Identiﬁcation and Expression Analysis of NAC Gene Family in Weeping Trait of Lagerstroemia indica

: Lagerstroemia indica is a widely used ornamental plant in summer gardens because of its desirable plant shape. The weeping traits of plants are related to secondary cell wall thickness and hormone signaling. NAC (NAM-ATAF1/2-CUC2), as one of the plant-speciﬁc transcription factors, is a switch for the secondary cell wall and also involved in leaf senescence, phytohormone signaling, and other growth processes. We identiﬁed a total of 21 LiNAC genes from the transcriptome data, which we divided into 14 subgroups and 2 groups. The physicochemical characteristics of amino acids, subcellular localization, transmembrane structure, GO and KEGG enrichment, and expression patterns were also examined. The qRT-PCR analysis showed that the expressions of LiNAC8 and LiNAC13 in upright L. indica ‘Shaoguifei’ and weeping L. indica ‘Xiariwuniang’ were signiﬁcantly higher from the beginning to the end of growth stage (S1–S3), and the expressions of ‘Shaoguifei’ were always higher than those of ‘Xiariwuniang’. However, LiNAC2 showed a downward trend in S1–S3 and the relative expression level of ‘Shaoguifei’ was lower than that of ‘Xiariwuniang’. It is hypothesized that these LiNAC genes may be involved in the regulation of weeping traits in L. indica . The results of this study provide a basis for analyzing the functions of LiNAC genes and help to explore the molecular regulatory mechanisms related to the weeping traits in L. indica . stage, the growth stage, end of the stage. signiﬁcant


Introduction
Weeping is a result of drooping branch development. The buds first develop upward, then as the tree grows, the apex of the branches shifts to growing downward. The reasons for the formation of the drooping branches are relatively convoluted, nevertheless. The development of drooping branches may be influenced by genetic make-up, hormones, and secondary growth, but it may also be intimately linked to certain external conditions, such as gravity, light, and supporting forces [1]. The weep gene mutation that led to the gravity anomaly in Prunus persica is what causes the weeping characteristic [2]. The weeping characteristics of Salix babylonica were brought on by a lack of mechanical support, brought on by an excessive trunk elongation [3]. For the weeping of P. mume, the secondary cell walls of branches in the pendant extension zone are thinner than those of upright branches [4]. To modulate the weeping phenotype, IAA and GA 3 biosyntheticpathway-related genes were discovered in P. mume [5,6]. Recently, Zheng et al. found that genes related to cell division, cell development, and plant hormones played an important role in the tortuous-branch phenotype of P. mume [7]. Overexpression of a secondarywall-associated cellulose synthase gene (PtdCesA8) was shown to result in a weeping phenotype in Populus tremuloides [8]. Ornamental plants often attract attention because of their specific traits. A certain plant architecture, known as weeping, has significant aesthetic appeal. Therefore, exploring the causes of plant drooping is potentially valuable for cultivating ornamental plants.
The ability to synthesize proteins via biological activities is controlled by transcription factors and some cis-acting elements, which also govern the spatiotemporal expression of downstream genes and eventually have an impact on the growth and development of organisms [9]. The genes NAM (no apical meristematic tissue) from Petunia hybrida [10], ATAF1/2 (Arabidopsis transcription activation factor) found in Arabidopsis thaliana [11], and CUC2 (cup2 shaped cotyledon) from A. thaliana are the sources of the name NAC [11]. At the N-terminal end of the protein, 150 amino acids make up its conserved structural domain, including five substructural components (A, B, C, D, and E) [12]. According to multiple studies, NAC has a wide spectrum of regulatory effects on plant growth and development [13][14][15][16][17]. Therefore, it is potentially possible to probe the weeping properties of plants from NAC.
Many studies show that the NAC gene family affects the weeping of plants by controlling the synthesis of secondary cell walls [15,18]. Inhibiting the expression of SND1 and NST1 in Arabidopsis at the same time can limit the synthetic gene expression in the three components of secondary cell walls (cellulose, xylem, and lignin), causing the flower stem cells to droop and fail to create secondary walls [15]. Studies on SND4 and SND5 revealed similar functions [19]. While the secondary cell wall of the stem of the overexpressed PtrWND1B-s plants greatly thickened and allowed them to grow upright, the secondary cell wall of the overexpressed PtrWND1B-l plants was unable to do so and displayed a phenotype similar to that of PtrWND1B-RNAi [18]. NAC15 was significantly expressed in the poplar xylem and is most likely the primary factor for transgenic tobacco to develop tall stems [20]. Lignin is an important component in secondary cell walls and EjNAC1 is associated with fruit lignification by activating genes involved in lignin biosynthesis [21]. The main regulatory mechanism in melatonin in enhancing flower stem strength was also to promote lignin accumulation and changed S/G lignin ratio [22]. Co-expression analysis of Pm024213 showed that most of the related genes were involved in auxin and lignin biosynthesis [23].
Because of its lovely blossoms and exquisite branches, the crape myrtle (Lagerstroemia indica L.), a deciduous shrub and small tree, has considerable ornamental value [24]. L. indica branches were classified into flat, upright, and weeping branches by Tian [25]. The study of the molecular mechanisms of L. indica's weeping trait will benefit from an understanding of the NAC gene family function. However, nothing is known about the NAC gene studies in L. indica. In this study, the NAC genes in L. indica were discovered using a variety of bioinformatics techniques. Additionally, the physiological and biochemical properties of the LiNAC proteins were examined and a phylogenetic tree of the LiNAC proteins in L. indica and A. thaliana was created. To clarify whether LiNAC genes are involved in regulating the weeping trait of L. indica, we analyzed the relative expression levels of branches at different growth stages. Overall, the findings might support a theoretical hypothesis of NAC gene family in L. indica.

