INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

GRAM-NEGATIVE SEPSIS and its experimental counterpart, endotoxemia, elicit a complex series of pathophysiological circulatory and metabolic events that result in systemic hypotension, depressed myocardial contractility, and inadequate organ perfusion. These cardiovascular derangements culminate in multiorgan failure and patient death (28, 29, 36). Although advances in medical technology and development of more potent antibiotics against gram-negative bacteria continue, sepsis remains the most common cause of death in intensive care units in the United States (29). Studies using animal models of endotoxemia have demonstrated that myocardial depression is an early consequence of an immunologic cascade, which occurs in advance of the hemodynamic alterations associated with systemic hypotension (25, 27). Depressed myocardial contractility persists in in vitro preparations of isolated heart and heart tissues (25, 26) and has been demonstrated to be intrinsic to the ventricular myocyte isolated from endotoxemic animals (19, 35).

Force generation during contraction of the cardiac myocyte is directly related to 1) intracellular free calcium concentration ([Ca²⁺]_i) and 2) the Ca²⁺ sensitivity of the contractile proteins. Clearly, depressed myocardial contractility follows perturbations of Ca²⁺ transport and depressed Ca²⁺ fluxes (12). In addition, various animal models of cardiac disease indicate that decreased Ca²⁺ responsiveness of myofilament proteins also can be a major contributing factor to depressed contractile function of the heart (11, 15, 17). It is unknown whether the depressed myocardial contractility that occurs early during endotoxemia results from decreased [Ca²⁺]_i availability, decreased responsiveness of the myofilaments to Ca²⁺, or both. Therefore, the objective of the current study was to assess whether myocardial contractile dysfunction of endotoxin-treated guinea pigs results from decreased myofilament Ca²⁺ responsiveness or decreased [Ca²⁺]_i availability. Preliminary results from this study have been reported in abstract form (32, 33).

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Experimental animals. Male, albino, Hartley strain guinea pigs (300-400 g) were injected intraperitoneally with either sterile saline (Ctl) or sterile saline containing lipopolysaccharide [Escherichia coli lipopolysaccharide (LPS) (0127:B8, Sigma), 4 mg/kg]. Animals were maintained in the laboratory in a quiet, warm environment. Four hours after treatment, guinea pigs were injected intraperitoneally with heparin sodium (1,000 U) and then decapitated 15 min later. All animal protocols were reviewed and approved by the Institutional Animal Care and Use Committee of the University of Missouri, Columbia, MO.

Myocyte isolation. For determination of myofilament Ca²⁺ responsiveness using measures of cell shortening, myocytes were isolated enzymatically as previously described (35). Briefly, hearts were removed rapidly through a midthoracic incision and placed in ice-cold isolation medium [Earle's balanced salt solution (GIBCO) supplemented with (in g/l) 0.35 MgCl₂, 5.03 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), 0.30 glutamine, 1.10 glucose, 0.20 KH₂PO₄, and 1× amino acids and vitamins (as supplied by GIBCO)]. The aorta was cannulated, and the heart was perfused retrogradely with 50 ml isolation media (37°C, bubbled with 95% O₂-5% CO₂) followed by 50 ml isolation media containing 0.05-0.08% collagenase (collagenase B, Boehringer Mannheim). Coronary effluent (flow) was measured continuously, and the hearts were considered digested in ~12 min or when flow rate was doubled. Ventricles were isolated, minced, and incubated in fresh isolation media containing 0.04% collagenase and 50 µM Ca²⁺ for 1-5 min at 37°C. Myocytes were then mechanically dispersed with a large-bore fire-polished pipette, centrifuged (15 g), and resuspended in fresh isolation media containing 100 µM CaCl₂. Myocytes were centrifuged again, and cells were resuspended in isolation media containing 500 µM CaCl₂. This procedure was repeated, and the final cell pellet was resuspended in modified Krebs-Henseleit solution (HKH) (see Physiological solutions) containing 2.0 mM CaCl₂. These preparations consistently yielded 30-35% rod-shaped "Ca²⁺-tolerant" ventricular myocytes from both Ctl (Ctl myocytes) and endotoxemic (LPS myocytes) guinea pig hearts. For all functional studies (Ctl and LPS), rod-shaped myocytes were selected that had clear striations and edges and no membrane blebs (indicating impending sarcolemmal rupture). There are no physical differences between myocytes from control and endotoxemic guinea pigs (see Ref. 35 and RESULTS). The contractile dysfunction of isolated myocytes from endotoxemic guinea pigs corresponds to the contractile dysfunction of isolated papillary muscle strips and isolated heart preparations from endotoxemic guinea pigs (1, 24, 26, 35). Therefore, we are confident that the LPS myocytes we examined represent an accurate sample of the dysfunction in the endotoxemic guinea pig heart. Less than 0.5% of rod-shaped Ctl or LPS myocytes accumulated trypan blue.

