Abstract Background Carrot is a root vegetable abundant in numerous nutritional values. Sugar is one of the most important carbohydrates in horticultural products that play important roles in plant growth and development and response to biotic and abiotic stresses. However, the dynamics of the metabolites including sugar during carrot root development still remain unclear. Here, the differential metabolites in carrot roots at different developmental stages were measured using an UPLC-ESI-MS/MS system. The accumulation profiles of metabolites, especially sugars, as well as the transcript patterns of Sugars Will Eventually be Exported Transporter (SWEET) genes were intensively examined. Results The results identified 727 metabolites over all the samples detected, of which, 539 metabolites were found to be differential accumulated. A total of 34 differentially accumulated sugar metabolites were identified over the period of root development. Furthermore, 17 DcSWEET genes were detected to be specifically expressed in the roots, indicating a potential for root enlargement and sugar accumulation in carrot root. Conclusions The results from the current study would help carrot breeding focused on yield and quality improvement. Supplementary Information The online version contains supplementary material available at 10.1186/s12870-025-06497-8. Keywords: Daucus carota, Sugar accumulation, Root development, Metabolites, SWEET, Transcript profiles Background Sugar accumulation, transport, and distribution are critical events during plant growth and development [[36]1]. It can determine organ formation, affect accumulation of secondary metabolites, and adjust its transport and partitioning to respond to external environmental signaling [[37]2–[38]4]. In terms of horticultural crops, sugar accumulation and composition are major targets that determine their market values [[39]5]. Therefore, understanding how plants regulate sugar transport and storage during developmental processes and how this process affects quality formation is meaningful. It is reported that sucrose is the main compound translocated from leaves to various plant tissues, such as roots, fruits, and seeds, by sucrose transporters [[40]6]. Afterwards, disaccharide or monosaccharide transporters will work to help the uptake of sugars into cells [[41]7]. To date, the mechanism of sugar transport into cells has been widely investigated, and many sugar transporters have been explored in plants. The rice monosaccharide transporter, OsTMT1, located at the tonoplast, was capable of transporting glucose into vacuoles [[42]8]. A hexose transporter, MdHT1.2, from apple was demonstrated to absorb glucose from the rhizosphere and control carbohydrate allocation [[43]9]. In Arabidopsis, a total of 53 monosaccharide transporters, belonging to the major facilitator superfamily, have also been revealed [[44]10]. In addition to monosaccharide transporters reported, numerous disaccharide transporter genes have also been reported. StTST1, a tonoplast sugar transporter highly expressed in potato tubers during cold storage, was demonstrated to take part in transportation of sucrose from the cytosol to the vacuole [[45]11]. SlSUT4, a sucrose transporter in tomato, could interact with SlSUT1 and affect its presentation on the plasma membrane, resulting in altered sucrose export and flower development [[46]12]. The SWEET (Sugars Will Eventually be Exported Transporter) family, a kind of sugar transporter recently identified, can be involved in the transport of sugars across intracellular or plasma membranes [[47]13]. SWEETs can be divided into four subfamilies according to their phylogenetic relationships, and members from different groups present distinction in subcellular localization and recognition substrates [[48]14]. Most SWEETs identified consist of seven transmembrane helices (7-TMHs) with two conserved MtN3/saliva domains [[49]15]. The SWEET proteins are energy-independent uniporters, indicating that their capability for cellular uptake and efflux of sugars are not attributed to proton gradient or pH. SWEET-mediated sugar translocation and accumulation