Arabidopsis Cys2/His2 Zinc Finger Transcription Factor ZAT18 Modulates the Plant Growth-Defense Tradeoff

Plant defense responses under unfavorable conditions are often associated with reduced growth. However, the mechanisms underlying the growth-defense tradeoff remain to be fully elucidated, especially at the transcriptional level. Here, we revealed a Cys2/His2-type zinc finger transcription factor, namely, ZAT18, which played dual roles in plant immunity and growth by oppositely regulating the signaling of defense- and growth-related hormones. ZAT18 was first identified as a salicylic acid (SA)-inducible gene and was required for plant responses to SA in this study. In addition, we observed that ZAT18 enhanced the plant immunity with growth penalties that may have been achieved by activating SA signaling and repressing auxin signaling. Further transcriptome analysis of the zat18 mutant showed that the biological pathways of defense-related hormones, including SA, ethylene and abscisic acid, were repressed and that the biological pathways of auxin and cytokinin, which are growth-related hormones, were activated by abolishing the function of ZAT18. The ZAT18-mediated regulation of hormone signaling was further confirmed using qRT-PCR. Our results explored a mechanism by which plants handle defense and growth at the transcriptional level under stress conditions.


Introduction
As sessile organisms, plants must struggle against various pathogens and herbivores, including bacteria, fungi, viruses and insects. To fight these invasions, plants activate defense responses, including pattern-triggered immunity (PTI) and effector-triggered immunity (ETI), to protect themselves [1,2]. However, constitutive activation of defense responses is an energy-consuming process and generally results in growth penalties [3][4][5][6]. For example, prolonged treatment with pathogen-associated molecular patterns (PAMPs), which activates PTI responses, significantly inhibits Arabidopsis growth [7]. This phenomenon is called the growth-defense tradeoff. Understanding the mechanisms underlying the growth-defense tradeoff is critical for breeding disease-tolerant crops. A good example is the study of TL1-binding transcription factor 1 (TBF1). The translation of TBF1 is controlled by upstream open reading frames (uORFs), which allows TBF1 to only be translated to activate immune responses and repress growth upon pathogen infection [8]. The uORFs of TBF1 were later used as a key regulatory element to balance rice immunity and growth in molecular breeding [9].
Numerous works verified that hormone crosstalk plays fundamental roles in regulating the growth-defense tradeoff [10][11][12]. Among all hormones, the accumulation of salicylic acid (SA) is observed during both PTI and ETI and plays a key role in plant defenses against biotrophic and hemibiotrophic pathogens [13]. Upon pathogen infection, SA is synthesized through the ICS pathway [14,15]. Then, the accumulated SA is detected by NPR proteins, inducing the expression of SA-responsive genes through TGA transcription factors [16,17].

ZAT18 Mediated Plant Responses to SA
To further study SA signaling, we performed transcriptome analysis on samples that were treated with SA for 1 h. Three biological samples were collected for both CK and SA treatment and we observed that ZAT18 was significantly induced in two SA-treated samples ( Figure S1). The SA-induced expression pattern of ZAT18 was further confirmed using qRT-PCR. We found that SA upregulated the expression of ZAT18 at all time points we checked (1 to 24 h after the SA treatment) ( Figure 1A). As NPR1 is the SA receptor that controls the expression of most SA-responsive genes and TGA2/5/6 are transcription factors that work downstream of NPR1 [32,33], we examined whether SA activates ZAT18 in an NPR1-and TGA-dependent manner. As shown in Figure 1B, the SA-induced expression of ZAT18 showed no significant difference between the WT and npr1-2 or tga2/5/6 mutants, indicating that the SA induction of ZAT18 was independent of NPR1 and TGA2/5/6. zothiadiazole (BTH), which is a chemical analog of SA and induces plant defense that is dependent on SA signaling [33]. One day later, the plants were further infiltrated with Pseudomonas syringae pv.maculicola (Psm) ES4326. The pathogen levels in the WT and zat18-1 were measured at three days after the infection. We observed that pretreatment with BTH enhanced the plant resistance to Psm ES4326 in the WT, where the bacteria titer reduced from 10 6.02 to 10 5.04 , but not in the npr1-2 mutant as expected. However, the BTHtriggered resistance was also significantly compromised in the zat18-1 mutant ( Figure 1D).

