Disuse of rat muscle in vivo reduces protein kinase C activity controlling the sarcolemma chloride conductance

Abstract

Muscle disuse produced by hindlimb unloading (HU) induces severe atrophy and slow-to-fast fibre type transition of the slow-twitch soleus muscle (Sol). After 2 weeks HU, the resting ClC-1 chloride conductance (g_Cl) of sarcolemma, which controls muscle excitability, increases in Sol toward a value typical of the fast-twitch EDL muscle. After 3 days of HU, the g_Cl increases as well before initiation of fibre type transition. Since ClC-1 channels are acutely silenced by PKC-dependent phosphorylation, we studied the modulation of g_Cl by PKC and serine–threonine phosphatase in Sol during HU, using a number of pharmacological tools. We show that a fraction of ClC-1 channels of control Sol are maintained in an inactive state by PKC basal activity, which contributes to the lower g_Cl in control Sol compared to EDL. After 14 days of HU, PKC/phosphatase manipulation produces effects on Sol g_Cl that corroborate the partial slow-to-fast transition. After 3 days of HU, the early increase of g_Cl in Sol is entirely attributable to a reduction of PKC activity and/or activation of phosphatase, maintaining ClC-1 channels in a fully active state. Accordingly, we found that HU reduces expression of PKCα, ɛ, and θ isoenzymes in Sol and EDL muscles and reduces total PKC activity. Moreover, we show that the rheobase current is increased in Sol muscle fibres as soon as after 3 days of HU, most probably in relation to the increased g_Cl. In conclusion, Sol muscle disuse is characterized by a rapid reduction of PKC activity, which reduces muscle excitability and is likely to contribute to disuse-induced muscle impairment.

During development skeletal muscle fibres differentiate into slow and fast-twitch fibres by acquiring specific metabolic and contractile properties owing to motor nerve-dependent modulation of the expression of a specific set of genes (Schiaffino & Serrano, 2002). On a molecular basis slow and fast twitch fibres are distinguished by the myosin heavy chain (MHC) isoform they express. Slow twitch fibres are characterized by type I MHC expression, whereas fast twitch fibres express MHC type IIa, IIx and IIb. Although the skeletal muscle is a highly specialized tissue, it can adapt to respond to functional demands during adult life. For instance, a reduced functional demand of the antigravity slow-twitch muscles, as it occurs during space flight or prolonged bed rest, produces profound modification of the muscular apparatus resulting in a reduction of strength and making postural maintenance and locomotion more difficult (Fitts et al. 2000). Understanding the molecular mechanisms responsible for muscle impairment during disuse is fundamental to develop effective countermeasures.

The rat hindlimb unloading (HU) is a widely accepted model for muscle disuse (Morey-Holton et al. 2005). In this model, the postural slow-twitch soleus (Sol) muscle experiences a progressive and severe atrophy measurable as soon as after 3 days of HU, while a partial slow-to-fast phenotype transition, characterized by an increased expression of fast MHC, initiates on the fourth day of HU (Stevens et al. 1999; Frigeri et al. 2001; Pierno et al. 2002). After 15 days of HU, the proportion of fast MHC-positive fibres may reach ∼40% of the whole Sol muscle compared to < 15% in control animals (Pierno et al. 2002). These effects are slowly reversed after several weeks of reloading (Desaphy et al. 2005).

Recent genomic and proteomic studies of the HU model have shown that a number of proteins involved in Sol muscle function are down- or up-regulated as a consequence of muscle atrophy or fibre type transition (Isfort et al. 2002; Wittwer et al. 2002; Stevenson et al. 2003; Seo et al. 2006). Our own studies have focused on water and ion channels, showing that many of them change expression/activity levels during disuse in accordance with the slow-to-fast transition of Sol muscle fibres (Desaphy et al. 2001; Frigeri et al. 2001; Pierno et al. 2002; Fraysse et al. 2003; Tricarico et al. 2005). For instance, expression of the voltage-gated chloride channel ClC-1, which is normally lower in slow-twitch muscles than in fast-twitch muscles, increased in the Sol after 1–3 weeks of HU in parallel with the increased expression of fast MHC (Pierno et al. 2002). The ClC-1 channel provides the macroscopic chloride conductance (g_Cl) that is by far the largest component conductance of adult muscle at resting potential and consequently a key determinant of muscle excitability (Steinmeyer et al. 1991; Pierno et al. 1999). The g_Cl is lower in slow with respect to fast muscles, which may contribute, together with other electrical parameters, to the higher excitability and better resistance to fatigue of the slow fibres (Pedersen et al. 2004, 2005). Thus, the increased g_Cl in HU Sol muscle may contribute to disuse-induced muscle function impairment. A striking observation, however, was that the g_Cl was already greatly increased after 3 days of HU, a time point at which no modification in MHC expression can be detected (Pierno et al. 2002). Thus the g_Cl increase appeared as an early event that may participate in the mechanisms leading to muscle fibre type transition. Indeed a pharmacological or genetic reduction of g_Cl can trigger the opposite transition of fast muscle properties toward a slower phenotype (Salviati et al. 1986; Goblet & Whalen, 1995; Wu & Olson, 2002). After 3 days of HU, the g_Cl increases quite homogeneously in all the Sol muscle fibres and its recovery was rapidly completed after 4 days reloading, suggesting an acute biochemical modulation of ClC-1 channels in disuse (Desaphy et al. 2005). In fast muscles, activation of calcium and phospholipid-dependent protein kinase C (PKC) by phorbol esters reduces the g_Cl, most probably through phosphorylation of the ClC-1 channels (Bryant & Conte Camerino, 1991; Tricarico et al. 1991; Rosenbohm et al. 1999). Nevertheless, inhibition of PKC by staurosporine or chelerythrine had no effect on g_Cl in fast muscle, suggesting that ClC-1 channels are basically fully active, thereby maintaining the high g_Cl typical of fast muscles (Tricarico et al. 1991; Pierno et al. 2003). In contrast, chelerythrine was able to increase the g_Cl in Sol muscle, suggesting that partial PKC-dependent inactivation of ClC-1 channels in basal conditions may contribute to the lower g_Cl in slow-twitch muscles (Desaphy et al. 2005).

