Physiol. Genomics 22: 227-243, 2005.
First published April 19, 2005; doi:10.1152/physiolgenomics.00210.2004
1094-8341/05 $8.00
Received 10 September 2004;
accepted in final form 13 April 2005.
Physiological Genomics 22:227-243 (2005)
1094-8341/05 $8.00 © 2005 American Physiological Society
Comparison of gene expression of 2-mo denervated, 2-mo stimulated-denervated, and control rat skeletal muscles
Tatiana Y. Kostrominova1,2,
Douglas E. Dow1,3,
Robert G. Dennis1,3,
Richard A. Miller1,4 and
John A. Faulkner1,2,3
1 Institute of Gerontology
2 Department of Molecular and Integrative Physiology
3 Department of Biomedical Engineering, and
4 Department of Pathology, University of Michigan, Ann Arbor, Michigan
 |
ABSTRACT
|
|---|
Loss of innervation in skeletal muscles leads to degeneration, atrophy, and loss of force. These dramatic changes are reflected in modifications of the mRNA expression of a large number of genes. Our goal was to clarify the broad spectrum of molecular events associated with long-term denervation of skeletal muscles. A microarray study compared gene expression profiles of 2-mo denervated and control extensor digitorum longus (EDL) muscles from 6-mo-old rats. The study identified 121 genes with increased and 7 genes with decreased mRNA expression. The expression of 107 of these genes had not been identified previously as changed after denervation. Many of the genes identified were genes that are highly expressed in skeletal muscles during embryonic development, downregulated in adults, and upregulated after denervation of muscle fibers. Electrical stimulation of denervated muscles preserved muscle mass and maximal force at levels similar to those in the control muscles. To understand the processes underlying the effect of electrical stimulation on denervated skeletal muscles, mRNA and protein expression of a number of genes, identified by the microarray study, was compared. The hypothesis was that loss of nerve action potentials and muscle contractions after denervation play the major roles in upregulation of gene expression in skeletal muscles. With electrical stimulation of denervated muscles, the expression levels for these genes were significantly downregulated, consistent with the hypothesis that loss of action potentials and/or contractions contribute to the alterations in gene expression in denervated skeletal muscles.
gene array; long-term denervation; muscle atrophy; electrical stimulation
|
INTRODUCTION
|
|---|
INNERVATION IS A CRITICAL factor for the support of the structural and functional integrity of skeletal muscles. Nerve and spinal root damage associated with traumatic injuries leads to denervation of muscle fibers, fiber atrophy, and loss of function. In addition, a number of neurodegenerative disorders, as well as normal aging of skeletal muscles, are associated with denervation of muscle fibers. Depending on the type and severity of the denervation, either the entire muscle may be denervated, as with an immediate nerve transection, or partial muscle denervation/reinnervation processes might continue for years, similar to the processes that occur during normal aging. Skeletal muscles of both humans and animals show a progressive accumulation of denervated muscle fibers with aging (15, 58). Regardless of whether the timing of the denervation process is rapid or progressive, denervated muscles undergo a decline of muscle mass and force due to the atrophy and decreased myofibrillar content of the muscle fibers (16).
After denervation, the functional decline of muscles correlates highly with a number of modifications in morphological, biochemical, and physiological properties that reflect changes in the mRNA expression of a wide range of genes. Some of these genes are well known, and their expressions are often used as models for the studies of gene expression in denervated skeletal muscles. After denervation of skeletal muscles, the upregulation of mRNA expression of myogenic transcriptional factors [myoblast determination protein (MyoD), myogenin, myogenic regulatory factor-4 (MRF4)] and subunits of acetylcholine receptors (AChR) is well documented (1, 13, 27). Increases in mRNA expression of myogenic transcriptional factors correlate highly with the upregulation of their protein levels in denervated skeletal muscles (22, 58). Muscle atrophy and upregulation of the machinery for protein degradation is another well known characteristic of denervated skeletal muscles (37). During muscle atrophy multiple proteolytic pathways are activated (52, 61, 78, 87). For a number of genes, only one or two publications demonstrate their regulation of expression in response to denervation. After skeletal muscle denervation, the regulation of acute myeloid leukemia-1 (AML1) (99) and the regulation of 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase (14) are each supported by only one report. The goal of this study was to use gene array technology to identify genes not known previously to be regulated by denervation to provide an in-depth analysis of those genes in subsequent investigations. In addition, we tested the hypothesis that denervation of skeletal muscles upregulates a wide range of structural and functional genes and that electrical stimulation of denervated muscles maintains the levels of expression at values not different from those of innervated control muscles.
|
MATERIALS AND METHODS
|
|---|
Animals and muscle denervation.
Brown-Norway male rats, obtained from the National Institute on Aging colony at Harlan Sprague-Dawley (Indianapolis, IN), were used for the experiments. Animal housing, operations, and subsequent animal care were carried out in accordance with the guidelines of the Unit for Laboratory Animal Medicine at the University of Michigan. Experimental protocols used in this study were approved by the University of Michigan Animal Care and Use Committee. At 4 mo of age, the right legs of 11 rats were denervated by a high sciatic nerve section in the hip region of the hind limb (94). After pentobarbital sodium anesthesia, the right sciatic nerve was ligated tightly with silk in two places, and the nerve was cut between the sutures. Proximal and distal nerve stumps were implanted into muscular tissue as far away from each other as possible. This procedure results in a permanent denervation of the lower hind leg (94). Two months after denervation (6 mo of age), extensor digitorum longus (EDL) muscles were removed from the operated legs and frozen in liquid nitrogen. The animals were killed by an overdose of anesthetic. The EDL muscles from 11 age-matched nonoperated rats served as innervated controls. All tissues were stored at –80°C before RNA isolation. Ten additional rats that had undergone the denervation procedure at 4 mo of age had battery-operated stimulators implanted subcutaneously on their backs (24). The stimulators generated series of symmetrical and balanced bipolar pulses that produced tetanic contractions in the denervated EDL muscles (20 bipolar pulses/contraction, 100 Hz, 9 V, 200 contractions/day, with a 7.2-min interval between contractions). The design of the stimulators and implantation procedures were as described previously (24). This stimulation procedure preserves muscle mass and maximum force at levels not different from the innervated control EDL muscles of rats (24). The activity of the stimulators was monitored weekly. Muscle samples were taken for the analysis only if the stimulators were working continually without interruptions during the 2-mo period and the muscles showed no chronic inflammation due to the denervation and implantation. Three denervated rats received sham stimulators that lacked battery power. Two months after denervation-stimulation, EDL muscles were removed from the operated legs and were frozen in liquid nitrogen. For three of each denervated, stimulated-denervated, sham-stimulated, and control rats, midbelly parts of the dissected EDL muscles were removed, washed in PBS, and embedded in TBS medium (Triangle Biomedical Sciences, Durham, NC) for histological evaluation. Muscles used for histology were not used in the microarray analysis.
RNA isolation, synthesis of labeled cDNA targets, and hybridization to the membranes.
