Cytoskeletal Dynamics. CS2

The three cytoskeletal components differ in the way that they polymerise. Microtubules and microfilaments share many properties of assembly, but intermediate filaments are fundamentally different. MTs and MFs bind GTP and ATP respectively which hydrolyses when these proteins polymerise. However, many other proteins including IFs and bacterial flagellin (which produces the flagella), polymerise without a requirement for nucleotide hydrolysis. The nucleotide hydrolysis is the price the cell must pay in order to allow rapid remodelling of these cytoskeletal elements. For technical (and historic) reasons most in vitro information is available from the microfilament systems, whereas the microtubule field is well ahead in in vivo studies. Comparatively little is known about IF dynamics. This discussion will therefore concentrate on the actin and tubulin systems.

Actin Polymerises to produce Microfilaments

In physiological conditions, actin monomers (G-actin) spontaneously self associate to form microfilaments (F-actin). This polymerised form of actin is in constant equilibrium with G-actin. Monomers are able to add to the ends of filaments, form nuclei with another two monomers to create new filaments, and to leave filaments. The polarity of actin subunits within microfilaments means that the two ends are topologically different, the narrow face of actin subunits is exposed at the pointed end while the more bulbous face containing the NH₂ and COOH termini is exposed at the barbed end. Pure actin can be switched between these states in the test-tube by altering the salt concentration. Actin in low salt conditions is in the G-state, while adding salts causes the actin to polymerise.

Figure 6

Polymerisation is a multi-step process. The first two steps are highly unfavourable. Only the trimer is more likely to lead to the next step to the right than to disassemble. Because the early stages are improbable the time course of polymerisation is multi-phasic, starting off slowly accelerating and finally, as the amount of free G-actin becomes limiting, the rate decreases.

Figure 7. After addition of salts at time zero, there is a lag phase during which the rate limiting nucleus for polymerisation form. These nuclei consist of three actin monomers associating to form the smallest polymerising unit. This step is highly unfavourable, but after these have formed, polymerisation by extension is favoured and rapid. The rate of extension gradually lessens as the available monomers become exhausted. An equilibrium is then established when the number of actin monomers joining the filaments equals that of monomers leaving the filaments

Tubulin Polymerises to produce Microtubules

The assembly of MTs is less clear but is thought to involve a similar process to that of actin (above). Tubulin dimers are stable, tightly bound structures under normal conditions and are the polymerising subunit. The intermediate steps are not known, but the formation of a tubulin nucleus is much less favourable than is the case for actin and so the lag phase is longer. In fact the nucleation step is so unlikely that MTs in cells grow from specialised structures known as MicroTubule Organising Centres (MTOC).

Actin and Tubulin Hydrolyse Nucleotide during Polymerisation

In both cases hydrolysis takes place after the particular subunit has joined the polymer but in neither case is nucleotide hydrolysis required for polymerisation. Actin which has ADP bound polymerises albeit more slowly that is the case for ATP and polymerisation can also take place when actin has a non-hydrolysable analogue of ATP in the binding site. The same is true for GDP-bound tubulin. The energy dissipated in the hydrolysis is used to destabilise the polymers. In both cases the Di-phosphate nucleotide containing polymer is much less stable than the Tri-phosphate state and so is likely to depolymerise. Therefore nucleotide hydrolysis allows polymer formation at the same time and place as polymer disassembly. This is crucially important for a wide variety of cell movement which will be discussed in following lectures.

Figure 8. The polymer state as a function of the concentration of protein present. In polymerising conditions no polymer is observed until the monomer concentration (ordinate) exceeds that of the critical concentration (CC) of the barbed/plus end of the MF or MT. However the filament formed would not be stable until after the point indicated above, where the off rate at the pointed/minus end equals that of the barbed/plus end. The model is equally true for both MTs and MFs, however differences between the two systems mean that some properties are difficult to see in one system while being very apparent in the other. Treadmilling is discussed later.

