J Neurol DOI 10.1007/s00415-009-5362-5

ORIGINAL PAPER

Enhanced muscle shortening and impaired Ca2+ channel function in an acute septic myopathy model Oliver Friedrich • Ernst Hund • Frederic von Wegner

Received: 18 April 2009 / Revised: 29 September 2009 / Accepted: 15 October 2009 Ó Springer-Verlag 2009

Abstract Myopathies in critically ill patients are increasingly documented. Various animal models of chronic sepsis have been employed to investigate reduced membrane excitability or altered isometric contractility of skeletal muscle. In contrast, immediate changes occurring during acute sepsis are significantly under-characterised; L-type Ca2? channel function or isotonic shortening are examples. We recorded slowly activating L-type Ca2? currents (ICa) in voltage-clamped single intact mouse skeletal muscle fibres and tested the effects of acute challenge with serum fractions from critical illness myopathy patients (CIM). Using a high-speed camera system, we simultaneously recorded unloaded fibre shortening during isotonic contractions with unprecedented temporal resolution (*1,600 frames/s). Time courses of fibre lengths and Electronic supplementary material The online version of this article (doi:10.1007/s00415-009-5362-5) contains supplementary material, which is available to authorized users. O. Friedrich
F. von Wegner Medical Biophysics, Department of Systems Physiology, Institute of Physiology and Pathophysiology, University of Heidelberg, INF 326, 69120 Heidelberg, Germany O. Friedrich (&) School of Biomedical Sciences, University of Queensland, Brisbane, QLD 4072, Australia e-mail: [email protected]; [email protected] E. Hund Department of Neurology, University of Heidelberg, INF 400, Heidelberg, Germany F. von Wegner Brain Imaging Centre, Johann Wolfgang Goethe University, Schleusenweg 2-16, 60528 Frankfurt am Main, Germany

shortening velocity were determined from automated imaging algorithms. CIM fractions acutely induced depression of ICa amplitudes with no shifts in ICa–V-relations. Voltage-dependent inactivation was unaltered and ICa activation and inactivation kinetics were prolonged compared to controls. Unexpectedly, maximum unloaded speed of shortening was slightly faster following CIM serum applications, suggesting a direct action of CIM serum on weak-binding-state cross-bridges. Our results are compatible with a model where CIM serum might acutely reduce a fraction of functional L-type Ca2? channels and could account for reduced SR Ca2? release and force production in CIM patients. Acute increase in isotonic shortening velocity might be an early diagnostic feature suitable for testing in clinical studies. The acute challenge model is also robust against atrophy or fibre type changes that ordinarily would have to be considered in chronic sepsis models. Keywords Skeletal muscle
Critical illness myopathy
Patient serum fractions
Ca2? currents
Muscle shortening
High-speed imaging Abbreviations CIM Critical illness myopathy CS Control sera ICa L-type Ca2? current DHPR Dihydropyridine receptor MWCO Molecular weight cut-off RyR Ryanodine receptor SR Sarcoplasmic reticulum sdec ICa Inactivation time constant TTP ICa Time-to-peak vu Unloaded shortening speed

123

J Neurol

Introduction Acquired neuromuscular weaknesses are a major morbidity factor in critically ill patients worldwide [1–3]. They often become apparent only at late stages, e.g. weaning failure or paralysis/plegia when patients awake from sedation. Critical illness polyneuropathy (CIP) and myopathy (CIM) either occur alone or in combination [4, 5]. Their differentiation is difficult and requires special electrophysiology and biopsy tests [6–8]. Skeletal muscle is affected in a biphasic pattern: functionally, early membrane inexcitability [4] and, morphologically, late myosin hyperproteolysis with ‘ghost sarcomeres’ (thick filament myopathy) are manifested [3, 9]. Several risk factors trigger CIM: sepsis [4, 10, 11], steroids [7, 12, 13] or neuromuscular blockers [14]. For sepsis, CIM can be considered a form of organ failure of the peripheral nervous system [1]. During advanced sepsis, muscle force in animal models is reduced presumably at the level of the motor proteins [10, 15]. This might also apply to septic ICU patients when muscle weakness is apparent [11], but the precise mechanisms have yet to be clarified. As sepsis and multi-organ failure worsen, a deeper knowledge of muscle affection during early sepsis is crucial. Little is known about how sepsis initially triggers neuromuscular dysfunction. Skeletal muscle is highly specialised: membrane excitation activates motor proteins and force production through a tightly regulated cascade [excitation–contraction (ec) coupling [16]]. Dihydropyridine (DHPR) voltagesensors in transverse tubules activate ryanodine receptor release channels (RyR1) on the sarcoplasmic reticulum (SR). Ca2? released from SR binds to myofilaments and initiates cross-bridge cycles. Weak cross-bridge binding states are predominantly recruited during unloaded shortening and strong-binding states during contraction against external load [17]. Maximum shortening velocity is an important functional parameter in muscle physiology, reflecting maximum cross-bridge cycling rates [18]. Chronic sepsis animal models (cecum puncture and ligation, CLP) have showed marked isometric force drops in whole muscle [10] or permeabilised fibre bundles [15]. However, such recordings do not necessarily reflect the cellular response to sepsis in the intact individual muscle cell. Likewise, very little is known about immediate responses of muscle cells to septic loads [19]. For example, although the contractile proteins may show reduced Ca2? sensitivity during the course of sepsis [15], earlier steps in the ec-coupling cascade might compensate for such detriments. Intact fibre experiments in acute sepsis models are important because they more closely reflect the pathophysiology where intact muscle fibres are acutely exposed to serum capillary filtrate (from the systemic circulation)

