Identification of Chalcone Isomerase Family Genes and Roles of CnCHI4 in Flavonoid Metabolism in Camellia nitidissima

Camellia nitidissima is a woody plant with high ornamental value, and its golden-yellow flowers are rich in a variety of bioactive substances, especially flavonoids, that are beneficial to human health. Chalcone isomerases (CHIs) are key enzymes in the flavonoid biosynthesis pathway; however, there is a scarcity of information regarding the CHI family genes of C. nitidissima. In this study, seven CHI genes of C. nitidissima were identified and divided into three subfamilies by phylogenetic analysis. The results of multiple sequence alignment revealed that, unlike CnCHI1/5/6/7, CnCHI2/3/4 are bona fide CHIs that contain all the active site and critical catalytic residues. Analysis of the expression patterns of CnCHIs and the total flavonoid content of the flowers at different developmental stages revealed that CnCHI4 might play an essential role in the flavonoid biosynthesis pathway of C. nitidissima. CnCHI4 overexpression significantly increased flavonoid production in Nicotiana tabacum and C. nitidissima. The results of the dual-luciferase reporter assay and yeast one-hybrid system revealed that CnMYB7 was the key transcription factor that governed the transcription of CnCHI4. The study provides a comprehensive understanding of the CHI family genes of C. nitidissima and performed a preliminary analysis of their functions and regulatory mechanisms.


Introduction
Flavonoids are an important class of plant secondary metabolites that contain a characteristic phenylchromenone core [1]. They are widely distributed all over the plant kingdom and play an important role in the formation of plant pigments, allelopathy [2], signaling [3], and fertility [4]. Flavonoids have vital roles in plant development and are beneficial to human health owing to various bioactivities [5], including antioxidant [6], anti-inflammatory, and anti-cancer properties [7].
Flavonoids can be subdivided into different classes according to their chemical structures, including flavanones, flavones, flavonols, proanthocyanidins, and anthocyanins [8]. The flavonoid biosynthesis pathway comprises a cascade of enzymatic reactions; chalcone synthases (CHSs) and chalcone isomerases (CHIs) catalyze the first two committed steps of the pathway and provide essential precursors for downstream flavonoid synthesis [9]. CHSs catalyze the production of 2 ,4,4 ,6 -tetrahydroxychalcone [10], which is converted to the corresponding (2S)-flavanones following catalysis by CHIs [11]. Although the latter step can occur spontaneously, CHIs can elevate the rate of the reaction by 10 7 -fold [12].
Plant CHIs are categorized into four types, namely, types I, II, III, and IV. Of these, type I and type II CHIs are catalytically active and represent bona fide CHIs, while the

Identification of CHI Family Genes in C. nitidissima
The transcriptome sequence data used in this study were previously generated and validated in our laboratory (PRJNA909942) [32]. The transcriptome data of C. nitidissima were simultaneously obtained from seven tissue types, including roots, leaves, fruits, sepals, petals, stamens, and pistils. Also, the RNA-seq data were obtained from flower samples at different stages of development, including buds that were 10 mm, 20 mm, and 30 mm in diameter, denoted as stages B1, B2, and B3, respectively, and half-open (Fh) and completely open (Fc) flowers. Total RNAs were extracted by RNA Extraction Kit DP441 (TIANGEN, Beijing, China), and Agilent 2100 (Agilent Technologies, Santa Clara, CA, USA) was used to analyze the quality of RNA. mRNA was collected using oligo-attached poly-T magnetic beads. Random hexamer primers were used for cDNA synthesis; the sequencing library was created via polymerase chain reaction (PCR) amplification. Then, cDNA libraries were sequenced on Illumina HiSeq Xten (Illumina, San Diego, CA, USA), and clean data were acquired by removing adapter reads, and low-quality reads from raw data. Additionally, mRNA was reverse transcribed into full-length cDNA using a SMARTer™ PCR cDNA Synthesis Kit (Clontech, Mountain View, CA, USA), and PacBio RS II (Pacific Biosciences, Menlo Park, CA, USA) was used for RNA sequencing. ToFu pipeline with default parameters was used to convert raw reads into error-corrected reads of insert (ROIs). By looking for the polyA tail signal and the 5 and 3 cDNA primers in ROIs, fulllength non-chemiric (FLNC) transcripts were identified. To obtain consensus isoforms, ICE (Iterative Clustering for Error Correction) was conducted, and the results were polished by Quiver. TransDecoder v3.0.0 was used to predict the ORFs (open reading frames). Bowtie2 was used to align sequencing reads from second-generation sequencing against the reads of third-generation sequencing, and transcript abundance quantification was obtained using RSEM and presented as Fragments Per Kilobase of transcript per Million mapped reads (FPKM). The formula is shown as follows: cDNA Fragments/(Mapped Fragments (Millions) × Transcript Length (kb)).
Hidden Markov models of the CHIs were identified from the Pfam database (http: //pfam-legacy.xfam.org/, accessed on 17 November 2021) (accession IDs: PF02431, PF16035, and PF16036), and the models were used as query against the transcriptome database for identifying candidates belonging to the CHI gene family with an E-value of 1 × 10 −5 , using the TBtools program [33]. The sequences of the proteins encoded by the CHI family genes of A. thaliana (AtCHI, AtCHIL, AtFAP1, AtFAP2, and AtFAP3) were then subjected to BLASTp search using a threshold E-value of < 1 × 10 −20 .
The presence of the true CHI domain in these proteins was determined using the Conserved Domain (https://www.ncbi.nlm.nih.gov/cdd, accessed on 29 November 2021) and SMART (http://smart.embl-heidelberg.de/, accessed on 29 November 2021) databases.

