Epigenetic Activation of Enoyl-CoA Reductase By An Acetyltransferase Complex Triggers Wheat Wax Biosynthesis

The epidermal surface of bread wheat (Triticum aestivum) is coated with a hydrophobic cuticular wax layer that protects plant tissues against environmental stresses. However, the regulatory mechanism of cuticular wax biosynthesis remains to be uncovered in bread wheat. Here, we identified wheat Enoyl-CoA Reductase (TaECR) as a core component responsible for biosynthesis of wheat cuticular wax. Silencing of TaECR in bread wheat resulted in a reduced cuticular wax load and attenuated conidia germination of the adapted fungal pathogen powdery mildew (Blumeria graminis f.sp. tritici; Bgt). Furthermore, we established that TaECR genes are direct targets of TaECR promoter-binding MYB transcription factor 1 (TaEPBM1), which could interact with the adapter protein Alteration/Deficiency in Activation 2 (TaADA2) and recruit the histone acetyltransferase General Control Non-derepressible 5 (TaGCN5) to TaECR promoters. Most importantly, we demonstrated that the TaEPBM1-TaADA2-TaGCN5 ternary protein complex activates TaECR transcription by potentiating histone acetylation and enhancing RNA polymerase II enrichment at TaECR genes, thereby contributing to the wheat cuticular wax biosynthesis. Finally, we identified very-long-chain aldehydes as the wax signals provided by the TaECR-TaEPBM1-TaADA2-TaGCN5 circuit for triggering Bgt conidia germination. These results demonstrate that specific transcription factors recruit the TaADA2-TaGCN5 histone acetyltransferase complex to epigenetically regulate biosynthesis of wheat cuticular wax, which is required for triggering germination of the adapted powdery mildew pathogen.


51
The epidermal surfaces of aerial plant organs are coated with a hydrophobic layer, the cuticle, to 52 protect plant tissues against enormous environmental stresses such as desiccation, ultraviolet radiation, 53 excessive light, extreme temperatures, and even pathogen infections Samuels et al., 54 2008;Domínguez et al. 2017). Although the composition and structure of the cuticle vary among plant 55 species, organs, developmental stages, and even environmental conditions, plant cuticle generally 56 consists of a macromolecular scaffold of cutin impregnated by and covered with the cuticular wax 57 mixture Fernández et al., 2016;Domínguez et al., 2017). Increasing evidence 58 reveals that many microbial pathogens have acquired the capacity to utilize the plant cuticular wax 59 components to initiate their pre-invasion and infection processes Aragón et al., Hooker et al., 2002;Zheng et al., 2005;Bach et al., 2008;Beaudoin et al., 2009;Lee 71 et al., 2009;Weng et al., 2010;Haslam et al., 2012;Haslam and Kunst, 2013;Kim et al., 72 2013;Haslam et al., 2015). The elongated VLC acyl-CoAs are then modified into aldehydes, alkanes, 73 secondary alcohols, and ketones by an alkane-forming pathway, or into primary alcohols and wax esters 74 by an alcohol-forming pathway (Aarts et al., 1995;Chen et al., 2003;Rowland et al., 75 2006;Greer et al., 2007;Rowland et al., 2007;Bourdenx et al., 2011;Bernard et al., 2012;Yang et al., 76 2017;. As a core component of fatty acid elongase complex, enoyl-CoA reductase 77 (ECR) catalyzes the final step in the biosynthesis of VLC acyl-CoAs (Zheng et al., 2005). In Arabidopsis 78 thaliana, silencing of AtECR results in a reduction of all cuticular wax compositions such as VLC fatty 79 acids, alcohols, aldehydes, alkanes, and ketones, suggesting that Arabidopsis AtECR gets involved in the 80 VLC acyl-CoAs biosynthesis (Zheng et al., 2005). Increasing research in Arabidopsis reveals that AtECR 81 expression is governed by multiple transcriptional regulators. For instance, the AtECR transcription is 82 up-regulated by the MYB type transcription factors such as AtMYB30 and AtMYB94, but negatively 83 regulated by the AP2/ERF-type transcription factor DECREASE WAX BIOSYNTHESIS (AtDEWAX) 84 in Arabidopsis (Raffaele et al., 2008;Go et al., 2014;Lee and Suh 2014). However, the biological 85 function and transcriptional regulation of ECR remain to be uncovered in important cereal crops such as 86 bread wheat (Triticum aestivum).

