|Home | About | Journals | Submit | Contact Us | Français|
There is accumulating evidence during sepsis that cardiomyocyte (CM) homeostasis is compromised, resulting in cardiac dysfunction. An important role for complement in these outcomes is now demonstrated. Addition of C5a to electrically-paced CMs caused prolonged elevations of [Ca2+]i during diastole, together with appearance of spontaneous Ca2+ transients. In polymicrobial sepsis in mice, we found that three key homeostasis-regulating proteins in CMs were reduced in amounts: Na+/K+ ATPase, which is vital for effective action potentials in CMs; and two [Ca2+]i regulatory proteins; SERCA2 and the Na+/Ca2+ exchanger (NCX). Sepsis caused reduced mRNA levels and reductions in protein concentrations in CMs for all three proteins. The absence of either C5a receptor mitigated sepsis-induced reductions in the three regulatory proteins. Absence of either C5a receptor (C5aR1, C5aR2) diminished development of defective systolic and diastolic ECHO/Doppler parameters developing in the heart (cardiac output, LV stroke volume, isovolumic relaxation, E’sa, E/E’sa, LV diastolic volume). We also found in CMs from septic mice the presence of defective current densities for Ik1, L-type calcium channel and NCX. These defects were accentuated in the co-presence of C5a. These data suggest complement-related mechanisms responsible for development of cardiac dysfunction during sepsis.
Myocardial dysfunction in sepsis is correlated with high mortality rates . Plasma mediators elevated in sepsis include interleukin (IL)-1β, IL-6, tumor necrosis factor (TNF) and the complement anaphylatoxin, C5a, all of which have cardio-suppressive effects causing in vitro defects in contraction and relaxation of CMs (reviewed [2, 3]). Absence or blockade of either C5a receptor (C5aR1, C5aR2) substantially improved survival after CLP . In rat CMs, mRNA and proteins for C5a receptors were constitutively expressed in normal CMs and overexpressed after CLP. In vitro addition of C5a to rat CMs caused dysfunction involving both contractility and relaxation as well as release of proinflammatory cytokines such as IL-1β, IL-6, and TNF .
Maintenance of physiological levels of [Ca2+]i in the cytosol of CMs is regulated by phosphorylation and dephosphorylation of key proteins. Ca2+ entry into CMs via the voltage gated L-type calcium channel (ICa,L) activates Ca2+ release from the sarcoplasmic reticulum (SR) via the ryanodine receptor 2 (RyR2) during systole. Cytoplasmic clearance of Ca2+ during diastole occurs after activation of the sarcoplasmic/endoplasmic reticulum calcium ATPase 2 (SERCA2) and the Na+/Ca2+ exchanger (NCX) on T-tubules and on the sarcolemma. SERCA2a in humans mediates 70% of cytosolic Ca2+ clearance [6, 7]. In the failing heart, SERCA activity is reduced . Impairment of Ca2+ transport mechanisms in the myocardium occurs during sepsis . Numerous kinases and ion channel currents in CMs are activated by Ca2+/calmodulin-dependent protein kinase II (CaMKII) . Autophosphorylation of Thr286 is an essential activation response for this kinase in vivo . RyRs can also be activated by CaMKII, ultimately enhancing the contractile apparatus of CMs . In addition, CaMKII also regulates SR Ca2+ uptake by phosphorylation of phospholamban (PLN) , resulting in release of SERCA inhibition. Modulation of Ca2+ regulating proteins may have important functional consequences during heart failure, arrhythmias and sepsis.
Reactive oxygen species (ROS) are increased in the failing heart, being associated with reduced left ventricular function . After CM exposure to TNF, mitochondrial ROS production occurs . ROS modulate Ca2+ regulating proteins such as RyR by oxidizing the regulatory subunit, causing destabilization of the closed state of the channel, resulting in [Ca2+]i increases in the cytosol . ROS also affects ATP-binding sites on SERCA2, which may result in cytosolic Ca2+ overload . The current report describes the ability of C5a in vitro to cause defective clearance of Ca2+ in paced CMs. We describe during sepsis reduced levels of mRNAs and proteins involved in [Ca2+]i homeostasis (SERCA2 and NCX) as well as reductions in Na+/K+-ATPase, which directly regulates intracellular Na+ levels and the resting membrane potential (RMP) in CMs, and indirectly regulates Ca2+ levels by proximally controlling Ca2+ flux through NCX . Collectively, the data provide information that directly link complement to the loss of homeostatic control of [Ca2+]i in CMs during sepsis as well as the faulty contraction and relaxation responses in CMs during sepsis.