Phylogenetic Analysis and Multiple Sequence Alignment of CnCHI Proteins
The sequences of the CnCHI proteins of C. nitidissima were compared with the CHI protein sequences of other plants. The sequences of CHI proteins were retrieved from the NCBI (https://www.ncbi.nlm.nih.gov/, accessed on 10 January 2022). A multiple sequence alignment of the protein sequences was constructed using the MUSCLE program in MEGA11 software with default parameters, and a phylogenetic tree was constructed by the neighbor-joining method with 1000 bootstrap replications. The phylogenetic tree was visualized with iTOL (https://itol.embl.de/, accessed on 17 April 2022). The multiple sequence alignment was edited using GeneDoc, and Adobe Illustrator software was used for image processing.

Plant Materials and Sample Preparation
The flower and leaf samples of C. nitidissima used were obtained from the Germplasm Resource Center of the Institute of Subtropical Forestry, Chinese Academy of Forestry (Daqiao Road, Hangzhou, China), grown in the field and are about 10 years old. For RNA extraction and determination of total flavonoid content (TFC), flower samples at different development stages (B1, B2, B3, Fh, and Fc) were collected and immediately snap-frozen in liquid nitrogen and kept frozen at −80 • C for further use. For transient overexpression, the shoots with leaves about 20 cm were collected and cultivated with clean water.
For DNA extraction, the young leaves of C. nitidissima were obtained from the National Camellia Germplasm Resource Bank (Gecun Road, Nanning City, China). Leaf samples were collected and immediately snap-frozen in liquid nitrogen and kept on dry ice, back to the laboratory, and kept frozen at −80 • C.
Tobacco (N. tabacum) used in this experiment was grown at 25 • C with a 16-h/8-h (light/dark) cycle.

Determination of Total Flavonoid Content (TFC)
The TFC in fresh samples was quantified by UV-vis spectrophotometry. Tissues of C. nitidissima and tobacco were ground in liquid nitrogen and digested in 60% ethanol at 60 • C for 4 h, and TFC detection kit BC1330 (Solarbio, Beijing, China) was used.

Genes Isolation
RNA extraction was performed using RNA extraction reagent kits DP441 (TIANGEN, Beijing, China), and the cDNA library of C. nitidissima was built using a PrimeScript RT Master MIX (Takara, Dalian, China). According to transcriptome sequencing data, primer premier 5 was used for primer design, and the full-length coding sequences (CDS) of genes were amplified from the cDNA library and cloned into pMD19-T vectors (Takara, Dalian, China). The recombinant vector was transferred into DH5α competent cells of Escherichia coli (Takara, Dalian, China), and positive clones were confirmed by sequencing. Primers for gene cloning are listed in Supplementary Table S1.

Analysis of Subcellular Localization
The pCAMBIA1302 vector with a GFP-tag was used for analyzing the subcellular localization, and the empty vector and vector with the sequences of CnCHI4 and CnMYB7 were transformed into the GV3101 strain of Agrobacterium tumefaciens (Weidi Bio, Shanghai, China). The positive clones were selected and grown on LB liquid medium to an OD 600 of 0.6-0.8 and subsequently resuspended in activation buffer (10 mM MgCl 2 , 10 mM MES, and 100 mM acetosyringone; pH 5.8) to an OD 600 of 0.5. The solutions were injected into the abaxial side of Nicotiana benthamiana leaves with a needleless syringe. The tobaccos were cultured in darkness for 24 h (25 • C) and followed with 16 h-light/8 h-dark (25 • C) conditions for 48 h. The fluorescence was measured with a confocal microscope (Carl Zeiss, Oberkochen, Germany). Primer sequences used in vector construction are provided in Supplementary Table S2.

CnCHI4 Overexpression in N. tabacum
The full-length CDS of CnCHI4 was cloned into the pCAMBIA1302 vector, and tobacco plants were transformed with the GV3101 strain (Weidi Bio, Shanghai, China) using the leaf disc method [34]. The transformed tobacco plants were selected using 20 mg/L hygromycin, and transformation was confirmed by polymerase chain reaction (PCR) of the genomic DNA using primers specific for the hygromycin resistance gene (Supplementary Table S3). The levels of CnCHI4 transcript in the leaves of transgenic lines were measured by quantitative real-time PCR (qRT-PCR), the leaves were collected and snap-frozen in liquid nitrogen for analysis of TFC.