Identification of NAC Genes in L. indica
Using AtNAC proteins as queries, 60 LiNAC genes were discovered by BLASTP searches in the L. indica transcriptome database and 43 putative LiNAC genes were discovered using an HMMER3 search in the L. indica protein database using the NAC-type NAM model (PF02365). Finally, using SMART and the NCBI Conserved Domain to remove duplicated genes and proteins without NAM-conserved domains, we were able to obtain 21 members of the LiNAC family. For further searching, all genes were given the designations LiNAC1-LiNAC21 in the order of their assembled IDs (Table 1).  9.93 (LiNAC19), with an average value of 6.90. Instability index calculation predicted that 17 (81%) of the NAC proteins were unstable. The LiNAC3, LiNAC4, LiNAC14, and LiNAC15 predicted proteins had instability indices less than 40 and were classified as stable proteins. Aliphatic amino acid indices showed that the thermal stability of the proteins ranged from 55.89 (LiNAC9) to 75.47 (LiNAC14). All LiNACs predicted protein grand average of hydropathicity (GRAVY) values as negative, demonstrating that they were predominantly hydrophilic. The predicted transmembrane structure results showed that LiNAC7 and LiNAC16 proteins had one transmembrane structure at the C-terminal of the 21 LiNAC proteins ( Figure S1). Cell-PLoc subcellular localization predictions suggested that almost all LiNAC proteins were located in the Nucleus (Table 1).

Phylogenetic Analysis and Classification of LiNAC Proteins
We used the neighbor-joining (NJ) method to create a phylogenetic tree with 1000 bootstrap replications to shed light on the evolutionary relationships between proteins from L. indica and A. thaliana. Based on well-established Arabidopsis family classification and their homology with NAC proteins in Arabidopsis, 21 LiNAC proteins were classified into two groups, Group I and Group II, including 14 subgroups ( Figure 1): OsNAC7, ANAC011, NAM, NAC1, ATAF, ANAC3/NAP, SENU5, ONAC022, TIP, NAC2, OsNAC8, ANAC001, ONAC003, and ANAC063. In our analysis, no LiNAC members from the subgroups ANAC011, NAM, ATAF, AtNAC3/NAP, OsNAC8, ONAC003, and ANAC063 were identified. Subgroup TIP had the most members (6) of L. indica and subgroups NAC1 and OsNAC7 had the fewest (1).
The LiNAC proteins clustered in the same subgroup, according to phylogenetic relationships, and may have related roles. For instance, NAC1 members are crucial for auxin signaling and the growth of lateral roots in plants [26]. The NTL9 gene belonging to the TIP subgroup is heavily implicated in plant immune response [27]. The NAC family members in the ATAF, NAP, and AtNAC3 subgroups have a conserved role in stress response [28] and leaf senescence [17,29]. There are two distinct subgroups of NAC proteins that are involved in the formation of vascular vessels and the creation of secondary plant cell walls [30]. SND2, SND3, and other participants were placed into subgroup ONAC003, while NST1, NST2, NST3/SND1, and VNDs were grouped into subgroup OsNAC7.

