5-Hydroxymethylcytosine Loss in Conjunctival Melanoma

Abstract

Aims: Conjunctival and cutaneous melanoma partially share similar clinical and molecular backgrounds. As 5-hydroxymethylcytosine (5-hmC) loss has been demonstrated in cutaneous melanoma, we decided to assess if similar changes were occurring in conjunctival melanoma. Methods: 5-methylcytosine (5-mC), 5-hmC and TET2 were respectively identified by immunohistochemistry and RNA ISH in 40 conjunctival nevi and 37 conjunctival melanomas. Clinicopathological correlations were established. Results: 5-mC, TET2 and 5-hmC were respectively identified in 67.5%, 95% and 100% of conjunctival nevi and in 81.1%, 35.1% and 54% of conjunctival melanomas. A significant 5-hmC and TET2 loss was identified in conjunctival melanoma comparing to nevus, as well as a significant correlation between TET2 and 5-hmC expression. In the melanomas, 5-hmC expression was only significantly associated with local lymphatic invasion, but not with other clinicopathological parameters. There was a correlation between TET2 expression and the localization of the tumors. 5-mC expression was not associated with any clinicopathological parameters. Conclusions: We identified a significant 5-hmC loss in conjunctival melanoma similar to cutaneous melanoma. This loss may possibly be attributed to TET2 loss or IDH1 mutations. 5-hmC loss in conjunctival melanoma may help in the differential diagnosis between atypical conjunctival nevus and conjunctival melanoma.

Deep5hmC: Predicting genome-wide 5-Hydroxymethylcytosine landscape via a multimodal deep learning model

“…Beyond neurodegenerative diseases, 5hmC also plays a significant role in cancer development and treatment. Genome-wide mapping of 5hmC reveals that loss of 5hmC is an epigenetic hallmark of melanoma and medulloblastoma (Lian et al, 2012; Stahl et al, 2021; Zhao et al, 2021). Additionally, 5hmC in circulating cell-free DNA serves as diagnostic biomarkers for colorectal, gastric, pancreatic cancer, acute myeloid leukemia and other common human cancer types (Gao et al, 2019; W. Li et al, 2017; Shao et al, 2022).…”