The aim of this study was to evaluate the role of PKC in the effects of muscle disuse on the g_Cl. To this end, we evaluated the effects of a number of pharmacological tools able to modulate activities of PKC and serine–threonine phosphatases on the component resting conductances of control rat Sol and extensor digitorum longus (EDL) muscles and of disused Sol muscles of hindlimb-unloaded rats. In addition, we measured the protein expression of PKC isoenzymes as well as the total PKC activity in unloaded Sol and EDL muscles. The results demonstrate that PKC isoenzymes are early targets of HU-induced muscle disuse; the reduction of PKC activity modifies Sol muscle fibre excitability and is likely to contribute to disuse-induced muscle impairment.

Methods

Animal care and hindlimb unloading

The experiments complied with the Italian guidelines for the use of laboratory animals, which conforms with the European Community Directive of 1986 (86/609/ECC). Experiments in this study were specifically approved by the Italian Health Department (Art. 9 del Decreto Legislativo 116/92: Decreto no. 33/2000-B del Dipartimento degli alimenti e nutrizione e della sanità pubblica) and performed under the supervision of a local veterinary official.

Male Wistar rats weighing 250–350 g (Charles River Laboratories, Calco, Italy) were randomly assigned to control or HU groups. The total numbers of animals used in each experimental condition were as follows: control rats: 16; HU3 rats: 15; HU14 rats: 14. To induce muscle unloading, the animals were suspended individually in special cages for 3 days (HU3) or 2 weeks (HU14). A shoelace was linked at one extremity to the base of the tail by sticking plaster and at the other extremity to a trolley that can move on horizontal rails at the top of the cage (Desaphy et al. 2001). The length of the lace was adjusted to allow the animals to move freely on their forelimbs, while the body was inclined at 30–40 deg from the horizontal plane. Control and suspended animals had food and water ad libitum. Animals were inspected at least 3 times a day by laboratory personnel for fine-tuning of suspension and control of animal health (examination of behaviour and cleanliness of animals, aspect of hairs and eyes, daily food and water consumption, change of body weight, etc.). Deep post-mortem examination was also performed. No sign of skin inflammation was ever observed. In our 10 years of experience, the tail extremity becomes slightly dark in about 10% of the HU animals; in these cases, the rat was unfastened from lace and sticking plaster and the tail was examined for eventual lesion; no lesion was observed in this study and such animals were re-suspended using new lace and sticking plaster. No difference was observed between experimental data collected from these rats and those suspended for the entire period. After suspension, rats were unfastened from the lace and anaesthesia was induced by intraperitoneal injection of urethane (1.2 g (kg body weight)⁻¹). Then additional urethane was injected to produce deep anaesthesia to allow surgery. The Sol and EDL muscles were removed and used immediately for the electrophysiological experiments or stored at −80°C for PKC isoforms immunoblotting or PKC activity assays. After surgery, animals were killed by an overdose of urethane.

Ex vivo electrophysiological studies

Muscles tied at the end of each tendon were placed on a glass rod located in a 25 ml bath chamber maintained at 30°C and perfused with 95% O₂–5% CO₂ in normal and/or chloride-free physiological solution (Pierno et al. 2002). The normal physiological solution had the following composition (mm): NaCl 148, KCl 4.5, CaCl₂ 2.0, MgCl₂ 1.0, NaHCO₃ 12.0, NaH₂PO₄ 0.44, and glucose 5.5. A chloride-free solution was prepared by equimolar substitution of methylsulphate salts for NaCl and KCl and nitrate salts for CaCl₂ and MgCl₂. Tetrodotoxin (1 μm) was added to the chloride-free solution to suppress spontaneous fibre contraction. The pH of all the solutions was carefully maintained between 7.2 and 7.3 during each experiment. Using a computer-assisted two-intracellular-microelectrodes technique in current-clamp mode, the resting membrane potential, cable parameters, component conductances, and excitability parameters of Sol and EDL muscle fibres were measured in vitro.

Cable parameters, in both normal and chloride-free solutions, were calculated from the electrotonic potential elicited by a square-wave hyperpolarizing current pulse (100 ms duration) injected by the current electrode filled with 2 m potassium citrate. The membrane voltage responses were monitored by the voltage electrode, filled with 3 m KCl, at two distances from the current electrode. The current pulse generation, the acquisition of the voltage records and the calculation of the fibre constants were done in real time under computer control as described elsewhere (Bryant & Conte Camerino, 1991). The cable parameters, including fibre diameter and membrane resistance (R_m), were calculated from the experimentally determined values of input resistance and length constant and assuming a myoplasmic resistivity of 125 Ω cm. The reciprocal of R_m measured in normal physiological solution gives a value of the total membrane conductance (G_m) and the same parameter measured in chloride free solution is assumed to be the potassium conductance (g_K). The mean chloride conductance, g_Cl, is calculated as the mean G_m minus the mean g_K (Bryant & Conte Camerino, 1991).

To measure excitability parameters, the membrane potential was held to −80 mV by injecting a steady current, and 100 ms depolarizing current pulses of increasing amplitude were applied to elicit first a single action potential (AP) and then a train with the maximal number of action potentials. The excitability parameters, determined off-line on AP recordings, were the current threshold to elicit the first AP (or rheobase current, I_th), the latency of the AP (Lat), the AP amplitude (APA), the maximal number of elicitable AP (N spikes), and the ratio between I_th and the current threshold needed to elicit more than one AP (I_th/I₂).

The compounds tested were 4-β-phorbol-dybutyrate (4-β-PDB; Sigma Aldrich, St Louis, MO, USA), chelerythrine (Tocris Bioscience, Missouri, USA), okadaic acid (Tocris Bioscience, Ellisville, MO, USA) and insulin-like growth factor 1 (recombinant human IGF-1, Sigma). 4-β-PDB was dissolved in dimethylsulphoxide (DMSO) to produce concentrated stock solutions from which microlitre amounts were added to the muscle bath solution, as needed. The time of incubation of 4-β-PDB varied from 90 min (50 nm) to 30 min (100 nm) so as to reach a steady state of drug effect. Okadaic acid was dissolved in DMSO. The maximum DMSO concentration used (0.04%) was without effect on the parameters studied (not shown). Chelerythrine chloride was dissolved in distilled water. IGF-1 was reconstituted in a stock solution of 10 μg in 100 μl of 0.1 m acetic acid. The final concentration of IGF-1 to be tested in vitro was obtained with further dilution in normal or chloride free physiological solution.

