The GT factor ZmGT-3b mediates growth–defense tradeoff by regulating photosynthesis and defense response1

Plant growth and development face constant threat from various environmental stresses. Transcription factors (TFs) are crucial for maintaining balance between plant growth and defense. Trihelix TFs display multifaceted functions in plant growth, development, and responses to various biotic and abiotic stresses. Here, we explore the role of a trihelix TF, ZmGT-3b, in regulating the growth–defense tradeoff in maize (Zea mays). ZmGT-3b is primed for instant response to Fusarium graminearum challenge by implementing a rapid and significant reduction of its expression to suppress seedling growth and enhance disease resistance. ZmGT-3b knockdown led to diminished growth, but improved disease resistance and drought tolerance in maize seedlings. In ZmGT-3b knockdown seedlings, the chlorophyll content and net photosynthetic rate were strongly reduced, whereas the contents of major cell wall components, such as lignin, were synchronically increased. Correspondingly, ZmGT-3b knockdown specifically downregulated photosynthesis-related genes, especially ZmHY5 (encoding a conserved central regulator of seedling development and light responses), but synchronically upregulated genes associated with secondary metabolite biosynthesis and defense-related functions. ZmGT-3b knockdown induced defense-related transcriptional reprogramming and increased biosynthesis of lignin without immune activation. These data suggest that ZmGT-3b is a regulator of plant growth–defense tradeoff that coordinates metabolism during growth-to-defense transitions by optimizing the temporal and spatial expression of photosynthesis- and defense-related genes. One-sentence summary ZmGT-3b regulates photosynthesis activity and synchronically suppresses defense response.

Introduction growth in young seedlings (Fig. 2, A-E), but improved resistance to F. graminearum infection. Both the shoot and 215 root growth phenotypes of GT-KD seedlings were much better than CK seedlings following inoculation, and the 216 disease severity index (DSI) of inoculated G3 and G6 seedlings was also markedly lower than that of inoculated CK 217 seedlings (Fig. 3, B-C). However, following field inoculation of mature maize plants, the DSI of GT-KD plants was 218 similar to (or higher than) that of CK plants (Fig. 3, D-E). Taken together, the finding that ZmGT-3b is only highly 219 expressed in a few young tissues (Supplemental Fig. S1B-C), and that young GT-KD seedlings show retarded growth 220 and improved disease resistance, suggests that ZmGT-3b is a positive regulator of growth and a negative regulator of 221 disease resistance in maize seedlings. 222 Unexpectedly, when we stopped watering the growing seedlings after two-leaf stage, all CK seedlings severely 223 wilted at 25-DAG, while GT-KD seedlings wilted less (see G7 seedlings in Fig. 4B), indicating that GT-KD seedlings 224 were drought tolerant. Transgenic G3 and G6 seedlings were significantly shorter, but G7 seedlings were not 225 significantly shorter, compared with CK ( Fig. 4A). We then tested the leaf water loss rate (WLR) and survival rate 226 (SR) of G3 and G7 seedlings (the growth of G7 seedlings was similar to CK under normal watering conditions). The 227 leaf WLR and SR of both lines were significantly lower than those of CK (Fig. 4, C-D). We estimated the 228 transpiration rate (TR) at the center of the widest part of the newest expanded leaf of each 15-DAG seedling (with 229 three leaves and a heart leaf) (Fig. 4F) from 09:00 to 12:00. The estimated TR range of CK seedlings was 0.  0.96 µmol H 2 O m −2 s −1 , while that of GT-KD seedlings was 0.664-0.818 µmol H 2 O m −2 s −1 . The average TR of G7 231 and G3 seedlings was ~22.12% and ~25.73% less than that of CK seedlings, respectively (Fig. 4E). These results 232 indicate that GT-KD seedlings with knocked down ZmGT-3b expression were more drought tolerant than CK 233 seedlings. 234 235

