Tomato fruit cracking key regulatory gene identification method and application

Abstract

The invention discloses a tomato fruit cracking key regulation gene identification method, through the method, 41 key genes for regulating tomato cracking are determined and obtained for the first time, and the genes mainly participate in plant hormone synthesis, cell wall metabolism, wax and cuticle metabolism, absorption and transportation of mineral elements and water, and some transcription factors. Two regulation networks are established and comprise an ethylene-cell wall metabolic pathway and a transcription factor-water channel-mineral element. The research combines high-throughput sequencing and molecular biological techniques, enriches the cracking mechanism of tomato fruits on transcription and translation levels, and thus, has important scientific significance and application prospects.

Description

Technical Field

[0001] The invention belongs to the field of biotechnology, and in particular relates to a method for identifying key regulatory genes for tomato fruit cracking and an application thereof. Background Art

[0002] Tomato (Solanum lycopersicum), one of the most widely grown vegetable crops in the world, faces the problem of fruit cracking before and after harvest. Fruit cracking is a common physiological disease that occurs in various fruits, including tomatoes, sweet cherries, grapes, watermelons, blueberries, and nectarines. Fruit cracking affects the appearance and quality of the fruit and reduces its commercial value, resulting in huge economic losses.

[0003] In recent years, transcriptomics and other omics have been widely used in plant science, especially in fruit cracking. This type of research has made great progress in the study of the molecular mechanism of fruit cracking. Transcriptome sequencing of tomato crack-resistant and susceptible genotypes revealed 61 differentially expressed genes related to calcium, cell wall, cuticle structure, hormone metabolism, starch / sucrose metabolism, transcription factors and water transport. In the subtropical fruit litchi, seven co-expression module genes (1733 candidate genes) were found to be related to fruit cracking through transcriptome sequencing and weighted gene co-expression network analysis. Among them, genes with high expression levels, especially plant hormones (growth hormone and abscisic acid), starch / sucrose metabolism and sugar / water transport (Lc.0.190, Lc.0.938 and Lc.0.806, etc.) may increase the pressure of aril expansion, thereby preventing fruit cracking. Meanwhile, low expression level genes, especially phytohormones (auxin, gibberellin, and ethylene), transcription factors (WRKY, calcium transport and signaling), and wax synthesis genes (Lc.12.1668, Lc.7.1507, and Lc.8.658), may reduce the mechanical strength of the cracked litchi peel, which may lead to litchi cracking. Transcriptome studies of watermelon genotypes with different cracking levels showed that the expression of xylin transferase genes such as Cla002042 and Cla010096 was significantly different, which may lead to the occurrence of watermelon fruit cracking. In addition, transcriptome sequencing of interspecific cracking resistant (CR) and cracking prone (CS) genotypes was performed in plants such as grapes and tomatoes to identify genes associated with fruit cracking. Transcriptome sequencing data analysis of the grape cracking prone variety "Xiangfei" showed that genes related to cell wall metabolism and cuticle biosynthesis played an important role in regulating berry cracking. The discovery of genes related to fruit cracking in tomato is of great significance. Summary of the invention

[0004] The present invention provides a tomato fruit cracking regulatory gene and application thereof in fruit cracking prevention and control, so as to reduce the probability of tomato fruit cracking.

[0005] The present invention provides a method for screening and identifying key regulatory genes of tomato fruit cracking, wherein the method is based on Riob-seq and RNA-seq methods for screening and identifying, and specifically comprises the following steps:

[0006] 1) Ribosome imprint sequencing: tissue samples were frozen in liquid nitrogen, ground and added to lysis buffer, mixed and centrifuged, and then the supernatant was added to RNase I and DNase I to remove DNA and RNA, ribosomes were separated using MicroSpin S-400 columns, RFs were isolated, rRNA was removed, and magnetic beads were used to further purify RFs; these RNA fragments protected by ribosomes and extracted for sequencing are also called ribosome footprints (RFs); sequencing was performed on the Illumina HiSeq 4000 sequencing platform, and the quality of sequencing data reached Q20-Q30 for subsequent analysis; the number of RFs located around the start codon and stop codon of the CDS of the coding gene was calculated based on the comparative position of the 5' end of the RFs in the genome; the RFs compared with the CDS region were divided into three categories with a codon number of 1 to 3 bases according to the codon position corresponding to the 5' end of the RFs comparison position;

[0007] 2) RNA-seq data processing and analysis: Quality control of raw reads for downstream machines was performed using fastp to filter low-quality data and obtain clean reads; the steps for read filtering were as follows:

[0008] (1) Delete reads containing adapters;

[0009] (2) remove reads containing N ratio greater than 10%;

[0010] (3) delete all reads with the same base number as A;

