Abstract
With developing congestive heart failure (CHF), activation of the vasopressin V1a and angiotensin II type 1 (AT1) receptors can occur. In the present study, we examined the direct effects of V1a receptor blockade (V1a block), selective AT1 receptor blockade (AT1 block), and dual V1a/AT1receptor blockade (dual block) with respect to left ventricular (LV) function and contractility during the progression of CHF. LV and myocyte functions were examined in pigs with pacing CHF (rapid pacing, 240 beats/min, 3 weeks, n = 10), pacing CHF with concomitant V1a block (SR49059, 60 mg/kg,n = 8), pacing CHF with concomitant AT1block (irbesartan, 30 mg/kg, n = 7), or pacing CHF with dual block (n = 7). LV end-diastolic dimension and peak wall stress were reduced in all receptor blockade groups compared with CHF values. However, LV fractional shortening was increased only in the dual block group compared with CHF values (29 ± 3 versus 21 ± 2, P < .05). Basal LV myocyte percent shortening increased in the dual block group compared with CHF values (3.44 ± 0.23 versus 2.88 ± 0.11,P < .05). Although V1a or AT1 block reduced LV loading conditions, only dual block resulted in improved LV and myocyte shortening.
It has been demonstrated that angiotensin-converting enzyme (ACE) inhibition improved indices of left ventricular (LV) function and survival in patients with congestive heart failure (CHF) (The CONSENSUS Trial Study Group, 1987; The SOLVD Investigators, 1991). More recently, angiotensin II (Ang II) type 1 (AT1) receptor antagonists have been shown to be well tolerated in patients with CHF and may provide beneficial effects (Pitt et al., 1997). Thus, interruption of AT1 receptor activity is an important therapeutic target for the treatment of CHF. However, although ACE inhibition, presumably through a reduction in Ang II production and AT1 receptor activity, improved survival in patients with CHF, the rate of morbidity and mortality associated with this disease process remain significant. Thus, the development of therapeutic strategies that can operate in conjunction with interruption of AT1 receptor activity in the setting of CHF are warranted. The progression of the CHF process is invariably associated with heightened synthesis and the release of a number of vasoactive peptides (Szatalowicz et al., 1981;Riegger and Liebau, 1982; Goldsmith et al., 1983; Benedict et al., 1993; Naitoh et al., 1994; Wei et al., 1994). For example, past clinical and experimental studies have documented that increased plasma levels of the nonapeptide vasopressin accompany the progression and/or exacerbation of CHF (Szatalowicz et al., 1981; Riegger and Liebau, 1982; Goldsmith et al., 1983; Naitoh et al., 1994). Vasopressin has been implicated as exerting effects through two distinct receptor-mediated pathways (Manning et al., 1993; Bichet, 1994; Burrell et al., 1994). The first vasopressin receptor pathway is the vasopressin V1a receptor subtype, which is located on a number of cell types, including vascular smooth muscle. Activation of the V1a receptor has been shown to cause peripheral vasoconstriction (Riegger and Leibau, 1982; Manning et al., 1993; Bichet, 1994, Burrell et al., 1994). The second vasopressin receptor pathway, the V2 receptor, is located primarily in the distal tubule of the kidney and, when activated, results in water and sodium reabsorption (Manning et al., 1993; Bichet, 1994; Burrell et al., 1994). In experimental rodent models of CHF, it has been demonstrated that V2 receptor inhibition produced beneficial hemodynamic response primarily as a result of an aquaretic effect (Nishikimi et al., 1996; Burrell et al., 1998). However, it remains unclear whether and to what degree activation of the V1a receptor contributes to the progression and/or exacerbation of the CHF process. Accordingly, the overall goal of the present study was 3-fold: 1) to determine the direct effects of V1a receptor blockade during the progression of CHF with respect to LV function and systemic hemodynamics, neurohormonal system activity, and contractility; 2) to compare and contrast the relative effects of V1a receptor blockade with respect to AT1 receptor blockade with the development of CHF; and 3) to determine the potential interaction of combined V1a and AT1 receptor blockade during the progression of CHF.
