CMB CH4

JWCL151_ch04_117-172.
qxd 6/25/09 1:24 PM Page 117
4
The Structure and Function
of the Plasma Membrane
4.1
4.2
An Overview of Membrane Functions
A Brief History of Studies
on Plasma Membrane Structure
4.3 The Chemical Composition
T he outer walls of a house or car provide a strong, inflexible barrier that
protects its human inhabitants from an unpredictable and harsh external
world. You might expect the outer boundary of a living cell to be constructed
of an equally tough and impenetrable barrier because it must also protect its delicate
internal contents from a nonliving, and often inhospitable, environment. Yet cells
of Membranes are separated from the external world by a thin, fragile structure called the plasma
4.4 The Structure and Functions membrane that is only 5 to 10 nm wide. It would require about five thousand plasma
of Membrane Proteins membranes stacked one on top of the other to equal the thickness of a single page of
this book.
4.5 Membrane Lipids
Because it is so thin, no hint of the plasma membrane is detected when a section of
and Membrane Fluidity a cell is examined under a light microscope. In fact, it wasnt until the late 1950s that
4.6 The Dynamic Nature techniques for preparing and staining tissue had progressed to the point where the
of the Plasma Membrane plasma membrane could be resolved in the electron microscope. These early electron
4.7 The Movement of Substances
micrographs, such as those taken by J. D. Robertson of Duke University (Figure 4.1a),
portrayed the plasma membrane as a three-layered structure, consisting of darkly
Across Cell Membranes
staining inner and outer layers and lightly staining middle layer. All membranes that
4.8 Membrane Potentials were examined closelywhether they were plasma, nuclear, or cytoplasmic membranes
and Nerve Impulses (Figure 4.1b), or taken from plants, animals, or microorganismsshowed this same
The Human Perspective: Defects in ultrastructure. In addition to providing a visual image of this critically important cellular
Ion Channels and Transporters as a Cause of structure, these electron micrographs touched off a vigorous debate as to the molecular
Inherited Disease composition of the various layers of a membrane, an argument that went to the very
Experimental Pathways:
A model of a fully hydrated lipid bilayer composed of phosphatidylcholine molecules (each
The Acetylcholine Receptor containing two myristoyl fatty acids) penetrated by a single transmembrane helix composed
of 32 alanine residues. (FROM LIYANG SHEN, DONNA BASSOLINO, AND TERRY STOUCH, BRISTOL-MYERS
SQUIBB RESEARCH INSTITUTE, FROM BIOPHYSICAL JOURNAL, VOL. 73, P. 6, 1997.)
117
JWCL151_ch04_117-172.qxd 6/25/09 1:24 PM Page 118
118 Chapter 4 THE STRUCTURE AND FUNCTION OF THE PLASMA MEMBRANE
FIGURE 4.1 The trilaminar appearance of membranes. (a) Electron micro-

graph showing the three-layered (trilaminar) structure of the plasma mem-
brane of an erythrocyte after staining the tissue with the heavy metal osmium.
Osmium binds preferentially to the polar head groups of the lipid bilayer, pro-
ducing the trilaminar pattern. The arrows denote the inner and outer edges of
the membrane. (b) The outer edge of a differentiated muscle cell grown in cul-
ture showing the similar trilaminar structure of both the plasma membrane
(PM) and the membrane of the sarcoplasmic reticulum (SR), a calcium-storing
compartment of the cytoplasm. (A: COURTESY OF J. D. ROBERTSON; B: FROM AN-
DREW R. MARKS ET AL., J. CELL BIOL. 114:305, 1991; BY COPYRIGHT PERMISSION OF
THE ROCKEFELLER UNIVERSITY PRESS.)
