Sugar signaling-mediated hormonal regulatory networks: molecular mechanisms underlying fruit abscission and development in seedless litchi

Abstract

Background

Seedless litchi (Litchi chinensis ‘Wu He’) exhibits an excellent parthenocarpic trait, the physiological regulation of which during early fruit development is critical for ultimate yield. Sugar availability serves as the predominant factor governing both fruit abscission and development. However, the molecular mechanisms by which sugar signaling mediates hormonal networks to regulate these processes in seedless litchi remain unclear. This study explored the regulatory patterns in seedless litchi during the massive abscission stage (MA stage) and subsequent developmental stages (SD stages), through integrated physiological and transcriptomic analyses.

Results

At the MA stage (15 days after flowering, 15 DAF), the concentrations of glucose and fructose in fruits dropped to their lowest levels. Concurrently, auxin and abscisic acid (ABA) levels increased significantly, while the ethylene (ET) precursor (1-aminocyclopropane-1-carboxylic acid, ACC) content declined. Tissue section revealed that both cell size and cell number were at their lowest, indicating delayed fruit development. Upon entering the SD stages (22 DAF, 29 DAF, 36 DAF), fruit abscission progressively decreased as sugar availability improved. During these stages, cytokinin (iP-type-CK) and gibberellin (GA₃) levels increased, accompanied by synchronized increases in cell volume and number, leading to continuous fruit enlargement. Transcriptomic analysis demonstrated that differentially expressed genes were primarily enriched in pathways related to starch and sucrose metabolism, hormone metabolism and signal transduction. During the MA stage, expression of genes such as MAT, ACS, ACO, CTR1, ZEP,NCED, PP2C, and SnRK2 was upregulated, whereas EIN2 was downregulated. In contrast, the SD stages exhibited upregulation of IPT,B-ARR, KAO, GID1, GID2, TFs and DELLA genes, alongside downregulation of CKX, A-ARR, and GA2ox. These gene expression patterns were consistent with fluctuations in hormone levels and collectively regulated fruit abscission and development. RT-qPCR validation confirmed the accuracy of the RNA-Seq results.

Conclusions 

Deficiencies in fructose and glucose mediate ABA and ET accumulation while disrupting polar auxin transport, thereby exacerbating fruit abscission. Following massive fruit abscission, improved sugar availability enhanced the effects of CK and GA, which promoted cell division and expansion, resulting in gradual fruit enlargement. These findings elucidate the sugar signaling-mediated hormonal regulatory networks during seedless litchi abscission and development, providing a theoretical foundation for achieving high and stable yields in orchards.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-025-07557-9.

Introduction

Angiosperms complete their reproductive cycle through double fertilization, with fruit set commencing following successful pollination and fertilization [1]. However, insufficient nutrient supply and environmental stresses (e.g., drought, extreme temperatures) may disrupt plant reproductive processes, leading to flower, seed, or fruit abortion, thereby significantly reducing fruit yield [2, 3]. Certain horticultural crops exhibit parthenocarpy, a desirable trait that maintains fruit yield under unfavorable pollination or fertilization conditions [4]. Research indicates that parthenocarpy can be promoted through hormonal regulation, thereby enhancing fruit growth and development. For instance, natural parthenocarpy occurs when endogenous ovarian hormone concentrations exceed the threshold required to autonomously trigger growth and development in the absence of pollination or fertilization [5]. Su et al. (2021) demonstrated that the highly parthenocarpic line displays elevated cytokinin (CK) and gibberellin (GA) levels alongside reduced abscisic acid (ABA) during cucumber (Cucumis sativus) fruit development compared to the weakly parthenocarpic line [6]. To induce parthenocarpy, exogenous plant growth regulators are widely applied, including N-(2-chloro-4-pyridyl)-N′-phenylurea, 24-Epibrassinolide, 2,4-Dichlorophenoxy-acetic acid, 1-Methylcyclo-propene, and GA₃ [1]. Critically, recent studies have demonstrated that hormonal induction of parthenocarpy is closely associated with sugar signaling [7, 8].

Previous research has demonstrated that sugar signaling directly governs fruit physiological processes, whereas hormonal effects are mediated through sugar metabolic pathways. For example, ABA spraying-induced fruit abscission in ‘Bartlett’ and ‘D’Anjou’ pears (Pyrus communis) is mediated by carbohydrate deficiency [9, 10]. Benzyladenine spraying regulates apple fruit abscission and development through the induction of sugar stress in fruits [11]. As the primary product of photosynthesis, sugar supplies energy for plants under both normal and stress conditions while regulating key physiological processes [12, 13]. Studies have demonstrated that sucrose, fructose, and glucose function not only as carbon sources but also as signaling molecules, playing pivotal roles in parthenocarpic fruit development [7]. Kusano et al. (2022) demonstrated that enhanced sugar metabolism is essential for early fruit set induction in the parthenocarpic process [8]. In Citrus clementina, parthenocarpic capacity correlates strongly with efficient glucose and fructose mobilization, and is regulated by GA-auxin (IAA)-ABA crosstalk [14]. In cucumber, spermidine treatment has been shown to trigger parthenocarpy by enhancing sugar and photosynthate accumulation in fruits [15]. Similarly, treatment with sucrose, fructose, or glucose has been demonstrated to increase fruit set rates in weakly parthenocarpic cucumbers through enhanced IAA and CK signaling [7]. Furthermore, parthenocarpic fruits exhibit upregulated sugar biosynthesis that feeds back to hormone pathways, promoting fruit differentiation [16, 17]. Although the pivotal role of sugar signaling in parthenocarpy has been well-documented, the regulatory mechanisms underlying its function in seedless litchi fruit formation remain largely elusive.

