Overexpression of the Liriodendron tulipifera BOP2 Gene (LtuBOP2) Affects Leaf Margin Development in Transgenic Arabidopsis thaliana

BLADE-ON-PETIOLE 2 (BOP2) plays a pivotal role in leaf morphogenesis. Liriodendron tulipifera is a suitable model for exploring the molecular mechanisms underlying leaf serration formation, which are largely unknown. Here, we isolated the full-length LtuBOP2 gene and its promoter from L. tulipifera and characterized its function in leaf morphogenesis through multidimensional approaches. The spatiotemporal expression pattern of LtuBOP2 indicated the high expression of LtuBOP2 in stems and leaf buds. We constructed LtuBOP2 promoter, fused the promoter sequences to the β-glucuronidase (GUS) gene, and then transformed them into Arabidopsis thaliana. Histochemical GUS staining results indicated that GUS activity was higher in petioles and the main vein. LtuBOP2 overexpression in A. thaliana caused moderate serration in the leaf tip, owing to the increased number of abnormal lamina epidermal cells and defective vascular tissue, thus indicating a novel role of BOP2. The ectopic expression of LtuBOP2 in A. thaliana promoted the expression of the lateral organ boundary gene ASYMMETRIC LEAVES2 (AS2) and inhibited JAGGED (JAG) and CUP-SHAPED COTYLEDON2 (CUC2) expression to establish leaf proximal–distal polarity. Moreover, LtuBOP2 participated in leaf serration formation by promoting the antagonistic relationship between KNOX I and hormones during leaf margin development. Our findings revealed the role of LtuBOP2 in the proximal–distal polarity formation and development of leaf margin morphology, providing new insights into the regulatory mechanisms of the leaf formation development of L. tulipifera.


Introduction
The growth and propagation of plants are always inseparable from energy supply; since energy production requires the assistance of photoreception, respiration, and transpiration in the leaves, leaves are considered the main source organs for higher plants to produce energy by photosynthesis [1,2]. Accordingly, the leaf size and shape influence the growth condition and developmental quality of plants [3]. During the evolutionary process of leaves, the changes in leaf size and shape can enhance their adaptability to natural environments. For example, plants with leaf serrations can improve their adaptability to stressful conditions, such as drought, heat, and wind, by adjusting their irradiation area. Furthermore, serrated leaves can improve the light transmittance of the tree crown and have advantage in light interception for lower layer leaves in shade conditions [4,5]. Therefore, leaves are an ideal system to explore plant organ morphogenesis.
Leaf development is an iterative process of organ initiation from the meristem, which can be divided into three stages: primordium initiation (PI), primary morphogenesis (PM), and secondary morphogenesis (SM) [6]. First, the stem cells originate as the main lateral appendage of the stem in the peripheral zone (PZ) of the shoot apical meristem (SAM) [7], where cells divide rapidly and can differentiate at specific sites to form flower primordium, lateral branch primordium, and leaf primordium [8]. Further, during PM, a boundary region separates the proximal region of the lateral organ from SAM [9]. Leaves with vigorous differentiation potential grow along the three-dimensional axis (proximal-distal, adaxial-abaxial, and medio-lateral axes), thereby establishing of leaf polarity [10,11]. The cells divide and differentiate continuously to form adaxial-abaxial axis, developing from a mass cell to a flat structure [12,13]. Meanwhile, in the PM stage, local cells at the leaf margin proliferate constantly, then serrations occur at the mature leaf margin. [14]. Finally, the leaf undergoes cell expansion, division, and differentiation to expand the leaf area and develop stomata and epidermal hairs along the polarity axis, concluding the leaf shape development [15].
Determination of the leaf shape involves a complex genetic network and hormones [16,17]. KNOX I and boundary genes are reportedly involved in leaf development. The KNOX I family genes include SHOOT MERISTEMLESS (STM), BREVIPEDICELLUS (BP or KNAT1), KNOT-TED IN Arabidopsis thaliana 2 (KNAT2), and KNAT6. Initiation of leaf founder cells requires repressed expression of KNOX I; ASYMMETRIC LEAVES1 (AS1) transcription factor plays an essential role in repressing KNOX I expression [18]. Once cells begin to divide and differentiate, the new leaf primordium separates from the SAM to form a boundary region, and the involved genes are collectively called boundary genes. These boundary genes include ASYMMETRIC LEAVES2, BLADE-ON-PETIOLE 1/2 (BOP1/2), JAGGED (JAG), JAGGED LATERAL ORGANS (JLO), A. thaliana HOMEOBOX1 (ATH1), and CUP-SHAPED COTYLE-DON (CUC). When the leaf primordium originates at the PZ, CUC is specifically expressed in boundary positions and interacts with the KNOX I protein to promote boundary formation [19]. CUC expression is influenced by auxin maxima and determined by the expression pattern of PINFORMED1 (PIN1). The PIN1-CUC regulation pathway performs important roles in organ boundary establishment, determining the position and size of serration [20]. JAG overexpression results in a smaller leaf and a phenotype with leaf margin serration [21]. As a transcription inhibitor, JAG controls the distal axis development of leaves by regulating cell division and is itself regulated by BOP [22]. In A. thaliana leaf formation, BOP regulates cell activity in lateral organs and polarity establishment in proximal-distal axis [23]. BOP1/2 proteins can dimerize in vivo and associate directly with the AS2 promoter, thereby regulating leaf morphogenesis by repressing KNOX I and JAG transcription [24]. Conserved mechanisms for BOP-mediated leaf morphogenesis with patterning along the proximaldistal axis have now been identified in several species. Some examples include NOOT in Medicago truncatula and COCHLEATA/COCH in Pisum sativum (A. thaliana BOP orthologs genes), whose mutants have serrated phenotypes similar to bop [25]; SlBOP2 regulates leaf shape development in tomato (Solanum lycopersicum) by repressing KNOX I expression [26], suggesting a conserved function of BOP in leaf organ development. In summary, although the molecular mechanism between BOP and KNOX I has been reported in detail, it remains unclear whether there is a regulatory relationship between BOP and other genes involved in leaf development.
Liriodendron, a genus of the Magnoliaceae family, is a tertiary relict plant, of which only two species remain-L. chinense and L. tulipifera. Liriodendron plants are commonly used as ornamental trees, owing to their peculiar leaf shapes and tulip-like flowers. In recent years, with the advent of next-generation sequencing and release of the Liriodendron genome, research on leaf morphogenesis and development in Liriodendron plants has been increasing [27]. Yang and co-authors reported that nine NAC genes were involved in leaf senescence regulation [28]. Ma and co-authors defined four developmental stages of young leaves in Liriodendron based on the results from using paraffin sections and scanning electron microscopy (SEM); they determined that the leaf petiole develops in the P2 phase, indicating the establishment of proximal-distal polarity, and that morphological features such as leaf veins and serration could be observed in the P3 phase [29]. Further studies revealed that LcKNOX6, homologous to AtKNAT6 in A. thaliana, negatively regulates AtK-NAT1, AtGA20ox1, AtPIN1, and AtAS2 expression, which causes the emergence of leaf lobes and ultimately affects leaf shape development [30]. Li et al. verified that LcNAM was expressed in all parts of the leaf, and LcCUC1/2/3 was a specific expression factor for the leaf margin. Functional analysis via genetic transformation illustrated that CUC1 contributes to the formation of the crinkled leaves phenotype, and CUC2 increases the number of leaves and causes different degrees of serration [31]. Based on these results, Wen et al. elucidated the mechanism by which LcCUC2 affects leaf lobes, showing that the ectopic expression of LcCUC2 caused the leaf petiole cotyledons to become longer and narrower, upregulated the KNAT2/6 expression, and ultimately influenced the development of leaf margin serrations by regulating auxin content [32]. Wei et al. reported that LtuJAG overexpression altered leaf margin shape in A. thaliana, resulting in leaves with serration and asymmetric differentiation and growth in the medio-lateral and proximal-distal positions [22]. Briefly, although these studies have uncovered some aspects of the mechanisms underlying leaf shape development in Liriodendron, the precise molecular mechanism remains unclear. Moreover, growth and developmental processes in plants are controlled by the expression of genes, while gene transcription levels are influenced by promoters containing cis-acting elements [33]. In this study, to reveal the potential function of LtuBOP2 in the leaf margin development of Liriodendron plants, we isolated LtuBOP2 and its promoter from L. tulipifera and then ectopically expressed it in A. thaliana wild-type and bop2 mutants. Consequently, we obtained a 35S::LtuBOP2 transgenic plant with smaller and lobed leaves and confirmed the regulatory relationship between LtuBOP2 and other leaf development genes. Our results provided a potential genetic mechanism of leaf margin development, as well as insights for breeding ornamental cultivars in L. tulipifera.

