Lipopolysaccharide Depresses Cardiac Contractility and β-Adrenergic Contractile Response by Decreasing Myofilament Response to Ca2+ in Cardiac Myocytes
Abstract Lipopolysaccharide (LPS) plays a key role in the pathogenesis of sepsis. Cardiac function and the inotropic response to β-adrenergic stimulation are impaired in sepsis. We hypothesized that LPS, in clinically relevant levels (1 ng/mL), directly depresses contractility and β-adrenergic responses in cardiac myocytes. Cardiac myocytes were isolated from the left ventricle of adult rabbits using digestive enzymes (collagenase and protease). We depyrogenated the enzymes (LPS contamination lowered from 100 to 300 ng/mL to <0.7 ng/mL) to minimize development of LPS tolerance during cell isolation. After 6 hours of incubation with 1 ng/mL LPS, there was a decrease in the extent of active cell shortening with no change in Ca2+ transients (measured with indo 1 fluorescence), indicating decreased myofilament responsiveness to Ca2+. This was related to NO pathways, since cGMP (a second messenger of NO) increased in cardiac myocytes and LPS effects were completely reversed with a 1 mmol/L NG-monomethyl-l-arginine (L-NMMA, a NO synthase inhibitor). LPS did not alter the intracellular Ca2+ response to β-adrenergic stimulation with isoproterenol but attenuated the contractile response (maximal cell shortening, 15.5±1.0% versus 23.3±1.1% in control myocytes; P<.001). LPS attenuation of the contrac-tile response to isoproterenol was restored completely by L-NMMA and almost completely restored (to 86% of the control response) by an inhibitor of cGMP-dependent protein kinase. We conclude that LPS depresses cardiac contractility and the contractile response to β-adrenergic stimulation by a NO-cGMP–mediated decrease in myofilament responsiveness to Ca2+. The direct effects of low levels of LPS on cardiac myocytes may contribute to cardiac depression and hemodynamic decompensation during sepsis.
Lipopolysaccharide, an integral part of the outer layer of the gram-negative bacterial cell wall, triggers the release of a cascade of endogenous mediators and induces hypotension, multiorgan failure, and death from sepsis and septic shock.1,2 LPS depresses cardiac function, which contributes to the development of hypotension during sepsis.3,4 Several cytokines released in response to LPS are capable of depressing cardiac function.5,6 TNF-α, for example, causes cardiac depression and mediates many of the vascular and lethal effects of LPS. However, we found that serum TNF-α does not mediate LPS-induced cardiac depression in vivo.7 This may be related to the fact that LPS can directly depress cardiac myocyte function.8 Therefore, direct effects of LPS (independent from cytokines) are pathophysiologically relevant for cardiac depression during sepsis.
In disease conditions such as septic shock, β-adrenergic activation provides an important compensatory mechanism for increasing cardiac output to help maintain blood pressure and adequate tissue perfusion. However, the myocardial β-adrenergic response is impaired in patients with sepsis.9 Inflammatory cytokines uncouple the β-adrenergic receptor from adenylate cyclase in cardiac myocytes10,11 by activation of iNOS.12 Although LPS activates iNOS in cardiac myocytes,8,13,14 LPS does not cause receptor uncoupling.10 It is not known if LPS directly impairs the β-adrenergic response in cardiac myocytes.
We hypothesized that LPS, in clinically relevant levels (1 ng/mL), directly depresses contractility and β-adrenergic responses in cardiac myocytes. The rationale for this hypothesis was based on prior studies. We found that cell shortening was lower over a wide range of [Ca2+]o (0.5 to 16 mmol/L) in cardiac myocytes isolated from rabbits with LPS-induced left ventricular dysfunction compared with myocytes from control rabbits.15 This suggests that LPS decreases myofilament responsiveness to Ca2+. cGMP, a second messenger of NO, decreases myofilament responsiveness to Ca2+.16,17 LPS activates iNOS to increase cGMP in cardiac myocytes.8 Therefore, we hypothesized that LPS depresses contractil-ity and attenuates the β-adrenergic response by a NO-cGMP–mediated decrease in myofilament responsiveness to Ca2+.
In the present study, we investigated direct effects of LPS on cardiac contractility and the contractile response to β-adrenergic stimulation in cardiac myocytes. We examined the in vitro effects of 1 ng/mL LPS, a clinically relevant level well within the few nanograms per milliliter of LPS measured in the plasma of patients with sepsis.18 To appreciate the direct effects of such low doses of LPS on cardiac myocyte function, we depyrogenated the digestive enzymes used for cell isolation (collagenase and protease, which were contaminated with several hundred nanograms per milliliter of LPS) to minimize induction of acute LPS tolerance.19,20
The present study demonstrates that 1 ng/mL LPS depresses cardiac myocyte function and attenuates the contractile response to β-adrenergic stimulation by a NO-cGMP–mediated decrease in myofilament responsiveness to Ca2+. Thus, clinically relevant levels of LPS have direct effects on cardiac myocytes that contribute to cardiac depression during sepsis.
Materials and Methods
Experiments were performed in accordance with institutional guidelines and the Guide of the Care and the Use of Laboratory Animals, US Department of Health and Human Services. In order to limit LPS exposure, we used sterile disposable labware, baked all glassware to 180°C for 4 hours, and used sterile pyrogen-free water to make all solutions. All LPS levels were measured with a Limulus Amebocyte Lysate test (QCL-1000, BioWhittaker Inc). LPS contamination levels were <0.04 ng/mL in all solutions except the digestive enzymes (see “Enzyme Depyrogenation”).
Single cardiac myocytes were isolated as described previously.15 Briefly, New Zealand White rabbits (1.8 to 2.8 kg, both sexes) were anesthetized with pentobarbital sodium (50 mg/kg IV). The heart was rapidly excised and mounted on a Langendorff perfusion apparatus. The heart was perfused with nominally Ca2+-free Tyrode’s solution containing (mmol/L) NaCl 136, KCl 5.4, MgCl2 1, NaH2PO4 0.33, glucose 10, and HEPES 5, pH 7.4 at 37°C. After 4 to 5 minutes, the superfusate was switched to 50 μmol/L Ca2+ Tyrode’s containing 15 mg/kg (animal weight) of the combination of collagenase B (lot No. 14325222, Boehringer Mannheim Co) and protease (lot No. 84H0613, Sigma Chemical Co) in a 2:1 ratio. After 20 to 30 minutes, the left ventricle was mechanically dissected and filtered gently through a 250-μm nylon mesh. The cell suspension was rinsed while the Ca2+ concentration was gradually increased up to 2 mmol/L. Cardiac myocytes were stored at 22°C in MEM with 3% autologous serum taken from the same rabbit.