87
Chromatin modifications such as acetylation, methylation, and ubiquitination play important roles in 88 the regulation of transcriptional reprogramming associated with plant development and stress responses 89 (Jenuwein and Allis 2001;. As important epigenetic modifications, trimethylation 90 of histone H3 lysine 4 and deubiquitination of histone H2B could induce a permissive chromatin 91 structure for gene activation (Kurdistani and Grunstein, 2003;Daniel et al. 2004;Schmitz et al. 2009).

92
Similarly, acetylation of histone lysine residues catalyzed by histone acetyltransferases (HAT) also 93 promotes gene transcription (Kurdistani and Grunstein, 2003). As the first HAT linked to gene 94 transcriptional activation, General Control Non-derepressible 5 (GCN5) interacts with the adaptor 95 protein Alteration/Deficiency in Activation 2 (ADA2) in the HAT module of the transcriptional 96 co-activator Spt-Ada-Gcn5-acetyltransferase (SAGA) complex, which is engaged in histone acetylation, 97 histone deubiquitination, and even recruitment of the RNA polymerase II (RNA Pol II) (Grant et al., 98 1997;. The 99 Alteration/Deficiency in Activation 2-General Control Non-derepressible 5 (ADA2-GCN5) complex is 100 reported to function in concert with specific transcription factors (TFs) to regulate gene transcription 2018; Xing et al., 2018;Zou et al., 2018;Zheng et al., 2020). On the aerial surface of wheat, the first 113 contact between Bgt and wheat takes place at the cuticle, and the Bgt conidia germination is induced to 114 initiate the infection processes Wright et al., 2002). In bread wheat, silencing of 115 3-KETOACYL-CoA SYNTHASE (TaKCS6) and WAX INDUCER 1 (TaWIN1), two positive regulators in 116 wheat cuticular wax biosynthesis, results in a reduction of Bgt conidia germination, suggesting that the 117 cuticular wax biosynthesis is essential to stimulate the Bgt conidia germination in bread wheat (Kong and 118 Chang, 2018;. However, the function of other components responsible for the wheat 119 cuticular wax biosynthesis in modulating Bgt conidia germination needs to be characterized.

120
In this study, we showed that wheat enoyl-CoA reductase (TaECR) is a core component responsible 121 for the cuticular wax biosynthesis in bread wheat. TaECR promoter-binding MYB transcription factor 1 122 (TaEPBM1) recruits the TaADA2-TaGCN5 histone acetylatransferase complex to activate TaECR 123 transcription by potentiating histone acetylation and enhancing RNA Pol II enrichment at TaECR genes 124 and thus stimulate the cuticular wax biosynthesis required for stimulating Bgt conidia germination.

125
Besides, VLC aldehydes were identified as the wax signals provided by the 126 TaECR-TaEPBM1-TaADA2-TaGCN5 circuit for Bgt germination in bread wheat. Thus, we revealed 127 that the TaECR-TaEPBM1-TaADA2-TaGCN5 circuit regulates the wheat cuticular wax biosynthesis 128 essential for the germination of powdery mildew fungus.

132
In this study, we are interested in exploring the function of the wheat Enoyl-CoA Reductase (ECR) 133 gene in regulating the cuticular wax biosynthesis required for stimulating Bgt conidia germination. To 134 this end, we first identified the wheat TaECR genes based on the sequence of the Arabidopsis AtECR 135 gene (AT3G55360) and the reference genome of the hexaploid bread wheat (International Wheat 136 Genome Sequencing Consortium 2018). Three highly conserved homologous sequences of TaECR 137 genes separately located on chromosomes 3AS, 3BS and 3DS were isolated from the hexaploid bread 138 wheat cultivar Jing411, and were designated as TaECR-A, TaECR-B, and TaECR-D (Supplemental Fig. 139 S1). The open reading frames (ORFs) of these TaECR genomic sequences all contained four exons and 140 three introns, encoding proteins with over 99% amino acid sequence identity (Supplemental Figs. S1).

