Rac1 Is Required for Cardiomyocyte Apoptosis During Hyperglycemia

This investigation conforms with the guide for the care and use of laboratory animals published by the U.S. National Institutes of Health (NIH Publication No. 85–23). All experimental procedures were approved by the Animal Use Subcommittee at the University of Western Ontario, Canada. Breeding pairs of C57BL/6 mice, mice bearing the modified Rac1 gene containing loxP sites, and db/db mice were purchased from the Jackson Laboratory. Transgenic mice with cardiomyocyte-specific expression of Cre recombinase (Cre) under the control of α-myosin heavy chain (α-MHC) were generously provided by Dr. Dale Evan Abel (University of Utah, UT) (23). A breeding program for mice was implemented at our animal care facilities.

Adult male rats (Sprague Dawley, 200 g body weight) were purchased from Charles River Labs. Adult rat ventricle cardiomyocytes (ARVC) were isolated and cultured as described (24).

Adult male mice (2 months old) were intraperitoneally injected with a single dose of streptozotocin (STZ) at 150 mg/kg body weight, dissolved in 10 mmol/l sodium citrate buffer (pH 4.5). On day 3 after STZ treatment, whole blood was obtained from the mouse tail vein and random glucose levels were measured using OneTouch Ultra 2 blood glucose monitoring system (LifeScan, Mountainview, CA). Blood glucose content of 20 mmol/l or greater was chosen as hyperglycemia for the present study, whereas citrate buffer-treated mice were used as normoglycemia (blood glucose <12 mmol/l).

Cardiomyocytes were infected with adenoviral vectors containing a dominant-negative mutant Rac1 (Ad-RacN17, Vector Biolabs) or β-gal (Ad-gal, Vector Biolabs) as a control at a multiplicity of infection (MOI) of 100 PFU/cell. Adenovirus-mediated gene transfer was implemented as we previously described (24).

To knock down gp91^phox (Nox2) and p47^phox expression, a small interfering RNA (siRNA) against rat Nox2 or p47^phox was obtained (Santa Cruz Biotechnology, Santa Cruz, CA) and two different scramble siRNAs (Con1 and Con2) were employed as control. Transfection was performed using TransMessenger Transfection Reagent (Qiagen) according to manufacturer's protocol as described in our recent study (25).

Activated Rac1 was determined by p21-binding domain of p21-activated protein kinase 1 pull-down assay (Rac Activation Assay Kit; Cell Biolabs) according to the manufacturer's protocol.

NADPH oxidase activity was assessed in cell lysates by lucigenin-enhanced chemiluminescence (20 μg of protein, 100 μmol/l NADPH, 5 μmol/l lucigenin) with a multilabel counter (Victor3 Wallac) (25).

The formation of ROS was measured by using the ROS-sensitive dye, 2,7-dichlorodihydro-fluorescein diacetate (DCF-DA, Invitrogen), as an indicator. The assay was performed on freshly dissected heart tissues. Samples (50 μg proteins) were incubated with 10 μl of DCF-DA (10 μmol/l) for 3 h at 37°C. The fluorescent product formed was quantified by spectrofluorometer at the 485/525 nm. Changes in fluorescence were expressed as arbitrary unit.

As described in detail previously (26), caspase-3 activity in myocardial tissues and cardiomyocytes was measured by using a caspase-3 fluorescent assay kit (BIOMOL Research Laboratories).

To localize cells undergoing nuclear DNA fragmentation in the myocardium, in situ terminal deoxynucleotidyl transferase–mediated dUTP nick-end labeling (TUNEL) was performed using an in situ apoptosis detection kit (Roche Biochemicals) as described previously (26).

Mitochondria were isolated from the freshly harvested heart as previously described (27). Mitochondrial superoxide generation was measured by lucigenin-enhanced chemiluminescence with a multilabel counter (Victor3 Wallac) using NADH (250 μmol/l) or succinate (2 mmol/l) (27). The formation of ROS in freshly isolated mitochondria was measured on addition of pyruvate/malate by using DCF-DA as an indicator (28).

Mouse hearts were isolated and perfused on a Langendorff-system. Myocardial function was then determined as described in our previous study (25). Maximal and minimal first derivatives of force (+dF/dt_max and −dF/dt_min) as the rate of contraction and relaxation were analyzed by PowerLab Chart program (ADInstruments).

