Depressed Levels of Ca²⁺-Cycling Proteins May Underlie Sarcoplasmic Reticulum Dysfunction in the Diabetic Heart

The experimental protocol in this investigation was approved by the Animal Care Committee of the University of Manitoba and conforms with the guidelines of the Canadian Council on Animal Care Concerning the Use of Experimental Laboratory Animals.

Diabetes was induced in male Sprague-Dawley rats (175–200 g) by administering a single tail-vein injection (65 mg/kg) of STZ according to the procedure previously described (5), and the control group was injected with the vehicle only. Three weeks after the induction of diabetes, a group of animals was injected with 3 units per day of humulin N (human biosynthetic insulin isophane; Eli Lilly) for the next 3 weeks. The urine and blood glucose levels were tested by using Keto-Diastix reagent strips (Miles, Etobicoke, Canada) and the Sigma Kit (Sigma, St. Louis, MO), respectively. Plasma insulin was measured using a standard radioimmunoassay technique as described elsewhere (5). Six weeks after the injection, general characteristics and hemodynamic parameters were measured (5) (although these parameters were recorded in all animals, data are reported from five or six rats randomly selected from the group), and the animals were killed. The SR preparation was isolated by the method previously described (10,11,12,13); the whole left ventricle, including the septum, was used for the preparation of the SR membranes and cytosolic fractions. The homogenization buffer contained 10 mmol/l NaHCO₃, 5 mmol/l NaN₃, 15 mmol/l Tris-HCl (pH 6.8), and a mixture of protease inhibitors (1 μmol/l leupeptin, 1 μmol/l pepstatin, and 100 μmol/l phenylmethyl-sulfonylfluoride).

Ca²⁺-uptake activity of the SR vesicles was determined as described previously (11,12,13). The standard reaction mixture (total volume 250 μl) contained 50 mmol/l Tris-maleate (pH 6.8), 5 mmol/l NaN₃, 5 mmol/l ATP, 5 mmol/l MgCl₂, 120 mmol/l KCl, 5 mmol/l K-oxalate, 0.1 mmol/l EGTA, 0.1 mmol/l ⁴⁵CaCl₂ (12,000 cpm/nmol), and 25 μmol/l ruthenium red. The initial concentration of free Ca²⁺ in this medium was 8.2 μmol/l; ruthenium red was added to inhibit the Ca²⁺-release channel activity under these conditions. The reaction was initiated by the addition of SR vesicles (6 μg) to the Ca²⁺-uptake reaction mixture and was terminated 1 min later by filtering 200 μl aliquot of the reaction mixture. The filters were washed, dried at 60°C for 1 h, and counted in a β-liquid scintillation counter.

Ca²⁺-release activity of the SR vesicles was measured by the method previously described (11,12). In brief, the SR fraction (62 μg protein) was suspended in a total volume of 625 μl of loading buffer containing 100 mmol/l KCl, 5 mmol/l MgCl₂, 5 mmol/l K-oxalate, 5 mmol/l NaN₃, and 20 mmol/l Tris-HCl (pH 6.8). After incubation with 10 μmol/l ⁴⁵CaCl₂ (20 mCi/l) and 5 μmol/l ATP for 45 min at room temperature, Ca²⁺-induced Ca²⁺ release was carried out by adding 1 mmol/l EGTA plus 1 mmol/l CaCl₂ to the reaction mixture. The reaction was terminated at 15 s by Millipore filtration. The Ca²⁺-induced Ca²⁺ release was completely prevented (95–97%) by treatment of SR preparations with 20 μmol/l ryanodine.

Protein content for RyR, SERCA2a, and PLB in the SR preparations was determined according to the procedure previously described (12). SR (20 μg) samples were separated on gels, electrophoretically transferred to membranes, and incubated with the appropriate antibodies. Antigen-antibody complexes of all probed membranes were detected by the enhanced chemiluminescence detection system (Amersham, Buckinghamshire, U.K.). The intensity of each band was scanned by an imaging densitometer with the aid of Molecular Analyst Software Version 1.3 (BioRad, Hercules, CA).