There is abundant evidence for consumptive depletion of complement during sepsis both in rodents  and in humans . Initially, we developed a neutralizing rabbit antibody to rat and mouse C5a, which was found to be highly protective, producing improved survival [19, 20] and reduced evidence of multiorgan damage in CLP rodents . We found that antibodies directed against the midregion of C5a or the C terminal region of C5a  conferred substantial protection against the harmful effects of sepsis . Our studies have also shown that both C5aR1 and C5aR2 were involved in the harmful outcomes of sepsis [4, 22]. In the current study we have focused on the relationship between defective cardiac function during sepsis and the roles of C5a and its receptors being linked to loss of homeostasis in CMs.
All procedures conformed to the Guide of Care and Use of Laboratory Animals published by the US National Institutes of Health. The study was approved by the University Committee on Use and Care of Animals (UCUCA) and performed according to appropriate guidelines. Specific pathogen-free male Sprague Dawley rats (300–350g; Harlan Laboratories, Indianapolis, IN) and male C57BL/6 mice (6–10 wk, 25–30 g; Jackson Laboratories, Bar Harbor, ME) were anesthetized with the combination of ketamine (Hospira, Lake Forest, IL) and xylazine (Lloyd Laboratories, Shenandoah, IA) i.p. Some experiments were also carried out, using male C57BL/6 mice from our C5aR1−/− and C5aR2−/− breeding colonies at the University of Michigan. The generation of C5aR1−/− and C5aR2−/− mice has been described previously [23, 24].
Sepsis was induced by the CLP procedure as described previously in both rats and mice [25, 26]. Sham animals underwent the same procedure, including manipulation of the bowel in the absence of cecal ligation and puncture. The animals were euthanized 8, 12, 16, 24 or 48 hr after CLP.
The isolation of adult rat and mouse CMs was performed as previously described, using a Langendorff perfusion system [27–29]. The hearts were perfused through a cannula inserted into the ascending aorta using the Langendorff apparatus. After perfusing using the rinsing solution, the hearts were retrograde perfused with enzyme solution (Liberase, Hoffmann-La Roche, Mannheim, Germany), according to manufacturer’s directions. Following digestion, the heart was detached from the Langendorff apparatus, atria and vessels were removed, and the ventricles were cut into small pieces, which were gently triturated with a plastic transfer pipette. After isolation of the CMs, the Ca2+ concentration in the buffer fluid was gradually (in 6 steps) increased (to 1.8 mM) and the cells were cultured in M199 medium with 1% Insulin-Transferrin-Selenium-X (Gibco®, Life Technologies, Carlsbad, CA) and antibiotic-antimycotic (Life Technologies, Carlsbad, CA).
Whole-cell patch clamp procedures were used in rat CMs to measure the ICa,L, IK1 and NCX currents. For ICa,L, the bath solution contained (in mM): TEA-Cl 136, CsCl 5.4, MgCl2 0.8, KCL 4, CaCl2 2, HEPES 10, glucose 10 (pH 7.4 with CsOH). The pipette solution contained (in mM): CsCl 20, Cs-aspartate 110, MgCl2 1, MgATP 5, GTP 0.1, Na2Phosphocreatine 5, EGTA 10, HEPES 10, (pH 7.2, with CsOH). For IK1 the bath solution contained (in mM): NaCl 148, KCl 5.4, MgCl2 1, CaCl2 1.8, NaH2PO4 0.4, HEPES 15, Glucose 5.5, pH 7.4 with NaOH. The pipette solution contained (in mM): KCl 20, K-aspartate 90, KH2PO4 10, EDTA 5.0, K2ATP 1.9, HEPES 5.0 and Mg2+ 7.9; pH 7.2 (KOH). Nifedipine (5µM) was included in the external solution to eliminate calcium current. Barium-sensitive currents were eliminated by adding 1 mM BaCl2 to the external solution. For NCX  the bath solution contained (in mM/L): NaCl 145, MgCl2 1, HEPES 5, CaCl2 2, CsCl 5, and glucose 10 (pH 7.4, adjusted with NaOH). Ouabain (0.02 mM/L) and nifedipine (0.01 mM/L) were added to the solution. The pipette solution contained (in mM/L): CsCl 65, NaCl 20, Na2ATP 5, CaCl2 6, MgCl2 4, HEPES 10, tetraethylammonium chloride 20, EGTA 21, and ryanodine 0.05 (pH 7.2, adjusted with CsOH).
IK1 recordings were carried out at 37°C while the other currents were measured at room temperature using a MultiClamp 700B amplifier (Axon Instruments, Forest City, CA). After gigaseal formation and patch break, the junction potential was nulled and cell capacitive currents and series resistance were optimally (~80%) compensated.
Action potentials were recorded from individual myocytes using the current-clamp mode of the MultiClamp 700B amplifier after gigaseal formation and patch break. Stimulus pulses (1–2 ms duration) were generated using a World Precision Instruments DS8000 stimulator (Sarasota, FL). The bath solution contained (in mM): NaCl 148, KCl 5.4, MgCl2 1, CaCl2 1.8, NaH2PO4 0.4, HEPES 15, Glucose 5.5, pH 7.4 with NaOH. The pipette solution contained (in mM): KCl 20, K-aspartate 90, KH2PO4 10, EDTA 5.0, K2ATP 1.9, HEPES 5.0 and Mg2+ 7.9; pH 7.2 (KOH). Recordings were carried out at 37°C.