The seedless litchi (Litchi chinensis ‘Wu He’) represents the only commercially cultivated litchi variety capable of producing seedless fruits. This cultivar is characterized by larger fruit size, delicate flesh texture, elevated sweetness, and a higher edible proportion, conferring significant market value [18]. Anatomical studies have demonstrated that the litchi ovary contains two ovules. However, high embryo abortion rates due to post-pollination ovule developmental abnormalities frequently restrict ovary development to a single fruit, which is a key yield-limiting factor [19]. In contrast, seedless litchi circumvents ovule developmental constraints, enabling dual-fruit development from a single ovary and demonstrating significant yield enhancement potential (Fig. 1). However, seedless litchi exhibits markedly higher preharvest fruit abscission rates (peaking before 15 days after flowering, 15 DAF) compared to seeded varieties, severely limiting commercial production [18, 20]. This phenomenon parallels observations in other parthenocarpic species, including persimmon (Diospyros Kaki) [21] and tropical squash (Cucurbita moschata) [22]. Exogenous hormone-mediated parthenocarpy also leads to increased fruit abscission rates [23]. Currently, the molecular mechanisms by which sugar signaling mediates hormonal regulatory networks to coordinate both fruit abscission and development in seedless litchi remain unclear.

Fig. 1.

Fruit growth and abscission rate changes in seedless litchi. A Representative photographs of seedless litchi fruit growth. B Changes in relative fruit abscission rate during growth. C Changes in cumulative fruit abscission rate during growth. D Changes in fruit length during growth. E Changes in fruit width during growth. F Changes in fruit weight during growth

This study used ‘A4 seedless’ litchi as the experimental material. First, early fruit developmental traits and the critical abscission stage were characterized through comprehensive analysis of growth curves, histological sections, and abscission rate dynamics. Subsequently, physiological parameter quantification, integrated with transcriptomic profiling, was used to elucidate the molecular mechanisms underlying sugar signaling-mediated hormonal regulatory networks during both the massive abscission stage (MA stage) and subsequent developmental stages (SD stages).

Materials and methods

Fruit material, physical measurements and abscission rate statistics

The experimental site was located at the ‘A4 seedless’ litchi base (‘Miaoguoshu’; 109°55′E, 19°49′N) in Chengmai County, Hainan Province China. Chemical emasculation was applied during the flowering period of the litchi trees to ensure the production of seedless fruits. Eighteen healthy eight-year-old litchi trees with synchronized flowering were selected for this study. From nine of these trees, well-developed twin fruits were collected at four developmental stages: 15, 22, 29, and 36 days after flowering, designated as 15 DAF, 22 DAF, 29 DAF, and 36 DAF, respectively. At each sampling time point, 150 fruits were measured for length, width, and weight using a digital vernier caliper (CD-20APX, Mitutoyo, Japan) and an electronic balance (ME204/02, METTLER TOLEDO Instruments Shanghai Co., Ltd.) following the method described by Gonçalves et al. (2015) [24]. A subset of the collected fruits was immediately fixed in FAA solution (70% ethanol; G1103, Servicebio, Wuhan, China) for histological analysis, while the remaining samples were stored at −80°C for subsequent analysis. All measurements were performed in triplicate. For the remaining nine trees, four fruiting branches were selected from different cardinal directions on each tree to monitor fruit abscission. Relative and cumulative fruit abscission rates were calculated according to Guo et al. (2024), with three trees serving as one biological replicate [25].

Microscopic observation of fruit sections

Paraffin sectioning was performed to examine cellular morphology in longitudinal sections of litchi fruits at different developmental stages. Sections were stained with Safranin O-Fast Green (G1031, Servicebio, Wuhan, China) and observed under a Nikon Eclipse E100 microscope. Digital images were captured using CaseViewer software v2.4.0. Following the protocol of Wang et al. (2022), cell length, width, and area were measured within a standardized 150 μm × 150 μm region using ImageJ software (NIH, USA), and the total number of cells in each sample was quantified [26].

Sugar content determination

Four sugars (sucrose, fructose, glucose and galactose) were quantified following our previously established method [27]. Sugar concentrations were expressed as micrograms per gram fresh weight (µg/g FW).