[0011] (4) remove low-quality reads (bases with a quality value of Q≤20);

[0012] According to the HISAT2 comparison results, transcripts were reconstructed using String Tie, and the expression of all genes in each sample was calculated; expression values ​​were expressed as fragments per kilobase per million (FPKM); the data were then analyzed using DESeq2 software; based on the analysis results, genes with FDR < 0.05 and |log2FC| > 1 were screened as significant DEGs (differentially expressed genes); DEGs were further subjected to GO and KEGG enrichment analysis using ClusterProfiler89;

[0013] 3) Ribo-seq data processing and analysis: Before bioinformatics analysis, off-plate data were quality controlled and filtered; the raw reads obtained after initial filtering, after which reads containing connectors, all A bases, and more than 10% N and low quality (Q20 ≤ 50%) were removed;

[0014] Ribosomal RNA removal: To eliminate the impact of ribosome contamination on subsequent analysis, the short reads matching tool bowtie2 (2.2.8) was used to match the Clean Reads with the ribosome database and delete the reads that did not match the ribosome. The retained data were used for subsequent analysis.

[0015] Remove transfer RNA, etc.: Compare the reads after ribosomal RNA removal with the GenBank and Rfam databases and remove small nucleolar RNA (snoRNA), small nuclear RNA (snRNA) and micro RNA (miRNA): After removing rRNA and tRNA, snoRNA and snRNA are discovered and removed from the sample by blasting the GenBank and Rfam databases; miRNA sequences identified in miRBase are also removed from the data; Compare the filtered reads with the tomato reference genome ITAG 4.0 to obtain reads that match the expected length of the reference genome;

[0016] 4) Two-group correlation analysis

[0017] The Pearson correlation coefficient between the expression of translated genes and transcript abundance within reads matching the expected length of the reference genome was calculated, and a scatter plot was drawn to analyze the correlation between the two omics of translation and transcriptome. The number of reads at the Ribo-seq level in the ORF region of the coding gene was calculated using Rsem software and converted into FPKM values ​​to obtain the expression of genes at the translation level. The FPKM method was used to calculate gene expression, and the FPKM value can be directly used to compare the differences in gene expression between samples. FDR and log2FC were used to screen differentially translated genes, and edgeR software was used to perform differential translation analysis on genes between groups, with the screening conditions of FDR<0.05 and |log2FC|>1.

[0018] 5) Association analysis of all DEGs

[0019] Step 4) The selected genes were divided into five categories according to their variation patterns in the two omics, corresponding to: 1) Transcription: genes with significant differences in transcriptome; 2) Translation: genes with significant differences only in translation; 3) Homodirection: genes with significant differences in both omics, and the up- and down-regulation directions were the same; 4) Opposite: genes with significant differences in both omics, and the up- and down-regulation directions were opposite; 5) Unchanged: genes with no significant differences in both omics;

[0020] 6) Screening of DEGs related to fruit cracking

[0021] The DEGs screened in step 5) were classified according to GO functional analysis and KEGG pathway analysis, and the DEGs with high expression and significant differences were used as candidate genes regulating tomato fruit dehiscence; the DEGs with high expression and significant differences were further classified according to GO functional analysis and KEGG pathway analysis;

[0022] 7) Protein interaction network prediction

[0023] Use online websites (https: / / cn.string-db.org / and https: / / www.chiplot.online / ) to predict the interactions between the candidate genes screened in step 6) and the encoded proteins to map the protein interaction network; the network construction is based on confirmed interacting proteins, or gene neighborhoods, gene fusions, and gene co-occurrences;

[0024] 8) Translation efficiency analysis

[0025] Translation efficiency (TE) represents the total number of RNA molecules (usually mRNA) in a gene that are bound to ribosomes and translated in a sample; translation efficiency is calculated by the following formula:

[0026] TE=(Ribo-seq in FPKM / (RNA-seq in FPKM)

[0027] Ribodiff was used to analyze the differences in gene translation efficiency between groups, while FDR and log2FC were used to screen DTEGs (differential translation efficiency genes) using the screening conditions of FDR < 0.05 and |log2FC| > 1;

[0028] 9) qRT-PCR analysis and verification

[0029] Ten DEGs were randomly selected using the same samples for qRT-PCR (real-time fluorescence quantitative PCR) analysis, and correlation analysis and verification of qRT-PCR and RNA-seq data were performed.

[0030] The methods described in the present invention are conventional methods and means in the art that are not specifically described, and can be performed according to conventional operations of those skilled in the art. For example, in step 1), RNA Cleaner and Concentrator-25 Kit (ZymoResearch; R1017) can be used to separate FRs.