Using intracellular antibody labeling by flow cytometry, CMs were fixed for 30 minutes in 0.25% paraformaldehyde (Sigma-Aldrich, St. Louis, MO) at 4°C. The cells were permeabilized for 60 minutes with 1% saponin (Sigma-Aldrich) at room temperature. Incubation with the indicated antibody was carried out in the presence of 0.1% saponin for 30 minutes in the dark at room temperature.
Intracellular Ca2+ in CMs was measured by flow cytometry using fluo-3 AM (4 µm) labeled CMs.
For intracellular Ca2+ recordings, isolated rat CMs were loaded with fluo-3AM (4 µM/L, 10 min) and images were recorded using epi-fluorescent or confocal microscopy (Nikon A1R). Recordings were performed on a stage heated to 37°C.
Na+/K+-ATPase activity in whole mouse heart homogenates was determined by a fluorometric method described by Huang and Askari , modified by Fraser and McKenna  and Barr et al . Briefly, homogenates were prepared in the assay medium (in which 3-O-MFPase activity was measured) contained 250 mM sucrose, 2 mM EDTA, 1.25 mM EGTA, 5 mM NaN3, and 10 mM Tris (pH 7.4). Homogenates were freeze-thawed four times and diluted 1:5 in cold homogenate buffer. Protein lysates were incubated for 10 min at 37°C in a buffer containing 5 mM MgCl2, 1.25 mM EDTA, 100 mM Tris base (pH 7.4), 1 mM EGTA, and 5 mM NaN3, with or without 6 mM Ouabain (Sigma-Aldrich, St. Louis, MO). Protein content of the homogenate was determined using BCA (bicinchoninic acid) Protein Assay Kit from Thermo Scientific Pierce (Rockford, IL). To initiate the reaction, 160 µM 3-O-methylfluorescein phosphate (3-O-MFP) (Sigma-Aldrich) was added to lysates followed by incubation at 37°C for 1 min. Activated Na+/K+-ATPase hydrolyzes 3-O-MFP and forms a fluorescent compound 3-O-MFP. Then 10 mM KCl was added, as the optimal K+-stimulating concentration for 3-O-MFP, and fluorescence was recorded using the Infinite® 200 PRO multimode microplate reader (Tecan Group Ltd, Switzerland). The excitation and emission wavelengths were 485 and 535 nm, respectively. The amplitude of the emission was shown to be proportional to the concentration of 3-O-MFP, and a standard curve was created using varying concentration of 3-O-MF. To calculate the K+-stimulated 3-O-MFPase activity, the non-specific ATPase activity and the spontaneous hydrolysis present after addition of 10 mM KCl were subtracted from the activity obtained before addition as previously outlined [32, 34]. Specific Na+/K+-ATPase activity was determined by subtracting K+-dependent 3-O-MFPase activity without ouabain from the activity with ouabain and expressed as micromoles (or nanomoles) of liberated phosphate (Pi) per minute per milligram protein.
Western blot analysis was used to quantitate α1-subunit protein levels of the Na+/K+-ATPase. For homogenate preparation, frozen tissue was homogenized in a buffer containing 250 mM sucrose, 2 mM EDTA, 1.25 mM EGTA, 5 mM NaN3, and 10 mM Tris (pH 7.40). Equal amounts of homogenates (40 µg) were electrophoresed. Gels were electrophoretically transferred to nitrocellulose membrane (Bio-Rad Laboratories, Philadelphia, PA). The nonspecific binding sites were blocked 5% (wt/vol) nonfat dry milk in Tris-buffered saline (TBS, pH 7.5) with 0.1% Tween 20 (vol/vol), for 1 hr at room temperature before incubation. After washing the membrane, primary antibody [anti-alpha 1 Sodium Potassium ATPase antibody (ab7671, Abcam, Massachusetts)] (diluted in 0.1% TBS Tween 20 (TBS-T) with 5% BSA) was added at a dilution of 1:2000 and incubated overnight. After washing (3× 10 min) in TBS-T, peroxidase-labeled secondary antibody (donkey anti-mouse; Jackson ImmunoResearch), with dilution of 1:10,000 in TBS-T and 5% nonfat dry milk, was added and allowed to incubate for 1 hr, followed by rinsing in TBS-T. The final detection was done using the West Femto substrate (Thermo Scientific Pierce).