123

that contains inflammatory mediators, septic agents or toxic tissue factors. We recently introduced a now firmly established acute septic challenge assay during which blood serum fractions from septic myopathy patients were applied to animal skeletal muscle. This approach is suitable to test: (1) whether specific short-term reactions of single muscle fibres to serum challenges, e.g. membrane excitability or SR Ca2? release, could predict early force detriments; and (2) whether these effects were confined to specific molecular weight serum fractions [19]. Indeed, fibres were acutely depolarised, SR Ca2? release was impaired but Na? channel availability increased. However, in intact muscle subject to prolonged depolarisation, DHPRs also serve as Ca2? channel [20, 21]. Functional data on L-type Ca2? channel properties in muscle during early sepsis are not available. Here, we tested acutely applied human CIM serum on intact single mouse muscle fibres and recorded DHPR properties and fibre shortening kinetics.

Methods Patient serum samples and single mouse muscle fibre assays Serum samples from a previous patient cohort were used (five septic ICU patients, mechanically ventilated for 7 days) [19]. Sera were pooled to obtain maximum amounts of putative myotoxic activity [19]. This approach was chosen for the following reasons: available serum volumes were limited (\1.5 ml per patient) and only a few successful recordings would otherwise have been obtained patient-wise (using *150 ll per experiment); and we wanted to perform the experiments under septic load conditions identical to our previous study [19]. We stress that single muscle fibre experiments were difficult to perform given the combination of electrophysiology and high-speed imaging that limited the number of successful recordings. Patients receiving corticosteroid therapy [300 mg/day for various indications were excluded so that we could examine the sepsis aspect of CIM. In addition, high dose steroids are capable of inducing a steroid-myopathy [12]. All patients had clinical evidence of CIM by day four of onset of critical illness (evaluated by SOFA, APACHE scores). Diagnosis was based on clinical and electrophysiological examinations using probable CIM criteria [3, 5, 22]. Direct muscle stimulation [8] could not be routinely performed in our clinics. Prior consent for routine muscle biopsies could not be obtained. Collection of serum samples was approved by the Local Ethics Committee and therefore was performed in accordance with the ethical standards described in the 1964 Declaration of Helsinki.

J Neurol

Informed consent was obtained from all individuals. For some experiments, sera aliquots from healthy control individuals who were enrolled in our previous study were available (CS or control serum) [19]. Sera were centrifuged at different molecular weight cut-offs (MWCO 5–100 kDa) and stored in 150 ll aliquots at -20°C until time of use to avoid unnecessary freeze–thaw cycles. To assess serum effects on voltage-clamped single muscle fibres, serum fractions were added to the external solution (1:20 v/v dilution) and recordings were performed 5 min after incubation. Test solutions, pH buffered with 10 mM HEPES, did not change pH; osmolarity was also unchanged. Interossei muscles from adult BALB/c mice and single fibres obtained by enzymatical treatment were used in our animal model assay [19]. The ‘principle of laboratory animal care’ (NIH publication No. 86–23, revised 1985) and relevant national laws were followed. Control experiments refer to saline without serum addition (saline). Voltage-clamp recordings of slowly activating L-type Ca2? currents (ICa) L-type ICa were recorded with a two electrode voltageclamp technique in single intact fibres before (saline) and following application of CIM or, in some experiments, control sera fractions. The 10 mM Ca2? containing external solution [23] was isotonic for experiments with simultaneous high-speed imaging of shortening during voltage-clamp pulses. To assess the voltage dependence of ICa amplitudes, ICa activation and inactivation kinetics in the same fibre, contraction was blocked by raising external osmolarity with 300 mM sucrose. From a holding potential of -70 mV, ICa were activated by 0.9 s lasting depolarising potential steps starting from -30 mV in 10 mV increments. ICa amplitudes were normalised to fibre surface area. ICa–V-plots, time-to-peak (for ICa activation) and time constants of the exponential decay of ICa, sdec (for ICa inactivation) were analysed under ‘saline’ conditions (without serum) and *5 min following serum fraction applications. Steady-state inactivation, F?, of ICa was obtained from two-pulse protocols [23]. High-speed imaging of single fibre unloaded shortening To accelerate imaging of the initial shortening of muscle fibres, we increased our previous time resolution [24] by connecting a CMOS high-speed camera (1200 hs, PCO, Kehlheim, Germany, *640 fps at 1,280 9 1,024 pixels) to the microscope of the voltage-clamp setting. Resizing region-of-interest to fibre containing area increased frame rates up to *1,600 fps at 300 9 400 pixels. The CMOS sensor was TTL-triggered to a voltage step to 0 mV. Fibre imaging was pursued for 1.5 s during which between 1,500