pCa-shortening relationship of saponin-permeabilized myocytes. Membrane-intact Ca²⁺-tolerant myocytes isolated by enzymatic dissociation were transferred to a Lucite superfusion chamber mounted on a Nikon TMS microscope. Myocytes were allowed to settle to the bottom of the chamber (~30 s) that was formed by an uncoated glass coverslip. Only myocytes that were quiescent, rod shaped, and exhibited clear striations with no visible membrane blebs were chosen for this study. Myocytes were positioned horizontally on a video monitor, and length was measured by a video edge motion detection device (model VED 104, Crescent Electronics, Ogden, UT). Data output from the motion detection device was calibrated in micrometers and output to an analog-to-digital converter. Data points were collected every 20 ms and analyzed using CODAS computerized data acquisition software (DATAQ Instruments, Akron, OH).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Representative tracing showing a noncumulative Ca2+ response protocol for a saponin-permeabilized lipopolysaccharide (LPS) ventricular myocyte. Percent cell shortening (0.8 Hz) of intact myocyte was determined before permeabilizing myocyte with saponin (50 µg/ml), indicated at break marks. Shortening of permeabilized myocyte was initiated and maintained during superfusion with PIPES buffer solution containing known levels of free Ca2+ (pCa). Myocytes were rinsed with PIPES buffer solution between each pCa exposure to allow cell to relengthen before next Ca2+-induced shortening event. As Ca2+ concentration increased, myocytes failed to return to initial resting cell length. Saponin-permeabilized control (Ctl) myocytes responded to this protocol in a similar manner. Inset: in a subset of cells, cell length between each Ca2+ addition was measured, normalized to resting cell length (100%) and maximal shortening (0%), and plotted as a function of pCa. There was no difference between Ctl and LPS myocytes in their lack of ability to return to initial cell length. Values represent means ± SE for 21 Ctl and LPS myocytes.

pCa-tension relationship of single skinned fibers. Myofilament Ca²⁺ responsiveness of ventricular myocytes isolated from Ctl and endotoxemic guinea pigs was determined using mechanically disrupted, Triton X-100-skinned myocytes attached via micropipettes to a piezoelectric translator (model 173, Physik, Waldbronn, Germany) and force transducer (model 403, Cambridge, Watertown, MA) as previously described (16). To attach the myocyte, micropipette tips were coated with adhesive (Great Stuff, Insta-Foam Products, Joliet, IL) and gently touched to the ends of a myocyte. The micropipettes, with attached myocyte, were then lifted to the center of a drop of relaxing solution located on a chamber attached to the stage of an inverted microscope (Diaphot, Nikon, Melville, NY). The myocyte was stretched, and sarcomere length was adjusted to 2.1-2.3 µm using a filar micrometer. Initial force was measured during maximal activation at pCa 4.5 (pH 7.0) followed by force generated by contractions at randomly chosen pH and submaximal pCa values. Force was measured at pCa 4.5 every third contraction and at final contraction to assess any decline in performance of the cell. A decline in force generated from initial to final pCa 4.5 was corrected for by assuming linear degradation in force per contraction. Active tension was calculated for each pCa tested by subtracting passive tension (pCa 9.0) from total tension. Relative tension at various submaximal pCa values was calculated as active tension per maximum active tension in pCa 4.5.

pCa-tension relationship in the presence of protease inhibitors. To determine whether proteolytic destruction of myofilament proteins had a role in decreasing myofilament Ca²⁺ responsiveness of LPS myocytes, the pCa-tension relationship was determined for Ctl and LPS myocytes isolated in the presence and absence of protease inhibitors. Skinned myocytes were isolated as described for tension measurements, with the exception that each ventricle was cut in half before immersion in either relaxing solution or relaxing solution that contained the protease inhibitors 4-(2-aminoethyl)benzenesulfonylfluoride HCl (0.2 mM), leupeptin (19.6 µM), and pepstatin (7.7 µM). Each cell suspension was centrifuged (120 g for 30 s), and the pellet was resuspended in appropriate relaxing solution with or without protease inhibitors. After a second centrifugation, all pellets were resuspended in relaxing solution without protease inhibitors. Triton X-100 (0.3%) was added, and the cell suspension was incubated for 5 min to skin the cells and eliminate sarcoplasmic reticulum membrane function. Each pellet was centrifuged, and the skinned cells were resuspended in relaxing solution and stored on ice during the course of the experiment. pCa-tension measurements were made as described above.

