Gene expression changes after complete spinal cord transection
In female rats at 17 weeks after SCI there were was enrichment in genes involved in angiogenesis and blood vessel morphogenesis, protein folding and Wnt signaling. Alterations in genes involved in the formation of blood vessels might be expected given the decrease in skeletal muscle mass and presumed remodeling of capillary beds that must also occur. Blood flow to the lower extremities is reduced after SCI reflecting diminished skeletal muscle mass . Capillary density of the tibialis anterior after SCI is similar to or somewhat less than that for able-bodied controls , indicating that after SCI substantial remodeling of capillary networks must have occured during atrophy after SCI. The decrease in expression of genes supporting normal protein folding might also be explained as a reflection of an atrophying muscle that requires less de-novo synthesis of proteins.
More surprising was the enrichment in genes for Wnt signaling, particularly the marked upregulation of the Wnt inhibitors sFRP1, sFRP2 and sFRP4. The elevations in expression of Wnt inhibitors must be interpreted together with changes of Pitx2, a Wnt responsive gene . That Pitx2 expression was unaltered at 17 weeks after SCI may indicate that Wnt signaling overall was unchanged by SCI. One must also consider that Wnts function primarily as autocrine and paracrine regulators and that inhibition of Wnt signaling in specific microenvironments, such as the satellite cell niche, may not be reflected by Pitx2 levels in whole muscle. Although decreased Wnt signaling has not been shown to accelerate disuse atrophy of muscle, one might suggest that diminished Wnt signaling is detrimental to maintaining muscle mass, because Wnt signaling has been implicated in muscle hypertrophy. Specifically, ß-catenin, through which canonical Wnt signaling regulates gene expression, has been found in a transgenic mouse model to be necessary for overload-induced muscle hypertrophy [18, 19]. Findings that signaling by Wnt7a through Fzd7 activates Akt/mTOR signaling  suggests there is also ß-catenin independent Wnt signaling that underlies the anabolic activity of Wnts. Wnt signaling is also critical to fate commitment and differentiation of myogenic precursors, and essential for the repair of injured skeletal muscle in adulthood [31, 32], and one may view the hypertrophic response of atrophied muscle as a reparative process following SCI.
Calcineurin activity has been shown to be exquisitely sensitive to neuromuscular activity [20, 33]. As such, it was not surprising that levels of the NFAT-sensitive transcript RCAN1.4 were greatly reduced after SCI. An unexpected finding was that expression of the calcineurin inhibitors RCAN1 and RCAN2 was also diminished after SCI. One interpretation of this result is that expression of these genes is regulated in part by calcineurin in an inhibitory feedback loop, such that expression levels of RCAN1 and RCAN2 fall when calcineurin activity is low. In contrast to these findings in muscle after SCI, RCAN2 expression rises by 35 days after sciatic nerve transection in gastrocnemius muscle . Why RCAN2 expression differs in these two model systems is not clear, although gender may be one factor. Another may be a difference in spontaneous action potentials in motor neurons, which are electrically silent after nerve transection, whereas there is some evidence that motor neurons are periodically activated after SCI, presumably through spinal reflex arcs .
PGC-1α mRNA was not reduced at 17 weeks after SCI, while PGC-1ß was reduced; marked reductions in mRNA levels for MyH7, Tnn C-slow and Tnn I-slow were also observed. Consistent with these findings, in a previous report examining changes in gastrocnemius muscle at 56 days after SCI, decreases were observed for PGC-1ß protein, and for expression of mRNA and proteins for slow myofibrillar components . However, in contrast to the current study, mRNA levels for PGC-1α were reduced in gastrocnemius muscle by greater than 50% at 56 days after SCI in male rats associated with a similar reduction in PGC-1α protein in whole muscle and nearly complete loss of PGC-1α from nuclear fractions obtained by subcellular fractionation . Differences in PGC-1α mRNA levels between these two studies may relate to differences in muscle studied, because levels of PGC-1α are elevated in slow twitch muscle such as soleus compared to fast twitch muscles . Reasons for the differences in effects of SCI on PGC-1α levels after SCI between males and females could also be gender-related. In male rats, administration of a replacement dose of testosterone combined with nandrolone markedly increased PGC-1α mRNA and protein in muscle of SCI rats , indicating that PGC-1α is an androgen responsive gene.