and 3,000 frames were recorded. For offline analysis of image series, fibre length detection and tracking were performed with an automated imaging algorithm developed in IDL programing environment (IDL 6.0, RSI Research Systems Inc., Boulder, USA). After noise removal (3 9 3 median filter) from all images in a series, fibre edges were detected by thresholding and the images binarised. On each binary image, a 2-dimensional array was defined collecting all pixel positions along the fibre circumference. This was followed by extracting their minimum and maximum vertical positions and calculating their corresponding horizontal coordinate from the median of all x coordinates fulfilling the min/max y-coordinate criterion. From upper and lower fibre edges of pixel-pairs, apparent fibre length l(t) was calculated using elemental trigonometry; this procedure was automatically repeated for all sequence images. With the acquisition rate, the time course of unloaded speed of shortening vu(t) was derived either using: (a) three-point Lagrangian interpolation; or (b) onepoint forward differentials [24]. vu,max values derived from procedure (b) were taken as the upper speed limit for unloaded shortening [24]. Data presentation and statistics Data are presented as mean ± SD or ±SEM where indicated. Significance was assessed with Student’s t test or one-way ANOVA (p level 0.05).

Results Effects of CIM serum fractions on L-type Ca2? currents (ICa) in skeletal muscle Figure 1 shows ICa traces in two single fibres where peak ICa amplitudes were reduced 5 min after application of 50 or 5 kDa CIM serum fractions. ICa–V-plots show reduced ICa amplitudes but no shifts in the curves. Peak ICa reduction with no shifts in ICa–V-plots was confirmed for all CIM serum fractions compared to saline in surface normalised ICa–V-plots from several fibres (Fig. 2). One control serum recording (50 kDa MWCO) was obtained (Fig. 2b) that showed no change to ICa, suggesting that ICa reductions may be unique to CIM serum. Minimum values from the ICa–V-relations showed no systematic ICa,min– MWCO relation (Fig. 2c). Peak ICa at a given membrane potential depends on steady-state availability of Ca2? channels at the holding potential from which depolarisation is initiated, among other factors. Steady-state ICa inactivation curves, F?, were obtained from two-pulse protocols (Fig. 3). Figure 3a shows an example of traces in a single fibre after acute application

123

J Neurol

Fig. 1 Slowly activating L-type Ca2? currents (ICa) after acute application of CIM serum fractions. ICa example traces in single intact mouse interossei fibres under saline conditions (left) and 5 min after

application of either 50 kDa (a) or 5 kDa (b) CIM serum fractions (middle). In both cases, a marked reduction in peak ICa amplitudes without a shift in individual ICa–V-relations (right) can be seen

Fig. 2 CIM serum fractions reduce ICa amplitudes in intact muscle fibres. a ICa amplitudes were consistently reduced by addition of CIM serum fractions with no change in minimum amplitude potential (0 mV) or reversal potential. b Application of 50 kDa control serum (CS) fractions did not alter either peak ICa amplitudes nor ICa–Vrelation compared to saline conditions. CIM serum fractions, however, reduced peak amplitudes. c Summary of serum effects on normalised minimum peak ICa amplitudes, ICa,min, show a marked reduction for CIM fraction compared to saline

of 5 kDa CIM fraction. With more positive pre-pulse potentials, L-type Ca2? channel fractions still available during the subsequent 0 mV test pulse decrease. This fractional availability is presented in the middle panel for the

123

same fibre under saline and after acute 5 kDa CIM serum conditions. From the F0.5 potentials of the Boltzmann-fits in this and another fibre, F? was identical for CIM serum and saline. Figure 3b shows similar results in a fibre following

J Neurol

Fig. 3 Effects of CIM serum on voltage-dependent inactivation of ICa. a ICa traces from a double-pulse protocol (inset) in a single fibre 5 min after application of 5 kDa of CIM serum. The relative fraction of available channels during the test pulse, F?, for a given pre-pulse potential is shown before (saline) and after CIM serum application