Conserved Sequence Alignment and Motif Analysis
Multiple sequence alignments were created by Jalview for the 21 LiNAC proteins to investigate the presence and locations of conserved protein domains ( Figure 2) and Weblogo displayed the conserved domains with five subdomains (A-E) ( Figure 3). All LiNAC family members contained a NAM domain ( Figure 4C), which contains a highly conserved A-E subdomain. To further analyze the structural diversity in LiNAC proteins, conserved motifs were searched using the MEME program. In total, 20 distinct motifs were identified ( Figure 4B). Most LiNAC proteins contained motif 3 (representing subdomain A), motif 4 (representing subdomain B), motif 2 (representing subdomain C), motif 1 (representing subdomain D), and motif 5 (representing subdomain E) ( Figure S2). Additionally, the majority of the members in the phylogenetic tree that were closely linked displayed comparable motifs in the same alignment and location ( Figure 4A). Members in the TERN subgroup had nearly identical motifs. Interestingly, some motifs were identified in a certain subgroup. For example, compared with Group I, the number of motifs in Group II was small. Motif 14 was found only in the NAC1 subgroup and motif 16 in the ONAC022 subgroup. The results revealed that LiNAC proteins clustering in the same subgroup may have comparable biological roles and that specific motifs may be associated with specific functions of different subgroups.

Conserved Sequence Alignment and Motif Analysis
Multiple sequence alignments were created by Jalview for the 21 LiNAC proteins to investigate the presence and locations of conserved protein domains ( Figure 2) and Weblogo displayed the conserved domains with five subdomains (A-E) ( Figure 3). All LiNAC family members contained a NAM domain ( Figure 4C), which contains a highly conserved A-E subdomain. To further analyze the structural diversity in LiNAC proteins, conserved motifs were searched using the MEME program. In total, 20 distinct motifs were identified ( Figure 4B). Most LiNAC proteins contained motif 3 (representing subdomain A), motif 4 (representing subdomain B), motif 2 (representing subdomain C), motif 1 (representing subdomain D), and motif 5 (representing subdomain E) ( Figure S2). Additionally, the majority of the members in the phylogenetic tree that were closely linked displayed comparable motifs in the same alignment and location ( Figure 4A). Members in the TERN subgroup had nearly identical motifs. Interestingly, some motifs were identified in a certain subgroup. For example, compared with Group Ⅰ, the number of motifs in Group Ⅱ was small. Motif 14 was found only in the NAC1 subgroup and motif 16 in the ONAC022 subgroup. The results revealed that LiNAC proteins clustering in the same subgroup may have comparable biological roles and that specific motifs may be associated with specific functions of different subgroups.

Annotation and Enrichment Analysis of LiNAC Genes
GO functional annotation of 21 LiNAC genes revealed that 18 of them are involv in biological processes (BPs), cellular components (CCs), and molecular functions (MF We examined the GO enrichment data of 18 LiNAC genes to forecast their biological rol The immunological effector mechanism has the highest degree of enrichment, as show in Figure 5A, scoring 54.4, followed by the response of cells to cold stress, with an enric ment score of 51.72. Additionally, we found that many genes have roles in a variety developmental processes (such as leaf development and shoot development).
As shown in Figure 5B, five LiNAC genes were enriched in the Nucleocytoplasm Transport (ko03013), Phenylpropanoid Biosynthesis Pathway (ko00940), and Signal Pa way (ko04020), according to a KEGG pathway analysis of 21 LiNAC genes.

