Diabetic cardiomyopathy is associated with defective myocellular copper regulation and both defects are rectified by divalent copper chelation

Protocols for the induction of diabetes in rats and TETA treatment were as previously described [8, 49]. Diabetes was induced by a single intravenous injection of STZ (55 mg/kg body-weight; Sigma) into the tail-veins of adult male Wistar rats (6–7 weeks of age; 220–250 g); control animals received a saline injection instead of STZ. TETA was administered via the drinking water (20 mg/day per rat, Fluka), beginning at 8 weeks after saline or STZ injection. LV tissues from each treatment group (non-diabetic control, diabetic, TETA treated-control and TETA treated-diabetic) were collected after 8-weeks’ treatment. All experimental protocols were approved by the Animal Ethics Committee of the University of Auckland. The study was performed according to the ‘Guide for the Care and Use of Laboratory Animals’ [51], and this manuscript is consistent with the ‘ARRIVE guidelines for the reporting of animal research’ [52].

The dosage used here was based on those employed in known clinical applications of TETA (as TETA dihydrochloride or trientine) in the treatment of patients with Wilson’s disease, and for the experimental therapy of diabetes [10]. In brief, dosages employed for the treatment of Wilson’s disease in adults typically vary from 750–2000 mg/day (equivalent to ~11-29 mg/kg-day in 70-kg adults) [53]. Here, we administered TETA dihydrochloride in the drinking water to diabetic rats at 20 mg/day (equivalent to ~68 mg/kg-day of trientine in 250-g rats). This dosage is supported by our published dose-rising phase-1 clinical trial, where we showed that dosages of 1200 and 3600 mg/day (equivalent to ~17 and 51 mg/kg-day in 70-kg adults) were effective and well tolerated in healthy adult human volunteers [54], and also by our phase-2 trial where 1200 mg/day of trientine (~17 mg/kg-day in 70-kg adults) administered for 12 months markedly improved LV mass in T2D patients with LV hypertrophy [48].

We have extensively validated the STZ-based model we chose to employ here by showing that these diabetic rats respond both to untreated diabetes and following TETA treatment, in ways that closely reflect responses in patients with T2D [8, 10, 16, 45, 48, 49, 55]. To date, we have shown that TETA treatment exerts substantial effects to restore LV mass in patients with LV hypertrophy [48], similar to its effects in rats [8, 49], and that the dosage-responsiveness of TETA-mediated removal of copper from the body is similar in rats [8] and human volunteers [54, 56]. These findings provide strong support for our use of this model for the current studies.

We measured cardiac function in isolated, ex-vivo perfused working hearts, as previously detailed [8, 49]. On the experimental day, rats were anesthetized (isoflurane), heparinized (1,000 IU/kg i.v.), and hearts excised and immersed in 4°C Krebs-Henseleit bicarbonate buffer (KHB). Retrograde (Langendorff) perfusion was established (KHB, 37°C, gassed with O₂:CO₂ 95:5 (vol/vol). Working-mode perfusion was then established (preload, 10 cmH₂O; afterload, 55.9 mmHg) with pacing (300 bpm; Digitimer). Intra-chamber LV pressure (SP855; AD Instruments), aortic pressure (PX23XL, Stratham Gould), and aortic (Transonic T206) and coronary flows were measured; pressure and flow data were recorded (Powerlab16s, ADI); and the maximum rate of ventricular pressure development (+dP_LV/dt) and minimum rate of relaxation (−dP_LV/dt) were derived. Atrial filling pressure was decreased (to 5 cmH₂O) and then increased (in seven equal steps of 2.5 cmH₂O to 20 cmH₂O [final]), and 1-min averages were extracted. Filling pressure was then fixed at 10 cmH₂O, and afterload at 75 mmHg.

Copper concentrations were determined in dry ex vivo LV-tissue by using a reference method, PIXE coupled with RBS [57]. The calibration, measurements, and limits of detection were based on the areas of the K_α x-ray peaks as measured by the software package GUPIX Elemental, and concentrations were extracted from PIXE spectra using GUPIX software (University of Guelph, ON, Canada), as previously described [10, 46].

