|Home | About | Journals | Submit | Contact Us | Français|
Heart failure is a complex disease process with various aetiologies and is a significant cause of morbidity and death world-wide. Post-translational modifications (PTMs) alter protein structure and provide functional diversity in terms of physiological functions of the heart. In addition, alterations in protein PTMs have been implicated in human disease pathogenesis. Small ubiquitin-like modifier mediated modification (SUMOylation) pathway was found to play essential roles in cardiac development and function. Abnormal SUMOylation has emerged as a new feature of heart failure pathology. In this review, we will highlight the importance of SUMOylation as a regulatory mechanism of SERCA2a function, and its therapeutic potential for the treatment of heart failure.
Congestive heart failure is a major cause of morbidity and mortality throughout the world [1, 2]. Over 23 million people world-wide and more than 5.8 million adults in the United State are living with heart disease. Heart failure accounts for over one million hospitalisations, nearly 300,000 deaths, and approximately 40 billion dollars in Medicare expenses annually with five-year survival being less than 50% in the United States despite advances in pharmacological treatment and device therapy [1, 3, 4]. Thus, there is an ultimate need for novel therapies to treat heart failure. Over the last decade, many studies have focussed on the elucidation of new molecular mechanisms associated with heart failure to identify novel drug targets. Positive inotropic agents, which stimulate and increase muscle contractility, have become the primary focus of new therapeutic approaches for the treatment of heart failure [5-7].
Post-translational modification with small ubiquitin-like modifier (SUMO) protein, SUMOylation, plays a significant role in the functional regulation of proteins by altering structure, enzymatic activity, stability or degradation, localisation, protein-protein interactions, and diverse signalling cascades . SUMOylation is responsible for the dynamic reaction to environmental stimuli in physiological and pathological states. Altered SUMOylation of specific proteins has been linked to diverse human diseases and disorders [9-11]. In addition, recent studies support an important role of SUMOylation in pathogenic mechanisms involved in heart failure such as oxidative stress and hypertrophic stimuli .
In this review, we summarise the role of SUMOylation in the heart. In particular, we focus on the role of SUMO1 in SERCA2a-mediated heart function and discuss the need for targeting SERCA2a SUMOylation as a new therapeutic approach to treat heart failure.
Ubiquitin (Ub) and ubiquitin-like proteins (Ubls) can covalently bind to substrates and subsequently regulate their function . SUMO belongs to a large family of ubiquitin-related proteins. There are four SUMO isoforms (SUMO1, SUMO2, SUMO3 and SUMO4) in mammalian cells. SUMO2 and SUMO3 are almost identical (only three N-terminal residues are different), whereas SUMO1 is approximately 48% identical at the amino acid level to SUMO2/3. SUMO4 shares 85% identity with SUMO2/3, however it is not clear whether SUMO4 can be processed and conjugated to substrates [14-16]. SUMO isoforms are functionally distinct as they modify different substrates [17-19] and they differ in their subcellular localisation patterns and dynamics [17, 20]. SUMO reversibly attaches to lysine residues via enzymatic cascade reaction called SUMOylation . Similar to ubiquitin, all SUMO proteins need to be processed by specific cysteine proteases, SENP (Sentrin-specific Protease), into their mature form. The mature SUMO is then adenylated by SUMO E1 activating enzyme (SAE1/SAE2) in an ATP dependent reaction and is subsequently transferred to a SUMO E2 conjugating enzyme (Ubc9) [22, 23]. In some cases, Ubc9 can directly recognise substrate proteins and catalyse the transfer of SUMO to the substrates [24-27]. SUMO E3 ligase functions as an adaptor protein that stimulates efficacy of SUMOylation [28-30] by facilitating the transfer of SUMO from Ubc9 to the substrate, a process particularly important for the substrates lacking SUMO recognition sites. The covalently linked SUMO proteins can be removed by SENPs, a process referred to as deSUMOylation . To date, six members of the SENP family have been identified in humans, which have distinct subcellular localisation and different preference for SUMO isoforms. Interestingly, some SENPs have dual functionality in that they both process SUMO to its mature form and also cleave the isopeptide bond between SUMO and its substrate protein . This functional diversity of SENPs suggests that SUMOylation is a dynamic process, which needs to be tightly regulated in the cell. Moreover, unlike ubiquitination, which targets proteins to proteasomal degradation, SUMOylation leads to changes in stability, activity and/or sub-cellular localisation of the modified proteins that modulates a diverse range of cellular processes [8, 32-34]. However, it is still unknown how substrate specificity of SUMOylation is achieved by the mechanisms constituted with a limited number of enzymes, such as a single SUMO E1 enzyme, a single SUMO E2 enzyme and few SUMO E3 ligases are known.
