Integrative Metabolomic and Transcriptomic Analysis Elucidates That the Mechanism of Phytohormones Regulates Floral Bud Development in Alfalfa

Floral bud growth influences seed yield and quality; however, the molecular mechanism underlying the development of floral buds in alfalfa (Medicago sativa) is still unclear. Here, we comprehensively analyzed the transcriptome and targeted metabolome across the early, mid, and late bud developmental stages (D1, D2, and D3) in alfalfa. The metabolomic results revealed that gibberellin (GA), auxin (IAA), cytokinin (CK), and jasmonic acid (JA) might play an essential role in the developmental stages of floral bud in alfalfa. Moreover, we identified some key genes associated with GA, IAA, CK, and JA biosynthesis, including CPS, KS, GA20ox, GA3ox, GA2ox, YUCCA6, amid, ALDH, IPT, CYP735A, LOX, AOC, OPR, MFP2, and JMT. Additionally, many candidate genes were detected in the GA, IAA, CK, and JA signaling pathways, including GID1, DELLA, TF, AUX1, AUX/IAA, ARF, GH3, SAUR, AHP, B-ARR, A-ARR, JAR1, JAZ, and MYC2. Furthermore, some TFs related to flower growth were screened in three groups, such as AP2/ERF-ERF, MYB, MADS-M-type, bHLH, NAC, WRKY, HSF, and LFY. The findings of this study revealed the potential mechanism of floral bud differentiation and development in alfalfa and established a theoretical foundation for improving the seed yield of alfalfa.


Introduction
The development of floral buds encompasses the initiation of plant reproductive growth, and they directly affect the yield and quality of seeds [1,2].Researchers usually classify the stages of flower bud development by observing the morphological changes during flower bud growth [3,4].Additionally, studying the dynamic changes in the physiological indexes and gene expression in each stage of flower bud growth can provide a theoretical basis for uncovering the potential mechanism of flower organ formation, improving the seed yield [5,6].
Alfalfa (Medicago sativa) is a perennial leguminous plant cultivated worldwide [7].It is the most important forage in stockbreeding, and it has good quality, high yield, and strong resistance to stress [8,9].However, low seed yields have limited the long-term development of the alfalfa industry.Alfalfa inflorescences are typical racemes with small flowers growing on the peduncle.Previous studies have revealed that the floret count determines the number of seeds, and the number of seeds per inflorescence is positively correlated with the seed yield [10,11].Moreover, the floral bud growth affects the quantity and quality of florets and seeds in plants [12].However, a few investigations have been performed on the inherent mechanisms of floral bud development in alfalfa.
The external environment and internal factors commonly influence the development of floral buds [13][14][15].Endogenous phytohormones play an important role in regulating Plants 2024, 13, 1078 2 of 17 flower bud growth [16,17].Various endogenous hormones, such as auxins (IAA), cytokinin (CK), abscisic acid (ABA), jasmonate (Ja), gibberellin (GA), and ethylene (ETH), coordinate with each other, thereby regulating flower bud development by jointly controlling the metabolism of plants.Yan et al. (2019) found that a higher IAA content was beneficial to the differentiation of licorice flower buds, eventually improving the seed yield of licorice [18].Before floral bud growth, increasing the levels of ABA and GA3 in stem apical meristems may promote flower induction and floral bud growth [19].Fang et al. (2018) explored the effect of growth regulators on cotton floral buds and concluded that the ratio of trans-Zeatin-riboside (ZR)/IAA and GA3/IAA was significantly elevated in the seeds after pretreatment with GA3 and N6-benzyladenine (6-BA), resulting in an increase in the number of floral buds [20].During the process of grape floral bud differentiation, GAs inhibited the primordium differentiation of grape inflorescences, and CTK could promote grape flowering [21,22].Understanding the dynamic change in endogenous phytohormones in the developmental stages of floral buds can establish a theoretical foundation for conducting cultivation management and improving the seed yield of alfalfa.
In recent years, with the development of high-throughput sequencing, omics technology has become an important method to explore the potential mechanism of floral bud growth.For example, Qu et al. demonstrated that GA 3 played a crucial role in regulating the floral bud development of Cyclocarya paliurus based on transcriptome analysis [23].Moreover, Xie et al. analyzed the gene expression and identified candidate genes associated with female and male floral bud development in Carya illinoensis by using RNA sequencing [24].In this work, we investigated the gene expression and phytohormone accumulation in three developmental stages of floral bud development in alfalfa, which provided a theoretical basis for molecular breeding.