[0031] The reagents used in the present invention can be purchased from the market, for example, the RNA extraction kit Buffer RL was purchased from Huapu Biotechnology (Changchun, China); the reverse transcription reagent kit PrimeScriptTM RT reagent kit was purchased from abm (Vancouver, Canada); and the qRT-PCR kit TOROGreen qPCR Master Mix was purchased from Toroivd (British Virgin Islands).

[0032] In some embodiments, in step 1), the tissue samples are divided into maternal and paternal germplasm, the maternal germplasm is tomato resistant to fruit cracking, and the paternal germplasm is tomato susceptible to fruit cracking. The tissue samples can also be obtained by gene editing or mutagenesis, that is, genes obtained by gene editing or mutagenesis can also be screened.

[0033] The present invention also provides genes related to tomato fruit cracking obtained by the screening and identification method of the key regulatory gene for tomato fruit cracking.

[0034] The present invention also provides application of the gene related to tomato fruit cracking in breeding tomato varieties resistant to fruit cracking.

[0035] The present invention also includes a method for cultivating fruit cracking-resistant tomatoes by overexpressing or knocking out genes related to fruit cracking in tomatoes.

[0036] The advantages of the present invention over the prior art are:

[0037] (1) Using the method described in the present invention, 41 genes are finally provided in the present embodiment that can coordinately regulate tomato fruit cracking in transcription and translation. These genes are related to plant hormone synthesis, cell wall metabolism, wax and cuticle metabolism, mineral elements, water absorption and transportation, and some genes encoding transcription factors, and a regulatory network related to tomato fruit cracking is established from the transcription and translation levels, including ethylene-cell wall and TFs-aquaporin-boron transporter. In addition, the translatable uORF may reduce the TE of the downstream mORF.

[0038] (2) This study enriched the mechanism of tomato cracking at the transcriptional and translational levels, which can be used to create new germplasm that is resistant to cracking.

[0039] (3) The present invention provides valuable gene resources for the genetic improvement of tomato fruit cracking resistance and enriches the mechanism of tomato fruit cracking at the transcriptional and translational levels, thus having important scientific significance and application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 : Experimental design of transcriptome and translationome experiments on CR and CS tomato genotypes.

[0041] Figure 2 :.Ribo-seq data features, A: Distribution of RFs in coding genes, B: Gene coverage statistics, C: Distribution of RFs around codons (red represents the first base of the codon, green represents the second base of the codon, and blue represents the third base of the codon), D): RFs align to codon positions.

[0042] Figure 3 : Figure of the combined data analysis of the two omics, A: Correlation analysis between the two omics of the four treatments, B: Analysis of DEGs, C: Comparison of translation and transcription differences, D: Consistency analysis of DEGs at the transcriptional and translation levels.

[0043] Figure 4 : Screening of differentially expressed genes related to fruit cracking, A: Differentially expressed genes in tomato fruit cracking in RNA-seq and Ribo-seq, B: Differentially expressed transcription factor genes in tomato fruit cracking in RNA-seq and Ribo-seq, C: Expression of differentially expressed genes in the same pathway after KEGG analysis.

[0044] Figure 5 : Construction of the protein interaction network encoded by differentially expressed genes, A: The protein network encoded by all differentially expressed genes that may be related to fruit dehiscence, B: Plant hormone-cell wall metabolism protein regulatory network, C: Transcription factor-aquaporin-mineral element absorption and transport protein regulatory network.

[0045] Figure 6 : Comparison of genotype differences between the two varieties before and after irrigation, A: DTEG between CR and CS under saturated irrigation at 0 and 30 h, B: Comparison of transcription levels of CR and CS with TE at saturated irrigation at 0 and 30 h.

[0046] Figure 7 : Prediction of the effect of ethylene-regulated cell wall metabolism model on tomato fruit cracking. (Solid arrows indicate regulatory pathways, red arrows indicate regulatory pathways found in this study, and dotted arrows indicate predicted regulatory pathways).

[0047] Figure 8: Regulatory network of tomato fruit dehiscence (red arrows indicate pathways supported by this experiment, black arrows indicate regulatory pathways supported by references, and dotted arrows indicate predicted regulatory pathways).

[0048] Fig. 9 :GO annotation classification statistics of DEGs transcripts.

[0049] Fig.10 : Enrichment scatter plot of the top 20 KEGG pathways enriched by DEGs.

[0050] Fig.11 : Translation efficiency analysis diagram, A: Translation efficiency calculation, B: GO classification of DTEG transcripts, C: Top 20 KEGG pathways enriched in DTEG.

[0051] Fig.12 :Correlation analysis diagram of qRT-PCR and RNA-seq. DETAILED DESCRIPTION

[0052] The embodiments of the present invention will be described in detail below in conjunction with the examples, but those skilled in the art will appreciate that the following examples are only used to illustrate the present invention and should not be considered to limit the scope of the present invention. If no specific conditions are specified in the examples, they are carried out according to normal conditions or the conditions recommended by the manufacturer. If the manufacturer is not specified for the reagents or instruments used, they are all conventional products that can be obtained commercially.

