Genome-Wide Characterization of Chrysanthemum indicum Nuclear Factor Y, Subunit C Gene Family Reveals the Roles of CiNF-YCs in Flowering Regulation

Nuclear Factor Y, Subunit C (NF-YC) transcription factors are conserved in most plants, and play essential roles in plant growth and development, especially in flowering regulation. Chrysanthemums are important commercial plants, and their market value is strongly impacted by flowering time. Until now, no details regarding the NF-YC family in the Chrysanthemum genus have been available. In this study, five NF-YC genes were cloned from Chrysanthemum indicum. Multiple alignments showed that CiNF-YCs had the highly conserved characteristic regions. Phylogenetic analyses identified a pair of paralogue NF-YC proteins in chrysanthemums. Gene structure and conserved motifs were also analyzed for functional understanding. According to the results of the expression experiments, CiNF-YC1 and CiNF-YC5 were mainly expressed in leaves or flowers, and their expression levels varied greatly from the seedling to flower bud differentiation stage. Arabidopsis overexpressing CiNF-YC1 and CiNF-YC3 showed significantly delayed flowering, accompanied by other morphological alterations. RT-qPCR analysis revealed that genes associated with photoperiod, vernalization, aging, and gibberellin pathways were downregulated in CiNF-YC1-OX lines, relative to the wild type, whereas in CiNF-YC3-OX lines, only SHORT VEGETATIVE PHASE (AtSVP), the key factor in the ambient temperature pathway, was upregulated. Taken together, these findings suggest that CiNF-YC1 and CiNF-YC3 negatively regulate flowering in Arabidopsis via different flowering pathways.


Introduction
Flowering is a major developmental transition in the life cycle of all angiosperms, and is determined by complex internal and external factors [1]. Many plants have evolved several mechanisms that control this fundamental process. Previous studies on Arabidopsis thaliana have confirmed that flowering is controlled by five major pathways: photoperiod, vernalization, gibberellin (GA), autonomous, and aging pathways [2]. These pathways interact with each other to form a complex regulatory network, and converge on a common set of floral integrators, such as FLOWERING LOCUS T (FT) and SUPPRESSOR OF OVEREXPRESSION OF CO1 (SOC1), and ultimately induce the floral meristem identify genes, including APETALA1 (AP1), LEAFY (LFY), and so on [2][3][4].
Among these flowering pathways, the photoperiod and vernalization pathways control flowering in response to seasonal changes in day length and temperature [2]. CON-STANS is a conserved central component of the photoperiod pathway. In Arabidopsis, CO different tissues and different developmental stages. In addition, we showed that CiNF-YC1 and CiNF-YC3 negatively regulated the flowering time when overexpressed in Arabidopsis. These results will provide a good basis for the further functional characteristics of CiNF-YC genes and their utilization for the genetic improvement of chrysanthemums.

Identification of the NF-YC Family Members
In order to identify NF-YC family members in wild chrysanthemums, the conserved NF-YC domains from Arabidopsis were used to screen for NF-YCs in the Chrysanthemum nankingense genome, resulting in the identification of five unique nucleotide sequences. Using the obtained CmNF-YC sequences as references, primers were designed, and five CiNF-YCs were amplified by PCR in Chrysanthemum indicum (Data S1, Figure S2). These CiNF-YCs were named according to the closest homologues of them in Arabidopsis (Figures 1b and S1). Detailed information is listed in Table 1. The gene lengths of CiNF-YCs ranged from 360 bp to 796 bp, and the CiNF-YC proteins ranged from 120 aa to 264 aa in length, whereas the corresponding protein molecular weights (MWs) ranged from 13.10 kDa to 29.44 kDa. The predicted theoretical isoelectric points (pI) ranged from 4.82 to 6.89, indicating that all NF-YC proteins were located in acidic environments, such as the nucleus. Previous studies have shown that NF-YC is a key factor in regulating flowering time, but the function of NF-YC in chrysanthemums has not been reported. In this study, we identified five NF-YC family genes in wild chrysanthemums (Chrysanthemum indicum), and investigated the characteristics, phylogenetic relationships, gene structures, and conserved motifs of these genes. We also analyzed the expression patterns of CiNF-YC genes in different tissues and different developmental stages. In addition, we showed that CiNF-YC1 and CiNF-YC3 negatively regulated the flowering time when overexpressed in Arabidopsis. These results will provide a good basis for the further functional characteristics of CiNF-YC genes and their utilization for the genetic improvement of chrysanthemums.

