Overexpression of Liriodendron tulipifera JAG Gene (LtuJAG) Changes Leaf Shapes in Transgenic Arabidopsis thaliana

In Arabidopsis thaliana, JAGGED (JAG) is a transcription inhibitor that controls the development of leaf polarity and regulates the expression of genes controlling lateral organ formation. Liriodendron tulipifera is an ornamental tree with extraordinary tulip-shaped flowers and goose web-like leaves, this is one of the suitable plants for morphological development research. To investigate the potential functions of the LtuJAG gene, we isolated the full-length LtuJAG from L. tulipifera, transferred it into A. thaliana via agrobacterium-mediated transformation, and monitored its expression pattern. Subcellular localization showed that LtuJAG was located in the nucleus. RT-qPCR assays indicated that LtuJAG was expressed mainly in leaf buds and flowers, but not in mature leaves and stems. GUS staining results showed that LtuJAG was expressed in the shoot apical meristem (SAM). Overexpressing LtuJAG changed A. thaliana leaf shapes, causing a moderate serration and a slight asymmetric distribution in the medio-lateral and proximal-distal axes. Ectopic expression of LtuJAG induced the expression of lateral organ boundary suppressors JAGGED LATERAL ORGANS (JLO) and ARABIDOPSIS THALIANA HOMEOBOX1 (ATH1). It also repressed the expression of the apical meristem suppressor class-1 KNOX gene (KNOX I) and altered endogenous hormone levels. Our results suggest that LtuJAG plays a role in negatively regulating leaf polarity formation in L. tulipifera.


Introduction
The organ development of seed plants usually undergoes two opposite modes: (1) root and stem development, which displays overall radial symmetry; (2) development of lateral organs such as leaves and flowers, which show an obvious asymmetry [1]. The leaves of seed plants evolved from lateral branches, initiating from the peripheral zone (PZ) of the shoot apical meristem (SAM) and developing into flattened structures [2]. Genetic evidence has revealed that leaves underwent four morphological changes during their evolution: curling into a cone, developing short and flexible petioles as well as serrated edges, forming compound pinnate leaves, and stacking with each other along the branches to adjust the phyllotaxis and reduce drag [3]. Due to their morphological plasticity, leaves are ideal research subjects for under-standing organ morphogenesis and environmental adaptability.
The development of the leaves includes three processes: leaf initiation, leaf polarity establishment, and leaf morphology modulation [4]. Leaves emerge from a mass of stem cells at the SAM via lateral differentiation, thereby forming leaf primordia. The stem cells differentiate and proliferate along the asymmetric polarity of the adaxial-abaxial, medio-lateral, and proximal-distal axes to establish leaf polarity [5]. After the establishment of leaf polarity, its shape is formed by continues cell division and expansion. Many genes involved in leaf development have been identified, including KNOX I family genes,

Cloning and Sequence Analysis of LtuJAG
Full-length LtuJAG (1344 bp; ORF, 801 bp) encoded a protein comprising 267 amino acid residues with a molecular weight of 29.50 kDa and a theoretical isoelectric point (pI) of 8.05. The hydrophilicity of the protein was −0.856 with an instability coefficient of

Cloning and Sequence Analysis of LtuJAG
Full-length LtuJAG (1344 bp; ORF, 801 bp) encoded a protein comprising 267 amino acid residues with a molecular weight of 29.50 kDa and a theoretical isoelectric point (pI) of 8.05. The hydrophilicity of the protein was −0.856 with an instability coefficient of 58.05, indicating that it was likely to be unstable and hydrophobic. The amino acid sequence of LtuJAG was used as a query to perform a BLASTP search to explore its structural conservation and phylogenetic position. Thirteen homologous protein sequences from other plant species showed that the N-terminus comprising the C 2 H 2 zinc-finger domain was conserved while the EAR motif was not, indicating that LtuJAG belongs to the C 2 H 2 zinc-finger family (Figure 2A). The alignment of these sequences showed a high degree of conservation relative to LtuJAG, varying from 91.35% (Magnolia wufengensis, QAY29215.1) to 51.05% (Populus euphratica, XP_011007835.1). The phylogenetic tree showed that JAG proteins from L. tulipifera and M. wufengensis were clustered together ( Figure 2B).  58.05, indicating that it was likely to be unstable and hydrophobic. The amino acid sequence of LtuJAG was used as a query to perform a BLASTP search to explore its structural conservation and phylogenetic position. Thirteen homologous protein sequences from other plant species showed that the N-terminus comprising the C2H2 zinc-finger domain was conserved while the EAR motif was not, indicating that LtuJAG belongs to the C2H2 zinc-finger family (Figure 2A). The alignment of these sequences showed a high degree of conservation relative to LtuJAG, varying from 91.35% (Magnolia wufengensis, QAY29215.1) to 51.05% (Populus euphratica, XP_011007835.1). The phylogenetic tree showed that JAG proteins from L. tulipifera and M. wufengensis were clustered together ( Figure 2B).
The highly conserved core sequence Homeodomain is represented by a red box. The EAR motif is indicated by a yellow box; (B) Phylogenetic relationships among JAG proteins from different plant species. LtuJAG is indicated by the red box.