Annotation and Enrichment Analysis of LiNAC Genes
GO functional annotation of 21 LiNAC genes revealed that 18 of them are involved in biological processes (BPs), cellular components (CCs), and molecular functions (MFs). We examined the GO enrichment data of 18 LiNAC genes to forecast their biological roles. The immunological effector mechanism has the highest degree of enrichment, as shown in Figure 5A, scoring 54.4, followed by the response of cells to cold stress, with an enrichment score of 51.72. Additionally, we found that many genes have roles in a variety of developmental processes (such as leaf development and shoot development).

Determination and Analysis of Lignin Content
The lignin content in L. indica was measured by collecting the branches of 'Shaoguifei' and 'Xiariwuniang' at the early growth stage, the growth stage, and the end of the growth stage. The lignin content in the branches was analyzed separately for different stages and plants. Figure 6C shows that the lignin content in the branches of both 'Shaoguifei' and 'Xiariwuniang' increased gradually with the maturity of the growth period and the lignin content in 'Shaoguifei' was always higher than that in 'Xiariwuniang'. We found that the difference in lignin content between the two branches was not obvious at the early growth stage, but there was a significant difference between the lignin content in 'Shaoguifei' and 'Xiariwuniang' at the growth stage and the end of the growth stage. As shown in Figure 5B, five LiNAC genes were enriched in the Nucleocytoplasmic Transport (ko03013), Phenylpropanoid Biosynthesis Pathway (ko00940), and Signal Pathway (ko04020), according to a KEGG pathway analysis of 21 LiNAC genes.

Determination and Analysis of Lignin Content
The lignin content in L. indica was measured by collecting the branches of 'Shaoguifei' and 'Xiariwuniang' at the early growth stage, the growth stage, and the end of the growth stage. The lignin content in the branches was analyzed separately for different stages and plants. Figure 6C shows that the lignin content in the branches of both 'Shaoguifei' and 'Xiariwuniang' increased gradually with the maturity of the growth period and the lignin content in 'Shaoguifei' was always higher than that in 'Xiariwuniang'. We found that the difference in lignin content between the two branches was not obvious at the early growth stage, but there was a significant difference between the lignin content in 'Shaoguifei' and 'Xiariwuniang' at the growth stage and the end of the growth stage.

Expression Analysis of LiNAC Genes in Regulating the Weeping Trait
To identify which LiNAC genes were involved in regulating the weeping trait of L. indica, we used real-time PCR to analyze the expression patterns of LiNAC genes in branches of 'Shaoguifei' and 'Xiariwuniang' during the early growth stage, the growth

Expression Analysis of LiNAC Genes in Regulating the Weeping Trait
To identify which LiNAC genes were involved in regulating the weeping trait of L. indica, we used real-time PCR to analyze the expression patterns of LiNAC genes in branches of 'Shaoguifei' and 'Xiariwuniang' during the early growth stage, the growth stage, and the end of the growth stage. The expression pattern analysis is shown in Figure 7. The relative expression levels of LiNAC3 and LiNAC20 showed the same trend in the early growth stage (S1), growth stage (S2), and the end of the growth stage (S3) for 'Shaoguifei' and 'Xiariwuniang', which showed an increase first and then a decrease, and the expression level was higher in S2. The relative expression levels of five genes, LiNAC7, LiNAC8, LiNAC14, LiNAC19, and LiNAC11, in 'Shaoguifei' gradually increased from the S1 stage and reached the highest in S3. We found that, except for LiNAC8 and LiNAC11, the relative expression of LiNAC7, LiNAC14, and LiNAC19 showed a trend of increasing and then decreasing in 'Xiariwuniang', with the highest expression in the S2. Interestingly, the expression trend of LiNAC8 in 'Xiariwuniang' is consistent with that in 'Shaoguifei', while the relative expression level of 'Shaoguifei' is slightly higher than that of 'Xiariwuniang' in three stages and the relative expression level is about twice that of 'Xiariwuniang'. We noticed that the relative expression of nine genes (LiNAC1, LiNAC9, LiNAC10, LiNAC16, LiNAC4, LiNAC5, LiNAC9, LiNAC21, and LiNAC15) decreased significantly from S1 to S2 in 'Shaoguifei' and increased gradually in S3. Except for four genes, LiNAC1, LiNAC10, LiNAC16, and LiNAC15, whose relative expression decreased from S1 to S3 in 'Xiariwuniang', all the other five genes showed an increasing and then decreasing trend and their relative expression peaked at S2. The relative expression levels of LiNAC2, LiNAC6, LiNAC12, LiNAC18, and LiNAC13 in 'Shaoguifei' showed a trend of gradual decline and the lowest expression level in S3 (except LiNAC13), especially LiNAC12, in S3 period, the expression level of LiNAC12 was 0.1-times that of 'Shaoguifei'. Interestingly, LiNAC2, LiNAC18, and LiNAC13 also showed the same expression trend as that of 'Shaoguifei' in 'Xiariwuniang'. Notably, the relative expression of LiNAC2 in 'Xiariwuniang' was always higher than that of 'Shaoguifei', especially in the S1-S2 stages. Further, the expression of LiNAC13 was always higher than that of 'Xiariwuniang' in all stages; the relative expression was 9.7-times higher than that of 'Xiariwuniang' in S3.
x FOR PEER REVIEW 11 of 17