All data were given as means ± SD. Differences between two groups were compared by unpaired Student's t test. For multigroup comparisons, ANOVA followed by Student Newman-Keuls test was performed. A value of P < 0.05 was considered statistically significant.

As shown in Fig. 1, hyperglycemia significantly increased Rac1 activity in STZ-treated compared with citrate buffer–treated hearts (Fig. 1,A). In cultured ARVC exposed to normal (5.5 mmol/l) or high glucose (33 mmol/l) for 1, 6, and 24 h, Rac1 activity was significantly upregulated by high glucose compared with normal glucose (Fig. 1 B).

Mice with cardiomyocyte-specific Rac1-ko were generated by crossing the floxed Rac1 mice with mice overexpressing Cre under the control of α-MHC as we recently described (29). PCR analysis identified mice carrying both the homozygous loxP-Rac1-loxP gene and the α-MHC-cre transgene (Fig. 1,C) as Rac1-ko mice. Deficiency of Rac1 decreased Rac1 mRNA expression in the heart (Fig. 1,D). Isolated cardiomyocytes of Rac1-ko mice exhibited a significant reduction of Rac1 protein expression as compared with their wild-type littermates (Fig. 1 E). These animals grow normally into adulthood without any health problems. Under normal condition, there were no differences in heart weight, body weight, ratio of heart weight over body weight, heart rate (data not shown), and cardiac contractility between Rac1-ko mice and their wild-type littermates. Histological analysis showed no pathological changes in Rac1-ko hearts. These observations suggest that cardiomyocyte-specific deletion of Rac1 does not affect basal myocardial function and morphology of the heart, thus excluding the possible effects of Cre expression on the heart of Rac1-ko mice.

Hyperglycemia was induced in Rac1-ko mice and their corresponding wild-type littermates including wild-type, floxed Rac1, and Cre transgenic (Cre+) mice by STZ. The levels of blood glucose were monitored on day 0, 3, and 9 after STZ injection. To eliminate possible confounding effects of STZ, insulin was immediately given using a long-term insulin preparation (Lantus Insulin glargine injection; sanofi-aventis Canada, Laval, Canada) at a dose of 5–15 units per mouse per day to maintain the blood glucose levels between 5–12 mmol/l when hyperglycemia was greater than 20 mmol/l on day 3 after STZ treatment in wild-type mice. In sham- or STZ-injected mice, blood glucose was not different between Rac1-ko mice and their corresponding wild-type littermates including wild-type, Rac1 floxed, and Cre+ mice (data not shown).

We first determined Rac1 activity in the heart. There was no difference of Rac1 activity in between Rac1-ko corresponding wild-type littermates under basal or diabetic conditions. Therefore, Rac1 floxed mice were presented as littermate wild-type controls for Rac1-ko mice. Rac1 activity was much lower in Rac1-ko compared with wild-type hearts during hyperglycemia (Fig. 1,F). We then examined the membrane translocation of Rac1 and p67^phox, as well as NADPH oxidase activation. As shown in Fig. 2,A, the protein levels of Rac1 and p67^phox in plasma membranes were significantly upregulated in hyperglycemic compared with normoglycemic wild-type hearts, which were correlated with the increases in NADPH oxidase activity and ROS production in hyperglycemic hearts (Fig. 2,B and C). However, myocardial ROS production was not increased in insulin supplemented mice (data not shown). There were no changes in protein levels of NADPH oxidase subunits, gp91^phox, p67^phox, and Rac1 in hyperglycemic hearts (data not shown). Deletion of Rac1 decreased membrane Rac1 and p67^phox, NADPH oxidase activation, and ROS production in the hyperglycemic heart (Fig. 2,A–C). Finally, we analyzed the effect of NADPH oxidase activation on ROS production during hyperglycemia. Adult wild-type mice were injected with STZ. Two days later, apocynin (30 mg · kg⁻¹ · day⁻¹) or vehicle was given in drinking water for 7 days. The levels of blood glucose were similar between apocynin- and vehicle-treated mice on day 9 (data not shown). Administration of apocynin blocked NADPH oxidase activity and ROS production in hyperglycemic hearts (Fig. 2 B and C). These results demonstrate that Rac1 is required for NADPH oxidase activation and ROS production during hyperglycemia.

