Effects of shading treatment on fruit quality
In the tomato fruit, Shading treatment inhibited coloration compared to the control, which was supported by carotene, lycopene and Xanthophyll contents (Fig. 1A, B C). However, Shading treatment increased the chlorophyll content (Fig. 1D) and the total acid content (Fig. 1H). The treatment also affected the fruit firmness which is reflected in the slowing of the fruit softening (Fig. 1G). All of the above indicated that shading treatment retarded the tomato fruit ripening process. Similar to tomato, shading treatment reduced the contents of anthocyanins (Fig. 2A) and UDP-glucose: flavonoid 3-O-glucosyltransferase (UFGT) (Fig. 2G), which resulted in suppressed coloring of grape fruits. At the same time, shading treatment also increased chlorophyll content (Fig. 2B) and the total acid (Fig. 2C).
Common differences were found in sugar metabolism at tomato and grape fruits, shading treatment increased the starch content compared to the control fruit. The difference of sucrose content in fruits was the result of the synergistic effect of various enzymes in sucrose metabolism. The contents of acid invertase (AI) (Fig. 1K, 2J), neutral invertase (NI) (Fig. 1L, 2K) and sucrose phosphate synthase (SPS) (Fig. 1I, 2H) decreased, while the content of sucrose synthase (SS) (Fig. 1J, 2K) increased. These changes inhibit the synthesis and transformation of sucrose (Fig. 1G), thus delaying the accumulation of fructose and glucose (Fig. 2E, F). In summary, shading delayed the ripening process of different fruit types, tomato and grape may respond to darkness stress through some similar pathways.
Transcriptome profiling in response to darkness
Transcriptome sequencing of tomato fruit (TW) and control group (TR), grape pulp (PulpD) and control group (PulpL), and grape skin (SkinD) and control group (SkinL) were performed after dark treatment. The three replicates for RNA-seq in TW group yielded 6.59, 6.66, and 5.98G clean reads after quality control and raw read filtering per sample and the TR group obtained 6.55, 6.19, 6.56 G clean reads, respectively. The Q30 of all samples were greater than 94%, and the lowest reads sequence is 67.85% which fits for comparison with reference genome. In addition, we performed transcriptional profiling of the pulp and skin of grapes from treatments described above, three replicates per library in four cDNA libraries were constructed. After removing impurities and linker sequence, approximately 6.4 G-7.2 G clean reads of these libraries were obtained. Most of the reads (> 93.5%) were highly matched with grape genome sequences indicating that the clean reads were good enough for subsequent analysis.
DEGS analysis
All of the treatments were divided into three comparison groups named TW vs TR, SkinD vs SkinL, PulpD vs PulpL. The construction of cluster analysis clearly showed the expression content of DEGs in different samples (Fig. 3A), and the quantitative relationship of DEGs in different control groups was statistically analyzed by histogram (Fig. 3B). In addition, we compared the relationship between the number of up-regulated and down-regulated gene DEGs in the treatment group and the control group by drawing volcanic maps (Fig. 3C).
The GO and KEGG analysis of DEGs
As the largest comprehensive database in the world, the Gene Ontology (GO) knowledge base is widely used for annotation analysis of gene or protein function. We used the Gene Ontology (GO) knowledge base to predict and analyze the function of gene expression in the dark processing group, and selected the most significant 30 notes for DEGS rich in biological processes, cellular components, and molecular functions (Fig. 4).
In the TW vs TR group, GO analysis indicated that Shading treatment affected multiple metabolic pathways including 10 cellular components GO terms,10 biological process terms, and 10 molecular function terms.
In the SkinD vs SkinL group, GO analysis indicated that shading treatment affected multiple metabolic pathways including 13 cellular components GO terms, with cytoplasm being the most enriched terms; 9 biological process terms, with organonitrogen compound biosynthetic process being the most enriched; and 9 molecular function terms, with transporter activity being the most enrich.
In the PulpD vs PulpL group, GO analysis indicated that Shading treatment affected multiple metabolic pathways including 10 cellular components GO terms, with membrane protein complex being the most enriched terms. Then, we used KEGG pathway data base to examine the DEGs-associated pathways, KEGG pathway enrichment analysis showed that shading treatment was involved in photosynthesis and the biosynthesis of secondary metabolites (Fig. 5). These terms are more important to fruit quality.