Total RNA was isolated by homogenization of muscles in TRIzol (Gibco-BRL, Grand Island, NY) followed by the single-step purification method described by the manufacturer's protocol. DNA contamination was removed by digestion with RNase-free DNase I by use of the DNA-free kit (Ambion, Austin, TX). RNA concentrations were estimated by spectrophotometer. Labeled cDNA was prepared with an ATLAS cDNA Expression Array kit (Clontech Laboratories, Palo Alto, CA), following the manufacturer's suggested protocol with small modifications. Briefly, 5 µg of total RNA were used to synthesize 32P-labeled first-strand cDNA by reverse transcription with Superscript II (Invitrogen, Carlsbad, CA). Labeled cDNA was purified by CHROMA SPIN-200 column chromatography (Clontech Laboratories). cDNA fractions with highest radioactivity were pooled and hybridized to the Rat Atlas 1.2 Array II membranes (Clontech Laboratories). Each membrane has 1,176 spotted cDNA fragments of various rat genes along with negative and positive controls. After a 30-min prehybridization at 68°C in ExpressHyb buffer (Clontech Laboratories) supplemented with 100 µg/ml sheared salmon sperm DNA (Invitrogen), denatured 32P-labeled cDNA was added to the hybridization mix. Hybridization was performed overnight at 68°C with continuous agitation. Membranes were washed four times for 30 min in buffer 1 (2x SSC/1% SDS) and two times for 30 min in buffer 2 (0.1x SSC/0.5% SDS) at 68°C, sealed in a bag (Wallac), and exposed to phosphor screen for 3–5 days. Labeling and hybridization of RNA from each of the eight denervated and eight control EDL muscles were performed twice.
Data acquisition and normalization.
Digital images were obtained with a phosphorimager (Molecular Dynamics, Sunnyvale, CA) and processed with Array Vision software (Imaging Research, St. Catharine, ON, Canada) as described by Dozmorov et al. (25). Background-subtracted pixel volumes were generated for each of the spots on the membrane. Each value was transformed to its common logarithm and used in subsequent steps of the analysis. For the analysis, a previously described linear regression method of data normalization was used (25). In brief, a "standard array" was calculated as the medial level for each of the 1,176 genes across the duplicate sets of data for all 16 EDL muscles that were analyzed. A linear regression method was employed to adjust set of measurements for each EDL muscle to obtain the same slope and intercept as in the standard array. After normalization of data for duplicate hybridizations of each RNA sample, average values were taken for statistical analysis.
In compliance with the minimum information about microarray experiments (MIAME) standards, a complete data set was submitted to the National Center for Biotechnology Information's Gene Expression Omnibus database (accession no. GSE1741).
Assessment of statistical significance.
Statistical significance of the results was assessed with the significance analysis of microarrays (SAM) algorithm developed by Tusher et al. (93). The SAM algorithm computes a statistic for each gene and measures the strength of the relationship between gene expression and the response variable. Repeated permutations of the data determine whether the expression of a specific gene was significantly different between test groups. The criterion for presentation in this report was a false discovery rate (FDR) 
0.05. We also calculated Student's t-statistic (2-tailed distribution, equal variance) for each gene in the data set.
Reverse transcription and real-time PCR.
Quantitative RT-PCR (QRT-PCR) was performed as follows: 1 µg of total RNA was reverse transcribed using a ThermoScript RT kit (Invitrogen). Real-time PCR was performed with an Opticon rapid thermal cycler system (MJ Research, Waltham, MA). Amplifications were performed in a 20-µl total volume with 50 pmol of primers and MgCl2 concentration optimized between 2 and 5 µM. Nucleotides, Taq DNA polymerase, and buffer were included in the Master SYBR Green mix (Qiagen, Valencia, CA). An amplification protocol incorporated an initial incubation at 95°C for 10 min for the activation of FastStartTaq DNA polymerase followed by 40 cycles, with a 94°C denaturation for 15 s, 58–60°C annealing for 30 s, and 72°C extension for 30 s. Detection of the fluorescent product was performed at the end of the 72°C extension period. To confirm the amplification specificity, the PCR products from each primer pair were subjected to a melting curve analysis and subsequent agarose gel electrophoresis. Relative quantification was performed based on the threshold cycle (CT value) for each of the PCR samples (63). Because the mRNA expression of muscle creatine kinase (MCK) does not change after denervation (1, 21), this gene was used for normalization. Student's t-test analysis and standard error were calculated. The sequences of the primers used for the QRT-PCR analysis are summarized in Table 1.
Western blotting.
Rat muscles were dissected, frozen in liquid nitrogen, pulverized, and homogenized in solution containing 20 mM Tris·HCl (pH 6.8), 4% (wt/vol) SDS, 1 mM phenylmethylsulfonyl fluoride (PMSF), and 1 µm each of leupeptin and pepstatin A. Protein concentrations were determined using the Bio-Rad DC protein assay (Hercules, CA). Protein samples were mixed with loading buffer, subjected to SDS-PAGE, and transferred electrophoretically to Immobilon-P membranes (Millipore, Bedford, MA). Gels with identical samples were stained with Coomassie brilliant blue and used as an additional control of equivalence of protein loading. After transfer, Immobilon-P membranes were blocked in buffer containing 5% dry milk in phosphate-buffered saline-0.2% Tween 20 (PBST) and then incubated overnight at 4°C with primary antibodies. The following primary antibodies were used: mouse monoclonal antibody recognizing all myosin heavy chain isoforms (clone MF-20; obtained from the Developmental Studies Hybridoma Bank, The University of Iowa, Iowa City, IA), rabbit anti-SK3 channel antibody (Sigma, St. Louis, MO), rabbit anti-S6 ribosomal protein and rabbit anti-phospho-S6 ribosomal protein (Cell Signaling Technology, Beverly, MA), and mouse anti-GAPDH (Chemicon International, Temecula, CA). Depending on the source of primary antibody, immunodetection was done using peroxidase-conjugated anti-mouse or anti-rabbit antibody (Jackson ImmunoResearch Laboratory, West Grove, PA) with subsequent chemiluminescence (Pierce, Rockford, IL). Band intensity was quantified by scanning densitometry.
Histochemical and immunohistochemical analysis.
EDL muscles were embedded in TBS medium (Triangle Biomedical Sciences, Durham, NC) and were sectioned on a cryostat (12 µm). Type I, type IIA, and type IIB fibers were differentiated on the basis of myofibrillar ATPase activity at pH 4.5 as described previously (12). Interstitial connective tissue was identified in muscle sections by Trichrome Masson staining (65).
Cytochrome c oxidase (COX) and succinate dehydrogenase (SDH) enzymatic activities were visualized in muscle sections using previously described protocols (26). SDH and COX activities were used to assay the proportion of oxidative fibers as well as to assess functional activity of the mitochondria in EDL muscles. Oxidative fibers have a higher number of mitochondria than glycolytic muscle fibers. High SDH but low COX activities in the same muscle fiber suggest damage in the mitochondrial electron transport system.
