Metabolic analysis of the regulatory mechanism of sugars on secondary flowering in Magnolia

Background Magnolia, a traditional and important ornamental plant in urban greening, has been cultivated for about 2000 years in China for its elegant flower shape and gorgeous flower color. Most varieties of Magnolia bloom once a year in spring, whereas a few others, such as Magnolia liliiflora Desr. ‘Hongyuanbao’, also bloom for the second time in summer or early autumn. Such a twice flowering trait is desirable for its high ornamental value, while its underlying mechanism remains unclear. Methods Paraffin section was used to show the flowering time and phenotypic changes of M. liliiflora ‘Hongyuanbao’ during the twice flowering periods from March 28 to August 25, 2018. Gas chromatography-mass spectrometry (GC-MS) was then performed to explore the chemical metabolites through the twice flower bud differentiation process in ‘Hongyuanbao’, and the metabolites were screened and identified by orthogonal projection to latent structures discriminant analysis (OPLS-DA). Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis (KEGG) was used to reveal the relationship between the sugar metabolites and twice-flowering characteristic. To further investigate the potential role of sucrose and trehalose on flowering regulation of ‘Hongyuanbao’, the plants once finished the spring flowering were regularly sprayed with sucrose and trehalose solutions at 30 mM, 60 mM, and 90 mM concentrations from April 22, 2019. The flower bud differentiation processes of sprayed plants were observed and the expression patterns of the genes involved in sucrose and trehalose metabolic pathways were studied by quantitative reverse transcription PCR (qRT-PCR). Results It showed that ‘Hongyuanbao’ could complete flower bud differentiation twice in a year and flowered in both spring and summer. The metabolites of flower bud differentiation had a significant variation between the first and second flower buds. Compared to the first flower bud differentiation process, the metabolites in the sucrose and trehalose metabolic pathways were significantly up-regulated during the second flower bud differentiation process. Besides that, the expression levels of a number of trehalose-6-phosphate synthase (TPS) genes including MlTPS1, MlTPS5, MlTPS6, MlTPS7 and MlTPS9 were substantially increased in the second flower differentiation process compared with the first process. Exogenous treatments indicated that compared to the control plants (sprayed with water, CK), all three concentrations of trehalose could accelerate flowering and the effect of 60 mM concentration was the most significant. For the sucrose foliar spray, only the 60 mM concentration accelerated flowering compared with CK. It suggested that different concentration of trehalose and sucrose might have different effects. Expression analysis showed that sucrose treatment increased the transcription levels of MlTPS5 and MlTPS6, whereas trehalose treatment increased MlTPS1, showing that different MlTPS genes took part in sucrose and trehalose metabolic pathways respectively. The expression levels of a number of flowering-related genes, such as MlFT, MlLFY, and MlSPL were also increased in response to the sprays of sucrose and trehalose. Conclusions We provide a novel insight into the effect of sucrose and trehalose on the flowering process in Magnolia. Under the different sugar contents treatments, the time of flower bud differentiation of Magnolia was advanced. Induced and accelerated flowering in response to sucrose and trehalose foliar spray, coupled with elevated expression of trehalose regulatory and response genes, suggests that secondary flower bud formation is a promoted by altered endogenous sucrose and trehalose levels. Those results give a new understanding of sucrose and trehalose on twice-flowering in Magnolia and provide a preliminary speculation for inducing and accelerating the flowering process in Magnolia. Supplementary Information The online version contains supplementary material available at 10.1186/s12860-022-00458-x.