(a) 50 nm
PM
SR
(b)
heart of the subject of membrane structure and function. As tion, their relative positions cannot be stabilized and their
we will see shortly, cell membranes contain a lipid bilayer, and interactions are dependent on random collisions. Because
the two dark-staining layers in the electron micrographs of of their construction, membranes provide the cell with an
Figure 4.1 correspond primarily to the inner and outer polar extensive framework or scaffolding within which compo-
surfaces of the bilayer (equivalent to the yellow atoms of the nents can be ordered for effective interaction.
chapter-opening image). We will return to the structure of 3. Providing a selectively permeable barrier. Membranes
membranes below, but first we will survey some of the major prevent the unrestricted exchange of molecules from one
functions of membranes in the life of a cell (Figure 4.2). side to the other. At the same time, membranes provide
the means of communication between the compartments
they separate. The plasma membrane, which encircles a
4.1 AN OVERVIEW cell, can be compared to a moat around a castle: both
serve as a general barrier, yet both have gated bridges
OF MEMBRANE FUNCTIONS that promote the movement of select elements into and
1. Compartmentalization. Membranes are continuous, un- out of the enclosed living space.
broken sheets and, as such, inevitably enclose compart- 4. Transporting solutes. The plasma membrane contains the
ments. The plasma membrane encloses the contents of machinery for physically transporting substances from
the entire cell, whereas the nuclear and cytoplasmic mem- one side of the membrane to another, often from a region
branes enclose diverse intracellular spaces. The various where the solute is present at low concentration into a re-
membrane-bounded compartments of a cell possess gion where that solute is present at much higher concen-
markedly different contents. Membrane compartmental- tration. The membranes transport machinery allows a cell
ization allows specialized activities to proceed without to accumulate substances, such as sugars and amino acids,
external interference and enables cellular activities to be that are necessary to fuel its metabolism and build its
regulated independently of one another. macromolecules. The plasma membrane is also able to
2. Scaffold for biochemical activities. Membranes not only transport specific ions, thereby establishing ionic gradi-
enclose compartments but are also a distinct compart- ents across itself. This capability is especially critical for
ment themselves. As long as reactants are present in solu- nerve and muscle cells.
4.2 A BRIEF HISTORY OF STUDIES ON PLASMA MEMBRANE STRUCTURE 119
(5) FIGURE 4.2 A summary of membrane functions in a plant cell. (1) An

Hormone example of membrane compartmentalization in which hydrolytic en-
zymes (acid hydrolases) are sequestered within the membrane-bounded
vacuole. (2) An example of the role of cytoplasmic membranes as a site
of enzyme localization. The fixation of CO2 by the plant cell is catalyzed
by an enzyme that is associated with the outer surface of the thylakoid
IP3 membranes of the chloroplasts. (3) An example of the role of mem-
(4) branes as a selectively permeable barrier. Water molecules are able to
Ca2+ penetrate rapidly through the plasma membrane, causing the plant cell
H+
(2) to fill out the available space and exert pressure against its cell wall.
CO2
(4) An example of solute transport. Hydrogen ions, which are produced
+ by various metabolic processes in the cytoplasm, are pumped out of plant
RuBP cells into the extracellular space by a transport protein located in the
PGA plasma membrane. (5) An example of the involvement of a membrane in
the transfer of information from one side to another (signal transduc-
tion). In this case, a hormone (e.g., abscisic acid) binds to the outer sur-
(3) face of the plasma membrane and triggers the release of a chemical
H2O message (such as IP3) into the cytoplasm. In this case, IP3 causes release
of Ca2! ions from a cytoplasmic warehouse. (6) An example of the role
(6) (1)
of membranes in cellcell communication. Openings between adjoin-
Acid hydrolases ing plant cells, called plasmodesmata, allow materials to move directly
from the cytoplasm of one cell into its neighbors. (7) An example of the
role of membranes in energy transduction. The conversion of ADP
to ATP occurs in close association with the inner membrane of the
mitochondrion.
