Membrane Cytoskeleton Interactions CS5

The cytoskeleton, especially the actin cytoskeleton interacts with cell membranes. The cell cortex forms a variety of structures such as cell to substrate adhesions, cell to cell adhesions, and large protein aggregations involved is signal transduction. The dual responsibilities of the cell cortex to both provide a physical and signalling communication between cells makes it crucial to cell behaviour including the cancerous process. Many oncogene products are constituents of the cortex. These interactions are structural, signalling, and sometimes to orient the internal cytoskeleton. Microtubules and intermediate filaments also form interactions with membranes, but these are at very specialised structures in certain cell types only. These will be discussed later.

Actin and Actin Binding Protein interactions with Membranes.

The actin-rich cortex of cells lies directly under the plasma-membrane. Although some reports have suggested a direct interaction of actin with membranes, these have been conducted under very artificial circumstances. It is assumed by the majority that these direct interactions do not happen in cells but rather the linkage between the microfilaments and the cell membrane are mediated by specific groups of actin binding proteins. Inhibition of the function of these proteins (by mutagenesis for example) leads to the separation of the actin-rich cortex and the membrane. “Ponticulin”, an actin binding protein from Dictyostelium membranes provides a good example. A variety of actin binding proteins (ABPs) link the cortex with the membrane both directly, and through other protein intermediaries. These interactions are reciprocal, not only do membranous proteins immobilise cytoskeletal domains at adhesions (both cell-substrate and cell-cell), but the cytoskeleton immobilises inter-membranous proteins. Broadly, there are three types of ABP-membrane interactions. ABPs that bind to the surface of membranes by interacting with lipids (see 1 below), those ABPs that are also integral membrane proteins (2), and ABPs that bind to other proteins that are membrane associated (3).

Figure 31. Three modes of interaction typically found between ABPs, and the plasma-membrane. 1) direct association of the ABP with the lipid either covalently or non-covalently. Many ABPs bind to PIP₂ (see later). 2) ABP can be transmembranous. 3) some ABPs associate with other membranous proteins. Each ABP type then links the membrane with the actin rich cortex.

1). ABP-lipid interactions. In 1984, lassing and Linberg, raised a few eyebrows by their discovery that the well documented protein profilin, bound specifically to phosphatindylinositol 4, 5-bisphosphate, (PIP₂) a highly charged phopsholipid found at the cytoplasmic leaflet of the plasma-membrane. This report was quickly followed by others who found that not only did other ABPs bind these lipids, but that in doing so, their actin-binding properties were modulated (e.g. Gelsolin see below).

PIP₂

Gelsolin

Figure 32. Gelsolin is present in inactivated cells on the barbed end of the filament, preventing further polymerization. Upon stimulation, gelsolin is removed from

the filament end by the production of PIP₂. Further actin polymerization can now take place at the fast growing end, driving the cell forwards possibly by the Brownian ratchet mechanism. PIP₂also leads to the dissociation of the profilin :actin, and the cofilin:actin complexes so that more actin would be available for polymerization.

Other ABPs bind to other lipid type non-covalently (e.g. Talin binds to phosphatidylserine) but most lipid binding ABPs bind to PIP₂.

Protein Actin Binding Interaction Effect of PIP₂

Profilin Binds G-actin and a host of other proteins Dissociates porfilin from G-actin

Gelsolin Binds and severs F-actin capping the end Removes gelsolin from the F-actin

a-actinin Bundles F-actin filaments Prevents actin binding

Filamin Gelates filaments forming gels Prevents actin binding

Dystrophin Binds filaments to plasma-membrane Prevents actin binding

ADF/cofilin Binds G-actin and severs filaments Prevents actin binding

Table The effect of PIP₂ on actin binding function of various ABP’s.

