The multiple facets of the SMC1A gene

Abstract

Structural Maintenance of Chromosomes (SMCs) are part of a large family of ring complexes that participates in a number of DNA transactions. Among SMCs, SMC1A gene is unique. It encodes a subunit of the cohesin-core complex that tethers sister chromatids together to ensure correct chromosome segregation in both mitosis and meiosis. As a member of the cohesin ring, SMC1A takes part in gene transcription regulation and genome organization; and it participates in the DNA Damage Repair (DDR) pathway, being phosphorylated by Ataxia Telangiectasia Mutated (ATM) and Ataxia Telangiectasia and Rad3 Related (ATR) threonine/serine kinases. It is also a component of the Recombination protein complex (RC-1) involved in DNA repair by recombination. SMC1A pathogenic variants have been described in Cornelia de Lange syndrome (CdLS), a human rare disease, and recently SMC1A variants have been associated with epilepsy or resembling Rett syndrome phenotype. Finally, SMC1A variants have been identified in several human cancers. In this review, our current knowledge of the SMC1A gene has been summarized.

1. Introduction

The Structural Maintenance of Chromosomes (SMC) family has long been known to play roles in both mitotic and meiotic macroscale chromosome organization in a dynamic manner (Yatskevich et al., 2019). In eukaryotes there are four SMC complexes: cohesin, condensin, the Smc5/6 complex, and the dosage compensation complex (DCC). Cohesin is required to hold sister chromatids together (Yatskevich et al., 2019), whereas condensin provides compaction and elasticity to chromosomes in mitosis, and also regulates sister chromatid resolution (Yuen and Gerton, 2018). The Smc5/6 complex plays a role in DNA repair, genome stability, and in the transcription inhibition of hepatitis B virus genomes (Aragon, 2018). Finally, DCC is best known for heterochromatin formation and gene silencing of the X chromosome of Caenorhabditis elegans (Chuang et al., 1994; Ercan et al., 2007; Kramer et al., 2015).

SMC1A (also called SMC1L1, MIM #300040) is evolutionarily conserved and codes a structural component of the cohesin complex, which ensures correct chromosome segregation (Yatskevich et al., 2019). As a cohesin subunit, it also plays a role in 3D genome organization and gene expression (Rao et al., 2014; Schwarzer et al., 2017). In addition, ample experimental evidence indicates that SMC1A participates in the DNA repair pathway and genome stability maintenance (Jessberger et al., 1993; Jessberger et al., 1996b; Musio et al., 2003; Musio et al., 2005). The identification of SMC1A pathogenic variants in both human rare diseases and cancer (Musio et al., 2006; Deardorff et al., 2007; Barber et al., 2008; Cucco et al., 2014; Sarogni et al., 2019b) further supports its pivotal role in fundamental cellular biological processes. This review focuses on SMC1A structure and function and its involvement in human disorders.

2. SMC1A gene and protein structure

The amazing story of SMC1A gene began in 1985 with the isolation of a mutant in Saccharomyces cerevisiae that was impaired in the maintenance of an artificial centromeric minichromosome. This gene was named Smc1 due to its role in the stability of mini chromosomes (Larionov et al., 1985). Once cloned, the sequence revealed that homologous proteins were evolutionarily conserved in bacteria, archaea and eukaryotes, raising the possibility that SMC belonged to a superfamily of proteins (Notarnicola et al., 1991; Strunnikov et al., 1993). In mammals, SMC1A was first identified during the screening of a cDNA library deriving from human lymphocytes (Lafreniere et al., 1991) and then was also isolated in the mouse homolog (Sultana et al., 1995). SMC1A maps to Xp11.22 in a region that escapes X inactivation (Brown et al., 1995), while Smc1a mouse gene is located at XF band and is subjected to inactivation (Sultana et al., 1995). According to Ensembl (http://www.ensembl.org/Homo_sapiens/Gene/Summary?g=ENSG00000072501;r=X:53374149-53422728), SMC1A has six transcripts (Fig. 1). Five of them are protein coding (SMC1A-201, SMC1A-202; SMC1A-203 and SMC1A-206), although one encodes a very short protein (SMC1A-204), while the sixth is a lncRNA (SMC1A-206). The transcript SMC1A-201 encodes a protein, highly conserved during evolution, comprising 1233 amino acids (aa) (Fig. 1, Table 1). At present, it is not clear whether the other transcripts have a biological role. SMC1A gene is ubiquitously expressed (Bioproject PRJEB4337) with higher expression in lymph nodes (read per kilobase of exon per million fragments mapped (RPKM) 13.2)), appendix (RPKM 13.1) and colon (RPKM 9.06) (Fagerberg et al., 2014).

