Previous studies have demonstrated the presence of myocardial depression in clinical and experimental septic shock. This response is mediated, in part, through circulating TNF-α-induced, nitric oxide-dependent, depression of basal myocyte contractility. Other mechanisms of early myocardial dysfunction involving decreased response to adrenergic stimulation may exist. This study evaluated the presence and nitric oxide dependence of impaired adrenergic response to TNF-α in in vitro cardiac myocytes. The contraction of electrically paced neonatal rat cardiac myocytes in tissue culture was quantified using a closed-loop video tracking system. TNF-α induced depression of baseline contractility over the first 20 min of cardiac myocyte exposure. This effect was blocked by N-methyl-arginine (NMA), a nitric oxide synthase inhibitor, in all studies. Contractile and cAMP response to increasing concentrations of isoproterenol was deficient in cardiac myocytes exposed to TNF-α regardless of the presence of NMA. In contrast, increasing concentrations of forskolin (a direct stimulant of adenylate cyclase) and dibutyryl cAMP (a metabolically active membrane-soluble analog of cAMP) completely reversed TNF-α-mediated depression, though only in the presence of NMA. Forskolin-stimulated cAMP generation remained intact regardless of NMA. Increasing concentrations of exogenous calcium chloride, unlike other inotropic agents, corrected TNF-α-mediated defects of contractility independent of the presence of NMA. These data suggest that TNF-α exposure is associated with a second nitric oxide-independent but calcium-dependent early depressant mechanism that is manifested by reduced contractile and cAMP response to β-adrenergic stimulation.
- septic shock
- myocardial depression
- myocardial depressant substance
the normal cardiac response to septic shock involves biventricular dilatation and decreased ejection fractions (14, 24, 25). Clinically, septic myocardial depression is also reflected by a suboptimal stroke volume response to fluid resuscitation (22) and infusion of exogenous catecholamines (15, 20, 32).
The causal role of circulating factors in myocardial depression in human septic shock is now well established (21, 27–29, 31). Measurements of isolated cardiac myocyte contraction in the presence of serum from patients with acute septic shock demonstrate a depression of maximum extent and peak velocity of myocyte shortening that correlates quantitatively and temporally with depression of the patient's left ventricular ejection fraction as measured by radionuclide ventriculography (27, 31). We have demonstrated that this depressant activity represents a synergistic combination of low-circulating concentrations of the proinflammatory cytokines TNF-α and IL-1β (21). Each individual cytokine, as well as combinations of the two, reproduce the depressant effects of septic serum. Furthermore, elimination of TNF-α and IL-1β eliminates the myocardial depressant activity of septic serum (21). This depressant effect appears to be mediated through a nitric oxide (NO)- and cGMP-dependent mechanism (17). Coincubation of TNF-α, IL-1β, synergistic combinations of the two, or human septic serum with competitive inhibitors of NO synthase (NOS) attenuates or prevents depressant responses in isolated cardiac myocytes. TNF-α also blunts generation of cAMP and contractile response to catecholamine stimulation in these myocytes (18). This observation is congruent with evidence of sepsis-associated cardiovascular resistance to adrenergic stimulation noted in a variety of clinical and experimental studies of sepsis and septic shock (15, 16, 20, 26, 33).
This study was designed to further examine the relationships between defects of adrenoreceptor signaling and NO generation in early (<1 h) myocyte depression induced by TNF-α. Specifically, we sought to determine the subcellular location of the defect in β-adrenoreceptor signaling induced by TNF-α and whether or not the defect is related to NO synthesis.
MATERIALS AND METHODS
The methods utilized were a modification of those previously described (21, 27). The basic myocyte assay is designed to measure the depression or enhancement of contractility of beating cardiac myocytes in cell culture in response to agents that are introduced into the growth medium.
Myocyte isolation and culture.