Identification of the NF-YC Family Members
In order to identify NF-YC family members in wild chrysanthemums, the conserved NF-YC domains from Arabidopsis were used to screen for NF-YCs in the Chrysanthemum nankingense genome, resulting in the identification of five unique nucleotide sequences. Using the obtained CmNF-YC sequences as references, primers were designed, and five CiNF-YCs were amplified by PCR in Chrysanthemum indicum (Data S1, Figure S2). These CiNF-YCs were named according to the closest homologues of them in Arabidopsis (Figures 1b and S1). Detailed information is listed in Table 1. The gene lengths of CiNF-YCs ranged from 360 bp to 796 bp, and the CiNF-YC proteins ranged from 120 aa to 264 aa in length, whereas the corresponding protein molecular weights (MWs) ranged from 13.10 kDa to 29.44 kDa. The predicted theoretical isoelectric points (pI) ranged from 4.82 to 6.89, indicating that all NF-YC proteins were located in acidic environments, such as the nucleus.

Multiple Alignment and Phylogenetic Analysis of NF-YC Proteins in Wild Chrysanthemum
Multiple alignments of the deduced amino acid sequences of CiNF-YCs were performed ( Figure 1a). All CiNF-YC proteins had structural similarities with H2A histone, and contained conserved NF-YA/B interaction domains. Most required amino acids were highly conserved in all CiNF-YCs proteins (Figure 1a). The characteristic intra-chain arginine-aspartate (R156-D163) bidentate pair, which was required for the stability of the NF-YB/C heterodimer, was completely conserved. Another specific feature, tryptophan (W148) at the end of α2, which allowed the specific interaction of NF-YC and NF-YB, was also completely conserved in all CiNF-YCs. The highly conserved peptide motifs suggested that all CiNF-YCs retain the ability to bind to NF-YB in evolution.
To reveal the phylogenetic relationship and potential function of the CiNF-YCs, an unrooted phylogenetic tree was generated with the neighbor-joining (NJ) method using the full length of NF-YC proteins identified in wild chrysanthemums and Arabidopsis (Figure 1b). CiNF-YC1 was the closest homologue to AtNF-YC1. A pair of paralogues, CiNF-YC3 and CiNF-YC4, showed high similarity in sequence, and were the homologues of AtNF-YC3 and AtNF-YC9 (Figure 1b).

Gene Structure and Conserved Motif Analyses of NF-YC in Wild Chrysanthemum
To predict the function of the CiNF-YC genes, the exon-intron structures of CiNF-YCs and the conserved motifs in the ORF (Opening Reading Frame) region of CiNF-YCs were analyzed ( Figure 2). Based on the genome sequences and corresponding coding sequences of CiNF-YC genes, we found that no CiNF-YCs contained introns. The MEME analysis discovered 10 putative motifs in total, namely motif 1 to motif 10 (Supplementary Figure  S3). Motif 1-5 were shared by all CiNF-YCs, indicating that these conserved motifs were essential for group-specific functions. Except for CiNF-YC5, the majority of CiNF-YCs featured motifs 6-10. These results suggested that CiNF-YC5 might have different functions from other CiNF-YCs.

Multiple Alignment and Phylogenetic Analysis of NF-YC Proteins in Wild Chrysanthemum
Multiple alignments of the deduced amino acid sequences of CiNF-YCs were performed ( Figure 1a). All CiNF-YC proteins had structural similarities with H2A histone, and contained conserved NF-YA/B interaction domains . Most required amino acids were highly conserved in all CiNF-YCs proteins (Figure 1a). The characteristic intra-chain arginine-aspartate (R156-D163) bidentate pair, which was required for the stability of the NF-YB/C heterodimer, was completely conserved. Another specific feature, tryptophan (W148) at the end of α2, which allowed the specific interaction of NF-YC and NF-YB, was also completely conserved in all CiNF-YCs. The highly conserved peptide motifs suggested that all CiNF-YCs retain the ability to bind to NF-YB in evolution.
To reveal the phylogenetic relationship and potential function of the CiNF-YCs, an unrooted phylogenetic tree was generated with the neighbor-joining (NJ) method using the full length of NF-YC proteins identified in wild chrysanthemums and Arabidopsis (Figure 1b). CiNF-YC1 was the closest homologue to AtNF-YC1. A pair of paralogues, CiNF-YC3 and CiNF-YC4, showed high similarity in sequence, and were the homologues of AtNF-YC3 and AtNF-YC9 (Figure 1b).