Transient Overexpression of CnCHI4 in C. nitidissima
The leaves of C. nitidissima were used for transient overexpression. The GV3101 strain of A. tumefaciens (Weidi Bio, Shanghai, China) carrying the pCAMBIA1302 or pCAMBIA1302-CnCHI4 vectors was injected into the left and right blades of each leaf with a needleless syringe, the specific method is the same as that of subcellular localization. Then, the shoots were kept in the dark for 24 h (25 • C) and maintained under 16 h light/8 h-dark (25 • C) conditions for 96 h. The leaves were collected after 5 days of injection and snap-frozen in liquid nitrogen for qRT-PCR and analysis of TFC.

qRT-PCR Analysis
The total RNA from the leaves of C. nitidissima and N. tabacum was extracted using RNA extraction reagent kits DP441 (TIANGEN, Beijing, China). The total RNA was reverse transcribed to cDNA using a PrimeScript RT Master Mix (Takara, Dalian, China). qRT-PCR analyses were conducted as previously described [35], and the 2 −∆∆CT method was used for quantitative analysis. The sequences of the primers used for qRT-PCR are depicted in Supplementary Table S4. 2.11. Weighted Gene Co-Expression Network Analysis (WGCNA) and Identification of Differentially Expressed Genes (DEGs) WGCNA analysis was performed with the WGCNAShiny plugin in TBtools [33], and a total of 7993 genes were selected for subsequent analysis after data filtering (sample percentage = 0.9, expression cutoff = 1, filter method = MAD, and number of reserved genes = 8000). A soft-thresholding power of 16 was used to obtain a scale-free topology index greater than 0.8. The minimum module size was set to 30, and the modules were merged at a cut height of 0.25. The FPKM values of CnCHI4 at different developmental stages were used for calculating the Pearson correlation with the modules. The modules with a high correlation coefficient were selected for subsequent analysis, and genes with a Module Membership (MM) ≥ 0.75 and Gene Significance (GS) ≥ 0.75 were identified as hub genes. DEGs between stages B2 and B3 were determined with DESeq, and the genes were considered to be differentially expressed if Fold Change ≥ 2 and FDR (False Discovery Rate) < 0.01. The transcription factors were predicted using the PlantTFDB database (http://planttfdb.gao-lab.org/index.php, accessed on 10 April 2022), and the data were visualized using the ggplot2, PerformanceAnalytics, and ggvenn packages in R studio.

Analyses of Promoter Clones and Cis-Elements
The genomic DNA was extracted using the modified CTAB method. Nested PCR was performed using primers specific for the CDS sequences of CnCHI4 and the promoter sequences of Camellia japonica CHI. The primers used for nested PCR are enlisted in Supplementary Table S5. The PCR products were ligated into a pMD19-T vector (Takara, Dalian, China), and the resulting plasmids were subsequently transformed into DH5α competent cells of E. coli (Takara, Dalian, China). The positive clones were selected and analyzed by DNA sequencing. The PlantCARE website was used for cis-elements analysis (http://bioinformatics.psb.ugent.be/webtools/plantcare/html, accessed on 7 March 2022).

Dual-Luciferase Assay
The full-length CDS of CnMYB7 was cloned into a pGreenII62-SK plasmid, and the promoter sequence of CnCHI4 was incorporated into a pGreenII0800-LUC plasmid; the primer sequences used to construct vectors are shown in Supplementary Table S6. The empty pGreenII62-SK vector was used as the control. The recombinant vectors were transferred into the GV3101 strain of A. tumefaciens (Weidi Bio, Shanghai, China) and prepared using the same method used for analyzing subcellular localization. The GV3101 cell suspensions containing the pGreenII62-SK and pGreenII0800-LUC vectors were mixed at a ratio of 1:1 and injected into the epidermis of tobacco leaves with a needleless syringe. The enzymatic activities of firefly luciferase (LUC) and renilla luciferase (REN) were determined using luciferase assay reagent kits (Beyotime, Shanghai, China) after three days of injection.

Yeast One-Hybrid Assay
The CnMYB7 gene was integrated into the pGADT7 plasmid, and the promoter sequences of CnCHI4 were incorporated into the pHIS2 plasmid; primer sequences are listed in Supplementary Table S7. The pGADT7-Gus vector was used as the control. The pHIS2 and pGADT7 vectors were co-transformed into the Saccharomyces cerevisiae Y187 strain. The cells were grown on SD/-Trp/-Leu/-His medium for 3 days at 30 • C, positive colonies were transferred to the SD/-Trp/-Leu/-His plates supplemented with 275 mg/L 3-Amino-1,2,4-triazole (3-AT).

Statistical Analyses
The statistical analyses were performed with GraphPad Prism 9 (Graphpad, San Diego, CA, USA). All the experiments were performed in triplicate, and the error bars represented the standard error (SE). Statistically significant differences were calculated using Student's t-test at a confidence level of 95.0% (* p < 0.05) and 99.0% (** p < 0.01).

