CB₁ cannabinoid receptor antagonism promotes remodeling and cannabinoid treatment prevents endothelial dysfunction and hypotension in rats with myocardial infarction

Animals

Adult female Wistar rats were used. Coronary artery ligation was performed in a standardized model as previously described (Fletcher et al., 1981). Sham-operated rats underwent the same surgical procedure except coronary ligation.

Pharmacological interventions

In a dose-finding study, blood pressure was measured under ether anesthesia via a polyethylene catheter inserted into the right carotid artery and connected to a microtip manometer (Millar). The CB₁ antagonist AM-251 (N-(piperidin-1-yl)-5-(4-iodophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide) (Gatley et al., 1996), 0.5 mg kg⁻¹ d⁻¹ i.p. (but not 0.25 mg kg⁻¹ d⁻¹), prevented the peak hypotensive effect of the endogenous cannabinoid anandamide, given 4 mg kg⁻¹ intravenously (−5±10 mmHg vs −45±1 mmHg, n=3, P<0.01). AM-251 alone had no significant effects on heart rate and blood pressure.

There is no published information about the effects of chronic (>2 weeks) CB₁ stimulation on cardiovascular parameters. We used HU-210 in a dose of 50 μg kg⁻¹/d⁻¹ i.p., 25 and 0.5 times the i.v. ED₅₀ for its acute hypotensive and bradycardic effects, respectively (Lake et al., 1997).

Rats were housed in polyethylene cages with a 12-h light–dark cycle, fed with standard chow and water ad libitum. The three treatment groups were: (a) AM-251, 0.5 mg kg⁻¹ d⁻¹, in vehicle (1 vol ethanol:1 vol emulphor:18 vol saline), in 0.5 ml i.p.; (b) HU-210, 50 μg kg⁻¹ d⁻¹, in 1:1:18 ethanol/emulphor/saline vehicle, in 0.5 ml i.p.; and c) control rats, who received 0.5 ml vehicle (1:1:18 ethanol/emulphor/saline) i.p. every day. Treatment started 24 h after surgery and was given for 12 weeks and stopped at least 24 h prior to the hemodynamic measurements.

According to the extent of histologic infarct size, three additional subgroups in each treatment group were established: (a) rats were considered to be sham-operated if the necrotic area was <8% of the left ventricle, (b) small/moderate infarcts measured 40% and (c) large infarcts >40% of the left ventricle (Pfeffer et al., 1985).

Hemodynamic studies

At 12 weeks after coronary ligation, polyethylene cannulas were inserted into the right carotid artery for pressure measurements and into the trachea for artificial ventilation. Pressures were measured through a fluid-filled PE50 tubing connected to a microtipped manometer (Millar). The carotid cannula was briefly advanced into the left ventricle and then withdrawn to the aortic arch, while pressure was recorded. Left-ventricular systolic (LVSP) and end-diastolic pressures (LVEDP), maximal rate of rise of left-ventricular systolic pressure (dP/dt_max), mean arterial pressure (MAP), and heart rate (HR) were measured under light ether anesthesia and spontaneous breathing.

Later, during positive pressure ventilation and after midsternal thoracotomy following isoflurane anesthesia, a calibrated flow probe (2.5 mm, Gould Statham) was placed around the ascending aorta for continuous measurement of aortic blood flow. Mean aortic blood flow was obtained electronically and taken as the cardiac index (CI, ml min⁻¹ per kg body weight). Total peripheral resistance index (TPRI) was calculated as MAP/CI, stroke volume index (SVI) was calculated as CI/HR (μl per kg body weight).

The flow probe was removed and the arterial catheter was advanced into the left ventricle. The ascending aorta was occluded around the catheter by a suture to produce contractions that are isovolumetric except for coronary flow. Measurements were made of left-ventricular peak systolic and end-diastolic pressures. Maximal left-ventricular developed pressure was calculated as peak systolic minus end-diastolic pressure. These measurements defined the maximal pressure-generating ability of the left-ventricle (Fletcher et al., 1981).

Left-ventricular volume

The heart was arrested by potassium chloride and a double-lumen catheter (PE 50 inside PE 200) was inserted into the left-ventricle through the ascending aorta. The right-ventricular free wall was incised to avoid fluid accumulation. The atrioventricular groove was ligated, and isotonic saline was infused at a rate of 0.76 ml min⁻¹ through one lumen, while intraventricular pressure was continuously recorded via the other lumen from negative pressure to 30 mmHg. Three reproducible pressure–volume curves were obtained within 10 min of cardiac arrest, well before the onset of rigor mortis. Stiffness constants were defined as described (Fletcher et al., 1981). Ventricular volumes at pressures of 0, 2.5, 5, 10, 15, 20 and 30 mmHg were calculated from the pressure–volume curve (Fletcher et al., 1981). The operating left-ventricular end-diastolic volume was derived from the left-ventricular pressure–volume curve (Pfeffer et al., 1985) and was defined as the volume on the pressure–volume curve corresponding to a filling pressure equal to in vivo end-diastolic pressure.

