INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

CONGESTIVE HEART FAILURE (HF) activates a variety of neurohumoral systems, including the sympathetic nervous system, renin-angiotensin-aldosterone system, and pituitary vasopressin system. Numerous experimental approaches have been employed in attempts to clarify the pathophysiological role of sympathetic nervous system activation in HF, which is implied by elevated plasma catecholamine levels in HF patients (8, 22). Plasma norepinephrine concentration ([NE]) represents the balance between spillover of NE into plasma and the clearance of NE from plasma. These events, in turn, are dependent on various processes that include the release, reuptake, and breakdown of NE by the myriad tissues of the body (14). Hence, the interactive and dynamic nature of the determinants of plasma [NE] limits the utility of this parameter as an indicator of sympathetic nervous system activity. Targeted assessments of the nature and extent of sympathetic activation in clinical HF have employed microneurographic monitoring of muscle sympathetic nerve traffic (16, 28) and quantitative assessment of sympathetic activity by radiotracer methodologies (13, 22, 38). The results of these investigations suggest that HF evokes a regional (rather than generalized) pattern of sympathetic activation in humans.

The opportunity to more fully address questions concerning the source and distribution of the elevated sympathetic drive in HF is provided through the use of experimental models, such as the coronary artery ligation model of HF in the rat. As in humans with HF, these animals display elevated plasma catecholamine levels (27, 33). Moreover, renal sympathetic nerve activity measured by multifiber recordings appears to be elevated relative to normal rats (10, 15, 27), suggesting that increased central sympathetic outflow accompanies HF. However, the validity of quantitative comparisons between groups using this technique remains uncertain, because measurement of integrated voltage from a multifiber recording is dependent on the position of the electrode and the nerve preparation (23). It is also important to note that these animals had been subjected to surgery (to implant recording electrodes) only 6-8 h before measuring nerve activity and thus may have still been under some degree of postsurgical stress (10, 15).

NE turnover can be used as a useful indicator of noradrenergic function in specific tissues harvested from experimental animals (5, 31, 39); however, the effect of HF on this parameter has only been assessed in the heart and the results appear to vary between HF models. Specifically, cardiac NE turnover is increased in the cardiomyopathic hamster (41), unaltered or slightly increased in pressure overload hypertrophy (17), and increased in the failing heart of the diabetic rat (19). The literature contains no NE turnover data obtained simultaneously in multiple tissues of conscious rats with HF induced by coronary artery ligation, information that would expose the specific targets of sympathetic activation in the HF state. In addition to targeted sympathetic activation of specific organs, the intriguing possibility exists that HF may alter sympathetic outflow to different regions of individual organs such as the kidney, where independent regulation of cortical and medullary processes can have substantial hemodynamic and excretory consequences (11).

The purpose of the present study was to examine the hypothesis that experimental HF provokes a centrally mediated, regionally heterogeneous increase in sympathetic tone. To address this hypothesis, basal sympathetic tone in a variety of peripheral tissues (as well as in different regions of the kidney) was assessed by measurement of NE turnover in conscious rats studied 6-8 wk after coronary artery ligation (or sham surgery). This model represents the most common mechanical cause of HF seen clinically, and the time course for development of HF mimics the clinical situation (18, 36). In additional experiments, ganglionic blockade was employed to assess the contribution of central nervous system-mediated events to the alterations in NE turnover that were evident during HF.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals. Experiments were conducted in conformity with the "Guiding Principles for Research Involving Animals and Human Beings" after approval was received by the University of Nebraska Medical Center Institutional Animal Care and Use Committee. Male Sprague-Dawley rats weighing 250-280 g were purchased from Sasco Breeding Laboratories (Omaha, NE). The rats were assigned randomly to one of two groups: HF and sham operated. The method of producing HF was similar to that previously described (18, 36). Briefly, each rat was anesthetized with pentobarbital sodium (50 mg/kg ip), and the trachea was cannulated to facilitate ventilation. A left thoracotomy was performed, and the heart was gently lifted out of the thorax. In the HF rats, the left coronary artery was ligated between the pulmonary artery outflow tract and left atrium as it exits the aorta. Sham-operated rats underwent thoracotomy and manipulation of the heart, but the coronary artery was not ligated. After these maneuvers were completed, the heart was returned to its original position and the thorax was closed. The air within the thorax was removed, allowing the rats to resume spontaneous respiration. The trachea was sutured, the neck incision was closed, and the rats were allowed to recover from the anesthesia. The rats were caged individually thereafter, receiving analgesics (Nubain-Stadol, 1 ml/kg sc) on each of the first 2 days after surgery. The animal quarters had a 12:12-h light-dark cycle with ambient temperature maintained at 22°C and humidity at 30-40%. Laboratory chow (Purina) and tap water were available ad libitum. Six to eight weeks after surgery, the rats were used to measure turnover of NE.