Chronic rapid pacing in animals has been previously demonstrated to produce changes in LV function and systemic hemodynamics, neurohormonal system activity, and contractility similar to those of the clinical spectrum of CHF (Spinale et al., 1992, 1995; Spinale, 1995; Travill et al., 1992). Specifically, pacing-induced CHF is accompanied by LV pump dysfunction and activation of several neurohormonal systems, including the renin-angiotensin pathway and vasopressin (Riegger and Liebau, 1982; Travill et al., 1992; Spinale et al., 1995, 1997a; Spinale, 1997b; Krombach et al., 1998). Moreover, the institution of ACE inhibition in this model of CHF produces beneficial effects on LV function and neurohormonal systems similar to those achieved in past clinical studies (Spinale et al., 1995, 1997a; Spinale, 1997b; Krombach et al., 1998). Thus, the pacing model may be a useful substrate to examine the effects of specific pharmacologic interruption of receptor pathways in the setting of developing CHF. Accordingly, the present study used a porcine model of pacing CHF that has been previously described, (Spinale et al., 1992, 1995, 1997a; Spinale, 1997b) to examine the effects of V1a receptor blockade, AT1 receptor blockade, and combined receptor blockade.
Materials and Methods
Rationale.
The overall goal of the present study was to examine the effects of V1a receptor blockade, AT1 receptor blockade, and combined receptor blockade (V1a/AT1 receptor blockade) on LV function and myocyte contractility in a pig model of CHF. The first objective was to obtain a dosage for each receptor antagonist as well as for combined receptor blockade that would significantly blunt the appropriate receptor agonist challenge. In addition to identification of single blockade and dual blockade treatment, the second objective of the present study was to determine the effects of chronic single receptor blockade and combined receptor blockade on LV function, hemodynamics, and myocyte contractility in developing pacing CHF.
Dose Selection Studies.
Nine Yorkshire pigs (25 kg, male) were chronically instrumented to measure aortic blood pressure in the conscious state as previously described (Clair et al., 1998; Krombach et al., 1998). After a recovery period of 7 to 10 days, the animal was returned to the laboratory for initial baseline hemodynamic assessment and pressor response studies. For these studies, the animals were sedated with diazepam (20 mg Valium p.o.; Hoffmann-La Roche, Nutley, NJ) and placed in a custom-designed sling that allowed the animal to rest comfortably. The vascular access port was entered using a 12-gauge Huber needle (Access Technologies, Skokie, IL) containing dual access sites, and basal, resting arterial pressure and heart rate were recorded. Pressures from the fluid-filled aortic catheter were obtained using an externally calibrated transducer (Statham P23ID; Gould, Oxnard, CA). The pressure waveforms were recorded using a multichannel recorder (Hewlett Packard, Houston, TX) as well as digitized on computer for subsequent analysis at a sampling frequency of 250 Hz (80386 processor; Zenith Data Systems, St. Joseph, MO). After these baseline measurements, a bolus infusion of Ang II (10 μg; Sigma Chemical Co., St. Louis, MO) was administered, and measurements were repeated 5 min after the Ang II infusion. This dose of Ang II was determined previously to yield a near-maximal blood pressure effect (Spinale et al., 1997a). After a 30-min recovery period, in which hemodynamics had returned to baseline levels, a bolus infusion of vasopressin (40 ng/kg; Fluka, Milwaukee, WI) was administered, and measurements were recorded as described previously. This dose of vasopressin was determined previously to yield a near-maximal blood pressure effect (Serradeil-Le Gal et al., 1993). Furthermore, previous dose-ranging experiments (0–80 ng/kg) in the conscious pig preparation showed that 40 ng/kg produced a near-maximal blood pressure response. The order of Ang II and vasopressin administration was alternated with each study.
After the baseline and pressor response studies, the pigs were randomized to receive either the V1a receptor antagonist SR49059 (60 mg/kg b.i.d.; n = 3) or the AT1 receptor antagonist irbesartan (30 mg/kg b.i.d.; n = 3) for 3 days. The doses for these compounds were determined from preliminary dose-ranging studies. Higher doses of the V1a receptor antagonist resulted in tachycardia and unstable hypotension. Based on initial pharmacokinetic studies, it was determined that a 3-day period provided adequate time for drug levels to reach steady-state plasma levels. After the morning dose on the 3rd day, the pigs were brought back to the laboratory for pressure response studies in the same manner outlined previously. From the dose-ranging studies, a dual blockade dose (dual block, 60 mg/kg SR49059 b.i.d. and 30 mg/kg irbesartan b.i.d.; n = 3) was used. This dual receptor blockade dose was used in the pacing protocol. According to Serradeil-Le Gal et al. (1993), SR49059 is a very potent and specific antagonist of the V1areceptor and does not appear to exhibit any relevant effects on other G protein-coupled receptors. Also, it has been shown previously that the AT1 receptor antagonist irbesartan displays high specificity for the AT1 receptor and negligible affinity for other receptor subtypes (Cazaubon et al., 1993).