Figure 1.
ZAT18 is an SA-induced gene that encodes a protein that is important for SA responses. (A) Two-week-old wild-type (WT) seedlings were dipped with 1 mM SA, and samples were collected at the indicated time points to examine the expression levels of ZAT18 with UBQ5 as a reference. Significant differences were detected using Student's t-test. Data are shown as mean ± s.d. (n = 3, n indicates the number of biological replicates). (B) The expression levels of ZAT18 in the WT and npr1-2 and tga2/5/6 mutants after sterile dH2O (CK) or 1 mM SA treatment for 1 h. Experiments were performed as described in (A). Significant differences were detected using two-way ANOVA. Data are shown as mean ± s.d. (n = 3, n indicates the number of biological replicates). (C) The expression levels of WRKY54 and WRKY70 in the WT and zat18-1 mutant after sterile dH2O (CK) or 1 mM SA treatment for 1 h. Experiments were performed as described in (A). n = 3, n indicates the number of biological replicates. (D) Three-week-old WT and zat18-1 and npr1-2 mutants were pre- Figure 1. ZAT18 is an SA-induced gene that encodes a protein that is important for SA responses. (A) Two-week-old wild-type (WT) seedlings were dipped with 1 mM SA, and samples were collected at the indicated time points to examine the expression levels of ZAT18 with UBQ5 as a reference. Significant differences were detected using Student's t-test. Data are shown as mean ± s.d. (n = 3, n indicates the number of biological replicates). (B) The expression levels of ZAT18 in the WT and npr1-2 and tga2/5/6 mutants after sterile dH 2 O (CK) or 1 mM SA treatment for 1 h. Experiments were performed as described in (A). Significant differences were detected using two-way ANOVA. Data are shown as mean ± s.d. (n = 3, n indicates the number of biological replicates). (C) The expression levels of WRKY54 and WRKY70 in the WT and zat18-1 mutant after sterile dH 2 O (CK) or 1 mM SA treatment for 1 h. Experiments were performed as described in (A). n = 3, n indicates the number of biological replicates. (D) Three-week-old WT and zat18-1 and npr1-2 mutants were pre-treated with dH 2 O (CK) or 120 µM BTH for 1 day before infiltration with Psm ES4326 (OD 600nm = 0.001). The bacterial titer was checked at 3 days post-inoculation (dpi). cfu, colony forming unit. n = 8, n indicates the number of biological replicates. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001; ns, no significant difference. Then, we wondered whether ZAT18 could mediate the plant responses to SA. For this purpose, a T-DNA insertion ZAT18 knockout mutant was identified ( Figure S2). Twoweek-old seedlings of the WT and zat18 mutant (zat18-1) were treated with 1 mM SA for 1 h to measure the expression of the SA-responsive genes, namely, WRKY54 and WRKY70. We found that the ZAT18 mutation significantly dampened the SA-induced expression of these genes ( Figure 1C), indicating that ZAT18 was required for the plants to fully respond to SA. Next, we tested whether ZAT18 modulates SA-triggered plant resistance to pathogens. For this purpose, three-week-old WT and zat18-1 were pretreated with 120 µM benzothiadiazole (BTH), which is a chemical analog of SA and induces plant defense that is dependent on SA signaling [33]. One day later, the plants were further infiltrated with Pseudomonas syringae pv.maculicola (Psm) ES4326. The pathogen levels in the WT and zat18-1 were measured at three days after the infection. We observed that pretreatment with BTH enhanced the plant resistance to Psm ES4326 in the WT, where the bacteria titer reduced from 10 6.02 to 10 5.04 , but not in the npr1-2 mutant as expected. However, the BTH-triggered resistance was also significantly compromised in the zat18-1 mutant ( Figure 1D).
To confirm the role of ZAT18 in SA signaling, we further mutated ZAT18 using the Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)-associated 9 (CRISPR/Cas9) genome editing tool [34]. A 20 nt specific spacer against the first exon of ZAT18 with the AGG protospacer-adjacent motif (PAM) sequence was chosen using the CRISPR program (http://crispor.tefor.net/, accessed on 13 December 2020) ( Figure S2A). Successfully edited zat18 mutants were identified in the T 1 generation. A homozygous line with an A insertion that resulted in early termination of ZAT18 translation was chosen for further experiments and was termed zat18cr (Figure 2A,B). We observed that the SA induction of WRKY54 and WRKY70 was dampened in zat18cr ( Figure 2C). Furthermore, BTH-triggered resistance to Psm ES4326 was also abolished in this mutant ( Figure 2D). All these results indicated that ZAT18 played an important role in the SA signaling. treated with dH2O (CK) or 120 μM BTH for 1 day before infiltration with Psm ES4326 (OD600 nm = 0.001). The bacterial titer was checked at 3 days post-inoculation (dpi). cfu, colony forming unit. n = 8, n indicates the number of biological replicates. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001; ns, no significant difference.
To confirm the role of ZAT18 in SA signaling, we further mutated ZAT18 using the Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)-associated 9 (CRISPR/Cas9) genome editing tool [34]. A 20 nt specific spacer against the first exon of ZAT18 with the AGG protospacer-adjacent motif (PAM) sequence was chosen using the CRISPR program (http://crispor.tefor.net/, accessed on 13 December 2020) ( Figure S2A). Successfully edited zat18 mutants were identified in the T1 generation. A homozygous line with an A insertion that resulted in early termination of ZAT18 translation was chosen for further experiments and was termed zat18cr (Figure 2A,B). We observed that the SA induction of WRKY54 and WRKY70 was dampened in zat18cr ( Figure 2C). Furthermore, BTH-triggered resistance to Psm ES4326 was also abolished in this mutant ( Figure 2D). All these results indicated that ZAT18 played an important role in the SA signaling.  Figure 1A. n = 3, n indicates the number of biological replicates. (D) BTH-induced resistance in the WT and zat18cr and npr1-2 mutants. Experiments were performed as described in Figure 1D. n = 8, n indicates the number of biological replicates.**** p < 0.0001.

