Cancer Letters

Cancer Letters

Volume 316, Issue 2, 28 March 2012, Pages 113-125
Cancer Letters

Mini-review
SUMOylation in carcinogenesis

https://doi.org/10.1016/j.canlet.2011.10.036Get rights and content

Abstract

SUMOylation is a post-translational modification characterized by covalent and reversible binding of small ubiquitin-like modifier (SUMO) to a target protein. In mammals, four different isoforms, termed SUMO-1, -2, -3 and -4 have been identified so far. SUMO proteins are critically involved in the modulation of nuclear organization and cell viability. Their expression is significantly increased in processes associated with carcinogenesis such as cell growth, differentiation, senescence, oxidative stress and apoptosis. Little is known about the role of SUMOylation in cancer development. Therefore the present review focuses on possible implications of SUMOylation in carcinogenesis highlighting its impact as an important regulatory cell cycle protein. Moreover, novel opportunities for therapeutic approaches are discussed. The differential expression levels, the target protein preferences and the function of the SUMO pathway in different cancer subtypes raises unexpected issues questioning our understanding of the implication of SUMO in carcinogenesis.

Introduction

SUMOylation is a post-translational modification characterized by covalent and reversible binding of small ubiquitin-like modifier (SUMO) to a target protein. Initially, four isoforms of SUMO termed SUMO-1, -2, -3 and -4 were identified [1], [2]. These proteins are ∼12 kDa in size [3], and have a similar three-dimensional structure compared to ubiquitin [4], [5], [6]. Unlike ubiquitin SUMO proteins have a 10–25 amino-acid tail at their N-terminal domains [7]. SUMO, a member of the Ubl (ubiquitin-like proteins) family shares under 20% identity with ubiquitin, and SUMO-2/3 are 95% identical to each other and 50% identical with SUMO-1 [7]. SUMO was first detected in mammals as a protein covalently bound to the GTPase activating protein RanGAP1 [8], [9]. The present review summarizes the major characteristics of SUMOylation and focuses on some more recent developments regarding the impact of SUMOylation on cancer pathogenesis.

Since SUMO competes with ubiquitin for substrate binding, SUMOylation of proteins does not lead to proteasomal degradation [10]. SUMO is essential for the maintenance of protein stability, transcriptional regulation, modification of particular transcription factors such as glucocorticoid receptor (GR), Myb, CAAT/enhancer binding protein (C/EBP) and SP3 [11]. It is well established that SUMO efficiently reduces transcriptional activity [11]. Induction of several gene promoters (e.g. for heat shock factors Oct4 and Smad4) through SUMO requires histone deacetylase (HDAC) activity [12], [13]. SUMO is involved in cellular processes like mitosis, cell development and differentiation, senescence and apoptosis [14], [15], [16]. Alteration of SUMO-1 expression caused by SUMO-1 haploinsufficiency or disruption of the SUMO-1 locus was linked to the presence of cleft lip and palate [17]. Recent evidence suggests that SUMO-2/3 is involved in regulating mitosis and proliferation (Fig. 1) [11]. Borealin, the inner centromere protein (INCENP), Survivin and Aurora B are components of the chromosomal passenger complex (CPC). Aurora B kinase plays an important role during cytokinesis by connecting the mitotic spindle to the centromere. The activity of Aurora B correlates with cell cycle progression and is linked to microtubules through chromosome movement and segregation [18]. Survivin negatively regulates apoptosis. The CPC coordinates crucial mitotic events such as spindle assembly checkpoint and cytokinesis [19]. Borealin is a main target of SUMO-2/3 and its modification by SUMO-2/3 is regulated particularly during early mitosis (G2/M phase). Importantly, the SUMO E3 ligase RanBP2 interacts with the CPC and mediates SUMO-2/3 binding to Borealin. Removal of SUMO-2/3 from Borealin is catalyzed by the cysteine-protease Sentrin-specific protease 3 (SENP3). SUMOylation is essential for the formation and stabilization of the Borealin-SUMO-2/3 complex in early mitosis during connection of the CPC with the centromere [19]. Silencing of SUMO-2/3 hinders conjugation of SUMO-2/3 to its substrates, and blocks cell division in SUMO-2/3 miRNA expressing cells [20]. Recently, single-nucleotide polymorphism screens identified human SUMO-4 to be potentially involved in progression of type I diabetes [21], [22].