is demonstrated to participate in various physiological processes, ultimately contributing to crop phenotype, output, and quality. In Camellia oleifera, CoSWEET10 promoted seed development and conferred drought resistance by modulating soluble sugar production and allocation [[50]16]. HvSWEET11b from barley (Hordeum vulgare) had the ability to transport not only sucrose and glucose, but also cytokinin to affect grain size as well as starch and protein generation [[51]17]. Similarly, poplar PtaSWEET1c, a glucose and sucrose transporter, was essentially required for ectomycorrhizal symbiosis [[52]18]. CsSWEET2, an energy-independent hexose/H^+ uniporter in cucumber, enhanced plant cold tolerance by regulating glucose and fructose metabolism [[53]19]. In maize, two SWEET family members ZmSWEET15a and ZmSWEET15b were confirmed as sucrose transporters and could contribute to seed germination and salt stress resistance [[54]20]. These findings suggested that the SWEET proteins had different transport substrates and might be involved in diverse processes. Carrot (Daucus carota L.), belonging to the Apiaceae family, is rich in pigments and dietary fibers [[55]21, [56]22]. It can be eaten as vegetables as well as salads. Carrot also possesses many medicinal properties and has the effect of preventing and controlling major diseases [[57]23]. Owing to high nutritional values, its fleshy root is popular and consumed worldwide. Sugar accumulation within carrot taproot is one of the most important indicators reflecting carrot quality. However, the sugar types and their changes during fleshy root development, and their underlying regulatory mechanisms are still unclear. In the present study, carrot roots at different developmental stages were harvested and analyzed. Their dynamics on metabolic substances were determined using an UPLC-ESI-MS/MS system. Differential accumulated metabolites, especially sugars, were unearthed and extensively investigated over a prolonged growth process. Furthermore, SWEET genes and their expression patterns during carrot root growth were determined by genome-wide and transcriptome analysis to reveal the roles of SWEETs in carrot sugar transport and accumulation. Although the specific roles of the SWEETs in carrot still remained unearthed, the results from the current work could pave way for understanding the SWEET-mediated sugar transport and accumulation in carrot. Results Metabolomics analysis during Carrot taproot development In order to investigate the metabolic dynamics during taproot development in carrot, roots were harvested at three time points after sowing for 25 (stage 1), 40 (stage 2) and 90 (stage 3) d, respectively, and metabolic profiling was determined by ultra-performance liquid chromatography (UPLC) coupled with tandem mass spectrometry (MS/MS) (Fig. [58]1). A total of 727 metabolites were identified across all samples detected. The Principal Component Analysis (PCA) analysis was introduced to assess the degree of variation among or within group samples (Fig. [59]2). As shown, the two principal components (PC) made up 53.8% (PC1) and 28.48% (PC2) of the total variance, respectively. The 9 samples analyzed were obviously classified into 3 different subgroups by PC1 and PC2 (Fig. [60]2). Fig. 1. [61]Fig. 1 [62]Open in a new tab Carrot taproots utilized for metabolomics analysis. Carrot roots were harvested at 25 (stage 1), 40 (stage 2) and 90 (stage 3) days after sowing. The scale bars represented 5 cm Fig. 2. [63]Fig. 2 [64]Open in a new tab PCA plot of the root samples at three stages. Stage 1, 2, and 3 indicated carrot taproot samples collected at 25, 40, and 90 days after sowing, respectively Differential metabolite analysis To characterize the differential metabolites during the process of carrot taproot development, comparison of the variation of metabolites among the three samples (stage1, stage2, and stage3) was carried out. There were 112, 116, and 112 up-regulated and 203, 291, and 255 down-regulated metabolites detected in “stage1 vs stage2”, “stage1 vs stage3”, and “stage2 vs stage3” comparisons, respectively (Fig. [65]3). Interestingly, the