Hormone content determination

Phytohormone content in litchi fruits was analyzed using MetWare (http://www.metware.cn/) based on the AB Sciex QTRAP 6500 LC-MS/MS platform following established extraction protocols [28, 29]. Chromatographic separation was performed on a Waters ACQUITY UPLC HSS T3 C18 column (100 mm × 2.1 mm, 1.8 μm) using a binary solvent system of 0.04% acetic acid in water (A) and 0.04% acetic acid in acetonitrile (B) with a 12-min gradient (0–1 min: 5% B; 1–8 min: increased to 95% B; 8–9 min: 95% B; 9–12 min: ramped back to 5% B) at 0.35 mL/min and 40 °C, with a 2 µL injection volume [30–32]. Mass spectrometric detection was conducted using a QTRAP 6500 + system (SCIEX, USA) with an ESI Turbo IonSpray interface operating in both positive and negative ion modes (ion source temperature at 550℃, spray voltage + 5500 V [positive] and − 4500 V [negative], curtain gas at 35 psi), employing scheduled multiple reaction monitoring (MRM) with individually optimized declustering potentials and collision energies for each phytohormone through direct infusion experiments using authentic standards [33–35]. Data acquisition and quantification were performed using Analyst 1.6.3 and MultiQuant 3.0.3 software (SCIEX), respectively, with MRM transitions monitored during specific elution windows for each metabolite. Subsequently, a total of 89 metabolites associated with the biosynthesis pathways of IAA, CK, GA, ABA, and ethylene (ET) were analyzed using the orthogonal partial least squares discriminant analysis (OPLS-DA; https://cloud.metware.cn/#/home) to characterize the hormonal changes (variable important in projection, VIP >1) in fruits during the MA and SD stages.

Transcriptome sequencing and identification of differentially expressed genes (DEGs)

Total RNA was extracted from samples using the RNAprep Pure Plant Plus Kit (DP441, Tiangen Biotech, Beijing, China) following the manufacturer’s protocol. RNA concentration and integrity were assessed using a Qubit 4.0 Fluorometer (Thermo Fisher Scientific) and Qsep400 Bioanalyzer (Bioptic Inc.), respectively. High-quality RNA samples were submitted to Wuhan Maiwei Metabolism Biotechnology Co., Ltd. for cDNA libraries construction. Following quality assessment of the constructed library by both Qubit dye methods and qPCR, sequencing was performed on the Illumina HiSeq platform. Raw sequencing data were processed by removing adapter sequences and low-quality reads (>50% bases with Q ≤ 20 per read) to generate clean reads. Gene expression levels were quantified using fragments per kilobase of transcript per million mapped reads (FPKM). Differential gene expression analysis was conducted according to the method described by Zhang et al. (2021), with significance thresholds set at |log2 (fold change)| ≥ 1 and adjusted p-value < 0.05 [36].

RT-qPCR validation of gene expression

To validate the reliability of the RNA-Seq data, selected DEGs involved in hormone metabolic pathways were analyzed by RT-qPCR. Gene-specific primers were designed using Beacon Designer 7.9 (Premier Biosoft). Total RNA was extracted from fruits at different developmental stages using the Magnetic Hi-Plant RNA Kit (DP772, Tiangen), followed by cDNA synthesis using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific). RT-qPCR was performed in MicroAmp Fast Optical 96-Well Reaction Plates (0.1 mL, Applied Biosystems) with three biological replicates × three technical replicates per reaction. Relative gene expression levels were normalized using the reference gene GAGA-25 [37] and calculated using the 2^−ΔΔCT method.

Statistical analysis

Statistical analysis of the experimental data was performed using WPS Office software, while significance analysis was conducted using ANOVA in SPSS software (version 27.0.1). Data are presented as means ± standard deviation. Correlation analysis was conducted on the Metware Cloud (https://cloud.metware.cn/#/home).

Results

Fruit abscission and developmental characteristics in seedless litchi

The ‘A4 seedless’ litchi typically reaches maturity at 120–130 DAF. To investigate the fruit abscission and developmental characteristics during the MA stage and SD stages, this study measured and statistically analyzed the fruit abscission rate, size parameters (individual fruit weight, transverse diameter, and longitudinal diameter), and cellular morphology (cell length, width, area, and number) at 15 DAF, 22 DAF, 29 DAF, and 36 DAF (Fig. 1). For abscission rate calculations, the rate at 8 DAF was set as a 0% baseline. Results showed that the relative fruit abscission rate peaked at 15 DAF (67%), then gradually decreased to 6% by 36 DAF (Fig. 1B). Meanwhile, the cumulative fruit abscission rate showed its most pronounced increase at 15 DAF, reaching 67%, after which the rate of increase slowed, with the cumulative rate reaching 87% by 36 DAF (Fig. 1C). Over two-thirds of fruit abscission occurred at 15 DAF, indicating that this stage was the MA stage. In the SD stages (22 DAF, 29 DAF and 36 DAF), fruit abscission gradually decreased. During seedless litchi development, the longitudinal diameter, transverse diameter, and single fruit weight all showed progressive increases (Fig. 1D, E, F). Notably, accelerated growth rates were observed at 22 DAF following the MA stage. The trends in cellular parameters paralleled those of fruit size (Fig. 2A, B). Throughout fruit development, cell length, width, and area progressively increased (Fig. 2C, D, E), with particularly pronounced growth occurring at 22 DAF. Concurrently, cell number increased continuously during fruit development, however, the proliferation rate appeared to decelerate after the MA stage (22 DAF, Fig. 2F).