Cell preparations were evaluated by counting>200 to 300 cells per dish from >5 to 10 fields visualized on a light microscope. The percentage of rod-shaped myocytes was calculated as an estimate of cell viability. The number of nonmyocytes was counted. The cell preparations typically contained <12% (mean, 4%) nonmyocytes, which were primarily endothelial cells. The identity of endothelial cells was confirmed by fluorescence-labeled monoclonal antibodies directed against the von Willebrand factor (factor VIII, Dako Co). We detected CD14 RNA (by reverse-transcriptase polymerase chain reaction) during the early washes in the cell isolation procedure, indicating the presence of monocytes and/or macrophages. However, CD14 was not detected in the late washes of cell isolation or in the final cell culture. The degree of nonmyocyte contamination did not contribute to the effects of LPS as evaluated in pilot studies. In myocytes incubated for 6 hours with and without LPS (10 ng/mL), the degree of LPS-induced contractile dysfunction did not correlate with the level of nonmyocyte contamination (r=.004, P=.99, n=20 experiments).
The collagenase and protease used for cell isolation contained 100 to 300 ng/mL LPS (measured by the Limulus Amebocyte Lysate test, BioWhittaker Inc). These original or “raw” enzymes were depyrogenated by a series of washes with Triton X-11421 and polymyxin B using recently described methods.22 The raw enzymes were dissolved in nominally Ca2+-free Tyrode’s solution, supplemented with Triton X-114 (1:100 volume), and stirred on ice for 2 hours. The enzymes were warmed to 37°C to separate the Triton X-114 layer and centrifuged at 3000 rpm for 5 minutes, and the supernatant was recovered. Polymyxin B was added to the supernatant (1:100 volume); the solution was stirred at 22°C for 2 hours, warmed to 37°C, and centrifuged at 3000 rpm for 5 minutes; and the supernatant was recovered. Finally, the enzymes were washed through SM-4 bio beads (Bio-Rad Laboratories), which were soaked beforehand with 0.1N NaOH for 24 hours. Depyrogenation removed ≈99.7% to 99.9% of LPS from the enzymes, lowering LPS levels to 0.3 to 0.7 ng/mL. This was associated with ≈50% to 70% loss of total protein. Since the amount of protein (and thus enzyme) loss varied with each depyrogenation, we used the same total amount of depyrogenated enzyme (15 mg/kg animal weight) for all cell isolations.
The necessity of enzyme depyrogenation was documented in a recent study.22 Cardiac myocytes (isolated with depyrogenated enzymes) exposed to 100 ng/mL LPS for 1 hour had an attenuated response to subsequent challenge with the same dose (100 ng/mL) of LPS. Thus, acute tolerance to LPS developed with brief LPS exposure under conditions comparable to isolating cells with untreated raw enzymes.
Cardiac Myocyte Function
Myocytes were placed in sterile 2-mL microscope study dishes with a 0.5-mm glass bottom and superfused at 2 mL/min with 2 mmol/L Ca2+ Tyrode’s solution at 22°C using a syringe pump (Pump 33, Harvard Apparatus Inc). Myocytes were stimulated with platinum electrodes connected to a stimulator (S44, Grass Instruments) at 0.5 Hz, a rate comparable to that found in several ventricular myocyte studies. The stimulator generated a 2-millisecond square-wave pulse set at 50% above threshold. Pulse polarity was alternated every minute by using a custom-made signal processor. Rod-shaped cells with clear striations with no spontaneous contractions were chosen for study. Myocytes were visualized using an inverted microscope (Nikon Diaphot) with a Panasonic GP-CD60 CCD camera attached. Longitudinal cell lengths were measured on-line with a video motion detector (Crescent Electronics) sampling at 60 to 240 Hz with on-line A/D conversion at 60 Hz using a 486 computer with a Codas Data Acquisition System (DI-220) and WinDaq software (Dataq Instruments). Customized software was used to average four cardiac cycles to measure resting myocyte length, minimum cell length, and percent cell shortening (from resting to minimum). Images also were recorded on VHS videotape.
Ca2+ Fluorescence Measurements
For assessment of intracellular Ca2+, myocytes were loaded with indo 1-AM). Indo 1-AM was solubilized in dimethyl sulfoxide containing Pluronic F-127. The cells were loaded with indo 1-AM at a final concentration of 6 μmol/L for 15 minutes at 22°C and then washed with 2 mmol/L Ca2+ Tyrode’s solution to remove the extracellular dye. The experiments were performed at 22°C to minimize the loss of fluorescent indicators from the cell. The cells were plated on 2-mL superfusion chambers (Bioptechs Inc) with laminin-coated coverslips placed on the inverted microscope.
Indo 1 fluorescence was measured with a PTI Alphascan system (Photon Technology International Inc). This system provided and controlled an ultraviolet light (360 nm) via a 75-W xenon arc lamp (Ushio Inc) with a monochromator. The ultraviolet light was reflected toward a fluorescence objective (×40, Nikon Neofluor) by a 380-nm dichroic mirror. The fluorescence emission from the cells crossed a dichroic mirror and then was reflected by a prism toward a second dichroic mirror (455 nm), where the light beam was split. Wavelengths of 405 and 485 nm were selected by band-pass interference filters placed in front of two photomultiplier tubes. The microscope emission field was restricted to a single myocyte with the aid of an adjustable window. The background fluorescence was recorded from a similar-sized field at both wavelengths and then subtracted from the signals recorded from the cell before the fluorescence ratio (405/485) was calculated.