Activation of SUMO precursors is required prior to protein SUMOylation by cleavage of the immature SUMO at the C-terminal residue. This initial step, mediated by the cysteine proteases ubiquitin-like-protein-processing (Ulps) enzymes in yeast and SENPs in mammals results in generation of a free carboxy terminal glycine residue [23]. These residues are necessary for the interaction between SUMO and its target proteins. Similar to ubiquitination the SUMOylation cascade contains a SUMO-activating enzyme (E1), a SUMO-conjugating enzyme (E2) and different SUMO ligases (E3s) [23]. The SUMO specific E1 activating enzyme has been described in yeast as well as in humans [23] and is composed of the two proteins small ubiquitin-like-modifier-activating enzyme E1 (SAE1) and SAE2, both known as activator of SUMO-1 (AOS1) and ubiquitin-like modifier activating enzyme 2 (Uba2) in yeast [23]. The E1 enzyme mediates an ATP-dependent activation of a mature SUMO protein and forms a SUMO-adenylate conjugate whereas the C-terminal carboxyl group of SUMO binds to adenosine monophosphate (AMP) releasing pyrophosphate [7], [23]. Due to the reactive thiol group of the cysteine in the E1 enzyme a high energy thiolester bond between the SUMOadenylate and the E1 enzyme is generated paralleled by the release of AMP [23]. Subsequently, SUMO is linked to the catalytic cystein of the E2 conjugating enzyme Ubc9 by E1 forming a transitional thiolester between Ubc9 and SUMO [23]. Ubc9 is the only known E2 conjugating enzyme in the SUMO pathway and has been found in yeast, invertebrates and vertebrates [24], [25], [26]. This clearly differs from the ubiquitin pathway where several E2 enzymes mediate substrate ubiquitination [23]. Ubc9 drives the transfer of SUMO to its target proteins by generating an isopeptide bond between the C-terminal glycine residue of SUMO and a lysine side chain of the substrate [7]. This step is co-catalyzed by specific E3 ligating enzymes which promote the translocation of SUMO from Ubc9 to a target protein (Fig. 1) [27]. In contrast to ubiquitination, SUMOylation of target proteins needs only one E2 enzyme and a small number of known E3 ligating enzymes. SUMO E3 ligases are divided into three classes [Siz/PIAS-RING (SP-RING), Ran-binding protein (RanBP2) and Human polycomb 2 (Pc2)] [23].

Three different groups of SUMO E3 ligases have been identified so far (Table 1). The largest group consists of E3 ligases containing a SP-RING motif which is crucial for their function [27]. One of these SP-RING containing E3 ligases is represented by protein inhibitor of activated STAT (PIAS) family proteins, termed Siz proteins in yeast. Next to the SP-RING domain, these proteins contain a conserved ∼400 amino acid long N-terminal domain and a SAP domain (SAR, Acinus, PIAS) which is involved in binding AT-rich DNA sequences [28], [29], [30], [31], [32]. At present, four PIAS family members [Siz1, Siz2, zipper protein (Zip) 3 and methyl methanesulphonate-sensitivity protein (Mms) 21] are known in Saccharomyces cerevisiae and five in mammals (PIAS1, PIAS3 and the splice variants PIASxα, PIASxβ and PIASy) [30], [33], [34], [35], [36], [37], [38], [39]. PIAS proteins might presumably SUMOylate several substrates as shown by an analysis of Siz1 mediated SUMOylation of Septins and Proliferating cell nuclear antigen (PCNA) [23]. Furthermore, PIAS proteins do not share the same substrate. For example PIAS1 and PIASxβ, but not PIASxα, catalyze SUMOylation of Mdm2 [40]. Overexpression of different PIAS proteins such as PIAS1, PIAS3 and PIASy promoting SUMOylation of vertebrate-derived proteins like p53 has been demonstrated in vitro [33], [38].

