Cell Locomotion - The Cytoskeleton and Cell Crawling
Anatomy of a crawling vertebrate animal cell - In vertebrates,
only the spermatozoa use undulipodia (flagella) for locomotion, when
somatic cells need to get about they move by amoeboid crawling (though
generally at much lower speeds than most amoebae). Cells that may need
to move about include embryonic cells during development, white blood cells
when fighting infection, cells in growing and regenerating tissues and cells
that maintain tissues, such as fibroblasts (which deposit structural proteins
like collagen and elastin). In pathology, cell movement is important as it
allows cancer cells to (abnormally) travel to other parts of the body. To put a
scale on this diagram: the width of the lamellipod at the front (leading edge)
of the cell is about 3-4 micrometres. A typical animal cell is about 10
micrometres diameter when spherical, but when crawling over a surface they
flatten and spread out to 20 or more micrometres diameter.
References/Bibliography

Christoph Ballestrem, Bernhard Wehrle-Haller, Boris Hinz, and Beat A. Imhof, 2000. Actin-dependent Lamellipodia Formation
and Microtubule-dependent Tail Retraction Control-directed Cell Migration.
Molecular Biology of the Cell 11: 2999–3012.

D. Boal, 2002. Mechanics of the Cell. Cambridge University Press.

Andrey Efimov, Natalia Schiefermeier, Ilya Grigoriev, Ryoma Ohi, Michael C. Brown, Christopher E. Turner, J. Victor Small and
Irina Kaverina, 2008. Paxillin-dependent stimulation of microtubule catastrophes at focal adhesion sites.
Journal of Cell Science
121: 196-204.

Ezratty et al. (2005). Nature Cell Biol. 7: 581-590.

Elaine Fuchs and Iakowos Karakesisoglou, 2001. Bridging cytoskeletal intersections.
Genes Dev. 15: 1-14.

Sabine le Gouvello, Valérie Manceau and André Sobel, 1998. Serine 16 of Stathmin as a Cytosolic Target for
Ca2+/Calmodulin-Dependent Kinase II After CD2 Triggering of Human T Lymphocytes.
J Immunol 161;1113-1122.

Margaret L. Gardel, Ian C. Schneider, Yvonne Aratyn-Schaus, and Clare M. Waterman, 2010.
Mechanical Integration of Actin and Adhesion Dynamics in Cell Migration.
Annu. Rev. Cell Dev. Biol. 26:315–33.

J. Howard, 2001. Mechanics of Motor Proteins and the Cytoskeleton. Sinauer Associates, Inc.

Reina E. Itoh, Etsuko Kiyokawa, Kazuhiro Aoki, Teruko Nishioka, Tetsu Akiyama and Michiyuki Matsuda, 2008. Phosphorylation
and activation of the Rac1 and Cdc42 GEF Asef in A431 cells stimulated by EGF.
Journal of Cell Science 121: 2635-2642.

Keith C, DiPaola M, Maxfield FR, Shelanski ML., 1983. Microinjection of Ca++-calmodulin causes a localized depolymerization of
microtubules.
J Cell Biol. 1983 97:1918-24.

Thomas Kuntziger, Olivier Gavet, Vale´rie Manceau, Andre´ Sobel, and Michel Bornens, 2001. Stathmin/Op18 Phosphorylation
Is Regulated by Microtubule Assembly.
Molecular Biology of the Cell 12: 437–448.

Edward A. Lemke and Jurgen Klingauf, 2005. Single Synaptic Vesicle Tracking in Individual Hippocampal Boutons at Rest and
during Synaptic Activity.
The Journal of Neuroscience, 25:11034 –11044.

Verena Niggli, 2002. Microtubule-disruption-induced and chemotacticpeptide-induced migration of human neutrophils:
implications for differential sets of signalling pathways.
Journal of Cell Science 116: 813-822.

Susanna Rydholm, Gordon Zwartz, Jacob M. Kowalewski, Padideh Kamali-Zare, Thomas Frisk,
and Hjalmar Brismar, 2010. Mechanical properties of primary cilia regulate the response to fluid flow.
Am J Physiol Renal Physiol
298: F1096–F1102.

Jaewon Shim, Sun-Min Lee, Myeong Sup Lee, Joonsun Yoon, Hee-Seok Kweon, and Young-Joon Kim, 2010. Rab35 Mediates
Transport of Cdc42 and Rac1 to the Plasma Membrane during Phagocytosis.
Mol. and Cell. Biol.  30: 1421–1433.

J.V. Small, T. Stradal, E. Vignal and K. Rottner, 2002. The lamellipodium: where motility begins.
TRENDS in Cell Biology, 12:
112-116.

Kelly J. Veale, Carolin Offenhäuser and Rachael Z. Murray, 2011. The role of the recycling endosome in regulating lamellipodia
formation and macrophage migration. Communicative & Integrative Biology 4: 44-47.

Clare M. Waterman-Storer, Rebecca A. Worthylake, Betty P. Liu, Keith Burridge and E.D. Salmon, 1999. Microtubule growth
activates Rac1 to promote lamellipodial protrusion in fibroblasts.
Nature cell Biiology, 1: 45-50.

Torsten Wittmann and Clare M. Waterman-Storer, 2001. Cell motility: can Rho GTPases and microtubules point the way?
Journal of Cell Science 114: 3795-3803.

Jingsong Xu, Fei Wang, Alexandra Van Keymeulen, Maike Rentel, and Henry R. Bourne, 2005. Neutrophil microtubules
suppress polarity and enhance directional migration. PNAS 102: 6884–6889.

Under construction...
When crawling the cell becomes polarised, meaning it has distinct ends or poles. The front end is limited by the leading edge,
which may forms advancing extensions of the cell, called pseudopods. In vertebrates these are typically either flattened
sheets of lamellipodia and/or spikes called filopods. Often filopods (filopodia) accompany the lamellipod, in which case the
name filopod is perhaps misleading as these structures are then very short and they are more like microspikes. The
lamellipodia ruffles when active - it ripples, transiently lifting up off the surface as it does so. Behind the lamellipod, is the
thicker, though still flattened part of the cell, called the lamellum. Behind the lamellum is the rounded hump or cell body
containing the nucleus and at the rear is a tail, which may be very long and tapered, short and broad or almost absent and is
called the uropod. Cells are dynamic and they can change direction either by turning or by reabsorbing the lamellipod and
lamellum and forming a new one along another edge.

The cytokeleton, acting as the supporting framework and muscles of the cell is crucial in cell crawling. Thinner microfilaments
(about 7 nm diameter), especially rods of the protein actin, termed filamentous or
F-actin, but also smaller rods
(minifilaments) of the protein myosin and thicker microtubules (about 22 nm diameter) are mainly involved. (The third
component of the cytoskeleton, the intermediate filaments (about 10 nm diameter) are involved more in supporting and
protecting organelles, like the nucleus, and tissues, like skin/hair.)