Background
Flowering is an prominent feature and essential part of reproductive process in flowering plants, which constitute the largest and most diverse group of the plant kingdom [1,2]. The flowering time is critical in the life cycle of a plant to ensure maximum reproductive success [3][4][5]. During the process of cultivation and breeding, ornamental plants bear flowers with rich colors and changeable flowering time, and longer flowering periods are generally favored [2]. Among the flowering plants, most bloom once a year and these plants are called as once flowering plants [6]. While some plants can bloom again within a year as twice flowering plants or continue to bloom under favorable conditions as continuous flowering plants [6].
In recent years, there have been a number of in-depth studies on the molecular mechanisms of plant flowering, especially in model plant Arabidopsis thaliana. It has found that six major pathways including photoperiod, vernalization, thermosensory, gibberellin, autonomous, and aging regulate flower development process [7,8]. Many genes have been found to integrate the signals received via these pathways, such as FT (FLOWER LOCUS T), AP1 (APETALA1) and LFY (LEAFY) [9]. In A. thaliana, FT serves as the florigen molecule that transmits this integrated flowering signal to the shoot apical meristem to trigger flowering [10]. AP1 and LFY integrate the signals of multiple floral induction pathways and activate floral organ development [11]. These integrative genes are highly conserved in plants, including apple [12], longan [13], and poplar [14].
In addition, sugars and metabolic pathways are known to influence flowering, and there are changes in metabolites associated with flowering process [15,16]. Sugars were found to act as signaling molecules interacting with flowering-control pathways, as exemplified by the transcription factor INDETERMINATE DOMAIN8 (AtIDD8) which controls flowering by regulating sucrose transport and metabolism in A. thaliana [17]. Additionally, trehalose is involved in many important metabolic and developmental processes regulation in flowering plants [15,[18][19][20]. Trehalose-6-phosphate (T6P) is synthesized from UDP-glucose and glucose-6-phosphate by T6P synthase (TPS) and then converted to trehalose by T6P phosphatase [21,22]. Since T6P exists in plants in trace amounts, it is thought to act as a signaling molecule for the regulation of sugars [23]. In many plants, the content of T6P is closely related to that of endogenous sucrose [15]. This significant correlation suggests that T6P may be a signal of sucrose availability and a negative feedback regulator of sucrose accumulation [18]. In studies of A. thaliana, it shows that the T6P pathway can directly regulate flowering at two sites: First, TPS1 activity is required for the induction of FT, which has a central role in flowering time control by integrating signals from the six major pathways [24]; Second, the T6P pathway affects the expression of important flowering-related genes such as SQUAMOSA PROMOTER BINDING PROTEIN LIKE (SPL) via the age pathway [15]. Taken together, sugars may be involved in the regulation of flowering through multiple flowering pathways. To our knowledge, the role of sugars in twice flowering has not been studied, which prompted our current investigation on the interactions between sugars and twice flowering in Magnolia.
Magnolia is a traditional and important ornamental plant in urban greening, which has been cultivated for about 2000 years in China for its elegant flower shape and gorgeous flower color [25,26]. Most species of Magnolia bloom once a year in spring, while a few can flower again in summer or early autumn [27]. The characteristic of twice flowering of Magnolia has greatly enhanced its ornamental and research value, and it is meaningful to understand the molecular mechanisms controlling secondary flowering in Magnolia. Magnolia liliiflora 'Hongyuanbao' is one of the twice flowering Magnolia varieties that bloom in both spring and summer [28]. In order to explore the factors affecting the secondary flowering, we investigated the biochemical metabolism during the flower buds differentiation in 'Hongyuanbao' and identified a variety of metabolic compounds that might influence the flowering process. Additionally, the expression levels of genes involved in the metabolic pathways were studied, which provided important clues to understanding the metabolic regulation of twice flowering and its underlying molecular mechanisms.

Identification of two distinct periods of flower bud differentiations in M. liliiflora 'Hongyuanbao'
To reveal the flowering time and phenotypic changes of M. liliiflora 'Hongyuanbao' , the entire flower bud differentiation process was observed by paraffin sections at regular intervals. Many Magnolia bloom once in either spring or summer, after completing the flower bud differentiation process during the previous year. Conversely, we found that 'Houngyuanbao' completed the differentiation process twice each year and flowered in both spring and summer (Fig. 1A). Based on the morphological features of floral bud differentiation in 'Hongyuanbao' , we divided the process into the following six stages (also represented by different colors) (Fig. 1B): undifferentiated, early flower bud differentiation, sepal primordium differentiation, petal primordium differentiation, stamen primordium differentiation, and pistil primordium differentiation. Plant phenotype observation revealed that the spring flowers, which had completed flower bud differentiation in the previous year, were mainly produced at the top of the last year branches. As the falling of the spring flowers, new branches grown, together with the first time progress of bud differentiation on top the branches from April 22 to May 30. Then the first time flower buds turned to be bigger and bloomed on June (summer flower), as the second flower bud differentiation would begin and finished at the end of August. The flower bud of the second time would bloom in the next spring. (Fig. 1). This phenomenon appeared that variation existed in the development and opening process of flower buds between spring and summer, which possibly be attributed to variation in plant nutritional status and environmental conditions [9,29].