Gene Structure and Conserved Motif Analyses of NF-YC in Wild Chrysanthemum
To predict the function of the CiNF-YC genes, the exon-intron structures of CiNF-YCs and the conserved motifs in the ORF (Opening Reading Frame) region of CiNF-YCs were analyzed ( Figure 2). Based on the genome sequences and corresponding coding sequences of CiNF-YC genes, we found that no CiNF-YCs contained introns. The MEME analysis discovered 10 putative motifs in total, namely motif 1 to motif 10 (Supplementary Figure  S3). Motif 1-5 were shared by all CiNF-YCs, indicating that these conserved motifs were essential for group-specific functions. Except for CiNF-YC5, the majority of CiNF-YCs featured motifs 6-10. These results suggested that CiNF-YC5 might have different functions from other CiNF-YCs.

Expression Profiles of CiNF-YC Genes
To investigate the biological roles of CiNF-YCs in wild chrysanthemums, the expression profiles of CiNF-YC genes were evaluated in different organs (leaf, stem, root, and flower) of the plants under short day (SD) conditions. As shown in Figure 3a, the transcript abundance of CiNF-YC1 in leaves was 10 times higher compared to other organs. The expression level of CiNF-YC5 in flowers was significantly higher than that in other organs. CiNF-YC3 and CiNF-YC2 had the highest expression in leaves, followed by flowers. The expression of CiNF-YC4 was the highest in flowers, followed by leaves. a negative factor in floral transition. Both CiNF-YC3 and CiNF-YC4 started to decrease on the ninth day, and then return to the initial level detected at 3d. The similar change trend suggested that CiNF-YC3 and CiNF-YC4 might have functional redundancy. The expression of CiNF-YC5 was increased sharply on the 21st day, and then rapidly decreased to a very low level, implying that it might play a role in flower development. The expression of CiNF-YC2 remained consistent from day 9 to day 27. Taken together, the expression of CiNF-YC1 decreased continuously from the seedling stage to flowering stage, and the expression of CiNF-YC5 changed significantly before flowering, whereas CiNF-YC2, 3, and 4 changed mildly throughout the development stage. These results indicate that CiNF-YC1 might play an important role in flower induction, and CiNF-YC5 might be related to flower organ development.  RT-qPCR was performed to evaluate expression of CiNF-YCs, using Actin (GenBank accession AB770470) as the control. For CiNF-YC1, each tissue was compared to the stem. For other CiNF-YC genes, each tissue was compared to the root. (b) Transcript abundance of CiNF-YCs in leaves of differently aged wild chrysanthemums. Plants were grown under SD conditions. Samples were harvested every 6 days from 3 days after propagation. RT-qPCR was performed, and each time point was compared to 3d. Three independent experiments were performed, and error bars indicate standard deviation. * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001 indicate significant differences.
To further evaluate the functions of CiNF-YCs in floral transition, we measured the expression level of the CiNF-YCs every 6 days in the leaves of plants grown in SD until the plants flowered (day 27). The expression of CiNF-YC1 peaked in young seedlings, and subsequently declined with development ( Figure 3b), indicating that CiNF-YC1 might be a negative factor in floral transition. Both CiNF-YC3 and CiNF-YC4 started to decrease on the ninth day, and then return to the initial level detected at 3d. The similar change trend suggested that CiNF-YC3 and CiNF-YC4 might have functional redundancy. The expression of CiNF-YC5 was increased sharply on the 21st day, and then rapidly decreased to a very low level, implying that it might play a role in flower development. The expression of CiNF-YC2 remained consistent from day 9 to day 27. Taken together, the expression of CiNF-YC1 decreased continuously from the seedling stage to flowering stage, and the expression of CiNF-YC5 changed significantly before flowering, whereas CiNF-YC2, 3, and 4 changed mildly throughout the development stage. These results indicate that CiNF-YC1 might play an important role in flower induction, and CiNF-YC5 might be related to flower organ development.