Fig. 2.

Microscopic observation and statistical analysis of seedless litchi fruit sections. A Fruit sections. B Cellular morphological changes during fruit growth. C Changes in cell length during growth. D Changes in cell width during growth. E Changes in cell area during growth. F Changes in cell number during growth

Sugar accumulation during MA and SD stages

To evaluate the sugar accumulation in seedless litchi during the MA and SD stages, we measured the levels of soluble sugars (sucrose, fructose, glucose and galactose) in fruits (Fig. 3). The results showed that the fruit sucrose content was 1199.81 µg/g at 15 DAF, decreased to 1040.23 µg/g at 22 DAF, and then gradually increased to peak at 1281.41 µg/g by 36 DAF (Fig. 3A). In contrast, glucose and fructose concentrations were at their lowest levels (1597.91 µg/g and 1313.68 µg/g) at 15 DAF, then increased to a peak (3257.24 µg/g and 2034.78 µg/g) at 36 DAF (Fig. 3B, C). Glucose and fructose contents increased by 103.84% and 54.89%, respectively, from 15 to 36 DAF. Galactose content peaked at 15 DAF (105.09 µg/g), decreased to its lowest level at 29 DAF (66.93 µg/g), and then increased slightly (Fig. 3D). Glucose and fructose concentrations were lowest during the MA stage compared to the SD stages, while sucrose levels, although moderately elevated, remained below those at 36 DAF. During the SD stages, concentrations of sucrose, fructose, and glucose showed an increasing trend.

Fig. 3.

Changes in sucrose (A), fructose (B), glucose (C), and galactose (D) content during MA and SD stages

Five key hormonal changes during MA and SD stages

Based on the OPLS-DA model, we identified the key differential hormone metabolites in fruit during the MA and SD stages. The OPLS-DA score plot revealed a clear separation between the MA and SD stages (Supplementary file Fig. S1A). By integrating VIP values and OPLS-DA loading S-plot analysis (Supplementary file Fig. S1B), a total of 17 metabolites with significant effects on fruit abscission and development were screened, including four biologically active hormones [IAA, ABA, N6-(∆2-isopentenyl)-adenine (iP-type CK), GA₃] and one key precursor (1-aminocyclopropane-1-carboxylic acid, ACC) for ET synthesis. Figure 4 showed that during the MA stage, the concentrations of IAA and ABA peaked, while iP, GA₃, and the ACC reached their lowest levels. In contrast, the SD stages showed significantly reduced IAA and ABA levels but markedly elevated iP, GA₃, and ACC concentrations.

Fig. 4.

Changes in IAA (A), iP (B), GA₃ (C), ABA (D) and ACC (E) content during MA and SD stages

Transcriptome sequencing and gene expression analysis during MA and SD stages of seedless Litchi

We utilized RNA-seq to examine gene expression changes during the MA and SD stages of seedless litchi development and calculated intragroup and intergroup correlations based on FPKM values in the samples. Intragroup sample R² values were ≥ 0.985 (Fig. 5A). During fruit development, the R² values between groups gradually decreased, with the R² value between 15 DAF and 36 DAF dropping to ≤ 0.901. Principal component analysis (Fig. 5B) showed that the intragroup samples formed tight clusters, while intergroup samples were more dispersed. DEG clustering analysis across four developmental stages of litchi fruit (Fig. 5C) showed that 15 DAF clustered with (22 DAF + 29 DAF) as one group, while (15 DAF + 22 DAF + 29 DAF) clustered with 36 DAF as another group. This clustering pattern was consistent with the developmental process of litchi fruit. To explore gene expression differences during the MA and SD stages in litchi fruit, we established three comparison groups: 15 DAF vs. 22 DAF, 15 DAF vs. 29 DAF, and 15 DAF vs. 36 DAF (Fig. 5D). When 15 DAF was used as the control group, the number of common DEGs among the 15 DAF vs. 22 DAF, 15 DAF vs. 29 DAF, and 15 DAF vs. 36 DAF comparisons reached 2727, accounting for 77.98%, 64.51%, and 36.88% of the total DEGs in each respective group. Common DEGs between the MA stage and the three SD stages on average exceeded 50% of total DEGs, indicating substantial gene expression differences during fruit development.

Fig. 5.

