Chapter 4 - The Cytoskeletal Model
The Cytoskeletal Model
The cytoskeleton is cellular structure composed largely of muscle type proteins. The theories of motility that are drawn from this structure are discussed and their inadequacy explained. The reasons they continue to be taken seriously is also discussed.
4.1 Introduction to Chapters 4-7
4.2 Recognition-Response and Entrainment Models
4.3 What is the Cytoskeleton?
4.4 The Cytoskeleton is not Precisely Arranged
4.5 What are the Axioms of the Cytoskeletal Model?
4.6 Predictions of the Cytoskeletal Model
4.7 The Original Cytoskeletal Model should be Rejected
4.8 Why is the Cytoskeletal Model taken Seriously?
4.9 The Cytoskeletal Model Today
Summary
4.1 Introduction to Chapters 4-7
Chapters 4 - 7 together briefly review the scientific details of the models proposed to explain capping and particle movement. Chapter 4 discusses the original cytoskeletal model while Chapters 5, & 6 review the flow and cytoskeletal flow models respectively and lists arguments against these ideas. "A Habit of Lies" supports the wave model, presented in Chapter 7, and summarises evidence supportive of it.
This section, shows how possible explanations of capping can be classified into recognition-response and entrainment mechanisms, a classification which is helpful as the two have fundamentally different implications. A recognition-response model requires the cell first to recognise the patch or particle and then respond to its presence by moving it. In a recognition-response model, whatever activity the cell engages in to move objects, it does so only after the membrane comes into contact with the object and in response to that contact. The alternative type of model is an entrainment model. Here the model proposes some ongoing cellular process, the particle or patch becomes caught up in this process and moves because it entrains or becomes associated with it.
The original form of the cytoskeletal model is a recognition- response model. The three models discussed in succeeding chapters are entrainment models, proposing that the cell is doing something all the time. The movement of surface particles and patches is a purely adventitious consequence of the ongoing activity.
"Cytoskeleton" means cell skeleton but, in fact, amoeboid cells have no rigid skeleton. In reality, the cytoskeleton certainly is not a cell skeleton, though this erroneous belief seems widely held and is encouraged by an inappropriate name. The confusion over its function can be gauged by the fact that Chapter 6 discusses the even less fitting, but still widely believed, idea of the "cytoskeleton" flowing beneath the membrane.
The cytoskeleton was first treated as a discrete system, in the early 1970's, although knowledge of some of its components is older. Following its discovery workers began to look for phenomena it could be causing. Capping was also newly discovered and the hope was that here was a convenient handle to grasp the functions of the newly discovered structure. Incidentally, the cell has a second organelle similar to the cytoskeleton based on a different pair of proteins, tubulin and dynein. However, drugs are available to disrupt the two structures separately. Disruption of the actin based cytoskeleton normally disrupts capping and accordingly that is the structure focused on.
The cytoskeleton is a matrix or scaffold of fibrous proteins in the cytoplasm of a cell. Two major proteins of the cytoskeleton are actin and myosin, proteins most noted for being found in muscle where they form a regular array that can shorten itself, contracting the muscle - hence they are contractile proteins. Muscle contraction takes place in response to a nerve impulse triggering a series of events, including a sharp rise in muscle calcium concentration. Other proteins are found associated with the cytoskeleton, some will have functions there, while others will merely stick adventitiously to the actin and myosin. Since it is made from muscle protein, it seems better to regard the cytoskeleton as the cell's muscle rather than its skeleton and some other name really should be chosen. Cytosinew would more sensibly describe its likely functions. However, this text will employ the commonly used word cytoskeleton, to ensure the reader knows what is being discussed. As the cell's muscle, the cytoskeleton would be expected to play a major role in motility and this much of its function is undisputed.
It is true that the cytoskeleton determines cell shape but it does so not by functioning as a skeleton but as a muscle. Just as the body shape of many invertebrate organisms is determined by their muscles, so the shape of an amoeboid cell is determined by its cytosinew. It is possible to disrupt cells and extract component organelles, including fibres of cytoskeleton. Such in vitro extracts can be shown to have some biochemical activities but, like all isolated organelles, the cytoskeleton does little on its own and its activity is much impaired. For its full functioning the cytoskeleton needs to be in its natural environment, the cell, where it depends upon other organelles. In particular, it is well established that its activity depends upon calcium ion concentration, which is in turn regulated by the permeability to calcium of the cell membrane. This, in its turn, is controlled by the electrical potential difference across the membrane.