141
To analyze the evolution of ECR in land plants, we employed protein sequences of TaECR as query 142 sequences to search the genomes of representative land plant species from the Joint Genome Institute 143 (JGI) Phytozome v12.1 database. As shown in Supplemental Figure S2, highly homologous ECR ( Fig. 1A). In addition, we expressed TaECR-HA in the wheat protoplast and performed a sucrose 153 density-gradient fractionation to validate the ER-localization of TaECR in wheat cells. Since TaECR-A,   154 TaECR-B, and TaECR-D share more than 99% amino acid sequence identity, TaECR-A was selected as 155 a representative TaECR in this experiment. As shown in Supplemental Figure S3, TaECR-HA   156 cofractionated with the ER marker BiP and exhibited the same Mg 2+ -dependent density shift as the ER 157 marker BiP, further confirming the ER-localization of TaECR in bread wheat. Thereafter, the expression 158 profiles of TaECR were analyzed in different tissues of wheat cultivar Jing411 using reverse 159 transcription quantitative PCR (RT-qPCR). As shown in Figure 1B, TaECR exhibits the lowest 160 expression level in roots and the highest expression levels in epidermis of leaves and stems. Its 161 expression levels were much higher in leaves and stems than in roots (Fig. 1B).

162
To explore whether TaECR involved in the wheat cuticle biosynthesis is required for stimulating Bgt  Fig. S5). As shown in Figure 1C, the total cutin load was not significantly changed 170 by the silencing of TaECR, but the wax load decreased from 12.1 μg cm -2 on wild-type Jing 411 leaves to 171 a significant level of 2.6 μg cm -2 on TaECR-silenced wheat leaves. Further quantitative analysis of wax 172 constituents revealed that VLC fatty acids and their derivatives such as aldehydes, alcohols, alkanes, 173 ketone, and even C 46 -C 50 esters showed a remarkable decrease in the TaECR-silenced plants (Fig. 1D).

174
Thereafter, the conidia germination of Bgt strain E09 was examined on leaves of wild-type and 175 TaECR-silenced plants using light microscopy. Compared with the mock control, inoculation with 176 BSMV-γ has no significant effect on Bgt conidia germination in bread wheat (Supplemental Fig. S5

183
To identify the transcriptional regulator that directly binds to TaECR promoters, we performed yeast 184 one-hybrid screening against a wheat leaf cDNA library using TaECR promoter regions as baits. One assay (EMSA) showed that TaEPBM1 exclusively associated with the wild-type MBS but not mutant 194 MBS fragments, suggesting that TaEPBM1 has the MBS-binding activity (Supplemental Fig. S7). Yeast 195 one-hybrid assays revealed that TaEPBM1 could bind to the wild-type TaECR promoters, but not the 196 MBS cis-element mutated TaECR promoters, suggesting that TaEPBM1 could recognize the MBS 197 cis-element and directly bind to TaECR promoters in yeast cells ( Fig. 2A).

198
To examine whether TaEPBM1 could bind to TaECR promoters in plant cells, we employed the wheat 199 protoplast transfection system, in which luciferase (LUC) reporters containing wild-type or mutant 200 TaECR promoter regions were cotransfected with effector constructs over-expressing TaEPBM1 (Fig.   201 2B). As shown in Figure 2C, the LucA ratio obtained from LUC reporters containing wild-type TaECR 202 promoters increased to a significant level of above 4.8 in the presence of TaEPBM1, compared with the 203 basal LUC activity of the Gal4 DNA binding domain (DBD). In contrast, the LucA ratio obtained from 204 LUC reporters containing the MBS-mutated TaECR promoters was not significantly changed by the 205 addition of TaEPBM1 (Fig. 2C). This result suggests that TaEPBM1 could bind to TaECR promoters and 206 activate their expression in wheat cells.

207
To determine the association of TaEPBM1 with TaECR promoters in bread wheat, we generated an 208 antibody specifically against TaEPBM1 and performed chromatin immunoprecipitation (ChIP) assay in 209 Jing 411 leaves (Supplemental Fig. S8A). The wheat elongation factor 1 (TaEF1) gene was employed as 210 control. As shown in Figure 2D and E, two genomic regions (represented by 1 and 2) containing the MBS 211 cis-element in TaECR promoters were subjected to the ChIP assay and found to be enriched in DNA 212 samples precipitated with the α-TaEPBM1 antibody, suggesting that TaEPBM1 associated with TaECR 213 promoters in bread wheat. At the same time, nuclear run-on assays revealed that TaECR

