Anthony Brown

Neurons and glia form the circuitry of our nervous system, allowing nerve cells to communicate with each other and other cell types throughout the body in a highly specific manner to integrate sensory, motor, and cognitive functions. The ability of these cells to do this is profoundly dependent on an elaborate and dynamic cytoskeleton which allows these cells to adopt complex shapes and specialized structures such as myelin sheaths, the bloodbrain barrier, axonal growth cones, dendritic spines, nodes of Ranvier, synapses, and elaborate sensory organs. The cytoskeleton is also critical for morphogenetic movements, the migration of neuronal precursors, and axon growth, which together establish the organization and wiring pattern of the central and peripheral nervous system during development. This review provides a primer on our current understanding of the cytoskeleton of neurons and glia, with an emphasis on how the protein polymers of the cytoskeletonmicrotubules, actin filaments, intermediate filaments, and septinswork together to orchestrate the remarkable shapes and functions of these cells. We also highlight key questions and challenges for the field and technological advances that may help address them.

It is demonstrated that neurofilament length and transport are regulated by a dynamic cycle of neurofilament polymer severing and end-to-end annealing. An enzymatic severing mechanism for neurofilaments involving N-terminal phosphorylation of their constituent polypeptides is proposed.

Constrained by published data, a computational model is presented that can explain the radial growth of myelinated axons in vivo. It is found that the radial expansion of these axons is driven primarily by an increase in the influx of neurofilaments at early times and by a slowing of neurofilament transport at later times.

Studies in cultured neurons have shown that neurofilaments are cargoes of axonal transport that move rapidly but intermittently along microtubule tracks. However, the extent to which axonal neurofilaments movein vivohas been controversial. Some researchers have proposed that most axonally transported neurofilaments are deposited into a persistently stationary network and that only a small proportion of axonal neurofilaments are transported in mature axons. Here we use the fluorescence photoactivation pulse-escape technique to test this hypothesis in intact peripheral nerves of adult malehThy1-paGFP-NFMmice, which express low levels of mouse neurofilament protein M tagged with photoactivatable GFP. Neurofilaments were photoactivated in short segments of large, myelinated axons, and the mobility of these fluorescently tagged polymers was determined by analyzing the kinetics of their departure. Our results show that >80% of the fluorescence departed the window within 3 h after activation, indicating a highly mobile neurofilament population. The movement was blocked by glycolytic inhibitors, confirming that it was an active transport process. Thus, we find no evidence for a substantial stationary neurofilament population. By extrapolation of the decay kinetics, we predict that 99% of the neurofilaments would have exited the activation window after 10 h. These data support a dynamic view of the neuronal cytoskeleton in which neurofilaments cycle repeatedly between moving and pausing states throughout their journey along the axon, even in mature myelinated axons. The filaments spend a large proportion of their time pausing, but on a timescale of hours, most of them move.

Neurofilaments are abundant space-filling cytoskeletal polymers that are transported into and along axons. During postnatal development, these polymers accumulate in myelinated axons causing an expansion of axon caliber, which is necessary for rapid electrical transmission. Studies on cultured nerve cells have shown that axonal neurofilaments move rapidly and intermittently along microtubule tracks in both anterograde and retrograde directions. However, it is unclear whether neurofilament transport is also bidirectionalin vivo. Here, we describe a pulse-spread fluorescence photoactivation method to address this in peripheral nerves dissected fromhThy1-paGFP-NFMtransgenic mice, which express a photoactivatable fluorescent neurofilament protein. Neurofilaments were photoactivated in short segments of myelinated axons in tibial nerves at 2, 4, 8, and 16 weeks of age. The proximal and distal spread of the fluorescence due to the movement of the fluorescent neurofilaments was measured over time. We show that the directional bias and velocity of neurofilament transport can be calculated from these measurements. The directional bias was ∼60% anterograde and 40% retrograde and did not change significantly with age or distance along the nerve. The net velocity decreased with age and distance along the nerve, which is consistent with previous studies using radioisotopic pulse labeling. This decrease in velocity was caused by a decrease in both anterograde and retrograde movement. Thus, neurofilament transport is bidirectionalin vivo, with a significant fraction of the filaments moving retrogradely in both juvenile and adult mice.

Charcot–Marie–Tooth disease (CMT) is one of the most common inherited neurological disorders and can be caused by mutations in over 100 different genes. One of the causative genes is NEFL on chromosome 8 which encodes neurofilament light protein (NEFL), one of five proteins that co‐assemble to form neurofilaments. At least 34 different CMT‐causing mutations in NEFL have been reported which span the head, rod, and tail domains of the protein. The majority of these mutations are inherited dominantly, but some are inherited recessively. The resulting disease is classified variably in clinical reports based on electrodiagnostic studies as either axonal (type 2; CMT2E), demyelinating (type 1; CMT1F), or a form intermediate between the two (dominant intermediate; DI‐CMTG). In this article, we first present a brief introduction to CMT and neurofilaments. We then collate and analyze the data from the clinical literature on the disease classification, age of onset and electrodiagnostic test results for the various mutations. We find that mutations in the head, rod, and tail domains can all cause disease with early onset and profound neurological impairment, with a trend toward greater severity for head domain mutations. We also find that the disease classification does not correlate with specific mutation or domain. In fact, different individuals with the same mutation can be classified as having axonal, demyelinating, or dominant intermediate forms of the disease. This suggests that the classification of the disease as CMT2E, CMT1F or DI‐CMTG has more to do with variable disease presentation than to differences in the underlying disease mechanism, which is most likely primarily axonal in all cases.

We used long-term multi-field time-lapse imaging to analyze the movement, severing and annealing of single neurofilaments along axons of cultured neurons. All filaments were capable of rapid movement. However, long filaments paused and reversed more often, resulting in little net movement, whereas short filaments moved persistently for long distances, pausing and reversing less often. Long filaments severed more frequently, generating shorter filaments, and short filaments annealed more frequently, generating longer filaments. Site-directed mutagenesis to mimic phosphorylation at known phosphorylation sites in the neurofilament protein L head domain increased the severing rate, generating shorter neurofilaments that moved more frequently. A dephospho-mimic mutant had the opposite effect. Treatment with activators of protein kinase A increased filament severing, and this effect was blocked by the dephosphomimic. Thus, neurofilament length and transport are regulated by a dynamic cycle of severing and annealing. We propose that focal destabilization of intermediate filaments by N-terminal phosphorylation may be a general enzymatic mechanism for severing these cytoskeletal polymers. SUMMARY Uchida & Brown demonstrate that neurofilament length and transport are regulated by a dynamic cycle of neurofilament polymer severing and end-to-end annealing and propose an enzymatic severing mechanism for neurofilaments involving N-terminal phosphorylation of their constituent polypeptides.