ZAT18 Enhanced Plant Resistance to Psm ES4326
SA signaling mediates plant defense against biotrophic and hemibiotrophic pathogens, such as Psm ES4326 [35]. As ZAT18 regulates SA signaling, we hypothesized that ZAT18 also manages plant basal defense against Psm ES4326. To test this speculation, the  Figure 1A. n = 3, n indicates the number of biological replicates. (D) BTH-induced resistance in the WT and zat18cr and npr1-2 mutants. Experiments were performed as described in Figure 1D. n = 8, n indicates the number of biological replicates.**** p < 0.0001.

ZAT18 Enhanced Plant Resistance to Psm ES4326
SA signaling mediates plant defense against biotrophic and hemibiotrophic pathogens, such as Psm ES4326 [35]. As ZAT18 regulates SA signaling, we hypothesized that ZAT18 also manages plant basal defense against Psm ES4326. To test this speculation, the expression of ZAT18 after Psm ES4326 inoculation was first examined. Compared with untreated samples, a 2000-fold induction of ZAT18 was observed at 1 h post-infection (hpi) and the expression peak showed a 14,000-fold induction at 6 hpi ( Figure 3A). Then, to examine whether the ZAT18 mutation dampens plant defense against Psm ES4326, a higher dosage of Psm ES4326 (OD 600nm = 0.01) was sprayed onto the three-week-old WT and zat18 mutants. As shown in Figure 3B, a higher bacterial titer was detected in the zat18-1 and zat18cr mutants than in the WT, indicating that ZAT18 was required for the plant basal defense against Psm ES4326.
ZAT18-OE lines. Consistent with expectations, the expression of WRKY54 and WRKY70 was upregulated in the ZAT18-OE lines ( Figure 3D). Moreover, the expression of PR genes, which are SA late responsive genes, was also upregulated in the ZAT18-OE lines ( Figure S3). Then, three-week-old ZAT18-OE lines were inoculated with Psm ES4326 to determine whether ZAT18 overexpression could enhance the plant basal resistance to pathogens. As shown in Figure 3E, ZAT18 overexpression reduced the pathogen titer from 10 6.17 to 10 4.66 and 10 4.82 , respectively, in the two transgenic lines. These data further confirmed that ZAT18 positively regulated the plant basal defense responses.  Three-week-old wild-type (WT) seedlings were infiltrated with Psm ES4326 (OD 600nm = 0.001) and samples were collected at indicated time points to examine the expression levels of ZAT18 with UBQ5 as a reference. n = 3, n indicates the number of biological replicates. (B) Three-week-old WT and npr1-2 and zat18 mutants were sprayed with Psm ES4326 (OD 600nm = 0.01) and the third to sixth true leaves were collected to measure the bacterial titer at 3 dpi. cfu, colony forming unit. n = 8, n indicates the number of biological replicates. (C,D) The expression levels of ZAT18 (C) and WRKY54 and WRKY70 (D) in three-week-old WT and ZAT18-OE seedlings without any treatment. n = 3, n indicates the number of biological replicates. (E) Three-week-old WT and ZAT18-OE lines were sprayed with Psm ES4326 (OD 600nm = 0.1) and the third to sixth true leaves were collected to measure the bacterial titer at 3 dpi. cfu, colony forming unit. n = 8, n indicates the number of biological replicates. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001.
To further confirm that ZAT18 positively regulates plant basal defense, we made a construct to express ZAT18 driven by the 35S promoter and transformed it into the WT. Two independent ZAT18 overexpression (ZAT18-OE) lines (ZAT18-OE #9 and ZAT18-OE #12), with 3.3 × 10 4 -fold and 1.6 × 10 4 -fold expressions of ZAT18 compared with that in the WT, were identified and used in the following experiments ( Figure 3C). As ZAT18 positively regulates SA signaling, the expression of SA-responsive genes was tested in the ZAT18-OE lines. Consistent with expectations, the expression of WRKY54 and WRKY70 was upregulated in the ZAT18-OE lines ( Figure 3D). Moreover, the expression of PR genes, which are SA late responsive genes, was also upregulated in the ZAT18-OE lines ( Figure S3). Then, three-week-old ZAT18-OE lines were inoculated with Psm ES4326 to determine whether ZAT18 overexpression could enhance the plant basal resistance to pathogens. As shown in Figure 3E, ZAT18 overexpression reduced the pathogen titer from 10 6.17 to 10 4.66 and 10 4.82 , respectively, in the two transgenic lines. These data further confirmed that ZAT18 positively regulated the plant basal defense responses.