Subcellular Localization of LtuJAG
Subcellular localization using LOCALIZER (http://localizer.csiro.au/) (accessed on 2 July 2020) [24] predicted that LtuJAG would be found in the nucleus. Green fluorescent protein (GFP) signals of the 35S::eGFP construct were inserted into the nucleus and cell membrane, while the signal of LtuJAG::eGFP was only localized in the nucleus of N. benthamiana leaves ( Figure 3B). DAPI (4 ,6 -diamidino-2-phenylindole) dye was utilized to stain the nucleus to stimulate a fluorescent blue signal. The resulting fluorescent cyan signal further confirmed that LtuJAG was located in the nucleus, which is consistent with transcription factor patterns ( Figure 3B). to stain the nucleus to stimulate a fluorescent blue signal. The resulting fluorescent cyan signal further confirmed that LtuJAG was located in the nucleus, which is consisten with transcription factor patterns ( Figure 3B).

Tissue Expression Pattern of LtuJAG
To verify the expression pattern of LtuJAG in different tissues of L. tulipifera, we examined the mRNA of LtuJAG in eight L. tulipifera tissues ( Figure 4A). Expression in flower buds was significantly higher than that of the other tissues, followed by expres sion in the stamen, petals, calyx and leaf buds, with stems and leaves showing the no expression. This suggests LtuJAG may play an important role in the development of flower organs.
The time-specific expression pattern of LtuJAG during leaf development determined by sampling the leaf bud growth stage (stages 1 and 2), young leaf stage (stage 3), mature stage (stages 4 to 6), and senescence stage (stage 7) showed that LtuJAG expression in creased with the gradual expansion of leaf buds at the bud growth stage ( Figure 4B). The expression of LtuJAG decreased with the growth of leaves at the leaf-spreading stage LtuJAG was not expressed during leaf maturation and senescence, indicating that it may be expressed during vigorous meristem growth of L. tulipifera but does not participate in the regulation of leaf growth and development at maturity.
To further confirm the leaf space-specific expression pattern of LtuJAG, we divided the leaf margin of L. tulipifera into tooth tips and sinuses. LtuJAG expression was mark edly higher in the petiole of the leaf than in the other parts, and it was generally higher in leaf tooth sinuses (d) than in leaf tooth tips (a, c, and e) ( Figure 4C). These results sugges that LtuJAG may play a direct regulatory role in leaf margin development as well.

Tissue Expression Pattern of LtuJAG
To verify the expression pattern of LtuJAG in different tissues of L. tulipifera, we examined the mRNA of LtuJAG in eight L. tulipifera tissues ( Figure 4A). Expression in flower buds was significantly higher than that of the other tissues, followed by expression in the stamen, petals, calyx and leaf buds, with stems and leaves showing the not expression. This suggests LtuJAG may play an important role in the development of flower organs.
The time-specific expression pattern of LtuJAG during leaf development determined by sampling the leaf bud growth stage (stages 1 and 2), young leaf stage (stage 3), mature stage (stages 4 to 6), and senescence stage (stage 7) showed that LtuJAG expression increased with the gradual expansion of leaf buds at the bud growth stage ( Figure 4B). The expression of LtuJAG decreased with the growth of leaves at the leaf-spreading stage. LtuJAG was not expressed during leaf maturation and senescence, indicating that it may be expressed during vigorous meristem growth of L. tulipifera but does not participate in the regulation of leaf growth and development at maturity.
To further confirm the leaf space-specific expression pattern of LtuJAG, we divided the leaf margin of L. tulipifera into tooth tips and sinuses. LtuJAG expression was markedly higher in the petiole of the leaf than in the other parts, and it was generally higher in leaf tooth sinuses (d) than in leaf tooth tips (a, c, and e) ( Figure 4C). These results suggest that LtuJAG may play a direct regulatory role in leaf margin development as well.

LtuJAG Promoter Is Leaf-Specific in Transgenic A. thaliana
To detect tissue specificity, the LtuJAG promoter fragment (named pLtuJAG) located approximately 2Kb upstream of the JAG gene in L. tulipifera, was cloned, sequenced and analyzed for potential regulatory elements (Table S2), including light-responsive cis-elements, such as Box 4, G-box and I-box, as well as TCCC-motifs. The abiotic stress response elements included LTR (low temperature), MYC (cold), STRE (thermal induction), TC-rich repeats (defense and stress responsiveness), and WRE3 (damage). Furthermore, a few hormone-responsive elements were identified, such as ABRE (ABA), AuxRE (Auxin), AuxRR-core (Auxin), ERE (ethylene), TATC-box (GA), and TCA-element (salicylic acid). Sequence analysis of the LtuJAG promoter region implies that the LtuJAG gene may be implicated in regulating a range of processes, including responses to light signals, abiotic stresses, and hormones.
To confirm the core elements responsible for meristem development, the pLtuJAG sequence was fused to the GUS reporter gene (pLtuJAG::GUS) and transformed into A. thaliana, pLtuJAG-induced GUS expression was observed during the two-leaf, four-leaf, six-leaf and flowering stages of T2 overexpression lines using histochemical staining ( Figure 5A,B). After germination for 2 days, GUS activity of seeds and seedlings was not expressed. At the stage of leaf bud differentiation (4-12 days), GUS expression was not detected in the radicle. Concurrently, the LtuJAG promoter drove GUS to be stably expressed in leaves and hypocotyls. With the growth of seedlings, GUS activity in leaf buds gradually decreased, while gradually increasing in newly differentiated leaves ( Figure  5B). It was only expressed in leaves, but not in inflorescences, pods, and stems at the reproductive stage ( Figure 5B). Furthermore, we observed that the promoter activity of