Identification and Characterization of CnCHI Proteins
A total of 14 putative proteins were identified using the Hidden Markov models of CHIs (accession IDs: PF02431, PF16035, and PF16036) as a query against the transcriptome database of C. nitidissima. The putative proteins were confirmed by conserved domain search against the SMART and NCBI conserved domain databases, and the sequences that contained the complete conserved domain of CHIs were selected. The redundant sequences were finally removed, and seven proteins were classified as members of the CHI family (Table 1). The number of amino acids, pI, molecular weight, hydropathicity, and instability index were calculated using the Expasy web server, and the results are provided in Table 1. The CnCHI proteins comprised 209-452 amino acids. The molecular weight of these proteins ranged from 23,193.41 to 50,117.33 KDa, of which the molecular weight of CnCHI1 was the lowest owing to relatively fewer amino acid residues, while the molecular weight of CnCHI7 was the highest. The pI of the predicted protein sequences ranged from 4.83 to 8.73. The hydropathicity index of all the CnCHI proteins was less than 0, indicating that all the proteins were hydrophobic. The instability index of all the proteins was higher than 40, indicating that the majority of CnCHI proteins were unstable.
The signal peptides of the seven CnCHI proteins were also predicted using the SignaIP-5.0 server, and the findings revealed that the proteins lacked an N-terminal signal peptide. The secondary structure composition of the CnCHI proteins was predicted with SOPMA ( Figure 1A), and the results demonstrated that the proteins secondary structures contained alpha helices, beta turns, random coils, and extended strands.
In order to obtain a better insight into the conservation and diversification of CnCHI proteins, the deduced sequences were submitted to the MEME web server for determining the motif composition. A total of ten motifs, denoted as motifs 1-10, were predicted for each protein, and they are depicted in Figure 1B. The findings revealed that the number of motifs ranged from 2 to 7, and the motifs comprised 15-50 amino acid residues. SOPMA ( Figure 1A), and the results demonstrated that the proteins secondary structures contained alpha helices, beta turns, random coils, and extended strands.
In order to obtain a better insight into the conservation and diversification of CnCHI proteins, the deduced sequences were submitted to the MEME web server for determining the motif composition. A total of ten motifs, denoted as motifs 1-10, were predicted for each protein, and they are depicted in Figure 1B. The findings revealed that the number of motifs ranged from 2 to 7, and the motifs comprised 15-50 amino acid residues.

Phylogenetic Relationships, Multiple Sequence Alignment, and Classification of CnCHIs
As aforementioned, the CHI proteins of plants can be categorized into four types. A phylogenetic tree was constructed using the neighbor-joining method for evaluating the phylogenetic relationships among the CHI proteins of different species and classifying the CnCHI proteins. As depicted in Figure 2, the proteins were clustered into four clades in the phylogenetic tree, which was consistent with the results of previous studies. The seven CnCHI proteins identified in this study were classified into three types, namely, type I, type III, and type IV proteins. Of these, CnCHI2, CnCHI3, and CnCHI4 were categorized

Phylogenetic Relationships, Multiple Sequence Alignment, and Classification of CnCHIs
As aforementioned, the CHI proteins of plants can be categorized into four types. A phylogenetic tree was constructed using the neighbor-joining method for evaluating the phylogenetic relationships among the CHI proteins of different species and classifying the CnCHI proteins. As depicted in Figure 2, the proteins were clustered into four clades in the phylogenetic tree, which was consistent with the results of previous studies. The seven CnCHI proteins identified in this study were classified into three types, namely, type I, type III, and type IV proteins. Of these, CnCHI2, CnCHI3, and CnCHI4 were categorized as type I, while CnCHI5, CnCHI6, and CnCHI7 were categorized as type III, and only CnCHI1 was classified as a type IV protein.
The results of multiple sequence alignment ( Figure 3) revealed that type I CnCHIs contained all the active site and critical catalytic residues, and the residues were wellconserved in type I and type II CHIs. These findings were consistent with previous studies, which reported that Ser/Ile in type I proteins and Thr/Met in type II proteins determine the differences in substrate preference between these classes [36]. The findings of this study revealed that three homologous type III proteins (CnCHI5, CnCHI6, and CnCHI7) and a type IV protein (CnCHI1) carried the substitutions at nearly all the critical catalytic sites. contained all the active site and critical catalytic residues, and the residues were wellconserved in type I and type II CHIs. These findings were consistent with previous studies, which reported that Ser/Ile in type I proteins and Thr/Met in type II proteins determine the differences in substrate preference between these classes [36]. The findings of this study revealed that three homologous type III proteins (CnCHI5, CnCHI6, and CnCHI7) and a type IV protein (CnCHI1) carried the substitutions at nearly all the critical catalytic sites.   The red and blue asterisks represent the active site and critical catalytic residues, respectively. The residues that affect substrate preference are indicated by the red box.

