Deregulation of SYCP2 predicts early stage human papillomavirus-positive oropharyngeal carcinoma: A prospective whole transcriptome analysis

Abstract

This study was designed to identify significant differences in gene expression profiles of human papillomavirus (HPV)-positive and HPV-negative oropharyngeal squamous cell carcinomas (OPSCC) and to better understand the functional and biological effects of HPV infection in the premalignant pathway. Twenty-four consecutive patients with locally advanced primary OPSCC were included in a prospective clinical trial. Fresh tissue samples (tumor vs. matched normal epithelium) were subjected to whole transcriptome analysis and the results validated on the same cohort with RT-quantitative real-time PCR. In a separate retrospective cohort of 27 OPSCC patients, laser capture microdissection of formalin-fixed, paraffin-embedded tissue allowed RNA extraction from adjacent regions of normal epithelium, carcinoma in situ (premalignant) and invasive SCC tissue. The majority of patients showed evidence of high-risk HPV16 positivity (80.4%). Predictable fold changes of RNA expression in HPV-associated disease included multiple transcripts within the p53 oncogenic pathway (e.g. CDKN2A/CCND1). Other candidate transcripts found to have altered levels of expression in this study have not previously been established (SFRP1, CRCT1, DLG2, SYCP2, and CRNN). Of these, SYCP2 showed the most consistent fold change from baseline in premalignant tissue; aberrant expression of this protein may contribute to genetic instability during HPV-associated cancer development. If further corroborated, this data may contribute to the development of a non-invasive screening tool. This study is registered with the UK Clinical Research Network (ref.: 11945).

Keywords: Diagnosis by tumor markers and biomarkers; Rb/p16-related genes; human papillomavirus; mRNA expression analysis; oropharyngeal carcinoma.

Figure 1

Clinical and histopathological data for the prospective cohort. Human papillomavirus (HPV)‐positive status was defined as evidence of HPV16 L1/E6/E7 DNA or HPV16 E6/E7 mRNA +/− p16^{INK 4A} expression (>70% tumor cell staining). †Non‐malignant cause of death. ††Malignant cause of death. §Locoregional recurrence. BoT, base of tongue; CRT, chemoradiotherapy; F, female; f/u, follow‐up; Histol., histology; IHC, immunohistochemistry; M, male; Mod, moderately differentiated squamous cell carcinoma (SCC); mths, months; n/a, not applicable; Poor, poorly differentiated SCC; PS, Eastern Co‐Operative Group⁽⁴⁸⁾ physiological performance status; RT, radiotherapy; Well, well differentiated SCC.

Figure 2

(a) Two‐way anova identified target sequences with significant differential expression between normal and tumor tissue and further stratified by human papillomavirus (HPV) status. In the volcano plot, the x‐axis is the log₂ fold change value (tumor vs. normal epithelium) and the y‐axis is the −log10 odds value (HPV‐positive vs. HPV‐negative cohort). The two vertical lines represent 1.5 log2 (×15) fold change as the threshold cut‐off, both downregulated (left side) and upregulated (right side). The horizontal line represents a log odds value, P < 0.01 (false‐discovery rate of 0.5%), as the threshold cut‐off. (b) Hierarchical clustered heatmap of 223 genes displaying significant differential expression. As expected, CDKN2A ranked highly in the full transcriptome analysis. Five other genes that may have clinical relevance in HPV‐positive malignant disease are SYCP2,SFRP1, CRCT1, CRNN, and DLG2. (c) Validation graph for expression analysis (Illumina vs. RT–real‐time quantitative PCR [RT‐qPCR]; Pearson correlation coefficient, r = 0.905; P = 0.013; Kolmogorov–Smirnov test of normality, P > 0.10).

Figure 3

(a) H&E staining showing the junction between normal epithelium and severe dysplasia in human papillomavirus (HPV)‐positive tonsil squa‐mous cell carcinoma (top left). Although some flattening of cells at the apical surface may exist, the majority of the abnormal epithelium has a poorly differentiated dense basaloid appearance with nuclear pleomorphism. MCM7 (top right) and Ki67 (bottom left) staining are present in more than two‐thirds of the abnormal epithelium. In situ hybridization reveals an elevated level of HPV 16 DNA (bottom right) in the dysplastic epithelium when compared to the normal region (magnification, ×100; inset, ×150; Infinity capture software). ISH, in situ hybridization. (b) Reverse transcription–real‐time quantitative PCR (RT‐qPCR) analysis for all HPV‐positive sample types showing significant differential expression of SYCP2 (log2 fold change, 3.1 [95% confidence interval, 1.8–4.4]; P < 0.01) in premalig‐nant (in situ) tissue.

Figure 4

(a) Disease‐free survival calculations from Kaplan–Meier survival analysis of patients with human papillomavirus (HPV)‐positive oropharyngeal carcinoma. (b) A multivariate model was developed using Cox regression to investigate the effect of clinical parameters on disease‐free survival. Strati‐fication by HPV16, SYCP2, physiological perfor‐mance status, and smoking were all retained in the final model [bold text p < 0.05]. +ve, positive; −ve, negative; BoT, base of tongue; CRT, chemoradiotherapy; n/a, not applicable; SCC, squamous cell carcinoma; (SCC); mths, months; Perf, physiological performance status; Poor diff, poorly differentiated SCC; RT, radiotherapy.

Figure 5

Hypothetical model for SYCP2 with regard to human papillomavirus‐positive oropharyngeal squamous cell carcinomas. SYCP2 is a protein that is involved in linkage chromosomes through the synaptonemal complex, binding DNA at the scaffold attachment regions and driving the prophase of meiosis. Alterations in the gene expression of SYCP2 have been associated with impaired meiosis.