9. Microtubules and Memory Encoding

Encoding

Figure 9.1. A model for memory encoding in the microtubular network. A blue line represents a microtubule molecule. (A) The "CLOSED" state. A microtubular track is interrupted before reaching its destination (e.g., a spine). Transport along this track is inefficient. (B) The "OPEN" state. A microtubular track is continuous up to its destination.

It has been observed that, for vertebrate neurons, microtubules in axons are unidirectional, with their plus ends distal to the cell body (plus-end-out). In dendrites, however, their polarity is mixed (reference 1; reference 2), and age-dependent (reference).

The microtubular polarity in axons can easily be understood. Most neurotransmitter receptors are produced at the cell body while their destination in an axon is its terminal. Transport of neurotransmitter receptors by microtubules is predominantly mediated by the motor protein kinesin (reference), which walks toward the plus end. This explains why the microtubular polarity in axons should be plus-end-out.

In dendrites, the destination of neurotransmitter receptors is spines (small membranous protrusions which form synapses with axon terminals), located away from the cell body. For kinesin to carry neurotransmitter receptors from the cell body to spines, the microtubular polarity in dendrites should also be plus-end-out. Then, why do dendrites need minus-end-out microtubules? A possible answer lies in another motor protein dynein, which walks toward the minus end. Dynein is responsible for the retrograde transport of membranous organelles from the axon terminal to the cell body. It can also mediate the transport of endosomes and the Golgi complex (review). Recently, dynein has been demonstrated to play a crucial role in dendritic branching (reference). This result is consistent with another finding that, in Drosophila (fly) dendrites, microtubular orientations are predominantly minus-end-out. These minus-end-out microtubules are used by dynein for dendritic growth (reference).

Although memory appears to be stored in synapses, it is the microtubular network which is calling the shots. Without dynein, dendrites cannot grow or branch out. Without kinesin, no neurotransmitter receptors are available for modifying synaptic strength. The two motor proteins walk along microtubular tracks. Therefore, memory could be encoded in the opening or closing of these microtubular tracks.

In axons, microtubules are continuous from the cell body to the axon terminal, but in dendrites they are interrupted. Figure 9.1 presents a simple binary model which may be used for memory encoding and storage. A microtubular track to its destination (e.g., a spine) is "OPEN" if a microtubule is continuous up to the destination. If a track is interrupted, namely, consisting of two or morel shorter microtubules, then the track is "CLOSED" because transport along this track will be inefficient.

This model is based on the observation that, in apical dendrites of mice, the average microtubular length increases with age (reference). It seems that experience can leave two types of memory traces: (1) insertion of neurotransmitter receptors into postsynaptic membranes (see Chapter 7) and (2) modification of microtubular tracks. The length of a microtubule is directly controlled by microtubule-associated proteins (MAPs) which, in turn, is regulated by calcium ions. Memory traces generally increase the length of microtubules so that motor proteins can walk continuously toward their destination.

In addition to their effect on the length of microtubules, memory traces are expected to make microtubules more rigid as MAPs may fix microtubular tracks to specific destinations. The next chapter will provide further evidence that the portion of short and free microtubules declines with age.

 

Author: Frank Lee
First Published: September 10, 2009