Comprehensive analysis of the lysine succinylome and protein co-modifications in developing rice seeds

: Lysine succinylation has been recognized as a post-translational modification (PTM) in recent years. It is plausible that succinylation may have a vaster functional impact than acetylation due to bulkier structural changes and more significant charge differences on the modified lysine residue. Currently, however, the quantity and identity of succinylated proteins and their corresponding functions in cereal plants remain largely unknown. In this study, we estimated the native succinylation occupancy on lysine was between 2% to 10% in developing rice seeds. Eight hundred fifty-four lysine succinylation sites on 347 proteins have been identified by a thorough investigation in developing rice seeds. Six motifs were revealed as preferred amino-acid sequence arrangements for succinylation sites, and a noteworthy motif preference was identified in proteins associated with different biological processes, molecular functions, pathways, and domains. Remarkably, heavy succinylation was detected on major seed storage proteins, in conjunction with critical enzymes involved in central carbon metabolism and starch biosynthetic pathways for rice seed development. Meanwhile, our results showed that the modification pattern of in vitro nonenzymatically succinylated proteins was different from those of the proteins isolated from cells in Western blots, suggesting that succinylation is not generated via nonenzymatic reaction in the cells, at least not completely. Using the acylation data obtained from the same rice tissue, we mapped many sites harboring lysine succinylation, acetylation, malonylation, crotonylation, and 2-hydroxisobutyrylation in rice seed proteins. A striking number of proteins with multiple modifications were shown to be involved in critical metabolic events. Given that these modification moieties are intermediate products of multiple cellular metabolic pathways, these targeted lysine residues may mediate the crosstalk between different metabolic pathways via modifications by different moieties. Our study exhibits a platform for extensive investigation of molecular networks administrating cereal seed development and metabolism via PTMs.

residue under physiological pH (2,3). Since succinylation results in a bulkier structural change and more significant charge difference on lysine, one could postulate that it generates a grander impact on the substrate protein's structures and functions in comparison to well-studied lysine acetylation and methylation (1,2).
Rice is one of the most significant cereals as it serves as the staple food for over half of the world's population (26). In rice grain, the majority of nutrients are stored in the form of starch, lipid, and protein, which extensively contribute to grain nutritional value, milling properties, appearance, and cooking quality (27). The content and composition of storage starch and protein are directly associated with seed development. Recently, PTMs of lysine acetylation (28), malonylation (29), and 2-hydroxyisobutyrylation (30) have been reported in developing rice seeds.
In this report, we successfully identified 854 lysine succinylation sites across 347 proteins with a false discovery rate (FDR) of ≤ 1% in developing rice seeds. Our results indicate that lysine succinylation is a highly conserved modification. It frequently occurs in the rice proteome with a preference on carbon metabolic pathways, starch biosynthetic pathways, and the major seed storage proteins. Furthermore, a large number of proteins involved in crucial metabolic processes were revealed to embrace various modifications in lysine residues. This analysis provides a comprehensive view of lysine modifications in developing rice seeds.
were purified by Sep-Pak C18 Plus Short Cartridge (WAT020515, Waters), and were lyophilized in a speed vacuum and stored at -80 °C before proper usage. All succinyl-CoA sodium salt treatments were performed with three technical replicates.

Western blot analysis
Protein samples extracted from different rice organs/tissue were separated by 12% SDS-PAGE and transferred onto PVDF membrane (EMD Millipore) for Western blot. Succinylated proteins were detected by utilizing rabbit-derived pan anti-succinyl lysine antibody (PTM-401, PTM Biolabs) in a 1:1000 (v/v) dilution overnight at 4°C, accompanied by gentle shaking in reference to the supplier's instruction.

Trypsin digestion
The lyophilized 15 dpa rice seed protein was re-dissolved in a buffer (8 M urea, 100 mM NH4HCO3, pH 8.0), and the protein content was determined with 2-D Quant kit (GE Healthcare). Before digestion, the protein was reduced with 10 mM DTT (Dithiothreitol) for 1 hour at 37 °C, and subsequently alkylated with 20 mM IAA (Iodoacetamide) for 45 minutes in dark setting at room temperature. For trypsin digestion, the protein sample was diluted with an addition of 100 mM NH4HCO3 for urea reduction to the desired concentration of less than 2M. Sequencing-grade trypsin (V5111, Promega Corporation) was supplemented at a 1:50 (w/w) enzyme-to-substrate mass ratio for an overnight digestion period and likewise at a 1:100 (w/w) enzyme-to-substrate mass ratio for an additional 4 hour digestion period at 37 °C.