Here, RT-qPCR was performed according to ‘The Minimum Information for Publication of Quantitative Real-Time PCR Experiments (MIQE)’ guidelines [58]. Total RNA was extracted from rat LV tissues that had previously been preserved (RNAlater, Ambion), by using an RNeasy mini kit (Qiagen), and was reverse-transcribed into cDNA using a first-strand cDNA synthetase kit (Roche Diagnostics). Real-time qPCR protocols were designed based on the MIQE guidelines and performed using the SYBR Green technology on a LightCycler rapid thermal cycler (Roche Diagnostics), in accordance with the manufacturer’s instructions. Quantification results for mRNA levels for each target gene were normalized to the geometric mean of the expression of three optimal reference genes for the study of gene expression in cardiac tissue (Rpl13a, Tbp and Ndc1). Evaluation of reference genes and PCR primers employed is described in Additional file 1.

Protein extracts from LV tissues were prepared by adding ice-cold RIPA buffer supplemented with a protease inhibitor cocktail (Roche, Basle, Switzerland), and homogenised using a TissueLyser II (Qiagen). After centrifugation (20 min, 16,500 × g, 4°C), protein concentrations of supernatants were determined by BCA assay (Thermo Scientific). Western blotting was performed as previously described [9]. Briefly, 25 μg of each sample was loaded onto a 12% Bis-Tris-polyacrylamide gel (Invitrogen) and transferred to nitrocellulose membranes. Blots were first probed with specific antibodies as follows: CTR1 (sc-66847, Santa Cruz Biotechnology Inc); CTR2 (sc-104852, Santa Cruz Biotechnology Inc); MT1/2 (sc-11377, Santa Cruz Biotechnology Inc), and then incubated with an HRP-conjugated IgG, (dilution 1:10,000, Jackson Laboratories). Specific protein bands were detected by using an enhanced chemiluminescent substrate (ECL Plus, GE Healthcare, Buckinghamshire, UK) and imaged using an LAS 3000 image reader (Fuji Photo Film Co. Ltd, Tokyo, Japan). Ponceau-S staining was used as the reference for loading controls, and band intensities were measured using MultiGauge software v2.0 (Fuji Photo Film).

LV tissues were homogenized in ice-cold 0.1 M Tris/HCl, pH 7.4 containing 0.5% Triton X-100, 5 mM mercaptoethanol and 0.1 mg/ml PMSF using a TissueLyser II (Qiagen). The crude tissue homogenates were then centrifuged (14,000×g, 5 min, 4°C) and the supernatants collected: these contained total SOD activity from combined cytosolic and mitochondrial sources. Following incubation with 0.4 vol of a solution containing ethanol and chloroform (25:15 vol/vol) to inactivate the mitochondrial SOD2 [59], mixtures were centrifuged at 5000 × g for 15 min and aliquots of the supernatant used to measure protein concentration by BCA assay, and Cu/Zn-SOD activity using the Superoxide Dismutase Activity Assay kit (Abcam). The latter works by measuring the inhibition of the reduction of a water-soluble tetrazolium salt (WST-1) which produces a water-soluble formazan dye upon reduction by O₂^●- radicals, which are generated by xanthine oxidase and inhibited by SOD. One unit of SOD activity is the amount necessary to inhibit the xanthine oxidation by 50%. The IC₅₀ (50% inhibition activity of SOD) was detected by a colorimetric method at 450 nm using an absorbance microplate reader (SpectraMAX 340, Molecular Devices) and the results expressed per mg of total protein.

Immunofluorescent staining was performed using standard techniques as described. Briefly, 6-μm LV sections frozen in OCT were cut, fixed in acetone, and blocked with 5% (vol/vol) normal donkey serum (NDS) in PBS. They were then incubated overnight at 4°C with primary antibody in 5% NDS (vol/vol). Primary antibodies used in these experiments were: anti-CTR1 (at 1:200 vol/vol final dilution, Abnova); anti-CTR2 (1:50, Santa Cruz); MT1/2 (1:100, Abcam); CCS (1:200, Novus Biologicals); ATOX1 (1:100, Santa Cruz); ATP7A (1:250, Abcam), ATP7B (1:150, Novus Biologicals), and N-cadherin (1:50, Abcam). Following 3 × washes with PBS, sections were incubated with appropriate fluorescently-labeled (−FITC or -Rhodamine Red, as required) donkey anti-rabbit/mouse/chicken IgG secondary antibodies (Jackson ImmunoResearch). After incubation, sections were washed and incubated overnight with WGA-Oregon Green 488 (1:5000; Molecular Probes) in PBS at 4°C. Finally, sections were washed again and mounted with Prolong Gold anti-fade reagent with DAPI (Invitrogen). The immunofluorescent images were captured at 40×, 60× or 100× magnifications using a confocal fluorescent microscope (Olympus FV1000, Fuji).

