Functional architecture of biomembranes
Early microscopists who were contemplating the marvelous diversity of life forms
have noticed a slim line delimiting cells, but actually membranes are not directly
visible under a microscope ~ what we see is a difference in optical density or
refractive index between the interior and exterior of the cell.
In 1900, Ernst Overton, professor of pharmacology at Zürich, found that lipid
soluble compounds show the fastest penetration into cells, water penetrates
more slowly and ions even slower, concluding that the cell membrane contains
lipids. Irwing Langmuir, a renowned physical chemist, professor in New Jersey
and the founder of surface chemistry, describes the structure of lipids as bimodal
molecules ~ amphiphilic or amphipathic ~ composed of a hydrophilic head and a
hydrophobic tail.
In 1925, Görter and Grendel, professors at the University of Leyden, perform a
crucial experiment: they extract the lipids from erythrocyte ghosts, an easily
obtainable preparation, with acetone, and dispose them in a monolayer, following
Langmuir’s method. Measuring the surface area of this monolayer, they inferred
it is double compared to the surface of erythrocytes, therefore the logical
conclusion is that the membrane should be a lipid bilayer. Their experiment
contained two mistakes, which fortunately compensated each other: first,
acetone does not extract all lipids, and second, they considered a smaller
erythrocyte surface (100 m2 instead of 140 m2).
In 1936, J.F. Danielli achieves at Buffalo another experiment: he obtained a
lecithin bilayer and measured its surface tension, finding it is approximately one
order of magnitude smaller than the surface tension of cell membranes. By
adding a protein to the bilayer (a solution of ovalbumin), the surface tension
raises and becomes equal to that of the membrane. Therefore, Danielli
developed the sandwich model of cell membrane, adding two protein layers,
one on each side of the lipid bilayer (Fig. 1).
Fig. 1 The sandwich model of membrane proposed by
J.F. Danielli in 1936: the phospholipid bilayer is covered on
both sides by proteins.
In 1955-1958, Jay David Robertson, an American researcher established at
University College, London, used electron microscopy to study the myelin sheath
secreted by Schwann cells by wrapping around their own membrane. He
distinguished three layers: a middle electron transparent layer, which appears
brighter on microphotographs, and two dense layers (an inner and an outer layer)
which appear in black. Robertson’s theory, unit membrane, states that all
membranes in all cells have the same thickness ( 75 Å) and the same trilamellar
structure. Robertson’s main mistake was a too categorical formulation of his
theory. Antirobertsonists made also a mistake during the 1960s-1970s,
presenting strange models without too much success.
In 1972, Jonathan Singer and G. Nicholson published their successful model in
two journals simultaneously: Nature (London) and Science (Washington). The
Singer-Nicholson model, fluid mosaic, compares the membrane with a sea
where proteins float like icebergs.
The lipid bilayer is a bidimensional asymmetrical heterogeneous fluid. Lipids
represent the ideal material to build a barrier between two aqueous
environments, because polysaccharides cannot be packed too tight, while
proteins would cost too much. Thus, the properties of lipids render them the ideal
material. Lipids self-assemble spontaneously in aqueous environments in
bilayers, just for thermodynamic reasons, to achieve a stable conformation.
The lipid bilayer resembles a liquid, where molecules perform limited motions of
four types:
1. flexions of the arms
2. rotation around their own axis
3. lateral translations at high velocities (10-7 s)
4. flip-flop (passing from one monolayer into the other) ~ these motions are
thermodinamically quite improbable, they occur seldom (once at several
hours, days, weeks, one year)
There are 5 main types of phospholipids: lecithine (phosphatidylcholine),
phosphatidylethanolamine,
phosphatidyserine,
phosphatidylinositol,
sphyngomyelin (having as alcohol sphyngosine). The bilayer is asymmetrical:
lecithine is the main component of the outer bilayer, phosphatidylethanolamine
and phosphatidylinositol of the inner bilayer. This asymmetry is maintained by the
low frequency of flip-flop transitions. It ensures a vectorial, oriented insertion of
transmembrane proteins. Fluidity of the bilayer is important for cell functions (e.g.
for cell division), therefore there is a system of fluidity control, there are fluidity
diseases and remedies aimed to cure them.
An important physical factor in fluidity control is temperature. Chemical factors
are intrinsic, like cholesterol or the degree of fatty acids unsaturation, or extrinsic,
physiological or pharmacological. Among physiological factors there are several
hormones. An ideea of Julius Axelrod was that cathecolamines stimulate
methyltransferases, transforming phosphatidylethanolamine into lecithine, and
thus increasing membrane fluidity. General anaesthesia functions by
hyperfluidifying cell membranes, but in 1983 it was noticed that at high
temperatures general anaesthesia functions less satisfactorily.