Intracellular Ca2+ Calibration
The myocytes loaded with indo 1-AM were superfused with 2 mmol/L Ca2+ Tyrode’s solution. Once fluorescence intensities at 405 and 485 nm were recorded at 0.5-Hz stimulation, the cells were superfused with the same buffer supplemented with 2,3-butanedione monoxime (40 mmol/L) and the nonfluorescent Ca2+ ionophore BrA-23187 (10 μmol/L) to measure the maximum value of the fluorescence ratio (Rmax). The 2,3-butanedione monoxime was used to completely inhibit Ca2+-induced force development, preventing an alteration in fluorescence due to Ca2+-induced hypercontracture of the myocytes.23 The cells then were superfused with zero Ca2+ buffer made with 10 mmol/L EGTA and nominally zero Ca2+ to evaluate the minimum value of the fluorescence ratio (Rmin). [Ca2+]i was estimated by the equation of Grynkiewicz et al24: [Ca2+]i=Kd · β · (R−Rmin) · (Rmax−R), where Kd is the dissociation constant for indo 1 (taken to be 250 nmol/L24), β is the ratio of free to bound indo 1 fluorescence at 485 nm, and R is the ratio of the two fluorescence intensities measured at 405 and at 485 nm. Neither Rmin nor Rmax differed between the control and LPS-treated myocytes.
The degree of compartmentation of indo 1 was assessed by using digitonin to permeabilize the cell sarcolemma and release cytosolic dye without disrupting the mitochondrial or sarcoplasmic reticulum membranes.25 Indo 1–loaded myocytes were superfused with nominally Ca2+-free Tyrode’s solution containing 50 μmol/L digitonin. The fluorescence intensities at both wavelengths declined to the zero-dye level (background level) within 3 minutes, with a similar time course in both control and LPS-treated cardiac myocytes. This finding suggests that the majority of the intracellular fluorescence dye was within the cytosol under the experimental conditions of the present study and that LPS did not significantly alter the degree of dye compartmentation.
The resting or diastolic [Ca2+]i values calculated under these conditions were 293±22 nmol/L (mean±SEM), with peak (systolic) levels of 665±33 nmol/L in freshly isolated rabbit cardiac myocytes (n=8). These values are comparable to prior measurements in the same species.26,27
Simultaneous Measurements of Cell Length and Ca2+ Fluorescence
We modified the PTI Alphascan system so that all measurements of Ca2+ fluorescence were accompanied by simultaneous measurements of cell length. The cells were transilluminated with red light via the 50-W xenon arc lamp passed through a 650-nm band-pass filter (Omega Optical). This wavelength was long enough not to interfere with fluorescence detection at 405 and 485 nm. The cell image was monitored with the Panasonic CCD camera and Crescent Electronics video motion detector. Data were sampled at 120 Hz with on-line analog-to-digital conversion using a 486 computer with a PTI Alphascan system (OSCAR) and FeliX software (Photon Technology International Inc). After a seven-point smoothing process, the following indexes were calculated from the cell-length recordings: myocyte length at rest (Lmax) and minimum cell length (Lmin) were measured to calculate percent cell shortening (100×[Lmax−Lmin]/Lmax). The peak rates of cell shortening (−dL/dt) and lengthening (+dL/dt) were measured. The following indexes were calculated from the Ca2+ fluorescence recordings: diastolic fluorescence ratio (Rd), peak systolic fluorescence ratio (Rs), percent transient amplitude (%R=100×[Rs−Rd]/Rd), and integral of the transient above the diastolic level. Data from five consecutive steady-state beats were averaged.
The role of cGMP was assessed by measuring cGMP levels in cardiac myocytes incubated for 1 or 6 hours in the absence or presence of 1 ng/mL LPS. For the last 20 minutes of the incubation period, 3-isobutyl-1-methylxanthine (1 mmol/L), a phosphodiesterase inhibitor, was added to the cell cultures to inhibit cGMP breakdown. The medium was removed, and the cells were lysed with ice-cold 65% ethanol. The supernatants were recovered after centrifugation and dried in a SpeedVac System (Savant Instruments Inc). The cGMP content of cell extracts was determined by enzyme immunoassay after acetylation using the Biotrak system (Amersham Life Science Inc). The cGMP content was normalized to milligrams protein per well, which was determined by a dye-binding assay (Pierce Chemical Co) with bovine serum albumin used as a standard.
Cardiac myocytes were incubated in sterile Petri dishes at a density of ≈60 000 cells/mL at 22°C. Separate dishes were used to measure cardiac myocyte function for each condition and time period. Each dish was coded so that the individual making measurements would know when to study myocytes from that dish but would not know if LPS or other substances were present or absent. Each dish was studied only once to minimize the introduction of LPS contaminants.
Protocols were performed with myocytes from at least 3 or 4 rabbits, with a similar number of myocytes studied from each rabbit. Each rabbit contributed a similar number of myocytes to all treatment and control groups, so that treatment effects could be compared in myocytes taken from the same animal. With this study design, each rabbit contributed equally to the results, avoiding unequal weight given to any individual animal.
The effects of low doses of LPS on cell function were evaluated by incubating myocytes with or without 1 ng/mL LPS (Escherichia coli 055, LPS No. B5, lot No. 2039F, List Biological Laboratories Inc). Cell shortening was measured within 1 hour and at 6 hours (5.5 to 6.5 hours) after LPS incubation. In an identical protocol, cardiac cGMP was measured in myocytes exposed to 1 ng/mL LPS or control medium for 1 or 6 hours. The role of NO pathways was evaluated by incubating cardiac myocytes for 6 hours with or without 1 ng/mL LPS in the presence or absence of 1 mmol/L L-NMMA, a NOS inhibitor. Cell function in these four groups (±LPS, ±L-NMMA) was measured 6 hours after LPS incubation.
The effects of LPS on intracellular Ca2+ handling were evaluated in cardiac myocytes incubated for 6 hours with or without 1 ng/mL LPS. Myocytes were loaded with indo 1 to simultaneously measure cell shortening and indo 1 fluorescence. Since indo 1 binds to intracellular Ca2+, we first evaluated the extent to which indo 1 loading decreases cell shortening in control and LPS-exposed myocytes.
