The first isoform of SUMO proteins was identified as sentrin in 1996 (18). Sentrin is demonstrated to form a covalent conjugate with proteins by attaching SUMO to specific lysine residue and regulate protein functions, suggesting a novel PTM mechanism (19, 20). Thus, SUMOylation is proposed to describe the modification process of various proteins via reversibly attaching to the lysine residues through a multi-step enzymatic reaction cascade (Figure 1).

To date, five SUMO isoforms (SUMO1, 2, 3, 4, and 5) have been identified (21). SUMO1 and 2/3 are widely distributed in the body, whereas SUMO4 and SUMO5 appear to be tissue-specific and less investigated though. The SUMO4 was only found in the kidney, spleen, and lymph nodes (9). What we know so far: SUMO4 is discovered as an intron-free pseudogene in the study of single nucleotide polymorphism in type 1 diabetes (22). As a recently discovered family member, SUMO5 was identified in primates and showed high tissue specificity and might be involved in the regulation of promyelocytic leukemia nuclear bodies (21, 23). Based on the primary structure, SUMO1 shows 48% sequence identity with both SUMO2 and SUMO3, while the latter two isoforms are highly similar and with 97% of identity (24). SUMO4 shares 85% identity with SUMO2/3 although no evidence supports the conjugation with substrates (4, 22). The SUMO5 is highly homologous to SUMO1 (25). On the other hand, SUMOylation occurs at selective lysine residue within the canonical/consensus motif of Φ-K-X-E/D, known as SUMO-interacting motif (SIM), where Φ is a hydrophobic residue, K is lysine for SUMO to form conjugate, X is any amino acid residue, and E/D is an acidic residue (26).

Nevertheless, SUMO isoforms have overlapping functions and may compensate each other under certain circumstances.

SUMOylation and deSUMOylation are dynamically balanced under physiological conditions. As shown in Figure 1, SUMOylation describes the covalent modification of protein substrates by SUMOs while deSUMOylation removes SUMOs from SUMOylated proteins. The SUMO proteases in the SUMO pathway are listed in Table 1. For forming covalent conjugates with protein substrates, SUMOs undergo maturation via proteolytic cleavage of the C-terminal Glycin -Glycin (~GG) dipeptide by SENP (28). SUMOs are attached to protein substrates through three reaction steps: ATP-dependent activation, SUMO- E2(UBC9) conjugation, and SUMO-protein substrate ligation. Firstly, the mature SUMO is activated by the E1 activating enzyme (SAE1/2) in an ATP-dependent manner. Secondly, the E2 conjugating enzyme UBC9 takes over SUMO moiety via forming covalent conjugate. Finally, the E3 ligase brings protein substrate and E2/UBC9 SUMO complex into the closer proximity and yields the covalent conjugate of the SUMOylated proteins including a protein inhibitor of activated STAT (PIAS) family, RanBR2, polycomb2, mitochondrial-anchored protein ligase (MAPL), and much other protein substrates (20, 29).

The deSUMOylation describes the removal of SUMO moiety from target proteins by the SENP and deSUMOylating isopeptidases (19, 30). The SENP family includes six members, namely SENP 1-3, 5-7 in humans. Different SENP members act on specific SUMOylated proteins (31). SENP1 and SENP2 catalyze the deSUMOylation of all types of SUMO proteins (32), whereas SENP3, 5, 6, and 7 prefer SUMO2/3-derived SUMOylated proteins (33, 34). Ultimately, deSUMOylation eliminates the effects of SUMOylation on protein functions (35).

SUMOylation plays important role in the various molecular events and processes (36). First, SUMOylation may promote or block the association of molecules that interact with SUMOylated substrates. Vivek et al. showed that SUMOylation mediated WRKY33 phosphorylation while disruption of WRKY33 SUMO sites inactivated WRKY33-mediated defense (37). Waizenegger et al. found that Mms4 was engaged by (poly)SUMOylation and targeted for proteasome degradation (38). Moreover, SUMOylation preserves substrate stability by competing with ubiquitination for the lysine residue or by recruiting the SUMO-targeted ubiquitin ligase (STUBL) family of proteins to the SUMOylated substrates. SUMOylation of hyperphosphorylated tau at K340 inhibits its ubiquitylation and the subsequent proteasome-dependent degradation (39). Gao et al. showed that PKCδ was SUMOylated at lysine 473, and the SUMOylation increased PKCδ stability by repressing its ubiquitination (40). Besides, SUMOylation induces conformational changes in proteins and thereby regulates protein functions (41–43). SUMOylation of PKM2 lysine 270 (K270) triggered conformation change from tetrameric to dimeric of PKM2, and reduced PKM activity (44). Furthermore, SUMOylation provides a new mechanism to recruit proteins with SUMO-interacting motif. Blondel et al. identified a non-covalent interaction between SUMO and β-arr2, via a SUMO interaction motif (SIM), that is, required for β-arr2 cytonuclear trafficking (45). Thus, SUMOylation and de-SUMOylation coordinate the regulation of various cellular signaling pathways (42).