For the phosphorylation experiments, the SR preparation was isolated in the presence of 1 mmol/l Na pyrophosphate, a phosphatase inhibitor, to prevent any dephosphorylation occurring during the isolation procedure (10,13). Both the homogenization buffer and the phosphorylation assay medium contained phosphatase inhibitor. SR protein phosphorylations by the endogenous CaMK and exogenously added PKA were determined by the previously described method (10,13). The kinases and phosphatase activities of SR and cytosolic fractions were measured using Upstate Biotechnology assay kits as previously described (10,13).

The results are expressed as means ± SE and were statistically evaluated using analysis of variance, followed by Student’s t test. P < 0.05 was considered the threshold for statistical significance between the control and experimental groups.

STZ-injected diabetic animals showed a significant decrease in body weight and an increase in heart/body weight ratio compared with controls (Table 1). The development of diabetes was confirmed by a marked elevation in the plasma glucose level and a decrease in the plasma insulin level in STZ-injected animals. The results in Table 1 also show that the rate of pressure development and the rate of pressure fall were significantly decreased in the 6-week diabetic animals. Furthermore, the SR Ca²⁺-uptake and -release activities were significantly decreased in the hearts of diabetic animals compared with those of controls. To test whether the depressed SR activities in the diabetic hearts were caused by a decrease in the content of SR Ca²⁺-cycling proteins, the levels of RyR, SERCA2a, and PLB were estimated in the control and diabetic hearts. Figure 1 shows that the contents of all of these proteins were decreased (P < 0.05) in the diabetic hearts. Treatment of 3-week diabetic animals with insulin was found to significantly reverse the diabetes-induced changes in the general characteristics, cardiac performance, SR function, and levels of SR Ca²⁺-cycling proteins (Table 1 and Fig. 1).

Because SR Ca²⁺-uptake and -release activities are regulated by CaMK-mediated phosphorylation of the Ca²⁺-cycling proteins (11), the endogenous SR CaMK phosphorylation of SR proteins was determined. Figure 2 shows the SR protein profile (panel A), the corresponding autoradiogram depicting SR protein phosphorylation (panel B), and the analysis of the RyR, SERCA2a, and PLB phosphorylations in the control, diabetic, and insulin-treated diabetic hearts, respectively (panel C). CaMK phosphorylation of all three proteins was significantly increased in the hearts of diabetic animals compared with those of controls, and this increase was not attenuated on insulin treatment; the identity of the phosphorylated proteins has been previously established (13,14). There was no clear decrease in the protein content of the bands representing RyR and SERCA2a in the diabetic group compared with the control group (Fig. 2A). This is in contrast to the significant decrease detected by Western blot analysis (Fig. 1). This discrepancy may be a result of the fact that the bands observed in Fig. 2A may contain other proteins comigrating with the RyR and SERCA2a proteins. Moreover, Western blot analysis uses specific monoclonal antibodies recognizing RyR and SERCA2a and is therefore an accurate method for determining the content of proteins. To understand the reason for the increase in the endogenous CaMK phosphorylation in the diabetic hearts, the SR and cytosolic CaMK activities were assessed. Table 2 shows that the SR CaMK activity was significantly increased in the hearts of diabetic animals, whereas the cytosolic CaMK activity was unaltered when compared with the hearts of controls. Insulin treatment did not reverse the diabetes-induced increase in SR CaMK activity.