Cloning and Bioinformatics Analysis of LtuBOP2
The BOP2 protein was analyzed using the NCBI Conserved Domain software and obtained as a member of the BTB family with BTB-POZ (Poxvirus Zinc Finger Protein) domain and ankyrin-repeats (Ank_2) domain ( Figure 1). BOP proteins are members of the small NON-EXPRESSOR OF PATHOGENESIS-RELATED GENES (NPR) family, characterized by a BTB-POZ domain at the N-terminus and ankyrin-repeats (ANK) located near the C-terminus. LtuBOP2 has sequence similarity with the BOP2 of other species in protein alignment. Multiple alignment results showed that these sequences were highly homologous, and all contained conserved N-terminal and polymorphic C-terminal sequences. L. tulipifera BOP2 has the closest genetic relationship with Tripterygium wilfordii BOP2 and is relatively close to Theobroma cacao BOP2. However, it is relatively far from Pisum sativum BOP2, Medicago truncatula BOP2, and A. thaliana BOP2, which are more highly evolved, indicating that this amino acid sequence is highly conserved in the BTB domain and belongs to the BTB family, named LtuBOP2.

Spatiotemporal Expression Pattern of LtuBOP2
As shown in Figure 2A, the quantitative analysis results revealed that the BOP2 gene had different expression levels in the eight tissues of L. tulipifera. The expression in stem was significantly the highest among the other tissues. The expression in the flower bud and leaf bud was higher than that in the other tissues (excluding the stem), and a lower expression was observed in the petals, calyx, stamens, pistils, and leaves. Leaf buds and flower buds originated from leaf primordium and flower primordium, respectively, and both contained the shoot apical meristem, while the stems contained the lateral meristem, suggesting that LtuBOP2 is involved in meristem development.

Spatiotemporal Expression Pattern of LtuBOP2
As shown in Figure 2A, the quantitative analysis results revealed that the BOP2 gene had different expression levels in the eight tissues of L. tulipifera. The expression in stem was significantly the highest among the other tissues. The expression in the flower bud and leaf bud was higher than that in the other tissues (excluding the stem), and a lower expression was observed in the petals, calyx, stamens, pistils, and leaves. Leaf buds and flower buds originated from leaf primordium and flower primordium, respectively, and both contained the shoot apical meristem, while the stems contained the lateral meristem, suggesting that LtuBOP2 is involved in meristem development. In L. tulipifera, BOP2 expression was the highest in the leaf base, followed by leaf middle, and lowest in the leaf margin ( Figure 2B). In L. chinense, BOP2 expression was also highest in the leaf base, followed by leaf margin, and lowest in leaf middle. Collectively, these results indicate that BOP2 was mainly expressed in the leaf base in L. tulipifera and L. chinense.
The time-specific expression condition of LtuBOP2 was surveyed at 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) ( Figure 2C). The LtuBOP2 expression decreases in correspondence with the expansion of leaves in the growth stage of leaf bud, and this expression is higher than that in the mature stage. This suggests that BOP2 may be expressed mainly during leaf bud differentiation to control organ development.
To further detect the leaf spatial-specific expression of LtuBOP2, we subjected the tooth tips (a, c, and e), tooth sinuses (b and d), petiole (f), and middle part (g) of L. tulipifera leaves to quantitative reverse transcription-PCR (RT-qPCR). The LtuBOP2 transcript level was visibly highest in the leaf petiole and substantially higher in the leaf tooth sinuses (b and d) than in the leaf tooth tips (a, c, and e) ( Figure 2D). These results indicate that LtuBOP2 may be involved in the formation of petiole development.
middle, and lowest in the leaf margin ( Figure 2B). In L. chinense, BOP2 expression was also highest in the leaf base, followed by leaf margin, and lowest in leaf middle. Collectively, these results indicate that BOP2 was mainly expressed in the leaf base in L. tulipifera and L. chinense.
The time-specific expression condition of LtuBOP2 was surveyed at 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) ( Figure 2C). The LtuBOP2 expression decreases in correspondence with the expansion of leaves in the growth stage of leaf bud, and this expression is higher than that in the mature stage. This suggests that BOP2 may be expressed mainly during leaf bud differentiation to control organ development.
To further detect the leaf spatial-specific expression of LtuBOP2, we subjected the tooth tips (a, c, and e), tooth sinuses (b and d), petiole (f), and middle part (g) of L. tulipifera leaves to quantitative reverse transcription-PCR (RT-qPCR). The LtuBOP2 transcript level was visibly highest in the leaf petiole and substantially higher in the leaf tooth sinuses (b and d) than in the leaf tooth tips (a, c, and e) ( Figure 2D). These results indicate that LtuBOP2 may be involved in the formation of petiole development. tulipifera. a, c, and e: leaf tooth protrudent, b and d: leaf tooth sinus, g: leaf middle, f: petiole. The one-way ANOVA Tukey's test was applied for statistical analysis. The error bars indicated the standard deviation between the three biological repeats. The capital letters indicate significant differences at p < 0.01 level, the lowercase letters indicate significant differences at p < 0.05 level.

