The purpose of the study was to test the hypothesis that neutrophils can injure cultured skeletal myotubes. Human myotubes were grown and then cultured with human blood neutrophils. Myotube injury was quantitatively and qualitatively determined using a cytotoxicity (51Cr) assay and electron microscopy, respectively. For the 51Cr assay, neutrophils, under non-in vitro-stimulated andN-formylmethionyl-leucyl-phenylalanine (FMLP)-stimulated conditions, were cultured with myotubes at effector-to-target cell (E:T) ratios of 10, 30, and 50 for 6 h. Statistical analyses revealed that myotube injury was proportional to the E:T ratio and was greater in FMLP-stimulated conditions relative to non-in vitro-stimulated conditions. Transmission electron microscopy, using lanthanum as an extracellular tracer, revealed in cocultures a diffuse appearance of lanthanum in the cytoplasm of myotubes and a localized appearance within cytoplasmic vacuoles of myotubes. These observations and their absence in control cultures (myotubes only) suggest that neutrophils caused membrane rupture and increased myotube endocytosis, respectively. Myotube membrane blebs were prevalent in scanning and transmission electron micrographs of cultures consisting of neutrophils and myotubes (E:T ratio of 5) and were absent in control cultures. These data support the hypothesis that neutrophils can injure skeletal myotubes in vitro and may indicate that neutrophils exacerbate muscle injury and/or delay muscle regeneration in vivo.
- chromium assay
- electron microscopy
- muscle injury
- muscle regeneration
skeletal muscle injury, induced by eccentric contractions (15), muscle trauma (22, 23), or the loading of atrophic muscle (7, 25), is associated with increased muscle neutrophil concentrations. The biological function of neutrophils in muscle injury and subsequent regeneration, however, is unclear. Previous investigators have reported that muscle neutrophil concentrations are increased within 2 h postinjury and remain above control concentrations for at least 48 h of recovery (7, 22,25). During this time, the muscle undergoes further degeneration (secondary injury) (5, 23, 29). Because of the temporal relationship between muscle neutrophil concentrations and secondary injury and because neutrophils can release potentially injurious reactive oxygen and nitrogen intermediates (ROIs and RNIs, respectively) and lysosomal enzymes (reviewed in Refs. 4and 28), neutrophils have been suggested to exacerbate skeletal muscle damage in the hours to days following the injury. Direct evidence supporting this contention, however, is lacking.
Ascertaining the biological function of neutrophils in vivo is difficult because neutrophils are present when injured muscle shows signs of both injury and regeneration (7, 10, 22, 26, 27,29). Tissue-cultured skeletal myotubes offer a distinct advantage over in vivo experimentation because the ability of neutrophils to damage previously uninjured myotubes can be evaluated. We reasoned that if neutrophils exacerbate skeletal muscle injury, then they should be capable of damaging previously uninjured myotubes.
To determine whether neutrophils injure skeletal muscle myotubes, we cultured human neutrophils in a non-in vitro-stimulated and an in vitro-stimulated state with human myotubes. Myotube injury was quantitatively and qualitatively determined using a cytotoxicity (51Cr) assay and electron microscopy (transmission and scanning), respectively. For transmission electron microscopy, lanthanum was used as an extracellular tracer to qualitatively determine whether neutrophils caused myotube membrane rupture and/or increased myotube membrane permeability. We expected that neutrophils would cause myotube membrane rupture and thus hypothesized that lanthanum would be predominantly found diffusely in the cytoplasm of myotubes. This hypothesis was based on the fact that neutrophils can release ROIs, RNIs, and lysosomal enzymes that may cause myotube membrane rupture and necrosis (reviewed in Refs. 4 and28). Because previous investigators have reported that traumatic injury to cultured cells (e.g., neurons and fibroblasts) increases the rate of endocytosis (3, 30), we also postulated that neutrophils would increase myotube endocytosis. Support for this hypothesis would be revealed if lanthanum was found localized to cytoplasmic vacuoles of myotubes when cultured with neutrophils.
