Epigenetic Upregulation of the O-Linked Beta-N-Acetylglucosamine Transferase (OGT) in Response to Dronabinol Results in Antileukemic Efficacy In Vivo

The O-linked β-N-acetyl glucosamine (O-GlcNAc) transferase (OGT) is a master regulator enzyme adding O-GlcNAc to serine or threonine residues in a multitude of target proteins. Aberrant O-GlcNAc modification is implicated in pathologies of metabolic and neurodegenerative diseases as well as cancers and autoimmunity.

We herein provide evidence that dronabinol (FDA approval as Marinol), the natural (−)-Δ⁹-Tetrahydrocannabinol, is a potent inducer of OGT via epigenetic hypomethylation of the transcription start site –thereby exerting antileukemic activity in acute leukemia in vivo.

We have recently shown, that dronabinol, the natural (−)-Δ⁹-Tetrahydrocannabinol, has growth-inhibiting antitumor efficacy - including acute leukemia. We now reveal a novel mechanism-of-action via epigenetic modulation of OGT, an enzyme linked to genes involved in leukemogenesis such as AKT, MLL5, TET2 or ASXL1, releasing leukemia blasts from differentiation blockage in vivo and sensitizing cells towards induction of apoptosis.

gDNA methylation gene arrays using Jurkat leukemia cells revealed global modulation of methylation patterns upon dronabinol treatment. OGT was identified as the highest altered gene (-42%, pval 3,68E-38) - correlating with an increase of OGT protein expression in Western immunoblots. Consistently, hypomethylation of the transcription start site of OGT and induction of OGT protein expression upon dronabinol were confirmed in an independent array using native patient samples. To study whether epigenetic activity is driven via the cannabinoid receptors, Jurkat cells were pretreated with CB1 (LY320135) and/or CB2 (JTE-907) antagonists, and exposed to dronabinol. Importantly, upregulation of OGT protein expression upon dronabinol was suppressed by inhibition of either receptor. Even more, inhibition of CB1 and/or CB2 reduced induction of apoptosis – and was most profound when inhibiting both receptors simultaneously. Similarly, retroviral knockdown of OGT in Jurkat and native leukemia blasts rendered cells less susceptible towards induction of apoptosis.

Furthermore, we have evidence, that OGT has lead to release of the differentiation block in leukemia cells in vivo. Supportive treatment with dronabinol of an unfit patient with secondary acute myeloid leukemia resulted in direct disease control: Tantalizingly, besides a proapoptotic effect, the leukemic clone was maturing – with loss of CD34 and upregulation of CD11c, CD14 and CD15. Remarkably, immunophenotypic and genotypic (using NGS) profiling of the predominant monocyte population present two months after start of treatment, revealed that these mature monocytes derived from the leukemia clone (presenting mutations in EZH2 and ASXL1 among others – both known candidate genes of OGT).

Mimicking this observation, we treated cells of this and other patients as well as defined leukemia models such as MOLM14 with dronabinol ex vivo and revealed upregulation of differentiation markers, such as CD11c, CD15 or CEBPA by flow cytometry and immunoblots – which was abrogated by lentiviral OGT-interference.

Our findings provide a strong rationale for further exploring dronabinol as an agent with remarkable antileukemic efficacy achievable in vivo. In specific, overriding the differentiation blockage in leukemia cells may open up alternative therapeutic approaches similar to promyelocytic leukemia.