(middle). Half-inactivation potentials in this and another fibre post 5 kDa CIM serum application were identical to the saline value (right). b Example of F? in another fibre after applying 30 kDa CIM fractions. c F0.5 values for different fractions tested

application of 30 kDa CIM fractions. Figure 3c summarises F0.5 values for all CIM conditions and one available control condition. Although F0.5 values appear more negative with increasing MWCO, i.e. a left-shift of F? curves, this was not significant (p [ 0.5). The results clearly show that reduced ICa after CIM serum applications cannot be due to changes in voltage-dependent Ca2? channel inactivation at -70 mV for which all channels are fully available. Another determinant of ICa amplitudes is given by ICa activation and inactivation kinetics [25]. Time-to-peak (TTP) and ICa decay time constants (sdec) for all traces are summarised (Table 1). TTP decreased (faster ICa activation) with positive potentials but sdec values were more variable. TTP values were consistently larger after CIM serum applications compared to saline with a tendency for larger TTP values at lower MWCO. A similar trend was observed for sdec values (larger for CIM fractions vs. saline; a decrease with larger MWCO). This suggests CIM serum fractions slow down channel activation/inactivation kinetics. Assuming first order kinetics, ICa amplitudes are determined by sdec:TTP with larger ratios representing slower inactivation and, thus, larger amplitudes [25]. For example, at 10 mV, ratios were *2.3 for saline and 2.5–4.5 for CIM fractions. The still reduced ICa amplitudes following acute CIM serum application suggest that CIM sera act at the single channel level, affecting either conductance or open probability [26], and that the number of functional channels would be acutely decreased by CIM fractions.

Simultaneous unloaded shortening and ICa recordings after serum applications Figure 4 shows a high-speed recording in a single fibre after applying 5 kDa CIM serum. The image series in Fig. 4a contained 1,860 images during the *1.5 s recording period and ICa was simultaneously recorded during the 0 mV pulse. The first image shown (#02) is taken 887 ls after the initial resting frame and highlights the unprecedented recording speed. The binarised images from the imaging algorithm show reliable circumference detection and the maximum and minimum position of the fibre at its edges (horizontal lines) from which the length was calculated. l(t) normalised to resting fibre length L0 is shown in Fig. 4b, unloaded speed of shortening vu(t) in Fig. 4c and time course of ICa(t) in Fig. 4d. Minimum fibre length during shortening was *0.5 l/L0 in this example. vu(t) traces are shown for two differentiation procedures (‘‘Methods’’). Figure 5 summarises the results of several fibres under saline conditions and after applications of 5 kDa control or CIM serum fractions. Minimum shortening length was unchanged under all conditions (p = 0.53, one-way ANOVA). l(t) was described by a double-exponential fit with time constants s1 and s2 [24]. The fast s1 was not different between groups (p = 0.29, one-way ANOVA). However, s2 was significantly smaller for CIM than for CS fractions (p \ 0.01, unpaired t test). Interestingly, vu,max was similar between groups (p = 0.81,

123

J Neurol Table 1 Activation (Time-to-peak) and inactivation kinetics (time constant, sdec) of L-type Ca2? currents (ICa) after applying control or CIM serum fractions MWCO (kDa)

Saline

Serum

[

TTP, sdec (ms)

Membrane potential -10 mV

0 mV

10 mV

20 mV

TTP

114 ± 8

91 ± 6

83 ± 8

77 ± 15

sdec

372 ± 257

322 ± 164

194 ± 25

240 ± 57

5

CIM

TTP sdec

237 ± 38 710 ± 185

156 ± 6 626 ± 145

139 ± 14 470

121 ± 24 756

30

CIM

TTP

189 ± 42

132 ± 28

129 ± 19

135 ± 28

50

CIM

100

sdec

569 ± 376

554 ± 281

607 ± 234

n.a.

TTP

194

92

116

79

sdec

n.a.

264

373

500

CS

TTP

182 ± 24

130 ± 12

127 ± 11

188 ± 41

sdec

455 ± 148

510 ± 158

401 ± 68

481

CIM

TTP

197 ± 68

112 ± 27

109 ± 24

120 ± 43

sdec

122 ± 25

302 ± 168

276 ± 107

259 ± 82

Saline: no serum present, n up to 13 single fibres for each condition, mean ± SD MWCO molecular weight cut-off, CS control sera, CIM critical illness myopathy, TTP time-to-peak Fig. 4 High-speed recordings of unloaded shortening and ICa during voltage-clamp activation. a Selection of fibre images from a high-speed recording of fibre length at the time points indicated in a single fibre after application of 5 kDa CIM serum fraction to the bath. The image number in the sequence is denoted by the # sign. The acquisition rate of *1.12 kHz in this fibre allowed each image to be collected at *890 ls. Binarised images from the IDL algorithm that detected the fibre circumference and vertical minimum and maximum position are also shown. The time course of fibre length l(t), unloaded speed of shortening vu(t) and normalised ICa(t) are plotted in b–d for the same time points. vu(t) is calculated from a symmetrical differential environment (black circles) and a one-step forward differential environment (grey circles) (‘‘Methods’’)

