Background: Cervical cancer is the third most prevalent cancer in women worldwide. Human papillomavirus (HPV) infection has been shown to positively correlate with invasive disease and has been implicated in upwards of 65% of these cases. Screening for HPV and morphological abnormalities has reduced the number of deaths; however, the overall five year survival rate remains relatively low, at 52%. Not all preinvasive lesions progress to invasive disease. Investigation of genes expressed in preinvasive CIN III lesions compared to those expressed by normal cervical epithelium may yield insight to early stage disease. Furthermore, an understanding of gene expression changes that parallel the progression stages will identify novel targets for developing treatment. Objectives: The key objectives of this study are (1) to identify transcripts that show increased expression in CIN III cervical specimens and (2) identify the genes and biological pathways disrupted in disease. This is achieved by profiling staged specimens using the Long-SAGE technology, which enables the quantitative analysis of mRNA transcripts without prior knowledge of the genes. Materials and Methods: Two normal tissue and two CIN III cervical LEEP cone biopsy specimens were collected at the British Columbia Cancer Agency in Vancouver and graded by a pathologist. Specimens were stored in RNAlater and once homogenised the resulting lysates were used for Long-SAGE library construction. Libraries were sequenced and individual sequence tags were mapped to genes. Scaled tags that showed three-fold or greater expression in CIN III or Normal tissue derived libraries were compiled. A Modified Z-Test was used to identify tags which showed statistically significant expression levels differences (p<.05) between CIN III expression profiles and normal cervix tissue profiles. Results: SAGE libraries were sequenced to between 165,000 and 200,000 tags deep. According to the modified Z-test 14 tags showed to have higher expression in the CIN III cases whereas 18 tags were found to have increased expression in the normal specimens. When comparing raw scaled tag counts, 1059 tags showed to have three fold or greater increased expression in CIN III when compared to normal specimens including those that mapped to EGFR and KRT4. Conclusion: This study is the first description of the transcriptome of normal cervix tissue. This in turns provided a baseline for comparison against progressive stages of cervical cancer. A detailed comparison between CIN III gene expression profiles against those of normal tissue yielded differential expression in a number of gene classes including cell cycle regulators and developmental genes. This study provides novel candidate genes for the development for therapeutics for preinvasive cervical cancer.

[Proc Amer Assoc Cancer Res, Volume 47, 2006]