Metabolome analysis
In order to study the metabolic response of different climacteric plants to dark stress, the metabolites of the treatment and the control group were compared and analyzed. Between TW-TR groups, 538 metabolites were identified across broad chemical classes from the spectra after quality validation when 201 metabolites were significantly upregulated and 300 were significantly down-regulated. And between SkinD vs SkinL, 298 metabolites were identified across broad chemical classes from the spectra when 161 metabolites were significantly up-regulated and 137 were significantly down-regulated. In PulpD vs PulpL group, 180 metabolites were found, 106 of which were up-regulated and 74 down-regulated. We conducted principal component analysis (PCA) of the metabolic differences between samples within each group and the differences between samples within the group (Fig. 6). P-value is the symbol of the result of significance test and p-value < 0.05 was defined as differentially expressed metabolites. The biological repetition of the three groups (TW-TR, SkinD-SkinL and Pulp-Pulp) met the requirements when there were significant differences between the treatment and the control group. Meanwhile, it was found that the expression of metabolites was mostly down-regulated in tomato, while the expression was mostly up-regulated in grape. The effects of dark stress on fruits of different climacteric were different.
Correlation analysis between transcriptome and metabolome
The KEGG pathway enrichment of the metabolome and DEGs in different climacteric fruits were integrated to evaluate the complementary analysis of transcriptomic and metabolomic data (Fig. 7). The results showed that the metabolic pathways related to Photosynthesis, hormone signal transduction, Starch and sucrose metabolism were more abundant in different species. At the same time, there are abundant pathways related to fruit coloring, such as Carotenoid biosynthesis and flavonoid biosynthesis-related pathways.
Shading treatment affects chlorophyll metabolism pathway
A significant enrichment degree of commonality in all comparison combinations has been observed in “Porphyrin and chlorophyll metabolism” according to KEGG pathway enrichment results. When berries underwent shading treatment, a series of enzyme activities all reduced related to chlorophyll biosynthesis in the tomato and grape. such as delta-aminolevulinic acid dehydratase (HEMB), uroporphyrinogen-III synthase (HEMC), uroporphyrinogen decarboxylase (HEME), coproporphyrinogen-III oxidase 1 (HEMF), protoporphyrinogen oxidase (HEMY), ChlH (encoding magnesium-chelatase subunit H), ChlI (ending magnesium-chelatase subunit I), CHLD (encoding magnesium chelatase subunits D) and protochlorophyllide reductase (POR2) were highly inhibited, therefore affecting the biosynthesis of chlorophyll. Comparative transcriptome analysis of the two contrasting species allows us to identify some candidate genes involved in chlorophyll degradation for example senescence -inducible chloroplast stay-green protein (SGR1), encoding protein stay-green 1 gene and ending red chlorophyll catabolite reductase gene transcript are inhibited in tomato and grape respectively (Table S1). Chlorophyll biosynthesis is a complex regulatory process, which is not only affected by external environmental factors, but also the enzymes involved in chlorophyll biosynthesis are regulated by genes(Nagata et al., 2005). Shading treatment inhibited the expression of several chlorophyll biosynthesis genes and the differential expression of genes encoding chlorophyll degrading enzymes in the two contrasting species, revealed that tomato and grape berries experience similar physiological mechanisms and parallel programs of cytological change following abiotic stress. Finally, it reflects the increase of chlorophyll content. The influence of light intensity on chlorophyll content showed a distinct performance in various types of plants. Previous studies have pointed out that in dark conditions, in case to ensure the normal growth of plants and promote net photosynthesis, Lilium auratum would capture light energy by synthesizing large amounts of chlorophyll (Zhang et al., 2015). Evidence indicates similar regulatory mechanisms for chlorophyll within both ripening physiologies, suggesting common regulators of climacteric and non-climacteric ripening physiology.
Shading treatment affects Sugar metabolism and signal transduction
In TW vs TR, RMATS software was used to compare the results for AS classification and differential AS analysis. In total, 4,026 AS events were identified in the transcriptome datasets including 223 Mutually exclusive exon (MXE) events and 3,803 Skipped exon (SE) events. Among them, 2 MXE events and 8 SE events were significantly enriched. In the MXE event, a sugar transporter (SUT,Solyc08g048290.3) was found (Table S2). Data suggests that INT1-transported myo-inosito not only plays a crucial role in the regulation of cell elongation in a sucrose dependent manner, but also participates in the regulation of sucrose and other carbohydrate intermediates as a metabolic signal (Strobl et al., 2018). In Arabidopsis, SUT belongs to the monosaccharide transporter (MST) -like gene family. The MST(-like) family includes 53 genes with significant homology to known monosaccharide transporters (Buttner, 2007).