HPLC fractionation and affinity enrichment of lysine succinylated peptides
The peptides were fractionated into 12 fractions by high pH reverse-phase HPLC with Agilent 300 by guest on March 6, 2020 https://www.mcponline.org
Peptide analysis was conducted by nano-HPLC/MS/MS, utilizing an EASY-nLC 1000 UPLC system (Thermo Fisher Scientific) attached to a Q-Exactive Plus Hybrid Quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific). Briefly, the peptides were eluted onto a reversedphase analytical column (2 μm particles, 50μm ID, 15cm length, Acclaim PepMap RSLC, Thermo Fisher Scientific). The utilization of a consistent flow rate of 280 nl/min with a 16-minute linear gradient from 7% to 18% of the solvent buffer of 0.1% FA in 98% ACN was the initial response.
A concentration increase of 18% to 22% for an 8 minute period and a concentration increase of 22% to 35% for another 8 minutes was achieved. The ascending concentration grasped 80% in another 5 minutes and immobilized at this value of the solvent buffer for the remaining 3 minutes.
The resulting peptides were subjected to a NanoSpray Ionization (NSI) source, succeeded by tandem mass spectrometry in Q Exactive TM Plus (Thermo Fisher Scientific) connected online to the UPLC. Intact peptides were detected at a resolution of 70,000 in the Orbitrap, and peptides were selected for MS/MS with the NCE setting to the preferred quantity of 30. Ion fragments were detected at a resolution of 17,500. A data-dependent procedure that alternated between one MS scan and 20 MS/MS scans was applied for the highest 20 precursor ions exceeding a threshold ion count of 1.0×10 4 in the MS survey scan with 15.0s dynamic exclusion. The operated electrospray voltage was 2.0 kV, and automatic gain control (AGC) was applied to avoid ion trap overcapacity.
Ions with charge states of 2 to 5 were permitted, and 5×10 4 ions were collected for the generation of MS/MS spectra. The m/z scan range was approximately 350 to 1800 for MS scans.
The protein and succinylation sites were identified by MaxQuant software (http://www.maxquant.org/) coupled with Andromeda search engine (v.1.4.1.2) (35). The mass spectra of the raw data were searched against UniProt_Oryza sativa_japonica database (63,195 sequences, released July 2014) concatenated with reverse decoy database. Trypsin/P was specified as a cleavage enzyme, allowing a maximum of 4 missing cleavages, 5 modifications per peptide, and 5 charges. Carbamidomethylation on cysteine was distinguished as a fixed modification.
Oxidation on methionine, succinylation on both lysine and protein N-terminal were specified as variable modifications. The mass error was programmed to 10 ppm for precursor ions and 0.02 Da for fragment ions. The false discovery rate (FDR) thresholds for protein, peptide, and modification site were adjusted to a quantity of 1% (36), and the minimum peptide length was changed to a quantity of 7. Lysine succinylation sites identified with a localization probability of < 0.75 were by guest on  eliminated. The remaining parameters in MaxQuant were set to default values.

Bioinformatics analysis of lysine succinylated peptides and proteins
Gene Ontology (GO) annotation was obtained from the UniProt-GOA database (www. http://www.ebi.ac.uk/GOA/). If some identified proteins were not annotated by UniProt-GOA database, the InterProScan software was used based on protein sequence alignment method. Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used to annotate protein pathway. Conserved amino-acid sequence motifs of succyl-21-mers (ten amino acids upstream and downstream of the succinylation site) were analyzed by motif-x (http://www.motifx.med.harvard.edu). The entire protein sequence of Oryza sativa L. japonica was utilized as the background database. For motif logo-based clustering analysis, the −log10 (p-value) were ztransformed for each category (p-value < 0.05) and gathered by one-way hierarchical clustering in Genesis. Cluster membership was visualized by heat map through the "heatmap.2" function, located in the "gplots" R-package. by guest on  A BLAST analysis was accomplished by operating the two-directional BlastP software with BLOSUM62 matrix, with the objective of evaluating succinylated protein conservation across selected succinylomes of diverse organisms. BLAST outcomes that encompassed a corresponding e value < 1E-10, a score > 200, and an identity > 30 were taken as the best-scoring homologous proteins.
Protein-protein interaction networks were constructed against the STRING database (version 10.5). Evidence for protein-protein interaction was obtained from seven sources: experiments, databases, text mining, co-expression, neighborhood, co-occurrence, and gene fusion. Identified interactions with confidence scores ≥ 0.7 (high confidence) were obtained. Interaction networks from STRING were visualized with Cytoscape software.

