Clogging of axons by tau, inhibition of axonal traffic and starvation of synapses

Abstract

Loss of synapses and dying back of axons are considered early events in brain degeneration during Alzheimer’s disease. This is accompanied by an aberrant behavior of the microtubule-associated protein tau (hyperphosphorylation, aggregation). Since microtubules are the tracks for axonal transport, we are testing the hypothesis that tau plays a role in the malfunctioning of transport. Experiments with various neuronal and non-neuronal cells show that tau is capable of reducing net anterograde transport of vesicles and cell organelles by blocking the microtubule tracks. Thus, a misregulation of tau could cause the starvation of synapses and enhanced oxidative stress, long before tau detaches from microtubules and aggregates into Alzheimer neurofibrillary tangles. In particular, the transport of amyloid precursor protein is retarded when tau is elevated, suggesting a possible link between the two key proteins that show abnormal behavior in Alzheimer’s disease.

Introduction

The loss of synapses and dying-back of axons is a characteristic feature of neurodegenerative diseases such as Alzheimer’s. They appear at early stages and correlate with the incipient loss of memory and brain functions [6], [7], [32]. A number of triggering events have been implicated, such as oxidative stress, inflammatory cytokines, lack of growth factors, or the toxic Aβ peptide which may lead to the decay of the axon or the neuron as a whole [25], [27]. These features may develop gradually and appear to precede the more overt pathological changes in the brain, such as deposition of protein aggregates in the form of amyloid plaques and neurofibrillary tangles.

Since neurons are highly elongated cells they depend on an efficient transport system for delivering proteins, lipids and other cell components from the cell body to the synapse. This system is based on microtubules which serve as tracks, motor proteins which represent the engines, vesicles and organelles which are the cargoes, and microtubule-associated proteins (MAPs) which serve as ties for the stabilization of the microtubule tracks [21]. In axons, tau protein is one of the predominant MAPs. It stabilizes microtubules and promotes neurite outgrowth. This apparently beneficial role of tau contrasts with its anomalous behavior in several neurodegenerative diseases, most prominently Alzheimer’s disease, where it occurs in a highly phosphorylated form, detaches from microtubules, and aggregates. It has been hypothesized that the detachment of tau from microtubules is caused by some imbalance in intracellular signaling which favors excessive phosphorylation. This in turn would detach tau from microtubules, prompt their decay, so that axonal transport would be interrupted by the breakdown of the tracks. In addition, the detached, soluble tau would aggregate and thus cause a generalized clogging of cytosolic space. One advantage of this hypothesis is that it relates neuronal degeneration to the abnormalities in tau, in agreement with observations (Braak stages [1], [5]). However, the assumptions on the underlying molecular mechanisms are not well supported at present; for example, it is not clear whether detachment of tau and its aggregation and the breakdown of microtubules are early causes of degeneration or later consequences. We, therefore, searched for other cellular roles of tau which might explain a more subtle and early defect in neuronal physiology. This lead us to study the interplay between motor proteins and MAPs, both of which have to interact with the microtubule surface. Unexpectedly, the “ties” are able to inhibit the “engines”. This is in contrast to a real railroad where ties and engines are spatially segregated on opposite sides of the tracks. This effect has the potential of blocking diverse transport processes, especially the anterograde processes from the cell body down the axon towards the synapse. It could be caused, for example, by a local elevation of tau which in its initial stages would not affect microtubule numbers or tau aggregation.

Human tau is encoded by a single gene (on chromosome 17) which can be spliced into six main isoforms in the central nervous system containing between 352 and 441 residues; a further higher molecular weight isoform is enriched in the peripheral nervous system (Fig. 1). The domain of the repeats (three or four, each about 31 residues) is essential for microtubule binding, and in addition it forms the core of Alzheimer PHFs [24]. However, domains flanking the repeats have to be present in order to ensure tight binding to microtubules. The N-terminal “projection domain” protrudes from the microtubule surface and presumably interacts with other cell components (e.g. anchor for kinases or phosphatases). Tau has a very hydrophilic composition which makes it highly soluble; the aggregation into insoluble fibers in disease states is therefore counterintuitive. Native tau also has a mostly unfolded structure, contrary to a typical globular protein. It is likely that some of this structure becomes folded when interacting with microtubules. Similarly, the repeat domain adopts a cross-beta structure when tau aggregates into pathological fibers, similar to other amyloids [35]. A notable feature of tau is its many potential phosphorylation sites, targeted by a variety of protein kinases. Most sites outside the repeat domain are of the type Ser–Pro or Thr–Pro and can therefore be phosphorylated by proline-directed kinases (MAP kinase, GSK-3β, cdk5). Other sites within the repeats are targets of non-proline-directed kinases, in particular the KXGS motifs in the repeats which are targets of microtubule-affinity regulating kinases (MARK [8]). Certain phosphorylated motifs are important diagnostic tools because they are among the first signs of neuronal degeneration [2], but their physiological roles are uncertain. Phosphorylation of the KXGS motifs in the repeats has the effect of detaching tau from microtubules and could therefore contribute to microtubule instability.

The results reported here were prompted by long-term observations of CHO cells transfected with GFP-labeled tau. The protein showed apparently normal behavior, i.e. it was attached to microtubules, and the microtubule affinities of different forms of tau were in general agreement with the values measured in vitro. However, there was a very slow rearrangement of cell components such as mitochondria which tended to accumulate in the cell center; the cells rounded up and divided less frequently [11]. Inspection of the movements of mitochondria and exocytotic vesicles by live-cell time-resolved microscopy revealed that movements were inhibited, especially those towards the cell periphery. Thus, the dominance of inward-directed movement lead to the retraction of mitochondria, cells lost their extended shape and rounded up [34]. The effects were clearly related to the binding of tau to the microtubule surface and were explained by an interference between microtubule-dependent motor proteins and tau [26]. Here we report on the consequences of this effect for neuronal cells, especially for their cell processes. We show that traffic inhibition by tau has drastic implications for the distribution of cell components, which leads to energy deprivation, lowered defence against oxidative stress, and dying-back of neurites.