ADP
ATP
(7)
5. Responding to external signals. The plasma membrane when energy in sunlight is absorbed by membrane-bound
plays a critical role in the response of a cell to external pigments, converted into chemical energy, and stored in
stimuli, a process known as signal transduction. Mem- carbohydrates. Membranes are also involved in the trans-
branes possess receptors that combine with specific mole- fer of chemical energy from carbohydrates and fats to
cules (or ligands) having a complementary structure. ATP. In eukaryotes, the machinery for these energy con-
Different types of cells have membranes with different versions is contained within membranes of chloroplasts
receptors and are, therefore, capable of recognizing and and mitochondria.
responding to different ligands in their environment. The
We will concentrate in this chapter on the structure and
interaction of a plasma membrane receptor with an exter-
functions of the plasma membrane, but remember that the
nal ligand may cause the membrane to generate a signal
principles discussed here are common to all cell membranes.
that stimulates or inhibits internal activities. For example,
Specialized aspects of the structure and functions of mito-
signals generated at the plasma membrane may tell a cell
chondrial, chloroplast, cytoplasmic, and nuclear membranes
to manufacture more glycogen, to prepare for cell divi-
will be discussed in Chapters 5, 6, 8, and 12, respectively.
sion, to move toward a higher concentration of a particu-
lar compound, to release calcium from internal stores, or
possibly to commit suicide.
6. Intercellular interaction. Situated at the outer edge of
4.2 A BRIEF HISTORY OF STUDIES
every living cell, the plasma membrane of multicellular ON PLASMA MEMBRANE STRUCTURE
organisms mediates the interactions between a cell and its
The first insights into the chemical nature of the outer bound-
neighbors. The plasma membrane allows cells to recog-
ary layer of a cell were obtained by Ernst Overton of the
nize and signal one another, to adhere when appropriate,
University of Zrich during the 1890s. Overton knew that
and to exchange materials and information.
nonpolar solutes dissolved more readily in nonpolar solvents
7. Energy transduction. Membranes are intimately involved than in polar solvents, and that polar solutes had the opposite
in the processes by which one type of energy is converted solubility. Overton reasoned that a substance entering a cell
to another type (energy transduction). The most funda- from the medium would first have to dissolve in the outer
mental energy transduction occurs during photosynthesis boundary layer of that cell. To test the permeability of the
outer boundary layer, Overton placed plant root hairs into protected from contact with the aqueous environment (Figure
hundreds of different solutions containing a diverse array of 4.3c). Thus, the polar head groups would face the cytoplasm
solutes. He discovered that the more lipid-soluble the solute, on one edge and the blood plasma on the other. Even though
the more rapidly it would enter the root hair cells (see p. 144). Gorter and Grendel made several experimental errors (which
He concluded that the dissolving power of the outer boundary fortuitously canceled one another out), they still arrived at the
layer of the cell matched that of a fatty oil. correct conclusion that membranes contain a lipid bilayer.
The first proposal that cellular membranes might contain In the 1920s and 1930s, cell physiologists obtained evi-
a lipid bilayer was made in 1925 by two Dutch scientists, dence that there must be more to the structure of membranes
E. Gorter and F. Grendel. These researchers extracted the than simply a lipid bilayer. It was found, for example, that lipid
lipid from human red blood cells and measured the amount of solubility was not the sole determining factor as to whether or
surface area the lipid would cover when spread over the sur- not a substance could penetrate the plasma membrane. Simi-
face of water (Figure 4.3a). Since mature mammalian red larly, the surface tensions of membranes were calculated to be
blood cells lack both nuclei and cytoplasmic organelles, the much lower than those of pure lipid structures. This decrease
plasma membrane is the only lipid-containing structure in the in surface tension could be explained by the presence of pro-
cell. Consequently, all of the lipids extracted from the cells can tein in the membrane. In 1935, Hugh Davson and James