ZAT18 Suppressed Plant Growth by Repressing Auxin Signaling
During our experiments, we observed that ZAT18-OE lines looked smaller than the WT. To determine the role of ZAT18 in plant growth, we carefully observed the morphological phenotypes of ten-day-old, three-week-old and eight-week-old plants of the WT, zat18 mutants and ZAT18-OE lines. No significant differences were observed between the WT and zat18 mutants at these stages ( Figure S4). At all these stages, the ZAT18-OE lines were smaller than the WT ( Figure 4A-D). ZAT18 overexpression significantly decreased the leaf cell size and cell number in each leaf ( Figure 4E-G). The mean cell size was 2642 µm 2 for the WT leaves and there were an average of 6.9 × 10 4 cells in each WT leaf, while in the ZAT18-OE lines, the mean cell sizes were reduced to 400 µm 2 and 604 µm 2 and there were only 4.3 × 10 4 and 3.7 × 10 4 cells per leaf, respectively. These results indicated that ZAT18 may regulate cell division and expansion during development. measure the bacterial titer at 3 dpi. cfu, colony forming unit. n = 8, n indicates the number of biological replicates. (C,D) The expression levels of ZAT18 (C) and WRKY54 and WRKY70 (D) in threeweek-old WT and ZAT18-OE seedlings without any treatment. n = 3, n indicates the number of biological replicates. (E) Three-week-old WT and ZAT18-OE lines were sprayed with Psm ES4326 (OD600 nm = 0.1) and the third to sixth true leaves were collected to measure the bacterial titer at 3 dpi. cfu, colony forming unit. n = 8, n indicates the number of biological replicates. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001.

ZAT18 Suppressed Plant Growth by Repressing Auxin Signaling
During our experiments, we observed that ZAT18-OE lines looked smaller than the WT. To determine the role of ZAT18 in plant growth, we carefully observed the morphological phenotypes of ten-day-old, three-week-old and eight-week-old plants of the WT, zat18 mutants and ZAT18-OE lines. No significant differences were observed between the WT and zat18 mutants at these stages ( Figure S4). At all these stages, the ZAT18-OE lines were smaller than the WT ( Figure 4A-D). ZAT18 overexpression significantly decreased the leaf cell size and cell number in each leaf ( Figure 4E-G). The mean cell size was 2642 μm 2 for the WT leaves and there were an average of 6.9 × 10 4 cells in each WT leaf, while in the ZAT18-OE lines, the mean cell sizes were reduced to 400 μm 2 and 604 μm 2 and there were only 4.3 × 10 4 and 3.7 × 10 4 cells per leaf, respectively. These results indicated that ZAT18 may regulate cell division and expansion during development.  Meanwhile, ZAT18-OE lines produced short primary roots and fewer lateral roots ( Figure 5A). Compared with the WT, the primary root length of the two-week-old seedlings was reduced by approximately 3 cm and the number of lateral roots was reduced from 27 to 9 and 14, respectively, in the two ZAT18-OE lines ( Figure 5B,C). All these phenotypes of ZAT18-OE lines are reminiscent of auxin-insensitive or auxin-deficient mutants, indicating that ZAT18 may regulate auxin signaling [36][37][38]. Thus, we further checked the expression levels of the auxin synthesis gene YUC2 and auxin-responsive gene IAA4 in the ZAT18-OE lines. Consistent with their phenotypes, the expression levels of these genes were significantly reduced when ZAT18 was overexpressed in plants ( Figure 5D). To exclude the possibility that the ZAT18-OE lines mimicked auxin-deficient mutants due to elevated SA signaling, we further checked whether auxin signaling is affected in the zat18 mutant. Even though the zat18 mutant showed no morphological differences from the WT plants, the expression of auxin-related genes was upregulated in the zat18 mutants ( Figure S5), confirming the repression of auxin signaling by ZAT18.
( Figure 5A). Compared with the WT, the primary root length of the two-week-old seedlings was reduced by approximately 3 cm and the number of lateral roots was reduced from 27 to 9 and 14, respectively, in the two ZAT18-OE lines ( Figure 5B,C). All these phenotypes of ZAT18-OE lines are reminiscent of auxin-insensitive or auxin-deficient mutants, indicating that ZAT18 may regulate auxin signaling [36][37][38]. Thus, we further checked the expression levels of the auxin synthesis gene YUC2 and auxin-responsive gene IAA4 in the ZAT18-OE lines. Consistent with their phenotypes, the expression levels of these genes were significantly reduced when ZAT18 was overexpressed in plants (Figure 5D). To exclude the possibility that the ZAT18-OE lines mimicked auxin-deficient mutants due to elevated SA signaling, we further checked whether auxin signaling is affected in the zat18 mutant. Even though the zat18 mutant showed no morphological differences from the WT plants, the expression of auxin-related genes was upregulated in the zat18 mutants ( Figure S5), confirming the repression of auxin signaling by ZAT18.