Different studies implicate a mechanistically similar function of the SP-RING domain compared to the RING domain in ubiquitin RING E3 ligases [41], [42], [43]. SP-RING ligases directly conjugate their substrates to Ubc9 and bind SUMO non-covalently by a SUMO-interacting motif (SIM/SBM) [7]. SIMs has been found in E1, different E3 ligases, SUMO substrates, SUMO binding proteins and SUMO targeted ubiquitin ligases [44], [45]. The functional relevance of E1 SIMs is still undefined despite resemblance to other SUMO-SIM complexes [46], [47].

Beside binding to SIMs, the Siz and PIAS E3 ligases bind the E2 conjugating enzyme Ubc9 and SUMO via their SP-RING and Siz/PIAS C-terminal domain (SP-CTD) [39]. The study by Yunus and Lima suggests that both, the SP-RING and SP-CTD domains, are needed to form the thioester bond between Ubc9 and SUMO, and thereby bringing Ubc9 and the substrate in a catalytic favourable position. Alternatively binding the substrate to Siz and PIAS E3 ligases is mediated by the proline, isoleucine, asparagine, isoleucine, threonine (PINIT) domain as shown for PCNA [43].

The second type of SUMO E3 ligases is the vertebrate specific RanBP2 (also known as Nup358) [48], [49], [50]. The internal repeat (IR) is the E3 domain of RanBP2 which includes two repeats of a ∼50 residue sequence (IR1 and IR2). This sequence shows no homology with any of the well established E3 ligases of the ubiquitination cascade. The E3-activity of the RanBP2 domain is located within a 33 kDa domain of RanBP2 that itself contains no RING finger domains and differs from that of the PIAS family [51]. As a functional E3 ligase in SUMOylation the IR-domain has the ability to catalyze SUMOylation of several proteins such as RanGAP1. Together with RanGAP1 and Ubc9 the IR-domain builds a trimeric complex mediating localization of SUMO-RanGAP1 to the nuclear pore [52], [53]. SUMOylation of proteins like HDAC4, Sp100, an interferon stimulated antigen, and RanGAP1, is mediated by the IR-domain in vitro [51], [52]. Currently, only one RanBP2 target has been identified in vivo [54]. Conversely, in vitro RanBP2 mediates SUMOylation of proteins like HDAC4, Sp100, Borealin and PML [19], [51], [52], [53].

Human polycomb 2 (Pc2) represents the third class of E3 ligases. Polycomb group (PcG) proteins form large multimeric complexes (PcG bodies) within the nucleus which may interact with DNA binding proteins such as Rb and E2F [55], [56]. PcG bodies are normally found in the vicinity of centromeres and probably result from formation of several DNA related groups of PcG complexes. Furthermore, Pc2 complexes are able to suppress DNA activity by broaden along the DNA template. They may also facilitate the interaction between remote DNA bound complexes by bringing them in close proximity [57]. It has been shown that PcG complexes promote gene suppression via histone methylation activities [58], [59]. Pc2 belongs to the group of human PcG proteins. Pc2 interacts with both Ubc9 and CtBP2, and stimulates the transfer of SUMO to CtBP [57]. In addition Pc2 leads to enhanced CtBP2 SUMOylation suggesting that PcG bodies might be a major site of SUMOylation [57], [60]. Currently it is not clear whether Pc2 serves as a catalyst in the transfer of SUMO from E2 to its target or if Pc2 mediates modification for example of its binding partner CtBP [57].

DeSUMOylation is mediated by specific cysteine (Cys) proteases which have been found in yeast, plant and mammalian cells [61]. These enzymes carry out two important functions. They deconjugate SUMO from the substrate and thus provide a pool of free SUMO which can be used to modify different substrates. Hence, the source of free SUMO is composed of both synthesized, immature SUMO and deconjugated SUMO.

