Genome-Wide Analysis of ZAT Gene Family in Osmanthus fragrans and the Function Exploration of OfZAT35 in Cold Stress

Osmanthus fragrans is a popular ornamental and odorant plant with high commercial value, but its cultivation and exploitation are limited by low temperature. The ZAT (zinc finger of Arabidopsis thaliana) genes as a subclass of the C2H2-type zinc finger proteins (C2H2-ZFP) family play essential roles in various abiotic stresses. However, their roles in cold stress response in O. fragrans remain unclear. This study identified 38 OfZATs, which could be divided into 5 subgroups based on the phylogenetic tree, with OfZATs in the same subgroup harboring similar gene structures and motif patterns. In addition, 49 segmental and 5 tandem duplication events were detected among OfZAT genes, while some OfZAT genes exhibited specific expression patterns in different tissues. Furthermore, two OfZATs were induced in salt stress and eight OfZATs responded to cold stress. Interestingly, OfZAT35 showed a continuously increasing expression trend under cold stress, while its protein showed nucleus localization with no transcriptional activation activity. Transiently transformed tobacco overexpressing OfZAT35 exhibited a significantly higher relative electrolyte leakage (REL) level and increased activities of superoxide dismutase (SOD), peroxidase (POD), and Ascorbate peroxidase (APX), while there was significantly decreased activity of catalase (CAT). Moreover, CAT, DREB3, and LEA5, which are associated with cold stress, were dramatically decreased after cold treatment in transiently transformed tobacco, suggesting that overexpression of OfZAT35 negatively regulated cold stress. This study provides a basis for exploring the roles of ZAT genes and contributes to uncovering the mechanism of ZAT-mediated cold stress response in O. fragrans.


Introduction
Osmanthus fragrans is an ever-green, small ornamental tree or shrub, which is famous for its fragrant flowers and high commercial value. The O. fragrans flowers can be processed into food additives in pastries and drinks, such as tea, while its flower extracts are known to produce one of the best natural fragrant essences [1]. Due to its commercial benefits, the color and flower fragrance have become trending research topics in O. fragrans [2][3][4]. Osmanthus mostly grows in the warm temperature region of Asia [5]. The cultivation and commercial exploitation of the species O. fragrans is however limited by low temperature [6]. Therefore, low temperature is a key factor in osmanthus breeding. Several molecular studies on the O. fragrans response to cold stress have recently been reported. The genes of the bHLH transcription factor (TF) in O. fragrans induced by cold stress were screened through quantitative real-time PCR (qRT-PCR), while the functions of the NAC TF family in response to cold stress in O. fragrans were also screened using qRT-PCR [7,8].
which demonstrated that the ZAT gene family predominantly originated from segmental and tandem duplication events.
The physicochemical evaluation of OfZATs revealed that the predicted protein molecular weight (MW) ranged from 19.50 to 55.83 kDa in OfZAT13 and OfZAT37, respectively, while the isoelectric point (pI) varied from 6.11 in OfZAT17 to 9.73 in OfZAT4. Moreover, the prediction of subcellular localization exhibited that all predicted OfZATs are located in the nucleus (Table S1).
The physicochemical evaluation of OfZATs revealed that the predicted protein molecular weight (MW) ranged from 19.50 to 55.83 kDa in OfZAT13 and OfZAT37, respectively, while the isoelectric point (pI) varied from 6.11 in OfZAT17 to 9.73 in OfZAT4. Moreover, the prediction of subcellular localization exhibited that all predicted OfZATs are located in the nucleus (Table S1).

Phylogenetic Analysis of ZAT Genes
The 38 predicted OfZATs along with 20, 20, and 11 homologous genes retrieved from Arabidopsis (A. thaliana), rice (Oryza sativa), and black cottonwood (P. trichocarpa), respectively, were used for phylogenetic tree construction with the neighbor-joining (NJ) Plants 2023, 12, 2346 4 of 17 method to explore their evolutionary relationships. According to the types of motifs ( Figure S2), the 89 genes could be clustered into 5 subgroups, including ZAT-A, ZAT-B, ZAT-C, ZAT-D, and ZAT-E ( Figure 2). The sequences of motifs are displayed in Table S2. Notably, ZAT-A was the largest subgroup containing 25 genes, while ZAT-C was the smallest subgroup with 11 members.
Plants 2023, 12, x 4 of 18 method to explore their evolutionary relationships. According to the types of motifs (Figure S2), the 89 genes could be clustered into 5 subgroups, including ZAT-A, ZAT-B, ZAT-C, ZAT-D, and ZAT-E ( Figure 2). The sequences of motifs are displayed in Table S2. Notably, ZAT-A was the largest subgroup containing 25 genes, while ZAT-C was the smallest subgroup with 11 members.