The effects of β-adrenergic stimulation were evaluated in cardiac myocytes incubated for 6 hours with or without 1 ng/mL LPS. Baseline cell function (before β-adrenergic stimulation) was measured in 2 mmol/L Ca2+ Tyrode’s solution with myocytes stimulated at 0.5 Hz. Electrical stimulations were stopped, and the superfusate was switched to an otherwise identical solution with isoproterenol added. Three minutes later (chamber volume was completely exchanged within 1 minute), electrical stimulations were resumed. We measured myocyte function continuously for 20 minutes after isoproterenol exposure to establish the time period during which cell function would be stable. Thereafter, we used this time period to assess the relationship between cell shortening and isoproterenol dose (from 10−10 to 10−5 mol/L) in cardiac myocytes incubated for 6 hours, with or without 1 ng/mL LPS. The relationship between cell shortening and isoproterenol dose was assessed by the maximal contractile response, and EC50 was calculated with a sigmoidal fit using a software program (Prism, GraphPad Software).
To evaluate alterations in intracellular Ca2+ handling with β-adrenergic stimulation, cardiac myocytes were incubated with or without 1 ng/mL LPS. At 1 and 6 hours, we simultaneously measured cell shortening and indo 1 fluorescence before (baseline) and after 1 μmol/L isoproterenol stimulation. To evaluate whether β-adrenergic receptor uncoupling10,11 contributes to the impaired contractile response to isoproterenol, we incubated cardiac myocytes with or without 1 ng/mL LPS. At 1 and 6 hours, we measured the response to direct stimulation of adenylate cyclase with 30 μmol/L forskolin.
We evaluated the role of NO on the β-adrenergic response by coincubating control or LPS-treated cardiac myocytes for 6 hours in the presence or absence of 1 mmol/L L-NMMA. The role of cGMP-PK on the β-adrenergic response was evaluated by using a cGMP-PK inhibitor, KT5823 (1 and 10 μmol/L). Control or LPS-treated cardiac myocytes were coincubated in the presence or absence of KT5823 for 6 hours. We attempted to evaluate the effects of guanylyl cyclase using the inhibitor LY83583 in doses of 0.1, 1.0, and 10 μmol/L. However, 6-hour incubation with LY83583 at all three doses significantly decreased cell viability and baseline cell function. Methylene blue, another potential inhibitor of guanylyl cyclase, is also toxic to cardiac myocytes with prolonged coincubation.28
Indo 1-AM and Pluronic F-127 were obtained from Molecular Probes, Inc. Laminin and MEM were from GIBCO BRL. BrA-23187, L-NMMA, LY83583, and KT5823 were from Calbiochem-Novabiochem Co. Pentobarbital sodium was from Abbott Laboratories. Other drugs were from Sigma Chemical Co. BrA-23187, KT5823, and forskolin were prepared as stock solution in dimethyl sulfoxide.
Comparisons between two groups were made by unpaired Student’s t test. Comparisons among three or more groups were carried out by ANOVA. When a significant difference among groups was indicated by the initial analysis, individual paired comparisons were made using a Bonferroni post hoc t test. Differences were considered significant at P<.05. Data are presented as mean±SEM.
Cell viability after 1 hour of LPS treatment (63.6±2.6%) was similar to cell viability in the control group (63.2±2.9%). After 6-hour incubation, cell viability did not change in either LPS-treated cardiac myocytes (63.4±2.4%) or control cardiac myocytes (63.6±2.6%). These data indicate that cell viability remained stable throughout the study and that exposure to 1 ng/mL LPS for 6 hours had negligible cytotoxic effects.
LPS Effects on Contractile Function
Fig 1⇓ shows the effects of 1 ng/mL LPS on contractile function. Each bar represents mean percent cell shortening (+SEM) for 97 to 99 myocytes. LPS did not alter cell function within 1 hour but significantly decreased cell shortening after 6 hours (P<.05). This time course is consistent with LPS-induced activation of iNOS. To evaluate this possibility, cardiac cGMP was measured with an identical protocol. Fig 2⇓ shows that 1 ng/mL LPS did not alter cardiac cGMP after 1 hour but increased cardiac cGMP significantly after 6 hours (P<.05). Thus, there was a similar time course for LPS-induced decrease in cell shortening (Fig 1⇓) and increase in cardiac cGMP (Fig 2⇓).
To confirm the role of NO and cGMP in LPS-induced cardiac depression, cardiac myocytes were incubated with or without 1 ng/mL LPS, with or without 1 mmol/L L-NMMA (an inhibitor of iNOS). Measurements were obtained from the four groups (n=42 myocytes in each group) after 6 hours. Cell shortening decreased in LPS-exposed myocytes (11.0±0.4%) compared with control myocytes (12.9±0.5%, P<.05). L-NMMA alone had no effect on cell shortening (12.3±0.5%, P=NS versus control) but blocked LPS-induced depression in cell shortening (12.6±0.4% with L-NMMA plus LPS, P=NS versus control, P<.05 versus LPS alone).
LPS Effects on Intracellular Ca2+
Intracellular Ca2+ was measured with indo 1 fluorescence. The effect of indo 1 loading on cell shortening was examined in myocytes incubated with or without 1 ng/mL LPS for 6 hours. Cell shortening was measured before and after indo 1 loading (n=7 myocytes for each group). Cell shortening was higher in control myocytes (15.1±0.1%) than in LPS-exposed myocytes (13.4±0.1%, P<.05). Indo 1 loading depressed cell shortening significantly (P<.05, ANOVA) both in control myocytes (10.6±0.1%) and in LPS-exposed myocytes (10.0±0.1%).
The effects of LPS on simultaneous cell shortening and intracellular Ca2+ were measured in myocytes exposed to 1 ng/mL LPS for 6 hours. Cell shortening was significantly lower in LPS-exposed myocytes (9.5±0.2%, n=120) than in control myocytes (10.4±0.2%, n=110, P=.01). However, the percent transient amplitude did not differ between LPS-exposed myocytes (38.8±1.4%) and control myocytes (37.5±1.1%, P=NS). A decrease in cell shortening without change in Ca2+ transients indicates a decrease in myofilament responsiveness to Ca2+.