Discovered in 1996, SUMOylation has been found to be highly relevant in signal transduction, particularly in response to cellular stress and disease . Importantly, a number of studies suggested that SUMOylation also plays a role in human disease pathogenesis [11, 36-38]. Indeed, critical regulatory proteins for human disease states, including neurodegenerative diseases and cancer, are substrates of SUMOylation . There is also growing evidence that the SUMOylation pathway is involved in cardiac physiology and pathology. Studies employing transgenic and knockout approaches have suggested the roles of SUMOylation in the heart. Global SUMO1 knockout mice developed congenital heart disease, including atrial and ventricular septal defects , and cardiac SUMO1 knockdown mice showed progression of cardiac dysfunction and sudden death . However, SUMO2 knockout mice showed defects in embryonic development without any specific cardiac phenotype . A deSUMOylating enzyme, SENP2 knockout mice led to embryonic lethality with defects in the embryonic heart . In addition, cardiac specific overexpression of SENP2 or SENP5 in mice led to dilated cardiomyopathy demonstrating the negative effects of excessive de-SUMOylation on cardiac function [43, 44]. On the other hand, increased SUMO2/3 conjugation together with elevated expression levels of SENP1 and SENP5 was recently observed in human failing hearts caused by idiopathic cardiomyopathy [44, 45]. These studies suggest the importance of SUMOylation/deSUMOylation in normal cardiac development and function.
A more direct link has been proposed between defective SUMOylation and human heart disease based on a target-specific study. For instance, a mutation analysis of Lamin A, which is a critical scaffolding protein important for nuclear structure, revealed that its deficiency causes human familial dilated cardiomyopathy. Importantly, Lamin A is SUMOylated at lysine 201, however, Lamin A E203G and E203K mutants, located in the SUMOylation consensus sites, have been shown to have reduced levels of SUMOylation and are associated with familial dilated cardiomyopathy . Mutation in TRPM4 (E7K), which is a calcium activated non-selective cation channel, has been associated with familial atrial fibrillation as a result of defective SUMOylation , which can lead to congestive heart failure and stroke.
Furthermore, several transcriptional factors/co-factors (GATA4, Nkx2.5, Myocardin, SRF1), mitochondrial protein Drp1, cardiac ion channel Kv2.1 and cardiac metabolic proteins (PPAR, PGC1 alpha, AMPK) have been reported as SUMOylation substrates , however their roles in pathologic conditions need to be elucidated.
Calcium cycling is a central mechanism of cardiomyocyte function and the sarcoplasmic reticulum (SR) is the primary calcium storage organelle in muscle cells. In cardiomyocytes, removal of cytosolic calcium is mainly accomplished by the cardiac isoform of the SR calcium ATPase (SERCA2a). SERCA2a actively transports calcium from the cytosol to the SR during myocardial relaxation. Abnormal regulation of SR calcium cycling due to SERCA2a dysfunction is a common feature of human heart failure [49-52]. Indeed, SERCA2a expression and activity are significantly reduced in human heart failure [52, 53]. Animal models of heart failure and failing human ventricular myocytes isolated from patients with end-stage heart failure showed that SERCA2a gene transfer could restore calcium uptake into the SR and improve velocity of muscle contraction and relaxation [54-58]. Additional beneficial effects of SERCA2a gene transfer were also observed in diverse animal models such as improvements in cardiac function, restoration of cardiac energetics, reduction of ventricular arrhythmias, and enhancement of survival [59-64]. Phase 1 and 2a trials using adeno-associated vector type 1 encoding SERCA2a (AAV1.SERCA2a) in patients with end stage heart failure showed potential clinical benefits [65-68]. Although AAV1-based Phase 2b trial did not meet its endpoints, SERCA2a continues to be a promising therapeutic target. SERCA2a activity and its expression are regulated by interacting with cardiac proteins and by microRNAs. In addition, there are several post-translational modifications that also regulate the transport activity of the SERCA2a protein. Recent review articles overviewed studies of the SERCA2a regulatory mechanism and new therapeutic approaches to restore SERCA2a-mediated calcium cycling [69, 70].