Sarcomere length. Freshly isolated myocytes in HKH solution were placed on a microscope slide, covered with a glass coverslip, and mounted on a Nikon Optiphot photomicroscope equipped with a Nikon Nomarski ×100 objective (numerical aperture 1.40) with differential interference contrast optics. Illumination was calibrated to an Image-1 AT image analysis and processing system (Universal Imaging, West Chester, PA). The myocyte image was collected with a high-resolution video monitor (Sony PVM122 or Sony PVM 1342Q), digitally enhanced, zoomed ×2, and stored by the analysis system. The caliper function of the Image-1 analysis system was used to measure the total length of three sarcomeres at 30 different locations on each myocyte image (10 locations at each end and 10 locations in the center of the cell). Care was taken to ensure that both ends of the caliper were aligned on ipsilateral sides of the sarcomere bands being measured. Average sarcomere length was calculated and converted to micrometers.

Fura 2 measurements. Myocytes were loaded with fura 2 before beginning an experiment using procedures similar to those of Laughlin et al. (21) with modifications. Briefly, freshly isolated ventricular myocytes from either endotoxemic or control guinea pigs were incubated with the cell membrane-permeant form of fura 2 [fura 2 acetoxymethyl ester (AM); Molecular Probes] diluted from a 1 mM stock solution into HKH solution to a final concentration of 2.5 µM fura 2-AM. In the current study, myocytes were incubated in fura 2-AM for 10 min at room temperature, washed twice with HKH solution (without fura 2-AM), and resuspended in HKH solution 1 h before subsequent measurement of fura 2 ratios and cell contraction. Myocytes were incubated at room temperature rather than 37°C to minimize problems associated with fura 2 sequestration into subcellular organelles and to avoid overloading the cells leading to buffering of intracellular Ca²⁺ transients. The stock fura 2-AM solution was made in 100% dimethyl sulfoxide (DMSO). DMSO at 0.25% had no effect on contractile function of either Ctl or LPS myocytes.

Physiological solutions. Intact myocytes were suspended in HKH solution that contained (in mM) 118 NaCl, 4.7 KCl, 2.0 CaCl₂, 1.2 MgSO₄, 1.2 KH₂PO₄, 11 glucose, 10 HEPES, and 13.5 NaHCO₃. pH was adjusted to 7.2-7.3 with NaOH. Saponin-permeabilized myocytes were perfused with a Ca²⁺-free/EGTA, PIPES buffer (PIPES buffer) that consisted of (in mM) 130 potassium propionate, 20 PIPES, 4 MgCl₂, 2 potassium EGTA, 4 Na₂ATP, 10 creatine phosphate, 11.4 glucose, and 0.02 ruthenium red, pH 7.0 adjusted with KOH. The desired free Ca²⁺ concentration was calculated using software developed by Dr. Robert Moreland. This program calculates exact amounts of CaCl₂, MgCl₂, and ATP necessary to obtain the desired free Ca²⁺, free Mg²⁺, and Mg²⁺-ATP concentrations, respectively. Free Ca²⁺ concentrations calculated using this program have been measured and verified by Ca²⁺-sensitive electrodes (23). Ionic strength was maintained at 200 mM. To ensure that mitochondrial and sarcoplasmic reticulum Ca²⁺ uptake and release were inhibited, 0.01 µM carbonyl cyanide p-trifluoromethoxyphenylhydrazone and 0.2 µM thapsigargin (Calbiochem) were added to the PIPES buffer solution.