Model of CHF and Experimental Design.
Thirty-two age- and weight-matched pigs (Yorkshire, 25 kg) were anesthetized, and a left thoracotomy was performed as described earlier. In addition, a shielded stimulating electrode was sutured onto the left atrium, connected to a modified programmable pacemaker (8329; Medtronic, Inc., Minneapolis, MN), and buried in a subcutaneous pocket.
Researchers at our laboratory have demonstrated previously that chronic rapid atrial pacing reliably causes LV dilation and pump dysfunction within a 21-day period (Tomita et al., 1991; Spinale et al., 1992). Therefore, after a 14- to 21-day recovery period from the surgical procedure, the animals were randomly assigned to one of four groups: 1) chronic rapid pacing at 240 beats/min for 21 days (n = 10), 2) chronic rapid pacing at 240 beats/min for 21 days with concomitant V1a receptor blockade (60 mg/kg;n = 8), 3) chronic rapid pacing at 240 beats/min for 21 days with concomitant AT1 receptor blockade (30 mg/kg; n = 7), or 4) chronic rapid pacing at 240 beats/min for 21 days with concomitant dual blockade (V1a block 60 mg/kg and AT1block 30 mg/kg; n = 7). LV function, systemic hemodynamics, and neurohormonal activity were measured at the normal baseline state and then every 7 days of the pacing protocol. At the end of the pacing protocol, all pigs were brought to the laboratory for pressure response studies to Ang II and vasopressin as outlined previously. All animals used in this study were treated and cared for in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (National Research Council, Washington, DC, 1996).
LV Function and Plasma Collection.
Indices of LV pump function were obtained from simultaneously recorded pressure and echocardiographic measurements previously described at our laboratory (Tomita et al., 1991; Spinale et al., 1992, 1995). LV peak circumferential wall stress was computed using a spherical model of reference: ς (grams per square centimeter) = [PD/4 h(1 + h/D)] × 1.36, where P is aortic systolic pressure, D is minor axis dimension at end-diastole, and h is wall thickness (Tomita et al., 1991). LV end-systolic wall stress was computed with the same formula but with the appropriate systolic dimensions. After these measurements, 40 ml of blood was drawn from the arterial access port into chilled tubes containing EDTA (1.5 mg/ml). A separate sample was collected into chilled tubes containing appropriate serine protease inhibitors for additional neurohormonal studies. The blood samples were immediately centrifuged (2000g, 10 min, 4°C), and the plasma was decanted into separate tubes, frozen in a dry ice/methanol bath, and stored at −80°C until the time of assay. An additional 5-ml blood sample was taken for subsequent serum osmolality and electrolyte analysis. To more carefully examine LV ejection performance, a normal control state LV fractional shortening afterload relationship was determined as described previously (Colan et al., 1984; Tomita et al., 1991). This approach provides a relatively load-independent index of ejection performance and does not require theoretical muscle models and the development of LV pressure-volume loops (Ross, 1976). Using this normal relationship, steady-state LV fractional shortening-peak wall stress values were plotted at the end of each of the treatment protocols.
Neurohormones and Analytes.