pLtuBOP2 Sequence Analysis
To determine tissue specificity, the LtuBOP2 promoter fragment (pLtuBOP2), situated approximately 2 Kb upstream of BOP2 in L. tulipifera, was separated and analyzed for potential regulatory elements (Supplementary Materials Table S2), including light-responsive cis-elements, G-box, GT1-motif, ACE, AE-box, GTGGC-motif, and (B) Expression pattern of BOP2 in the leaf base, leaf center and leaf margin of L. tulipifera and L. chinense, respectively; (C) Expression pattern of BOP2 in developmental stage of leaf in L. tulipifera. Stage1-Stage2: leaf bud germination stage; Stage3: young leaf stage; Stage4-Stage6: mature stage; Stage7: senescence stage; (D) Expression pattern of BOP2 in different parts of leaf of L. tulipifera. a, c, and e: leaf tooth protrudent, b and d: leaf tooth sinus, g: leaf middle, f: petiole. The one-way ANOVA Tukey's test was applied for statistical analysis. The error bars indicated the standard deviation between the three biological repeats. The capital letters indicate significant differences at p < 0.01 level, the lowercase letters indicate significant differences at p < 0.05 level.

pLtuBOP2 Sequence Analysis
To determine tissue specificity, the LtuBOP2 promoter fragment (pLtuBOP2), situated approximately 2 Kb upstream of BOP2 in L. tulipifera, was separated and analyzed for potential regulatory elements (Supplementary Materials Table S2), including lightresponsive cis-elements, G-box, GT1-motif, ACE, AE-box, GTGGC-motif, and chs-Unit 1 m1. The abiotic stress response elements included STRE (thermal induction), LTR (low temperature), and as-1 (cold). Moreover, a few hormone-responsive elements were identified, including ABRE (ABA), P-box (GA), CGTCA-motif (Methyl Jasmonate), TATC-box (GA), and TGACG-motif (Methyl Jasmonate). In addition, pLtuBOP2 specifically contained meristem expression regulatory elements, i.e., the CAT-box. The sequence analysis of the LtuBOP2 promoter region suggested that LtuBOP2 may be involved in regulating a range of processes, including abiotic stresses, responses to hormones, and meristem development. Figure 3B shows GUS gene expression driven by pLtuBOP2 at the two-leaf, four-leaf, six-leaf, and flowering stages of A. thaliana. GUS gene expression driven by pLtuBOP2 was not detected in 2d, 4d, and 6d seedlings. After leaf bud differentiation at 8-12d, GUS activity was strongest in the newly differentiated petiole, but it was almost nonexistent in the radicle, hypocotyl, and leaf. The LtuBOP2 promoter drove GUS expression in leaves and stems but not in inflorescences and pods. In addition, we observed that LtuBOP2 promoter activity was highest in petiole and the main vein, followed by the secondary vein. Among different tissues, pLtuBOP2 activity was found to be the highest in the stem, indicating that LtuBOP2 plays a role in the apical meristem at the early stage of leaf bud differentiation; further, it was strongly expressed in the vascular bundle of mature tissues.
LTR (low temperature), and as-1 (cold). Moreover, a few hormone-responsive elements were identified, including ABRE (ABA), P-box (GA), CGTCA-motif (Methyl Jasmonate), TATC-box (GA), and TGACG-motif (Methyl Jasmonate). In addition, pLtuBOP2 specifically contained meristem expression regulatory elements, i.e., the CAT-box. The sequence analysis of the LtuBOP2 promoter region suggested that LtuBOP2 may be involved in regulating a range of processes, including abiotic stresses, responses to hormones, and meristem development. Figure 3B shows GUS gene expression driven by pLtuBOP2 at the two-leaf, four-leaf, six-leaf, and flowering stages of A. thaliana. GUS gene expression driven by pLtuBOP2 was not detected in 2d, 4d, and 6d seedlings. After leaf bud differentiation at 8-12d, GUS activity was strongest in the newly differentiated petiole, but it was almost nonexistent in the radicle, hypocotyl, and leaf. The LtuBOP2 promoter drove GUS expression in leaves and stems but not in inflorescences and pods. In addition, we observed that LtuBOP2 promoter activity was highest in petiole and the main vein, followed by the secondary vein. Among different tissues, pLtuBOP2 activity was found to be the highest in the stem, indicating that LtuBOP2 plays a role in the apical meristem at the early stage of leaf bud differentiation; further, it was strongly expressed in the vascular bundle of mature tissues.