Photosynthesis-related Genes are Downregulated in the ZmGT-3b Knockdown Lines 236
To investigate the biological processes and genes regulated by ZmGT-3b, RNA-seq was used to compare the 237 transcriptomes of 7-DAG seedlings of GT-KD (referred to as GT) with CK. The correlation coefficient (R) for the 238 expression profiles of all transcripts between GT and CK was 0.87, suggesting that knockdown of ZmGT-3b affects 239 overall gene transcription. We compared the transcriptional responses of GT-KD with CK seedlings and identified 240 Of the downregulated DEG encoding proteins, 36 encode cellular components located in photosystems, 20 in 244 plastoglobules, 41 in photosynthetic membranes, and 37 in the plastid thylakoid. Gene Ontology (GO) and Kyoto 245 Encyclopedia of Genes and Genomes (KEGG) analysis revealed that the downregulated DEGs were significantly 246 enriched in photosynthesis-related functional categories, such as photosynthesis (44 DEGs), photosynthesis light 247 reaction (30 DEGs), photosynthesis light harvesting (17 DEGs), and photosynthesis antenna proteins (16 DEGs). 248 This suggests that ZmGT-3b is a positive regulator of photosynthesis-related processes (Fig. 5,Supplemental 249 Fig. S3). Therefore, the retarded growth of GT-KD seedlings might be caused by the transcriptional repression of 250 growth-promoting (photosynthesis-related) genes caused by ZmGT-3b knockdown. 251 Among the significantly downregulated genes induced by ZmGT-3b knockdown, the bZIP TF gene ZmHY5 252 (ELONGATED HYPOCOTYL5) showed an identical expression profile to ZmGT-3b; this gene was significantly 253 downregulated in non-inoculated GT-KD seedlings and upregulated in inoculated GT-KD seedlings (Fig. 6A). 254 Consistently, seven GT1 CONSENSUS (S000198) sites are discovered within 1500bp upstream of the start codon 255 ATG of ZmHY5, which is the conserved DNA binding site of GT factors (Supplemental Table S1). HY5 is a central 256 regulator of seedling development and light responses that promotes photomorphogenesis and mediates the positive 257 effect of shoot illumination on root growth Gangappa and Botto, 2016). This TF modulates 258 photosynthetic capacity by controlling the expression of chlorophyll biosynthesis-and photosynthesis-related genes 259 (Kobayashi et al., 2012;Toledo-Ortiz et al., 2014). 260 To verify the RNA sequencing data, we performed qRT-PCR to compare gene expression levels between 261 GT-KD and CK seedlings. ZmGT-3b, ZmHY5, and randomly selected photosynthesis-related genes such as ZmPSII3, 262 ZmLHCII, and ZmLHC117 were significantly downregulated in various GT-KD seedlings (Fig. 6B). Moreover, the 263 primary roots of CK seedlings grown in the light were significantly longer than those grown in the dark, whereas the 264 primary root lengths of GT-KD seedlings, which showed dramatically reduced 265 were similar in plants grown in the light and in the dark (Fig. 6, C-D). These results indicate that the knockdown of 266 ZmGT-3b led to reduced ZmHY5 expression, which disrupted the promotion of root growth via shoot illumination in 267 GT-KD seedlings. These findings are consistent with the reduced chlorophyll contents and photosynthetic rates of 268 GT-KD seedlings (Fig. 2, D-H). These results suggest that ZmGT-3b regulates seedling growth by modulating 269 chlorophyll biosynthesis and photosynthetic activity via the transcriptional regulation of photosynthesis-related 270

Activation 274
Most proteins encoded by the upregulated DEGs were associated with the membrane, cell periphery, or cytoplasmic 275 vesicle, and with molecular functions including catalytic activity, oxidoreductase activity, transferase activity, metal 276 ion binding, cation binding, and iron ion binding. The upregulated DEGs were enriched in oxidation-reduction 277 processes, secondary metabolic processes, plant-type cell wall organization, and reactive oxygen species (ROS) 278 metabolic processes. These upregulated DEGs significantly contribute to the biosynthesis of secondary metabolites, 279 especially the biosynthesis of phenylpropanoid, stilbenoid, diarylheptanoid, gingerol, benzoxazinoid, flavonoid, 280 diterpenoid, flavone, flavonol, and carotenoid (Fig. 5,. Almost all of these functional categories support basal 281 defense responses to various biotic/abiotic stresses, suggesting that ZmGT-3b acts as a negative regulator of plant 282 defense response-related biological processes. 283 Dramatic transcriptional reprogramming occurs during the induction of plant immune responses, allowing the 284 plant to prioritize defense-over growth-related cellular functions. The correlation coefficient (R) for the expression 285 profiles of all transcripts between the inoculated CK (CKi) and CK (CKi/CK) was 0.86, i.e., close to the value of 286 0.87 between GT and CK (GT/CK). This indicates that inoculation and ZmGT-3b knockdown (GT/CK) have similar 287 effects on general gene transcription. Compared with untreated CK seedlings, inoculation induced 1,026 DEGs, 288 including 239 DEGs that were simultaneously induced by ZmGT-3b knockdown. This indicates that overlapping 289 events or defense signaling pathways control gene expression in response to inoculation (CKi/CK) or ZmGT-3b 290 knockdown (GT/CK) (Fig. 5B). 291 The upregulated DEGs induced by inoculation (CKi/CK) were also significantly enriched in GO terms 292 associated with biosynthesis of secondary metabolites, especially biosynthesis of phenylpropanoid, stilbenoid, 293 diarylheptanoid, and gingerol, benzoxazinoid, flavonoid, diterpenoid, flavone and flavonol, carotenoid, 294 phenylalanine metabolism and plant-pathogen interaction. These results are similar to the transcriptome 295 reprogramming induced by ZmGT-3b knockdown (GT/CK), with many more DEGs enriched in photosynthesis, 296 RNA transport, and mRNA surveillance pathway (Supplemental Fig. S4). Therefore, considerable commonalities 297 were detected in the transcriptional reprogramming induced by ZmGT-3b knockdown or inoculation. ZmGT-3b 298 knockdown induced defense-related transcriptional reprogramming without immunity activation to upregulate or 299 activate basal defense-related genes. Therefore, ZmGT-3b might act as a transcriptional repressor of genes involved 300 in basal defense responses. 301 Furthermore, 1,049 genes were differentially expressed in inoculated GT-KD versus inoculated CK seedlings 302 (GTi/CKi) (Supplemental Fig. S5A). Of these, 254 DEGs were shared between the inoculated transcriptome pair 303 GTi/CKi and the non-inoculated transcriptome pair GT/CK (Supplemental Fig. S5B). The most highly enriched GO 304 functional categories were similar, and most highly enriched KEGG pathways were shared between the two 305 transcriptome pairs. The transcriptional differences between the non-inoculated GT/CK transcriptome emphasized in 306 downregulated photosynthesis-related processes. Most of the shared functional categories are involved in basal 307 defense responses to various biotic/abiotic stresses (Supplemental Fig. S5, C-D), pointing to considerable 308 commonalities in their transcriptional responses. 309 However, the R for the expression profiles of all transcripts between CKi and CK was 0.86, and that between 310 GTi and GT was 0.63, indicating that inoculation affected overall gene transcription to different degrees in the two 311 genotypes. Compared with non-inoculated seedlings, inoculation induced the expression of 1,026 and 1,340 DEGs in 312 CK and GT-KD, respectively, with 462 genes induced in both genotypes (Supplemental Fig. S5, C-D). This suggests 313 that overlapping signaling pathways control gene expression in these two types of seedlings in response to 314 inoculation. The transcriptional reprogramming was much stronger between inoculated and non-inoculated GT-KD 315 seedlings (GTi/ GT) than between inoculated and non-inoculated CK seedlings (CKi/CK). Many more DEGs were 316 enriched in almost all top defense-related GO and KEGG functional categories, except for biosynthesis of 317 brassinosteroids and zeatin, which had more DEGs enriched in CKi/CK (Supplemental Fig. S5, E-F). These results 318 are consistent with the notion that inoculated GT-KD have better disease resistance and lower DSI than CK 319 seedlings. 320 321