SUMOylation is an important PTM in the regulation of cell division, DNA repair, genetic and epigenetic transcriptional regulation, and cellular metabolism (4, 14, 20). Consequently, dysregulation of SUMOylation is closely related to the pathogenesis of various diseases, such as cancer, diabetes, epilepsy, and cardiovascular diseases (46, 47). Many recent studies suggest that SUMOylation is a therapeutic target in the cardiovascular system. Importantly, SUMOylation not only affects cardiac function and development (11, 48–50) but also controls the capacity of the heart to adapt to various pathological stresses (51) (Figure 2). Specific SUMO isoforms (e.g., SUMO1, 2, 3) are required and play a different role in the heart. SUMO-1 KO mice suffered from congenital heart diseases such as atrial and septal defects, progressive cardiac dysfunction and sudden death (52). Specifically, the over-expression of SUMO1 enhances cardiac function in mice with heart failure and increases contractility in isolated cardiomyocytes (10). SUMO-2 KO mice exhibited severe developmental delay, without any specific cardiac phenotype, and died at embryonic stage E10.5 whereas SUMO-3 KO mice were viable (53). Another study showed that the dynamics of the SUMOylation/deSUMOylation was altered during MI/R injury correlating with the decrease in SENPs activities, especially that of SENP3 (54). Presumably, the SUMOylation system may control the response of the heart to hypoxic/ischemic stress (55). Nevertheless, little is known about the regulation of the SUMOylation system and the potential target proteins during MI.

Myocardial ischemia is caused by a decrease in coronary blood flow or failure to meet the demand of cardiac tissue for oxygen. Ischemia indicates that the environment oxygen is not enough to maintain the oxidative phosphorylation in the mitochondria (60). During ischemia, anaerobic glycolysis of glucose produces excessive lactic acid in the ischemic cells and ultimately causes acidification in the heart (61). When oxidative phosphorylation is not supported by sufficient oxygen, ATP production is largely reduced, leading to the dysfunction of Na⁺-K⁺-ATPase and the elevation of intracellular calcium, sodium, and hydrogen concentration (62, 63). Subsequently, cells swell while the activities of cytoplasmic enzymes are impaired. Prolonged ischemia causes progressive and irreversible injury in the heart. Morphologically, such irreversible injury is characterized by glycogen depletion, margination of nuclear chromatin, mitochondrial swelling, and sarcolemma breaks (64).

Timely and complete reperfusion is critical to limit infarct size and subsequent ventricular remodeling (65). Reperfusion also causes irreversible injury to the myocardium and the coronary circulation, contributing to final infarct size (66, 67). During reperfusion, the mitochondrial permeability transition pore opening is triggered by multiple factors including mitochondrial calcium. Subsequently, oxidative stress and calcium overload increase the release of cytochrome C and thereby induce myocardial injury (68). Reperfusion not only salvages ischemic myocardium from infarction but also induces irreversible injury, leading to increased infarct size and microvascular dysfunction (69, 70).

Myocardial ischemia-reperfusion (MI/R) injury is a complex pathological process involving several signaling pathways (71, 72). As a key type of PTM, SUMOylation and deSUMOylation appear to affect different mechanisms toward cardiac damage. Several studies suggest that SUMOylation determines the fate of perfused hearts (73, 74). Ischemia increased SUMOylation levels, especially involving SUMO-2/3, while reperfusion further increased SUMOylation of various proteins in animal models and cell culture systems (75). For example, SUMO-2/3 conjugation was increased in failing human hearts (11). SUMOylation levels by SUMO-1 and SUMO-2/3 showed apparent differences in the mouse and rat models of ischemia-reperfusion injury (76, 77). In rats, a dramatic increase in SUMOylation by both SUMO-1 and SUMO-2/3 was observed at 6 h and 24 h in the striatal infarct area and hippocampus. In mice, no changes in SUMOylation occurred at 6 h but there was increased SUMO-1 conjugation in the cortical infarct after ischemia-reperfusion injury. Collectively, attention should be directed to the increase of SUMO2/3 conjugation in MI. Further work should be designed to clarify whether SUMOylation occurs mainly in the process of reperfusion or the stage of initial ischemia (41, 75, 78). Although the role of SUMOylation in MI/R injury remains elusive (54, 79, 80), SUMOylation may be a promising target for drug discovery against MI (30).

