MUSCLE wasting and persistent muscle weakness occur in critically ill patients who are treated in the intensive care unit. 1,2The incidence of this complication varies between 47 and 96%. 3The etiology of this muscle weakness is multifactorial 4,5and includes the disease itself, the associated immobilization, 6,7and the concomitant use of muscle relaxants alone or in combination with corticosteroids. 8,9An enquiry by the US Food and Drug Administration concluded that factors that increase susceptibility to myopathy and muscle weakness are the administration of high doses over several days primarily of steroidal relaxants (pancuronium and vecuronium) to patients who may also be receiving corticosteroids, aminoglycoside antibiotics, or both. 10Other recent reports have concurred with this notion. 11,12Although these reports of myopathy implicated steroidal relaxants as an important contributing factor, 9–12other reports implicate benzoylisoquinolinium-type relaxants also to have myopathic effects when combined with corticosteroids. 13,14It has been postulated that steroidal relaxants can behave like corticosteroids or, by some other interaction, potentiate the glucocorticosteroid actions of endogenous steroids. Conversely, steroids are known to inhibit nicotinic receptors 15and can also allosterically potentiate neuromuscular blocking effects of vecuronium in functional acetylcholine receptor channels expressed in vitro . 16
Denervation itself results in an up-regulation of glucocorticoid receptors in skeletal muscle that might make the muscle more susceptible to the effects of steroids. 17Muscle relaxants, when used for prolonged periods, by binding to the acetylcholine receptors on the muscle membrane and preventing neurotransmission induce a chemical denervation confirmed by the up-regulation of acetylcholine receptors. 18,19All muscle relaxants, therefore, have the potential to enhance the interaction of corticosteroids, steroidal relaxants, or both on the muscle steroidal receptors. Whether steroidal relaxants do in fact behave like or potentiate corticosteroids has not been tested because it has been difficult to differentiate the denervation effects induced by the steroidal muscle relaxants from those of steroids.
Corticosteroids activate transcription of target genes by binding to the glucocorticoid receptor, 20which belongs to an important class of ligand-regulated transcription factor. Regulation of the transcription of genes plays a central role in a variety of cellular processes and is tightly controlled. The glucocorticoid receptor, when activated by steroids, translocates to the nucleus and recognizes and binds to the specific DNA motif in the regulatory elements of target genes, termed glucocorticoid-responsive element (GRE). 21,22The sequence of the canonical GRE is 5′-GGTACA nnn TGTTCT-3′. 21,22When a ligand (corticosteroid) is bound, glucocorticoid receptor complexes with GRE and alters transcription of target genes.
An in vitro system using muscle cell cultures is more amenable to elucidate the interaction of corticosteroids and steroidal relaxants on muscle steroidal receptors because the confounding effect of nerve denervation does not come into play; muscle cells in culture devoid of the nerve input behave like denervated muscle cells. 23The current study, using C2C12 cultured skeletal myocytes and luciferase assay, compares the intrinsic steroidal effects of prototypical steroid relaxants, pancuronium, and vecuronium to d-tubocurarine and dantrolene, representing nonsteroidal compounds, which also have actions on muscle. The interaction of these compounds with dexamethasone was also examined.
Materials and Methods
Dexamethasone, dantrolene, d-tubocurarine, and pancuronium were purchased from Sigma (St. Louis, MO). Vecuronium was provided by Akzo Nobel (Chicago, IL). Mouse C2C12 myoblasts (American Tissue Culture Collection, Manassas, VA) were maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum (Sigma). To avoid the influence of glucocorticoids contained in fetal bovine serum, the medium was replaced with serum-free medium, Opti-MEM (Invitrogen, Carlsbad, CA), and then C2C12 myoblasts were seeded onto a 12-well plate. At 24 h after plating, the cells were transfected with pMMTV-Luc (0.8 μg), p6RGR (0.1 μg), and pRL-TK (0.05 μg) (Promega, Madison, MI) using Lipofectamine 2000 (Invitrogen). p6RGR is a mammalian expression vector containing complementary DNA (cDNA) for full-length glucocorticoid receptor driven by viral promotor, RSVLTR. The pRL-TK (Promega) contains cDNA for luciferase of sea pansy (Renilla reniformis ) that is linked to the constitutive eukaryotic promoter.