ZAT18 Oppositely Regulated Defense-and Growth-Related Pathways
Because ZAT18 is a Cys2/His2-type transcription factor [31], we further performed transcriptome analysis to deeply understand its function in the growth-defense tradeoff. Two-week-old WT and zat18-1 seedlings were collected and three biological replicates for each genotype were sent out for sequencing. The sequencing metrics of all RNAseq libraries are shown in Table S1. Compared with the WT, 320 differently expressed genes (DEGs) were upregulated and 389 DEGs were downregulated in the zat18-1 mutant ( Figure 6A). We further analyzed which biological processes were regulated by ZAT18. As shown in Figure 6B,C, the auxin-related pathway was significantly enriched in the 320 upregulated DEGs, as presented by genes involved in the auxin-activated pathway, response to auxin, auxin polar transport and regulation of auxin polar transport. Moreover, the cytokinin-activated signaling pathway and negative regulation of cytokinin-activated signaling pathway was enriched in the upregulated DEGs and downregulated DEGs, respectively, indicating that ZAT18 suppressed cytokinin (CK) signaling ( Figure 6B-D). Other biological pathways, including regulation of organ growth, photosynthesis, light harvesting in photosystem II, positive regulation of cell growth, and root and cotyledon development, which are related to plant growth, were also enriched in the upregulated genes ( Figure 6B). These results indicated that the regulation of ZAT18 on plant growth is a complex process. SA signaling was clearly downregulated in the zat18-1 mutant because we observed enrichment in biological processes such as the response to salicylic acid, systemic acquired resistance and salicylic acid-mediated signaling pathways ( Figure 6C,D). These results further confirmed that ZAT18 was a positive regulator of SA signaling. Except for SA, the genes related to biotic and abiotic defense were enriched, which was represented by the response to chitin, defense response to bacterium, incompatible interaction, defense response, cellular response to UV and response to oxidative stress. The downregulation of abiotic stress-related genes was consistent with previous data showing that ZAT18 is induced by abiotic stresses and enhances plant resistance to drought [31]. The regulation of biotic and abiotic stresses by ZAT18 may be achieved through phytohormones because in addition to SA signaling, the signaling of ABA and ethylene (ET), which are two other defense-related hormones, was also downregulated in the zat18-1 mutant ( Figure 6C,D).
To verify the RNA-seq results, qRT-PCR was performed to confirm the expression of these hormone-related genes in two zat18 mutants. Three genes from each hormone pathway were selected. As shown in Figure 7A, the genes related to SA, ABA and ET signaling were downregulated, while genes in the auxin pathways were upregulated in both zat18 mutants. As CK-activated genes and the negative regulators of CK response were enriched in upregulated and downregulated DEGs, respectively, qRT-PCR was performed on both categories. The CK activated genes were slightly upregulated ( Figure S6), while SA signaling was clearly downregulated in the zat18-1 mutant because we observed enrichment in biological processes such as the response to salicylic acid, systemic acquired resistance and salicylic acid-mediated signaling pathways ( Figure 6C,D). These results further confirmed that ZAT18 was a positive regulator of SA signaling. Except for SA, the genes related to biotic and abiotic defense were enriched, which was represented by the response to chitin, defense response to bacterium, incompatible interaction, defense response, cellular response to UV and response to oxidative stress. The downregulation of abiotic stress-related genes was consistent with previous data showing that ZAT18 is induced by abiotic stresses and enhances plant resistance to drought [31]. The regulation of biotic and abiotic stresses by ZAT18 may be achieved through phytohormones because in addition to SA signaling, the signaling of ABA and ethylene (ET), which are two other defense-related hormones, was also downregulated in the zat18-1 mutant ( Figure 6C,D).
To verify the RNA-seq results, qRT-PCR was performed to confirm the expression of these hormone-related genes in two zat18 mutants. Three genes from each hormone pathway were selected. As shown in Figure 7A, the genes related to SA, ABA and ET signaling were downregulated, while genes in the auxin pathways were upregulated in both zat18 mutants. As CK-activated genes and the negative regulators of CK response were enriched in upregulated and downregulated DEGs, respectively, qRT-PCR was performed on both categories. The CK activated genes were slightly upregulated ( Figure S6), while the negative regulators of CK were significantly downregulated in the zat18 mutants ( Figure 7A). Based on these results, we speculated that ZAT18 regulated the expression of CK signaling mainly through modulating the expression of these repressors.  To further confirm the regulation of ZAT18 on the signaling of these hormones, we checked the expression levels of these hormone-related DEGs in two ZAT18-OE lines and observed opposite expression patterns from them compared with the zat18 mutants ( Figure 7B). These results suggested that ZAT18 promoted the plant defense and inhibited plant growth under biotic and abiotic stress conditions by differently regulating the expression of defense-related hormones (SA, ABA and ET) and growth-related hormones (auxin and CK).