The data are expressed as means ±s.e.m. The estimates of s.e.m. and n of normalized g_Cl values were obtained as previously described (Bryant & Conte Camerino, 1991). Comparison of mean values was performed by using Student's unpaired t test (considering P < 0.05 as significant).

PKC expression/activity determination

For the preparation of cell extracts, Sol and EDL muscles were homogenized in a buffer containing 20 mm Tris-HCl (pH 7.5), 2 mm EDTA, 10 mm EGTA, 2 mm phenylmethylsulphonyl fluoride, 100 μg ml⁻¹ leupeptin and 0.2% Triton X-100 using a glass–glass homogenizer (Moraczewski et al. 2002). Homogenates were sonicated, kept on ice for 1 h and centrifuged at 100 000 g for 1 h. Supernatant aliquots were used for Western blot analysis and PKC activity assays.

For gel electrophoresis and immunoblot analysis, 50 μg of cell extracts was boiled for 10 min in Laemmli sample buffer and proteins were separated on a 12% SDS-polyacrylamide gel, then transferred onto Immobilon-P membranes. Membranes were blocked in Tris-buffered saline–Tween (TBS-T) with 3% bovine serum albumin (BSA) at room temperature for 1 h, and then incubated overnight at 4°C with antibodies raised against different PKC isozymes (PKC sampler kit, BD Biosciences) or creatine kinase-M (Santa Cruz Biotechnology, Inc.). After three washes in TBS-T–BSA, membranes were incubated with anti-mouse or anti-goat IgG conjugated with horseradish peroxidase (Sigma). Protein bands were visualized using SuperSignal West Pico Chemiluminescent substrate (PIERCE). The following dilutions of anti-PKC antibodies were used: anti-PKCα (1 : 1000); anti-PKCɛ (1 : 1000); anti-PKCθ (1 : 250). Anti-creatine kinase-M (CK-M) was used at 1 : 200 dilution. In order to quantify PKC expression, protein bands were scanned and densitometric analysis was performed using Quantity One software (Bio-Rad). Results for PKC bands were normalized for the CK-M content and expressed as a percentage of the values obtained from control muscles. Note that the content of muscle CK-M was different in the control Sol and EDL muscles, but remained quite constant within each muscle after HU. Since HU had no effect on CK-M, the observed variations in normalized PKC isoenzymes reflected a true effect of HU on PKC expression.

PKC activity was measured with a MESACUP protein kinase assay kit (MBL, Nagoya, Japan), based on an enzyme-linked immunosorbent assay, according to the manufacturer protocol. Briefly, cell extracts were incubated in the presence of phosphatidylserine and 1 mm ATP in 96-well microplates coated with a peptide substrate for PKC. All experimental conditions were carried out in the presence of 30 μm H89 to exclude any contribution by the PKA. After 20 min of incubation at 25°C, wells were washed and a biotinylated monoclonal antibody, which binds to the phosphorylated form of the substrate peptide, was added and incubated at 25°C for 60 min. After three washes, peroxidase-conjugated streptavidin was added and incubated for 60 min; after three washes, a peroxidase substrate was added to each well and the reaction was carried out for 10 min at room temperature in the dark. PKC activity associated with each sample was determined by measuring the absorbance at 492 nm using a microplate reader (Bio-Rad, model 550).

Results

Effects of hindlimb unloading-induced disuse on Sol and EDL muscle properties

In the present study, Sol muscle underwent significant atrophy as soon as after 3 days of hindlimb unloading, as evidenced by the reduction of the Sol muscle to body weight ratio and of Sol muscle fibre diameter (Table 1). Atrophy was more pronounced after 14 days of HU. In contrast, no change in the EDL muscle to body weight ratio was observed, confirming lack of atrophy for the fast-twitch EDL muscle. As expected (Pierno et al. 2002), the resting chloride conductance of sarcolemma (g_Cl) increased by 30–50% in the Sol muscle after 3–14 days of HU, whereas no significant change occurred in the resting potassium conductance (g_K). Note that, in a previous study, no significant change in g_Cl and g_K was found in EDL muscle fibres after 21 days of HU (Pierno et al. 2002).

Table 1.

Effects of hindlimb unloading (HU) on Sol and EDL muscle properties

	CTRL	HU3	HU14	ANOVA
EDL muscle to body weight ratio (mg g−1)	0.51 ± 0.07	0.51 ± 0.08	0.52 ± 0.02	F < 0.001
	(3)	(3)	(3)	P = 0.9915
Sol muscle to body weight ratio (mg g−1)	0.56 ± 0.03	0.47 ± 0.01*	0.35 ± 0.02*†	F = 25.1
	(9)	(10)	(10)	P < 0.0001
Sol muscle fibre diameter (μm2)	56.9 ± 1.0	52.4 ± 0.9*	43.6 ± 0.8*†	F = 54.5
	(82/9)	(127/10)	(124/10)	P < 0.0001
Sol muscle fibre gCl (μS cm−2)	1430 ± 32	1853 ± 36*	2162 ± 55*†	F = 67.3
	(74/7)	(107/8)	(83/6)	P < 0.0001
Sol muscle fibre gK (μS cm−2)	328 ± 13	351 ± 10	354 ± 8	F = 1.69
	(71/7)	(72/8)	(51/6)	P = 0. 1881

Each value is the mean ±s.e.m. from N rats or from n fibres/N rats (in brackets). Analysis of variance (ANOVA) followed by Bonferroni's test was used to compare control rats (CTRL) to 3-days' hindlimb unloaded rats (HU3) and 14-days' hindlimb unloaded rats (HU14).

^*P < 0.005 versus CTRL

^†P < 0.005 versus HU3.