The Characteristics of NAC Gene Family in L. indica
In this study, we identified 21 complete LiNAC genes with Open Reading Frames based on L. indica transcriptome data and analyzed their basic information. We predicted and analyzed 21 LiNAC proteins, such as isoelectric power, relative molecular weight, and transmembrane structure, LiNAC7, and LiNAC16 have a transmembrane structure at the C-terminal, and it is speculated that they are likely to be NAC-membrane-binding transcription factors. Almost all LiNAC genes are predicted to be localized in the nucleus (Table 1); LiNACs are probably mostly functional in the nucleus. The conserved structure and motif of LiNACs were also examined ( Figure 4B) and it was shown that all of them Figure 7. The qRT-PCR analysis of 21 LiNAC genes in Upright and Weeping L. indica at different growth stages. S1, S2, and S3 indicate the early growth stage, the growth stage, and the end of the growth stage; significant differences are identified by SPSS with Duncan's test (p < 0.05) and are represented by different letters above the error bars.

The Characteristics of NAC Gene Family in L. indica
In this study, we identified 21 complete LiNAC genes with Open Reading Frames based on L. indica transcriptome data and analyzed their basic information. We predicted and analyzed 21 LiNAC proteins, such as isoelectric power, relative molecular weight, and transmembrane structure, LiNAC7, and LiNAC16 have a transmembrane structure at the C-terminal, and it is speculated that they are likely to be NAC-membrane-binding transcription factors. Almost all LiNAC genes are predicted to be localized in the nucleus (Table 1); LiNACs are probably mostly functional in the nucleus. The conserved structure and motif of LiNACs were also examined ( Figure 4B) and it was shown that all of them possessed the normal NAM structure ( Figure 4C). Nearly all LiNAC genes have motif 3, motif 4, and motif 1, although motif 2 is missing from Group II, which may be related to the loss of assembled data. The motifs found in the gene members of various evolutionary branches vary and motif 3, motif 4, motif 2, motif 1, and motif 5 may be the conserved motifs in the NAC family ( Figure S2). Phylogenetic trees of L. indica and Arabidopsis divide into 2 groups with 14 subgroups (Figure 1) and LiNAC genes were not found in many subgroups (e.g., ANAC011, ATAF, and AtNAC3/NAP). Overexpression of AtNAP causes early senescence in A. thaliana and members of the ATAF subgroup have a significant role in leaf senescence and other characteristics of plants, indicating that these subgroups may be involved in the process of plant senescence. LiNAC8 is closely connected to AT1G32770.1, AT3G61910.1, and AT2G46770, for instance, leading one to hypothesize about the function of the LiNAC gene based on genetic similarities. AT1G32770.1 (SND1) is a key gene that regulates the synthesis of cell secondary walls. In Arabidopsis, SND1 is specifically expressed in xylem fibers, and functional silencing led to secondary wall thinning and the inability of pedicel fiber cells to form secondary walls and drooping. The other subgroups of the LiNAC gene may play different biological functions.
Further, 18 of the 21 LiNAC genes were shown to be engaged in the two biological processes of GO, leaf growth and branch creation, according to KEGG and GO enrichment analyses ( Figure 5). While lignin, a phenylpropane-derived polymer, together with cellulose and hemicellulose, forms the cell wall of plant vascular tissue and provides mechanical support for plant upright growth, KEGG enrichment results revealed that LiNAC genes were involved in the phenylpropanoid biosynthesis pathway. As a result, it is hypothesized that the LiNAC genes may influence the production of secondary cell walls in plants to regulate the weeping trait.