123

J Neurol Fig. 5 Shortening parameters in single fibres after 5 kDa CS or CIM serum applications. Minimum relative fibre length (lmin/L0), fast (s1) and slow (s2) time constant of shortening kinetics and maximum shortening speed vmax are summarised for several single fibres following either 5 kDa control (CS) or CIM serum fractions compared to saline conditions in a–d. #P \ 0.05. e Relationship between vmax and acquisition rates

one-way ANOVA). To estimate a putative upper limit for vu,max at increased time resolution, individual values were plotted against the acquisition rate (Fig. 5e). The similar vu,max values were pooled but are still colour coded according to experimental origin (saline, CS, CIM). The data points at 0.5 kHz were previously determined [24]. The data suggest an upper limit for vu,max around 19 L0 s-1 when fitting the data with a sigmoidal (not shown in Fig. 5). Figure 6 shows individual ICa recordings during shortening after 5 kDa CS or CIM serum applications. In contrast to recordings under hypertonic conditions (Fig. 1), almost all ICa traces showed a second outward component during shortening. As these might affect absolute peak current amplitudes, peak ICa was not compared under isotonic conditions. Expanded views of the early ICa traces scaled to their peak (Fig. 6) allowed comparison of activation and inactivation kinetics under isotonic conditions between CS and CIM (Fig. 6c). Similar to previous studies [23], kinetics were much more rapid in isotonic compared to hypertonic conditions (Table 1). Both activation (p [ 0.22) and inactivation kinetics (p [ 0.34) were not significantly different between CS and CIM fractions but showed a tendency for slightly larger values following CIM serum applications.

Discussion Changes to intact muscle on the single fibre level in sepsisinduced myopathies remain poorly characterised. Much of our knowledge on CIM pathophysiology comes from animal models using chronic sublethal CLP [15, 27, 28] or LPS-induced sepsis [29] with experiments performed over hours or days. Here, we specifically addressed acute changes to muscle contractile activation by serum fractions from septic CIM patients in an established intact single muscle fibre assay [19]. This study adds two major new

Fig. 6 ICa currents simultaneously recorded during high-speed imaging in single fibres after 5 kDa control (CS) or CIM sera applications. ICa traces for all fibres analysed in Fig. 5 after acute CS (a) or CIM fraction (b) application. To better compare activation and inactivation kinetics, all traces are also shown scaled to peak (right). c Activation and inactivation was slightly prolonged after application of CIM serum

findings to our understanding of the pathophysiological mechanism at very early stages of septic challenge in skeletal muscle: (1) L-type Ca2? channel function is acutely impaired; and (2) isotonic shortening kinetics is acutely accelerated.

123

J Neurol

Acute serum effects on muscle contractility To our knowledge, cross-bridge kinetics has not been studied in septic muscle. Accelerated isotonic shortening after CIM serum challenge is surprising, as Ca2? regulation and isometric force development were impaired in septic animal models [10] or using the same acute assay [19]. In fact, isotonic unloaded shortening even showed improved function, e.g. larger shortening fractions, faster shortening time constants and slightly larger maximum shortening velocities after acute CIM serum incubation. This finding suggests that cross-bridges seem to be differentially affected by acute and chronic exposure to septic loads. Only a small fraction (1–5%) of cross-bridges that would be firmly attached during isometric contractions occupies strong-binding states during maximum unloaded shortening [17], while the majority cycle through weak-binding configurations. Low molecular weight CIM serum fractions may acutely recruit weak-binding states in single fibres. This is a novel finding that contrasts with isometric contractions, e.g. tetanic stimulation of whole intact muscles [10] or Ca2? activation of permeabilised muscle bundles in sepsis models [15, 19]. Our results imply that cross-bridge cycling might increase even though force production decreases following acute CIM serum applications. Isometric contractions in animal models have solely been used to explain subcellular mechanisms of force loss seen in critically ill patients [11]. For example, force transient amplitudes in skinned mouse muscle fibre bundles decreased after the first three days following a CLP procedure, although they ultimately recovered [15]. Although force in sham-operated groups behaved similarly at day 2 (when clinical sepsis was most pronounced), after correcting for Ca2? sensitivity of the myofilaments, SR Ca2? release was already significantly reduced in CLP compared to sham animals [15]. However, it is not clear from these data whether the decrease in force solely depends on SR Ca2? release and Ca2? sensitivity, as a kinetics analysis of the force transients was not conducted to unravel differences in SR Ca2? pump (SERCA) activity [15]. Different loading contents of the store can also account for the observed differences. In contrast, in our previous study using acute serum application, both caffeine-induced force transient amplitudes and myofibrillar Ca2? sensitivity were generally reduced by CIM serum compared to control serum, with the exception of small molecular weight fractions, i.e. 5 kDa, where Ca2? sensitivity was even increased [19]. Based on similar decays of Ca2? transients, we concluded that myoplasmic Ca2? removal by SERCA was not primarily affected by CIM serum [30]. The acute septic load model has crucial advantages over chronic models using septic muscle. During sepsis and critical illness, muscle can undergo atrophy [31] that can