ZmGT-3b Knockdown Leads to the Upregulation of Defense-related Genes 322
ZmGT-3b knockdown dramatically upregulated genes that function in defense responses to various biotic/abiotic 323 stress, including genes encoding six PR proteins, three bZIPs, 18 MYBs, six WRKYs, eight NACs, eight 324 ethylene-response factors (ERFs, the largest group of the AP2/EREBP family), and three bHLHs, which were all 325 significantly upregulated in GT-KD versus CK seedlings (Supplemental Fig. S6). Many members of these TF 326 families reportedly function in plant responses to various biotic or abiotic stresses. Members of the MYB, ERF, bZIP, 327 and NAC TF families are well-known transcriptional regulators of genes required for drought tolerance, and MYBs 328 are important regulators of both secondary cell wall biosynthesis and abiotic stress tolerance (Mizoi et al., 2012;329 Baldoni et al., 2015). Among the TF genes upregulated by ZmGT-3b knockdown, some were significantly 330 upregulated in GT-KD seedlings without inoculation and in CK seedlings with inoculation, contrasting to 331 including ZmWRKY11, ZmWRKY69, ZmMYB36, ZmMYB93, ZmIBH1 (a basic helix-loop-helix, bHLH), ZmbHLH28, 332 ZmNAC67, ZmMYB8, ZmbZIP53, ZmbZIP7, and ZmbZIP8 (Fig. 7A). The upregulated expression of these TF genes 333 and a few well-known defense-related genes such as ZmDRR206, ZmPR1, and ZmPR-STH21 was confirmed by 334 qRT-PCR in different GT-KD seedlings with knocked down ZmGT-3b expression compared with CK seedlings (Fig.  335 7B). Therefore, various TF genes and defense-related genes were upregulated by ZmGT-3b knockdown. 336 337

ZmGT-3b Knockdown Increases the Biosynthesis of Cell Wall Components Independently of Immune 338
Activation 339 Based on the analysis of transcriptome reprogramming induced by ZmGT-3b knockdown (GT/CK), many proteins 340 encoded by the DEGs are involved in the biosynthesis of secondary metabolites, especially phenylpropanoid (Fig. 5, 341 C-D), which is the first critical step for lignin biosynthesis. Lignin is one of the most important secondary 342 metabolites and defense-induced lignin biosynthesis plays a major role in basal immunity. We therefore measured the 343 contents of the major components of the plant cell wall, such as cellulose, semi-cellulose, and lignin in seedlings. 344 Compared with CK seedlings, the contents of cellulose (~5.4% increase), semi-cellulose (~7.3% increase), acid 345 soluble lignin (ASL, ~22.7% increase), and lignin (~4.64% increase) were markedly higher in 12-DAG GT-KD 346 maize seedlings, whereas the content of acid-insoluble lignin was not (AIL) (Fig. 8A). Arabinose levels are 347 positively associated with lignin levels, and high concentrations of xylose are important in defense responses (Li et 348 al., 2015). The levels of both arabinose (~6.31% increase) and xylose (~6.7% increase) were significantly higher in (with FPKM>20) showed elevated expression levels in GT-KD seedlings, and four of these genes (with FPKM>50) 359 had FC>1.5 in GT-KD versus CK seedlings (Fig. 8E). These data suggest that ZmGT-3b knockdown promoted 360 secondary metabolite biosynthesis, especially lignin and cellulose biosynthesis, which occurred independently of 361 immune activation in maize seedlings. 362 Many genes upregulated in ZmGT-3b knockdown plants encode proteins involved in metal ion binding, cation 363 binding, or iron ion binding (Fig. 5, C-D). Thus, we measured the contents of mineral elements in CK and GT-KD 364 seedlings at 12-DAG. Compared with CK seedlings, potassium (K + ), phosphorus (P), and copper (Cu) levels were 365 significantly higher, while aluminum (Al) and iron (Fe) levels were significantly lower, in GT-KD seedlings. 366 However, Zn, Mg, and Na levels did not significantly differ in GT-KD vesus CK seedlings (Supplemental Fig. S7A). 367 Consistent with the different contents of various mineral elements, genes encoding transporters of these elements 368 showed considerably different transcript levels in GT-KD versus CK seedlings, including genes encoding phosphate, 369 potassium, copper, vacuolar iron, and zinc transporters. Finally, two sulfate transporter genes were significantly 370 upregulated in GT-KD versus CK seedlings (Supplemental Fig. S7B). These results suggest that ZmGT-3b 371 knockdown altered cellular osmotic conditions by increasing/decreasing the contents of various mineral elements by 372 affecting the take-up of inorganic ions from the environment using the corresponding transporters. 373