LiNACs Are Closely Associated with the Weeping Trait of L. indica
It has been suggested that the weeping trait may be associated with abnormal GA signaling [31]. The drooping phenomenon also occurs when the gravity response changes [32]. Changes in the distribution of plant hormones often affect the formation of plant xylem and bast, which is another factor contributing to the phenomenon of weeping plants. Lignin is a phenylpropanoid-derived polymer that, together with cellulose and hemicellulose, forms the cell wall of plant vascular tissue and provides mechanical support for upright plant growth [33,34]. The NAC family of transcription factors is the master switch in the regulation of secondary cell wall thickening in plants and influences the plant phenotype by regulating the growth of secondary cell walls in plants. The cell wall thickness of upright branches in P. mume was significantly higher than that of weeping branches. In terms of overall expression, PmWND1(PmNAC082) [35] regulated downstream secondary-wallsynthesis-related genes, which were all significantly less expressed in the weeping branches than in the upright branches, indicating that NAC may play an important role in regulating the production of weeping traits in plants.
In this study, qRT-PCR analysis of upright L. indica 'Shaoguifei' and weeping L. indica 'Xiariwuniang' at three different growth stages showed that LiNACs played an important role in the weeping trait of L. indica and different LiNACs expression patterns were different during this process. The expression of LiNAC13, a member of the NAC1 subgroup, did not change significantly in the three growth stages of 'Shaoguifei', but decreased gradually in 'Xiariwuniang' and was much lower than that of 'Shaoguifei'. The expression of LiNAC13 in 'Xiariwuniang' decreased gradually and was much lower than that of 'Shaoguifei'. NAC1 is normally involved in hormone signaling in plants to regulate plant growth and development [26] and it is speculated that LiNAC13 may be involved in regulating weeping traits in plants by participating in the phytohormone signaling pathway, but this remains to be verified. The relative expression of LiNAC2 in S1-S3 is always higher in 'Xiariwuniang' than in 'Shaoguifei', especially in the S1-S2 stages, and there is not much difference between them in the S3 stage. LiNAC2 is closely related to AT3G105001/NTL4, so we hypothesized that LiNAC2 began to negatively regulate the weeping trait of L. indica at S1. This still needs to be further verified by subsequent experiments. LiNAC12 is closely related to AT5G13180.1 (VNI2), a VNI2 transcription factor that plays a molecular linkage role between plant response to environmental stress and regulation of leaf longevity [36]. In contrast, the expression of LiNAC12 in 'Xiariwuniang' was always higher than that of 'Shaoguifei' and reached a peak during S2. It is speculated that 'Xiariwuniang' may have an advantage over 'Shaoguifei' in regulating plant growth and coping with environmental stress during development, which may also be a reason for its weeping branch formation. The relative expression of LiNAC8 (which is closely related to AT1G32770.1/SND1) gradually increased in the three growth stages (S1-S3) of 'Shaoguifei' and 'Xiariwuniang' branches and the expression of 'Shaoguifei' was always higher than that of 'Xiariwuniang'. In our study, by measuring the lignin content in 'Shaoguifei' and 'Xiariwuniang' branches at different growth stages, we found that the lignin content in 'Shaoguifei' was always higher than that of 'Xiariwuniang'. Further, in the early stage of growth, there was not much difference in the content between the two, but with the development, especially at the end of the growth stage, the lignin content in 'Shaoguifei' was much higher than that of 'Xiariwuniang' (Figure 6). Inhibition of SND1 expression in Arabidopsis leads to a dramatic decrease in the expression of all three secondary wall synthesis genes and pedicel fiber cells are unable to form secondary walls and droop grow [15]. In addition, dominant inhibition of PtrWNDs in poplar caused significant thinning of secondary walls of transgenic poplar stem fibers and failure of plants to grow upright [18]. Lignin is an important component in the secondary cell wall of plants and when the lignin content increases, the fruit of Eriobotrya japonica also becomes hard [21]. Therefore, we speculate that 'Shaoguifei' has more lignin content than 'Xiariwuniang' and the branches of 'Shaoguifei' provide more mechanical support, while 'Xiariwuniang' has less lignin content and, therefore, droops. SND5 in Arabidopsis is closely connected to SND2/3, which is involved with the secondary wall and its direct homolog in poplar, according to Zhong et al. [19]. Similar findings were found in our study, LiNAC8 showed a consistent trend with lignin content at S1-S3 in different L. indica species, which indicated that LiNAC8 may be involved in affecting the synthesis of secondary walls and determining the weeping trait of L. indica. However, this conclusion has to be further investigated.