number of up-regulated metabolites was consistently less than that of down-regulated metabolites during carrot root growth. Totally, 539 differential metabolites were identified, which was approximately 74.14% of all metabolites determined. To further understand the dynamics of metabolites in response to growth periods, the venn diagram of 539 differential metabolites was established, revealing that 122 metabolites were common among the three comparisons, i.e. stage1 vs. stage2, stage1 vs. stage3, and stage2 vs. stage3 (Fig. [66]4). Fig. 3. Fig. 3 [67]Open in a new tab Comparison of differential metabolites between samples. Blue and orange represented up- and down-regulated metabolites, respectively Fig. 4. Fig. 4 [68]Open in a new tab Venn diagram of metabolites among three comparison groups Metabolic processes involved in Carrot root development According to the results of KEGG pathway enrichment analysis, the differential metabolites with pathway annotation were primarily related to metabolic pathways, ABC transporters, biosynthesis of amino acids, 2-oxocarboxylic acid metabolism, purine metabolism, pyrimidine metabolism, aminoacyl-tRNA biosynthesis, and galactose metabolism (Fig. [69]5). A heat map consisting of all differential metabolites and their metabolic information was constructed (Fig. [70]6). It was noted that most lipids were highly enriched at the first stage, whereas phenolic acids intended to accumulate at a high level at the second stage. Sugar compounds were classified into the saccharides and alcohols subgroup in the group “others”. All 539 identified differential metabolites were further classified into 6 clusters on the basis of alterations during taproot development (Fig. [71]7). Of them, 177 (32.84%) metabolites showed a continuous decrease during root enlargement, whereas 121 (22.45%) metabolites displayed an opposite trend. Fig. 5. [72]Fig. 5 [73]Open in a new tab KEGG pathway analysis of differential metabolites identified during carrot development. Top 20 pathway enrichment categories were displayed Fig. 6. [74]Fig. 6 [75]Open in a new tab Heatmap map visualization of the differential metabolites detected. Red indicates high abundance, whereas low relative metabolites are marked in green Fig. 7. [76]Fig. 7 [77]Open in a new tab Clustered metabolite change profiles from developing carrot roots Sugar accumulation during Carrot taproot enlargement A total of 34 differentially accumulated sugar metabolites in the subgroup were further identified during carrot root development (Fig. [78]8). Among them, 24 (70.59%) sugar-related metabolites, accounting for the majority of differential sugar compounds, showed a continuous increase trend with the development of fleshy roots. Meanwhile, they were also grouped into the fifth subclass by K-means clustering (Fig. [79]7; Additional file 1). These metabolites consisted of d-sucrose, d-glucose, d-fructose, which were commonly recognized as transport substrates for SWEETs family members. In the remaining 10 compounds, d-gucose-1,6-bisphosphate accumulated a high amount at the second stage when the root began to change color and enlarge, and was separately classified into the subclass 4 by K-means clustering. The accumulation of d-glucose 6-phosphate and d-(-)-threose continuously decreased over the period of root growth. The levels of the remaining seven substances, such as d-fructose-6-phosphate and glucose-1-phosphate, were not significantly changed at the first two stages, followed by a rapid decrease at the last stage. Fig. 8. [80]Fig. 8 [81]Open in a new tab Differential accumulated sugar metabolites identified during carrot taproot development. Red and green represent high and low accumulation, respectively Identification of DcSWEETs from Carrot Based on carrot genome, a total of 30 non-redundant SWEET genes were identified in carrot. All of the determined DcSWEETs harbored at least one conserved MtN3/saliva domain. The deduced DcSWEET proteins ranged from 201 (DcSWEET6a/DcSWEET6b) to 313 (DcSWEET9b/DcSWEET9c) amino acids in length, with theoretical pI value between 5.15 (DcSWEET17) and 9.64 (DcSWEET4b). The largest