Cardioprotective effects of Cu(II)ATSM in human vascular smooth muscle cells and cardiomyocytes mediated by Nrf2 and DJ-1

Cu(II)ATSM was developed as a hypoxia sensitive positron emission tomography agent. Recent reports have highlighted the neuroprotective properties of Cu(II)ATSM, yet there are no reports that it confers cardioprotection. We demonstrate that Cu(II)ATSM activates the redox-sensitive transcription factor Nrf2 in human coronary artery smooth muscle cells (HCASMC) and cardiac myocytes (HCM), leading to upregulation of antioxidant defense enzymes. Oral delivery of Cu(II)ATSM in mice induced expression of the Nrf2-regulated enzymes in the heart and aorta. In HCASMC, Cu(II)ATSM increased expression of the Nrf2 stabilizer DJ-1, and knockdown of Nrf2 or DJ-1 attenuated Cu(II)ATSM-mediated heme oxygenase-1 and NADPH quinone oxidoreductase-1 induction. Pre-treatment of HCASMC with Cu(II)ATSM protected against the pro-oxidant effects of angiotensin II (Ang II) by attenuating superoxide generation, apoptosis, proliferation and increases in intracellular calcium. Notably, Cu(II)ATSM-mediated protection against Ang II-induced HCASMC apoptosis was diminished by Nrf2 knockdown. Acute treatment with Cu(II)ATSM enhanced the association of DJ-1 with superoxide dismutase-1 (SOD1), paralleled by significant increases in intracellular Cu(II) levels and SOD1 activity. We describe a novel mechanism by which Cu(II)ATSM induces Nrf2-regulated antioxidant enzymes and protects against Ang II-mediated HCASMC dysfunction via activation of the Nrf2/DJ-1 axis. Cu(II)ATSM may provide a therapeutic strategy for cardioprotection via upregulation of antioxidant defenses.


Cu (II) ATSM induces antioxidant protein expression in human cardiomyocytes and in vivo.
). Cu (II) ATSM was delivered by oral gavage at a dose of 30 mg/kg, which has previously been reported to confer protection against oxidative stress in vivo [13][14][15] . A significant increase in HO-1, Prx1, GCLM and NQO1 protein expression was observed in heart and aortic tissue at 24 h after oral administration of Cu (II) ATSM. These findings provide the first evidence that Cu (II) ATSM enhances Nrf2-regulated antioxidant protein expression in HCASMC and HCM in vitro and in the murine heart and aorta in vivo.
Since Ang II is known to increase superoxide generation through mitochondrial activity 26 , we examined the effect of acute Cu (II) ATSM treatment on Ang II-induced mitochondrial superoxide generation using MitoSOX red fluorescence in HCASMC ( Fig. 5C and Fig. S5). Ang II treatment (200 nM, 4 h) significantly increased mitochondrial superoxide generation, which was significantly reduced following acute Cu (II) ATSM (1 µM, 30 min) treatment. Treatment with Cu (II) ATSM alone did not alter basal levels of mitochondrial superoxide generation. Cu (II) ATSM protects against angiotensin II-induced apoptosis via Nrf2 and DJ-1. As Ang II has been reported to elicit apoptosis in VSMC 27 , we used annexin V binding as an index of apoptosis to examine whether Nrf2 or DJ-1 mediates protection afforded by Cu (II) ATSM against Ang II in HCASMC ( Fig. 5D and Fig. S6). Pre-treatment with Cu (II) ATSM (1 µM, 12 h) prior to Ang II (200 nM, 12 h) significantly reduced levels of Ang II-induced apoptosis. Cu (II) ATSM treatment alone did not enhance apoptosis (Fig. 5D), further demonstrating that cell viability was unaltered. Cu (II) ATSM-mediated protection against Ang II-induced apoptosis was attenuated following Nrf2 or DJ-1 knockdown, establishing that both DJ-1 and Nrf2 are required for Cu (II) ATSM-mediated protection against Ang II-induced HCASMC apoptosis. Cu (II) ATSM reduces angiotensin II-mediated cell proliferation. As Ang II enhances proliferation of VSMC 28 , we assessed the effect of Cu (II) ATSM pre-treatment on Ang II-mediated HCASMC proliferation (Fig. 5E). Cu (II) ATSM treatment (1 µM, 72 h) did not affect HCASMC number compared to control, providing further evidence that Cu (II) ATSM alone did not alter cell viability. Ang II treatment (200 nM, 72 h) increased proliferation 2.5 fold, which was significantly attenuated in cells pre-treated with Cu (II) ATSM (1 µM, 12 h). Cu (II) ATSM increases protein association of DJ-1 with SOD1 and intracellular Cu (II) levels.  has been demonstrated to act as a Cu (II) chaperone, which has been directly associated with an increase in its association with SOD1 and its enzyme activity 11 . We therefore hypothesized that acute treatment with Cu (II) ATSM increases the association of DJ-1 with SOD1 in HCASMC. Immunoprecipitation experiments confirmed that treatment of HCASMC with Cu (II) ATSM (1 µM, 30 min) significantly increased the association of DJ-1 with SOD1 (Fig. 6A). Our data clearly demonstrate that DJ-1 is not only involved in the induction of Nrf2-regulated antioxidant enzymes, but can also enhance SOD1 association with DJ-1 following acute Cu (II) ATSM treatment. This suggests that Cu (II) binding by DJ-1 may mediate both SOD1 and Nrf2 activation. Furthermore, the increased association between DJ-1 and SOD1 suggests that DJ-1 was enriched with Cu (II) through an increase in intracellular Cu (II) levels 11 . We determined whether Cu (II) ATSM increases intracellular Cu (II) using both inductively coupled plasma-mass spectrometry (ICP-MS, Fig. 6B) and Phen Green SK (PGSK) fluorescence (Fig. 6C). A significant increase in intracellular Cu (II) was observed following acute Cu (II) ATSM (1 µM, 30 min) treatment, suggesting that augmented Cu (II) levels may mediate the effects of Cu (II) ATSM to increase SOD1 activity through DJ-1 association and antioxidant enzyme expression via DJ-1/Nrf2 signaling. We further report a significant increase in ERK1/2 phosphorylation (Fig. 6D), which has been implicated in the dissociation of Cu (II) from ATSM 29 , suggesting bioavailable Cu (II) is increased in HCASMC treated acutely with Cu (II) ATSM (1 µM, 15 min). In addition to the increased association between SOD1 and DJ-1, we also observed a 2-fold increase in SOD1 activity (Fig. 6E), providing further evidence that acute Cu (II) ATSM activates SOD1 activity, thereby acutely reducing superoxide generation. However, the acute protection afforded by Cu (II) ATSM does not affect the cytoprotection observed following Cu (II) ATSM pre-treatment, as protection against Ang II-induced apoptosis remains unaltered after SOD1 knockdown (Fig. 6F), suggesting that Cu (II) ATSM provides protection via two independent pathways.