For confocal imaging, CMs were isolated from a healthy control mouse and 8 hr after CLP according to the method we described before [5, 29]. CMs were plated on sterile glass coverslips coated with natural mouse Laminin (Life Technologies). The CMs then fixed using 4% formaldehyde before staining to prevent entering the first antibody into the cytosol, in this way we could predominantly detect the extracellular part of the Na+/K+-ATPase plasma membrane. Cells were incubated with anti-alpha 1 Sodium Potassium ATPase antibody (ab7671, Abcam) for 30 minutes and then washed in PBS. Goat anti-mouse IgG-AlexaFluor488 (Jackson Immunoresearch) was considered as the secondary antibody for immunofluorescent labeling. CMs were mounted on slide with ProLong Gold anti-fade reagent containing DAPI (Life Technologies). Digital monochromatic images were acquired on a Zeiss LSM 510 Confocal microscope (Zeiss USA).
Mouse hearts were obtained 0, 8, 12, 16, 24 and 48 hr after CLP or sham procedure. Total RNA was isolated from mouse heart homogenates by TRIZOL® method (Life Technologies) according to manufacturer’s instructions. cDNA was then obtained and amplified (SYBR®) using reagents from Life Technologies. Amplification was performed on a 7500 Real-Time PCR System (Applied Biosystems, Carlsbad, CA). Calculation of the relative quantitative results was done with the 2−ΔΔCt algorithm. The following primers were used: mouse NCX (Na+/Ca2+-exchanger) 5′ CCTTGTGCATCTTAGCAATG 3′ (forward), mouse NCX 5′ TCTCACTCATCTCCACCAGA 3′ (reverse); mouse Atp2a2 (SERCA2) 5′ CTGTGGAGACCCTTGGTTGTAC 3′ (forward), mouse Atp2a2 (SERCA2) 5′ CAGAGCACAGATGGTGGCTAAC 3′ (reverse). For housekeeping gene mouse HPRT (hypoxanthine phosphoribosyltransferase) 5′ CGAGGAGTCCTGTTGATGTTGC 3′ (forward), 5′ CTGGCCTATAGGCTCATAGTGC 3′ (reverse) were used.
Mouse hearts were obtained 8, 12, 16, 24 and 48 hour after CLP and homogenized and lysed by using 1x RIPA Lysis Buffer (EMD Millipore) containing complete Mini-protease inhibitor and PhosSTOP protease inhibitor cocktail (Roche). Protein concentrations were determined in homogenates using Pierce® BCA Protein Assay Kit (Thermo Scientific). After centrifugation, pellets were resuspended in Laemmli Sample Buffer (Bio-Rad Laboratories) and boiled. After a last centrifugation step the samples were loaded under reducing conditions onto a 7.5% (NCX-1 and SERCA2) Mini-Protean®TGX™ Precast Gels (Bio-Rad Laboratories). After electrophoresis proteins were transferred by a Trans-Blot Turbo Transfer System using Mini Transfer Packs (both from Bio-Rad). The blots were blocked with 5% milk in TBS for 1 hour at RT and then incubated with antibodies (see below) overnight at 4°C. For analysis of the mouse heart homogenates, anti-SERCA2 ATPase (Abcam) and mouse anti-Na+-Ca2+-Exchanger (NaCX-1, EMD Millipore) were used. After washing, HRP-conjugated anti-mouse or anti-rabbit antibody (GE Healthcare, Buckinghamshire, UK) was used as secondary antibody (1:10,000) at RT for 1 hr followed by an additional washing step. Chemiluminescent HRP Hy Glo™ (Denville Scientific Inc, South Plainfield, NJ) was used for developing.
Mouse Na+/K+-ATPase, SERCA2 and NCX ELISA kits were purchased from Antibodies-online and levels of these proteins in heart hemogenates were quantified according to the instructions of the manufacturer. Protein levels in the heart hemogenates were determined using the bicinchoninic acid (BCA) assay. Bovine serum albumin was used as a standard. The ELISA results were expressed in pg/mg protein.
The following reagents were used: recombinant rat C5a (rrC5a) was produced in our lab as described previously [35, 36], chemicals used for preparation of solutions for CM isolation were purchased from Sigma-Aldrich.
Echocardiograms were performed as previously described . All echocardiograms were performed by a registered echocardiographer who was blinded to mouse genotype. C5aR1−/− and C5aR2−/− mice, together with Wt C57BL/6 mice were weighed and anesthetized with inhaled isoflurane. Imaging was performed using a Vevo 770 High-Resolution In Vivo Imaging System (Visualsonics Inc., Toronto, ON, Canada) equipped with a RMV-707 30 MHz RMV (Real-Time Visualization) (up to 45MHz) scanhead. Left ventricular volumes were measured from the parasternal long axis view at the level of the tips of the leaflets of the mitral valve at end systole (VolS) and end diastole (VolD) and used to calculate stroke volume (SV = VolD − VolS) and ejection fraction (EF % = endocardial SV / endocardial VolD x100). Cardiac output was calculated from stroke volume and heart rate (CO = SV x heart rate). Mitral valve E and A wave inflow velocities were sampled at the tips of the leaflets of the mitral valve from the apical four chamber view. Doppler tissue imaging was performed with acquisition of peak E’ velocity from the lateral (E’la) and septal annulus (E’sa) of the mitral valve imaged from the apical four chamber view. Isovolumic relaxation time (IVRT), from the closure of the aortic valve to the opening of the mitral valve, was measured from the apical five chamber view using Doppler flow imaging. Imaging was performed at 8 hr after CLP.