Structure and Motif Composition of OfZATs
To further analyze the evolutionary relationships among OfZATs, their structural and motif composition were analyzed ( Figure 3). A varied number of exon-intron distributions were observed among OfZAT genes. Interestingly, introns were absent in most OfZATs, with only five genes (OfZAT1/3/6/29/30) having one intron in their sequences. Motif identification using the MEME tool revealed 15 conserved OfZATs motifs, with OfZAT proteins in the same subgroup harboring similar motifs in both position and type. For example, members of the ZAT-B subgroup all contained Motifs 3, 2, 1, 10, 4, and 9 with similar permutations, while most members of the ZAT-A subgroup contained Motifs 5, 3, 2, 1, 4, and 9 with a similar permutation. Notably, Motif 5 was only absent in OfZAT1. In addition, most OfZATs contained Motif 1, Motif 2, and Motif 3. Motif 1 contains the conserved sequence CX2CX3FX2GQALGGHX3H, which was reported as first zinc finger domain, and the conserved sequence CX2CX3FX3QALGGHX3H in Motif 2 (Table S3) was reported as the second zinc finger domain (X represents arbitrary amino acid, and the

Structure and Motif Composition of OfZATs
To further analyze the evolutionary relationships among OfZATs, their structural and motif composition were analyzed ( Figure 3). A varied number of exon-intron distributions were observed among OfZAT genes. Interestingly, introns were absent in most OfZATs, with only five genes (OfZAT1/3/6/29/30) having one intron in their sequences. Motif identification using the MEME tool revealed 15 conserved OfZATs motifs, with OfZAT proteins in the same subgroup harboring similar motifs in both position and type. For example, members of the ZAT-B subgroup all contained Motifs 3, 2, 1, 10, 4, and 9 with similar permutations, while most members of the ZAT-A subgroup contained Motifs 5, 3, 2, 1, 4, and 9 with a similar permutation. Notably, Motif 5 was only absent in OfZAT1. In addition, most OfZATs contained Motif 1, Motif 2, and Motif 3. Motif 1 contains the conserved sequence CX 2 CX 3 FX 2 GQALGGHX 3 H, which was reported as first zinc finger domain, and the conserved sequence CX 2 CX 3 FX 3 QALGGHX 3 H in Motif 2 (Table S3) was reported as the second zinc finger domain (X represents arbitrary amino acid, and the number represents the quantity of amino acid) [18]. Motif 3 contains the conserved  (Table S3), while the conserved sequence has not been reported. The  sequences of the 15 motifs are displayed in Table S3. number represents the quantity of amino acid) [18]. Motif 3 contains the conserved sequence AX2LX2L (Table S3), while the conserved sequence has not been reported. The sequences of the 15 motifs are displayed in Table S3.

Cis-Elements Analysis in OfZATs
The promoter screening of OfZATs was performed to facilitate the prediction of their potential biological functions. After excluding the universal, incomplete, and unannotated cis-elements, a total of 32 cis-elements, which were associated with four categories, including hormones, stress, development, and light responses, were identified ( Figure S3). The hormone category contained 11 cis-elements, which were associated with the regulation of abscisic acid (ABA), methyl-jasmonate (MeJA), gibberellin (GA), Auxin, and salicylic acid (SA). Of the 38 predicted OfZATs, 32 (84.21%) contained 109 ABRE cis-elements, which are associated with the ABA response pathway, while 30 OfZATs (78.95%) contained 76 CGTCA and 76 TGACG-motifs, which are related to MeJA regulation. The development and light categories contained 8 cis-elements, and 31 genes (81.58%) harbored 109 G-boxes that are associated with light response. The stress group contained six ciselements related to anaerobic induction, drought, wound, and low temperature stress. These results suggested that most members of the ZAT family in O. fragrans might potentially be involved in the ABA, MeJA, and light response processes, which provide a reference for further investigation.