Infarct size

As previously described (Pfeffer et al., 1985), the hearts were fixed in 10% buffered formalin for 24 h and then dissected into the left ventricle plus interventricular septum and right-ventricular free wall. Transverse serial sections of 10-μm thickness were obtained in 1-mm intervals from apex to base, stained with Sirius red so that necrotic tissue stained red. Infarct size was calculated for each section by dividing the sum of the planimetered necrotic endocardial and epicardial circumferences by the sum of the total epicardial and endocardial circumferences of the left ventricle using a calibrated digitizer (Numonics Digitizer 2200).

Vascular reactivity studies

The descending thoracic aorta was dissected following removal and cleaned. Rings of 3 mm were mounted in an organ bath (Föhr Medical Instruments, Seeheim, Germany) for isometric force measurement. The rings were equilibrated for 30 min under a resting tension of 2 × g in oxygenated (95% O₂; 5% CO₂) Krebs–Henseleit solution containing diclofenac (1 μmol l⁻¹). Rings were repeatedly contracted with KCl (100 m mol l⁻¹) until reproducible responses were obtained. Thereafter, the rings were preconstricted with phenylephrine (0.3–1 μmol l⁻¹) to 80% of maximum constriction level and the relaxant response to the cumulative application of acetylcholine and sodium nitroprusside (SNP) was assessed.

CB₁ receptor immunoactivity

For Western Blot analysis of CB₁, 100 mg of fresh frozen heart tissue obtained from two groups was tested: (i) viable left-ventricular myocardium without scar tissue of five untreated rats 12 weeks post-large MI (MI size: 49.5±2.7%) or (ii) myocardium of five untreated control rats. The tissue was Dounce-homogenized in radioimmunoprecipitation assay (RIPA) buffer (20 mmol l⁻¹ MOPS, 150 mmol l⁻¹ NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS, 1 mmol l⁻¹ EDTA, pH 7.0) containing the protease inhibitors phenylmethylsulfonyl fluoride, leupeptin, aprotinin (Roche, Basel), pepstatin, E-64 (trans-epoxysuccinyl-L-leucylamido-(4-guanidino)butane) (Roche, Basel), and p-aminobenzamidine. The homogenate was centrifuged (10 min, 14000 × g, 4°C) and the pellet was discarded. Aliquots of the supernatant were used to measure protein concentrations with a Bio-Rad protein assay (Bradford) according to the manufacturer's instructions (Bio-Rad Laboratories, München, Germany). The remainder was mixed with an aliquot of × 2 Laemmli SDS-sample buffer, boiled at 95°C for 10 min and used for Western Blot analysis as described previously (Laser et al., 2000). Antibodies for the CB₁ receptor (N-15) were obtained from Santa Cruz Biotechnology Inc., Santa Cruz, U.S.A. actin and GAPDH antibodies were from Chemicon, Temecula, CA, U.S.A. The chemiluminescence signal was digitally recorded with a 12-bit CCD-camera (ChemiImager 5500 from AlphaInnotech, San Leandro, CA, U.S.A.) using ChemiGlow reagent (Biozym, Hess. Oldendorf, Germany). Quantitation of the digital signal (average signal intensity) was performed with AlphaEase software from AlphaInnotec, San Leandro, CA, U.S.A. Unless otherwise stated, all materials were obtained from Sigma (Deisenhofen, Germany).

Data analysis

ANOVA was used in multiple comparisons to find differences between the various infarct and treatment groups were used, and corrected by the Bonferroni test. Differences in mortality among treatment groups were determined by χ² and Fisher's exact test. P<0.05 indicated statistical significance.

Drugs

AM-251 (N-(piperidin-1-yl)-5-(4-iodophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide) and HU-210 ({-}-11-OH-Δ⁹ tetrahydrocannabinol dimethylheptyl) were obtained from Biotrend. For vascular reactivity studies, phenylephrine, acetylcholine and SNP were obtained from Sigma (Germany). For Western Blot analysis, the reagents were obtained as stated above.

Mortality

The Kaplan–Meier curve of a total of 140 rats subjected to the surgical procedure did not show significant differences among the three treatment groups (Figure 1).

MI size, behavior, body and heart weights

Treatment with HU-210 or AM-251 did not change the extent of MI of any of the rats that underwent coronary artery ligation, as compared with respective controls (30.1±4.3% vs 36.4±3.1% vs 29.5±3.4%, respectively, P=0.31). MI sizes were similar among the treatment subgroups (Table 1).

Rats with infarcts treated with HU-210 gained significantly less weight than rats with infarcts treated with vehicle or AM-251 (Table 2). While AM-251 did not cause obvious side effects, HU-210-treated rats showed enhanced grooming with consequent hair loss. They were listless and anxious and some of them developed short-lasting seizures. Right-ventricular weight and lung weight increased in rats with large MI compared with sham rats throughout all treatment groups.

Hemodynamic measurements before thoracotomy

As shown in Table 3, left-ventricular systolic pressure decreased in untreated and AM-251-treated rats with large infarcts compared with sham-operated animals, but not in HU-210-treated rats. Compared to AM-251, HU-210 prevented the decline in dP/dt_max in rats with large infarcts. With growing infarct size, rats showed enhanced LVEDP in all treatment groups. HU-210 further increased LVEDP in the subgroup with large infarcts.