Overview of transcriptomic data. A Inter-samples correlation heatmap analysis. B Principal component analysis (PCA) of transcriptome samples. C Cluster Analysis of DEGs (Color bar: Relative expression level of gene; Red/blue colors indicate up-/down-regulated DEG respectively). D Venn diagram of DEGs

GO functional annotation and enrichment analysis of DEGs

GO functional annotation and enrichment analysis was performed on DEGs with p-values ≤ 0.05 (Fig. 6). Most DEGs in the 15 DAF vs. 22 DAF, 15 DAF vs. 29 DAF, and 15 DAF vs. 36 DAF comparisons were enriched in the categories of molecular function (MF) and biological process (BP), indicating that active cellular metabolism occurs during the early stages of seedless litchi fruit development. DEGs enriched in the cellular component (CC) category mainly originated from the comparisons of 15 DAF vs. 29 DAF and 15 DAF vs. 36 DAF, with higher enrichment in the latter. This indicates progressive changes in cellular components during the seedless litchi fruit development. In the MA vs. SD comparisons, DEGs were most enriched in BP, including polysaccharide catabolic process, diterpenoid biosynthetic process, and the ET-activated signaling pathway, etc. DEGs enriched in MF were mainly associated with glycosyl compounds hydrolysis and hormone metabolism regulation. The DEGs in the CC category were primarily enriched in the “anchored component of membrane”.

Fig. 6.

GO function annotation and classification of three comparison groups. X-axis: Number of genes. Y-axis: GO terms. The padj values labeled after bar chart

KEGG pathway annotation and analysis of DEGs

KEGG functional annotation and enrichment analysis was performed on DEGs to elucidate their metabolic functions. Figure 7 displays the pathways associated with “starch and sucrose metabolism” and “hormone metabolism and signal transduction”. Among them, the pathways of tryptophan metabolism, zeatin biosynthesis, diterpenoid biosynthesis, carotenoid biosynthesis, and cysteine and methionine metabolism are closely associated with the biosynthesis of IAA, CK, GA, ABA, and ET, respectively. In the MA vs. SD comparisons, 80% of KEGG pathways showed increasing DEG numbers during development. To investigate the expression differences of hormone biosynthesis and signal transduction-related genes between the MA and SD stages, we identified 140 DEGs (including protein-encoding and transcription factor genes) from the transcriptome. Compared with the SD stages, the MA stage displayed the lowest expression levels of IAA biosynthesis genes [N-hydroxythioamide S-beta-glucosyltransferase, UGT74B1; indole-3-pyruvate monooxygenase, YUCCA; and amidase, amiE (LITCHI001981, novel.1607); Fig. 8A], while most auxin-responsive protein IAA (AUX/IAA) and auxin response factor (ARF) family genes in the IAA signaling pathway were upregulated (Fig. 9A). ABA biosynthesis genes zeaxanthin epoxidase (ZEP) and 9-cis-epoxycarotenoid dioxygenase (NCED) reached peak expression (Fig. 8D), accompanied by activation of core ABA signaling components protein phosphatase 2 C (PP2C) and serine/threonine-protein kinase SRK2 (SnRK2; Fig. 9D). In ET biosynthesis, S-adenosylmethionine synthetase (MAT), 1-aminocyclopropane-1-carboxylate synthase (ACS), and aminocyclopropanecarboxylate oxidase (ACO, LITCHI017644) showed maximal expression (Fig. 8E). In contrast, ET signal transduction components showed divergent regulation: serine/threonine-protein kinase CTR1 (CTR1) was up-regulated while mitogen-activated protein kinase kinase 4/5 (SIMKK) and ethylene-insensitive protein 2 (EIN2) were down-regulated (Fig. 9E). Compared with the MA stage, the SD stages showed upregulation of the CK biosynthesis key gene adenylate dimethylallyltransferase (IPT) and downregulation of the catabolic gene cytokinin dehydrogenase (CKX; Fig. 8B). Concurrently, the CK signaling pathway showed activation of two-component response regulator ARR-B family (B-ARR) genes (LITCHI016915, LITCHI020996) and suppression of two-component response regulator ARR-A family (A-ARR) gene (Fig. 9B). For GA metabolism, the biosynthetic gene ent-kaurenoic acid monooxygenase (KAO) was upregulated while gibberellin 2beta-dioxygenase (GA2ox) and gibberellin 3beta-dioxygenase (GA3ox) were downregulated (Fig. 8C), with activation of the signaling components gibberellin receptor GID1 (GID1; LITCHI011965, LITCHI007943, LITCHI007944, LITCHI023882), F-box protein GID2 (GID2), phytochrome-interacting factor (TF) and most DELLA protein (DELLA) family genes (Fig. 9C). Interestingly, 55% of the aforementioned genes exhibited significant or extremely significant correlations between their expression patterns and the concentrations of fructose and glucose during fruit abscission and development (Supplementary file Fig. S2).

Fig. 7.

KEGG pathway enrichment analysis of three comparison groups. X-axis: Number of genes. Y-axis: KEGG terms. The P values labeled after bar chart

Fig. 8.

