Identification of a Gene Expression Signature to Predict the Risk of Abdominal Aortic Aneurysm in Psoriasis Patients

Abstract

Background : Psoriasis is an immune-mediated, hereditary condition that presents itself in the skin or joints, or even both. Increasing evidence indicates that psoriasis is connected to an elevated risk of abdominal aortic aneurysm (AAA), owing to their shared inflammatory pathogenesis. Nevertheless, the interplay between psoriasis and AAA lacks sufficient documentation.

Methods: Through WGCNA and DEGs, psoriasis and AAA phenotype-related genes were identified. Identifying risk genes involved in both psoriasis and AAA involved generating candidate genes by finding the common intersection of hub genes, followed by using LASSO regression. Following this, a nomogram was created to forecast the development of psoriasis alongside AAA, and was then assessed through a ROC curve, DCA, calibration curve, and PR curve. Five algorithms, namely CIBERSORT, ssGSEA, ESTIMATE, MCPcounter, and QuanTIseq, were utilized to assess immune infiltration differences between high and low-risk groups. Simultaneously, we verified the differential gene expression in different tissues.

Results: A total of 1073 psoriasis hub genes and 128 AAA hub genes were generated. A Venn diagram revealed 20 candidate genes that were common to both hub genes of psoriasis and AAA. Of these, six genes (CCR7, CD3D, GBP5, HCLS1, IL7R, and ITGAL) were identified as risk genes. The gene signature generated by these genes demonstrated high accuracy in predicting psoriasis and AAA. Using five algorithms for immune infiltration analysis, an abundance of inflammatory cells was observed in high-risk subgroups. The above six genes were found to be highly expressed in both psoriasis tissue and abdominal aortic aneurysm tissue.

Conclusions : The study resulted in the identification of a novel gene signature, including six high-risk genes, that has enhanced our knowledge of the common causes and control mechanisms of psoriasis and AAA. These findings are anticipated to pave the way for promising therapeutic targets in mitigating the comorbidities of cardiovascular disease.

Keywords: abdominal aortic aneurysm; nomogram; psoriasis; signature.

Figure 1

Landscape of DEGs in psoriasis. (A) Differential gene expression analysis showing up-and down-regulated genes across ten datasets. An adjusted P value < 0.05 and log2FC ≥ 1 is indicated in red, while an adjusted P value < 0.05 and log2FC ≤ 1 is indicated in blue. (B) Upset diagram showing the intersection across ten datasets.

Figure 2

Functional enrichment of DEGs. The top 30 GO enrichment analyses of BP (A), MF (B), and CC (C). The q value is scaled by color and gene count is represented by the size of the circle. (D) The top 12 KEGG pathway enrichment of GSEA analysis with P value < 0.05.

Figure 3

Identification of hub genes in psoriasis. (A) The cluster dendrogram of gene modules in psoriasis samples. (B) The correlation heatmap indicates module-trait relationships. The first-row number in each block indicates the correlation coefficient, which is scaled by color. The number in parentheses represents the P value. (C) Scatter diagram of yellow, pink, turquoise, blue, and green modules. The x-axis represents module membership in each module, and the y-axis represents gene significance for psoriasis. (D) Venn diagram indicates the intersecting gene between DEGs and module hub genes in psoriasis. Gene count is scaled by color. Gene percentages are displayed below the gene count.

Figure 4

Identification of hub genes between psoriasis and AAA. (A) PCA diagram visualizes the two-dimensional distribution of the principal component of AAA and control samples. (B) The volcano plot of GSE57693 indicates the distribution of DEGs. AAA, abdominal aortic aneurysm. Control, normal aortic arteries. Genes in red represent significantly up-regulated, blue represents significantly down-regulated, brown represents P value < 0.05, gray represents no significance, light red represents log2FC ≥ 1, and light blue represents log2FC ≤ 1. (C) The cluster dendrogram of gene modules in AAA samples. (D) The correlation heatmap indicates module-trait relationships. The first-row number in each block indicates the correlation coefficient, which is scaled by color. The number in parentheses represents the P value. (E) Venn diagram indicates the intersecting gene among DEGs in AAA, module hub genes in AAA, and hub genes in psoriasis. Gene count is scaled by color.

Figure 5

Establishment of gene signature by LASSO regression. (A and B) Heatmap of 20 hub genes in GSE226244 (A) and GSE57691 (B). Gene expression is scaled by the Z-score. The bar plot beside the heatmap reveals the log2FC of genes. (C) The PPI network constructed by STRING reveals the interaction between these 20 genes. (D) Friend analysis discovers the rank of 20 genes. The x-axis indicates the gene similarity. The y-axis represents genes. (E) The log (lambda) sequence was used to construct a coefficient profile diagram. (F) LASSO coefficient profiles of the 20 genes in psoriasis. (G and H) ROC curve of CCR7, CD3D, GBP5, HCLS1, IL7R, ITGAL, and the gene signature. (I) The visible nomogram for diagnosing psoriasis.

Figure 6

Diagnostic value evaluation of gene signature. (A–C) The ROC curves of testing sets (GSE79704, GSE13355, and GSE63741) demonstrate satisfactory accuracy of gene signature predicting psoriasis. (D) Robust accuracy of gene signature in predicting AAA risk emerges in the training set (GSE57691) and testing set (GSE47472), followed by ROC curves (D and H), DCA curves (E and I), calibration curves (F and J) and PR curves (G and K). AUC: area under the curve. CI: 95% confidence interval. (L) Histogram indicates the differential immune infiltration of GSE226244 between high-risk and low-risk subgroups. The y-axis represents the percentage of immune cells.

Figure 7

Investigation of candidate gene expression across diverse tissue types. (A) Protein extraction from normal abdominal aorta (n=3) and abdominal aortic aneurysm specimens (n=5), followed by Western blot analysis. (B) Protein extraction from normal skin tissue (n=3) and psoriatic plaque specimens, followed by Western blot analysis (n=5). (C) Utilizing RT-qPCR experiments to analyze RNA extracted from normal abdominal aorta (n=3), abdominal aortic aneurysm (n=5); normal skin (n=3), and psoriatic plaque tissues (n=5).

Figure 8

Analysis of gene expression in different tissues. (A) Perform immunohistochemical staining on normal abdominal aorta, abdominal aortic aneurysm, normal skin, and psoriatic plaque tissues to observe the expression differences of six candidate genes (scale bars: IHC, 250μm; n = 3 per group).