Regulation of Subcellular Organization in Skeletal Muscle

在过去的一年里，我们已经成功地将编码几种与GFP相关的靶蛋白的cdna表达到骨骼肌中。我们使用了鼠标脚垫的肌肉，即短屈指肌（FDB）来进行这些实验。FDB很容易进入；DNA可以被注射并电穿孔到肌肉中，而不需要打开皮肤。我们已经测试了编码形成微管核心的蛋白质（微管蛋白- gfp），与微管相关的蛋白质（MAP4-GFP， enconsin微管结合位点- gfp）和涉及微管前端的蛋白质（EB1-和EB3-GFP）的结构。我们已经能够获得大量表达这些结构的纤维，更好的是，其中一些结构再现了内源性微管的模式。我们已经观察到，在肌肉细胞培养物中用于跟踪微管的结构不一定在肌肉纤维中起作用，反之亦然。

During the past year we have successfully expressed cDNAs encoding several target proteins linked to GFP into skeletal muscles. We have used the muscle of the mouse footpad, the Flexor Digitorum Brevis (FDB) for these experiments. The FDB is easy to access; DNA can be injected and electroporated into the muscle without need to open the skin. We have tested constructs encoding proteins that form the core of the microtubules (tubulin-GFP), proteins associated with microtubules (MAP4-GFP, ensconsin microtubule-binding site-GFP) and proteins involved in the plus-end tip of the microtubules (EB1- and EB3-GFP). We have been able to obtain good numbers of fibers expressing the constructs and, better, several of the constructs reproduce the pattern of the endogenous microtubules. We have observed that constructs that are useful for tracking microtubules in muscle cell cultures are not necessarily working well in muscle fibers, and vice versa. As a first approach, we have taken single fibers of the FDB obtained by collagenase dissociation of the muscle one week after cDNA injection. The fibers are plated on a chamber suitable for confocal microscopy and observed within 24h of plating. Time-lapse recordings of single FDB fibers plated after collagenase dissociation of the muscle reveal the dual character to the microtubule network of muscle fibers. There is a stable lattice, composed of bundles of microtubules of different polarities, forming both transverse and longitudinal paths. Along this network, individual microtubules move constantly, generally in one direction. Microtubules starting their movement obliquely generally shift direction when they hit one of the lattice lines and then follow the line. We propose that the interaction with dystrophin (demonstrated by Prins et al., 2009) is responsible for the alignment of microtubules with the lattice. Dystrophin forms transverse ribs, the costameres of muscle fibers. In the mdx mouse which lacks dystrophin, microtubules do not form an orthogonal lattice. This stable microtubule lattice which, through dystrophin, would be associated with the other cytoskeletal systems of muscle fibers, may provide structural support to the cell. The very dynamic microtubules that course along the lattice may, in contrast, play the traditional roles of microtubules in protein and organelle transport. We have observed nucleation of microtubules during recovery from nocodazole treatment, and at steady-state, which could only be observed on live fibers. Nucleation is manifested by the formation of asters of short microtubules which resemble those observed in proliferating cells such as myoblasts. However, whereas nucleation in myoblasts starts from the centrosome, in muscle fibers nucleation starts from the multiple Golgi complexes (Golgi elements). Nucleation in muscle fibers is also associated with the nuclear membrane, as is the case in myotubes. The association with the Golgi complex is being further tested by expressing cDNAs to knock down Golgi complex proteins hypothesized to be involved in the process. This mechanism gives muscle fibers a very local control of microtubule assembly and appears to conjugate the nucleation mechanisms found in myoblasts and in myotubes, a "management" system unique to muscle and likely well adapted to the huge size of the individual cells of muscle. Our interest in applying microscopy techniques for the detection and analysis of muscle defects in myopathies, coupled with our collaboration with Dr. Raben on Pompe disease has led us into developing new software for the quantitation of muscle anomalies. Various forms of microscopy reveal muscle defects such as changes in the periodicity of the contractile proteins, inclusion of non-contractile bodies etc. It is however very difficult to quantitate these defects and therefore to compare the severity of disease between different biopsies or the evolution of muscle health following treatments such as enzyme replacement therapy in the case of Pompe disease. The new software based on the Matlab platform uses different filters and the concept of image "texture" to quantitate muscle defects without prior assumption of the type of defects. We are in the process of testing its ruggedness and applying it to samples from Pompe patients.