Plasma samples were assayed for norepinephrine, endothelin, vasopressin, Ang II, and atrial natriuretic peptide (ANP) levels. Plasma samples were also assayed for drug levels of SR49059 and the AT1 receptor antagonist. Plasma norepinephrine was measured using HPLC and normalized to picograms per milliliter of plasma; this assay had a less than 4% coefficient of variation (Anilytics, Bethesda, MD). For the endothelin determinations, the plasma was first eluted over a cation exchange column (C18 Sep-Pak; Waters Associates, Milford, MA) and then dried by vacuum-centrifugation. The samples were reconstituted in 0.02 M borate buffer, and a high-sensitivity radioimmunoassay was performed (RPA545; Amersham, Arlington Heights, IL). The recovery from the extraction procedure was 75 ± 5% based on spiked plasma standards (4–20 fmol/ml). The interassay variation was 10% and the intra-assay variation was 9% for the endothelin radioimmunoassay procedure. For the Ang II determinations, the plasma was eluted with methanol, and a high-sensitivity radioimmunoassay was performed (NR 79980, Angiotensin II Pasteur; Sanofi Diagnostics Pasteur, Montpellier, France). For lysine-vasopressin determinations, plasma was eluted with ethanol, and a high-sensitivity radioimmunoassay for lysine-vasopressin was performed (NR 23065; DIASORIN, Stillwater, MN). ANP levels were determined from eluted plasma by radioimmunoassay. Serum Na+ was measured by ion selective potentiometry. Plasma levels of SR49059 and the AT1 receptor antagonist were determined by a mass spectrophotometry method (ESI liquid chromatography-mass spectrometry/mass spectrometry) and HPLC, respectively. Serum blood urea nitrogen (BUN) and creatinine were determined by measurements of a chromogenic substrate (Vitros; Johnson & Johnson, Rochester, NY). Serum osmolality was measured by freezing-point osmometry.
Myocyte Isolation and Contractile Function Studies.
After the final set of LV function measurements and plasma collection, the animals were anesthetized as described earlier, a sternotomy was performed, and the heart was quickly extirpated and placed in a phosphate-buffered ice slush. The region of the LV free wall incorporating the circumflex artery (5 × 5 cm) was excised and prepared for myocyte isolation. The region of the LV free wall composing the left anterior descending coronary artery (3 × 5 cm) was cannulated and prepared for perfusion fixation.
Myocytes were isolated from the LV free wall and examined using methods described by at laboratory previously (Spinale et al., 1992, 1995). Viable myocytes were defined as those cells that retained a rod shape, were calcium tolerant, and responded to electrical stimulation. In addition to basal measurements of contractility, myocyte function was determined after β-adrenergic receptor stimulation with 25 nM isoproterenol (Sigma Chemical Co.) and 8 mM Ca2+. These doses have been demonstrated previously to result in near-maximal response for this myocyte preparation (Spinale et al., 1992; Tanaka et al., 1993).
Data Analysis.
Indices of LV function, systemic hemodynamics, and neurohormonal profiles were compared among the treatment groups by ANOVA for repeated measures. An ANOVA with a randomized-block split-plot design was used for the myocyte function studies. For these myocyte studies, each pig was a block and drug treatment was the parameter for the split-plot design. If the ANOVA revealed significant differences, pairwise tests of individual group means were compared by the use of Bonferroni's probabilities. All statistical procedures were performed with the BMDP statistical software package (BMDP Statistical Software Inc., Los Angeles, CA). Results are presented as mean ± S.E. Values of P< .05 were considered to be statistically significant.
Results
All of the animals entered into the individual protocols successfully completed the study.
Pressor Response Studies.
The results of the pressor responses to Ang II and vasopressin in normal pigs is summarized in Fig.1. The 60-mg/kg dose of the V1a receptor antagonist resulted in a slight blunting of the response to Ang II compared with control levels and a greater than 50% reduction in the response to vasopressin compared with control values. The 30 mg/kg dose of the AT1receptor antagonist resulted in a significant blunting of the Ang II pressor response compared with control levels. However, there was a significant potentiation of the vasopressin pressor response after AT1 receptor blockade compared with control measurements. The dual blockade treatment yielded a significant blunting of both the Ang II and the vasopressin response compared with control values.
Plasma Compound Levels.
The plasma levels (drawn at the midpoint between the morning and evening doses) were determined for each receptor antagonist on the final day of the chronic CHF study. For the V1a receptor antagonist, SR49059, plasma levels were 8.93 ± 2.69 mg/l in the monotherapy group. The plasma levels for the V1a receptor antagonist were similar to monotherapy values in the dual blockade group (10.65 ± 0.75 mg/l, P = .56). Based on past in vitro studies, the plasma concentration of SR49059 reflects an approximate 2000-fold higher concentration than necessary to inhibit V1a-mediated vessel contractility (Serradeil-Le Gal et al., 1993).