For immunostaining, sections were incubated in blocking buffer (20% calf serum in PBST) for 1 h and then in the solution of primary antibody overnight at 4°C. The following primary antibodies were used: mouse anti-myogenin (clone F5D), mouse anti-slow myosin (clone A4.84), and mouse anti-fast IIa myosin (clone A4.74) (all three obtained from the Developmental Studies Hybridoma Bank); mouse anti-GAPDH, rabbit anti-laminin, and rabbit anti-neural cell adhesion molecule (anti-N-CAM) (all three from Chemicon International); and rabbit anti-SK3 potassium channel (Sigma). Depending on the source of primary antibody, 1 h of room temperature incubation with anti-mouse or anti-rabbit Cy3- and Cy2-conjugated secondary antibodies (Jackson ImmunoResearch Lab) was used for visualization. Nuclei were stained by 5-min incubation with a bis-benzimide solution (Sigma) in PBST. The sections were examined and photographed with a Zeiss Axiophot-2 microscope (Carl Zeiss).
|
RESULTS
|
|---|
Morphological alteration and changes in fiber type composition in denervated and stimulated-denervated compared with innervated control EDL muscles of rats.
Compared with control EDL muscles, muscles denervated for 2 mo showed substantial atrophy of fibers (Fig. 1, Aa, Ab, Ba, Bb, Bd, and Be) and an increased accumulation of interstitial connective tissue (Fig. 1, Ad and Ae). Despite these changes, the expression of myosin ATPase isoforms was preserved (Fig. 1Ab). The preservation of the isoforms permitted identification of type I, type IIa, and type IIb fibers (Fig. 1Ab). Compared with innervated fibers (Fig. 1Aa), after denervation, type IIa and IIb fibers showed the greatest amount of atrophy, and type I fibers the least (Fig. 1Ab). Electrical stimulation prevented atrophy of type I, type IIa and type IIb fibers (Fig. 1Ac). The interstitial connective tissue content in stimulated-denervated muscles was maintained at levels not different from those of control EDL muscles (Fig. 1Af).
View larger version (70K):
[in this window]
[in a new window]
|
Fig. 1. Evaluation of changes in fiber type composition and connective tissue accumulation after denervation and stimulation-denervation of extensor digitorum longus (EDL) muscles of rats. A: histological evaluation of fiber type composition using histochemical staining for myosin ATPase activity (a, b, and c) and interstitial connective tissue accumulation using Trichrome Masson staining (d, e, and f) in innervated (a and d), 2-mo denervated (b and e), and 2-mo stimulated-denervated (c and f) EDL muscles. Black fibers in a, b, and c represent slow (type I) fibers, light-brown fibers represent fast IIa fibers, and dark brown fibers represent fast IIb muscle fibers. Bar = 200 µm. B: immunohistochemical evaluation of slow (a, b, and c) and fast IIa (d, e, and f) myosin heavy chain (MHC) expression in innervated (a and d), 2-mo denervated (b and e), and 2-mo stimulated-denervated (c and f) EDL muscles, shown in red. Immunostaining for laminin, shown in green, was used for visualization of fiber outlines (a–f). Bar = 300 µm. C: estimation of MHC mRNA expression levels in control, 2-mo denervated, and 2-mo stimulated-denervated EDL muscles of rats using quantitative RT-PCR (QRT-PCR). MHC I, slow myosin; MHC IIa, MHC IIb, and MHC IIx, fast IIa, fast IIb, and fast IIx isoforms, accordingly. Data are expressed as means ± SE; n = 3. Expression level of muscle creatine kinase in each sample was used for normalization. *Significant difference compared with the expression levels in control EDL muscles at p(t) 0.05. #Significant difference compared with the expression levels in stimulated-denervated EDL muscles at p(t) 0.05. D: representative Western blotting and histogram of relative levels of MHC protein expression (% from total MHC expression in the sample) in control (C), 2-mo denervated (D), and 2-mo stimulated-denervated (S) EDL muscles of rats. Data are expressed as means ± SE; n = 3. Significant difference compared with the expression levels in control EDL muscles: p(t) 0.05 (*) and p(t) 0.01 (**).
|
|
Immunostaining, using antibodies against myosin heavy chain slow (type I) and fast IIa isoforms (Fig. 1B), showed changes similar to those detected with myosin ATPase staining (Fig. 1A). After denervation, an increase in the proportion of slow (Fig. 1Bb) and fast IIa (Fig. 1Be) fibers was observed. In electrically stimulated muscles, slow fibers were easily identified (Fig. 1Bc). Fast IIa fibers were not recognized as readily (Fig. 1Bf), since they did not show bright staining as observed in control (Fig. 1Bd) and denervated (Fig. 1Be) muscles.
In denervated, stimulated-denervated, and control EDL muscles, levels of mRNA expression of four different myosin heavy chain isoforms were compared by QRT-PCR (Fig. 1C). In denervated muscles, mRNA expression of myosin heavy chains was upregulated 
6-fold for slow myosin, 10-fold for fast IIa myosin, and 2-fold for fast IIb and fast IIx myosins. In stimulated-denervated EDL muscles, mRNA expression of myosin heavy chains was upregulated 2-fold for slow and fast IIb myosins, not changed for fast IIx myosin, and decreased 10-fold for fast IIa myosin. The increase in the proportion of slow and fast IIa fibers in denervated muscles and decrease in the proportion of fast IIa fibers in stimulated-denervated EDL muscles were confirmed by Western blotting, using an antibody recognizing all myosin heavy chain isoforms (Fig. 1D). Stimulated-denervated muscles expressed predominantly the myosin heavy chain fast IIb isoform.
In denervated EDL muscles, analysis of the expression of GAPDH, one of the glycolytic pathway enzymes often used as a marker of the glycolytic capacity (64), provided additional evidence of the decrease in fast glycolytic (fast IIb) fibers (Fig. 2). QRT-PCR analysis did not detect a statistically significant decrease in GAPDH mRNA in denervated compared with control EDL muscles [60% of control value, n = 8, p(t) = 0.06]. In contrast, Western blotting identified a 10-fold decrease in GAPDH protein expression in denervated compared with control EDL muscles (Fig. 2B) that suggested a posttranscriptional regulation of expression. In stimulated-denervated muscles, the level of GAPDH protein expression was not different from the control values. Immunostaining with anti-GAPDH antibodies when compared with data on control muscles detected a substantial decrease of this protein in denervated muscle fibers (Fig. 2Ab), as well as an increased proportion of muscle fibers expressing high GAPDH levels in stimulated-denervated muscles (Fig. 2Ac).
Compared with control EDL muscles (Fig. 3, A and D), the area of fibers that showed a high intensity of SDH and COX activities relative to the overall muscle area was increased in denervated muscles (Fig. 3, B and E). In both control and denervated muscles, the difference between fibers that showed high or low intensity of SDH and COX activities was much higher than the difference in stimulated-denervated muscles (Fig. 3, C and F). High magnification of SDH stained fibers showed differences in the distribution of mitochondria in denervated (Fig. 3H) compared with control (Fig. 3G) EDL muscles. In control muscles, mitochondria formed elongated fibrillar structures running along the myofibrils, whereas in both oxidative (dark blue staining) and glycolytic (lighter blue staining) fibers of denervated muscles, the mitochondria formed globular conglomerates. Sections of neonatal muscles (Fig. 3J) and adult livers (Fig. 3K) of rats also had a globular distribution of mitochondria.