[0053] Embodiment 1:

[0054] 1. Materials and methods

[0055] 1.1 Plant materials and treatments

[0056] A cherry tomato that is extremely resistant to cracking (CR / cherry tomato from the United States) and a tomato genotype that is prone to cracking (CS / LA2661 from the Tomato Genetics Resource Center) were selected as plant materials. The plant materials were self-pollinated for six consecutive generations and had stable cracking traits. The experiment was conducted at Nanjing Agricultural University (Nanjing, Jiangsu Province, China) in March 2019. Seeds were first soaked in warm water for 6 h, then sterilized and washed, and then the seeds were placed on wet filter paper to promote germination. The germinated seeds were sown in 72-well cavity trays. Seedlings with five to six true leaves were grown in a plastic greenhouse at Jiangsu Agricultural Expo Park. Plant spacing followed a 40 × 50 cm pattern, and single-branch pruning was set. When the fruits were ripe, saturated irrigation was used to induce cracking. Fruits at the red ripening stage before and 30 h after irrigation treatment (CR0, CS0, CR30, and CS30, respectively) were collected and RNA-seq and ribosome profiling were performed with three biological replicates ( Figure 1 ).

[0057] 1.2 Data generation and ribosome profiling

[0058] From CR0, CS0, CR30, and CS30, 175.8 million, 210.5 million, 195.89 million, and 193.055 billion bases were obtained from Ribo-seq, and 625 million, 622.9 billion, 652.7 million, and 674.1 million bases were obtained from RNA-seq, respectively. More than 97% of the fragments were distributed in the CDS region, with a smaller number in the UTR and Intron regions ( Figure 2 .A). Gene coverage analysis (i.e., the percentage of each gene covered by reads) was performed ( Figure 2 .B). The results showed that the abundance of translation upstream of the start codon or downstream of the stop codon was extremely low ( Figure 2 .D), the RFs signal gradually decays with the distance from the start codon and the stop codon, which may be due to the assembly of the 60S large subunit and the 40S small subunit at the start codon and the dissociation of the ribosome at the stop codon, resulting in the stagnation of RFs and the accumulation ( Figure 2 .D). Comparison of codon positions by identified RFs ( Figure 2 .C). The bar graph shows that the ribosome spends the longest time at the first base of the codon ( Figure 2 .C). Analysis of RF lengths showed that they were mainly distributed around 29 to 32 bp ( Figure 2 .C).

[0059] 1.3 Correlation analysis between the two groups

[0060] Pearson correlation coefficient R between the two groups of CR0 before irrigation 2 =0.7113, R 2 =0.7099, R of CR30 after irrigation 2 =0.845, R 2 =0.7145( Figure 3 .A). Saturated irrigation treatment during the fruiting stage caused stress to the plants and promoted fruit cracking. By comparing the Pearson coefficient between the two omics before and after irrigation, we found that the correlation between the two omics increased after irrigation treatments of CR and CS. This correlation showed that the number of transcripts was consistent with the abundance of mRNA-bound ribosomes, indicating that adversity promoted a synergistic response between transcription and translation, and there was a certain correlation between transcription and translation.

[0061] 1.4 DEGs analysis

[0062] With CR as the control group and CS as the experimental group, before irrigation treatment, 1753 up-regulated genes and 1010 down-regulated genes were detected at the transcriptional level, and 1447 up-regulated genes and 679 down-regulated genes were detected at the translational level ( Figure 3 .B). After irrigation treatment, 2933 up-regulated genes and 747 down-regulated genes were detected at the transcriptional level ( Figure 3 .B). At the same time, 2853 up-regulated genes and 656 down-regulated genes were detected at the translation level ( Figure 3 .B).

[0063] The expression of DEGs before and after irrigation in CR and CS was studied. We found that compared with before irrigation, 359 up-regulated genes and 1106 down-regulated genes were detected at the transcriptional level in CR 30 h after irrigation, and 210 up-regulated genes and 733 down-regulated genes were detected at the translational level. In CS 30 h after irrigation, 153 up-regulated genes and 278 down-regulated genes were detected at the transcriptional level; 57 up-regulated genes and 241 down-regulated genes were detected at the translational level ( Figure 3 .B). In both CR and CS tomatoes, the number of down-regulated genes after saturated irrigation treatment was higher than that of up-regulated genes. Comparing the two tomato genotypes, CR fruits showed more DEGs after saturated irrigation at both transcriptional and translational levels.

