UBA2 (Ubiquitin-Like Modifier-Activating Enzyme 2)
Posttranslational modification of eukaryotic cellular proteins with small ubiquitin-like modifier (SUMO) proteins, i.e., sumoylation (GO: 0016925), has a variety of functional effects, including the modification of protein structure and the regulation of the intracellular or intranuclear localization, among others. In this manner, sumoylation contributes to the regulation of signal transduction pathways and gene expression systems related to the cell cycle control, apoptosis, cell differentiation, and the stress response, among others (Saitoh and Hinchey 2000; Johnson 2004; Schulman and Harper 2009).
Ubiquitin and ubiquitin-like protein conjugation requires the sequential action of three enzymatic activities (namely E1, E2, and E3) that are conserved across the eukaryotic evolution and were initially characterized in rabbit reticulocytes (Ciechanover et al. 1982). The overall mechanism basically consists in an initial ATP-dependent activation of the modifier by E1, and its subsequent ligation to a target protein, which is carried out by the conjugating enzyme E2 assisted by one of the several E3 ligases that confer specificity.
The E1 enzymes were initially isolated from rabbits, mice, humans, yeast, and wheat, showing a remarkable heterogeneity. The SUMO-activating enzyme subunit 2 (SAE2) was formerly identified as UBA2 (ubiquitin-like modifier-activating enzyme 2) during the initial characterization of the E1 enzymatic activity in wheat (Hatfield and Vierstra 1992). Human UBA2 (EC: 126.96.36.199) contains 640 residues and shares a high degree of sequence identity with SAE1, despite the latter being only 346 aa. These proteins form a functional heterodimeric enzyme that activates SUMO proteins in a manner analogous to the single E1 ubiquitin-activating enzymes in yeast.
Ubiquitin-like protein modifiers, collectively termed Ubls, are posttranslationally attached to substrate proteins by similar enzymatic reactions. Some Ubls may display a high degree of sequence similarity to ubiquitin, but others like SUMO-1 have roughly 18% sequence identity. Despite that their tertiary structures are very similar; SUMO and ubiquitin have distinct functions and specificities (Dohmen 2004). Four different SUMO isoforms have been detected in mammals. From these, SUMO-1 (MIM 601912) seems the most prominently conjugated isoform under normal conditions, while SUMO-2 and SUMO-3, which are very similar in sequence, appear to be preferentially conjugated to proteins under stress conditions (Saitoh and Hinchey 2000). SUMO-4 has a particular expression profile, restricted to certain tissues such as kidney. In general terms, sumoylation appears to be a highly selective process both at the substrate level, at temporal distribution, and at subcellular localization. For instance, the sumoylation/desumoylation cycle is dynamic, and for many substrates it appears to be synchronized with the cell cycle. Moreover, the spatial distribution of the conjugating and deconjugating enzymes is consistent with a role in the cytosol/nucleus transit (Dohmen 2004).
Role of UBA2 in the Sumoylation Pathway
Protein modification by the addition of SUMO proteins requires the ATP-dependent initial activation of SUMO by the ubiquitin-activating enzyme E1 (reviewed in Schulman and Harper 2009), which in human is constituted by a heterodimer of SAE1 and UBA2 (Okuma et al. 1999). This heterodimeric enzyme can activate SUMO1, SUMO2, SUMO3, and probably SUMO4.
Once SUMO is activated, E2 enzymes like human UBC9 (codified by gene UBE2I) are then recruited to transfer the E1-SUMO thioester adduct to a conserved cysteine in E2, forming an E2-SUMO thioester adduct via transesterification (Fig. 1b). E2s are ubiquitin-conjugating enzymes that directly ligate SUMO to the lysine residues of the target protein by amide (isopeptide) bonds, with or without the assistance of E3 protein ligases. The majority of SUMO-accepting lysine residues in the target proteins lie within the consensus sequence ΨKXE, where Ψ corresponds to an aliphatic residue, preferably leucine, isoleucine, or valine, and X represents any residue (Johnson 2004).
Interestingly, UBA2 is autosumoylated at lysine 236 and also at several lysines surrounding the catalytic cysteine by UBC9. This modification does not affect the formation of the E1-SUMO thioester but inhibits the transfer of SUMO from E1 to E2, thus affecting the downstream conjugation pathway. This mechanism is proposed to generate pools of E1s with distinct capabilities to respond to environmental changes, such as heat shock (Truong et al. 2012).
Biological Function of UBA2
UBA2 is an essential element of the sumoylation machinery, as it participates specifically at the activation step of SUMO proteins. Hence, the biological processes indirectly affected by UBA2 activity are numerous, as a wide variety of target proteins and pathways are regulated by SUMO modifications. Unlike ubiquitination, which mainly targets proteins for degradation, sumoylation is involved in signal transduction regulation, nuclear transport, and transcriptional regulation. In addition, a number of sumoylation targets that participate in DNA repair and chromosome segregation have been identified.
