INTRODUCTION

Top Abstract Introduction Materials Results Discussion References

SINCE THE INITIAL observation in which the potent cardioprotective effect of ischemic preconditioning (IPC) was demonstrated by Murry et al. (15), a number of studies have focused on the signaling mechanisms involved. G_i/o proteins have been implicated in the cardioprotective effect of IPC in a few species (8, 11, 14, 16, 17, 19). Thornton et al. (19) used an in vivo rabbit model and Miura et al. (14) used a canine model to study the involvement of G proteins in the protective effects of IPC; isolated heart or cardiac membrane preparations were used in most other studies (8, 11, 16, 17). Furthermore, Liu and Downey (12) were unable to demonstrate a role of G_i/o proteins in the cardioprotective effect of IPC in an intact rat model of myocardial infarction. Therefore, we decided to reevaluate the involvement of G_i/o proteins in IPC in the intact rat model of myocardial infarction. A relatively low dose (10 µg/kg) of pertussis toxin (PTX) was used, and its effectiveness in blocking G_i/o proteins was assessed by determining its ability to block the bradycardic response to ACh and adenosine (Ado) in vivo and by its ability to ADP-ribosylate G proteins in vivo.

	METHODS AND MATERIALS

Top Abstract Introduction Materials Results Discussion References

In Vivo Studies

General surgical procedure. Male Wistar rats weighing 350-450 g were anesthetized by administration of thiobutabarbital (Inactin, 100 mg/kg ip), a long-acting barbiturate. A tracheotomy was performed, and the rat was intubated with a cannula that was connected to a rodent ventilator (model 683, Harvard Apparatus). The rats were ventilated with room air at 65-70 breaths/min. Atelectasis was prevented by maintaining a positive end-expiratory pressure of 1-2 cmH₂O with a trap. Arterial pH, PO₂, and PCO₂ were monitored at specific intervals by a blood gas system (model AVL 995) and were maintained within a normal physiological range (pH 7.35-7.45, PO₂ 80-120 mmHg, PCO₂ 25-40 mmHg) by adjustment of the respiratory rate and/or the tidal volume. Normal rat body temperature was monitored (Yellow Springs Instrument Tele-Thermometer) and maintained at 37 ± 1°C by using a heating pad.

Drugs. Inactin, a thiobutabarbital sodium salt, and PTX were purchased from Research Biochemicals International (Natick, MA). 2,3,5-Triphenyltetrazolium chloride (TTC), Ado, and ACh were purchased from Sigma Chemical (St. Louis, MO). Inactin was dissolved in distilled water. Each 50-µg vial of PTX was reconstituted with 500 µl of sterile distilled water. TTC was dissolved in 100 mM phosphate buffer (pH 7.4). Ado and ACh were dissolved in 0.9% saline.

Experimental protocol. The control group (group I) consisted of rats subjected to 30 min of occlusion and 2 h of reperfusion (Fig. 1). Group II was subjected to the 1× 5 min IPC protocol, which was elicited by one 5-min occlusion period followed by 10 min of reperfusion before the 30-min occlusion and 2-h reperfusion period (1× PC). Group III was subjected to the multiple IPC protocol, which was elicited by three 5-min occlusion periods interspersed with 5 min of reperfusion before the 30-min occlusion and 2-h reperfusion period (3× PC). Rats in groups IV, V, and VI were pretreated with PTX (10 µg/kg ip) for 48 h before the 30 min of occlusion, 1× 5 min PC, or 3× 5 min PC, respectively. The 10 µg/kg dose of PTX was based on the protocol utilized by Endoh et al. (4). The PTX-treated animals did not appear to be visibly sick; however, when a left thoracotomy was performed, a yellowish-to-clear fluid was observed in the chest cavity of some animals. In addition, to demonstrate that PTX inhibited G protein function, the changes in HR induced by ACh and Ado, G_i-mediated responses (7), were measured.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Experimental protocol depicting study to test role of Gi/o proteins in cardioprotective effect of ischemic preconditioning (PC) in rat heart. PTX, pertussis toxin; Occ, occlusion; Rep, reperfusion.

