Anat Embryol (2001) 203:193–201
© Springer-Verlag 2001
O R I G I N A L A RT I C L E
Frank Hoover · Iacob Mathiesen · Bjørn S. Skålhegg Terje Lømo · Kjetil Taskén
Differential expression and regulation of the PKA signalling pathway in fast and slow skeletal muscle
Accepted: 6 November 2000
Abstract To identify intracellular signalling pathways that transduce muscle electrical activity, we have investigated the Protein Kinase A (PKA) pathway in fast and slow skeletal muscle. The slow soleus muscle (SOL) displayed approximately twice as much PKA catalytic activity and cAMP-binding compared to the fast Extensor Digitorum Longus (EDL) muscle. These results were confirmed by Western blot analysis using antibodies directed against the catalytic or regulatory subunits of PKA. PKA subunits were concentrated at the neuromuscular junction in innervated and denervated muscle fibers demonstrating that PKA is expressed post-synaptically. In addition, we also detected PKA subunits outside the junctional area, suggesting that PKA functions outside of the synaptic regions. Following denervation, levels of cyclic AMP, PKA C activity, R cAMP-binding and RI alpha protein levels increased significantly in the SOL, in contrast to the EDL where only elevated levels of RI alpha protein were observed. These observations demonstrate that PKA levels in skeletal muscle are subject to control at several levels and suggest that some of the differences may be in the pattern of electrical activity that motoneurons impose on the SOL and EDL. Keywords cAMP · Electrical activity · Denervation
F. Hoover (✉) · I. Mathiesen · T. Lømo Department of Physiology, Institute of Basic Medical Sciences, University of Oslo, PO Box 1103, Blindern, 0317 Oslo, Norway B.S. Skålhegg · K. Taskén Department of Medical Biochemistry, Institute of Basic Medical Sciences, University of Oslo, PO Box 1112, Blindern, 0317 Oslo, Norway Present address: F. Hoover, University of Bergen, High Technology Center, Dept. of Virology, PO Box 7800, 5200 Bergen, Norway e-mail:
[email protected] Tel.: +47-55-58 45 02, Fax: +47-55-58 45 12
Introduction The patterns of activity that motoneurons discharge onto muscle fibers vary greatly. In adult rats, fast motoneurons activate fast muscle fibers with brief, relatively infrequent bursts of impulses at high frequencies (near 100 Hz). Slow motoneurons stimulate slow muscle fibers more often and for long continuous periods at around 20 Hz (Hennig and Lømo 1985). These fast and slow activity patterns can operate molecular switches that control specific and reversible gene expression programs in skeletal muscle (Ausoni et al. 1990; Buonanno et al. 1998; Buonanno and Fields 1999; Hughes 1998). The signalling pathways that mediate the effects of electrical muscle activity on muscle gene expression are not well understood. The following evidence suggests that in rat protein kinase A (PKA, a serine-threonine kinase) may be involved in such signalling. First, denervation results not only in muscle inactivity but also in increased levels of cAMP and PKA (Carlsen 1975; Chahine et al. 1993) as well as nicotinic acetylcholine receptor (AchR) and sodium channel subunit mRNAs and proteins (Goldman et al. 1985; Offord and Catterall 1989). Second, in cultured muscle cells, pharmacological manipulations that increase the levels of cAMP induce expression of AChR and sodium channel subunits (Chahine et al. 1993; Sherman et al. 1985). Third, at neuromuscular junctions (NMJs), cAMP counteracts the destabilization of AChRs induced by denervation (Xu and Salpeter 1995). Finally, inhibition of the calcineurin (a serine-threonine phosphatase) pathway, has been shown to promote slow to fast fiber transformation (Chin et al. 1998). Cyclic AMP activates PKA by binding to the regulatory subunits (R) of the tetrameric PKA holoenzyme. This liberates the serine/threonine kinase activity of the catalytic subunits (C), resulting in phosphorylation of cellular proteins. Functional diversity and specific targeting of PKA can be achieved in different ways. First, the PKA holoenzyme can be assembled from different isoforms of R (RI alpha, RII alpha, RI beta and RII
194
beta) and C (C alpha, C beta and C gamma) subunits (Skålhegg and Taskén 1997). These subunits are encoded by different genes and associated splice variants. Second, the site of expression of these genes can vary. The alpha forms are considered to be ubiquitous and constitutive, whereas the beta forms are expressed primarily in the nervous system (Brandon et al. 1997). Finally, the location of RI and RII subunits within the cell vary. While it is generally observed that RI subunits tend to be found in the cytosolic fraction, the RII subunits predominate in the particulate fraction (Skålhegg and Taskén 1997), where AKAPs (A-kinase anchoring proteins) may target their actions to particular subcellular sites (Scott and McCartney 1994; Burton et al. 1997; Huang et al. 1997a,b; Gray et al. 1998). The aim of the present work was to identify properties of PKA that could be regulated by electrical activity. First, we examined the PKA signalling system in fast Extensor Digitorum Longus (EDL) and slow Soleus (SOL) muscles of adult rats to identify differences associated with the strikingly different patterns of motor unit activity on fast and slow fiber types in EDL and SOL, respectively (Gundersen et al. 1988). Second, we denervated EDL and SOL to examine the effects of muscle inactivity on PKA signalling, and to assess whether PKA expression at the neuromuscular junction was pre- or postsynaptic. Third, we compared components of the PKA pathway in endplate-containing and endplate-free regions of the muscle in an attempt to separate synaptic and non-synaptic functions of PKA.