A small number of ABPs are also covalently bound to lipids. One such modification is “myristoylation” that being the covalent linkage of the fatty acid myristate (CH₃(CH₂)₁₂COO^-). An example of an ABP that is myristylated is MARKS (which stands for Myrsitolated alanine rich C-kinase substrate). MARKS cross-links actin filaments at the cytoplasmic face of the plasma-membrane but this activity is inhibited by phosphorylation by protein kinase C. Another myristolated ABP is the oncegenic c-Able tyrosine kinase.

Other linkages include geranyl, palmitoyl, farnesyl

2). ABPs that are also integral membrane proteins. Examples of this type include the epidermal growth factor receptor (EGFr), Lymphocyte specific phosphoprotein (LPS1), ponticulin, and some integrins. EGFr is a tyrosine kinase, activated by extracellular ligation by EGF through homotypic dimerization. In the cytoplasmic tail of the receptor is a sequence motif that is similar to a motif in profilin that is responsible for actin-binding. LPS1 is similar but contains a motif that is similar to another abundant actin binding protein caldesmon.

3). ABPs that bind to membrane associated proteins. This seems to be the most common method by which ABPs are connected to the plasma-membrane. ERM proteins bind to a host of membrane associated proteins in addition to binding PIP₂.So ERM proteins are examples of both a type 1 and a type 3 membrane interaction.

Cell Adhesions

Many types of cell to cell adhesions and cell-substrate adhesions exist, each with their particular constellation of actin binding proteins, intermediate filament proteins and other cytoskeletal components. Although these are amongst the earliest cytoplasmic structures discovered, it is just becoming apparent that they communicate with the nucleus regulating gene expression. Their importance has many implications for human conditions such as cancer and skin disease.

	Transmembranous Linker	Connected on outside by:-	Connected on inside by:-	Via:-
Cell-Cell Adherens	Cadherin	Cadherin on other cell	Microfilaments	a & b-catenin, vinculin, a-actinin, plakoglobin.
Cell-Matrix Adherens	Integrin	Extracellular matrix	Microfilaments	Tensin, talin, vinculin, a-actinin
Desmosomes	Cadherin	Cadherin on other cell	Intermediate filaments	Desmoplakins, placoglobins
Hemidesmosomes	Integrin	Extracellular matrix	Intermediate filaments	Desmoplakin-like proteins.

Another type of junction is the gap-junction. These form form specific pores across the membranes of neighboring cells in some tissues. They are particularly important in the lives of plants. Gap junctions, however are not typically associated or regulated by cytoskeletal components and so will not be discussed further here.

Figure 33. The various types of adhesion made by cells to each other and to the extra-cellular matrix. Some involve Intermediate filaments whereas others do not but involve microfilament instead. Not involve microtubules as permanent components.

Cell –Cell Adherens-

Many distinct type of cell-cell adherens exist, but they all share the same overall structure.

Figure 34 Cell-cell Adherens. Cadherins bind to other cadherins (from opposite cells) across the extracellular space. The C-terminus of cadherin binds a complex of cytoskeletal proteins so that forces are transmitted from one cell to the other. Signalling also occurs at this junction

The C-terminus of the caderin binds b-cadherin, which in turn binds a-catenin (a vinculin homolog) that binds to a-actinin, that binds vinculin. Both a-actinin and vinculin bind actin. Cells in tissues are normally in intamate contact with each other, through cell-cell adhesion. Cancerous cells however, often metastasize, that is, they leave their site of origin and invade the surrounding healthy tissue by active locomotion. If they contact blood vessels, they may travel through the blood stream to target distant tissues to invade. The first step in this process is to become detached from the neighbouring cells which means that the cell-cell contacts must be broken. In addition to forming a link in cell-cell adhesion, b-catenin is also part of the Wnt pathway (see CS6, the last lecture).

Cell-Matrix Adherens – Focal Adhesions.