Determination of infarct size. After each experiment, the left coronary artery was reoccluded and patent blue dye was injected intravenously to stain the normal area of the LV. The rat was killed with 15% KCl. The heart was excised, and the LV was removed and sliced into five cross-sectional pieces. This procedure allowed for visualization of the nonischemic area and area at risk (AAR). The AAR was separated from the nonischemic area under a dissecting microscope (Cambridge Instruments). Both tissue areas (nonischemic and AAR) were incubated for 15 min with 1% TTC in 100 mM phosphate buffer (pH 7.4) at 37°C. TTC stain is an indicator of viable and nonviable tissue (10). TTC is reduced by dehydrogenase enzymes, which are present in viable myocardium, resulting in a formazan precipitate, thereby turning the tissue a deep red color. Because the nonviable, infarcted myocardium does not retain the dehydrogenases, the tissue remains a pale gray color. The tissue was stored overnight in 10% formaldehyde. With use of the dissecting microscope, the infarct size (IS) and AAR were determined by gravimetry. IS was expressed as a percentage of the AAR (IS/AAR) and as a percentage of the total LV weight (IS/LV).

Exclusion criteria. Animals were omitted from further data analysis if 1) blood gases, most notably pH, indicated metabolic acidosis and sodium bicarbonate administration was unsuccessful in raising the pH and bicarbonate levels to the normal physiological range; 2) marked hypotension [mean arterial blood pressure (MBP) <30 mmHg] was observed; or 3) intractable ventricular fibrillation occurred in which the experiment could not be continued successfully for the duration of the protocol.

Statistical analysis of data. Values are means ± SE. Differences between groups in hemodynamics at various time points were compared using a two-way ANOVA for time and treatment with repeated measures. If significant F ratios were obtained, Fisher's least significant difference test was performed. A one-way factorial ANOVA with a Fisher's post hoc test was used to determine differences among groups for LV, AAR, and IS weights and IS/AAR. Statistical differences were considered significant if P < 0.05.

In Vitro Analysis

Preparation of homogenate from rat heart. Hearts from non-PTX-treated (control) and PTX-treated (10 µg/kg ip for 48 h) rats were removed from the thoracic cavity quickly and immediately frozen with liquid nitrogen. The hearts were weighed, and the appropriate volume of sucrose buffer was added [1(wt of heart):4 (vol of sucrose buffer)]. Sucrose buffer (pH 7.4) consisted of 20 mM HEPES, 1 mM EDTA, 255 mM sucrose, and 100 µM phenylmethylsulfonyl fluoride. Sucrose buffer was added to heart tissue in a glass homogenizer and placed in ice. The ice-chilled heart tissue was homogenized with an electrical homogenizer (Glas-Col). The homogenate was centrifuged at 20,000 rpm for 20 min at 4°C (model J2-MI centrifuge and model JA-20.1 rotor, Beckman). The supernatant was collected and retained for determination of protein concentration via a spectrophotometer (model SD-1, Spectrodonic). Protein concentration of the rat heart homogenate was determined using the following calculation where sample is supernatant from heart homogenate, BSA is bovine serum albumin, SD is absorbance value, and 5 is the microliter volume of sample. The homogenate (supernatant fraction) was aliquoted and stored at 80°C until further use.

Urea-gradient SDS-PAGE. Electrophoresis was performed using a slab system (0.75 mm × 8.3 cm × 10.2 cm, Hoefer Scientific Instruments) with Laemmli running buffer. The resolving gel (5-6 cm) consisted of 12% acrylamide, 5× buffer A [Tris base (2.59 M, pH 8.9), 1% SDS, 0.5 ml N,N,N',N'-tetramethylethylenediamine], 10% ammonia persulfate, and 0.1% bromphenol blue in a linear gradient of 4.8 (top) to 7.3 (bottom) M urea. The stacking gel (2 cm) consisted of 6% acrylamide, stacking buffer [625 mM Tris · HCl (pH 6.9) and 1% SDS], 10% ammonia persulfate, and 36 µl of N,N,N',N'-tetramethylethylenediamine.