In general terms, sumoylation affects the localization of proteins, either by regulating their cytosol-nucleus trafficking or by targeting proteins to subnuclear structures, such as nuclear dots, promyelocytic leukemia (PML) nuclear bodies, among others. Frequently, SUMO modification of transcription factors and coregulators results in a negative regulatory effect, either by directly affecting their structure and promoter binding or by modulating their nuclear distribution. Transcription factors such as p53, Elk-1, the androgen receptor, STAT-1, LEF1, and Sp3 are regulated by sumoylation. Examples that suggest possible positive regulatory mechanisms of sumoylation include the observed enhancement of the in vitro DNA-binding activity of heat shock factors HSF1 and HSF2, which correlates with their sumoylated status and activation in vivo, and the observed decrease of beta-catenin-dependent TCF-4 transcriptional activity when the latter lacks the SUMO attachment sites (Hong et al. 2001; Yamamoto et al. 2003). Many more examples of targets and pathways regulated by sumoylation have been extensively reviewed elsewhere (Dohmen 2004; Johnson 2004). Moreover, it has been shown that E1 and E2 enzymes are essential for the maintenance of epigenetic silencing, possibly by increasing histone deacetylase 1 (HDAC1) activity via sumoylation (Poleshko et al. 2014).
Yeast mutants deficient in E2 or Smt3 (the yeast SUMO protein) display defects in the cell cycle progression, apparently due to their inability to degrade mitotic cyclins and the increase of securin Pds1 (Dieckhoff et al. 2004). These defects were partially suppressed in mutants of the deconjugationg enzyme Ulp2, indicating that the sumoylation state of the targets is precisely regulated (Schwienhorst et al. 2000). Other works with yeast mutants have implicated sumoylation in the DNA damage response. The replication factor PCNA is sumoylated during the S phase of the cell cycle and upon exposure to lethal doses of methyl methanesulfonate, and this modification apparently activates PCNA for participation in DNA repair processes mediated by the translesion DNA polymerases (Hoege et al. 2002). Interestingly, the acceptor lysine residue for SUMO is the same that is used for ubiquitin modification, suggesting that sumoylation may antagonize the ubiquitin degradation pathway in specific cases.
UBA2 Mutations in Human Pathogenesis
Taking into account the biological functions affected by sumoylation, mutations in different SUMO pathway elements have been implicated in several neoplastic diseases and developmental disorders. Specifically, UBA2 mutations have been described in mammary tumors, hepatocellular carcinoma, and small cell lung carcinoma. At the molecular level, the E1 activity has been related to Myc-dependent tumor growth in a human mammary epithelial cell model. Genetic inactivation of UBA2 switches the Myc oncogenic transcriptional program from activated to repressed, leading to alterations in the mitotic spindle, mitotic catastrophe, and subsequent apoptosis selectively upon Myc hyperactivation. These data indicate that the sumoylation pathway is one of several unexplored underlying processes supporting an oncogenic program, and its inhibition can impair Myc-driven tumors. Therefore, sumoylation agents are suggested as potential therapeutic targets to be investigated (Kessler et al. 2012). In an apparently independent signaling mechanism, the growth of Notch-activated breast epithelial cells is as well decreased by inhibition of the E1 heterodimer with ginkgolic acid (Licciardello et al. 2015). In addition, thorough multi-omics analyses of the MCF10, isogenic model of breast cancer progression revealed MYC amplification, and previously undescribed UBA2–PDCD2L expressed in-frame fusion genes in MCF10Ca1h malignant cells, corresponding to the invasive carcinoma stage (Maguire et al. 2016). The molecular significance of such genetic fusion remains unclear.
The observation that UBA2 is highly expressed in liver cancer cells and clinical samples was investigated in HepG2 cells. Silencing UBA2 resulted in some extent of suppression of tumor cell proliferation, and further identification of the sumoylation targets revealed that TFII-I transcriptional activity, which is critical to promote cell proliferation and colony formation, is increased by sumoylation via a decreased binding to its repressor HDAC3 (Tu et al. 2015). UBA2 is highly expressed in small cell lung cancer as well, and its high expression levels correlate with tumorigenesis idem. RNAi-mediated silencing of UBA2 expression in H446 cells inhibited migration and invasion, simultaneously increasing the sensitivity to etoposide and cisplatin (Liu et al. 2015).
Several transcription factors related to mammalian development such as the androgen receptor, SOX9, and SF1 are targets of sumoylation. Hence, a possible role of UBA2 mutations in human congenital malformations has been suggested. For instance, the heterozygous deletion of UBA2 and its consequent haploinsufficiency has been associated to genitourinary phenotypes such as hypospadias, which have been associated with the 19q13.11 microdeletion syndrome in male patients, a condition mainly characterized by cutis aplasia and intellectual disability (Venegas-Vega et al. 2014).
Posttranslational modification of proteins with small ubiquitin-like modifier (SUMO) proteins is an important event that allows the regulation of many cellular processes. The UBA2 enzyme was originally isolated from wheat, and together with SAE1, it forms the heterodimeric enzyme E1, which carries out the first stage of the sumoylation process that consists in the activation of the SUMO proteins. This is achieved in two steps: the initial adenylation of the C-terminal glycine of SUMO, followed by the formation of a thioester bond between such glycine and a conserved cysteine residue at position 173 of UBA2. Once activated, the SUMO protein is transferred to the E2 conjugating enzyme that ligates the modifier to a target protein, with the assistance of one of several E3 ligases that may confer specificity. Sumoylation affects protein conformation and localization, and several signaling pathways and transcription factors related to the cell cycle control, cell differentiation, apoptosis, and DNA repair, among other biological processes, are regulated by this mechanism. The activities of UBA2 and the overall sumoylation pathway have been shown to support Myc- and Notch-driven neoplastic diseases. Hence, intense research on the utility of sumoylation machinery elements as cancer biomarkers or potential therapeutic targets is currently under development. In addition, haploinsufficiency of UBA2 has been related to genitourinary malformations in males affected by the 19q13.11 microdeletion syndrome. The increasing interest in understanding the role of SUMO modification in mammalian and plant development will demand future research.