LPS Effects on Contractile Function With β-Adrenergic Stimulation
The effects of LPS on the β-adrenergic response were evaluated in cardiac myocytes incubated with 1 ng/mL LPS for 6 hours. Fig 3⇓ shows the fold increase in cell shortening with 1 μmol/L isoproterenol in LPS-exposed (n=6) compared with time-matched control myocytes (n=8). Isoproterenol increased the extent of cell shortening 2.0-fold in control myocytes but only 1.4-fold in LPS-treated cardiac myocytes.
The isoproterenol response had a similar time course in the two groups. After restarting stimulation of myocytes, cell shortening increased within 1 to 2 minutes, remained stable over the next 7 minutes, and then gradually decreased. Thus, in all subsequent protocols, the effects of isoproterenol were measured between 3 to 10 minutes after restarting cell stimulation, when steady-state contractions were maintained in both cell groups.
The dose-response relation to isoproterenol was compared between control and LPS-treated cardiac myocytes after a 6-hour incubation (n=11 to 14 at each dose). The dose-response data were fit by a sigmoidal curve with R2=.997 in control cardiac myocytes and R2=.990 in LPS-treated cardiac myocytes. Fig 4⇓ shows that 1 ng/mL LPS decreased the maximal response of percent cell shortening to isoproterenol (15.5±1.0% [LPS] versus 23.3±1.1% [control], P<.001) and increased the EC50 value (−6.64±0.46 log10 mol/L [LPS] versus −7.11±0.24 log10 mol/L [control], P<.01).
LPS Effects on Intracellular Ca2+ With β-Adrenergic Stimulation
Simultaneous cell shortening and Ca2+ transients were measured before and after 1 μmol/L isoproterenol in control and LPS-treated cardiac myocytes. One-hour exposure to LPS did not alter contractility or Ca2+ transients at baseline or in response to isoproterenol. At 1 hour, 1 μmol/L isoproterenol increased the percentage of cell shortening from 13.1±0.4% to 19.0±0.5% in LPS-treated cardiac myocytes (n=15) and from 13.2±0.3% to 19.7±0.6% in control cardiac myocytes (n=15).
In contrast, the response to isoproterenol differed after a 6-hour incubation with LPS. Representative tracings at 6 hours are shown in Fig 5⇓. At baseline (dotted lines), all measurements were similar in the control and LPS-treated cardiac myocytes. After isoproterenol (solid lines), intracellular Ca2+ increased similarly, but the increase in cell shortening was blunted in the LPS-treated cardiac myocytes. Group data for the effects of isoproterenol at 6 hours are shown in Table 1⇓. In both control and LPS-treated myocytes (n=23 in each group), isoproterenol increased intracellular Ca2+ similarly, as measured by diastolic and peak systolic fluorescence ratios, percent transient amplitude, and the integral of the Ca2+ transients. However, the contractile response to isoproterenol was significantly attenuated in LPS-treated cardiac myocytes compared with control cardiac myocytes with respect to percent cell shortening, peak rate of cell shortening (−dL/dt), and the peak rate of cell relengthening (+dL/dt) (all P<.001).
Table 2⇓ shows the effects of direct adenyl cyclase activation with 30 μmol/L forskolin in cardiac myocytes incubated for 6 hours, with and without 1 ng/mL LPS (n=13 in each group). Forskolin increased intracellular Ca2+ similarly in the two groups. However, the contractile response to forskolin was significantly blunted in LPS-treated cardiac myocytes with respect to percent cell shortening, peak −dL/dt, and peak +dL/dt (all P<.001). These findings indicate that LPS (1 ng/mL for 6 hours) attenuates the contractile response to β-adrenergic stimulation by decreasing the myofilament response to Ca2+ and not by β-adrenergic receptor uncoupling or attenuation of the Ca2+ transients.
Role of NO Signaling Pathways in β-Adrenergic Response
The role of NO signaling pathways in the contrac-tile response to isoproterenol was investigated using L-NMMA to inhibit iNOS. Cardiac myocytes were incubated for 6 hours with or without 1 ng/mL LPS, with or without 1 mmol/L L-NMMA. L-NMMA did not affect cell viability or baseline cell function before isoproterenol stimulation. Percent cell shortening and percent transient amplitude were measured before and after 1 μmol/L isoproterenol stimulation in the same cardiac myocyte (Fig 6⇓). At baseline, there were no significant differences among the four groups in the resting myocyte length or diastolic fluorescence ratio. The diastolic fluorescence ratio was 0.325±0.013 in control myocytes, 0.321±0.011 in myocytes treated for 6 hours with LPS alone, 0.314±0.012 in myocytes treated with L-NMMA alone, and 0.314±0.011 in myocytes treated with both LPS and L-NMMA. LPS depressed the contractile response to isoproterenol but not the augmentation in the Ca2+ transients. L-NMMA alone had no effect on the positive inotropic response to isoproterenol in control myocytes. However, L-NMMA restored the contractile response to isoproterenol in LPS-treated cardiac myocytes (P<.01 for LPS+L-NMMA versus LPS alone). In the presence of L-NMMA, the response to isoproterenol in LPS-treated cardiac myocytes was 96% of the control response with L-NMMA alone (P=NS).
The role of cGMP in the inotropic response to isoproterenol was investigated by using KT5823, a relatively specific inhibitor of cGMP-PK. Fig 7 shows the effects of coincubating myocytes with 1 μmol/L KT5823 for 6 hours. KT5823 at 1 μmol/L had no effect on cell viability or baseline cell function before isoproterenol stimulation. There were no significant differences among the four groups in myocyte length at rest or diastolic fluorescence ratio. The diastolic fluorescence ratio was 0.319±0.008 in control myocytes, 0.323±0.011 in myocytes treated with LPS alone, 0.332±0.011 in myocytes treated with KT5823 alone, and 0.333±0.010 in myocytes treated with both LPS and KT5823. LPS depressed the contractile response to isoproterenol but not the augmentation in the Ca2+ transients. The 1 μmol/L dose of KT5823 alone did not affect the positive inotropic response to isoproterenol in control myocytes. KT5823 restored the contractile response to isoproterenol in LPS-treated cardiac myocytes (P<.01 for LPS+KT5823 versus LPS alone), but not completely. In the presence of KT5823, the response to isoproterenol in LPS-treated cardiac myocytes was only 86% of the control response with KT5823 alone (P<.05).