Phenotypic Analysis of LtuBOP2 Transgenic Plants
To monitor the effect of LtuBOP2 on plant leaf morphogenesis, we transformed the LtuBOP2 into A. thaliana wild-type and bop2 mutants. To facilitate the further investigation of the leaf shape, we used a stereoscopic microscope to observe the leaf structure and found that 35S::LtuBOP2-1 and 35S::LtuBOP2-2 had shorter leaf lengths at the proximal-distal axis than the wild-type and 35S::LtuBOP2/bop2 ( Figure 4). Moreover, the leaf base morphology was unchanged, and there was a clear serration at the tip of the plant leaves. These results suggest that BOP2 affects the establishment of leaf proximal-distal polarity.
The network system connecting the leaf veins determines the leaf shape and size, and the leaf vein formation is related to polarity establishment. An observation of leaf veins with the serrated phenotype revealed a reduction in the number of lateral veins and slender veins differentiated in 35S::LtuBOP2 and 35S::LtuBOP2/bop2. LtuBOP2 may have altered the vascular development of leaves to affect leaf shape, and leaf tip serration may be attributed to the incomplete differentiation of the plant's distal axis.

Phenotypic Analysis of LtuBOP2 Transgenic Plants
To monitor the effect of LtuBOP2 on plant leaf morphogenesis, we transformed the LtuBOP2 into A. thaliana wild-type and bop2 mutants. To facilitate the further investigation of the leaf shape, we used a stereoscopic microscope to observe the leaf structure and found that 35S::LtuBOP2-1 and 35S::LtuBOP2-2 had shorter leaf lengths at the proximaldistal axis than the wild-type and 35S::LtuBOP2/bop2 ( Figure 4). Moreover, the leaf base morphology was unchanged, and there was a clear serration at the tip of the plant leaves. These results suggest that BOP2 affects the establishment of leaf proximal-distal polarity.
The network system connecting the leaf veins determines the leaf shape and size, and the leaf vein formation is related to polarity establishment. An observation of leaf veins with the serrated phenotype revealed a reduction in the number of lateral veins and slender veins differentiated in 35S::LtuBOP2 and 35S::LtuBOP2/bop2. LtuBOP2 may have altered the vascular development of leaves to affect leaf shape, and leaf tip serration may be attributed to the incomplete differentiation of the plant's distal axis.

Leaf Morphology of Transgenic A. thaliana Lines
To investigate the effect of LtuBOP2 ectopic expression on leaf development in A. thaliana, we measured petiole length, lamina length, lamina width, lamina area, number of leaf, plant height, and the ratio of lamina length to width at 32 d after germination ( Figure 5). The 35S::LtuBOP2 plants were shorter and smaller than the wild-type in terms of petiole length, leaf length, leaf width, and leaf area; 35S::LtuBOP2/bop2 plants were also shorter and smaller in all aspects than bop2 plants, which suggested that BOP2 overexpression inhibited normal leaf development. LtuBOP2 expression in bop2 mutants was not very different from that observed in the wild-type. Although the leaf-width values of 35S::LtuBOP2/bop2 were initially larger than those of the wild-type and mutant, other aspects such as petiole length, leaf length, and aspect ratio were smaller. The loss and restoration of BOP2 had little effect on leaf shape size; thus, BOP2 had a certain inhibitory effect on the leaf size and shape development. However, BOP2 alone was not enough to cause large changes in leaf shape and may coordinate with other genes to jointly control leaf development.

Leaf Morphology of Transgenic A. thaliana Lines
To investigate the effect of LtuBOP2 ectopic expression on leaf development in A. thaliana, we measured petiole length, lamina length, lamina width, lamina area, number of leaf, plant height, and the ratio of lamina length to width at 32 d after germination ( Figure 5). The 35S::LtuBOP2 plants were shorter and smaller than the wild-type in terms of petiole length, leaf length, leaf width, and leaf area; 35S::LtuBOP2/bop2 plants were also shorter and smaller in all aspects than bop2 plants, which suggested that BOP2 overexpression inhibited normal leaf development. LtuBOP2 expression in bop2 mutants was not very different from that observed in the wild-type. Although the leaf-width values of 35S::LtuBOP2/bop2 were initially larger than those of the wild-type and mutant, other aspects such as petiole length, leaf length, and aspect ratio were smaller. The loss and restoration of BOP2 had little effect on leaf shape size; thus, BOP2 had a certain inhibitory effect on the leaf size and shape development. However, BOP2 alone was not enough to cause large changes in leaf shape and may coordinate with other genes to jointly control leaf development. The error bars represent the standard deviation among the three biological repeats. The 5 th, 6 th, 7 th, 8 th, 9 th, 10 th, 11 th and 12 th represent the fifth leaf, the sixth leaf, the seventh leaf, the eighth leaf, the ninth leaf, the tenth leaf, the eleventh leaf and the twelfth leaf of A. thaliana. The one-way ANOVA Tukey's test was applied for statistical analysis. Values are means ± SD (n ≥ 30). The capital letters indicate significant differences at p < 0.01 level, the lowercase letters indicate significant differences at p < 0.05 level. Values are means ± SD (n ≥ 30). The capital letters indicate significant differences at p < 0.01 level, the lowercase letters indicate significant differences at p < 0.05 level.

SEM Observation of Transgenic A. thaliana Lines
To better explore the mechanism of leaf margin serration formation, cells in the leaf epidermal and margin lobe sites of 35S::LtuBOP2 and wild-type A. thaliana were observed using SEM. Leaf epidermal cells of the wild-type were irregular in shape, with small size differences. They were essentially closely arranged in an inlaid manner. Palisade cells on the leaf margins were arranged tightly and neatly in an elongated shape ( Figure 6A-C). The leaf margin of A. thaliana is defined by elongated boundary cells. Compared with the relatively smooth leaf margin of the wild-type A. thaliana, the leaf margin boundary cells of the 35S::LtuBOP2 lines disappeared, followed by the leaf margin formation via an arrangement of exuberantly dividing mesophyll cells. Notably, the epidermal cells at the leaf tooth margins proliferate abnormally, with many filamentous appendages derived from the roots of the epidermal hairs ( Figure 6D-F). These results indicated that LtuBOP2 affected cell proliferation and expansion in the early stages of leaf serration development, and played a prominent role in promoting cell division.