be assumed to have resided in the cells plasma membranes. Danielli proposed that the plasma membrane was composed
The ratio of the surface area of water covered by the extracted of a lipid bilayer that was lined on both its inner and outer sur-
lipid to the surface area calculated for the red blood cells from face by a layer of globular proteins. They revised their model
which the lipid was extracted varied between 1.8 to 1 and 2.2 in the early 1950s to account for the selective permeability of
to 1. Gorter and Grendel speculated that the actual ratio was the membranes they had studied. In the revised version (Fig-
2:1 and concluded that the plasma membrane contained a bi- ure 4.4a), Davson and Danielli suggested that, in addition to
molecular layer of lipids, that is, a lipid bilayer (Figure 4.3b). the outer and inner protein layers, the lipid bilayer was also
They also suggested that the polar groups of each molecular
layer (or leaflet) were directed outward toward the aqueous en-
vironment, as shown in Figure 4.3b,c. This would be the ther- R
modynamically favored arrangement, because the polar head O P O

groups of the lipids could interact with surrounding water O

HCH H
molecules, just as the hydrophobic fatty acyl chains would be
HC CH
O O
C O C O
HCH HCH
HCH HCH
HCH HCH
(b)
HCH HCH
HCH HCH
HCH HCH
HCH HCH
CH HCH
CH
HCH
Stationary Movable HCH
barrier barrier HCH
HCH
Lipids HCH
HCH
HCH HCH
HCH HCH
HCH HCH
H H
(a) (c)
FIGURE 4.3 The plasma membrane contains a lipid bilayer. (a) Calcu- and Grendel concluded that red blood cells contained enough lipid to
lating the surface area of a lipid preparation. When a sample of phos- form a layer over their surface that was two molecules thick: a bilayer.
pholipids is dissolved in an organic solvent, such as hexane, and spread (b) As Gorter and Grendel first proposed, the core of a membrane con-
over an aqueous surface, the phospholipid molecules form a layer over tains a bimolecular layer of phospholipids oriented with their water-
the water that is a single molecule thick: a monomolecular layer. The soluble head groups facing the outer surfaces and their hydrophobic fatty
molecules in the layer are oriented with their hydrophilic groups bonded acid tails facing the interior. The structures of the head groups are given
to the surface of the water and their hydrophobic chains directed into in Figure 4.6a. (c) Simulation of a fully hydrated lipid bilayer composed
the air. To estimate the surface area the lipids would cover if they were of the phospholipid phosphatidylcholine. Phospholipid head groups are
part of a membrane, the lipid molecules can be compressed into the orange, water molecules are blue and white, fatty acid chains are green.
smallest possible area by means of movable barriers. Using this type of (C: FROM S.-W. CHIU, TRENDS IN BIOCHEM. SCI. 22:341, 1997, COPYRIGHT
apparatus, which is called a Langmuir trough after its inventor, Gorter 1997, WITH PERMISSION FROM ELSEVIER SCIENCE.)
4.2 A BRIEF HISTORY OF STUDIES ON PLASMA MEMBRANE STRUCTURE 121
penetrated by protein-lined pores, which could provide con- The structure and arrangement of membrane proteins in
duits for polar solutes and ions to enter and exit the cell. the fluid-mosaic model differ from those of previous models
Experiments conducted in the late 1960s led to a new in that they occur as a mosaic of discontinuous particles that
concept of membrane structure, as detailed in the fluid- penetrate the lipid sheet (Figure 4.4b). Most importantly, the
mosaic model proposed in 1972 by S. Jonathan Singer and fluid-mosaic model presents cellular membranes as dynamic
Garth Nicolson of the University of California, San Diego structures in which the components are mobile and capable of
(Figure 4.4b). In the fluid-mosaic model, which has served as coming together to engage in various types of transient or
the central dogma of membrane biology for more than three semipermanent interactions. In the following sections, we
decades, the lipid bilayer remains the core of the membrane, will examine some of the evidence used to formulate and sup-
but attention is focused on the physical state of the lipid. Un- port this dynamic portrait of membrane structure and look
like previous models, the bilayer of a fluid-mosaic membrane at some of the recent data that bring the model up to date
is present in a fluid state, and individual lipid molecules can (Figure 4.4c).
move laterally within the plane of the membrane.