Fig. 1.

SMC1A gene is located on Xp11.22, from 53,374,149 to 53,422,728, and spans about 48.5 kb (assembly: GRCh38.p13, sequence: NC_000023.11). According to Ensembl, it has six transcripts: SMC1A-201, SMC1A-202; SMC1A-203, SMC1A-204, SMC1A-205 and SMC1A-206. The transcript SMC1A-201 is composed of 25 coding exons (black boxes).

Table 1.

Features of SMC1A transcripts.

Name	Transcript ID	Transcript length (bp)	Protein length (aa)	Translation ID	Biotype
SMC1A-201	ENST00000322213.8	9784	1233	ENSP00000323421.3	Protein coding
SMC1A-202	ENST00000375340.10	9930	1211	ENSP00000364489.7	Protein coding
SMC1A-203	ENST00000428014.1	868	262	ENSP00000413509.2	Protein coding
SMC1A-204	ENST00000463684.1	880	54	ENSP00000476958.1	Nonsense mediated decay
SMC1A-205	ENST00000469129.1	847	No protein		lncRNA
SMC1A-206	ENST00000470241.2	838	279	ENSP00000476416.1	Protein coding

SMC1A protein consists of five distinct regions: two coiled-coil domains, the hinge domain and the N- and C-terminal domains. The N-terminal domain includes a nucleoside triphosphate (NTP) binding motif (P-loop or Walker A box), and is responsible for binding ATP. Instead, the C-terminal holds the DNA binding domain DA box (36 aa, also called Walker B). Between the N- and C-terminal domains are two long coiled-coil motifs, constituted by 362 and 354 aa respectively, whereas the hinge domain of 115 aa joins the two coiled-coil domains (https://www.uniprot.org/uniprot/Q14683).

3. Canonical role of cohesin

The hinge domain is thought to contribute to SMC1A flexibility and to allow interaction with other proteins, in particular with SMC3 which shares with SMC1A the same protein organization.

The SMC1A and SMC3 proteins fold back upon themselves through interaction of the antiparallel coiled-coil domains assuming a rod-shaped conformation with the ATP binding head domain at the one end of the rod and the hinge domain positioned at the other end of the rod.

(Fig. 2A). The cohesin complex is formed when the C-terminal domain of RAD21, belonging to the SMC-interacting kleisin protein superfamily, binds to the SMC1A head while its N-terminal domain binds to the proximal coiled-coil domain of SMC3 (Schleiffer et al., 2003; Haering et al., 2004; Gligoris et al., 2014; Huis in 't Veld et al., 2014). Finally, the RAD21 subunit associates with SA1/2 (Fig. 2B). The binding and release of the cohesin complex is performed by a number of regulator factors including NIPBL (Nipped-B like protein) which is required for cohesin loading onto chromatin (Ciosk et al., 2000).

Fig. 2.