Spontaneously beating newborn rat myocardial cells were established using a modification of the technique described by Harary and Farley (10). Hearts from 2-days-old Sprague-Dawley rat pups were removed (using sterile technique), pooled, and minced into small blocks. The cells were disaggregated with 0.15% bovine pancreatic trypsin (Sigma, St. Louis, MO) in a modified Hank's balanced salt solution (potassium, calcium, and magnesium-free) (GIBCO Laboratories, Grand Island, NY). Suspensions were centrifuged at 500 g for 15 min and plated into 35 × 10-mm petri dishes at a density of 300,000 cells/ml. Plating media (standard control media) consisted of 90% HEPES-buffered RPMI 1640 (GIBCO Laboratories) and 10% heat-inactivated newborn calf serum (GIBCO Laboratories) supplemented with glutamine (Sigma), penicillin (Sigma), and streptomycin (Sigma). The solution pH was 7.35 in room air and ∼7.30 in 5% CO2. A second, custom-made modified version of this media was used for specified experiments involving calcium chloride as an inotropic agent. This medium was calcium depleted (0.20 mM Ca2+) compared with normal media (1.57 mM Ca2+)(GIBCO Laboratories). Previous experience has demonstrated that this modified medium can be used in place of standard RPMI media and provides a comparable support for cardiac myocyte contractions.
After isolation, myocytes were incubated at 37°C in 5% CO2. Medium was changed every 48 h. After plating, cells became confluent and began spontaneously beating within 4 days. Latex microbeads (Polysciences, Worthington, PA) of 3 μM diameter were introduced into the culture after 4 days and affixed themselves to cell membranes. Beating cells were used for the assay between 5 and 10 days after plating. At this time, cardiac myocytes accounted for ∼75–80% of the viable cells based on cell contraction in response to an electrical pulse.
All pipettes, plates, and other equipment used for preparation, culture, or testing of cardiac myocytes were endotoxin-tested and disposable. All liquid media and other utilized solutions contained <1 pg/ml endotoxin, with the exception of newborn calf serum, which contained 0.48 ng/ml endotoxin. Human recombinant TNF-α derived from yeast (Sigma) with 20,000 IU/μg L929 cytotoxicity assay activity was utilized for experiments. Recombinant TNF-α contained <50 pg endotoxin/μg. Culture media and other test solutions were tested for endotoxin content using a quantitative, chromogenic Limulus amebocyte lysate assay [Whitaker Bioproducts, Walkersville, MD].
Each petri dish containing beating myocardial cells was fastened to the heated (37°C) stage of an inverted-optics, phase-contrast microscope attached to a sidearm video camera. A television monitor displayed the image of the target cells. A custom-built electronic tracking system was used to quantify the rhythmic displacement of a latex bead selected from the many beads attached to the membranes of beating myocytes. The tracking system produced a signal proportional to the maximum extent of each contraction. To ensure a fixed contraction frequency, myocytes were paced using a custom-built alternating current electrical pulse generator (12 V, maximum 40 mA, 0.7- to 7-ms pulse duration). The minimum current and pulse duration required to effectively pace cardiac myocytes was utilized for each experiment.
After plates were removed from the incubator, fresh growth media/standard control media was applied. Following 15-min equilibration to ambient Pco2 and Po2, plates were mounted on the microscope stage and myocytes were paced to 60 contractions/min. An appropriate bead was located, and the maximum extent of myocyte shortening was measured for 5 min (baseline contractility). If the extent of cell contraction was stable (maximum 2.5% variation over 5 min), 1 ml of control media or control media containing either TNF-α (50 ng/ml) or both TNF-α (50 ng/ml) and N-methyl-arginine (NMA) (10 μM) was added. Subsequently, measurements of maximum extent of myocyte shortening were obtained every 5 min. After 20 min observation, 100 μl of control media containing one of the inotropic agents, isoproterenol, forskolin (a direct activator of adenylate cyclase), dibutyryl cAMP (dBcAMP) (a membrane-permeable, metabolically active analog of natural cAMP), or calcium chloride was added and allowed to diffuse through the medium. Measurements of maximum extent of cardiac myocyte shortening were obtained every 5 min for an additional 15–20 min. The duration of observation for each test agent was based on preliminary experiments examining the time point of maximum response. Plates were not reused for additional experiments.
Isoproterenol (Abbott Laboratories, Chicago, IL) in 50 μl media was added to final (plate) concentrations of 0, 0.005, 0.025, 0.125, 0.5, and 2.5 μM (n = 10 each group). Similarly, final concentrations for other inotropic agents were as follows: forskolin (Sigma) 0, 10−5, 10−3, 10−1, 10 μg/l (n = 10 each group); dBcAMP (Sigma) 0, 0.001, 0.1, 1, and 10 mM (n = 10 each group); CaCl2 (Sigma) 0, 0.35, 0.7, 1.4, 2.8 mM (n = 8 each group).