223
To explore the molecular mechanism by which TaEPBM1 regulates TaECR transcription, we 224 performed a yeast two-hybrid screening against a wheat leaf cDNA library to identify 225 TaEPBM1-interacting proteins. One of the isolated interacting proteins was homologous to the rice 226 ADA2 (LOC_Os03g53960) and was designated as TaADA2. Three highly-homologous sequences of 227 TaADA2 genes located on chromosomes 5AL, 5BL and 5DL were obtained from Jing411 and encode 228 TaADA2-A, TaADA2-B, and TaADA2-D with more than 99% amino acid sequence identity, among 229 which, TaADA2-A was selected as a representative TaADA2 in the following experiments 230 (Supplemental Fig. S10). As shown in Figure 3A, the interaction between TaEPBM1 and TaADA2 was 231 detected in the EGY48 yeast cells. Further yeast two-hybrid analysis with truncated TaEPBM1 and 232 TaADA2 revealed that the C-terminal region of TaEPBM1 and the N-terminal region of TaADA2 were 233 responsible for their interaction (Fig. 3A).
To validate the TaEPBM1-TaADA2 interaction, we performed both in vitro and in vivo protein 235 interaction assays. As shown in Figure 3B, GST-TaEPBM1, but not GST alone, could retain 236 TaADA2-His instead of TaGCN5-His in the GST pull-down assay, suggesting that TaEPBM1 directly 237 interacts with TaADA2 but not TaGCN5 in vitro. In the bimolecular fluorescence complementation 238 (BiFC) assay, YFP was reconstituted in the nucleus only in the co-expression pair of nYFP-TaEPBM1 239 and cYFP-TaADA2, but not in control pairs (Fig. 3C). In addition, we generated an antibody specifically 240 against TaADA2 and performed co-immunoprecipitation (co-IP) assay to analyze the 241 TaEPBM1-TaADA2 association in Jing 411 leaves (Supplemental Fig. S8B). As shown in Figure 3D, 242 TaEPBM1 was found co-immunoprecipitated with TaADA2, which was not detected in the TaEPBM1-243 or TaADA2-silenced plants, suggesting that TaEPBM1 interacts with TaADA2 in bread wheat.

244
In yeast and plants, ADA2 interacts with GCN5 to constitute the ADA2-GCN5 complex (Grant et al., 245 1997;Zhou et al., 2017; 246 Castroverde 2019). The finding that TaEPBM1 interacts with TaADA2 prompted us to 247 ask whether TaEPBM1, TaADA2, and TaGCN5 could form a complex. To this end, we first identified 248 the wheat TaGCN5

266
Having already demonstrated that TaEPBM1 directly binds to TaECR promoters and TaEPBM1, 267 TaADA2 and TaGCN5 could form a complex, we next ask whether the TaEPBM1-TaADA2-TaGCN5 268 ternary protein complex associates with TaECR promoters in bread wheat. To test this hypothesis, we of TaADA2 decreased the accumulation of TaGCN5-HA but not TaEPBM1 at TaECR promoters (Fig. 5

282
In yeast and other plants, the ADA2-GCN5 complex was reported to activate gene transcription 283 through enhancing histone acetylation and recruitment of the RNA polymerase II (Grant et al., 1997; 284 Zhou et al., 2017;

347
Increasing evidence revealed that the cuticular wax provides signals for stimulating conidia 348 germination of powdery mildew fungus Blumeria graminis Wright et al., 2002; 349 Weidenbach et al., 2014;Kong and Chang, 2018;. The finding that  Fig. S13)., suggesting that the chemical composition of cuticular wax was responsible for the difference 380 in Bgt germination rate between wild-type and TaECR, TaEPBM1, TaADA2 or TaGCN5-silenced plants.