relative molecular weight was 35.22 kDa (DcSWEET9c), whereas DcSWEET6a displayed the least pI value (22.29 kDa) (Table [82]1). According to the results from TMHMM Server v2.0, 21 DcSWEET proteins harbored seven TMHs, accounting for 70% of the DcSWEETs identified. Besides, six DcSWEET proteins contained six TMHs, whereas only DcSWEET14d, DcSWEET14e, and DcSWEET16 had five TMHs. A phylogenetic tree containing 68 SWEET proteins from carrot, rice, and Arabidopsis was established to determine the evolutionary relationships of DcSWEET family members (Fig. [83]9). The results revealed that DcSWEET proteins were obviously classified into four different groups. Clade III is the largest team harboring 14 DcSWEETs (DcSWEET9a/9b/9c/10a/10b/11a/11b/11c/14a/14b/14c/14d/14e/15), whereas clade IV contained only two members (DcSWEET16/17). The distribution of conserved motifs within DcSWEET proteins was determined using the MEME program (Fig. [84]10). The results revealed that ten conserved motifs were identified in the DcSWEETs selected. Motifs 3 and 4 were observed in 29 DcSWEET proteins, whereas motif 9 was just present in DcSWEET9b and DcSWEET9c. Table 1. Characteristics of the DcSWEET genes in Carrot Gene Name Gene ID in carrot genome Clade Chromosomal Location ORF length (bp) Protein length (aa) pI Molecular weight (kDa) TMHs DcSWEET1a lcl|[85]CM004281.1_mrna_14392 I chr4: 22114959.22116637 (+) 768 255 9.22 28.2 7 DcSWEET1b lcl|[86]CM004278.1_mrna_772 I chr1: 11004719.11006286 (+) 729 242 9.22 26.43 7 DcSWEET1c lcl|[87]CM004280.1_mrna_10122 I chr3: 17457973.17459691 (+) 744 247 9.11 26.92 7 DcSWEET2a lcl|[88]CM004278.1_mrna_3581 I chr1: 43522694.43524854 (‒) 708 235 9.11 25.82 7 DcSWEET2b lcl|[89]CM004279.1_mrna_7729 I chr2: 38183455.38185302 (‒) 708 235 9.02 25.73 7 DcSWEET2c lcl|[90]CM004282.1_mrna_19802 I chr5: 40572339.40574123 (+) 705 234 7.6 26.43 7 DcSWEET3 lcl|[91]CM004280.1_mrna_10506 I chr3: 26359449.26360745 (‒) 642 213 9.16 23.93 6 DcSWEET4a lcl|[92]CM004279.1_mrna_6231 II chr2: 26196024.26197862 (+) 765 254 9.42 28 7 DcSWEET4b lcl|[93]CM004286.1_mrna_30551 II chr9: 14535029.14537866 (+) 639 212 9.64 23.2 6 DcSWEET5 lcl|[94]CM004282.1_mrna_16542 II chr5: 4032551.4033988 (‒) 708 235 8.78 25.97 7 DcSWEET6a lcl|[95]CM004280.1_mrna_11345 II chr3: 35732532.35733718 (‒) 606 201 8.89 22.29 6 DcSWEET6b lcl|[96]CM004280.1_mrna_10090 II chr3: 16679145.16680062 (+) 606 201 8.99 22.34 6 DcSWEET6c lcl|[97]CM004284.1_mrna_23726 II chr7: 1039981.1041357 (‒) 708 235 9.49 26.3 7 DcSWEET6d lcl|[98]CM004280.1_mrna_9640 II chr3: 10955889.10959666 (+) 696 231 8.33 25.62 6 DcSWEET9a lcl|[99]CM004281.1_mrna_14987 III chr4: 27069945.27071275 (‒) 816 271 8.71 30.23 7 DcSWEET9b lcl|[100]CM004280.1_mrna_12369 III chr3: 46735785.46737124 (+) 942 313 6.3 35.12 7 DcSWEET9c lcl|[101]CM004280.1_mrna_12368 III chr3: 46726435.46728006 (‒) 942 313 5.51 35.22 7 DcSWEET10a lcl|[102]CM004283.1_mrna_21674 III chr6: 21,111,221. 21,112,626 (+) 891 296 9.23 32.99 7 DcSWEET10b lcl|[103]CM004282.1_mrna_16410 III chr5: 2596668.2598112 (+) 885 294 9.05 32.46 7 DcSWEET11a lcl|[104]CM004283.1_mrna_21673 III chr6: 21100753.21102320 (+) 816 271 8.4 30.77 7 DcSWEET11b lcl|[105]CM004282.1_mrna_16409 III chr5: 2587385.2588747 (+) 813 270 9.15 30.29 7 DcSWEET11c lcl|[106]CM004284.1_mrna_26737 III chr7: 34,508,183. 34,509,774 (‒) 819 272 8.6 30.76 7 DcSWEET14a lcl|[107]CM004279.1_mrna_7298 III chr2: 34911760.34913165 (‒) 885 294 9.25 33.23 7 DcSWEET14b lcl|[108]CM004284.1_mrna_26777 III chr7: 34959313.34960798 (+) 780 259 9.1 28.97 7 DcSWEET14c lcl|[109]CM004284.1_mrna_26735 III chr7: 34495268.34496358 (‒) 858 285 9.3 32.33 6 DcSWEET14d lcl|[110]CM004284.1_mrna_26736 III chr7: 34500054.34501811 (‒) 753 250 8.2 27.64 5 DcSWEET14e lcl|[111]CM004284.1_mrna_26778 III chr7: 34964329.34965701 (+) 762 253 9.24 28.2 5 DcSWEET15 lcl|[112]CM004284.1_mrna_25018 III chr7: 17574227.17575693 (+) 894 297 6.3 33.26 7 DcSWEET16 lcl|[113]CM004278.1_mrna_1719 IV chr1: 25,902,347. 25,905,485 (‒) 651 216 7.75 23.75 5 DcSWEET17 lcl|[114]CM004286.1_mrna_31260 IV chr9: 26,111,289. 