NE turnover. Turnover of NE was calculated on the basis of the disappearance rate of NE after tyrosine hydroxylase inhibition (5). All rats were kept in their home cages during the NE turnover study. Animals in each group (sham operated and HF) were randomly assigned to three subgroups. Rats in the first subgroup were used to establish baseline tissue NE concentrations ([NE]₀). These animals were killed by decapitation at time 0. The second subgroup received the tyrosine hydroxylase inhibitor -methyl-DL-p-tyrosine methyl ester HCl (AMPT, 300 mg/kg ip; Sigma Chemical, St. Louis, MO) at time 0 and were killed 4 h later. The third subgroup received AMPT at both time 0 and time 4 h and were killed at time 8 h. The following tissues were harvested from each animal immediately after death: the noninfarcted apical region of the cardiac left ventricle, the entire left kidney, separate samples of cortex and outer medulla from the right kidney, a 2-cm length of duodenum, and representative samples of hindlimb skeletal muscle, spleen, and liver. These samples were weighed, homogenized in cold perchloric acid containing EDTA, and centrifuged at 4°C for 15 min at 15,000 g. The supernatants were stored at 70°C until they were assayed for catecholamines. [NE] were assayed using alumina extraction followed by separation and analysis with HPLC and electrochemical detection (Bioanalytical Systems, West Lafayette, IN). Dihydroxybenzylamine (Sigma) was used as an internal standard in each tissue sample. Details and validation of these procedures in our laboratory have been published previously (31).

Effect of ganglionic blockade. Additional experiments were performed to assess more directly the involvement of centrally mediated neural activation in eliciting the changes in NE efflux from peripheral tissues observed in animals with HF. Animals from each group (sham operated and HF) received injections of both AMPT (300 mg/kg ip) and hexamethonium bromide (2 mg/kg sc; Sigma) at time 0. AMPT treatment was repeated at 4-h intervals, whereas hexamethonium treatment was repeated at 2-h intervals. The animals were killed at time 8 h, after which tissues were harvested and processed for measurement of [NE] as detailed above. The data obtained from these animals were assumed to reflect tissue NE processing in the absence of sympathetic neural input (31).

Myocardial histology. Hearts from sham-operated and HF rats were sectioned into three major segments located from the atrium to the apex: A, B, and C. Segment C was used for NE turnover measurement, as detailed above. Segment B (representing the bulk of the left ventricle) was subjected to graded dehydration with ethanol, embedded in paraffin, cut into sequential 10-µm-thick sections, and stained with hematoxylin-eosin. Sections were projected onto a screen, and the outline of the tissue was diagrammed. The scale drawing thus obtained was used for measuring left ventricular outer (epicardial) and inner (endocardial) circumferences as well as the arc length of the infarcted region. The infarcted fraction of the left ventricular wall was calculated on the basis of these measurements.

Data analysis. Linear regression analysis of the log[NE] vs. time relationship was performed using individual data points obtained at 0, 4, and 8 h after tyrosine hydroxylase inhibition. The slope (m) and standard error of the regression coefficient (SE_R) were computed by the least squares method. Between-group comparison of slopes was performed using a t-test for parallelism (20). The rate constant for NE disappearance (k_NE; defined as m/0.434), NE turnover time (1/k_NE), and NE turnover rate ([NE]₀ × k_NE) were calculated as described by Brodie et al. (5). A t-test was employed to compare tissue [NE] measured 8 h after AMPT in hexamethonium-treated sham-operated and HF groups. Unless otherwise noted, all data are reported as means ± SE. P values < 0.05 were considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cardiac histology. After histological data were examined, rats in the HF group that exhibited infarcts involving <35% of inner ventricular circumference were discarded from the study. At the time of the NE turnover studies, the body weight of rats retained in the HF group (344 ± 13 g, n = 26) was similar to that of sham-operated rats (318 ± 8 g, n = 28). The HF rats displayed transmural infarcts of the left ventricle involving 42.7 ± 1.3% of the inner circumference and 33.5 ± 1.5% of the outer circumference. Sham-operated rats had no observable damage to the myocardium. The minimum ventricular thickness was significantly less in HF rats (0.50 ± 0.04 mm) than in sham-operated rats (1.38 ± 0.15 mm; P < 0.05), suggesting a reduction in myocardial volume in the HF group. Although hemodynamic function was not assessed in these animals (to avoid compromising the NE turnover measurements), we and others previously documented that rats with similar large-to-moderate myocardial infarcts exhibit left ventricular end diastolic pressures exceeding 20 mmHg (18, 34-36).