Membrane proteins can be divided in two cathegories, extrinsic or peripheral
and intrinsic or integral, according to several criteria. Extrinsic proteins interact
only with the hydrophylic heads of phospholipid bilayers, they are hydrophilic,
water-soluble, and can be easily extracted by rinsing membranes with saline
solutions. In contrast, intrinsic proteins present hydrophobic domains in the
membrane crossing regions, they are amphiphilic, and can be extracted form
membrane preparations with detergents, like Triton X-100. Their functions are
extremely diverse: receptors for various chemical signals, enzymes, channels
and transporters, recognition molecules involved in intercellular interactions and
immunity (for example, the major histocompatibility complex class I and II
antigens).
The contemporary representation of the cell membrane completed the SingerNicholson model with two components: the glycocalix and the cytoskeleton.
The glycocalix is the outermost layer of the cell membrane. It is formed by
oligosaccharides (2 to 15 monomers) attached to the ectodomain of membrane
proteins or on the polar head of membrane phospholipids. On average, 1 of 10
phospholipids in the outer layer is glycosylated with a single polysaccharide, and
all transmembrane proteins have several polysaccharides attached to their
ectodomain (their weight can reach up to one third of the molecular weight of
such a protein). These oligosaccharides are generally branched and built up of
only 9 different types of monomers:
- galactose
- N-acetyl-galactosamine
- manose
- fucose
- glucose
- N-acetyl-glucosamine
- NANA (N-acetyl-neuraminic acid) or sialic acid
and, with lower frequencies:
- arabinose
- xylulose.
Protein glycosilation can be stopped by the antibiotic tunicamycin. Sialic acid is
always located in terminal positions, and, being negatively charged, creates a
negative surface charge of the cell, while glucose is never placed in a terminal
position. Manose is included only in the polysaccharides attached to proteins, not
to phospholipids. The branching of oligosaccharidic trees is always dichotomic.
The glycocalix creates an extracellular microenvironment which may play a role
of selectivity filter, based on size and electric charge. Besides, it is important in
the phenomenon of cellular recognition. For example, the main blood group
system (ABO), is represented by cell surface antigens having in the terminal
position N-acetylgalactosamine for group A and galactosamine for group B.
Therefore, via an enzymatic treatment that removes these terminal
monosaccharides, the eritrocytes of any blood group can be transformed in
group O eritrocytes.
Fig. 2 The erythrocyte cytoskeleton formed by spectrin dimers joined by bands
4.1, 4.9 and actin. Band 4.2 and ankyrin ensure connection with transmembrane
protein band 3. Functional proteins, like glycaraldehide 3-phosphate
dehydrogenase and hemoglobin are attached to the cytoskeletal network.
Protein-attached polysaccharides forming the glycocalix are present on the
extracellular side.
The membrane-associated cytoskeleton is an irregular three-dimensional
network connected to the cytoskeleton of the cell. Its average thickness is 5-10
nm, and its structure is that of a colloidal crystal. The molecular composition of
the cytoskeleton can be studied with the SDS-PAGE method. Cell membrane
preparations, e.g. erythrocyte ghosts (the cytoskeleton is abundant in
erythrocytes) are treated with sodium dodecyl sulphate (SDS), which creates a
layer of negative charge and then subjected to polyacrylamide gel
electrophoresis, thus separating 15 distinct bands. The first two bands represent
the  and  subunits of spectrin, a protein identified in the 1970s by Marchesi, a
student of George Emil Palade. The third band is a chloride-bicarbonate anionic
channel, the 4.1 band contains two subunits (a and b), and the fifth band
contains actin, the monomer of microfilaments. The second band contains
several sub-bands: 2.1 - ankyrin, and 2.2-2.6 - syndeins.
The architecture of the erythrocyte membrane-associated cytoskeleton is
represented in Fig 2. The  and  spectrin subunits create a fibrillar network of
dimers, which are connected at their ends to actin microfilaments by 4.1 a and b.
Ankyrin anchors the spectrin network to band 3 dimers or tetramers, thus
ensuring the specific shape of erythrocyte membranes, which accounts for the
special properties of flexibility and rigidity of these cells.
Genetic diseases where spectrin synthesis is stopped lead to hereditary
spherocytosis, characterized by fragile erythrocytes and haemolythic anemia. Its
frequency is 1:5000 in the white (caucasian) USA population. Other erythrocyte
cytoskeletal disorders involve the 4.1 band. An imperfect interaction between 4.1
and spectrin leads to poikylocytosis