Increased expression of neurofilament subunit NF-L produces morphological alterations that resemble the pathology of human motor neuron disease

Premature arrest of myelin formation in transgenic mice with increased proteolipid protein gene dosage

Proteolipid protein (PLP) is an integral membrane protein of CNS myelin. Mutations of the X chromosome-linked PLP gene cause glial cell death and myelin deficiency in jimpy mice and other neurological mutants. As part of an attempt to rescue these mutants by transgenic complementation, we generated normal mouse lines expressing autosomal copies of the entire wild-type PLP gene. Surprisingly, increase of the PLP gene dosage in nonmutant mice with only 2-fold transcriptional overexpression results in a novel phenotype characterized by severe hypomyelination and astrocytosis, seizures, and premature death. This demonstrates that precise control of the PLP gene is a critical determinant of terminal oligodendrocyte differentiation. Dysmyelination of PLP transgenic mice provides experimental evidence that Pelizaeus-Merzbacher disease, previously associated with a partial duplication of the human X chromosome, can be caused by doubling of the PLP gene dosage.

Neurofilament-deficient axons and perikaryal aggregates in viable transgenic mice expressing a neurofilament-beta-galactosidase fusion protein

Interactions between neurofilament side arms may modulate axon caliber. To investigate this hypothesis, we derived transgenic mice expressing a fusion protein in which the carboxyl terminus of the high molecular weight neurofilament protein (NFH) was replaced by beta-galactosidase. The transgene, regulated by NFH sequences, was expressed in projection neurons. However, the fusion protein remained in perikarya precipitating large filamentous aggregates. Axons were not invested with neurofilaments and developed only small calibers. Perikaryal aggregates, with similar structural features, are associated with neurodegenerative diseases, but these mice showed few ill effects and their neurons rarely degenerated. We conclude that an organized neurofilament cytoskeleton is required by axons to achieve large calibers but is not essential for neuronal function or extended survival.

Making heads and tails of intermediate filament assembly, dynamics and networks

Thus far, intermediate filaments (IFs) have been the least understood of the three cytoskeletal filament systems with regard to their structure, assembly, network formation, and dynamics. This picture is now slowly but definitely changing, as recent in vivo and in vitro experiments, including generation of transgenic animals, have yielded important new data shedding light on the following areas: the molecular architecture of IFs; the role of the highly variable end domains during IF assembly and network formation; the factors that govern whether IF proteins are involved in de novo filament formation or are incorporated into a pre-existing IF network; and the effects of post-translational modifications, such as phosphorylation and glycosylation of IF polypeptides, on filament assembly, dynamics and turnover.

Neurofilament function and dysfunction: involvement in axonal growth and neuronal disease

Neurofilaments make up the major intermediate filament system in mature neurons. Recent studies demonstrate that neurofilaments in vivo are obligate heteropolymers and are required for proper radial growth of axons. Furthermore, forced over-expression of neurofilament subunits in transgenic mice shows that abnormal accumulation and assembly of neurofilaments, similar to that commonly found in human motor neuron disease can directly cause motor neuron dysfunction.

Degeneration of skeletal muscle, peripheral nerves, and the central nervous system in transgenic mice overexpressing wild-type prion proteins

Prion diseases of humans and animals are known to be caused by infection with prions containing PrPSc or mutation of the prion protein (PrP) gene. During transgenetic studies, we discovered that uninoculated older mice harboring high copy numbers of wild-type (wt) PrP transgenes derived from Syrian hamsters (SHa), sheep (She), and PrP-B mice developed truncal ataxia, hindlimb paralysis, and tremors. These transgenic (Tg) mice exhibited a profound necrotizing myopathy involving skeletal muscle, a demyelinating polyneuropathy, and focal vacuolation of the central nervous system. Development of disease was dependent on transgene dosage. For example, half of all Tg(SHaPrP+/+)7 mice homozygous for the SHaPrP transgene array developed disease by approximately 460 days of age, while no hemizygous Tg(SHaPrP+/o)7 mice became ill before 650 days. The novel neurologic syndrome found in older Tg(wtPrP) mice implies that overexpression of wtPrPC is pathogenic and widens the spectrum of prion diseases.