LtuJAG Promoter Is Leaf-Specific in Transgenic A. thaliana
To detect tissue specificity, the LtuJAG promoter fragment (named pLtuJAG) located approximately 2Kb upstream of the JAG gene in L. tulipifera, was cloned, sequenced and analyzed for potential regulatory elements (Table S2), including light-responsive cis-elements, such as Box 4, G-box and I-box, as well as TCCC-motifs. The abiotic stress response elements included LTR (low temperature), MYC (cold), STRE (thermal induction), TCrich repeats (defense and stress responsiveness), and WRE3 (damage). Furthermore, a few hormone-responsive elements were identified, such as ABRE (ABA), AuxRE (Auxin), AuxRR-core (Auxin), ERE (ethylene), TATC-box (GA), and TCA-element (salicylic acid). Sequence analysis of the LtuJAG promoter region implies that the LtuJAG gene may be implicated in regulating a range of processes, including responses to light signals, abiotic stresses, and hormones.
To confirm the core elements responsible for meristem development, the pLtuJAG sequence was fused to the GUS reporter gene (pLtuJAG::GUS) and transformed into A. thaliana, pLtuJAG-induced GUS expression was observed during the two-leaf, four-leaf, six-leaf and flowering stages of T 2 overexpression lines using histochemical staining ( Figure 5A,B). After germination for 2 days, GUS activity of seeds and seedlings was not expressed. At the stage of leaf bud differentiation (4-12 days), GUS expression was not detected in the radicle. Concurrently, the LtuJAG promoter drove GUS to be stably expressed in leaves and hypocotyls. With the growth of seedlings, GUS activity in leaf buds gradually decreased, while gradually increasing in newly differentiated leaves ( Figure 5B). It was only expressed in leaves, but not in inflorescences, pods, and stems at the repro-ductive stage ( Figure 5B). Furthermore, we observed that the promoter activity of LtuJAG was the strongest in the main leaf vein, followed by the secondary leaf vein, and weak GUS expression was detected in mesophyll cells, indicating that the LtuJAG promoter is exclusively expressed in leaf organs. From these observations, it can be inferred that pLtuJAG is a leaf bud meristem-specific promoter. LtuJAG was the strongest in the main leaf vein, followed by the secondary leaf vein, and weak GUS expression was detected in mesophyll cells, indicating that the LtuJAG promoter is exclusively expressed in leaf organs. From these observations, it can be inferred that pLtuJAG is a leaf bud meristem-specific promoter.

Phenotype of Transgenic A. thaliana
The coding sequence of LtuJAG was fused to the 35S promoter (35S::LtuJAG) and transferred into A. thaliana. Approximately 32 T1 generation positive plants were produced. Seven of these overexpression lines had the phenotype of small leaf area (LtuJAG-OE11), including two with a phenotype producing serration, LtuJAG-OE7 and LtuJAG-OE8 ( Figure 6A,C). The transgenic line 35S::LtuJAG-OE7 exhibited a phenotype causing small leaf area, we also measured lamina length, width, and area in wild type and overexpression lines. Compared with the wild type, the petiole of the overexpression strain became longer ( Figure 6E), the lamina length, width, and area were reduced (Figure 6F,G,I) The ratio of lamina length to width remained unchanged ( Figure 6I), indicating that the overexpression line 35S::LtuJAG-OE7 altered leaf size by simultaneously reducing the rate of leaf cell expansion in both the medio-lateral and proximal-distal axis directions. A range of defects was observed, including irregular curls and lobes at the distal ends of the leaves ( Figure 6D), thinner mesophyll tissue in the lobed part of the leaves, abnormal vein growth, and narrowing of the proximal ends forming a wedge. This latter also induced some petiole elongation, and only petioles remained in some severely defective leaves. This phenotypic change in the absence of leaf cells suggests suppression of leaf polarity had occurred.
The transgenic line 35S::LtuJAG-OE8 exhibited a phenotype causing petiole elongation and leaf serration at the four-leaf stage ( Figure 6B) while showing dorsal leaf margin curling at the mature stage ( Figure 6A,C). This was caused by accelerated proliferation and expansion of mesophyll cells on the adaxial blade, while the cell division speed on the abaxial blade remained unchanged. To a greater extent, the leaves showed a moder-