Protein molecule Polar pore
Lipid molecule
(b)
(a)
Oligosaccharide Glycoproteins
Glycolipid
Integral
Hydrophobic proteins
helix Cholesterol
Peripheral
protein Phospholipid
(c)
FIGURE 4.4 A brief history of the structure of the plasma membrane. external surface of most membrane proteins, as well as a small percent-
(a) A revised 1954 version of the Davson-Danielli model showing the age of the phospholipids, contain short chains of sugars, making them
lipid bilayer, which is lined on both surfaces by a monomolecular layer of glycoproteins and glycolipids. Those portions of the polypeptide chains
proteins that extends through the membrane to form protein-lined that extend through the lipid bilayer typically occur as ! helices com-
pores. (b) The fluid-mosaic model of membrane structure as initially posed of hydrophobic amino acids. The two leaflets of the bilayer con-
proposed by Singer and Nicolson in 1972. Unlike previous models, the tain different types of lipids as indicated by the differently colored head
proteins penetrate the lipid bilayer. Although the original model shown groups. The outer leaflet may contain microdomains (rafts) consisting
here depicted a protein that was only partially embedded in the bilayer, of clusters of specific lipid species. (A: FROM J. F. DANIELLI, COLLSTON
lipid-penetrating proteins that have been studied span the entire bilayer. PAPERS 7:8, 1954; B: REPRINTED WITH PERMISSION FROM S. J. SINGER AND G.
(c) A current representation of the plasma membrane showing the L. NICOLSON, SCIENCE 175:720, 1972; COPYRIGHT 1972, AMERICAN
same basic organization as that proposed by Singer and Nicolson. The ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE.)
REVIEW
?
1. Describe some of the important roles of membranes in
the life of a eukaryotic cell. What do you think might
be the effect of a membrane that was incapable of per-
Myelin
forming one or another of these roles? sheath
2. Summarize some of the major steps leading to the current
{
model of membrane structure. How does each new model
retain certain basic principles of earlier models?
Axon
4.3 THE CHEMICAL COMPOSITION

OF MEMBRANES
Membranes are lipidprotein assemblies in which the compo-
nents are held together in a thin sheet by noncovalent bonds.
As noted above, the core of the membrane consists of a sheet
of lipids arranged in a bimolecular layer (Figure 4.3b,c). The
lipid bilayer serves primarily as a structural backbone of the
membrane and provides the barrier that prevents random
movements of water-soluble materials into and out of the cell. FIGURE 4.5 The myelin sheath. Electron micrograph of a nerve cell axon
The proteins of the membrane, on the other hand, carry out surrounded by a myelin sheath consisting of concentric layers of plasma
membrane that have an extremely low protein/lipid ratio. The myelin
most of the specific functions summarized in Figure 4.2. Each
sheath insulates the nerve cell from the surrounding environment, which
type of differentiated cell contains a unique complement of increases the velocity at which impulses can travel along the axon (dis-
membrane proteins, which contributes to the specialized ac- cussed on page 162). The perfect spacing between the layers is main-
tivities of that cell type (see Figure 4.32d for an example). tained by interlocking protein molecules (called P0) that project from
The ratio of lipid to protein in a membrane varies, depend- each membrane. (FROM LEONARD NAPOLITANO, FRANCIS LEBARON, AND
ing on the type of cellular membrane (plasma vs. endoplasmic JOSEPH SCALETTI, J. CELL BIOL. 34:820, 1967; BY COPYRIGHT PERMISSION
reticulum vs. Golgi), the type of organism (bacterium vs. plant OF THE ROCKEFELLER UNIVERSITY PRESS.)
vs. animal), and the type of cell (cartilage vs. muscle vs. liver).