Modified Flowering Time of CiNF-YC1 Overexpressing Arabidopsis
According to the results of gene expression in different tissues and different development stages, CiNF-YC1 was selected to validate its relationship with flowering. We generated transgenic Arabidopsis homozygous lines overexpressing CiNF-YC1 under the control of the CaMV35S promoter. After verification of the transcript presence of CiNF-YC1 using RT-PCR, three CiNF-YC1-OX lines (OX1-2, OX1-5, and OX1-9) were selected for further analysis (Figure 4c). Compared to WT plants, the CiNF-YC1-OX plants significantly delayed flowering under LD conditions, and the differences were recorded at 21d, 27d, and 45d, respectively (Figure 4a). At 20 days, flower bud emergence was observed in WT plants, whereas it was not observed in CiNF-YC1-OX lines 2, 5, or 9 until days 28, 30, and 29, respectively (Figure 4b). At 27 days, the number of rosette leaves in WT plants was 10 the plants at full flowering stage were sampled. RT-qPCR was performed to evaluate expression of CiNF-YCs, using Actin (GenBank accession AB770470) as the control. For CiNF-YC1, each tissue was compared to the stem. For other CiNF-YC genes, each tissue was compared to the root. (b) Transcript abundance of CiNF-YCs in leaves of differently aged wild chrysanthemums. Plants were grown under SD conditions. Samples were harvested every 6 days from 3 days after propagation. RT-qPCR was performed, and each time point was compared to 3d. Three independent experiments were performed, and error bars indicate standard deviation. * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001 indicate significant differences.

Modified Flowering Time of CiNF-YC1 Overexpressing Arabidopsis
According to the results of gene expression in different tissues and different development stages, CiNF-YC1 was selected to validate its relationship with flowering. We generated transgenic Arabidopsis homozygous lines overexpressing CiNF-YC1 under the control of the CaMV35S promoter. After verification of the transcript presence of CiNF-YC1 using RT-PCR, three CiNF-YC1-OX lines (OX1-2, OX1-5, and OX1-9) were selected for further analysis (Figure 4c). Compared to WT plants, the CiNF-YC1-OX plants significantly delayed flowering under LD conditions, and the differences were recorded at 21d, 27d, and 45d, respectively (Figure 4a). At 20 days, flower bud emergence was observed in WT plants, whereas it was not observed in CiNF-YC1-OX lines 2, 5, or 9 until days 28, 30, and 29, respectively (Figure 4b). At 27 days, the number of rosette leaves in WT plants was 10   In addition, CiNF-YC1-OX plants also displayed other morphological alterations, such as smaller leaves, decreased crown width, and shorter plant height (Figure 4b,d). The height of WT plants was 48.5 cm and the height of CiNF-YC1-OX lines was 42-44 cm. The plant height mainly depends on the number and length of internodes. Compared with WT plants, CiNF-YC1-OX Arabidopsis plants showed an obvious decrease in the internode length. Taken together, the results indicated that the negative regulation of flowering time by CiNF-YC1 did not depend on the day length, and might be related to development.

Effects of CiNF-YC3 on Flowering and Growth of Arabidopsis
It has been reported that AtNF-YC3 and AtNF-YC9 positively regulate flowering time under LD conditions. CiNF-YC3 and CiNF-YC4 are the homologue genes of AtNF-YC3 and AtNF-YC9, and CiNF-YC3 was chosen to further investigate its role in flowering. We obtained ten lines of CiNF-YC3 overexpressing Arabidopsis, and three homozygous lines (OX3-2, OX3-8, and OX3-9) were confirmed by RT-PCR and selected for further analysis (Figure 5c). CiNF-YC3-OX plants flowered 2-3 days later than WT plants under LD conditions (Figure 5b), and the delayed flowering was also observed under SD conditions (Figure 5e). In addition, CiNF-YC3-OX plants also displayed other morphological alterations, such as increased leaf size, crown width, and plant height (Figure 4b,d). The height of WT plants was 48.5 cm, and the height of CiNF-YC3-OX lines was 50.8-53.5 cm. Compared with WT, CiNF-YC3-OX plants showed an increase in the internodes number and a decrease in internode length. These results indicate that CiNF-YC3 plays a role in flowering and plant growth, but the regulatory mechanism is different from that of CiNF-YC1. In addition, CiNF-YC1-OX plants also displayed other morphological alterations, such as smaller leaves, decreased crown width, and shorter plant height (Figure 4b,d). The height of WT plants was 48.5 cm and the height of CiNF-YC1-OX lines was 42-44 cm. The plant height mainly depends on the number and length of internodes. Compared with WT plants, CiNF-YC1-OX Arabidopsis plants showed an obvious decrease in the internode length. Taken together, the results indicated that the negative regulation of flowering time by CiNF-YC1 did not depend on the day length, and might be related to development.