381
To characterize the role of single wax components including aldehydes, fatty acids, alcohols, alkanes, 382 and even esters (which are reduced by silencing of TaECR, TaEPBM1, TaADA2, and TaGCN5), in 383 stimulating Bgt germination, we employed glass slides covered with Formvar/wheat cuticular wax 384 supplemented with corresponding synthetic chemicals and examined the Bgt conidia germination. As 385 shown in Figure 8B, a supplement of C 26 , C 28 or C 30 -aldehyde could restore the Bgt germination penalty 386 on the glass slides coated with Formvar/cuticular wax isolated from the TaECR, TaEPBM1

403
In previous studies, Enoyl-CoA Reductase (ECR) was revealed to function in the endoplasmic 404 reticulum (ER) to catalyze the reduction of the enoyl-CoA, the final step of VLC acyl-CoA elongation 405 (Kohlwein et al., 2001). Arabidopsis cer10 mutant disrupted in the AtECR gene exhibited reduced 406 cuticular wax load, indicating that ECR participates in VLC acyl-CoA's elongation reactions in 407 Arabidopsis (Zheng et al., 2005). However, the biological function of ECR in other plants, especially the 408 important crops, is still unknown. In this study, we characterized the function of wheat TaECR, which 409 had more than 88% amino acid identities with Arabidopsis AtECR. Both confocal microscopy imaging 410 and sucrose density-gradient fractionation assay showed that TaECR was localized to the ER, the site of Arabidopsis, suggesting that some cuticle biosynthetic components have acquired divergent functions in 427 the evolution of dicots and monocots   (Avato et al., 1987;Raffaele et al., 2008;Liu et al., 442 2012;Zhu et al., 2014;Lee and Suh, 2015b;Lee et al., 2016). Here, we found that

460
Increasing evidence revealed that the ADA2-GCN5 complex complex is recruited to the target 461 promoters through interaction with specific transcription factors, in which ADA2 proteins serve as an 462 adapter to bridge the association between ADA2-GCN5 complex and transcription factors. For instance, 463 the rice homeodomain transcription factor WOX11 directly interacts with rice ADA2 and recruits the 464 ADA2-GCN5 complex to target root-specific genes involved in cell proliferation of crown root meristem 465 (Zhou et al., 2017). Similarly, the Arabidopsis AP2 transcription factor CBF directly interacts with 466 ADA2 to activate transcription of cold-responsive genes . Recent research in P.

479
It has been demonstrated that the fungal pathogen Blumeria graminis could utilize the plant cuticular 480 wax components to initiate its pre-penetration processes Wright et al., 2002; 481 Weidenbach et al., 2014;Kong and Chang, 2018;. Additionally, VLC 482 aldehydes promote the in vitro conidia germination and appressorial development of B. graminis in a 483 dose-dependent manner (Hansjakob et al., 2010;Hansjakob et al., 2011;Kong and Chang, 2018;  TaECR coding regions were amplified using primers listed in Supplemental Table S1 and cloned into   517 vector pCAMBIA1300-YFP via the pENTRY-TaECR construct using the GATEWAY cloning 518 technology and then transformed into the Agrobacterium tumefaciens strain GV3101. The 519 pCAMBIA1300-derivatives expressing the endoplasmic reticulum (ER) marker mCherry-HDEL were 520 also transformed into the A. tumefaciens strain GV3101. Nicotiana benthamiana plants used in this study 521 were grown in a growth chamber at 22°C with a 14/10 h light /dark photoperiod. N. benthamiana leaves 522 co-infiltrated with A. tumefaciens strain GV3101 expressing TaECR-YFP and mCherry-HDEL were 523 imaged using the confocal microscope (Leica TCS SP5) at 48 hours post-Agro-infiltration.

524
For the subcellular localization analysis in wheat protoplast, the TaECR coding region was amplified 525 using primers listed in Supplemental Table S1 and cloned into the pCAMBIA1300-HA vector via the 526 pENTRY-TaGCN5 construct using GATEWAY cloning technology to generate the fusion protein 527 TaECR trifluoroacetamide, and then subjected to GC-MS analysis as described previously (Kong and Chang, 555 2018). The oven temperature program was set at an initial temperature 75°C, increased to 200°C at 15°C 556 min -1 , then increased to 280°C at 1.5°C min -1 . Methyl nonadecanoate was added as the internal standard 557 for the FID peak-based quantification.

558
The cuticular wax composition analysis was performed as described by Hansjakob et al. (2010).

559
Briefly, wheat leaves from at least 5 BSMV-VIGS wheat plants were dipped into chloroform (Merck).  Table S1 and cloned into the pCAMBIA1300-HA vector via the 576 pENTRY-TaGCN5 construct using GATEWAY cloning technology to generate the fusion protein 577 TaGCN5-HA. 10 µg plasmids for RNAi and pCAMBIA1300-TaGCN5-HA constructs were 578 co-transfected into wheat protoplasts as previously described by Liu et al (2019). The transformed 579 protoplasts were cultured in W5 solution for at least 48 hours for the next gene expression analysis or 580 ChIP-qPCR assays.