Statistical analysis. For saponin-permeabilized pCa-shortening experiments, data from each preparation were normalized to maximum shortening, and the pCa that resulted in half-maximal shortening response (pCa₅₀) was calculated. For the tension-pCa experiments of skinned myocytes, data were collected from cells that 1) retained >65% of maximum active tension from initial pCa 4.5 to final pCa 4.5, 2) had a change in sarcomere length of <0.2 µm from pCa 9.0 to pCa 4.5, and 3) had a passive-to-active tension ratio of <0.35. Hill plot analysis of the pCa-tension relationship was calculated as described by Hofmann and Moss (18). Statistical differences between Ca²⁺ response curves for Ctl and LPS myocytes were analyzed using analysis of variance; pCa₅₀ values were compared using an unpaired two-tailed Student's t-test. P value <0.05 was considered significantly different. All data shown represent means ± SE.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Experimental animals. Four hours after administration of LPS, guinea pigs appeared lethargic; control animals exhibited no adverse effects. Survival of control and endotoxemic guinea pigs at 4 h was 100%. Previous characterization of this guinea pig model of endotoxemia indicates that at 4 h after LPS injection, animals are hypothermic and exhibit a decrease in stroke volume and cardiac output, although heart rate and systemic blood pressure are normal (26, 27, 37). At 18 h after LPS injection, this model exhibits 20-30% mortality and is characterized by central nervous system depression, lethargy, minimal changes in arterial blood gases, and marked decreases in systolic and diastolic blood pressure (24, 26).

Myocytes. Ca²⁺-tolerant myocytes isolated by enzymatic dissociation from hearts of Ctl and endotoxemic guinea pigs were morphologically indistinguishable as previously described (35). Resting cell length of Ctl myocytes was 129.0 ± 2.0 µm (n = 57 cells) and was not different from that of LPS myocytes (126.9 ± 2.7 µm; n = 57 cells). After permeabilization with saponin, resting length of Ctl myocytes increased by 3.0 ± 0.6 µm to 132.0 µm, whereas LPS myocytes increased in cell length by 3.7 ± 0.8 µm to 130.6 µm. Similarity in resting cell length both before and after permeabilization of the sarcolemma suggests that internal load-bearing structures important in determining resting cell length (34) are not different between LPS and Ctl myocytes.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Contractile responses of sarcolemma-intact Ctl and LPS myocytes. Myocytes were field stimulated at 0.8 Hz for 3-4 min to achieve steady-state responses. Percent cell shortening as well as maximal rates of contraction (dLmax/dt) and relaxation (+dLmax/dt) were measured. Data represent average of 33 LPS myocytes and 33 Ctl myocytes. Values represent means ± SE. * Significantly different at P < 0.05.

Shortening-pCa relationship. After saponin permeabilization, increasing the concentration of free Ca²⁺ in the PIPES buffer solution resulted in decreased cell length (shortening) in both LPS and Ctl myocytes. The Ca²⁺-stimulated decrease in cell length reached steady state within 5 min of adding Ca²⁺ to the superfusate and was maintained for as long as Ca²⁺ was present in the superfusate (Fig. 1). As the concentration of free Ca²⁺ was increased, extent of shortening increased similarly in both Ctl and LPS myocytes (Fig. 3A). As the free Ca²⁺ in the superfusate was increased above 2 µM (5.69 pCa units), cells progressively failed to relengthen to initial resting cell length when rinsed with 0 Ca²⁺ PIPES buffer (Fig. 1). However, the extent to which myocytes failed to return to initial resting cell length was not statistically different between Ctl and LPS myocytes (Fig. 1, inset). Therefore, percent shortening at each pCa tested was calculated relative to initial resting cell length. When data were normalized to maximal shortening (pCa before myoball formation) for each myocyte (Fig. 3B), pCa₅₀ of LPS myocytes (5.72 ± 0.02) was not different from Ctl (5.78 ± 0.04) (Fig. 3B). These data suggest that intrinsic myofilament Ca²⁺ responsiveness of LPS myocytes is not different from Ctl myocytes.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   pCa-shortening relationship of saponin-permeabilized Ctl and LPS myocytes. Myocytes were superfused with PIPES buffer solution containing increasing concentrations of Ca2+, and cell length was measured. A: myocyte shortening at each pCa was similar for Ctl and LPS myocytes. Threshold pCa for shortening was 6.60 ± 0.11 for Ctl and 6.42 ± 0.11 for LPS myocytes and was not statistically different. At high Ca2+, both LPS and Ctl myocytes lost normal conformation and rounded into myoballs. Average pCa that invoked myoball formation in Ctl myocytes (5.45 ± 0.02) was not different from that in LPS myocytes (5.50 ± 0.03). Maximal shortening of LPS myocytes before myoball formation (21.49 ±1.89%) was not different from Ctl (22.38 ± 1.92%). B: values from A normalized to maximum percent shortening for each myocyte. Values represent means ± SE; n = 33 cells in each group.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   pCa-shortening relationship of Ctl and LPS myocytes at pH 7.0 and 6.6. Saponin-permeabilized myocytes were superfused with PIPES buffer solution containing increasing Ca2+ concentration at either pH 7.0 or 6.6. Maximal shortening of either Ctl or LPS myocytes was not different at pH 7.0 or 6.6 (Ctl: pH 7.0, 22.02 ± 1.58% vs. pH 6.6, 25.70 ± 1.35%; LPS: pH 7.0, 16.07 ± 3.96% vs. pH 6.6, 21.02 ± 2.63%). pCa value that resulted in half-maximal shortening (pCa50) for Ctl myocytes at pH 7.0 was 5.72 ± 0.02 and shifted to 5.04 ± 0.05 at pH 6.6. pCa50 for LPS myocytes at pH 7.0 was 5.72 ± 0.03 and shifted to 4.98 ± 0.01 at pH 6.6. There was no statistical difference in pCa50 between Ctl and LPS at either pH 7.0 or 6.6. Values represent means ± SE; Ctl myocytes, n = 6 cells; LPS myocytes, n = 9 cells.

