Whole genome expression profiling reveals a significant role for immune function in human abdominal aortic aneurysms

Abstract

Background: Abdominal aortic aneurysms are a common disorder with an incompletely understood etiology. We used Illumina and Affymetrix microarray platforms to generate global gene expression profiles for both aneurysmal (AAA) and non-aneurysmal abdominal aorta, and identified genes that were significantly differentially expressed between cases and controls.

Results: Affymetrix and Illumina arrays included 18,057 genes in common; 11,542 (64%) of these genes were considered to be expressed in either aneurysmal or normal abdominal aorta. There were 3,274 differentially expressed genes with a false discovery rate (FDR) </= 0.05. Many of these genes were not previously known to be involved in AAA, including SOST and RUNX3, which were confirmed using Q-RT-PCR (Pearson correlation coefficient for microarray and Q-RT-PCR data = 0.89; p-values for differences in expression between AAA and controls for SOST: 4.87 x 10-4 and for RUNX3: 4.33 x 10-5). Analysis of biological pathways, including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG), indicated extreme overrepresentation of immune related categories. The enriched categories included the GO category Immune Response (GO:0006955; FDR = 2.1 x 10-14), and the KEGG pathways natural killer cell mediated cytotoxicity (hsa04650; FDR = 5.9 x 10-6) and leukocyte transendothelial migration (hsa04670; FDR = 1.1 x 10-5).

Conclusion: Previous studies have provided evidence for the involvement of the immune system in AAA. The current expression analysis extends these findings by demonstrating broad coordinate gene expression in immunological pathways. A large number of genes involved in immune function were differentially expressed in AAA, and the pathway analysis gave these results a biological context. The data provide valuable insight for future studies to dissect the pathogenesis of human AAA. These pathways might also be used as targets for the development of therapeutic agents for AAA.

Figure 1

Heatmap of the 100 most differentially expressed genes in AAA. To show consistency of the results between the two platforms, both Illumina (15 left columns) and Affymetrix (4 right columns) results are presented. Sample IDs are on top (see Table 1 for details) and gene symbols on the right. A listing of these 100 genes with expression values in both conditions as well as the FDR values is provided [see the additional file 2: Table II].

Figure 2

Q-RT-PCR results for RUNX3 and SOST in AAAs (n = 10) and control abdominal aortas (n = 10). Larger values of Max – ΔCT represent larger amounts of RNA present, and a change of this value by 1 represents a two-fold increase in the amount of mRNA. Medians are indicated by horizontal lines. See Table 1 for details on samples analyzed.

Figure 3

Interactions of KEGG biological pathways for genes differentially expressed in AAA. The 21 most significant pathways are shown (see Table 2 for details). Areas represent pathway size (number of genes) and darker shades indicate the proportion of differentially expressed genes of the total number of genes in the pathway. See Methods section for details on how the interactions were identified.

Figure 4

Modified "natural killer cell mediated cytotoxicity" (hsa04650) pathway from KEGG. Protein symbols were replaced by gene symbols to reflect gene-centric data. See key for explanation of colors and symbols. Gene symbols were not italicized for better legibility [see additional file 4: Table III for expression values of the genes in both conditions as well as the FDR value for those that were significantly different].