The Role of Actin in the Lamellipod

The actin cytoskeleton is the main engine driving the lamellipod. The lamellipod is almost completely free of microtubules but
contains a dense actin network. Most papers state that this actin network is branched, though a detailed study in which a
series of sections were taken (on samples rapidly frozen) and examined under the electron microscope (cryoelectron
microscopy) failed to show any branching (ref?). However, there is a difference between branching and crosslinking and the
filaments are said to be crosslinked. Clearly additional studies are needed to clarify this issue. In our diagram we have
represented the filaments as unbranched (and no cross-links are shown). The actin in this region is so dense that its electric
charges cause the water molecules to align in semi-crystalline fashion (so-called 'ordered water') and the cytoplasm in the
lamellipod partially solidifies into a gel. This actin is thought to be responsible for extending the cell membrane at the front of
the cell, the so-called
leading edge.

Actin filaments have the ability to form or dissolve in seconds whenever they are needed and wherever in the cell they are
needed. The cell keeps a pool of soluble actin monomers (called
G-actin, since the monomers are small globular molecules)
which, when activated, can link together, end-to-end, to form a tubular actin filament (F-actin). The actin filaments are
more-or-less perpendicular to the cell membrane and grow/extend by adding more monomers predominantly at the tip closest
to the cell membrane. This is thought to push the cell membrane forwards as the filaments extend. The current model is a
Brownian Ratchet Model.

Brownian motion is the random movement of molecules, due to their thermal diffusion - specifically it is the movement of
large particles, like soot, in liquid, such as water, as seen under a microscope. For example, if smoke particles in water are
viewed under the microscope then they will be seen to jostle about. The particles are too large to diffuse by such large
amounts of their own accord, but tiny water molecules which are diffusing rapidly collide with the particles. So many collide
every second, that the smoke particles jostle about. Similarly, in a cell, large molecules exhibit diffusive movements, due to
their own diffusion and jostling by water molecules. For example, membranes vibrate according to the amount of thermal
energy they have (which they acquire mainly from colliding water molecules) and this is a type of Brownian motion. This type
of motion in membranes is predicted by mathematical models which determine the most stable configurations for membranes
at different thermal energies. [Interestingly, the description of the bending of these membranes uses essentially the same
mathematics (geometric tensors) as that used to describe bending of spacetime by energy in Einstein's General Theory of
Relativity.] The idea is, that when the membrane vibrates outward, that is away from the end of the actin filaments, more
monomers can stick onto the ends of the filaments to fill the space, preventing the membrane from vibrating back into its
original position and forcing it forwards. By way of this biased or constrained vibration of the membrane, it edges forwards
and the cell moves with it. This is a ratchet mechanism - a ratchet is a mechanical device which allows movement in one
direction only, like a cable tie or a toothed cog passing over a hinged (and spring-loaded) pawl which allows the teeth to slide
past it in one direction only.
Retrograde Actin Flow

As actin filaments extend at the front, pushing out against the membrane at the leading edge, it is clear to understand that
relative to the cell front the actin network as a whole is flowing backwards (retrograde flow) carrying with it anything
embedded in the gel. This is a necessary reaction to new actin polymerising at the tips of actin filaments, causing the older
parts of each filament to move further back behind the tip, until they depolymerise at the rear. Thus we have
actin
treadmilling
- new actin monomers are added to the front of an existing filament, these then flow backwards to be removed
at the rear of the filament.

Role of Microtubules

Microtubules are thicker than actin filaments and more rigid and consist of 'hollow' cylinders which are polymers of a protein
called
tubulin. Like actin, tubulin can alternate rapidly between a polymerised and a dissolved (soluble) monomer state. We
shall tackle the role of microtubules (MT) in cell crawling by asking and attempting to answer a series of questions. To
understand this information, some preliminaries are needed.

The Role of Small GTPases as Molecular Switches

G-proteins are proteins that break-down (by the addition of a water molecule, or hydrolysis) the molecule guanosine
trisphosphate (GTP) into guanosine diphosphate (GDP) and inorganic phosphate and so are also called GTPases. GTP is a
nucleotide phosphate similar to ATP (adenosine trisphosphate) and like ATP is also a source of energy. These proteins are
involved in intracellular cell signalling events in passing a signal from one component of a cell to another. One sub-group,
the small GTPases are small and so water soluble proteins that are free to diffuse in the cytosol. Small GTPases act as
molecular switches, specifically binary switches with two states, on and off. Three main small GTPases we will considere
here are: Rac, Ras and Rho. (Each of these is actually a family of closely related GTPases).

The mechanism of these switches is similar for each. Consider
Rho, which switches on tail (uropod) retraction. When Rho
binds to a molecule of its substrate, GTP, forming Rho-GTP, then it is switched on. To flip the switch to off, another protein, a
GTPase-activating protein (GAP) has to activate the GTPase activity of Rho, which will then break down its substrate to
GDP. The GDP remains bound to the Rho, as a Rho-GDP complex, and the switch is off. To flip the switch on, yet another
protein, a
guanine-nucleotide exchange factor (GEF), removes the GDP and replaces it with a GTP, forming Rho-GTP
again. In this way, Rho can cycle between its on and off states as required. This is summarised in the diagram below (Pi is
inorganic phosphate):
The Role of Actin in the Lamellum, Cell Body and Uropod

In the lamellum, the actin may take one of several states, the main state depending on the cell type. It may form a branched
network (
network-contraction arrays) of single actin filaments (loosely associated with the protein myosin) that may be
backed by arcs of actin filaments bundled together (
dorsal arcs) and it may form straight rod-like of actin bundles called
contractile bundles, with myosin bound to them, in which several actin filaments are cross-linked together by a protein
called
alpha-actinin. These contractile bundles are located toward the top (dorsum) of the cell and are called dorsal stress
fibres
. Some cells also have ventral stress fibres, though these seem to be better developed in slow-moving cells which
strongly adhere to the surface, like fibroblasts, and are much reduced in faster-moving cells like neutrophils. These stress
fibres are intimately connected to adhesive spots that form on the underside (ventral surface) of the crawling cell. In order to
move a cell has to grip the surface, to gain traction, as simply extending the frontal membrane is inefficient. These adhesive
spots are transient, as they must shortly detach again as the cell advances. They form in the lamellum, just behind the
lamellipod as
nascent adhesions and mature as they progress to the rear of the cell (which they do as the cell moves on
forward).