Transcriptome Analysis
By observing the morphological changes in floral bud development of alfalfa, floral bud growth is divided into three developmental stages (D1, D2, and D3) (Figure 1A,B).At the early bud stage (D1), the floral buds are detectable as small swellings, surrounded by the leaf primordium.At the mid bud stage (D2), the floral bud is larger and easier to detect.Small swellings differentiate into multiple florets along with floral buds enlarging, and many white trichomes grow on the floret primordium.At the late bud stage (D3), each floret primordium becomes mature along with the bud expanding and lengthening rapidly.After the late bud stage, the florets will gradually bloom.These three stages are the most typical stages in the developmental process of floral buds.
Floral buds at the three stages (D1, D2, and D3) were utilized in transcriptome sequencing to elucidate the potential mechanism of floral bud development in alfalfa.As a result, a total of 59.21 Gb clean data were identified, and the clean data of each sample reached 5 Gb.Moreover, the Q30 base percentage was above 94%, and more than 80% of the clean reads were mapped to the Medicago sativa reference genome (Table S1).Principal component analysis (PCA) suggested that the three samples were obviously separated (Figure 2A).These results indicated that the transcriptome datasets were reliable and accurate for further investigation.Furthermore, a total of 1736 up-regulated and 407 down-regulated genes, 3131 up-regulated and 1604 down-regulated genes, and 1465 up-regulated and 1373 downregulated genes were identified in the D2 vs. D1, D3 vs. D1, and D3 vs. D2 comparisons, respectively (Figure 2B).In addition, nine genes were selected to identify the reliability of the transcriptome datasets.The results showed that the FPKM value of most genes had a similar tendency to the relative expression levels in the three groups, revealing that the transcriptome data in this work can be trusted for further investigation (Figure S1).

GO and KEGG Enrichment Analysis
Subsequently, we conducted a GO and KEGG enrichment analysis of differentially expressed genes (DEGs) in the D2 vs. D1 and D3 vs. D2 comparisons.In the D2 vs. D1 comparison, the GO enrichment results revealed that most DEGs were mainly enriched in biological processes, cellular components, and molecular functions, including the anatomical structure formation involved in morphogenesis (GO:0048646), secondary metabolite biosynthetic process (GO:0044550), plant-type cell wall organization or biogenesis (GO:0071669), integral component of plasma membrane (GO:0005887), glucosyltransferase activity (GO:0046527), inorganic anion transmembrane transporter activity (GO:0015103), and some flower-development-related terms (Figure S2).In the D3 vs. D2 comparison, most DEGs were mainly enriched in biological processes and molecular functions, including anatomical structure formation involved in morphogenesis (GO:0048646), secondary metabolite biosynthetic process (GO:0044550), cell wall biogenesis

