The Role of FAS Receptor Methylation in Osteosarcoma Metastasis

Abstract

Osteosarcoma is the most frequent primary malignant bone tumor with an annual incidence of about 400 cases in the United States. Osteosarcoma primarily metastasizes to the lungs, where FAS ligand (FASL) is constitutively expressed. The interaction of FASL and its cell surface receptor, FAS, triggers apoptosis in normal cells; however, this function is altered in cancer cells. DNA methylation has previously been explored as a mechanism for altering FAS expression, but no variability was identified in the CpG island (CGI) overlapping the promoter. Analysis of an expanded region, including CGI shores and shelves, revealed high variability in the methylation of certain CpG sites that correlated significantly with FAS mRNA expression in a negative manner. Bisulfite sequencing revealed additional CpG sites, which were highly methylated in the metastatic LM7 cell line but unmethylated in its parental non-metastatic SaOS-2 cell line. Treatment with the demethylating agent, 5-azacytidine, resulted in a loss of methylation in CpG sites located within the FAS promoter and restored FAS protein expression in LM7 cells, resulting in reduced migration. Orthotopic implantation of 5-azacytidine treated LM7 cells into severe combined immunodeficient mice led to decreased lung metastases. These results suggest that DNA methylation of CGI shore sites may regulate FAS expression and constitute a potential target for osteosarcoma therapy, utilizing demethylating agents currently approved for the treatment of other cancers.

Keywords: 5-azacytidine; DNA methylation; FAS receptor; osteosarcoma; prognosis.

Figure 1

Variant FAS expression in osteosarcoma tumor samples is associated with outcome. (A) Comparison between standard deviation among tumor samples of mRNA expression (log2 intensity) of FAS versus all other genes on the expression array. (B–E) Kaplan–Meier survival curves comparing high and low FAS mRNA (stratified at the 25th percentile). Overall survival and event-free survival are calculated using (A,B), only FAS mRNA high or low expression status and (D,E), after including metastasis at diagnosis status as a covariate. Significant differences in survival calculated using a Wald test. Only the p-value of FAS expression is reported in the case of multivariate models.

Figure 2

FAS promoter methylation negatively correlates with mRNA expression. (A) Relative location of all HM450 probes overlapping FAS promoter and start of coding region. Red indicates methylation probes significantly correlated with FAS mRNA expression in a negative manner. The red bar indicates the coding region. (B–D) Plots showing correlation between methylation beta value and expression log2 intensity for three sites with significant negative correlation.

Figure 3

Functional effects of FAS demethylation. (A) Bisulfite sequencing validation results from SaOS-2 and LM7 cell lines. Site 2* represents a CpG site 6bp upstream of Site 2. (B) Western blot of FAS protein expression shows TR-LM7 cells have higher protein expression in comparison with UT-LM7 cells. (C) TR-LM7 cells at 5 µM and 10 µM for 12 days were compared with UT-LM7 cells in a wound healing assay.

Figure 4

Orthotopic injection of 5-aza TR and UT-LM7 cells. (A) Primary bone tumor and lung tissues with and without metastasis from OT-injected mice. Mice injected with UT-LM7 cells developed both primary tumor and lung metastasis while injection with TR-LM7 cells resulted in fewer metastases, even after development of bone tumor. Scale bar in zoomed out image = 50 µm, zoomed in image = 25 µm. (B) Mice injected with TR-LM7 cells trend towards a low number of pulmonary metastases. (C) Mice injected with TR-LM7 have a significantly improved overall survival rate compared to those injected with UT-LM7 based on a Wald test (p-value = 0.036).