Phenotype of Transgenic A. thaliana
The coding sequence of LtuJAG was fused to the 35S promoter (35S::LtuJAG) and transferred into A. thaliana. Approximately 32 T 1 generation positive plants were produced. Seven of these overexpression lines had the phenotype of small leaf area (LtuJAG-OE11), including two with a phenotype producing serration, LtuJAG-OE7 and LtuJAG-OE8 ( Figure 6A,C). The transgenic line 35S::LtuJAG-OE7 exhibited a phenotype causing small leaf area, we also measured lamina length, width, and area in wild type and overexpression lines. Compared with the wild type, the petiole of the overexpression strain became longer ( Figure 6E), the lamina length, width, and area were reduced ( Figure 6F,G,I) The ratio of lamina length to width remained unchanged ( Figure 6I), indicating that the overexpression line 35S::LtuJAG-OE7 altered leaf size by simultaneously reducing the rate of leaf cell expansion in both the medio-lateral and proximal-distal axis directions. A range of defects was observed, including irregular curls and lobes at the distal ends of the leaves ( Figure 6D), thinner mesophyll tissue in the lobed part of the leaves, abnormal vein growth, and narrowing of the proximal ends forming a wedge. This latter also induced some petiole elongation, and only petioles remained in some severely defective leaves. This phenotypic change in the absence of leaf cells suggests suppression of leaf polarity had occurred.
The transgenic line 35S::LtuJAG-OE8 exhibited a phenotype causing petiole elongation and leaf serration at the four-leaf stage ( Figure 6B) while showing dorsal leaf margin curling at the mature stage ( Figure 6A,C). This was caused by accelerated proliferation and expansion of mesophyll cells on the adaxial blade, while the cell division speed on the abaxial blade remained unchanged. To a greater extent, the leaves showed a moderate serration and slight asymmetric distribution along the medio-lateral axis, which was consistent with the heteromorphic leaves of a mutant [20]. ate serration and slight asymmetric distribution along the medio-lateral axis, which was consistent with the heteromorphic leaves of a mutant [20].  The asterisks indicate significant difference from the WT using the Student's t-test (** p < 0.01, * p < 0.05).

Expression of Genes Related to Leaf Development in Transgenic A. thaliana
Since the formation of leaf morphology occurs during the transition from leaf primordia initiation to polarity establishment, we selected 22 genes that have been reported to play a part in this process. The heterologous overexpression of LtuJAG resulted in a significant increase in the expression of LtuJAG in overexpression lines (OE7, OE8, and OE11), indicating that we successfully identified the overexpression lines of T 3 generation (Figure 7). Compared with the wild type, 35S::LtuJAG increased the expression of IPT7 and decreased the expression of GA20ox1 at the leaf primordia, suggesting that JAG promotes antagonism between GA20ox1 and IPT7, thus maintaining the dynamic balance of hormones during development. 35S::LtuJAG downregulated the expression of STM, KNAT2, SBP-LIKE9 (SPL9), and CUC2/3, and upregulated the expression of AS2, JLO, BOP1, SAW1/2, ATH1, AUX1 and TEOSINETE BRANCHED1/CYCLOIDEA/PROLIFERATING CELL FACTOR 4 (TCP4) at the leaf primordia. The expressions of KNAT1, KNAT6, PIN1, AS1, BOP2, TCP2, and ER were not significantly different between WT and the overexpression lines ( Figure 7). Therefore, we speculate that JAG may participate in leaf primordia differentiation and leaf polarity formation through multiple regulatory pathways, but determining by which regulatory pathways would require further investigation.

Expression of Genes Related to Leaf Development in Transgenic A. thaliana
Since the formation of leaf morphology occurs during the transition from leaf primordia initiation to polarity establishment, we selected 22 genes that have been reported to play a part in this process. The heterologous overexpression of LtuJAG resulted in a significant increase in the expression of LtuJAG in overexpression lines (OE7, OE8, and OE11), indicating that we successfully identified the overexpression lines of T3 generation (Figure 7). Compared with the wild type, 35S::LtuJAG increased the expression of IPT7 and decreased the expression of GA20ox1 at the leaf primordia, suggesting that JAG promotes antagonism between GA20ox1 and IPT7, thus maintaining the dynamic balance of hormones during development.  Figure 7). Therefore, we speculate that JAG may participate in leaf primordia differentiation and leaf polarity formation through multiple regulatory pathways, but determining by which regulatory pathways would require further investigation.  Quantitative RT-PCR analysis of LtuJAG and genes in relation to leaf development in the ten-day-old wild type (WT) and three independent LtuJAG overexpression lines, OE7, OE8 and OE11. ACTIN2 served as the reference gene. The value of genes expression level in WT was set to "1" as a control, the error bars represent the standard deviation among three biological replicates. Data were analyzed statistically using the one-way ANOVA Tukey's test. Different letters indicate significant differences at p < 0.01 and p < 0.05 level.

Domain Function in LtuJAG
A 1344 bp cDNA full-length sequence named LtuJAG encoding 267 amino acid proteins was identified within the leaf bud cDNA of L. tulipifera. Sequence analysis of Ltu-JAG showed that LtuJAG had similar conserved domains with JAG in Ricinus communis, Populus trichocarpa, Jatropha curcas, Helianthus annuus, and Brassica rapa. A single 31 amino acid C 2 H 2 type Zinc-finger motif (amino acid position 60-91), which may have a role in DNA binding, and a short nine amino acid ERF-associated amphiphilic expression (EAR) motif (amino acid positions 250-258) at the amino acid end were observed to function as repressors. These results demonstrate that JAG belongs to the C 2 H 2 type zinc-finger transcription factor family. Homology analysis showed that the amino acid sequence of LtuJAG was 51.05-91.35% similar to that of other species, especially the woody plant M. wufengensis. LtuJAG is predicted to function similarly to AtJAG, a class of transcriptional inhibitors that regulate processes involved in organ morphological development and polar growth [25].
Promoters are central to processes of transcriptional regulation and fundamental to the study of gene expression and regulation. Analysis of cis-acting elements showed that their most common products were hormones, such as aux (AuxRE), GA (TATC-box), abscisic acid (ABRE), salicylic acid (TCA-element), and ethylene (ERE). The functions of auxin and GA are associated with the developmental function of JAG in regulating the three-dimensional morphology of plant organs. Moreover, plants synthesize abscisic acid to initiate defense mechanisms against stress [26]. Salicylic acid and ethylene are endogenous signaling molecules that activate plant hypersensitivity and systemic acquired resistance [27,28]. MYB and WRKY are involved in regulating the dual response of plants to biotic and abiotic stresses [29,30]. Therefore, we speculate that the LtuJAG gene is a transcription factor that participates in regulating development and resistance to stress.