581
For the wheat protoplast transactivation assay, the TaEPBM1 coding region was amplified using the 582 primers listed in Supplemental Table S1 and cloned into the vector pIPKb004 vector via the 583 pENTRY-TaEPBM1 construct using the GATEWAY cloning technology. Similarly, TaECR promoters 584 were amplified using the primers listed in Supplemental Table S1 and ligated into the vector 585 5XGAL4-LUC, which was then co-transfected with pIPKb004 derivatives and internal control pPTRL 586 into the wheat protoplast as previously described by Liu et al (2019). LUC activity was measured at 48 587 hours post-transfection using a Promega dual-luciferase reporter assay system (Promega, E1910) according to the manual. In wheat protoplast transactivation assay, three biological replicates were 589 statistically analyzed.

596
After RNA extraction using Trizol, the nascent RNA was enriched by streptavidin magnetic beads 597 (Invitrogen) and subjected to the RT-qPCR assay using the primers listed in Supplemental Table S1. For 598 the RT-qPCR assay, total RNA was extracted using Trizol reagent and treated with Dnase I for the 599 gDNA removal. 2μg RNA was then employed to synthesize the first-strand cDNA template using the 600 cDNA synthesis supermix (Transgen) according to the manual. RT-qPCR was performed using the 601 qPCR Master Mix (Invitrogen) under the following programs: 95°C for 3 min, 40 cycles at 95°C for 20 s, 602 55°C for 20 s, and 72°C for 15 s, followed by 72°C for 1 min. The expression levels of TaGADPH,

603
TaECR, TaEPBM1, TaADA2, and TaCHR729 were analyzed using the primers listed in Supplemental 604 Table S1 and the TaGADPH, whose expression is stable among various treatments, was used as the 605 internal control for reference. For the nuclear run-on and RT-qPCR assays, three biological replicates 606 were statistically analyzed.

608
In yeast one-hybrid analysis, the TaEPBM1 coding region was amplified using the primers listed in 609 Supplemental Table S1 and cloned into the vector pGADT7 via the pENTRY-TaEPBM1 construct using 610 GATEWAY cloning technology to generate protein fusions to the GAL4 transcription-activating domain 611 (AD). Similarly, TaECR promoters were amplified using the primers listed in Supplemental Table S1 612 and cloned into the vectors pHIS2, which were then co-transformed with pGADT7 derivatives into 613 competent cells of yeast strain Y187 according to the manual. Yeast transformants were then grown on 614 the SD/-Trp-Leu-His plate with 15% (w/v) 3-amino-1,2,4-triazole (3-AT) to test for HIS2 expression. In 615 yeast two-hybrid analysis, the coding fragments of TaEPBM1, TaADA2, and TaGCN5 were amplified 616 using primers listed in Supplemental Table S1 and separately cloned into the vectors pLexA and 617 pB42AD via the pENTRY-TaEPBM1, pENTRY-TaADA2, and pENTRY-TaGCN5 constructs using 618 GATEWAY cloning technology; these were co-transformed into competent cells of yeast strain EGY48 619 according to the Clontech Yeast Protocols Handbook. For the truncated TaEPBM1 and TaADA2 used in 620 the yeast two-hybrid assay, the TaEPBM1-NT(1-120), TaEPBM1-CT(121-314), TaADA2-NT(1-244), 621 and TaEPBM1-NT(245-568) were amplified using primers listed in Supplemental Table S1 and cloned 622 into the vectors pLexA and pB42AD. The pB42AD-derived prey wheat cDNA library was constructed 623 and screened as previously described by Liu et al (2019). Yeast transformants were grown on the SD/-Ura-Trp-Leu-His plate with X-gal to test for expression of LEU2 and LacZ. The yeast strains Y187 625 and EGY48 were maintained on YPAD and SD/-Ura medium, respectively. For the yeast one-and 626 two-hybrid experiments, at least three independent biological replicates were performed with consistent 627 results.

