Neural Agrin Changes the Electrical Properties of Developing Human Skeletal Muscle Cells
- First Online:
- Cite this article as:
- Jurdana, M., Fumagalli, G., Grubic, Z. et al. Cell Mol Neurobiol (2009) 29: 123. doi:10.1007/s10571-008-9304-z
- 149 Views
Recent investigations suggest that the effects of neural agrin might not be limited to neuromuscular junction formation and maintenance and that other aspects of muscle development might be promoted by agrin. Here we tested the hypothesis that agrin induces a change in the excitability properties in primary cultures of non-innervated human myotubes. Electrical membrane properties of human myotubes were recorded using the whole-cell patch-clamp technique. Cell incubation with recombinant chick neural agrin (1 nM) led to a more negative membrane resting potential. Addition of strophanthidin, a blocker of the Na+/K+ ATPase, depolarized agrin-treated myotubes stronger than control, indicating, in the presence of agrin, a higher contribution of the Na+/K+ ATPase in establishing the resting membrane potential. Indeed, larger amounts of both the α1 and the α2 isoforms of the Na+/K+ ATPase protein were expressed in agrin-treated cells. A slight but significant down-regulation of functional apamin-sensitive K+ channels was observed after agrin treatment. These results indicate that neural agrin might act as a trophic factor promoting the maturation of membrane electrical properties during differentiation, confirming the role of agrin as a general promoter of muscle development.
KeywordsAgrinMyogenesisElectrical membrane propertiesSkeletal muscleNa+/K+ pump
Sodium dodecylsulfate-polyacrylamide gel electrophoresis
Agrin is a heparan sulfate proteoglycan that is best known for its involvement in the organization and maintenance of postsynaptic structures at the neuromuscular junction (NMJ, Burden 2002; Kummer et al. 2006; McMahan 1990). In accordance with this role, mice deficient of agrin die perinatally due to breathing failure (Gautam 1996). It was also reported that agrin plays an important role in both peripheral and central nervous system (CNS) development, modulates posterior development in the zebrafish embryos (Kim et al. 2007), and participates in lymphocyte activation (Khan et al. 2001; Zhang et al. 2006). Taken together, these findings further widen the role of agrin in vertebrate living organisms and suggest that even in skeletal muscle its effects may not only be synaptogenetic. Recently, we have shown that in cultured human myotubes, neural agrin favors the maturation of the excitation-contraction (E-C) coupling mechanism mimicking the effect of co-culturing human muscle cells with rat spinal cord explants. We also demonstrated that such effect is due to the modulation of the activity of ryanodine receptors and dihydropyridine-sensitive L-type Ca2+ channels (Bandi et al. 2008). In skeletal muscle tissue, there are extrasynaptic proteins known to be regulated by the nerve activity. In rat skeletal muscle, the α1 subunit of Na+/K+ ATPase appears to be the predominant isoform during the early stages of development. At birth, the expression of the α2 subunit of Na+/K+ ATPase is increased and this form predominates in mature muscle cells (Sharabani-Yosef et al. 1999). It has recently been found that in the CNS, agrin has a direct effect on the Na+/K+ ATPase. It binds to the α3 subunit of the Na+/K+-ATPase and inhibits the activity of the pump in cortical neurons, resulting in membrane depolarization and increased action potential frequency (Hilgenberg et al. 2006).
It has been also shown that the innervation controls the expression of membrane proteins as K+ channels. A down-regulation of small-conductance K+ (SK) channel activity has been reported during development after innervation (Schmid-Antomarchi et al. 1985).
In the present article, we investigated the possible effects of recombinant full-length neural agrin in modulating the expression of Na+/K+ ATPase subunits and SK K+ channel activity during in vitro myogenesis of human muscle cells. Using the whole-cell patch clamp technique, we found changes in the resting membrane potential following treatment of the muscle cells with agrin; the changes were correlated to quantitative modifications in the amount of the α1 and α2 subunits of the Na+/K+ ATPase. We also observed a significant decline in the SK K+ channel activity after agrin application in the culture medium. We suggest a new role for neural agrin in controlling the changes of electrical membrane properties of skeletal muscle cells during differentiation.