Expression Pattern of LtuJAG during Leaf Shape Formation
Our study showed that LtuJAG was located in the nucleus of tobacco, while by observing AtJAG::GFP fusion protein under control of AP1 promoter in the epidermal cells of pedicels of flower primordia, AtJAG was located in the nucleus. This suggests that JAG has the characteristics of a typical transcription factor [21]. LtuJAG was specifically expressed in flower organs, such as flower buds, petals, and stamens. High concentrations of JAG RNA in rice have been detected in the inflorescence meristem, axial meristem, flower organ primordia, stamen, pistil, anther, and integument [31][32][33], demonstrating that JAG has a certain impact on the development of each floral organ. LtuJAG was only expressed in leaf buds at the vegetative stage, and not in mature leaves or stems, which is consistent with other observations of A. thaliana by Dinneny et al. [21]. To gain further insight into the expression pattern of LtuJAG, we observed that the LtuJAG promoter was stably expressed in leaves and hypocotyls at the vegetative stage of Arabidopsis, and its expression in the leaves gradually increased over time. AtJAG promoter-driven GUS gene expression was highest in the radicles and hypocotyls 12 h after germination, gradually decreased in the cotyledons and hypocotyls at 3 and 4 days, and increased in the radicles [34], which further explains the specific expression of JAG at the early stage of apical meristem differentiation.
In conclusion, we demonstrated that JAG is a transcription factor that regulates leaf and floral organ development.

Overexpression of LtuJAG Results in Leaf Polarity Defect
The adaxial axes of leaf primordia face towards the meristem while abaxial axes face away from it. The formation of the adaxial-abaxial axis is fundamental for establishing the medio-lateral axis and proximal-distal axis of the leaf [1]. Overexpression of LtuJAG in A. thaliana resulted in the curling of adaxial tissue towards the abaxial axis. Cytological observations indicate that defects within the thick-walled cells on the abaxial axis of leaves are responsible for curling [35], indicating that LtuJAG affects cell development. Intrigu-ingly, we found that LtuJAG was ectopically expressed in leaves with deep serrations at the tip and margin. Study of KIP-RELATED PROTEIN2 (KRP2) revealed that the appearance of deep serration was caused by a proliferation defect in total cells, and the degree of lobe formation was related to the number of cells [36]. Moreover, the SHALLOT-LIKE1 (SLL1) gene in rice controls the development of leaf adaxial thick-walled cells by regulating programmed cell death (PCD) to establish leaf adaxial polarity [35]. Thus, the rate of cell division and proliferation can directly affect the final morphology of an organ, but the mechanisms by which cell division is controlled are not well understood. On the other hand, the establishment of adaxial-abaxial, proximal-distal, and medio-lateral axis polarities appears to be coordinated rather than independent of each other. How the development of these three polarities is coordinated remains to be discovered.