View larger version (146K):
[in this window]
[in a new window]
|
Fig. 3. Succinate dehydrogenase (SDH; A–C, G–I) and cytochrome c oxidase (COX; D–F) enzymatic activities in serial sections of control (A, D, and G), 2-mo denervated (B, E, and H), and 2-mo stimulated-denervated (C, F, and I) EDL muscles of rats. Note formation of elongated fibrillar mitochondrial structures running along the myofibrills in innervated EDL muscles (G) and accumulation of mitochondria in globular clusters in denervated EDL muscles (H). Section of muscle from 3-day neonatal rat (J) and liver section from adult rat (K) also have globular mitochondrial structures. A–F: bar = 150 µm. G–K: bar = 25 µm.
|
|
Microarray analysis of gene expression in denervated and control rat EDL muscles.
For the 2-mo denervated compared with control muscles, SAM analysis of the normalized microarray data of gene expression (FDR 0.05; Fig. 4, Table 2, and Supplemental Table S1; Supplemental Material is available at the Physiological Genomics web site)1
identified 128 genes with differences in expression. For the 128 genes, of which 121 were upregulated and 7 were downregulated by denervation, the magnitude of the difference varied from 1.6-fold to 39-fold. On the basis of the Student's t-statistics, 33 genes were identified with 0 < p(t) 0.001, 59 genes with 0.001 p(t) 0.01, and 36 genes with 0.01 p(t) 0.05. The genes were subsequently classified into eight groups on the basis of what is known of their functions (Table 2 and Supplemental Table S1) and included transcriptional factors (9 genes), signaling cascades (23 genes), structural and cell adhesion proteins (26 genes), protein synthesis and degradation (21 genes), and metabolism (21 genes). The number of genes identified in each of the groups correlated with the total number of genes from that group represented in the microarrays. About 15% of the genes had no known function and consequently could not be assigned to any of these groups. These genes were placed together arbitrarily as a separate group (Supplemental Table S1).
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 4. Graphical outcome for significance analysis of microarrays (SAM) analysis [false discovery rate (FDR) 0.05] of microarray hybridization data for 2-mo denervated and control EDL muscles of rats.
|
|
View this table:
[in this window]
[in a new window]
|
Table 2. Genes with altered mRNA expression in 2-mo denervated compared with control EDL muscles of rats (FDR 0.05)
|
|
Confirmation of the results of the microarray analysis using QRT-PCR.
For each of the 13 genes studied by QRT-PCR, the inferences derived from the microarray analysis were confirmed, although the magnitudes of changes in the levels of mRNA expression in microarray and QRT-PCR analysis were typically different, with higher ratios seen in the QRT-PCR data (Table 3). The greater magnitude of the difference with the QRT-PCR method reflects the greater sensitivity of the QRT-PCR, since this method compares linear amplification of the transcripts, while saturation of cDNA spots in microarrays does not provide quantitative evaluation of highly expressed transcripts. Among all the genes analyzed after denervation by QRT-PCR, the transcription factors AML1 (273-fold) and myogenin (108-fold) showed the highest levels of upregulation (Table 3).
Comparison of gene and protein expression in denervated, stimulated-denervated, and control rat EDL muscles.
Levels of gene expression in denervated, stimulated-denervated, and control EDL muscles were measured by QRT-PCR for genes selected from a number of functional groups identified by the microarray analysis (Tables 2 and 3). The results are summarized in Fig. 5. Compared with denervated muscles, electrical stimulation downregulated the expression of each of the 10 genes selected for the analysis. Nevertheless, in stimulated-denervated muscles, six of the genes were expressed at levels from two- to sevenfold higher than in control muscles. Compared with the denervated state, stimulation of the denervated muscles produced the greatest changes in the level of expression of the genes for AML1, with a decrease of 40-fold, and myogenin, with a decrease of 28-fold (Fig. 5). Sham-operated EDL muscles had mRNA expression levels not different from the denervated muscles (data not shown).
In denervated muscles, immunostaining for myogenin (Fig. 6), one of the genes that is highly upregulated by denervation and downregulated by electrical stimulation, showed that this protein was localized in myonuclei and in a lesser quantity in the cytoplasm of the small-size atrophic muscle fibers (Fig. 6E). In stimulated-denervated muscles, myogenin immunoreactivity was restricted to the small-size fibers (Fig. 6F). All of the large-size, optimally stimulated muscle fibers were negative.
View larger version (84K):
[in this window]
[in a new window]
|
Fig. 6. Increased expression of myogenin in atrophic denervated fibers in EDL muscles of rats. Immunostaining with antibodies against laminin, shown in green, was used for visualization of fiber outlines in innervated (A, D), 2-mo denervated (B, E), and 2-mo stimulated-denervated (C, F) EDL muscles. Immunostaining with antibodies against myogenin showed bright red nuclear staining in small-size fibers in 2-mo denervated (E) and 2-mo stimulated-denervated (F) but not innervated (D) EDL muscles. D, E, and F show higher magnification of EDL muscle sections than is shown in A, B, and C, respectively. Bar = 100 µm.
|
|
In denervated compared with control EDL muscles, microarray analysis detected a fourfold upregulation in the mRNA expression of N-CAM (Table 2). In innervated muscles, nerves and satellite cells had strong N-CAM expression (Fig. 7, A, D, and G). In denervated muscles, bright immunostaining for N-CAM was observed on the sarcolemma of many muscle fibers and in satellite cells (Fig. 7, B, E, and H). Myonuclei of N-CAM-positive muscle fibers were also surrounded by a layer of N-CAM-positive material (Fig. 7, E and H). In stimulated-denervated muscles, N-CAM-positive immunostaining was restricted to the small-size atrophic fibers and satellite cells (Fig. 7, C, F, and I).
View larger version (72K):
[in this window]
[in a new window]
|
Fig. 7. Increased expression of neural cell adhesion molecule (N-CAM) in denervated muscle fibers of EDL muscles of rats. Immunostaining with antibodies against N-CAM showed bright red staining of satellite cells (arrow) in control (D, G), 2-mo denervated (E, H), and 2-mo stimulated-denervated (F, I) EDL muscles. N-CAM antibodies also recognized nerve fibers in control EDL muscles (large arrowheads in A, D, and G). In 2-mo denervated EDL muscles, bright sarcolemmal and cytoplasmic expression of N-CAM was detected in a large number of fibers (B, E, and H). Often, bright rim of N-CAM-positive staining surrounded the myonuclei (small arrowheads in E and H). In 2-mo stimulated-denervated EDL muscles, bright sarcolemmal and cytoplasmic N-CAM immunostaining was restricted to the small atrophic muscle fibers (C, F, and I). Immunostaining with antibodies against laminin, shown in green, was used for visualization of fiber outlines in control (G), 2-mo denervated (H), and 2-mo stimulated-denervated (I) EDL muscles. A–C: bar = 100 µm. D–I: bar = 50 µm.
|
|
Immunostaining and Western blotting with antibodies against small-conductance calcium-activated channel SK3 (Fig. 8) confirmed previously detected (by microarray) upregulation of SK3 mRNA expression in denervated EDL muscles (Table 2). Innervated EDL muscles had extremely low SK3 protein expression (Fig. 8, Aa, Ad, and B). After denervation, expression of SK3 protein was upregulated approximately sixfold (Fig. 8B). Most of the fibers in denervated muscles showed a variable degree of SK3 immunoreactivity on the plasma membrane (Fig. 8, Ab and Ae). No correlation was observed between fiber type and SK3 expression. In denervated muscles, both slow and fast muscle fibers expressed SK3 (Fig. 8, Aj and Ak). Some of the SK3-positive fibers also expressed embryonic and developmental myosin heavy chain isoforms (Fig. 8, Ag and Ah). Many of the fibers in denervated muscles coexpressed slow and fast IIa (Fig. 8, Aj and Ak) or embryonic/developmental and fast IIa (Fig. 8, Ag, Ah, and Ak) myosin heavy chain isoforms. Satellite cells in denervated muscles also had bright SK3-positive staining on the plasma membrane, while areas surrounding myonuclei were SK3 negative (Fig. 8Al). Electrical stimulation suppressed SK3 protein expression in denervated muscle to levels approximately twofold higher than in innervated muscles (Fig. 8B). The SK3-positive immunoreactivity was restricted to the small-size fibers (Fig. 8, Ac and Af), many of which also expressed embryonic myosin (Fig. 8Ai).