Turnover of NE. In most tissues, [NE]₀ (baseline levels measured in the absence of tyrosine hydroxylase inhibition) was similar in sham-operated and HF rats (Table 1); however, renal [NE]₀ was elevated significantly in HF rats. Figure 1 illustrates the decline in tissue [NE] during the 8-h after tyrosine hydroxylase blockade. After blockade of synthesis, [NE] in each tissue declined exponentially with time as evidenced by a linear decrease in log[NE] vs. time (correlation coefficient > 0.90). The slope of this relationship, indicating the rate of NE disappearance, was significantly greater in muscle tissues of HF rats (m ± SE_R: left ventricle 0.066 ± 0.019 h¹, skeletal muscle 0.059 ± 0.007 h¹) than in sham-operated rats (left ventricle 0.021 ± 0.006 h¹; skeletal muscle 0.027 ± 0.005 h¹). HF rats also displayed accelerated NE disappearance from the kidney, with the slope values averaging 0.033 ± 0.007 h¹ in sham-operated and 0.076 ± 0.010 h¹ in HF rats (P < 0.05). This effect was largely indicative of a change in cortical NE disappearance rate (sham 0.046 ± 0.010 h¹, HF 0.075 ± 0.009 h¹; P < 0.05), as medullary NE disappearance rate was not altered significantly (sham 0.048 ± 0.016 h¹, HF 0.062 ± 0.036 h¹). Sham-operated and HF groups did not differ with regard to slope values calculated for the intestine, liver, or spleen (Fig. 1). Table 2 summarizes the effect of HF on the fraction of total NE that disappears per unit time (the rate constant, k_NE) and the total time required for dissipation of tissue NE (turnover time). By virtue of the impact of HF on the slope of the log[NE] vs. time relationship, k_NE was elevated significantly in both cardiac and skeletal muscle, as well as in kidney (cortex, but not medulla). Rate constants for duodenum, liver, and spleen did not differ between groups.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Disappearance of norepinephrine (NE) from peripheral tissues after tyrosine hydroxylase inhibition (time 0).  with solid regression lines, data from sham-operated rats;  with dashed regression lines, data from heart failure (HF) rats. Values represent means ± SE (n = 18-21/group). [NE]0, initial NE concentration. * P < 0.05 vs. sham-operated [comparison of regression line slopes using a t-test for parallelism (20)].

NE turnover rates for peripheral tissues in the two groups of rats are summarized in Fig. 2. Compared with sham-operated rats, NE turnover was significantly higher in left ventricle, skeletal muscle, whole kidney, and renal cortex of HF rats, consistent with the accelerated rate of NE disappearance from these tissues. Although there was no significant effect of HF on the slope of the log[NE] vs. time relationship in duodenum (Fig. 1), there was increased NE turnover. HF failed to alter NE turnover in the renal medulla, liver, or spleen.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   NE turnover in peripheral tissues of sham-operated and HF rats. A, muscle; B, kidney; C, visceral organs. Values represent means ± SE (n = 6-8/group). * P < 0.05 vs. sham operated.

Assessment of central component by ganglionic blockade. Figure 3 illustrates tissue [NE] (% of mean [NE]₀) measured 8 h after inhibition of NE synthesis in rats subjected to ganglionic blockade. There were no significant differences between sham-operated and HF rats with regard to relative NE levels remaining in any tissue studied 8 h after inhibition of NE synthesis. Thus elimination of the neurally mediated central component of NE turnover by ganglionic blockade abrogated the effect of HF to promote the disappearance of NE from muscle and kidney. These data suggest that the tissue-specific impact of HF on NE turnover requires activation from the central nervous system.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Tissue NE levels 8 h after tyrosine hydroxylase inhibition in rats subjected to ganglionic blockade. Values represent means ± SE (n = 6 or 7/group). There was no significant difference between sham-operated and HF groups in any tissues examined.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present study provides two principal new findings related to the status of the sympathetic nervous system in HF. First, conscious undisturbed rats with HF display tissue-specific increases in peripheral NE turnover, i.e., increases in some regions but not all. Second, ganglionic blockade abolishes the differences in tissue NE handling between HF and sham-operated rats. These data support the hypothesis that centrally mediated events underlie sympathetic activation of selected tissues in chronic HF.

