At team 5 in Jingpai Village, Chengmai County, Hainan Province, five 16-year-old ‘Feizixiao’ litchi trees with uniform growth potential and no clear pest or health issues were selected for the experiment. ‘Feizixiao’ litchi from the garden entered the physiological fruit drop period in early April, the fruit expansion period in late April, and the fruit maturity period in mid-May (Liao et al., 2022). Sampling was performed during the fruit expansion period and fruit ripening period, that is, starting from 35 days after anthesis (DAA) (19, Apr, 2020, 18, Apr, 2021), and five fruit of the same size in the middle and periphery of each tree were selected. The size and coloring condition of the five fruits were used as the reference for each sampling. Highly similar fruit were selected as the materials, and thirty fruits were taken from each tree. The sampling times were 35, 42, 50, 56, 63, and 69 DAA in 2020 and 35, 42, 49, 56, 63, and 70 DAA in 2021. After harvesting the samples, they were placed in liquid nitrogen for quick freezing, and then taken back to the laboratory and stored in an ultralow temperature freezer (-80°C).

The method referenced in Wang et al. (2006) was performed with some modifications. We weighed 0.5 g of pulp in a mortar, heated it in a microwave oven for 30 s, added 5 mL of 90% ethanol, ground the sample thoroughly, centrifuged it at 10,000 g _n for 15 min, aspirated the supernatant and added 5 mL of 90% ethanol for another extraction. Then, we combined the two supernatants, which were evaporated in a 90°C water bath, Then, the remaining sample was fixed with 10 mL of deionized water, and a small amount was aspirated with a syringe and filtered through a 0.45 μm membrane for testing. The sugar contents were determined on a Waters 2695 high-performance liquid chromatograph with an evaporative light scattering detector and a Boston Green Amino Column (4.6×250 mm, 5 μm). The mobile phase ratio was acetonitrile:water = 8:2, the flow rate was 1 mL min^-1, the column temperature was 35°C, and the injection volume was 10 μL. High-purity glucose, fructose and sucrose (Beijing Tanmo Quality Inspection Technology Co., Ltd.) were used as the standards, and high-purity acetonitrile (Sinopharm Chemical Reagent Co., Ltd.) was used as the mobile phase. The sum of the fructose, glucose and sucrose contents was considered the total soluble sugar contents.

The pulp sucrose synthase cleavage (SS-C) and synthesis (SS-S) and AI, NI, and SPS activities were measured by double antibody sandwich-enzyme-linked immunosorbent assay kits (Catalog Nos: KT8013-A, KT50452-A, KT5045-A, KT8107-A, KT5044-A, Jiangsu Kete Biotechnology Co., Ltd., Yancheng, Jiangsu, China) following the manufacturer’s protocol.

Data statistics and graphing were performed with Excel, and data analysis was performed with SAS software, in which the ANOVA program was used for variance analysis and Duncan’s new multiple range test was used for multiple comparisons. The differential gene heatmap and Venn plot were drawn with TBtools software (Chen et al., 2020).

The high-performance liquid chromatograms of the pulp sugar components of the samples harvested at 63 and 69 DAA in 2020 are shown in Supplementary Figures S1A, B . There was no significant change in fructose or glucose content and a significant decrease in sucrose content during the “sugar receding” period. The high-performance liquid chromatograms of the sugar content in the pulp at 63 and 70 DAA in 2021 are shown in Supplementary Figures S1C, D and the changes in the sugar content during the “sugar receding” period are consistent with those in 2020. The results from both years indicated that there was no significant change in reducing sugar content and a significant decrease in sucrose content in the “sugar receding” period.

The dynamic changes in fructose, glucose and sucrose contents in litchi pulp over the two years are shown in Figures 1A, B . In 2020, the fructose content increased significantly from 35 to 50 DAA and became stable after 56 DAA. In 2021, the fructose content increased significantly from 35 to 56 DAA and then stabilized. The dynamics of fructose contents for both years were basically the same, indicating that pulp fructose continues to accumulate and increase with the expansion of the fruit and then stabilizes during maturity. Therefore, fructose is not the main component causing the “sugar receding” phenomenon during maturity.

In 2020, the glucose content increased significantly from 35 to 50 DAA and remained stable from 50 to 69 d with no significant difference between any two stages. In 2021, the glucose content increased significantly from 35 to 42 DAA and then stabilized. The dynamics of glucose contents for the two years were basically the same, indicating that glucose accumulates rapidly during the rapid expansion period of the fruit and tends to become stable during maturity. Therefore, glucose is not the main component causing the “sugar receding” phenomenon during maturity.