For statistical analysis GraphPad Prism version 6 (GraphPad Software, La Jolla, CA) was used. All values are expressed as means ± SEM. Data were analyzed by two-tailed Student t test, paired t test or one-way ANOVA, as appropriate. Differences were considered significant when p ≤ 0.05.
CMs were electrically paced and, where indicated, exposed to rrC5a at 37°C. Effects of C5a presence on single CM Ca2+ transients were quantified using fluorescence microscopy techniques as previously reported . Figure 1A shows the fluorescent image of an adult rat CMs loaded with the fluorescent calcium indicator, fluo-4AM. White areas along the red line represent cytosolic Ca2+. CMs were electrically stimulated (0.5 Hz, 30 V, the field stimulation being denoted by the short vertical black lines in frames B-G) in the absence (frames B and C) or presence of recombinant C5a (45 ng/ml) after 2 min (frames D and E) and after 5 min (frames F and G). The selection of 45 ng rrC5a/ml was based on a dose response profile (EC50) of responding rat or mouse CMs, with the values being 10 µM, (data not shown). C5a-induced changes in [Ca2+]i-related fluorescence were recorded using a high speed CCD camera (200 fps) mounted on an inverted microscope (Nikon, Ti). Frames B, D and F depict the changes in fluorescence over time along the red line shown in frame A, commonly referred to as “time-space plots” (time on the x-axis and distance on the y-axis). Frames C, E and G are graphical representations of the corresponding time-space plots (frames B, D and F). Arrows (frames D-G) revealed prolonged diastolic elevations in [Ca2+]i after systole and also showed spontaneous Ca2+ transients (frame G, arrow). This is the first demonstration that C5a receptor ligation with C5a alters CM intracellular Ca2+ homeostasis, leading to elevated cytosolic Ca2+ during diastole, together with spontaneous Ca2+ release events and a buildup of [Ca2+]i.
The density-voltage relationship in Supplemental Figure 1A summarizes the effects of C5a on the normal repolarizing inward rectifier potassium current (IK1) in CMs, using patch clamp technology. Consistent with the reduced RMP and the prolongation of action potential duration (APD), the peak outward IK1 current density was reduced at voltages between −80 and −50 mV following C5a application to CMs (Supplemental Figure 1A, inset). C5a also reduced peak inward L-type calcium current (LTCC), ICa-L, by approximately 34% (p < 0.05, Supplemental Figure 1B). Defects in L-type Ca2+ currents in CMs have been described during endotoxemia and in CMs exposed to IL-1β . Supplemental Figure 1C shows that C5a markedly reduced the NCX current in non-septic CMs by nearly 47% (Supplemental Figure 1C, white versus black bars), while septic CMs showed a greatly diminished NCX current (nearly 59%) (white bars, absence of C5a). In the co-presence of C5a, the NCX current in septic CMs fell even further.
Collectively, these results indicate that C5a induces a variety of electrophysiological defects related to impaired clearance of cytosolic Ca2+ during diastole, resulting in reduction of sarcolemmal Ca2+ extrusion currents. These modifications can lead to heart dysfunction and arrhythmogenic outcomes of sepsis.
For these studies, CMs were obtained from nonseptic and septic rats or mice (24 hr after CLP). CMs were evaluated by patch-clamp techniques. In frame A, representative action potentials in CMs were induced in CMs before and after exposure to rrC5a (45 ng/ml). In the presence of buffer, the RMP values were approximately −76 mv in CMs. In the presence of C5a, the values fell to −69.5 mv and returned to −74 mv after washout of C5a. In frame B, CMs from rats 24 hr after CLP showed early after depolarizations (EADs) (frame B, upper region) and widened and prolonged action potentials. Addition of C5a (45ng/ml) accentuated the abnormalities (lower frames) accentuating the EAD profile. In frame C, RMP (mV) values were measured in the presence of buffer or C5a as determined by patch-clamp technology. CMs from septic and nonseptic rodents had the same RPM values (white bars), whereas addition of C5a caused an 11% reduction in RMP values for nonseptic CMs (frame C). In the case of septic (24 hr CLP) CMs, addition of C5a (black bars) caused an 18% reduction in RMP values (frame C). In frame D, APD 90% values for repolarization suggested a modest msec increase (22%) in nonseptic CMs exposed to C5a but, when septic CMs were employed, the differences between buffer and C5a exposed CMs were dramatically increased (3 fold), indicating profound prolongation in action potentials as a result of sepsis together with presence of C5a (frame D). When maximum upstroke velocity (dv/dt) was evaluated (frame E), CMs were obtained from nonseptic mice (white bars) or from CLP (24 hr) hearts (black bars). In nonseptic CMs, addition of C5a reduced maximum upstroke velocity by 53%, but CMs from CLP hearts exposed to C5a showed an 85% reduction in upstroke velocity (frame E). Accordingly, exposure of CMs from septic hearts to C5a results in dramatically altered action potentials, with defective values for RMP, maximal upstroke velocities, APD prolongation and the appearance of EADs.