MATERIALS AND METHODS
Human myoblasts were obtained from a 19-yr-old female donor and were negative for mycoplasma, hepatitis B virus, hepatitis C virus, and human immunodeficiency virus (Clonetics, San Diego, CA). Myoblasts were seeded at a density of 10,000 cells/cm2 in either gelatin-coated microtiter plate wells (24 well; Becton Dickinson, Lincoln Park, NJ) or on gelatin-coated Thermanox coverslips (Fisher Scientific, Pittsburgh, PA). Myoblast proliferation occurred in growth medium (Clonetics) that was supplemented with 10 ng/ml epidermal growth factor (EGF), 10 μg/ml insulin, 50 μg/ml fetuin, 50 μg/ml bovine serum albumin, 375 ng/ml dexamethasone, 50 μg/ml gentamicin, and 50 ng/ml amphotericin-B. Myoblasts were maintained in a humidified, 37°C, and 5% CO2 atmosphere. At ∼90% confluence, the growth medium was exchanged for a differentiation medium that consisted of Dulbecco's modified Eagle's medium (Sigma Chemical, St. Louis, MO), 2 ng/ml EGF, and 2% heat-inactivated fetal bovine serum (FBS). The differentiation medium was changed every 2 days for 4 days. On the 5th day, myotubes were cultured with neutrophils. Light microscopy observations of hematoxylin-and-eosin-stained cultures revealed numerous multinucleated myotubes.
Human neutrophils were obtained from heparinized venous blood (50–70 ml) of healthy male volunteers (n = 6) after obtaining verbal and written consent in accordance with institutional guidelines. Each subject's neutrophils were used for both conditions (non-in vitro-stimulated and FMLP-stimulated) and for all effector-to-target cell (E:T) ratios.
Blood neutrophils were isolated from other cells using density gradient centrifugation [neutrophil isolation medium (NIM); Cardinal Assoc. Sante Fe, NM]. Briefly, blood was layered on the NIM, centrifuged, and the polymorphonuclear (PMN) cell layer was aspirated. Cells were then washed with calcium- and magnesium-free Hanks' balanced salt solution (HBSS) and centrifuged. The remaining red blood cells were lysed with an ammonium chloride solution and centrifuged, and the PMN cells were washed again with HBSS. The PMN cells were resuspended in Earle's balanced salt solution (EBSS) supplemented with 2% FBS and 400 μM ofl-arginine (coculture medium) to yield the desired E:T ratio. l-Arginine was included in the coculture medium to provide a substrate for nitric oxide synthesis and to enhance neutrophil degranulation (33). The final neutrophil preparation routinely yielded >98% neutrophils with cell viability >98% as determined by trypan blue exclusion.
For the cytotoxicity assay, neutrophils in a non-in vitro-stimulated and an in vitro-stimulated state were cultured with myotubes at E:T ratios of 10, 30, and 50. In vitro stimulation was accomplished by adding N-formyl-methionyl-leucyl-phenylalanine (FMLP; 2.0 × 10−6 M final concentration 0.02% vol/vol DMSO) to neutrophil suspensions just before culturing them with myotubes. FMLP was used because it is a physiologically relevant stimulus that activates neutrophils by receptor-mediated G protein signal transduction (33).
Cytotoxicity against allogeneic myotubes was measured in a 6-h51Cr release assay. Myotubes were rinsed twice with EBSS and then labeled with 51Cr in EBSS (Na2 51CrO4; 36 μCi/well) for 1 h. After the wells were washed twice with EBSS, neutrophils suspended in coculture medium were added to the appropriate wells. To facilitate neutrophil-myotube adhesion and to minimize neutrophil aggregation, the plate was then centrifuged (50 g) for 1 min. After centrifugation, the plate was incubated for 6 h in a humidified, 37°C, 5% CO2 atmosphere.
Each plate contained wells for maximal 51Cr release, background 51Cr release, non-in vitro-stimulated neutrophils, and FMLP-stimulated neutrophils. Maximal 51Cr release was induced with a 4% Triton X-100 solution. Because preliminary experiments demonstrated that FMLP did not influence51Cr release, background release wells contained only coculture medium (data not reported). After the 6-h incubation, an aliquot was collected from each well and the radioactivity was counted using a gamma counter. An injury index was calculated using the mean of triplicates using the following equation: injury index = [(e − b)/(m − b)], where e is mean experimental release, b is mean background release, and m is maximal release.
Transmission and scanning electron microscopy were performed on several control cultures (myotubes only) and on cultures containing both neutrophils (non-in vitro-stimulated) and myotubes (E:T ratio of 5). Because of the severity of the neutrophil-mediated myotube injury at E:T ratios greater than five, an E:T ratio of five was used for electron microscopy experiments to ensure that an adequate quality and quantity of monolayers was obtained for analysis.
