Cytoplasmic dynein consists of two homodimerized heavy chains and multiple accessory proteins, including intermediate, light intermediate, and light chains [9], and is a major motor driving retrograde transport in cells. Dynein is involved in a variety of intracellular motile processes, including mitosis, maintenance of the Golgi apparatus, and the trafficking of membranous vesicles and other intracellular particles [5,10,11]. Two dynein mutant mouse strains, Legs at odd angles (Loa) and Cramping1 (Cra1), were identified in a screen for genes involved in late onset MND [7]. Heterozygous mice carrying these mutations exhibit an age-related progressive loss of muscle tone and locomotor ability [7]. The mechanisms by which these mutations cause neurodegeneration are still discussed. Although Hafezparast et al. demonstrated that the Loa mutation causes axonal transport defect by using an assay for the visualization and quantitation of axonal retrograde transport of motor neuron cultures from Loa/Loa mouse based on a fluorescent fragment of tetanus toxin (TeNT HC), heterozygous knock out of dynein did not exacerbate, but ameliorated, axonal transport in a mutant superoxide dismutase 1 (SOD1) mouse model of ALS [7,12]. Moreover, Ori-McKenney et al. reported that neurodegeneration in the Loa^+/− mutant mouse is unlikely to result from dynein subunit dissociation or dissociation of dynein from its cargo, but from an altered interaction between dynein and microtubules [13]. On the other hand, another mutation of dynein, sprawling (Swl), induces an early-onset sensory neuropathy, but not progressive motor neuron loss, in the heterozygous knock out mice for this gene, indicating the perplexing relationship between dynein mutations and axonal transport [14]. In addition to the mutations in dynein, a G59S mutation in the P150^Glued subunit of dynactin (dynactin-1) has been identified in families with a slowly progressive, autosomal dominant form of lower motor neuron disease [8]. Dynactin is a multiprotein complex associated with dynein [15] and is required to attach dynein to its cargos [16]. Indeed, the dynactin depletion model of Drosophila exhibited disrupted axonal transport [17]. The G59S substitution occurs in the highly conserved CAP-Gly motif of dynactin-1, a domain that binds directly to microtubules [8] and slows transport of organelles in vitro [18]. The G59S-mutated protein also has an enhanced propensity to misfold and forms aggregates in which trapped organelles, such as mitochondria, are found. Although early degeneration of axons and neuromuscular junctions are the common features of three different mouse models of mutant G59S [19,20], whether an axonal transport defect is present in these models has yet to be elucidated. Lai et al. reported that the retrograde transport of neurofilaments and synaptophysin were affected in motor neurons of heterozygous knock-in mice with G59S mutant dynactin-1 [21]. However, Chevalier-Larsen et al. demonstrated the absence of axonal transport defects in their transgenic mouse model by performing a double ligation assay on sciatic nerves [19]. To resolve this discrepancy, further experiments, such as in vivo imaging of axons, are needed to analyze the transport of cargos, and it is also important to generate and analyze dynactin-1 depletion models.

The kinesin superfamily proteins (KIFs) comprise three major groups: N-terminal motor domain KIFs (N-KIFs), middle motor domain KIFs (M-KIFs), and C-terminal motor domain KIFs (C-KIFs), and are mainly responsible for the anterograde transport of the various materials to nerve terminals [22–24]. KIF5A encodes a neuron-specific, kinesin-1 heavy chain member of the kinesin superfamily, which has been implicated in MND. Mutations in the gene for the KIF5A subunit of kinesin 1 (formerly SPG10) are associated with an early-onset, pure form of HSP, exhibiting a dying-back neuropathy characterized by progressive weakness and spasticity of the legs [5,25]. The pathological mechanism underlying this mutant is the disruption of KIF5A’s ability to bind with microtubules, leading to a failure of adequate anterograde transport [26]. Because KIF5A motor proteins transport many cargos that are important for neurons, including neurofilaments, mitochondria, and SNARE proteins essential for synaptic vesicle docking [27,28], disruption of KIF5A motility results in dysregulation of axonal homeostasis. In addition to mutations in the kinesin family itself, modifications of the motor domain are also potential causes of MNDs [29].

