MacroH2A – An epigenetic regulator of cancer

Abstract

Epigenetic regulation is one of the most promising and expanding areas of cancer research. One of the emerging, but least understood aspects of epigenetics is the facultative and locus-specific incorporation of histone variants and their function in chromatin. With the characterization of the first loss of function phenotypes of the macroH2A histone variants, previously unrecognized epigenetic mechanisms have now moved into the spotlight of cancer research. Here, we summarize data supporting different molecular mechanisms that could mediate the primarily tumor suppressive function of macroH2A. We further discuss context-dependent and isoform-specific functions. The aim of this review is to provide guidance for those assessing macroH2A’s potential as biomarker or therapeutic intervention point.

Introduction

The term cancer encompasses a large number of diverse diseases of hyperproliferative character. A common feature is that cells acquire capabilities that allow them to divide uncontrollably, ultimately disrupting the homeostasis of the tissue and even the whole organism via metastasis (reviewed in [1]). From the identification of the first oncogenes onwards, cancer has been considered a genetic disease for several decades. Today, we know that alterations of the mechanisms that establish and maintain stable gene expression programs also participate in cancer development and progression (reviewed in [2], [3], [4]). These mechanisms are often globally referred to as epigenetics. Since the inception of the term in the mid-40s by Waddington [5], epigenetics has been redefined and reinterpreted in several ways. As reviewed by Danny Reinberg, many argue that epigenetic changes should only refer to heritable variation that does not rely on the DNA sequence [6]. However, others subscribe to the loosening of this definition in order to extend it to non-dividing cell as proposed by Adrian Bird who defined epigenetics as “the structural adaptation of chromosomal regions so as to register, signal or perpetuate altered activity states” [7].

In the cell nucleus, all chromosomal DNA is stored as chromatin with the nucleosome as universal and structural unit. The nucleosome consists of a stretch of DNA double helix wrapped twice around a core of eight histone proteins (reviewed in [8]). Modifications of both DNA and histones contribute to epigenetic regulation (discussed in [9], [10]). Other less understood epigenetic mechanisms involve non-coding RNAs and the regulation of higher order chromatin structures (reviewed in [11], [12]), which in the simplest case are loops formed by enhancer–promoter contacts.

Epigenetics is one of the most promising and expanding fields in the current biomedical research landscape. The reversible nature of epigenetic modifications provides the rationale for the development of inhibitory drugs (discussed in [13], [14]). At the same time, cancer-associated chromatin alterations can serve as biomarkers for prognosis, diagnosis and drug responses (summarized in [4]). Three decades ago, changes in DNA methylation were the first chromatin alteration identified in cancer [15]. Today, several “epigenetic” drugs targeting chromatin-modifying enzymes have gained approval for the treatment of different types of leukemia and the first epigenetic biomarkers have reached clinical application (reviewed in [4]). The new technical possibilities opened by massive parallel sequencing technologies are now further accelerating the discovery of cancer-relevant epigenetic mechanisms and provide the basis for their translation into tools for the management of cancer.

Yet another, much less understood, epigenetic mechanism is based on the exchange of canonical histones for histone variants. Mammals possess two major variants of H3 and five of H2A plus several tissue-restricted proteins. While some variants only contain minor sequence variations, others significantly differ in structure and sequence (reviewed in [16], [17], [18]). Among all histone variants, macroH2A differs most from its canonical counterpart. MacroH2A is the only histone with a tripartite structure consisting of a N-terminal histone-fold, an intrinsically unstructured linker domain and a C-terminal macro domain (reviewed in [19]). The macro domain is approximately twice the size of the histone domain. The linker protrudes out of the compact structure of the nucleosome core and places the macro domain in an accessible position outside the nucleosome proximal to its symmetrical axis (Fig. 1) [20]. Given the size of the macro domain, the exchange of the canonical H2A with macroH2A may be one of the most extensive chromatin alterations occurring at the level of the nucleosome. Two genes encode the proteins macroH2A1 and macroH2A2, which share the same overall domain structure and 68% identity on the amino acid sequence level [21], [22]. MacroH2A1 and macroH2A2 differ in genomic localization and patterns of expression among cell types [22]. Alternative splicing of the macroH2A1-encoding transcript further gives rise to the two isoforms macroH2A1.1 and macroH2A1.2 [23] that differ in a short amino acid stretch affecting the size and hydrophobicity of the binding pocket of the macro domain [24]. This difference determines the capacity of macroH2A1.1 to interact with NAD+ derived metabolites such as ADP-ribose, whereas macroH2A1.2 is unable to do it so [24]. Binding to ADP-ribose is an evolutionarily conserved yet poorly understood function of macro domains (extensively discussed in [25]). Whenever possible, we will point out isoform-specific observations. But when no distinction can be made, we will collectively refer to them as macroH2A1.

During the last few years, a series of publications have brought major advances to our knowledge of the physiological and pathophysiological functions of this particular histone variant. Here, we will summarize and discuss the recent data establishing a role for macroH2A in the pathogenesis of cancer.