26,113,342 (+) 720 239 5.15 25.97 7 [115]Open in a new tab Fig. 9. [116]Fig. 9 [117]Open in a new tab The phylogenetic tree of SWEET family members from rice, Arabidopsis, and carrot. At, Arabidopsis thaliana; Os, Oryza sativa; Dc, Dacus carota. AtSWEET, OsSWEET, and DcSWEET proteins are represented by blue triangles, red squares, and green dot, respectively Fig. 10. [118]Fig. 10 [119]Open in a new tab The distribution of conserved motifs of DcSWEET family members. The colored bars correspond to ten different conserved sequences Expression profiles of DcSWEET genes during Carrot root development To understand the roles of DcSWEET genes in carrot root enlargement, the transcript abundance of DcSWEET genes at different developmental stages were measured according to the digital data from RNA-seq results. Totally, there were 17 DcSWEET genes detected to express during carrot root development, and most genes examined were differentially expressed to some extent (Fig. [120]11). To further validate the accuracy from the digital data, the relative expression levels of eight DcSWEET genes were determined by qRT-PCR technique (Fig. [121]12). The expression patterns of most genes were well correlated with the digital data. Specially, transcript levels of DcSWEET15 showed a continuous increase with the extension of time, whereas totally opposite results were observed for DcSWEET2a. DcSWEET17 was highly expressed at the first developmental stage, followed by extremely low mRNA abundance at the remaining stages. Transcription of DcSWEET2b and DcSWEET2c first decreased at the second and third stages and then increased. DcSWEET1b and DcSWEET6d were lowly and highly expressed at the second and fourth stages, respectively, displaying similar expression profiles during carrot root development. The transcript levels of DcSWEET1a had been declining at the first three stages and sharply increased at the fourth stage, followed by a sudden drop (Fig. [122]12). Fig. 11. Fig. 11 [123]Open in a new tab The heatmap of expression levels for DcSWEET genes during carrot development. Red and green represent high and low expression, respectively Fig. 12. [124]Fig. 12 [125]Open in a new tab Expression profiles of DcSWEET genes during carrot root development. The data were expressed as mean ± standard deviation Discussion Taproot development is a complex process that involving a series of alterations, such as anatomical enlargement, gene regulation, and especially metabolic changes [[126]24–[127]26]. Unlike fruits and other organs on the ground, underground roots are difficult to observe, study, and obtain without damage. Therefore, there is limited information about metabolic and molecular progress for root vegetables. With the introduction and rapid development of modern sequencing technologies, it has been possible to achieve huge amounts of molecular data in a short time to investigate taproot enlargement and quality formation in root vegetables [[128]27, [129]28]. Carrot, a typical root vegetable, is renowned for its rich nutritional values [[130]29]. However, to date, the accumulation profiles of a large number of metabolites during carrot root development have not been investigated. Over the period of root development, significant changes occur in the expression of a large number of genes and the accumulation of metabolites. Genome and transcriptome results indicated that pathways associated with carbohydrate metabolism were predominantly triggered during radish fleshy root enlargement, especially in cell proliferating tissues [[131]30]. The accumulation of ascorbate, an essential antioxidant and enzyme cofactor, was observed to continually decrease in radish flesh and skin with root growth [[132]31]. During turnip tuberous root development, plant hormone signal transduction pathway may contribute largely to root initiation, whereas secondary thickening process was well correlated with starch and sucrose accumulation [[133]32]. In carrot, the distribution and accumulation of anthocyanidins, lignins, and carotenoids have been investigated over the period of root enlargement [[134]27, [135]33, [136]34]. Contrary to the enlargement process, lignin content in carrot was demonstrated to progressively