Two longan cultivars with contrasting sucrose contents, namely, ‘Songfengben’ (high sucrose accumulating type, SFB) and ‘Qingkebaoyuan’ (low sucrose accumulating type, QKBY), were used in this study for RNA-seq. The cultivars were collected at Fruit Research Institute, Fujian Academy of Agricultural Sciences, Fuzhou, China. Fruits of uniform size and free from visible defects were collected from three trees, with ten fruits from one tree were regarded as one replication, at 60, 90, and 120 days after flowering (DAF), respectively. After removing the peel and seed coat, the pulp was immediately preserved in liquid nitrogen and stored at −80°C until further analysis. The ‘Hongyan’ strawberry cultivar used for transient transformation were grown in a greenhouse with 16 h/8 h light conditions at 22°C.

Sugar content was measured using high performance liquid chromatography (HPLC) according to our reported protocol (Fang et al., 2020a). In brief, 0.5 g of fruit sample powder was dissolved in 6 mL deionized water and ultrasonic treatment for 15 min. After centrifugation at 5,000 rpm for 15 min at 4°C, the supernatant was collected and then filtered through a 0.22 μm Sep-Pak filter (ANPEL, Shanghai, China). Sugar content was detected using a 1260 HPLC system (Agilent, USA) with refractive index detector (RID). The mobile phase was deionized water and elution flow rate was 0.5 mL/min.

The RNAprep Pure PlantPlus Kit (Tiangen, Beijing, China) was used to extract RNA in accordance with the manufacturer’s instructions. Subsequently, the quality of the extracted RNA was assessed using both a NanoDrop 2000 spectrophotometer (Thermo, USA) and an Agilent 2100 Bioanalyzer (Agilent, USA). For library preparation, 1 μg of RNA from each sample was used as the starting material. The NEBNext Ultra™ RNA Library Prep Kit for Illumina (NEB, USA) was then utilized to construct the sequencing librariesaccording to the manufacturer’s guidelines. Sequencing was carried out on the Illumina NovaSeq 6000 platform, yielding paired-end RNA-seq reads. Raw data underwent preprocessing to eliminate adapter sequences, reads with poly-N sequences, and those of poor quality. The clean reads were then aligned to the ‘Honghezi’ longan genome (Lin et al., 2017) using HISAT2 with standard settings (Kim et al., 2015). The expression levels of genes were calculated using fragments per kilobase of transcript per millionmapped reads (FPKM). The differentially expressed genes (DEGs) analysis was performed using the DEseq2 R package, with an absolute log2 (fold change) ≥1 and false discovery rate ≤ 0.05 (Love et al., 2014). Both Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were facilitated by BMKcloud (https://www.biocloud.net/).

The WGCNA was conducted using the WGCNA plug-in (https://github.com/ShawnWx2019/WGCNA-shinyApp). The modules were constructed using the automatic network build function with the following parameters: power=22 and minModuleSize=30. Module-trait associations were considered statistically significant at p<0.05. Genes that were significantly associated with traits were identified within the modules, and their relationships were analyzed. The co-expression network was mapped using Cytoscape V3.8.2 software.

The qRT-PCR analysis were performed on the LightCycler 96 Real-Time PCR Systems (Roche, USA) with the TB Green Premix Ex Taq II (supplied by Takara in Dalian, China) and the primer sequences listed in Supplementary Table S1 . Actin was chosen as the reference gene (Jue et al., 2018). The relative gene expression levels were calculated using the 2^-ΔΔCt method (Livak and Schmittgen, 2001).

The coding region of DlSPS gene was amplified from the cDNA of ‘SFB’ fruit and inserted into the site of pSAK277 vector between Hind III and XbaI enzyme sites using ClonExpress^® IIOne Step Cloning Kit (Vazyme, China). Primer sequences are listed in Supplementary Table S1 . The recombinant constructs pSAK277-DlSPS and the empty vector pSAK277 were individually transferred into Agrobacterium tumefaciens EHA105 with the freeze-thaw method. Transient overexpressed in strawberry fruits was conducted as previously described (Pi et al., 2019).

Values are means ± standard deviation (SD) of three biological replicates. The significant differences between mean values was determined by the Student’s t test (*p<0.05 and **p<0.01) calculated using SPSS 21.0 (IBM, USA).