To verify the physiological relevance of such changes in g_Cl, we measured the excitability parameters in Sol muscles using two intracellular microlectrodes (Table 2). In control animals, with respect to EDL muscle fibres, Sol muscle fibres displayed a resting membrane potential (RMP) slightly but significantly more negative; the rheobase current (I_th) was far smaller; the latency of the AP, which is inversely related to the g_Cl, was twofold longer; the AP amplitude (APA) was significantly lower; fewer action potentials could be elicited (N spikes); and the I_th/I₂ ratio was smaller. These effects indicate that the Sol muscle needs less excitation than the EDL muscle to elicit an action potential, but can generate only a maximum of two APs under intense stimulation. This agrees with the antigravitational role of the plantarflexor Sol muscle, which needs to be tonically active.

Table 2.

Effects of disuse on excitability parameters in control EDL and Sol muscles and Sol muscles after 3 or 14 days' hindlimb unloading (HU3 and HU14)

Muscle	RMP (−mV)	Ith (nA)	Lat (ms)	APA (mV)	N spikes	Ith/I2(nA)
EDL	67.2 ± 1.8*	206 ± 22*	6.1 ± 0.4*	111.2 ± 1.3*	5.3 ± 0.4*	0.76 ± 0.03*
(N = 3)	(n = 12)	(n = 12)	(n = 12)	(n = 12)	(n = 12)	(n = 10)
Sol	73.3 ± 0.9	75.8 ± 4.5	13.0 ± 0.6	101.0 ± 3.2	1.2 ± 0.1	0.54 ± 0.02
(N = 5)	(n = 23)	(n = 23)	(n = 23)	(n = 23)	(n = 23)	(n = 5)
HU3	75.4 ± 1.4	96.4 ± 4.9*	6.6 ± 0.4*	90.2 ± 3.2*	1.5 ± 0.1	0.40 ± 0.03*
(N = 3)	(n = 13)	(n = 13)	(n = 13)	(n = 13)	(n = 13)	(n = 7)
HU14	69.6 ± 2.4	108 ± 9.8*	7.3 ± 0.3*	85.7 ± 2.4*	1.2 ± 0.2	0.48
(N = 2)	(n = 6)	(n = 6)	(n = 6)	(n = 6)	(n = 6)	(n = 1)

Excitability parameters were measured with intracellular microelectrodes in n fibres from N rats (in brackets). The parameters were the resting membrane potential (RMP), rheobase current (I_th), latency of first action potential (Lat), amplitude of first action potential (APA), maximum number of action potentials (N spikes), and ratio between I_th and current threshold for eliciting more than two action potentials (I₂). Statistical analysis was performed with Student's unpaired t test versus control Sol muscle (

^*P < 0.05).

After 3 days of HU in the Sol muscle, the rheobase current was significantly greater with respect to control Sol, although it did not reach a value similar to EDL muscle fibres; the latency was shorter, as expected from an increased g_Cl; the APA was smaller; and the I_th/I₂ ratio was significantly smaller. After 14 days of HU, Sol muscle fibres were slightly depolarized with respect to control Sol muscle fibres; all the other parameters were modified in a similar way as after 3 days of HU. Thus as soon as after 3 days of HU, the disused Sol muscle was less excitable.

Effects of protein kinase C modulators on the resting g_Cl and g_K of Sol muscle after hindlimb unloading

We have previously shown that the g_Cl in EDL muscle fibres is inhibited after activation of protein kinase C by 4-β-phorbol 12,13-dibutyrate (4-β-PDB), whereas the PKC-inactive α-isomer leaves g_Cl unaltered (De Luca et al. 1998). In contrast, the other component conductance, g_K, is essentially unaffected by 4-β-PDB (De Luca et al. 1994). In the present study, we measured the effect of 50 and 100 nm 4-β-PDB on the total resting conductance, assuming that the observed effects are essentially due to modulation of g_Cl (Fig. 1A). The large inhibitory effect of 4-β-PDB on g_Cl was confirmed in the EDL muscle fibres of control rats. The percentages of block are reported in Table 3. In contrast, the g_Cl of Sol muscle fibres was far less sensitive to PKC activation. After hindlimb unloading, the sensitivity to 4-β-PDB of Sol muscle fibre g_Cl increased to an intermediate level between control Sol and EDL muscle fibres.

Figure 1.

Effects of disuse on the protein kinase C-dependent modulation of sarcolemma chloride conductance in fast-twitch and slow-twitch muscles A, effects of 50 or 100 nm 4-β-phorbol-dibutyrate (PDB), an activator of PKC, on the total resting conductance of sarcolemma measured in control EDL muscle, control Sol muscle, and Sol muscles after 3 or 14 days' hindlimb unloading (HU3 or HU14). B, effects of 1 μm chelerythrine, an inhibitor of PKC, on the resting chloride conductance of sarcolemma measured in the same experimental conditions as in A. Each bar represent the mean value ±s.e.m. from at least 11 fibres/2 muscles. Symbols indicate significant difference (P < 0.05, Student's unpaired t test) between treated fibres and control (*), between fibres treated with 50 and 100 nm PDB (#), between Sol and EDL muscle fibres (†), and between HU and control Sol muscle fibres (‡).

Table 3.

Effects of PKC and phosphatase modulators on g_Cl in control EDL and Sol muscles and Sol muscles after 3 or 14 days' hindlimb unloading (HU3 and HU14)

	PDB 50 nm	PDB 100 nm	Chelerythrine	Okadaic acid	IGF-1
EDL	−62.4 ± 4.2%	−89.0 ± 1.4%	+6.6 ± 1.6%	−27.6 ± 4.1%	−2.2 ± 3.2%
	P < 0.0001	P < 0.0001	P < 0.01	P < 0.0001	NS
Sol	−11.8 ± 4.1%	−32.7 ± 2.5%	+40.9 ± 7.6%	−5.4 ± 2.7%	+31.0 ± 5.1%
	P < 0.05	P < 0.0001	P < 0.0001	NS	P < 0.0001
Sol HU3	−31.9 ± 4.4%	−50.6 ± 2.6%	−4.8 ± 3.0%	−38.6 ± 5.5%	−6.5 ± 3.8%
	P < 0.0001	P < 0.0001	NS	P < 0.0001	NS
Sol HU14	−33.0 ± 3.0%	−59.9 ± 3.1%	+22.4 ± 4.5%	−13.0 ± 4.1%	+0.9 ± 3.3%
	P < 0.0001	P < 0.0001	P < 0.0001	P < 0.05	NS
Theoretical value Y	−32.0%	−55.2%	+27.2%	−14.3%	+17.7%