Changes in metabolite levels in IAA (A), CK (B), GA (C), ABA (D), and ET (E) biosynthesis pathways and differential gene expression F. Four squares/circles represent 15 DAF, 22 DAF, 29 DAF, and 36 DAF (left to right). Color bars represent metabolite abundance or gene expression level, with warm colors indicating high levels/upregulation and cool colors indicating low levels/downregulation

Fig. 9.

Signaling transduction pathways of IAA (A), CK (B), GA (C), ABA (D), and ET (E) with differential gene expression. Four squares represent 15 DAF, 22 DAF, 29 DAF, and 36 DAF (left to right). Color bar represents the relative expression level of gene, with red and blue representing upregulation and downregulation, respectively

Quantitative RT-qPCR analysis

RT-qPCR analysis showed that the expression patterns of nine genes related to hormone biosynthesis/signal transduction were consistent with the transcriptome sequencing results (Fig. 10). During the MA stage, the YUCCA (IAA biosynthesis gene) and ABA2 (ABA biosynthesis gene) showed downregulated expression, likely due to feedback inhibition caused by elevated hormone levels. Concurrently, key ET biosynthesis genes (MAT and ACO) were significantly upregulated, while the signal transduction gene SIMKK in fruits was downregulated. These transcriptional changes may enhance ET sensitivity in the abscission zone, consequently promoting fruit abscission. During the SD stages, the coordinated upregulation of KAO and downregulation of GA2ox synergistically promoted the accumulation of active GA₃. Furthermore, the activation of GID2 enhanced GA signal transmission. These regulatory changes likely contribute to fruit development by stimulating cell elongation and growth.

Fig. 10.

RT-qPCR analysis (A) and RNA-seq heatmap (B) for nine genes. B: Color bar represents the relative expression level of gene, with red and blue representing upregulation and downregulation, respectively

Discussion

Fruit set involves the ovary transitioning from quiescence to active growth, typically initiated by pollination and fertilization of the ovules, followed by subsequent ovary development. However, certain species exhibit parthenocarpy, whereby fruits develop without pollination or fertilization. Recent studies have characterized parthenocarpic fruit development by analyzing phenotypic traits and cytological changes in ovary tissues, such as pomelo (Citrus grandis) [38], red raspberry (Rubus idaeus) [39], and tomato (Solanum lycopersicum) [40]. Investigating abscission and developmental patterns in seedless litchi can provide valuable insights into parthenocarpic mechanisms and high-yield cultivation.

Sugar deficiency mediates fruit abscission and promotes sugar redistribution to retained fruits

Fruit development is controlled by endogenous ovarian hormones; and, this physiological process requires adequate carbohydrate supply. As heterotrophic organs, developing fruits rely predominantly on photoassimilates derived from leaf photosynthesis for growth. When fruit load exceeds the tree’s carrying capacity, this creates a carbohydrate supply-demand imbalance, thereby triggering severe physiological fruit abscission [41]. Carbohydrate allocation strategies, such as main-stem girdling and fruit thinning, enhance carbohydrate allocation to fruits and reduce inter-fruit competition, thereby significantly improving fruit size and development rates in various species including cherry [42] and pomegranate [43]. Conversely, artificial defoliation or branch girdling disrupts the tricarboxylic acid cycle and carbohydrate metabolism in young fruits, causing carbon starvation and energy deficiency that severely inhibits normal growth and exacerbates abscission [44, 45]. The A4 seedless litchi produces a huge quantity of flowers during flowering, and even after artificial flower thinning, a large number of fruits still remain on the tree [46]. The results of this study demonstrate that the A4 seedless litchi exhibits a remarkably high cumulative abscission rate of 87% during the early fruit development stage, with over two-thirds of this abscission occurring within the 15 DAF. Therefore, the starvation stress caused by fruit numbers exceeding the tree’s carrying capacity may be the primary driver of severe early fruit abscission, which aligns with previous reports in litchi (Litchi chinensis Sonn. cv. Wuye [47]) and longan [48].