Effects of ES on gene expression in soleus muscle
The current study showed that over the first 7 days after initiating ES in a female rat model of SCI, alterations in soleus gene expression were notable for genes were involved in the Notch and Wnt signaling pathways, neuromuscular junctions, vascular remodeling, and calcium transport, and for a lack of change in PGC-1α or of slow myofibrillar proteins. Several studies of the effects of ES on gene expression of individuals with SCI have been conducted. Twenty four hours after resistance exercise in humans with SCI, increased levels for IGFBP4, MyoD, myogenin, p21-Waf1, and IGF-1 were observed [37, 38]. In separate studies of individuals with SCI, prolonged periods of exercise training using ES promoted gains in skeletal muscle mass and strength that were associated with increased expression of citrate synthase, GLUT1 and GLUT4 ; improved insulin sensitivity was reported in one study  and a trend toward improvements in insulin responsiveness was found in a second .
The finding in the present study that ES did not alter expression of slow myofibrillar proteins in soleus was unexpected because in normally innervated muscle, neuromuscular activity rapidly increases expression levels of genes for mitochondrial proteins , and because in humans, long-term studies of ES have shown increased expression of slow-twitch fiber genes and genes encoding proteins involved in metabolism of carbohydrates and lipids to generate ATP . Expression of slow fiber genes is upregulated by calcineurin/NFAT signaling and PGC-1α  but neither PGC-1α expression nor calcineurin/NFAT signaling assessed by RCAN1.4 levels appeared to have changed after ES, suggesting one explanation for the lack of effect of ES on such genes. Our data do not permit one to exclude the possibility that RCAN1.4 mRNA has a short half-life, which could preclude detection of transient rises in levels of this transcript at the time tissues were harvested, which was approximately 24 hours after the last session of ES. Evidence of a delay in fiber type changes after initiating ES has also been reported in humans in whom 4 weeks of ES did not alter fiber types in the tibialis anterior muscle . Why upregulation of slow myofibrillar genes is delayed when muscle is exercised after SCI by ES is unclear. One possibility is that during the early period after initiating ES, ES injured muscle fibers, as discussed further below. Such injury might limit the ability of muscle fibers to respond appropriately to increased neuromuscular activity. It is also possible that one or more factors necessary for full activity of PGC-1α and/or calcineurin/NFAT in upregulating such genes is absent early after initiating ES, or not fully activated by 7 days of ES.
Recent observations that activation of mTOR, a requisite step for muscle hypertrophy, may not involve growth factor/PI3 kinase/Akt signaling [41–43] has led to the exciting question of what other signals might stimulate overload-induced hypertrophy. In this regard, a novel aspect of the above findings was that in the soleus muscle of rats with SCI, expression of both Hey1 and Pitx2 was elevated after 7 days of ES, suggesting that there were increases in both Notch and Wnt/ß-catenin signaling. Signaling through Notch and Wnt is necessary for muscle repair after injury [31, 32], and has been implicated in signaling for muscle hypertrophy in some experimental models but not others. Increased expression of Notch has been implicated in muscle hypertrophy resulting from testosterone in mice and humans [16, 17]. Expression of a Notch-responsive gene, Hey2, was found to be increased in female rats at 4 and 24 hours after a bout of resistance exercise but was unaltered in male rats . As noted above, activation of ß-catenin and upregulation of Wnt signaling genes has been found to be associated with overload-induced muscle hypertrophy [18, 28], although specifically how Wnt signals to stimulate hypertrophy remains unknown.
A quite unexpected effect of ES was the upregulation of genes encoding proteins that form the neuromuscular junction, specifically subunits of the nicotinic acetylcholine receptor (e.g., Chrnd). One interpretation of this finding is that by 17 weeks after SCI, an insufficient number or density of such receptors is present at motor endplates to support normal neuromuscular transmission, the failure of which is one potential mechanism for reduced contractile response to neural inputs. Skeletal muscle membrane potential has been reported to be reduced over the first 30 days after spinal cord transection in rats, associated with diffusion of nicotinic acetylcholine receptors away from motor endplates . Alternative explanations should certainly be considered and may include aborization of axons and subsequent formation of additional neuromuscular junctions.