LtuJAG Overexpression Alters the Expression of Genes Related to Leaf Morphogenesis
Numerous studies have confirmed the inhibitory effect of JAG on the development of petal polarity, and that petals are derivatives of the apical meristem and homologues of leaves [20][21][22][23]. Therefore, we speculated that JAG has an effect on the development of leaf polarity. The mutation of JAG in Arabidopsis resulted in edge serration of leaves and petals [21]. The overexpression of JAG promoted the fusion of cotyledons to varying degrees and had no effect on the growth of mature leaves [20]. We confirmed that LtuJAG influenced the establishment of three polar axes of leaves, and further examined the expression of related genes. We observed that LtuJAG significantly affected other developmental genes at the differentiation stage of leaf primordia. We further confirmed that LtuJAG significantly downregulated the expression of the KNOX I gene, STM, and KNAT2, but had little effect on KNAT1 and KNAT6. This expression pattern is also consistent with the finding that KNAT1 and KNAT6 do not participate in the regulation of early morphogenesis in leaf primordia [7,8]. Previous studies have confirmed that AS1/2, BOP1/2, and SAW1/2 act as negative regulators of KNOX I in the development of leaf primordia [9,11]. Our study found that LtuJAG can upregulate the expression of the KNOX I negative regulator, indicating that JAG may indirectly downregulate the expression of KNOX I during the development of lateral meristematic tissue. LtuJAG also accelerated the expression of the meristem boundary inhibitor genes ATH1 and JLO. ATH1 induces the expression of KNOX I downstream genes by forming heterodimers with STM and BP [37]. Similarly, JLO can activate the expression of STM and BP [16]. In conclusion, we speculate that JAG may synergize or promote KNOX I upstream activators in leaf primordia differentiation and participate in the positive regulation of KNOX I upstream inhibitors.
Other boundary inhibitors, such as CUC1/2/3, have been found to be activated by STM during the development of apical meristem, and CUC2 is regulated by STM and restricted to the SAM boundary region [38]. The inactivation of TCP function contributes to the excessive proliferation of leaf margins, resulting in leaf expansion and curled or serrated leaflets [39][40][41]. Regulation of leaf margin development by TCP transcription factors is accomplished by the negative regulation of CUC1/2/3. TCP4 forms dimers with CUC2 and CUC3, respectively, inhibiting the expression of CUC2/3 [42]. SPL9 competitively binds TCP to form dimers with CUC [42], and it is likely that the expression of SPL9 also affects the expression of CUC. This study found that LtuJAG can increase the expression of TCP4 and downregulate the expression of SPL9 and CUC2/3, indicating that JAG can also participate in the TCP4-CUC2/3 regulation pathway, but whether it is relative by JAG regulating SPL9 expression with this pathway requires further investigation.
Gibberellin and brassinolide control leaf growth by promoting cell proliferation and expansion [43,44]. PIN1 is also required for organ initiation, boundary restriction, and growth [45]. ERECTA (ER) can control cell proliferation by inhibiting auxin expression [46]. In LtuJAG overexpression lines, the expression of PIN1 and ER was not significantly different from that of the wild type, indicating that JAG may not be directly related to auxin biosynthesis. As a downstream target gene of KNOX I, IPT7 in A. thaliana was induced to be expressed by STM during apical meristem differentiation [47]. STM also inhibited gibberellin activity by reducing the level of GA20ox1. In addition, the expression of GA20ox1 in tomato was negatively correlated with the degree of leaf margin serration, whereas IPT7 was positively correlated with it [48]. LtuJAG significantly upregulated the expression of IPT7 and decreased the expression of GA20ox1, indicating that JAG can also participate in the antagonism between hormones, thereby contributing to the differentiation of meristem organs. Whether it regulates hormones through the KNOX I pathway requires further research. In conclusion, we speculate that the boundary limiting factors JLO and ATH1, the upstream activators of the KNOX I gene, were directly activated by JAG, resulting in leaf primordia with disorganized polarity that sometimes fail to complete morphological development (Figure 8), but further studies are needed to determine the validity of this hypothesis. The mechanism by which genes controlling leaf morphogenesis interact at different developmental stages is unclear. Investigating this will be a focus of our future research.
Gibberellin and brassinolide control leaf growth by promoting cell proliferation and expansion [43,44]. PIN1 is also required for organ initiation, boundary restriction, and growth [45]. ERECTA (ER) can control cell proliferation by inhibiting auxin expression [46]. In LtuJAG overexpression lines, the expression of PIN1 and ER was not significantly different from that of the wild type, indicating that JAG may not be directly related to auxin biosynthesis. As a downstream target gene of KNOX I, IPT7 in A. thaliana was induced to be expressed by STM during apical meristem differentiation [47]. STM also inhibited gibberellin activity by reducing the level of GA20ox1. In addition, the expression of GA20ox1 in tomato was negatively correlated with the degree of leaf margin serration, whereas IPT7 was positively correlated with it [48]. LtuJAG significantly upregulated the expression of IPT7 and decreased the expression of GA20ox1, indicating that JAG can also participate in the antagonism between hormones, thereby contributing to the differentiation of meristem organs. Whether it regulates hormones through the KNOX I pathway requires further research. In conclusion, we speculate that the boundary limiting factors JLO and ATH1, the upstream activators of the KNOX I gene, were directly activated by JAG, resulting in leaf primordia with disorganized polarity that sometimes fail to complete morphological development (Figure 8), but further studies are needed to determine the validity of this hypothesis. The mechanism by which genes controlling leaf morphogenesis interact at different developmental stages is unclear. Investigating this will be a focus of our future research.

Plant Materials and Growth Conditions
We found that the number of leaf serration increased with time in L. tulipifera and ended at senescence. Moreover, in a previous study, we had determined the timing and location of leaf initiation, leaf polarity establishment, leaf serration formation and leaf morphogenesis in L.chinense and L. tulipifera using SEM and paraffin section observation [49]. Our focus was on whether LtuJAG genes were involved in the process of altering the number of leaf serrations. Therefore, L. tulipifera trees were studied, originating from South Carolina, USA, were planted in a provenance trial plantation in the Xiashu Forest Farm in Jurong County, Jiangsu Province, China (119 • 13 20 E, 32 • 7 8 N) in 1993 [50].
To investigate the expression pattern of LtuJAG in different tissues, leaves, stems, stamens, pistils, flower buds, sepals, leaf buds, and petals of L. tulipifera were collected in April 2018. To verify the time-specific expression pattern, leaves at different stages of development (leaf bud germination, young leaves, mature leaves, and senescence) were collected from March to August 2018 ( Figure 9A). As our focus lies in studying the leaf polarity establishment, leaf serration formation and leaf morphogenesis, we collected leaf buds and leaf from the corresponding stages, namely leaf bud germination stage (P1 and P2), young leaf stage (P3), leaves with three lobes, five lobes and six lobes, divided into mature stage (P4 and P6) and senescence stage(P7). Current studies emphasized that the generation of leaf margin serrations requires hormonal convergence at the leaf tooth tip, while boundary genes are confined at the leaf tooth sinus [45]. To explore the effect of JAG genes as boundary suppressors on different parts of leaf in L. tulipifera, leaves were categorized into the protruding parts of the leaf margin (a, c, and e), concave parts of the leaf margin (b and d), petiole (f), and the middle part of the leaf (g) (Figure 9B), three replicates from each of the seven samples (approximately 1 g each) were separated using sterilized scissors. All samples were frozen immediately in liquid nitrogen and stored at −80 • C for RNA extraction, subsequent molecular cloning and real-time quantitative PCR. stamens, pistils, flower buds, sepals, leaf buds, and petals of L. tulipifera were collected in April 2018. To verify the time-specific expression pattern, leaves at different stages of development (leaf bud germination, young leaves, mature leaves, and senescence) were collected from March to August 2018 ( Figure 9A). As our focus lies in studying the leaf polarity establishment, leaf serration formation and leaf morphogenesis, we collected leaf buds and leaf from the corresponding stages, namely leaf bud germination stage (P1 and P2), young leaf stage (P3), leaves with three lobes, five lobes and six lobes, divided into mature stage (P4 and P6) and senescence stage(P7). Current studies emphasized that the generation of leaf margin serrations requires hormonal convergence at the leaf tooth tip, while boundary genes are confined at the leaf tooth sinus [45]. To explore the effect of JAG genes as boundary suppressors on different parts of leaf in L. tulipifera, leaves were categorized into the protruding parts of the leaf margin (a, c, and e), concave parts of the leaf margin (b and d), petiole (f), and the middle part of the leaf (g) ( Figure  9B), three replicates from each of the seven samples (approximately 1 g each) were separated using sterilized scissors. All samples were frozen immediately in liquid nitrogen and stored at −80 °C for RNA extraction, subsequent molecular cloning and real-time quantitative PCR.
Transgenic and wild-type plants were grown in a Columbia-0 (Col-0) background with Nicotiana benthamiana, and planted in a 6:3:1 mixture of potting soil: vermiculite: perlite in a 23 °C illumination incubators under long days (16 h light/8 h dark, 80% humidity).