Analysis of CnCHI Expression and Measurement of TFC
Analysis of gene expression levels can provide important information for elucidating gene functions. In order to decipher the function of CnCHI genes during flower development, the expression levels of seven CnCHI genes at different stages of flower development (B1, B2, B3, Fh, and Fc) are displayed in a heatmap. As depicted in Figure 4A, the expression levels of type I CnCHI genes varied markedly across the different stages of flower development. The expression levels of CnCHI2 and CnCHI3 were very low and negligible compared to that of CnCHI4, which indicated that CnCHI2/3 do not play an important role during the different stages of flower development. The expression level of CnCHI4 rapidly peaked at the B2 stage. Interestingly, the expression levels of CnCHI4 reduced markedly during the transition from B2 to B3 stages. The red and blue asterisks represent the active site and critical catalytic residues, respectively. The residues that affect substrate preference are indicated by the red box.

Analysis of CnCHI Expression and Measurement of TFC
Analysis of gene expression levels can provide important information for elucidating gene functions. In order to decipher the function of CnCHI genes during flower development, the expression levels of seven CnCHI genes at different stages of flower development (B1, B2, B3, Fh, and Fc) are displayed in a heatmap. As depicted in Figure 4A, the expression levels of type I CnCHI genes varied markedly across the different stages of flower development. The expression levels of CnCHI2 and CnCHI3 were very low and negligible compared to that of CnCHI4, which indicated that CnCHI2/3 do not play an important role during the different stages of flower development. The expression level of CnCHI4 rapidly peaked at the B2 stage. Interestingly, the expression levels of CnCHI4 reduced markedly during the transition from B2 to B3 stages.
The expression of type III CnCHI genes altered during the different stages of flower development. The expression of CnCHI5 altered in a bimodal manner over time in that the expression of CnCHI5 was upregulated during the transition from B1 to B2, reduced sharply during the transition from B2 to B3, upregulated from B3 to Fh and declined to nearly 0 during the transition from Fh to Fc. The expression level of CnCHI6 increased gradually from B1 to Fh, and declined smoothly at the end of flower development. Interestingly, the changes in the expression levels of CnCHI7 were the complete reverse of those of CnCHI6.Analysis of the expression of type IV CnCHIs revealed that the expression of CnCHI1 decreased substantially from B1 to B3 and was maintained at relatively low levels in the subsequent stages.
Analysis of the expression of CnCHI genes in the different tissues, including leaves, roots, fruits, petals, sepals, stamens, and pistils, revealed that the patterns of gene expression were different among the different tissues ( Figure 4B). It was evident that the expression of CnCHI3 was very low in all the tissues, while the expression of CnCHI4 was generally high across all tissue types. The expression of CnCHI4 was higher in the leaves, followed by fruits and roots, and was lowest in the sepals and stamens. The expression of CnCHI2 was higher in the roots compared to that in the other tissues. Analysis of the expression levels of type III CnCHIs revealed a high similarity between the expression of CnCHI5 and CnCHI6, which were relatively high in the sepals, stamens, and petals, and low in the roots and fruits. The type IV CHI gene, CnCHI1, was abundantly and variably expressed across the different tissues. The expression levels of CnCHI1 were highest in the fruits and lowest in the stamens. CnCHI1 expression in the fruits was nearly 23-fold that in the stamens. The expression levels of genes were highly associated with gene function, and CnCHI1 and CnCHI4 possibly play a major role in the flavonoid biosynthesis pathway of C. nitidissima. The expression of type III CnCHI genes altered during the different stages of flower development. The expression of CnCHI5 altered in a bimodal manner over time in that the expression of CnCHI5 was upregulated during the transition from B1 to B2, reduced sharply during the transition from B2 to B3, upregulated from B3 to Fh and declined to nearly 0 during the transition from Fh to Fc.. The expression level of CnCHI6 increased Flavonoids can form complexes with aluminum ions in alkaline nitrite solution, which form characteristic absorption peaks at 470 nm. The TFC in C. nitidissima petal was therefore measured using a colorimetric method ( Figure 4C,D). The TFC levels increased from B1 to B3, peaked at B3, and declined from B3 to Fc. The TFC was highest at B3, reaching a value of 27.02 mg/g. TFC meet a rapid growth during the transition from B2 to B3, which could correspond to a crucial period of flavonoid biosynthesis in the petals of C. nitidissima. Moreover, the TFC at the different stages of flower development appeared to reflect the expression of CnCHI4, with some lag.