A ∼200 amino acid long C-terminal domain, the so called ubiquitin-like protein-specific protease (Ulp) domain is responsible for SUMO deconjugation and the cleavage of SUMO precursors to produce the mature SUMO peptides [6]. Two cysteine proteases, Ulp1 and Ulp2, have been characterized in yeast. Ulp1 localizes to the nuclear pore complex and is needed for cleaving the SUMO precursor but also to remove SUMO from its substrates [62], [63], [64]. Ulp2 which is exclusively located in the nucleus is not able to cleave the SUMO precursor, and seems to deSUMOylate conjugates different from Ulp1 [62], [63], [64], [65], [66]. Moreover, it has been demonstrated that Ulp1 is needed for proper cell cycle progression. Ulp2 in contrast has an important function in maintaining chromosome stability and renewal from cell cycle arrest [63]. Deletion of Ulp1 in yeast strains leads to enhanced cellular lethality caused by cell cycle arrest, whereas deletion of Ulp2 increases sensitivity against cellular stress thereby derogating genome integrity [63], [67]. This might indicate that Ulp1 and Ulp2 are not able to functionally compensate for each other [68]. Substrate specificity of Ulp1 is mediated by its N-terminal regulatory domain whereas absence of this regulatory domain causes non-specific deSUMOylation of Ulp2 substrates and incomplete deSUMOylation of Ulp1 target proteins [68].

Six Ulp homologues were found in humans, termed SENP1-3 and SENP5-7 (Table 2). SENP members display different activities in mediating SUMO maturation and cleavage of SUMO isoforms. It has been shown that SENP1 and SENP2 specifically bind to SUMO-1 and SUMO-2/3, and mediate both processing and deconjugation of SUMO-1 and SUMO-2/3 [69], [70] whereas SENP3 and SENP5 support solely deconjugation of SUMO-2/3 from their substrates [71], [72]. Beside different preferences in modifying SUMO paralogues, distinct intracellular locations of SENP have been described. Like Ulp1 in S. cerevisiae, mammalian SENP2 is mainly located at nuclear pore complexes [68], [73], [74]. Increased amounts of SENP5 have been detected in the nucleolus [71], [75] whereas only low amounts of SENP5 were found in the cytoplasm. SENP5 is necessary for mitochondrial splitting and fusion [76]. SENP1 appears to shuttle between the cytoplasm and the nucleus [1]. Current evidence suggests that SENP1 is located in the nucleoplasm but not in the nucleolus [77], probably mediated by a concerted action of a nuclear localization signal in the N-terminus and a nuclear export sequence near the C-terminus integrated in SENP1 [78], [79]. A shared conserved catalytic domain with different N-termini, mainly influencing its intracellular distribution, is characteristic for all SENPs [80].

Three SENP families are distinguished. The first one, represented by SENP1 and SENP2, specifically binds to mammalian SUMO-1 to -3. SENP3 and SENP5 (there is no SENP4) belong to the second family and preferentially target SUMO-2/3 and are located in the nucleolus. SENP6 and SENP7 are members of the third family, and contain an extra loop inserted in the catalytic domain and, similar to SENP3 and SENP5, predominantly bind SUMO-2/3 [61]. On the other hand SENP6 and SENP7 are moderately engaged in deconjugation of monomeric SUMO-2/3 from their substrates whereas both of them considerably conjugate/deconjugate poly-SUMO-2/3 chains. Recently, the study by Alegre and Reverter [81] identified a single sequence in SENP6 and SENP7 which is mainly responsible for their catalytic activity. This motif supports formation of an effective linkage with SUMO through the deconjugation process. Another specific motif in the SUMO surface determinant seems to be necessary for proper binding of SUMO-2/3 to SENP6 or SENP7.

SENP1 has both N-terminal localization signals (NLSs) and nuclear export signals (NESs), although SENP1 is mainly localized in the nucleus [77], [78]. Cell culture experiments with GFP-SENP1 revealed a possible shuttling of SENP1 between nucleus and cytoplasm. It has, therefore, been assumed that SENP1 localization is regulated by extracellular signalling [79], [82].

The catalytic domains of SENP1 and SENP2 are able to distinguish between the different SUMO proteins and are responsible for their substrate specificity. SENP1 was demonstrated to more likely process SUMO-1 than SUMO-2 [83], [84]. Xu and Au showed, by substituting the C-terminal sequence after the ‘GG’ region that the C-terminal fragment is essential for efficient maturation of SUMO-1 and -2 precursors. Furthermore, they revealed that the two residues located directly after the ‘GG’ region (HS, VY and VP), are responsible for the maturation of SUMO paralogues [84].