GO and KEGG Enrichment Analysis
Subsequently, we conducted a GO and KEGG enrichment analysis of differentially expressed genes (DEGs) in the D2 vs. D1 and D3 vs. D2 comparisons.In the D2 vs. D1 comparison, the GO enrichment results revealed that most DEGs were mainly enriched in biological processes, cellular components, and molecular functions, including the anatomical structure formation involved in morphogenesis (GO:0048646), secondary metabolite biosynthetic process (GO:0044550), plant-type cell wall organization or biogenesis (GO:0071669), integral component of plasma membrane (GO:0005887), glucosyltransferase activity (GO:0046527), inorganic anion transmembrane transporter activity (GO:0015103), and some flower-development-related terms (Figure S2).In the D3 vs. D2 comparison, most DEGs were mainly enriched in biological processes and molecular functions, including anatomical structure formation involved in morphogenesis (GO:0048646), secondary metabolite biosynthetic process (GO:0044550), cell wall biogenesis
In addition, KEGG enrichment showed that many genes were enriched in metabolic pathways, biosynthesis of secondary metabolites, phenylpropanoid biosynthesis, protein processing in the endoplasmic reticulum, and flavonoid biosynthesis in the D2 vs. D1 comparison (Figure 3A).In the D3 vs. D2 comparison, most DEGs were enriched in metabolic pathways, biosynthesis of secondary metabolites, plant hormone signal transduction, and phenylpropanoid biosynthesis (Figure 3B).These results suggested that these pathways might play a critical role in floral bud development.(GO:0042546), glucosyltransferase activity (GO:0046527), UDP-glucosyltransferase activity (GO:0035251), dioxygenase activity (GO:0045543), and some flower-development-related terms (Figure S2).
In addition, KEGG enrichment showed that many genes were enriched in metabolic pathways, biosynthesis of secondary metabolites, phenylpropanoid biosynthesis, protein processing in the endoplasmic reticulum, and flavonoid biosynthesis in the D2 vs. D1 comparison (Figure 3A).In the D3 vs. D2 comparison, most DEGs were enriched in metabolic pathways, biosynthesis of secondary metabolites, plant hormone signal transduction, and phenylpropanoid biosynthesis (Figure 3B).These results suggested that these pathways might play a critical role in floral bud development.
To detect the content of phytohormones in floral buds at the three stages (D1, D2, and D3), a targeted phytohormone metabolome analysis was conducted.The PCA results showed that the three groups had a good separation, suggesting that the experimental results were reliable for further study (Figure 4A).Subsequently, a total of 17, 31, and 25
To detect the content of phytohormones in floral buds at the three stages (D1, D2, and D3), a targeted phytohormone metabolome analysis was conducted.The PCA results showed that the three groups had a good separation, suggesting that the experimental results were reliable for further study (Figure 4A).Subsequently, a total of 17, 31, and 25 differentially accumulated metabolites (DAMs) were identified in the D2 vs. D1, D3 vs. D1, and D3 vs. D2 comparisons, respectively (Figure 4B).Moreover, seven categories of plant hormones were differentially accumulated in the three groups, including cytokinin (CK), auxin (IAA), jasmonic acid (JA), salicylic acid (SA), gibberellin (GA), abscisic acid (ABA), and ethylene (ETH) (Figure 4C).These results suggested that these phytohormones might participate in the floral bud development.

Quantitative Analysis of Differentially Accumulated Phytohormones
Furthermore, we performed a quantitative analysis of DAMs in the D2 vs. D1 and D3 vs. D2 comparisons to investigate the dynamic changes in seven phytohormones in the developmental process of floral bud (Tables 1 and 2).For GA, GA8 (Log 2 fold-change value = Inf; Log 2 fold-change value = 1.07) was up-accumulated in the D2 vs. D1 and D3 vs. D2 comparisons, and the content of GA7 (Log 2 fold-change value = 1.13) was higher in the D3 group compared to the D2 group (Tables 1 and 2).   1 and 2).   1 and 2).  1 and 2).Based on the results above, we concluded that the content of one GA (GA8), six CKs (pT9G, iP9G, tZ, DHZ7G, tZR, and tZRMP), and one JA (MEJA) continuously increased with development, and some auxins were differentially accumulated in the D3 vs. D2 comparison.Therefore, we speculated that GA8, pT9G, iP9G, tZ, DHZ7G, tZR, and tZRMP, and MEJA might be involved in floral bud growth, and auxins played a vital role in the mid and late bud stages.

Key Genes Involved in GA, IAA, CK, and JA Biosynthesis Pathways
To screen the candidate genes involved in crucial phytohormone synthesis, the GA, IAA, CK, and JA biosynthesis pathways were examined, respectively (Figure 5).The FPKM value of DEGs in each group is shown in Table S2.In GA biosynthesis, we identified that one CPS (MsG0780036356.01),three GA20ox

Key Genes Involved in GA, IAA, CK, and JA Signaling Pathways
Simultaneously, some DEGs were identified in the GA, IAA, CK, and JA signaling pathways in the D2 vs. D1 and D3 vs. D2 comparisons (Figure 6).The FPKM value of DEGs in each group is shown in Table S3

Discussion
The development of floral buds determines the seed yield.To explore the gene expression patterns and phytohormone accumulation of the floral bud differentiation and development in alfalfa, we performed a comparative transcriptome and metabolome analysis for the three developmental stages of floral buds.
Plant hormones are the key regulatory factors during plant morphogenesis [16,17].Spraying growth regulators on a plant is a conventional cultivation method to increase the crop yield during the flower bud developmental stage; therefore, exploring the accumulation of endogenous phytohormones in floral buds can provide a theoretical support for guiding the production and improving the seed yield.Cytokinin plays a key role in reproductive growth [25].Gibberellin can affect flower bud differentiation by regulating the expression of downstream genes [26].Auxin is a crucial plant hormone involved in the development of flower organs [27].It is involved in regulating the gibberellin signaling