Expression Pattern of OfZAT Genes in Different O. fragrans Tissues
The fragments per kilobase million (FPKM) values of OfZATs in four tissues, including the root, stem, leaves, and flowers, were obtained from previously reported transcriptome data (Table S4) which have been published in NCBI (SRP143423) [4]. The heat map was generated using TBtools according to the FPKM values from the transcriptome (Figure 4). Generally, FPKM values between 0 and 1 were considered as low expression levels

Cis-Elements Analysis in OfZATs
The promoter screening of OfZATs was performed to facilitate the prediction of their potential biological functions. After excluding the universal, incomplete, and unannotated cis-elements, a total of 32 cis-elements, which were associated with four categories, including hormones, stress, development, and light responses, were identified ( Figure S3). The hormone category contained 11 cis-elements, which were associated with the regulation of abscisic acid (ABA), methyl-jasmonate (MeJA), gibberellin (GA), Auxin, and salicylic acid (SA). Of the 38 predicted OfZATs, 32 (84.21%) contained 109 ABRE cis-elements, which are associated with the ABA response pathway, while 30 OfZATs (78.95%) contained 76 CGTCA and 76 TGACG-motifs, which are related to MeJA regulation. The development and light categories contained 8 cis-elements, and 31 genes (81.58%) harbored 109 G-boxes that are associated with light response. The stress group contained six cis-elements related to anaerobic induction, drought, wound, and low temperature stress. These results suggested that most members of the ZAT family in O. fragrans might potentially be involved in the ABA, MeJA, and light response processes, which provide a reference for further investigation.

Expression Patterns of OfZATs under Salt Stress
For salt treatment, the seedlings were planted in 1/2 Hoagland's nutrient solution with 250 mM NaCl solution and set at four time points (S0, S6, S24, and S72 h) to collect samples for transcriptome sequencing. In addition, the seedlings of the control group for salt stress were soaked in 1/2 Hoagland nutrient solution and also set at four time points (S0, CK6, CK24, and CK72 h) to collect samples for transcriptome sequencing. The values of FPKM were extracted from unpublished transcriptome data (Table S5) and submitted to TBtools to generate the heat map ( Figure S4). As a result, four genes (OfZAT8/25/35/36) showed higher expression levels (FPKM > 10), while the remaining exhibited low (FPKM < 1) or no expression (FPKM = 0) under salt stress ( Figure S4). The expression trends of OfZAT25 and OfZAT35 were decreased overall during salt treatment. However, the expression trends of OfZAT25 and OfZAT35 were also decreased in the control. In contrast, the levels of OfZAT8 and OfZAT36 exhibited an increasing trend under salt treatment, but with a decreasing pattern in the control treatment, which suggested that OfZAT8 and OfZAT36 play pivotal roles in salt stress response ( Figure 5).

Expression Patterns of OfZATs under Salt Stress
For salt treatment, the seedlings were planted in 1/2 Hoagland's nutrient solution with 250 mM NaCl solution and set at four time points (S0, S6, S24, and S72 h) to collect samples for transcriptome sequencing. In addition, the seedlings of the control group for salt stress were soaked in 1/2 Hoagland nutrient solution and also set at four time points (S0, CK6, CK24, and CK72 h) to collect samples for transcriptome sequencing. The values of FPKM were extracted from unpublished transcriptome data (Table S5) and submitted to TBtools to generate the heat map ( Figure S4). As a result, four genes (OfZAT8/25/35/36) showed higher expression levels (FPKM > 10), while the remaining exhibited low (FPKM < 1) or no expression (FPKM = 0) under salt stress ( Figure S4). The expression trends of OfZAT25 and OfZAT35 were decreased overall during salt treatment. However, the expression trends of OfZAT25 and OfZAT35 were also decreased in the control. In contrast, the levels of OfZAT8 and OfZAT36 exhibited an increasing trend under salt treatment, but with a decreasing pattern in the control treatment, which suggested that OfZAT8 and OfZAT36 play pivotal roles in salt stress response ( Figure 5).

Expression Patterns of OfZATs under Cold Stress
During the 4 °C cold treatment, the samples were collected at seven timepoints (C0, C3, C12, C24, C72, and C120 h during cold treatment, and after recovery Cr72 h) for transcriptome sequencing. The values of FPKM were extracted from unpublished transcriptome data (Table S6) and submitted to TBtools to generate the heat map. Strong expression profiles (FPKM > 10) were only observed in eight OfZAT genes under cold treatment. In contrast, 29 genes (76.32%) were unresponsive to cold stress (FPKM ≤ 1), of which 10 genes (26.32%) were not expressed (FPKM = 0) in all stages ( Figure S5). Under cold treatment, OfZAT7 and OfZAT38 were only upregulated at 0-3 h, and then downregulated in later stages. Five genes, including OfZAT8/25/26/29/32, were continuously upregulated 0-24 h