To assess the effects of muscle relaxants on glucocorticoid receptor and GRE-mediated transcription, the mouse mammary tumor virus (MMTV) promoter that contains multiple GREs was fused upstream of the firefly luciferase. The expression of firefly luciferase in this plasmid, pMMTV-Luc, is strongly activated in the presence of glucocorticoid ligand and glucocorticoid receptor. Luciferase assay is one of reporter gene assays to study gene transcription and is currently in more extensive use for promoter analysis because of better sensitivity than chloramphenicol acetyltransferase or other commonly used reporter gene assays. Luciferase enzyme of the firefly (Photinus pyralis ) 24catalyzes the reaction for the luciferase assay; adenosine triphosphate + luciferin + oxygen, when catalyzed by luciferase, → adenosine monophosphate + oxyluciferin + pyrophosphate + light (560 nm). Because posttranslational modification is not required for its activity, the enzyme activity, as measured by the intensity of light output, is a sensitive surrogate marker of transcription. Luciferases of the firefly and sea pansy use different substrates and possess distinct biochemical properties. 25,26The activity of luciferase of sea pansy derived from pRL-TK construct, which reflects transfection efficiency, was used as an interassay standardization control. GRE-mediated transcription was determined by the activity of firefly luciferase normalized to that of sea pansy luciferase.
At 24 h after the transfection, the various concentrations of muscle relaxants (pancuronium: 10−12~10−4m; vecuronium: 10−12~10−4m; dantrolene: 10−6m; d-tubocurarine: 10−6m) and dexamethasone (10−12~10−9m ) were added to the culture media, and the cells were incubated for an additional 24 h. Then, the cells were harvested for luciferase assay. Luciferase activities were measured using the Dual-Luciferase Reporter System (Promega) and a luminometer (Turner Designs Instrument, Sunnyvale, CA). The treatment with these agents for 24 h did not affect the viability of the cells, as shown by trypan blue exclusion test.
Glucocorticoid-responsive element–mediated transcription was compared using one-way analysis of variance followed by the Scheffé F test. The null hypothesis was rejected when P was less than 0.05. All values are reported as mean ± SEM.
First, pancuronium effects on the transcription mediated by GRE were examined by luciferase assay in mouse C2C12 myoblasts transfected with the glucocorticoid receptor expression vector and the reporter genes. Dexamethasone was used as a positive control, and d-tubocurarine and dantrolene were used as negative controls. The treatment with dexamethasone resulted in a marked induction of luciferase activity in a dose-dependent manner. In contrast, pancuronium did not induce luciferase activity at the wide range of concentrations tested (fig. 1). Likewise, neither dantrolene nor d-tubocurarine affected luciferase activity. The other prototypical steroidal relaxant, vecuronium, also did not increase GRE-mediated transcription of luciferase in C2C12 myoblasts (fig. 2).
Next, to address whether pancuronium or vecuronium modulates dexamethasone-induced GRE-mediated transcription, we examined the effects of combined administration of varying concentrations of muscle relaxants and dexamethasone on the induction of luciferase. Dexamethasone caused luciferase induction in a dose-dependent manner, whereas the superimposition of neither pancuronium nor vecuronium altered dexamethasone-induced, GRE-mediated luciferase activities at all the concentrations of dexamethasone tested (10−12~10−8m figs. 3A and B).Dantrolene (10−6m) and d-tubocurarine (10−6m) also failed to show any effects on dexamethasone-induced GRE-mediated transcription (fig. 3A).