Transgenic and knock-out mice: models of neurological disease

Besides providing useful model systems for basic science, studies based on modification of the mammalian germ line are changing our understanding of pathogenetic principles. In this article, we review the most popular techniques for generating specific germ line mutations in vivo and discuss the impact of various transgenic models on the study of neurodegenerative diseases. The "gain of function" approach, i.e., ectopic expression of exogenous genes in neural structures, has deepened our understanding of neurodegeneration resulting from infection with papova viruses, picorna viruses, and human retroviruses. Further, inappropriate expression of mutated cellular molecules in the nervous system of transgenic mice is proving very useful for studying conditions whose pathogenesis is controversial, such as Alzheimer's disease and motor neuron diseases. As a complementary approach, ablation of entire cell lineages by tissue-specific expression of toxins has been useful in defining the role of specific cellular compartments. Modeling of recessive genetic diseases, such as Lesch-Nyhan syndrome, was helped by the development of techniques for targeted gene deletion (colloquially termed "gene knock-out"). Introduction of subtle homozygous mutations in the mouse genome was made possible by the latter approach. Such "loss of function" mutants have been used for clarifying the role of molecules thought to be involved in development and structural maintenance of the nervous system, such as the receptors for nerve growth factor and the P0 protein of peripheral myelin. In addition, these models are showing their assets also in the study of enigmatic diseases such as spongiform encephalopathies.

Genetic basis of inherited peripheral neuropathies

Progress in the elucidation of the genetic basis for inherited peripheral neuropathies has been remarkable over the last years. In particular, the molecular mechanisms underlying the autosomal dominantly inherited disorders Charcot-Marie-Tooth disease type 1A (CMT1A), Charcot-Marie-Tooth disease type 1B (CMT1B), and hereditary neuropathy with liability to pressure palsies (HNPP) have been determined. While mutation in the gene encoding the major myelin protein, P0 has been associated with CMT1B, CMT1A and HNPP have been shown to be associated with reciprocal recombination events leading either to a large submicroscopic duplication in CMT1A, or the corresponding DNA deletion in HNPP. Available evidence is consistent with the hypothesis that one or more genes within the relevant rearranged segment of 1.5 Mb on chromosome 17 is sensitive to gene dosage providing a novel mechanism for inherited human disorders. It is likely that the gene encoding the peripheral myelin protein PMP22 is at least one of the genes involved since the PMP22 gene maps within the CMT1A duplication (or HNPP deletion), and point mutations within it have been shown to cause a CMT phenotype in humans and comparable neuropathies in rodents (trembler and tremblerJ). The mechanism(s) by which gene dosage and point mutations affecting the same gene might lead to a similar phenotype are currently unknown but recent transgenic mouse experiments suggest that similar mechanisms may also underlie other genetic diseases.

Transient expression of the intermediate filament nestin during skeletal muscle development

It has previously been established that skeletal muscle development is accompanied by changes in the composition of intermediate filaments: vimentin is expressed predominantly in myoblasts and desmin in adult myotubes. We show that the intermediate filament transitions during muscle development are more complex, and involve a transient expression of the recently discovered intermediate filament nestin. Nestin RNA is expressed predominantly early, in a biphasic pattern, and is markedly downregulated in adult rat muscle, whereas desmin RNA becomes more abundant throughout development. Nestin protein was found up to the postnatal myotube stage, where it colocalized with desmin in Z bands. The intracellular distribution of nestin, vimentin and desmin was analysed in the human myogenic cell line G6 before and after in vitro differentiation. Despite its more distant evolutionary and structural relationship to the other two intermediate filaments, nestin formed a cytoplasmic filamentous network indistinguishable from that of desmin and vimentin, both in undifferentiated myoblasts and after differentiation to multinuclear myotubes. In conclusion, our data suggest that nestin is an integrated component of the dynamic intermediate filament network during muscle development and that nestin copolymerizes with desmin and vimentin at stages of coexpression.

Neurofilaments are obligate heteropolymers in vivo

Neurofilaments (NFs), composed of three distinct subunits NF-L, NF-M, and NF-H, are neuron-specific intermediate filaments present in most mature neurons. Using DNA transfection and mice expressing NF transgenes, we find that despite the ability of NF-L alone to assemble into short filaments in vitro NF-L cannot form filament arrays in vivo after expression either in cultured cells or in transgenic oligodendrocytes that otherwise do not contain a cytoplasmic intermediate filament (IF) array. Instead, NF-L aggregates into punctate or sheet like structures. Similar nonfilamentous structures are also formed when NF-M or NF-H is expressed alone. The competence of NF-L to assemble into filaments is fully restored by coexpression of NF-M or NF-H to a level approximately 10% of that of NF-L. Deletion of the head or tail domain of NF-M or substitution of the NF-H tail onto an NF-L subunit reveals that restoration of in vivo NF-L assembly competence requires an interaction provided by the NF-M or NF-H head domains. We conclude that, contrary to the expectation drawn from earlier in vitro assembly studies, NF-L is not sufficient to assemble an extended filament network in an in vivo context and that neurofilaments are obligate heteropolymers requiring NF-L and NF-M or NF-H.