increase after maturity and post-harvest [[137]35, [138]36]. Over the early period of carrot growth, the concentrations of soluble solids, which were mainly composed of free sugars, were detected to continuously increase [[139]37]. However, a broad research of metabolite accumulation during carrot root enlargement is lacking. Here, a total of 539 metabolites were identified to be differentially accumulated with carrot root development. Interestingly, the levels of most lipid showed a continuous decreasing trend as the root development process prolonged (Fig. [140]6). A possible explanation is that carrot roots develop from seeds, and the lipid content at the early seedling stage is still high. As growth progresses, lipid decomposes to provide energy for seedling growth; meanwhile, with the enhancement of photosynthesis, more and more metabolic substances accumulate, resulting in a relative decrease in lipid content. The accumulation pattens of the differential metabolites would help to understand carrot root development, therefore providing new insights into regulation of carrot yield and quality formation. Sugar accumulation is one of the major intrinsic quality standards for vegetables and fruits [[141]38–[142]40]. Varieties with higher sugar content are increasingly popular and favored by people. In recent years, people’s pursuit of high-quality horticultural crops has become increasingly high. Therefore, strengthening the research and development of high sugar carrot varieties can meet people’s personalized needs. However, it is not yet clear what sugar types are present in carrot fleshy roots and how they accumulate. In this study, we found that there were 34 sugar metabolites differentially accumulated during carrot root development, suggesting a potential role of sugar in root development [[143]41]. In a previous study, we found that the sucrose levels in four carrot cultivars kept increasing during carrot root development, accompanied by elevated total soluble sugar accumulation [[144]42]. Here, most of the sugar metabolites including sucrose, fructose, and glucose, displayed a continuous increase trend, conforming to the results observed in the previous study. However, in a black carrot cultivar, soluble solids presented a marked peak 13 weeks after sowing, followed by an evident decrease [[145]37]. One possible reason is that 13 weeks may already be a relatively mature stage for carrots, and the decrease in soluble solids during the later growth stages may be to meet the needs of other growth and developmental processes. Additionally, it could also be due to factors related to the variety and growth conditions. The sugar accumulation profiles detected in the present study shed light on sugar metabolism regulation in carrot taproot development. SWEET proteins are the newly discovered type of sugar transporters in plants [[146]43]. Tomato SlSWEET5b was demonstrated to function in pollen mitosis and maturation by mediating hexose import [[147]44]. AtSWEET17 from Arabidopsis could make difference on carbohydrate allocation to affect branch elongation, especially under limited carbon supply [[148]45]. In rice, mutation of OsSWEET1b resulted in enhanced sugar starvation, reduced photosynthesis, and aggravated leaf cell death, thus causing yield loss [[149]46]. MeSWEET15a/b, two cell membrane-localized protein from cassava, were involved in the response to water and salt stress by regulating sugar accumulation and allocation [[150]3]. These findings further suggested that the various roles of SWEET proteins in plants. In the present study, a total of 17 DcSWEETs were found to be expressed in carrot roots, suggesting a potential role for sugar accumulation and root development. In particular, both digital gene expression and qRT-PCR results revealed that the transcript levels of DcSWEET15, DcSWEET2b, and DcSWEET6d gradually increased during carrot root growth (Fig. [151]10), correlated well with the accumulation of sucrose, glucose, and fructose accumulation (Fig. [152]8). More experiments and efforts need to be performed to further identify the roles of the candidate SWEET proteins. The