Cohesin structure. (A) Coiled-coil structure, ATPase head and hinge domain of the SMC1A-SMC3 heterodimer. SMC1A (black protein) and SMC3 (red protein) fold back on themselves at a hinge, forming rod-like antiparallel coiled-coils that hold the N and C termini together. The two coiled-coil regions are thought to be involved in protein-protein interactions, whereas it has been proposed that the N- and C-terminal sequences of the proteins are brought together, thus forming a functional ATPase domain (Saitoh et al., 1994). The function of the hinge domain is more enigmatic. Experimental evidence suggests that cohesin loading and translocation are mediated by conformational changes in the hinge of cohesin, driven by cycles of ATP hydrolysis (Srinivasan et al., 2018). (B) Multi-subunit ring like structure of the cohesin complex. The kleisin subunit RAD21 (blue protein) links the SMC1A head and the proximal coiled-coil domain of SMC3 while the SA1 or SA2 (green protein) cohesin subunit interacts with RAD21. (C) Localization of the 46 SMC1A missense/nonsense variants identified in CdLS patients and reported in HGMD. (D) Localization of SMC1A missense/nonsense variants detected in human colorectal cancer samples (Barber et al., 2008; Cucco et al., 2014; Sarogni et al., 2019b). CdLS and cancer samples do not share any SMC1A variant, with the exception of E493A. The various boxes representing the various protein domains do not reflect their actual sizes.

Cohesin ensures proper chromosome segregation in mitosis and meiosis by holding sister chromatids together using a topological mechanism (Yatskevich et al., 2019). The loading of DNA into the cohesin complex requires the transient opening of a ring interface in cohesin likely mediated by the coiled-coil domains of SMC heterodimer (Gruber et al., 2006; Murayama and Uhlmann, 2015; Burmann et al., 2017; Chapard et al., 2019). It has been proposed that the coiled-coil arm of SMC acts as a single functional unit (Burmann et al., 2017). This notion is supported by the observation that aa sequences of SMC coiled coil were found to be conserved well beyond the levels observed for spacer rods (White and Erickson, 2006).

Cohesin is also involved in meiotic chromosome segregation. Mammalian meiocytes express two distinct versions of cohesin complexes, which feature either SMC1A or SMC1B (also called SMC1L2). SMC1B maps on human chromosome 22, at 22q13.31, and encodes a protein involved in sister chromatid cohesion, chromosome synapsis, and protects telomeres from rearrangement (Novak et al., 2008; Adelfalk et al., 2009; Biswas et al., 2013; Murdoch et al., 2013; Biswas et al., 2018). SMC1B is thought to be meiotic-specific; however, it has been shown that it is also expressed in mitotic cells and is mutually exclusive with SMC1A (Mannini et al., 2015). It is unclear why two variants of SMC1 evolved. In meiosis, SMC1B appears to be more abundant than the SMC1A variant. At meiotic entry, SMC1B becomes expressed and remains present on the centromeres until the metaphase/anaphase II transition, whereas SMC1A is more prominent on the chromatin (Revenkova et al., 2001; Revenkova et al., 2004). Though it is possible that some functions of SMC1B could be supplied by SMC1A, it has recently been demonstrated that SMC1A cannot protect telomeres from damage (Biswas et al., 2018).

4. Non-canonical functions of the cohesin complex

4.1. Genome organization and gene expression

In addition to entrapping DNA, the ability to translocate along DNA is a further peculiar feature of the cohesin complex. Specifically, cohesin creates chromosome domain structure partitioned at a submegabase scale into a sequence of self-interacting regions, called topologically associating domains (TADs) or domain boundaries, interchromatid contacts mediating distant-element interactions for the purposes of transcriptional regulation and intrachromatid contacts for the purposes of sister chromatid cohesion (Wendt et al., 2008; Hadjur et al., 2009; Dixon et al., 2012; Nora et al., 2012; Seitan et al., 2013; Sofueva et al., 2013). Most TADs or domain boundaries are strongly enriched for CTCF (CCCTC-binding factor), an 11-zinc finger DNA-binding protein that serves to insulate boundary elements (Hark et al., 2000; Kanduri et al., 2000; Ghirlando and Felsenfeld, 2016). In silico simulations suggest a model according to which chromosomal domains are formed when loop extrusion factors, such as cohesin, translocate along DNA until they encounter a boundary element, as well as CTFC, that inhibits further translocation (Alipour and Marko, 2012; Fudenberg et al., 2016; Goloborodko et al., 2016). It has been proposed that cohesin facilitates interactions between enhancers and promoters (Kagey et al., 2010; Merkenschlager and Nora, 2016). In this process, loop domains prevent enhancers from forming incorrect interactions with targets that are placed in a different loop domain (Lupianez et al., 2015; Flavahan et al., 2016). The removal of cohesin or the depletion of NIPBL results in abnormal DNA domain topology as loop domains spanning multiple compartment intervals lead to mixing among loci in different compartments (Rao et al., 2017; Schwarzer et al., 2017). Loop loss led to gene expression dysregulation though most of the change in transcription was modest. Interestingly, more genes were down-regulated (61%) rather than up-regulated (39%), suggesting that cohesin-associated loops may play a dual role: favoring the enhancer-promoter interaction and blocking the activation of promoters by inappropriate enhancer (Rao et al., 2017).