The fluorescently-labeled signals from each image were processed and quantified using a custom-written IDL analysis program based on signal-thresholding, as previously described [9, 10]. Briefly, the original image was masked to exclude non-labeled areas. The threshold for quantifying fluorescent labeling was then established, based on information in the masked image’s histogram compared with the original image. The criteria for the thresholding process were maintained equivalent across all experimental groups. Results were expressed as percentages of corresponding total cross-sectional areas. To avoid any possible confounding effect of orientation and origin of the muscle fibers on the signal quantification, at least five transverse optical-sectional images from each section of the LV-myocardium, two sections from each animal, and four animals per group, were analyzed.

Data have been expressed as mean ± SEM unless stated otherwise. Significance of between-group differences was determined by unpaired Student’s t-tests, one-way ANOVA, or two-way ANOVA, as appropriate. The primary a priori null hypothesis in each contrast was that there was no difference between signals from TETA-treated diabetic and untreated diabetic states. A second between-group comparison was performed in each case, with the a priori null hypothesis being that there was no difference between signals from control and diabetic animals: this latter contrast was included to ensure that effects of diabetes were present as a starting condition for comparison with TETA-treated samples. Where stated in the figure legends, correction for multiple-comparisons has been applied. All contrasts were two-tailed, and values of P < 0.05 have been considered significant.

LV-copper levels were markedly deficient (decreased to ~50% of normal) after 16-weeks’ untreated diabetes. Strikingly, LV-copper levels were fully restored to normal (control) concentrations when TETA treatment was instituted after 8-weeks’ diabetes and maintained for a further 8 weeks (Figure 1A): this effect was accompanied by substantial improvements in cardiac function, including in cardiac output, and indexes of systolic and diastolic function (Table 1). However, TETA treatment did not change any index of cardiac function in non-diabetic control rats (Table 1), consistent with previous reports [8, 60].

Immunoblotting for the high-affinity copper transporter, CTR1, indicated the presence of dimeric (~45 kD) and trimeric (~75 kD) forms, both of which were diminished in diabetic LV myocardium as compared to control (Figure 1B and 1C). TETA treatment did not restore CTR1 expression (Figure 1B and 1C), and Ctr1 mRNA levels were unchanged in all treatment groups: furthermore, CTR1 protein levels were also unchanged in TETA-treated controls (data not shown).Confocal imaging was used to further investigate the role of CTR1 in copper-deficient diabetic LV. CTR1 was found to co-localize predominantly with the myocardial T-tubule membrane system in control LV (Figure 2A); however, CTR1 staining appeared to be diminished and irregular in diabetic LV, consistent with possible disruption of the T-tubules themselves. TETA-treated diabetic tissues may have a slightly more organized CTR1 staining localization as compared to untreated diabetes, although this effect is subtle. In addition, quantitative immunofluorescent analysis of CTR1 confirmed the diminished staining intensity in both diabetic and TETA-treated diabetic LV (Figure 2B). These results are consistent with CTR1 immunoblotting data, and cumulatively support the down-regulation of CTR1 as being a potential mechanism of impaired copper uptake and subsequent copper deficiency in diabetic LV.