Neurofilaments are made up of three components, neurofilament light chain (NF-L), middle chain (NF-M), and heavy chain (NF-H). Neurofilament assembly in the axon is essential for establishing proper axon diameter, and its aggregates have long been recognized as a characteristic pathological change in MNDs [30–32]. Neurofilaments are transported along microtubules by kinesin and the dynein/dynactin-1 complex, and dysfunctions of these motor proteins lead to accumulation of neurofilaments in axons and somata, as observed in the KIF5A mutant mouse and dynactin-1 mutant mouse [20,30,33,34]. Moreover, mutations in neurofilaments also cause motor neuron degeneration. Mersiyanova et al. reports that the A998C transversion in the first exon of NF-L results in Charcot-Marie-Tooth (CMT) type 2 [35]. The missense mutations in NF-L observed in CMT lead to disruption of axonal transport of neurofilaments and to the subsequent assembly and aggregation of NF-L in neuronal cells and in primary cultured neurons [36,37]. These studies strongly suggest that disruption of the axonal transport of neurofilaments is one of the underlying pathogenic mechanisms in the development of MNDs.

Growing evidence suggests that mitochondrial damage is involved in motor neuron degeneration in CMT, HSP, and ALS. Local energy synthesis by mitochondria is essential to axonal homeostasis and its altered distribution and dysfunction play critical roles in axonal degeneration. Because mitochondria are transported by kinesin and the dynein/dynactin complex [24], defects in axonal transport result in abnormal accumulation of mitochondria. It has been shown that mitochondria accumulate in the axons of spinal motor neurons in mouse mutant SOD1 models [38–40] and in sporadic ALS (SALS) patients [41], suggesting an impairment of axonal transport in ALS. Indeed, in mutant SOD1 models, a defect in the axonal transport of mitochondria in sciatic nerves was directly observed during the presymptomatic stage [39]. Moreover, the presence of mitochondria with abnormal morphology, observed in both ALS models and SALS patients, indicates a poor recycling or degradation of abnormal mitochondria due to impaired retrograde transport [42–44].

It is important to assess whether the activation of mitochondrial mobility in axons has a beneficial effect on axonal degeneration in MNDs. It is proposed that increased mitochondrial transport might aid the efficient delivery of healthy mitochondria to axons and/or the removal of damaged mitochondria from the periphery. However, Zhu et al. reported that mutant SOD1 mice crossed with syntaphilin knock-out mice that have a higher proportion of axonal mitochondrial mobility, displayed no observable improvement in the deterioration of motor function or in disease progression and lifespan [45]. They suggested that mitochondrial dysfunction, rather than impaired mobility, may be more important for mutant SOD1 toxicity. On the other hand, a mouse model of mutant HSPB1-induced CMT exhibits impaired axonal transport of mitochondria, and treatment with histone deacetylase 6 inhibitors restored mitochondrial transport and attenuated the motor phenotype of this model [46].

The autophagy-lysosome pathway is responsible for the highly regulated recycling of intracellular contents, and it is well known that dysfunction of this pathway causes neurodegeneration [47,48]. Abnormal intraneuronal accumulation of autophagosomes is observed in various neurodegenerative diseases, including Alzheimer’s disease [49], Parkinson’s disease [50], and Huntington’s disease [51]. Because the autophagosome is a cargo that moves bidirectionally along microtubules powered by the kinesin family of motor proteins and the dynein/dynactin complex [52–55], the autophagosomes accumulates in those motor neurons with altered axonal transport. Human pathology and animal models of motor neuron disease also exhibited the same accumulations of autophagosomes, for example in SALS patients [56], the mutant dynein model (Loa mouse) [47], and in the mutant dynactin-1 mouse model [20]. Abnormal organelles, such as damaged mitochondria and dilated endoplasmic reticulum, are also observed in these motor neurons, and organelle accumulations are also observed in the neuritis of the Atg7^−/− mouse, which cannot form autophagosomes [48]. Therefore, disruption of autophagosomal transport resulting in the dysfunction of the autophagy system in axons is construed to be a causative mechanism of axonal degeneration.

Amyotrophic lateral sclerosis is a fatal neurodegenerative disease characterized by progressive loss of motor neurons [67]. Approximately 10% of ALS cases are familial (FALS) with inheritance patterns, while 90% are sporadic (SALS) with no known genetic defect [68,69]. There is at present no obvious consensus understanding of the pathogenic mechanism of SALS; however, the abnormal accumulation of neurofilaments, damaged mitochondria, and autophagosomes present in motor neurons of SALS patients [41,44,56,70], suggests that axonal transport defects might be implicated in the pathogenesis, but the molecular mechanisms causing axonal transport defects are still unclear.