In separate experiments using an identical protocol, the cAMP content of cardiac myocytes was determined. These additional experiments were performed without observations of myocyte contractility (n = 4 each group). At the time point at which maximum contractile response to isoproterenol or forskolin was observed, myocytes were washed with calcium and magnesium-free phosphate-buffered saline, frozen by direct exposure to liquid nitrogen, and then stored at −70°C.
cAMP concentrations were measured in duplicate using the Biotrak cAMP enzyme-immunoassay system (dual-range) according to the protocol of the manufacturer (Amersham). Frozen tissue culture samples in 35-mm tissue-culture petri dishes were rapidly thawed by the application of a 1-ml volume of room temperature assay buffer (0.05 M sodium acetate, pH 5.8) containing 4 mM EDTA for phosphodiesterase inhibition. Cells were scraped (as an intact tissue sheet) and immersed in boiling water for 10 min to denature and precipitate protein (7). Following 3000 g centrifugation (4°C) for 10 min, 100 μl of supernatant were utilized according to the nonacetylation protocol instructions of the manufacturer. Results were normalized to cell count.
By comparing the maximum extent of myocyte shortening at each 5-min interval to the baseline value, changes in contractility were referenced to initial contractility. To confirm the existence of TNF-α-induced cardiac myocyte depression, the change in maximum extent of cardiac myocyte shortening over the first 20 min (immediately before the specified inotropic agent was introduced) was examined. Percentage change from baseline was used as the dependent variable and was analyzed as a function of time for control and test solutions using ANOVA. Paired comparisons at specific time points were performed using least-square means.
To examine the inotropic response to each agent (isoproterenol, forskolin, dBcAMP, and calcium chloride), the concentration-response relationship at the point of maximum response (for each inotrope) was assessed. Response curves (generated from the point of maximum response) for cells exposed to TNF-α ± NMA were compared with curves for myocytes exposed to control medium by ANCOVA. Second-degree polynomial regression analysis was used to assess the existence of a concentration-dependent cAMP response to isoproterenol and forskolin and to compare the concentration-response curves for TNF-α, TNF-α with NMA, and control media.
In each experimental set, TNF-α alone generated depression of extent of shortening of cardiac myocytes over the first 20 min of exposure (i.e., prior to the introduction of any inotropic agent) (each group P < 0.05, ANOVA). The presence of NMA, in each experimental group, abrogated this response.
Effects of isoproterenol on contractility and cAMP generation response of TNF-α ± NMA-exposed cardiac myocytes.
Figure 1A shows cardiac myocyte contractility (expressed as maximum extent of myocyte shortening) for cells exposed to control or TNF-α (50 ng/ml) ± NMA (10 μM)-containing media stimulated with 0.005 μM isoproterenol. Figure 1B shows the similar cells stimulated with 2.5 μM isoproterenol. The inotropic effect of isoproterenol peaked at ∼5 min following introduction.
Figure 1C shows the response (change in maximum extent of cardiac myocyte shortening) of TNF-α ± NMA-exposed cardiac myocytes to increasing concentrations of isoproterenol (at 5 min postintroduction). A significant isoproterenol concentration-cardiac myocyte contractility relationship was demonstrated for control and TNF-α-exposed cardiac myocytes (P < 0.0001 and P = 0.021, respectively, ANCOVA). The response curve for TNF-α + NMA-exposed myocytes did not reach statistical significance. Despite maximally tolerated concentrations of isoproterenol, the response curve for TNF-α-exposed cardiac myocytes remained significantly offset from the response curve for control media alone (P < 0.0001, ANCOVA). TNF-α-exposed cardiac myocytes demonstrated less of an inotropic response to any given concentration of isoproterenol (P < 0.05 at any time point). Even at the highest concentration of isoproterenol tested (2.5 μM), TNF-α-exposed cardiac myocytes fail to demonstrate maximum extent of cardiac myocyte shortening equal to myocytes not exposed to TNF-α (P = 0.0002). Higher concentrations of isoproterenol resulted in significant toxicity (rhythmic perturbations and impaired myocyte contractility).