Expression Patterns of OfZATs under Cold Stress
During the 4 • C cold treatment, the samples were collected at seven timepoints (C0, C3, C12, C24, C72, and C120 h during cold treatment, and after recovery Cr72 h) for transcriptome sequencing. The values of FPKM were extracted from unpublished transcriptome data (Table S6) and submitted to TBtools to generate the heat map. Strong expression profiles (FPKM > 10) were only observed in eight OfZAT genes under cold treatment. In contrast, 29 genes (76.32%) were unresponsive to cold stress (FPKM ≤ 1), of which 10 genes (26.32%) were not expressed (FPKM = 0) in all stages ( Figure S5). Under cold treatment, OfZAT7 and OfZAT38 were only upregulated at 0-3 h, and then downregulated in later stages. Five genes, including OfZAT8/25/26/29/32, were continuously upregulated 0-24 h before their overall expression levels decreased in the later periods. Both OfZAT35 and OfZAT36 were dramatically upregulated 0-72 or 0-120 h during cold treatment, and then decreased to a pre-treatment (0 h) level ( Figure 6). Subsequently, qRT-PCR analysis was performed in five genes (OfZAT25/26/29/32/35). (Figure 7). As a result, consistent expression patterns were observed between the selected genes and those of transcriptome before their overall expression levels decreased in the later periods. Both OfZAT35 and OfZAT36 were dramatically upregulated 0-72 or 0-120 h during cold treatment, and then decreased to a pre-treatment (0 h) level ( Figure 6). Subsequently, qRT-PCR analysis was performed in five genes (OfZAT25/26/29/32/35). (Figure 7). As a result, consistent expression patterns were observed between the selected genes and those of transcriptome data in all stages, which demonstrated the reliability of RNA-seq data. Based on its unique expression level dynamics at different intervals after stress exposure, the OfZAT35 gene was selected as a candidate for further investigation.

Subcellular Localization and Transcriptional Activation Activity
The coding sequence of OfZAT35 was amplified and fused into 35S::GFP plasmids. The empty 35S::GFP vector and 35S::GFP-OfZAT35 vector were transfected into leaves of 30 d tobacco (N. benthamiana), respectively. The fluorescent signals from the leaves revealed that the OfZAT35 protein was only located in the nucleus, indicating its involve- Figure 7. The relative expression levels of five selected OfZATs during cold treatment. The histograms indicating the data from qRT-PCR and FPKM values from transcription data are marked by a red line. The qRT-PCR data were statistically assessed using one-way ANOVA followed by Duncan's test (p < 0.05), and the error bar represents ± SE (standard error) (n = 3). The letters (a, b, c) indicated the statistically differences based on Duncan's test (p < 0.05).

Subcellular Localization and Transcriptional Activation Activity
The coding sequence of OfZAT35 was amplified and fused into 35S::GFP plasmids. The empty 35S::GFP vector and 35S::GFP-OfZAT35 vector were transfected into leaves of 30 d tobacco (N. benthamiana), respectively. The fluorescent signals from the leaves revealed that the OfZAT35 protein was only located in the nucleus, indicating its involvement in nucleus functions (Figure 8a). Moreover, pGBKT7-OfZAT35 was also constructed using homologous recombination and then introduced to the AH109 yeast strain for a transcriptional activation activity assay. The yeast strains of the negative control (empty vector pGBKT7) and recombinant plasmids pGBKT7-OfZAT35 grew in a different element deficient culture medium of SD/-Trp, SD/-Trp-Ade, and SD/-Trp-Ade + X-α-gal medium. OfZAT35 and the negative control grew well in SD/-Trp medium, but not in other media, which indicated the absence of its transcriptional activation activity (Figure 8b).

Analysis of Physiological Parameters
The fused Super1300-OfZAT35 plasmids and empty Super1300 vector (EV) were transfected into leaves of 30 d tobacco (N. benthamiana) by an Agrobacterium-mediated transient expression method. Then, the transiently transformed tobaccos of OfZAT35 and EV were treated for 6 hours under 4 °C. The OfZAT35 was overexpressed according to the result of semi-quantitative RT PCR (sqRT-PCR) (Figure 9a). After 6 h of cold treatment at 4 °C, the relative electrolyte leakage (REL) of transiently transformed tobacco was significantly increased compared to that of the empty vector (EV) (Figure 9b), and the REL level is widely used as an indicator of cell membrane damage [34,35]. Thus, we speculated that the transiently transformed tobacco suffered more intense stress than the EV after 6 h cold The transcriptional activation activity of OfZAT35. The empty PGBKT7 vector (negative control) and recombinant plasmids pGBKT7-OfZAT35 were transformed into yeast strain AH109. The yeast strain of negative control and OfZAT35 (pGBKT7-OfZAT35) grew on different screening culture media (SD/-Trp, SD/-Trp-Ade, and SD/-Trp-Ade + X-α-gal medium). The number of 10 0 , 10 −1 , 10 −2 , 10 −3 , and 10 −4 represents the concentration of dilution ratio of original yeast culture.