[0064] 1.5 Comparison of translational and transcriptional differences

[0065] Before irrigation treatment, 1156 genes were differentially expressed only at the transcriptional level, accounting for 35.22% of all differentially expressed genes; 660 genes were differentially expressed only at the translational level, accounting for 20.11%; a total of 1466 genes (44.67%) showed significant differences in both transcription and translation, and were regulated up and down in the same direction. After irrigation treatment, among all differentially expressed genes, 1155 genes were differentially expressed only at the transcriptional level, accounting for 24.76% of all differentially expressed genes; 1154 genes were differentially expressed only at the translational level, accounting for 24.74%; a total of 2352 genes (50.43%) were significantly differentially expressed at both the transcriptional and translational levels, and the up and down regulation directions were the same; finally, only 3 genes (0.06%) were significantly different at both the transcriptional and translational levels, and the up and down regulation directions were opposite ( Figure 3 .C). By analyzing all DEGs before and after irrigation, it was found that only about half of the genes were up-regulated or down-regulated at both the transcriptional and translational levels before and after irrigation treatment ( Figure 3 .D). At the same time, after watering, the proportion of genes with significant differences in both transcription and translation levels and the same up- and down-regulation direction increased, indicating that plants respond to stress at both the transcription and translation levels after being subjected to external stress ( Figure 3 .D).

[0066] After obtaining DEGs, the DEGs were classified and counted by GO terms. Fig. 9 As shown in the figure, the horizontal axis represents three GO (ontology): molecular function, cellular component and biological process. At the transcriptional level, the DEGs of CR and CS were mainly enriched in metabolic process, cellular process, single organic matter process, cell, cellular component, catalytic activity and integration, and the number of up-regulated genes in each function was slightly higher than the number of down-regulated genes. After irrigation treatment, the main enriched functions of DEGs in CR and CS did not change, but the number and proportion of up-regulated genes increased significantly. At the translational level, the functional enrichment before and after irrigation treatment was consistent with the transcriptional level, and after irrigation treatment, the number and proportion of up-regulated genes in CR and CS increased significantly ( Fig. 9 ). Combining RNA-seq and Ribo-seq data, GO analysis was performed on genes with significant differences and the same up- and down-regulation directions in the two omics, and it was found that the main enriched functions were consistent with any single omics ( Fig. 9 ). KEGG Pathway analysis showed that DEGs were mainly enriched in metabolic pathways and biosynthetic pathways of secondary metabolites at the transcriptional and translational levels. The combined analysis of the two omics showed that both omics were significantly different, and genes with the same up- and down-regulation directions were mainly enriched in metabolic pathways and biosynthetic pathways of secondary metabolites, which was consistent with the results of single omics ( Fig.10 ).

[0067] 1.6 Screening of DEGs related to fruit cracking

[0068] In order to obtain genes that regulate tomato fruit cracking, we combined RNA-seq and Ribo-seq data and screened DEGs of CR and CS using FDR<0.05 and |log2FC|>1. After screening the DEGs of these pathways, it was found that there were 11 DEGs, such as Solyc04g071070.2 (EXT3) and Solyc03g123630.4 (PMEU1), which were related to cell wall modification and degradation, mainly involving the modification and degradation of major cell wall components such as xyloglucan, pectin, mannan, and extensin. There are 8 DEGs related to plant hormones, mainly involving the synthesis and regulation of auxin, ethylene, gibberellin, and abscisic acid. A total of 5 DEGs were screened that were related to mineral elements, namely calcium and boron. A total of 3 DEGs were found to be related to aquaporins ( Figure 4 .A).

[0069] By analyzing the data from both omics, a total of 10 differentially expressed transcription factors were identified. The differentially expressed TFs mainly covered the WRKY and ERF / AP2 families ( Figure 4 .B). Detailed information of all 41 DEGs is provided in Table 1.

[0070] Table 1

[0071]

[0072]

[0073] Among the differentially expressed genes, three genes with cell wall metabolism functions were found to be enriched in the same KEGG pathway (ko00040 / / Pentose and gluconate interconversion). Searching the KEGG official website (https: / / www.kegg.jp / ), the results showed that the three differentially expressed genes were located upstream of the pathway and encoded proteins involved in the pectin degradation pathway.

[0074] 1.7 Prediction of gene interaction networks associated with fruit cracking

[0075] Predict the functions of the screened DEGs and the interaction networks of the encoded proteins. Figure 5 .A). Two potential pathways that may regulate tomato fruit dehiscence come from large protein regulatory networks, namely the hormone-cell wall metabolism pathway and the transcription factor-water channel-mineral element pathway ( Figure 5 .B, 5.C).