Materials and methods Animals and surgical procedures Adult male Long Evans or Wistar rats (150–300 g/body weight) were used in this study. Surgical operations were performed under Equithesin anesthesia (42.5 mg chloral hydrate and 9.7 mg pentobarbital in 1 ml solution, 0.4 ml/100 g body weight, i.p.). Unilateral denervations were performed by reflection of a 2 mm segment of the sciatic nerve at the level of the trochanter. These experiments were approved under the guidelines established by the Norwegian Experimental Board and Ethical Committee for Animal Experiments and were overseen by the responsible veterinarian for the animal house. The animals were allowed to survive for up to 10 days following denervation. At the respective time point, the SOL and EDL muscles on both the denervated and innervated contralateral side were removed from deeply anesthetized animals. The muscles were either homogenized immediately for use in protein activity assays, frozen in liquid nitrogen and stored at –70°C for further use (cAMP level determination, protein assays or RNA preparation), or placed in ice-cold 4% paraformaldehyde/ PBS (100 mM phosphate buffer, 145 mM NaCl) fixative. Following muscle extraction the animals were killed by cervical dislocation.
Samples were resuspended in assay buffer and assayed in duplicate. Three to six muscles were used for each time point. Differences between animals were avoided by using paired comparisons; each denervated muscle (right leg) was compared to the control innervated muscle (left leg). The resulting numbers were normalized to wet weight of the muscle. Protein extracts Fresh or frozen muscle whole extracts from innervated or denervated SOL and EDL muscles were prepared by mechanical disruption in ten vol. of ice-cold Homogenization Buffer (HB+T) containing 10 mM potassium phosphate (pH 6.8), 1 mM EDTA, 250 mM sucrose, 0.5% Triton X-100, and a cocktail of protease inhibitors (chymostatin, leupeptin, antipain and pepstatin A; Peninsula Laboratories). The homogenate was centrifuged at 4°C for 30 min at 15,000 g. Following centrifugation, the supernatant was decanted, collected and the protein concentration was determined by using a modified Bradford assay (Sedmak and Grossberg 1977). Soluble proteins were prepared by homogenization in ten vol. of homogenization buffer without Triton X-100 (HB) and the samples were centrifuged as above. The supernatant was collected, the protein content quantified and the pellet was resuspended, washed twice with five vol. of HB and centrifuged as above. The remaining pellet was re-extracted in five vol. of (HB+T) and centrifuged as above. This Triton X-100-soluble supernatant was decanted and quantified for further use. Phosphotransferase activity of PKA Catalytic activity of PKA in muscle extracts was assayed by phosphorylating a PKA-specific substrate ((Kemp et al. 1977), Kemptide, Peninsula Laboratories) using [gamma-32P]-ATP (specific activity 6,000 Ci/mMol, Amersham) in an assay mixture described by Roskoski (Roskoski 1983). Calculation of the molar concentration of C was based on the specific activity of homogenous bovine heart C subunit (15 µmol/min/mg). Phosphotransferase activity was measured both in the presence and absence of cAMP (5 µM) and PKI (1 µM). The addition of PKI virtually abolished all phosphotransferase activity. Additional controls consisted of using substrates for other kinases: CamK I (synapsin site I), CamK II (Autocamtide 3, BRL) and protein tyrosine kinases (Promega). In each of these instances minor or background levels of phosphorylation were observed. Catalytic activity of PKA in muscle extracts was confirmed using PhosphoSpots (Jerini, BioTools, GmbH). Briefly strips containing covalently bound substrate peptide were reacted with muscle extracts (normalized for the same protein concentration) under similar conditions to above (Roskoski 1983). The reactions were stopped with extensive washes at 37°C in: (1) 1 M NaCl; (2) 10 mM TRIS pH 7.4; 1% Tween 20, 1% SDS; (3) Water; (4) 100% ethanol before exposing to film. Cyclic AMP binding measurements Quantification of specific [3H]cAMP binding of solublized PKA regulatory subunits was performed as described by Cobb and Corbin (Cobb and Corbin 1988) in a mixture containing [2,8-3H]cAMP (3 µM; specific activity of 5 Ci/mMol in the reaction; Du Pont-New England Nuclear). Molar ratios of R subunits were calculated based on two cAMP binding sites on each regulatory subunit monomer.