CnCHI4 Is Localized in the Cytoplasm and Nucleus
Analysis of the subcellular localization of proteins provides vital information for understanding protein function. The subcellular localization of CnCHI4 was therefore investigated in this study ( Figure 5). The findings demonstrated that CnCHI4 is located in the cytoplasm and nucleus. expressed across the different tissues. The expression levels of CnCHI1 were highest in the fruits and lowest in the stamens. CnCHI1 expression in the fruits was nearly 23-fold that in the stamens. The expression levels of genes were highly associated with gene function, and CnCHI1 and CnCHI4 possibly play a major role in the flavonoid biosynthesis pathway of C. nitidissima.
Flavonoids can form complexes with aluminum ions in alkaline nitrite solution, which form characteristic absorption peaks at 470 nm. The TFC in C. nitidissima petal was therefore measured using a colorimetric method ( Figure 4C, 4D). The TFC levels increased from B1 to B3, peaked at B3, and declined from B3 to Fc. The TFC was highest at B3, reaching a value of 27.02 mg/g. TFC meet a rapid growth during the transition from B2 to B3, which could correspond to a crucial period of flavonoid biosynthesis in the petals of C. nitidissima. Moreover, the TFC at the different stages of flower development appeared to reflect the expression of CnCHI4, with some lag.

CnCHI4 is Localized in the Cytoplasm and Nucleus
Analysis of the subcellular localization of proteins provides vital information for understanding protein function. The subcellular localization of CnCHI4 was therefore investigated in this study ( Figure 5). The findings demonstrated that CnCHI4 is located in the cytoplasm and nucleus.

Overexpression of CnCHI4 in C. nitidissma and N. tabacum
In order to further elucidate the biological functions of CnCHI4, CnCHI4 was transiently overexpressed in the leaves of C. nitidissma. The relative expression levels of CnCHI4 in the leaves following infiltration were analyzed by qRT-PCR. As depicted in Figure 6A, the relative expression levels of CnCHI4 increased seven-fold compared to that of the respective control groups. Correspondingly, the TFC in fresh leaves was also increased ( Figure 6B).

Overexpression of CnCHI4 in C. nitidissma and N. tabacum
In order to further elucidate the biological functions of CnCHI4, CnCHI4 was transiently overexpressed in the leaves of C. nitidissma. The relative expression levels of CnCHI4 in the leaves following infiltration were analyzed by qRT-PCR. As depicted in Figure 6A, the relative expression levels of CnCHI4 increased seven-fold compared to that of the respective control groups. Correspondingly, the TFC in fresh leaves was also increased (Figure 6B).
CnCHI4 was also overexpressed in tobacco plants to verify its role in flavonoid metabolism ( Figure 6C). The findings revealed that the TFC of the three independent transgenic lines overexpressing CnCHI4 was significantly higher than that of the wild-type plants ( Figure 6D).

Identification of Key Regulatory Factors of CnCHI4
As depicted in Figure 4B, the expression levels of CnCHI4 declined immediately during the transition from B2 to B3 stage, which indicated that CnCHI4 is possibly more tightly regulated in this phase. A total of 199 transcription factors were found differentially expressed between B2 and B3 ( Figure 7A) and could play an essential role in the CnCHI4 was also overexpressed in tobacco plants to verify its role in flavonoid metabolism ( Figure 6C). The findings revealed that the TFC of the three independent transgenic lines overexpressing CnCHI4 was significantly higher than that of the wild-type plants ( Figure 6D).

Identification of Key Regulatory Factors of CnCHI4
As depicted in Figure 4B, the expression levels of CnCHI4 declined immediately during the transition from B2 to B3 stage, which indicated that CnCHI4 is possibly more tightly regulated in this phase. A total of 199 transcription factors were found differentially expressed between B2 and B3 ( Figure 7A) and could play an essential role in the regulation of CnCHI4 expression. factors at different stages of flower development were calculated. As depicted in Figure  7F, all the transcription factors were significantly associated, either positively or negativelywith the expression of CnCHI4, with correlation coefficients ranging from 0.79 to 0.93. Notably, an MYB transcription was identified, and the PlantTFDB database suggested that this MYB transcription factor shared a high similarity to AtMYB7.

CnMYB7 was Localized in the Nucleus
In this study, CnMYB7 was identified as a key factor in the transcriptional regulation of CnCHI4. The subcellular localization of CnMYB7 was analyzed to obtain insights into the molecular function of CnMYB7. The results demonstrated that the GFP fluorescence of the empty vector was distributed throughout the cells of tobacco leaves, while the The potential interactions of the genes were determined by WGCNA, and 7993 genes were selected for hierarchical clustering following data filtering. A total of 25 gene modules were identified and indicated by different colors based on the similarity in gene expression patterns( Figure 7B). The modules were subsequently correlated to the trait phenotype (expression level of CnCHI4 at different stages of flower development). Calculation of the correlation and significance between the characteristics of the modules and CnCHI4 expression revealed that the brown module was highly correlated to CnCHI4 expression (r = 0.94, p = 2.1 × 10 −7 ) ( Figure 7C). The genes in the brown module were therefore subjected to subsequent analyses in this study. The Gene Significance (GS) was determined for measuring the correlation between trait phenotype and the genes in the brown module, and the importance of the genes within the module was evaluated based on the Module Membership (MM) ( Figure 7D).
By setting the thresholds for GS and MM to 0.75, a total of 26 pivotal transcription factors were identified in the brown module for subsequent analysis. The common transcription factors determined by differential expression analysis and WGCNA were identified by plotting the results with a Venn diagram ( Figure 7E). The Venn diagram revealed that five transcription factors identified by differential expression analysis were encoded by the genes in the brown module and were subjected to further analyses. The Pearson correlation coefficients between the expression levels of CnCHI4 and the five transcription factors at different stages of flower development were calculated. As depicted in Figure 7F, all the transcription factors were significantly associated, either positively or negativelywith the expression of CnCHI4, with correlation coefficients ranging from 0.79 to 0.93. Notably, an MYB transcription was identified, and the PlantTFDB database suggested that this MYB transcription factor shared a high similarity to AtMYB7.

