Quantitative comparison of the crotonylome of VSMCs before and after PDGF-BB treatment
It is demonstrated that PDGF-BB could regulate VSMC differentiation at the concentration of 10 ng/ml. Therefore, VSMCs were incubated with PDGF-BB (10 ng/ml) for 1, 3, 6, 12, 24, 48 and 72 h. As shown in Fig. 1f, the pan-crotonylated level increased at 3 h after PDGF-BB-stimulated and peaked at 24 h. Moreover, the expression of the smooth muscle contractile proteins SM α-actin and SM22α decreased at 12 h, indicated VSMC transformation toward a proliferative phenotype at 12 h after PDGF-BB-stimulating (Fig. 1g). For other types of post-translational modifications, with the prolonged time of PDGF-BB stimulation, there is no consistency in the level of post-translational modification and the phenotypic transformation (Fig. 1a-e).
Therefore, the proteins derived from VSMCs before and after PDGF-BB treatment were separately identified using liquid chromatography and mass spectrometry (LC-MS/MS)-based quantitative analysis within three random repetitions according to the experimental flowchart (Fig. 2a).
For biological replicates or technical replicates, we test whether the quantitative results of biological replicates or technical replicates are statistically consistent. The heat map shows an obvious positive correlation between sample replicates and a negative correlation between different treatment groups (Fig. 2b).
In this work, 2386 crotonylation sites corresponding to 570 proteins were identified and 2138 crotonylation sites from 534 proteins were quantified by the comparison between VSMCs with and without PDGF-BB treatment (Supplementary information Table 1).
The heat maps show that there is a high similarity between duplicate samples and the protein expression and modification sites have changed significantly after PDGF-BB treatment (Fig. 2c/d).
To validate the MS data, we then checked the mass error of the identified peptides. The distribution of mass errors was near zero, and most errors were < 0.02 Da, indicating the accuracy of the MS data (Fig. 2e).
Among the 534 quantifiable proteins, 245 crotonylation sites from 125 proteins changed over 1.3 folds (200 sites from 94 proteins up-regulated and 45 sites from 31 proteins down-regulated) (Fig. 2f). We also performed statistics on the proteins and sites corresponding to other fold differences (Supplementary information Table 2).
Of these 570 crotonylated proteins, 233 (40.9%) have only one crotonylation site, 94 (16.5%) contain two sites, and 67 (11.8%) possess three. 58 (10.2%) proteins get 10 more crotonylation sites (Fig. 2g), such as pyruvate kinase, which includes 16 crotonylated sites (Supplementary information Table 3).
To determine whether there are any specific amino acid biases adjacent to crotonylation sites, we analyzed the flanking amino acids of these sites, from − 10 to + 10 positions of lysine. This analysis showed that there were a total of 2731 peptides containing crotonylation sites. Further characterization of the motifs surrounding lysine crotonylation sites showed a series of distinct patterns of crotonylation motifs (Supplementary information Table S4). Among them, the peptides included 10 specific and conserved motifs, such as *********KcrL**********(368 peptides), **********KcrE********* (334 peptides), *********EKcr********** (331 peptides) and so on (Figure S1). Crotonylation is the meaning of crotonylated lysine, L leucine, E glutamic acid, respectively. “*” represents any single amino acid. Motif enrichment heat map showed glutamic acid (E) and aspartic acid (D) residues were over-enriched at the − 1 and + 1 positions surrounding the lysine crotonylation sites (Fig. 2h). The distribution of other residues is as follows: alanine (A) is at -1/+4 positions, phenylalanine (F) is at -1/+1, isoleucine (I) and leucine (L) are at -3/-4, arginine (R) is at + 2, and tyrosine (Y) is at -1. The crotonylation of lysine neutralizes its positive charge and weakens its electrostatic interaction with surrounding negatively charged amino acids, such as E and D. These data showed for the first time that numerous non-histone proteins are modified by crotonylation in VSMCs.
The prediction and classification statistics of the subcellular structure of the protein corresponding to the differential modification site and the functional classification of the protein
To further evaluate crotonylated proteins, we classified these proteins according to their subcellular location, cellular component, molecular function, and associated biological processes.
In the up-regulated crotonylated proteins, 46% were in the cytoplasm, 21% were in the nucleus, and 13% were in the mitochondria (Fig. 3a). On the contrary, in the down-regulated crotonylated proteins, 32% were located in the cytoplasm, 26% were located in the nucleus, and 13% were in the mitochondria (Fig. 3b).
The cellular component of crotonylated proteins was classified according to the gene ontology (GO) annotation. The up-regulated crotonylated proteins were distributed in the cell (22%), the organelles (22%), the macromolecular complexes (14%), membranes (13%), and extracellular region (11%) (Fig. 3c). However, the down-regulated crotonylated proteins were distributed in the cell (23%), the organelles (23%), membranes (16%), extracellular region (16%), and the macromolecular complexes (9%) (Fig. 3d).
The molecular function of these crotonylated proteins was analyzed based on the GO annotation. Crotonylated proteins were assigned to several groups. In the up-regulated crotonylated proteins, 53% were involved in binding, 19% were related to structural molecule activity, and 18% were associated with catalytic activity (Fig. 3c). In the down-regulated crotonylated proteins, 60% were involved in binding, 15% were associated with catalytic activity, and 9% were related to structural molecule activity (Fig. 3d).