LV Function and Hemodynamics.
Complete LV function and hemodynamic measurements at week 3 of rapid pacing are summarized in Table 1. Ambient resting heart rate was increased in all of the rapid pacing groups compared with baseline values. In the AT1 blockade group and the dual blockade group, ambient resting heart rate was lower than that of untreated rapid pacing values. In all rapid pacing groups, mean aortic pressure was lower than baseline values and was further reduced in the dual blockade group. Moreover, in the AT1blockade group, mean aortic pressure was significantly reduced from baseline, pacing-only, and all other receptor blockade groups. LV wall stress patterns increased approximately 3-fold in the rapid pacing-only group and was significantly reduced in all receptor blockade groups. LV wall stress was reduced to the greatest degree in the AT1 blockade group and the dual blockade group compared with untreated pacing values.
To more carefully determine LV ejection performance with respect to rapid pacing and receptor blockade, the stress-shortening relationship was determined with a phenylephrine infusion. The LV stress-shortening relation fell in an inverse relationship with phenylephrine infusion in the normal control group as described previously (Fig.2) (Colan et al., 1984; Tomita et al., 1991). The steady-state LV stress-shortening values for the rapid pacing groups were plotted with respect to this normal relationship. In the rapid pacing-only group, the stress-shortening value moved downward and to the right, which is indicative of intrinsic defects in LV myocardial performance (Colan et al., 1984; Tomita et al., 1991). In all of the receptor blockade groups, the stress-shortening values moved upward and to the left, suggesting improved LV myocardial ejection performance.
Neurohormones and Analytes.
Plasma neurohormone values are summarized in Table 1. An approximately 5-fold increase in plasma norepinephrine was observed in the untreated pacing group compared with baseline values. Plasma norepinephrine was significantly reduced in all receptor blockade groups compared with pacing-only values. In the AT1 blockade and dual blockade groups, plasma norepinephrine was further reduced from rapid pacing-only values. Plasma endothelin levels are also summarized in Table 1. As with plasma norepinephrine, a 5-fold increase in plasma endothelin was observed in the untreated pacing group compared with baseline values. Plasma endothelin values were reduced from the untreated pacing group in the AT1 blockade and the dual blockade groups. Plasma Ang II levels were increased from baseline values in all pacing groups but were lower in the V1a and dual receptor blockade groups. Plasma lysine-vasopressin was increased in the pacing-only and in the dual blockade groups. Plasma ANP was increased from baseline values in all pacing groups but was reduced from pacing-only values in both the AT1 and dual receptor blockade groups.
The changes in serum Na+, osmolality, blood urea nitrogen (BUN), and creatinine from baseline values are summarized in Fig. 3. Serum Na+significantly fell in all rapid pacing groups. This observation is consistent with increased plasma water content. In all the receptor blockade groups, the change in serum Na+ was reduced from pacing-only values. The serum osmolality significantly decreased from baseline values in the dual blockade group. Serum BUN significantly increased in the rapid pacing-only group from baseline values and was reduced from rapid pacing-only values in all receptor blockade groups. Serum creatinine levels were increased from baseline values in all rapid pacing groups regardless of treatment protocol.
Myocyte Contractility.
Myocyte contractile function was examined in more than 1000 LV myocytes from all treatment groups, and this analysis is summarized in Table 2. Indices of myocyte contractile function were reduced by approximately 50% in all rapid pacing groups compared with normal control values. There was no significant improvement in basal contractile performance in the V1a blockade or the AT1 blockade groups. However, a significant improvement in myocyte contractile function was observed in the dual blockade group compared with rapid pacing-only values.
To examine the capacity of the myocyte to respond to an inotropic stimulus, myocyte contractile function was examined after the addition of 25 nM isoproterenol and 8 mM Ca2+. The results are summarized in Table 2. In the presence of isoproterenol, myocyte contractile function was significantly increased in all groups compared with baseline values. However, myocyte function remained reduced in the presence of isoproterenol in all rapid pacing groups compared with normal control values. In the AT1 blockade group and the dual blockade group, myocyte contractile function with isoproterenol was higher than rapid pacing-only values. In the presence of 8 mM Ca2+, myocyte contractile function was increased significantly in all rapid groups compared with baseline values. However, myocyte function remained reduced in the presence of 8 mM Ca2+ in all rapid pacing groups compared with normal control values.