Analysis of Physiological Parameters
The fused Super1300-OfZAT35 plasmids and empty Super1300 vector (EV) were transfected into leaves of 30 d tobacco (N. benthamiana) by an Agrobacterium-mediated transient expression method. Then, the transiently transformed tobaccos of OfZAT35 and EV were treated for 6 hours under 4 • C. The OfZAT35 was overexpressed according to the result of semi-quantitative RT PCR (sqRT-PCR) (Figure 9a). After 6 h of cold treatment at 4 • C, the relative electrolyte leakage (REL) of transiently transformed tobacco was significantly increased compared to that of the empty vector (EV) (Figure 9b), and the REL level is widely used as an indicator of cell membrane damage [34,35]. Thus, we speculated that the transiently transformed tobacco suffered more intense stress than the EV after 6 h cold treatment at 4 • C. The activities of superoxide dismutase (SOD), peroxidase (POD), and Ascorbate peroxidase (APX) were significantly increased in the transiently transformed tobacco compared to the EV under cold stress (Figure 9d). However, the activity of catalase (CAT) was significantly reduced (Figure 9d). Notably, the expression levels of NbAPX and NbCAT were consistent with the activities of APX and CAT, respectively (Figure 9c). To analyze the response mechanism of transiently transformed tobacco to cold stress, the expression profiles of cold-related genes were determined using qPCR. The results showed that the levels of NbDREB3 and NbLEA5 were significantly decreased (Figure 9c). Overall, transiently transformed tobacco overexpressing OfZAT35 displayed more cold stress effects than EV plants.

Discussion
In this study, 38 members of the ZAT family in O. fragrans were identified, and the number is higher than the number reported in the herbaceous A. thaliana (20) and O. sativa (20) plants or in the woody plant P. trichocarpa (11) [11,14,36]. Synteny analysis detected 49 segmental and 5 tandem duplication events in the OfZAT genes ( Figure S1). Large genomic duplication events are known to drive the evolution and expansion of gene families [37][38][39]. For example, the A. thaliana genome has undergone at least 4 major duplication events between 100 and 200 million years ago (MYA), with segmental and tandem duplication contributing to the generation and maintenance of gene families [37,40]. Correspondingly, the O. fragrans genome has experienced 2 duplication events that occurred approximately 14 MYA, which might have contributed to the larger size of the OfZAT gene family [4].
The 38 candidate OfZAT genes were divided into 5 subgroups, and the genes within the same subgroup displayed similar motif types and arrangement, indicating that The activities of superoxide dismutase (SOD), peroxidase (POD), and Ascorbate peroxidase (APX) were significantly increased in the transiently transformed tobacco compared to the EV under cold stress (Figure 9d). However, the activity of catalase (CAT) was significantly reduced (Figure 9d). Notably, the expression levels of NbAPX and NbCAT were consistent with the activities of APX and CAT, respectively (Figure 9c). To analyze the response mechanism of transiently transformed tobacco to cold stress, the expression profiles of cold-related genes were determined using qPCR. The results showed that the levels of NbDREB3 and NbLEA5 were significantly decreased (Figure 9c). Overall, transiently transformed tobacco overexpressing OfZAT35 displayed more cold stress effects than EV plants.