CnMYB7 was Localized in the Nucleus
In this study, CnMYB7 was identified as a key factor in the transcriptional regulation of CnCHI4. The subcellular localization of CnMYB7 was analyzed to obtain insights into the molecular function of CnMYB7. The results demonstrated that the GFP fluorescence of the empty vector was distributed throughout the cells of tobacco leaves, while the fluorescence due to CnMYB7-GFP was only observed in the nucleus ( Figure 8A). The findings demonstrated that CnMYB7 is localized in the nucleus and is likely to play a role in regulating gene transcription.

CnMYB7 Bound and Inhibited the CnCHI4 Promoter
The promoter sequence of CnCHI4 was cloned to elucidate whether CnCHI4 is transcriptionally regulated by CnMYB7, and the findings revealed the presence of MYB binding sites in the promoter sequence of CnCHI4 (Supplementary Table S8). A dual-luciferase reporter assay was subsequently performed in which CnMYB7 was incorporated into a pGreenII62-SK vector, and the promoter sequence of CnCHI4 was integrated into a pGreenII0800-LUC vector. The findings demonstrated that CnMYB7 significantly repressed the expression of the CnCHI4 promoter by approximately three-fold compared to that of the control ( Figure 8B).
The mechanism underlying the effect of CnMYB7 on the transcription of CnCHI4 was determined by yeast one-hybrid assays. The findings revealed that CnMYB7 could directly bind to the promoter of CnCHI4, indicating that CnMYB7 suppresses CnCHI4 expression by binding to the promoter region of CnCHI4 ( Figure 8C).
Biomolecules 2023, 13, x 14 of 18 fluorescence due to CnMYB7-GFP was only observed in the nucleus ( Figure 8A). The findings demonstrated that CnMYB7 is localized in the nucleus and is likely to play a role in regulating gene transcription.

CnMYB7 Bound and Inhibited the CnCHI4 Promoter
The promoter sequence of CnCHI4 was cloned to elucidate whether CnCHI4 is transcriptionally regulated by CnMYB7, and the findings revealed the presence of MYB binding sites in the promoter sequence of CnCHI4 (Supplementary Table S8). A dual-luciferase reporter assay was subsequently performed in which CnMYB7 was incorporated into a pGreenII62-SK vector, and the promoter sequence of CnCHI4 was integrated into a pGreenII0800-LUC vector. The findings demonstrated that CnMYB7 significantly repressed the expression of the CnCHI4 promoter by approximately three-fold compared to that of the control (Fig. 8B).
The mechanism underlying the effect of CnMYB7 on the transcription of CnCHI4 was determined by yeast one-hybrid assays. The findings revealed that CnMYB7 could directly bind to the promoter of CnCHI4, indicating that CnMYB7 suppresses CnCHI4 expression by binding to the promoter region of CnCHI4 (Fig. 8C).

Discussion
Flavonoids are the main active chemical constituents of C. nitidissima and are beneficial to human health. However, flavonoid metabolism in C. nitidissima remains poorly understood to date. CHI is a key enzyme in the flavonoid biosynthesis pathway. In order to obtain further insights into the mechanism of flavonoid biosynthesis in C. nitidissima flowers, the CHI family genes of C. nitidissima were identified. Previous transcriptomic studies