629
The TaEPBM1 coding region was amplified using the primers listed in Supplemental Table S1 and   630 cloned into the vector pET32, and TaEPBM1

642
For pull-down assays, the TaEPBM1 coding region was amplified using the primers listed in 643 Supplemental Table S1 and cloned into the vector pGEX4T-1 to generate the fusion protein 644 GST-TaEPBM1, while the coding regions of TaADA2 and TaGCN5 were amplified using the primers 645 listed in Supplemental Table S1 and cloned into the vector pET32 to create proteins fusions to the 646 His-tag. The pull-down assay was performed as described by . Briefly, the recombinant 647 proteins with GST and His tags were expressed and purified from E.coli using glutathione sepharose and 648 Ni-NTA resin according to the manual. Recombinant proteins with GST and His tags were mixed as 649 pairs indicated and incubated with glutathione sepharose, and then subjected to the 650 centrifugation-assisted precipitation. After being washed five times with PBS buffer, the precipitates 651 were subjected to SDS-PAGE separation, and the co-precipitation of TaADA2-His or TaGCN5-His with 652 GST or GST-TaEPBM1 was resolved by immunoblotting with α-His antibody (CWBIO, CW0286). For 653 the pull-down assays, at least three independent biological replicates were performed with similar 654 results.

656
For BiFC assays, coding regions of TaEPBM1, TaADA2, and TaGCN5 were amplified using primers 657 listed in Supplemental Table S1 and cloned into vectors pCAMBIA1300-YN and pCAMBIA1300-YC 658 via the pENTRY-TaEPBM1, pENTRY-TaADA2, and pENTRY-TaGCN5 constructs using GATEWAY respectively. Similarly, the TaADA2 coding region was amplified using primers listed in Supplemental 661 Table S1 and cloned into vectors pCAMBIA1300 to express TaADA2 alone. The BiFC assay was 662 performed as described by Liu et al (2019). The interaction was imaged using a confocal microscope 663 (Leica TCS SP5) at 48 hours post-Agro-infiltration. All BiFC images were collected on a Leica TCS SP5 664 confocal laser scanning system (Leica, Mannheim, Germany) connected to an inverted motorized 665 microscope with the following settings: pinhole 1 airy unit, scan speed 400 Hz bidirectional. DAPI and 666 YFP were excited with a 405 nm diode laser and a 514 nm argon laser, respectively. Fluorescence 667 emissions were collected using the following wavelengths: 420-480nm (for DAPI) and 529-540nm (for 668 YFP). Digital confocal images were analyzed using Adobe Photoshop (Version CS5) and adjusted with 669 ImageJ (Version 1.38) for the optimized intensity projection. At least three independent biological 670 replicates were performed for this BiFC assay.

683
The ECL chemiluminescence kit (Pierce Biotechnology) was employed for the immunoblot 684 visualization. For the Co-IP assay, at least three independent biological replicates were perfomed with 685 similar results.

687
For LCI assay, coding regions of TaEPBM1, TaADA2, and TaGCN5 were amplified using primers 688 listed in Supplemental Table S1 and cloned into vectors pCAMBIA-nLUC and pCAMBIA-cLUC via the 689 pENTRY-TaEPBM1, pENTRY-TaADA2, pENTRY-TaGCN5 constructs using GATEWAY cloning 690 technology to express protein fusions to the N-terminal or C-terminal domain of firefly LUCIFERASE, 691 respectively. The LCI assay was performed as described by Kong and Chang (2018). The luminescent 692 signal was collected at 60 hours post-Agro-infiltration trough using a cooled CCD camera (iXon, 693 Andor Technology). At least three independent biological replicates were performed for this LCI assay.

708
The cuticle hydrophobicity on the wheat leaf surface and glass slide was analyzed by measuring the 709 contact angle of 1 μL water droplets on the indicated surface. Angles from at least 50 water droplets were 710 separately measured for 5s using the contact angle system (SDP-300, Sindin) and five independent 711 surface samples were statistically analyzed using Student's t-test. The cuticle wax on the wheat leaf 712 surface and glass slide were manipulated as described previously (Hansjakob et al., 2011, Wang et al.,   Formvar/cuticular wax (480 μg ml -1 ) and pure waxes (7×10 -5 mol l -1 ), or with Formvar/cuticular wax 873 (480 μg ml -1 ) only (control treatment). For A and B, more than 500 Bgt conidia were analyzed in one 874 experiment, and three biological replicates were statistically analyzed for each treatment, and data are 875 presented as the mean±SE (Student's t-test; ** P<0.01). 876 877