The isoproterenol concentration-contractile response for TNF-α + NMA-exposed myocytes was relatively flat across increasing concentrations compared with control myocytes (P = 0.0310, ANCOVA) (Fig. 1C). The inclusion of 10 μM NMA with TNF-α primarily affected contractility at absent or low concentrations of isoproterenol (basal contractility) (Fig. 1, A and C). Extent of contraction (shortening) in myocytes incubated with TNF-α + NMA 5 min after control and introduction of 0 and 0.005 μM isoproterenol was unchanged compared with control and media-exposed myocytes. However, increasing isoproterenol concentrations failed to elicit increases in extent of contraction shortening comparable to those seen in control and media-exposed myocytes (Fig. 1C). Correction of TNF-α-induced depression by NMA was deficient at even maximally tolerated concentrations of isoproterenol (P = 0.0008), and the contractile response was quantitatively similar to myocytes exposed to TNF-α alone (Fig. 1, B and C).
Intracellular cAMP concentrations of cardiac myocytes from the preceding experiment are shown in Fig. 1D. Cells were harvested at 5 min following the introduction of isoproterenol. All groups (control media, TNF-α alone, TNF-α ± NMA) exhibited a significant positive isoproterenol concentration-dependent increase in cAMP content (each P < 0.001, second-order polynomial regression). However, the cAMP response of TNF-α- and TNF-α + NMA-exposed cardiac myocytes to increasing isoproterenol concentrations was significantly blunted compared with controls (both P < 0.02, polynomial regression).
Effects of forskolin on contractility and cAMP generation response of TNF-α-exposed cardiac myocytes.
Maximum response to forskolin occurred ∼10 min after introduction. Increasing concentrations of forskolin generated a distinct contractile and cAMP response relative to isoproterenol (Fig. 2). Figure 2A shows the response (change in maximum extent of cardiac myocyte shortening) of cardiac myocytes 10 min following the introduction of increasing concentrations of forskolin. Forskolin caused a concentration-dependent increase in myocyte contraction shortening across all groups (control, TNF-α alone, TNF-α + NMA; each P < 0.02, ANCOVA). Similar to the isoproterenol experiments, TNF-α alone caused significant depression of cardiac myocyte contractile response to forskolin throughout all tested forskolin concentrations (P < 0.0001, ANCOVA) (Fig. 2A). In contrast to the results with isoproterenol, the presence of NMA with TNF-α completely corrected the TNF-α-induced defect of myocyte contractility. With the addition of NMA to TNF-α, increasing concentrations of forskolin resulted in a response curve similar to control media alone.
Intracellular cAMP concentrations of cardiac myocytes from the forskolin studies are shown in Fig. 2B. Cells were harvested at 10 min following the introduction of forskolin. Similar to the results with isoproterenol, all groups demonstrated a high forskolin concentration-dependent increase in intracellular cAMP concentration (P < 0.001, second-order polynomial regression). However, unlike isoproterenol, there was no attenuation of the response in the presence of TNF-α. All three response curves (control, TNF-α alone, TNF-α + NMA) were similar to each other.
Effects of dibutyryl CAMP on contractility response of TNF-α-exposed cardiac myocytes.
Peak contractile response to dBcAMP occurred 20 min after introduction. dBcAMP produced a maximal response pattern that was similar to forskolin (Fig. 3). A concentration-dependent increase in the extent of myocyte contraction was seen across all three groups (each P < 0.005, ANCOVA). TNF-α again caused a highly significant depression of cardiac myocyte contraction response to dBcAMP across all concentrations (P < 0.001, ANCOVA). In contrast to the response to isoproterenol and similar to forskolin, the presence of NMA with TNF-α resulted in complete correction of the TNF-α-induced defect of cardiac myocyte contraction. The contraction-response curve to increasing concentrations of dBcAMP in the presence of TNF-α + NMA was similar to the response curve in control media.
Effects of calcium chloride on contractility response of TNF-α-exposed cardiac myocytes.
The contractile response to calcium chloride was maximal at 10 min postintroduction. Calcium chloride generated a significant concentration-dependent increase in cardiac myocyte shortening in all groups (control, TNF-α, TNF-α + NMA; each P < 0.01, ANCOVA) (Fig. 4). TNF-α alone was associated with depression of myocyte contraction. This effect existed only at absent or the lowest concentrations of exogenously introduced calcium (P < 0.05) (Fig. 4). Higher concentrations of exogenous calcium chloride completely reversed TNF-α-mediated myocyte depression regardless of the presence of NMA. The response curve to exogenous calcium for TNF-α-exposed cardiac myocytes was significantly different than that for myocytes exposed to control media (P = 0.0466, ANCOVA). The inclusion of NMA in TNF-α-containing media resulted in correction of depressed basal contractility so that the calcium-chloride contractility response curve was similar to that generated by myocytes in control media alone (Fig. 4).