Discussion
The goal of the present study was to examine the effects of V1a receptor blockade, AT1receptor blockade, and dual receptor blockade in a pacing model of CHF. Using a single or a combined dose of the V1a and AT1 receptor antagonists that inhibited the pharmacologic response to the appropriate agonist (vasopressin, Ang II, or both) during the development of pacing CHF, two important observations were made. First, V1a receptor blockade, AT1 receptor blockade, and dual receptor blockade reduced LV wall stress and plasma norepinephrine from untreated CHF values. However, only dual receptor blockade resulted in a positive effect on LV pump function. Second, at the level of the myocyte, basal contractile function was increased from untreated CHF values in the dual receptor blockade group only. Thus, the present study demonstrated a potential interaction with V1a and AT1 receptor blockade in a model of CHF.
Pressor Response Studies.
In the present study, the doses selected for the V1a and the AT1 receptor antagonists provided adequate blockade of the appropriate receptor in response to agonist infusion. However, there was a significant potentiation of the vasopressin pressor response after AT1 receptor blockade compared with control measurements. It has been shown previously in a conscious rat preparation that Ang II binding of the AT1 receptor influences the release of vasopressin (Yamaguchi et al., 1982). Therefore, although remaining speculative, inhibition of the AT1 receptor may have reduced local vasopressin release and increased the number of unoccupied V1a receptors. This may have resulted in heightened activity of the V1a receptor system after exogenous vasopressin administration and yielded a potentiated vasopressin response in the AT1 receptor blockade group. It must be recognized that these vasopressin studies were performed in only the normal pig preparation. Interpretation of results from pressor studies performed in the setting of CHF can be problematic due to endogenous neurohormonal system activation.
LV Function and Hemodynamics.
In the present study, V1a receptor blockade did not significantly affect resting heart rate and aortic blood pressure from untreated CHF values. However, a small, but significant, reduction in the degree of LV dilation and wall stress occurred with V1areceptor blockade, suggesting some favorable effects on LV loading conditions. Nevertheless, this was not translated into an improvement in LV pump function as determined by LV fractional shortening. In the AT1 receptor blockade group, a significant reduction occurred in mean aortic blood pressure and LV wall stress patterns. However, AT1 receptor blockade was not accompanied by significantly improved LV pump function. Dual receptor blockade reduced LV loading conditions (as defined by LV wall stress), and this effect was translated into an overall improvement in global LV pump performance. Interestingly, mean aortic blood pressure was higher in the dual blockade group compared with the AT1receptor blockade group. The increase in blood pressure in the dual blockade group was likely due to heightened plasma norepinephrine and plasma endothelin levels observed in this group compared with the AT1 group. The reduction in resting aortic blood pressure with AT1 receptor blockade was not apparently associated with significant hemodynamic compromise as evidenced by a reduction in resting heart rate compared with untreated pacing values. Nevertheless, the significant reduction in resting blood pressure that occurred in the AT1 receptor blockade group makes comparisons of load-dependent indices of LV pump function difficult with respect to other treatment interventions.
Neurohormonal Systems.