The validity of these observations is dependent on the sensitivity of the technique to measure NE turnover and the reliability of this parameter as an indicator of noradrenergic function (5, 31, 39). The "synthesis inhibition" or AMPT method of measuring NE turnover has been validated by comparison with other turnover methods (5). A linear decline in log[NE] is evident in a variety of tissues for at least 8 h after inhibition of NE synthesis (5, 31). The rate constant k_NE , which is independent of tissue NE content, can be viewed as a functional index of mean neuronal activity. NE turnover has been shown to change in direct relation to the frequency of electrical stimulation of sympathetic nerves (5, 39). Neurogenic activation by baroreceptor deafferentation increases NE turnover in various peripheral tissues (31), whereas ganglionic blockade reduces NE turnover to a level innate to each individual tissue. These observations indicate a major dependency of NE turnover on neural input from the central nervous system. The innate decline in tissue [NE] after synthesis inhibition and ganglionic blockade reflects the rate of spontaneous leakage of NE from storage vesicles and subsequent metabolism.

There is considerable evidence indicating sympathetic dysfunction in the hearts of humans with HF, as well as in animals with experimentally induced cardiac disease. Decreases in tissue [NE] and NE storage capacity are evident in enlarged hearts, whereas cardiac NE turnover has been reported to be unaltered or slightly increased in pressure overload hypertrophy (6, 17). Failing hearts do not respond adequately to sympathetic nerve stimulation (9) and are reported to have diminished levels of tyrosine hydroxylase (37), the rate-limiting enzyme for NE biosynthesis. Both NE content and turnover are high in failing hearts of rats with streptozotocin-induced diabetes mellitus (19). Cardiac NE turnover is increased in cardiomyopathic hamsters (41), although turnover normalized for heart weight is unaltered; however, no measurements of NE turnover have been reported for rats with HF induced by coronary artery ligation. The present study restricted its assessment of cardiac NE turnover to tissue from the relatively healthy apical portion of the heart, specifically, ventricular muscle not directly influenced by the infarct. Although the results failed to indicate an effect of HF on basal [NE] in this region of the heart, there was an accelerated rate of NE disappearance that was largely responsible for an increase in NE turnover. Consistent with these observations, clinical studies of regional NE kinetics have indicated an increased cardiac NE spillover in humans with HF (13, 29). Eisenhofer et al. (13) showed that increased neuronal release of NE and decreased efficiency of NE uptake both contribute to increased cardiac turnover of NE in humans with congestive HF. Furthermore, they reported that decreased NE stores in the failing human heart result from chronically increased NE turnover and reduced efficiency of NE reuptake and storage, rather than insufficient tyrosine hydroxylation (13). The increase in NE turnover may contribute to diminished myocardial -adrenergic receptor expression in patients with congestive HF (4), as these receptors are downregulated by prolonged agonist exposure.

Few tissues other than the heart have been evaluated for NE turnover in experimental models of congestive HF. Similar to the effect on cardiac NE handling, the results of the present study indicate that both the rate of NE disappearance and the calculated NE turnover were increased in skeletal muscle and kidney of rats with HF. NE turnover was also increased in the duodenum of HF rats, reflecting the tendency of both the rate constant and basal [NE] to be elevated in these animals. In contrast, there was no effect of HF on basal [NE], its rate of disappearance, or NE turnover in the liver and spleen. Consistent with these results, splenic NE turnover has been reported to be normal in the cardiomyopathic hamster (41). These observations provide strong evidence for a regionally heterogeneous pattern of sympathetic activation that appears to be targeted to specific tissues in experimental HF. Increased NE spillover from heart and kidney (but not liver; Refs. 13 and 22) and increased burst number in postganglionic sympathetic efferents to the vasculature of skeletal muscle (16) have been described in clinical HF. The close parallel between the results of the present study and the clinical literature strengthens the contention that sympathetic activation of some (but not all) tissues accompanies HF and indicates that coronary artery ligation in the rat represents an appropriate experimental model for studying sympathoexcitation in HF.