All biological activity of actin and myosin, the chief cytoskeletal components, are membrane controlled this way. Even in muscle, extensions of the cell membrane, called sarcoplasmic reticulum, pass between the bundles of actin and myosin filaments which comprise muscle. This internal muscle membrane is heavily supplied with proteins whose function is to control calcium flux, thereby regulating the contractile state of the muscle. Thus, for normal function, the actomyosin cytoskeleton operates in harmony with a cell membrane. In particular, it is very clear that capping and particle movement on cells certainly does involve the membrane. Since they occur on the surface of the cell, it could hardly be otherwise. It is equally clear that the cytoskeleton, the cell's muscle, provides the driving force for the movements involved.
To this author, it seems clear that the cytoskeleton will be controlled by the membrane during capping, just as muscle is at other times. However, the various models which go by the name cytoskeletal model assign all control to the cytoskeleton and downplay any role for the membrane, relegating it to a purely passive role. Viewed in terms of evolution and Occam's razor, this is a highly dubious step to take. Bayesian hypothesis testing (section 2.24) should give such ideas a low antecedent probability because they abandon the controlling role the membrane plays elsewhere. This abandonment groups a series of otherwise dissimilar models together under the same title, the cytoskeletal model. Two such models are described in some detail in chapters 4 and 6, while another two are mentioned very briefly in the final chapter. The use of the same name for what are essentially dissimilar concepts causes additional problems which will be mentioned later in this chapter and in chapter 6.
Our own skeleton is a precise arrangement of bones, everyone has a tibia and a fibula, connected at the knee to the femur etc. The skeleton gives us our recognisable and, within limits, stable shape. Our muscles, or those of invertebrate animals, are also precise in this way. However, the cytoskeleton is not precise like this, it is a chaotic, continuously changing arrangement of filaments - and so it must be as the shape of the cell itself continuously changes. This chaotic character leads to a deeper distinction between entrainment and recognition-response models and another reason to reject the original cytoskeletal model. The interior of the cell should not be seen as a piece of machinery, in the manner of a car engine, with each component precisely located, but as a loose assortment of molecular structures.
An unrealistic implication of recognition-response models is that the motile organelles of the cell are themselves arranged precisely. They imply that some subtle guiding hand places each filament of the cytoskeleton in an ordained location. It is true some individual components of the cell are machine like in their nature - enzymes, ribosomes, the various filaments etc. but it does not follow that they group together in a precise arrangement. Motor cars are machines that work together in a sort of harmony on the roads, though their occupants neither know nor care about the positions or actions of most other cars. The city possesses no central brain guiding the actions of each car, or monitoring their positions. The arrangement of cars on a road system is chaotic. So it is with the cytoskeleton, each filament has some properties of a machine but there is only a chaotic ordering of them into the matrix.
To reiterate, there is no doubt the cytoskeleton powers motility but that does not mean the cell needs to plan and organise every detail of its arrangement. The arrangement of cytoskeletal filaments is not like the arrangement of wires on a microcircuit chip, or the components in a motor car and the cytoskeleton is not a machine. The precise arrangement of filaments is adopted is irrelevant, at that level, the structure is chaotic.
The cell changes shape all the time and the cytoskeletal arrangement changes with it, both causing the changes and responding to them. Cells can be pushed and prodded, even breached with microprobes without killing them. Two cells can be fused together. It is even possible to take the nucleus from one cell and replace it with another. All these things are well demonstrated techniques - all would smash to ribbons an intricate, precise cytoskeletal arrangement. Even so this wrong idea of precise circuitry has never been abandoned and, though rarely stated in so many words, is implicitly assumed by much discussion on cell movement and morphology.
Answering this question is not nearly so easy as might be imagined. The traditional image of scientific model building with clearly stated postulates and consequential predictions has not been followed. No single paper states the axioms of the cytoskeletal model. Equally, and for the same reasons, it is difficult to list its predictions, though the next section attempts the task.
Broadly the original idea was that the interior of a cell, its cytoskeleton, would recognise the presence of a patch or particle on its surface. How the recognition is achieved goes unspecified but, having performed it, the cytoskeleton would catch the object and reel it in on a fibre much as an angler might land a fish. Later it was discovered that actin and myosin could accumulate beneath caps and patches and this was interpreted as very strong support for the cytoskeletal model.
The model proposes that the cytoskeleton controls the position of patches and particles much as a puppeteer controls the position of a marionette, pulling it one way or the other in response to the plan of action formed in his own mind. By implicit extension, the cytoskeleton was assumed to be controlling the position not only of surface patches and particles but also itself and the positions of intracellular organelles too. In other words this original cytoskeletal model asserts that the whole cell is organised in response to some prior plan present somewhere in the cell and expressed in the cytoskeleton. The cytoskeleton is taken to be precisely arranged and so capable of information processing. This implication of information recording, and probably of information processing, makes these ideas very questionable.
For some, this picture has greatly changed with time and this model has been transformed into the later cytoskeletal flow model. "A Habit of Lies" uses the term "cytoskeletal model", to refer to these two ideas generically. The "original cytoskeletal model" will refer to the model described in this chapter though, occasionally, to bring out the information processing implications of the model, it may be described as the "puppet" or "marionette" model.