Adhesion Dynamics

The adhesion dynamics of these structures follows the following sequence (see review by Gardel et al. 2010) as they mature.
Of course each is fixed in position, attached to the surface, but as the cell advances, each adhesion moves further back
relative to the cell, beginning just behind the lamellipod at stage 1, and detaching somewhere in the cell body or uropod at
stages 3 or 4:

1.
Nascent adhesions

These are the most immature adhesions that form near the leading edge. They are less than 0.25 micrometres in diameter.
Many of these adhesions will disassemble, but those persisting for a minute or so will mature into focal complexes.
Fast-moving cells, like neutrophils, tend to form nascent adhesions that mature no further. The formation of nascent
adhesions is as follows:

i) These develop when sticky proteins called
integrins, which are diffusing in the cell membrane adhere to the surface.
Several of them (
integrin clusters of less than 10 integrin molecules) must cluster together to trigger the formation of a
nascent adhesion.
ii) A protein called
talin then binds to the integrins from inside the cell (integrins span the membrane projecting into and out
of the cell) and this increases the stickiness of the integrins which bind more tightly to the extracellular matrix (ECM). The
ECM is the extracellular materials secreted by cells in tissues, and forms a scaffold for the tissue to which cells can attach.
iii) Talin also binds F-actin and so attached the integrins to actin filaments. If loosely bound, then the integrins possibly
move/slide along the actin filament until they find a suitable attachment site and bind more tightly.
iv) The protein paxillin then binds, which completes formation of the nascent adhesion and recruits more integrins to the site.

2.
Focal complexes

These are 0.5 micrometres in diameter and last about 5 minutes before becoming mature focal adhesions. Their formation is
as follows:

i) alpha-actinin binds, cross-linking actin filaments together to form an actin bundle, attached to the adhesion.
ii) This maturation is driven by increasing tension and can be mediated by rho and
myosin or by external forces. Myosin is a
protein that binds to actin bundles (forming
actomyosin) and force generated between the actin and myosin then causes the
actin bundle to contract as the actin filaments slide past one-another. This cellular structure is developed to an extreme in
muscle cells, which use myosin type I, whereas non-muscle cells, such as our crawling cell, use myosin II. Myosin is an
example of a molecular motor. The contraction of these anchored bundles possibly helps pull the cell forward and may also
help to push the actin network forwards (and ultimately helping to extend the leading edge). However, it is responsible for
pulling the actin filaments backwards, but 10x slower than they polymerise and elongate forwards. This causes older parts of
the actin filaments to move towards the cell body and tail where they depolymerise, allowing recycling of the actin to the
leading edge. These contractile bundles are the ventral stress fibres mentioned earlier. Focal complexes form preferentially
in slower moving cells like fibroblasts.

3.
Focal adhesions

These are 1-5 micrometres in diameter and last about 20 minutes, after which they may disassemble or develop into an even
more mature form called a fibrillar adhesion. Their formation is as follows:

i) The small actomyosin bundles, bound also to the proteins vinculin and zyxin, form mature stress fibres.
ii) A protein called tensin, which acts on integrins bound to the ECM, increases teh stickiness of integrins for fibronectin (a
glycoprotein in the ECM) and inducing the recruitment, phosphorylation and activation of FAK (focal adhesion kinase). FAK is
a kinase, that is an enzyme that phosphorylates its substrate and sets of signalling networks within the cell by
phosphorylating target proteins (paxillin and p130cas).
iii) The tension pulling on the adhesions partially unfolds the protein talin and in this state the binding of vinculin to talin is
increased. Vinculin links F-actin to integrins, and binding to talin activates the vinculin and so strengthens the binding of actin
filaments to the adhesion.

4.
Fibrillar adhesion

These are the most mature and stable form of focal adhesion and are elongated and over 5 micrometres long.

A General Model of Cell Crawling

Thus, our general model of cell crawling can be divided into three steps:

1. Actin polymerisation pushing the leading edge forward to form a lamellipod whilst integrins become engaged (stuck)
loosely and reversibly at first, to the ECM..
2. The polymerisation of actin at the front, depolymerisation at the rear, and contraction of actomyosin, causes the actin
filaments to move backwards relative to the cell, such that actin monomers are added at the front of the filament and
removed at the back - a phenomenon called
treadmilling.
3. The role of mysoin in pushing forwards the actin network is unclear. Actomysoin contraction in the uropod is required to
draw in the tail as the cell advances. However, filaments bound to adhesions may push forwards under the influence of
myosin.

The question remains: how do some free-living amoebas move so fast? Some free-living amoebae do use flagella for
locomotion, but even those that do not can be much faster than animal cells. A typical mammalian neutrophil, for example,
crawls at about 10 micrometres per minute, whilst amoebae can move at 5-10 micrometres per second. Size is one factor,
with many of the faster amoebae being very large cells, 100 or more micrometres in diameter. There is also evidence of a
different mechanism, especially in the larger amoebae. These larger amoebae tend to use blunt, rounded finger-like
pseudopods called
lobopods, though some amoeba form lamellipods or filopods, and some form different pseduopod types
depending on the nature of the substrate they are moving over or through, with lamellipods being more common on flat
surfaces. (Even mammalian cells may form lobopods when crawling through a 3D mesh, however, crawling is most easily
observed on a flat surface). These lobopods probably make more use of contractile actomyosin, with large amoebae
functioning more like muscle cells. In particular, the cytoplasm tends to form an actin gel beneath the cell membrane, possibly
forming a contractile tube of peripheral cytoplasm, called ectoplasm, around the whole cell, and the more fluid endoplasm
can be seen to flow inside this tube. This endoplasm flows to the front of the lobopod where it either solidifes into ectoplasm
or bursts through the ectoplasm in a fountain before gelling (so-called explosive pseudopods which really look as if they are
explosively rupturing at their tips). In the largest amoeboid cells, the
plasmodial slime moulds, this form of movement is taken
to another extreme. In these multinucleate 'cells' (plamodia) tubes of contractile actin-rich ectoplasm, pump endoplasm
forwards (and sometimes backwards) through the cell.

In smaller amoebae, like
Dictyostelium, the mechanism seems more like that seen in animal cells and the sizes and speeds of
these cells are comparable. However, I have seen very fast small amoebae that did not seem to be using flagella (though
flagella can be hard to detect under the light microscope) and I suspect that the full range of mechanisms remains to be
discovered. The picture below illustrates a crawling cell as seen from above:
Filopodia (filopods) or spikes also form at the leading edge and may become considerably developed in cells that lack other
pseudopod type, or else they may form small spikes protruding from the lamellipod. They contain bundles of F-actin,
cross-linked together by a protein called alpha-actinin. [Expand on filopods and microspikes.]
Q.1 Are MTs required for cell crawling?