DISCUSSION 375
The growth-defense tradeoff in plants is associated with the limited availability of resources, which requires the 376 plant to prioritize growth or defense, depending on dynamic external and internal factors. This balance is important 377 for agriculture and natural ecosystems due to the vital importance of these processes for plant survival, reproduction, 378 and plant fitness (Huot et al., 2014). Numerous studies have revealed the crucial roles of TFs in regulating diverse 379 plant growth, development, and biotic or abiotic stress responses, including the trihelix TFs (Kaplan-Levy et al., 380 2012). Here, we showed that ZmGT-3b transcript levels were significantly decreased during the interaction of maize 381 seedlings with F. graminearum. In addition, ZmGT-3b knockdown in maize seedlings led to retarded growth, 382 improved resistance to F. graminearum, and enhanced drought tolerance ( Fig. 2-4). Moreover, ZmGT-3b is 383 associated with the positive regulation of photosynthesis-related genes and the negative regulation of defense-related 384 genes, as ZmGT-3b knockdown severely downregulated the expression of photosynthesis-related genes and 385 significantly upregulated the expression of defense-related genes at the same time ( 5-6). These findings suggest that ZmGT-3b involves in pathogen attack induced suppression of photosynthesis 387 activity, and coordinates metabolism during growth-to-defense transitions by optimizing the temporal and spatial 388 expression of photosynthesis-and defense-related genes, uncovering a molecular mechanism underlying the growth-389 defense tradeoff. 390 391

ZmGT-3b Modulates Seedling Growth via the Transcriptional Regulation of Photosynthesis-related Genes 392
The conserved DNA-binding domains of plant-specific trihelix TFs are often found at the N terminus. By contrast, 393 the C terminus, which harbors a hydrophilic region, is less conserved and probably acts as the activation domain 394 (Qin et al., 2014;Kaplan-Levy et al., 2014). Trihelix TFs control the transcription of light-regulated genes as well as 395 genes involved in plant development and stress responses (Kaplan-Levy et al., 2012). During these transcriptional 396 regulatory processes, trihelix TFs form homodimers or heterodimers by participating in interactions with other 397 classes of TFs . Consistent with the light-inducible expression profile of ZmGT-3b, 398 we identified 19 light-responsive elements in the promoter of this gene (Fig. 1A, Supplemental Table. S1). 399 We obtained GT-KD mutants with severely reduced ZmGT-3b levels by transforming maize inbred line 400 B73-329 with a cDNA fragment encoding the C-terminal 149 aa of ZmGT-3b under the control of the maize 401 Ubiquitin promoter. This mutant might produce protein without the conserved N-terminal DNA-binding domains, 402 ultimately leading to ZmGT-3b knockdown. These plants exhibited diminished growth at the seedling stage, with 403 shorter primary roots and reduced seedling height (Fig. 2, A-D). However, ZmGT-3b knockdown had no effect on 404 the growth of mature maize plants in the field (Supplemental Fig. S2). The effect of ZmGT-3b knockdown on 405 seedling growth might be associated with the expression profile of ZmGT-3b (primarily expressed in a few specific 406 young tissues, such as primary roots and ear primordium) and the elements for meristem and root expression in its 407 promoter (Supplemental Fig. S1, Table S1). 408 Light perception activates many TFs from various families, such as bZIP, bHLH, MYB, GATA, and GT1. 409 These TFs bind to various LREs such as G, GT1, GATA, and MREs, leading to massive transcriptional 410 reprogramming (Gangappa andBotto, 2014, 2016). Among these TFs, the role of HY5 as a master transcriptional 411 regulator is conserved across plant species. HY5 mediates the light-responsive coupling of shoot growth and 412 photosynthesis with root growth and nitrate uptake, and it functions as the center of a transcriptional network hub 413 connecting different processes such as hormone, nutrient, abiotic stress (abscisic acid, salt, cold), and ROS signaling 414 pathways Gangappa and Botto 2016;. HY5 activates its own expression and 415 is a critical player in seedling development and responses to light by turning on or off many genes involved in 416 fundamental developmental processes such as cell elongation, pigment accumulation (chlorophyll and anthocyanin), interacting partners, as HY5 does not have its own activation or repression domain. The primary function of HY5 in 420 promoting transcription may depend on other, likely light-regulated, factors .