The sucrose content trends in both years were consistent, with a large amount of accumulation in the fruit expansion stage and a large amount of cleavage in the late stage of fruit ripening, indicating that sucrose is the main component causing the “sugar receding” phenomenon during maturity.

The dynamic changes in soluble sugar content in pulp showed the same trend. The soluble sugar contents all increased significantly in the early stage of fruit expansion, tended to become stable in the late stage of fruit expansion, peaked at 63 DAA, and then began to decline; then, the phenomenon of “sugar receding” occurred.

In summary, the content of reducing sugars continued to increase with fruit growth and development; the dynamic trends of sucrose and soluble sugar are highly similar, and their linear correlation is significant for two years (r₂₀₂₀ = 0.8611, r₂₀₂₁ = 0.8818), which indicates that the pulp of ‘Feizixiao’ litchi mainly accumulates reducing sugars, and the “sugar receding” phenomenon in the pulp is caused by the decrease in the sucrose content in the pulp.

The dynamic changes in AI activity are shown in Figures 2A, B . In 2020, the activity of AI increased significantly from 35 to 50 DAA, decreased significantly at 56 DAA, and then decreased significantly after reaching the maximum value at 63 DAA. In 2021, there was a rising-falling-rising-falling trend, the difference was significant between any two stages from 35 to 63 DAA, and there was no significant change after 63 DAA. The dynamic changes in AI activity in the two years were different, which may have been related to the different temperatures, amounts of light and other factors that occurred within the two years. However, in general, the overall activity of AI was low for both years, possibly indicating that AI had less of an effect on sucrose metabolism in ‘Feizixiao’ litchi pulp.

The dynamic changes in NI activity are shown in Figures 2C, D . In 2020, the activity of NI decreased, increased, and then decreased again, and the difference in activity between any two stages was significant. In 2021, NI activity increased and decreased twice, showing an “M”-type change trend, and reached maximum values at 35 d and 56 DAA, with a significant decline after 56 DAA. NI activity in both years showed a significant downward trend after 56 DAA.

The dynamic changes in SS-C activity are shown in Figures 2E, F . In 2020, the activity of SS-C showed an “M”-type change trend, reaching two peaks at 42 and 63 DAA. There was no significant difference between them, and there were significant differences in the rest of the stages, with significant declines from 42 to 56 DAA and 63 to 69 DAA. In 2021, there was also an “M”-type change trend, reaching two peaks at 42 and 52 DAA, with significant declines from 42 to 49 DAA and 56 to 70 DAA. The activity of SS-C decreased significantly in the “sugar receding” period in both years, but the activity was still higher than the maximum INV activity, which may indicate that the cleavage of sucrose is mainly catalyzed by SS.

The dynamic changes in SS-S activity are shown in Figures 2G, H . In 2020, the activity of SS-S showed a dynamic trend of inclines and declines, and the difference in activity between any two stages was significant. The enzyme activity decreased and then increased significantly from 35 to 49 DAA, reached the maximum value at 63 DAA and was significantly higher than that at the other three stages after 49 DAA. The dynamic trend of SS-S activity in the two years was basically the same, and SS-S activity in the “sugar receding” period decreased significantly.

The dynamic changes in SPS activity are shown in Figures 2I, J . In 2020, SPS activity showed a trend of a significant increase and then a significant decrease from 35 to 50 DAA. In 2021, it showed a trend of a significant decline and then a significant increase from 35 to 49 DAA. In the middle and late stages of fruit growth and development, there was a significant decrease, increase and then decrease trend in both years, and it decreased significantly in the “sugar receding” period.

The dynamic changes in the net activity of sucrose metabolism enzymes (difference between the cleavage and synthesis activities of sucrose metabolism enzymes) are shown in Figures 2K, L . In 2020, there was a unimodal trend, reaching a maximum value at 50 DAA, decreasing significantly at 56 DAA, and continuing to gradually decrease. In 2021, there was a rising-falling-rising-falling-rising trend with significant differences between any two stages. The net activity of sucrose metabolism enzymes for both years remained at a high level in the “sugar receding” period, which may have caused the significant decrease in sucrose content in the “sugar receding” period.

In summary, the dynamic changes in sucrose metabolism enzyme activities were similar for both years. In the “sugar receding” period, except for the activity of AI, which did not change significantly in 2021, the activities of the other enzymes decreased significantly. During the growth and development of ‘Feizixiao’ litchi pulp, the activity of SS-C was the highest among sucrose-cleaving enzymes, followed by NI, and that of AI was the lowest, which may indicate that SS plays a major role in cleaving sucrose. The activity of SS-C was consistently higher than that of SS-S, which may indicate that SS mainly cleaves sucrose.