Discussion
Flavonoids are the main active chemical constituents of C. nitidissima and are beneficial to human health. However, flavonoid metabolism in C. nitidissima remains poorly understood to date. CHI is a key enzyme in the flavonoid biosynthesis pathway. In order to obtain further insights into the mechanism of flavonoid biosynthesis in C. nitidissima flowers, the CHI family genes of C. nitidissima were identified. Previous transcriptomic studies did not reveal major specific roles for CHI genes in the flavonoid metabolism of C. nitidissima [30,31], although recently, it was shown that these genes are upregulated in the petals during the early stages of flower development [30]. A total of seven genes belonging to the CHI gene family were identified in this study. CHI family genes have been intensively studied in several plants, including A. thaliana and G. max. The CHI genes of A. thaliana can be categorized into four types, of which AtCHI is a type I CHI gene, AtFAP1/2/3 are type III CHI genes, and AtCHIL is a type IV gene [18]. The 12 CHI genes of G. max are categorized into four types, ranging from types I to IV [37]. Only two types of CHI genes, types III and IV, have been identified in Physcomitrella patens [1]. Analysis of the phylogenetic tree (Figure 2) revealed that the CnCHIs were divided into three types, namely, type I (CnCHI2/3/4), type III (CnCHI5/6/7), and type IV (CnCHI1) genes. Type II CHI genes are relatively common in leguminous plants [38], which is absent in C. nitidissima. Interestingly, the CHI genes of C. nitidissima exhibited high similarity to those of Antirrhinum majus and always shared the same branch in the phylogenetic tree.
Previous studies have demonstrated that type I and type II CHIs are bona fide enzymes that contain several of the critical catalytic residues [39]. The results of multiple sequence alignment demonstrated that CnCHI2/3/4 possess all the critical catalytic residues, including Ser and Ile at positions 190 and 191, respectively (Figure 3). The finding illustrates that CnCHI2/3/4 are type I CHIs that are indispensable to the flavonoid biosynthetic pathway. Nearly all the critical catalytic sites of type III (CnCHI5/6/7) and type IV (CnCHI1) CHIs were substituted, which suggested the loss of catalytic activity. Previous studies have demonstrated that type III CHIs of A. thaliana are FAPs that are associated with fatty acid metabolism [18]. CnCHI5 clustered with AtFAP1, and CnCHI6/7 shared the same clade as AtFAP2, which suggested that CnCHI5/6/7 are likely to function as FAPs [1,18]. Type IV CHIs are also known as CHIL proteins, and recent studies have demonstrated that CHIL proteins can bind to CHSs and function as rectifiers that affect flavonoid metabolism [10,19,40]. The type IV CHI identified in this study (CnCHI1) could have a similar function in C. nitidissima; however, further studies are necessary to verify the conjecture.
It has been reported that the expression patterns of CHI genes differ across plant tissues owing to distinct functions in the different tissues. The CHI genes of Scutellaria lateriflora are mostly expressed in the roots [41], which is similar to the expression patterns observed in leguminous plants and could be attributed to the interactions between signal molecules and rhizobia [11,13,42]. The expression levels of the CHI genes of Fagopyrum dibotrys are high in petals and low in stems [43]. In this study, the CnCHI1 gene of C. nitidissima was mainly expressed in the roots, fruits, and leaves, while the expression of CnCHI4 was relatively low in the sepals and stamens ( Figure 4B). Previous studies on other plants have demonstrated that CHIs partake in the formation of pericarp color and influence the bulb color in onion [44] and seed coat color in A. thaliana. [13]. The high content of flavonoids in the leaves of C. nitidissima could be correlated to the high expression levels of CnCHI1/4 [27].
Notably, the transcription of CnCHI4 (type I CnCHI) altered drastically during the transition from B1 to B3, and the expression of CnCHI4 was extremely high during the transition from B1 to B2 but decreased sharply from B2 to B3 ( Figure 4A). The same has been reported in A. majus [45], citrus plants [46] and in a recent study involved in C. nitidissima [30], in which the CHI genes are expressed slightly earlier during floral development. In this study, the TFC increased acutely from B2 to B3 and began decreasing after B3 during flower development. The findings revealed that the changes in the expression of TFC lag behind the changes in the expression of CnCHI4 and could be attributed to the various processes between gene expression and metabolite synthesis, including protein translation and transport [37,47]. Additionally, CHIs act as early biosynthetic genes in the flavonoid biosynthesis pathway and provide precursors for the biosynthesis of downstream flavonoids. These findings explain the expression pattern of CnCHI4 observed in this study. The overexpression of CnCHI4 increased the TFC in C. nitidissima and tobacco. The findings indicated that CnCHI4 plays a prominent role in flavonoid biosynthesis in C. nitidissima.
Transcription factors, especially MYB transcription factors, are known to play essential roles in regulating the flavonoid biosynthesis pathway. For instance, flavonol biosynthesis in A. thaliana is regulated by MYB11, MYB12, and MYB111 [48,49]. The MYB90, MYB113, and MYB114 transcription factors are related to the regulation of anthocyanin biosynthesis [50,51]. MYB6, an R2R3 MYB transcription factor, promotes anthocyanin and proanthocyanidin biosynthesis in Populus tomentosa [52]. Previous studies have reported that MYB10 [53] and MYB308L [22] induce anthocyanin accumulation in apple. CHI genes are also regulated by several MYB transcription factors. The ectopic expression of EsAN2 in tobacco upregulates the expression of CHS, CHI, and ANS and significantly enhances anthocyanin biosynthesis [54]. The MYB11, MYB12, and MYB111 transcription factors in A. thaliana play a vital role in promoting the expression of CHI genes [55][56][57]. However, there is a scarcity of information regarding the mechanism of regulation of CHI expression in C. nitidissima. In this study, WGCNA and differential expression analysis revealed that several transcription factors act as pivotal regulators of CnCHI4 expression. Using dualluciferase reporter and yeast one-hybrid assays, the study further verified that the CnMYB7 transcription factor is a major regulator of CnCHI4 expression and negatively regulates the expression of CnCHI4.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

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