Discussion
Activation of defense responses under unfavorable conditions is usually accompanied by growth inhibition, probably because of the competition for resources [3,10]. To balance growth and defense, plants use multiple strategies, including hormone crosstalk. Defenserelated hormones are synthesized under stress conditions, and their signaling is activated to defend against environmental stresses [39][40][41][42]. Meanwhile, they also negatively regulate pathways of growth-related hormones to inhibit growth [19,43,44]. Compared with enzymes that function in protein modification, transcription factors that directly change the transcriptome have been less explored in the crosstalk of growth and defense hormones. In this study, we found that ZAT18 was upregulated by the defense-related hormone SA and pathogens to promote SA signaling for pathogen resistance. In addition, ZAT18 inhibited auxin signaling, indicating that ZAT18 was a transcription factor regulating the growth-defense tradeoff upon pathogen infection by modulating the crosstalk of SA and auxin. Further transcriptome analysis suggested that ZAT18 was also a positive regulator of the ABA and ET pathways and a negative regulator of CK signaling. Combined with ZAT18 being activated by multiple abiotic stresses and enhancing plant defense to drought in a previous report [31], we proposed that ZAT18 was upregulated by both abiotic and biotic stresses to balance growth and defense by fine-tuning the antagonistic interactions between the signaling of growth-related hormones (auxin and CK) and defense-related hormones (SA, ET and ABA) at the transcriptional level ( Figure 8). mechanisms by which ZAT18 differentially regulates growth and defense may help us to find a way to enhance defense without growth penalties for crop molecular breeding.

Plant Materials, Growth Conditions and Phytohormone Treatments
The Arabidopsis thaliana wild type (WT), mutants and transgenic plants used in this study were all from the Colombia-0 ecotype background. The zat18-1 (SALK_027144C) was an Arabidopsis T-DNA insertion mutant and was obtained from the Arabidopsis Biological Resource Center (ABRC). Homozygotes were identified via PCR using the primers Among the zinc finger proteins, the ZAT family is the better-characterized subclass in defense responses. Even though most studies on ZATs are on the controls of abiotic stresses [24,26,30,45], the connections of ZATs with biotic stresses were also reported. Shi et al. found that ZAT6 is induced by pathogen infection and enhances plant immunity by modulating SA synthesis and signaling [46]. ZAT6 is also a target of mitogen-activated protein kinase 6 (MPK6), which can be activated by pathogen infection, indicating that plants may activate ZAT6-mediated defense responses at both the transcriptional and posttranslational levels [47,48]. During the preparation of this manuscript, ZAT18 was reported to be a regulator of immunity by modulating the crosstalk of SA and pathogensecreted coronatine [49]. In this Gao et al. report, ZAT18 enhanced susceptibility instead of resistance to pathogens and suppressed SA signaling during pathogen infection through coronatine [49]. We hypothesize that these inconsistencies may have been due to different pathogen strains (Psm ES4326 vs. Pst DC3000) or different infection methods (spray inoculation vs. infiltration) used in the two studies. Under their conditions, the antagonistic function of coronatine on SA may be stronger, which concealed the promotion of SA signaling by ZAT18.
Our qRT-PCR results showed that SA could induce the expression of ZAT18 in the npr1-2 mutant ( Figure 1B), indicating that ZAT18-mediated auxin repression was NPR1independent. This is consistent with previous findings showing that SA suppresses plant growth that relies on SA-binding proteins (SABPs) other than NPRs [20,50,51]. Thus, other SABPs may bind SA to induce the expression of ZAT18 after SA treatment [52]. Moreover, the expression of ZAT18 can be upregulated by many elicitors, such as pathogens, salt, drought and heat [31], and ZAT18 promotes plant defense against biotic and abiotic stresses, indicating that ZAT18 is a general stress-responsive gene for defense activation and growth inhibition. It will be interesting to investigate whether stresses elevate the transcription of ZAT18 through different mechanisms or whether a common regulator exists to combine stress signaling for ZAT18 activation.
In summary, we identified a C2H2-type transcription factor, namely, ZAT18, which oppositely regulated SA and auxin signaling to promote defense and inhibit growth. Moreover, our transcriptome data suggested that ZAT18 also regulated the signaling of other defense-and growth-related hormones. It is interesting that different expression levels and/or tag-fusing forms (GFP-ZAT18 or ZAT18-GFP) of ZAT18-OE lines change the growth phenotype but not defense responses [31,49]. Thus, further exploration of the mechanisms by which ZAT18 differentially regulates growth and defense may help us to find a way to enhance defense without growth penalties for crop molecular breeding.

