Targeted proteomics reveals immune protein patterns in serum from patients with calcific aortic stenosis

Figures

• 

  Figure 1. Boxplots showing the distribution of the Normalized Protein eXpression units for proteins, which significantly differ between groups, individuals with aortic valve calcification (CAVD, in dark blue), and controls (CTR, in light blue) See also Table 2 for specifications and paired comparisons. Data are represented as boxplots (median and interquartile range, IQR). Each box extends from the first quartile to the third quartile (interquartile range, IQR), while the horizontal line within each box represents the median; whiskers indicate values within 1.5× IQR from the quartiles. p values were computed using the Limma method and adjusted with the Benjamini-Hochberg procedure for multiple testing correction (Adj - p). PRDX1, peroxiredoxin 1; PRDX5, peroxiredoxin 5; AREG, amphiregulin; GLB1, beta-galactosidase; TRIM5, tripartite motif-containing protein 5; TRIM21, E3 ubiquitin-protein ligase; PLXNA4, Plexin-A4; ICA1, islet cell autoantigen 1; BTN3A2, butyrophilin subfamily 3 member A2; and CLEC4C, C-type lectin domain family 4 member C.

• 

  Figure 2. Protein-protein network illustrating the physical interactions between seven proteins that were significantly downregulated in patients with CAVD Larger nodes represent the candidate protein biomarkers, while smaller nodes denote interacting proteins. Node colors indicate the clusters to which they belong, as determined by MCL clustering. Dashed edges represent physical interactions connecting proteins from different MCL-defined clusters, as derived from the original STRING network. These edges were not added manually but reflect genuine cross-cluster interactions.

• 

  Figure 3. Validation of amphiregulin and Plexin-A4 levels by ELISA Histograms show serum concentrations of amphiregulin (AREG, left) and Plexin-A4 (PLXNA4, right) measured by quantitative ELISA in control subjects (CTR) and patients with calcific aortic valve disease (CAVD). Both proteins were significantly reduced in patients compared to controls (∗∗∗∗p < 0.0001 by unpaired two-tailed Student’s t test). Bars indicate mean values with standard deviation (mean ± SD). Asterisks indicate levels of statistical significance: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

• 

  Figure 4. The heatmap displays mRNA expression levels of the candidate biomarkers across various human tissues Dark blue colors represent higher expression levels, while light yellow colors indicate lower levels. Genes and tissues were arranged using unsupervised hierarchical clustering.

• 

  Figure 5. Identification of the three clusters (A) UMAP plot where each node represents a sample. Dark blue nodes correspond to individuals with aortic valve disease, while light blue nodes represent controls. This color scheme is consistent across all three images. (B) Heatmap displaying scaled NPX values. Red indicates high levels, while blue represents low levels. Both samples and genes are ordered using Ward’s hierarchical clustering, based on Euclidean distance. (C) Expression profiles of the three identified clusters. The black line represents the mean levels trend of each protein within the cluster.

• 

  Figure 6. Analysis of the biomarker potential using ROC methodology for patients with CAVD; ROC curves were ordered according to descending values of AUC. Statistical significance of each AUC is indicated by p-values (p).