Discussion
The development of floral buds determines the seed yield.To explore the gene expression patterns and phytohormone accumulation of the floral bud differentiation and development in alfalfa, we performed a comparative transcriptome and metabolome analysis for the three developmental stages of floral buds.
Plant hormones are the key regulatory factors during plant morphogenesis [16,17].Spraying growth regulators on a plant is a conventional cultivation method to increase the crop yield during the flower bud developmental stage; therefore, exploring the accumulation of endogenous phytohormones in floral buds can provide a theoretical support for guiding the production and improving the seed yield.Cytokinin plays a key role in reproductive growth [25].Gibberellin can affect flower bud differentiation by regulating the expression of downstream genes [26].Auxin is a crucial plant hormone involved in the development of flower organs [27].It is involved in regulating the gibberellin signaling pathway and eventually promoting flower formation [28].Jasmonic acid can participate in floral organ development via promoting SlMYB21 expression in tomatoes [29].In agreement with our findings, we found that the content of one GA (GA8), six CKs (pT9G, iP9G, tZ, DHZ7G, tZR, and tZRMP), and one JA (MEJA) continuously increased in the floral bud developmental process, and auxins were significantly differentially expressed in the D2 vs. D1 and D3 vs. D2 comparisons.These results indicated that these phytohormones might play a vital role in regulating the downstream genes related to flower organ development.Notably, previous studies have shown that auxin polar transport is closely related to plant morphological construction [30,31]; therefore, identifying the accumulation in specific cells of auxin will be important for exploring the mechanism of auxins involved in floral bud development in the future.
Moreover, some candidate genes related to GA, IAA, CK, and JA biosynthesis were identified based on the integrated transcriptome and metabolome results.CPS and KS are upstream enzymes of GA biosynthesis, catalyzing geranylgeranyl diphosphate to ent-Kaurene.The inhibition of CPS and KS enzymes influenced the GA biosynthesis and limited plant organ growth [32,33].A previous study reported that the high expression levels of the SoGA20ox1 gene in shoot tips enhanced GA biosynthesis [34].Additionally, overexpression of the GA2ox gene facilitated pollen growth in transgenic Arabidopsis plants [35].In this work, we detected that the expression of one CPS, two KS, four GA20ox1, and five GA2ox genes maintained a high level in the D2 and D3 groups, suggesting that these genes might play an essential role in GA8 accumulation.Moreover, Liu et al. found FvYUCCCA6 played a critical role in vegetative and reproductive development in woodland strawberry [36].Consistent with our results, one YUCCA6 gene maintained high expression levels in the D3 group compared to the D1 and D2 groups, and it might be involved in the IAA accumulation in the D3 group.IPT was the first enzyme participating in CK biosynthesis, catalyzing ATP, ADT, and AMP to iPRTP, iPRDP, and iPRMP, respectively [37,38].Then, CYP735 converted iP-nucleotide to tZ-nucleotide [39].In the present work, the expression levels of one IPT and two CYP735 genes were high in the D2 and D3 groups, suggesting that these genes may be closely related to CK accumulation in the floral bud developmental process.In addition, we identified that two LOX, one AOC, four OPR, one MFP2, and three JMT genes were highly expressed in the D2 and D3 groups.Previous studies have demonstrated that these genes have vital functions in JA biosynthesis [40,41].An enhancement or inhibition of gene expression in these genes directly affects JA synthesis, eventually changing the plant physiological activities [42][43][44][45][46]. Therefore, these DEGs related to JA biosynthesis might play an essential role in regulating the floral bud growth.
The phytohormone signaling pathway was associated with plant organ development [47].In accordance with our study, many DEGs were enriched in the plant hormone signal transduction pathway.The findings of a previous study revealed that GID1 and TF positively promote flower bud differentiation [48].Consistent with our results, several GID1 and TF genes in the D2 or D3 group were highly expressed in the GA signaling pathway.Additionally, most studies demonstrated that the DELLA protein inhibits plant development and growth by binding to transcription factors [49][50][51].In this work, the expression level of most DELLA genes was low in the D3 group, suggesting that the DELLA protein played a crucial role in participating in floral bud differentiation.In the auxin signaling pathway, the AUX1, TIR, AUX/IAA, ARF, GH3, and SAUR genes played a pivotal role in regulating plant growth [52].In the present work, most differentially expressed AUX1, AUX/IAA, ARF, and SAUR genes in the auxin signaling pathway maintained high expression levels in the D2 and D3 groups compared to the D1 group, revealing that the auxin signaling pathway may participate in regulating floral bud development.A-ARR was a key gene involved in the CK signaling pathway and affected flower organ development and growth [53,54].In this work, the expression level of four A-ARR genes was significantly higher in the D3 group than in the D1 and D2 groups, suggesting A-ARR genes might play a crucial role in regulating floral bud development in the D3 group.In the JA signaling pathway, JAR1 can induce the JA converted to the biologically active JA-Ile by responding to environmental stress [40], and the JAZ protein family usually regulates JA responses by interacting with the MYC family [55].As we are currently aware, one JAR1, one JAZ, and eight MYC genes were differentially expressed in the three developmental stages, and these genes might be positively involved in floral bud development.
In this work, many AP2/ERF-ERF, MYB, MADS-M-type, bHLH, NAC, and WRKY genes were identified in the three developmental stages, which was consistent with previous studies [56,57].AP2/ERF-ERF transcription factor family proteins are mainly involved in the development of sepals and petals in reproductive organs [58].The overexpression of RcAP2 results in the transformation of stamens into petals, thereby increasing the number of petals in Arabidopsis, while silencing RcAP2 decreases number of petals [59].MYB transcription factors play a role in the flowering time, pollen development, flower color, and sex differentiation of flower organs [60].During the developmental stage of Arabidopsis reproductive organs, AtMYB125 and AMYB98 are involved in the development of male and female gametes, respectively [61,62].Most studies indicated that the MADS family is an important factor regulating flowering time and flower organ development in plants [63][64][65].bHLH proteins CIB1, CIB2, CIB4, and CIB5 commonly regulate flowering initiation, which facilitates FT transcription by binding to the FT promoter in plants [66].Additionally, AtWRKY75 may be a new member regulating flowering in the GA signaling pathway, and WRKY71 induces early flowering by activating FT and LFY in Arabidopsis [67,68].The NAC transcription factor family can regulate flower growth by participating in the JA and GA signaling pathways [69,70].Simultaneously, we also discovered that the expression level of some TFs, such as AP2/ERF-ERF, MYB, MADS-M-type, bHLH, NAC, WRKY, HSF, and LFY, is continuously increased or decreased in the three developmental stages, suggesting that these TFs might play a pivotal role in regulating floral bud development.