Tension-pCa relationship. To further compare myofilament Ca²⁺ responsiveness of Ctl and LPS myocytes and to validate data obtained from the shortening experiments, tension was measured as a function of pCa in Triton X-100-skinned myocyte preparations from hearts of Ctl and endotoxemic guinea pigs. In contrast to the lack of difference in maximum shortening between Ctl and LPS (Fig. 3A), maximum tension generated by LPS myocytes was depressed compared with Ctl myocytes (Fig. 5A, Table 1). Although tension generated by LPS myocytes was significantly reduced, there was no difference in myofilament sensitivity to Ca²⁺ between Ctl and LPS myocytes (Fig. 5B, Table 1). The effect of acidic pH (6.6) on myofilament Ca²⁺ sensitivity of Ctl and LPS myocytes also was examined. In agreement with the shortening-pCa results, reducing pH from 7.0 to 6.6 resulted in a rightward shift of the pCa-tension relationship of both Ctl and LPS myocytes, and the decrease in myofilament Ca²⁺ sensitivity due to decreased pH was similar for LPS and Ctl myocytes (Fig. 6, Table 1). Reducing pH from 7.0 to 6.6 also increased the slope of the tension-pCa relationship for both Ctl and LPS myocytes and decreased maximum tension generated by Ctl myocytes (Table 1). Maximum tension also declined for LPS myocytes at pH 6.6, but the difference did not reach statistical significance (Table 1).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   pCa-tension relationship for skinned Ctl and LPS myocytes. A: active tension as a function of pCa was significantly depressed in LPS myocytes when compared with Ctl myocytes (P < 0.05; analysis of variance). At maximally activating pCa (4.5), LPS myocytes generated 2.89 ± 0.23 g/mm2 of tension compared with 3.75 ± 0.34 g/mm2 generated by Ctl myocytes. B: mean relative tension as a function of pCa for both Ctl and LPS myocytes. Neither pCa50 nor slope of tension-pCa relationship was significantly different between LPS and Ctl myocytes. Values represent means ± SE; n = 8 cells in each group.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   pCa-tension relationship at pH 6.6 and 7.0 for Ctl and LPS myocytes. Mean relative tension as a function of pCa at pH 7.0 and pH 6.6 for both Ctl (A) and LPS (B) myocytes. At pH 6.6, pCa50 of Ctl myocytes decreased by 0.50 units to 5.23 ± 0.02 (A), whereas pCa50 of LPS myocytes decreased by 0.52 units to 5.24 ± 0.03 (B). Values represent means ± SE; n = 8 cells in each group. Values for pCa50 and slope of these relationships are described in Table 1.

pCa-tension relationship in the presence of protease inhibitors. To determine whether the decrease in tension generated by LPS but not Ctl myocytes, at pH 7.0, was the result of proteolytic breakdown of myofilament proteins, the pCa-tension relationship was reassessed for myocytes that were isolated with or without protease inhibitors present during the isolation process. Neither maximal tension nor the pCa₅₀ of Ctl myocytes was affected by isolating the myocytes in the presence of protease inhibitors (Fig. 7). In contrast, maximal tension generated by LPS myocytes was significantly improved when myocytes were isolated in the presence of protease inhibitors. Maximal tension of LPS myocytes isolated in the presence of protease inhibitors (3.53 ± 0.98 g/mm²) was significantly greater than maximal tension generated by LPS myocytes from the same hearts but isolated in the absence of protease inhibitors (2.85 ± 0.18 g/mm²). Furthermore, maximal tension of LPS myocytes isolated in the presence of protease inhibitors was no longer different from maximal tension of Ctl myocytes (Fig. 7A). The presence of protease inhibitors in the isolation solution had no effect on the pCa₅₀ of LPS myocytes (Fig. 7B). These results suggest that the decrease in maximal tension generated by LPS myocytes was most likely due to proteolytic destruction of myofilament proteins during the isolation process. Proteolytic breakdown does not appear to occur in Ctl hearts.