[0076] In the protein interaction network in hormone and cell wall metabolism, the functions of the covered proteins were found to include 1-aminocyclopropane-1-carboxylic acid oxidase 1, a key enzyme in ethylene synthesis, and 1-aminocyclopropane-1-carboxylic acid oxidase homolog. Figure 5 .B)

[0077] In the transcription factor-aquaporin-mineral element absorption and transport protein interaction network, it was found that the covered proteins included a WRKY family, two ethylene transcription factors ( Figure 5 .C).

[0078] 1.8 Analysis of translation efficiency (TE) between the two varieties

[0079] Translation efficiency (TE) is an important indicator of the translation process. We calculated the TE of genes in the two species before and after irrigation using the formula ( Fig.11.A). It was found that there were no more differential translation efficiency genes (DTEGs) before and after one irrigation, while there were a large number of DTEGs between the CR and CS genotypes before irrigation. Compared with the CR genotype before irrigation, there were 68 up-regulated DTEGs and 13 down-regulated DTEGs in the CS genotype, and compared with the CR genotype after irrigation treatment, there were 286 up-regulated DTEGs and 81 down-regulated DTEGs in the CS genotype ( Figure 6 .A). After obtaining DTEGs, TE differences were compared with transcription differences, and it was found that before irrigation, 10 genes had significant differences in TE and transcription levels in the same direction, and 16 genes had differences in the opposite direction. After irrigation, 38 genes had significant differences in TE and transcription levels in the same direction, and 54 genes had significant differences in the opposite direction. ( Figure 6 .B) Then, we performed functional enrichment analysis on the screened DTEGs. Through GO term analysis, we found that DTEGs were mainly enriched with functions consistent with previous single-omics analysis and two-omics association analysis ( Fig.11 .B). KEGG pathway analysis showed that DETGs were mainly enriched in ribosome, spliceosome and RNA transport pathways ( Fig.11 .C)

[0080] 1.9 qRT-PCR analysis

[0081] Validation of qRT-PCR showed that the relative expression of the tested genes had the same trend and difference as the RNA-seq data. Correlation analysis of qRT-PCR and RNA-seq data showed R2 = 0.807, and R2>0.8 proved the reliability of RNA-seq data ( Fig.12 ).

[0082] 2. Results

[0083] 2.1 Transcription and translation coordinately regulate tomato fruit dehiscence

[0084] After calculating the Pearson correlation coefficient between translation gene expression and transcript abundance, it was found that the Pearson correlation coefficient between the two omics of the two genotypes before and after irrigation was >0.7, indicating that there was a synergistic effect of transcription and translation. This was later confirmed by the analysis of DEGs, which showed genes that were differential at both transcriptional and translational levels. Before irrigation treatment, genes that were up-regulated and down-regulated in the same direction between CR and CS accounted for 44.67% of all DEGs. After irrigation treatment, genes that were differential at both transcriptional and translational levels and up-regulated in the same direction accounted for 50.43% of all DEGs. Nearly half of the DEGs were able to show consistency at the transcriptional and translational levels. Before and after irrigation, the number of up-regulated genes at both transcriptional and translational levels in CS was greater than the number of down-regulated genes compared with CR, indicating that CS fruits are recruiting more candidate genes to cope with stress.

[0085] About 50% of DEGs showed single-omics differences between the two species, indicating that transcription and translation are independent. Before watering, 660 genes differed only at the translation level between CR and CS, and 1154 after watering. This indicates that the translation level is translated on the original mRNA and does not require the production of new mRNA.

[0086] 2.2 Prediction of the regulatory network of tomato fruit cracking at the transcriptional and translational levels

[0087] The ethylene metabolic pathway includes ethylene synthesis and ethylene signaling, and ACO and ethylene response factors (ERFs) play important roles in the pathway. In addition, hormone and ROS signals can coordinately regulate the modification of fruit cell walls, which in turn affects fruit cracking, and some transcription factors may be involved in this network. Ethylene response factors can regulate the expression of cell wall-related genes. In this study, four differentially expressed ERFs in CR and CS fruits were screened. It has been shown that several enzymes regulate the softening process of fruits, such as pectin methylesterase (PME), β-galactosidase (β-gal), and cellulase (Cx). Jiang et al. found that silencing PG and EXP significantly reduced the cracking rate of tomato fruits. Studies have shown that ethylene is an important plant hormone that may be involved in regulating fruit ripening and softening by affecting the expression of genes related to cell wall metabolism. Treatment of apples with the ethylene inhibitor 1-MCP significantly reduced the expression of PG in the peel, reduced the fruit cracking rate, and extended the shelf life. The DEGs screened in this study include some genes related to ethylene synthesis and cell wall metabolism. By analyzing the protein interaction network, we obtained the equivalent relationships between ethylene-related enzymes and pectin-degrading enzymes as well as tensile proteins, especially ACO and pectin metabolic enzymes ( Figure 7). In addition, some genes with potential regulation of tomato fruit dehiscence were discovered, whose functions involved plant hormones, keratin wax synthesis, mineral element absorption and cellular water absorption. IAA3 has been shown to be a molecular bridge between the auxin and ethylene signaling pathways in tomatoes; while IAA17 plays a key role in controlling fruit quality, and the fruit skin of the IAA17 silent line is thicker than that of the wild-type line. In addition, ethylene promotes programmed cell death of epidermal cells in response to ROS signals; while growth hormone induces the production of ROS, and OH- produced by H2O2 decomposes complexes on the cell wall. In our study, two significantly different and highly expressed reactive oxygen species-related genes were identified, further supporting our prediction of the tomato fruit dehiscence pathway map.