View larger version (76K):
[in this window]
[in a new window]
|
Fig. 8. Expression of SK3 channel in control, 2-mo denervated, and 2-mo stimulated-denervated EDL muscles of rats. A: immunohistochemical evaluation of SK3 channel expression in control (a and d), 2-mo denervated (b, e, and l), and 2-mo stimulated-denervated (c and f) EDL muscles is shown in red. e, g, h, j, and k: immunostaining of serial sections of the same 2-mo denervated EDL muscle at the same magnification. Immunostaining with antibodies against embryonic (red staining in g), neonatal (red staining in h), slow (green staining in j), and fast IIa (green staining in k) MHC isoforms is presented. f and i: immunostaining of serial sections of the same 2-mo stimulated-denervated EDL muscle at the same magnification. Immunostaining with antibodies against embryonic MHC (green staining in i) is presented. Two large arrowheads show the same muscle fibers positive for immunostaining with antibodies against SK3 channel (b and e), embryonic (g), neonatal (h), and fast IIa (k) MHC isoforms but negative for immunostaining with antibodies against slow myosin (j). Two small arrowheads show the same muscle fibers positive for immunostaining with antibodies against SK3 channel (e) and fast IIa (K) MHC but negative for immunostaining with antibodies against embryonic (g), neonatal (h), and slow (j) MHC isoforms. *Fibers positive for staining with antibodies against SK3 channel (e), slow (j), and fast IIa (k) MHC isoforms. **Fibers positive for immunostaining with antibodies against SK3 channel (e) and fast IIa (k) MHC isoforms but negative for immunostaining with antibodies against embryonic (g), neonatal (h), and slow (j) MHC isoforms. In 2-mo denervated EDL muscle, antibodies against SK3 channel stain sarcolemma of the muscle fibers (red staining in b, e, and l) as well as plasma membrane of some of the satellite cells (3 small arrows in l) but not myonuclei (2 small arrows in l). Staining with DAPI was used for visualization of myonuclei and satellite cell nuclei (blue staining in l). In 2-mo stimulated-denervated EDL muscles, antibodies against SK3 channel brightly stained only sarcolemma of the small-size muscle fibers (small and large arrows in c and f), some of which were positive (large arrows in i) and some negative (small arrows in i) for embryonic MHC immunostaining (green staining in i). a–c: bar = 200 µm. d–k: bar = 100 µm. l: bar = 50 µm. B: representative Western blotting and histogram of relative levels of SK3 channel expression in control (C), 2-mo denervated (D), and 2-mo stimulated-denervated (S) EDL muscles of rats (% from the expression levels in innervated control EDL muscles). Data are expressed as means ± SE; n = 3. *Significant difference compared with the expression levels in control muscles at p(t) 0.01. #Significant difference compared with the expression levels in stimulated-denervated EDL muscles at p(t) 0.01.
|
|
The mRNA expression of a large number of ribosomal proteins was upregulated approximately twofold after muscle denervation (Supplemental Table S1). The amount of total and phosphorylated S6 ribosomal protein was increased 2- and 3.5-fold, respectively, in denervated compared with control EDL muscles (Fig. 9). In stimulated-denervated muscles, levels of both total and phosphorylated S6 protein were not different from the innervated control levels (Fig. 9).
|
DISCUSSION
|
|---|
Denervated muscles undergo a rapid decline in mass and in the development of force due to the decrease in the myofibrillar content of muscle fibers (16, 37). In addition, data presented in this paper and by others (5, 23) demonstrate that denervation of fast EDL muscles of rats leads to changes in fiber type composition and an increase in the proportion of oxidative compared with glycolytic fibers. Rapid declines in the mass and force of denervated muscles occur despite a denervation-induced activation of satellite cells and generation of some new muscle fibers (10). As reported by us (16) and others (8), degenerative changes are greater for fast than for slow fibers. The diverse processes associated with degeneration in fibers lead to changes in the mRNA expression of a large number of genes (62). On the basis of the PubMed database search for denervated skeletal muscles, 21 genes detected in the current study had been reported previously to be altered by denervation (Table 4). The other 107 genes identified in the present study had not been associated with denervation previously. Of the genes linked previously with denervation, the upregulation of MyoD, myogenin, MRF4, embryonic isoform of myosin heavy chain, and N-CAM was expected. These genes are intimately involved with the activation of satellite cells and regenerative processes that occur in denervated skeletal muscles (10). Denervation-induced upregulation of the mRNA expression of a number of genes, including biglycan, osteonectin, and Selenoprotein P, was unexpected, although both the unloading (86) and regeneration (38) of skeletal muscles upregulate the expression of these genes.
As reported previously (24), the electrical stimulation of denervated muscles preserves muscle mass and maximal force at levels not different from values in innervated control muscles. Although electrical stimulation decreased substantially the mRNA expression of the genes upregulated by denervation, for several genes the expression remained higher than in the control muscles. A small number of atrophic denervated muscle fibers were observed in cross sections of the stimulated-denervated EDL muscles, mostly at the periphery in close proximity to the epimysium. Similar to fibers in denervated EDL muscles, atrophic fibers in stimulated-denervated muscles expressed high levels of denervation-induced proteins (myogenin, N-CAM, SK3, embryonic myosin). The presence of the small atrophic fibers may be responsible for slightly higher levels of gene expression in stimulated-denervated than in control EDL muscles.
In the rat, 96% of the fibers in the soleus muscles are slow type, whereas 95% of the fibers in the EDL muscles are fast (84). After denervation of both soleus and EDL muscles, an increase occurs in the number of fast IIa fibers (5), although denervated soleus still has >70% slow fibers and denervated EDL >70% fast fibers. Despite the significant differences in control and denervation-induced fiber type composition of soleus and EDL muscles, the current study of mRNA expression in EDL muscles after denervation displays a number of similarities with changes in gene expression in soleus muscles in response to hindlimb unloading (7, 86).
Expression of transcription factors.
After denervation, an increase in the expression of the members of myogenic regulatory factors (MyoD, myogenin, MRF4) is well characterized (1, 13, 58). Expression of MyoD, myogenin, and MRF4 has also been studied after several months of denervation of skeletal muscles of rats (1). Expression of MyoD (22), myogenin (22, 58), and MRF4 (95) proteins was localized to both activated satellite cells and the myonuclei of denervated muscle fibers. After denervation, the increased mRNA and protein expression of myogenic regulatory factors detected in the current study is in good agreement with these previous results.