Rapid Amplification of cDNA Ends (RACE), Sequencing, and Sequence Analysis of LtuJAG
The JAG EST number (gnl|Liriodendron|b4_c58305) was obtained from the transcriptome database of L. tulipifera (http://jlmwiki.plantbio.uga.edu/aagp/) (accessed on 24 July 2018). RACE cloned primers were designed using Oligo 7 software (Table S1). RACE was conducted using a SMARTer ® RACE 5/3 kit (Takara Biomedical Technology, Beijing, China) according to the manufacturer's instructions for amplifying full-length LtuJAG, and the amplified PCR product was then ligated into the pEASY vector using a

Rapid Amplification of cDNA Ends (RACE), Sequencing, and Sequence Analysis of LtuJAG
The JAG EST number (gnl|Liriodendron|b4_c58305) was obtained from the transcriptome database of L. tulipifera (http://jlmwiki.plantbio.uga.edu/aagp/) (accessed on 24 July 2018). RACE cloned primers were designed using Oligo 7 software (Table S1). RACE was conducted using a SMARTer ® RACE 5/3 kit (Takara Biomedical Technology, Beijing, China) according to the manufacturer's instructions for amplifying full-length LtuJAG, and the amplified PCR product was then ligated into the pEASY vector using a pEASY-Blunt Zero cloning kit (Transgen Biotech, Beijing, China) and verified by sequencing (GenScript Biomedical Technology, Nanjing, China).
LtuJAG Open Reading Frames (ORFs) were predicted using the NCBI ORF finder. The physicochemical characteristics of LtuJAG were analyzed using Expasy Protparam (https: //web.expasy.org/protparam/) (accessed on 16 October 2021). Multiple alignments of JAG amino acid sequences from various plant species were edited using the DNAMAN 8.0 software. A phylogenetic tree of the aligned sequences was created with MEGA 5.1 software using the neighbor-joining method based on 1000 bootstrap replicates [51]. The promoter was analyzed using the online software PlantCARE to predict putative cis-elements.

Subcellular Localization
To investigate localization where LtuJAG functions in the leaf cell, its subcellular localization was performed using the Agrobacterium injection method as described earlier by Lin et al. [52]. Briefly, the CDS of LtuJAG (without a stop codon) was inserted into the pBI121 vector containing the 35S promoter with an eGFP protein, according to the CloneExpress ® Ultra One Step Cloning Kit instructions (Vazyme Biotech, Nanjing, China). Next, the constructed 35S::LtuJAG-GFP and the positive control 35S::eGFP plasmids were inserted into Agrobacterium GV3101, which was then injected into the N. benthamiana leaves [53]. To monitor the location of LtuJAG more precisely, the nucleus-specific dye DAPI was utilized after transfection at 22 • C for 24 h in the dark. We observed the green fluorescent protein (eGFP) in the tobacco leaves at 2 days after transfection using a Zeiss LSM 710 confocal microscope (Carl Zeiss AG, Jena, Germany). Green fluorescence signals may appear in the cell membrane, nucleus, and cytoplasm. DAPI is a blue fluorescence signal that represents the nucleus location.