  • In neutrophils (a type of white blood cell) apparently not, though in some amoebae they are, some amoebae are much
    larger than neutrophils and It may be that microtubules (MT) are required or more important for cell crawling in larger cells,
    like amoebae and fibroblasts. Ballestrem et al. (2000) found that the drug nocodazole causes MT depolymerisation and
    reduces ruffling in B16 melanoma and Melb-a melanoblasts (types of skin cancer cells). (Methods used: GFP-actin, GFP-
    tubulin, time-lapse; kymograph analysis to quantify ruffling (changes in the shape and position of the leading edge)). These
    cell types will undergo crawling when stimulated by a suitable chemical stimulus: PMA (4-phorbol 12-myristate 13-acetate)
    normally induces crawling in B16 cells and SCF (stem cell factor) in Melb-a cells.
  • Transfection with dominant-negative Rac (N17rac, which knocks out the function of Rac) and the PMA signal to B16 cells
    results in NO ruffling - the cells do not crawl as they would with PMA alone, so Rac appears essential for crawling.
  • Transfection with constitutively active Rac increases ruffling (abolishes PMA response which is no longer needed).
  • Rac is essential for ruffling [ Nobes & Hall (1995) Cell 81: 53-62 ].
  • Nocodazole and PMA added to B16 cells induces ruffling, but focal contacts + stress fibres remain stable, no MT form and
    the cell tends to tear-itself apart!
  • Nnocodazole and the SCF signal added to Melb-a cells reduced migration despite lamellipod formation. The cells were
    unable to retract their tails and so could not move.
  • Destruction or stabilisation of MT blocks lamellipod formation (as in fibroblasts) but PMA or SCF induces them.

  • In a study by Xu et al. (2005), MT disruption (e.g. by nocodazole) increased polarisation in neutrophils and HL-60 cells ( a
    white blood cell cancer cell line).
  • These neutrophils migrate at normal speed & orient toward chemotactic source, but they pursue abnormal circuitous paths.
  • In fibroblasts, some amoebae, melanoma cells and others MT disruption induces a reduction in ruffling/lamellipod fromation.

Conclusions from these studies:

1. MT are NOT essential for actual ruffling / lamellipod formation and operation.
2. They are essential for tail retraction and to regulate adhesion.
3. They are, however, ordinarily
required to initiate polarisation - without which meaningful lamellipod function is impossible.
4. Neutrophils (and keratocytes) remain polarised despite MT disruption – these cells do not depend on focal adhesions (FA). In
contrast, fibroblasts are FA forming and do depend on MT for polarisation. This suggests that MT are required for cell
polarisation in cells that form focal adhesions during cell crawling.
5. FA forming cells depend on MT for polarisation.

Q.2 How do MTs affect cell polarisation?

Verkhovsky et al. (1999) found that in discoid fragments of fish skin keratocytes the actin polymerisation and retrograde (in this
case centripetal, towards the cell-fragment centre) flow is radially symmetric. Myosin II also forms radially symmetric ribbons. If one
side is pushed (micromanipulation) that edge retracts as myosin here condenses. This becomes the tail, the remainder of the
margin becomes a crescentic lamellipod. Thus an asymmetry in contracile actomyosin (induced by prodding in this case) seems
sufficient to cause polarisation. Such asymmetry could be brought about by FA and stress fibres. The tendency of MT to detach
FA at the centre/rear/sides of the cell may enable retraction and contraction, inducing polarity.

MT might polarise cells / assist their migration in a number of ways (see review by Wittmann & Waterman-Storer, 2001):

1.
MT serving as tracks: organelles travel along MT in cells, which function as 'monorails', thus MT may be needed to transport
organelles forwards in crawling cells, especially in larger cells. However, a low concentration of nocodazole inhibits MT assembly
but does not disrupt existing MT, yet still inhibits migration. This suggests that the role of MT as tracks may not be their only role,
since existing MT could serve as tracks, at least for a time (until new tracks were needed) but that new MT must be formed. [The
key here is how quickly migration is inhibited - check this]. Kinesin antibodies interfere with migration like MT depolymerisation.

2.
Growing MT could promote lamellipod protrusion. MT regrowth after nocodazole removal results in an increase in ruffling. MT
growth seems to trigger lamellipod formation (Waterman-Storer et al., 1999).

3.
MT locally regulate adhesion and contraction. MT depolymerisation in the rear of the cell increases contractility by triggering
formation of focal adhesions and actin stress fibres. Involves kinesin?

Model: MT assembly near to the leading edge activates Rac1 (a member of the Rac subfamily) which stimulates lamellipod
formation (ruffling). MT disassembly in the cell body activates RhoA which increase formation of actin stress fibres, focal contacts
and cell contraction (tension?).

Q.3 How do MTs guide migrating cells?

By polarising cells - see the answer to Q.1 and Q.2. MT depolymerisation (in some cells, e.g. fibroblasts?) reduces lamellipodial
area and activity and results in a random lamellipod distribution. Small
et al. (2002): reducing membrane-delivery (Golgi
disruption with brefeldin A or by blocking kinesins) induces similar changes to MT depolymerisation. For details of how they
polarise cells see Q.6.

Q.4 Do MT affect the velocity of cell crawling?

The velocity of lamellipod extension is inversely proportional to membrane stiffness or proportional to membrane slack (Veale et
al
., 2011, in macrophages). In a resting spherical cell, the cell surface membrane is folded (for example as tiny finger-like or spike-
like projections called microvilli in neutrophils and as membrane folds or ruffles in macrophages). Of all shapes, the sphere has
the smallest surface area enclosing a given volume, and so when a cell flattens and spreads-out to crawl, it's surface area would
have to increase if it began as a smooth sphere. Initially, additional membrane is supplied by the smoothing of the cell membrane
folds and then by the continued incorporation of
vesicles (spherical containers made of membrane) into the lamellipod tip to
create the membrane slack necessary to maintain membrane extension rates. If vesicle incorporation is blocked, then the cell
gradually slows down and stops as its membrane stiffens and can no longer be extended. Vesicle translocation across the cell,
from the rear, is required for exocytosis at the lamellipod tip, effecting membrane delivery and possibly Rac1 delivery? [Small et
al. (2002)].

Q.5 Are these vesicles transported by MT, actin or both?

Do MT support transport in the lamellipod?
How can vesicles travelling along MT reach the lamellipod tip?
How do vesicles cross the actin-rich lamellipod cortex?