123

skew reduced isometric force levels [15]. Indeed, when relating isometric force to cross-sectional area (CSA), an increase in specific force was seen in extensor digitorum longus (EDL) muscle from septic rats following CLP [10]. In our acute model, no change in CSA can occur in the same muscle fibre between saline and 5 min post-serum application recordings. Interestingly, in the case of crossbridge properties, passive force increased in EDL muscle from chronically septic rats (10 days after CLP) compared to controls or sham-operated animals [10]. This could be attributed to a decrease in series elasticity (e.g. changes in properties/amount of connective tissue or intracellular titin) or impaired cross-bridge properties, i.e. larger fraction of cross-bridges in the rigor state due to impaired detachment rates. Tissue fibrosis was not detected in histological sections of muscles from chronically septic rats [10]. In a recent study in chronically septic CLP rats, isometric twitch contraction and relaxation phases were also significantly accelerated [10]. Direct comparison to our results is difficult in that SR Ca2? release kinetics, crossbridge properties and cytoplasmic Ca2? removal rates will likely differ during a single isometric twitch [10] versus unloaded shortening under maintained depolarisations. Speed of shortening is further crucially determined by myosin ATPase activity and phosphorylation states [32]. For example, it is known that during muscle atrophy in bed rest and in muscle unloading models, fibre type transitions from slow type I to fast type II fibres occur [33, 34]. In critically ill patients with sepsis who develop CIM, this transition is significantly less characterised [35], but a preferential decrease in type II fibre diameters is observed [9]. In addition, exhaustability of muscle during chronic sepsis increases, probably due to a larger proportion of type II fibres [10]. Again, our acute septic challenge model is independent of atrophy or fibre type transitions in the time range of experimentation. Therefore, our results of accelerated contraction kinetics during unloaded shortening after CIM serum challenge suggest an early direct action of septic mediators on the cross-bridge kinetics level. Effects of CIM serum fractions on L-type Ca2? channel function Studies on early or late sepsis affecting L-type Ca2? channel (DHPR) function in skeletal muscle have not been conducted to date. This is surprising given that proper DHPR function is needed to convey tubular membrane excitation to Ca2? release via the RyR1. This interaction is regulated by many intracellular factors, e.g. Mg2?, ATP, Ca2?, Ca2?-binding proteins [36], but the RyR1 itself is also a target for many drugs [37] and putatively also for myotoxic factors in CIM serum [19]. Long-term blockade

J Neurol

of L-type Ca2? channels upregulates membrane DHPR expression [38]. To gain initial insight into the effects of acute septic challenge on DHPR function, we recorded L-type Ca2? currents in our model. Both activation and inactivation were slightly prolonged by 5 kDa CIM serum fractions. Compared to controls, ICa–V-relations suggested depressed peak amplitudes with no shifts in ICa–V-relation. Likewise, steady-state inactivation was unchanged (Fig. 3). This suggests that CIM fractions may acutely impair single channel properties of functional L-type channels. Possible serum factors that alter ec-coupling in septic muscle The question ’what factors are responsible for myotoxic activity in serum from septic CIM patients?’ has not been decisively answered to date. In a recent study on elementary Ca2? release events (ECRE) in cultured adult skeletal muscle fibres, ECRE increased in dedifferentiating fibres, but only when cultured in serum-containing medium. It was suggested that an unidentified serum factor could relieve inhibition of the DHPR on RyR1 [39]. In cardiomyocytes, acute application of inflammatory cytokines had differential effects on ICa. IL-6, a cytokine produced by working muscle, did not change basal ICa amplitudes [40]. Besides, IL-6 is also markedly higher in patients with septic-shock [41] and one of its biological roles is believed to inhibit detrimental tissue effects of other pro-inflammatory cytokines, e.g. TNF-a [40]. TNF-a, in turn, depressed ICa amplitudes within a few minutes after application to rat cardiomyocytes while SR Ca2? release increased [42]. Interestingly, a recent study that also applied septic-shock patient serum fractions to adult rat cardiomyocytes found a fast contractile depression progressing over 5–60 min [41], thus corroborating the validity of our acute serum application approach to study early effects of septic challenge. Therefore, our technique provides a tool to also study specific effects of other inflammatory processes on muscle function at the single cell level. Acknowledgments This study was supported by a Faculty Grant from the Medical Faculty of the Ruprecht-Karls-University (F.203694) and through an Australian Research Council International Fellowship (ARCIF) awarded to one of the authors (OF).