We employed both Western blot analysis and an enzymatic assay for measuring Na+/K+-ATPase activity using various protocols for left ventricular (LV) homogenates. The results are shown in Figure 3. ATPase activity was defined as potassium-elicited and ouabain-inhibitable release of phosphate (micro or nano moles/mg protein/min). Frame A shows a standard activity curve in the presence of increasing concentrations of 3-O-MFP. In frame B, Na+/K+-ATPase activity progressively increased as a function of time (30–360 sec, solid line). When Na+/K+ activity was measured in LV homogenates from mice obtained either 8 hr or 18 hr after CLP, there were profound reductions in enzyme activity. In frame C, enzymatic activity in LV homogenates 360 sec after addition of substrate revealed striking reductions (75–85%) in ATPase activity both 8 hr and 18 hr after CLP. Frame D shows the levels of Na+/K+-ATPase enzymatic activity in homogenates from Wt mouse LV (8 hr after CLP) and in LV from Wt, C5aR1 or C5aR2 KO mice. In LV homogenates from Wt mice after CLP, enzyme activity was reduced by nearly 70% (first black bar). In the absence of C5aR1, there was almost no reduction in Na+/K+-ATPase activity after CLP in C5aR1−/− mice. In C5aR2−/− mice, there appeared to be less protection against the CLP-induced fall in Na+/K+-ATPase, whereas in C5aR1−/− mice the sepsis-induced fall was virtually abolished. Frame E shows Western blots from LV homogenates of Wt (control) mice and from mice 8 hr after CLP. For each Western blot, there were 2 separate samples (from two Wt mice) before or 8 hr after CLP. It is clear that, 8 hr after CLP, profound reductions in intensity of LV blots for Na+/K+-ATPase (β chain) occurred after CLP, which is consistent with reductions in enzymatic activity [Na+/K+ ATPase (frames B, C)]. The intensity of staining for GAPDH in frame E suggested roughly equivalent loading of samples. Frame F represents immunostaining of isolated control CMs for Na+/K+-ATPase, and CMs 8 hr after CLP (frame G). Consistent with the Western blots (frame E), the intensity of the Na+/K+-ATPase on the surfaces of CMs before (frame F) and after CLP (frame G) was substantially diminished after onset of sepsis. Together with the diminished IK1 (Supplemental Figure 1), the reduced Na+/K+-ATPase activity in CLP CMs (Figure 3) likely contributed to intracellular sodium accumulation and reduction in RMPs (Figure 2).
Based on defective clearance of cytosolic [Ca2+]i after CM exposure to C5a (Figure 1), we investigated the levels of SERCA2 and NCX in LV homogenates, from CLP Wt mice and in LV homogenates from mice lacking C5a receptors. These two enzymes are required for maintaining physiological levels of cytosolic [Ca2+]i . After CLP in Wt mice, there was a time-dependent reduction in mRNAs for both SERCA2 (A) and for NCX (B) in LV homogenates from Wt mice. When Western blots were obtained at different time points after CLP, there were substantial reductions in LV homogenate proteins for both SERCA2 (frame C) and NCX proteins (frame D) 16 hr after CLP. Wt mice 16 hr after CLP had approximately a 70% drop in mRNA for SERCA2 and NCX in LV homogenates. Reductions in SERCA2 were dramatically attenuated in mice lacking either C5aR1 or C5aR2 (frame E), indicating that CLP-induced reductions in these Ca2+ regulatory proteins were C5a receptor-dependent (similar data have been obtained for mRNA for NCX [data not shown]).
To determine whether C5aR deficiency prevented loss of Na+/K+-ATPase, SERCA2 and NCX protein levels, additional experiments by ELISA were done on heart homogenates to measure protein levels quantitatively. Based on the western blot data in Figure 4, time point of 16 hours after CLP, in which the lowest levels of the reduction for SERCA2 and NCX was observed, was selected for measuring protein levels in heart homogenates. As shown in Figure 5, there were significant reductions in heart homogenate proteins in Wt for Na+/K+-ATPase (frame A), SERCA2 (frame B) and NCX proteins (frame C) 16 hr after CLP. Wt mice after CLP had a 43% drop in protein levels for Na+/K+-ATPase, 24% for SERCA2 and 55% for NCX in heart homogenates compared to sham controls (p<0.05 for all three proteins). Reductions in these proteins were dramatically attenuated in mice lacking either C5aR1 or C5aR2 (frames A-C), indicating that CLP-induced reductions in these proteins were C5a receptor-dependent (similar data have been obtained for mRNA and western blot in Figures 3 and and44).