These vesicles reach the lamellipod tip by travelling along the cytoskeletal filaments, which act as 'monorails', along which the
vesicles are carried by
molecular motor proteins. It is generally assumed that this is the role of the MT, though vesicles can
also be transported along actin filaments (different molecular motors are involved in each case). Very few MT penetrate far into
the lamellipod. Some adventurous MT do, but they became bent over at their tips due to the retrograde flow of the actin network,
so vesicles traveling along MT have a problem when they arrive here - how do they get across the lamellipod to the leading edge
membrane? According to a review by Wittmann & Waterman-Storer (2001, review) kinesin (MT motor, see Q.6b) antibodies
interfere with cell crawling, but a low concentration of nocodazole inhibits MT assembly but does not disrupt existing MT yet still
inhibits cell migration. MT regrowth after nocodazole removal causes an increase in ruffling (lamellipod activity).

According to research by Shim
et al. (2010) Rab35 (a Rab family small GTPase) mediates/triggers transport of Cdc42 and Rac1
to the cell-surface membrane during phagocytosis ('cell eating'). Rab GTPases regulate vesicle trafficking (>60 types in
mammalian cells). In
Drosophila Rab35 regulates actin in the cell-surface membrane, filopod and lamellipod formation and is
delivered via MT tracks. However, the MT tracks end about 5-10 um from leading edge.

Hypothesis 3: vesicles travel along MT and translocate to F-actin in the lamellipod.

In neural growth cones (the advancing fronts of grwoing or regenerating nerve cells) a direct interaction between the actin motor
myosin (MyoVA) and a MT motor (KhcU) has been hypothesised [Huang
et al., 1999]. To test this hypothesis we could track
individual vesicles. Techniques like FRAP, FFS can be used, but with these methods it is hard to distinguish individual vesicles. In
FRAP (fluorescence recovery after photobleaching) a small area within a cell loaded with fluorescent probe is illuminated
and then the time taken for fluorescence to recover in this region, by diffusion of fresh probe from outside areas in the cytosol, is
measured. From such measurements it is possible to measure 2D (membranes) or 3D (cytosol) diffusion within the cell and obtain
the diffusion coefficient of whatever substance our fluorescent probe is attached to. For example, we could label vesicles with the
probe and obtain
diffusion coefficients of vesicles (indicating how rapidly they diffuse) in different regions of the cell. Clearly if
the vesicles are translocating along cytoskeltal filaments then the diffusion will be bias in a given direction. For 3D analysis a
confocal microscope can be used. A confocal microscope which can scan different layers within the cell using a laser beam,
essentially illuminating only a 'point' within the cell rather than the whole cell. In this way a confocal microscope becomes a laser-
scanning microscope (see e.g. Lemke & Klingauf, 2005 ). In
FFS (fluorescence fluctuation spectroscopy) a tiny region of a
cell, again loaded with fluorescent probes, is sampled and the fluctuations due to Brownian motion (see diffusion) are measured
and this enables us to estimate the concentration and size of the particles.

Alternatively one could label Rac1 in the vesicles (e.g. with GFP) and observe its association with actin and MT, both of which
could be labeled with fluorescent probes of different colours. We could block the function of vesicle-carrying proteins that use
actin, e.g. MyoVA, and see whether or not the vesicles can reach the lamellipod tip. Interestingly, disrupting the MT has been
shown to result in an increase in actin-based transport to compensate [Niggli, 2002] so both actin and MT do seem to have
shared transport functions.

In neural growth cones, bundles of microtubules (and actin filaments) tarverse the long extending axon, carrying presynaptic
vesicles to the growing tip, whilst some endocytic vesicles move in the opposite direction, from the front back along the axon.
However, these tips have a large region of actin-rich and MT-poor cortex and so it has been hypothesised that endocytic vesicles
moving along actin tracks skip to the MT [Fuchs & Karakesisoglou, 2001]. This would involve the vesicles carried along actin by
the motor protein Myosin VI (MyoVI) being transfered to the MT motor protein dynein (as part of a synactin/dynein complex).
Similarly, pre-synaptic vesicles would have to transfer in the opposite direction. Direct links between MT and actin filaments have
been observed, such that one actin filament links to the (+) end of each MT (1:1), raising the possibility of such direct cargo
transfers from MT to actin and vice versa. Some +TIPs proteins (proteins that track to the +ends of MT) also bind F-actin, e.g.
CLASP.

External link - Vesicle tracking:
http://youtu.be/CDWoOOd_VHo

Q.6 How do MT polarise cells?

Q.6b Can MT regulate activation of Rac1 induced by external signals, e.g. GF (growth factors)?

Hepatocyte GF (HGF) activates Rac1 in human breast cancer cells (MDA-MB-23 cell-line) and this increases WAVE2 delivery (to
the leading edge) in vesicles. (Takahashi & Suzuki, 2008). Rac1 is associated with the heavy-chain of
kinesin (determined by
immunoblot assays). The WAVE2 protein increases lamellipod activity/formation, MT growth, actin polymerisation and membrane-
targeting of proteins. The WAVE group of proteins (members of the WASP protein family) are activated by Rac and are involved
in organising the actin cytoskeleton.

Kinesin is a molecular motor that carries cargo along MT. MT grow from the microtubule organising centre (MTOC) of the cell,
which in animal cells is the centrosome which contains a pair of centrioles. MT are polar, the two ends of a MT are chemically
distinct. One end, called the +end (plus-end), is furthest from the centriole at the cell periphery, the other -end (minus-end) is
closest to the centrosome. Molecular motors travel along MT in one direction only. Kinesins usually travel towards the +end. Each
kinesin molecule is composed of two heavy polypeptide chains (KHC) bound to two light chains (KLC). Each KHC has a globular
head, so each kinesin has two heads (or 'feet') that bind to the MT and walk along it in an alternating step-like fashion. Attached
to each head on the KHC is a stem, the two stems coil together and each attaches at its tail end to a KLC. The KLCs attach to
cargo that the molecular motors carry along the MT tracks. Thus, MT are involved in delivering Rac1 to the leading edge (though
experiments suggets that blocking MTs increases transport along the actin filaments, which rely on different molecular motors,
such as the myosin protein-family, so transport of Rac1 along actin also seems likely).

This covers the transport of Rac1 to the leading edge, but what about activating Rac1? To look for an answer we can examine
MT dynamics - remember that MT can grow and shrink, polymerise and dissolve. Lamella MT exhibit more growth (fewer pauses,
shrinkages, catastrophes). A catastrophe is when the MT suddenly dissolves (depolymerises). The reasons why MT can
depolymerise so catastrophically will be examined below. However, this increased stability of MT in the lamella may be important
(other than for providing transport tracks).

It has been hypothesised that growing MT promote cell polarisation / lamellipod protrusion. MT regrowth during recovery from
nocodazole treatment increases ruffling and MT growth seems to trigger lamellipod formation (Waterman-Storer
et al., 1999).

