Ian R. Griffiths

Mutations of the major myelin gene, proteolipid protein (Plp), cause Pelizaeus‐Merzbacher disease and some forms of spastic paraplegia in man and dysmyelinating phenotypes in animals. The clinical severity is markedly heterogeneous, ranging from relatively mild to severe and fatal. Point mutations, or frame shifts, which are predicted to result in translation of structurally altered proteins account for many of these cases, including 3 of the allelic murine conditions. Plp^jp‐rsh, Plp^jp‐msd, and Plp^jp represent an increasing severity of clinical and pathological phenotypes, respectively. In this study we determined whether there was any correlation between the severity of phenotype and the transport of the predicted abnormal protein. We examined the ability of the two products of the Plp gene, PLP and DM20, to insert into the plasma membrane of transfected BHK or COS‐7 cells, and into the myelin sheath of oligodendrocytes. With these complementary in vitro and in vivo approaches we find that proteins of Plp^jp‐rsh, associated with the mildest phenotype, have a far greater ability to insert into the cell membrane or myelin than those associated with the severe phenotypes. Additionally, altered DM20 is more readily transported to the cell surface and to myelin than the PLP isoprotein. Interestingly, the two clonal cell lines chosen for transient transfection differ in their ability to fold DM20 from Plp^jp‐rsh and Plp^jp‐msd mice correctly, as inferred by staining for the conformation‐sensitive O10 epitope. In the case of Plp^jp, which is associated with the most severe phenotype, no PLP or O10 staining is present at the cell surface or in myelin. The perturbation in trafficking observed for altered Plp^jp PLP and DM20 in oligodendrocytes does not extend to other myelin membrane proteins, such as MAG and MOG, nor to wild type PLP co‐expressed in the same cell, all of which are correctly inserted into myelin. As Plp‐knockout mice do not have a dysmyelinating phenotype it seems unlikely that absence of PLP and/or DM20 in the membrane is responsible for the pathology. It remains to be determined whether the perturbation in protein trafficking is associated with the dysmyelination, or if the altered product of the mutant alleles acquire a novel function which is deleterious to myelin production by oligodendrocytes. GLIA 20:322–332, 1997.

Spontaneous mutations that perturb myelination occur in a range of species including man, and together with engineered mutations have been used to study disease, normal myelination and axon/glial inter‐relationships. Only a minority of the currently defined mutations have an apparently simple pathogenesis due to lack of a functional protein. Mutations in the myelin basic protein gene lead to a lack of protein, resulting in changes in the structure of myelin, which can be rescued by transgenic complementation. The pathogenesis of autosomal dominant and X‐linked mutations affecting either oligodendrocytes or Schwann cells is more complex. Point mutations may act in a dominant negative manner and gene dosage is clearly linked to phenotypic change. Mutations in regulatory genes, such as those encoding transcription factors, can also disturb myelination by selected cell types. Other less‐well studied and unexpected consequences of myelin mutations, such as seizures in mutations affecting genes expressed in Schwann cells and axonal changes associated with dysmyelination, are also considered. With the major developments in gene mapping and cloning it is now relevant to study mutations in a variety of species with the real prospect of defining their molecular basis. Examples are given of unusual, but potentially useful, uncharacterised mutations in dog and bovine.

The X chromosome‐linked PLP/DM‐20 gene is the CNS myelin gene most frequently associated with mutations, resulting in dysmyelination in several species including man (Pelizaeus‐Merzbacher disease, X‐linked Spastic Paraplegia). The pathology of most PLP gene mutations is characterized by hypomyelination, glial cell proliferation, increased numbers of microglia, and premature oligodendrocyte death. In most mutants, residual myelin structures have an abnormal ultrastructure and periodicity. Surprisingly, transgenic mice which carry extra copies of the wild type PLP gene show dysmyelination, demonstrating that the PLP gene is dosage sensitive. Pathological changes of transgenic mice vary from the phenotype of natural mutants. Specifically, many Golgi saccules of oligodendrocytes are vacuolated and the cytoplasm contains autophagic vacuoles hinting at a perturbation in protein trafficking. In fact, upon transgenic overexpression PLP becomes a prominent peripheral myelin protein, whereas in normal Schwann cells PLP is restricted from entering the myelin compartment. Surprisingly, transgenic animals which overexpress PLP/DM‐20 at a low level appear normal during early development, but later spontaneously demyelinate. The mechanisms underlying this demyelination phenotype is unknown but an immune‐mediated process has been suggested. All attempts to correct the phenotype of natural PLP mutants, such as jimpy mice, with a wild type transgene have had little effects, indicating a dominant‐negative effect of the mutant gene product. On the other hand, mice with a targeted disruption of the PLP/DM‐20 gene have suprisingly minor clinical signs. This suggests that the lethal phenotype associated with the majority of PLP gene mutations is a complex combination of loss and gain‐of‐function effects of a mutant myelin protein.