PTX-induced ADP-ribosylation. To demonstrate the level of PTX-induced ADP-ribosylation of G_i/o proteins in vivo, a PTX assay was performed using [³²P]NAD. The PTX-induced ADP-ribosylation assay (final volume 25 µl) contained 0.1 M Tris · HCl buffer (pH 7.6), 20 mM dithiothreitol (fresh), 1.1 µM [³²P]NAD (specific activity 10 Ci/mmol), 0.1 mM ATP, rat heart homogenate (20 µg protein), and PTX (1.5 µl = 0.15 µg protein). The reaction mix was incubated for 1 h at room temperature and stopped with the addition of 12.5 µl of SDS sample buffer with -mercaptoethanol followed by boiling at 100°C for 5 min. Equal amounts of sample were subjected to SDS-PAGE with use of 12% acrylamide gel with a urea gradient (29-44%, 4.8-7.3 M; electrophoresed at 80 V for 25 min and 200 V for 60 min). The gel was fixed for 1 h, stained for 10 min using 0.1% Coomassie brilliant blue, and destained for 2-4 h. The gel was air dried overnight, and autoradiography was subsequently performed for 18 h.

	RESULTS

Top Abstract Introduction Materials Results Discussion References

Whole Animal Study

Exclusions. A total of 49 animals were assigned to this study. Two animals in the control group, two in the 3× 5 min PC group, and one in the 1× 5 min PC group were excluded because of intractable ventricular fibrillation. In our initial studies, six animals were pretreated with PTX (25 µg/kg ip) 48 h before the 3× 5 min PC protocol, and these rats died because of ventricular fibrillation (n = 4) or marked hypotension (n = 2). Because of the difficulty in animal survival when this high dose of PTX was used, a 10 µg/kg dose was chosen on the basis of previous studies of Endoh et al. (4) for all subsequent experiments. A total of 38 animals completed the study.

Hemodynamics. Hemodynamic parameters including HR, MBP, and rate-pressure product (RPP) are summarized in Table 1. There were no significant differences in HR in the groups measured at any of the time points. MBP was significantly lower in the PTX control and PTX-treated IPC groups (PTX + 1× PC and PTX + 3× PC) at baseline and 2 h of reperfusion. The 1× PC group had a significantly lower MBP at 2 h of reperfusion. Similarly, MBP was significantly lower in the PTX and PTX + 3× PC groups at 30 min of occlusion. Similarly, RPP was significantly lower at baseline as well as at 2 h of reperfusion in the PTX + 3× PC group. Furthermore, RPP was significantly lower in the PTX and PTX + 1× PC groups at 2 h of reperfusion.

IS and AAR. Table 2 shows the weights of the LV and ischemic area (AAR), IS, and IS/AAR. The LV weights were not significantly different among the six groups; however, AAR were significantly larger in the PTX and PTX + 3× PC groups than in the control group. IS was significantly smaller in the 1× PC and 3× PC groups. Furthermore, IS was significantly larger in the nonpreconditioned, PTX-treated group; however, IS/AAR for this group was not statistically different from the control group. Figure 2 depicts IS/AAR for the individual rat hearts and the means for each group. Figure 2 demonstrates that the mean IS/AAR for the control group was 46.2 ± 5.6%. One 5-min ischemic insult followed by 10 min of reperfusion or three 5-min occlusion periods interspersed with 5 min of reperfusion markedly reduced IS/AAR to 6.4 ± 1.1 and 8.2 ± 1.2%, respectively (P < 0.05 vs. control). Pretreatment with PTX (10 µg/kg ip) for 48 h completely abolished the cardioprotective effect of single or multiple, brief ischemic periods (45.5 ± 5.4 and 38.3 ± 3.9%, respectively; Fig. 2). PTX alone had no effect on infarct size (53.3 ± 9.3%).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Infarct sizes (IS) in rat hearts subjected to control (Con, 30 min of ischemia and 2 h of reperfusion); ischemic preconditioning elicited by one or three 5-min ischemic periods interspersed with 5 min of reperfusion (1 × PC or 3 × PC); PTX (10 µg/kg ip), an inhibitor of Gi/o proteins, administered 48 h before 30 min of ischemia (PTX); and PTX given 48 h before 1 × PC or 3 × PC (PTX + 1 × PC and PTX + 3 × PC). , IS of individual hearts; , group mean IS (mean ± SE). AAR, area at risk. * P < 0.05 vs. Con.

