O-GlcNAcylation in Chronic Lymphocytic Leukemia and Other Blood Cancers

Abstract

In the past decade, aberrant O-GlcNAcylation has emerged as a new hallmark of cancer. O-GlcNAcylation is a post-translational modification that results when the amino-sugar β-D-N-acetylglucosamine (GlcNAc) is made in the hexosamine biosynthesis pathway (HBP) and covalently attached to serine and threonine residues in intracellular proteins by the glycosyltransferase O-GlcNAc transferase (OGT). O-GlcNAc moieties reflect the metabolic state of a cell and are removed by O-GlcNAcase (OGA). O-GlcNAcylation affects signaling pathways and protein expression by cross-talk with kinases and proteasomes and changes gene expression by altering protein interactions, localization, and complex formation. The HBP and O-GlcNAcylation are also recognized to mediate survival of cells in harsh conditions. Consequently, O-GlcNAcylation can affect many of the cellular processes that are relevant for cancer and is generally thought to promote tumor growth, disease progression, and immune escape. However, recent studies suggest a more nuanced view with O-GlcNAcylation acting as a tumor promoter or suppressor depending on the stage of disease or the genetic abnormalities, proliferative status, and state of the p53 axis in the cancer cell. Clinically relevant HBP and OGA inhibitors are already available and OGT inhibitors are in development to modulate O-GlcNAcylation as a potentially novel cancer treatment. Here recent studies that implicate O-GlcNAcylation in oncogenic properties of blood cancers are reviewed, focusing on chronic lymphocytic leukemia and effects on signal transduction and stress resistance in the cancer microenvironment. Therapeutic strategies for targeting the HBP and O-GlcNAcylation are also discussed.

Keywords: O-GlcNAc transferase (OGT); O-GlcNAcase (OGase); O-linked β-D-N-acetylglucosamine (O-GlcNAc); cancer; chronic lymphocytic leukemia; cytokines; metabolism; signal transduction.

Figure 1

Schema of the hexosamine biosynthetic pathway (HBP). The amino sugar UDP-GlcNAc is generated in the HBP by combining fructose-6-phosphate, glutamine, nucleotides, and acetyl-CoA. More detailed description of the sequential enzyme reactions in the pathway is provided in the text. O-GlcNAc transferase (OGT) uses UDP-GlcNAc to transfer O-GlcNAc moieties onto target proteins that can be recognized by RL2 antibodies and removed by O-GlcNAse (OGA). OGT and OGA inhibitors mentioned in the text are shown in the yellow boxes.

Figure 2

Schema of O-GlcNAcylated oncogenic processes in the CLL microenvironment. Major signaling and system modules that drive growth and progression of CLL cells are organized in the boxes along with important auxiliary pathways like cytokine- and AKT-signaling. Genes frequently mutated in CLL are colored in red. Proteins and cells discussed in the text that can be affected by O-GlcNAcylation are indicated by the yellow circled G’s. The figure emphasizes that O-GlcNAcylation increases the tumor-suppressing activity of wild-type p53 but enhances many of the signaling processes that support tumor growth in the presence of mutant p53 and an impaired p53 axis.

Figure 3

Correlation of OGT mRNA and protein levels in primary blood cancer cells with clinical course. (A) Survival of patients with high (n=66) or low (n=40) expression of OGT in CLL cells (left panel) and survival of diffuse large cell lymphoma patients with rituximab, cyclophosphamide, vincristine, adriamycin, and prednisone (R-CHOP) with high (n=215) or low (n=192) OGT expression in lymph node biopsies (right panel) were respectively compared by in silico analysis of the Herold (72) and Lenz (73) databases using the DRUGSURV bioinformatics analysis tool (http://www.bioprofiling.de/GEO/DRUGSURV). P-values for differences between the curves are 0.001 (left) and 0.0002 (right). (B) Expression of OGT (left) and OGA (right) proteins in 44 patients with aggressive CLL cells more likely to have an impaired p53 axis as indicated by unmutated IGHV genes (U) and 47 patients with more indolent CLL cells and mutated IGHV genes (M) were obtained with the R Shiny app (http://mozi.embl.de/public/proteomExplorer) provided for exploration of the CLL proteome (74). P-values and adjusted P-values are shown in the graphs and suggest more aggressive CLL cells have a lower OGT/OGA ratio, possibly consistent with relatively lower global O-GlcNAcylated protein levels.