In the present study, V1a receptor blockade reduced plasma norepinephrine levels from that of untreated pacing CHF values. This was likely due to a secondary effect of improved systemic hemodynamics achieved with V1a receptor blockade. In the AT1 receptor blockade group, the significant reduction in plasma norepinephrine levels was probably due to a favorable effect on systemic hemodynamics, as well as direct Ang II-mediated inhibition of sympathetic activation (Zimmerman et al., 1972; Brasch et al., 1993). Another neurohormonal system that is activated with the progression of CHF is the endothelin system (Wei et al., 1994; Clair et al., 1998; Krombach et al., 1998). Consistent with the clinical phenotype of CHF, pacing CHF resulted in increased plasma endothelin. In both the AT1 and dual receptor blockade groups, plasma endothelin levels were reduced from untreated pacing values. The reduction in endothelin levels in both the AT1 and dual receptor blockade groups was a probable contributory factor for the improvements in myocyte contractility because endothelin has been shown previously to influence contractile performance (Thomas et al., 1997). Ang II levels were reduced with V1a receptor blockade compared with CHF values. This reduction in Ang II levels was likely due to favorable hemodynamic effects observed in the V1a and dual receptor blockade groups. However, in the AT1receptor blockade group, Ang II levels were not different from pacing-only values, which was probably due to a loss of receptor feedback inhibition. In the present study, lysine-vasopressin levels were examined because this is the predominant form of vasopressin in pigs (Stebbins et al., 1994). In the AT1 receptor blockade group, plasma vasopressin levels were lower than pacing-only values, which was likely due to favorable hemodynamic effects. In the V1a and dual receptor blockade groups, plasma vasopressin was not different from pacing-only values. One contributory factor for this effect may be an interaction between both the V1a and the AT1 receptors (Yamaguchi et al., 1982). Plasma ANP was reduced in the AT1 receptor blockade groups from pacing-only values. This was likely a result of the reduced systemic pressures and LV chamber dimensions observed in these treatment groups.
Consistent with severe CHF, serum BUN and creatinine levels were increased in the pacing CHF group. In all treatment modalities, BUN was reduced from pacing CHF values, suggesting improved organ perfusion. However, plasma creatinine levels were not significantly reduced in any of the receptor blockade groups compared with pacing CHF values. This is of particular relevance in light of the fact that in both groups undergoing AT1 receptor blockade, a reduction in blood pressure occurred; however, this was not associated with a compromise in renal perfusion. Moreover, the relative BUN/creatinine ratio was reduced in all the treatment groups, providing further evidence to suggest that there was no compromise in renal function.
LV Myocyte Function and Inotropic Response.
To more carefully examine inherent myocyte contractile performance in the absence of external loading conditions and neurohormonal system activity, isolated myocyte contractility studies were performed. Consistent with past reports from this laboratory, the development of CHF resulted in inherent defects in LV myocyte contractile function (Spinale et al., 1992, 1995; Spinale, 1997b). V1a receptor blockade alone or AT1 receptor blockade alone did not change basal myocyte contractile function from CHF values. Dual receptor blockade significantly increased indices of myocyte contractile function compared with untreated pacing CHF values. Therefore, the improvement in indices of LV ejection performance observed in the dual receptor blockade group was likely due to an inherent improvement in myocyte contractile function.
In the present study, the capacity of the myocyte to respond to an inotropic stimulus was examined through β-receptor stimulation or increased extracellular Ca2+. Consistent with past reports (Tanaka et al., 1993; Spinale et al., 1995, 1997b), the diminished response to either inotropic stimulus occurred in pacing CHF myocytes. It has been demonstrated previously that the contributory mechanisms for the reduced inotropic response to β-receptor stimulation include reduced β-receptor density and alterations in β-receptor transduction (Spinale et al., 1995). The diminished inotropic response to extracellular Ca2+ was likely due to inherent defects in Ca2+homeostatic processes that have been reported to occur with pacing CHF (Spinale, 1997b). In the present study, V1areceptor blockade did not significantly improve inotropic response to either β-adrenergic stimulation or extracellular Ca2+. In the AT1 receptor blockade group, myocyte contractility with β-adrenergic stimulation was increased, but myocyte contractile response to exogenous Ca2+ was unchanged from CHF values. A likely contributory mechanism for this effect was an inherent protection on the β-adrenergic receptor system due to the reduction in plasma norepinephrine levels that occurred with AT1receptor blockade. In the dual receptor blockade group, a similar selective effect to that of AT1 receptor blockade only was observed with respect to myocyte β-adrenergic response. These findings suggest that V1a receptor blockade or AT1 receptor blockade did not provide a protective effect on myocyte intracellular Ca2+homeostatic processes with the development of CHF.
Vasopressin and CHF.