As shown in Table 1 , the transcriptome sequencing data were filtered to obtain an average of 43,676,652 clean reads and 6.55 Gb of clean base. The proportion of clean reads in each sample was greater than 96%, Q20 was above 98%, Q30 was above 94%, and the GC content was higher than 45%, indicating that the sequencing quality was favorable for subsequent analyses.

As shown in Figures 3A–D , 18,061 genes were differentially expressed (8783 genes were downregulated, and 9278 genes were upregulated) in 35 d vs. 63 d; 19,575 genes were differentially expressed (9762 genes were downregulated, and 9813 genes were upregulated) in 35 d vs. 69 d; and 985 genes were differentially expressed (712 genes were downregulated, and 273 genes were upregulated) in 63 d vs. 69 d. Figure 3E shows that 254 differentially expressed genes were common to all three groups, and 263 differentially expressed genes were unique to 63 d vs. 69 d. In 35 d vs. 69 d, the number of differentially expressed genes was the highest, and the number of differentially expressed genes was the lowest in 63 d vs. 69 d. The number of differentially expressed genes between the two groups was significantly different, indicating that some genes are differentially expressed in the various fruit growth and development stages and that the number of differentially expressed genes increases as fruit growth and development progress.

GO enrichment analysis was performed on the differentially expressed genes in 35 d vs. 63 d, 63 d vs. 69 d, and 35 d vs. 69 d, and the 50 GO terms with the lowest q value were selected for representation. The results are shown in Figures 4A–C . Differentially expressed genes in all groups were divided into 3 major categories: biological process, cellular component and molecular function. The GO enrichment analysis results of differentially expressed genes in the three groups showed partial differences. Notably, seven PFK genes were enriched in the biological process and molecular function categories in 63 d vs. 69 d. It was speculated that the PFK genes influenced the changes in sugar content during the “sugar receding” period.

KEGG enrichment analysis was performed on the differentially expressed genes in 35 d vs. 63 d, 63 d vs. 69 d, and 35 d vs. 69 d, and the 20 pathways with the most significant enrichment were selected for representation with a bubble chart. The results are shown in Figures 5A–C . The plant hormone signaling pathway was enriched in all groups. Fructose and mannose metabolism, glycolysis/gluconeogenesis, and pentose phosphate pathways were found in 63 d vs. 69 d, and it was speculated that the differentially expressed genes of these pathways may have an impact on “sugar receding”.

To further analyze the differentially expressed genes related to the “sugar receding” phenomenon in ‘Feizixiao’ litchi pulp, a total of 91 genes related to sucrose metabolism, EMP and PPP were screened from the transcriptome data. No FK, SWEET and sucrose transporter (SUT) genes were screened due to low expression and no significant change in the “sugar receding” period. The log₂(fold change) was calculated to plot the heatmap shown in Figure 6 . In 63 d vs. 69 d, four SS genes were screened out and all were downregulated; three AI genes were screened out, of which two were downregulated; two NI genes were screened out, of which one was downregulated; three SPS genes were screened out, of which two were downregulated; seven HK genes were screened out, of which five were upregulated; Twenty-five PFK genes were screened out, of which twenty-three were upregulated; twenty PK genes were screened out, of which twelve were upregulated; one G6PD gene was screened out and it was upregulated; and three 6PGD genes were screened out, of which one was upregulated. The results showed that during the “sugar receding” period, seventy-five percent of the sucrose metabolism enzyme genes were downregulated. In EMP, seventy-seven percent of the genes were upregulated. In PPP, two genes were upregulated, and two genes were downregulated.

In summary, most of the sucrose metabolism enzyme genes were downregulated during the “sugar receding” period, most of the key genes in EMP were upregulated, and most of them were related to PFK genes. There was no significant difference in the key genes in PPP, indicating that reducing sugars may be mainly consumed through EMP and that PFK is a key rate-limiting enzyme in EMP in ‘Feizixiao’ litchi pulp.

Figures 7A, B show that there was a significant linear correlation between the expression of 15 genes detected by real-time PCR and the transcriptome sequencing results (r₂₀₂₀ = 0.9139, r₂₀₂₁ = 0.8912), which proved the reliability of the transcriptome sequencing results.