The mean g_Cl±s.e.m. (G_m for 4-β-PDB) was calculated from at least 11 fibres/2 muscles, first in control condition then in the presence of drug (50 or 100 nm 4-β-PDB, 1 μm chelerythrine, 0.25 μm okadaic acid, or 3.3 nm IGF-1). The percentage ±s.e.m. of g_Cl change induced by the drugs is reported in each column. Statistical significance of the effects was assessed by Student's unpaired t test (NS: not significant). The theoretical value Y is the percentage of effect expected after 14 days' hindlimb unloading, if 40% of Sol muscle fibres have acquired fast-twitch phenotype, as determined by expression of fast myosin heavy chain (Pierno et al. 2002). Y is calculated following the equation Y = ((1 − 0.40) ×X_Sol) + (0.40 ×X_EDL), X_Sol is the percentage of effect in the control Sol muscle, and X_EDL is the percentage of effect in the control EDL muscle.

The PKC inhibitor chelerythrine (1 μm) did not significantly affect the resting g_K in the four experimental groups (Table 4). In EDL muscle fibres, application of chelerythrine produced a small but significant increase of g_Cl (Fig. 1B, Table 3). In contrast, chelerythrine increased g_Cl by ∼40% in Sol muscle fibres of control animals, suggesting a higher basal activity of PKC in slow-twitch muscle than in fast-twitch muscle. After 3 days of HU, chelerythrine had however, no more significant effect on Sol muscle g_Cl. After 14 days of HU, the chelerythrine effect was halfway between effects observed in control Sol and EDL muscles.

Table 4.

Effects of PKC and phosphatase modulators on g_K in control EDL and Sol muscles and Sol muscles after 3 or 14 days' hindlimb unloading (HU3 and HU14)

Muscle	Condition	Chelerythrine	Okadaic acid	IGF-1
EDL	CTRL	307 ± 36 (18)	324 ± 14 (16)	336 ± 19 (17)
	DRUG	288 ± 13 (17)	333 ± 19 (18)	327 ± 21 (19)
Sol	CTRL	336 ± 21 (11)	309 ± 26 (22)	373 ± 26 (22)
	DRUG	340 ± 21 (14)	345 ± 28 (31)	364 ± 14 (20)
Sol HU3	CTRL	380 ± 22 (21)	356 ± 14 (13)	365 ± 15 (23)
	DRUG	367 ± 14 (23)	345 ± 17 (12)	361 ± 18 (24)
Sol HU14	CTRL	371 ± 24 (16)	350 ± 10 (14)	354 ± 21 (13)
	DRUG	384 ± 14 (19)	333 ± 23 (18)	394 ± 16 (13)

Each column reports the g_K value in μS cm⁻² measured in n fibres (in brackets) from 2 rats, before and after acute application of 1 μm chelerythrine, 0.25 μm okadaic acid, or 3.3 nm IGF-1. No significant difference was found between DRUG and relative CTRL by Student's unpaired t test.

Effects of serine–threonine phosphatase modulators on the resting g_Cl and g_K of Sol muscle after hindlimb unloading

We previously showed that the inhibitory effects of PKC activation on the g_Cl can be prevented or reversed by activation of an okadaic acid-sensitive serine–threonine phosphatase (De Luca et al. 1998). Thus we measured the effects of in vitro application of 0.25 μm okadaic acid and 3.3 nm IGF-1 on g_Cl as inhibitor and activator of muscle serine–threonine protein phosphatase, respectively. No significant effect of the two drugs was found on the resting g_K in all four experimental conditions (Table 4).

As previously described, okadaic acid reduced g_Cl by ∼25% in EDL muscle fibres (Fig. 2A, Table 3). Together with the little effect of chelerythrine on g_Cl in the same condition, this suggests that a basal phosphatase activity is able to compensate for the basal activity of PKC on the conductance in EDL muscle fibres of adult rats, thereby maintaining chloride channels in an activated state. In the same animals, no significant effect of okadaic acid was observed on g_Cl of Sol muscle fibres, thus suggesting that either phosphatases are not active at rest in slow-twitch muscle or basal phosphatase activity cannot compete against the high basal activity of PKC evidenced by the rough activating effect of chelerythrine on g_Cl. In contrast, after 3 days of HU, okadaic acid greatly reduced the g_Cl by ∼40% in Sol muscle fibres. This indicates that phosphatase activity is vigorous in such condition, being able to counteract the low PKC activity revealed by lack of chelerythrine effect. After 14 days of HU, okadaic acid produced a ∼13% reduction of g_Cl in Sol muscle fibres, which is significantly more than in control but significant less than after 3 days of HU.

Figure 2.

Effects of disuse on the modulation of sarcolemma chloride conductance by okadaic-sensitive protein phosphatase in fast-twitch and slow-twitch muscles A, effects of 0.25 μm okadaic acid, an inhibitor of phosphatase, on the resting chloride conductance of sarcolemma measured in control EDL, control Sol muscle, and Sol muscles after 3 or 14 days' hindlimb unloading (HU3 or HU14). B, effects of 3.3 nm IGF-1, an activator of protein phosphatase, on the resting chloride conductance of sarcolemma in the same experimental conditions as in A. Each bar represents the mean value ±s.e.m. from at least 22 fibres/2 muscles. Symbols indicate significant difference (P < 0.05, Student's unpaired t test) between treated fibres and control (*), between Sol and EDL muscle fibres (†), and between HU and control Sol muscle fibres (‡).