Untargeted metabolomic comparison of primary metabolites across the first and second flower bud differentiation
To explore the mechanisms of the twice flowering trait, we harvested buds in April and June with early stage (undifferentiated, S1 & S4), middle stage (sepal primordium and petal primordium differentiation, S2 & S5), and late stage (stamen primordium and pistil primordium differentiation, S3 & S6) (Fig. 1). The primary metabolites of the two bud differentiation periods were analyzed using gas chromatography-mass spectrometry (GC-MS). In most metabolomic data analyses, confounding factors orthogonal to the variables of interest may obscure the intended class separation. To analyze these data, we compared two different analyses: principal components analysis (PCA) and orthogonal projection to latent structures discriminant analysis (OPLS-DA). For PCA, our classification variables was obscured, and the variables were dispersed across several principal components. Therefore, better visualization and subsequent analysis could not be carried out (Additional file 1). Thus, we tested OPLS-DA as an alternative method, which could filter out the disrelated orthogonal variables in classification variables. We were then able to analyze both the nonorthogonal and orthogonal variables (Fig. 2). The value of the Q2 is deemed to represent the prediction ability of OPLS-DA model, and for biological samples, Q2 ≥ 0.4 is ideal [30]. It showed that the OPLS-DA score plot for each comparison between the primary and secondary flower buds had overall cross-validation coefficients, and Q2(y) was 58, 78 and 77% respectively. It suggested that OPLS-DA model was more suitable than PCA for sample category prediction of 'Hongyuanbao' . The coordinates value for to1 and t1 showed a clear separation of the first predicted component between the two groups ( Fig. 2A-C), showing the inter-group and intra-group differences.
To investigate the relationship between metabolites and the twice flowering phenotype, the metabolic profiles of both the primary and the secondary flower buds were used to conduct hierarchical cluster analysis (Fig. 2). A total of 41 differential metabolites were identified (Additional file 2). Based on the Mass Bank and Kyoto Encyclopedia of Genes and Genomes (KEGG) and LECO-Fiehn Rtx5 Database analyses [31,32], all 41 metabolites were analyzed and divided into six groups: Carboxylic acids and derivatives (including 12 compounds), Organooxygen compounds (13), Fatty Acyls (5), Prenol lipids (3), Indoles and derivatives (2), and other compounds (6). Compared to first flower bud differentiation, the second flower bud showed higher levels of most sugars and organic acids, while the levels of glutamate, pyroglutamate, malate, citrate and 2-oxoglutarate were decreased. It revealed that in second flower buds, sugars including D-Talose 1 and 6-deoxy-D-glucose 1 were increased at the early stage and the levels of sucrose, trehalose, galactose 2, xylose 1 and some other sugars at the middle stage were higher. The second flower buds had higher levels of certain sugars than first flower buds at the later stage, including trehalose, erythrose 2, and d-glucoheptose 1. The data suggested that many metabolites, especially sugars, are necessary for the differentiation of the second flower bud.

Correlation among metabolites in M. liliiflora 'Hongyuanbao'
To study the correlation between the metabolites in different flower bud differentiation stages of M. liliiflora 'Hongyuanbao' , Pearson correlations were used to calculate the degree of similarity within metabolite profiles in the three different developmental stages. MetaboAnalyst 4.0 (http:// www. metab oanal yst. ca) [33] was used to map the correlation network among metabolites. S1, S2, S3 and S4, S5, S6 represented the early (undifferentiation), middle (sepal differentiation and petal differentiation), and later (stamen differentiation and pistil differentiation) stages of the first and second bud differentiation respectively. It revealed that between the first and second flower buds, 22, 41, and 25 metabolites differed across the early, middle, and later stages ( Fig. 3A-C). In order to screen key differential metabolites and metabolic pathways by analyzing the differential metabolite interactions at each flower bud differentiation stage, correlation networks were built. These networks showed that in the differential metabolites at the early stage, Methylmalonic acid, 4-hydroxycinnamic acid, 4-Hydroxybenzoic acid and Hydroxypyruvic acid interacted the most closely with other differential metabolites. Sucrose, Oxalic acid and Glutamic acid were found to interact the most closely with other differential metabolites at the middle stage. Glutaric acid, Palmitic acid, Myristic acid, Trehalose and Linolenic acid interacted the most closely with other differential metabolites at the later stage (Fig. 3). Besides that, it was found that the activity and association of metabolites were more prominent during the comparison of S2 and S5 than the comparison of S1 and S4, or S3 and S6 (Fig. 3). The results suggested that among those differential metabolites, the sugar metabolites were closely associated with other metabolites, and the sugar metabolism pathway might play a vital role during the first and second flower bud differentiation processes.