Past clinical and experimental studies have documented that increased plasma levels of vasopressin accompany the progression and/or exacerbation of CHF (Szatalowicz et al., 1981;Riegger and Liebau, 1982; Goldsmith et al., 1983; Naitoh et al., 1994). Increased vasopressin results in the activation of not only the V1a receptor, causing increased vascular resistance, but also the V2 receptor, resulting in increased sodium and fluid retention at the level of the kidney (Manning et al., 1993; Bichet, 1994; Burrell et al., 1994, 1998;Nishikimi et al., 1996). It must be recognized that in the present study, we used V1a receptor blockade. Moreover, V2 receptor blockade, both acute and chronically, has been demonstrated to have a significant aquaretic effect as demonstrated by a reduction in organ weight accumulation and a decrease in urine osmolality (Nishikimi et al., 1996; Burrell et al., 1998). In addition, both acute and chronic administration of nonselective vasopressin receptor blockade has been demonstrated to provide beneficial hemodynamic and aquaretic effects in experimental models of CHF as demonstrated by a reduction in systemic blood pressure and increased urinary output (Mulinari et al., 1990; Naitoh et al., 1994;Wang et al., 1991). Thus, future studies in which combined vasopressin receptor blockade is chronically administered in this model of pacing CHF are warranted. The present study is the first to examine the potential interrelationship between the V1areceptor and the AT1 receptor with respect to LV pump function and myocyte contractile performance. The signal transduction pathways for both the AT1 receptor and V1a receptor involve the activation of phospholipase C with subsequent increases in intracellular Ca2+ and protein kinase C activation (Burrell et al., 1994). The present study provides additional evidence to suggest that an interrelationship exists between activation of the V1a receptor and the AT1receptor in the progression of a CHF process.
Study Limitations and Future Directions.
In the present study, we used a model of chronic rapid pacing that produces a phenotype consistent with clinical CHF. However, only one dose of each inhibitor was used, and these doses were predicated on initial dose-ranging studies for both V1a receptor blockade and AT1 receptor blockade in normal animal preparations. Therefore, dose-response relationships with V1a receptor blockade in the setting of CHF remain to be established. In the animal preparation used in the present study, higher doses of the V1a receptor antagonist resulted in tachycardia and unstable hypotension. These results suggest that a narrow therapeutic window may exist for SR49059. Based on past in vitro studies (Serradeil Le-Gal et al., 1993), the plasma concentration of SR49059 reflects an approximate 2000-fold higher concentration than necessary to inhibit V1a-mediated vessel contractility. Thus, the plasma levels achieved in the CHF preparation significantly exceed that which is necessary to inhibit V1a receptor activity based on in vitro computations. It must be recognized that the plasma levels of the V1a receptor antagonist achieved in the present study have been reported to bind to the receptor (Serradeil Le-Gal et al., 1993). Therefore, the possibility that the dose of SR49059 used in the study might have also inhibited V2 activity must be considered. Thus, the present study should be considered an initial study, and future studies are necessary to more carefully define the role of the V1a receptor in developing CHF. Furthermore, although the present study evaluated V1a receptor blockade in combination with AT1 receptor blockade, a more appropriate comparison may have been with that of an ACE inhibitor; past studies have compared acute administration of V1a receptor blockade with that of an ACE inhibitor (Arnolda et al., 1991). Thus, future studies that evaluate V1a receptor blockade in the background of ACE inhibition in developing CHF are warranted.
Acknowledgments
This work formed the thesis for the Master's degree for M.J.C., and the advice from the Research Committee members, Drs. J. G. Ondo, G. E. Tempel, R. Mukherjee, and A. Ergul, is greatly appreciated. We would like to acknowledge Terese Patterson for her excellent technical assistance during the course of this project.
Footnotes
-
Send reprint requests to: Francis G. Spinale, M.D., Ph.D., Cardiothoracic Surgery, Room 625, Strom Thurmond Research Bldg., 770 MUSC Complex, 114 Doughty St., Charleston, SC 29425.
-
↵1 This work was supported in part by National Institutes of Health Grants HL59165 and HL57952 (F.G.S.) and an unrestricted Basic Research Grant from Sanofi-Synthelabo/Bristol Meyers-Squibb (F.G.S.). F.G.S. is an Established Investigator for the American Heart Association.
- Abbreviations:
- ACE
- angiotensin-converting enzyme
- AT1
- angiotensin II type 1
- CHS
- congestive heart failure
- LV
- left ventricular
- ANP
- atrial natriuretic peptide
- BUN
- blood urea nitrogen
- Received August 20, 1999.
- Accepted February 4, 2000.
- The American Society for Pharmacology and Experimental Therapeutics