If the cytoskeletal model were correct it would be expected that:-
- Transmembrane interactions between patch, or particle, should be a necessary condition of capping or particle movement. In other words an object should not cap or move unless it actually can make contact with the cytoskeleton by going right across the membrane to the inside of the cell.
- Some specificity in capping or particle movement might be expected. There are many types of protein in the cell membrane - some would be expected to make the required contact while others would not. Thus some receptors would cap well, others not at all.
- The cytoskeletal model does not obviously predict any size dependency in capping. (For example, it is not clear why a transmembrane protein, making the necessary contacts, should cap only after being patched.)
All three predictions are testable and have been tested. None of them are experimentally satisfied. Just as damaging for the original cytoskeletal model are its information processing implications. Any recognition-response model implies information processing capabilities in the cell - the original cytoskeletal model implies that the cytoskeleton can calculate what it should do next, that it is preprogrammed with the appropriate response to a new situation - detecting a patch or particle. Proposing a recognition-response model is quite different from proposing an entrainment model because of these implications of recognition and response.
Arguments based on evolution require that, for any recognition- response model to be true, capping and particle movement must be cell functions, activities which help the cell survive or reproduce. Consider the following arguments :-
- If the cell recognises a patch or particle on its surface then it must possess sensory devices to do this. Such devices evolve in the natural world because living things derive a competitive benefit from them. Pursuing the puppet on a string analogy, these sensors would be analogous to the eyes of the puppeteer.
- If the cell selects out a patch or particle and moves it, in preference to other components on the cell surface, it must possess mechanical devices and proteins whose job it is to do this (analogous to the puppetmaster's hands and fingers). Again, such devices could arise only if they conferred on their possessors some competitive advantage.
- Given that these mechanical devices and proteins must have evolved over time in response to selective pressures, it follows that capping and particle movement, are themselves cellular functions. Some as yet unidentified benefit is conferred upon the cell by being able to move patches or pieces of carbon around on its surface.
This simply does not seem to be the case. Not even advocates of the cytoskeletal model make any suggestion about why the cell should want to cap or move adherent particles. It seems the correct description of capping will be come via an entrainment model.
Many workers now seem to concentrate on two of the entrainment models. Even so the original cytoskeletal model with all its elaborate implications is not explicitly rejected. Rather it has become transformed by imperceptible degrees into the cytoskeletal flow model. Forms of it still persist during discussions of cell motility which treat cytoskeletal filaments as like railway lines with engines of myosin running along them.
One "strong" argument was taken by many to be a virtual proof of the cytoskeletal idea. Cap formation was seen to be associated with the movement of actin and myosin of the cytoskeleton into the same region of the cell as the cap, (see for example Koch and Smith (1978)). In the systems studied, an association between the cap outside the cell and the cytoskeleton within the cell evidently occurs. They demonstrated that capping could occur for membrane components capable of associating with the cytoskeleton. However, they did not demonstrate this contact to be a necessary condition for capping; other papers show capping occurs even when such association seems impossible. For example, there are observations of lipids capping, and of proteins capping even though they do not traverse the membrane (Schroit and Pagano (1981), Révész and Greaves (1975)). In short, contact between the patch and the cytoskeleton is not a necessary condition for capping.
Such observations seem to disprove the original cytoskeletal model and instead entrainment models are adopted. These escape the requirement for a recognition of the patch or particle; the cell need not change its ongoing behaviour because it does not respond to a particle or patch.
These arguments summarise the grounds for rejecting the cytoskeletal model but others are listed in the literature, often by Bretscher. The criticisms of the cytoskeletal model are valid and many workers seem to reject the cytoskeletal model in its original form - but not in so many words. Some retain the original cytoskeletal model by ridding it of any predictive value, saying simply that the cytoskeleton has something to do with capping but not specifying what. Others reject the model by axiomatic drift, its postulates being gradually altered, a change made possible because they were never stated precisely. So marked is the drift that today's description of the cytoskeletal model is quite different from earlier ones. The difference is so marked as to form a separate model, the cytoskeletal flow model, discussed in Chapter 6, but first the original competitor to the cytoskeletal model, the membrane flow model, will be described.
This Chapter has :-
- Explained the classification of models into recognition- response and entrainment types and noted that recognition- response implies the cell is organised, or arranged according to some plan.
- Introduced the cytoskeleton and described how its function is normally linked to the activity of the membrane.
- Described how the cytoskeleton was thought to be involved in capping and particle movement and classified this model as a recognition-response.
- Rejected both the original cytoskeletal model and all other recognition-response models.
- Noted how, in the literature, the original cytoskeletal model has not been rejected but progressively altered.
© Copyright John A Hewitt.
For contact information, see copyright page.
Last Modified 22 October 2005