Myocardial caspase-3 activity was not different between Rac1-ko corresponding wild-type littermates including wild-type, Rac1 floxed, and Cre+ mice under either normal or diabetic conditions (data not shown). Therefore, Rac1 floxed mice were chosen as littermate wild-type controls for Rac1-ko mice in the following experiments. Compared with citrate buffer–treated mice, hyperglycemia significantly increased myocardial caspase-3 activity (Fig. 3,A) and the number of TUNEL- positive cardiomyocytes in STZ-treated wild-type mice (Fig. 3,B and C), whereas myocardial caspase-3 activity was not increased in insulin-supplemented mice (data not shown). These results exclude the direct possible effects of STZ on myocardial apoptosis, which is in agreement with a recent report (4). Deficiency of Rac1 decreased caspase-3 activity and the number of TUNEL-positive cells in the hyperglycemic heart (Fig. 3). Similarly, inhibition of NDAPH oxidase by apocynin blocked caspase-3 activation and reduced the number of TUNEL-positive cells in hyperglycemic hearts (Fig. 3 D and E). These data demonstrate that Rac1-NADPH oxidase activation induces apoptosis in the hyperglycemic heart.

Mitochondria are known to constitutively generate superoxide radicals as a by-product of electron transport and are also important sources for intracellular ROS production in hyperglycemia (30,31). Inhibition of mitochondrial ROS production has been shown to prevent the development of diabetic cardiomyopathy (32,33). We explored the impact of Rac1-NADPH oxidase on mitochondria in producing ROS. Myocardial mitochondria were isolated from Rac1-ko mice, wild-type littermates, and apocynin-treated mice on day 9 after STZ treatment. We first determined superoxide generation in the isolated frozen-thawed mitochondria. In the normoglycemic hearts, addition of NADH but not NADPH slightly increased superoxide production and there was no difference between Rac1-ko mice and wild-type littermates. However, in the hyperglycemic hearts, mitochondria superoxide production was significantly increased (Fig. 4,A). Deficiency of Rac1 or apocynin administration decreased mitochondrial superoxide production in the hyperglycemic heart (Fig. 4 A). Addition of succinate did not elicit superoxide production in mitochondrial preparations (data not shown). This suggests that mitochondrial complex I and/or III but not complex II is involved in superoxide generation.

To further confirm the role of Rac1 and NADPH oxidase in mitochondrial ROS production, we analyzed the release of ROS from mitochondria by using DCF-DA as an indicator. Similarly, ROS production was increased in intact mitochondria from the hyperglycemic compared with the normoglycemic heart on addition of pyruvate/malate (5/5 mmol/l). Deletion of Rac1 or apocynin administration blocked the increase of ROS production in mitochondria from the hyperglycemic heart (Fig. 4 B). Thus, these data demonstrate that Rac1 and NADPH oxidase contribute to mitochondrial ROS generation in the hyperglycemic heart.

To determine whether Rac1 activation directly contributes to apoptosis in cultured ARVC exposed to high glucose, we used adenoviral vector Ad-RacN17 expressing a dominant negative mutant of Rac1, which specifically blocks Rac1 activation (22,34). ARVC were infected with Ad-RacN17 or Ad-gal and then incubated with normal (5.5 mmol/l) or high glucose (33 mmol/l) for 24 h. High glucose significantly induced NADPH oxidase activity in ARVC (Fig. 5,A). In contrast, equal amount of mannitol did not affect NADPH oxidase activity (data not shown). Infection of Ad-RacN17 blocked NADPH oxidase activity in high glucose–treated ARVC (Fig. 5,A). Similarly, high glucose significantly increased caspase-3 activity compared with normal glucose, whereas, as an osmotic control, equal amount of mannitol did not induce caspase-3 activity in ARVC (data not shown). Ad-RacN17 blocked high glucose–induced caspase-3 activity (Fig. 5,B). In association with upregulation of caspase-3 activity, high glucose increased the percentage of annexin V positive cells, which was significantly reduced by infection of Ad-RacN17 (Fig. 5,C and E). The increased percentage of annexin V positive cells was also significantly reduced by a selective caspase-3 inhibitor, AC-DEVD carbohydrate (up to 90% reduction). These results suggest that Rac1 is important in high glucose–induced apoptosis. Cardiomyocyte death was also measured by phosphatidylinositol nuclear staining. Compared with normal glucose, high glucose increased the percentage of phosphatidylinositol positive cells, which was significantly decreased by Ad-RacN17 (Fig. 5 D and F). Although phosphatidylinositol is a viability exclusion dye and does not enter cells with intact membrane, in the late stage of apoptosis there is also a loss of membrane integrity as the cell dies (35). As such, the population of phosphatidylinositol-positive cells is a mixture of necrotic and apoptotic cells.