Identification and Sequence Analysis of NAC Genes from L. indica
As a reference sequence, the NAC protein sequence was acquired from the Arabidopsis

Characterization of LiNAC Proteins
The website ExPasy (https://www.expasy.org/) (accessed on 18 March 2022) predicted and examined the physicochemical characteristics of all LiNAC potential proteins, including molecular weight and theoretical points. TMHMM-2.0, a website that can be accessed online (https://services.healthtech.dtu.dk/service.php?TMHMM-2.0) (accessed on 18 March 2022), examined the transmembrane structure of LiNAC protein.

Sequence Alignment and Phylogenetic Tree Construction
Using MUSCLE, the sequences of the found 21 LiNAC proteins and 80 AtNAC protein sequences were aligned. In MEGA 6.0, the outcome was utilized to create a neighborjoining (NJ) phylogenetic tree using 1000 bootstrap repetitions. P-distance and pairwise were the predetermined parameters. Phylogenetic tree were illustrations created with iTOL (https://itol.embl

Annotation and Enrichment Analysis in GO and KEGG Databases
Online software EggNOG-Mapper (http://eggnog-mapper.embl.de/) (accessed on 20 March 2022) [38] and KEGG database (https://www.kegg.jp/) (accessed on 20 March 2022) were used to annotate the GO and KEGG functions of LiNAC gene. The results were collated using the eggNOG-mapper Helper function of TBtools v1.098661 and the text files for downstream analysis were output into GO Enrichment and KEGG Enrichment Analysis for enrichment analysis, respectively. Finally, the data were seen and examined using the online charting tool HIPLOT (https://hiplot.com.cn/) (accessed on 20 March 2022).

Determination of Lignin Content
The plant materials were collected from the greenhouse of Zhejiang Agriculture and Forestry University, which is situated at 30 • 15 2 N/119 • 43 37 E, by upright L. indica 'Shaoguifei' ( Figure 6A) and weeping L. indica 'Xiariwuniang' (Figure 6B). They were relocated into a habitat with stable temperature and favorable growing circumstances. By keeping track of processes since the time of budding, the growth condition of branches was studied.
The whole branches of 'Shaoguifei' and 'Xiariwuniang' were weighed 0.5 g at the early growth stage, growth stage, and the end of the growth stage, respectively. Further, three samples of each were collected and ground in liquid nitrogen and the ground powder was packed into 5 mL of an eluent. We then, centrifuged at 12,000 rpm for 20 min after 30 min of shaking the sample at 28 • C, then discarded the supernatant. After adding 100% methanol, the mixture was shaken for 30 min before being centrifuged at 12,000 rpm for 20 min, with the supernatant being discarded. There were four iterations of this stage. After that, they were dried in an oven set to 80 • C for an overnight period.
Accurately weigh 10 mg (can be recorded repeatedly) of the powder (washed and dried) into a 10 mL tube (the total weight of the powder after drying should also be recorded): first add 1 mL of 2 M HCL, then add 0.1 mL of thioglycolic acid, mix it upside down and evenly, then place it in a boiling water bath and heat it for 8 h. Then it was cooled on ice and centrifuged at 12,800 rpm/4 • C for 20 min and the supernatant was discarded. The precipitate was washed twice with distilled water and the precipitate was dried overnight after centrifugation. Then the precipitate was resuspended in 2 mL of 1 M NaOH, mixed evenly, and slightly shaken at 28 • C to react for 18 h. The precipitate was centrifuged at 12,800 rpm for 20 min. Next, 0.5 mL of the supernatant was put into a new glass test tube, 100 µL of concentrated hydrochloric acid was added to each tube, and the solution was placed in a refrigerator at 4 • C for 4 h (this operation was to precipitate the thioglycolate-bound lignin). The solution was centrifuged at 12,800 rpm/4 • C for 20 min to precipitate 1 mL of 1 M NaOH.
After dilution, a UV spectrophotometer was used to determine the absorbance at A280 nm. Each sample underwent three biological replicates, with NaOH solution serving as the blank control. This lignin determination method is referenced by Xu et al. [21].