current work provided substantial for understanding SWEET-mediated sugar accumulation and transport, although it was a preliminary work. In subsequent research, we will employ physiological and biochemical experiments, along with in vivo and in vitro approaches, to conduct precise analysis and identification of the functions of these candidate genes in sugar transport and related processes. Conclusions In the present study, 539 differentially accumulated metabolites were detected during carrot root development. Of them, 34 sugar related metabolites were further analyzed. Also, characteristics, phylogenetic tree, conserved motif discovery, and expression patterns of DcSWEETs were carried out to reveal their roles in differential sugar accumulation. The results from the current work would help carrot breeding and cultivation aimed at yield and quality improvement. Materials and methods Plant materials and growth conditions The seeds of carrot cultivar ‘Kurodagosun’ were sown in the pots containing a mixture of organic matter and vermiculite. The plants were cultivated in a greenhouse in Huaiyin Institute of Technology. According to morphological traits and growth time, carrot plants were classified into different developmental stages and harvested at 25, 40, 60, 75, and 90 days after sowing (DAS). Carrot taproots from different development stages were respectively collected, immediately immersed into liquid nitrogen, and preserved at an ultra-low temperature refrigerator for subsequent analysis. Metabolite profiling using UPLC-MS/MS The samples collected were vacuum freeze-dried in a freeze drier and ground to a powder state using a grinding miller (MM 400, Retsch) at 30 Hz for 1.5 min. Approximately 0.1 g powder was dissolved in 1.2 mL of 70% methanol extraction solution. The mixture was vortexed for 30 s every 30 min, a total of 6 times, and then placed in a refrigerator at 4 °C overnight. After centrifugation at a speed of 12,000 rpm for 10 min, the supernatant was extracted and filtered through a microporous filter membrane (0.22 μm pore size) and then stored in the injection bottle for UPLC-MS/MS analysis. The data acquisition instrument system mainly consists of Ultra Performance Liquid Chromatography (UPLC) (SHIMADZU Nexera X2) and Tandem mass spectrometry (MS/MS) (Applied Biosystems 4500 QTRAP). The samples were injected into an Agilent SB-C18 column (1.8 μm, 2.1 mm * 100 mm) working at 40 °C and a flow rate of 0.35 mL/min. The mobile phase system was made up of ultra-pure water with 0.04% acetic acid as solvent A and acetonitrile containing 0.04% acetic acid as solvent B. During elution gradient process, the proportion of phase B was 5% at 0 min, followed by a linear increase to 95% within 9 min, and remained at 95% for 1 min. Within the next 70 s, the proportion of phase B was reduced to 5% and equilibrated at 5% until 14 min. The effluent was then alternatively linked with an ESI-triple quadrupolelinear ion trap (Q TRAP)-MS. Linear ion trap (LIT) and triple quadrupole (QQQ) scans were achieved on a triple quadrupole linear ion trap mass spectrometer (QTRAP), AB4500 Q TRAP UPLC/MS/MS system. The system was furnished with an ESI Turbo Ion-Spray interface and can be carried out in both positive and negative ion modes with the Analyst 1.6.3 software (AB Sciex). The operation parameters of ESI source were as follows: ion source, turbo spray; source temperature, 550 °C; ion spray voltage 5 500 V; ion source gas I, gas II, and curtain gas were separately adjusted to 55, 60, and 25.0 psi; the collision gas was high. 10 and 100 µmol/L polypropylene glycol solutions were respectively introduced to implement instrument tuning and mass calibration. QQQ scans were accomplished by using multiple reaction monitoring (MRM) mode with collision gas (nitrogen) set to medium. De-clustering potential (DP) and collision energy (CE) were applied and optimized for MRM ion transitions. A specific set of MRM transitions were monitored within each period on the basis of the metabolites eluted. The metabolites were qualitatively determined according to the secondary spectrum information in the self-built