Experimental design and statistical rationale
Proteome-wide succinylation of rice (Oryza sativa L. japonica cv. Nipponbare) was investigated by using 15 dpa developing seeds selected from 5 individual plants. Succinylated peptides were enriched using immunoaffinity enrichment strategies from 12 HPLC fractions and analyzed by high accuracy nano-HPLC/MS/MS. Since the current research aims to detect the modification sites instead of performing quantitative analysis among different tissues or treatments, no analyses with replicates were involved. Instead, we used high stringent screening criteria for mass spectrometry.
The mass error was programmed to 10 ppm for precursor ions and 0.02 Da for fragment ions. FDR thresholds for proteins, peptides, and modification sites were set at 1% using MaxQuant software (36). Minimum peptide length was set at 7. Lysine succinylation sites identified with a localization probability of < 0.75 were eliminated. Succinylated proteins were annotated by GO, KEGG, and by guest on March 6, 2020 domain enrichment analyses. A two-tailed Fisher's exact test was used to examine the identified succinylated proteins. Multiple testing correction was performed using the Benjamini-Hochberg FDR control method. The GO term, KEGG pathway term and domain categories with a corrected p-value < 0.05 and enrichment fold ≥ 2 were determined as significantly enriched. For motif logobased clustering analysis, the −log10 (p-value) were z-transformed for each category (p-value < 0.05) and gathered by one-way hierarchical clustering in Genesis. Protein in vitro nonenzymatic succinylation was achieved by succinyl-CoA sodium salt treatments. Quantitative comparative analysis of succinylation level was revealed by Western blot with three technical replicates.

Protein lysine succinylation in different rice tissue/organs
To detect the status of lysine succinylation in rice proteomes, proteins from nine different tissue/organs were examined by Western blot. These distinctive tissue/organs included cultured cells, roots, leaves, flowers, pollens, 7 dpa seeds, 15 dpa seeds, 21 dpa seeds, and mature dry seeds.
Specific antibodies for succinylated lysine residue were utilized under equal protein loading conditions ( Fig. 1 and Supplementary Fig. S1). Multiple protein bands possessing different sizes and smears were detected from all tested samples (Fig. 1b). It was evident that different tissue/organs possessed different patterns of protein succinylation modification. These discoveries demonstrated that lysine succinylation is a highly specific modification system and widely presented in the rice plant.