Sarcoplasmic reticulum (SR) Ca²⁺ ATPase pump (SERCA2a) is an important ATP hydrolase and highly expressed in the heart for the control of Ca²⁺ reuptake and replenishment to SR in excitation-contraction coupling (30). Phospholamban phosphorylation increases the ATPase activity of SERCA2a and enhanced Ca²⁺ transport into the endoplasmic reticulum for the next round of the contraction cycle (127). Dysregulation of SERCA2a activity and Ca²⁺ cycling hallmark the pathology of heart failure and may drive the development of other cardiac dysfunctions (128). A recent study found that SUMOylation of SERCA2a increased the expression and activity of SERCA2a partly through SUMO1, thus improving ΔΨm and reducing apoptotic cells in vitro and promoting the recovery of heart function and reducing infarct size in vivo (129). Mouse SERCA2a could be SUMOylated at the sites of lysine 585, 480 and 571, whereas SUMOylation at lysine 585 enhanced SERCA2a stability (129, 130). The protein levels of SERCA2a, SUMO1, and the SUMOylated SERCA2a (S-SERCA2a) were decreased in myocardial injury (14). Interestingly, SUMO1 knockdown exacerbated while SUMO1 overexpression reversed the decline of SERCA2a function and SUMOylated SERCA2a (10, 129). Moreover, SUMO1 overexpression decreases cardiomyocyte apoptosis, reduces infarct size, and increases cardiac function (48). Collectively, SUMO1 mediated-SUMOylation of SERCA2a appears to be an important cardioprotective mechanism against MI/R injury (129).

Mitochondrial dysfunction is implicated in MI/R injury (131, 132). Dynamin-related protein (Drp) 1 is an important mitochondrial protein in the regulation of mitochondrial morphology and fission (15). It was reported that SUMOylated Drp1 was the key to mediate zinc-induced cardioprotection against I/R injury, possibly via activating the mitophagy of the mitochondria and suppressing ROS production (133). SUMOylation might inhibit Drp1 translocation from the cytosol to mitochondria, maintained the mitochondrial morphology, restrained mitochondrial fission, and protected the heart against MI/R injury (134). Interestingly, SUMOylation of Drp1 prevents MI/R injury, suggesting a protective mechanism against stress (80). SUMOylation of Drp1 increased mitophagy during reperfusion, leading to the prevention of ROS, myocardial apoptosis, and myocardial injury (135). With the increase in SUMO1-mediated SUMOylation of Drp1, the mitochondrial translocation of Drp1 is prevented (15). MI/R appears to increase the expression of Drp1 in mitochondria and the SUMOylations of Drp1 by both SUMO1 and SUMO2/3 (74).

On the other hand, the biological functions of SUMOylated proteins are reversed by deSUMOylation. For deSUMOylation, SENP2 removes SUMO1 from Drp1 whereas SENP3 and SENP5 remove SUMO2/3 from Drp1 (74, 136). SENP3 and SENP5 preferentially remove SUMO2/3 from Drp1 and inhibits SUMO2/3-induced mitochondrial translocation of Drp1(137). Pharmacological inhibition of Drp1 by mitochondrial division inhibitor-1(Mdivi-1) significantly attenuated the effects of SENP3 on mitochondrial membrane potential, mitochondrial swelling and cardiac injury (54). Others found the decrease in SUMO2/3-mediated SUMOylation of Drp1 and the reduction of mitochondrial fission level, ROS generation, and cell death in SENP5 transgenic mice (49). Collectively, Mdivi-1 may protect the myocardium by inhibiting the SENP3 mediated-deSUMOylation pathway. These studies suggest that the balance of SUMOylation and deSUMOylation may be a potential therapeutic target for the treatment of MI.