As TETA treatment did not restore CTR1 protein expression in diabetes, it seemed unlikely that this could be a mechanism by which TETA restores LV-copper levels. Interestingly, immunoblotting for the lower affinity-copper transporter, CTR2, showed increased expression of both dimeric (~32kD) and tetrameric (~64 kD) CTR2 forms in diabetic LV (Figure 1D). CTR2 is known to form functional homo-multimers and the specificity of bands was verified here by antibody blocking (data not shown). Furthermore, Ctr2 mRNA was also increased in diabetic LV as compared to control (Figure 1E). Further studies showed that, unlike CTR1, TETA treatment significantly elevated CTR2 expression, although rather than merely restoring levels, it further increased both Ctr2 mRNA and protein expression as compared to untreated diabetic LV (Figure 1D and 1E), whereas by contrast, it has no effects on CTR2 expression in non-diabetic controls (data not shown). These results are consistent with an increase of copper uptake via CTR2 as a key part of the mechanism by which TETA restores myocardial copper content.We employed confocal imaging to further examine the role of CTR2. Transverse sections of LV showed CTR2 localized to the nucleus and intracellular vesicles of myocytes, along with subtle staining of the sarcolemmal membrane in controls. This staining pattern was more intense in diabetic LV, particularly in the intracellular small vesicles, from which copper is known to translocate to the cytosol (Figure 3A). This response correlates well to enhanced CTR2 protein expression, which could represent an endogenous, self-compensating response towards maintenance of cytosolic copper in diabetes. Interestingly, LV from TETA-treated animals showed further increases in CTR2 staining at or near the outer sarcolemmal membrane: increased expression of CTR2 in this location could well mediate enhanced uptake of extracellular copper into cardiomyocytes. Moreover, longitudinal LV sections confirmed enhanced staining of CTR2 at the cardiomyocyte periphery, particularly at the intercalated discs of TETA-treated diabetic LV (as compared to control and untreated-diabetic LV) (Figure 3B), consistent with increased transport of copper between neighboring myocardial cells. Co-staining of N-cadherin with anti-CTR2 confirmed the intercalated disc localization (Figure 3C). In addition, the enhanced localization of CTR2 at the outer sarcolemmal membrane in TETA-treated LV is a candidate mechanism by which TETA could enhance copper uptake and restore LV copper content.

Deficient myocardial-copper levels could be associated with altered expression or function of intracellular copper-regulating proteins. We therefore examined the expression of the copper-binding proteins, MT1 and MT2, which act to maintain low levels of free transition-metal ions (particularly Cu and Zn) within cells, and play essential roles in copper-ion detoxification [61]. RT-qPCR demonstrated a significant reduction in mRNA levels of Mt1 and Mt2 in diabetic LV compared with control tissue (Figure 4A and 4B), and TETA treatment did not reverse these changes. The lower levels of Mt1 and Mt2 mRNAs observed in diabetic LV tissue are consistent with the observed decreases in myocardial copper content, since expression of Mt1 and Mt2 is known to be primarily regulated by intracellular metal concentrations at the level of transcription [61].Confocal micrographs of control LV stained for combined MT1/2 showed intense sarcoplasmic and nuclear localization, with lower-intensity staining of the sarcolemmal membrane (Figure 4C). MT1/2 staining intensity was notably diminished in both untreated and TETA-treated diabetic cardiomyocytes, with some stronger staining persisting near the sarcolemmal surface. Quantification of the fluorescent signal area confirmed diminished staining intensity of MT1/2 in diabetic and TETA-treated diabetic LV as compared to control (Figure 4D). Furthermore, no notable change in the localization of MT1/2 signal intensity was detected in diabetic or corresponding areas of TETA-treated LV. Together, these results are consistent with decreased capacity of cytosolic copper binding and thus decreased protection against copper toxicity in diabetes.

Oxidation of the sulfhydryl groups in MT can lead to their polymerization and altered copper-binding properties. Western blotting of LV did not detect monomeric MT1/2, but instead several higher molecular-weight bands of ~30, 45 and 70 kD in all 3 treatment groups (Figure 5A) (specificity of MT bands was verified by antibody blocking studies, data not shown). Combined densitometry of all 3 polymerized MT bands indicated a reduction of ~34% expression in diabetes which TETA treatment did not significantly alter (Figure 5A), consistent with the lowering in the MT immunofluorescent signal shown above.However, densitometry of the respective MT band sizes revealed that ~70 kD MT was significantly increased, by ~50%, in diabetic LV tissue (Figure 5B), whereas the ~45kD and ~30 kD bands were significantly reduced, by 48% and 40% respectively, in diabetic LV as compared to control (Figure 5C and 5D). Interestingly, TETA treatment partially corrected diabetes-induced changes in the expression of all 3 polymerized MT bands (Figure 5B, 5C and 5D). These results provide clear evidence that TETA treatment could result in decreased polymerization of MT, which is consistent with decreased oxidative stress in the TETA-treated diabetic heart.

