Search

Although cytokine-induced nuclear factor kappaB (NF-kappaB) pathways are involved in muscle wasting subsequent to disease, their potential role in disuse muscle atrophy has not been characterized. Seven days of hind limb unloading led to a 10-fold activation of an NF-kappaB-dependent reporter in rat soleus muscle but not the atrophy-resistant extensor digitorum longus muscle. Nuclear levels of p50 were markedly up-regulated, c-Rel was moderately up-regulated, Rel B was down-regulated, and p52 and p65 were unchanged in unloaded solei. The nuclear IkappaB protein Bcl-3 was increased. There was increased binding to an NF-kappaB consensus oligonucleotide, and this complex bound antibodies to p50, c-Rel, and Bcl-3 but not other NF-kappaB family members. Tumor necrosis factor alpha (TNF-alpha) and TNF receptor-associated factor 2 protein were moderately down-regulated. There was no difference in p38, c-Jun NH(2)-terminal kinase or Akt activity, nor were activator protein 1 or nuclear factor of activated T cell-dependent reporters activated. Thus, whereas several NF-kappaB family members are up-regulated, the prototypical markers of cytokine-induced activation of NF-kappaB seen with disease-related wasting are not evident during disuse atrophy. Levels of an anti-apoptotic NF-kappaB target, Bcl-2, were increased fourfold whereas proapoptotic proteins Bax and Bak decreased. The evidence presented here suggests that disuse muscle atrophy is associated with activation of an alternative NF-kappaB pathway that involves the activation of p50 but not p65.

Disuse atrophy of skeletal muscle leads to an upregulation of genes encoding sarcoplasmic reticulum (SR) calcium-handling proteins. Because many of the proteins that are induced with endoplasmic reticulum (ER) stress are ER calcium-handling proteins, we sought to determine whether soleus muscle atrophy was associated with a prototypical ER stress response. Seven days of rat hindlimb unloading did not alter expression of ubiquitous ER stress proteins such as Grp78, calreticulin, and CHOP/GADD-153, nor other proteins that have been shown to be activated by ER stressors such as vinculin, the type I D-myo-inositol 1,4,5-trisphosphate receptor, or protein kinase R, a eukaryotic initiation factor 2 alpha kinase. On the other hand, expression of heme oxygenase-1 (HO-1), an antioxidant ER stress protein, was significantly increased 2.2-fold. In addition, unloading led to an increase in calsequestrin, the muscle-specific SR calcium-binding protein, at both the mRNA (68%) and protein (24%) levels. Although disuse atrophy is associated with a significant remodeling of muscle-specific proteins controlling SR calcium flux, it is not characterized by a prototypical ER stress response. However, the upregulation of HO-1 may indicate ER adaptation to oxidative stress during muscle unloading.

The skeletal muscle sarco(endo)plasmic reticulum calcium ATPase (SERCA1) gene is transactivated as early as 2 days after the removal of weight-bearing (Peters, D. G., Mitchell-Felton, H., and Kandarian, S. C. (1999) Am. J. Physiol. 276, C1218-C1225), but the transcriptional mechanisms are elusive. Here, the rat SERCA1 5' flank and promoter region (-3636 to +172 base pairs) was comprehensively examined using in vivo somatic gene transfer into rat soleus muscles (n = 804) to identify region(s) that are both necessary and sufficient for sensitivity to weight-bearing. In all, 40 different SERCA1 reporter plasmids were constructed and tested. Several different regions of the SERCA1 5' flank were sufficient to confer a transcriptional response to 7 days of muscle unloading when placed upstream of a heterologous promoter. Two of these regions were analyzed further because they were necessary for the unloading response of -3636 to +172, as demonstrated using internal deletion constructs. Deletion analysis of these regions (-1373 to -1158 and -330 to +172) suggested that unloading responsiveness corresponded to CACC sites and E-boxes. Mutagenesis of cis-elements in the first region showed that a specific CACC box (-1262) was involved in SERCA1 transactivation and a nearby E-box (-1248) was also implicated. Constructs containing trimerized CACC sites and E-boxes showed that the presence of both elements is required to activate transcription. This is the first identification of specific cis-elements required for the regulation of a Ca(2+) handling gene by changes in muscle loading condition.

Direct injection of plasmid DNA into muscle allows the study of promoters in a physiological environment. Because of the variability of reporter gene activity, attempts have been made to normalize activity to muscle plasmid uptake by coinjection of a second "control" plasmid whose reporter gene is constitutively expressed by a viral promoter. The purpose of this study was to evaluate the use of a control plasmid vs. Southern blot to normalize for differences in uptake of plasmids containing promoter fragments of the skeletal muscle-specific sarco(endo)plasmic reticulum Ca(2+)-ATPase (SERCA1) gene. Results showed that the correlation of luciferase activity from control vs. SERCA1 plasmids is poor and that normalization by a virally driven control plasmid increased variability of SERCA1 luciferase activity. In several cases, the presence of a control plasmid inhibited SERCA1 reporter expression. When Southern blot analysis was used to normalize for differences in plasmid uptake there was less variability than with coinjection, and correlations between plasmid uptake and SERCA1 luciferase activity were better. Moreover, there were no inhibitory effects of a control plasmid allowing for optimization of injection conditions of the SERCA1 deletion constructs. The use of Southern analysis is suggested to determine whether plasmid uptake is differentially affected by physiological stimuli, muscle types, or plasmid sizes under study.

Previous work showed that protein and mRNA levels of the "fast" isoform of the sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA1) are markedly increased in unloaded slow-twitch soleus muscles, suggesting pretranslational control of gene expression [L. M. Schulte, J. Navarro, and S. C. Kandarian. Am. J. Physiol. 264 (Cell Physiol. 33): C1308-C1315, 1993]. However, because of the difficulty of measuring transcription rates from whole muscle, transcriptional activation of the SERCA1 gene with unloading has not been confirmed. Because SERCA1 pre-mRNA levels can reflect transcriptional activity, in the present study SERCA1 introns were sequenced to allow intron-directed RT-PCR measurement of SERCA1 pre-mRNA. These data were then compared with changes in SERCA1 mRNA expression in control and unloaded soleus muscles. After 2, 4, and 10 days of unloading, SERCA1 pre-mRNA and mRNA transcript levels increased significantly by two-, three-, and sevenfold, respectively (P < 0.01). Parallel increases in SERCA1 pre-mRNA and mRNA suggest transcriptional activation of the endogenous SERCA1 gene by muscle unloading. SERCA2, the cardiac/slow-twitch skeletal muscle isoform, was not markedly increased by unloading, and RNase protection assays showed no change in alternative splicing of SERCA1 or SERCA2 primary transcripts. With use of in vivo plasmid injection, the activity of a reporter gene driven by 3.6 kb of the SERCA1 5'-flanking region increased fivefold in 7-day-unloaded soleus muscles. Comparison of the magnitude of transcriptional activation of endogenous and constructed SERCA1 genes by unloading confirms the fidelity of using intronic RT-PCR to examine muscle gene transcription rates and suggests that cis-acting elements sufficient for regulating unloading-induced transcriptional activation are contained in this promoter construct.