Plant Materials
Alfalfa seedlings were planted at the Institute of Grassland Research, Chinese Academy of Agricultural Sciences, Hohhot (40 • 58 N, 111 • 78 E).The samples of the three developmental stages were collected in the early flowering period.We observed the monological traits of floral buds and identified the developmental stages of floral buds by using a dissecting microscope (Dian Ying, Shanghai, China).The floral buds of the three developmental stages were named D1, D2, and D3.Three biological replicates were obtained for the three samples, and the mass of each biological replicate was more than 3 g.All samples were stored in liquid nitrogen.

Transcriptome Sequencing and Data Analysis
By utilizing ethanol precipitation and CTAB-PBIOZOL, the total RNA of floral buds (D1, D2, and D3) was obtained.Total RNA was analyzed by utilizing a Qubit fluorescence quantifier and a Qsep400 high-throughput biofragment analyzer (AUTO Q BIO-SCIENCES, San Diego, CA, USA).Subsequently, all cDNA libraries were sequenced on the Illumina platform.After reads with adapters were removed by using fastp software (fastp v0.19.4), all non-redundant transcripts were mapped with Medicago sativa reference genome (https://figshare.com/articles/dataset/Medicago_sativa_genome_and_annotation_files/12623960 (accessed on 1 November 2023)).Novel genes were screened by using StringTie.FPKM (Fragments Per Kilobase Million) values were calculated in accordance with the gene length.Differentially expressed genes (DEGs) were identified between comparisons by using DESeq2.p-values, and log 2 fold changes were set as criteria for obvious differential expression.In accordance with the hypergeometric test, with pathway-based hypergeometric distribution checking for Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) term-based profiles for GO, enrichment analysis was conducted.We thank Wuhan Metware Biotechnology Co., Ltd.(Wuhan, China) for assistance with sequencing.