Of all the transcriptional factors identified by the current microarray study of muscle denervation, AML1 showed the highest level of upregulation. AML1 is a transcription factor that, in cooperation with Ets-1, binds DNA as part of T cell receptor-
(TCR) enhanceosome (50). Gattenlohner et al. (36) showed that AML1 was bound to the promoter region of N-CAM, and both AML1 and N-CAM mRNA expression were upregulated after ischemia of the heart. In denervated EDL muscles, the increase in N-CAM mRNA and protein expression detected in the current study indicates a possible role of AML1 in the upregulation of N-CAM transcription. A 50- to 100-fold upregulation of AML1 mRNA expression in skeletal muscles after denervation was reported by others (99). In addition, recent microarray studies found the upregulation of AML1 in skeletal muscles of patients with Duchenne muscular dystrophy (DMD) (43) and in muscles of rats after acute exercise (18). Apparently, AML1 plays an important role in the adaptations of striated muscles to changes in physiological conditions, although the mechanism of its action is not clear.
The upregulation of Kruppel-associated box zinc finger-1 (Kzf1) transcription factor after muscle denervation has not been reported previously. Nonetheless, a Kruppel-like factor-15 is highly expressed in myocytes in vivo where it interacts specifically with myocyte enhancer factor-2A (MEF2A) and induces GLUT4 expression (42). Kzf1 expression in skeletal muscles is most likely associated with activation of satellite cells and may have regulatory mechanisms similar to those of Kruppel-like factor-15. Interestingly, expression of the closely related Kzf2 gene after EDL muscle denervation was unchanged (Gene Expression Omnibus database, accession no. GSE1741).
Transcription factor Ngfi-A binding protein-1 (NAB1), which showed increased expression in EDL muscles after denervation, has not been studied previously in skeletal muscles. NAB1 is a ubiquitously expressed transcriptional corepressor that is able to block Egr-1 mediated transcription (89). Egr-1 is a zinc finger-containing transcriptional activator involved in the regulation of myoblast growth and differentiation. Egr-1 activates transcription of inhibitor of differentiation (Id) gene (31). In proliferating myoblasts, differentiation is inhibited due to the high levels of expression of Egr-1 and Id proteins. In differentiating myoblasts, expression of Egr-1 and Id proteins is decreased, allowing the upregulation of mRNA expression of myogenic regulatory factors and, subsequently, muscle-specific genes. Consequently, increased NAB1 mRNA expression detected in the present study may regulate the expression of myogenic regulatory factors (MyoD, myogenin, MRF4) after muscle denervation by suppressing the inhibitory effect of Egr-1 and Id proteins.
Expression of structural, extracellular, and cell adhesion proteins.
The denervation-induced upregulation of mRNA and protein expression of the embryonic myosin heavy chain and N-CAM detected in the present study is consistent with data reported by Schiaffino et al. (82) and Muller-Felber et al. (68). The upregulation in the expression of embryonic myosin heavy chain was also reported in skeletal muscles of patients with DMD and -sarcoglycan deficiencies (17), as well as in muscles of mice after acute ischemia (72). Expression of the cardiac isoform of troponin T, detected in our study, was upregulated in skeletal muscles after both denervation and cold injury (80), and in muscles of DMD patients (6) and mdx mice (91). Troponin T was upregulated during myoblast differentiation (67, 80). The upregulation of cardiac troponin T expression was linked to the activation of satellite cells during skeletal muscle regeneration (6). Similarly, the upregulation of biglycan after denervation of EDL muscles most likely is related to the activation of the satellite cells and regenerative processes. Biglycan, a small leucine-rich proteoglycan located in the extracellular matrix, interacts with -dystroglycan, one of the components of the dystrophin-glycoprotein complex (11). Increased expression of ubiquitous actin-binding protein tropomyosin-3 (TPM3) in EDL muscles after denervation might be related to the stabilization of muscle structure. Tropomyosins are localized to the thin filament of the sarcomeres, and their function includes stabilization of thin filaments (76). There are three striated muscle-encoded isoforms: -TPM, ß-TPM, and TPM3. Switching from -TPM to TPM3 is associated with a decrease in myofilament Ca2+ sensitivity and an attenuation of length-dependent activation (76). TPM3 isoform is predominantly found in the slow-twitch muscles (75), and the upregulation of TPM3 mRNA expression correlates with the increase in the number of slow fibers in denervated EDL muscles seen in the present study.
The increased expression of ninjurin, neurofilament M, and NrCAM detected by the current study is most likely related to the Schwann cell remodeling after skeletal muscle denervation. Ninjurin (nerve injury-induced protein) is upregulated after axotomy in neurons and in Schwann cells surrounding the distal nerve segment (4). NrCAM is an Ig superfamily transmembrane protein expressed in Schwann cells and localized specifically at the node of Ranvier (20). Neurofilament M is upregulated in developing Schwann cells when they acquire myelin-forming phenotype (54).
Protein synthesis and degradation.
Activation of protein degradation pathways after denervation leads to muscle atrophy (37, 62). The present study identified upregulation in the expression of five enzymes involved in protein degradation. Upregulation of calpains found in the present study is critical for the disassembling of the sarcomeric structures in muscles after denervation (47, 51). Calpains are a family of calcium-activated proteases that are ubiquitously expressed and cleave a large variety of substrates (19). Calpain-2 is expressed in proliferating myoblasts and is further upregulated during myoblast differentiation in vitro (3, 66). Increased expression of calpains after skeletal muscle denervation was previously reported (28, 59). Expression of calpains was also upregulated in muscles of rats after unloading (87). The twofold upregulation in the expression of matrix metalloproteinase-12 in denervated EDL muscles reflects its important role in the remodeling of the extracellular matrix. Previously, the upregulation of metalloproteinases has been reported in muscles in response to unloading (96) and during regeneration (38).
Coordinate upregulation in the expression of two serine proteases, tonin and kallikrein, after denervation of EDL muscles may have considerable physiological significance. Angiotensin II is necessary for an optimal response of skeletal muscles to overload, and inhibiting angiotensin II attenuates muscle hypertrophy (41). Tonin releases angiotensin II directly from angiotensinogen and angiotensin. Tissue kallikrein prevents diabetic microangiopathy in rats by attenuating endothelial cell apoptosis and promoting capillarization (29). Consequently, kallikrein likely participates in the maintenance of the level of blood flow in denervated muscles, facilitating the supply of nutrients and oxygen to the muscle fibers.
The ubiquitin-proteasome pathway plays a critical role in muscle atrophy (9, 40, 62, 86). Consequently, after denervation of the EDL muscles, the lack of any change in the mRNA levels of the ubiquitin-conjugated enzymes UBC7 and UBC8A and in the S1 subunit of the 26S proteasome complex was surprising. The limited number of genes of the ubiquitin-proteasome pathway that were present on the Clontech arrays used in the current study is one of the possible explanations of this inconsistency. Key players in the ubiquitin-proteasome pathway were not within the scope of the current study.