The Ran-GTPase controls several important cellular processes, such as nucleocytoplasmic transport, spindle assembly during mitosis, cell cycle control and rearrangement of the nuclear envelope after termination of mitosis [85], [86], [87]. In vertebrates Ran-GTP formation is facilitated by a guanine nucleotide exchange factor, termed RCC1 in vertebrates whereas hydrolysis of Ran-GTP is mediated by a GTPase activating protein, known as RanGAP1 in vertebrates.

SENP1 de-SUMOylates both, RanGAP1-SUMO-1 and RanGAP1-SUMO-2, with similar efficiency [83]. This is presumably caused by disruption of its nuclear localization under in vitro conditions. Since SENP1 shuttles between the nucleus and cytosol due to its nuclear localization and nuclear export sequences, it might not mediate SUMOylation of RanGAP1 in vivo, in case it is indeed restricted to the cytoplasmic fibrils of the nuclear pore complex [78], [79].

Another important regulator of transcription, genome integrity, apoptosis, reaction to viral infection and tumor suppression are the promyelocytic leukemia (PML) nuclear bodies (NBs). It is assumed that generation of NBs depends on PML [88]. This hypothesis is supported by cell culture experiments with Pml null cells where no formation of NBs was detected. Shen and colleagues hypothesized that PML-NB generation requires an interaction between both, namely a noncovalent interaction between the SUMO paralogues on SUMO modified PMLs and the SUMO binding motifs [88]. The latter are found in PMLs, and SENPs are essential for removal of SUMO [72], [89], [90], [91], [92], [93], [94].

SENP1 de-SUMOylates HDAC1, thereby diminishing its acetylase activity and thus enhances transcriptional activity. SENP1 may also play an important role as a regulator of androgen receptor (AR)-dependent transcription [95]. The AR is a member of the nuclear receptor superfamily and is a ligand-regulated transcription factor [96], [97]. Furthermore, the AR is involved in regulating important cellular processes such as cell growth, differentiation and protection of male reproductive functions [96], [97]. It has been reported that the AR binds to an androgen-response element in the SENP1 promoter and is thus believed to directly regulate SENP1 transcription [98]. Interestingly, the AR has been shown to be altered in its N-terminal domain at two conserved lysine residues, also in prostate cancer. Replacement of SUMO-1 binding sites leads to significant AR transcription activity, and SUMOylation is proposed to be a modulator of progesterone, glucocorticoid and mineralocorticoid receptors [99]. Overexpression of SENP1 in prostate cancer (PCa) cells leads to enhanced transcriptional activity of AR caused by deSUMOylation of the coregulatory protein HDAC1. Repression of endogenous SENP1 in these cells is directly linked to a significant reduction of AR expression levels [98].

SENP1 has been reported to be overexpressed in human prostate cancer. Recently an enhanced expression of SENP1 was shown to directly correlate with a permanent androgen or interleukin (IL)-6-stimulated status in prostate cancer cells. Early progression of prostatic intraepithelial neoplasia was demonstrated in SENP1 overexpressing transgenic mice, denoting SENP1 as oncogenic factor for prostate cancer [100].

Similar to HDAC1, SENP1 also facilitates deSUMOylation of SIRT1, a class III HDAC and regulator of members of the p53 family, thereby reducing its deacetylase activity [61], [101]. Moreover, SENP1 has been shown to be involved in initiating stress dependent apoptosis by deSUMOylation of sirtuin (SIRT) 1. Additional evidence implicates SENP1 to have a role in TNF signalling triggered by reactive oxygen species (ROS) [101]. During TNF signalling SENP1 translocates to the nucleus, leading to deSUMOylation of homeodomain-interacting protein kinase (HIPK) 1 and its cytoplasmic translocation, enhancing activity of the mitogen-activated protein kinase kinase kinase (MAP3K) and apoptosis signal-regulating kinase (ASK) 1. ASK1 activates important mediators of stress/apoptotic pathways such as JNK and p38 via different MAP2Ks [101].