To examine if NADPH oxidase activation mediated apoptosis, we first incubated ARVC with normal or high glucose in the presence of apocynin (100 μmol/l), NAC (2 mmol/l), or vehicle. Twenty-four hours later, coincubation with apocynin or NAC inhibited caspase-3 activity and decreased the percentage of annexin-V positive and phosphatidylinositol-positive cells in high glucose–induced ARVC (Fig. 6,A–C). To clarify the contribution of gp91^phox-NADPH oxidase, we transfected ARVC with gp91^phox-specific siRNA and then incubated these cells with normal or high glucose for another 24 h. Two different scramble siRNAs (Con1 and Con2) were used as controls. Transfection with gp91^phox siRNA significantly decreased gp91^phox protein compared with either Con1 or Con2 siRNA in ARVC (Fig. 6,D). To confirm the specific knockdown by gp91^phox siRNA, we also measured Nox4 expression because our recent study has shown that cardiomyocytes express both Nox2 and Nox4 under physiological conditions (25). Gp91^phox siRNA did not alter Nox4 expression in ARVC (data not shown). In response to high glucose, gp91^phox siRNA blocked caspase-3 activation and reduced annexin V–staining positive cells in ARVC (Fig. 6,E and F), compared with either Con1 or Con2 scramble siRNA. These results demonstrate that gp91^phox-NADPH oxidase activation induces apoptosis in high glucose–stimulated ARVC. The role of NADPH oxidase in high glucose–induced apoptosis was further demonstrated by using p47^phox siRNA. Transfection of p47^phox siRNA blocked NADPH oxidase activation in high glucose–stimulated ARVC (data not shown). Similarly, caspase-3 activation and annexin V–staining positive cells were significantly decreased in p47^phox siRNA–transfected ARVC (Fig. 6 G and H).

To investigate the functional significance of Rac1, we assessed myocardial function in Rac1-ko and wild-type mice after 8 weeks of STZ injection. Although there was no change in heart rate, rate of contraction/relaxation was significantly reduced in diabetic wild-type mice compared with sham animals (Fig. 7). Lack of Rac1 restored the rate of contraction and relaxation without affecting heart rate in diabetic Rac1-ko mice (Fig. 7). These results demonstrate that Rac1 contributes to myocardial dysfunction in diabetic mice.

To further demonstrate the role of Rac1 in cardiac apoptosis in diabetes, we used type 2 diabetic model of db/db mice. A highly selective Rac1 inhibitor NSC23766 (36) was given to db/db mice starting at age of 12 weeks (2.5 mg · kg⁻¹ day⁻¹ i.p.) for 2 weeks. Rac1 activity, NADPH oxidase activity, caspase-3 activity, and TUNEL-positive cells were significantly increased in the heart of db/db mice compared with wild-type mice (Fig. 8,A–D). Administration of NSC23766 inhibited NADPH oxidase activation, blocked caspase-3 activity, and reduced TUNEL-positive cells in db/db mice (Fig. 8B–D), whereas blood glucose levels were similar between NSC23766 and vehicle-treated db/db mice (20–28 mmol/l). Finally, myocardial dysfunction in db/db mice was slightly but not significantly (P = 0.1418 and 0.0648 for the rate of contraction and relaxation, respectively) attenuated by NSC23766 treatment (Fig. 8 E and F).