Influx of calcium across the cytoplasmic membrane and release of calcium from the sarcoplasmic reticulum and subsequent reuptake of calcium are critical to excitation-contraction coupling. Failure of excitation-contraction coupling is believed to be an important mechanism underlying the easy fatigability observed after SCI . Information as to how calcium storage and transport is altered in muscle after SCI is limited. In one report, SCI was found to induce a mismatch between the slow and fast isoforms of the calcium transporter SERCA and fiber type . It is thus notable that calcium transport was also an enriched GeneGo theme over time in soleus muscle after ES. Among genes related to calcium binding and transport were those of calsequestin 2 and S100a9. Calsequestin is thought to be the major protein responsible for storing calcium within the lumen of the sarcoplasmic reticulum [47, 48], and release of calcium from calsequestin results in increased cytoplasmic calcium levels, thereby initiating contraction. Two isoforms of calsequestin have been identified, with calsequestin 2 found in slow-twitch muscle and heart. Upregulation of calsequestin 2 in response to ES might improve the capacity of muscle for excitation-contraction coupling by increasing the reservoir of calcium binding sites within the SR.
Comparison of gene expression changes in the soleus muscle after ES or GA
The comparison of gene expression changes elicited in soleus muscle by ES and GA revealed a small number of genes regulated by both forms of exercise, but overall showed very different profiles, and marked differences in the functional categories for which differentially expressed genes were over-represented. Notably under-represented among the genes regulated by ES were those associated with slow-twitch fibers, metabolism of substrates in energy metabolism and oxidative phosphorylation. One might infer that the genes regulated by both ES and GA play critical roles in adaptation of skeletal muscle to resistance exercise, and that the overall poor overlap in genes differentially expressed after ES or GA reflects early differences in the ability of muscle to adapt to resistance exercise, or in the hitherto unexplored mechanisms by which such adaptation occurs in the two paradigms.
The absence of changes after ES in PGC-1α, slow-fiber genes, and many PGC-1α target genes of energy metabolism suggest impairments or delays in the response of paralyzed and atrophied muscle to neuromuscular activity elicited in our study by ES. When considering this interpretation of the data it should be noted that the isometric exercise achieved in our study by ES for 1 hour each day is of shorter duration than GA, which overloads soleus during every periods of activity; in addition, GA produces non-isometric overloading of the soleus. However, other reports have also found impaired exercise responses of such genes in skeletal muscle after SCI, supporting the view that differences observed between ES and GA were, at least in part, attributable to underlying properties of muscle rather than the nature of the exercise. Studies in man suggest that changes in expression of genes for slow-twitch fibers and energy metabolism do not occur at 4 weeks after initiating ES . Importantly, longer periods of exercise by ES stimulate increases in slow oxidative type I and fast oxidative type IIA fibers, increased capacity for oxidative phosphorylation, and increased contractile force associated with diminished fatigability [1–3]. Why there is a delay in upregulation of such genes occur is unknown.
Several reports suggest that during periods of disuse, the normal protective mechanisms that minimize tissue injury from the mechanical or metabolic demands of exercise are diminished or lost. A single bout of resistance exercise in persons with SCI produced greater muscle injury than a similar exercise in controls ; similarly, studies of muscle from rats ambulating after periods of muscle disuse due to spaceflight or hindlimb suspension have shown increased myofiber damage and inflammation . The identity of the genes involved in altering gene expression programs to protect skeletal muscle against the demands of exercise are not well understood but are likely to include PGC-1α and its downstream target genes. Loss of PGC-1α from skeletal muscle reduces exercise tolerance , and PGC-1α has been reported to favorably alter levels of antioxidants, and thus to reduce oxidant stress [51, 52].
Genes altered by both ES and GA included two with reported roles in muscle biology: Ankrd1 and ATF3. ATF3 is a transcription factor upregulated by cellular stress  with an uncertain role in adaptation of muscle to inactivity or exercise; it has been associated with pathological cardiac hypertrophy and dysfunction . Ankrd1 is one of 3 muscle ankyrin repeat domain proteins (MARPs) and appears to function in part at least by dissociating from its typical binding sites at the I band to undergo nuclear translocation and participate in transcriptional regulation thereby serving as a link between the contractile apparatus and nucleus . Ankrd1 does not appear to be essential to normal function, because mice lacking Ankrd1 appear to have a normal phenotype, but mice lacking all MARPs have subtle abnormalities in muscle contractile properties and stiffness . Whether Ankrd1 is needed for proper response to muscle overloading, or optimal recovery of muscle after periods of inactivity, is unknown. A role for Ankrd1 in muscle pathology is suggested by findings that dilated cardiomyopathies may be linked to Ankrd1 mutations [57, 58].