ZmGT-3b Knockdown Induces Defense Responses without Immune Activation by Regulating Lignin 436 biosynthesis and Defense-related Gene Expression 437
Diminished growth is thought to be an integral facet of induced resistance and a molecular mechanism involved in 438 the crosstalk between growth and defense responses. This process involves optimization of the temporal and spatial 439 expression of defense genes. Pathogen infection usually affects primary metabolism, reduces plant growth, limits 440 photosynthesis, and modifies secondary metabolism towards defense responses . Consistently, 441 ZmGT-3b knockdown led to diminished seedling growth and reduced Pn, but improved disease resistance and 442 drought tolerance, and increased lignin contents in young seedlings (Fig. 2, 3 Many upregulated genes encoded multiple members of the MYB, bZIP, WRKY, NAC, ERF, and bHLHs TF 448 families (Supplemental Fig. S6). Some members of these TF families are well-known regulators of lignin 449 biosynthesis, such as NAC-MYB-GRN, which regulate lignin biosynthesis in both dicot and monocot species (Yoon 450 et al., 2015). Five maize MYB TFs (ZmMYB2, ZmMYB8, ZmMYB31, ZmMYB39, ZmMYB42) function in lignin 451 biosynthesis by controlling ZmCOMT expression . Overexpression of ZmMYB167 increased 452 lignin content to 13% in maize without affecting plant growth or development . Besides MYB 453 TFs, some NAC  and WRKY TFs also regulate lignin biosynthesis by modulating the expression 454 of cell wall synthesis-related genes (Gallego- Giraldo et al., 2016). 455 The plant cell wall is a highly organized structure composed of lignin, cellulose, semi-cellulose, pectin, proteins,  ZmGT-3b knockdown led to the significant upregulation of a subset of secondary metabolite biosynthesis-related 466 genes, especially genes encoding lignin biosynthesis enzymes, including PALs, 4CLs, HCTs, CoCOMTs, CCR, 467 COMTs, LACs, DPs, PODs, and CASPs, in GT-KD versus CK seedlings (Fig. 8, Supplemental Fig. S6). The 468 mutation of ZmCOMT, encoding a caffeic acid O-methyl transferase, leads to the brown midrib3 phenotype (Fornale 469 et al. 2006). ZmCOMT transcript levels were high in maize seedlings, with FPKM values over 1,200; its transcript 470 level increased 1.57-fold in GT-KD versus CK seedlings. ZmMYB19/Zm1 is an ortholog of the Arabidopsis SG3-type 471 R2R3-MYB genes MYB58 and MYB63. These TFs transactivate the promoters of monolignol pathway genes, and 472 their overexpression specifically activates the monolignol pathway and lignin accumulation at the expense of biomass 473 production ). Among the 18 MYB genes with significantly upregulated expression in GT-KD versus 474 CK seedlings, ZmMYB19/Zm1 expression increased 2.04-fold in response to ZmGT-3b knockdown (Supplemental Fig.  475 S6). The upregulation of these genes might contribute to the increased lignin and ASL contents of GT-KD seedlings 476 (Fig. 8A). 477 Lignin biosynthesis is coordinately regulated with the biosynthesis of other cell wall components. The deposition 478 of cellulose in cell walls is vital for controlling cell growth Ohtani and Demura 479 2019). Consistent with the retarded growth of young GT-KD seedlings, the cellulose and semi-cellulose contents were 480 also higher in these seedlings compared with CK seedlings. Indeed, the expression levels of all 22 Cesa-encoding 481 genes were higher GT-KD versus CK seedlings, although not all differences were significant (Fig. 8E). These results 482 suggest that ZmGT-3b knockdown induces defense responses without activating the immune system by 483 transcriptionally regulating basal defense-and lignin biosynthesis-related genes in maize seedlings. Therefore, 484 ZmGT-3b might be a negative regulator of biological processes related to plant defense responses, thus acting as a 485 molecular hub connecting developmental/environmental signals with lignin biosynthesis during maize seedling 486 growth. to various abiotic or biotic stresses, including drought stress, as increased K + uptake confers higher levels of drought 501 tolerance . Phosphorus (P) is an indispensable nutrient for plant growth and 502 development, as it is a constituent of many important molecules such as nucleic acids, phospholipids, and ATP (Guo 503 et al., 2015). Cu is critical for electron transport and for scavenging ROS produced in chloroplasts during 504 photosynthesis under stress conditions . 505 ZmGT-3b knockdown significantly increased K, P, and Cu contents in GT-KD seedlings, and the corresponding 506 transporter genes were upregulated in these seedlings compared with CK seedlings (Supplemental Fig. S7). Two 507 sulfate transporter genes were also significantly upregulated in GT-KD seedlings. Sulfate transporters are important 508 for plant drought and salinity tolerance, as sulfate accumulation in leaves enhances ABA biosynthesis, leading to 509 stomatal closure (Gallardo et al., 2014). The increased accumulation of sulfate, P, K, and Cu in the absence of stress 510 might contribute to the enhanced drought tolerance of GT-KD seedlings. Fe is involved in various chelation and 511 oxidation/reduction steps that affect ROS production because it is a component of all photosystems and a critical 512 redox-active metal ion in photosynthetic electron flow. Therefore, Fe homeostasis must be fine-tuned in plants (Briat 513 et al., 2007). Al is toxic to plants and seriously affects plant growth and productivity . Finally, since 514 lignin is a component of the cell wall and the first barrier for metal ions, lignin biosynthesis is associated with heavy 515 metal absorption, transport, and tolerance in plants. Lignin binds to multiple heavy metal ions and reduce their entry 516 into the cytoplasm due to its numerous functional groups (e.g., hydroxyl, carboxyl, and methoxyl) (Dalcorso et al., 517 2010). The significantly reduced Al and Fe contents in GT-KD versus CK seedlings might be associated with 518 increased lignin biosynthesis in response to ZmGT-3b knockdown, although one vacuolar iron transporter gene was 519 significantly upregulated in these seedlings, and no Al-related transporter genes were identified in GT-KD seedlings. ). MYB15 plays a role in the complex regulatory relationship between lignin, growth, and defense (Chezem et 532 al., 2017). 533 Consistent with our observation of a reduced TR and increased lignin contents in GT-KD seedlings (Fig. 4E, Fig.  534 8A), ZmGT-3b knockdown induced the significant upregulation of multiple TF genes in GT-KD seedlings, including 535 multiple members of the bZIP, MYB, WRKY, NAC, ERF, and bHLH families (Fig. 7). Considering that these TFs 536 are critical in the complex regulatory relationship between growth, lignin, and defense, their elevated expression 537 might be associated with the improved disease resistance and drought tolerance of GT-KD seedlings. Collectively, 538 the reinforced cell wall with increased lignin content, the increased accumulation of various mineral elements, and 539 the significantly upregulated expression of various defense-related TFs might all contribute to the drought tolerance 540 of GT-KD seedlings. 541 542