Results from different mouse models have shown that SENP1, in addition to its apoptosis-inducing properties, is critical during embryogenesis. Reduced expression of SENP1 caused by a random retroviral insertion within the first intron leads to embryonic death at day E13.5 resulting from a pathologically changed placental blood supply [64]. Senp1 deficient mouse embryonic fibroblasts (MEFs) express significantly decreased levels of hypoxia inducible factor (HIF1)α-dependent genes such as vascular endothelial growth factor (VEGF) and glucose transporter-1 (GLUT-1) under hypoxia. These data suggest an important regulatory function of SENP1 in HIF1α signalling that is involved in erythropoiesis, angiogenesis and glycolysis [102].

Three different isoforms of SENP2 that originate from alternative mRNA splicing exist [89], [103], [104]. The SENP2/Axam isoform is characterized by a N-terminal tail which mediates the binding of Axam to the nucleoplasmic site of the nuclear pore complex [73], [74] whereas Axam2/SMT3IP2 contains a different N-terminus and is located in the cytoplasm. The third isoform of SENP2, small ubiquitin-related modifier-specific protease 1 (SuPr1) has no N-terminal domain and is situated in the PML nuclear bodies [89]. Thus SENP1 and SENP2 functionally differ. They are not able to compensate for each other in knockout embryos. Under hypoxia HIF1α was upregulated in SENP1−/− MEFs while this phenomenon did not occur in SENP2−/− MEFs, indicative of substrate specificity of SENP1 and SENP2 [102].

SENP3 and SENP5 mainly deSUMOylate SUMO-2/3 substrates [71], [72]. Both SENP3 and SENP5 are located in the nucleolus [61], [71] but small amounts of SENP5 are detectable in the cytoplasm. SENP5 is essential for mitochondrial fission and fusion [76]. Since knockdown of SENP5 via siRNA abolishes cell proliferation and promotes generation of defects in nuclear morphology, paralleled by development of binucleated cells, SENP5 might also have an impact on mitosis and cell cycle regulation [71]. In addition, overexpression of SENP5 removes SUMO from several mitochondrial substrates and prevents SUMO-1 mediated mitochondrial fragmentation but leads to the generation of non-functional mitochondria [76]. SENP5 is primarily expressed in the nucleolus and mainly deconjugates SUMO-2/3. It is unknown at present, whether SENP5 is a direct regulator of mitochondrial morphology and metabolism maintenance [76].

SENP6 and SENP7 are clearly distinct from the other SENP/ULP protease family members with regard to their conserved catalytic domain. However, both of them are located in the nucleoplasm and favourably cleave SUMO-2/3 isoforms [81], [92], [94], [105]. Moreover, it has been shown that a motif within the SUMO surface is responsible for the SUMO-2/3 isoform specificity of SENP6 and SENP7 [81]. In addition, recent studies argue for a regulatory function of SENP6 and SENP7 in PML body formation and that deletion of them cannot be compensated by the other [91], [94].

Several transcription factors like AP-2, AR, c-Jun, c/EBP, c-Myb, CREB, Erm, Ets-1, GATA-2, HSF1, IκBα, IRF-1, Pdx1, p53, Sp3, STAT1, TEL have been demonstrated to be SUMOylated in cancer (Table 1).

Section snippets

SUMO in diseases, especially in carcinogenesis

Post-translational modification is known to be critically involved in carcinogenesis as it regulates protein activity. Protein expression of SUMO isoforms has been extensively studied in various tumor entities and protein expression profiling turned out to be a useful strategy for evaluating the possible relevance of any chosen protein in cancer development. A number of different receptors or intracellular signalling molecules has been demonstrated to be modified by SUMOylation, thereby to

SUMOylation and histone deacetylases (HDACs)

Generally, histone acetyltransferases (HATs) and histone deacetylases (HDACs) control chromatin acetylation equally. At the transcriptional level, acetylation of histones is closely related to gene transcription whereas deacetylated histones are associated with transcriptionally inactive chromatin. HDACs are enzymes that remove acetyl groups from ε-N-acetyl-lysine amino acid on histones [192]. Moreover, initiation of differentiation of tumor cells, cell cycle arrest and maintenance of apoptosis

SUMO targeted therapy of various tumor entities

SUMO has been proposed as future target in cancer therapy [187]. As Ubc9 is overexpressed in several tumor entities like malignant melanoma, lung and ovarian carcinomas Ubc9 is currently tried to be experimentally targeted. Either the active sites or the binding to the E1 enzyme or the binding to specific proteins may be targeted. Arsenic trioxide which promotes the binding of SUMO-1 and -2 chains to PML-Rasα can induce degradation of PML-RARα [112], [212], [213]. This substance has already