Human Muscle Culture
Experiments were performed on cultures of human myocytes prepared as previously described (Askanas et al. 1987; Mars et al. 2001, 2003). All studies reported here were approved by the Ethical Commission at the Ministry of Health of the Republic of Slovenia (permit No: 63/01/99) and in accordance with the Declaration of Helsinki. Briefly, human muscle cells were derived from satellite cells obtained from muscle abductor hallucis routinely discarded at orthopedic operations on four patients aged from 5 to 17 years, without muscular disease as determined by clinical, electrophysiological, and morphological criteria. The muscle tissue was cleaned of adhering connective and fat tissue, cut to 0.5–1 mm pieces, and trypsinized to release muscle satellite cells. Cells were grown at clonal density in 100-mm Petri dishes in advanced Minimum Essential Medium (aMEM, Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS, Gibco) at saturated humidity in a mixture of 5% CO2-enriched air at 37°C. Myoblast colonies identified by morphological characteristics and devoid of fibroblast contamination were trypsinized and further expanded. Confluent myoblast cultures were harvested prior to myoblast fusion by trypsinization and were plated on round glass coverslips coated with a 1:2 mixture of 1.5% gelatin (Sigma, St Louis, MO, USA) and human plasma in 35-mm six-well dishes. To induce myoblast differentiation into myotubes, myoblasts were cultured in a differentiation medium (DM), aMEM medium supplemented with 2% FBS. Thereafter, myocytes were maintained in culture up to 15 days by renewing the medium every 3 days. After myoblasts differentiation into myotubes (which usually occurred in the first three days after myoblast replating), some cultures were incubated from day 3 with recombinant full-length chick neural agrin (cAgrin7A4B8, 1 nM, Denzer et al. 1995). Neural agrin was purified from the conditioned media of stably transfected HEK 293 cells (gift to G.F. from Dr. M.A. Ruegg, University of Basel, Basel, Switzerland) using mono Q-Sepharose Fast Flow Beads (Amersham Pharmacia Biotech, Piscataway, NJ, USA) as previously described (Bandi et al. 2008) by a modified method (Bezakova et al. 2001) and added each time with the media changes.
For SK3 immunolabeling, cells grown on cover slips were fixed with 3% paraformaldehyde in PBS buffer at room temperature for 15 min and permeabilized in PBS containing 0.4% Triton X-100 at room temperature for 5 min. Incubations with primary and secondary antibodies were carried out in PBS containing 0.1% bovine serum albumin, at room temperature. The rabbit polyclonal antibody specific for SK3 (Alomone Labs, Jerusalem, Israel) was used at 1:1000 dilution; controls included omission of the primary antibody and the use of non-immune rabbit serum. After rinsing, slides were incubated with Cy3-conjugated goat anti-rabbit antibodies (Amersham; 1:1000 dilution). For desmin immunostaining, cells were fixed in freshly prepared 3.7% paraformaldehyde in PBS for 15 min at 37°C, permeabilized with 100% methanol and then incubated for 60 min at 37°C with a mouse monoclonal anti-desmin antibody (D33, DAKO, Denmark) used at 1/50 dilution. For visualization of bound mab, cells were incubated for 60 min at room temperature with Rhodamine-conjugated goat anti-mouse IgG (Jackson ImmunoResearch Laboratories, PA, USA) used at 1/200 dilution. For nuclear staining, methanol permeabilized cells were incubated with bis-Benzimide H-33258 (1/1000, Sigma) for 5 min at room temperature. Slides were observed with standard fluorescence microscope or (for SK3) with Zeiss LSM510 confocal microscope.
