Cystatin M, originally described as a putative tumor suppressor, whose expression is often diminished or completely lost in metastatic breast cancers  has been clearly shown to be epigenetically regulated by strong hypermethylation of the CST6 gene promoter in breast cancer cell lines , in breast cancer and metastatic lesions in the lymph nodes , in malignant gliomas , in cervical  and prostate cancer . Because promoter hypermethylation does not account for the loss of CST6 expression in all tumors alternative modes of CST6 repression are likely, such as histone deacetylation and repressive chromatin structure may be involved , since silencing of CST6 has been associated with repressive trimethyl-H3K27 and dimethyl-H3K9 histone marks .
Recently, CST6 was also identified among 10 hypermethylated genes that distinguish between cancerous and normal tissues according to the extent of methylation . Moreover, a whole-genome approach using a human gene promoter tiling microarray platform to identify genome-wide and gene-specific epigenetic signatures of breast cancer metastasis to lymph nodes led to functional associations between the methylation status and expression of genes CDH1, CST6, EGFR, SNAI2 and ZEB2 associated with epithelial-mesenchymal transition . In addition, a recent functional epigenetic study of renal cell carcinoma (RCC) cell lines and primary tumors by high-density gene expression microarrays identified CST6 as one of eight genes that showed frequent (>30%) tumor-specific promoter region hypermethylation associated with transcriptional silencing (epigenetically inactivated candidate RCC TSGs). According to this study, re-expression of BNC1, CST6, RPRM and SFRP1 suppressed the growth of RCC cell lines . All these recent studies are in support of the importance of CST6 promoter methylation in metastasis.
Our group has shown for the first time the prognostic significance of CST6 promoter methylation in patients with operable breast cancer . According to our findings, the diagnostic sensitivity and specificity of CST6 methylation as a biomarker for prediction of relapses and deaths in operable breast cancer seems to be quite promising . Moreover, we have recently shown that CST6 promoter was methylated in Circulating Tumor Cells (CTC) isolated from peripheral blood of breast cancer patients, in both groups of early disease and verified metastasis . A recent study has also shown that cystatin M loss may be associated with the losses of ER, PR, and HER4 in invasive breast cancer .
Based on all these studies, we strongly believe that the reliable and easy detection of CST6 methylation in clinical samples will be of great importance for cancer research. For this reason we decided to develop a closed tube, highly sensitive, cost-effective, rapid and easy-to-perform assay for CST6 promoter methylation based on methylation-sensitive high resolution melting analysis (MS-HRMA). Resolution of DNA methylation by melting analysis relies on the fact that the Tm of a PCR product generated from bisulfite-treated DNA reflects the methylation status of the original DNA template . Because unmethylated cytosines will be converted into uracil during bisulfite treatment and subsequently amplified as thymine, whereas methylcytosines will remain as methylcytosine and be amplified as cytosine, the methylated sequence will have a higher G:C content, and hence a higher Tm, than the corresponding unmethylated sequence. After amplification with primers that will not differentiate between methylated and unmethylated molecules, the melting properties of the PCR products can be examined in the thermal cycler by slowly elevating the temperature under continuous or step-wise fluorescence acquisition. The melting curves or derived melting peaks provide a profile of the methylation status of the entire pool of DNA molecules in the sample .
Many reports have already clearly illustrated the great potential of melting analysis for sensitive and highthroughput assessment of DNA methylation in inherited disorders and cancer [6, 11–13, 30–33, 44]. Compared with current gel-based assays MS-HRMA has the important advantage of the closed-tube format, which simplifies the procedure, decreases the risk of PCR contamination, and decreases analysis time. In addition, melting analysis resolves heterogeneous methylation, detects methylated and unmethylated alleles in the same reaction, and requires only standard, inexpensive PCR reagents. In addition, the design of individual assays is simple [45–47].
The developed assay is highly specific and sensitive since it can detect the presence of low abundance CST6 methylated DNA sequences (down to 1%). Moreover to the best of our knowledge, this is the first assay reported so far that provides additionally a semi-quantitative estimation of CST6 promoter methylation. When compared to MSP, the developed MS-HRMA gives comparable but not identical results. The discrepancies between MS-HRMA and MSP can be explained by the different principles on which these methods are based. In MSP we get a positive signal only when the specific CpG island that the primers are designed for is methylated. However it is known that different samples can vary in the methylation sites in specific positions in their CpG islands. In this way if a sample is methylated in positions 3, 6 and 7 and the MSP primers are designed to recognize methylation in positions 4, 5 and 8, MSP will give a negative result, while MS-HRMA will give a positive result since it is affected by the presence of any methylated CpG island that is located between the primers. In the opposite way, if the methylation sites that are recognized by the MSP primers are not included in the region amplified by MS-HRMA primers a sample found positive by MSP will be negative by MS-HRMA.
This is the first time that methylation levels for CST6 are reported in clinical samples. Based on our findings, we can definitely say that these levels vary significantly among samples. An interesting finding is that a histologically “non-cancerous” tissue that was adjacent to a highly methylated (50%) tumor sample was also found to be methylated, at a lower percentage (1%). CST6 methylation is an early event in breast cancer, since methylation of the CST6 promoter has already been reported in 7 out of 28 corresponding normal tumor-adjacent breast tissues samples . This could possibly indicate that some “normal” cells surrounding the tumor tissue have already a malignant transformation, not detected by conventional immunohistochemistry. In our study we have used whole tissue sections containing more than 80% of tumour cells. However, we can speculate that the percentage of contaminating normal cells affect the level of methylation seen in our samples. For this reason, we believe that laser capture microdissection could ensure a higher proportion of lesional cells in clinical samples to be studied.