Using transmission electron microscopy, myotube membrane rupture and/or increased membrane permeability were qualitatively determined utilizing lanthanum as an extracellular tracer. For transmission electron microscopy, cultures were rinsed once with 0.1 M cacodylate buffer (pH 7.4) and then fixed using a 3% glutaraldehyde solution that contained 1% lanthanum chloride. Cultures were kept at room temperature for 2 min and then transferred to 4°C for 2 h. After primary fixation, cultures were washed with cacodylate. Secondary fixation took place at 4°C in S-collidine-buffered 2% osmium tetroxide/1% lanthanum for 1 h followed by three changes with S-collidine. To enhance the contrast of lanthanum against cellular structures, samples were not stained with uranyl acetate and lead citrate. This practice facilitated the determination of lanthanum within myotubes. After dehydration, monolayers were transferred to vertical molds and infiltrated with 100% epoxy (Embed 812/Araldite). Following polymerization, thin silver-to-gold (70–90 nm) sections were cut on a Reichart OM-U3 ultramicrotome using a diamond knife. Sections were viewed at 80 kV on a Phillips CM10 transmission electron microscope.
For scanning electron microscopy, coverslips were washed five times with EBSS without FBS (37°C) to remove soluble proteins and detritus. Fixation was performed for 30 min using sodium cacodylate-buffered 3% glutaraldehyde at 4°C. Cultures were then washed with cacodylate buffer three times before dehydration with graded changes of ethanol. Following changes of absolute ethanol, coverslips were removed from culture dishes and critically point dried in a Polaron E3000 (Watford) bomb. Coverslips were mounted to stubs using silver paste and coated with gold using a Polaron E5100 sputter coater. Samples were viewed at 5 kV on a Cambridge Stereoscan 180 (Cambridge) scanning electron microscope. All pictures were taken at a 65° tilt.
The injury index was statistically analyzed using a repeated measures analysis of variance to analyze the main effects and the interaction effect. The Huynh-Feldt Epsilon was applied to degrees of freedom to account for violation of the sphericity assumption. The Newman-Keuls post hoc test was used to locate the differences between means when the observed F ratio was statistically significant (P < 0.05).
Statistical analysis revealed a significant main effect for E:T ratio and condition (non-in vitro-stimulated and FMLP-stimulated neutrophils) with no interaction detected (Fig. 1). Post hoc analysis revealed significant differences between E:T ratios of 10 and 30, 10 and 50, and 30 and 50. The injury index was also significantly greater for the FMLP-stimulated conditions relative to the non-in vitro-stimulated conditions across all E:T ratios.
To demonstrate that FMLP activated neutrophils, we quantified superoxide anion production (O ·) using the cytochrome c assay (24) in a separate set of experiments. These experiments (n = 15) demonstrated that FMLP resulted in a fourfold increase (P < 0.05) in neutrophil-derived O · (means ± SE; 16.0 nmol · 10 min−1 · 2 × 106 neutrophils−1 ± 0.7) relative to non-in vitro-stimulated neutrophils (3.7 nmol · 10 min−1 · 2 × 106neutrophils−1 ± 2.0).
Electron microscopy observations.
In control cultures (myotubes only), lanthanum was found on the extracellular surface of myotube membranes and was not found in the cytoplasm of myotubes (Fig. 2). In contrast, cultures consisting of both neutrophils and myotubes showed two different patterns of lanthanum's appearance within myotubes.
The most frequently observed pattern was a localization of lanthanum to cytoplasmic vacuoles of myotubes (Fig.3). In most cases, these vacuoles consisted of a larger vacuole with smaller vacuoles surrounding and appearing to be associated with the larger vacuoles. Some of the larger and smaller vacuoles contained lanthanum in their lumen and others were rimmed with lanthanum. In most cases, the lumina of the larger vacuoles contained no internal structures visible by electron microscopy. Small isolated lanthanum-containing vacuoles were also found near myotube membranes (Fig. 3). The second observed pattern of lanthanum's appearance was one in which it was found diffusely within myotubes (Figs. 4 and5).