Conclusion

The mechanisms by which SUMO influences carcinogenesis are poorly understood. However, SUMOylation has been repeatedly demonstrated as being critically involved in processes driving pathways towards tumor formation like cell growth, differentiation, senescence, apoptosis and autophagy [215], [216], [217]. Increasing evidence suggests that SUMO influences various pathways towards cancer formation. Among others self-sufficiency in growth signals, insensitivity to growth-inhibitory (anti-growth)

Disclosure

Herewith all authors state that regarding this study there are no conflicts of interest.

Acknowledgements

This work was supported by an Innovative Medicines Initiative Joint Undertaking (IMI) grant (OncoTrack) (J.H.) and a grant from the Steirische Kinder-Krebs-Hilfe (M.B.).

References (238)

  • M. Ihara et al.

    Noncovalent binding of small ubiquitin-related modifier (SUMO) protease to SUMO is necessary for enzymatic activities and cell growth

    J. Biol. Chem.

    (2007)
  • W. Yang et al.

    Gene expression and cell growth are modified by silencing SUMO2 and SUMO3 expression

    Biochem. Biophys. Res. Commun.

    (2009)
  • K.M. Bohren et al.

    A M55V polymorphism in a novel SUMO gene (SUMO-4) differentially activates heat shock transcription factors and is associated with susceptibility to type I diabetes mellitus

    J. Biol. Chem.

    (2004)
  • J.M. Desterro et al.

    Ubch9 conjugates SUMO but not ubiquitin

    FEBS Lett.

    (1997)
  • T. Hayashi et al.

    Ubc9 is essential for viability of higher eukaryotic cells

    Exp. Cell Res.

    (2002)
  • E.S. Johnson et al.

    Ubc9p is the conjugating enzyme for the ubiquitin-like protein Smt3p

    J. Biol. Chem.

    (1997)
  • M. Hochstrasser

    SP-RING for SUMO: new functions bloom for a ubiquitin-like protein

    Cell

    (2001)
  • L. Aravind et al.

    SAP – a putative DNA-binding motif involved in chromosomal organization

    Trends Biochem. Sci.

    (2000)
  • J.A. Tan et al.

    Protein inhibitors of activated STAT resemble scaffold attachment factors and function as interacting nuclear receptor coregulators

    J. Biol. Chem.

    (2002)
  • T. Kahyo et al.

    Involvement of PIAS1 in the sumoylation of tumor suppressor p53

    Mol. Cell.

    (2001)
  • K. Nakagawa et al.

    PIAS3 induces SUMO-1 modification and transcriptional repression of IRF-1

    FEBS Lett.

    (2002)
  • T. Nishida et al.

    PIAS1 and PIASxalpha function as SUMO-E3 ligases toward androgen receptor and repress androgen receptor-dependent transcription

    J. Biol. Chem.

    (2002)
  • Y. Miyauchi et al.

    Sumoylation of Mdm2 by protein inhibitor of activated STAT (PIAS) and RanBP2 enzymes

    J. Biol. Chem.

    (2002)
  • E.S. Johnson et al.

    An E3-like factor that promotes SUMO conjugation to the yeast septins

    Cell

    (2001)
  • Y. Takahashi et al.

    A novel factor required for the SUMO1/Smt3 conjugation of yeast septins

    Gene

    (2001)
  • A.A. Yunus et al.

    Structure of the Siz/PIAS SUMO E3 ligase Siz1 and determinants required for SUMO modification of PCNA

    Mol. Cell.

    (2009)
  • J.J. Perry et al.

    A SIM-ultaneous role for SUMO and ubiquitin

    Trends Biochem. Sci.

    (2008)
  • A. Pichler et al.

    The nucleoporin RanBP2 has SUMO1 E3 ligase activity

    Cell

    (2002)
  • M.M. Dawlaty et al.

    Resolution of sister centromeres requires RanBP2-mediated SUMOylation of topoisomerase IIalpha

    Cell

    (2008)
  • M.H. Kagey et al.