Discussion
In this study, 38 members of the ZAT family in O. fragrans were identified, and the number is higher than the number reported in the herbaceous A. thaliana (20) and O. sativa (20) plants or in the woody plant P. trichocarpa (11) [11,14,36]. Synteny analysis detected 49 segmental and 5 tandem duplication events in the OfZAT genes ( Figure S1). Large genomic duplication events are known to drive the evolution and expansion of gene families [37][38][39]. For example, the A. thaliana genome has undergone at least 4 major duplication events between 100 and 200 million years ago (MYA), with segmental and tandem duplication contributing to the generation and maintenance of gene families [37,40]. Correspondingly, the O. fragrans genome has experienced 2 duplication events that occurred approximately 14 MYA, which might have contributed to the larger size of the OfZAT gene family [4].
The 38 candidate OfZAT genes were divided into 5 subgroups, and the genes within the same subgroup displayed similar motif types and arrangement, indicating that OfZATs in the same subgroup may have a similar function, and the classification of OfZATs was reliable (Figure 3). Motif 1, Motif 2, and Motif 3 are present in most OfZAT members, and Motif 1, Motif 2, and Motif 3 were also found in most ZAT proteins of A. thaliana, O. sativa, and P. trichocarpa ( Figure S2). Similarly, Motif 1, Motif 2, and Motif 3 have been reported in most members of the ZAT protein family in herbaceous plants, such as Fragaria × ananassa [17]. This indicated that the three ZAT motifs might be conserved in different plant species, and their similar clustering in same subgroup demonstrated the reliability of phylogenetic tree clusters. The structural analysis revealed the absence of introns in most OfZAT genes (Figure 3c), which was consistent with the observations made in G. hirsutum and P. trichocarpa [14,16]. Introns are predicted to delay regulatory responses, and genes with fewer introns are rapidly activated during stress [41]. In addition, most OfZATs contained only one exon, indicating their capacity to rapidly respond to environmental stresses. Cis-elements enable the binding of TFs and gene transcription [42]. Most OfZATs contained four types cis-elements associated with hormones, stress, development, and light responses ( Figure S3), which suggested their potential functions in correlated reaction pathways.
Normally, the tissue-specific expression of genes implies their potential roles. For example, AtERF102, AtERF103, AtERF104, and AtERF105, which are predominantly expressed in root tissues, are cold stress regulator genes [43]. In this study, the differentially expressed ZATs in O. fragrans tissues might crucially be involved in the plant growth and developmental processes. The genes involved in abiotic stress usually exhibited a differential expression profile in stress. For example, AtZAT10 (STZ) responded to salt stress obviously and improved the resistance for salt stress [44]. OfZAT8 and OfZAT36 were strongly induced in salt stress and have a close phylogenetic relationship with AtZAT10, suggesting that the two genes are crucial for salt stress.
The expression of OfZAT35 was upregulated under long-term cold treatment ( Figure 8). Interestingly, AtAZF2, AtZAT10, and AtZAT6, which are cold stress responsive and regulators in Arabidopsis, were phylogenetically closely clustered with the candidate OfZAT35 gene (Figure 1) [18,28]. Taken together, these results strongly suggested that OfZAT35 might also play pivotal roles in cold stress response in O. fragrans. The OfZAT35 protein was shown to be localized in the nucleus, which indicated its in-nucleus activity during cold stress response (Figure 8a). However, OfZAT35 exhibited no transcriptional activation activity (Figure 8b). A similar phenomenon was also discovered in other ZFPs. For example, FaZAT10 exhibited no transcriptional activation activity in strawberry [17]. In Capsicum annuum, the full-length CAZFP1 protein had no transcriptional activation activity [45]. Previous research has revealed that ZFPs regulate gene transcription and expression by interacting with other ZFPs to bind to other DNA sequences [44,46,47]. Thus, the OfZAT35 protein may interact with other proteins to regulate the expression of related genes and reduce cold tolerance. In addition, transiently transformed tobacco has been verified that overexpressed OfZAT35 (Figure 9a) and showed a significantly higher REL level than in EV (p < 0.05) (Figure 9b). The REL is a decisive parameter for predicting the damage to the membrane system [48]. For example, the ectopic expression of MdCDPK1a could improve cold stress tolerance in N. benthamiana by reducing the REL value after cold stress exposure [49]. Similarly, the ectopic expression of RmICE1 from Rosa multiflora enhanced cold stress tolerance in N. benthamiana and reduced the REL value after cold treatment [50]. Thus, the observed higher levels of REL indicated that OfZAT35 overexpression reduced cold stress tolerance in O. fragrans. In addition, the activities of SOD, POD, and APX were all significantly increased, while that of CAT was dramatically decreased (Figure 9d). Correspondingly, the expression levels of NbSOD, NbAPX, and NbCAT were consistent with the activities of SOD, APX, and CAT ( Figure 9c). Generally, the activities of antioxi-dant enzymes are induced when the plant is exposed to abiotic stress and could represent plant responses to adverse conditions [51]. The more significantly increased activities of SOD, POD, and APX in transiently transformed tobacco of OfZAT35 may indicate that the overexpression of OfZAT35 makes antioxidant enzymes more active to respond to cold stress. Furthermore, previous research has revealed that the CAT gene can reduce ROS levels to enhance cold tolerance [52,53]. Consequently, the decreased activity of CAT and expression level of NbCAT suggested that the transiently transformed tobacco of OfZAT35 may accumulate an ROS level and exhibit more sensitivity to cold stress. DREB/CBF and LEA are essential genes that modulate cold stress, and the accumulation of their transcripts is positively correlated with enhanced cold tolerance [54][55][56]. The expression levels of NbDREB3 and NbLEA5 were significantly decreased ( Figure 9c) and implied that OfZAT35 is a likely negative regulator of cold stress tolerance in O. fragrans.