In Wt mice, a marked decrease (53%) in calculated cardiac output (using echocardiographic/Doppler imaging) occurred 8 hr after CLP, compared to sham controls (cardiac output: 6.4 ml/min [Wt CLP] vs. 13.6 ml/min [Wt sham] [Figure 6A]; p < 0.05). There was a reduction in heart rate by 8 hr (mean HR: 423 bpm [Wt CLP] vs 486 bpm [Wt sham]; p < 0.05) (data not shown), but most of the difference in cardiac output could be attributed to a reduction in LV stroke volume (11.3 µl/beat [Wt CLP] vs 29.8 µl/beat [Wt sham] [panel B, Figure 6B]; p < 0.05). LV ejection fraction was modestly enhanced in response to CLP (LV ejection fraction: 69.3% [Wt CLP mice] vs 55.2% [Wt sham]; p < 0.05) (data not shown). When CLP was performed in the C5aR1−/− and C5aR2−/− mice, minimal detrimental effects on cardiovascular performance were noted. Cardiac output (A) and LV stroke volume (B) were markedly higher in C5aR1−/− and C5aR2−/− mice when compared to Wt mice 8 hr after CLP (cardiac output: 6.4 ml/min [Wt], 12.8 ml/min [C5aR1−/−], and 13.1 ml/min [C5aR2−/−]; LV stroke volume: 11.3 µl/beat [Wt], 24.2 µl/beat [C5aR1−/−], 25.7 µl/beat [C5aR2−/−]; both p < 0.05). 8 hr after CLP, cardiac output and stroke volume in the C5aR1−/− and C5aR2−/− mice closely approximated those noted in the sham-operated controls. As in Wt control mice, CLP did not affect systolic function in the C5aR1−/− and C5aR2−/− mice as measured by the LV ejection fraction at 8 hr post-procedure (data not shown).
Diastolic function (as measured by isovolumic relaxation time and Doppler tissue imaging) were prolonged 8 hr after CLP induction in Wt but not in C5aR1−/− and C5aR2−/− mice (Figure 6C). Tissue Doppler Imaging demonstrated a reduction in annular velocities of the mitral valve at the lateral (E’la) (data not shown) and septal annulus (E’sa) (Figure 6D) suggesting reduced diastolic function after CLP that is at least partially reversed in the setting of C5aR1 or C5aR2 K.O. Correcting E wave velocities for diastolic function (E/E’ ratio) revealed that CLP resulted in a reduction of LV filling pressures, likely as the result of reduced preload in Wt animals. In C5aR1−/− and C5aR2−/− mice, the E/E’ ratios were similar to sham operated Wt controls, suggesting normalization of ventricular filling pressures which was supported by the assessment of ventricular volumes (Figure 6E). Similarly, the significantly decreased end diastolic volumes noted after CLP in Wt mice were less pronounced in the C5aR1−/− and C5aR2−/− mice (Figure 6F). These data suggest that CLP mice have reduced cardiac performance that was muted in mice lacking either C5a receptor.
We investigated the mechanisms by which sepsis and C5a disrupt calcium balance in cardiomyocytes. Our earlier studies in the CLP model of sepsis showed the appearance of C5a in plasma [19, 41] and demonstrated that in vivo neutralization of C5a greatly attenuated acquired defects in innate immunity in blood neutrophils , accompanied by markedly improved survival . Absence or blockade of either C5aR1 or C5aR2 improved survival in CLP sepsis in mice, but the best results were obtained when both C5a receptors were unavailable . In CMs isolated after CLP-sepsis, levels of C5aR mRNA and protein were increased and these CMs were more susceptible to C5a induction of defects in contractility and relaxation in vitro .
