Unravelling the Complexity and Functions of MTA Coregulators in Human Cancer

Da-Qiang Li*,†,‡,§,1; Rakesh Kumar¶,||,#,1    * Fudan University Shanghai Cancer Center and Institutes of Biomedical Sciences, Shanghai Medical College, Fudan University, Shanghai, China
† Department of Oncology, Shanghai Medical College, Fudan University, Shanghai, China
‡ Key Laboratory of Breast Cancer in Shanghai, Shanghai Medical College, Fudan University, Shanghai, China
§ Key Laboratory of Epigenetics in Shanghai, Shanghai Medical College, Fudan University, Shanghai, China
¶ Department of Biochemistry and Molecular Medicine, School of Medicine and Health Sciences, George Washington University, Washington, DC, USA
|| Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas, USA
# Department of Molecular and Cellular Oncology, University of Texas M.D., Anderson Cancer Center, Houston, Texas, USA
1 Corresponding authors: email address: daqiangli1974@fudan.edu.cn, bcmrxk@gwu.edu

1 Introduction

Distant metastasis represents a hallmark of human cancer and accounts for about 90% of cancer-related deaths (Hanahan & Weinberg, 2000). The metastatic cascade is a series of highly orchestrated biological processes that are driven by numerous gene products, including those important for cancer cell migration and invasion, angiogenesis, and cell survival and colonization of tumor cells at distant target organs (Hanahan & Weinberg, 2000; Nguyen, Bos, & Massague, 2009; Psaila & Lyden, 2009). One of such family of gene products with predominant roles in cancerous and metastatic process is the metastasis-associated proteins (MTA proteins). The MTA protein family consists of six members in vertebrates, including MTA1, MTA1s (MTA1 short form), ZG29p (zymogen granule 29 kDa protein), MTA2, MTA3, and MTA3L, which are coded by three distinct transcripts, namely, MTA1, MTA2, and MTA3 (Bowen, Fujita, Kajita, & Wade, 2004; Futamura et al., 1999; Kumar, Wang, & Bagheri-Yarmand, 2003; Manavathi & Kumar, 2007; Fig. 1).

MTA1, the founding member of the MTA1 gene family, was originally identified in 1994 by differential cDNA library screening prepared from the 13762NF rat mammary adenocarcinoma metastatic system (Toh, Pencil, & Nicolson, 1994). Subsequent studies mapped MTA1 to human chromosome 14q 31.2 (Martin et al., 2001). As MTA1’s expression level closely correlates with the metastatic potential of breast cancer cells, the newly discovered gene was named as metastasis-associated protein 1 (Toh et al., 1994). Following this, two naturally occurring variants ZG29p and MTA1s were discovered (Kleene, Zdzieblo, Wege, & Kern, 1999; Kumar et al., 2002). The ZG29p encodes an N-terminally truncated form of MTA1 due to alternative transcription initiation and is exclusively expressed in pancreatic acinar cells (Kleene et al., 1999). In contrast, the MTA1s is a naturally occurring C-terminal truncated version of MTA1 generated by alternative splicing followed by addition of 33 novel animo acids due to a frame shift (Kumar et al., 2002).

In contrast to MTA1, MTA2 was identified in 1999 by sequencing of a 70-kDa polypeptide during analysis of the nucleosome remodeling and histone deacetylase (NuRD) complex (Zhang et al., 1999). Sequence alignment revealed that human MTA2 is 65% identical to human MTA1 (Yao & Yang, 2003; Zhang et al., 1999) and that both proteins are highly homologous in the N-terminal region but divergent in the C-terminal region (Yao & Yang, 2003). In general, MTA family members form distinct histone deacetylase (HDAC) containing complexes, and thus, potentially could target different genes (Yao & Yang, 2003).

Mouse MTA3 was initially cloned in 2001 (Simpson, Uitto, Rodeck, & Mahoney, 2001), followed by identification of human MTA3 as an estrogen-dependent component of the NuRD complex with a role in breast cancer invasion (Fujita et al., 2003). While MTA1 and MTA2 are primarily nuclear proteins, the expression of MTA3 is somewhat diffused. This suggests that the biologic functions of MTA3 are likely distinct as compared to MTA1 and MTA2 (Yao & Yang, 2003). In this context, the expression of MTA1 and MTA2, but not MTA3, increases during breast cancer progression in a MMTV-PyV-mT mouse model, wherein different stages of mammary tumorigenesis could be easily recognized (Zhang, Stephens, & Kumar, 2006).

During the past two decades, a large body of experimental and clinical studies demonstrates that MTA proteins are frequently deregulated in a wide range of human cancers and play a central role in tumor progression and metastasis (Li, Pakala, et al., 2012; Sen, Gui, & Kumar, 2014a). Moreover, MTA proteins localize at the nexus of multiple upstream oncogenic signaling pathways and regulate gene expression through a NuRD-dependent and -independent mechanism (Lai & Wade, 2011; Li, Pakala, et al., 2012).

2 Domain Architectures of MTA Proteins

The fundamental unit of a protein is its structural domain, which is the building block of protein structure as well as a potential determinant for its putative function or functions. Accumulating structural and biochemical studies reveal that the MTA proteins, with the exception of two truncated version of MTA1, ZG29p, and MTA1s (Kleene et al., 1999; Kumar et al., 2002), contain three conserved structural domains, including the bromo-adjacent-homology (BAH) domain; the EGL-27 and MTA1 homology 2 (ELM2) domain; and the SWI3, ADA2, N-CoR, and TFIII-B (SANT) domain (Manavathi & Kumar, 2007; Fig. 1).

The BAH domain is commonly found in chromatin-associated proteins and plays critical roles in regulatory protein–protein interactions, nucleosome binding, and recognition of methylated histones (Kuo et al., 2012; Norris, Bianchet, & Boeke, 2008; Onishi, Liou, Buchberger, Walz, & Moazed, 2007; Yang & Xu, 2013). The ELM2 domain functions as a transcriptional repression domain via recruiting HDACs (Ding, Gillespie, & Paterno, 2003; Wang, Charroux, Kerridge, & Tsai, 2008). The ELM2 domain is also found transcription corepressors with proteins containing SANT domains (Ding et al., 2003). More recently, the ELM2-SANT combined domain has been shown to act as a binding scaffold for various transcriptional cofactors, and thus, allow proteins to modify chromatin structures leading to modulation of gene transcription (Wang et al., 2008). Consistent with this notion, recent structural studies of the HDAC1–MTA1 complex reveal that MTA1 makes extensive contacts with the HDAC through the ELM2 as well as SANT domains (Millard et al., 2013). The presence of multiple functional domains including the BAH domain, the ELM2, and the SANT domains in the MTA family members suggests a potential role of these coregulatory proteins in the regulation of target gene transcription.

3 Regulation of Gene Expression by MTA Proteins

Cancer development and progression are tightly regulated by cell context-dependent transcriptional programs, which involves a dynamic fine balance between the packaging of regulatory sequences into chromatin and allowing transcriptional regulators to gain access to such sequences (Cairns, 2009; Voss & Hager, 2014). By its very nature, the highly condensed structure of chromatin generally limits the accessibility of transcription factors to the DNA, and thus, inhibits gene transcription. To overcome these barriers, cells use two major mechanisms...