G_i/o protein-mediated bradycardia. To determine whether the dose of PTX (10 µg/kg ip) was eliciting a pharmacological response in these animals, changes in HR induced by ACh and Ado were measured (Fig. 3). It was previously demonstrated that the bradycardic effects of ACh and Ado are mediated via PTX-sensitive G proteins (7). In four control rats, administration of ACh (0.15 mg/kg iv) and Ado (1 mg/kg iv) produced marked decreases in HR from 425 ± 56 to 265 ± 26 and from 428 ± 31 to 248 ± 33 beats/min, respectively (P < 0.05 vs. control). These responses to ACh and Ado were completely abolished in six PTX-treated rats: 456 ± 12, 429 ± 8, and 431 ± 10 beats/min in PTX control, PTX + ACh, and PTX + Ado, repsectively. These data indicated that administration of PTX at 10 µg/kg was sufficient to uncouple G protein-mediated signal transduction in this animal model.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Bradycardic effects of ACh and adenosine (Ado) are mediated via PTX-sensitive G proteins. Administration of ACh (0.15 mg/kg iv) and Ado (1 mg/kg iv) produced marked decreases in heart rate (HR;  and , respectively) * P < 0.05. Change in HR produced by ACh and Ado was completely abolished in PTX-treated rats ( and , respectively).

G_i/o Protein Assay

ADP-ribosylation. Experiments were performed to determine the extent of G protein ADP-ribosylation by PTX in vivo. The basis for the assay is that G proteins that have been ADP-ribosylated in vivo will not be available for subsequent ADP-ribosylation in vitro. Figure 4 represents an autoradiogram of the ADP-ribosylation via PTX of G_i/o proteins in rat heart. Coomassie blue staining of the Western blot demonstrated that comparable amounts of total heart tissue protein (20 µg) were electrophoresed in each lane. Transducin, a PTX-sensitive G_t protein located in the retina, was ADP-ribosylated via PTX (Fig. 4, lane 5), demonstrating that the assay contained the appropriate experimental conditions. G proteins in the control (non-PTX-treated) rat hearts were ADP-ribosylated in vitro by PTX (Fig. 4, lane 3), and the PTX-treated rat hearts showed the same level of availability for in vitro ADP-ribosylation by PTX (Fig. 4, lane 1). This indicated that PTX treatment did not ADP-ribosylate the bulk of the total G protein in the hearts. This is consistent with a model where PTX has ADP-ribosylated only a subpopulation of G proteins in the myocardial membrane rather than the entire population of G proteins in the heart. In Fig. 4 the radiolabeled band that migrated faster than the ADP-ribosylated 41-kDa G_i represents the activity of an endogenous eukaryotic enzyme radiolabeling another eukaryotic protein. Because the labeling appears in the absence of added PTX, either in vivo or in vitro, the activity is a constitutive activity of the extract and is not related to the study.

View larger version (68K): [in this window] [in a new window]   Fig. 4.   Autoradiogram of PTX-catalyzed incorporation of [32P]ADP-ribose from [32P]NAD into unreacted Gi/o proteins in rat heart homogenate. Substrate for PTX had a relative molecular weight of 39,000-41,000 (Gi/o). Lane 1, PTX-treated hearts in vivo + PTX in vitro; lane 2, PTX-treated hearts in vivo + no PTX in in vitro assay; lane 3, non-PTX-treated hearts in vivo + PTX in in vitro assay; lane 4, non-PTX-treated hearts in vivo + no PTX in in vitro assay; lane 5, transducin (TND) + PTX in in vitro assay; lane 6, TND + no PTX in in vitro assay. TND is a positive control for ADP-ribosylation.