Myotube membrane blebs (bubble-like protrusions on the plasma membrane) were frequently observed on myotubes cultured with neutrophils (see Figs. 5, 6, and 8) and were absent in control cultures (Fig. 2). Most of the membrane blebs were ruptured as indicated by the diffuse appearance of lanthanum within membrane blebs and the cytoplasm of myotubes (Figs. 4 and 5). At some of the sites of membrane rupture (Fig. 4), we observed neutrophils in close proximity to vacuoles that presumably were contained within membrane blebs before their rupture. Whether neutrophils caused the membrane bleb rupture or the membrane rupture resulted in the chemotaxis of neutrophils cannot be determined from our observations. We also observed intact membrane blebs and blebs that were presumably shed from myotubes (Fig.6). Transmission electron micrographs revealed that membrane blebs (ruptured, intact, and shed blebs) contained vacuoles that either were rimmed with or contained lanthanum (Figs. 4-6). The presence of lanthanum within vacuoles of intact and shed membrane blebs suggests that lanthanum was incorporated into membrane bleb vacuoles before bleb rupture. The vacuoles in membrane blebs contained no visible cellular structures.
Neutrophils, the first inflammatory cell type to appear in injured muscle (7, 22, 23), have been suggested to contribute to the exacerbation in skeletal muscle injury in the hours to days following the injury. Neutrophils could contribute to muscle injury by damaging uninjured areas of damaged fibers, damaging adjacent uninjured fibers, and/or delaying muscle regeneration by injuring muscle precursor cells and myotubes. Our results provide support for this hypothesis by demonstrating that neutrophils injure cultured skeletal muscle myotubes. The myotube injury was proportional to neutrophil number (E:T ratio) and to their state of activation (non-in vitro-stimulated vs. FMLP-stimulated; Fig. 1). The neutrophil-mediated injury was confirmed via transmission and scanning electron microscopy. The presence of lanthanum in the cytoplasm of myotubes (Figs. 3-5) and myotube membrane blebs (Figs. 4-6 and 8) in cultures consisting of both neutrophils and myotubes support our quantitative data. In addition, lanthanum's localized and diffuse appearance in the cytoplasm of myotubes when cultured with neutrophils suggests that neutrophils are capable of increasing myotube endocytosis and rupturing myotube membranes, respectively.
Our results represent the first report to provide quantitative evidence supporting the contention that neutrophils are capable of damaging previously uninjured myotubes (Fig. 1). The greater myotube injury in the FMLP-stimulated condition relative to the non-in vitro-stimulated condition is attributable to a greater state of neutrophil activation. In separate experiments, we demonstrated that FMLP resulted in a fourfold increase in neutrophil-derived O · relative to non-in vitro-stimulated neutrophils. Previous investigators have reported similar effects of FMLP on neutrophil ROIs (reviewed in Ref. 4) and have also reported FMLP-mediated release of lysosomal enzymes from neutrophil granules (33).
Myotube injury in the non-in vitro-stimulated condition is most likely the result of a basal state of neutrophil activation and/or an adhesion-dependent increase in neutrophil activation. We and others have demonstrated that non-in vitro-stimulated blood neutrophils produce low concentrations of ROIs in vitro (reviewed in Ref.4). Thus, in the non-in vitro-stimulated conditions, neutrophils were in a state of activation before culturing them with myotubes. Because neutrophil adhesion to extracellular matrix proteins is known to activate neutrophils (20), neutrophils likely became further activated when they adhered to myotubes.
Our quantitative data are consistent with other in vitro studies that have demonstrated that neutrophils injure cultured endothelial cells (8, 11, 17, 18, 32) or cardiac myoblasts (3). Neutrophil-derived superoxide anion, hydrogen peroxide, hydroxyl radical, hypochlorous acid, and lysosomal proteases have all been demonstrated to cause varying degrees of injury to cultured endothelial cells (8, 11, 17, 18, 32). In addition, blocking neutrophil adhesion and ROI production using anti-CD18 and intracellular adhesion molecule-1 antibodies has been reported to ameliorate neutrophil-mediated injury to endothelial cells and cardiac myoblasts (3, 21). Future experiments are needed to investigate whether similar mechanisms are operating in our neutrophil and myotube skeletal muscle cultures.