The present study generated cardiomyocyte-specific Rac1-ko mice and investigated the role of Rac1 in ROS production and apoptosis in the hyperglycemic hearts. We demonstrated for the first time that Rac1, through gp91^phox-NADPH oxidase activation and ROS production, induces apoptosis in both hyperglycemic hearts and high glucose–stimulated cardiomyocytes. Rac1 also plays a role in cardiac apoptosis in type 2 diabetic db/db mice. Furthermore, Rac1 and NADPH oxidase activation is associated with mitochondrial ROS production in the hyperglycemic hearts. Finally, Rac1 contributes to myocardial dysfunction in STZ-induced type 1 diabetes. Thus, Rac1/NADPH oxidase/mitochondrion axis is important in ROS production and cardiomyocyte apoptosis in hyperglycemia, which may contribute to the development of diabetic cardiomyopathy.

Accumulating evidence suggests that hyperglycemia induces ROS (6,11) and overproduction of ROS is associated with apoptosis in the diabetic heart (4,6). However, the pathophysiologically relevant source of increased myocardial ROS production in diabetes remains to be further characterized. Recent studies have demonstrated that myocardial NADPH oxidase activity is increased in diabetes (16,17). Hyperglycemia has been shown to activate NADPH oxidase in cardiomyocytes (18). These studies suggest the pathophysiological relevance of NADPH oxidase activation for cardiomyocyte apoptosis in hyperglycemia. An important step for the assembly of active NADPH oxidase is the interaction between cytosolic subunits (p67^phox, p47^phox, and Rac1) and membrane subunits (gp91^phox and p22^phox) (19). Rac1 is required to anchor cytosolic p67^phox to the membrane for the assembly of active NADPH oxidase, leading to superoxide generation (21). In the present study, deficiency of Rac1 inhibited the relocalization of NADPH oxidase subunits to membrane and diminished NADPH oxidase activation and ROS production in the hyperglycemic heart. Either Rac1 deletion or inhibition of NADPH oxidase prevented myocardial apoptosis in hyperglycemia. These findings suggest that Rac1 contributes to apoptosis through NADPH oxidase activation in the hyperglycemic heart. To characterize whether the role of Rac1 in apoptosis could be reproduced by high glucose levels, we extended our analyses to cardiomyocytes. Direct exposure of RAVC to high glucose, through Rac1, promoted NADPH oxidase activation and induced apoptosis. Selective inhibition of Rac1, knockdown of gp91^phox or p47^phox, or scavenging ROS prevented apoptosis in high glucose–stimulated ARVC. These data support the notion that Rac1 is critical for gp91^phox-NADPH oxidase activation, which is the major source of ROS production responsible for apoptosis in cardiomyocytes under hyperglycemic conditions. Rac1 was also reported to induce NADPH oxidase activation and ROS production in angiotensin-II stimulated cardiomyocytes (21). Thus, the results reported herein also suggest that Rac1 may be a universal mechanism for NADPH oxidase activation in the heart under various stresses, which may have potential therapeutic implications.