The results of the present study have intriguing implications regarding the impact of HF on renal function. Numerous observations suggest that sympathetic dysfunction contributes to the salt and water retention in HF. For example, acute volume expansion evokes diuresis and natriuresis associated with a reduction in renal sympathetic nerve activity (renal sympathoinhibition) (2). The renal excretory and sympathoinhibition responses to acute volume expansion are blunted in rats with HF (11), and renal denervation has been shown to normalize the excretory response (25, 34). Other evidence obtained using renal denervation or pharmacological inhibition of renal nerves also suggests an exaggerated effect of renal nerves during HF (11, 25, 34). One possibility is that there is increased basal renal sympathetic nerve activity in rats with HF. Because it is difficult to make chronic electrophysiological recordings in conscious rats, a few studies have attempted to examine this issue by recording renal nerve activity in normal and HF rats 6-8 h after surgical implantation of recording electrodes (10, 27). These results suggest increased postganglionic integrated voltage of efferent renal sympathetic nerve activity in rats with chronic congestive HF (10, 27). Others have evaluated cycle activity from multifiber recordings to provide evidence indicative of increased basal renal sympathetic nerve activity in HF rats (15). Our observation of increased renal NE turnover in conscious rats with HF provides further evidence for elevated renal sympathetic nerve activity under these conditions. Consistent with this observation, clinical studies of regional NE kinetics have indicated that renal NE spillover is increased in HF (22). It is also of interest to note that the increase in renal NE turnover was evident in the cortex, but not in the medulla. Thus the regional heterogeneity of HF-induced alterations in NE turnover is evident both between various organs and within specific organs. The intrarenal heterogeneity in NE turnover responses to HF can be expected to impact on specific aspects of renal function that are known to be under the influence of adrenergic mechanisms in the cortex. It is well known that adrenergic influences on the renal cortex promote vasoconstriction, renin release, and reabsorption of sodium and water by the proximal tubules (11, 21). Consistent with our observation of increased NE turnover in the renal cortex, rats with HF display sodium and water retention as well as increased plasma renin activity (11, 25, 34). On the basis of this reasoning, it seems reasonable to infer that increased sympathetic activation targeted to the renal cortex likely contributes to the alterations in renal function commonly observed in HF (11).

Because NE turnover is dependent in part on neural input from the central nervous system, the effect of HF to increase NE turnover in specific tissues suggests a regional heterogeneity in sympathetic neural activation mediated by central mechanisms. However, NE turnover can be influenced by any event that affects synthesis, release, uptake, or breakdown of NE at the noradrenergic nerve terminal. For example, increased circulating or local ANG II levels in HF could stimulate NE release from peripheral sympathetic nerve endings (7, 12), thus promoting increased NE turnover. Because ganglionic blockade (hexamethonium) ameliorated differences in NE handling between the sham-operated and HF rats in all tissues examined, it is unlikely that the observed increase in NE turnover involves altered intraneural metabolism of NE that leaks from storage vesicles. Thus noradrenergic processes at the level of nerve terminal that are independent of neural activity appear to be unaltered in rats with HF, which strongly implicates centrally mediated neural mechanisms in causing the regional increases in NE turnover revealed by the present study. Consistent with these studies, the rate constant of NE turnover was substantially increased in failing human heart and most of the increase was attributed to increased release of NE (13).

The effect of HF to differentially influence NE turnover in specific tissues is interesting in light of the nonuniform distribution of sympathetic activity to various peripheral vascular beds during activation of baroreceptor reflexes (24), which are known to be altered in HF (43). Indeed, HF impairs a variety of autonomic reflexes that traverse the central nervous system (30). Studies in both animals and humans have suggested that impairment of cardiopulmonary and arterial baroreflex sensory mechanisms may be important in the sympathoexcitation in chronic HF. However, sinoaortic denervation does not completely abolish the increased sympathoexcitation or elevated plasma [NE] observed in dogs with pacing-induced HF, suggesting that central mechanisms engender these events (3). Central dysfunction has also been implicated on the basis of studies demonstrating altered neural activity (35, 42) and altered noradrenergic activity in various central sites in rats with HF (40) as well as activation of central sites in doxorubicin-induced congestive cardiomyopathic rabbits (1). Consistent with these observations, recent evidence indicates that release of central nervous system monoaminergic neuronal transmitters is increased in clinical HF and that the degree of activation of aminergic brain neurons is linked to the magnitude of peripheral sympathetic activation (26). We have shown that experimental HF is associated with alterations in hexokinase activity (a marker of neuronal activity) in regions of the hypothalamus that contain vasopressin-producing neurons, as well as in sympathoexcitatory sites such as the paraventricular nucleus and the locus ceruleus (35, 42). In view of these observations, we postulate that increased neuronal activity in these central sites known to be involved in regulation of autonomic outflow may contribute to the increased sympathoexcitation observed with HF (30, 32, 35, 42).

In summary, NE turnover is elevated in specific peripheral organs of conscious undisturbed rats with HF. This observation indicates that HF provokes increased basal sympathetic activation of muscle (both cardiac and skeletal) and kidney (cortex, but not medulla), whereas liver and spleen were unaffected. The ability of ganglionic blockade to abrogate HF-induced disparities in tissue NE handling fuels the contention that the central nervous system engenders this phenomenon. Thus, although centrally mediated sympathetic activation accompanies chronic HF in the rat, the effect demonstrates a regional heterogeneity (between and within organs) that likely contributes to the functional consequences of this pathophysiological state.