Abstract:Introduction: The promoter methylation status of O-6-methylguanine-DNA methyltransferase (MGMTp) is an established predictive and prognostic marker in GBM. Previous studies showed that the expression of MGMT based on immunohistochemistry was variable and lacked association with survival. This in part is because non-tumor cells including endothelial cells and macrophages express MGMT. Advanced technologies such as single-cell RNA (scRNA) sequencing have helped to elucidate the cellular composition of cancer and its microenvironment. scRNA sequencing allows to assess gene expression level in tumor cells specifically. Methods: We used publicly available data from two recent GBM scRNA studies that included MGMTp methylation status data for patients to explore and uncover details about MGMT expression at the single-cell level: CPTAC (13 primary samples) and Neftel (20 primary samples). Results: In the CPTAC study, MGMT expression ranged from 0.19%-1.43% in the MGMTp methylated group (median 0.82%), and from 2.17%-28.36% in the MGMTp unmethylated group (median 5.7%). It therefore appears that 2% is a reasonable expression cutoff to predict the MGMTp methylation status based on scRNA data. In the Neftel study, MGMT expression ranged from 0-1.26% in the MGMTp methylated group (median 0.59%), and from 0.3-27.67% in the MGMTp unmethylated group (median 12.44%). Three unmethylated samples (out of 16) did not follow the 2% rule. It remains unclear if this is due to technical inaccuracies as the Neftel paper did not specify the method used to detect MGMTp methylation or even mere typos. Alternatively, could it be that truly MGMTp unmethylated samples can have low MGMT expression? Could this explain why some unmethylated MGMTp GBM patients surpass the expected survival? Interestingly, gene set enrichment analysis shows that MGMT expressing cells are enriched with mesenchymal genes, whereas MGMT negative cells are enriched with proneural genes. Conclusion: Fewer than 2% of GBM cells express MGMT when MGMTp is methylated.

What scRNA sequencing taught us about MGMT expression in glioblastoma