Because a defect in PKA-mediated phosphorylation may also contribute to depressed SR function (10,13), PKA phosphorylation of SR proteins was examined in the control, diabetic, and insulin-treated diabetic hearts. Figure 3 shows the SR protein profile (panel A), the corresponding autoradiogram depicting SR protein phosphorylation (panel B), and the analysis of RyR and PLB phosphorylation by PKA in the control, diabetic, and insulin-treated diabetic groups (panel C). PKA-mediated phosphorylation of RyR and PLB was significantly increased in the hearts of diabetic animals compared with those of controls, and this increase was not attenuated on insulin treatment; the identity of these phosphorylated proteins has been previously established (11,13). To determine the mechanisms for the increase in PKA-mediated phosphorylation in the diabetic hearts, the SR and cytosolic PKA activities were assessed. Table 2 shows that the SR PKA activity was significantly increased in the hearts of diabetic animals compared with those of controls, whereas the cytosolic PKA activity was unaltered. Insulin treatment did not reverse the diabetes-induced increase in SR PKA activity.

In view of the association of CaMK- and PKA-mediated phosphorylations of the SR proteins with dephosphorylation, the SR and cytosolic phosphatase activities were assessed in the diabetic hearts. A significant increase in both the SR and cytosolic phosphatase activities was observed in the hearts of diabetic animals compared with those of controls (Table 2). Insulin treatment did not attenuate the diabetes-induced increase in SR phosphatase activity.

Our results show impaired cardiac function and decreased SR Ca²⁺-uptake and -release activities in the diabetic heart. Although a reduction of SR Ca²⁺-uptake, Ca²⁺-release, and ryanodine binding has been previously demonstrated in the diabetic heart (5,6), the mechanisms underlying alterations in SR function were not fully established. In this regard, the decrease in protein contents of RyR, SERCA2a, and PLB observed in the diabetic heart in this and other studies (15,16) could explain the depression in SR Ca²⁺-uptake and -release activities. However, it should be pointed out that the depression of the Ca²⁺-cycling protein levels may represent one of several factors, such as increased phospholipid and fatty acid content in the SR vesicles (5,17), which have been proposed to explain SR dysfunction in the diabetic heart. Nonetheless, in view of the role of SR in cardiac contraction and relaxation processes, the observed reduction in SR function may explain the depressed cardiac performance in diabetic animals. Such a view does not exclude other mechanisms, such as depressed myofibrillar ATPase (18), sarcolemmal Na⁺-K⁺ ATPase (19), and Ca²⁺-pump and Na⁺-Ca²⁺ exhange activities (20), which have been shown to contribute to cardiac dysfunction in the diabetic heart.

Current evidence favors enhanced SR Ca²⁺-uptake under physiological conditions as a consequence of phosphorylation of the SERCA2a by CaMK (14) and phosphorylation of PLB by both CaMK and PKA (21,22,23). Phosphorylation of PLB by both CaMK and PKA has been shown to relieve the inhibitory action of PLB on SERCA2a, resulting in an increased affinity of SERCA2a for Ca²⁺ (21,22,23). CaMK phosphorylation of SERCA2a has been reported to increase V_max of SR Ca²⁺ transport (14). Although Odermatt et al. (24) failed to demonstrate the CaMK-mediated stimulation of V_max of Ca²⁺ transport, the data from our previous study (10) supported the observation of other investigators (14). SR Ca²⁺-release was also shown to be regulated by RyR phosphorylation directly because both CaMK and PKA were reported to promote SR calcium release (25,26,27). Thus, the impairment in the SR Ca²⁺-uptake and -release can be seen as a result of a decrease in SR protein phosphorylations. However, we observed an increase in the in vitro endogenous CaMK-mediated phosphorylation of RyR, SERCA2a, and PLB, as well as the PKA-mediated phosphorylation of RyR and PLB in the diabetic heart. A recent study (28) has linked RyR hyperphosphorylation by PKA to defective Ca²⁺-release channel function in the failing heart. Because CaMK and PKA have been observed to be endogenous to SR (29,30), the observed increase in susbstrate phosphorylation may be explained on the basis of an increase in endogenous CaMK and PKA activities. Moreover, these increases in SR protein kinase activities were specific because the cytosolic CaMK and PKA activities were unaltered in the diabetic heart. Although other investigators (31) did not observe any change in the levels of CaMK and PKA activities in the hearts of diabetic rats, it is noted that whole-heart homogenate was used in their study (31), whereas SR vesicles were used in the present investigation.