Our group recently discovered that small ubiquitin-like modifier type 1 (SUMO1) conjugates to SERCA2a at lysine 480 and lysine 585 which are located within nucleotide domain . The SUMOylation sites are canonical consensus sequences and direct interaction between SUMO E2 enzyme, Ubc9 and SERCA2a was detected. Biochemical analysis showed that SUMOylation increased SERCA2a ATPase activity as well as its protein stability . The cardio protective effect of SUMO1 was further observed. In human failing hearts, levels of SUMO1 protein and SERCA2a SUMOylation were significantly decreased . Cardiotrophic adeno-associated vector (AAV9)-mediated short hairpin knockdown of SUMO1 decreased expression and activity of SERCA2a and accelerated animal mortality whereas gene delivery of SUMO1 rescued cardiac dysfunction in mice induced by pressure overload . In addition, SUMO1 protected the SERCA2a function against hydrogen peroxide-induced oxidative stress and hypertrophic stimuli in cardiomyocytes . These observations suggest that post-translational modifications of SERCA2a caused by the toxic environment of the hypertrophied and failing myocardium can be prevented by the expression of SUMO1. The crosstalk between SUMOylation and other post-translational modifications, such as acetylation, ubiquitination and phosphorylation, is considered to be a part of a coordinated mechanism for differential and context-related regulation [71, 72]. Our recent findings show that SERCA2a is closely linked to lysine acetylation/deacetylation pathways. We observed that SERCA2a SUMOylation is decreased in failing hearts while acetylation is increased in the setting of heart failure that can be decreased by targeting deacetylation enzyme, sirtulin 1 (unpublished data).
The potential of SUMO1 gene therapy for heart failure has been studied in several murine models. Adenoviral gene delivery of SUMO1 to cardiomyocytes isolated from both sham-operated and transverse aortic constriction (TAC)-induced failing mouse hearts resulted in increased cell shortening and a faster calcium re-uptake . A more prominent enhancement in contractility was observed when failing cardiomyocytes were overexpressed with SUMO1. The overall inotropic effect of SUMO1 overexpression was comparable with that induced by SERCA2a overexpression. Chronic benefits of SUMO1 delivery using AAV vectors were observed in TAC-induced mouse model of heart failure. AAV9-mediated cardiac overexpression of SUMO1 rescued cardiac dysfunction despite constant pressure-overload. Haemodynamic performance was significantly improved by SUMO1 gene delivery along with prevention of cardiac remodelling. In addition, the recovery induced by SUMO1 overexpression resulted in significant improvement in animal survival. Importantly, SUMO1 gene therapy restored SERCA2a-mediated calcium cycling. AAV9-mediated SUMO1 overexpression in TAC-induced mice down-regulated abnormally high protein levels of sodium-calcium exchanger, whereas levels of phospholamban and ryanodine receptor 2 did not change. The beneficial effects of SUMO1 overexpression no longer existed in the absence of SERCA2a protein, suggesting that SERCA2a is a relevant substrate for SUMO1-mediated cardiac protection. After TAC surgery, tamoxifen-induced cardiac-specific SUMO1 overexpression in transgenic mice significantly reversed failing phenotypes resulted in less left ventricular dilation, enhanced cardiac contractility and improved animal survival . The improvements were associated with increased expression and activity of SERCA2a. These findings are consistent with the result from SUMO1 based gene therapy. Recently, we reported the anti-oxidative role of SUMO1. We demonstrated that SUMO1 gene transfer ameliorates development of cardiac hypertrophy in both isolated cardiomyocytes and in adult hearts. Furthermore, we found that SUMO1 overexpression can influence redox signalling molecules such as mitogen activated protein kinases, nicotinamide-adenine dinucleotide phosphate oxidase 4, and manganese superoxide dismutase. In addition, we observed the inhibitory effects of SUMO1 pre-treatment on SERCA2a oxidation, suggesting the potential role of SUMO1 in the ageing heart. These studies were the first to demonstrate proof-of-concept and long-term therapeutic efficiency of SUMO1 gene therapy in a small animal model, supporting its therapeutic potential. The therapeutic potential of SUMO1 gene therapy is also being evaluated in other diseases such as muscular dystrophy and ageing in our group.