Cellular proteins displayed a distinct succinylation pattern compared to in vitro non-enzymatically succinylated proteins in Western blots
Succinylation could occur via a chemical mechanism dependent on succinyl-CoA concentration and pH (4,5). Using pan anti-succinyl lysine antibody, we observed that incubating proteins isolated from developing rice seeds with 0.5 mM or 1mM of succinyl-CoA sodium salt under high pH condition (pH7.2) caused a significant increase in lysine succinylation compared with native rice proteins without succinyl-CoA treatment (Fig. 2a, compare lanes 1 and 2 with lanes 3 and 4).
We confirmed via Western blot that succinyl-CoA could non-enzymatically succinate rice proteins in vitro under certain pH. Western blot signals for the succinyl-CoA treated samples tended to be saturated (Fig. 2a, lane 3 and 4), and we speculate that all proteins were fully succinylated in the artificially treated samples.
To estimate global protein succinylation occupancy in developing rice seeds, the initial quantity of 1mM succinyl-CoA sodium salt-treated protein sample was reduced to 60%, 45%, 30%, 15%, 10%, 5%, 2%, and 1% for Western blot analysis ( and lane 10 ( Fig. 2a), which was 10% and 5% amount of the protein sample saturated with succinylation ( Fig. 2b), respectively. With the removal of the background, the succinylation intensities of the lower band of native developing rice seed proteins were slightly higher or close by guest on  to the intensity presented in lane 11 (Fig. 2a), which was 2% amount of the developing rice seeds protein saturated with succinylation ( Fig. 2b). Three technical repeats were conducted and shown in Supplementary Fig. S2. This observation indicated that the native succinylation occupancy was between 2% to 10% in developing rice seeds.
Comparing the band patterns in lanes 1 and 2 with lanes 11 in the Western blot, it is clear that lane 11 is less intense than lanes 1 and 2 in the thick upper protein band ( Fig. 2a and b). However, multiple weaker bands with larger protein size can be seen in lane 11 but not in lanes 1 or 2.
Therefore, the results demonstrated that chemical mediated succinylation occurs more evenly to all protein bands, while native succinylation occurs more selectively -only occurring on specific protein bands. Understanding the global succinylation occupancy of lysine residues is essential for understanding succinylation functions in rice plant. Using SWATH-MS2 methodology, the majority of lysine succinylation modifications in wild-type E.coli proteins showed less than 2% occupancy (37). Our approximation of succinylation occupancy estimated in developing rice seeds was between 2% to 10%.

Global profiling of succinylated peptides and proteins
To examine the succinylated peptides on a global scale, the 15 dpa developing rice seeds were collected for protein extraction. After trypsin digestion, the proteins were fractioned into 12 fractions to reduce the complexity of the peptide samples. Then, the succinylated peptides were To investigate the coverage of succinylated lysine residues in a modified protein, the frequency of succinylated lysines within each modified protein was calculated by dividing the number of succinylated lysine residues with a total number of lysine residues and the results were recorded in Supplementary Figure S4. Most identified proteins had fewer than 20% of the lysine residues experiencing succinylation. However, 22 proteins with succinylated lysine residues exceeded 40% of the total lysines, and 2 proteins obtained succinylated lysine rates surpassing 70% of the total lysine residues. One of the two exceptional proteins is non-specific lipid-transfer protein 1(Q7XBA6). Observations in Mycobacterium tuberculosis and mice suggested a by guest on  connection between lysine succinylation and fatty acid metabolism (7,17,38). Sirt5-deficient mice exhibited accumulated lysine succinylation level and defective fatty acid metabolism (38). It will be interesting to investigate the connection between non-specific lipid-transfer protein succinylation and fatty acid metabolism in plants.

Characterization of lysine succinylated peptides
To understand the regulation and amino acid residue preference at the sites surrounding the succinylated lysine, we carried out succinylation site motif analysis by examining the sequences from -10 to +10 of the 854 succinylation sites. Six distinguished motifs were identified: KsuXXXXR, KsuXI, KsuXXXR, LKsu, KsuXXXXXXXR, and KXXXXKsu (Ksu indicates the succinylated lysine, and X indicates a random amino acid residue) (Fig. 3a). The frequency of amino acid residues flanking succinylated lysine was analyzed to investigate the enrichment or depletion of various amino acids (Fig. 3b). Arginine at positions from +4 to +9, especially at +4, +5 and +7, and isoleucine at +2 position, leucine at -1 position, were substantially preferred and these patterns agreed with the identified conserved motifs reported in this study. Interestingly, lysines with an arginine positioned at -1 or another lysine located from -2 to + 2 have the lowest frequency to be succinylated according to flanking sequence analysis (Fig. 3b). Remarkably, motif LKsu had also been identified in the succinylomes of E.coli BW25113, T.gondii, mouse liver mitochondria, and S.cerevisiae (4,7,15). Motif KsuXI was also reported in B.subtilis succinylome A noteworthy motif preference was discovered in proteins associated with different biological processes, molecular functions, cellular components, pathways, and domains (Fig. 3c-e). For example, the proteins expressing binding activities prefer to harbor motifs of KXXXXKsu and KsuXI (Fig. 3c). The storage-related proteins in developing rice seeds habitually adopt cupin 1 domain, RmlC-like cupin domain, and RmlC-like jelly roll fold domain. They prefer to harbor motifs containing arginine (KsuXXXR, KsuXXXXR, KsuXXXXXXXR) (Fig. 4e). The recognition of conserved motifs and motif preferences in different protein clusters verifies that protein lysine succinylation is a highly regulated modification process instead of non-specific modifications, and its function is influenced by the environment of neighboring amino acid residues.