KEGG enrichment analysis of sugar metabolites and the expression of MlTPS genes
The results of KEGG enrichment analysis showed that sucrose and trehalose in the sucrose and starch metabolic pathways were significantly up-regulated in the second flower buds than the first flower buds (Additional file 3, Fig. 4A). T6P is a signal of sucrose availability and is synthesized from UDP-glucose and glucose-6-phosphate by TPS genes [18,21]. But only five TPS genes were found in transcriptome from the bud mixtures during twice flowering process, which named as MlTPS1, MlTPS5, MlTPS6, MlTPS7 and MlTPS9 respectively. The expression levels of MlTPS genes in twice flowering process were then analyzed by qRT-PCR. It showed that the expression levels of MlTPS genes had significantly differences during the twice differentiation process. Compared to the first flower bud differentiation, MlTPS1, MlTPS7 and MlTPS9 were the most prominent in expression, and MlTPS5 was barely discernible in the second flower bud differentiation process (Fig. 4B). The overall trends demonstrated that the expression of MlTPS genes were increased significantly at the middle stage in the second flower bud differentiation (Fig. 4), suggestive of potential function role of MlTPS genes in the second flower bud differentiation.

Trehalose promoted early flowering
In order to confirm the effects of sucrose and trehalose on the twice flowering trait in M. liliiflora 'Hongyuanbao' , the plants grown in the nursery of Zhejiang A&F  (Fig. 5B).

Expression of TPS genes and flowering-related genes under sucrose treatments
To more clearly understand whether trehalose and sucrose influence the twice flowering phenotype, we explored the expression of both sugar-responsive genes and genes involved in flowering regulation. Specifically, we focused on the TPS family of MlTPS1, MlTPS5, MlTPS6, MlTPS7 and MlTPS9, and key floral integrator genes MlFT, MlLFY, MlCO, and MlAP1. Previous studies have revealed that plants respectively sprayed with 60 mM of trehalose and sucrose could obviously promote flower bud differentiation process relative to the control and other treatment plants (Fig. 5B). Thus, the expression patterns of MlTPS genes and floral integrator genes in 'Hongyuanbao' leaves under treatments with 60 mM of trehalose and sucrose and CK condition were examined (Fig. 6).
It was apparent that the MlTPS genes were widely expressed throughout the flower bud differentiation period. In the treatment with a solution of 60 mM trehalose, relative to CK, MlTPS1 and MlTPS5 were increased during the middle (25d) and later stage (40d), and the expression of MlTPS6 was decreased following an initial increase. While the expression of MlTPS7 and MlTPS9 remained unchanged. The results indicated that MlTPS1 and MlTPS5 might be response to trehalose treatment to regulate flowering. Under the treatment with 60 mM sucrose, the expression level of MlTPS1 was barely detectable, and the expression of MlTPS5 decreased following an initial increase. The expressions of MlTPS6 and MlTPS7 was always higher than that of the CK, and the MlTPS9 expression remained unchanged, indicating that MlTPS6 and MlTPS7 might have an important function in sucrose treatment accelerating flowering. Moreover, the flowering integrators also responded to sugar treatment (Fig. 6). The expression of MlFT, which correlated with the timing of differentiation, was higher with trehalose and sucrose treatments than CK, suggesting that  Analysis of the expression of transcription factor SPL following sugar treatment SPL genes are regulated by diverse flowering signals and involved in age pathway to regulate flowering [34]. In other species, SPL3 has been suggested as a participant in the T6P pathway and to affect the process of flowering [15]. Thus, we wanted to know whether SPL3 homologous genes in 'Hongyuanbao' could respond to trehalose and sucrose treatments. We identified two MlSPL3 genes (MlSPL3-1 and MlSPL3-2) from the 'Hongyuanbao' transcriptome and determined whether they could respond to trehalose and sucrose treatments on 20th day. MlSPL3-1 showed significantly higher expression relative to CK under both sugar treatments, with the trehalose treatments showing relatively higher expression than the sucrose treatment (Fig. 7A). While MlSPL3-2 had significant expression under trehalose treatments, and the overall expression of MlSPL3-2 was higher following spraying with trehalose than sucrose (Fig. 7B). These results suggest that the T6P pathway can affect the expression of flower-patterning gene MlSPL3 via the age pathway to promote the floral differentiation process.