Although SUMO1 gene therapy is effective in a small animal model of heart failure, it is important to validate this approach in a pre-clinical setting. We used a Yorkshire pig chronic ischaemia/reperfusion model that is a well-established heart failure model to evaluate new therapeutics . Myocardial SUMO1 protein expression was significantly decreased in control (with heart failure) groups compared to sham-operated pigs. SUMO1 expression and amount of SUMOylated SERCA2a were restored by gene therapy. AAV1-mediated myocardial delivery of SUMO1 improved contractile function and relieved the deterioration of the left ventricle. SUMO1 treatment was able to increase SERCA2a expressions at both mRNA and protein levels. Beneficial effect of SUMO1 was greater and more consistent in a high dosage group than in a low dosage group, suggesting the concentration-dependent effects of the treatment. The effect of SUMO1 gene transfer on cardiac proteins was identical to that of the SERCA2a gene transfer indicating that the working mechanism of SUMO1 as a positive inotropic agent is through SERCA2a. There was no significant off-target toxicity in the animals treated with AAV1-SUMO1. Interestingly, a combined gene delivery of SUMO1 and SERCA2a suggested synergistic effects on a molecular level and on cardiac function. Increased expression of SERCA2a together with enhanced SUMOylation, potentially stabilises the ATPase transporter and its activity. This approach may be more effective in human patients because the fundamental cardiac pathophysiology is not always corrected even after SERCA2a gene transfer. This study provides critical aspects of clinical applications of the SUMO1 gene transfer, which includes feasibility and safety of the agent.
Given the important role of SUMOylation in SERCA2a-mediated cardiac function and the subsequent demonstration of therapeutic value in pre-clinical animal models, we considered alternative therapeutic approaches such as small molecule drugs. Through in vitro high-throughput screening assays and series of the step evaluating SERCA2a SUMOylation and effects on cardiac contractility and calcium transient, we identified a small molecule N106, which could enhance SUMOylation of SERCA2a . In this study, we found that the N106 compound directly enhanced activity of a SUMO E1 enzyme, thereby triggering intrinsic SUMOylation. N106 treatment significantly increased cardiac cell contractility and calcium cycling in a dose-dependent manner. Acute treatment of N106 significantly improved left ventricular systolic function in TAC-induced heart failure mice. Molecular modelling analysis identified potential binding pocket of N106 on SUMO E1 that exists only in an active state of the enzyme, indicating that N106 is likely to be able to stabilise the active state of SUMO E1 enzyme. In addition, N106 had no significant effect on the in vitro ubiquitin E1 activity at concentrations up to 100 μM suggesting the specificity of N106 for the SUMO E1 enzyme. Several lines of evidence indicated that these actions of N106 are associated with SUMOylation of SERCA2a. For example, the beneficial effects of N106 did not exist in cardiac specific SERCA2 knockout mice. Neither protein levels nor phosphorylation states of phospholamban were altered by N106 treatment, suggesting the specificity of N106 compound to SERCA2a and its SUMOylation. N106 had no effect on tumourigenesis as determined by NCI 60 human tumour cell line screen. In addition, N106 exhibited a pharmacokinetic profile suitable for in vivo evaluation in mice. In our pilot study, haemodynamic improvement was observed in TAC mice orally treated with N106. Moreover, the beneficial effects of N106 translated into the increased survival of mice under prolonged pressure overload (unpublished data). Further long-term safety and efficacy of N106 activator in small and large animal models needs to be further determined. This study proposes a new pharmacological potential of SUMO activators.
Post-translational modifications of critical proteins have offered new information regarding the regulation of cardiac diseases. In the failing heart, SUMOylation of SERCA2a has been found to be reduced along with low SUMO1 expression, both of which contribute to cardiac dysfunction as the disease progresses. Restoration of SERCA2a activity through modulating its SUMOylation status, such as overexpression of SUMO1 via gene transfer or small molecule activation of a SUMO E1 enzyme, may be a new therapeutic strategy for the treatment of heart failure (Figure 1). However, several important questions still remain. For instance, what is the mechanism of regulating expression levels of SUMO1 in the heart? Can we define in vivo kinetics and extent of SERCA2a SUMOylation? In addition, what is the optimal dose and time window for global activation of SUMOylation to improve cardiac function? These studies will lead to a more in-depth understanding of the role of SUMOylation-SERCA2a axis in the progression of heart disease and to a novel platform for treatment of heart failure.
This work was supported by NIH R00 HL116645, R01 HL117505, HL093183, P50 HL112324 and a NHLBI Program of Excellence in Nanotechnology (PEN) Award, contract number HHSN268201000045C.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
CONFLICT OF INTEREST