Effects of CiNF-YC3 on Flowering and Growth of Arabidopsis
It has been reported that AtNF-YC3 and AtNF-YC9 positively regulate flowering time under LD conditions. CiNF-YC3 and CiNF-YC4 are the homologue genes of AtNF-YC3 and AtNF-YC9, and CiNF-YC3 was chosen to further investigate its role in flowering. We obtained ten lines of CiNF-YC3 overexpressing Arabidopsis, and three homozygous lines (OX3-2, OX3-8, and OX3-9) were confirmed by RT-PCR and selected for further analysis (Figure 5c). CiNF-YC3-OX plants flowered 2-3 days later than WT plants under LD conditions (Figure 5b), and the delayed flowering was also observed under SD conditions (Figure 5e). In addition, CiNF-YC3-OX plants also displayed other morphological alterations, such as increased leaf size, crown width, and plant height (Figure 4b,d). The height of WT plants was 48.5 cm, and the height of CiNF-YC3-OX lines was 50.8-53.5 cm. Compared with WT, CiNF-YC3-OX plants showed an increase in the internodes number and a decrease in internode length. These results indicate that CiNF-YC3 plays a role in flowering and plant growth, but the regulatory mechanism is different from that of CiNF-YC1.

Modified Expression of Genes Involved in Flowering in CiNF-YC1 and CiNF-YC3 Overexpressor Lines
To investigate the regulation pathway of CiNF-YCs, RT-qPCR was performed to analyze the transcript abundance of key genes in five major flowering pathways, including AtSVP (ambient temperature pathway), AtCO (photoperiod pathway), AtSPLs (age pathway), AtGA20oxs (GA pathway), and AtFLC (vernalization pathway). Relative to WT plants, the expression of AtCO, AtSPL3, AtSPL4, AtGA20ox3, and AtGA20ox4 was clearly downregulated in the CiNF-YC1-OX lines, whereas the expression of AtFLC was dramatically upregulated. The expression of AtSPL5, AtSVP, AtGA20ox1, and AtGA20ox2 showed no obvious differences between CiNF-YC1-OX Arabidopsis, and WT plants. In CiNF-YC3-OX lines, only AtSVP, the key gene in the ambient temperature pathway, was upregulated. In addition, the expression of flowering integrators, AtFT, AtLEAFY, and AtSOC1, was significantly decreased in CiNF-YC1-OX lines. However, in CiNF-YC3-OX lines, only the expression of AtFT was significantly decreased.
These results suggest that the influence of CiNF-YC1 on flowering time may be related to multiple flowering pathways directly or indirectly, whereas CiNF-YC3 has a close relationship with the ambient temperature pathway.