[0088] Fruit ripening is also affected by the interaction between ethylene and abscisic acid, which promotes fruit cracking by increasing water flow into the fruit. Gibberellic acid promotes the deposition of cuticle substances, thereby increasing the elasticity of the cuticle and reducing the activity of several cell wall degradation-related enzymes, thereby preventing fruit cracking.

[0089] Mineral elements may affect fruit cracking, and spraying mineral salt solution before harvest can effectively avoid fruit cracking. Calcium was found to play an important role in maintaining cell wall structure and strength, regulating cell osmotic pressure, and reducing fruit cracking. The application of elements such as boron and iron before the pit hardening reduced the rate of sweet cherry cracking. In this study, 5 DEGs were screened, and the gene functions were annotated as boron transport and calcium binding. The cuticle is attached to the epidermal cells of plants and is composed of keratin, wax and other components. Its structure and function have an important influence on the mechanism of fruit cracking. It has been proven that epidermal wax is effective in preventing water loss from plant tissues, and changes in osmotic pressure caused by fruit water absorption are also one of the factors leading to tomato fruit cracking. In this study, a total of four DEGs related to keratin wax synthesis were identified.

[0090] Water is also an important factor affecting fruit cracking. The absorption of a large amount of water by the fruit through the root system will cause cell swelling and increased swelling pressure, leading to cell and tissue rupture. In this experiment, differentially expressed genes related to water channel proteins located on the cell membrane were screened through irrigation treatment, which may be crucial for the large amount of water absorption by cells and the impact on fruit cracking.

[0091] Subsequently, the promoter cis-acting elements of the screened DEGs related to cell wall metabolism, cuticle metabolism, and mineral elements were analyzed, and the promoters of the DEGs were found to contain ethylene, abscisic acid, growth hormone, and gibberellin response elements.

[0092] Based on DEGs, protein interaction analysis and previous studies, a hypothesis for the tomato fruit dehiscence network was proposed, involving ethylene cell wall and transcription factor osmotic pressure ( Figure 8). For detailed information on DEGs, see Table 2. Significantly different traits of fruit dehiscence, two potential regulatory networks related to fruit dehiscence were established, including ethylene-cell wall and TFs-aquaporin-boron transporter. Analysis of TEs showed that there were significant differences in the expressed genes, such as Solyc03g111690.4 and Solyc09g089580.4. Analysis of ORFs showed that translatable uORFs may reduce the TE of downstream mORFs.

[0093] Table 2

[0094]