The results from a higher dose of KT5823 (10 μmol/L) were identical. The 10 μmol/L dose of KT5823 did not affect the positive inotropic response to isoproterenol in control myocytes and restored the isoproterenol response in LPS-treated cardiac myocytes to 87% of the control response with KT5823 alone. These findings indicate that attenuation of the contractile response induced by 6-hour exposure to 1 ng/mL LPS is primarily mediated by activation of NO and cGMP-PK to decrease myofilament Ca2+ responsiveness.
The major finding in the present study is that cardiac myocytes are sensitive to the direct effects of low levels of LPS (1 ng/mL), comparable to those found circulating in patients with sepsis.18 LPS depresses cardiac contractility and the contractile response to β-adrenergic stimulation. LPS did not alter Ca2+ transients, indicating a decreased myofilament responsiveness to Ca2+. LPS induces cardiac depression by inducing the NO signaling pathway to increase cGMP and subsequent activation of cGMP-activated protein kinase in cardiac myocytes.
The Experimental Model
The pathological sequelae of sepsis are largely related to LPS.1,2 LPS initiates the release of several mediators, including cytokines, which can mimic many LPS effects. However, several instances indicate that specific cytokines are not requisite for the pathophysiological response to LPS.7,29–31 These results suggest that the direct effects of LPS on cardiac myocytes are relevant. Direct LPS effects are difficult to evaluate in multicellular preparations, because LPS can stimulate nonmyocyte components to release substances capable of depressing cardiac function in an endocrine or paracrine fashion.5,6 Thus, we used an isolated myocyte model to elucidate the direct effects of LPS on contractile function and its response to β-adrenergic stimulation.
The cardiac myocyte preparation is not entirely pure and may contain nonmyocyte contamination, which averaged 4% in our preparations. However, in pilot studies, there was no significant correlation between the level of nonmyocyte contamination and the degree of LPS-induced cardiac depression (r=.004, P=.99). Thus, it is unlikely that contaminating nonmyocyte cells contributed significantly to the LPS-induced cardiac depression in this model.
In order to study low (1 ng/mL) levels of LPS, we minimized the exposure of cardiac myocytes to LPS, particularly during the cell isolation procedure. Prior exposure of cells to LPS diminishes the response to secondary challenges with LPS, including the induction of iNOS.32 This phenomenon, known as LPS tolerance,20 may occur after primary exposure to LPS doses above 1 ng/mL.19 Therefore, we depyrogenated the digestive enzymes used to isolate the cardiac myocytes. This lowered the level of LPS contamination from 100 to 300 ng/mL in the original or raw enzymes to 0.3 to 0.7 ng/mL. We recently found that exposing cardiac myocytes to 100 ng/mL LPS for 1 hour (as would occur when isolating myocytes with untreated enzymes) was sufficient to induce acute LPS tolerance.22 Depyrogenation minimized acute LPS tolerance, which facilitated evaluating the direct effects of low doses of LPS on cardiac myocyte function. This novel feature in our model makes it more suitable for studying direct LPS effects than using cardiac myocytes isolated with raw enzymes that can induce LPS tolerance.
Role of NO
LPS depression of cardiac contractility and attenuation of the inotropic response to β-adrenergic stimulation were prevented when cardiac myocytes were coincubated with L-NMMA. This indicates involvement of NO signaling pathways. Cardiac myocytes possess both cNOS and iNOS.6,12,33,34 Activation of cNOS causes the release of NO within seconds to minutes,35 which can acutely attenuate the contractile response of cardiac myocytes to isoproterenol.34 In contrast, cardiac iNOS expression increases after several hours of exposure to LPS or cytokines, with a peak after 6 hours.13,33,36 Sustained release of large quantities of NO contributes to contractile dysfunction in the heart.6,12,28 In the present study, baseline function, the β-adrenergic response, and cGMP production were unaltered during the first hour of LPS exposure. Thus, cNOS activity did not play an important role in the present study. However, after 6 hours of LPS exposure, cardiac contractility and the β-adrenergic inotropic response were attenuated in association with enhanced NOS activity, as evidenced by increased production of cGMP in cardiac myocytes. These effects were blocked by L-NMMA, indicating that the effects of LPS were mediated by the induction of cardiac iNOS.
Inflammatory cytokines also induce iNOS and reduce the β-adrenergic inotropic response in rat cardiac myocytes12 by uncoupling of the β-adrenergic receptor from adenylate cyclase.10,11 However, this mechanism does not mediate LPS-induced depression of the β-adrenergic response. LPS did not alter augmentation of the Ca2+ transient by isoproterenol (Table 1⇑), suggesting that β-adrenergic stimulation of cAMP in cardiac myocytes was unimpaired. This is consistent with studies showing that LPS (>10 ng/mL) alone does not inhibit the isoproterenol-stimulated increase of cAMP in cardiac myocytes.10 Direct stimulation of adenylate cyclase with forskolin (Table 2⇑) produced a response similar to that seen with isoproterenol. Thus, LPS impaired the inotropic response to β-adrenergic stimulation by an iNOS-mediated mechanism, but at a site distal to the generation of cAMP.
Decreased Myofilament Ca2+ Responsiveness by LPS
LPS depressed baseline contractility and the inotropic response to isoproterenol without affecting Ca2+ transients (Fig 5⇑, Table 1⇑). This indicates a reduction in myofilament response to Ca2+. Isoproterenol potentiated the myofilament effects of LPS, since β-adrenergic stimulation decreases myofilament responsiveness to Ca2+ in normal cardiac myocytes.37 The effects of LPS on the myofilament response to Ca2+ were blocked by coincubating cardiac myocytes with L-NMMA, an iNOS inhibitor (Fig 6⇑). NO can mediate these effects by activating soluble guanylyl cyclase to increase cGMP in the heart. cGMP decreases myofilament responsiveness to Ca2+ by stimulating cGMP-PK in skinned cardiac fibers.16 The stable lipid-soluble analogue of cGMP, 8-bromo-cGMP, induces negative inotropic effects in intact cardiac myocytes without any significant change in Ca2+ transients through the activation of cGMP-PK.17 cGMP-PK causes phosphorylation of the inhibitory subunit of troponin (troponin I),16,38,39 which reduces the Ca2+ affinity of the Ca2+ binding subunit (troponin C).
