p16INK4A polymorphism and ANRIL expression in human breast cancer

Abstract

Overview: Breast cancer is the second most prevalent cancer in New Zealand with a largely unclear genetic etiology. Recent progress has demonstrated that expression of tumour suppressor proteins is regulated by long non-coding RNA. At chromosome position 9p21 the tumour suppressor p16INK4A and the long non-coding RNA antisense non-coding RNA in the INK4 locus (ANRIL) are both encoded. Inherited variants in p16INK4A including the minor allele of the rs11515 (C/G500) polymorphism at the 3’UTR region and are more frequent in breast cancer patients. The rs11515 allele was associated with increased ANRIL and reduced p16INK4A expression. This study aims to develop a method to type the rs11515 alleles in formal fixed normal associated breast cancer tissues to determine the frequency of the minor allele in a cohort of 166 breast tissues acquired from mastectomy at Dunedin Hospital. The study also aims to identify the cell type responsible for increased ANRIL in breast cancer tissues and lastly aims to investigate ANRIL expression in a wide range of tissues to obtain a better understanding of where ANRIL is expressed.

Hypothesis: The rs11515 CG genotype is associated with higher ANRIL in malignant cells in breast cancer. Expression of ANRIL will vary between tissues and a better understanding of ANRIL expression in multiple tissues will identify future tissue types to investigate the effect of rs11515 genotypes. Methods: Formalin fixed tissue samples were genotyped for the rs11515 polymorphism using a modified protocol to that for frozen tissue. Tumour and non-neoplastic breast tissues were analysed for p16INK4A and ANRIL expression using RT-qPCR in n=23 tissue samples each for p16INK4A and ANRIL. The effects of rs11515 genotype, and clinical factors on the co-expression of p16INK4A and ANRIL were assessed. Cellular location of ANRIL was determined in n=50 breast cancer tissues using RNAscope® In Situ Hybridisation. Results: Nine percent of samples from Dunedin were heterozygous (CG) and the rest were homozygous (CC) for the rs11515 genotype. Expression of ANRIL was higher in those with the GC compared to the CC genotype (p=0.004). A positive correlation was observed between p16INK4A and ANRIL expression in both genotypes (r=0.89 and r=0.68 respectively) for the tumour tissue whereas a non-significant positive correlation was observed amongst the normal associated tissue. Additionally, a tentative correlation between ER and PR status and co-expression of ANRIL and p16INK4A was noted with ER+ breast cancer cases had higher ANRIL and p16INK4A whereas PR+ cases expressed a higher ANRIL with lower p16INK4A level compared to ER/PR- cases. However, further studies and larger cohort size are required to confirm this. The in situ analysis of ANRIL revealed ANRIL in malignant cells depended on the invasiveness of the cancer cells and was highly expression in the nucleolus compared to the normal associated cells. Conclusion: Increased ANRIL and p16INK4A were found in rs11515 heterozygotes whereas the presence of ER/PR status influenced the co-expression in both genotypes. The increased ANRIL with the CG genotype was consistent with that found earlier; however, the positive correlation between ANRIL and p16INK4A differed from the inverse correlation found previously. In Situ Hybridisation showed presence of ANRIL in malignant cells whereas the multiple tissue based analysis was suggestive of a tissue specific regulatory co-expression profile.