View larger version (15K): [in this window] [in a new window]   Fig. 7.   pCa-tension relationship of Ctl and LPS myocytes mechanically isolated in presence of protease inhibitors. A: maximal tension of Ctl myocytes isolated without protease inhibitors (not shown, 3.65 ± 1.03 g/mm2) was not different from maximal tension of myocytes from same heart but isolated in presence of protease inhibitors (3.01 ± 0.80 g/mm2). Maximal tension of LPS myocytes isolated in presence of protease inhibitors (3.53 ± 0.98 g/mm2) was significantly greater than maximal tension generated by LPS myocytes isolated from same hearts but not containing protease inhibitors in isolation media (not shown, 2.85 ± 0.18 g/mm2). B: mean relative tension for Ctl and LPS myocytes. pCa50 of Ctl myocytes isolated without inhibitors (5.70 ± 0.01) was not different from that in presence of protease inhibitors (5.71 ± 0.03). Similarly, pCa50 was not different between LPS myocytes isolated with (5.73 ± 0.01) or without (5.66 ± 0.01) protease inhibitors. Values represent means ± SE; n = 10 Ctl, 7 LPS myocytes.

Intracellular fura 2 ratios. Peak systolic fura 2 ratios from both Ctl and LPS myocytes increased in a frequency-dependent manner; however, the amplitude of the peak fura 2 ratio of LPS myocytes was significantly reduced compared with Ctl. The disparity between peak fura 2 ratios in LPS and Ctl myocytes increased with increasing frequency (Fig. 8A). Basal, diastolic fura 2 ratios of LPS myocytes were not different from Ctl either at rest (no stimulation) or during low-frequency stimulation. However, as stimulation frequency increased, basal fura 2 ratios of Ctl myocytes increased significantly more than those of LPS myocytes and correlated with the differences in peak systolic ratios (Fig. 8B). These results suggest that reduced cell shortening of the sarcolemmal-intact LPS myocyte is more likely the result of reduced [Ca²⁺]_i availability during systole.

View larger version (13K): [in this window] [in a new window]   Fig. 8.   Fura 2 ratios of Ctl and LPS myocytes. Intracellular Ca2+ was assessed as peak fura 2 ratio (A) and basal ratio (B) in Ctl and LPS myocytes. Fura 2-loaded myocytes were superfused with modified Krebs-Henseleit solution containing 2 mM CaCl2. Maximal peak fura 2 ratio attained by LPS myocytes was 0.68 ± 0.05 nm at a frequency of 1 Hz, whereas Ctl myocytes reached a peak ratio of 1.13 ± 0.07 nm under same conditions. Values represent means ± SE; n = 18 cells from 6 animals in each group. * Significantly different from Ctl at P < 0.05.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Septicemia and endotoxemia produce a progressive cardiodynamic dysfunction in human patients and experimental animal models. This dysfunction is intrinsic to the ventricular myocyte, although the cellular mechanisms remain unresolved. We tested the hypothesis that decreased contractile function of ventricular myocytes isolated from endotoxemic guinea pigs was associated with reduced Ca²⁺ responsiveness of the myofilament proteins. We report that in vivo administration of LPS and induction of endotoxemia without development of endotoxemic shock does not alter myofilament Ca²⁺ responsiveness of ventricular myocytes. However, intracellular Ca²⁺ transients of myocytes isolated from this model of endotoxemia were significantly less than control, suggesting that decreased available Ca²⁺, rather than decreased myofilament Ca²⁺ responsiveness, leads to decreased contractile function of the heart during endotoxemia.