TNF-α contributes to myocardial depression in a wide variety of pathological conditions, including sepsis/septic shock. Studies of human septic shock have suggested the existence of a circulating myocardial depressant substance(s) whose early (<1 h) in vitro depressant activity quantitatively correlates to measures of cardiac contractility (i.e., ejection fraction) documented in vivo (27). We have previously demonstrated that TNF-α and IL-1β are responsible for the in vitro myocardial depressant activity of serum from patients with septic shock (21). This depressant activity (inclusive of TNF-α) has been shown to be mediated, in part, by NO (generated by a calcium-dependent NOS) and cGMP (17). An adrenoreceptor signaling defect in response to catecholamine stimulation has also been implicated in early TNF-α-mediated dysfunction of isolated cardiac myocyte contraction (3, 18). In this study, we sought to determine whether TNF-α may act through multiple mechanisms to generate depression of cardiac myocyte contraction and whether these mechanisms are NO-dependent or independent.
Our data suggest that at least two discrete mechanisms of TNF-α-induced myocyte depression exist. As we have previously documented, exposure of cardiac myocytes to TNF-α generates an early (<1 h) NO-dependent depression of basal contractility (Fig. 1C) (17, 21). This depressant response appears to be abrogated in the presence of NMA (17). However, in this study, increasing concentrations of isoproterenol added to TNF-α-exposed cardiac myocytes failed to correct the depressant effect, even at the highest tolerated concentrations. Although addition of NMA to TNF-α-corrected depression at absent or low concentrations of isoproterenol (basal contractility), there was no impact at higher concentrations (Fig. 1C). This experiment suggests that a second, early NO-independent defect of β-adrenoreceptor signal transduction is induced by exposure of cardiac myocytes to TNF-α. This possibility is supported by the demonstration that cAMP generation in response to isoproterenol is also substantially blunted by TNF-α and is not improved in the presence of NMA (Fig. 1D).
The location of this second early mechanism of TNF-α-induced myocardial depression is addressed by the experiments with forskolin and dBcAMP. β-adrenoreceptor signaling is known to involve a coupled G protein complex that activates adenylate cyclase, which generates cAMP (2). Defects of β-adrenoreceptor signal transduction could involve the receptor itself (e.g., receptor internalization), G protein subunits, or adenylate cyclase.
In myocytes exposed to NMA and TNF-α, dBcAMP and forskolin cyclase generated a contractile response distinct from that seen with isoproterenol (Figs. 1C, 2A, and 3). At absent or low concentrations of isoproterenol, TNF-α-induced decreases in basal myocyte contractility were prevented by NMA (Fig. 1C). However, maximally tolerated concentrations of isoproterenol failed to fully correct TNF-α-induced myocyte depression, even in the presence of NMA, i.e., at maximally tolerated concentrations of isoproterenol, myocyte contraction was still deficient relative to control myocytes. The isoproterenol-stimulated increase in extent of contraction of these TNF-α- and NMA-exposed myocytes was similar to myocytes exposed to TNF-α without NMA.
As with isoproterenol, TNF-α-induced decreases in basal myocyte contractility were prevented by NMA at absent or low concentrations of forskolin or dBcAMP, (Figs. 1C, 2A, and 3). However, in the presence of NMA, maximally tolerated concentrations of forskolin or dBcAMP completely corrected TNF-α-induced myocyte depression (Figs. 2A and 3). This contrasts with the findings seen with isoproterenol.
These data suggest that the location of the defect in β-adrenoreceptor signaling induced by TNF-α is proximal to adenylate cyclase. This proposition is further supported by the demonstration that concentration-dependent forskolin-mediated cAMP generation was not adversely affected by TNF-α exposure whether or not NMA was present (unlike the response seen with isoproterenol) (Figs. 1D and 2B). The cAMP response curves for control, TNF-α-, and TNF-α + NMA-containing media in response to increasing concentrations of forskolin were all similar. This would only be expected if the TNF-α-induced β-adrenoreceptor signal transduction defect was proximal to adenylate cyclase. Although G protein uncoupling is likely, these observations cannot rule out the possibility of direct β-adrenoreceptor responses, e.g., rapid receptor internalization.