SEM Observation of Transgenic A. thaliana Lines
To better explore the mechanism of leaf margin serration formation, cells in the leaf epidermal and margin lobe sites of 35S::LtuBOP2 and wild-type A. thaliana were observed using SEM. Leaf epidermal cells of the wild-type were irregular in shape, with small size differences. They were essentially closely arranged in an inlaid manner. Palisade cells on the leaf margins were arranged tightly and neatly in an elongated shape ( Figure 6A-C). The leaf margin of A. thaliana is defined by elongated boundary cells. Compared with the relatively smooth leaf margin of the wild-type A. thaliana, the leaf margin boundary cells of the 35S::LtuBOP2 lines disappeared, followed by the leaf margin formation via an arrangement of exuberantly dividing mesophyll cells. Notably, the epidermal cells at the leaf tooth margins proliferate abnormally, with many filamentous appendages derived from the roots of the epidermal hairs ( Figure 6D-F). These results indicated that LtuBOP2 affected cell proliferation and expansion in the early stages of leaf serration development, and played a prominent role in promoting cell division.

LtuBOP2 Affects the Expression of Genes Related to Leaf Shape Development
Upon analyzing the plant appearance, an incision phenotype at the leaf tips was found in the LtuBOP2 transgenic strain. To explore how the LtuBOP2 gene regulates leaf shape development in A. thaliana, we first used the NCBI online database to identify the reported associated genes, mainly the genes associated with BOP2 gene regulation from previous studies. The expression of these genes in A. thaliana seedlings overexpressing LtuBOP2 was detected using RT-qPCR. As shown in Figure 7, we obtained three transgenic lines (OE1, OE2, and OE5) that overexpressed LtuBOP2 at the T3 generation. There was little change in AtSTM and AtPIN1 expression in the bop2 and overexpression lines

LtuBOP2 Affects the Expression of Genes Related to Leaf Shape Development
Upon analyzing the plant appearance, an incision phenotype at the leaf tips was found in the LtuBOP2 transgenic strain. To explore how the LtuBOP2 gene regulates leaf shape development in A. thaliana, we first used the NCBI online database to identify the reported associated genes, mainly the genes associated with BOP2 gene regulation from previous studies. The expression of these genes in A. thaliana seedlings overexpressing LtuBOP2 was detected using RT-qPCR. As shown in Figure 7, we obtained three transgenic lines (OE1, OE2, and OE5) that overexpressed LtuBOP2 at the T3 generation. There was little change in AtSTM and AtPIN1 expression in the bop2 and overexpression lines compared to that in the wild-type. AtBOP1 was upregulated both in bop2 and LtuBOP2-OE lines, demonstrating that BOP1 is redundant with BOP2. In contrast, AtCUC2, AtJAG, and AtIPT7 were downregulated in bop and LtuBOP2-OE lines, indicating that BOP may not act through one pathway but in combination with other genes to co-repress their expression. In LtuBOP2-overexpressing plants, AtBP expression was slightly higher than that in the wild-type, but AtBP expression was significantly increased in the bop2 mutants, indicating that BOP2 may have an inhibitory effect on BP. Furthermore, AtKNAT2, AtAS2, AtGA20ox1, and AtCUC3 were upregulated in the LtuBOP2-OE and downregulated in the bop2 lines, demonstrating that BOP2 is positively correlated with the expression of these genes. In addition, AtAS1, AtKNAT6, and AtTCP2/4 expression did not change in bop2 lines compared to that in the wild-type but was increased in the LtuBOP2-OE lines, indicating that BOP2 may affect their expression indirectly rather than directly. In conclusion, these results indicate that there may be multiple pathways for LtuBOP2 to regulate different genes related to leaf shape development, thus affecting the morphological development of plant leaf margin from different aspects.
LtuBOP2-OE lines, demonstrating that BOP1 is redundant with BOP2. In contrast, At-CUC2, AtJAG, and AtIPT7 were downregulated in bop and LtuBOP2-OE lines, indicating that BOP may not act through one pathway but in combination with other genes to co-repress their expression. In LtuBOP2-overexpressing plants, AtBP expression was slightly higher than that in the wild-type, but AtBP expression was significantly increased in the bop2 mutants, indicating that BOP2 may have an inhibitory effect on BP. Furthermore, AtKNAT2, AtAS2, AtGA20ox1, and AtCUC3 were upregulated in the LtuBOP2-OE and downregulated in the bop2 lines, demonstrating that BOP2 is positively correlated with the expression of these genes. In addition, AtAS1, AtKNAT6, and AtTCP2/4 expression did not change in bop2 lines compared to that in the wild-type but was increased in the LtuBOP2-OE lines, indicating that BOP2 may affect their expression indirectly rather than directly. In conclusion, these results indicate that there may be multiple pathways for LtuBOP2 to regulate different genes related to leaf shape development, thus affecting the morphological development of plant leaf margin from different aspects. Figure 7. RT-qPCR was used to analyze the expression of genes in the 10-day-old wild-type plants (WT), three independent LtuBOP2 overexpression lines (OE1, OE2, and OE5), and bop2. AtACTIN2 served as the reference gene. The gene expression of WT as a control, the error bars indicated the standard deviation between the three biological repeats. The one-way ANOVA Tukey's test was applied for statistical analysis. The error bars indicated the standard deviation between the three biological repeats. The capital letters indicate significant differences at p < 0.01 level, the lowercase letters indicate significant differences at p < 0.05 level. Figure 7. RT-qPCR was used to analyze the expression of genes in the 10-day-old wild-type plants (WT), three independent LtuBOP2 overexpression lines (OE1, OE2, and OE5), and bop2. AtACTIN2 served as the reference gene. The gene expression of WT as a control, the error bars indicated the standard deviation between the three biological repeats. The one-way ANOVA Tukey's test was applied for statistical analysis. The error bars indicated the standard deviation between the three biological repeats. The capital letters indicate significant differences at p < 0.01 level, the lowercase letters indicate significant differences at p < 0.05 level.