Western Blot Analysis
Muscle cultures were briefly washed in pre-warmed PBS and incubated in cold lysis buffer (20 mM Tris–HCl, pH = 7.4, 1 mM EDTA, 10% sucrose, 0.1% Triton X-100, and 10 μg leupeptin, aprotinin, and pepstatin per ml; all from Sigma) for 10 min at 4°C. After incubation in lysis buffer, cell extracts were scraped from the culture dishes, transferred to microcentrifuge tubes, and homogenized with a Dounce homogenizer. Particulate insoluble components were removed by centrifugation of the homogenate in a microfuge (13,000g at 4°C for 10 min). For each sample, 10 μg of protein extract was separated in 10% Bis-Tris gel by using XCell SureLock MiniCell electrophoresis system (Invitrogen, Scotland, UK) and transferred to a PVDF membrane (Immobilon, Millipore, MA, USA). Membranes were incubated with mouse antibodies to α1 and with goat antibodies to α2 subunits of the Na+/K+ ATPase (catalog no. sc-21712 for α1 and sc-16049 for α2 subunit, both purchased from Santa Cruz Biotechnology, CA) diluted 1:1000 in blocking buffer (PBS with 0.1% Tween 20, Sigma and 5% I-block Tropix Applied Biosystems, Bedford, MA, USA). Polyclonal goat anti-actin antibody (sc-1616; Santa Cruz Biotechnology, CA) was used in control experiments. Blots were washed 3 × 10 min with blocking buffer. This step was followed by incubation with alkaline phosphatase-conjugated secondary antibody, diluted 1:5000 in blocking buffer. The blots were developed by in NBT-BCIP solution (Roche Applied Science, Germany) prepared in developing buffer (0.1 M Tris–HCl, 0.1 M NaCl and 0.05 M MgCl2). Membranes were quantitated with Chemi Genius BioImaging System (Syngen, Cambridge, UK).
Electrophysiological Recording and Data Analysis
All experiments were carried out under a phase contrast inverted microscope (Zeiss Axiovert 135, Oberkochen, Germany) at 400× magnification. Passive membrane properties of myotubes were recorded using the whole-cell patch-clamp technique in normal external solution (NES) containing (in mM): 100 NaCl, 2.8 KCl, 2 CaCl2, 2 MgCl2, 10 HEPES, 10 glucose, pH = 7.3 at room temperature (22–24°C). The patch pipette solution contained (in mM): 140 K-aspartate, 10 NaCl, 2 MgCl2, 10 HEPES, pH = 7.3. To record the resting membrane potential, access to the cytosolic compartment using the perforated patch-clamp method was preferred to conventional whole-cell recording, in order to provide exchange of small ions only and to avoid washout of intracellular second messengers. Thus, membrane potentials were recorded with 150 μg/ml amphotericin B-filled patch pipettes (Rae et al. 1991). Only cells with a stable resting potential were accepted for analysis.
Small-conductance K+ currents were recorded in NES solution with added tetrodotoxin (TTX, 1 μM) to block voltage-dependent Na+ channels and tetraethyl ammonium (TEA, 1 mM) to minimize big conductance K+ channels. For these experiments, patch pipettes were filled with the following solution (in mM): 130 K-gluconate, 10 KCl, 1 MgCl2, 10 HEPES, 1 Mg2ATP, 0.3 Na2GTP, 0.05 EGTA, pH = 7.3. All data were acquired at room temperature (22–24°C) with 3–5 MΩ patch pipettes using an Axopatch 200B (Axon Instruments, Foster City, CA) amplifier, digitized through a Digidata 1321A interface (Axon Instruments) and stored on a PC-compatible hard disk. Currents were acquired at a sampling time of 200 μs and low-pass filtered at 2 kHz. For data acquisition and analysis, the pCLAMP software suite (v. 8.0, Axon Instruments) was used. Membrane input resistance (Rm) was estimated from the approximately linear portion of the steady-state current–voltage relationships obtained by measuring the amplitude of the voltage response (10 mV) to hyperpolarizing current pulses. Membrane capacitance (Cm) was determined by integrating the area of the average of initial membrane current responses to a 5 mV hyperpolarizing command under voltage clamp.