Cells in culture often form focal adhesions (also known as focal contacts). The focal contact is a specialised and discrete region of the plasma membrane that forms molecular contact with the extra-cellular matrix of the underlying substrate. Focal adhesions can be seen most clearly in living cells by Interference Reflection Microscopy (IRM) which involves bouncing light off the ventral surface of the cell to interfere with light reflected off the coverslip that the cell is attached to. The result is that areas close to the glass coverslip appear very dark whereas regions further away appear progressively lighter. At the E.M. level, the focal contact appears as a dense plaque from which microfilaments emanate, tending to point towards the nucleus. Many specialised proteins are found in the focal contact, many of these are structural, but it is increasingly evident that the focal contact is an area of cell signalling. Clustering of integrins at the focal contact leads to tyrosine phosphorylation of p125^FAC, a focal adhesion associated kinase. Some cytoskeletal proteins that are concentrated elsewhere in the cell membrane (e.g. ERM proteins) are excluded from the focal contact.

	a-actinin	talin	tensin	vinculin	paxillin	b-integrin	zyxin	actin
a-actinin	l	m		l		l	l	l
talin	m			l				l
tensin				l				l
vinculin	l	l	l	l	l			l
paxillin				l
b-integrin	l	l
zyxin	l						m
actin	l	l	l	l				l

Major structural protein-protein interactions at the focal contacts (Evidence for, l and against m an interaction).

As cells such as fibroblasts and endothelial cells crawl along they make focal adhesions close to the leading edge but they must break down focal adhesions at the rear of the cell. The turnover of focal contacts is controlled by a cascade of kinases and G-proteins. Rho for example stimulates the formation of actin polymerization at the focal contact and activates myosin light chain kinase leading to contractility and the development of tension.

Figure 35. Desmosomes and Hemi-desmosomes. Cell-matrix adherens. Many other proteins are assoicated with focal contacts. These include signalling proteins; FAK Src another kinase, proteases, utrophin

Desmosomes

It seems very likely that the intermediate filaments exist to provide strength to tissues but not play a major role in the lives of cells as individuals. Desmosomes and hemidesmosomes have been described as being rivet-like, holding cells together within tissues. “Pemphigus” is a serious skin disorder in which the patient makes antibodies to his/her own cadherin (Demoglein, or Desmocilin), this disrupts the continuity of the skin producing blisters and increases the risk of secondary infection considerably.

Muscular Dystrophy

There are many different forms of muscular dystrophy, many of these are due to mutations in the group of proteins that connect the muscle cell membrane to the underlying cytoskeleton. The most common dystrophies (Becker’s, Duchenne’s) is due to mutations in the gene encoding the giant protein Dystrophin. The disease is X-linked and so very much more common in boys rather than girls (males only have one allelic copy of the X-chromosome). The frequency is about 1 in 3500 live male births. Less severe symptoms are described as being Becker’s, more severe Duchenne’s. This protein is encoded by one of the largest human genes so far characterised, this one gene is the same size as a sixth of the entire genome of E.coli! The dystrophin gene is also very complex having very many intron/exon boundaries and this one gene gives rise to very many different proteins through alternate splicing.

Dystrophin is expressed in muscle and nerve tissue (some Duchenne’s patients also have neurological defects), whereas Utrophin, a protein quite like dystrophin is expressed in all tissues. Hopes of a treatment for the disease by for example introduction of the complete gene are dogged by several complications. The protein is very large and so naturally is the encoding cDNA, and as the patient lacks the protein entirely (in many cases), even if the gene was to be successfully placed and producing functional dystrophin, the patient’s immune system would surely make antibodies against this “foreign” protein. It is hoped that the utrophin gene might be used instead to perform the same role as dystrophin in affected individuals. Indeed recent work with the mdx mouse which lacks dystrophin has concluded that a truncated utrophin transgene ameliorates the dystrophic phenotype.

References

Cowin, P. & Burke, B. 1996 Cytoskeleton-membrane interactions. Curr.Op.Cell Biol. 8, 56-65.

Mangeat, P., Roy, C. and Martin, M. (1999) ERM proteins in cell adhesion and membrane dynamics. Trends Cell Biol. 9, 187-192.

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