What causes this reduction in catastrophe frequency? Research, summarised in the diagram below, reveals a possible answer.
The protein stathmin (S) sequesters (binds to and stores) tubulin dimers (T2), which are then not available for polymerisation and
MT formation. Rac1, bound to GTP in its ON state, increases phosphorylation of PAK-1, a kinase enzyme which phosphorylates
stathmin (at the serine-16 residue). In its phosphorylated state, stathmin releases its sequestered tubulin which increases the
pool of free tubulin, promoting MT polymerisation. This will in turn promote the switching on of Rac1, so a positive-feedback circuit
promotes rapid MT polymerisation. The increasing activation of Rac will also increase lamellipod formation, which coupled with the
symmetry induced by the MT system, causes the cell to polarise.
The level of Rho can also be regulated by another protein called a GDP-dissociation inhibitor (GDI) which binds to Rho in its
off state, as Rho-GDP, and sequesters it, preventing it from being switched on again until it releases it. When switched on,
Rho will trigger the contractility of the actomyosin bundles which bring about tail retraction, and also triggers peeling away of
the adhesions, so that the rear of the cell detached from the surface and retracts.

Rac is another GTPase molecular switch and triggers lamellipod formation at the front of the cell. Thus, to some extent Rho
and Rac are antagonistic since Rho is switched on at the rear and Rac switched on at the front. Ras is also switched on at
the front and triggers the extension of microspikes/filopods.

Methods Used to Study the Cytoskeleton

Fluorescence microscopy. The array of tools and techniques used to study the cytokeleton is quite amazing. Some of
the main methods used will be mentioned here. The cytokeletal filaments are mostly too fine to be seen directly with the light
microscope (bundles of microtubules are visible in some cells such as ciliates like
Euplotes). One way around this is to tag
the filaments with a flourescent protein and then view them with a
fluorescent microscope. In the most commonly
encountered form of fluorescence, a molecule or atom absorbs one photon and then shortly afterwards releases some of
this energy as a photon of longer (redder) wavelength (shorter frequency or lower energy). Typically, an ultraviolet photon
is absorbed and then a visible photon emitted. Fluorescence is a form of luminescence (emission of light). Different
flourescent tags have been designed to flouresce at different wavelengths (colours). Originally a
green fluorescent
protein (GFP)
isolated from a jellyfish was used, but now modified forms exist which fluoresce in a variety of colours
(external link:
https://en.wikipedia.org/wiki/File:FPbeachTsien.jpg) allowing more than one to be used to label different
features within a cell (for example actin could be labeled red and microtubules blue). These fluorescent proteins can be
chemically bonded to actin or tubulin, forming e.g. GFP-actin or GFP-tubulin, causing different parts of the cytoskeleton to
fluoresce. Care has to be taken, since both the light used to observe cells and the activity of flourescent markers can
change cell behaviour and the changes induced should be understood. As most crawling cells are slow moving, time-lapse
photomicrography must be used. This involves periodically illuminating the cells at pre-set intervals and recording them with
a (digital) camera. This also reduces the photobleaching of the fluorescent markers - they lose their ability to flouresce after
a few minutes of illumination. Many flourescent microscopes are equipped with
environmental cabinets - enclosing
heated perspex boxes which can be kept at the desired temperature (such as 37C for human cells).

Kymograph analysis. A kymograph is a device that plots spatial position over time. Classically they were revolving drums
covered in a sheet of paper with a stylus to leave a track of the recording, e.g. of blood pressure, on the drum as it rotated.
These days, digital cameras can feed a sequence of images to image analysis software (a process which is often laborious
since the software is to-date poor at object recognition and requires some manual input in a semi-automated process). A
graph can then be plotted showing, for example, the position of the leading edge over time.

Genetic Engineering. There are several aspects to this, but a common approach is to transfect cells with new genetic
information. For example, it is possible to add mutant DNA of a gene that affects or replaces the function of the healthy
copies of the gene in the cell. The new DNA may be
dominant, meaning that it will override any healthy gene copies
already in the cell, and it may be
negative, meaning that it abolishes the normal function of the gene. It is also possible to
permanently switch a gene on by transfecting with a faulty gene that is always being transcribed into mRNA and is said to be
constitutively expressed. Another technique is gene knockdown, in which the normal expression (transcription and
translation) of a gene is reduced. For example, a sequence of RNA that is complementary to the mRNA of the target gene
may be added to a cell (by microinjection) which will prevent the mRNA being translated into the protein by ribosomes, and
after some time the function of the gene becomes impaired. These small RNAs that are injected into cells are called
small-interfering RNAs (siRNA). (See also ribosomes).

Cell Lines. Many studies on basic cell biology make use of cancer or tumour cell lines. These are cells that have ceased
behaving normally and become immortalised. These cells are active and it is easy to induce them to crawl and they
constantly reproduce by cell division, making them easy to culture in the lab. The use of cell-lines also helps standardise
research, since many labs can work on the same cell line with the same genetic make-up (though cell-lines may change
over time). The use of cell-lines also makes it easier to justify research funding - many agencies will not offer forth funds for
bioscience research unless they think it is tackling an important disease, especially a disease that is of more concern in
wealthy nations. Many scientists accept that the most crucial issue is not studying cancer,
per se, important though that is,
but studying cells to understand how they work (which will impact on all diseases and nanotechnology and many other
endeavours). Unfortunately, the people who supply the funding can seldom see the bigger or more esoteric picture. We
don't like politics on Cronodon, but I feel this is a very important issue as I believe that the current funding arrangements are
having a sub-optimal impact on research. The moral is:
never underestimate the value of discovering knowledge, for until
you know what that knowledge is how can you ascertain its worth
. Let's have more basic cell biology research please.
Critics of the use of cell-lines will point-out that they often behave differently from normal healthy cells, and some
researchers prefer to use fresh healthy cells, such as white blood cells. In truth, both have their place and so long as
scientists are aware of the differences and characteristics of each cell-line this should not pose problems.

Some more sophisticated and specific methods will be mentioned were applicable, below.