Sugars serve as critical regulators of metabolic and physiological processes in fruit cells. As primary photosynthetic products, soluble sugars (e.g., sucrose) are transported to fruits via the phloem and hydrolyzed into fructose and glucose. These sugars not only provide energy and carbon skeletons for fruit growth and development but also participate in abiotic stress responses [49]. Studies have demonstrated that the concentration and metabolism of sucrose, fructose, and glucose in fruits determine the abscission patterns during fruit development [45, 50]. Yi et al. (2023) demonstrated that hexoses, particularly glucose, play a crucial role in reducing fruit abscission, with exogenous glucose application significantly decreasing cumulative fruit abscission rates in litchi [51]. Yang et al. (2015) found that girdling-defoliation treatment reduces the concentrations of glucose and fructose in longan fruits, leading to substantial abscission [52]. Guo et al. (2020) also reached a similar conclusion that the main cause of fruit abscission in sweet cherry in low-temperature regions is the failure of glucose and fructose contents in the fruits to reach the required thresholds [53]. Consistent with these findings, we observed the lowest glucose and fructose concentrations during the MA stage. This indicates that the sharp reduction in available sugars may represent the initial event triggering fruit abscission under starvation stress, which is consistent with the conclusions of Yang and Xiang (2022) [45]. Notably, sucrose concentration remained relatively high at 15 DAF, significantly exceeding levels observed at 22 DAF and 29 DAF. During periods of heavy fruit abscission, sugar starvation may induce reactive oxygen species (ROS) accumulation [11], with elevated sucrose levels potentially representing an adaptive response to both sugar starvation and oxidative stress [11, 54]. Following substantial fruit abscission, remaining fruits sustained growth accompanied by elevated sucrose, fructose, and glucose concentrations, with the accelerated growth rates at 22 DAF. Cytological observations showed that during SD stages, substantial abscission promoted rapid cell growth and sustained division in the remaining fruits. These results showed that sugar allocation in seedless litchi may be actively redirected to retained fruits following physiological fruit abscission, sustaining subsequent growth and development. The large number of significant correlations between sugars (glucose and fructose) and the genes involved in hormone metabolism and signal transduction also demonstrate this viewpoint.

Sugar deficiency activates ABA and ET signaling and disrupts polar IAA transport resulting in fruit abscission

In plants, the crosstalk between sugar and hormone signaling pathways plays a crucial role in regulating fruit abscission [54]. ET and ABA are key endogenous signaling molecules in the process of plant organ abscission [45]. Research has indicated that ET and ABA biosynthesis are closely related to nutritional stress responses. Under carbohydrate scarcity, ET and ABA levels in tissues accumulate rapidly, suggesting their function as sugar-starvation sensors and participation in fruit abscission regulation [45]. ET can be generated via the cysteine and methionine metabolic pathway, wherein the key enzymes ACS and ACO both depend on a stable S-Adenosylmethionine (SAM) supply generated by MAT catalysis [55]. During the MA stage, MAT, ACS and ACO transcription was upregulated in fruits, while the ACC content markedly decreased, suggesting the ET biosynthesis pathway activation. Notably, CTR1, a core negative regulator of ET signaling, was significantly upregulated in fruits during the MA stage, consequently repressing expression of the downstream signaling component EIN2 [56]. This expression pattern suggests that the fruit may not be the primary target organ for ET action. ABA is primarily synthesized through the carotenoid pathway, with the levels regulated by key biosynthetic enzymes including ZEP and NCED [57]. In this study, the transcript levels of four ZEP and two NCED genes showed significant upregulation during peak abscission, coinciding with the ABA accumulation. Meanwhile, transcription of the core ABA signaling pathway genes PP2C and SnRK2 [58] was activated at the 15 DAF. ABA accumulation induces the closure of the pericarp stomata, thereby promoting ET diffusion toward the abscission zone, which may represent a primary mechanism underlying ABA- and ET-mediated fruit abscission [54]. Correspondingly, the expression of multiple key genes involved in IAA biosynthesis (e.g., YUCCA, TDC, and amiE; [59]) was downregulated which is inconsistent with the high IAA content in the fruits. Research has indicated that sugar starvation stress may trigger ROS bursts in plant tissues, leading to cellular senescence and death; this process can subsequently impede IAA efflux from organs, resulting in IAA accumulation [45]. Reduced IAA transport to the abscission zones may enhance ET sensitivity, thereby exacerbating fruit abscission [45]. Therefore, the accumulation of ET and ABA coupled with perturbed polar IAA transportation under sugar-deficient conditions triggers the process of fruit abscission, which is similar to that reported in apples [54].

Enhanced sugar availability promotes fruit development through GA-CK signaling-mediated cell division and expansion

Adequate sugar availability is crucial for the growth of parthenocarpic fruits [60]. Research has demonstrated that sugar accumulation promotes parthenocarpic fruit formation by modulating hormone signaling to sustain cell division and expansion [7]. During SD stages, fructose and glucose levels increased. The metabolic changes in ACC, ABA, and IAA exhibited opposite trends compared to the MA stage. This indicates that the improved sugar availability suppresses fruit abscission, which aligns with the observed changes in abscission rates. Meanwhile, both GA₃ and iP levels increased, which are known to promote fruit set and development [61, 62].