ZmGT-3b Mediates the Growth-defense Tradeoff in Maize Seedlings 543
Based on our results, we propose a model for the mode of action of ZmGT-3b, as shown in Fig 9. According to this 544 model, under normal growth conditions, light induces the expression of ZmGT-3b. ZmGT-3b may act upstream or as 545 an activator of ZmHY5 to activate various photosynthesis-related genes, thereby promoting photosynthesis and 546 seedling growth; ZmGT-3b synchronically acts as a transcriptional repressor of the expression of multiple TF genes, 547 including MYBs, bZIPs, NACs, bHLHs, and ERFs, which in turn repress the expression of various defense-related 548 genes. When plants are exposed to pathogen attack, ZmGT-3b and ZmHY5 expression were dramatically decreased, 549 thus relieving the repressive effects of these TF genes and enhancing the expression of defense-related genes, 550 including genes encoding PR-proteins and various enzymes involved in the biosynthesis of secondary metabolites 551 especially lignin, thereby activating the defense response. Reduced ZmGT-3b expression also leads to a decrease in 552 photosynthesis-related activities to benefit defense-related biological processes. Therefore, perhaps ZmGT-3b and 553 ZmHY5 coordinately regulate the light response and photosynthesis during maize seedling growth, and the 554 interaction of ZmGT-3b with other TFs might be important in its diverse regulatory functions. In conclusion, we 555 propose that ZmGT-3b functions as a transcriptional regulator that calibrates the balance between plant growth and 556 defense responses by coordinating metabolism during growth-to-defense transitions by optimizing the temporal and 557 spatial expression of photosynthesis-and defense-related genes. Therefore, ZmGT-3b might serve as a molecular 558 hub connecting developmental or environmental signals with secondary metabolite biosynthesis. 559

Plant Growth, F. graminearum Inoculation and Disease Severity Scoring. 563
The fungal pathogen Fusarium graminearum preparation and inoculation with F. graminearum in the field were 564 done according to Yang et al. (2010); young seedling inoculation on primary roots was done according to Ye et al. 565 (2013); and disease severity scoring was done according to Ye et al. (2018). Three replicates were set for each 566 genotype with about 25 plants per replicate. The primary roots with typical symptoms were scored at 48 hours after 567 inoculation (hai). The length of the 7-days after germination (DAG) young seedling primary roots cultured with 568 paper-rolling were measured and used for comparing seedling root growth rate, and shoot growth rate was measured 569 with the soil-growth young seedlings at 12 or 15 days after germination. The seedlings (CK and GT-KD) were 570 cultured in controlled growth room conditions of 28/22°C (day/night) at a light intensity of 500 μmolm −2 s −1 571 (16-h-light/8-h-night) and 40-50% relative humidity. Statistical analysis was conducted using Student's t-test 572 between the control (CK) and mutant GT-KD seedlings to determine statistical significance in three independent 573 experiments. 574 575

Generation of the Transgenic Knockdown Lines of ZmGT-3b 576
According to the maize genome sequence RefGen V3.22 in 2014, the third exon (encoding 149aa) of ZmGT-3b was 577 obtained by RT-PCR and was put under the control of maize Ubiquitin promoter in a pBXCUN-derived binary vector 578 to generate pUbi::cZmGT-3b (for primer sequences see Supplemental Table S2), the construct was transformed into 579 Agrobacterium strain EHA105, and then into the immature embryos of the Zea mays L. B73-329 inbred lines (used 580 as the control, CK, in the afterward experiments). Six transgenic events of the construct were obtained, self-crossed 581 and enough homogenous seeds were harvested. 582 583

Plasmid Construction and Subcellular Localization Analysis 584
The full coding sequence (cds) of ZmGT-3b was obtained by RT-PCR with gene-specific primers (designed 585 according to the B73 reference genome RefGen V4.32, primer sequences see Supplemental Table S2) amplified with 586 reverse-transcription cDNA templates from maize young seedlings at 7-DAG. Subsequently, the sequenced clone 587 was used for constructing p1300-35S: ZmGT-3b-GFP vector without the stop codon, with a pCAMBIA1300-derived 588 binary vector by the introduction of the cds to fuse to GFP driven by the Camv35S promoter. Agrobacterium strain 589 EHA105 containing p1300-35S: ZmGT-3b-GFP vector was cultured at 28°C overnight. Bacteria cells were harvested 590 by centrifugation and resuspended with buffer (10mM MES, pH 5.7, 10mM MgCl2, and 200 mM acetosyringone) at 591 OD 600 = 0.6. Leaves of 3-week-old soil-grown N. benthamiana were infiltrated with Agrobacterium cultures carrying 592 a binary vector p1300-35S::GFP or p1300-35S::ZmGT-3b-GFP expressing GFP or ZmGT-3b-GFP. The plants were 593 incubated under 16h light/8h dark at 25°C in growth chambers. The GFP fluorescence signals were detected 2-days 594 post-infiltration (dpi). The Agrobacterium cultures carrying the binary vectors were also used to transform onion 595 epidermal cells. Fluorescence images were examined and taken with LSM 880 confocal laser microscope systems 596 and images were processed using LSM microscope imaging software. The excitation laser of 488 nm was used for 597 imaging GFP signals. 598 599