The role of microtubules (MT) in cell crawling is not well understood and so we will ask some questions and refer to results
of experiments in the published literature to attempt to answer these questions, raise further questions, and formulate a
model. This is a good example of how science works for the science student!
However, Pak1 phosphorylation is not sufficient in vivo (Wittman et al., 2004). It is known that calcium ions bound to the calcium
sensor
calmodulin, CaM (a protein) or Ca2+ / CaMKII activates stathmin serine-16 phosphorylation in neuronal cells and T cells
[ Ohkawa
et al., 2007]. Calmodulin is a calcium sensor and is triggered by an increase in cytosolic calcium-ion concentration,
either by an influx of calcium-ions across the cell membrane from the intracellular tissue fluid, or from internal stores with the cell
(part of the endoplasmic reticulum). We formulate a hypothesis:


Hypothesis 1: Ca2+/CaM phosphorylates stathmin to help drive cell polarisation

Evidence for hypothesis 1. Calcium-ion pulses originate from the endocytic cycle. That is, as a result of membrane-recycling,
especially important in cell migration, in which membrane is endocytosed from the rear of the cell, transported across the
MT/actin network and exocytosed at the leading edge, causes small spikes or transient increases in calcium-ion concentration in
the cytosol. [ Oberleither
et al., 1993, using MDCK cells]. In crawling vascular SMC (smooth muscle cells) calcium-ion waves
travel from the rear of the cell to the front and control actomyosin contractility [ Espinosa-Tanguma
et al., 2011 ]. Recall that this
contractility is important in regulating tail retraction and adhesion. We should not be surprised by the possible involvement of
calcium, calcium-ions act as intracellular messages for a very large variety of cell processes, including, for example, ciliary-
locomotion in
Paramecium. It is reasonable to assume that calcium is vital in regulating cell crawling.

How could we test hypothesis 1? We could use a CaM inhibitor (e.g. KN-93, R24571) combined with calcium imaging to assess
the levels of calcium-ions in the cytosol, e.g. using the fluorescent calcium-indicator fura-2 loaded into cells (and fluorescent
microscopy) and a chelator (BAPTA) to bind the calcium and prevent activation of CaM. However, research has shown that a
PKC inhibitor blocks stathmin phosphorylation, but R24571 does not [Duraj
et al., 1995 in pre-B leukemia cells]. (PKC or
phosphokinase C is another kinase).

Q.6c How does MT polymerisation activate Rac1?

A model has been proposed, in which polymerisation of MT activates Rac1 via a Rac-specific GEF (gunaine-nucleotide exchange
factor).
Asef (a Rac1-specific GEF) switches activity on binding to APC [Wittmann & Waterman-Storer, 2001 review].


Hypothesis 2: Asef is required for MT-mediated Rac1 activation.

Recall that a GEF is a small protein that switches on a small GTPase switch. When activated, Asef switches on Rac1. APC is a
protein which binds MT directly and also indirectly (via the protein EB1). APC moves along MT to +ends, it is a
+TIPs (MT +end
tracking protein)
. Recall that the +end of MT are situated at the cell periphery, and so Rac1 activators accumulate at the
+ends (+tips) of stable MT, at the cell periphery (and as we shall see later, specifically at the front pole of the cell).

Has anybody blocked Asef, e.g. with antibodies or siRNA in this context? Itoh
et al. (2008) conducted a siRNA Asef knockdown (in
A431 cells) and concluded that
Asef and other GEFs activate Rac1. Thus, in addition to delivering Rac1 to the lamellipod, the
MT may also deliver factors that activate Rac1, switching on lamellipod formation at the cell periphery. We shall see later that the
MT dynamics at the cell rear are very different, so MTs are thus able to direct the positions of lamellipod and tail formation.

Q.7 Do MT affect retrograde flow in lamellipodia?

Summarising Waterman-Storer & Salmon (1999): They appear to be passively effected by it. MTs that enter the lamellipod
become bent parallel to the leading edge and the tubulin moves backwards as MT growth continues to maintain its position.
These MTs may break and depolymerise. This retrograde flow is thought to be due to actomyosin contraction as the free minus-
ends depolymerise near the cell centre or where detached? Retrograde flow speeds (about 4 um/min) match the backward flow
of membrane particles and lipids (though some proteins flow forwards).

Q.8 How does oscillatory locomotion/migration affect MT dynamics?

Neutrophils, like Dictyostelium amoebae, respond to chemoattractant waves by oscillatory behaviour – pushing forward with a
lamellipod and then depolarising on the back of a wave-crest (producing a number of lamellipodia in random directions). This
could be useful to study changes in MT dynamics.

Q.9 What are the mechanisms of retrograde flow?

Cramer (?) showed that retrograde flow requires an intact actin cytoskeleton. Dorsal and ventral rates differ in fibroblasts. Hence
this flow could be due to:
1. Actin-based motors.
2. Actomyosin contraction.
3. Mechanical stresses in a cross-linked actin network due to loss of actin filaments at the rear of the lamellipod.
4. Tension-driven surface lipid-flow: lipids flow retrograde (e.g. as seen in chick neurites) relative to the substrate. This is
probably driven by exocytosis at the lamellipod tip displacing the membrane rearwards.

Receptor plasma-membrane proteins travel forwards to the front of the cell. These are surface-membrane proteins that are
involved in sensing the environment and move forwards through the membrane (and so against the bulk flow of phospholipids) to
reach the lamellipod so that the cell can react to stimuli with more sensitivity.
Nobody seems to have considered actin movement along MTs. Could it be that actin filaments move along MT and so help push
the leading-edge mebrane forwards?

Salmon
et al. (2008) using newt-lung endothelial cells demonstrated that:
  • F-actin flows backwards rapidly in the lamellipod, slower in the lamellum; and flows anterograde in the cell body; no flow
    occured at the lamellum-cell body junction (where the velocities change direction).
  • MTs move with the same velocity as F-actin in the lamellum and cell body; grow along F-actin bundles; quiescent MTs
    move with F-actin bundles.
Thus, F-actin primarily drives the flow but in turn drives the movement of MT and any cargo attached to them.

Q.10 Are MTs implicated in tail retraction?

Indirectly, by detaching FAs. The MTs are polarised in polarised cells: those running toward the leading edge (lamellal MTs)
exhibit more growth and fewer shrinkage/pause events. (But oscillate on reaching the base of the lamellipod, Waterman-Storer &
Salmon, 1999). MTs approaching FAs are prone to more frequent catastrophes.

Q.11 What role do MT play in cell adhesion?

MT disassemble focal adhesions (FA), resulting in their detachment from the surface. This process is FAK (focal adhesion
kinase) and dynamin (a GTPase) dependent [Ezratty
et al., 2005 ]. MT growing toward FA display increasing catastrophes and
increases pauses/shrinkages at their distal ends (+ends pointing to the rear of the cell behind the centrosome).

In fibroblasts – MT depolymerisation activates
Rho-GTP, a small GTPase molecular switch which turns on tail retraction.
What might be causing these catastrophic depolymerisations at the MT ends? (What are the catastrophe factors?). This could
be due to cap removal near the FA. These caps ordinarily cover the +ends of Mt and stabilise them. Stathmin?

Interestingly, there are spikes in intracellular calcium-ion levels upon FA peeling, possibly due to actomyosin bundles pulling on
the FA (or surrounding sites) and opening stretch-activated calcium channels in the cell-surface membrane [Lee
et al., 1999].
Thus, detachment of the rear cell-margin is accompanied by increases in intracellular calcium levels. What further active roles
these calcium spikes might play is uncertain.