Plant Materials and Treatments
Two-year-old cuttings of O. fragrans cv. 'Rixianggui' originated from Nanjing Forestry University in a previous study. Before application of abiotic treatments, seedlings with good development and consistent growth were selected and transferred to the growth chamber for one week. For cold treatment, the growth chamber temperature was adjusted to 4 • C, while other parameters were maintained [9]. During the cold treatment, we set seven timepoints (0, 3, 12, 24, 72, and 120 h during cold treatment, and after recovery 72 h) to collect the first two leaf pairs from seedlings as samples, and each timepoint set three biological replicates. For salt treatment, with other conditions maintained, the seedlings were planted in 1/2 Hoagland's nutrient solution with 250 mM NaCl solution, and at four time points (0, 6, 24, and 72 h), we collected the first two leaf pairs from each biological replication as a sample for subsequent analyses. In addition, the seedlings of control group for salt stress were set by soaking in 1/2 Hoagland nutrient solution as previously described [32]. The samples were stored at −80 • C in a freezer.
The chromosomal locations of OfZAT genes were analyzed using TBtools based on the O. fragrans genome annotation files. The duplication events in ZAT family were determined using MCScanX (Multiple Collinearity Scan toolkit) program in TBtools, and the results were collected and then visualized using the advanced Circos program [61,62].

Expression Profiles of OfZATs in Tissues under Abiotic Stress
The FPKM values of OfZATs in four tissues, including the root, stem, leaves, and flowers, were obtained from previously reported transcriptome data which have been published in NCBI (SRP143423) [4]. In addition, the expression patterns of OfZATs during cold and salt stress were also analyzed based on the unpublished transcriptome data, and the expression results were visualized with heat maps in TBtools [63].

RNA Extraction and qRT-PCR
The extraction of total RNA and the synthesis of cDNA from the frozen leaf samples of O. fragrans were performed as in a previous study [9]. Primer 5.0 [66] was utilized to design specific primers for OfZAT25, OfZAT26, OfZAT29, and OfZAT35 (Table S7). OfRNA was used to normalize the data analysis of qRT-PCR [7,67]. The reaction system of qRT-PCR was composed of 0.4 µL primer (forward and reverse primer, respectively), 5 µL SYBR, 0.2 µL ROX, 3 µL ddH 2 O, and 1 µL cDNA. The reaction condition was set as in a previous study [68].
The protocols of total RNA isolation from transiently transformed tobacco leaves and synthesis of cDNA were performed as in a previous study [9]. A semi-quantitative (sqRT-PCR) analysis was performed to verify the positive overexpression of OfZAT35 in tobacco, and Nbactin was selected as reference gene (Table S7). The mix reaction solution of sqRT-PCR was composed of 1 µL cDNA, 1 µL primer, 10 µL 2 × Rapid taq, and 7 µL ddH 2 O. The primers of NbSOD, NbCAT, NbP5CS, NbDREB3, and NbLEA5 for qRT-PCR are displayed in Table S8 [9]. Meanwhile, Nbactin was used as a normalizer for the qRT-PCR analysis. The qRT-PCR data were analyzed using the 2 −∆∆CT method, and the results were statistically analyzed in SPSS 20.0 [7].

