Figure 1.
Histologically, a mixture of keratinous material and stratified squamous epithelium is required to diagnose cholesteatoma. The epithelium displays stratified squamous epithelium with keratinization (matrix). The subepithelial region is occupied by connective tissue. Abbreviations: C: stratum corneum; G: stratum granulosum; S: stratum spinosum; B: stratum basale; D: dermis. Magnification, 600x.
Figure 2.
Up-regulated and down-regulated genes in cholesteatoma.
(A) mean log expression value plot. The most prominent hits are shown in red. Microarray analysis revealed that the expression of genes, such as PI3, SPRR1B, LCN2 (lipocalin 2), MMP1, MMP9, MMP10, MMP12, SPP1, GJB2, Bcl-xl (BCL2L1), cIAP2 (BIRC3), S100A7A (koebnerisin), S100A9, SERPINB3, CEACAM6, KRT6B and SERPINB4 is higher in cholesteatoma compared to external auditory canal skin. Genes like TFAP2B, CDH18, CDH19, PRPH, ID4, PAX3, LAMC2, SP5 and FGFBP2 were significantly down-regulated. (B) Up-regulated and down-regulated genes in cholesteatoma, according to biological function. Biological processes, which are denoted with ‘others’ are processes with less than 10 genes involved.
Figure 3.
Heatmap of log expression values.
For the heatmap generation, significantly regulated genes with a fold change of five were used. 116 up-regulated and 29 down-regulated genes are shown. The clustering reveals that the mRNA profiles of control samples (EAS1-EAS7) show differences to cholesteatoma samples (Chole1-Chole7) relating to the expression of inflammation related genes such as metalloproteinases (e.g. MMP9). The strong co-expression of MMP9 and its known substrate SPP1 (osteopontin 9) might be important. Furthermore, there was a group of patients with low expression of both MMP9 and SPP1, but high expression of genes such as S100A and GJB2.
Figure 4.
Real-time analysis of candidate genes.
(A) The Real-time PCR analysis of potential transcripts that seem to be involved in major processes in cholesteatoma. For normalization the housekeeping gene GAPDH was used. The real-time analysis and comparison of the expression of different genes in external auditory canal skin and cholesteatoma confirms the results of the microarray data. In the present study, genes that encode for MMPs and their substrate were highly up-regulated in cholesteatoma; transcripts that encode e.g. for tumor suppressors were down-regulated. (B) Real-time analysis of apoptosis-related genes. In cholesteatoma genes for the anti-apoptotic proteins Bcl-xl (BCL2L1), A20 (TNFAIP3) and cIAP2 (BIRC3) show a somewhat higher expression. In which cIAP2 (BIRC3) appear to be down-regulated and genes for the proteins p65 (RELA), cFlip (CFLAR) and Bcl-2 show the same expression pattern in cholesteatoma and external auditory skin.
Figure 5.
Immunohistochemical staining profiles obtained with CK 5/6 (A) and CK14 (B) antibodies.
Using antibodies against Cytokeratin 5/6 and Cytokeratin 14, strong immunoreactivity was detected in the stratum basale and adjacent suprabasal structures, whereas CK5/6 is increased in the pars tensa of cholesteatoma. CK14, which is used as a marker of keratinizing squamous epithelium is synthesized in the basal epithelia cells. (C) Statistical evaluation of CK5/6 expression in cholesteatoma stratum granulosum, stratum spinosum and stratum basale compared to the expression within external auditory skin revealing significantly higher expression in cholesteatoma. p<0.05. (D) Comparison of CK14 expression in cholesteatoma and healthy auditory skin. The basal epithelia cells of cholesteatoma show significantly higher expression of CK14 compared to external auditory skin. p<0,04; Abbreviations: C: stratum corneum; G: stratum granulosum; S: stratum spinosum; B: stratum basale; D: dermis; PM: peri matrix; Magnification: 400x.
Figure 6.
Bioinformatical network analysis for cytokeratin.
Proteins are the nodes and the edges are protein activation/inactivations, such as phosphorylation and dephosphorylation across a set of proteins. The cytokeratin analysis reveals an up-regulation of Keratin 6 A and B, Keratin18, Keratin 19 and Keratin 81. In our data, genes for Keratin 15, FAM107A, BBS1 and Peripherin were down-regulated in cholesteatoma. Colors: red: up-reglated genes within cholesteatoma compared to external auditory skin; green: down-regulated genes within cholesteatoma compared to external auditory skin; grey: the gene for the protein is not in the microarray; Node-Size: The “Node Ranking” describes the number of interactions. The bigger the nodes, the more interactions are represented for the protein in the network.
Figure 7.
Bioinformatical network analysis for S100.
Proteins are the nodes and the edges are protein activation/inactivations, such as phosphorylation and dephosphorylation across a set of proteins. The S100 Network shows an up-regulation of S100A7, S100A8 and S100A9 as well as ILK and IκB, USF2 and ARRB2. In which NUAK1 seems to be down-regulated. Different colors showing the expression of genes with the cholesteatoma compared to external auditory skin: red: up-regulated; green: down-regulated; grey: the gene for the protein is not in the microarray; Node-Size also called “Node Ranking” describes the number of interactions. The bigger the nodes, the more interactions are represented for the protein in the network.
Figure 8.
Bioinformatical network analysis for MMPs
. Proteins are the nodes and the edges are protein activation/inactivations, such as phosphorylation and dephosphorylation across a set of proteins. The network presented here reveals an up-regulation of the matrix metalloproteinase family members MMP1, 7, 9, 10, 12 as well as LCN2, IL-8 and CXCL1. Down-regulated genes within this network are RECK, COL4A5 and COL4A6. The colors of the nodes describe the regulated expression of genes within cholesteatoma compared to external auditory skin: red: up-regulated genes within cholesteatoma; green: down-regulated genes within cholesteatoma; grey: the gene for the protein is not in the microarray; The “Node Ranking” (Node-Size) describes the number of interactions. The bigger the nodes, the more interactions are represented for the protein in the network.