III-b (EgWRKY41, EgWRKY106, and EgWRKY104). Notably, EgWRKY116, EgWRKY62, EgWRKY107, EgWRKY109, EgWRKY38, EgWRKY41, and EgWRKY106 were expressed at higher levels in mycorrhizal plants compared to non-mycorrhizal plants. This trend was observed not only in the absence of the pathogen (at 0 h of Rs infection), but also at other time points of the pathogenic interaction. These results suggest that the expression of these seven genes is influenced by AMF symbiosis and might enhance E. grandis resistance against Rs.

Additionally, during Rs infection, EgWRKY116, EgWRKY62, EgWRKY107, EgWRKY41, and EgWRKY106 showed a significant decrease in expression at 24 hpi, regardless of mycorrhization, followed by an increase at 48–96 hpi; this trend was observed for EgWRKY26 only in the mycorrhizal roots of E. grandis. In the absence of mycorrhizal treatment, EgWRKY57 and EgWRKY114 showed slowly upregulated expression before 24 hpi followed by downregulation at 48 hpi and reached the lowest expression level at 96 hpi. EgWRKY27 expression increased after infection with Rs and was quickly and significantly upregulated at 96 hpi. EgWRKY109 was quickly upregulated at 48 hpi with Rs, and at 96 hpi, its expression level returned to that at 24 hpi. These results indicate that these EgWRKYs have different expression patterns following infection by Rs. EgWRKY47, EgWRKY84, EgWRKY13, and EgWRKY25 generally showed sustained increases in expression during Rs infection. Additionally, the expression of EgWRKY37, EgWRKY42, and EgWRKY90 significantly decreased after Rs infection compared to the uninfected controls.