Sugar, as photosynthetic products, plays an important role in nutrition supplement and signal molecules throughout the life cycle of higher plants (Zhang et al., 2020a). MSTs are important sugar transporters that were functioned in carbohydrate flux for maintaining the balance between source and sink tissues(Zheng et al., 2014). The number and proportion of different subtypes of SUT were significantly different from those before shading treatment, and the expression of SUT was decreased. In view of that plant sugar transporters play an important regulatory role in the transport balance of sucrose which was one of the important photosynthetic products. We speculate that AS of SlSUT after dark treatment, inhibited the transport and accumulation of sugar, which affected the fruit quality.
As previously reported, the accumulation and transport of sugar in grape berries has been well understood. Carbohydrates produced during leaf photosynthesis are transported as sucrose to the phloem, where they will be transported to the berry cluster. In fruits ripeness and senescence course, a series of sugar transporters guide the transport of sugars through different organelles (Conde et al., 2006, Swanson and El-Shishiny). Following berries ripening, sugars accumulate in vacuoles in the form of glucose and fructose and the polysaccharides are broken down by enzymes. Then these sugars will be transported through different organelles in the direct of monosaccharide. The three encoding sucrose phosphate synthase (VIT_218s0089g00410, VIT_211s0118g00200, VIT_205s0029g01140) and the gene encoding sucrose phosphatase (VIT_208s0032g00840) were significantly down-regulated in shading fruits (Table S3). At the same time, a decrease in sucrose content was detected in the metabolites (Table S4). In grape berries, sucrose is either degraded by sucrose synthase into uridine 5 '-diphosphate (UDP) glucose and fructose or cleaved by invertase into glucose and fructose for subsequent metabolism and biosynthesis (Stitt, 1993, Koch and K., 1996).
Plant hormone biosynthesis
In climacteric fruits like tomato, ripening and senescence are generally controlled by ethylene. While in the ripening process of grape, a non-climacteric fruit, may be facilitated by the accumulation of ABA. However, numerous studies implicating ABA has a hand in in the ripening of climacteric fruits (Chernys and Zeevaart, 2000, Leclercq et al., 2002). The cleavage of 9-cisviolaxanthin to xanthoxin (C15) and C25-epoxy-apo-aldehyde is necessary for ABA biosynthesis, which is catalyzed by 9-cis-epoxycarotenoid dioxygenase (NCED). When grape was shading stressed, a rise in ABA biosynthesis via expression of NCED (VIT_219s0093g00550, VIT_210s0003g03750), and strong ABA degradation correlated with a higher decrease of abscisic acid 8'-hydroxylase 4 (Hyd1)(VIT_218s0001g10500). However, in tomato, it was observed that ABA biosynthesis and degradation correlated with up-regulation of NCED1 (solyc07g056570.1) and abscisic acid 8'-hydroxylase (CYP707A1) (Solyc08g005610.3, Solyc04g078900.3) separately. Furthermore, the metabolomics data showed that accumulation of ABA was repressed and induced respectively in tomato and grape processing group. During maturation, environmental factors such as dark stress accelerated the induced expression of these genes to different levels., under illumination deficit. This characteristic pattern of expression is shared with many Chlorophyll metabolism pathway genes. These observations regarding correlation of expression present a common regulatory mechanism in both climacteric (tomato) and the non-climacteric (grape) throughout the ripening process.
In the GA-biosynthesis pathway, the activity of GA20- oxidase is considered to be one of the main regulatory points. It has been proved in spinach, Arabidopsis and potato that the expression of GA 20-oxidase was regulated by photoperiod with significantly higher levels of GA 20-oxidase in long-compared with short-day conditions (Wu et al., 1996, Xu et al., 1995, Carrera et al., 1999). The GA biosynthetic key enzyme GA20ox1 in StSUT4-RNAi plants was reduced indicating that SUT4 can promote GA biosynthesis(Kuehn, 2011). Ethylene stimulates phytochrome-mediated shade avoidance responses to shade treatment by enhancing GA action (Pierik et al., 2004).
In higher plants, methionine is a biosynthetic precursor of ethylene, 1-aminocyclopropane-1-carboxylic Acid Synthases (ACS) process S-adenosine l-methionine (SAM) to 1-aminocypropane-1-carboxylic Acid (ACC), Finally, ACC is oxidized to produce ethylene. (Yang, 1979). The two most important steps in this biosynthesis process are the catalytic processes of ACS and ACO.(Yang and Hoffman, 2003). The transcripts encoding ACS6 (Solyc08g008087.1), ACS2 (Solyc01g095080.3), ACO5 (Solyc07g026650.3) and ACO4 (Solyc07g049550.3) were promoted under shading stress. At the same time, in grape, a gene encoding 1-aminocyclopane-1-carboxylate synthase (VIT_202s0025g00360) and ACO1 (VIT_200s2086g00010) was inhibited (Table S5).