qRT-PCR Analysis
We extracted the total RNA by using the RNA pure plant kit (Tb Green ® Premix Ex Taq™ II (TAKARA, Beijing, China)).Then, we obtained the first strand of the reversetranscribed cDNA by using the specifications of the Monad first-strand cDNA Synthesis Kit.The primers were designed by utilizing PRIMER-BLAST (Table S5).The ABI7500 quantitative PCR instrument was adopted to perform real-time fluorescence quantitative PCR.The gene (MsG0180001288.01)was selected as an actin gene for high and stable levels of expression in nine samples according to the FPKM value, and each group was identified from three repetitions.

Conclusions
In this work, we elucidated the molecular mechanism of floral bud development in alfalfa based on the phenotypic, metabolome, and transcriptome in three stages (D1, D2, D3).The transcriptome results revealed that the phytohormone biosynthesis and signaling pathway were closely associated with floral bud growth.The metabolomic results indicated that GA, IAA, CK, and JA were the critical phytohormones involved in floral bud differentiation and development.Notably, many key genes participating in phytohormone biosynthesis and signaling pathways might play a crucial role in regulating floral bud growth.Finally, we uncovered that many TF family members were closely correlated with floral bud development, such as AP2/ERF-ERF, MYB, MADS-M-type, bHLH, NAC, and WRKY.This work established regulatory networks related to phytohormones regulating floral bud development, providing potential leads for the molecular breeding of alfalfa.

Supplementary Materials:
The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/plants13081078/s1,Table S1: Data filtering and comparison of reference tables; Table S2: FPKM of candidate genes related to phytohormones biosynthesis in each group; Table S3: FPKM of candidate genes related to phytohormones signaling pathway in each group; Table S4: FPKM of TFs with continuously increased or decreased expression levels in each group; Figure S1: primers used in qRT-PCR in this work (Table S5); verification of nine DEGs by qRT-PCR; Figure S2: GO enrichment analysis in the D2 vs. D1 and D3 vs. D2 comparison.
Author Contributions: X.H. is the first author.L.L. and X.Q. are the corresponding authors.X.H. conceived and designed the experiments.X.H. performed the experiments.X.H. analyzed the data and wrote the original manuscript.L.L., X.Q., Y.M., Z.L. and F.H. revised and approved the final version of the paper.All authors have read and agreed to the published version of the manuscript.Funding: This work was supported by the projects from the Supplementary list and its assessment of the rare and endangered plant species in Inner Mongolia (Grant 2021MS03074); the secure conservation of forage germplasm resources in north of China (Grant 19230874); Study on Special Characteristics in Various Flower Color in Alfalfa (Medicago L.) (Grant 31402122).

Plants 2024 , 19 Figure 1 .
Figure 1.(A) The phenotypes of the three developmental stages (D1, D2, and D3) of floral buds in alfalfa.(B) Width and height of D1, D2, and D3; X and Y represent values of width and height of floral bud, respectively.

Figure 1 . 19 Figure 1 .
Figure 1.(A) The phenotypes of the three developmental stages (D1, D2, and D3) of floral buds in alfalfa.(B) Width and height of D1, D2, and D3; X and Y represent values of width and height of floral bud, respectively.

19 Figure 4 .
Figure 4. (A) PCA plot analysis of metabolome; the x-axis and y-axis represent principal component 1 (PC1) and principal component 2 (PC2), respectively.(B) Venn map of the three comparisons.(C) Differentially accumulated phytohormones in the three groups.

Figure 4 .
Figure 4. (A) PCA plot analysis of metabolome; the x-axis and y-axis represent principal component 1 (PC1) and principal component 2 (PC2), respectively.(B) Venn map of the three comparisons.(C) Differentially accumulated phytohormones in the three groups.

Figure 7 .
Figure 7. (A) Transcription factors in the three groups.(B) TFs with continuously increased and decreased expression levels in the developmental process of floral buds.

Figure 7 .
Figure 7. (A) Transcription factors in the three groups.(B) TFs with continuously increased and decreased expression levels in the developmental process of floral buds.