Characterization of the NF-YC Gene Family in Chrysanthemum Indicum
NF-Y is a heterotrimeric transcription factor that includes three subunits: NF-YA, NF-YB, and NF-YC, and is conserved in nearly all eukaryotes. In the past two decades, many genes encoding NF-YC proteins have been identified in several species, including 20 NF-YC genes in Solanum lycopersicum [27] [45], and 3 NF-YC genes in Brassica napus [46]. In this study, five NF-YC coding genes were obtained by PCR amplification in wild chrysanthemums, which was less than that reported in most plants. This might be explained by a contraction of the NF-YC family in wild chrysanthemums, as well as by the limitation of reference genome data used in the study [47]. Notably, CiNF-YC5 was different from other members in several aspects (Table 1), including protein length, molecular weight, theoretical isoelectric point, and expression pattern, which suggested the unique functions of CiNF-YC5.
In general, proteins can be classified by their conserved characteristic regions. Multiple alignments showed that the characteristic region of the NF-YC family was highly conserved in all CiNF-YC proteins, and the required amino acids previously reported in mammals were also conserved ( Figure 1a) [32,38]. This high conservation suggested that CiNF-YCs retained functions related to the complex formation in mammals. Notably, the amino acid sequences of CiNF-YC3 and CiNF-YC4 were highly consistent, especially in the conserved region ( Figure 1a). Moreover, phylogenetic analysis also indicated that CiNF-YC3 and CiNF-YC4 were a pair of paralogues and had homologous genes in Arabidopsis (Figure 1b). Therefore, the function of CiNF-YC 3 and 4 can be predicted according to the function of their homologous genes in Arabidopsis. In addition, CiNF-YCs were absent on several branches, and only AtNF-YCs were presented.
Gene organization plays an important role in the evolution of gene families, and is also an important basis for predicting gene function. Gene structural analysis of CiNF-YC gene families showed that all CiNF-YC genes lacked introns (Figure 2). The result was consistent with the gene structure of MdNF-YC and StNF-YC families [36,48]. Intriguingly, a few NF-YC members in other reported plant species had introns, such as poplar [31], sorghum [30], common bean [39], and Foxtail Millet [49]. Previous studies have reported that introns were ubiquitous in eukaryotes and had many functions, such as alternative splicing, the reduction of gene damage, and the regulation of gene expression [48,[50][51][52]. In general, alternative splicing always led to the rapid expansion of genes. The lack of the intron of CiNF-YCs may limit the expansion of the CiNF-YC family in evolution.
Another key piece of evidence related to the gene family's function is the variety of motifs. According to the motif analysis of CiNF-YCs, motif 1-3 functioned as DNA interaction, and motif 4-7 functioned by interacting with others ( Figure 2). The motif patterns of CiNF-YCs were associated with their phylogenetic relationship. These findings suggested that CiNF-YCs belonging to the same clade may have a similar function.
It was shown that CiNF-YC2, 3, and 4 had relatively higher expression levels in flowers and leaves, suggesting that they were involved in flower or leaf development. It was possible that these three members had functional redundancy in evolution. The highest expression levels of CiNF-YC1 and CiNF-YC5, which were dozens of times higher than those in other tissues, were found in leaves or flowers, respectively. This suggested that CiNF-YC1 and CiNF-YC5 might be essential and irreplaceable to the development of leaves or flowers. Previous studies have confirmed that NF-YCs regulated flowering time mainly through the photoperiodic pathway. To investigate the roles of CiNF-YCs in flowering, their expression was analyzed from the young seedling stage (three days after propagation) to the flower bud differentiation stage (27 days after propagation). It is widely known that the leaf is the major organ for flowering signal transduction. Combined with the results of tissue-specific expression, the leaf was chosen for the expression experiments. Interestingly, CiNF-YC2, CiNF-YC3, and CiNF-YC4 maintained high expression levels before flower bud differentiation. One explanation was that they played diverse functions at various developmental stages through forming various NF-Y trimers with NF-YA and NF-YB subunits. The expression of CiNF-YC1 continued to decrease before day 27, indicating that it might play a role in flowering time regulation. CiNF-YC1 was closely placed with AtNF-YC1 and AtNF-YC4, which were known to associated with the regulation of flowering [22]. These data led us to the conclusion that CiNF-YC1 might play a negative role in flowering. CiNF-YC5 peaked on day 21 and then decreased on day 27. However, CiNF-YC5 had no close relatives in Arabidopsis, and it was difficult to predict its function. The expression data indicated that it might be related to flowering or flower organ development.

Function of CiNF-YC Genes in Flowering
The roles of several NF-YC genes in flowering have been well investigated in Arabidopsis and rice. AtNF-YC3, AtNF-YC4, AtNF-YC9, and their homologous genes in rice had conserve functions, i.e., regulating flowering time via the photoperiodic pathway [21,22]. Notably, OsNF-YC2 and OsNF-YC4 acted as inhibitors of flowering under long-day conditions, whereas OsNF-YC6 acted as the opposite [21], indicating that the NF-YC family had functional differentiation. In the present study, both CiNF-YC1-OX and CiNF-YC3-OX plants delayed flowering to varying degrees (Figures 4 and 5). These findings were contrary to the functions of AtNF-YC3, AtNF-YC4, and AtNF-YC9, but were consistent with ZmNF-YC14 [25]. Additionally, we noticed additional phenotypic alterations in CiNF-YC1-OXand CiNF-YC3-OX-overexpressing plants, including changes in leaf size, plant height, crown width, internode length, and internode number (Figure 4b,d and Figure 5b,d). Similarly, nf-yc3/4/9 displayed a decreased crown width compared to WT [22]. Increased plant height was observed in ZmNF-YC14-overexpressing Arabidopsis [25]. These data revealed that NF-YCs were involved in flowering time regulation, and also participated in plant growth and development. The mechanisms by which NF-YC regulates plant growth and development will be an interesting question in the future.