Our findings that endotoxemia does not alter myofilament Ca²⁺ sensitivity are consistent with reports that intraperitoneal sepsis (rats) does not alter the pCa₅₀ of cardiac myofibrillar ATPase activity (30). Although our results initially appear to conflict with a report of depressed myofibrillar ATPase activity in a canine model of endotoxemia (7), the discrepancy was subsequently resolved by demonstration that depressed myofibrillar ATPase activity did not occur in endotoxemic dogs if coronary blood flow was maintained (20). Thus it would appear that inadequate coronary blood flow and resulting ischemia in the hypotensive phase of endotoxemia-induced shock was responsible for the decrease in myofibrillar ATPase activity (7, 20). Myocardial blood flow was not measured in the present study. However, we previously demonstrated that cardiac dysfunction in this guinea pig model of endotoxemia precedes changes in systemic blood pressure (27). Moreover, in hearts isolated from endotoxemic guinea pigs, myocardial contractile dysfunction was not surmounted by maximally effective increases in coronary flow (1).

Although endotoxemia did not mediate a decrease in Ca²⁺ sensitivity of myofilaments, we observed a significant reduction in maximal Ca²⁺-activated tension in LPS myocytes (Fig. 5). Depressed Ca²⁺-activated tension could result from 1) reduced numbers of cycling cross bridges or 2) reduced force generated by activated cross bridges. Increased proteolytic breakdown of myofibrillar proteins is one mechanism by which such changes in cross-bridge activity could occur, and evidence exists for skeletal muscle, which suggests that endotoxemia and sepsis increase intracellular proteolysis and catabolism of contractile proteins. Skeletal muscle catabolism (5, 22) and myosin heavy chain destruction (2) were evident as early as 24 h after bacterial inoculation in a septic rat model. The role of myocardial proteolytic damage in endotoxin-mediated myocardial contractile dysfunction remains to be determined.

In the present study, we examined the pCa-tension response of both Ctl and LPS myocytes that were isolated with protease inhibitors present during the isolation process. We reasoned that if endotoxemia resulted in a net destruction of myofilament proteins during the 4 h the animal was ill, then the presence of the protease inhibitors during the isolation process would not improve the decreased maximal tension of LPS myocytes. However, Ca²⁺-activated tension of LPS myocytes isolated with protease inhibitors present, was in fact improved and was no longer different from Ctl (Fig. 7A). The pCa-tension response of Ctl myocytes was not altered by the presence of protease inhibitors during the isolation process. These data indicate that proteolytic destruction of myofilaments occurs during the isolation process rather than during the disease and was responsible for the decrease in maximal tension of LPS myocytes. Had proteolysis occurred in vivo, adding protease inhibitors to the isolation solution would not have reversed the dysfunction in tension-generating ability. Interestingly, proteolytic activity appeared to be significant only in LPS preparations, suggesting that proteolytic enzyme activity is increased in the heart during endotoxemia relative to control hearts. We suspect extracellular proteases, possibly from inflammatory cells present in the hearts of endotoxemic animals, were introduced to the myofilament proteins during mechanical isolation. A decrease in maximal shortening was not detected because myocytes were isolated as intact cells, washed of all extracellular and cellular debris, and then permeabilized. Thus extracellular proteases were removed before permeabilization of myocytes used for experiments that measured cell shortening.

It is well documented that the pCa-tension relationship can be regulated by changes in pH. Exposing Ctl myocytes to acidic pH (6.6) altered the pCa-tension relationship similar to that originally described by Fabiato and Fabiato (9) in that both maximum tension and Ca²⁺ sensitivity were reduced. Furthermore, the slope of the pCa-tension relationship increased with acidic pH. Acidic pH had similar effects on the pCa-tension relationship of LPS myocytes, although the reduction in maximum tension did not reach statistical significance. The relative ineffectiveness of acidic pH at reducing maximum tension could suggest myofilament proteins were modified in LPS myocytes. However, our results with protease inhibitors indicate that the relatively smaller decrease in maximum tension of LPS myocytes at pH 6.6 most likely occurred because maximum tension was already depressed in LPS myocytes due to the proteolytic breakdown of myofilaments.

Decreasing pH from 7.0 to 6.6 caused a similar rightward shift in the pCa-shortening relationship of saponin-permeabilized myocytes. These results were particularly important, since sarcomere length of both Ctl and LPS myocytes was 1.9 µm in these unloaded myocytes. Although this resting sarcomere length was within the physiological range of the in situ myocyte (13), we were concerned that sarcomere length of 1.9 µm might be sufficiently short that length-dependent Ca²⁺ sensitivity of myofilament proteins would be too low for us to detect a subtle shift in myofilament Ca²⁺ responsiveness of the LPS myocytes. However, the shift in pCa₅₀ values from pH 7.0 to pH 6.6 for the pCa-shortening relationship was remarkably similar to the shift measured for the pCa-tension relationship. The ability to demonstrate a similar right shift with acidic pH using both methodologies suggests that the shorter sarcomere length of myocytes undergoing Ca²⁺-activated shortening did not significantly reduce Ca²⁺ responsiveness of the myofilament proteins. Thus measures of the pCa-tension relationship and pCa-shortening relationship provide similar values for pCa₅₀ even under conditions with different pH.