Multiple mechanisms contribute to flowering in M. liliiflora 'Hongyuanbao'
As the flowering time is critical for successful sexual reproduction, crop productivity and yield [34], numerous models have been established in A. thaliana and some crop plants. These models have also been used to study the molecular mechanisms that underlie the control of flower bud differentiation in ornamental plants. Nevertheless, the specific requirements for multiple flowering in ornamental plants have only recently become a popular research topic due to the increasing commercial interest [35,36]. Flower bud differentiation is a physiological and morphological sign of the transformation from vegetative growth to reproductive growth [36]. The whole process consists of the induction stage before flower bud differentiation and the specific process of flower differentiation [37]. 'Hongyuanbao' , a M. liliiflora variety, was introduced to China in 2001 [5]. Its ornamental features have been relatively stable for many years, especially the twice flowering phenotype. The phenomenon of twice flowering in 'Hongyuanbao' was found in one previous study [5], but the in-depth understanding on its flowering mechanism is still lacking. In this study, we made observations on the flowering time and the flower buds differentiation throughout the entire flowering period, and revealed that 'Hongyuanbao' had two distinct flower bud differentiations, the first and second differentiations in spring and summer respectively. In recent years, increasing amounts of attention have been paid to plant metabolomics, and sugars have been deemed as signaling molecules that regulate a variety of genes involved in plant developments. In order to further uncover the potential mechanism underlying the twice flower bud differentiation in 'Hongyuanbao' , the whole flower bud differentiation process was categorized into six stages. A metabolomic analysis of two different flowering buds across these six stages was performed and it suggested that the content of sugars increased significantly during the whole process in second flower bud differentiation. Sucrose and trehalose primarily occurred in the middle stage (sepal differentiation and petal differentiation) and trehalose also accumulated during the later stage (stamen differentiation and pistil differentiation). This indicated that sucrose and trehalose might play important roles during the process of flower bud differentiations, especially the secondary flower bud differentiation.

Sucrose and trehalose play important roles during twice flower bud differentiation and blossom
Sugars, as the basic molecules of carbon metabolism, can be used as energy substances or signaling molecules in the entire life cycle of plants. A number of plant systems, such as Chrysanthemum [38], A. thaliana [15,39], and apple [40,41] have been well established for studying sugar metabolism. Recent studies found that sugars could influence the transition from vegetative growth to reproductive growth in plants, by acting as a signaling molecule which could interact with other inorganic regulatory networks [42,43]. Consistent with previous research, metabolomics analysis of flower buds in 'Hongyuanbao' verifies the importance of sugars during flower bud differentiation process. In order to investigate the functional role of sugars during the twice flower bud differentiation in M. liliiflora 'Hongyuanbao' , solutions of sucrose or trehalose were sprayed on plant leaves. It revealed that both sucrose and trehalose treatments could promote flower formation, with trehalose showing a more prominent effect than sucrose (Fig. 5).
As illustrated in Fig. 8, T6P acted as a signaling molecule that relayed information about carbohydrate availability to other signaling pathways, and played an important role in the sugar metabolism pathway [15]. The T6P pathway responds to sugar signaling and regulates flowering based on two different aspects. In the plant leaves, TPS1 is activated by increasing daylength in spring, which produces T6P and induces the expression of florigen FT to regulate flowering [15,42]. Besides that, the T6P pathway affects the expression of SPL3 via the age pathway directly at the SAM independently of the daylength (Fig. 8).
Therefore, we screened and characterized five MlTPS and two MlSPL3 genes from M. liliiflora 'Hongyuanbao' . The expression of MlTPS genes in various stages were verified using qRT-PCR. It showed that MlTPS1 、MlTPS6、MlTPS7 were increased significantly under sugar treatment, indicating the possible function of MlTPS genes during flowering regulation. Previous studies in A. thaliana have also found that AtTPS1 can promote flowering, while knocking out AtTPS1 gene will result in delay in flowering [44]. However, the role of AtTPS6 and AtTPS7 in flowering regulation has not been studied further, and AtTPS9 is involved in stress resistance [45]. Photoperiod is a daily recurring pattern of light and dark periods and can regulate sugar accumulation [46]. It is attempting to assume that the longer photoperiod during the timeframe of second flower bud differentiation leads to the accumulation of polysaccharides, which promote the expression of MlTPS1 and the secondary flowering (Fig. 8). The T6P pathway may also function by the pathway of aging to promote flowering through the expression of SPL3 [15]. In this study, we found that the activation of MlSPL3 expression under treatments of sugars might induced by exogenous sugars treatments. In conclusion, our study demonstrated that M. liliiflora 'Hongyuanbao' could go through two cycles of flower bud differentiations in spring and summer, distinct from the continuous flowering in roses and strawberry. Sucrose and trehalose are important players during the secondary flower bud differentiation and bloom likely by activating expression of TPS and SPL genes.