We previously demonstrated that application of IGF-1 in vitro to rat EDL muscle fibres does not induce any change in g_Cl, but is able, through activation of an okadaic acid-sensitive phosphatase, to antagonize the inhibitory effect of PKC on g_Cl that is produced by prior application of phorbol esters (De Luca et al. 1998). Accordingly, we found here that IGF-1 has no significant effect on resting g_Cl in EDL muscle fibres (Fig. 2B, Table 3). In contrast, IGF-1 appeared to increase the g_Cl dramatically by ∼30% in Sol muscle fibres of control rats. Again, this effect was mediated by phosphatase activation, since no more effect was observed after prior incubation with okadaic acid; indeed in two rat Sol muscles, the g_Cl was 1516 ± 63 μS cm⁻² (n = 11 fibres) in control, 1503 ± 47 μS cm⁻² (n = 9) in the presence of 0.25 μm okadaic acid, and 1580 ± 58 μS cm⁻² (n = 9) in the presence of 0.25 μm okadaic acid plus 3.3 nm IGF-1 (no significant difference was found using ANOVA, F = 0.4829, N = 2, N-k = 26, P = 0.622). After 3 or 14 days of HU, no significant effect of IGF-1 was found on the g_Cl of Sol muscle fibres (Fig. 2B).

Effect of hindlimb unloading on PKC isoenzymes expression and total PKC activity in Sol and EDL muscles

Expression of PKC isoforms in Sol and EDL muscles of control, HU3 and HU14 rats (two animals for each condition, three independent experiments for each muscle) was determined by immunoblotting analysis as described in Methods. In Sol and EDL muscles of the various experimental groups, protein expression was found for the conventional PKCα and the novel PKCɛ and PKCθ. No significant immunoreactivity was found against the other anti-PKC antibodies included in the kit, which should recognize β, γ, δ and ι/λ isoenzymes. Abundance was normalized for creatine kinase-M content and expressed as percentage of control muscles (Fig. 3). In the Sol muscle, PKCα expression was reduced to 75.9 ± 1.7% after 3 days of HU and to 63.1 ± 7.9% after 14 days of HU, with respect to control rats. In EDL muscle, a smaller but still significant reduction to 80.5 ± 0.32% and to 75.8 ± 2.6% was observed after 3 and 14 days of HU, respectively. Effects of hindlimb unloading were more marked for PKCɛ, with reduction of expression to 43.9 ± 6.9% (HU3) and to 48.2 ± 7.0% (HU14) in the Sol muscle, and with reduction to 60.8 ± 5.1% (HU3) and 55.4 ± 6.6% (HU14) in the EDL muscle. Also expression of PKCθ was reduced in the Sol muscle to 67.4 ± 4.2% after HU3 and to 47.5 ± 2.9% after HU14, and in the EDL muscle to 62.7 ± 4.7% after HU3 and to 64.0 ± 7.5% after HU14. Significant difference between protein expression levels at HU3 and HU14 was observed only for PKCθ in the Sol muscle (P < 0.05, Student's t test for paired data).

Figure 3.

Effects of disuse on protein expression levels of PKCα, PKCɛ, and PKCθ isoenzymes in fast-twitch and slow-twitch muscles Cell extracts were prepared from EDL and Sol muscle of control rats and after 3 (HU3) or 14 (HU14) days' hindlimb unloading (2 rats for each condition). Western blots were immunolabelled using antibodies against PKCα (A), PKCɛ (B), and PKCθ (C) isoenzymes, as well as muscle creatine kinase. For each muscle, the measure was repeated in three independent experiments. PKC protein levels were quantified by densitometry, normalized with respect to creatine kinase signal, and expressed as percentage ±s.e.m. of control muscle. Symbols indicate significant difference (P < 0.05, Student's paired t test) between HU and control (*), and between HU3 and HU14 (#),

The total PKC activity was measured in cell extracts from Sol and EDL muscles of control, HU3 and HU14 rats (two animals for each condition, three independent experiments for each muscle) using a commercial assay kit. Data obtained after HU were expressed as the percentage of the activity found in control muscles. PKC activity was significantly reduced to 39.7 ± 6.5% at HU3 and to 20.2 ± 9.9% after 14 days of HU in the Sol muscle. In the EDL muscles, PKC activity was reduced to 55.5%± 9.2 at HU3 and 59.0 ± 2.4% after 14 days (Fig. 4). Accordingly, a reduced phosphorylation of PKCα-Ser657 has been found in Sol muscles of young rats after 1–3 weeks' HU, suggesting a reduced activity of this isoenzyme (Choi et al. 2005). When the total PKC activities of Sol versus EDL were compared in control rats, no statistical difference was observed. This, however, does not exclude that one or more PKC specific isoforms might have a differential activity in the two muscle types.

Figure 4.

Effects of disuse on total PKC activity in fast-twitch and slow-twitch muscles PKC total activity was evaluated in cell extracts from EDL muscle and Sol muscle of control rats and after 3 (HU3) or 14 (HU14) days' hindlimb unloading (2 rats for each condition). For each muscle, the measure was repeated in three independent experiments. Data are expressed as percentage ±s.e.m. of the activity detected in control muscles. Symbols indicate significant difference (P < 0.05, Student's paired t test) between HU and control (*), and between HU3 and HU14 (#).

Discussion

The protein kinase C (PKC) family plays important roles in many intracellular signalling events, such as cell growth and differentiation. The family is composed of a number of individual isoforms that belong to three distinct categories: conventional isoforms (α, β, γ) activated by Ca²⁺, phorbol esters and diacylglycerol; novel isoforms (δ, ɛ, η, θ) which are also activated by phorbol esters and diacylglycerol but not by Ca²⁺; and atypical isoforms (ζ, λ, ι) which are not activated by Ca²⁺, phorbol esters or diacylglycerol (Sampson & Cooper, 2006). It is well established that in mammalian fast-twitch skeletal muscle, PKC can alter the resting chloride conductance through modulation of the main voltage-gated chloride channel (Bryant & Conte Camerino, 1991; Tricarico et al. 1991; Chen & Jockusch, 1999; Rosenbohm et al. 1999). Indeed, PKC activation by phorbol esters reduces chloride currents supported by hClC-1 channels expressed in a mammalian cell line and decreases the resting g_Cl in fast-twitch muscle fibres of mouse, rat and goat. Such a modulation is involved in the changes of muscle electrical properties observed during ageing or after various pharmacological treatments (De Luca et al. 1994; Pierno et al. 2003; Liantonio et al. 2007). In the present work, we confirm that, in normal resting conditions, PKC activity on g_Cl is quite low or at least compensated by phosphatase activity in the fast-twitch muscle, because the PKC inhibitors chelerythrine and IGF-1, which activates an okadaic acid-sensitive phosphatase, have little effect on the g_Cl, whereas okadaic acid reduces the g_Cl by 25%. We now demonstrate that the situation is exactly the opposite in slow-twitch muscle fibres, which show a lower sensitivity to phorbol esters and okadaic acid, whereas chelerythrine and IGF-1 greatly affect g_Cl. Thus the lower g_Cl observed in slow-twitch muscle fibres with respect to fast-twitch muscle fibres is due in part to an inverted PKC/phosphatase activity ratio, which maintains a fraction of ClC-1 channels in an inactive state. A schematic diagram drawn to illustrate the possible modulation of g_Cl by PKC-dependent phosphorylation in various muscle types and use conditions is shown in Fig. 5. We propose that the higher calcium ion content in resting Sol muscle may contribute at least in part to the higher basal activity of conventional PKC (Fraysse et al. 2003).