Dynamic Instability

This is a phenomenon known to occur in microtubules which has been observed in vivo as well as in vitro. Although it is theoretically possible that this also occurs in microfilaments it should become clear that the differences between the systems make this less likely and/or less important than is the case for MTs. Individual MTs can be seen in very thin processes in cells, and in most cells using fluorescence microscopy because of their relative large diameter and due to their general lack of abundance. It is only possible to see individual MFs with fluorescence microscopy using very dilute preparations. In cells some MTs are seen to grow, yet other MTs close by are seen to dramatically shorten and sometimes to disappear. In other cases a rapidly shortening MT may suddenly start to grow. This is “dynamic instability” and is explained (surprisingly easily perhaps) by the state of bound nucleotide. It is known that MTs depolymerise very rapidly if they are composed of GDP- bound a-b dimers while being much more stable if composed of GTP dimers. As dimers (usually GTP-dimers) may add to either end of an MT this means that both ends are likely to have GTP dimers. If however the rate of GTP hydrolysis exceeds the rate of dimer addition even transiently, GDP-dimers would be exposed and the filament would rapidly disassemble.

Figure 9

Treadmilling.

The barbed end of microfilaments is preferred for monomer addition over the pointed end, where net disassembly is favoured. Within the region labelled “treadmilling possible” in figure 8 page 6, a net flux of monomers occurs along the length of the filament. Monomers add to the barbed end, travel the length of the filament as other monomers add behind, and leave from the pointed end, this is called this "treadmilling". There is evidence that treadmilling may indeed occur close to the leading edge of motile lamellipodium of fibroblasts, neuronal growth cones and this is probably true for all expanding lamellae. Interestingly, a microfilament in the process of treadmilling may actually do work, this will be discussed later.

Figure 10. If the overall filament size remains constant, addition at the barbed end and loss at the pointed end means that any particular monomer is conveyed from barbed to pointed end. There is good evidence that such a mechanism actually takes place in cells and is responsible for the protrusion of the leading edge of moving cells as well a type of vesicular transport in cells, called "rocketing".

Microfilaments are controlled by Actin Binding Proteins (ABPs)

Actin is bound by about 48 different types of actin binding proteins. Many of these are present in cells as a range of iso-types with subtly different actin binding properties which makes the situation very complex indeed. Some of the major types are diagrammed below:-

Figure 11 The Actin Binding Proteins

Some ABP cannot simply be classified as belonging to just one of the above broad classes, because they have properties found in more than one category. Villin for example, a member of the gelsolin severing proteins could be described as a severing protein, a capping protein and even a bundling protein since all three properties are displayed under the influence of a variety of second messengers. Indeed a common theme in the ABPs is that their actin binding is modulated or switched by second messengers. Many are calcium sensitive (gelsolin, a-actinin), some (capZ, filamin, gelsolin, profilin) are inhibited by polyphosphoinositides such as PIP₂, and a few (cofilin, EF1a,) are pH dependent.

Microtubules are controlled by Microtubule Associated Proteins (MAPs)

Less is known about these proteins than the ABPs, but many of the known MAPs perform the same type of function as ABPs. Interestingly, some ABPs are also MAPs. EF1a for example bundles microfilaments in a unique parallel, squared configuration and the same protein severs microtubules, this is all in addition to its more traditional function as an elongation factor in translation of mRNA! Four main types of MAPs have so far been characterised. Those which bind MTs and stabilise them (such as MAP2 and Tau), those which nucleate assembly of MTs, MT severing proteins and those which are MT dependent motor proteins. The motor proteins will be discussed later. MT nucleating proteins include another member of the tubulin family - g-tubulin, which is associated with MTOCs.

References:-

General Molecular Biology of the Cell chapter 16, p787. (for third edition)

Microtubules and MAPs McNally, F.J. (1996). Current Opinion in Cell Biology. March 8, 23-29.

MF and MT dynamics Mitchison, T.J. (1992). Molecular Biology of the Cell. 3, 1309-1315.

Please direct any questions to me at:-

Room 444 or lab 446 fourth floor HRB. Tel (0131) 650 3714 or 3712. E-mail SKM@srv4.med.ed.ac.uk