Effects of light deficit on carotenoid metabolism in tomato berries
At the transcriptional level, change of genes that associated with the carotenoid biosynthesis were firstly observed. The prominent gene enriched in up-regulated treatment-responsive transcripts was Lycopene cyclase protein (Lcy) (Solyc12g008980.2). Suppression of LCYe increased enhanced the accumulation of β-carotene both in transgenic and sweet potato (Diretto et al., 2006, Kim et al., 2013). In the pericarp of SlLCYe-RNAi fruit, transcript level of SlPSY1 was elevated, indicating that there was a negative regulatory feedback loop (Wang et al., 2020). In addition, the unigene of Squalene/phytoene synthase (PSY) (Solyc03g031860.3) was up-regulated and flavin containing amine oxidoreductase (Solyc03g123760.3) known as 15-cis-phytoene desaturase (PDS) was down-regulated (Table S6). The results showed that the content of lycopene and total carotenoids was positively correlated with the expression of PSY and PDS during the ripening of tomato fruits (Fraser et al., 2002). At the same time, the increase of geranylgeranyl pyrophosphate (GGPP) which known as precursor of carotenoid content was detected in the metabolites.
The up-regulation of PSY increased the synthesis of phytoene, while the decrease of PDS was the rate-limiting step of lycopene biosynthesis. LCYe was a key regulatory step for lycopene cyclization to β-carotene. The downregulation of LCYe gene expression inhibited the biological process of lycopene conversion into beta-carotene, which directly affected the ripening and color transformation of tomato fruits.
Effects of light deficit on flavonoid metabolism in grape berries
The GO and KEGG analysis results indicated that DEGs involved in flavonoid biosynthesis-related pathway consist of anthocyanins and Flavone difference between SkinD and SkinL. Therefore, we further studied DEGs participate in flavonoid biosynthesis in detail and established a predicted flavonoid biosynthesis pathway (Fig. 8). Nine DEGs were identified in flavonoid biosynthesis pathway. Interestingly, most them were down-regulated in SkinD vs SkinL, except PAL (VIT_216s0039g01110). These down regulated genes include, structural genes 4CL (VIT_216s0039g02040), CHS (VIT_214s0068g00920, VIT_205s0136g00260), CHI (VIT_213s0067g03820, VIT_213s0067g02870), and F3'H (VIT_217s0000g07200, VIT_217s0000g07210), F3H (VIT_204s0023g03370); DFR (VIT_218s0001g12800) and LDOX (VIT_202s0025g04720). Simultaneously, two metabolites (Laricitrin, Syringetin) showed significant down-regulated expression differences between SkinD and SkinL (Table S4,7). In our study, flavonols content was positively correlated with the transcripts of genes involved in the flavonoid biosynthesis pathway and their corresponding enzymatic activities. The down-regulation of these key genes, including the downstream genes for anthocyanin synthesis (DFR, UFGT), affects proanthocyanidin biosynthesis (cyanidin, pelargonidin, delpinidin), which largely explains the low accumulation of anthocyanins in SkinD.
Shading treatment affects the pathway of α-linolenic acid metabolism in grape berries
Compared with the control group, lipoxygenase (LOXO; VIT_209s0002g01080 LOXA; VIT_206s0004g01510) was significantly upregulated in berries of Shading treatment group. Whereas, fatty acid hydroperoxide lyase (HPL1; VIT_212s0059g01060) were dramatically reduced in Shading treatment group. The expression levels of alcohol dehydrogenase (ADH; VIT_204s0044g01110, VIT_218s0001g15450) were significantly higher than the control group (Fig. 9, Table S8).
DEGs for RNA-seq results verifified by qRT-PCR analysis
To verify the accuracy and repeatability of RNA-Seq data in this study, we analyzed the expression levels of related genes in tomato and grape berries by qPCR after dark treatment.We selected 13 DEGS as qPCR genes, which are involved in different biological pathways. These include ‘flavonoid metabolism’, ‘carotenoid metabolism’, ‘Plant Hormone biosynthesis’, ‘Sugar metabolism’, ‘Chlorophyll metabolism’. According to the qPCR results, we found that the expression levels of most genes were similar to the transcriptome data results (Fig. 10).