Plant Materials, Growth Conditions and Phytohormone Treatments
The Arabidopsis thaliana wild type (WT), mutants and transgenic plants used in this study were all from the Colombia-0 ecotype background. The zat18-1 (SALK_027144C) was an Arabidopsis T-DNA insertion mutant and was obtained from the Arabidopsis Biological Resource Center (ABRC). Homozygotes were identified via PCR using the primers ZAT18TM-LP, ZAT18TM-RP and LBb1.3, which are listed in Table S2. The npr1-2 and tga256 mutants were described previously [53,54]. The zat18cr and ZAT18-OE #9/#12 were transgenic plants created in our lab.
Plants grown in soil were cultivated at 22 • C in a greenhouse with a 12 h light/12 h dark cycle. For plant seedling cultivation, seeds were sterilized with 75% ethanol for 1 min, rinsed once with sterile dH 2 O and then sterilized with 1% sodium hypochlorite for 8-10 min before being rinsed three times with sterile dH 2 O. Sterile seeds were vernalized at 4 • C for 3 days. After that, plants were grown on solid 1/2 MS medium with 15 g/L sucrose and were maintained at 22 • C with a 16 h light/8 h dark cycle.
For the benzothiadiazole (BTH) treatment in the pathogen infiltration assays, threeweek-old plants were sprayed with 120 µM BTH and were maintained in the dark for 24 h under normal growth conditions. For the total RNA extraction assay, two-week-old seedlings were dipped in 1 mM SA solution or control solution for a specified time before collection for RNA extraction.

Pathogen Inoculation
In this study, Pseudomonas syringae pv.maculicola (Psm) ES4326 was used for the pathogen inoculation. Psm ES4326 was grown at 28 • C in King's B medium containing 50 mg/L streptomycin for approximately 36 h. Bacterial cultures were collected and suspended with 10 mM MgSO 4 and then diluted to a final optical density of 0.01 or 0.1 at 600 nm (OD 600nm ). The diluted bacterial suspension with 0.02% Silwet L-77 was sprayed on the leaves of three-week-old Arabidopsis. Three days later, the inoculated third to sixth true leaves were weighed and collected for grounding in 10 mM MgSO 4 and then diluted to five concentrations in a tenfold decreasing gradient for the bacterial biomass detection assay.

Construction for Creating Transgenic Plants
For the CRISPR/Cas9 mediated genome editing in plants, a 20 nt specific spacer against the exon of ZAT18 with the AGG protospacer-adjacent motif (PAM) sequence (ZAT18gRNA-R) and its reverse complementary sequence (ZAT18gRNA-F) were artificially synthesized (Table S2). The annealed products of these two primers were ligated into vector pHEE401E [55] and transformed into competent cells of E. coli strain DH5α. The correct construct (pHEE401E::ZAT18gRNA) was then transformed into Agrobacterium strain GV3101 using a heat shock transition.
For creating the ZAT18 overexpression plants, the full-length coding sequence of ZAT18 was ligated into vector pART27 with a GFP tag [56]. The correct construct pART27::ZAT18-GFP was transformed into Agrobacterium strain GV3101 using a heat shock transition for further experiments. The primers used in this experiment are listed in Table S2.
The Arabidopsis thaliana ecotype Colombia-0 was used for transformation with Agrobacterium tumefaciens carrying correct constructs (pHEE401E::ZAT18gRNA or pART27::ZAT18-GFP), as described previously [56,57]. Briefly, the transformed Agrobacterium tumefaciens were incubated overnight in Luria-Bertani medium with 50 mg/L kanamycin and 50 mg/L rifampicin for pHEE401E::ZAT18gRNA, or 100 mg/L spectinomycin and 50 mg/L rifampicin for pART27::ZAT18-GFP, at 28 • C and 200 rpm. Then, the bacteria were pelleted and resuspended in an infiltration buffer (10 mM 2-(N-morpholino) ethane sulphonic acid (MES), 10 mM MgCl 2 and 200 mM acetosyringone, pH 5.6). The flowering Arabidopsis was dipped into a suspension mixture for 10 to 30 s and was cultivated overnight in the dark before growth under normal conditions. The positive transformants of ZAT18 overexpression in T1 generation were screened on solid 1/2 MS medium with 50 mg/L kanamycin. Then, the T1 plants with a single T-DNA insertion were used to obtain ZAT18-OE homozygous lines (#9 and #12) in the T3 generation. Plants carrying the pHEE401E::ZAT18gRNA construct were screened on solid 1/2 MS medium with 25 mg/L hygromycin. The homozygous genome-edited plant was identified via Sanger sequencing using primer ZAT18-F.