Mitochondrion is another important source of ROS production contributing to apoptosis (37,38). Increased mitochondrial ROS generation has been shown in diabetic hearts (4,33,39). However, relatively few studies have directly measured mitochondrial ROS production in the isolated mitochondria from diabetic hearts, and the cross talk of other ROS sources to mitochondrion has not been fully elucidated. In the present study, we measured ROS production directly from the isolated mitochondria and analyzed the role of Rac1-NADPH oxidase in mitochondrial ROS production. In agreement with a recent report of Boudina et al. (39), our data showed that ROS production was increased in the isolated mitochondria from the hyperglycemic hearts. The increased mitochondrial ROS was achieved by using mitochondrial complex I substrates, NADH or pyruvate/malate, but not succinate, a complex II substrate; this is different from the result of Boudina et al. (39), which demonstrated that mitochondrial ROS is induced on addition of succinate in the isolated mitochondria from the diabetic heart. The cause of this discrepancy is unknown but may result from different diabetic models used. Interestingly, cardiomyocyte-specific deletion of Rac1 or inhibition of NADPH oxidase significantly attenuated ROS production in isolated mitochondria from the hyperglycemic heart, suggesting that Rac1 and NADPH oxidase activation functions upstream of mitochondrion in generating mitochondrial ROS in cardiomyocytes exposed to hyperglycemia. This finding is supported by a recent observation demonstrating that NADPH oxidase–derived superoxide anion (O₂^-) directly activates the mitoK_ATP channels, presumably through a direct action on the sulfhydryl groups of this channel, or reacts with nitric oxide to form peroxynitrite in vascular endothelial cells in response to angiotensin-II (40). Both the increase in K⁺ influx through opening of the mitoK_ATP channels and the peroxynitrite formation can damage respiratory complexes, leading to mitochondrial ROS production (41). Alternatively, one-electron transfer and/or rapid dismutation of superoxide result in substantial H₂O₂ generation. H₂O₂ is a quantitatively important signaling species (42), which may mediate mitochondrial ROS production. Indeed, direct addition of H₂O₂ to isolated mitochondria significantly increased superoxide generation (E.S. and T.P., unpublished data). Whether these mechanisms operate in the diabetic heart is currently unknown but merits further investigations. On the other hand, H₂O₂ converted from superoxide in the mitochondria can diffuse to the cytoplasm, where it could activate NADPH oxidase producing ROS. Recent studies have suggested that mitochondrial ROS activate PI3K and PKC. Activation of PI3K in turn promotes Rac1 and subsequent NADPH oxidase activation (43,44), and PKC induces NADPH oxidase activation via p47^phox phosphorylation (45). In the heart of type 1 diabetic mice, scavenging ROS inhibits NADPH oxidase expression and activation (46). Thus, a mutual functional relationship between NADPH oxidase and mitochondria in a feed-forward fashion may favor sustaining ROS production, which promotes apoptosis in the hyperglycemic heart. Future studies will be required to investigate this functional cross talk between Rac1-NADPH oxidase and mitochondrion.

The present study also found that deficiency of Rac1 attenuated myocardial dysfunction in diabetic mice, suggesting a pathophysiological significance of Rac1 activation in diabetic hearts. This protection may result from inhibition of cardiomyocyte apoptosis by Rac1 blockade. Rac1-mediated ROS has been shown to induce cardiac hypertrophy in response to angiotensin-II stimulation (21). A recent study also indicated that Rac1 may be associated with cardiac inflammatory responses in hyperglycemia (47). Both cardiac hypertrophy and inflammation are important in the development of cardiac complications of diabetes. Thus, it is possible that Rac1 via NADPH oxidase activation and ROS production contributes to the diabetic cardiac hypertrophy, inflammatory responses, mitochondrial dysfunction, in addition to apoptosis, leading to myocardial dysfunction in diabetes (Fig. 9), which requires further investigations. However, Rac1 may also play a role in cardiac hypertrophy in diabetes through NADPH oxidase–independent pathways (48).

It is worthwhile to mention that Rac1 activity was decreased by about 40% in the whole heart of Rac1-ko mice compared with their wild-type littermates. The exact deletion of Rac1 in cardiomyocytes from Rac1-ko mice is currently unknown. A recent study showed that the Cre-loxP system could achieve ∼70% deletion of Rac1 in cardiomyocytes in transgenic mice (21). Thus, the deletion of Rac1 in cardiomyocytes from our Rac1-ko mice is estimated to be around 40–70%, which may lead to some limitations of the interpretation of results from the present study. For example, it is unclear whether further down- regulation of Rac1 provides a better protection in diabetic hearts. However, complete deletion of Rac1 may have adverse effects since recent studies have suggested that Rac1 is also required for cardiovascular development and physiological functions (49,50).

In conclusion, Rac1 plays a critical role in hyperglycemia-induced apoptosis in cardiomyocytes through NADPH oxidase activation and ROS production, leading to myocardial dysfunction in diabetes. Furthermore, Rac1 and NADPH oxidase activation contributes to mitochondrial ROS production in the hyperglycemic heart. Given excessive ROS and cardiac apoptosis significantly contributes to diabetic cardiomyopathy (1,–3,6,32,33,51), these findings suggest that myocardial Rac1 may serve as a potential target for the treatment of cardiac complications in diabetes.

This study was supported by grants from the Heart & Stroke Foundation of Ontario (NA 5940) and the Canadian Institutes of Health Research (MOP 93657) awarded to T.P. T.P. is a recipient of a New Investigator Award from Heart & Stroke Foundation of Canada and Rick Gallop Award for research excellence from Heart & Stroke Foundation of Ontario (2006).

No potential conflicts of interest relevant to this article were reported.