Expression profiling of cutaneous squamous cell carcinoma with perineural invasion implicates the p53 pathway in the process

Abstract

Squamous cell carcinoma (SCC) is the second most common cancer worldwide and accounts for approximately 30% of all keratinocyte cancers. The vast majority of cutaneous SCCs of the head and neck (cSCCHN) are readily curable with surgery and/or radiotherapy unless high-risk features are present. Perineural invasion (PNI) is recognized as one of these high-risk features. The molecular changes during clinical PNI in cSCCHN have not been previously investigated. In this study, we assessed the global gene expression differences between cSCCHN with or without incidental or clinical PNI. The results of the analysis showed signatures of gene expression representative of activation of p53 in tumors with PNI compared to tumors without, amongst other alterations. Immunohistochemical staining of p53 showed cSCCHN with clinical PNI to be more likely to exhibit a diffuse over-expression pattern, with no tumors showing normal p53 staining. DNA sequencing of cSCCHN samples with clinical PNI showed no difference in mutation number or position with samples without PNI, however a significant difference was observed in regulators of p53 degradation, stability and activity. Our results therefore suggest that cSCCHN with clinical PNI may be more likely to contain alterations in the p53 pathway, compared to cSCCHN without PNI.

Figure 1. Expression profiling of cSCCHN, cSCCHN with incidental PNI or clinical PNI.

(a) Genes found to be differentially expressed between the three groups using one-way ANOVA, P < 0.05, were clustered using Euclidian distance metric and Ward’s linkage, for visualisation. Red – cSCCHN, blue – cSCCHN with Incidental PNI, maroon – cSCCHN with Clinical PNI. Scale below shows difference in fold expression. (b) Venn diagram comparison of differentially expressed genes for pairwise analysis using t-test, P < 0.05 with Benjamini and Hochberg False Discovery Rate. (c) Mechanistic network of upstream regulator signature for differences observed in p53 signaling pathway. (d) Mechanistic network of upstream regulator signature for differences observed in p53 signaling pathway. Both mechanistic network figures were taken from Ingenuity Pathway Analysis software.

Figure 2. Differential expression of known p53- or MYC-target genes.

Selected known target genes of (a) p53 or (b) MYC were taken from the expression profiling dataset, and plotted as box and whisker (maximum and minimum) plots. Groups are as indicated, showing normalized gene expression (log2) of each of the tumors included in the DASL analysis. Data was analyzed using one-way ANOVA with multiple comparisons. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Figure 3. Immunohistochemical detection of p53 protein in cSCCHN with clinical PNI.

Representative images of p53 staining. (a) A tumor with a negative pattern of p53 expression; no p53 expression is visualized in any of the tumor cells nuclei; (b) The same pattern shown at higher magnification; (c) A tumor with a focal pattern of p53 expression; strong p53 expression is evident, but within restricted areas of the tumor tissue; positively stained cells are typically immersed in a background of negative or weakly positively stained cellular nuclei; (d) The same pattern shown at a higher magnification; (e) A tumor displaying a diffuse pattern of p53 protein expression; very strong positive p53 expression scattered extensively throughout the tissue; (f) The same pattern shown at a higher magnification. Nuclear p53 is visualized in red. Scale: a, c, e: 500 μM, 40×; b, d, f: 50 μM, 400×.

Figure 4. Differential expression of regulators of p53 degradation, stability and activity in cSCCHN with clinical PNI.

Selected target genes known to play roles in degradation, stability and activity of p53 were taken from the expression profiling dataset, and plotted as box and whisker (maximum and minimum) plots. Groups are as indicated, showing normalized gene expression (log2) of each of the tumors included in the analysis. Data was analyzed using one-way ANOVA with multiple comparisons. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.