Muscle growth and maturation depend on numerous factors, including uptake of glucose through the glucose transporter GLUT4 
. A role for GLUT4-mediated glucose uptake during muscle regeneration has also been proposed 
. Here we analyze components of the GLUT4 membrane traffic pathway and their expression in regenerating human muscle, as well as assess a role for the GLUT4 pathway in muscle regeneration in a transgenic mouse model. We present evidence that membrane traffic of GLUT4 is increasingly active during human and mouse muscle regeneration and show that in CHC22-transgenic mice with impaired GLUT4 membrane traffic in their skeletal muscle, regeneration is delayed. The myoblasts from CHC22-mice do not increase their proliferation in response to increased glucose, which promotes proliferation of myoblasts from WT mice, suggesting a connection between the impaired GLUT4 pathway in the CHC22-transgenic animals and their impaired muscle regeneration. The skeletal muscle of aged CHC22-mice also had an increased percentage of glycolytic fibers, as seen in some patients with type 2 diabetes 
. Together these findings support a role for GLUT4 function during muscle regeneration in humans and mice and in maintenance of fiber type, at least in mice. In addition, we show that markers of the GLUT4 pathway, including GLUT4, CHC22 and VAMP2 are diagnostic for regenerating human myofibers.
The GLUT4 transporter is sequestered in a GSC until it is released in response to insulin stimulation, so its membrane traffic is highly specialized. In humans, GSC formation in skeletal muscle and adipocytes involves the CHC22 isoform of clathrin that is missing from mice, defining a species-restricted aspect to GLUT4 membrane traffic 
. Following up the demonstration that CHC22 clathrin is a necessary component of the human GLUT4 pathway, we investigated whether the increased CHC22 levels, previously observed in regenerating rat muscle 
, indicated upregulation of the GLUT4 pathway during regeneration. Analysis of regenerating muscle fibers in patients with four different human myopathies revealed that levels of CHC22, GLUT4 and VAMP2 are elevated during regeneration. These proteins also have an altered internal distribution in regenerating fibers compared to mature fibers in the same tissue, which can be explained by repositioning of the nuclei and associated perinuclear regions in regenerating muscle 
. The elevation of GLUT4 pathway markers, combined with more visible staining of the GSC region, suggests that active membrane traffic of GLUT4 and regeneration of the GSC is occurring during human muscle regeneration. High levels of CHC22 and GLUT4 in Pax7-positive cells suggested that this pathway is amplified early in the regenerative process.
The mechanism by which the GLUT4 pathway components are stimulated is not clear. Earlier studies of rat muscle regeneration revealed an interaction between the regulatory GLUT4 enhancer sequence and transcription factors that are induced during muscle regeneration 
, but the regulatory sequences that affect CHC22 expression are not known. Here we investigated whether signaling via inflammatory cytokines that are present in regenerating muscle might influence expression of CHC22 and we observed no effect of TNF-α, IL-1β or IFN-γ on expression of CHC22 in cultured human myotubes. Although TNF-α was reported to influence GLUT4 expression in rat muscle 
, no obvious effect was observed in cultured human myoblasts and myotubes, nor did other cytokines affect GLUT4 expression in these cultures. Thus, we suggest that inflammation is not responsible for the observed increase in GLUT4 transport components in regenerating human muscle and propose that the increase results from some muscle-intrinsic program during regeneration. The time course of GLUT4 pathway expansion in regenerating mouse muscle coincides with the time course for innervation (data not shown), suggesting that the pathway might expand to meet the increased energy needs that accompany contraction.
We have previously demonstrated that transgenic mice expressing CHC22 in their skeletal muscle and fat, under the control of its human promoter, display clear defects in GLUT4 trafficking. In their skeletal muscle, presence of the CHC22 transgene causes excessive intracellular sequestration of GLUT4 and VAMP2 and a reduction of GLUT4 at the sarcolemma and T-tubules. These changes are accompanied by age-dependent hyperglycemia in the CHC22-mice 
. With their impairment of GLUT4 traffic and mild diabetic condition, the CHC22-mice qualified as a model system to test the role of the GLUT4 pathway in muscle regeneration. This approach was validated by showing that CHC22 expression is increased upon cardiotoxin injury of skeletal muscle in the CHC22-mice, and that GLUT4 and VAMP2 levels increase in both WT and CHC22-mice after injury, as predicted by our staining of human regenerating muscle. Consistent with previously observed trafficking defects in the GLUT4 pathway of the CHC22-mice 
, the kinetics of VAMP2 expression in their muscle was prolonged after injury compared to injured WT muscle. Importantly, analysis of their skeletal muscle after cardiotoxin injection revealed that the CHC22-mice have a delay in muscle regeneration, characterized by a delay in myofiber maturation. As CHC22, expressed in a tissue-specific fashion from its endogenous human promoter, is not expressed in motor neurons of the transgenic mice (data not shown), the defective regeneration is most likely due to the presence of CHC22 in the muscle fibers themselves and their associated defective GLUT4 traffic. Consistent with this interpretation, we observed that isolated myoblasts from the CHC22-mice were capable of fusing and forming myotubes, but did not proliferate in response to an increase in glucose, as observed for myoblasts from WT mice. Compared to WT myoblasts, the myoblasts from CHC22-mice also had higher fusion rates, typical of myoblasts under stress 
. We therefore propose that the delayed maturation of CHC22 myofibers during regeneration may be a consequence of the defective GLUT4 pathway in the transgenic mice.
Not all of the regenerating fibers in the CHC22-mice were diminished in size relative to the WT mice, so we compared the fiber types in skeletal muscle of CHC22-mice with WT mice. We found an increase in glycolytic fibers compared to oxidative fibers in aged CHC22-mice, compared to age-matched WT mice, which coincided with onset of hyperglycemia in the former 
. We also observed that in the injured muscle of CHC22-mice, the glycolytic fibers had decreased cross-sectional area compared to regenerating fibers in WT muscle. Since glycolytic fibers have lower levels of GLUT4 
and are more glucose-dependent, regeneration of the surviving glycolytic fibers could be more sensitive to perturbation of the GLUT4 pathway, accounting for their reduced size compared to regenerated oxidative fibers. Conversely, glycolytic fibers are more reliant on GLUT1 than GLUT4 for glucose import, while oxidative fibers are more dependent on GLUT4 
. Thus over time, oxidative fibers may be less viable due to the impairment of the insulin-responsive GLUT4 pathway that occurs in CHC22 mice, resulting in the alteration of fiber type composition we observed, which mimics that reported for type 2 diabetes. Hyperglycemia is characteristic of aged (>20 weeks old) CHC22-mice 
, so the reduction of oxidative fibers observed in aged CHC22-mice could also be a result of secondary effects of their hyperglycemia. On the other hand, the abnormality in muscle regeneration that we observed following muscle injury of 8-week-old mice is feasibly a direct cause of their impaired GLUT4 pathway rather than due to long term diabetic symptoms. Though we did show defective behavior of myoblasts from CHC22 mice in response to glucose, there may be additional features of the impaired GLUT4 pathway that influence muscle regeneration. For example, the GLUT4 pathway also mobilizes the multi-functional insulin-responsive amino peptidase (IRAP) 
. Interestingly, the canonical transient receptor potential 3 (TRPC3), a non-selective cation channel, was shown to be involved in insulin-responsive glucose uptake 
and was also implicated in muscle regeneration 
. Establishing how the GLUT4 pathway plays a role in muscle regeneration remains a task for the future, but our combined studies of human tissue and a mouse model in which the pathway is defective further support that this pathway contributes to the regenerative process.