Plant Materials, RNA Extraction, and qRT-PCR Analysis
Take the entire branches of 'Xiariwuniang' and 'Shaoguifei' in the early growth stage, growth stage, and the end of the growth stage to examine the expression patterns of LiNAC genes. Three samples of each were collected and the sampling time was 10:00. All samples were immediately frozen in liquid nitrogen and stored at −80 • C until needed for RNA isolation.
Total RNA was extracted from plants according to the Instructions of FastPure ® Plant Total RNA Isolation Kit (Vazyme, Nanjing, China). Reverse transcription reference HiScript ® III All-in-one RT SuperMix perfect for qPCR (Vazyme, Nanjing, China) was used. Then, quantitative real-time PCR (qRT-PCR) analysis was performed on an ABI 7300 real-time PCR instrument (Applied Biosystems, Foster City, CA, USA). The reaction system is as follows: SYBR ® Premix Ex TaqTM (TaKaRa, Dalian, China) 5 µL, cDNA 2 µL, forward and reverse primers 0.4 µL each, ddH 2 O 2.2 µL. qRT-PCR reaction procedure is: pre-denaturation at 95 • C for 30 s, denaturation at 95 • C for 5 s, denaturation at 60 • C for 30 s, 40 cycles. Three biological replicates were performed for each sample. The relative expression levels of LiNACs genes were analyzed by the 2 −∆∆Ct method [39] and the experimental data were analyzed by Excel 2010 and SPSS Statistics 20.0 software (IBM Corporation, Armonk, NY, USA). Finally, 21 pairs of gene-specific primers are shown in Table S1 and LiEF-1α [40] was used as the internal reference gene.

Conclusions
In this study, we identified a total of 21 LiNAC genes from the transcriptome data of L. indica. We analyzed the physicochemical properties of the 21 LiNAC proteins and all LiNAC genes were localized in the nucleus, and LiNAC7 and LiNAC16 also had a transmembrane structure at the C-terminus. Amino acid sequence alignment showed that almost all LiNAC genes contain a conserved NAM structure, consisting of five substructures, A, B, C, D, and E. The phylogenetic tree constructed with A. thaliana classified LiNAC proteins into 2 groups and 14 subgroups. GO and KEGG enrichment analysis also indicated that LiNAC genes are involved in plant growth development and metabolic pathways. The lignin content in the branches of upright L. indica 'Shaoguifei' and weeping L. indica 'Xiariwuniang' were also measured and it was found that the lignin content in the two branches differed at different growth and development stages. The qRT-PCR analysis showed that the expression of LiNAC13 and LiNAC8 in 'Shaoguifei' was always higher than that in 'Xiariwuniang' during S1-S3, while the relative expression of LiNAC2 in 'Shaoguifei' was lower than that in 'Xiariwuniang' in S1-S3, which indicated that LiNAC2, LiNAC13, and LiNAC8 might regulate the weeping traits of L. indica through their respective regulatory pathways. However, the specific molecular mechanism of regulation and the downstream genes that jointly regulate the weeping traits need to be further verified. In conclusion, the current study improved our understanding of the role of the NAC gene family in weeping traits in L. indica.

Supplementary Materials:
The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/plants11162168/s1, Figure S1: The transmembrane structure of LiNAC7 and LiNAC16; Figure S2: Conserved motifs in five subdomains of LiNACs; Table S1: Specific primers for RT-PCR of LiNAC in this study.

Conflicts of Interest:
The authors declare no conflict of interest.