Enrichment analyses of lysine succinylated proteins
Enrichment analyses of GO annotations, KEGG pathways, and protein domains were performed to expand our knowledge on the biological regulations and functions of succinylated proteins in developing rice seeds ( Supplementary Fig. S5 and Supplementary Table S2).
GO enrichment analysis indicated that proteins related to molecular functions of oxidoreductase activities, nutrient reservoir activities, binding activities, translation elongation factor activity, antioxidant activity, etc. were significantly enriched in rice succinylome with corrected p-value < 0.05 and fold enrichment ≥ 2 ( Supplementary Fig. S5). Moreover, GO by guest on March 6, 2020 enrichment analysis suggested that the succinylated proteins in developing rice seeds are significantly enriched in various biological processes of biotic/abiotic stresses and responses to inorganic/ionic substances (Supplementary Fig. S5). Therefore, succinylation may have a fundamental engagement in stress regulatory procedures and defense responses.
KEGG pathway enrichment analysis substantiated the idea that most succinylated proteins were essential participants in carbon, energy, and amino acid metabolic pathways (corrected pvalue < 0.05 and fold enrichment ≥ 2) ( Supplementary Fig. S5). Outstandingly, succinylated proteins recognized in other reported eukaryotic and prokaryotic succinylomes were acknowledged as significant contributors in analogous pathways (7, 13-15, 17, 19-21). Perception of this data portion provided indications of a certain degree of conservation embraced by succinylation among mammals, microbes, and plants. Enzymes involved in fatty acid oxidation pathways were strikingly succinylated in mouse liver cells (4,7). However, the succinylation status of the homologous proteins in developing rice seeds is incompatible with mouse liver cells, suggesting that a particular extent of functional differentiation in succinylation may transpire between plants and mammals.
Protein domain enrichment analysis for succinylated proteins showed that domains of 11-S seed storage protein conserved site, RmlC-like jelly roll fold, RmlC-like cupin, plant 11-S storage protein, and cupin 1 were significantly enriched (corrected p-value < 0.05 and fold enrichment ≥ 2) (Supplementary Fig. S5), and these domains were consistently positioned in the nutrient reservoir proteins of seeds.

Lysine succinylated proteins in central carbon metabolism
by guest on  Extensive evidence revealed that lysine succinylation targets metabolic enzymes for carbon source utilization and metabolic flux coordination in both prokaryotes and eukaryotes (4,7,14).
Central carbon metabolism (TCA cycle and glycolysis/gluconeogenesis pathway) accomplishes oxidative degradation of fatty acids, amino acids, and monosaccharides for energy production.
Grain filling is mainly a process of carbohydrate accumulation. Thus, carbon metabolism is crucial for the development of cereal seeds. Gluconeogenesis pathway generates glucose as the primary substrate for starch biosynthesis. Proteomic and transcriptional studies revealed that protein expression in glycolysis and TCA cycle was less abundant in inferior spikelets than in superior spikelets during grain filling (39,40).
In developing rice seeds, the majority of enzymes engaged in TCA cycle were succinylated, and most were additionally modified by acetylation (28) Additionally, KGDHC stimulates fumarate hydratase activities and its succinylation levels, but it inhibits isocitrate dehydrogenase activity in the TCA cycle (6). The gathered investigation data indicates that the TCA cycle is the critical donor for global succinylation in cells through either a by guest on  succinyl-CoA-guided non-enzymatic mechanism or a KGDHC-guided enzymatic mechanism. In our examinations, all three KGDHC subunits in the TCA cycle were succinylated in plants. Given that conserved motifs and motif preferences were observed in different protein clusters of succinylated proteins and the patterns of artificially succinylated proteins were different from isolated cellular proteins in Western blot, our results suggest that protein lysine succinylation in developing rice is a highly regulated modification process instead of non-specific modifications.
Pyruvate dehydrogenase complex (PDC) catalyzes the conversion of pyruvate into acetyl-CoA.
Thus, PDC links the glycolysis metabolic pathway to the TCA cycle. We discovered 5 succinylation sites on the PDC E1 α subunit (PDHA, pyruvate dehydrogenase α subunit), 16 sites on the E2 subunit (dihydrolipoyl transacetylase, DLAT), and 11 sites on the E3 subunit (dihydrolipoyl dehydrogenase, DLD). Moreover, the E1 and the E3 subunits are additionally modified by acetylation (Fig. 4). In cancer cells, phosphorylation and acetylation play an imperative role in the joint regulation of PDHA activities. The varying acetylation status of PDHA directly changes PDC activities, which stimulates glycolysis in cancer cells (41,42). Furthermore, decreasing the succinylation levels by KGDHC inhibition leads to a substantial reduction of PDC activities and acetylation levels in neuronal cell lines (6). Down-regulation of succinylation levels may hinder PDC E2 (operating as acetyltransferase) activities, consequently preceding towards acetylation diminution (6). Broadening investigation is intriguing if succinylation and acetylation by guest on  work simultaneously in PDC and glycolysis pathway regulation in correspondence to phosphorylation and acetylation.