CKs are a group of adenine derivatives that play multiple important roles in plant growth, including promoting the development of the cambium and vascular system, stimulating the growth of stems and roots, and inhibiting the senescence of plant organs [7]. CK biosynthesis and degradation are regulated by two key rate-limiting enzymes, IPT and CKX, respectively [63]. During the post-abscission development stages, the fruits upregulated IPT gene expression while downregulating CKX gene expression. This suggests that enhanced CK biosynthesis coupled with reduced degradation may lead to iP accumulation in the fruits. The CK signal transduction constitutes a two-component system, with B-ARRs and A-ARRs playing pivotal roles as transcriptional activators and negative feedback regulators, respectively [64]. Glucose and fructose accumulation alleviated A-ARR suppression in CK signaling while promoting B-ARR expression (3/5). This indicates that the CK response may mediate sugar-induced parthenocarpic fruit formation. GA is a likewise pivotal phytohormone regulating plant reproductive development. Previous studies have demonstrated that GA can promote parthenocarpy and enhance fruit set rates by synergizing with sugar metabolism to ensure the supply of precursor substances and energy during cell wall biosynthesis in fruits [65]. GA biosynthesis is regulated by the key rate-limiting enzyme KAO in plants [66], while bioactive GAs are catalytically inactivated by GA2ox [67]. As glucose and fructose availability increased, KAO upregulation and GA2ox suppression in seedless litchi paralleled the GA₃ content increase. In the GA signaling pathway, GA-GID1 interactions induce DELLA protein degradation, thereby activating downstream transcription factor (TF) expression [68]. GID2 functions as a positive signaling regulator by controlling DELLA protein stability [69]. Improved sugar availability similarly upregulated the expression of core pathway components, including GID1, GID2, TFs and most DELLA members. Meanwhile, cytological observations revealed synchronous increases in both cell size and number. These changes suggest that improved sugar availability during the SD stages likely maintains active cell growth and proliferation capacity through activation of both CK and GA signaling pathways, thereby promoting parthenocarpy in seedless litchi. Similar regulatory mechanisms have also been reported in parthenocarpic cucumber [7] and tomato [8].

Conclusion

Our findings demonstrate that insufficient sugar availability during early development in seedless litchi leads to the accumulation of ET and ABA in fruits. Concurrently, perturbed polar IAA transport enhances the sensitivity of the abscission zone to ET, contributing to massive fruit abscission. This massive fruit abscission promoted fructose and glucose accumulation in the remaining fruits, thereby suppressing the increase in abscission rates during the SD stages. As sugar availability recovers, CK and GA biosynthesis and signaling pathways are activated, sustaining cellular growth competence and proliferation capacity to promote continuous fruit expansion. This study elucidates the molecular mechanisms through which sugar signaling modulates hormonal regulatory networks to determine fruit fate in seedless litchi, providing theoretical insights for achieving high and stable yields in orchard production. It is important to note that the seedless litchi is a parthenocarpic variety. Therefore, the generalizability of this molecular mechanism to seeded varieties requires further investigation.

Abbreviations

MA stage

Massive abscission stage

SD stages

Subsequent developmental stages

DAF

Days after flowering

FW

Fresh weight

IAA

Auxin.

ABA

Abscisic acid

ET

Ethylene

ACC

1-aminocyclopropane-1-carboxylic acid

CK

Cytokinin

iP

N6-(∆2-isopentenyl)-adenine

GA

Gibberellin

FPKM

Fragments per kilobase of transcript per million mapped reads

DEGs

Differentially expressed genes

MAT

S-adenosylmethionine synthetase

ACS

1-aminocyclopropane-1-carboxylate synthase

ACO

Aminocyclopropanecarboxylate oxidase

CTR1

Serine/threonine-protein kinase CTR1

ZEP

Zeaxanthin epoxidase

NCED

9-cis-epoxycarotenoid dioxygenase

PP2C

Protein phosphatase 2 C

SnRK2

Serine/threonine-protein kinase SRK2

SIMKK

Mitogen-activated protein kinase kinase 4/5

EIN2

Ethylene-insensitive protein 2

IPT

Adenylate dimethylallyltransferase

B-ARR

 Two-component response regulator ARR-B family

KAO

 Ent-kaurenoic acid monooxygenase

GID1

 Gibberellin receptor GID1

GID2

 F-box protein GID2

TFs

Phytochrome-interacting factors

DELLA

 DELLA protein

CKX

 Cytokinin dehydrogenase

A-ARR

 Two-component response regulator ARR-A family

GA2ox

 Gibberellin 2beta-dioxygenase

GA3ox

 Gibberellin 3beta-dioxygenase

UGT74B1

 N-hydroxythioamide S-beta-glucosyltransferase

YUCCA

 Indole-3-pyruvate monooxygenase

amiE

 Amidase

AUX/IAA

 Auxin-responsive protein IAA

ARF

 Auxin response factor

Authors’ contributions

K.Z. and Z.G. performed the experiments and wrote the manuscript. X.Z. reviewed and polished the manuscript. S.L., Q.H., H.L., F.L. and H.S. collected experimental materials and assisted with experiments. J.W. revised and proofread the manuscript. All authors have read and approved the final manuscript.

Funding

The research was supported by the Hainan Provincial Natural Science Foundation of China (324MS091), National Key R&D Program (2024YFD1600302; 2023YFD2300805-3).

Data availability

All raw transcriptome sequencing data generated in this study have been deposited in the NCBI Sequence Read Archive (SRA) under the BioProject accession number PRJNA1296845 and are publicly accessible.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.