In view of the role of endogenous phosphatases in dephosphorylation of membrane proteins, the increased endogenous CaMK- and PKA-mediated phosphorylations in the diabetic heart may be attributed to depressed endogenous phosphatase activities. On the contrary, we observed an increase in the endogenous and the cytosolic phosphatase activities in the diabetic heart. Such an increase in the total phosphatase activity may be caused by increases in protein phophatase-1 and -2 activity in the diabetic heart (32). It is likely that the enhanced endogenous phosphatase activities may compensate for the marked elevation in the endogenous CaMK and PKA activities in such a way that the SR phosphorylation/dephosphorylation cycle is unaffected in the diabetic heart. This view is consistent with an earlier report (33) suggesting that the SR regulatory mechanisms may not play any role in the cardiac SR dysfunction observed in the diabetic heart. It is possible that the diabetic heart may respond to the depressed SR function by enhancing the activities of the endogenous kinases; however, these compensatory mechanisms may be inadequate to maintain the SR function because of increased dephosphorylation.

The results in this study indicate that a depression in the content of SR Ca²⁺-cycling proteins may contribute to SR dysfunction in the diabetic heart. Because the treatment of STZ-induced diabetic rats with insulin has been reported to prevent diabetes-induced changes in the heart (3–6,17–20), it is likely that insulin deficiency may play a role in regulating cardiac protein content. Increased proteolysis shown to occur in the diabetic heart (34) is reduced by insulin (35). Our results showing partial or full recovery of the SR Ca²⁺-cycling protein contents in the diabetic heart on insulin treatment support the ideas mentioned above. However, it must be mentioned that insulin treatment did not completely normalize diabetes-induced alterations in general characteristics, hemodynamic parameters, and SR Ca²⁺-uptake and -release activities, and SERCA2a content. Earlier studies (5,36,37) have also reported an incomplete recovery of some of the above parameters in the insulin-treated diabetic animals. Although the insulin dose (3 units per day) used in this study was selected based on our previous experience (5,32,36), it is possible that a higher dose of insulin might recover all of the above parameters to control values. Furthermore, the 3-week duration of treatment with insulin in this study may not be sufficient for full recovery of all subcellular changes in the diabetic heart. These views are consistent with partial recovery of plasma insulin and glucose levels in diabetic animals treated with insulin. Nonetheless, it should be noted that the diabetes-induced changes in SR, RyR, and PLB contents were reversed completely by treatment of diabetic animals with 3 units of insulin per day for 3 weeks. On the other hand, insulin treatment did not affect the diabetes-induced increase in CaMK and PKA activities or phosphorylation of their substrates. A recent study has also reported the ineffectiveness of insulin treatment in reversing the diabetes-induced increase in CaMK and PKA activities in liver tissue homgenates (31). Thus, recovery of SR function as a result of insulin treatment may be mediated via protection of the SR Ca²⁺-cycling proteins rather than through regulatory mechanisms, and, therefore, SR proteins could be potential targets for the treatment of diabetes.

By using the hearts of diabetic rats, our study is the first to demonstrate enhanced in vitro CaMK and PKA phosphorylation of SR Ca²⁺-cycling proteins, specific increases in the endogenous CaMK and PKA activities, and increased SR and cytosolic phosphatase activities. Furthermore, we have shown partial or full recovery of SR Ca²⁺-cycling proteins, unlike CaMK- and PKA-mediated SR phosphorylations and their activities in the diabetic heart, upon treatment with insulin. Thus, these results provide new information regarding the mechanisms underlying cardiac dysfunction in chronic diabetes.

This study was supported by a grant from the Canadian Institutes of Health Research (CIHR Group in Experimental Cardiology). N.S.D. holds the CIHR/Pharmaceutical Research and Development Chair in Cardiovascular Research supported by Merck Frosst Canada, and R.M.T. is a predoctoral fellow of the Heart and Stroke Foundation of Canada.