To uncover this mechanism, we created a motor neuron-specific gene expression profile in patients with SALS using microarray technology combined with laser-captured microdissection, the results of which were further verified by in situ hybridization and quantitative wide-ranging research activities [71]. Among the genes with dysregulated expressions, dynactin-1 was markedly and widely downregulated in SALS motor neurons [71]. Furthermore, we found that the downregulation of dynactin-1 occurs prior to the accumulation of phosphorylated-neurofilament, which is, again, a marker of defective axonal transport and is considered a rather early marker of neurodegeneration (Figure 1) [72]. The dramatic change in dynactin-1 in patients with SALS seems to be specific for motor neurons, as dynactin-1 expression was preserved in neurons in the dorsal nucleus of Clarke and the intermediolateral nucleus in the spinal cord, Purkinje cells of the cerebellum, and cortical neurons in the occipital cortex. By taking into account that mutations in dynactin-1 cause MNDs, these results strongly suggest that the downregulation of dynactin-1 in motor neurons may play a significant role in motor neuron degeneration in SALS.

Spinal and bulbar muscular atrophy, or Kennedy’s disease, is a hereditary neurodegenerative disease characterized by a loss of bulbar and spinal motor neurons [63,73]. The causative molecular defect in SBMA is the expansion of a trinucleotide CAG repeat, which encodes a polyglutamine tract in the first exon of the AR gene [74]. The resulting mutant AR forms aggregations in the nucleus and affects the expression of various genes by inhibiting transcription factors and coactivators [75,76]. Polyglutamine-expanded AR inhibits axonal transport via a pathway involving cJun N-terminal kinase (JNK) activation, phosphorylation of kinesin-1 heavy chain subunits by JNK, and inhibition of kinesin-1 function [29]. Interestingly, we found that dynactin-1 is downregulated in the spinal motor neurons of SBMA patients and transgenic model mice (Figure 2) [62], which is also observed in SALS patients [71,72]. Fluoro-gold labeling revealed that retrograde axonal transport in SBMA mice was significantly impaired. Moreover, abnormal accumulations of neurofilaments and synaptic proteins were observed in the distal part of axons in these mice. The retrograde transport defects preceded the onset of motor symptoms, and were restored by hormonal interventions that prevent accumulations of mutant AR in the nucleus. The disruption of retrograde axonal transport was also documented in a knock-in mouse model of SBMA and in mice over-expressing wildtype AR in muscle [77], although this was not the case in another mouse model of SBMA [78].

Spinal muscular atrophy is an autosomal recessive disease that selectively affects lower motor neurons and is classified into 3 forms: a severe form (type I; Werdnig–Hoffmann disease), an intermediate form (type II), and a juvenile form (type III; Kugelberg–Welander disease). SMA occurs in 1/6000 live births and is the leading genetic cause of infantile mortality in the USA. SMA is caused by the mutation or deletion of the “survival motor neuron 1” gene (SMN1), which plays an important role in RNA metabolism [79]. Reduced levels of SMN1 lead to pathological changes at neuromuscular junctions, including decreased synaptic vesicle density, synaptic transmission defects, and neurofilament accumulation [80]. In a mouse model of SMA, fast anterograde axonal transport is disrupted, providing the molecular basis for the synaptic pathology [81]. The levels of dynein in the sciatic nerve are also reduced in this mouse model, suggesting the concomitant involvement of retrograde axonal transport [81].

The hereditary spastic paraplegias are a heterogeneous group of genetic disorders characterized by lower extremity spastic weakness due to retrograde degeneration of the corticospinal tracts and posterior columns, which is often accompanied by brisk reflexes, extensor plantar reflexes, and urinary urgency [82]. This disorder is classified into pure (uncomplicated) HSP and complicated HSP, depending on the presence of other neurological features in addition to spastic paraparesis. Spastic weakness is confined to the lower extremities in pure HSP, whereas complicated HSP is associated with mental retardation, ataxia, extrapyramidal signs, visual dysfunction or epilepsy. Among the causative proteins of HSPs, atlastin, spastin, and K1F5A regulate axonal transport [33,64–66]. While atlastin and spastin facilitate vesicle trafficking, KIF5A regulates anterograde axonal transport as described above.