Succinylation of starch biosynthesis enzymes and regulators
In rice grain, the majority of accumulated carbohydrates is in starch form, significantly contributing to its nutritional values, milling properties, appearances, and cooking qualities (27).
The constituents and contents of storage starch are predominantly determined during seed development. We found that the enzymes in the starch synthesis and regulatory pathways, including ADP-glucose pyrophosphorylases (OsAGPS2 and OsAGPL2), starch branching enzymes (OsBEI and OsBEIIb), starch debranching enzymes (OsPUL), starch phosphorylase (OsPHOL), sucrose synthase (SUS2 and SUS3), UDP-glucose pyrophosphorylase (UGP) and phosphoglycerate mutase (PGM) were succinylated in developing rice seeds (Table 1 and Fig. 4). Moreover, many of these starch biosynthesis proteins were found subjected to acetylation, crotonylation, malonylation, and 2-hydroxyisobutyrylation (Fig. 4). OsAGPs catalyzes the initial step of starch biosynthesis, which has correlations with rice grain filling rates and starch accumulations (43). OsBEs, OsPUL, and OsPHOL catalyze amylopectin formation. Starch quality regulator FLO4 was similarly succinylated (Table 1). FLO4 mutation induced the formation of smaller grain dimensions and floury-white endosperms (44). Since starch biosynthetic enzymes are succinylated, it is commendable to intensify upcoming evaluations that delve into the concept of potential regulatory responsibilities of succinylation during starch synthesis in seed development.
by guest on March 6, 2020

Storage proteins are targeted by succinylation
Storage proteins and starches are both core components of rice endosperm. Seed storage proteins are not only critical as a nutritional reservoir for rice grain, but they also participate as vital machinery for seed germination, initial seedling development, endosperm structure, and culinary value (27,45). For our investigations, 13 proteins with storage specialization were succinylated, including 9 rice glutelin proteins, 3 globulin proteins, and 1 patatin-like storage protein ( Table 2).

Comparison of succinylome in different organisms
Global proteome analysis for lysine succinylation in germinating rice embryo identified 665  Table 3 and   Supplementary Table S3). We discovered that proteins for carbon metabolism, amino acid biosynthesis, and oxidative phosphorylation pathways are conserved in plant succinylomes.
Orthologs of 131, 75, 39, and 68 succinylated proteins of developing rice seed were detected in the succinylomes of Animalia, Fungi, Protista, and Bacteria, respectively. Moreover, 18 developing rice seed proteins contain succinylated homologs in all examined kingdoms. Among these conserved proteins, 11 proteins are classified as ATP synthases and DnaK family proteins.
Two are classified as aldehyde dehydrogenases with involvement in the cellular response to oxidative stress. One is classified as adenylate kinase enzyme with specializations in energy homeostasis and cell growth. Three are TCA cycle enzymes (OGDH, DLST, and IDH) which by guest on March 6, 2020 catalyze succinyl-CoA formation; DLST additionally operates as succinyltransferase (6). Ortholog detection in the succinylomes of various kingdoms provides noteworthy evidence of succinylation conservation across diversified organisms. -modified by succinylation, acetylation, malonylation, crotonylation, and 2-hydroxyisobutyrylation Emerging evidence proposes that PTMs may incorporate crosstalk and substitutions between alterations, which contributes to fluctuating functional consequences of proteins (47). Certain lysine acylation regulatory enzymes could have influences on various lysine acylation categories in a comparable manner (1,8,12,50,51). A plausible notion is that the active acetyltransferases randomly utilize acyl-CoAs to modify lysine substrates, and diverse lysine modifications may experience competition for protein lysine sites during specific biological regulations. Proper investigations were performed to determine if rice proteins, as well as induvial lysine sites, could have the capability of sheltering distinctive PTMs. We achieved comparative examinations of the developing rice seed succinylomes by utilizing acetylome (28), malonylome (29), crotonylome (Meng et al., unpublished result), and 2-hydroxyisobutyrylome (30). A remarkable PTM overlapping was detected on proteins in rice (Fig. 5a). We discovered 169 existing proteins in 15 dpa developing rice seeds were modified by both lysine succinylation and acetylation. This obtained data was accountable for approximately 48.7% (169/347) of succinylome reported here.