	DISCUSSION

Top Abstract Introduction Materials Results Discussion References

It has been previously demonstrated that G_i/o proteins are involved in IPC in a number of species including the rat (8, 11, 14, 16, 17, 19). However, only Thornton et al. (19), using an in vivo rabbit model, demonstrated that PTX completely abolished the IPC-mediated reduction in infarct size. In addition, Miura et al. (14) used an in vivo canine model and showed that PTX abolished the effect of IPC on myocardial acidosis. Conversely, all other studies used isolated heart or cardiac membrane preparations (8, 11, 16, 17) to demonstrate an involvement of G_i/o proteins in the protective effects of IPC. The results of the present study agree with those of studies previously performed in isolated rat heart experiments and support a role for G proteins in the cardioprotective effect of IPC in the in vivo rat heart model.

PTX has been shown to ADP-ribosylate the -subunit of G_i/o proteins, thereby interfering with signal transduction between the G protein and its G protein-coupled receptor, which causes a complete loss of G_i/o protein function and an abolishment of the receptor-stimulated "inhibitory" activity (1, 3, 20). Pretreatment with PTX (10 µg/kg ip) for 48 h before single or multiple 5-min ischemic insults to induce PC completely abolished the cardioprotective effect: 45.5 ± 5.4% vs. 1× PC of 6.4 ± 1.1 and 38.3 ± 3.9% vs. 3× PC of 8.2 ± 1.2% (Fig. 2). The dose of PTX used in the present study was based on the protocol of Endoh et al. (4), in which PTX attenuated the inhibitory effects of atrial muscarinic receptor activity in a dose-dependent manner at 0.125-1.0 µg/100 g body wt in Wistar rats. Thornton et al. (19) showed a similar inhibition by PTX (48 h of pretreatment at 25 µg/kg iv) of IPC in the in vivo rabbit heart: 27.3 ± 4.3, 37.4 ± 4.6, and 5.2 ± 2.1% for PTX-treated, control, and PC IS/AAR, respectively. Similarly, in the isolated rat heart the enhanced recovery of function induced by IPC was blocked by PTX (8). In addition, in the isolated rat heart, IPC has been demonstrated to enhance the concentration of G_i proteins with a decrease in G_s proteins (17) or an increase the responsiveness of G_i proteins in canine sarcolemmal membranes (16). However, Liu and Downey (12) were unable to demonstrate a role for G_i/o proteins in the cardioprotective effect of IPC to reduce infarct size in the intact rat heart. Using a 3× 5-min IPC protocol, they showed that PTX could not block its cardioprotective effect even though the bradycardia produced by Ado and ACh, which produce their responses via G_i-coupled Ado and muscarinic receptors, respectively, was blocked (12). In the present study we observed that the reduction in IS/AAR, produced by the single or multiple IPC protocol, was abolished by PTX in the intact rat heart, indicating that G_i/o proteins were involved (Fig. 2). We also observed, like Liu and Downey, that the bradycardic responses elicited by ACh (0.15 mg/kg iv) and Ado (1 mg/kg iv) were blocked by 48 h of pretreatment with PTX (10 µg/kg ip; Fig. 3). It is difficult to reconcile the opposing results found in the present study and those of Liu and Downey. The major differences between the studies are the strain of rat, the dose of PTX, and the magnitude of the protective effects observed with PC. Whether any of these factors are responsible for the marked differences between these two studies remains to be resolved. Hu and Nattel (8) found that 25 µg/kg of PTX blocked PC in the isolated rat heart, and we recently showed that 10 µg/kg of PTX abolished the cardioprotective effect of a -opioid agonist, TAN-67, which is thought to act via G_i/o (18). All these results together strongly support a role for G_i/o in mediating IPC in the rat heart.