Electron micrographs of neutrophil-myotube cultures revealed unexpected and novel observations. We hypothesized that neutrophils would cause myotube membrane rupture, and thus we postulated that lanthanum would predominantly be found diffusely in the cytoplasm of myotubes. Although diffuse appearance of lanthanum was observed (Figs. 4 and 5), lanthanum was more frequently localized to cytoplasmic vacuoles of myotubes when cultured with neutrophils (Fig. 3). We interpret the localized appearance of lanthanum to indicate neutrophil-mediated myotube endocytosis. This interpretation is based on 1) the observation that lanthanum was found in the lumen of cytoplasmic vacuoles (Fig. 3) and in vacuoles of intact and shed membrane blebs (Fig. 6), 2) the appearance of clathrin-coated pits that contained lanthanum (Fig. 7),3) pitted regions on myotube membranes (Fig.8), 4) an apparent endocytotic lanthanum-containing vesicle near a membrane bleb (Fig. 5), and5) the morphological similarities between our lanthanum-containing cytoplasmic vacuoles and endocytotic vesicles previously reported in cultured myotubes (1, 13).
Several possibilities exist to explain the postulated neutrophil-mediated myotube endocytosis. One possibility is that, as a result of myotube injury, myotubes were internalizing injured membrane components. Another possibility is that the increased endocytosis in our injured myotubes may have been an attempt to reseal the membrane. This possibility is based on the work of previous investigators who have established that cytoplasmic vesicles that fuse and reseal injured membrane originate from endocytotic vesicles (2, 30). Further investigation is needed to test these hypotheses.
Plasma membrane blebs have been reported in a variety of cells and is considered a prominent feature of cell injury regardless of its etiology (9, 12, 14, 31). Membrane blebs have been reported in cells undergoing apoptosis or oncosis, and thus their mere formation provides no insight on the type of cell injury (i.e., apoptosis or oncosis) (31). Membrane blebs can rupture or be shed from the membrane and have been suggested to represent irreversible and reversible injury, respectively (9,12, 14). Membrane bleb formation appears to be the result of a disassociation of the plasma membrane lipid bilayer from the cytoskeleton (19). Previous investigators have postulated that ATP depletion, elevations in free cytoplasmic calcium concentrations, activation of calpains, altered thiol status, and ROIs cause the formation of membrane blebs (9).
Fidzianska and Kaminska (6) reported plasma membrane blebs in skeletal muscle of newborn rats 24 h after a myotoxin (bupivacaine) injection. Their plasma membrane blebs were filled with vacuoles, some of which contained presumed remnants of skeletal muscle organelles, an observation that is consistent with apoptosis (31, 34). Our myotube membrane blebs also contained vacuoles; however, these vacuoles contained no visible cellular structures (Figs. 6-8). The organelle-free vacuoles in our membrane blebs is consistent with changes in cell morphology associated with oncosis, a type of prelethal injury in which injured cells swell before necrosis (16, 31). Pyknotic nuclei and apoptotic bodies were also reported at 24 h in bupivacaine-injected rats (6). However, we observed no electron microscopy signs of neutrophil-mediated apoptosis of myotubes. Thus our observations may indicate that the myotube membrane blebs were the result of neutrophil-mediated oncosis.
The results from the present study demonstrate that neutrophils can injure cultured skeletal muscle myotubes. However, the results may not be applicable to the in vivo events associated with muscle injury for several reasons. First, in vitro conditions cannot mimic the complex interactions between the various cell types that are present in injured muscle or the vascular response to muscle injury. Second, our experiments were performed on developing muscle and not adult myofibers. The effect of neutrophils on myotubes may be different from their effect on adult myofibers, since it is likely that differences in muscle defense mechanisms exist between myotubes and myofibers. However, previous investigators have reported proliferation of muscle precursor cells and the formation of myotubes at times when neutrophil concentrations are elevated (10, 26, 27). Thus, based on our results, it is conceivable that neutrophils could exacerbate muscle injury and delay muscle regeneration in vivo by injuring muscle precursor cells and/or myotubes. If this scenario is true, then strategies to ameliorate neutrophil-mediated myotube injury may be important in myoblast transplantation and to enhance muscle regeneration following skeletal muscle injury. Further work is needed to determine the mechanism for the neutrophil-mediated myotube injury and whether similar mechanisms are operating in vivo.
We gratefully acknowledge Eleni Mylona and Susan Tsivitse for their assistance.
This project was supported by the University of Toledo deArce Memorial Research Endowment Fund.
Address for reprint requests and other correspondence: F. X. Pizza, Dept of Kinesiology, The Univ. of Toledo, 2801 W. Bancroft St., Toledo, OH 43606 (E mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2001 the American Physiological Society