Molecular genetic study of novel biomarkers for early diagnosis of oral squamous cell carcinoma

Abstract

Objectives: Early detection and treatment of an oral squamous cell carcinoma (OSCC) is critical because of its rapid growth, frequent lymph-node metastasis, and poor prognosis. However, no clinically-valuable methods of early diagnosis exist, and genetic analysis of OSCCs has yielded no biomarkers.

Study design: We investigated the expression of genes associated with inflammation in OSCCs via a quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) analysis of microarray data. Tumor and normal tissues from five patients with an OSCC were used for microarray analysis. Differentially-expressed genes, identified using permutation, local pooled error (LPE), t-tests, and significance analysis of microarrays (SAM), were selected as candidate genetic markers.

Results: Two groups corresponding to tissue identity were evident, implying that their differentially-expressed genes represented biological differences between tissues. Fifteen genes were identified using the Student's paired t-test (p<0.05) and the SAM, with a false discovery rate of less than 0.02. Based on gene expression, these 15 genes can be used to classify an OSCC. A genetic analysis of functional networks and ontologies, validated by using a qRT-PCR analysis of the tissue samples, identified four genes, ADAM15, CDC7, IL12RB2 and TNFRSF8, that demonstrated excellent concordance with the microarray data.

Conclusions: Our study demonstrated that four genes (ADAM15, CDC7, IL12RB2 and TNFRSF8) had potential as novel biomarkers for the diagnosis and the treatment of an OSCC.

Figure 1

Hierarchical clustering of microarray gene expression data. The dendogram at the bottom lists all samples arrayed and measures their degree of relatedness in terms of gene expression. All samples were coded with numbers, as shown in table 1. The colored bar beneath the sample identifiers marks samples from patients, where the Normal Group of patients with oral squamous cell carcinoma are in red and those from the Tumour Group are in blue (A). Genes identified by all four statistical tests (the permutation, LPE and t-tests and SAM) were selected as candidate genetic markers. The 15 genes passed Student’s paired t-test and SAM analysis (B).

Figure 2

Differential expression of 15 genes selected by microarray in oral squamous cell carcinoma and normal tissue. Five genes were downregulated, while the other 10 were upregulated. The full line corresponds to the median value for each group.

Figure 3

Quantitative reverse transcription polymerase chain reaction validation of microarray gene expression data from tested individuals from the Tumour Group. The quantitative reverse transcription polymerase chain reaction tests were performed with primer sets specific to ADAM15, CDC7, IL12RB2 and TNFRSF8 in 25 patients with oral squamous cell carcinoma and 20 normal matching samples. p-values (Student’s t-test) are presented. (*P < 0.05; **P < 0.01; ***P < 0.001)

Figure 4

Bar chart representation of the mRNA expression of selected genes in tumour and normal patient-matched specimens (tumour–normal-paired). Normalised expression was determined by 2-ΔΔCt, by comparing threshold cycle values of ADAM15 with β-actin. Overexpression of all genes in tumour tissues is clearly demonstrated. Number of cases analysed and corresponding p-values (paired t-test) are provided. (*P < 0.05; **P < 0.01; ***P < 0.001)

Figure 5

Receiver operating characteristic (ROC) curve for comparing diagnostic validity of the candidate genes. The area under the curve (AUC) of ADAM15, CDC7, IL12RB2 and TNFRSF8 were 0.699, 0.645, 0.561 and 0.602, respectively. Among them, only ADAM5 showed statistical significance (p < 0.05).