Sugar compositions and contents are key factors affecting the intrinsic quality of fruit (Borsani et al., 2009), and sucrose metabolism enzymes play an important role in sugar accumulation in fruit (Sun et al., 2020). In this study, contents of the reducing sugars in the pulp increased continuously during the growth and development process, but the content of sucrose decreased significantly at the ripening stage, resulting in an overall decrease in the total soluble sugar content ( Figures 1A, B ). Some studies have shown that AI plays an important role in sugar accumulation (Ruan, 2014). However, the activity of AI was relatively low in this study, which may suggest that AI does not play an important role in the “sugar receding” phenomenon ( Figures 2A, B ). Moreover, the activity of SS-C was consistently greater than that of SS-S ( Figures 2E–H ), and the same result was found in a study of sucrose accumulation in peaches (Vizzotto et al., 1996), which may indicate that SS plays an important role in sucrose cleavage. The activity of SS ( Figures 2E–H ) in the middle and late stages of fruit growth and development is consistent with the changes in sucrose content ( Figures 1A, B ), a previous study in citrus (Citrus unshiu Marc.) showed that the change in SS activity at the ripening stage was consistent with the change in total sugar content (Komatsu et al., 2002). It was speculated that SS played a greater role in the middle and late stages of fruit growth and development. The comprehensive effect of sucrose metabolism enzymes, that is, the net activity of sucrose metabolism enzymes, was an important factor affecting the sugar content of ‘Feizixiao’ litchi pulp. The reducing sugar content increased, while the activities of enzymes that cleave sucrose decreased during the “sugar receding” period ( Figures 2A, C–F ), this observation may have been related to the simultaneous decrease in the activities of enzymes that synthesize sucrose ( Figures 2G–J ), so that the net activity showed a strong sucrose cleaving activity during the “sugar receding” period ( Figures 2K, L ), resulting in a significant decrease in sucrose content ( Figures 1A, B ). Moreover, a previous study showed that sugar accumulation is related to photosynthesis (Wang et al., 2022), and the products of photosynthesis are mainly transported to the fruit in the form of sucrose. The “sugar receding” phenomenon is mainly caused by the decrease in sucrose content, and it is speculated that the photosynthetic rate of leaves may decrease during the “sugar receding” period. Therefore, to study the causes of the “sugar receding” phenomenon in ‘Feizixiao’ litchi, we should study not only the changes in sucrose metabolism enzyme activities and respiratory metabolism of pulp but also need to further study photosynthesis in leaves.

Figures 1A, B showed that the content of sucrose in ‘Feizixiao’ litchi pulp decreased significantly, but the content of reducing sugars did not increase significantly, indicating that the downstream pathway of sugar metabolism consumes reducing sugars. The pulp of ‘Shixia’ longan (Dimocarpus longan Lour), belonging to the Sapindaceae family, also had the “sugar receding” phenomenon, and the research results showed that upregulated EMP, tricarboxylic acid cycle, fermentation and energy metabolism promoted “sugar receding” in the pulp (Luo et al., 2021). In a study of potato (Solanum tuberosum L.) tubers, it was found that low temperature reduced the activity of PFK, resulting in the inhibition of EMP, the accumulation of hexose phosphates, and ultimately an increase in sugar content (Hammond et al., 1990), indicating that PFK has a significant effect on the metabolic rate and sugar accumulation of the pulp via EMP. Similar results were also found in this study. The GO enrichment analysis of the differentially expressed genes in the “sugar receding” period demonstrated that PFK genes were significantly enriched in the biological process and molecular function categories ( Figure 4B ). Additionally, the KEGG enrichment analysis found that glycolysis/gluconeogenesis and pentose phosphate were significantly enriched ( Figure 5B ). A total of 91 differentially expressed genes with high expression abundance, including sucrose metabolism genes and key genes in EMP and PPP, were screened from the transcriptome sequencing data ( Figure 6 ). Fifteen of these genes were selected for real-time PCR verification ( Figures 7A, B ), proving the reliability of the transcriptome sequencing results. Figure 6 showed that the expression of SS and SPS genes was downregulated during the “sugar receding” period, which was consistent with the change in enzyme activity ( Figures 2E-J ); Seventy-seven percent of the genes associated with EMP were upregulated, while PFK genes were expressed in high abundance, and the expression of key genes in PPP showed no significant changes ( Figure 6 ), indicating that reducing sugars mainly entered EMP and are consumed in large quantities and that PFK is the most critical rate-limiting enzyme in the glycolysis pathway in ‘Feizixiao’ litchi pulp. Since no sucrose transporter genes was screened out from transcriptome data, it was speculated that ‘sugar receding’ is mainly caused by the enzymes related to sucrose metabolism and EMP.