Although pCa-shortening, a novel method for assessing myofilament responsiveness to Ca²⁺, provides reliable information relative to pCa₅₀, there are limitations to this method not present when measuring pCa-tension. For example, it is necessary with measures of shortening to remove the sarcoplasmic reticulum and mitochondria. Thus, unlike the studies of Fabiato and Fabiato (10) in which skinned cells with an intact sarcoplasmic reticulum were used to study force development and shortening rates, rate of shortening could not be accurately determined because the rate of Ca²⁺ binding to troponin C was limited by both superfusion rate and the diffusion rate of Ca²⁺ through the cytosol of the permeabilized cell. Thus shortening of permeabilized myocytes in our study was not a Ca²⁺-induced Ca²⁺ release-driven spontaneous contraction but rather a tonic contraction dependent on the Ca²⁺ concentration that diffuses to the myofilament and binds troponin C. Extent of shortening in response to increasing free Ca²⁺ most likely depends on the number and Ca²⁺ affinity of troponin C molecules activated. A second limitation in the shortening methodology is the loss of normal cell conformation at high free Ca²⁺ concentrations. Myoball formation occurs at a free Ca²⁺ concentration that is considerably lower than that required for maximal tension generation (pCa 4.5). However, we emphasize that myoball formation did not occur until free Ca²⁺ became much higher (pCa 5.45 for Ctl; 5.50 for LPS) than that which is considered maximal in the intact myocyte (11).

Clearly, if one considers that a complete description of myofilament Ca²⁺ responsiveness includes 1) maximal activation, 2) pCa₅₀, and 3) slope of the pCa-response relationship, then measures of the pCa-shortening response do not completely describe myofilament Ca²⁺ responsiveness. For example, when pH was reduced to 6.6, the pCa-shortening relationship failed to demonstrate a decrease in maximal response in LPS myocytes, and neither cell group showed an increase in the Hill coefficient. Discrepancies between shortening results from results obtained using tension measurements may well be due to the inability to maintain normal restorative forces in unloaded myocytes exposed to increasing levels of Ca²⁺. However, using this nontraditional technique of myocyte shortening to determine myofilament Ca²⁺ responsiveness, we generated pCa values for shortening threshold and half-maximal shortening that were nearly identical to those generated using tension measurements of myocytes from both Ctl and LPS myocytes. With both methodologies, pCa values were within the range reported for skinned fiber bundles from dog hearts (31) and skinned myocyte preparations from rat hearts (14, 15). Thus measures of myocyte shortening describe many, but not all, functional characteristics of myofilament proteins.

In the intact unattached myocyte, cell shortening provides a measure of the maximal turnover rate of the contractile protein cross bridges (6) and is regulated by cytosolic Ca²⁺. As cytosolic Ca²⁺ declines, so does cross-bridge cycling rate and extent of shortening. In this study, LPS myocytes exhibited both depressed extent of cell shortening and depressed maximal rates of contraction and relaxation (Fig. 2), suggesting that cytosolic Ca²⁺ concentration may be depressed during this early, nonhypotensive stage of endotoxemia. In fact, peak [Ca²⁺]_i transients measured using fura 2 microfluorescence were significantly less in LPS myocytes compared with Ctl (Fig. 8). These data suggest that the contractile dysfunction exhibited by LPS myocytes most likely results from decreased cytosolic Ca²⁺ concentration rather than a decrease in myofilament Ca²⁺ responsiveness.

In conclusion, unlike other cardiac disease states in which depressed myofilament Ca²⁺ sensitivity contributes to contractile dysfunction (3, 11, 17), myofilament Ca²⁺ responsiveness of myocytes from nonhypotensive, endotoxemic guinea pigs was not depressed, whether assessed by comparing 1) the pCa-shortening relationship or 2) the pCa-tension relationship. Therefore, decreased myofilament Ca²⁺ responsiveness does not contribute to the contractile dysfunction of the heart during the early stages of endotoxemia.