Total RNA Extraction and qRT-PCR
Total RNA was extracted using a TRIzol reagent. For testing the expression pattern of ZAT18 induced by SA or the pathogen, two-week-old seedlings or leaves of three-week-old plants were collected at 0, 1, 3, 6, 12 and 24 h post-inoculation (hpi) respectively. For testing the expression levels of WRKY54 and WRKY70 after SA treatment, two-week-old seedlings were collected at 1 hpi, as WRKY54 and WRKY70 are SA early responsive genes. For testing the expression levels of marker genes in different hormone signaling pathways in the ZAT18-OE lines and zat18 mutants, leaves of three-week-old plants without any treatment were collected. First-strand complementary DNA was synthesized from total RNA using a reverse transcription kit [HiScript ® III RT SuperMix for qPCR (+gDNA wiper), R323-01, Vazyme Biotech Co., Ltd, Nanjing, China]. qRT-PCR was performed using a ChamQ SYBR qPCR Master Mix (Q311, Vazyme Biotech Co., Ltd, Nanjing, China) on a QuantStudio 5 Real-Time PCR System (Applied Biosystems QuantStudio 5, Thermo Fisher Scientific, Waltham, MA, USA). UBQ5 was used as an internal control and the primers UBQ5-qF/qR are listed in Table S2. The expression level of ZAT18 was determined using the primers ZAT18-qF and ZAT18-qR listed in Table S2. All primers for qRT-PCR are listed in Table S2.

Transcriptome Sequencing
Leaves of three-week-old Arabidopsis, both WT (Colombia-0) and zat18-1, were collected as three biological replicates for RNA extraction and RNA-seq. Transcriptome sequencing was performed using the Illumina NovaSeq6000 platform from Biomarker Bioinformatics Technology Co., Ltd. (Qingdao, China). The clean reads were mapped to the Arabidopsis genome (Tair11). Gene expression levels were expressed as fragments per kilobase of exon model per million reads mapped (FPKM). The DEGSeq R package (p value < 0.01 and Fold Change ≥ 1.5) was used for differential expression analysis. Gene Ontology (GO) term enrichments were analyzed with the BMKCloud platform (www.biocloud.net, accessed on 2 January 2021).

Statistical Analyses
Group comparisons in the SA or BTH induction experiments were calculated using two-way ANOVA analysis (experimental design: no matching, use regular two-way ANOVA, no repeated measures) in GraphPad Prism 7. The significant differences of all other experiments were determined using Student's t test (experimental design: non parametric tests, Mann-Whitney test, compare ranks) in GraphPad Prism 7. The sample size n can be found in the figure legends, where n means the number of biological replicates unless specifically stated. All data are shown as mean ± s.d. * p < 0.05; ** p < 0.01; *** and **** p < 0.001; ns, no significant difference.

The Measurements of Root Length, Fresh Weight, Cell Size and Cell Number
For measuring the fresh weight of seedlings, 100 ten-day-old seedlings were collected as 10 groups and each group was quantified by scale. For checking the root length, the primary root of each two-week-old seedling was measured from the stem base to the tip using Image J.
For the cell size measurement, the fully expanded fifth and sixth true leaves of threeweek-old plants were fixed and decolorized in the clearing solution (75% (v/v) ethanol and 25% (v/v) acetic acid). Next, the leaves were cleaned with different concentrations of ethanol solutions (60%, 40%, 20% and 10%) were applied in order. After that, the solution was replaced with the chloral hydrate solution (250 g chloral hydrate dissolved in 100 mL dH 2 O). Mesophyll cells were examined and photographed under a 40× objective lens and the cell size was measured using Image J. For counting the cell number per leaf, the leaf area was measured using Image J and the cell number was determined using the leaf area/cell size.

Supplementary Materials:
The following supporting information can be downloaded from https: //www.mdpi.com/article/10.3390/ijms232315436/s1. Figure S1: The relative expression levels of ZAT18 in the RNA-sequencing data. Figure S2: Identification of the T-DNA insertion mutant of ZAT18. Figure S3: The expression levels of PR genes in the ZAT18-OE lines. Figure S4: Abolishing the function of ZAT18 had no obvious effects on plant growth. Figure S5: Auxin signaling was activated in the zat18 mutants with no obvious phenotype. Figure S6: The expression levels of cytokinin-activated genes in the zat18 mutants. Table S1: Sequencing metrics of the 6 RNA-seq libraries. Table S2: The primers used in this study.