metware database. During the analysis, isotopic signals, as well as repetitive signals containing K^+, Na^+, NH4^+, and fragment ions of large molecular weight substances were eliminated. Metabolomics data analysis The data acquired from the metabolite profiling were transformed for the principal component analysis (PCA) and orthogonal partial least squares-discriminant analysis (OPLS-DA). Samples with a variable importance in project (VIP) ≥ 1 and a fold change ≥ 2 or ≤ 0.5 were considered to be differential metabolites. Then, hierarchical cluster analysis (HCA), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, and K-means clustering were applied to further identify the differential metabolites. All the data of each sample for metabolite measurements were obtained in three independent biological replicates. Identification of sweets from Carrot To identify SWEET gene family members, the SWEET amino acid sequences from Arabidopsis and rice were searched against the carrot reference genome [[153]47]. The exported non-redundant protein sequences were authenticated by InterProScan ([154]http://www.ebi.ac.uk/interpro/) and SMART ([155]http://smart.embl-heidelberg.de/) forward to verify the existence of SWEET domains. Genes were assigned names based on their evolutionary relationships among SWEET proteins from Arabidopsis and rice. The SWEET family genes and their predicted encoding proteins were displayed in Additional file 2. The sequence length, molecular weight, and theoretical isoelectric point (pI) values of all DcSWEET proteins were determined by using the ExPASy program ([156]http://web.expasy.org/protparam). The TMHMM Server v.2.0 ([157]http://www.cbs.dtu.dk/services/TMHMM) was applied to detect the presence of transmembrane helices (TMHs). The conserved motifs within SWEET proteins were revealed by MEME ([158]http://meme-suite.org/index.html) and optimized using TBtools software. A phylogenetic tree harboring SWEET proteins from carrot, Arabidopsis, and rice was established by using the Neighbour-Joining method. Furthermore, the RNA-seq results generated from carrot roots at different developmental stages were obtained to identify the transcript profiles of DcSWEET genes [[159]48]. A heat map reflecting expression patterns of DcSWEET genes was created with HemI 2.0 software ([160]https://hemi.biocuckoo.org/). qRT-PCR analysis Total RNA extraction was carried out on carrot taproots from different developmental stages using an RNAprep pure plant kit (Tiangen, Beijing, China) in compliance with the manufacturer’s regulations. The RNA isolated was then reverse-transcribed into cDNA using HiScript II Q RT SuperMix for qPCR (Vazyme Biotech, Nanjing, China). Based on Primer Premier 6, the primers utilized for qRT-PCR were generated (Additional file 3). The qRT-PCR procedures were carried out on a CFX96 Real-Time PCR Detection System (Bio-Rad, California, USA) according to the instructions of ChamQ SYBR qPCR Master Mix (Vazyme Biotech, Nanjing, China) kit. The raw data exported were applied to calculate gene relative expression levels across all developmental stages using the 2^–ΔΔCt method and reference gene DcACTIN [[161]49]. Electronic supplementary material Below is the link to the electronic supplementary material. [162]Supplementary Material 1^ (24KB, xlsx) [163]Supplementary Material 2^ (149.5KB, xls) [164]Supplementary Material 3^ (10.9KB, xlsx) Author contributions GW and AX initiated and designed the research. GW wrote the main manuscript text. GW, YX, and YA performed the experiments. GW, JW, and YC analyzed the data. GW, ZH, and AX revised the paper. All authors read and reviewed the final manuscript. Funding This study was funded by the National Natural Science Foundation of China (32372681, 32102369), Natural Science Foundation of Jiangsu Province (BK20211366). Data availability The data sets supporting the results of this article are included within the article and its additional files. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Guanglong Wang, Email: 11160021@hyit.edu.cn, Email: guanglongwang@hyit.edu.cn. Aisheng Xiong, Email: xiongaisheng@njau.edu.cn. References