Maize Seedling Drought Stress Analysis 600
After germination, the seedlings (CK and GT-KD) were cultured in controlled growth room conditions of 28/22°C 601 (day/night) at a light intensity of 500 μmolm −2 s −1 (16-h-light/8-h-night) and 40-50% relative humidity. They were 602 grown under well-watered conditions by maintaining soil water content close to field capacity for approximately 603 10 days until drought treatment. Drought stress (cessation of watering) was imposed on the growing seedlings after 604 two leaves; that is, at about 10 days after sowing by withdrawing water supply and keeping the plants under 605 observation for the following 15 days, when indications of severe withering symptom (all the leaves turned soft, 606 drooping and dried) were visible in almost all of the CK seedlings, then the seedlings were re-watered for 6 days. 607 Measurements were made at day 15 after the stress treatment and at day 6 following the start of re-watering. For 608 each treatment, the normally irrigated plants were used as controls. The phenotypes and physiological indexes of the 609 seedlings were detected, and the number of surviving seedlings and the total seedling number to obtain the survival 610 rate were checked. The water loss rate test was done with the third leaf of well-grown seedlings at 15 DAG; the 611 leaves were collected and put on to flat pallets, separately and individually under the same environment as the 612 seedlings were growing. The weight of the five leaves was measured as a group at given time; the weight of the lost 613 water was obtained by subtracting this weight from the fresh weight, and the water loss rate was calculated as a 614 percentage of the weight of the lost water to the initial fresh weight of the given group. Data are means ± SD of three 615 replicates. 616 617

Analysis of Cell Wall Components, Cellulose and Lignin Content 618
The 12-day-old seedlings were harvested, dried and homogenized to a fine powder using a mixer mill (MM400, 619 Retsch Technology, Haan, Germany) at 25 Hz for two minutes. One-hundred milligrams of powdered seedling tissue 620 was sequentially ultrasonicated for 15 min in a mixture of twice with 1 mL methanol, twice with phosphate buffered 621 saline pH (7.0) containing 0.1% (v/v) Tween 20, twice with 1 mL 95% ethanol, twice with 1 mL (1:1) chloroform: 622 methanol and twice with 1 mL acetone. The samples were centrifuged at 16,000 g for 10 min and the pellets dried at 623 50°C. The remaining cell wall extract was used for determination of total lignin content. The lignin content of 624 seedlings was quantified using the acetyl bromide soluble lignin method. Seedling tissue was macerated in 72% (v/v) 625 sulfuric acid for 2 h, diluted with 112 mL deionized water, and thereafter autoclaved at 121°C for 1 h. The 626 acid-insoluble lignin was quantified using pre-weighed medium coarse-ness glass crucibles, while a UV/VIS 627 Spectrometer was used to determine the acid-soluble lignin content at 205 nm with an ultraviolet spectrophotometer 628 (TU-1901). To hydrolyze the cell wall polysaccharides, 10 mg of destarched sample was mixed with 200 µl 72% 629 sulfuric acid and incubated at 60°C for 1h, and then the sulfuric acid was diluted to 3% with distilled water for 630 hydrolysis at 121°C for 1 h. After cooling to room temperature, the supernatant was collected, erythritol was added 631 as an external standard, and then was neutralized with barium carbonate. The sugars in the supernatant were 632 separated using an SP0810 column (Shodex) on a UHPLC system (Agilent-1260). The content of the detected sugars 633 was calculated based on standard curves of glucose, xylose, mannose, galactose and arabinose. Error bars indicate 634 SD of three biological replicates. 635 636

Leaf Chlorophyll Content, Net Photosynthetic Rate and Transpiration Rate Analysis 637
Leaf chlorophyll content was measured from the latest expanded leaves of the CK and GT-KD seedlings grew at 12-638 or 15-day after germination (DAG), with a SPAD meter (SPAD-502 Plus, Konica Minolta, Inc. Tokyo, Japan) under 639 a saturating actinic light (660 nm) with an intensity of 1100 µmol m -2 s -1 . The middle widest part of the latest 640 expanded leaf of every seedling, that is the second leaf of the12-DAG seedlings and the third leaf of the 15-DAG 641 seedlings, was used for SPAD value measurement. The net photosynthetic rate (Pn) and transpiration rate were 642 measured from the latest expanded leaf (the third leaf) of the 15-DAG seedlings (with three leaves and a heart leaf) 643 with a portable LI-6400XT Portable Photosynthesis System (LI-COR, USA), recorded at a saturating actinic light 644 (660 nm) with an intensity of 1100 µmol m -2 s -1 , at the time from 09:00 to 12:00 in the morning. All measurements 645 were conducted on the middle part of the latest expanded leaves following the manufacturer's instructions. Five 646 replicates were randomly taken for each genotype. 647