KT5823 is a relatively specific inhibitor of cGMP-PK, with an inhibition constant Ki of 0.23 μmol/L compared with Ki of >10 μmol/L for other protein kinases (eg, cAMP-PK).40 In the present study, coincubating myocytes with 1 μmol/L KT5823 for 6 hours restored the isoproterenol contractile response to 86% of the control response (Fig 7⇓). We also examined the effects of 10 μmol/L KT5823. Although this higher dose of KT5823 more closely approached the Ki for cAMP-PK, it reduced the contractile response to isoproterenol minimally (≈5%) in control cardiac myocytes. In LPS-treated cardiac myocytes, increasing the KT5823 dose from 1 to 10 μmol/L provided no additional restoration in the isoproterenol response, which remained at 87% of the control response. This indicates that the majority, but not all, of iNOS effects on cardiac myofilaments are mediated by cGMP-PK.
NO may also contribute to LPS-induced contractile dysfunction of cardiac myocytes by cGMP-independent effects.6 NO can cause the production of free radicals,41,42 which can denature myofibrillar proteins in cardiac myocytes. NO can inhibit the mitochondrial respiratory enzyme or ADP-ribosylation of glycolytic enzyme.35,43 These actions affect intracellular metabolism13 and contribute to contractile dysfunction.44
LPS may decrease the extent of cell shortening without changing Ca2+ transients by several mechanisms. LPS may lead to a decrease in Ca2+ binding to troponin C, a decrease in crossbridge turnover kinetics (decreased rate constant for transition of crossbridges from nonforce to force-generating states), and/or a decrease in force generated by individual crossbridges. Further investigations are required to determine which of these mechanisms is responsible for LPS-induced decreased myofilament responsiveness to Ca2+.
Patients with sepsis and septic shock develop myocardial dysfunction and decreased systemic vascular resistance. Increasing cardiac output in response to β-adrenergic activation provides an important compensatory mechanism to prevent hypotension and maintain adequate tissue perfusion. However, the myocardial response to β-adrenergic stimulation is impaired in patients with sepsis.9 We demonstrated that in cardiac myocytes exposed to LPS for 6 hours, baseline cell shortening was reduced by ≈10% and the maximal response to isoproterenol was reduced by >50%. Thus, LPS depresses cardiac function and inhibits normal compensatory mechanisms, which may contribute to the hemodynamic deterioration in sepsis and the development of septic shock.
Sepsis is associated with the release of several cytokines that may depress myocardial function in a paracrine or humoral manner.5 Although both LPS and cytokines attenuate the β-adrenergic response by inducing iNOS in cardiac myocytes, this occurs by different mechanisms. Cytokines may impair cardiac function by the uncoupling of β-adrenergic receptors,10–12 whereas LPS primarily depresses the myofilament response to Ca2+. These findings indicate that cytokines and LPS have unique effects on cardiac myocyte function, even though both involve iNOS-mediated pathways. Effective blockade of exogenous mediators (eg, cytokines) will not prevent LPS-induced cardiac depression related to the direct LPS effects.
The results of the present study indicate the importance of the activation of endogenous NOS in cardiac myocytes by clinically relevant levels of LPS. Since the intracellular distribution and local subcellular concentrations of NO may vary, the effects of LPS-induced endogenous NO may differ from exogenous NO produced by nonmyocyte cells or provided by pharmacological donors. We found that LPS alone can induce endogenous NO to decrease the cardiac myofilament response to Ca2+ by activation of cGMP-PK. Thus, additional therapeutic targets should be considered to ameliorate the direct effects of LPS on contractile function. For example, Ca2+-sensitizing agents and/or drugs that inhibit cGMP-PK may prove useful for improving cardiac function in patients with sepsis.
We conclude that clinically relevant levels of LPS directly depress contractility and attenuate the β-adrenergic response in cardiac myocytes. The direct cardiac effects of LPS may contribute to cardiac depression and hypotension in sepsis and septic shock.
Selected Abbreviations and Acronyms
|cGMP-PK||=||cGMP-dependent protein kinase|
|cNOS, iNOS||=||constitutive and inducible NOS|
|TNF-α||=||tumor necrosis factor-α|
This study was supported by the Office of Research and Development, Medical Research Service of the Department of Veterans Affairs, and Grant-in-Aid from the American Heart Association (No. 94–703). This work was performed during the tenure of a Grant-in-Aid from the American Heart Association and Astra-Merck, Inc (No. 9650584N) and during the tenure of an Established Investigatorship from the American Heart Association (Dr Lew). This work was done with support from the Heiwa-Nakajima Foundation and the Japanese Heart Foundation (Dr Yasuda). We thank Dr Evelyn Bayna for critical comments on the studies, Gary Firestone (deceased), Erminia Dalle Molle, and Maureen Lee for technical assistance, and Dr Yutaka Kagaya for technical advice on the Ca2+ transient study. We also thank Professor Kunio Shirato, Tohoku University School of Medicine, for providing Dr Yasuda with the opportunity to conduct research at the University of California, San Diego.
Reprint requests to Wilbur Y.W. Lew, MD, Cardiology Section 9111A, VA San Diego Healthcare System, 3350 La Jolla Village Dr, San Diego, CA 92161.
- Received March 11, 1997.
- Accepted September 2, 1997.
- © 1997 American Heart Association, Inc.
Bone RC. The pathogenesis of sepsis. Ann Intern Med. 1991;115:457–469.
Hung J, Lew WYW. Temporal sequence of endotoxin-induced systolic and diastolic myocardial depression in rabbits. Am J Physiol. 1993;265:H810–H819.
Kelly RA, Balligand J-L, Smith TW. Nitric oxide and cardiac function. Circ Res. 1996;79:363–380.
Nishikawa Y, Lew WYW. Serum tumor necrosis factor-α does not mediate endotoxin-induced myocardial depression in rabbits. Am J Physiol. 1996;270:H485–H491.