Determination of cAMP levels in skeletal muscle Antibodies The levels of cAMP were measured in rat SOL or EDL muscle using a commercially available kit (Amersham) based on a radioligand competition assay. Individual muscles were homogenized in ice-cold 6% trichloroacetic acid, extracted with multiple rounds of water-saturated ether and dried in either a speed vacuum centrifuge or lyophilized. The samples were stored at –70°C until use.
An anti-human C gamma polyclonal antibody (cat. no. SC-905, Santa Cruz Biotechnology) was used at a concentration of 0.1 ug/ml for immunoblot analysis and 1.0 ug/ml for immunocytochemistry. The corresponding competitor peptide was used at 750-fold molar excess. Although this antibody was generated
195 against a C gamma peptide, it recognized specifically C alpha and C beta recombinant proteins (Reinton et al., 2000). Monoclonal antibodies directed against human RI alpha and human RII alpha (cat. no. P53620, p55120, respectively, KT in collaboration with Transduction Laboratories, (Kubo et al. 1986; Eide et al. 1998; Collas et al. 1999; Keryer et al. 1999)) were used at a concentration of 1.0 µg/ml for immunoblot analysis. The RII alpha antibody was used at a concentration of 25 µg/ml for immunostaining, and competed with a 10-fold molar excess of human RII alpha fulllength recombinant protein (K. Taskén, unpublished). We used affinity purified rabbit anti-peptide antibodies to rat and human RI alpha (Taskén et al. 1993; Imaizumi-Scherrer et al. 1996) at a concentration of 5 µg/ml. Competition of staining by RIalpha antibodies was attempted with corresponding peptides at concentrations of 100 to 2,000-fold molar excess. Specificity of all antibodies have been previously established and was verified here by the use of purified/recombinant protein standards of the immunoblots and by competition experiments in the immunofluoresence studies.
transversely into a middle third segment containing all the NMJs and two lateral segments free of junctions (Brown et al. 1976; Taxt 1983). We detected similar levels of cAMP in these regions (Table 1). We did not examine the EDL in this manner since endplate-containing and endplate-free segments are less clearly delineated.
Western blotting Proteins were separated by SDS-PAGE and transferred by electroblotting to nitrocellulose membranes; immunoblot analysis was performed as described elsewhere (Skålhegg et al. 1992). Primary antibodies were detected by horseradish-peroxidase labelled protein A (dilution 1/25,000, Amersham; for polyclonal antibodies) or anti-mouse IgG (dilution 1/5,000, Transduction Laboratories, for monoclonal antibodies) in the second layer and developed using ECL (Amersham). Results were evaluated using visual inspection. The resulting exposures were scanned into a computer, manipulated only for brightness and contrast in Adobe Photoshop (V4.0, Adobe Systems) and imported into Adobe Pagemaker (V6.5, Adobe Systems) to create the figures.
Results Cyclic AMP in SOL and EDL muscles We assayed the concentration of cAMP in extracts of normal SOL and EDL muscles, and report that cAMP levels were similar in both (Table 1) and comparable to that reported for gastrocnemius and diaphragm muscles, which contain a mixture of slow and fast fibers (Carlsen 1975; Hopkins and Manchester 1981). These basal levels of cAMP correspond to the intracellular concentrations needed to partially activate PKA isoenzymes (Dostmann and Taylor 1991; Sandberg et al. 1991). Previous reports have suggested that PKA is localized to the neuromuscular junction (NMJ; Imaizumi-Scherrer et al. 1996). One prediction from this work, is that cAMP levels might be higher in junctional (J) versus non-junctional (XJ) regions. To test this notion, the SOL was cut Table 1 Levels of cAMP in SOL and EDL skeletal muscles. Total levels of cAMP in slow SOL, fast EDL, and in SOL extracts enriched for junctional and extra-junctional regions. Mean values±SEM are shown and are reported as fmol cAMP/mg wet weight muscle Muscle
Fraction
[cAMP]
Sample size
SOL EDL SOL SOL
Whole Whole Junctionally-enriched Extra-junctional
278±59.5 255±43.2 215±14.8 220±19.5
8 10 11 11
Fig. 1A, B Changes in cAMP levels following muscle denervation. A Rat muscles were denervated and isolated along with the contralateral innervated control muscles at 1 day (d), (n=4 SOL, n=3 EDL); 2–3 d, (n=8 SOL, n=11 EDL); 10 d, (n=3 SOL, n=5 EDL). The ratio of cAMP levels in denervated versus innervated muscle in individual animals was calculated. B Distribution of cAMP in naive (J, XJ), denervated (d), junctionally-enriched (J) and extra-junctional (XJ) regions of the SOL at selected time points 3 h (n=5), 1 d (n=3), 2 d, (n=3). Mean values±SEM are shown
196
Denervation significantly (Mann-Whitney rank comparison test, P