The traditional viewpoint is that, during skeletal muscle atrophy, protein synthesis is decreased (92). Consequently, after denervation, the upregulation in the mRNA expression of a large number of genes involved in protein synthesis, including all of the ribosomal subunits present in the arrays, as well as an increase of protein level and phosphorylation (activation) status of S6 ribosomal protein was unexpected. Nevertheless, Goldspink (39) demonstrated that after denervation of skeletal muscles, protein synthesis showed an initial decrease (1–2 days postdenervation) followed by a subsequent increase (7–10 days postdenervation). After denervation of the diaphragm, an increased phosphorylation level of S6 ribosomal protein was reported previously (70). Several recent studies also showed an upregulation of a number of ribosomal proteins and initiation and elongation factors in skeletal muscles undergoing atrophy due to immobilization and unloading (7, 85, 86, 96) as well as fasting, uremia, diabetes, and cachexia (62). The current microarray study is in good agreement with previous publications (16, 55, 56) in the detection of increased expression of the "developmental" isoform (eEF1A-1) of elongation factor-1A (eEF1A) in EDL muscles after denervation. The level of protein synthesis was not measured directly, but the upregulation in the mRNA expression of a large number of ribosomal proteins, as well as an increase of protein level and phosphorylation status of S6 ribosomal protein, suggested an enhanced protein synthesis in denervated EDL muscles. Nevertheless, increased protein synthesis does not compensate for the upregulated protein degradation that leads to the progressive atrophy of denervated skeletal muscles.
Metabolism.
Denervation induces dramatic changes in muscle metabolism and, consequently, changes in the expression of a large number of metabolic enzymes, such as stearoyl-CoA desaturase-1 and -2, which were both upregulated in denervated EDL muscles. Stearoyl-CoA desaturase is a key rate-limiting enzyme involved in the metabolism of unsaturated fatty acids. Stearoyl-CoA desaturase is expressed in rabbit slow muscles but is absent in fast muscles (81). The upregulated expression of stearoyl-CoA desaturases after denervation of EDL muscles correlates well with the increased accumulation of slow myofibers found in our study and reported by others (5, 8, 23, 69).
Calsequestrin-2 was another enzyme regulated by the increase in the proportion of slow fibers in EDL muscles after denervation, as observed in our study. Calsequestrin-2, a calcium-sequestering protein, was upregulated in muscles of DMD patients (6, 43) and mdx mice (91). Activation of this gene is related to changes from fast- to slow-twitch phenotype during skeletal muscle regeneration and dystrophy (6).
The increase in the expression of the -type of glutathione S-transferase (GST) and Selenoprotein P suggests denervation-induced activation of an antioxidant response in EDL muscles. The GSTs represent a major group of detoxification enzymes. GSTs are regulated in vivo by reactive oxygen species and a broad spectrum of toxic chemicals (44). Selenoprotein P is involved in antioxidant defense mechanisms. Selenoprotein P mRNA expression is upregulated in skeletal muscles after acute ischemia (72) and unloading (86) and in muscles of mdx mice (90). The upregulated expression of GST and Selenoprotein P in EDL muscles after denervation may represent a compensatory mechanism in response to denervation-induced oxidative stress.
The data on upregulated mRNA expression of methionine adenosyltransferase II in denervated EDL muscles correlate with findings reported by Hopkins and Manchester (46) on elevated activity of methionine adenosyltransferase and increased concentration of S-adenosylmethionine in diaphragm muscles after unilateral nerve section. Methionine adenosyltransferase catalyzes biosynthesis of S-adenosylmethionine from methionine and ATP. S-adenosylmethionine is critical for DNA methylation and regulation of gene expression (33). In addition, incubation of cardiac sarcoplasmic reticulum in the presence of S-adenosylmethionine increased Ca2+-stimulated ATPase activity (35). Upregulation of methionine adenosyltransferase expression is likely related to regulation of gene expression and ATPase activity in denervated EDL muscles.
Signaling cascades.
As a response to skeletal muscle denervation, the regulation in the expression of a large number of genes involved in signal transduction pathways indicated complex changes within the signaling cascades. Among them was CaM-kinase kinase-ß (CaMKKß). CaM-kinases are involved in mediation of contractile activity-dependent gene regulation and mitochondrial biogenesis (97). CaMKKß phosphorylates and activates CaM-kinases I and IV. CaMKKß is broadly expressed in rat tissues, and its activity is enhanced by elevation of intracellular Ca2+ concentration (2). During myoblast differentiation, activity of CaM-kinase IV was increased and myogenin expression was promoted through CaM-kinase IV-mediated activation of MEF2 transcription factor (98). The upregulation of CaMKKß mRNA expression might be required for maximal activation of CaM-kinases and upregulation of gene expression (myogenin) in skeletal muscles after denervation.
Before this study, no data were available on regulation of mRNA expression of protein kinase C receptor (RACK1) in skeletal muscles. RACK1 receptor binds to activated members of the protein kinase C and Src families and functions as a scaffolding, anchor, and adaptor protein in multiple intracellular signal transduction pathways. After skeletal muscle denervation, PKC levels increase 80% compared with control levels (45). In cardiac muscles, enhanced RACK1 expression and PKC
-RACK1 interactions are critical for coordinated PKC-mediated hypertrophic signaling (74). Increased RACK1 expression after denervation is probably involved in the enhancement of PKC-mediated signaling cascades.
Calcium-independent receptor of -latrotoxin (CIRL-2) expression after denervation of skeletal muscles has not been observed previously. CIRLs are G protein-coupled receptors that also contain domains characteristic of cell adhesion proteins (48). Roles of CIRLs and their natural ligands in vivo are unknown. CIRL-2 is ubiquitously expressed with high concentrations found in placenta, kidney, spleen, ovary, heart, and lung and low concentration in innervated skeletal muscles (48). The functional significance of this CIRL-2 upregulation in EDL muscles after denervation is unknown.
Mitochondrial proteins.
The upregulation in mRNA levels of four mitochondrial genes in EDL muscles after denervation, detected by our array study, indicated increased mitochondrial biogenesis and/or modification in mitochondrial structure/function relationships. In agreement with these data, histochemical staining evaluating SDH and COX activities showed higher intensity of enzymatic activity and changes in mitochondrial distribution in denervated compared with control EDL muscles. Among upregulated genes were two liver isoform subunits of COX: subunit VIa and subunit VIIIa. Muscle isoforms of subunits VI and VIII are expressed exclusively in heart and skeletal muscles, while liver isoforms are broadly expressed in different tissues and have very weak expression in adult skeletal muscles (83, 71). Subunits VI and VIII are nuclear encoded and developmentally regulated. A switch from liver to muscle isoform expression of COX VIa has been demonstrated during in vivo and in vitro muscle development (73). An increase in the expression of liver isoforms of subunits VI and VIII was previously reported during skeletal muscle regeneration and denervation (71). This study detected increased mRNA expression of COX VIIIa muscle isoform in denervated EDL muscles and downregulation of both isoforms, along with a decrease in the intensity of SDH enzymatic activity, after electrical stimulation of the denervated muscles. The upregulation in expression of liver isoforms of COX VIa and VIIIa subunits in denervated skeletal muscles might be the result of an adaptation to the constant high-energy demands due to the proliferation of satellite cells and an increase in regenerative and in protein synthesis/degradation processes.