Claims

1. A method for screening and identifying key regulatory genes for tomato fruit cracking, The following steps are involved: 1) Ribosome imprint sequencing: tissue samples were frozen in liquid nitrogen, ground and added to lysis buffer, mixed and centrifuged, and then the supernatant was added to RNase I and DNase I to remove DNA and RNA, ribosomes were separated using MicroSpin S-400 columns, RFs were isolated, rRNA was removed, and magnetic beads were used to further purify RFs; these RNA fragments protected by ribosomes and extracted for sequencing are also called ribosome footprints (RFs); sequencing was performed on the Illumina HiSeq 4000 sequencing platform, and the quality of sequencing data reached Q20-Q30 for subsequent analysis; the number of RFs located around the start codon and stop codon of the CDS of the coding gene was calculated based on the comparative position of the 5' end of the RFs in the genome; the RFs compared with the CDS region were divided into three categories with a codon number of 1 to 3 bases according to the codon position corresponding to the 5' end of the RFs comparison position; 2) RNA-seq data processing and analysis: Quality control of raw reads for downstream machines was performed using fastp to filter low-quality data and obtain clean reads; the steps for read filtering were as follows: (1) Delete reads containing adapters; (2) remove reads containing N ratio greater than 10%; (3) delete all reads with the same base number as A; (4) remove low-quality reads (bases with a quality value of Q≤20); According to the HISAT2 comparison results, transcripts were reconstructed using String Tie, and the expression of all genes in each sample was calculated; expression values ​​were expressed as fragments per kilobase per million (FPKM); the data were then analyzed using DESeq2 software; based on the analysis results, genes with FDR < 0.05 and |log2FC| > 1 were screened as significant DEGs (differentially expressed genes); DEGs were further subjected to GO and KEGG enrichment analysis using ClusterProfiler89; 3) Ribo-seq data processing and analysis: Before bioinformatics analysis, off-plate data were quality controlled and filtered; the raw reads obtained after initial filtering, after which reads containing connectors, all A bases, and more than 10% N and low quality (Q20 ≤ 50%) were removed; Ribosomal RNA removal: To eliminate the impact of ribosome contamination on subsequent analysis, the short reads matching tool bowtie2 (2.2.8) was used to match the Clean Reads with the ribosome database and delete the reads that did not match the ribosome. The retained data were used for subsequent analysis. Remove transfer RNA, etc.: Compare the reads after ribosomal RNA removal with the GenBank and Rfam databases and remove small nucleolar RNA (snoRNA), small nuclear RNA (snRNA) and micro RNA (miRNA): After removing rRNA and tRNA, snoRNA and snRNA are discovered and removed from the sample by blasting the GenBank and Rfam databases; miRNA sequences identified in miRBase are also removed from the data; Compare the filtered reads with the tomato reference genome ITAG 4.0 to obtain reads that match the expected length of the reference genome; 4) Two-group correlation analysis The Pearson correlation coefficient between the expression of translated genes and transcript abundance within reads matching the expected length of the reference genome was calculated, and a scatter plot was drawn to analyze the correlation between the two omics of translation and transcriptome. The number of reads at the Ribo-seq level in the ORF region of the coding gene was calculated using Rsem software and converted into FPKM values ​​to obtain the expression of genes at the translation level. The FPKM method was used to calculate gene expression, and the FPKM value can be directly used to compare the differences in gene expression between samples. FDR and log2FC were used to screen differentially translated genes, and edgeR software was used to perform differential translation analysis on genes between groups, with the screening conditions of FDR<0.05 and |log2FC|>1. 5) Association analysis of all DEGs Step 4) The selected genes were divided into five categories according to their variation patterns in the two omics, corresponding to: 1) Transcription: genes with significant differences in transcriptome; 2) Translation: genes with significant differences only in translation; 3) Homodirection: genes with significant differences in both omics, and the up- and down-regulation directions were the same; 4) Opposite: genes with significant differences in both omics, and the up- and down-regulation directions were opposite; 5) Unchanged: genes with no significant differences in both omics; 6) Screening of DEGs related to fruit cracking The DEGs screened in step 5) were classified according to GO functional analysis and KEGG pathway analysis, and the DEGs with high expression and significant differences were used as candidate genes regulating tomato fruit dehiscence; the DEGs with high expression and significant differences were further classified according to GO functional analysis and KEGG pathway analysis; 7) Protein interaction network prediction Use online websites (https: / / cn.string-db.org / and https: / / www.chiplot.online / ) to predict the interactions between the candidate genes screened in step 6) and the encoded proteins to map the protein interaction network; Network construction was based on confirmed interacting proteins, or gene neighborhoods, gene fusions, and gene co-occurrences; 8) Translation efficiency analysis Translation efficiency (TE) represents the total number of RNA molecules (usually mRNA) in a gene that are bound to ribosomes and translated in a sample; translation efficiency is calculated by the following formula: TE = (Ribo-seq in FPKM / (RNA-seq in FPKM)Ribodiff was used to analyze the differences in gene translation efficiency among groups, while FDR and log2FC were screened for DTEG (differential translation efficiency genes) using the screening conditions of FDR < 0.05 and |log2FC| > 1; 9) qRT-PCR analysis and verification Ten DEGs were randomly selected from the same samples for qRT-PCR (real-time fluorescence quantitative PCR) analysis, and correlation analysis and validation of qRT-PCR and RNA-seq data were performed.

2. The method for screening and identifying key regulatory genes for tomato fruit cracking according to claim 1, It is characterized in that In step 1), the tissue samples are divided into maternal and paternal parent, wherein the maternal parent for which the trait to be screened is easily expressed is a tomato germplasm for resistant fruit cracking, and the paternal parent is a germplasm for easy fruit cracking.

3. The method for screening and identifying key regulatory genes for tomato fruit cracking according to claim 1, It is characterized in that The method also includes editing or mutagenesis of the gene.

4. A gene related to tomato fruit cracking obtained by the screening and identification method according to any one of claims 1 to 3.

5. Use of the gene according to claim 4 in breeding tomato varieties resistant to fruit cracking.

6. A method for cultivating crack-resistant tomatoes, It is characterized in that The crack-fruit-resistant tomato is obtained by overexpressing or knocking out the gene described in claim 4.

Images