Recently, our laboratory showed that cardioprotection induced by opioid receptors (₁-receptors) was mediated by a PTX-sensitive mechanism in the rat heart (18). We showed that the cardioprotective effect of TAN-67, a selective nonpeptidic ₁-opioid receptor agonist, significantly reduced IS/AAR [27 ± 5 vs. 56 ± 2% (control), P < 0.05] and 48 h of pretreatment with PTX (10 µg/kg ip) completely abolished the protective effect of the ₁-agonist (61 ± 4%). It is known that opioid receptors belong to the family of G protein-coupled receptors (2, 6, 9, 13, 21). In addition, a number of other receptor systems, such as adenosine (A₁) and muscarinic (M₂), which are coupled to G_i/o proteins, have been implicated in the cardioprotective effect of IPC (11, 19). Not only are the adenosine A₁ and muscarinic M₂ receptors involved in cardioprotection, they also elicit a bradycardic response that is mediated via G_i/o proteins. We showed that the decrease in HR produced by Ado or ACh was abolished in PTX-treated animals (Fig. 3).

ADP-ribosylation of G_i/o proteins by PTX uncouples G protein-mediated signal transduction (1, 3, 20). In the present study we demonstrated that 48 h of pretreatment with PTX abolished the cardioprotective effect of IPC as well as the Ado- or ACh-mediated bradycardia. Because the dose of PTX (10 µg/kg ip) used in these experiments showed a pharmacological and/or functional change, we sought to determine the extent of PTX-mediated ADP-ribosylation of G_i/o proteins in vivo. Figure 4 shows the results of ADP-ribosylation via PTX. Lane 3 contained protein from non-PTX-treated hearts and demonstrated that PTX added to the assay mediated the ADP-ribosylation of substrates with relative molecular weights of 39,000 (G_o) to 41,000 (G_i). Lane 1 contained protein from PTX-treated hearts and showed that a band still was present after PTX was added to the assay. This demonstrates that the bulk of the G protein in the PTX-treated heart had not been ADP-ribosylated in vivo. In our model, pretreatment with PTX ADP-ribosylates a small pool of G_i/o proteins in vivo below the level of detection with autoradiography (although these G_i/o proteins are inhibited functionally) and there is a larger population of G_i/o proteins that were not ADP-ribosylated by PTX in vivo. Fleming and colleagues (7) stated that PTX-mediated ADP-ribosylation of G subunits measures a subpopulation or fraction of the total concentration of G_i/o proteins in tissue. Furthermore, it has been suggested that the discrepancy of G_i/o protein density may be due to the fact that some of the G protein is inaccessible for ADP-ribosylation in vivo and that treatment of tissue homogenates with detergents may expose more sites for ADP-ribosylation in vitro (7). In addition, the subtypes of G_i and G_o proteins were not resolved in this study, and further separation of these G protein subtypes may demonstrate a change in ADP-ribosylation. Finally, our dose (10 µg/kg) of PTX used in vivo was three times lower than that used in the study by Endoh and colleagues (5). This group showed that pretreatment of hearts with 30 µg/kg of PTX inhibited all the ADP-ribosylation sites for PTX in cardiac membrane preparations in vitro (5). In the present study it appears that significant functional and physiological changes were elicited by in vivo administration of PTX, demonstrating a role of G_i/o proteins in the cardioprotective effect of IPC and that this occurred by in vivo ADP-ribosylation of a small subpopulation of G proteins. Alternatively, it is possible that only a small portion of G_i/o proteins that could not be resolved by the in vitro assay was inactivated by PTX.

In conclusion, we have demonstrated a role for G_i/o proteins in IPC in the intact rat heart. The results of this study provide further understanding of the mechanism underlying the cardioprotective effect of IPC. In addition, our data are consistent with the ability of PTX to ADP-ribosylate a small population of G_i/o proteins to induce significant physiological effects that can result in a marked decrease in the IS reduction normally observed in IPC hearts. Further studies are needed to elucidate the involvement of the specific G_i/o protein subtype involved in mediating this potent cardioprotective effect.