Discussion
Current therapeutic strategies have had limited success in augmenting endogenous antioxidant defenses to counteract oxidative stress in cardiovascular diseases. Although recent findings have established that Cu (II) ATSM affords protection against oxidative stress in the brain 13,14 , the underlying molecular mechanisms remain to be elucidated. Our study provides novel evidence that both oral delivery of Cu (II) ATSM in mice, and in vitro Cu (II) ATSM treatment of HCASMC and HCM, significantly upregulates Nrf2 dependent antioxidant defenses which is likely to confer protection against cardiovascular diseases associated oxidative stress 7 .
Classically, modification of Keap1 cysteine residues by oxidative or electrophilic stress inhibits proteasomal degradation of Nrf2 30 . The electrophilic nature of dietary compounds such as sulforaphane and curcumin makes them suitable Nrf2 activators 31 , however, it remains to be determined whether Cu (II) ATSM, a neutral and lipophilic compound 15 (Fig. S1A) is able to activate Nrf2 via interactions with Keap1. Although Cu (II) can mediate Nrf2 activation via a redox-cycling mechanism 32 , the levels of free Cu (II) in our study are likely to be lower compared to previous reports using compounds that can release significantly higher levels of Cu (II) under normal cell culture conditions compared to levels achieved by Cu (II) ATSM 33 . The intracellular dissociation of Cu (II) ATSM has been shown to increase the phosphorylation of ERK1/2 within a hypoxic environment 29 . We demonstrate that Cu (II) ATSM treatment enriches Cu (II) , in HCASMC and enhances ERK1/2 phosphorylation, suggesting an increase in bioavailable Cu (II)29 . Moreover, ERK1/2 mediates Nrf2 phosphorylation at serine 40 and its activation 21 , providing an additional mechanism through which Cu (II) ATSM may enhance Nrf2 signaling.
Although the Parkinson's associated protein DJ-1 is required for Nrf2 stability 10,34,35 , the presence of conserved cysteine residues on DJ-1 suggests a role as a redox sensor [36][37][38][39] , which may additionally modulate Nrf2 activity 24 . Copper chaperone functionality of DJ-1 may further serve as a mechanism to activate Nrf2 following Cu (II) ATSM delivery 10 . It is possible that DJ-1 enriched by copper enhances Nrf2 activation, as the induction of antioxidant enzymes was only evident upon treating cells with the Cu (II) ATSM complex and not with the ATSM ligand alone. Notably DJ-1 has been shown to directly regulate SOD1 activity 11 . Cu (II) ATSM delivery in vivo has been reported to increase mutated SOD1 activity in the brain 13 , but to date, only experiments without the use of cells have established that copper enriched DJ-1 directly increases SOD1 activity 11,40,41 . In this study, we have identified increased DJ-1 and SOD1 protein interactions in HCASMC treated with Cu (II) ATSM, providing a possible mechanism by which SOD1 activity may be increased acutely. Furthermore, we also demonstrate that acute Cu (II) ATSM-mediated protection via SOD1 occurs in addition to the activation of Nrf2 and target antioxidant defense proteins conferring long term protection.
Hearts of DJ-1 deficient mice have been shown to exhibit increased cardiomyocyte apoptosis, excessive DNA oxidation and cardiac hypertrophy when subjected to trans-aortic banding, as well as increased oxidative stress in response to Ang II infusion 19 , suggesting an important role for DJ-1 in cardioprotection. Notably, renal depletion of DJ-1 in mice decreases Nrf2 expression and activity, leading to increased oxidative stress and elevated systolic blood pressure 42 . Knockdown of renal Nrf2 in mice increases systolic blood pressure without effecting DJ-1 expression, suggesting that Nrf2 activation is downstream of DJ-1 and is thus required for the maintenance of redox balance. Our data corroborate these findings, as Cu (II) ATSM was unable to induce HO-1 and NQO1 expression in HCASMC following DJ-1 knockdown. Furthermore, our observation that Cu (II) ATSM increases DJ-1 expression suggests that this multifunctional protein is involved in the therapeutic protection by Cu (II) ATSM against cardiovascular oxidative stress, in part through its ability to stabilise Nrf2 10 and enhancing the expression of endogenous antioxidant enzymes.