Possible Flowering Pathways That Involved by CiNF-YC1 and CiNF-YC3
It was first reported in Arabidopsis that NF-YC played a critical role in the photoperiodic pathway by interacting directly with CONSTANS (CO) [20,22]. Subsequently, the involvement of NF-YC in the photoperiodic flowering response was reported in rice [21], maize [25], poplar [23], and other plants. Recent studies have shown that the regulation of flowering by NF-Y family members may also be related to the GA and age pathway [7,25]. To explore the flowering pathways that CiNF-YC1 and CiNF-YC3 participated in, we detected the expression levels of key genes involved in five major flowering pathways.
In the CiNF-YC1-OX Arabidopsis, the expression of AtCO, AtSPL3, AtSPL4, AtGA20ox3, and AtGA20ox4 was clearly downregulated, whereas the expression of AtFLC was dramatically upregulated (Figure 6a). These results indicated that CiNF-YC1 participated in multiple flowering pathways, including the photoperiod pathway, vernalization pathway, age pathway, and GA pathway. A previous study showed that GA treatment could partially rescue the delayed flowering of ZmNF-YC14-overexpressing plants [25]. Another study showed that CmNF-YB8 negatively regulated flowering time through the aging pathway in chrysanthemums [7], but the NF-YCs with which it formed a trimer were still unclear. In previous studies, the histone methyltransferase CURLY LEAF (CLF) could mediate H3K27me3 deposition in FT in the photoperiodic pathway, and indirectly controls FT expression by epigenetically regulating FLC [24]. In Arabidopsis, NF-YC could bind to the CLF, and relieved the repression of FT transcription under LD conditions. However, whether NF-YC can epigenetically regulate FLC requires further investigation.

Identification, Alignments, and Phylogenetic Analysis of NF-YCs in Wild Chrysanthemum
The coding and protein sequences of AtNF-YCs were retrieved from the Arabidopsis Information Resource (TAIR, https://www.arabidopsis.org/) (accessed on 10 September 2020), and then the conserved sequences were utilized as queries to search the protein and nucleotide sequences of CmNF-YCs in the chrysanthemum genome database (http://www.amwayabrc.com) (accessed on 7 October 2020). SMART (http://smart.emblheidelberg.de/) (accessed on 10 October 2020) was used to check for the presence of the NF-YC domain in the obtained sequences. The repeated sequences and sequences without the NF-YC domain were removed. Using the nucleotide sequences of CmNF-YCs as the reference sequences, the primers were designed to amplify the genome and coding sequences of NF-YCs in wild Chrysanthemums (Table S1). In the CiNF-YC3-OX lines, the expression of AtSVP was clearly upregulated. Until now, there has been no relevant report on the involvement of NF-YC in the ambient temperature pathway. Interestingly, AtSPL5 was downregulated in CiNF-YC3-OX plants and upregulated in CiNF-YC1-OX plants. In Arabidopsis, SPLs were proved to modulate leaf morphology [59]. We concluded that AtSPL5 might be connected to changes in crown width and leaf size.
This evidence indicates that NF-YC members participate in flowering regulation directly or indirectly, but the regulatory mechanisms are not the same, which may be connected to the diversity of NF-Y heterotrimers.

Identification, Alignments, and Phylogenetic Analysis of NF-YCs in Wild Chrysanthemum
The coding and protein sequences of AtNF-YCs were retrieved from the Arabidopsis Information Resource (TAIR, https://www.arabidopsis.org/) (accessed on 10 September 2020), and then the conserved sequences were utilized as queries to search the protein and nucleotide sequences of CmNF-YCs in the chrysanthemum genome database (http: //www.amwayabrc.com) (accessed on 7 October 2020). SMART (http://smart.emblheidelberg.de/) (accessed on 10 October 2020) was used to check for the presence of the NF-YC domain in the obtained sequences. The repeated sequences and sequences without the NF-YC domain were removed. Using the nucleotide sequences of CmNF-YCs as the reference sequences, the primers were designed to amplify the genome and coding sequences of NF-YCs in wild Chrysanthemums (Table S1).

Plant Materials and Culture Conditions
The wild chrysanthemum (Chrysanthemum indicum) used in this study was propagated by in vitro culturing. Chrysanthemum shoots, including 1 node, were cultured on a medium comprising 1/2 Murashige and Skoog (MS) for 40 days in a culture room at 23 ± 1 • C, 40% relative humidity, and 100 µmol m −2 s −1 illumination with fluorescent lamps under a LD cycle (16 h light /8 h dark). Then, the plants were transplanted into 9 cm diameter pots containing a mixture of 1:1 (v/v) peat and pearlite, and grown in a culture room at 25/18 • C day/night, 60% relative humidity, and 250 µmol m −2 s −1 illumination with fluorescent lamps under a SD cycle (10 h light/14 h dark).
Arabidopsis thaliana ecotype Columbia (Col-0) was the genetic background of wild-type and transgenic lines used throughout the work. The plants were grown in a culture room at 25/18 • C day/night, 60% relative humidity, and 250 µmol m −2 s −1 illumination with fluorescent lamps under a LD cycle (16 h light/8 h dark) or SD cycle (10 h light/14 h dark).