All data are expressed as the mean ± SD with n being the number of cells tested, and statistical significance was assessed using the Student’s unpaired t-test at the P < 0.05 level. Origin 7 (Microcal Software, Northampton, MA) software was routinely used for graph plotting.
Human Skeletal Muscle Cells Develop in Culture
Agrin Treatment Affects the Resting Membrane Potential of Human Myotubes
We next explored the possibility that the more negative resting membrane potential promoted by agrin might be partially induced by the ionic gradient established by the Na+/K+ ATPase. We therefore recorded the resting membrane potential in myotubes after 6 days in DM in control and after superfusion of strophanthidin, a specific and reversible Na+/K+ ATPase inhibitor (Darbon et al. 2003). After 10–15 min superfusion of strophanthidin (4 μM), the resting membrane potential of human myotubes became significantly more depolarized (−50 ± 3.6 mV in control and −42 ± 6 mV in strophanthidin-treated myotubes; P < 0.05, n = 11) (16 ± 7% reduction); recovery from strophanthidin was observed, albeit extremely slowly (data not shown). Myotubes from the same set of cultures incubated in agrin had always a more negative resting potential (−63 ± 10 mV) than the untreated controls (−50 ± 3.6 mV, n = 11); in these cells, strophantidin superfusion had a stronger effect than in control and the resting membrane potential returned to more positive values (−50 ± 9 mV, n = 10; 24 ± 6% reduction). This effect indicates that the Na+/K+ pump has a larger relevance in creating the ionic gradients in agrin-treated cells (Fig. 2b).
Higher Expression of Na+/K+ ATPase Subunit Proteins in Agrin-Treated Cells
In Human Myotubes SK Channels are Modulated by Agrin
The SK membrane currents were revealed as tail currents under voltage clamp mode following depolarizing voltage steps to +10 mV (200 ms long) from a holding potential of −50 mV. Representative tail currents are shown in Fig. 4b. Depolarizing voltage steps to +10 mV followed by a final voltage step varying between −75 and −45 mV evoked tail currents with a reversal potential around −75 mV. A component of the tail currents was reduced after 10 min application of apamin (100 nM) both in control and agrin-treated myotubes. However, a difference was found in the percentage of reduction in agrin-treated myotubes; the mean apamin-sensitive component was 55 ± 25% in control (n = 39) and 32 ± 19% in myotubes incubated for 5 days in agrin (n = 16). In control cells, the tail current, measured at −50 mV, had a mean amplitude of 163 ± 120 pA (n = 30) and the mean current density (current amplitude/capacitance) of the apamin-sensitive component was 0.75 ± 0.4 pA/pF. In agrin-treated myotubes, the mean tail current amplitude was 186 ± 144 pA (n = 16) and the mean current density of apamin-sensitive component was significantly lower (0.5 ± 0.29 pA/pF, Fig. 4c). Altogether, the data indicate that in agrin-treated cells, SK currents were reduced by approximately 30%.
We show for the first time that agrin, besides its well-documented effects in the formation and maintenance of the NMJ, also controls some of the electrical properties characterizing skeletal muscle development. The present findings also complement our recent observation that neural agrin might not only aggregate AChRs but also be able to participate in skeletal muscle maturation by favoring the E-C coupling during differentiation (Bandi et al.2008), most likely by changing the density of dihyropyridine receptor/ryanodine receptor (DHPR/RyR) clusters (Flucher et al. 1994).