The soluble sugar component and content in two cultivars was mainly carried out at 60, 90, and 120 DAF. Throughout the developmental process, the sucrose content increased most rapidly and becomes the dominant soluble sugar at mature stage ( Figure 1 ). Notably, ‘SFB’ fruit exhibits a significantly higher sucrose content compared to ‘QKBY’ fruit at 120 DAF. However, no significant differences are observed at other developmental stages ( Figure 1A ). The glucose and fructose contents showed a slight increase from 60 DAF to 90 DAF, followed by a slight decrease from 90 DAF to 120 DAF. Differences in glucose content between the two cultivars are significant at three stages, while significant differences in fructose content are observed at 60 DAF and 120 DAF ( Figures 1B, C ). Based on the results, we infer that the higher sugar content and sweeter taste of ‘SFB’ longan fruit can be attributed to the higher accumulation of sucrose.

To study the transcriptional regulation of soluble sugar accumulation in longan, RNA-seq was conducted on two cultivars, ‘SFB’ and ‘QKBY’, during three critical developmental stages. An overview of the sequencing quality is provided in Supplementary Table S2 . After removing low-quality reads, the number of clean reads per library ranged from 20,418,767 to 81,405,449, with 72.57%-92.28% of the clean reads successfully mapped to the longan genome. The percentage of Q30 bases ranged from 94.60% to 95.63%, indicating the sequencing data was high quality. The GC content of the clean reads varied between 44.19% and 46.14%.

In total, 33,655 genes, including 4,960 novel genes, were identified across all fruit samples tested. For the annotation of these novel unigenes, we conducted a comprehensive search in public databases, including NR, Swiss-Prot, eggNOG, COG, GO, Pfam, KOG, and KEGG. The majority of the novel unigenes were annotated in the NR protein database (2539), followed by eggNOG (2204), Swiss-Prot (1465), KOG (1448), GO (1369), Pfam (1216), KEGG (669), and COG (363) ( Supplementary Table S3 ).

The gene expression levels among different fruit samples were assessed base on the FPKM value. Genes with FPKM values≥1.0 were considered expressed genes. Among the 33,655 identified genes, 17,055, 16,355, and 16,107 genes were expressed at 60 DAF, 90 DAF, and 120 DAF, respectively ( Figure 2A ). Of these genes, 14,015 were expressed throughout fruit development, while 1,458, 301, and 1,269 were specifically expressed at 60 DAF, 90 DAF, and 120 DAF, respectively ( Figure 2B ). A principal component analysis (PCA) of the expressed genes clustered them into six groups ( Figure 2C ). We further classified the expressed genes into five different groups based on their expression levels: extremely low-expressed genes with FPKM values less than 10, low-expressed genes with 10-30 FPKM, medium-expressed genes with 30-100 FPKM, and highly expressed genes with FPKM values above 100 ( Figure 2D ). Upon comparing the qRT-PCR evaluation of ten genes against the transcriptomic FPKM datasets, the expression patterns were similar to the RNA-Seq data ( Supplementary Figure S1 ).

To explore the expression-profile differences between two cultivars, the DEGs analysis was conducted. In ‘SFB’, 7,831 and 3,248 genes were found to be differentially expressed in S3 compared to S1 and S2 ( Figure 3A ). Of these, 3,512 genes were differentially expressed in S2 compared to S1 ( Figure 3A ). Specifically, 334 genes were up-regulated and 605 genes were down-regulated in S3/S1, S3/S2, and S2/S1 ( Figures 3B, C ). Similarly, in ‘QKBY’, 7,382 and 3,945 genes were differentially expressed in Q3 compared to Q1 and Q2 ( Figure 3D ). Among these, 2,987 genes were differentially expressed in Q2 compared to Q1 ( Figure 3D ). Notably, 346 genes were up-regulated and 347 genes were down-regulated in Q3/Q1, Q3/Q2, and Q2/Q1 ( Figures 3E, F ). When comparing ‘QKBY’ and ‘SFB’ at the same fruit development time points, 1,894 (Q1 vs. S1), 1,986 (Q2 vs. S2), and 2,667 (Q3 vs. S3) differentially expressed genes (DEGs) were identified ( Figure 3G ). After excluding redundant genes, a total of 12,350 unique DEGs were detected across various development stages and different longan varieties.

To gain insight into the functional properties of the DEGs, a comprehensive GO and KEGG analysis was conducted. The DEGs were categorized into three broad domains: molecular functions, cellular components, and biological processes. Within the cellular component category, DEGs were highly enriched in specific subcategories including the anchored component of the plasma membrane, the plant-type cell wall, the chloroplast envelope, the plasma membrane, and the chloroplast thylakoid membrane. In the biological process category, DEGs were further subclassified into a range of processes including response to light stimulus, regulation of cell size, cell proliferation, chlorophyll biosynthesis process, and polysaccharide biosynthetic process. Notably, within the molecular function category, the sole enriched function was ATP binding ( Figure 4A ).