4.2. SMC1A and centrosomes

SMC1A associates with centrosomes throughout the cell cycle and localizes at the midbody of the cytokinetic cells and at the base of cilia in ciliated epithelia (Guan et al., 2008). Experimental evidence indicates that SMC1A interacts with the RAE1 (RNA export factor 1) that is localized at the spindle poles during mitosis. The dysregulation of SMC1A leads to cytokinesis defects and multipolar spindles supporting a model where cohesin would function to embrace microtubules at the spindle pole (Wong, 2010a; Wong, 2010b).

4.3. DNA repair and genome stability

As a member of the cohesin complex, SMC1A plays a role in the DNA damage response pathway. The down-regulation of SMC1A by oligonucleotide antisense or RNA interference induces genome instability in human cells, and CdLS cells harboring SMC1A variants show a high degree of spontaneous chromosome aberrations (Musio et al., 2003; Musio et al., 2005; Revenkova et al., 2009; Mannini et al., 2012; Cukrov et al., 2018). Furthermore, following DNA damage induced by chemical treatment or ionizing radiation, SMC1A is phosphorylated on Ser957 and Ser966 residues. This phosphorylation is operated by the Ataxia Telangiectasia Mutated (ATM) kinase in presence of double strand breaks (DSBs) and by the Ataxia Telangiectasia and Rad3 Related (ATR) kinase after single strand breaks (Kim et al., 2002; Yazdi et al., 2002; Wakeman et al., 2004; Musio et al., 2005; Wakeman and Xu, 2006). Cohesin is recruited at DNA damage sites to establish de novo connection with DNA (Potts et al., 2006). Cohesin could stabilize DNA following damage and allow the recruitment of proteins involved in cell cycle checkpoints and of the enzymatic apparatus responsible for DNA repair (Watrin and Peters, 2009; Ochs et al., 2019). Failure of this pathway leads to genomic instability (Musio et al., 2005; Ochs et al., 2019).

SMC1A is also a component of the Recombination protein complex (RC-1). Unfortunately, RC-1 is little investigated. It has a molecular weight of about 550–600 kDa and is formed by SMC1A, SMC3, DNA ligase III, DNA polymerase ε and a 5′–3′ exonuclease. It is thought that RC-1 is involved in the repair of DNA deletion and DSBs by recombination (Jessberger et al., 1993; Jessberger et al., 1996a; Jessberger et al., 1996b; Stursberg et al., 1999).

5. SMC1A and human disorders

The idea that cohesin plays further roles beyond sister chromatid cohesion became clear in 2006 when pathogenic variants in SMC1A were identified in Cornelia de Lange syndrome (CdLS, MIM #300590) (Musio et al., 2006), whose subjects do not show sister chromatid cohesion defects (Castronovo et al., 2009; Revenkova et al., 2009). CdLS is a human rare developmental disease characterized by pre- and post-natal growth retardation, intellectual disability, dysmorphic features, and multiorgan abnormalities (Kline et al., 2018; Sarogni et al., 2019a). CdLS is caused mainly by variants in the cohesin core subunits (SMC1A, SMC3, RAD21) or in cohesin-associated factors (NIPBL, HDAC8, BRD4) (Kline et al., 2018). Interestingly, NIPBL interacts with BRD4 and BRD4 missense variants retain the ability to coimmunoprecipitate with NIPBL, but bind poorly to acetylated histones. This finding suggests that sequestration of NIPBL underlies the pathogenic mechanism (Olley et al., 2018).