Recent studies have shown that Cu (II) containing compounds have a therapeutic potential in inflammation, cancer, cardiac hypertrophy, PD and other neurodegenerative disorders 14,[43][44][45] , suggesting that Cu (II) ATSM may additionally exhibit cardioprotective properties. Studies where Cu (II) ATSM has been orally administered in rodent models of ALS and PD have reported improved neurological outcomes and increased survival through the reduction of oxidative stress [13][14][15]44 . Although it has been reported that acute Cu (II) ATSM treatment reduces lipid peroxidation in an isolated perfused rat heart model of ischemia-reperfusion 46 , the underlying mechanisms were not determined. Therefore, our study provides novel mechanistic insights for the actions of Cu (II) ATSM to mediate cardiovascular protection via activation of Nrf2/DJ-1 signaling and induction of Nrf2-regulated antioxidant defenses.
Ang II contributes to the development and progression of hypertension and cardiovascular pathologies via increases in superoxide generation, intracellular [Ca 2+ ] and cell proliferation 1, 2, 20 . Our findings in HCASMC strongly suggest that the observed protection against the pro-oxidant effects of Ang II on enhanced intracellular [Ca 2+ ] and proliferation are conferred through the activation of Nrf2/DJ-1 signaling. The attenuation of Ang II-induced increases in [Ca 2+ ] i , following pre-treatment of HCASMC with Cu (II) ATSM, is likely to decrease smooth muscle contractility associated with Ang II-mediated oxidative stress 1, 2 . Notably, DJ-1 deficient mice exhibit altered Ca 2+ homeostasis in skeletal muscle 47 , suggesting an additional role for DJ-1 in the redox regulation of [Ca 2+ ] i in HCASMC.
Smooth muscle apoptosis has been implicated in a number of processes contributing to cardiovascular diseases, including atherosclerotic plaque instability and rupture leading to myocardial infarction or cerebral stroke [48][49][50] . Ang II induces SMC apoptosis via activation of the Ang II type 2 receptor 27 , leading to enhanced caspase 3 activity, increased DNA fragmentation and oxidative stress 27,49,51 . We demonstrate that Cu (II) ATSM pre-treatment significantly attenuates Ang II-induced apoptosis in HCASMC, which was abolished following Nrf2 knockdown, suggesting that Nrf2-mediated upregulation of antioxidant enzymes may account for the protection afforded by Cu (II) ATSM. As DJ-1 knockdown also attenuated the protection afforded by Cu (II) ATSM against Ang II-induced apoptosis, it is likely that Nrf2-mediated antioxidant gene induction is also dependent on DJ-1 expression.
Oral delivery of Cu (II) ATSM in a mouse model of ALS markedly reduces levels of oxidatively modified protein carbonyls 15 . Cu (II) ATSM treatment in a mouse model of PD has also been linked to a significant reduction in oxidative stress, and thereby preventing aggregation of α-synuclein 14 . The similarities in the involvement of oxidative stress in both neurological and cardiovascular diseases highlights the therapeutic potential of Cu (II) ATSM in these pathologies. Although the protective properties of Cu (II) ATSM have been reported in rodent models of neurodegeneration, we now provide the first evidence that Cu (II) ATSM enhances cardiac and aortic expression of antioxidant proteins in vivo and provides protection against Ang II-mediated oxidative stress in HCASMC via Nrf2-regulated antioxidant defenses (Fig. 6).
Our study further confirm the potential therapeutic properties of Cu-containing compounds 52 and is the first to demonstrate that Cu (II) ATSM induces antioxidant enzymes in vivo and in HCASMC and HCM in vitro via Nrf2/DJ-1 axis to protect against Ang II-mediated oxidative stress. Therefore, Cu (II) ATSM represents a novel Nrf2 and DJ-1 activator with therapeutic potential to enhance endogenous antioxidant defenses, providing protection against cardiovascular diseases through ameliorating oxidative stress.