The KEGG pathway classification analysis revealed that the DEGs were significantly associated with several key metabolic pathways. These included fructose and mannose metabolism, starch and sucrose metabolism, glucosinolate biosynthesis, and glycolysis/gluconeogenesis ( Figure 4B ).

As shown in Figure 1 , the sucrose content in ‘SFB’ fruit only exhibit significantly higher than that in ‘QKBY’ at 120 DAF. Therefore, we focused on the DEGs between the two varieties at this stage. Venn analysis revealed that 1,531 genes, which show differential expression specifically at 120 DAF, have been designated as 120 DAF-preferential ( Figure 3G ). To identify the key regulatory pathways involved in longan sucrose accumulation, we conducted KEGG pathway analysis on the 120 DAF stage-preferential genes. The results indicate that 11 annotated unigenes related to sucrose metabolism (starch and sucrose metabolism) are involved, including genes encoding pectinesterase, beta-glucosidase, endo-(1, 4)-beta-D-glucanase, 1, 4-alpha-glucan-branching, nudix hydrolase, sucrose-phosphate synthase, galacturonosyltransferase, phosphoglucomutase, and UDP-glucuronic acid decarboxylase ( Supplementary Figure S2 and Supplementary Table S4 ). Additionally, the glucose content showed significant differences between ‘SFB’ and ‘QKBY’ during fruit development. A total of 571 common DEGs were identified ( Figure 3H ). KEGG pathway analysis revealed that 5 annotated unigenes related to glucose metabolism (Glucosinolate biosynthesis and Glycolysis/Gluconeogenesis) are involved, including genes encoding UDP-glycosyltransferase, alcohol dehydrogenase, ATP-dependent 6-phosphofructokinase, pyruvate kinase, and hexokinase ( Supplementary Figure S3 and Supplementary Table S5 ).

Moreover, the fructose content in ‘SFB’ and ‘QKBY’ fruits exhibits significant variations at 60 DAF and 120 DAF. Notably, 718 common DEGs were identified in these two stages. KEGG pathway analysis of these genes led to the identification of two genes involved in fructose metabolism (Fructose and mannose metabolism), including one encoding ATP-dependent 6-phosphofructokinase and another encoding hexokinase ( Supplementary Figure S4 and Supplementary Table S6 ).

We conducted WGCNA to identify the co-expressed gene modules significantly associated with the content of sucrose, glucose, and fructose. The result identified 10 distinct modules, with the number of genes per module ranging from 134 (magenta) to 4,793 (turquoise) ( Supplementary Figure S5 and Supplementary Table S7 ). Notably, the red (r=0.9, p=0.00000028) and magenta (r=0.74, p=0.00047) module was significantly positively correlated with sucrose content. The yellow module was significantly positively correlated with glucose (r=0.86, p=0.000004) and fructose (r=0.83, p=0.000018) content ( Figure 5A ). To further explore key genes involved in sugar accumulation, venn plot was used to reveal the overlapped filtered genes of significantly modules genes and DEGs. As shown in Figure 5B , one gene (DlSPS) was identified to be involved in sucrose content. Furthermore, one gene (DlHK) and two genes (DlADH and DlHK) were involved in glucose and fructose content, respectively ( Supplementary Figure S6 ).

To examine the possible role of DlSPS in fruit sucrose accumulation, we transiently transferred pSAK277-DlSPS vector into immature white fruit of strawberry with empty pSAK277 vector as control. The qRT-PCR analysis showed that DlSPS was significantly higher expressed in transgenic fruit than that in the control ( Figure 6A ). The concentrations of sucrose and total sugar in the DlSPS-overexpression were significantly higher than those in the control. However, the glucose and fructose showed no significant difference between fruits overexpressing DlSPS or empty vector ( Figure 6B ).

To identify the key transcription factors regulating sugar content in longan, we first identified the transcription factors present in the magenta, red and yellow modules. The results revealed that a total of 30 transcription factors exist in these three modules, including bHLH (3), bZIP (1), ERF (4), MYB (4), NAC (3), and WRKY (1) ( Figure 7A ). As shown in Figure 5A , DlSPS is key gene regulating sucrose synthesis in longan fruit. Therefore, we investigated the potential regulation pathway between TFs and DlSPS. Notably, five TF genes that were categorized in the magenta module ( Figure 7B ) had similar expression trends to DlSPS. Furthermore, many binding sites of this five TFs on the promoters of DlSPS genes were also identified ( Supplementary Table S8 ).