Function of Domain in the LtuBOP2 Gene
While our previous studies provided preliminary definitions for the developmental stage and morphological characteristics and identified candidate genes in relation to the leaf development of L. chinense based on anatomical observations and transcriptome assays [28,29], little is known about the regulatory pathway or molecular mechanisms related to leaf shape development. Therefore, we compared the expression of some genes in the database of Ma et al. and identified the key gene BOP2 that may affect leaf shape development [29]. Subsequently, the full-length sequence of LtuBOP2 was cloned from the transcriptome database of L. tulipifera, and its functions were analyzed. BOP2 encodes ANK and a BTB/POZ domain [34], which are widely expressed at the boundary of lateral organs and play a key role in plant morphogenesis [35]; the loss of the BOP2 function affects the structure of leaves, fruits, and flowers. Using the ClusterX software, the sequences of the BOP2 proteins of L. tulipifera and other species were compared. The results showed that these sequences had high homology, with highly conserved BTB and Ank_2 domains. It is speculated that LtuBOP2 may participate in the initial morphogenesis development of plants, similar to that in other species.

Expression Pattern of LtuBOP2 during Growth and Development
During plant development, the SAM gives rise to leaf primordium and lateral branching primordium at their flanks. These meristem processes are all completed at the lateral organ boundaries, which are restricted growth zones that separate the forming organ and meristem, giving rise to the axillary meristem [36]. Subsequently, the formation of internode and axillary meristems in Arabidopsis is inhibited, resulting in the generation of leaves that form a tight rosette. BOP has been extensively studied as a boundary defining gene; Norberg et al. performed in situ mapping of BOP2 mRNA and found that BOP2 was weakly expressed in early leaf primordium in Arabidopsis but more strongly expressed at the site of flower primordium. In our study, LtuBOP2 expression was highest in stems and relatively high in leaf buds and flower buds, indicating that BOP2 was specifically expressed in tissues with vigorous meristem activity, which is consistent with Ha et al.'s research [23]. At the later stages of Arabidopsis development, BOP2 expression is restricted to the base of developing leaves [37]. Our study also confirmed the expression specificity of BOP2 at the base of leaves in Liriodendron. The BOP2::GUS reporter verified the expression pattern of BOP2 at the base of mature embryonic cotyledons [38], and our GUS-staining results in Arabidopsis showed strong pLtuBOP2 expression in the petioles at 8 and 12 days after germination (DAG) and in the stem and main veins of leaves at 42 DAG. Taken together, these results indicate that BOP2 has distinct functions in SAM development and leaf proximal-distal polarity; however, the specific mechanism remains to be investigated.

Roles of LtuBOP2 during Leaf Margin Formation
The meristem originates in the boundary zone. However, in addition to forming the meristem, the boundary also differentiates the proximal parts of the leaflet in the compound leaf and separates individual leaflets from each other [39]. Notably, the boundary zone provides a versatile pattern to maintain morphology establishment. For example, the mechanisms related to leaf primordium initiation also have important functions in leaf development. Furthermore, the axillary and distal leaflet boundaries have pluripotent cell groups that retain the ability to form the meristem [40]. To observe the effects of LtuBOP2 on plant meristem, we ectopically expressed LtuBOP2 in A. thaliana and obtained a short plant phenotype with apical serration. In previous studies, in the Arabidopsis double mutants with bop1bop2 function loss, the leaves grew transversely along the petiole without obvious petioles, which may be caused by the ectopic growth of the germ cells along the petiole [41]. Overexpression of BOP1 or BOP2 led to a reduction in plant height [37], which is consistent with our results. Therefore, the role of BOP in leaf development may be to inhibit cell viability.
The process of leaf differentiation includes the formation of vasculature and specialized leaf cells (such as stomata and vessels). The secondary veins of our transgenic plants (35S::LtuBOP2 and 35S::LtuBOP2/bop2) all showed a reduction in varying degrees, indicating that LtuBOP2 expression resulted in the simplification of the leaf vein architecture system. The leaf blades of the Osbop1/2/3 triple mutant of Oryza sativa L. were shorter, darker, and curlier than those of the wild-type [42]. Defects were observed in the leaf sheath without obvious abnormality at the tip. OsBOPs overexpression led to high activity of the mesenchymal tissues and abnormal cell proliferation in the ligules [42]. Generally, boundary cells in A. thaliana wild-type often exhibit a characteristic elongated shape [43]. We have observed that the LtuBOP2 ectopic expression caused boundary cells to disappear at the leaf margin of A. thaliana and generated many abnormal epidermal hairs, suggesting that BOP2 inhibited the formation and development of boundary cells. In tomato, BOP inhibits leaflet formation. Therefore, BOPs may promote proximal axis development, and regulate the polar differentiation of proximal-distal axis, by inhibiting the differentiation of distal axis differentiation. Notably, A. thaliana has lobed leaf edges with the addition of a characteristic shape of protrusions (e.g., serrations) and indentations (e.g., sinuses). Our observation showed that ectopic meristem proliferation was evident at the leaf tip, where serrations were ectopically derived, whereas site growth at the sinuses of leaf margin sinuses was markedly inhibited. Thus, we hypothesized that LtuBOP2 may act through two different pathways to coordinate the growth and development of serrations at the leaf margin; however, the underlying mechanism is unclear.