Tao S, McKenna TM. In vitro endotoxin exposure induces contractile dysfunction in adult rat cardiac myocytes. Am J Physiol. 1994;267:H1745–H1752.
Gulick T, Chung MK, Pieper SJ, Lange LG, Schreiner GF. Interleukin 1 and tumor necrosis factor inhibit cardiac myocyte beta-adrenergic responsiveness. Proc Natl Acad Sci U S A. 1989;86:6753–6757.
Chung MK, Gulick TS, Rotondo RE, Schreiner GF, Lange LG. Mechanism of cytokine inhibition of β-adrenergic agonist stimulation of cyclic AMP in rat cardiac myocytes: impairment of signal transduction. Circ Res. 1990;67:753–763.
Balligand JL, Ungureanu D, Kelly RA, Kobzik L, Pimental D, Michel T, Smith TW. Abnormal contractile function due to induction of nitric oxide synthesis in rat cardiac myocytes follows exposure to activated macrophage-conditioned medium. J Clin Invest. 1993;91:2314–2319.
Luss H, Watkins SC, Freeswick PD, Imro AK, Nussler AK, Billiar TR, Simmons RL, del Nido PJ, McGowan FX Jr. Characterization of inducible nitric oxide synthase expression in endotoxemic rat cardiac myocytes in vivo and following cytokine exposure in vitro. J Mol Cell Cardiol. 1995;27:2015–2029.
LaPointe MC, Sitkins JR. Mechanisms of interleukin-1beta regulation of nitric oxide synthase in cardiac myocytes. Hypertension. 1996;27:709–714.
Hung J, Lew WYW. Cellular mechanisms of endotoxin-induced myocardial depression in rabbits. Circ Res. 1993;73:125–134.
Shah AM, Spurgeon HA, Sollott SJ, Talo A, Lakatta EG. 8-Bromo-cGMP reduces the myofilament response to Ca2+ in intact cardiac myocytes. Circ Res. 1994;74:970–978.
Virca GD, Kim SY, Glaser KB, Ulevitch RJ. Lipopolysaccharide induces hyporesponsiveness to its own action in RAW 264.7 cells. J Biol Chem. 1989;264:21951–21956.
Cheung JY, Tillotson DL, Yelamarty RV, Scaduto RC. Cytosolic free calcium concentration in individual cardiac myocytes in primary culture. Am J Physiol. 1989;256:C1120–C1130.
Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem. 1985;260:3440–3450.
Altschuld RA, Wenger WC, Lamka KG, Kindig OR, Capen CC, Mizuhira V, Vander Heide RS, Brierley GP. Structural and functional properties of adult rat heart myocytes lysed with digitonin. J Biol Chem. 1985;260:14325–14334.
Brady AJ, Poole-Wilson PA, Harding SE, Warren JB. Nitric oxide production within cardiac myocytes reduces their contractility in endotoxemia. Am J Physiol. 1992;263:H1963–H1966.
Eskandari MK, Bolgos G, Miller C, Nguyen DT, DeForge LE, Remick DG. Anti-tumor necrosis factor antibody therapy fails to prevent lethality after cecal ligation and puncture or endotoxemia. J Immunol. 1992;148:2724–2730.
Rothe J, Lesslauer W, Lotscher H, Lang Y, Koebel P, Kontgen F, Althage A, Zinkernagel R, Steinmetz M, Bluethmann H. Mice lacking the tumour necrosis factor receptor 1 are resistant to TNF-mediated toxicity but highly susceptible to infection by Listeria monocytogenes. Nature. 1993;364:798–802.
Fantuzzi G, Zheng H, Faggioni R, Benigni F, Ghezzi P, Sipe JD, Shaw AR, Dinarello CA. Effect of endotoxin in IL-1 beta-deficient mice. J Immunol. 1996;157:291–296.
Szabo C, Thiemermann C, Wu CC, Perretti M, Vane JR. Attenuation of the induction of nitric oxide synthase by endogenous glucocorticoids accounts for endotoxin tolerance in vivo. Proc Natl Acad Sci U S A.. 1994;91:271–275.
Balligand JL, Ungureanu-Longrois D, Simmons WW, Pimental D, Malinski TA, Kapturczak M, Taha Z, Lowenstein CJ, Davidoff AJ, Kelly RA. Cytokine-inducible nitric oxide synthase (iNOS) expression in cardiac myocytes: characterization and regulation of iNOS expression and detection of iNOS activity in single cardiac myocytes in vitro. J Biol Chem. 1994;269:27580–27588.
Balligand JL, Kobzik L, Han X, Kaye DM, Belhassen L, O’Hara DS, Kelly RA, Smith TW, Michel T. Nitric oxide-dependent parasympathetic signaling is due to activation of constitutive endothelial (type III) nitric oxide synthase in cardiac myocytes. J Biol Chem. 1995;270:14582–14586.
Gambassi G, Spurgeon HA, Lakatta EG, Blank PS, Capogrossi MC. Different effects of α- and β-adrenergic stimulation on cytosolic pH and myofilament responsiveness to Ca2+ in cardiac myocytes. Circ Res. 1992;71:870–882.
Blumenthal DK, Stull JT, Gill GN. Phosphorylation of cardiac troponin by guanosine 3′:5′-monophosphate-dependent protein kinase. J Biol Chem. 1978;253:324–326.
Lincoln TM, Corbin JD. Purified cyclic GMP-dependent protein kinase catalyzes the phosphorylation of cardiac troponin inhibitory subunit (TN-1). J Biol Chem. 1978;253:337–339.
Beckman JS, Beckman TW, Chen J, Marshall PA, Freeman BA. Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide. Proc Natl Acad Sci U S A.. 1990;87:1620–1624.
Stamler JS, Simon DI, Osborne JA, Mullins ME, Jaraki O, Michel T, Singel DJ, Loscalzo J. S-Nitrosylation of proteins with nitric oxide: synthesis and characterization of biologically active compounds. Proc Natl Acad Sci U S A. 1992;89:444–448.
Depre C, Vanoverschelde JL, Goudemant JF, Mottet I, Hue L. Protection against ischemic injury by nonvasoactive concentrations of nitric oxide synthase inhibitors in the perfused rabbit heart. Circulation. 1995;92:1911–1918