Glucocorticoid receptor is a member of the nuclear receptor family of ligand-dependent transcription factors. 22In the absence of its ligand, corticosteroid, the glucocorticoid receptor is inactive and forms a large heteromeric cytoplasmic complex with several other proteins, including heat shock proteins. The binding of corticosteroid to its receptor, however, leads to dissociation of heat shock proteins from the receptor, and homodimers of the ligand-glucocorticoid receptor are formed. Mobilization of the ligand-receptor complex to the nucleus, binding to GRE, and recruitment of coactivators, including histone acetyltransferases, to the receptors leads to activation of transcription; the whole process occurs in a matter of seconds to several minutes.
In our studies, we exposed the test drugs for 24 h, allowing sufficient time for translocation to nucleus and activation of GRE-mediated transcription. Our results demonstrate that neither pancuronium nor vecuronium modulated GRE-mediated transcription in the cultured skeletal muscle cells (figs. 1 and 2). In terms of GRE-mediated transcription, no difference was found between steroidal (pancuronium and vecuronium) and nonsteroidal (d-tubocurarine) muscle relaxant, as well as dantrolene, a ryanodine receptor antagonist in muscle. Based on our preliminary observation that pancuronium did not induce GRE-mediated transcription in simian kidney COS-7 cells and human Jurkat lymphoma cells (data not shown), the failure of pancuronium to induce GRE-mediated luciferase transcription should be considered not to be cell-type specific. It is important to note, however, that the main purpose of the study was to examine steroidal effects on muscle.
Because some previous studies suggested that treatment of critically ill patients with the combination of corticosteroids and steroidal nondepolarizing muscle relaxants was associated with severe muscle atrophy and delayed functional recovery of muscle, a question was raised whether or not steroidal muscle relaxants may enhance the adverse effects of corticosteroids. The current study, however, clearly demonstrates that the combined administration of pancuronium or vecuronium together with corticosteroid does not influence the effects of dexamethasone on the glucocorticoid receptor and GRE-mediated transcription. Both dantrolene and d-tubocurarine also did not enhance steroidal effects of dexamethasone (fig. 3). Therefore, the neuromuscular dysfunction associated with steroidal relaxants and steroids in combination is unrelated to its glucocorticoid receptor–mediated effects on transcription. In retrospect, this misconception is probably related to the fact that pancuronium and vecuronium were overwhelmingly the most commonly administered relaxants at one time because of the lack of histamine-releasing properties. In the current intensive care unit setting, with availability of other relaxant drugs, myopathy of critical illness has now been observed even with nonsteroidal relaxants and steroids in combination (e.g. , atracurium). 13,14
Steroids have a profound myopathic effect, 27–29an effect seen even in the absence of denervation. 29Similarly, muscle relaxants themselves induce a denervation-like state, 18,19which can induce muscle wasting and weakness. Critical illness itself and the attendant immobilization, denervation, or both independent of each other lead to insulin resistance, 7,30,31which can result in increased protein catabolism and decreased protein synthesis, partially accounting for the muscle wasting seen in critical illness. Finally, there is evidence that corticosteroids have additive effects on inhibition of neurotransmission block produced by muscle relaxants by allosteric actions on the nicotinic acetylcholine receptor itself. 15,16All of these different effects, to a greater or lesser extent, may account for the myopathy when muscle relaxants are used in combination with steroids.
The current study documents that neither pancuronium nor vecuronium by itself exerts corticosteroid-like action on transcription, and neither of them affect genomic actions of corticosteroids. The myopathy and prolonged muscle weakness observed in critically ill patients is not compounded by an effect specific to steroidal relaxants mimicking a steroidal effect on the glucocorticoid receptor. The fact that prolonged paralysis was seen in patients receiving nonsteroidal relaxants, atracurium and cisatracurium, 13,14confirms that neuro-muscular relaxants may have direct neuromuscular toxicity and that the myopathy with relaxants is unrelated to the corticosteroid effect of the steroidal relaxants.