Sweetness, which depends on the soluble sugar components and content, plays crucial role in determining fruit flavor and quality in longan. Various components of soluble sugar have been observed in plants, such as sucrose, fructose, glucose, raffinose, and galactose (Kotha et al., 2020; Zheng et al., 2022). In this study, we found that the main soluble sugars in longan fruit were sucrose, glucose, and fructose, which consistent with previous reports in other fruits such as apples (Ma et al., 2015) and peaches (Vimolmangkang et al., 2016). However, each soluble sugar component showed a distinct accumulation pattern, for instance, the sucrose content increased markedly during fruit development and exhibited significant differences between ‘SFB’ and ‘QKBY’ at the mature stage. While, the sucrose content was notably lower than previously research which conducted on the same two varieties in Guangdong province (Luo et al., 2019). This discrepancy may be caused by different weather and cultivation conditions, such as rainfall, light exposure, temperature, and fertilization. Furthermore, we observed a slight increase in glucose and fructose contents from 60 DAF to 90 DAF, followed by a slight decrease from 90 DAF to 120 DAF. At the mature stage, the contents of glucose and fructose were significantly lower than those of sucrose, consistent with previous research demonstrating that sucrose is the primary component of soluble sugar in longan.

The sugar accumulation during fruit development is regulated by a variety of biological processes, such as sugar metabolism and transportation (Chen et al., 2021). In previous studies, many important enzymes and transporters have been identified in longan, such as soluble acid invertases (SAI), vacuolar glucose transporters family (VGTs), and sugars will eventually be exported transporters (SWEET) (Luo et al., 2019; Fang et al., 2020a, 2022).

To gain a deeper understanding of the molecular mechanisms of sugar accumulation in longan, a comparative transcriptomic analysis was conducted and identified several potential genes involved in sugar metabolism, including SPS and HK. SPS is a key enzyme in sucrose metabolism. In melon, the activity of SPS is positively correlated with sucrose accumulation (Gao et al., 2024; Wang et al., 2023). In sugarcane, overexpression the SoSPS1 gene increased sucrose content in leaves (Anur et al., 2020). In this study, we found expression level of one SPS gene and sucrose content were significant differences between two cultivars at 120 DAF, suggesting that the SPS gene may be involved in sucrose metabolism in longan fruit. Furthermore, transient overexpression of the DlSPS gene enhances sucrose and total sugar content in strawberry fruit, demonstrating DlSPS gene is a strong candidate for the regulation of sugar accumulation in longan fruit. Hexokinase (HK), a multifunctional protein that serves as a sugar sensor for glucose signaling and a catalyst for glycolysis, has also been implicated in sugar metabolism. Overexpression of the PbHXK1 gene in tomato significantly reduced sugar content in leaves (Zhao et al., 2019). In this study, one HK gene showed significant expression levels between two cultivars during fruit development, indicating its potential function in longan soluble sugar metabolism.

Several studies have demonstrated that transcription factors play crucial roles in fruit sugar accumulation. For instance, in apple, MdbHLH3 modulates the soluble sugar content by activating phosphofructokinase gene expression (Yu et al., 2022), while MdMYB305 interacts with sugar-related genes, including MdCWI1, MdVGT3, and MdTMT2, thereby enhancing their activities and ultimately increasing the sugar content in fruits (Zhang et al., 2023). PuWRKY31 bound to the PuSWEET15 promoter and induced its transcription to increase sucrose content in pear (Li et al., 2020). In present study, three modules, including 30 transcription factors (TFs), were significantly positively correlated with soluble sugar content. Our RNA-seq results revealed that DlSPS expression was significantly different between the ‘SFB’ and ‘QKBY’ fruits ( Supplementary Table S4 ). The transient overexpression of the DlSPS gene in strawberry fruit validated its positive role in regulating sucrose accumulation. Previous studies revealed various TFs participated in sucrose synthesis via binding to promoters of SPS to alter its expression level. For instance, CdWRKY2 is a positive regulator in cold stress by targeting CdSPS1 promoters and activating its expression to mediate sucrose biosynthesis in bermudagrass (Huang et al., 2022). CmMYB44 repressed the transcriptional activation of CmSPS1, leading to lower sucrose content in melon (Gao et al., 2024). In present study, we found that five TFs exist in the same module as DlSPS gene and further analysis indicated thatseveral binding sites of this five TFs were found on the promoters of DlSPS genes, suggesting that these TFs might be involved in regulating sucrose pathways via interaction with DlSPS genes, thus affecting sucrose accumulation in longan. Taken together, these results partially elucidate the sucrose accumulation regulation process in longan and provide a theoretical foundation for fruit quality improvement. Obviously, more in-depth works needed to be carried out to identify our hypothesis.