In the current study, we focused on the ability of C5a to cause changes in [Ca2+]i handling in CMs. Here, we show that C5a interfered in paced CMs with the clearance of Ca2+ during diastole. The buildup of [Ca2+]i was accompanied by spontaneous calcium transients, prolonged durations of action potentials, reductions in outward IK1 and L-type calcium currents, and reduced Na+/K+-ATPase activity and NCX currents. Concerning the buildup of [Ca2+]i in CMs during diastole, a recent review reported an association with arrhythmias and defective contractility and relaxation . As an underlying mechanisms for the elevated [Ca2+]i, in CMs impaired clearance of [Ca2+]i after CLP-sepsis was linked to significant reductions in expression of mRNA and proteins for the two [Ca2+]i regulatory proteins: NCX and SERCA2, each of which plays an important homeostatic role in regulating cytosolic [Ca2+]i in CMs . Defective Ca2+ handling in the cytosol due to reduced Ca2+ sequestration into the sarcoplasmatic reticulum (SR) has also been noted in the failing human heart . In such cases, SERCA function was reduced . Similarly, in congestive heart failure in humans the expression level of mRNA for SERCA was reduced and inversely correlated with levels of the atrial natriuretic factor mRNA level, suggesting that the expression of these proteins may be decreased in severe heart failure . In the experimental setting, SERCA2 overexpressing mice showed evidence of increased transport of Ca2+ into the SR, resulting in improved cardiac contractility and relaxation [46, 47].
Besides, reduced SERCA2 and NCX protein levels as well as mRNA enzymatic activity of Na+/K+-ATPase was reduced after CLP in hearts. It is noteworthy, that such reductions in the three proteins after CLP were C5a receptor-dependent, although it is possible that other down-stream signaling pathways were also defective. Currently, we have in vitro data suggesting that exposure of CMs to C5a results in a reduction in protein content for the three homeostatic proteins that regulate [Ca2+]i. We are currently exploring the mechanism(s) involved (data not shown). Regarding the mechanism(s) of C5a evoking increases in [Ca2+]i in CMs, it is important to note that C5a significantly down-regulated Na+/K+-ATPase protein and reduced its enzyme activity in CMs. The intracellular sodium concentration ([Na+]i) is controlled by the balance between Na+ influx through various pathways, including NCX and Na+ channels, and Na+ extrusion via the Na+/K+-ATPase. In the heart, [Na+]i is also a key modulator of Ca2+ cycling through NCX. Therefore, it is reasonable to speculate that by inhibiting the Na/K-ATPase, C5a presence results in [Na+]i overload during pacing, which would be expected to decrease post-pacing NCX Ca2+ extrusion and lead to increased [Ca2+]i. The consequence of Ca2+ overload would be oscillatory Ca2+ release from the SR which would result in arrhythmogenic outcomes after depolarizations, as it did in the experiment shown in Figure 1. This could occur even if NCX protein was reduced.
Taken together, the triad of SERCA2, NCX and Na+/K+-ATPase function seems to be reduced in experimental sepsis, which in turn may account for buildup of [Ca2+]i and defective action potentials in CMs resulting in systolic and diastolic dysfunction.
A valid tool to access diastolic and systolic function and hemodynamics in septic and critically ill patients is provided by echocardiography . Echocardiographic evaluation of septic mice was obtained. The undesirable cardiovascular effects developing after CLP were largely attenuated in C5aR1−/− and C5aR2−/− mice, suggesting that C5a receptor signaling may be linked to deterioration in cardiac output and diastolic function observed following CLP. Changes in cardiac output, stroke volume and the cardiac Doppler measurements may be in part related to changes in preload levels due to third space fluid losses and intravascular volume depletion following CLP. The prolongation of isovolumic relaxation is more likely to be a direct deleterious effect on the myocardium. These results suggest in experimental polymicrobial sepsis in mice, that development of defective cardiac function may be linked to C5a signaling via C5a receptors, although down-stream signaling pathways such as MAPKs may also be linked to cardiac dysfunction in sepsis.
The undesirable cardiovascular effects developing after CLP-induced sepsis were largely ameliorated in C5aR1−/− and C5aR2−/− mice, suggesting that C5a receptor signaling is tied to reductions in cardiac output and the impaired systolic and diastolic function following CLP. The reduction in Doppler tissue velocities (E’sa and E’la) may be a direct effect on the myocardium. These results suggest that, in experimental polymicrobial sepsis in mice, development of cardiac dysfunction can be linked to C5a signaling via C5a receptors.
The authors are responsible for the content of this publication. We acknowledge support from the Microscopy and Image-analysis Laboratory (MIL) and the Department of Pathology Flow Cytometry Core facility (University of Michigan Medical School), the Endowment for the Basic Sciences, and the Department of Cell and Developmental Biology. We would like to thank Kimber Converso-Baran, research sonographer and echocardiographic specialist for excellent services provided. We thank Beverly Schumann, Melissa Rennells, Sue Scott, Michelle Possley, and Robin Kunkel for their excellent assistance in the preparation of this manuscript and figures.
This study was supported by grants from the National Institutes of Health, USA (GM-29507, GM-61656 to PAW; NHLBI-T32-HL007517-29 to JJG; and NHLBI-HL122352 to JJ), the Leducq Foundation (JJ), and the Deutsche Forschungsgemeinschaft (DFG) Fellowship (Project KA 3740/1-1 to MK).
The authors declare no commercial or financial conflicts of interest related to the studies.