Subcellular Localization and Transcriptional Activation Activity of OfZAT35
Specific primers were designed for cloning the full-length CDS sequence of OfZAT35 into the Super1300 vector between Hind III and Kpn I restriction sites to construct the 35S::OfZAT35-GFP fusion vector followed by a positive validation test (Table S9). The 35S::OfZAT35-GFP fusion vector and an EV were introduced into GV3101 (Agrobacterium tumefaciens), respectively, and incubated in Luria-Bertani culture with the following conditions: 28 • C, 200 rpm, 10 h. Then, OfZAT35 and EV were injected into the leaves of 30-day-old tobacco plants, respectively. Transformed tobacco plants were incubated for 2 d in a growth room [9,32]. Finally, the fluorescent signals in the transgenic leaves were detected through an LSM710 microscope (Zeiss, Jena, Germany). Moreover, the whole ORF of OfZAT35 was inserted into GAL4 DNA-binding domain between Smal I and Sal I restriction sites in pGBKT7 plasmid for transcriptional activation activity assay (primers are displayed in Table S8). The empty vector pGBKT7 (negative control) and fused vector pGBKT7-OfZAT35 were introduced into AH109 yeast strain, respectively. Then, the yeast strain of pGBKT7-OfZAT35 and pGBKT7 were respectively incubated in SD/Trp culture with the following conditions: 30 • C and 200 rpm until the value of OD600 reached 0.6. The yeast strains of pGBKT7-OfZAT35 and pGBKT7 in 2 mL SD/Trp culture were collected into 100 µL ddH 2 O, and then diluted into different dilution multiples (10 0 , 10 −1 , 10 −2 , 10 −3 , and 10 −4 ). The different dilution multiples solution of pGBKT7-OfZAT35 and pG-BKT7 was incubated in different nutrient deficient media (SD/-Trp, SD/-Trp-Ade, and SD/-Trp-Ade + X-α-gal medium) under 30 • C.
Transiently transformed tobacco was generated as described for subcellular location. Transiently transformed tobacco was exposed to cold stress treatment at 4 • C in the growth chamber for 6 h. Samples were then collected and frozen by liquid nitrogen. The samples were stored at −80 • C in a freezer for RNA isolation and the determination of physiological properties.

Analyses of Physiological Parameters
The REL value was determined to evaluate the degree of stress damage in the transgenic tobacco plants following a previously reported protocol [69]. Briefly, 0.2 g of shredded leaves was placed in 20 mL distilled water to measure the electrolyte leakage (C1) after 24 h, and the electrolyte leakage of pure distilled water (C0) was also determined. The shredded leaves in the pure distilled water were boiled for 30 min, and then the electrolyte leakage was determined (C2). The relative electrolyte leakage was calculated with the formula REL = (C1 − C0)/(C2 − C0) × 100%.
For enzyme activity assay, the crude enzyme solution was produced by 0.2 g leaf samples powder in 5 mL sodium phosphate buffer at pH 7.8. The assay for determining the activity of SOD referred to Beauchamp and Fridovich, while POD, CAT, and APX activities were quantified as previous in research with a slight modification [70][71][72]. The assay for POD activity was modified by adding 0.1 mL enzyme solution to the substrate solution and determining the reaction mixture at A470 nm. The assay for CAT activity was performed by adding 0.2 mL enzyme solution into substrate solution and determining the reaction mixture at A240 nm. The assay for APX activity was revised as adding 0.2 mL enzyme solution into substrate solution and determining the reaction mixture at A290 nm. The reaction mixture of POD, CAT, and APX was determined every 30 s during a total 180 s reaction time, respectively.

Conclusions
In summary, this study identified 38 OfZAT genes, which could be classified into 5 subgroups based on phylogenetic relationships and sequence structures. The members of each subgroup contained similar motifs arranged in consistent patterns. In addition, 49 segmental and 5 tandem duplication events were detected among OfZATs and might have contributed to the expansion of the ZAT gene family. The ZAT genes predominantly contained cis-elements associated with hormone, stress, light, and developmental processes. Numerous OfZAT genes showed tissue-specific expression patterns. Screening of the transcriptome data revealed two genes that could be induced by salt stress. Under cold stress, eight genes were strongly and differently expressed, of which OfZAT35 showed a continuously increasing expression trend. OfZAT35 is located in the nucleus, and it showed no transcriptional activation activity. The transient overexpression of OfZAT35 in tobacco resulted in significantly higher REL values and downregulated the expression of positive to cold tolerance genes, such as NbDREB3, NbLEA5, and NbCAT, which suggested that OfZAT35 is a negative regulator of cold tolerance in transgenic tobacco. Overall, this study not only expands the understanding of the ZAT gene family in O. fragrans but also provides a basis for the further functional evaluation of OfZATs under cold stress.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/plants12122346/s1, Figure S1: Synteny analysis of ZAT gene family across the O. fragrans genome, Figure S2: The phylogenetic classification and motif composition of 89 ZAT genes, Figure S3: The cis-element screening in the promoter regions (2-kb genome sequence) of OfZATs, Figure S4: The expression profiles of OfZATsin the salt stress treatment (S0 h, S6 h, S24 h, S72 h) and control (S0 h, CK6 h, CK24 h, CK72 h), Figure S5: The expression profiles of OfZATs during cold stress treatment,