Plant materials and growth conditions
The experiments were conducted with flower buds from three 10-year-old M. liliiflora 'Hongyuanbao' , which were grown in the Zhejiang Agricultural and Forestry University, Zhejiang, China. All mature trees were managed according to ordinary culture practices. During the entire flower bud differentiation process, the flower buds were collected from the middle and upper regions of the trees crown every 5 days at about 2 pm from April 22 to August 21, 2018 for the follow-up observation and analysis. The buds for analysis were stored at − 80 °C until extraction.

Morphological observation
To understand the flowering time and phenotypic changes of 'Hongyuanbao' , a total of 10 flower buds from the three plants were collected every 5 days and stored in FAA fixative (8:1:1 ratio of 50% ethanol: formaldehyde: glacial acetic acid) for paraffin section observation. Samples were sliced to a 10-μm thickness and subjected to safranin fast green staining for 30 s. The samples were sealed with neutral balata and then observed and photographed using an Axio Imager A2 positive fluorescence microscope (Carl Zeiss, Oberkochen, Germany).
At the same time, the development process of buds was observed using an anatomical microscope [47].

Gas chromatography-mass spectrometry analysis
A total of 50 mg of each sample was placed in a 2 mL polyethylene tube, mixed with 400 μL cold extraction liquid (VMethanol: VH2O = 3:1), and 20 μL adonitol as the internal standard. A steel ball was added to the tube to assist the grinding with a 40 Hz grinder (JXFSTPRP-24, Shanghai Jingxin Experimental Technology, Shanghai, China). Following grinding for 4 min, the samples were treated with sonication in an ice bath for 5 min. The samples were then mixed with 200 μL of chloroform and 400 μL of water and centrifuged at 13,500x g for 15 min at 4 °C. Finally, 200 μL of supernatant was collected and evaporated by vacuum drying in a glass sampling vial. The samples were analyzed using gas chromatographmass spectrometry (GC-MS) Agilent 7890 GC-TOFMS which was equipped with an Agilent DB-5 ms capillary column (30 M × 250 μm × 0.25 μm, J & W Scientific, Folsom, CA, USA).

Data processing and analysis
Chroma TOF 4.3X software of LECO Corporation and LECO-Fiehn Rtx5 database were used for raw peaks exacting, the data baselines filtering and calibration of the baseline, peak alignment, deconvolution analysis, peak identification and integration of the peak area. Both of mass spectrum match and retention index match were considered in metabolites identification. Remove peaks detected in < 50% of QC samples or RSD>30% in QC samples. SIMCA software (v.14) was used for multivariate pattern recognition and data normalization. Principal component analysis and the OPLS-DA method were used to discriminate the metabolic changes in the experimental group. The corresponding metabolic pathways were mapped in the KEGG database, and P values were calculated by Student's T test with the level of statistical significance set at p < 0.05. To shrink any possible variance and to improve the performance for downstream statistical analysis, metabolite data were checked for data integrity and normalized using MetaboAnalyst4.0's normalization protocols (selecting normalization by sum, log transformation, and auto-scaling) for statistical analysis.
The heat map was generated by OriginPro2015 (Origin-Lab Corporation, Northampton, MA, USA).