Hemodynamic measurements after thoracotomy

Table 4 shows that HU-210 increased MAP in rats with large MI. HU-210 improved CI and SVI and reduced TPRI in rats with small MI. Following aortic occlusion, the peak-developed pressure was decreased by chronic CB₁ antagonism in rats with large MI.

Left-ventricular volume

As expected, a rightward shift of the pressure–volume curves was found with increasing infarct size (Figure 2). Histologic left-ventricular diameter and cavity were increased (not shown). Importantly, in rats with large infarcts, AM-251 further shifted the passive left-ventricular pressure–volume curve to the right. Stiffness constants were not different among treatment groups (not shown). Left-ventricular operating volume increased with infarct size and was significantly higher in AM-251-pretreated rats (AM-251: 930±40 μl vs untreated: 820±40 μl vs HU-210: 790±50 μl; P<0.05).

Vascular reactivity

As indicated in Figure 3, in untreated or AM-251-treated rats with large MI, the endothelium-dependent vasodilator response of aortic rings to acetylcholine was reduced as compared with preparations from rats without MI (sham-operated). HU-210-treatment prevented endothelial dysfunction in rats with large MI. The relaxation to the endothelium-independent vasodilator SNP was unchanged in all treatment groups and independent from the existence of MI.

CB₁ receptor immunoactivity

Left-ventricular CB₁ receptor immunoactivity in rats with large MI was unaltered as compared with noninfarcted hearts (Figure 4). Statistical comparison with a Student's t-test (mean values±s.e.m) showed no differences in Western Blot signal intensity between control and infracted hearts for CB₁ (259±8 and 274±11, respectively) and GAPDH (1009±32 and 1005±11, respectively) as well as for actin (data not shown). Further, chronic CB₁ stimulation by daily intraperitoneally administered HU-210 (50 μg kg⁻¹) did not diminish the hypotensive and bradycardic response to intravenously administered HU-210 (10 μg kg⁻¹) as compared with non-pretreated controls challenged with HU-210 (−46±5 vs −49±4 mm Hg; −108±25 vs −85±19 beats min⁻¹, respectively, data obtained 30 min after HU-210 i.v., n=5), which rules out tolerance and receptor desensitization.

CB₁ receptor immunoactivity

Chronic heart failure stimulates several neurohumoral systems of which one might be endocannabinoids. CB₁ receptor stimulation did not alter CB₁ immunoactivity in the viable heart muscle tissue of rats with large MI (Figure 4). We report, for the first time, CB₁ receptor immunoactivity in the heart post-MI, which was not different from that of sham-operated animals. However, we measured cardiac CB₁ receptor expression 12 weeks post-MI, which leaves the possibility of transient CB₁ upregulation in the heart.

Study limitations

First, the dose of the CB₁ agonist HU-210 used in this study was high enough to cause severe side effects (Giuliani et al., 2000b). So far, HU-210 in a comparable dose had been used only for 1 week (Giuliani et al., 2000a). Second, if endocannabinoids are involved in cardioprotection in that model, one should expect clear opposite effects of control and AM-251-treated rats. CB₁ antagonism promoted remodeling (Figure 2) in rats with large infarcts and significantly reduced maximal developed pressure (Table 4). However, AM-251 treatment did not have a significant effect on other hemodynamic parameters improved by HU-210, such as LVSP, MAP, dP/dt_max, CI, and TPRI, which argues against a beneficial endocannabinoid ‘tone’ in this chronic model. However, the dose may have been too low. Third, peripheral CB₁ receptors mediate most of the cardiovascular effects of cannabinoids (Lake et al., 1997; Wagner et al, 1997; Varga et al., 1998; Wagner et al., 2001a,2001b), but not all. For example, some effects have been attributed to activation of CB₂ receptors (Lagneux & Lamontagne, 2001), Ca²⁺-activated potassium channels (White et al., 2001), an as-yet-unidentified ‘anandamide receptor’ (Jarai et al., 1999; Wagner et al., 1999), vanilloid receptors (Zygmunt et al., 1999), conversion to vasoactive prostanoids (Grainger & Boachie-Ansah, 2001) or effects on blood vessels were sensitive to atropine or β₂-adrenoceptor blockade (Gardiner et al., 2002). It is possible that HU-210 as a nonselective CB₁/CB₂ agonist may have achieved its effects by activating CB₂ receptors. However, it is unlikely, because HU-210 acutely elicits hypotension and bradycardia via CB₁ receptors (Lake et al., 1997). The use of CB₁ and CB₁/CB₂ knock-out mice (Jarai et al., 1999; Ledent et al., 1999) could further clarify the possible role of cannabinoid receptors in hemodynamic changes following MI.

In conclusion, the results of the present study show that cannabinoid receptor agonism prevents hypotension and endothelial dysfunction but increases LVEDP post-large MI. Although myocardial CB₁ expression remains unchanged, CB₁ receptor antagonism has negative consequences on cardiac remodeling.