Plasma protein profiling in patients undergoing coronary artery bypass grafting surgery and clinical significance

Figures

• 

  Figure 1. Heatmap from the results of iTRAQ This reflects the protein expression values in different groups and functional cluster analysis of different proteins. The protein expression was compared in male and female, respectively. Proteins with similar functions have a relative shorter Euclidean distance. The categorical annotation bar (left of heatmap) demonstrate the annotation of protein expression values. The right side shows the name of proteins. The red and green colors represent up- and down-regulated proteins, respectively. The numbers in the left lower coner gives the fold of changes of a protein, correlating to the color shown beside the number.

• 

  Figure 2. Protein ratio distribution in coronary artery disease, according to the result of iTRAQ X-axis presents differential ratio log-transformed with base 2. All differential proteins including up-regulation (triangles in red) and down-regulation (triangles in green) with the change greater than 1.2 fold compared to the control are included in the figure. However, among all these proteins, only 35 up-regulated and 41 down-regulated differential proteins were statistically significant (CABG vs. control, p < 0.05) as mentioned in the text. These proteins are listed in the supplemental Table S1 in details.

• 

  Figure 3. Protein-Protein interaction network STRING, a widely used biological database, was utilized as a search tool for the Retrieval of Interacting Genes/Proteins. The network model is generated with cytoscape web application, based on information gained up to 4 level of functional analysis: fold change of gene/protein, protein-protein interaction, KEGG pathway enrichment and biological process enrichment. Circle nodes showed the identified up- and down-regulated proteins displayed in red and green, respectively. Pathway were colored with gradient color from yellow to blue, yellow for smaller p-value, blue for bigger p-value. The gray arrows with solid and dashed lines indicated the interaction with bigger and smaller confident score, respectively between the molecules. The Figure suggests that complement and coagulation cascade, biosynthesis of amino acids, exec, are possibly linked with CAD-related protein changes.

• 

  Figure 4. ELISA validation ELISA test for platelet factor 4 (PF4, A), secreted frizzled-related protein 1 (sFRP1, B), and coagulation factor XIII B chain (F13B) in CABG group (n = 160) and control group (n = 160). The plasma level of PF4 and sFRP1 were up-regulated significantly (CABG vs. control, p < 0.05). However, there were no significant differences of F13B (C) between two groups (p = 0.145).

• 

  Figure 5. Comparison of plasma levels of differential proteins between males and females There is a significant difference of sFRP1 between two genders, i.e. female patients represented a relatively higher level of sFRP1 (B, also see Table 5).

• 

  Figure 6. ROC tests ROC tests for PF4 and sFRP1 for the diagnosis of the severity of CAD (n = 240 including 120 CABG patients and 120 controls). Areas under the curve are 0.858 and 0.788 for PF4 (A) and sFRP1 (B), respectively (p < 0.05). The best differential point of PF4 was 1021.00 with sensitivity of 77.4% and specificity of 93.3% (YI = 0.707). The best differential point of sFRP1 was 1581.625 (YI = 0.507). The sensitivity and specificity of sFRP1was 92.4%, 58.3%, respectively. ROC curves help selecting optimal differential point for a biomarker or a test during diagnosis.