CCS delivers copper to the apoenzyme of SOD1 to yield the active holoenzyme. RT-qPCR demonstrated Ccs mRNA levels were significantly down-regulated, by ~30%, in diabetic LV compared with control tissue (Figure 6A). Interestingly, TETA treatment reversed this decrease in mRNA expression to levels similar to non-diabetic control values. Furthermore, quantitative immunofluorescent analysis showed that diabetic LV had decreased CCS-specific signal areas, which were normalized by TETA treatment (Figure 6B). These results show that transcriptional and translational dysregulation of CCS occurs in diabetes and is rectified by TETA treatment.

As the down-regulation of CCS in diabetes could well impair its ability to deliver copper to SOD1, we investigated whether SOD1 activity was altered in diabetes. Indeed, SOD1 activity was significantly decreased in diabetic LV, whilst TETA-treated diabetic tissue showed restoration of SOD1 activity (Figure 6C). ELISA of SOD1 protein levels showed a significant decrease in diabetic myocardium compared to control values, which TETA treatment did not restore (Figure 6D), although no detectable between-group differences in Sod1 mRNA levels were present (data not shown). These results indicate that SOD1 is subject to post-transcriptional down-regulation in diabetic LV tissue. We also found that TETA had no detectable effects on SOD1 expression in diabetic LV tissue (Figure 6D). These results, when taken together, indicate that decreased expression of SOD1 and CCS contribute to the deficiency in LV-myocardial SOD1 activity in diabetes. They also indicate that TETA treatment activates SOD1 by increasing the supply of copper for incorporation into the enzyme via improved CCS function, without changing the expression of SOD1 protein itself.

Copper supply to ATP7A and ATP7B is mediated by the copper-chaperone protein ATOX1, which binds to and transports copper to compartments containing the target ATPases. In light of their roles in copper regulation, we investigated the myocardial expression of these three proteins. We found no significant changes in the mRNA levels of Atox1 or Atp7b in diabetic or TETA-treated diabetic LV compared to control (data not shown). However, quantitative immunofluorescence of signal areas of ATOX1 and ATP7B, demonstrated that both were decreased in diabetic LV, and remained so following TETA treatment (Figure 7A and 7B).Confocal micrographs of longitudinal LV sections from all treatment groups showed that ATP7B localizes in the nuclear/peri-nuclear region (presumably in the trans-Golgi network), and also with the T-tubules (Figure 8A) and intercalated disks (as indicated by co-staining with N-cadherin) (Figure 8B). Comparative staining of ATP7B in all 3 treatment groups showed that, in control transverse LV sections, ATP7B was once again detected in nuclear and peri-nuclear regions, but this nuclear staining was diminished in diabetic and TETA-treated diabetic LV. Instead, diabetic and TETA-treated diabetic LV had a more striking vesicular localization in the sarcoplasm, which in places was close to the sarcolemmal membrane (Figure 8C). The diminished expression and altered localization of ATP7B, combined with lowered ATOX1 expression, may cause impaired transport of copper to copper-requiring proteins, and indeed efflux from the cell/between cells, providing a putative mechanism for reversing the localized copper imbalances that may occur in diabetic myocardium.By contrast, the mRNA levels and quantitative immunofluorescent signal areas for ATP7A were both unaltered in diabetic LV as compared to control, whereas TETA-treated diabetic LV showed significantly increased ATP7A levels, implying transcriptional and translational up-regulation (Figure 7C and 7D), consistent with TETA-elicited alterations in rates of copper trafficking via ATP7A in the secretory pathway. Confocal micrographs of transverse sections showed that, in control LV, ATP7A was mainly localized to the peri-nuclear region and also showed vesicular staining throughout the sarcoplasm which may represent the intracellular trans-Golgi network membranes (Figure 7E). No notable changes in ATP7A localization in diabetic LV were apparent, whilst TETA-treated diabetic LV showed significantly enhanced vesicular and peri-nuclear staining intensity (correlating to quantitation). These results imply activation of copper translocation by ATP7A in the secretory pathway, which could lead to improve utilization of copper by cuproenzymes in the secretory compartments in response to TETA treatment.