    The polycomb protein Pc2 is a SUMO E3

    Cell

    (2003)
  • B. Czermin et al.

    Drosophila enhancer of Zeste/ESC complexes have a histone H3 methyltransferase activity that marks chromosomal Polycomb sites

    Cell

    (2002)
  • E.T. Yeh

    SUMOylation and De-SUMOylation: wrestling with life’s processes

    J. Biol. Chem.

    (2009)
  • L. Gong et al.

    Characterization of a family of nucleolar SUMO-specific proteases with preference for SUMO-2 or SUMO-3

    J. Biol. Chem.

    (2006)
  • J. Hang et al.

    Association of the human SUMO-1 protease SENP2 with the nuclear pore

    J. Biol. Chem.

    (2002)
  • L. Gong et al.

    Differential regulation of sentrinized proteins by a novel sentrin-specific protease

    J. Biol. Chem.

    (2000)
  • D. Bailey et al.

    Characterization of the localization and proteolytic activity of the SUMO-specific protease, SENP1

    J. Biol. Chem.

    (2004)
  • Y.H. Kim et al.

    Desumoylation of homeodomain-interacting protein kinase 2 (HIPK2) through the cytoplasmic-nuclear shuttling of the SUMO-specific protease SENP1

    FEBS Lett.

    (2005)
  • F. Melchior et al.

    SUMO: ligases, isopeptidases and nuclear pores

    Trends Biochem. Sci.

    (2003)
  • J.H. Kim et al.

    Emerging roles of desumoylating enzymes

    Biochim. Biophys. Acta

    (2009)
  • A. Arnaoutov et al.

    The Ran GTPase regulates kinetochore function

    Dev. Cell.

    (2003)
  • B.B. Quimby et al.

    The small GTPase Ran: interpreting the signs

    Curr. Opin. Cell. Biol.

    (2003)
  • T.H. Shen et al.

    The mechanisms of PML-nuclear body formation

    Mol. Cell.

    (2006)
  • J.L. Best et al.

    SUMO-1 protease-1 regulates gene transcription through PML

    Mol. Cell.

    (2002)
  • Y. Han et al.

    SENP3-mediated de-conjugation of SUMO2/3 from promyelocytic leukemia is correlated with accelerated cell proliferation under mild oxidative stress

    J. Biol. Chem.

    (2010)
  • N. Ohbayashi et al.

    The IL-6 family of cytokines modulates STAT3 activation by desumoylation of PML through SENP1 induction

    Biochem. Biophys. Res. Commun.

    (2008)
  • N.J. McKenna et al.

    Combinatorial control of gene expression by nuclear receptors and coregulators

    Cell

    (2002)
  • T. Bawa-Khalfe et al.

    Induction of the SUMO-specific protease 1 transcription by the androgen receptor in prostate cancer cells

    J. Biol. Chem.

    (2007)
  • J. Cheng et al.

    Role of desumoylation in the development of prostate cancer

    Neoplasia

    (2006)
  • V.G. Wilson

    SUMO Regulation of Cellular Processes

    (2009)
  • R. Geiss-Friedlander et al.

    Concepts in sumoylation: a decade on

    Nat. Rev. Mol. Cell. Biol.

    (2007)
  • Cited by (130)

    • Epigenetic modifications from arsenic exposure: A comprehensive review

      2022, Science of the Total Environment
      Citation Excerpt :

      Since DNA damage responses are discontinued in the cells in which Mus 81 (SUMo resistant) is expressed, SUMoylation of Mus81 due to chronic arsenic exposure has been proposed as a mediating factor on chromosomal alignment (Hu et al., 2017). The SUMoylation cycles help rapidly proliferating cells in terms of increased expressions of SAE1/SAE2 (SUMo activating enzyme 1/ 2), SUMo isopeptides, and Ube9 (human counter part of S. cerevisiae gene UBC9) that are found to be associated with continuous cell malignancies, metastasis and tumorigenesis (Bettermann et al., 2012; Kessler et al., 2012). Kessler et al. (2012) reported that SAE1 and SAE2 are significantly correlated with Myc-dependent oncogenic transcription.

    View all citing articles on Scopus
    View full text