Finally, experiments using exogenous calcium chloride as an inotropic agent suggest common distal mechanistic elements in defects of both basal and catecholamine-stimulated contractility. In this set of experiments, NMA again prevented TNF-α-induced depression of basal cardiac myocyte contraction (Fig. 4). However, in contrast to forskolin (Fig. 2A) and dBcAMP (Fig. 3), NMA was not required to eliminate TNF-α-induced depressant effects at higher concentrations of exogenous calcium chloride (Fig. 4). Exogenous calcium completely corrected TNF-α-induced myocyte depression regardless of the presence of NMA. This suggests that reduced calcium availability or sensitivity represents a final common pathway through which both mechanisms of TNF-α-associated depressant activity are mediated.
Most of the work on TNF-α-induced myocyte responses has focused on β-adrenoreceptor dysfunction associated with late (>1 h) depression (3, 4, 6, 9, 36). Early depressant effects (<1 h) of TNF-α on isolated myocyte contractility have been ascribed to various mechanisms. These include sequential NO and cGMP generation (via calcium-dependent NOS activation rather than calcium-independent NOS induction consistent with the short time frame to onset of myocardial depression) (5, 8, 17) and sphingomyelinase-dependent signaling (5, 23, 35). There is some suggestion that these mechanisms do not necessarily operate exclusively of each other (5, 19, 30).
Although significant data links calcium-dependent NOS to physiological modulation of adrenergic responsiveness (1, 11–13, 37), there has been little work on the question of β-adrenoreceptor signaling alterations in immediate/early myocyte depression induced by TNF-α. We have previously demonstrated that TNF-α exposure results in an early effect of attenuation of β-adrenoreceptor-mediated increases in myocyte contractility and cAMP generation (18). Bick et al. (3) have also shown that short-term neonatal rat cardiac myocyte exposure to TNF-α results in blunted contractility and cAMP response to isoproterenol but that responses to forskolin are preserved. Similarly, Sugishita et al. (35) have shown that TNF-α acutely suppresses β-adrenergic-induced increases in contractility and ICa transients.
There exist two major limitations to this model. First, the neonatal rat cardiac myocytes utilized have the disadvantage of not being terminally differentiated as are adult myocytes. Their characteristics change as they age. They exhibit metabolic differences from adult cells, including significant differences in membrane receptors (34). It can be argued that phenomena observed in neonatal rat myocytes may have limited relevance to human adult cardiac pathophysiology. However, this study was performed to better understand septic myocardial depression. In this context, we have already shown a significant relationship between septic serum-induced in vitro cardiac depression in this model and in vivo cardiac depression (decreased ejection fraction) in the septic shock patients from whom the serum is obtained (21, 27, 31). To the extent that TNF-α is a central component of septic serum-induced myocyte depression (21), our current findings provide insight into the pathophysiological mechanisms that may underlie septic myocardial depression. In addition, the use of serum-based media, although a necessary component of this study, also introduces additional variables. Bovine serum contains a variety of poorly understood, unidentified factors that can affect myocyte metabolism. However, the use of controls using the same 10% bovine serum for comparison purposes compensates for this variable.
Several conclusions can be drawn from our current data. First, early generation of NO (presumably by a calcium-dependent NOS) is clearly implicated in immediate/early depression of depression of basal cardiac myocyte contractility induced by TNF-α. In addition, TNF-α exposure is associated with a second independent early depressant mechanism that is manifested by reduced contractile and cAMP response to β-adrenergic stimulation. This second mechanism appears to be NO independent but calcium dependent. Furthermore, the signaling defect is located proximal to adenylate cyclase. Alterations in the β-adrenoreceptor itself or in the coupled G protein complex are likely.
These mechanisms, although similar in some respects to those described with more prolonged exposure of cardiac myocytes to TNF-α, represent distinct responses that may represent specific targets for intervention in pathological states of inflammatory myocardial dysfunction. Further clarification of the nature of these disturbances will be required to develop specific therapies. Future studies should examine β-adrenoreceptor density/localization on the myocyte cell membrane and integrity of coupled G protein subunit signaling activity following exposure to TNF-α.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society