Material and Methods
Treatment of animals. Male C57BL6 mice (6-8 weeks, Charles River, UK) were acclimatized for at least 1 week before treatment and maintained on a 12 h light/dark cycle. All procedures were approved by the UK Home Office and King's College London after a rigorous ethical review process and performed under the authority of Project Licence No. PPL70/6579, in accordance with the UK Animal (Scientific Procedures) Act 1986. A suspension of the compound was prepared in standard suspension vehicle [SSV; 0.9% (w/v) NaCl, 0.5% (w/v) Na-carboxymethylcellulose (medium viscosity), 0.5% (v/v) benzyl alcohol, and 0.4% (v/v) Tween-80]. Cu (II) ATSM in SSV was delivered by oral gavage at a dose of 30 mg/kg body weight and the heart and aorta were harvested after 24 h. Control mice received an equivalent volume of SSV only.

Culture of primary human coronary artery smooth muscle cells (HCAMSC) and cardiomyocytes (HCM).
Primary HCASMC from 4 male donors, and HCM from 2 male donors were obtained from PromoCell (Germany) or Lonza (USA). Cells were cultured in phenol red free basal medium (PromoCell, Germany) supplemented with fetal calf serum (5%), epidermal growth factor (0.5 ng/mL), basic fibroblast growth factor (2 ng/mL) and insulin (5 µg/mL). Confluent cultures at passage 4-8 were equilibrated in phenol red free basal medium supplemented only with 5% FCS (Sigma, UK), without growth factors for 24 h prior to treatments with Cu (II) ATSM (0.1 µM-10 µM), synthesised as previously described 53 . Replicate experiments were performed on cells from different donors where possible.
Superoxide dismutase activity and superoxide generation. SOD1 activity was assessed using a commercially available SOD activity assay kit (Cayman Chemicals, USA). Total cellular superoxide production was assessed by L-012 enhanced chemiluminescence in live HCASMC cultures, as previously described 58 . Cells were incubated at 37 °C in Krebs buffer containing L-012 (20 µM). Luminescence was monitored over 10 min after 30 min equilibration at 37 °C in a luminescence microplate reader (Chameleon V, Hidex, Finland).
Detection of mitochondrial superoxide generation. Mitochondrial superoxide production was measured using MitoSOX Red (Life Technologies, USA) as previously described 59 . Cells were loaded with MitoSOX Red (5 µM, 30 min) at 37 °C, fixed with 4% paraformaldehyde and visualised by fluorescence microscopy. Equivalent numbers of cells were imaged in each field. Fluorescence intensity per cell was corrected for background intensity and quantified using image analysis software (Image J, NIH, USA) 55, 59 .

Assessment of apoptosis.
Annexin V binding to phosphatidylserine can be used as a marker of early apoptotic events 60 . Binding of Cy5-conjugated annexin V to HCASMC was assessed using a kit (Biotium, USA). Cells were co-stained with Hoechst 33342 (Sigma, UK) to identify nuclei and visualised using a fluorescence microscope (Nikon, Japan) and images acquired using a cooled CCD camera (Hamamatsu, Japan). Equivalent numbers of cells were captured for each field. Fluorescence intensity was determined using analysis software (Image J, NIH, USA).

Statistical analysis.
Data denote mean ± S.E.M. of experiments. All experiments were performed in n = 4-8 different cultures of HCASMC (from 4 donors) or HCM (from 2 donors). Comparison of more than two conditions in the same experiment were evaluated using either a Student's t-test or one way or two-way ANOVA followed by Bonferroni post hoc test. P < 0.05 values were considered significant.