Efimov & Kaverina (2009): FAs account for ~5% of cell area but ~40% of catastrophes in ~90% of MTs approaching FAs
(probability of catastrophe increases 7x). In other words, MT growing towards FA experience massive increases in catastrophe
frequency at the ends nearest the FA. MTs captured proximally (attached to centrosome at their (-) end and so still extending
from here) undergo repeated catastrophes at their distal (+) end followed by intervals of rescue from catastrophe, where paxillin
phosphorylation is higher.
Paxillin is an adaptor protein that is thought to bind to integrins from inside the cell, to the portion of
the integrin molecule that extends inside the cell (the cytosolic domain). It appears to have a diverse set of functions, depending
on the level of phosphorylation, and further binds to signaling molecules, including kinase enezymes like FAK, and to structural
proteins in the FA (like vinculin).

Thus we have the idea of MT growing backwards towards the FA and their tips periodically dissolving (depolymerising) and
stabilising as they approach the FA.

Hypothesis 4: the release of MT-associated factors during the catastrophes triggers phosphorylation of paxillin
which when phosphorylated may dock with the MT stabilising it again.

The precise role of paxillin in regulating the adhesion process is complex and I have not yet teased out the important facts from
the research literature. However, phosphorylation of paxillin appears to be a crucial event for several possible reasons. [Expand
on this?]

Hypothesis 5: Stathmin de-phosphorylation in response to Ca2+ triggers the catastrophes.

Evidence for hypothesis 5:
  • Phosphorylation of stathmin reduces catastrophe frequency [ Kuntziger et al., 2001 ].
  • Stathmin ser-16 is phosphorylated by Ca2+/calmodulin-dependent kinase II in human T cells [ Gouvello et al., 1998 ].

Or do calcium waves themselves inhibit MT catastrophes? The fact that calcium-activated calmodulin activates a kinase enzyme
which phosphorylates stathmin, at least in T cells, contradicts our hypothesis. The role of calcium-ions and the generation of
calcium spikes at peeling FA is not clear, but I have a hunch that it will prove of crucial importance.

Ways to test hypothesis 5:
  • Calcium-ion microinjection
  • But: Microinjection of Ca2+/calmodulin or calmodulin then Ca2+ causes MT disruption in fibroblasts [ Keith et al., 1983. J.
    Cell Biol. ]

Q. 12 What role does the primary cilium play in cell crawling?

Most mammalian cells possess a primary cilium, and sometimes several, per cell. These are immotile, sensory antennae with a
(9+0) arrangement of MT doublets instead of the (9+2) arrangement seen in motile
undulipodia. Similar (9+0) arrangements are
seen in various sensory cilia throughout the animal kingdom. The primary cilium of a typical mammalian cell is short and not an
obvious feature in electron microscopy and for a long time was simply dismissed as an evolutionary relic (suggesting vertebrate
evolution from a line descended from colonial flagellates).

The structure of a motile and a primary cilium, as seen in cross-section, is illustrated below. In the primary cilium the nexin
bridges are missing; dynein arms are also missing, and they are said to have few radial spokes (one published EM section shows
none.) Not shown are intra-ciliary transport motors which form possible, but irregular, connections between the MT doublets and
the plasma membrane.
Summary

Cell crawling is regulated by a series of small GTPase molecular switches. Rac activates extension of the lamellipod and ruffling,
Rho controls molecular circuits concerned with adhesion and tail retraction and Cdc42 for controlling the formation of filopods
and microspikes.

Although the primary motive forces are thought to be due to actin polymerisation, extending the leading-edge membrane
forwards in a Brownian ratchet mechanism, and by the retraction of actomyosin bundles to brink up the cell-rear (pushing), the
microtubules play a crucial part in steering many cell types by polarising the cell. This they achieve in part by regulating Rac
activation. MT also play a role in recycling membrane, guiding vesicles to the leading edge, ensuring rapid migration.
According to the literature, the primary cilia project from the dorsal surface of crawling cells and curve forwards, e.g. they are
oriented toward the leading edge in 3t3 fibroblasts, pointing forwards. Since the primary cilium incorporates the mother centriole
(from which the ciliary basal body is derived) the primary cilium may be involved in cell polarisation, or perhaps its orientation is
dictated by cell polarisation. This is because the orinetation of the centrosome dictates the polarity of the MT network and hence
the polarity of the cell. It has also been suggested that when cells are crawling through 3D meshes, as is the norm in tissue
matrices, that the primary cilium may get snagged on the 3D matrix and that this may bend the cilium, activating stretch-sensitive
calcium-ion channels in the base of the primary cilium, resulting in calcium spikes and a mechanosensory response [Rydholm
et
al
., 2010].

In kidney tubule cells the primary cilium acts as a flow mechanosensor, allowing the cells to measure the flow of filtrate through
the kidney nephron. In fibroblasts, the primary cilium contains membrane receptors for growth factors (GF), e.g. the platelet-
derived growth-factor (PDGF-AA) ligand binds to the PDGFalpha receptor, triggering cytoskeletal rearrangement and cell
division (mitosis). Thus the primary cilium may serve a function in maintaining the cell in its differentiated (non-dividing) state.
[Note: this makes the primary cilium of prime importance in cancer studies, which demonstrates what i said earlier - it is more
important to understand how the cell works.]

Hypothesis 6: the primary cilium is a mechanotransducer in migrating cells (fibroblasts). (This maybe necessary to
maintain migration/crawling?)

Testing hypothesis 6:
  • Micromanipulation (deflecting the primary cilium) and measuring the resulting calcium currents;
  • Mechanotransduction to-date is best studied in the nematode Caenorhabditis elegans and involves mec genes and MT;
  • Using insensitive mutants for comparison;
  • Search for mec-homologues? E.g. MEC-4, MEC-10 stretch-activated calcium-ion channels similar to ENaC – 2
    transmembrane domains;
  • siRNA knock-downs, e.g. to remove the function of mec-homologues;
  • In the neamtode the function of these mechanotransducers is blocked by amilonide, so we could try testing the effects of
    amilonide on cell crawling using a migration assay.
  • But: stretch-activated channels are very diverse, so is the nematode model applicable?
  • Introducing identified mechanosensory genes into Xenopus oocytes is a method that has been used to study the
    nematode system and could be extended to the mammalian system.
cell spreading
For more information on the role of the actin
cytoskeleton in cell crawling see:
lamellipod.
migrating cell model
migrating cell model
A model of a migrating cell, as seen
from above (left), and moving to the
right, and advancing towards the
viewer (bottom).

The nucleus (in magenta) is near the
raised uropod at the left; cytoskeletal
rods (could be MTs or actin bundles)
shown in red with vesicles migrating to
the leading edge on the right.

Click images to enlarge...