Biological process analysis of crotonylated proteins showed that 18% of up-regulated crotonylated proteins were involved in cellular processes, 13% were in the single-organism process, 13% were in the metabolic process and 10% were in biological regulation (Fig. 3c). On the contrary, 17% of up-regulated crotonylated proteins were involved in cellular processes, 16% were in the single-organism process, 13% were in biological regulation and 9% were in response to the stimulus (Fig. 3d). These results indicate the relatively broad distributions of crotonylated proteins in subcellular location, cellular component, molecular function, and biological processes.
Functional enrichment-based clustering analysis for the quantified crotonylome
To investigate the roles of the crotonylation pathways, enrichment analysis was used to classify all the crotonylated protein groups based on the GO and KEGG databases. GO enrichment analysis of the molecular function showed that up-regulated crotonylated proteins were significantly enriched in the actin-binding, cell adhesion, molecule binding, and structural constituent of ribosome (Fig. 4a). The down-regulated crotonylated proteins were significantly enriched in calcium ion binding, metal ion binding, and cation binding (Fig. 4b). GO-based functional enrichment analysis of cellular components showed that up-regulated crotonylated proteins were significantly enriched in the structural constituents of ribosomes, cytoskeleton, and anchoring junction (Fig. 4a). The down-regulated crotonylated proteins were significantly enriched in the extracellular vesicle, extracellular exosome, and extracellular organelle (Fig. 4b). Enrichment analysis of biological processes indicated these up-regulated modified proteins were enriched in a variety of biological processes including protein folding, actin cytoskeleton organization, and actin filament-based process (Fig. 4a). These down-regulated modified proteins were enriched in a variety of biological processes including protein heterooligomerization, heterotetramerization, and oligomerization (Fig. 4b). GO annotation analyses suggest that crotonylated proteins participate in a variety of important cellular pathways and perform different functions in VSMCs.
KEGG enrichment analysis showed that the up-regulated crotonylated proteins were related to multiple metabolic pathways, including ribosome, HTLV-1 infection, biosynthesis of amino acids, glucagon signaling pathway, and glycolysis/gluconeogenesis (Fig. 5a). The pathways involved in down-regulated proteins include multiple metabolic pathways, including arginine and proline metabolism, endoplasmic reticulum, endocytosis, bacterial invasion of epithelial cells, and viral carcinogenesis (Fig. 5b). Among them, some proteins were found to be closely involved in the physiological process of VSMC phenotypic remodeling including glycolysis/gluconeogenesis, vascular smooth muscle contraction, or PI3K-Akt signaling pathway. Differential protein domain enrichment analysis results showed that many up-regulated proteins had the same amino acid sequence as the calponin homology (CH) domain (Fig. 5c). CH domain is the most common shared amino acid sequence of various actin-binding proteins, which is composed of 6 alpha-helices. The CH domain plays a stabilizing role in cytoskeletal dynamics and can activate downstream pathways in signal transduction[32]. In the present study, there are 6 proteins (Transgelin, Calponin, Microtubule-associated protein RP/EB family member 1, Plectin, Filamin A, IQ motif containing GTPase activating protein 1) containing CH domain were extensively crotonylated at multiple sites.
Through the KEGG pathway enrichment analysis, the abundances of 21 ribosomal proteins (12 large subunit ribosomal proteins and 9 small subunit ribosomal proteins) involved in the ribosome pathway were differentially expressed in the with and without PDGF-BB-stimulated VSMCs (Fig. 5d), which indicated their possible involvement in PDGF-BB induced VSMC phenotypic remodeling.
For the enrichment of protein complex, we found one statistically significant SDH-mABC1-PIC-ANT-ATPase complex (Fig. 5e), which was in the mitochondrial inner membrane and transported K+ with characteristics similar to mitoKATP. This suggested that differentially expressed proteins may be involved in cell mitochondrial activity and thus affected phenotypic transformation.
Cluster analysis of functions of proteins with different multiples of difference
To find the correlation of the functions of proteins with different differential expression fold changes, we divide them into 4 parts according to their differential expression multiples, including Q1 to Q4: Q1 (0 < Ratio ≤ 1/1.5), Q2 (1/1.5 < Ratio ≤ 1/1.3), Q3 (1.3 < Ratio ≤ 1.5) and Q4 (Ratio > 1.5) (Ratio = PDGF/Con) (Fig. 6a). The statistical results show that after PDGF-BB treatment, the number of modification sites in the up-regulated proteins increased significantly.
Then, we performed enrichment and cluster analysis of GO, KEGG, and protein domains for each Q group. GO enrichment for the Q3 group showed up-regulated proteins were involved in many biological processes, such as response to hypoxia, regulation of protein metabolic process, regulation of endocytosis, positive regulation of cell proliferation, and so on (Fig. 6b). And these up-regulated proteins are mainly distributed on the cell surface, cytoskeleton, cytoplasm, and ribosomes (Fig. 6c). The molecular function analysis of the up-regulated proteins showed they can perform corresponding functions by combining with cell adhesion factors, RNA, cadherin, ADP (Fig. 6d).