Epigenetic Bases of Aberrant Glycosylation in Cancer

Abstract

In this review, the sugar portions of glycoproteins, glycolipids, and glycosaminoglycans constitute the glycome, and the genes involved in their biosynthesis, degradation, transport and recognition are referred to as "glycogenes". The extreme complexity of the glycome requires the regulatory layer to be provided by the epigenetic mechanisms. Almost all types of cancers present glycosylation aberrations, giving rise to phenotypic changes and to the expression of tumor markers. In this review, we discuss how cancer-associated alterations of promoter methylation, histone methylation/acetylation, and miRNAs determine glycomic changes associated with the malignant phenotype. Usually, increased promoter methylation and miRNA expression induce glycogene silencing. However, treatment with demethylating agents sometimes results in silencing, rather than in a reactivation of glycogenes, suggesting the involvement of distant methylation-dependent regulatory elements. From a therapeutic perspective aimed at the normalization of the malignant glycome, it appears that miRNA targeting of cancer-deranged glycogenes can be a more specific and promising approach than the use of drugs, which broad target methylation/acetylation. A very specific type of glycosylation, the addition of GlcNAc to serine or threonine (O-GlcNAc), is not only regulated by epigenetic mechanisms, but is an epigenetic modifier of histones and transcription factors. Thus, glycosylation is both under the control of epigenetic mechanisms and is an integral part of the epigenetic code.

Keywords: DNA methylation; glycome; glycosyltransferases; miRNA targeting.

Figure 1

Structure of some core glycans. Monosaccharides are depicted according to this representation: blue square—GlcNAc, N-acetylglucosamine; yellow square—GalNAc, N-acetylgalactosamine; yellow circle—Gal, galactose; blue circle—Glc, glucose; green circle—Man, mannose; red triangle—Fuc, fucose; pink diamond—sialic acid, Sia. Anomers, linkage positions, and enzymes involved in relevant reactions are indicated. (A) N-glycans. As an example, the reactions are indicated using only the simple bi-antennary core structure as the substrate. Note that they are not alternative and can occur in various orders, because many of the indicated products can act as the substrate for several of the reported enzymes. An exception is represented by MGAT3 (bisecting enzyme) and MGAT5 (branching enzyme), whose reactions are mutually exclusive; (B) O-glycans; (C) Gangliosides. In both panels, the enzymes are indicated in the order in which they act.

Figure 2

Structure of some oligosaccharide chain terminations. Monosaccharides are depicted as in Figure 1. Anomers, linkage positions, and enzymes involved in relevant reactions are indicated. (A) Origin of type 1 and 2 chains, and termination by bioactive ends; (B) AB0 antigens.

Figure 3

Summary of the common epigenetic mechanisms of glycosylation control. A hypothetical glycogene is represented in which the activity of histone methyltransferase (HMT), histone deacetylase (HDAC), and DNA methyltransferase (DNA MT) results in chromatin condensation determining a transcriptionally inactive state; closely spaced red circles represent nucleosomes, black lollipops represent methylated cytosine residues (A). The gene can be turned into a transcriptionally active state by the activity of histone demethylase (HDM) and histone acetyltransferase (HAT), which reduce chromatin condensation (loosely spaced green circles), and by demethylation of the promoter region (white lollipops) (B). Transcription of the gene results in a mRNA comprised of a 5′UTR (yellow), a protein coding region (violet), and a 3′UTR (blue), which can be targeted by miRNA, resulting in translation inhibition; RISC, RNA-induced silencing complex (C). This hypothetical glycogene is a paradigm of many others controlled by epigenetic mechanisms. They are involved in a variety of biosynthetic steps that are grouped in the hypothetic N- and O-linked structures depicted in (D) and (E). The epitopes shown in (D) and (E) are often mutually exclusive, and are presented here as a single structure only for didactic purposes. Moreover, glycogenes encode molecules like galectins (F) involved in the biological roles of glycans.