Figure 5.

Schematic diagram of Sol muscle chloride conductance modulation by hindlimb unloading- induced disuse In control conditions, slow-twitch fibres (red dashed) in the Sol muscle express a lower amount of ClC-1 channels than the fast-twitch fibres (grey dashed) in the EDL muscle. In addition, a fraction of Sol muscle ClC-1 channels are in an inactive state due to basal PKC activity, which contributes to the lower g_Cl in the slow-twitch muscle. There are a number of pieces of evidence suggesting that the phosphorylation site is the ClC-1 channel protein itself, as shown in the diagram, but we cannot exclude an indirect effect of PKC on ClC-1 channel function (see Discussion). After 3 days of HU, PKC activity is reduced in the whole Sol muscle, thus leading to inactivation of ClC-1 channels and elevated g_Cl. A similar effect can be obtained acutely by application of chelerythrine (PKC inhibition) or IGF-1 (phosphatase activation) in control Sol muscle fibres. After 14 days of HU, part of the Sol muscle fibres acquire a fast phenotype characterized by expression of myosin heavy chain type II isoform. These fast fibres show an increased expression of ClC-1 channels, which are fully active owing to a low basal activity of PKC. The remaining slow-twitch fibres of HU14 Sol muscle conserve the characteristics of control Sol muscle fibres, with a low expression of ClC-1 channels and a partial PKC-dependent ClC-1 channel inactivation.

We previously showed that the g_Cl increases in the Sol muscle during disuse induced by 3 days or 14 days of hindlimb unloading (Pierno et al. 2002; Desaphy et al. 2005). The increased g_Cl observed after 14 days of HU was associated with an increase of ClC-1 mRNA content, being consistent with the partial slow-to-fast transition of Sol muscle. After 14 days of HU, ∼40% of the Sol muscle fibres expressed fast myosin heavy chain (MHC) isoforms, which correlated well with the fraction of Sol muscle fibres exhibiting an increased g_Cl similar to that measured in fast-twitch EDL muscle fibres (Pierno et al. 2002). The remaining ∼60% of HU14 Sol muscle fibres expressed exclusively the slow MHC type I, while about half of the fibres conserved a g_Cl typical of control Sol muscle. Interestingly, the mean effects of phorbol ester, chelerythrine and okadaic acid measured in the Sol muscle of HU14 rats overlapped those expected from a theoretical calculation taking into consideration the distribution of fast-MHC and slow-MHC fibres after HU14 and hypothesizing that the drug effects in each fibre type were similar to those measured in control Sol or EDL muscles (see Table 3). This suggests that the mean effects may simply reflect the partial slow-to-fast transition of the Sol muscle after 14 days of HU (Fig. 5). Nevertheless, we cannot fully exclude that the observed effects are specific to the disuse condition, independently of fibre phenotype. For instance, the lack of IGF-1 effect on Sol muscle g_Cl after 14 days of HU allows us to discard the first hypothesis, since the effect expected from Sol muscle fibre type composition should be an averaged g_Cl increment of about 15% (Table 3). Instead, the g_Cl of disused Sol muscle appeared no more responsive to IGF-1 stimulation. It is noteworthy that IGF-1 over-expression in transgenic mice failed to prevent HU-induced atrophy of gastrocnemius and tibialis anterior muscles, suggesting that loss of muscle sensitivity to IGF-1 may be a general aspect of muscle disuse (Criswell et al. 1998).

In contrast to HU14, no change in MHC expression was observed in the Sol muscle after 3 days of HU, whereas the g_Cl increased quite uniformly in the entire Sol muscle fibres (Pierno et al. 2002). Because the g_Cl rise is rapidly reversed upon return of rats to normal loading, we proposed that the increase of g_Cl after HU3 may be due to an acute biochemical regulation of ClC-1 channels (Desaphy et al. 2005). Here, we found that the g_Cl of HU3 Sol muscle lost its sensitivity to chelerythrine but acquires sensitivity to okadaic acid, which is exactly contrary to the g_Cl in control Sol muscle. In control Sol muscle, chelerythrine increases the g_Cl by ∼40% just as HU3 does, while in HU3 Sol muscle, okadaic acid decreases the g_Cl by ∼40% toward a level similar to that of control Sol muscle. This result strongly supports our hypothesis, suggesting that muscle disuse for 3 days favours phosphatase activity over PKC activity, which results in ClC-1 channel activation and increased g_Cl (Fig. 5).

It can be noted here that we have no direct demonstration that PKC phosphorylates ClC-1 channel protein. Nevertheless, experiments previously performed on heterologously expressed hClC-1 channels in HEK cells strongly suggest that the PKC phosphorylation site is the channel protein itself or a closely associated subunit, and that PKC-dependent phosphorylation affects functionally active channels (Rosenbohm et al. 1999). Other studies have also proposed that PKC inhibition may promote the trafficking of channels from an intracellular pool to the membrane surface (Papponen et al. 2005). Whatever the mechanism by which PKC decreases ClC-1 channel activity, our data clearly show that PKC inhibition by disuse results in an increased g_Cl that, in turn, affects muscle excitability.