Up to now, 104 SMC1A variants have been reported in the Human Genome Mutation Database and 60 of them are associated with CdLS (Human Gene Mutation Database (http://www.hgmd.cf.ac.uk/ac/index.php December 2019, Table 2) accounting for about 5% of CdLS cases. Most of the SMC1A variants (76.6%, 46 out of 60) identified in CdLS patients are missense/nonsense. In addition, the majority of variants are de novo and germinal mosaicism has been described in about 3–5% of cases (Slavin et al., 2012). Amino acid changes encompass all gene domains without any mutational hotspots (Fig. 2C). SMC1A variants cause a very mild clinical phenotype with absence of major structural anomalies typically associated with CdLS and a slight impairment in cognitive development (Mannini et al., 2013). Furthermore, female patients are less affected than mutated males (Borck et al., 2007; Deardorff et al., 2007). Since the SMC1A escapes X inactivation, it is likely that the mechanism in affected females is due to a dominant-negative effect of the altered protein. CdLS etiopathogenetic mechanisms are still poorly understood. Of note, CdLS cells carrying SMC1A variants show gene expression dysregulation at both mRNA and protein levels (Gimigliano et al., 2012; Liu et al., 2009; Mannini et al., 2015). Mutant SMC1A proteins likely produce functional cohesin complexes (Musio et al., 2006); however, it has been posited that their mutations may alter their chromosome binding dynamics (Deardorff et al., 2007; Marcos-Alcalde et al., 2017). In particular, it has been hypothesized that SMC1A variants at the boundary of the hinge domain disrupt ATP hydrolysis kinetics or DNA binding, whereas those mapping at the coiled-coil domains affect the binding of accessory proteins to the cohesin ring (Deardorff et al., 2007). According to the role of cohesin in gene transcription regulation, it is possible to speculate that SMC1A variants affect DNA topography domains. This notion is supported by the findings that the dynamic association between the cohesin complex and DNA is affected by SMC1A mutations in CdLS probands impairing the recruitment of RNA polymerase II (Mannini et al., 2015; Revenkova et al., 2009).

Table 2.

SMC1A variants associated with CdLS.

Mutation type	Number of mutations
Missense/nonsense	46
Splicing substitutions	2
Small deletions	8
Small insertions/duplications	1
Gross deletions	1
Complex rearrangements	2

Recently, exome sequencing has also shed light on SMC1A variants in individuals with phenotypes that do not resemble CdLS. In fact, SMC1A missense variants have been described in two brothers with clinical diagnosis of Wiedemann-Steiner syndrome (WDSTS, MIM # 605130) suggesting that different mutations in SMC1A contribute to a spectrum of phenotypes (Yuan et al., 2015). Furthermore, females with SMC1A truncation variants were described to have a clinical phenotype different from CdLS with moderate to severe neurological impairment and pharmaco-resistant epilepsy or resembling Rett syndrome phenotype with epileptic encephalopathy, profound intellectual disability and stereotypic movements (Goldstein et al., 2015; Lebrun et al., 2015; Jansen et al., 2016; Gorman et al., 2017; Huisman et al., 2017; Symonds et al., 2017; Chinen et al., 2019). However, the frame is further complicated by the identification of SMC1A missense variants in two related subjects (mother and daughter) characterized by periodic pharmaco-resistant cluster seizures (Oguni et al., 2019) and in another patient with an epileptic encephalopathy phenotype (Huisman et al., 2017). The finding that all affected patients are female indicates that identified SMC1A variants are likely nonviable in males.