References 1. Friedrich O (2008) Critical illness myopathy: sepsis-mediated failure of the peripheral nervous system. Eur J Anaesthesiol 25(suppl 42):73–82 2. Latronico N, Shehu I, Seghelini E (2005) Neuromuscular sequelae of critical illness. Curr Opin Crit Care 11:381–390

3. Bolton CF (2005) Neuromuscular manifestations of critical illness. Muscle Nerve 32:140–163 4. Khan J, Harrison TB, Rich MM, Moss M (2006) Early development of critical illness myopathy and neuropathy in patients with severe sepsis. Neurology 67:1421–1425 5. Bednarik J, Lukas Z, Vondracek P (2003) Critical illness polyneuromyopathy: the electrophysiological components of a complex entity. Intensive Care Med 29:1505–1514 6. Lefaucheur JP, Nordine T, Rodriguez P, Brochard L (2006) Origin of ICU acquired paresis determined by direct muscle stimulation. J Neurol Neurosurg Psychiatry 77(4):500–506 7. Kerbaul F, Brousse M, Collart F, Pellissier JF, Planche D, Fernandez C, Gouin F, Guidon C (2005) Combination of histopathological and electromyographic patterns can help to evaluate functional outcome of critically ill patients with neuromuscular weakness syndromes. Crit Care 8:R358–R366 8. Rich MM, Bird SJ, Raps EC, McCluskey LF, Teener JW (1997) Direct muscle stimulation in acute quadriplegic myopathy. Muscle Nerve 20:665–673 9. Sander HW, Golden M, Danon MJ (2002) Quadriplegic areflexic ICU illness: selective thick filament loss and normal nerve histology. Muscle Nerve 26:499–505 10. Rossignol B, Gueret G, Pennec JP, Morel J, Rannou F, GirouxMetges MA, Talarmin H, Arvieux CC (2008) Effects of chronic sepsis on contractile properties of fast-twitch muscle in an experimental model of critical illness neuromyopathy in the rat. Crit Care Med 36(6):1855–1863 11. Eikermann M, Koch G, Gerwig M, Ochterbeck C, Beiderlinden M, Koeppen S, Neuha¨user M, Peters J (2006) Muscle force and fatigue in patients with sepsis and multiorgan failure. Intensive Care Med 32:251–259 12. Horinouchi H, Kumamoto T, Kimura N, Ueyama H, Tsuda T (2005) Myosin loss in denervated rat soleus muscle after dexamethasone treatment. Pathobiology 72(3):108–116 13. Rich MM, Pinter MJ (2003) Crucial role of sodium channel fast inactivation in muscle fibre inexcitability in a rat model of critical illness myopathy. J Physiol 547.2:555–566 14. Murray MJ, Brull SJ, Bolton CF (2006) Brief review: nondepolarizing neuromuscular blocking drugs and critical illness myopathy. Can J Anesth 53(11):1148–1156 15. Zink W, Kaess M, Hofer S, Plachky J, Zausig YA, Sinner B, Weigand MA, Fink RH, Graf BM (2008) Alterations in intracellular Ca2?-homeostasis of skeletal muscle fibers during sepsis. Crit Care Med 36(5):1559–1563 16. Berchtold MW, Brinkmeier H, Mu¨ntener M (2000) Calcium ion in skeletal muscle: its crucial role for muscle function, plasticity and disease. Physiol Rev 80(3):1215–1265 17. Stehle R, Brenner B (2000) Cross-bridge attachment during highspeed active shortening of skinned fibers of the rabbit psoas muscle: implications for cross-bridge action during maximum velocity of filament sliding. Biophys J 78:1458–1473 18. Stamler JS, Meissner G (2001) Physiology of nitric oxide in skeletal muscle. Physiol Rev 81(1):209–237 19. Friedrich O, Hund E, Weber C, Hacke W, Fink RHA (2004) Critical illness myopathy serum fractions affect membrane excitability and intracellular calcium release in mammalian skeletal muscle. J Neurol 251:53–65 20. Bannister RA, Pessah IN, Beam KG (2008) The skeletal L-type Ca2? current is a major contributor to excitation-coupled Ca2? entry. J Gen Physiol 133(1):79–91 21. Kurebayashi N, Ogawa Y (2001) Depletion of Ca2? in the sarcoplasmic reticulum stimulates Ca2? entry in mouse skeletal muscle fibres. J Physiol 533:185–199 22. Lacomis D, Zochodne DW, Bird SJ (2000) Critical illness myopathy (editorial). Muscle Nerve 23:1785–1788