BOP2 Affects the Expression of Genes Related to Leaf Development
The lateral organ boundary is a specialized connection separating the initiating lateral organ from the meristem, and the separation of the lateral organ primordium and lateral organ formation is controlled by a complex network of genes [41]. BP and KNAT2 have been implicated in meristem maintenance [44]. CUCs are involved in the establishment of SAM by upregulating STM and other KNOX I [45]. CUCs are also repressed in organ primordium by the primordial marker genes AS1 and AS2 [46]. AS1/2 loss of function mutants are defective in the proximal-distal axis but do not affect adaxial-abaxial growth, which is associated with the ectopic expression of the KNOX gene (BP, KNAT2, and KNAT6) [47]. Furthermore, Ha et al. reported that there was no change in BOP expression in as1as2 double mutants as well as the reduced expression of AS2 in bop1bop2 mutants [23]. Our study also found that BOP2 overexpression led to AS2 upregulation, indicating that BOP contributes to AS2 expression during primordial development.
Boundary genes, a class of genes specifically expressed in the boundary region between the meristem and leaf primordium, function to define the boundary. Recently, several studies confirmed serration development in single leaves, and that boundary formation between leaflets of compound leaves was inseparable from the action of boundary genes [48]. Leaf morphogenesis goes through several stages, and the location of boundary regions varies among stages. Nevertheless, the boundary zone retains functions that control meristem genes, including the meristem-maintaining special genes STM, BP, and CUCs, which are primarily required for promoting KNOX I transcripts to form stem cells in primordium initiation. Notably, these meristem genes are repressed in leaf primordia [23] and flower primordia by BOP 1/2 [49]. However, BOP in turn 1/2 represses KNOX I to establish the lateral boundary of the stem organ and leaf polarity during differentiation [50]. This mechanism has been well demonstrated previously. For example, KNOX I was ectopically expressed in bop1bop2 mature leaves [35]. BP and JAG, which are upregulated in bop1bop2 [41], therefore promote petiole indeterminate, leading to ectopic leaflet and cotyledon formation [23]. JAG ectopic expression suppresses the premature differentiation of tissues, which is necessary for the formation of distal region [21]. Plants with JAG ectopic expression were phenotypically similar to the bop2 mutants, and AtJAG was downregulated in the LtuBOP2-OE lines. These results confirm that the BOP2 gene negatively regulates JAG. Hormones have long been considered to be the ultimate influencers of serration at the leaf margins. The phytohormone gibberellin (GA), which is thought to antagonize KNOX function, is actively degraded by KNOX I activity [51]. Moreover, cytokinin (CK) was positively correlated with KNOX I expression. Our previous studies confirmed that JAG affects the development of apical serrations in leaves by repressing GA20ox1 and promoting IPT7 expression. Conversely, we have now demonstrated that GA20ox1 expression was upregulated, while IPT7 expression was downregulated in BOP2-overexpressing strains, which further confirmed the negative regulatory relationship between BOP2 and JAG ( Figure 8). Altogether, we speculate that BOP2 contributes to normal petioles development and that its regulatory effects may be the partial involvement in the upregulation of the adaxial-abaxial polarity genes AS1/2 and downregulation of the proximal-distal polarity gene JAG during the leaf polarity establishment. More specifically, BOP2 stimulated the antagonism between IPT7 and GA20ox1 by controlling the expression range of KNOX I (BP, KNAT2/6) to repress or regulate hormone content and expression position and to form serrations at leaf margins during leaf morphology development. Although it remains to be determined how phytohormone signaling and gene expression profiles interact with each other, our study provides new insights into the regulatory mechanisms of leaf formation development in L. tulipifera.
which is thought to antagonize KNOX function, is actively degraded by KNOX I activity [51]. Moreover, cytokinin (CK) was positively correlated with KNOX I expression. Our previous studies confirmed that JAG affects the development of apical serrations in leaves by repressing GA20ox1 and promoting IPT7 expression. Conversely, we have now demonstrated that GA20ox1 expression was upregulated, while IPT7 expression was downregulated in BOP2-overexpressing strains, which further confirmed the negative regulatory relationship between BOP2 and JAG ( Figure 8). Altogether, we speculate that BOP2 contributes to normal petioles development and that its regulatory effects may be the partial involvement in the upregulation of the adaxial-abaxial polarity genes AS1/2 and downregulation of the proximal-distal polarity gene JAG during the leaf polarity establishment. More specifically, BOP2 stimulated the antagonism between IPT7 and GA20ox1 by controlling the expression range of KNOX I (BP, KNAT2/6) to repress or regulate hormone content and expression position and to form serrations at leaf margins during leaf morphology development. Although it remains to be determined how phytohormone signaling and gene expression profiles interact with each other, our study provides new insights into the regulatory mechanisms of leaf formation development in L. tulipifera.

Plant Materials
L. tulipifera and L. chinense samples with the same phenological stage were selected as the testing material in the Xiashu Forestry Field (119°13′20″ E, 32°7′8″ N), affiliated with the Nanjing Forestry University (Jurong County, China) [52].

Plant Materials
L. tulipifera and L. chinense samples with the same phenological stage were selected as the testing material in the Xiashu Forestry Field (119 • 13 20 E, 32 • 7 8 N), affiliated with the Nanjing Forestry University (Jurong County, China) [52].
To identify the expression pattern of BOP2 in various plant tissues at the same time, leaf, flower bud, stem, stamen, pistil, sepal, leaf bud, and petal tissues of L. tulipifera and L. chinense were sampled in April 2021. To distinguish the expression of the gene at each location in the leaf, three parts of young L. tulipifera and L. chinense leaves (leaf margin, leaf base, and middle of the leaf) in good condition were collected. To verify BOP2 expression in leaves at several specific periods, L. tulipifera leaves at different developmental stages (the sprouting of leaf bud, immature leaves, mature leaves, and senescence) were collected from April to August 2021. Based on previous studies reporting that the BOP2 is associated with leaf morphogenesis [23], the leaf polarity establishment, and leaf serration formation, we took leaf buds and leaves from the corresponding stages, including leaf bud germination stage (P1 and P2), immature leaf stage (P3), mature stage (P4 to P6), and senescence stage (P7). Previous studies have reported that the BOP2 gene is a regulator of lateral organ growth and development [39]; therefore, we categorized the leaves into different parts, namely the leaf margin (a, c, and e), hollow parts of the leaf margin (b and d), petiole (f), and the middle of the leaf (g), to be explored separately.
A. thaliana wild-type plants were the Columbia (Col) ecotype. The A. thaliana bop2 mutant (SALK_075879C) was purchased from the Non-profit Arabidopsis Share Center (FuZhou, China). A. thaliana seeds were stored at 4 • C until planting. Before sowing, the seed surfaces were disinfected using 75% alcohol and 10% sodium hypochlorite, germinated on 1/2 Murashige for 10 d, and transferred into the soil. Seedlings were precisely cultivated at 23 • C under a long-day photoperiod (14 h day/10 h night), with 70% relative humidity.

cDNA Synthesis, Sequencing, and Sequence Analysis of LtuBOP2
In terms of the known transcriptome database in other reports [29], EST sequences of the annotated BOP2 gene were identified in L. tulipifera. RACE-cloned primers were created using the Oligo 7 software. RACE was implemented through a SMARTer ® RACE 5/3 kit (Takara, Beijing, China) and Phanta ® Max DNA Polymerase (Vazyme Biotech Co. Ltd., Nanjing, China), according to the manufacturer's instructions for amplifying full-length LtuBOP2. Further, they were tested and validated by sequencing (GenScript, Nanjing, China) [53]. Full-length cDNA sequences were obtained through the stitch processing of DNAMAN software (Lynnon Biosoft, San Diego, CA., USA). The open reading frame (ORF) region was predicted by the NCBI ORF Finder (https://www.ncbi.nlm.nih.gov/orffinder/) (accessed on 16 April 2022) to verify the correctness of gene cloning [54]. Protein sequences were aligned from the Cluster W and DNAMAN 8.0 program. Then, a phylogenetic tree was reconstructed with the MEGA7 software, and the Neighbor-Joining (NJ) method based on 1000 bootstrap replicates [55]. The upstream promoter sequence of the BOP2 gene was found based on the validated CDS sequence. The Plant CARE (http://bioinformatics.psb. ugent.be/webtools/plantcare/html) (accessed on 20 May 2022) was operated to predict the core promoter elements of pLtuBOP2.