RNA Isolation and RT-qPCR Analysis
For the tissue-specific transcript profiles of CiNF-YCs, samples of roots, stems, leaves, and flowers were harvested from the full flowering wild chrysanthemums. Obvious flower bud differentiation could be observed in most chrysanthemums after 27 days of SD treatment. In order to detect the expression patterns of CiNF-YCs before flowering, the second fully expanding leaves from the top were sampled at 3 d, 9 d, 15 d, 21 d, and 27 d after transplanting into the SD conditions. To evaluate the transcript profiles of floweringrelated genes in Arabidopsis, the leaves of 14-day-old seedlings of WT and transgenic plants were sampled. After sampling, all of the collected materials were immediately frozen in liquid nitrogen and stored at −80 • C. Each treatment was replicated three times.
Total RNA was isolated from samples using an RNA Aprep Pure Plant Kit (Vazyme, Nanjing, China) according to the manufacturer's instructions. The first cDNA strand was synthesized from 1 µg of total RNA using a HiScript ® III 1st Strand cDNA Synthesis Kit (+gDNA wiper) (Vazyme, Nanjing, China) according to the manufacturer's instructions. RT-qPCR reactions (20 µL volume containing 1 µL cDNA as the template) were run using the CFX96 Real-Time PCR Detection System (Bio-Rad, CA, USA) in standard mode with the ChamQ Universal SYBR qPCR Master Mix Kit (Vazyme, Nanjing, China). Gene-specific primers were designed using the Primer-BLAST tool in NCBI (https://www.ncbi.nlm.nih. gov/tools/primer-blast/) (accessed on 10 October 2021), and the chrysanthemum Actin gene (GenBank accession AB770470) was used as an internal control. Relative transcript abundances were calculated via the 2 −∆∆Ct method. Three independent experiments were conducted. The primers for RT-qPCR are listed in Supplementary Table S1.

Generation of CiNF-YC1 and CiNF-YC3 Overexpressing Arabidopsis
To construct the CiNF-YC1 and CiNF-YC3 overexpression vector, the CDSs of CiNF-YC1 and CiNF-YC3 were amplified by primers containing restriction sites and homologous sequences of the vector, pCAMBIA1300. The PCR products were fused to the binary vector, pCAMBIA1300. The PCR primers are listed in Supplementary Table S1. The overexpression plasmids were transformed into A. thaliana using the floral dip method by A. tumefaciens strain GV3101. The transformants were screened on an MS basal medium containing 25 mg· L −1 Hygromycin B (Hyg), and confirmed by RT-PCR. T3 plants displaying 100% Hyg resistance were considered homozygous and were used in this study.

Functional Validation of CiNF-YC1 and CiNF-YC3 in Arabidopsis
Homozygous T3 Arabidopsis plants overexpressing the CiNF-YC1 or CiNF-YC3 were phenotypically characterized. The WT and Transgenic Arabidopsis plants were photographed at 21, 27, and 45 d after transplanting to soil. The number of rosette leaves and the crown width of the plants were measured at day 27. The plant height, internode number, and internode length were counted and measured after the pods matured. Three lines of CiNF-YC1-OX, CiNF-YC3-OX, and WT plants were selected, with five plants per line, and each experiment was repeated three times. All presented data were from the mean of 15 individual plants. The data were presented as the mean ± standard error of the mean, and were analyzed using GraphPad Prism version 8 (GraphPad Software Inc., San Diego, CA, USA). The error bars were calculated according to Tukey's multiple range test, and with * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001 being used to indicate statistically significant effects.

Conclusions
In this study, five CiNF-YC family genes were cloned from Chrysanthemum indicum. The characteristics, phylogenetic relationships, gene structures, and conserved motifs of CiNF-YCs were analyzed. The expression patterns of CiNF-YC genes in various tissues and developmental stages were also analyzed by RT-qPCR. Additionally, the negative roles of CiNF-YC1 and CiNF-YC3 in flowering regulation were demonstrated by comparing WT and CiNF-YCs overexpressing Arabidopsis. The regulatory mechanisms involved by CiNF-YC1 and CiNF-YC3 were preliminarily investigated through evaluating the expression of key genes in different flowering pathways. CiNF-YC1 regulated flowering time through the photoperiod, GA, vernalization, and aging pathway directly or indirectly, whereas CiNF-YC3 had a close relationship with the ambient temperature pathway. These results will provide a good basis for further functional analysis of CiNF-YC genes, and support their use in chrysanthemum genetic improvement.