We demonstrated that the activity and the expression of the Na+/K+ ATPase, which preserves ionic gradients and participates in the development of the transmembrane resting potential of skeletal muscle cells (Bannet et al.1984; Brodie et al.1985, 1987), is up-regulated by agrin. After 14 days in agrin, human myotubes in culture developed a significant increase in expression levels of the α1 and α2 subunits of the Na+/K+ ATPase and exhibited hyperpolarized membrane potentials. This is in accord with the finding that, when co-cultured with rat embryonic spinal cord, cultured human muscle fibers increase their resting membrane potential, compared with aneurally cultured human muscle (Askanas et al.1987). A depolarized resting potential has long been considered to be a feature of immature excitable cells favoring the Ca2+ influx, necessary for muscle growth at early stage of development. A change of membrane potential in negative direction makes the cell excitable only if stimulated and might decrease the fraction of Na+ and Ca2+ channels undergoing inactivation (Filatov et al. 2005; Sciancalepore et al. 2005).
The observed effects of agrin on the muscle Na+/K+ ATPase differ from what was recently found in the CNS. Here, agrin binds to the α3 subunit of the Na+/K+ ATPase of cortical neurons and inhibits the activity of the pump, inducing cell depolarization and facilitating repetitive firing (Hilgenberg et al.2006). In our system, agrin had an opposite effect and potentiated the activity of the pump. The difference in the total effect (inhibition in CNS, potentiation in muscle) may originate from differences in subunit expression. The α3 subunit of the Na+/K+ ATPase was in fact found in both heart and presynaptic motor terminals; the α2 subunit was predominant in skeletal muscle (Zahler et al. 1996). In addition, it is possible that a recently discovered non-pumping Na+/K+ ATPase (Liang et al.2007) might function as a substrate for agrin allowing the activation of an unknown metabolic cascade promoting firing facilitation or inhibition depending on the cellular context. We also show that agrin down-regulates the expression of SK K+ channels, previously reported to decline with development under nerve control (Schmid-Antomarchi et al. 1985). We cannot exclude that, together with nerve growth factor (Vigdor-Alboim et al. 1999), agrin participates to the modulation of such conductance during synaptogenesis acting as a general promoter of muscle development.
Agrin is present at nerve-muscle contacts during development (Fallon et al.1985) but is also critically expressed by neurons in the brain (O’Connor et al.1994) during synaptogenesis (Cohen et al.1997; Li et al.1997) and seems to be important not only in early stages of synaptogenesis but also for functional maturation of NMJ (Mars et al. 2003). Motoneurons express a neural-specific isoform of agrin that contains inserts at the z splice site (z+) endowed of high efficacy to induce NMJ formation, whereas muscles express an isoform that lacks these inserts (z−) (Sanes and Lichtman 1999) and is free of known biological activities on muscle cells. We cannot exclude a role of the muscle agrin isoform as a possible modulator of the properties of developing muscle cells. However, the changes in the electrical membrane properties induced by agrin7A4B8 suggest the involvement of the neural isoform on the non-synaptogenic effect of agrin identified in this study.
Further investigations will be required to elucidate the mechanisms used by agrin to promote muscle cell differentiation and in particular if neural agrin interacts with functional receptor systems producing diffusible factors or if it promotes the aggregation of critical components of locally acting signalling machinery as in the case of NMJ formation. Although MuSK does not bind agrin directly, it is part of the functional agrin receptor complex (Sanes and Lichtman 1999). mRNA for MuSK was found to be present in aneurally cultured human myotubes (Gajsek et al. 2006) but we have no direct evidence that such muscle-specific tyrosine kinase is also involved in the agrin effects described here. As recently suggested (Kim and Burden 2008), agrin may act independently from MuSK activity at locations other than the NMJ.
In conclusion, our results show that agrin can induce in muscle effects that are extended to aspects of muscle differentiation and that are different from the well-known effects related to NMJ formation and maintenance.
The authors are very grateful to Prof. Fabio Ruzzier (University of Trieste) for helpful discussions and to Dr. Andy Constanti (School of Pharmacy, London, UK) for carefully reading the manuscript. This work was supported by the Italian Ministero dell’Istruzione, dell’Università e della Ricerca (MIUR), PRIN 2006 and by the Slovenian Research Agency. We also gratefully acknowledge help from Dr. Janez Brecelj and Dr. Elisa Luin. Marzia Di Chio and Zvonka Frelih are acknowledged for technical assistance.