Importantly, the reduction of total PKC activity in Sol muscle after 3 and 14 days of HU was confirmed by direct in vitro assay measurements. A reduction of PKC activity was found also in the HU EDL muscle, which, however, has little consequence for the EDL muscle g_Cl, probably because most of the chloride channels of this muscle are already in a dephosphorylated, active state. In HU3 Sol muscle, a ∼60% reduction of PKC activity appeared sufficient to obtain a full loss of chelerythrine effect on g_Cl. On the other hand, although the PKC activity reduction was ∼70% after 14 days of HU, chelerythrine was still able to increase the mean Sol muscle g_Cl. Such a discrepancy may be well explained on the basis of the above hypothesis considering that PKC activity reduction may affect homogeneously the entire HU3 Sol muscle fibres, whereas the HU14 Sol muscle is composed of fast-MHC positive fibres fully lacking PKC sensitivity and of slow-MHC fibres that conserved basal PKC activity typical of control Sol muscle. There is also the possibility that a different regulation of phosphatase activity between HU3 and HU14 may allow a complete or partial compensation of the low PKC activity measured in these two conditions.

The possible molecular mechanisms contributing to the reduced PKC activity in disused muscle include a reduced expression of PKC isoenzymes, as we observed here for PKCα, PKCθ, and PKCɛ, and/or a decreased resting calcium concentration, as previously shown (Fraysse et al. 2003). Little is known about PKC expression and modulation in skeletal muscle. There is, however, growing evidence that in vivo muscle activity provoked by various exercise regimens or direct nerve stimulation, as well as in vitro muscle contraction, increases PKC activity in mammalian skeletal muscles (Richter et al. 1987; Cleland et al. 1989; Nielsen et al. 2003; Perrini et al. 2004; Rose et al. 2004). Alterations of activity/distribution/expression of PKC isoenzymes have also been observed in skeletal muscles of different models of insulin resistance (Sampson & Cooper, 2006). In addition, denervation may affect muscle PKC expression/activity, although these effects are not clearly defined yet (Hilgenberg et al. 1996; Sneddon et al. 2000; Lin et al. 2002). In the HU model, the reduced expression of PKC isoenzymes may be a consequence of Sol muscle atrophy that is caused mainly by an increased protein breakdown through the ubiquitin/proteasome pathway (Bodine et al. 2001). Nevertheless, we also observed a reduced PKC expression in the EDL muscle that shows little or no atrophy. Moreover the reduction of PKC expression in Sol muscles was very similar after the two HU periods, whereas atrophy is by far more pronounced after 14 days of HU. Another possibility for reduced expression/activity of PKC is the reduced electrical activity of the Sol muscle during hindlimb unloading. Indeed, Sol muscle EMG activity is quite zeroed immediately after HU and remains very low after 3 days of HU, which may explain a reduced PKC activity early during HU since nerve stimulation has been shown to increase PKC activity (Cleland et al. 1989; Ohira et al. 2002; De-Doncker et al. 2005). After 14 days of HU, the mean EMG activity of Sol muscle has recovered control levels and such a mechanism may not apply for the effects observed at HU14 (Ohira et al. 2002; De-Doncker et al. 2005). However, it remains a possibility that PKC expression depends also on the EMG pattern, which shifts from tonic in control Sol to phasic in HU14 Sol (De-Doncker et al. 2005). A phasic EMG is more similar to the EMG activity of fast-twitch muscles and may corroborate the slow-to-fast transition of Sol muscle as well as our hypothesis formulated above of a heterogeneous population of ‘slow-like’ and ‘fast-like’ fibres, with high and low basal PKC expression/activity, respectively. On the other hand, it is worth noting that, even in the presence of EMG activity, a passively shortened muscle, such as the HU Sol, does not develop tension. Thus, since in vitro muscle contraction increases PKC activity, the reduced PKC activity we recorded during HU may result from the lack of contractile activity (Richter et al. 1987).

What could be the consequences of a reduced PKC activity in the disused Sol muscle? Some data indicate that a reduction of PKC activity may directly affect muscle function. Indeed, previous results suggest that a reduced activity of PKCα may be involved in HU-induced changes of contractile proteins that determine muscle tension (Choi et al. 2005). Here we demonstrate that the reduced PKC activity is an early event of Sol muscle disuse able to modify the g_Cl. In turn, the increased g_Cl is likely to contribute to the observed reduction of muscle excitability, evidenced by the increased rheobase current, I_th, and the decreased I_th/I₂ ratio. Indeed it has been shown that Sol muscle acidification, which decreases the g_Cl, reduces the rheobase current (Pedersen et al. 2005). Thus the increased I_th induced by HU may result from the increased g_Cl. The same authors proposed that inhibition of g_Cl by the intracellular acidification developing during intense activity may increase excitability, thereby protecting the working muscle from fatigue (Pedersen et al. 2004, 2005). It is therefore reasonable to propose that the increased g_Cl as soon as after 3 days of HU may contribute to the impaired Sol muscle resistance to fatigue. Moreover, since PKCθ co-operates with calcineurin in the activation of slow muscle genes in cultured myogenic cells (D'Andrea et al. 2006), the reduced expression of PKCθ we observed in HU muscle may be a key event in the signalling cascade leading to slow-to-fast fibre type transition of Sol muscle. In addition, reduced muscle activity as during space flights, bed rest, or hindlimb unloading has been shown to induce insulin resistance, which likely contributes to the development of muscle atrophy (Stein et al. 1994; Stuart et al. 1988; Mondon et al. 1992). A number of studies support the idea that insulin may rapidly modulate PKC expression, while many PKC isoenzymes are able to modulate insulin signalling in a complex manner (Sampson & Cooper, 2006). It would thus be important to evaluate the possible link between PKC expression/activity and insulin resistance in models of reduced physical activity, such as the hindlimb unloading rodent.

In conclusion, our study demonstrates that the reduction of PKC activity is an early event occurring during muscle disuse, which directly affects excitability parameters of myofibers. These effects are likely to contribute to the adaptation of slow-twitch muscles to a reduced functional demand and render the muscle unadapted upon return to normal activity. Therapeutic interventions able to increase PKC expression/activity may represent a promising countermeasure against muscle damage induced by disuse, such as during bed rest or spaceflights.