Agrobacterium-Mediated Transformation
To generate the LtuJAG-overexpressing transgenic A. thaliana, the LtuJAG coding region was cloned into an improved pBI121-GUS vector downstream of the 35S promoter through the XbaI and BamHI sites to generate the plasmid 35S::pBI121-LtuJAG. This plasmid was inserted into Agrobacterium GV3101, which was transferred to A. thaliana (Col-0) using the floral-dip method [54]. The T 0 seeds of the transgenic plants were screened using a 1/2 MS solid medium containing 50 mg/L kanamycin. The positive seedlings could grow normally about 10 days after seed germination-also observed in transgenic lines of the T 1 generation. The selection of homozygous transgenic lines through T 2 and T 3 generations was performed as described by Mahmood et al. [55]. To monitor the transcript levels of transgene in LtuJAG transgenic lines, total RNA was isolated from the rosette leaves of WT (Col-0) and LtuJAG transgenic plants at 28 days after seed germination using a DP441-RNAprep Pure Plant Kit (Tiangen, Beijing, China) according to the manufacturer's instructions. Primer pairs, LtuJAG-qRT-F and LtuJAG-qRT-R, were used in quantitative RT-PCR. AtACTIN2 was used as an internal control to normalize the expression values of the transgene transcript level [56]. The used primers were listed in Table S1. The relative expression was calculated using the 2 −∆∆Ct method [57]. Three biological replicates and three technical replicates. Data were analyzed statistically using the one-way ANOVA Tukey's test.

GUS Histochemical Assay and Phenotypic Characterization
It is known that JAG is a class of transcription factor that is specifically expressed in the apical meristem. To investigate LtuJAG promoter expression pattern during differentiation of the apical meristem, we cloned promoter sequences. About 2Kb bp fragments upstream of the translational region of the LtuJAG gene were inserted into the plasmid pBI121-GUS at the corresponding restriction sites in place of the 35S promoter region, thereby forming a pBI121-pLtuJAG::GUS construct and transformed into A. thaliana. T 2 generation transgenic positive lines were used as research material after screening and identification, these seeds and seedlings at the two-leaf, four-leaf, and six-leaf stages were collected, respectively, at 0, 2, 4, 6, 8, 12 days after seed germination. To understand the tissue-expression pattern, we collected leaves, flowers, stems and siliques from the reproductive stage at 42 days after seed germination, of which GUS activity was detected using GUS histochemical analysis, specific method as previously described by Jefferson et al. [58]. First, tissue samples (1 g each) were submerged in 10 mL X-gluc solution (0.1% 5-bromo-4-chloro-3-indolyl-beta-Dglucuronic acid (X-gluc), 1% dimethyl formamide, 50 mM sodium phosphate [pH 7.0]) at 37 • C for 24 h. The treated tissues were then decolorized in 75-100% ethanol. The samples were photographed using a Leica DM500 microscope (Shanghai, China).
To explore the effect of the LtuJAG gene on leaf morphology, the 5-12th rosette leaves (except 1-4th cotyledons) of the T 3 generation homozygous WT and LtuJAG overexpression lines grown for 32 days, were photographed using a Fujifilm Finepix S5100 digital camera (Tokyo, Japan), and the JPEG images were used to calculate the lamina length, lamina width, petiole length, lamina area of 12 replicated individual lines using Digimizer (www. digimizer.com) (accessed on 12 November 2021) software, the specific method referring to Luo et al. [59]. We aimed at studying the leaf size differences between the wild-type and overexpression strains and therefore data were analyzed statistically using Student's t-test to pairwise comparison, and differences were considered significant when values p < 0.01 and p < 0.05.

Real-Time Quantitative PCR
To study the temporal expression pattern of genes associated with leaf serration, we examined the transcript levels of leaf serration related genes in wild-type A. thaliana leaves at different growth stages and confirmed all genes were specifically expressed at leaf primordia differentiation (10 days after seed germination), and to some extent, some genes were not even expressed at the maturation (25 days after seed germination), reproductive (40 days after seed germination), and senescent stages (50 days after seed germination). To explore the effect of LtuJAG on leaf serration related genes, triplicate seedling samples of the homozygous WT and LtuJAG overexpression lines (with the lobed leaf phenotype) at the initial leaf primordia stage (10 days after seed germination), were frozen in liquid nitrogen and stored at −80 • C in preparation for gene expression testing using real-time quantitative PCR.
Total RNA was extracted from different samples of A. thaliana and L. tulipifera using a total RNA extraction kit (Tiangen, Beijing, China). cDNA was synthesized using the PrimeScript II 1st strand cDNA Synthesis Kit (TaKaRa Biomedical Technology). Quantitative real-time PCR was performed in triplicate using a SYBR Premix Ex Taq kit (TaKaRa Biomedical Technology) on an Applied Biosystems real-time PCR system. AtACTIN2 of A. thaliana and LcActin97 of Liriodendron chinense were used as internal controls with the primers listed in Table S1 [56,60]. For PCR, initial denaturation was performed at 95 • C for 60 s, which was followed by 40 cycles of 5 s at 95 • C and 34 s at 60 • C [61]. Three biological replicates and three technical replicates. The relative expression was calculated using the 2 −∆∆Ct method [56]. Data were analyzed statistically using the one-way ANOVA Tukey's test.

Conclusions
In this study, we cloned and characterized the LtuJAG gene involved in the development of leaf polarity in L. tulipifera. This gene is a nuclear localized transcription inhibitor that is highly expressed in the lateral organs. We speculate that LtuJAG activates the boundary limiting factors JLO and ATH1 and upregulates the negative regulator of the KNOX I, thereby causing a downregulation of the KNOX I gene. This may disrupt the balance of endogenous hormones, affecting the development of leaf primordia and producing defective leaves. Our study provides evidence that JAG plays a role in the development of leaf polarity.