Considering the role of SMC1A in DNA repair and genome stability maintenance, it is not surprising that SMC1A variants have been identified in human cancers. In addition to SMC1A, many genes encoding for cohesin subunits or cohesin-associated factors are mutated in a wide range of human cancers (De Koninck and Losada, 2016). The first manuscript reporting SMC1A variants in cancer was published in 2008 (Barber et al., 2008). Thereafter, variants were detected in bladder, blood, brain and colon cancer (Balbas-Martinez et al., 2013; Ley, 2013; Cucco et al., 2014; Huether et al., 2014; Thol et al., 2014; Cessna et al., 2019; Sarogni et al., 2019b). In particular, SMC1A seems to play an important role in colorectal cancer development. Most of the SMC1A variants identified in colorectal cancer samples are missense (Fig. 2D). Colorectal tissue acquires extra-copies of SMC1A gene during tumorigenesis and its expression is significantly stronger in carcinomas than in normal mucosa (Sarogni et al., 2019b). In addition, the overexpression of SMC1A was identified as a predictor of poor prognosis in late-stage colorectal cancer (Wang et al., 2015). The contribution of SMC1A to tumorigenesis is not completely understood. It is possible to postulate that SMC1A variants affect the proper activity of cohesin in chromosome segregation and gene expression. This is corroborated by the findings that SMC1A variants trigger chromosome aneuploidy (Barber et al., 2008; Cucco et al., 2014) and cohesin-dependent regulation of gene expression is lost in leukemia cells (Sasca et al., 2019).

6. Conclusion and future perspectives

During the past few decades, key principles of chromosome segregation, chromatin organization and gene expression regulation have been progressively unveiled with the discovery of the cohesin complex. SMC1A variants can result in different phenotypes. Most missense variants, with rare exceptions, are responsible for the mild form of CdLS whereas the identification of loss-of-function variants in SMC1A expands the phenotype, number of genes, and mechanisms contributing to epilepsy or Rett-like syndrome. In addition, SMC1A variants have been identified in human cancers. Nevertheless, these studies have raised more questions that need to be answered. First, SMC1A variants, as well variants in other genes of the cohesin complex, cause gene expression dysregulation, which is a hallmark of CdLS cells. However, as gene expression changes are modest, it has been proposed that CdLS phenotype results from the collective consequences of multiple perturbations (Kawauchi et al., 2009; Liu et al., 2009; Mannini et al., 2015; Muto et al., 2011). Does mutated cohesin lead to global transcription disturbance or does it alter the expression of a “master” gene whose dysregulation triggers a cascade of events? c-MYC could represent this master gene in CdLS. In fact, the expression of c-MYC has been found to be regulated by cohesin, downregulated in CdLS and CdLS mouse models, and a convergent hub lying at the center of dysregulated pathways (Kawauchi et al., 2009; Rhodes et al., 2010; Gimigliano et al., 2012). Second, it is intriguing that, while CdLS cells carrying SMC1A variants display gene expression dysregulation, they show no defects in sister chromatid cohesion. Does this phenomenon uncover a dosage-sensitive functional hierarchy of cohesin? Is chromatid cohesion less cohesin dose sensitive than gene expression? Third, various members of cohesin complex, including the SMC1A subunit, are emerging actors in genome safeguarding (Mannini et al., 2010; Mannini and Musio, 2011; Cucco et al., 2014; Wang et al., 2015; Cucco and Musio, 2016; Litwin et al., 2018; Sarogni et al., 2019b). How exactly does cohesin participate in DDR? Is stabilizing DNA in proximity to damaged DNA the only role of cohesin in DDR, or do cohesin’s roles extend to joining sister chromatids? Fourth, how is cohesin dysfunction crucial for cancer development? Is it through changes in chromatin accessibility affecting transcription, sister chromatid defects or both? Answers to these questions will allow us to better understand the importance of cohesin in both human diseases and the preservation of genome stability.

Abbreviations:

AA

Amino acids

ATM

Ataxia telangiectasia mutated

ATR

Ataxia telangiectasia and rad3 related

CdLS

Cornelia de Lange syndrome

CTCF

CCCTC-binding factor

DCC

Dosage compensation complex

DDR

DNA damage repair

DSBs

Double strand breaks

HGMD

Human genome mutation database

NIPBL

Nipped-B like protein

RC-1

Recombination protein complex

RAE1

RNA export factor 1

RPKM

Read per kilobase of exon per million fragments mapped

SMC

Structural maintenance of chromosomes

SMC1A

Structural maintenance of chromosome 1A

TADs

Topologically associating domains

WDSTS

Wiedemann-Steiner syndrome