RNA Extraction and Transcriptome Sequencing 649
To compare the transcriptomes between GT-KD and CK (Control, B73-329) with or without inoculation, the 650 inoculated (the seedlings were inoculated at 5-DAG and sampled at 18 h after inoculation) and the non-inoculated 651 GT-KD and CK seedling samples (whole seedlings without the kernel) were collected at the same time (at 18 hai) 652 and used for RNA extraction and deep sequencing. Total RNA was extracted using RNAiso Plus (Takara Bio) 653 according to the user manual. Three micrograms of total RNA from each sample were used for transcriptome 654 sequencing at Novogene (http://www.novogene.com/). Sequencing was performed on each library to generate 655 100-bp paired-end reads employing the high throughput sequencing platform highseq3000. Read quality was 656 checked using FastQC and low quality reads were trimmed using Trimmomatic version 0.32 657 (http://www.usadellab.org/cms/?page=trimmomatic). The clean data for each sample amounted to ~6 Gb. The clean 658 reads were aligned to the masked maize genome database for mapping, calculation, and normalization of gene 659 expression (the updated Z. mays B73 reference genome AGPv4, http://ensembl.gramene.org/Zea_mays/Info/Index).  Table S2. PCR was performed with the following conditions: 94℃ for 2 min; 40 cycles of 94℃ for 30 683 s, 60℃ for 30 s, 72℃ for 30 s. The relative expression levels were calculated using the relative quantization 684 according to the quantification method (2 -ΔΔCt ) (Livak and Schmittgen, 2001) and plotted with standard errors. The 685 variation in expression was estimated using three biological replicates independently by comparative reverse 686 transcription-quantitative PCR (RT-qPCR). 687 688

Statistical Analyses 689
Statistical analysis was performed using the paired Student's t-test. All values represent the mean ± SD. Asterisks 690 indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001). All experiments were independently repeated 691 at least three times. ZmGT-3b, ZmLHC117, and ZmLHCII is induced by light. 'Dark' represents control seedlings (CK, maize inbred line 697 B73-329, wild type) that were grown in the dark for 5 days; L1 and L2 represent CK seedlings transferred to the light 698 for 1 h and 2 h after 5 days of growth in the dark, respectively. Error bars denote the mean ± SD of n = 3 replicates. maize seedlings is significantly reduced compared with CK (maize inbred line B73-329, wild type) at 7 days after 707 germination (DAG). G3, G4, G6, and G7 represent seedlings from four transgenic events in which maize was 708 transformed with the partial coding sequence of the C-terminal part of ZmGT-3b. Seedling phenotypes (B) and 709 average primary root lengths (C) of 7-DAG maize seedlings. The average seedling height (D) and shoot phenotype 710 (E for 12-DAG and F for 15-DAG) of soil-cultured maize seedlings grown in a greenhouse under natural light. 711 Twelve-DAG seedlings with two leaves and a heart leaf are denoted as CK-2, G3-2, and G6-2, and 15-DAG 712 seedlings with three leaves and a heart leaf are denoted as CK-3, G3-3, and G6-3 (used to measure net 713 photosynthetic rate). The SPAD values (G) of the above seedlings were obtained from the central widest part of the 714 newest expanded leaf, that is, the second leaf of 12-DAG seedlings or the third leaf of 15-DAG seedlings. (H) The 715 net photosynthetic rate (μmol CO₂·m −2 ·s −1 ) of the central widest part of the third leaf of 15-DAG seedlings (indicated 716 by an arrow in (F)) was measured from 09:00 to 12:00 under natural light in a greenhouse. Error bars denote the 717 mean ± SD of three biological replicates. The asterisk indicates a statistically significant difference between CK and 718 the GT-KD lines, as calculated by a paired Student's t-test (* at P < 0.05, ** at P < 0.01, *** at P < 0.001). Wilting (upper) and survival (bottom) of CK (maize inbred line B73-329, wild type) and G7 seedlings. Water was 733 withheld from growing seedlings at the two-leaf stage, and the plants were re-watered when all CK seedlings were 734 severely wilted (at 25-DAG). C, Leaf water loss rates of young maize (Zea mays) seedlings. The leaf water loss rates 735 are shown as the means of the percentage of leaf water loss ± SD (n = 5). Three independent experiments were 736 performed. (D-F) The survival rate (D), transpiration rate (TR, E), and seedling leaves (F) used for TR 737 measurements. The estimated TR was obtained from the central widest part of the third leaf, which is the newest 738 expanded leaf of 15-DAG CK or GT-KD seedlings (with three leaves and a heart leaf). The TR value was measured 739 from 09:00 to 12:00 under natural light in a greenhouse, TR values are the mean ± SD (n=6). The asterisks indicate a 740 statistically significant difference between CK and the GT-KD lines, as calculated by a paired Student's t test (* at p 741 < 0.05, ** at P < 0.01 and *** at P < 0.001). Values are the mean ± SD (n = 3). The asterisks represent a significant difference at * P < 0.05, ** P < 0.01 785 (according to a paired Student's t-test). (C) The general lignin biosynthesis pathway, including the related enzymes. were developed from a QTL qRfg1 on chromosome 10 that could explain 36.6% of the total variations of maize 810 resistance to Fusarium graminearum induced stalk rot . RCK (SCK) is the primary roots of R-NIL 811 (S-NIL) without inoculation, R6, R18, R48 (S6, S18, S48) was the primary roots of R-NIL (S-NIL) inoculated after 812 6h,18h and 48h, respectively . B and C, ZmGT-3b expressed highly in normal seedling roots, and only 813 expressed in a few kinds of young tissues, such as primary roots, ear primordium (2-8µm), embryo at 20 DAP and seven day-after-germination (7-DAG). Compared with CK seedlings, the content of Al and Fe was significantly 850 decreased, while the content of Cu, K and P was increased in the GT-KD seedlings. DW, Dry weight. Values are the 851 mean ± SD. The asterisk * represents a significant difference at P < 0.05 (according to a paired Student's t-test); NS, 852 not significant. B, The relative expression fold of the related mineral element transporter encoding genes from the 853 transcriptome sequencing with ZmGT-3b knock-down (GT) and CK (B73-329) maize seedlings at 7-DAG, with 854 (CKi, GTi) or without inoculation (CK, GT).