Proteins co
Additionally, 133, 250, and 300 succinylated rice proteins were modified by malonylation, crotonylation, and 2-hydroxyisobutyrylation, respectively. The quantity of proteins embraces two by guest on  and three particular PTM types are 1902 and 613, respectively (Fig. 5a). Remarkably, 67 rice proteins encompass the complete array of all the analyzed PTMs. These proteins are extensively involved in central carbon metabolism, amino acid biosynthesis, starch biosynthesis, and plant defense response mechanisms (Supplementary Table S4). Moreover, 85.1% of succinylated lysine sites identified in this study could also be modified by at least one of the other PTMs: acetylation, crotonylation, malonylation, and 2-hydroxyisobutyrylation (Fig. 5b). Interestingly, 14 lysine sites were found can harbor all the five types of PTMs in developing rice seed proteome (Fig. 5b). The PTM modification sites on representative protein phosphoglycerate kinase (Q6H6C7) are highlighted in Figure 5c.
Frequency of amino acid residues flanking lysines with at least three types of PTMs was analyzed to investigate the enrichment or depletion of various amino acids (Fig. 5d). Lysines that could be modified by multiple PTMs has a strong sequence preference for the charged amino acids of lysine (K) and glutamic acid (E) flanking the modification sites (Fig. 5d). Moreover, amino acids of D, I, V, H, F, Y, and R were found overly presented near to the modification sites as well (Fig. 5d). The discovery of multi-acylations at the same lysine sites and the patterned amino acid arrangement around the lysine residues ( Fig. 5b and d) suggested that the same or similar enzymes are involved in the establishment and removal of these different acylation moieties. A plausible notion is that active acetyltransferases utilize acyl-CoAs, based on their cellular concentrations and affinity, to modify lysine substrates; and the diverse lysine modifications, whose modification moieties come from the different metabolic pathway, may be subjected to competition for the lysine sites for specific biological regulations. The identification of substantial overlaps of by guest on  modifications at the same sites indicates that these modifications may interact and regulate biological processes via crosstalk. The possible competition, crosstalk, and functional interplay among different acylation moieties on the same sites of a protein require further investigation to validate.
To completely understand the protein regulatory mechanisms with multi-modifications, KEGG pathway enrichment analysis, and protein-protein interaction networks were created for rice proteins with at least three analyzed modifications (Fig. 5e-f). Proteins embracing multiple modifications are associated with numerous cell biological processes. The superior six proteinprotein interaction clusters are located in the highlighted portion of Figure 5f. Proteins engaging in interactions with ribosomes, protein processing in the endoplasmic reticulum, proteasome, glycolysis/gluconeogenesis, the TCA cycle, and oxidative phosphorylation provides evidence that cellular processes with strong connections to protein metabolism, carbon metabolism, and energy metabolism are subjected to various post-translational regulations.

CONCLUSION
Throughout this investigative study, 854 lysine succinylation sites on 347 proteins in developing rice seeds have been discovered. Western blot analysis provided significant evidence that lysine succinylation is a protein-specific reaction in developing rice seeds, and the global succinylation occupancy estimated from developing rice seed proteins was from 2% to 10%.