Epigenetic Profiling Identifies EGLN3 as a Frequent Target for Transcriptional Silencing in Plasma Cell Neoplasias.

Background: Multiple myeloma (MM) results from an excess of monoclonal plasma cells in the bone marrow. MM cells are interleukin (IL)-6 and stromal cell dependent. Although the molecular signature of MM cells has been identified, the signature of the stromal component of the BM microenviroment is not yet fully determined and holds important clues to the understanding of disease pathobiology and progression. It has been suggested that bone marrow hypoxia is lessened during myeloma progression and that myeloma-associated angiogenesis is functional. Hypoxia-inducible factor 1 (HIF-1) is a transcription factor and a master regulator of hypoxia responsive genes in cancer, which promote angiogenesis, vasodilation and glycolysis. As part of a programme to seek novel genes subject to epigenetic regulation in B cell neoplasia, we have identified the egl9 homologue 3 (EGLN3) as a frequent target for methylation-dependent transcriptional silencing in multiple myeloma (MM) and Waldenström’s macroglobulinemia (WM). EGLN3 is a prolyl hydroxylase (POH) that functions as a cellular oxygen sensor that targets HIF-1 alpha for proteasomal degradation via the von Hippel-Lindau (VHL) complex.

Patients and Methods: Thirty six MM patients (18 male, 18 female, age range 50–87 years), 4 patients with WM (2 male and 2 female, age range 47–75 years) and two patients with monoclonal gammopathy of undetermined significance (MGUS; 2 male age 56 and 67 years old) were included in the study. Twenty out of 36 MM patients had advanced stage disease (Durie Salmon stage≥II, IPI≥II). All samples were taken at diagnosis except for 5 that were taken when progression occurred from plateau phase. The Durie-Salmon staging system and the International Prognostic Index (IPI) were used for clinical and prognostic discrimination of patients. Bone marrow aspiration was performed with aseptic technique. Genomic DNA was isolated with a commercially available kit (QIAmp DNA mini kit, Qiagen).

Results: We identified a CpG island in the 5′ sequences of the EGLN3 gene. Using bisulphite sequencing and methylation specific PCR, we observed aberrant CpG methylation in 17/36 cases of MM and 1/4 cases of WM. Expression of EGLN3, assessed by quantitative PCR (qPCR) correlated inversely with methylation. In contrast, we detected no methylation in the related EGLN1 or EGLN2 in either MM or WM. Cases with methylated EGLN3 were more likely to have bone lytic lesions (r=0.39, p≤0.05), but there was no correlation with extramedullary disease, hemoglobin<10 mg/dl or advanced disease stage.

Conlusions: EGLN3 is frequently methylated in patients with multiple myeloma. We suggest that methylation of EGLN3 leading to decreased HIF-1a degradation could be a possible mechanism of increased angiogenesis and altered bone marrow microenviroment that is more supportive for survival and growth of MM cells. The role of POH in MM merits additional study.