Muscle precursor cells may act not only as a means of inserting normal genes into diseased muscle fibres, in order to correct or alleviate a genetically inherited myopathy, but recent demonstrations have shown they may prove an invaluable tool for the expression of, and systemic dissemination of, non-muscle gene products. If muscle precursor cells are proved to act as such widespread vectors in terms of gene therapy, then it is imperative that methods are properly elucidated to produce large populations of pure viable myogenic cells for such purposes. In the past, many methods of cell separation have been investigated but carry with them the problems of either a lack of myogenic purity of the population or poor percentage recovery of the original cell population. In the present work we have investigated two methods for segregating myogenic from non-myogenic cells and have critically reviewed the efficiency of separation of the two techniques used. To obtain a quantitative measure of separation efficiency, segregation was carried out on a 1:1 mixture of murine C2 myogenic and murine 3T3 fibroblastic cells. To distinguish between C2 and 3T3 cells, the latter were prelabelled with the fluorescent strain carboxyfluorescein diacetate succinimyl ester (CFSE). Once incorporated into the cell, CFSE remains there, thus preventing transfer of the label to C2 cells. Both methods of separation used depend on the affinity of myogenic cells for the monoclonal antibody Mab H28, which specifically binds to the mouse neuronal cell adhesion molecule N-CAM, but differ in that one method, “panning”, completes segregation by adherence of N-CAM positive cells to a dish precoated with secondary IgG antibody whereas in the other separation proceeds by the use of commercially available IgG-coated magnetic beads. Results indicate magnetic bead separation to be more efficient than panning if the beads are precoated with 0.1% gelatin.
With increasing interest in the idea of therapeutic implantation of normal muscle precursor cells into muscle lacking the protein product of the dystrophin gene, it has become important to obtain enriched populations of myogenic cells from biopsied muscle sources. Myogenic cells for implantation are highly favoured as they are the only cells that will fuse readily with host muscle fibres into which they are implanted, thus carrying the introduced gene into the target fibre with the maximum of efficiency. Second, myogenic cells appear less immunogenic than those of a non-myogenic nature; and third, the use of mononuclear myogenic cells may permit the introduction of multiple copies of a deficient gene into the patient's own cells. From a mixed population of cells obtained by the enzymic disaggregation of neonatal murine muscle we have selected, utilising a modification of the panning technique, for a cell population rich in myogenic cells. Segregation was accomplished using Mab H28, an antibody to the mouse neuronal cell adhesion molecule (N-CAM), derived from mouse/rat hybridoma cells. Following incubation with Mab H28, disaggregated muscle was applied to the surface of a bacteriological grade dish previously coated with anti-rat immunoglobulin. Cells segregated into two populations; those bearing N-CAM, and hence labelled with Mab H28, were adherent to the dish, whereas those not expressing N-CAM remained in suspension. Use of this technique, which involves minimal cell loss, resulted in the segregation of prefusion myogenic cells together with fibroblasts in the ‘non-adherent’ fraction, whereas cells in the adherent fraction consisted of a highly enriched population of actively dividing myogenic cells.