Screening and identification of metabolic differences
A total of 569 non-targeted metabolites were obtained based on the naming conditions of ion peak matching. All the differentially expressed compounds in the treated group were selected by comparing the compounds in the treated group with the control using the multivariate statistical method. For biologically duplicated metabolites, the combination of P-value and Variables with variable importance in the projection (VIP) value of the OPLS-DA (29 Thévenot E A, 2015) model was used to screen differential metabolites. The screening criteria were P < 0.05. VIP > 1.0 was considered relevant for group discrimination.

RNA extraction, Illumina sequencing, assembly and annotation
Total RNA was extracted and enriched for poly(A) mRNA using the NEBNext Poly(A) mRNA Magnetic Isolation Module. Synthesis of cDNA for sequencing followed the strand-switching protocol from Oxford Nanopore Technologies. The cDNA library was sequenced on an Illumina HiSeq 4000 sequencing platform (Illumina, San Diego, CA, USA) by BIOMARKER TECHNOLO-GIES (Beijing, China) to yield 2 × 150-bp paired-end raw reads. The sequenced raw reads were subjected to a quality check using FastQC [48]. The adapter sequences were removed from the raw reads. Reads with a ratio of ambiguous N nucleotides greater than 5% and those with low-quality sequences (quality score of less than 20) were removed. Sequencing reads were de novo assembled using Trinity software under default parameters and with a k-mer size of 25 [49]. The transcriptomes were assembled using pooled reads from all replications and stages. Assembly quality was critically assessed by BIOMARKER TECHNOLOGIES Company before subsequent analyses. The assembled transcriptome sequences were named 'unigenes' . All unigenes were queried against six commonly used databases using BLASTx search to identify homologs (E-value < 10 − 10 ). The databases used were Swiss-prot [50], Nr [4], KEGG [31], KOG [51], Pfam [52], and GO [53].

The induction of flowering by sucrose and trehalose treatments
Twenty-one plants of similar size and growth were selected to grow in the Pingshan base experimental station. Once the plants finished the spring flowering, the leaves were sprayed from top to bottom with different sucrose or trehalose solutions with concentrations of 30 mM, 60 mM, and 90 mM on April 22, 2019. The plants sprayed with water were used as the control (CK). Each treatments, including CK, had three plants. The leaves of all the plants were sprayed once every 5 days and the samples were taken for microscopy observation and genes expression analysis on the 20th and 35th day following the first spraying. The stem tips were collected for cytological observation when the shoot apical meristems began to expand, which was used as the indicator of floral bud initiation.

RNA extraction and the qRT-PCR analysis of MlTPS genes in response to sucrose treatment
Total RNA was extracted from 'Hongyuanbao' buds using the RNAprep Pure Plant Kit (TianGen, Beijing, China) and the quality was detected by nucleic acid analyzer (Implen Company in Germany). First-strand cDNA was synthesized by HiScript ® III Reverse Transcriptase (Vazyme, Nanjing, China). The expression of genes including MlTPS1, MlTPS5, MlTPS6, MlTPS7 and MlTPS9, and key floral integrator genes MlFT, MlLFY, MlCO, MlAP1, MlSPL3-1 and MlSPL3-2 with sequence similarity to the Arabidopsis genes were identified from the transcriptomic data of M. liliiflora 'Hongyuanbao' through the BlastX annotation, and detected by quantitative real-time polymerase chain reaction (qRT-PCR) on Light Cycller 480II Real Time PCR system (Roche, Basel, Switzerland). MlACTIN (MlACT ) was used as a reference gene and its stability was analyzed in Additional file 4 [27]. Reaction system was: SYBR Premix ExTaq 10 μL, cDNA 2 μL, upstream and downstream primer (10 μmol/L) 8 μL each one, ddH 2 O supplemented to 20 μL. The reaction procedure was 95 °C for 30 s, 95 °C for 5 s, 60 °C for 30 s, a total of 40 cycles: 95 °C for 5 s, 60 °C for 1 min, 95 °C for 15 s. All qRT-PCR experiments were conducted with three biological replicates. The expression levels of genes between the control and treatment groups were compared. Statistical significance was calculated by Anova analysis followed by TukeyHSD. The genes expression pattern were performed using Sigma-Plot 10.0 (Systat Software Inc., San Jose, CA, USA), and the relative transcript abundances were calculated using the 2 −∆∆Ct method [22].