Increased expression of outer mitochondrial membrane receptor rTOM20, a major component of the mitochondrial protein import complex, further supports increased mitochondrial biogenesis in denervated skeletal muscles. Most of the mitochondrial proteins are synthesized in the cytosol as precursor proteins and then are imported into the mitochondria through the protein import complex (30). This process has a critical role in transporting nuclear-coded COX subunits (including COX VI and COX VIII) through the mitochondrial membrane. rTom20 protein content increased in skeletal muscle mitochondria after chronic contractile activity (88).
Most proteins destined for the mitochondrial matrix are synthesized initially as larger precursor polypeptides carrying NH2-terminal extensions that are cleaved off by specific processing peptidases once each precursor has reached its final destination within the mitochondrion (34). Mitochondrial processing peptidase (MPP) removes the matrix-targeting NH2-terminal extensions from most mitochondrial protein precursors (50). The twofold increase in the expression of MMPß in denervated EDL muscles is consistent with the data on increased rTom20 expression and supports the increased mitochondrial protein import and processing in skeletal muscles after denervation.
Ion channels.
The threefold upregulation of mRNA and sixfold upregulation of protein expression of SK3 potassium channel in rat EDL muscles after denervation are in good agreement with the data of Pribnow et al. (77). Accumulation of high levels of SK3 mRNA and protein expression starts at 2–3 days (77) and, according to the present study, continues for at least 2 mo after denervation. Our observation, that both slow and fast fibers in denervated muscles expressed SK3, correlates well with the data of Kimura et al. (57), which showed slow and fast SK3-positive fibers in muscles of patients with myotonic dystrophy. In control skeletal muscles, SK3 immunoreactivity is present in presynaptic compartments of the mature neuromuscular junctions (79). In denervated muscle fibers, SK3 is localized in the extrajunctional as well as the junctional plasma membrane of the muscle fibers (79). SK3 appears to be involved in the hyperexcitability (spontaneous fibrillatory activity) and hyperpolarization observed in denervated muscles. In addition to the presence in denervated muscles, SK3 channels are present in most neurons and in several nonexcitable cell types (77). Our data indicate that some of the muscle satellite cells also express SK3 on their plasma membrane.
Other genes.
Twenty genes did not fit into any of the groups described above or had no known functions. The upregulation in the clathrin heavy chain mRNA expression in denervated EDL muscles, as seen in our array results, correlates with increased clathrin protein levels in denervated soleus muscles (60). Clathrin-coated vesicles are involved in receptor-mediated intracellular transport pathways related to lysosomal proteolysis. During muscle development, clathrin first appears in the sarcomeric structures after myoblast fusion (53). The upregulation in clathrin heavy chain expression after muscle denervation may be related to the activation of lysosomal protein degradation pathways and muscle atrophy.
Correlation between changes in gene expression and physiological adaptation of skeletal muscles to denervation and electrical stimulation.
The present study identified 128 genes that modified mRNA expression in response to the denervation. The genes that were identified belonged to several functionally distinct groups. The complexity of the changes in gene expression patterns reflected the diversity of the physiological processes activated by the denervation of the skeletal muscles. In addition to preventing the atrophy of muscle fibers, electrical stimulation decreased substantially the changes in mRNA expression. In stimulated-denervated EDL muscles, the expression of myogenin, N-CAM, SK3 channel, and embryonic myosin was restricted to muscle fibers with small cross-sectional areas that were probably not stimulated adequately. The majority of fibers in stimulated-denervated muscles were similar to the fibers in control muscles in both fiber cross-sectional area and mRNA expression pattern. Our data and data published previously (5) indicated that denervation of EDL muscles increased the proportion of slow and fast IIa (oxidative) muscle fibers, while the fraction of fast IIb (glycolytic) fibers decreased. The increase in the mRNA expression of a number of genes, normally found in slow/oxidative muscles, correlated with the denervation-induced changes observed in fiber type composition. The genes with increased expression included the following: TPM3, stearoyl-CoA, and calsequestrin-2. Upregulation of genes involved in mitochondrial biogenesis (CaMKKß, COX VIa, COX VIIIa, rTom20, MMPß) was related to the increase in the proportion of oxidative fibers. Activation of satellite cells and new muscle fiber formation were linked to the increased mRNA levels of myogenic transcription factors (myogenin, MyoD, myogenic factor-4), embryonic myosin, N-CAM, and elongation factor eEF1A-1. Denervation-induced increases in protein degradation and upregulation of protein synthesis were associated with the upregulated mRNA expression of calpain-2, matrix metalloproteinase-12, and a number of ribosomal proteins. On the basis of the observations made in the present study and on data published by others, a schematic representation of our current view of the complex processes that take place in EDL muscles in response to denervation and stimulation of denervated muscles is summarized in Fig. 10.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 10. Schematic representation of changes in gene expression in 2-mo denervated and 2-mo stimulated-denervated compared with control EDL muscles of rats. +++Large increase. +Small increase. N/C, no change. – – –Large decrease in expression relative to the innervated control levels.
|
|
|
GRANTS
|
|---|
This study was supported by the Contractility Core and Gene Expression Profiling Core of the Nathan Shock Center [National Institute on Aging (NIA) Grant P30-AG-13283]. T. Y. Kostrominova and D. E. Dow were supported by fellowships from NIA (T32-AG00116).
|
ACKNOWLEDGMENTS
|
|---|
We thank Cheryl Hassett for assistance with surgical procedures associated with this study.
Present address for R. G. Dennis: Dept. of Biomedical Engineering, Univ. of North Carolina, 2350 Hayward, Chapel Hill, NC 27514-7575.
|
FOOTNOTES
|
|---|
Address for reprint requests and other correspondence: T. Y. Kostrominova, Institute of Gerontology, Univ. of Michigan, 300 N. Ingalls St., Ann Arbor, MI 48109-2007 (e-mail: kostromi{at}umich.edu)
10.1152/physiolgenomics.00210.2004
1 The Supplemental Material for this article (Supplemental Table S1) is available online at http://physiolgenomics.physiology.org/cgi/content/full/00210.2004/DC1. 
|
REFERENCES
|
|---|
- Adams L, Carlson BM, Henderson L, and Goldman D. Adaptation of nicotinic acetylcholine receptor, myogenin, and MRF4 gene expression to long-term muscle denervation. J Cell Biol 131: 1341–1319, 1995.[Abstract/Free Full Text]
- Anderson KA, Means RL, Huang QH, Kemp BE, Goldstein EG, Selbert MA, Edelman AM, Fremeau RT, and Means AR. Components of a calmodulin-dependent protein kinase cascade. Molecular cloning, functional characterization, and cellular localization of Ca2+/calmodulin-dependent protein kinase kinase ß. J Biol Chem 273: 31880–31889, 1998.[Abstract/Free Full Text]
- Aragon B, Poussard S, Dulong S, Touyarot K, Dargelos E, Brustis JJ, Levieux D, Ducastaing A, and Cottin P. Protein kinase C is a calpain target in cultured embryonic muscle cells. Mol Cell Biochem 231: 97–106, 2002.[CrossRef][Medline]
TOP
ABSTRACT
INTRODUCTION