123

J Neurol 23. Friedrich O, Ehmer T, Fink RHA (1999) Calcium currents during contraction and shortening in enzymatically isolated murine skeletal muscle fibres. J Physiol 517(Pt 3):757–775 24. Friedrich O, Weber C, von Wegner F, Chamberlain JS, Fink RHA (2008) Unloaded speed of shortening in voltage-clamped intact skeletal muscle fibers from wt, mdx and transgenic minidystrophin mice using a novel high-speed acquisition system. Biophys J 94:4751–4765 25. Friedrich O, Kress KR, Hartmann M, Frey B, Sommer K, Ludwig H, Fink RHA (2006) Prolonged high-pressure treatments in mammalian skeletal muscle result in loss of functional sodium channels and altered calcium channel kinetics. Cell Biochem Biophys 45(1):71–83 26. Friedrich O, Both M, Gillis JM, Chamberlain JS, Fink RHA (2004) Mini-dystrophin restores L-type calcium currents in skeletal muscle of transgenic mdx mice. J Physiol 555.1:251–265 27. Rossignol B, Gueret G, Pennec JP, Morel J, Giroux-Metges MA, Talarmin H, Arvieux CC (2007) Effects of chronic sepsis on the voltage-gated sodium channel in isolated rat muscle fibers. Crit Care Med 35:351–357 28. Williams AB, Decourten-Myers GM, Fischer JE, Luo G, Sun X, Hasselgren PO (1999) Sepsis stimulates release of myofilaments in skeletal muscle by a calcium-dependent mechanism. FASEB J 13:1435–1443 29. Combaret L, Tilignac T, Claustre A, Voisin L, Taillandier D, Obled C, Tanaka K, Attaix D (2002) Torbafylline (HWA 448) inhibits enhanced skeletal muscle ubitquitin–proteasome-dependent proteolysis in cancer and septic rats. Biochem J 361:185– 192 30. Friedrich O, Hund E, Hacke W, Fink RHA (2004) Critical illness myopathy (CIM): evidence for a myotoxic low-molecular weight factor in the serum of patients after severe SIRS. In: Faist E (ed) Proceedings of the 6th world congress on trauma, shock, inflammation and sepsis pathophysiology, immune consequences and therapy. Medimond International Proceeding, pp 269–272 31. Li H, Malhotra S, Kumar A (2008) Nuclear factor-kappa B signalling in skeletal muscle atrophy. J Mol Med 86(10):1113–1126 32. Toniolo L, Maccatrozzo L, Patruno M, Pavan E, Caliaro F, Rossi R, Rinaldi C, Canepari M, Reggiani C, Mascarello F (2007) Fiber

123

33.

34.

35.

36.

37.

38.

39.

40. 41.

42.

types in canine muscles: myosin isoform expression and functional characterization. Am J Physiol Cell Physiol 292:C1915– C1926 Salanova M, Schiffl G, Pu¨ttmann B, Schoser BG, Blottner D (2008) Molecular biomarkers monitoring human skeletal muscle fibres and microvasculature following long-term bed rest with and without countermeasures. J Anat 212:306–318 Fraysse B, Desaphy JF, Pierno S, De Luca A, Liantonio A, Mitolo CI, Camerino DC (2003) Decrease in resting calcium and entry associated with slow-to-fast transition in unloaded rat soleus muscle. FASEB J 17(13):1916–1918 Diaz NL, Finol HJ, Torres SH, Zambrano CI, Adjounian H (1998) Histochemical and ultrastructural study of skeletal muscle in patients with sepsis and multiple organ failure syndrome (MOFS). Histol Histopathol 13(1):121–128 Haarmann CS, Dulhunty A, Laver DR (2005) Regulation of skeletal ryanodine receptors by dihydropyridine receptor II–III loop C-region peptides: relief of Mg2? inhibition. Biochem J 387:429–436 Zucchi R, Ronca-Testoni S (1997) The sarcoplasmic reticulum Ca2? channel/ryanodine receptor: modulation by endogenous effectors, drugs and disease states. Pharmacol Rev 49(1):1–51 Garcia MC, Faria JM, Escamilla J, Sanchez-Armass S, Sanchez JA (1999) A long-term blockade of L-type calcium currents upregulates the number of Ca2? channels in skeletal muscle. J Membr Biol 168:141–148 Brown LD, Rodney GG, Hernandez-Ochoa E, Ward CW, Schneider MF (2007) Ca2? sparks and T tubule reorganization in dedifferentiating adult mouse skeletal muscle fibers. Am J Physiol Cell Physiol 92:C1156–C1166 Pedersen BK, Steensberg A, Schjerling P (2001) Muscle-derived inteleukin-6: possible biological effects. J Physiol 536.2:329–337 Joulin O, Petillot P, Labalette M, Lancel S, Neviere R (2007) Cytokine profile of human septic shock serum inducing cardiomyocyte contractile dysfunction. Physiol Res 56:291–297 Li XQ, Zhao MG, Mei QB, Zhang YF, Cao W, Wang HF, Chen D, Cui Y (2003) Effects of tumor necrosis factor-alpha on calcium movement in rat ventricular myocytes. Acta Pharmacol Sin 24(12):1224–1230