In:
Cancer Research, American Association for Cancer Research (AACR), Vol. 73, No. 8_Supplement ( 2013-04-15), p. 4239-4239
Abstract:
Tumor tissues are always contaminated with normal cells. If we can assess the fraction of tumor cells in DNA samples, a great advancement will have been achieved, such as exclusion of samples with a low tumor cell content from a study, and analysis of tumor heterogeneity. Though it has been reported next generation sequencing data can provide such information, the method is complicated and expensive. Here, we focused on DNA methylation levels of specific CpG sites that were fully methylated in cancer cells and fully unmethylated in normal cells to develop a fraction marker of tumor cells. First, to find such specific CpG sites of esophageal squamous cell carcinomas (ESCCs), we performed a genome-wide methylation analysis of 28 cancerous mucosae from ESCC patients, a pooled sample of non-cancerous mucosae, a pooled sample of normal esophageal mucosae from four healthy volunteer, four ESCC cell lines, and one sample of blood, using an Infinium HumanMethylation450 BeadChip array that covered 485,577 CpG sites. We searched for CpG sites highly methylated in the four ESCC cell lines (β value & gt; 0.8) and hardly methylated in the pooled samples and blood (β value & lt; 0.1), and identified 53 CpG sites. From the 53 CpG sites, we selected three CpG sites frequently hypermethylated (β value & gt; 0.5) in ESCCs (TFAP2B; 24/28, ARHGEF4; 20/28, RAPGEFL1; 19/28). To measure methylation levels of these and neighboring CpG sites, we designed three primer sets for high-resolution melt analysis. Next, we analyzed methylation levels of the three genes in ESCC cells and non-cancerous cells purified from three samples by laser capture microdissection, and confirmed at least one of the three genes was almost fully methylated in the ESCC cells, and all the three genes were almost fully unmethylated in non-cancerous cells. Furthermore, to exclude an effect of DNA copy number (CN) changes, we analyzed CN changes of each gene by quantitative PCR of 15 ESCC samples. CN gains of ARHGEF4 or RAPGEFL1 (more than 1.5-fold change) were not observed, but a small gain of TFAP2B (1.6-fold gain) was observed at a low frequency (3/15). Therefore, the higher value of either ARHGEF4 or RAPGEFL1 was considered to better reflect a fraction of ESCC cells. Only when values of these two genes were extremely low, was the value of TFAP2B considered to reflect it. Finally, we analyzed methylation levels of 35 ESCC samples using the three markers, and estimated fractions of ESCC cells. In the biopsy samples (n=15), a mean fraction of ESCC cells was 46.5±11.4 % (range, 24.8-72.8%), and, in the surgical samples (n=20), it was 64.1±11.7 % (range, 38.4-83.7%). Pathological analysis of three samples showed a fraction of ESCC cells assessed by the analysis was almost equal to that obtained by the markers. These results demonstrate we can estimate a fraction of ESCC cells in a sample with the DNA methylation markers established here. The use of such methylation markers can be applied to other types of cancers. Citation Format: Takamasa Takahashi, Satoshi Yamashita, Yasunori Matsuda, Toshikazu Ushijima. Development of DNA methylation markers to estimate the fraction of tumor cells in DNA samples. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 4239. doi:10.1158/1538-7445.AM2013-4239
Type of Medium:
Online Resource
ISSN:
0008-5472
,
1538-7445
DOI:
10.1158/1538-7445.AM2013-4239
Language:
English
Publisher:
American Association for Cancer Research (AACR)
Publication Date:
2013
detail.hit.zdb_id:
2036785-5
detail.hit.zdb_id:
1432-1
detail.hit.zdb_id:
410466-3
Permalink