Recombinant Plasmid Construction and Genetic Transformation
To overexpress the LtuBOP2 gene in A. thaliana, the LtuBOP2 coding regions were amplified and inserted at the XbaI and BamHI sites of the transformed pBI121 vector downstream of the 35S (Cauliflower mosaic virus) promoter using QuickCut enzymes (Takara, Dalian, China); further, the plasmid 35S::pBI121-LtuBOP2 was generated using the One Step Cloning Kit (ClonExpress ® Ultra Vazyme Biotech Co. Ltd., Nanjing, China). We further imported the recombinant plasmids into Agrobacterium tumefaciens (GV3101) and transformed it into A. thaliana, applying the floral dipping approach [56]. Kanamycinresistant primary transformants were gained on 1/2 MS agar plates. Taking the genomic DNA of transgenic leaves in A. thaliana as templates, a PCR reaction was accomplished to identify T1 transgenic plants by exploiting specific primers. We acquired stable genetic transgenic plants after the continuous screening of three generations.

RT-qPCR
To explore the effect of LtuBOP2 on leaf shape formation, we chose some genes that reportedly regulated the leaf serration and designed their primers to verify their transcripts level in 35S::LtuBOP2 transgenic seedlings, bop2, and wild-type. Triplicate seedling samples of the these seedling were prepared at the leaf primordium differentiation stage (10 days after seed germination). Tissue RNA was isolated from various samples of A. thaliana and Liriodendron L using RNA Extraction Kit (Tiangen, Beijing, China). First-strand cDNA was conflated from 4 µg of total RNA through reverse transcription (PrimeScriptTM RT Master Mix, Takara, Dalian, China) [57]. Quantification of gene expression was performed utilizing a SYBR kit (Takara, Dalian, China) and a 96-hole plate (NEST Biotechnology Co. Ltd., Wuxi, China). AtACTIN2 from A. thaliana and LcActin97 from L. chinense were used as an internal control [58]. Gene expression was calculated using the 2 −∆∆CT method with three technical replications and three biological replicates [59].

SEM Observation
To observe cellular changes in the leaf epidermis and leaf serration sites of 35S::LtuBOP2 and wild-type A. thaliana leaves, we sampled the 32-day leaves in an SEM assay. The leaves were placed in a vial, vacuumed with 10 mL of FAA fixative and fixed for 24 h. The fixed material was placed in 50%, 70%, 90%, and 100% ethanol with forceps and dehydrated for 15 min each. The dehydrated material was placed in a 1:2 mixture of isoamyl acetate and alcohol, which was subsequently replaced with 100% isoamyl acetate after 15 min [60]. Then, the samples were treated and observed using SEM, as described by Hepworth et al. [61].

Phenotypic Characterization and Statistical Analysis
The leaves of the homozygous wild-type and LtuBOP2 transformed plants were photographed by using a digital single-lens reflexes camera (Fujifilm Finepix S5100, Tokyo, Japan) to explore the effect of the LtuBOP2 overexpression on leaf morphology. The PNG images were taken to calculate the lamina length, lamina width, petiole length, and ratio of lamina length to width of 30 individual leaves through the Digimizer software (www.digimizer.com) (accessed on 17 September 2022) [62]. Statistical analyses were performed using ANOVA Tukey's test for pairwise comparison. All data are expressed as mean ± standard deviation (SD) of three replicates. Statistical significance was set at p < 0.01 and p < 0.05, and significantly different values were denoted with different letters.

GUS Histochemical Assay and Venation Pattern Analysis
To identify the expression specificity of LtuBOP2 in early leaf development, we cloned the LtuBOP2 promoter sequence, substituted the 35S promoter region with approximately 2 Kb fragment upstream of the translation region of the LtuBOP2, and inserted it into the corresponding restriction endonucleatease site of the plasmid pBI121-GUS to form the pBI121-pLtuBOP2::GUS structure. The plasmid was then transformed into A. thaliana. The seedlings of T2 transgenic positive plants were collected 0, 2, 4, 6, 8 and 12 DAG. To investigate the tissue expression pattern of LtuBOP2, we collected cotyledons, young leaves, inflorescences, stems siliques, and seeds 42 days after seed germination at the reproductive stage. Moreover, we used GUS histochemical analysis to detect their GUS activity, and the experimental method was performed as described by Jefferson et al. [63]. High resolution images were taken using a Leica DM500 microscope (Shanghai, China). To compare the mature leaf vein morphology of transformants and mutants, the leaves were directly bleached with 75% (V/V) ethanol for 24 h without GUS staining and observed by using a Leica DM500 microscope (Shanghai, China).

Conclusions
In this study, we cloned and characterized LtuBOP2, which is associated with leaf shape development in L. tulipifera. Our results confirmed that LtuBOP2 is a pleiotropic gene, it activated the lateral organ boundary gene AS1/2 and suppressed the proximal-distal polarity gene JAG, and inhibited cell over-differentiation to establish polarity during the establishment of leaf polarity. Furthermore, BOP2 promotes ectopic meristem formation at the leaf margin by controlling the KNAT2/6-mediated upregulation of GA20ox1 expression during leaf margin expansion, resulting in serrated leaves during the leaf morphology development. Generally, multiple pathways acted in combination to disrupt the balance of endogenous hormones and affect leaf primordium development with subsequent effects on leaf shape. Our study provides evidence for a novel role of BOP2 in leaf polarity establishment and provides new insights into leaf shape development in L. tulipifera.