For example, the inner mitochondrial membrane has a very
high ratio of protein/lipid in comparison to the red blood cell
plasma membrane, which is high in comparison to the mem- triglycerides, which have three fatty acids (page 47) and are
branes of the myelin sheath that form a multilayered wrapping not amphipathic, membrane glycerides are diglyceridesonly
around a nerve cell (Figure 4.5). To a large degree, these differ- two of the hydroxyl groups of the glycerol are esterified to
ences can be correlated with the basic functions of these mem- fatty acids; the third is esterified to a hydrophilic phosphate
branes. The inner mitochondrial membrane contains the group. Without any additional substitutions beyond the phos-
protein carriers of the electron-transport chain, and relative to phate and the two fatty acyl chains, the molecule is called
other membranes, lipid is diminished. In contrast, the myelin phosphatidic acid, which is virtually absent in most membranes.
sheath acts primarily as electrical insulation for the nerve cell it Instead, membrane phosphoglycerides have an additional
encloses, a function that is best carried out by a thick lipid layer group linked to the phosphate, most commonly either choline
of high electrical resistance with a minimal content of protein. (forming phosphatidylcholine, PC), ethanolamine (forming
Membranes also contain carbohydrates, which are attached to phosphatidylethanolamine, PE), serine (forming phosphatidylser-
the lipids and proteins as indicated in Figure 4.4c. ine, PS), or inositol (forming phosphatidylinositol, PI). Each of
these groups is small and hydrophilic and, together with the
negatively charged phosphate to which it is attached, forms a
Membrane Lipids
highly water-soluble domain at one end of the molecule,
Membranes contain a wide diversity of lipids, all of which are called the head group. At physiologic pH, the head groups of
amphipathic; that is, they contain both hydrophilic and hy- PS and PI have an overall negative charge, whereas those of
drophobic regions. There are three main types of membrane PC and PE are neutral. In contrast, the fatty acyl chains are
lipids: phosphoglycerides, sphingolipids, and cholesterol. hydrophobic, unbranched hydrocarbons approximately 16 to
22 carbons in length (Figure 4.6). A membrane fatty acid may
Phosphoglycerides Most membrane lipids contain a be fully saturated (i.e., lack double bonds), monounsaturated
phosphate group, which makes them phospholipids. Because (i.e., possess one double bond), or polyunsaturated (i.e., pos-
most membrane phospholipids are built on a glycerol back- sess more than one double bond). Phosphoglycerides often
bone, they are called phosphoglycerides (Figure 4.6a). Unlike contain one unsaturated and one saturated fatty acyl chain.
4.3 THE CHEMICAL COMPOSITION OF MEMBRANES 123
Recent interest has focused on the apparent health benefits of

two highly unsaturated fatty acids (EPA and DHA) found at
high concentration in fish oil. EPA and DHA contain five and
six double bonds, respectively, and are incorporated primarily
into PE and PC molecules of certain membranes, most no-
tably in the brain and retina. EPA and DHA are described as
omega-3 fatty acids because their last double bond is situated
3 carbons from the omega (CH3) end of the fatty acyl chain.
With fatty acid chains at one end of the molecule and a polar O
head group at the other end, all of the phosphoglycerides ex- H2C O C (CH2)7CH=CH(CH2)7CH3
O
hibit a distinct amphipathic character. HC O C (CH2)7CH=CH(CH2)7CH3
O
Sphingolipids A less abundant class of membrane lipids, Dioleoyl O

P O CH2
phosphatidic acid O
called sphingolipids, are derivatives of sphingosine, an amino
alcohol that contains a long hydrocarbon chain (Figure 4.6b). Phosphatidic
acid H
Sphingolipids consist of sphingosine linked to a fatty acid
(R of Figure 4.6b) by its amino group. This molecule is a Phosphatidyl-
+
O
(CH3)3N CH2 CH2 H2C O C R
ceramide. The various sphingosine-based lipids have addi- choline (lecithin) O
tional groups esterified to the terminal alcohol of the sphingo- + HC O C R'
sine moiety. If the substitution is phosphorylcholine, the Phosphatidyl- H3N CH CH2 O
serine COO

O P O CH2
molecule is sphingomyelin, which is the only phospholipid of O
the membrane that is not built with a glycerol backbone. If the Phosphatidyl- +
substitution is a carbohydrate, the molecule is a glycolipid. If ethanolamine H3N CH2 CH2

(cephalin)
the carbohydrate is a simple sugar, the glycolipid is called a OH OH
cerebroside; if it is a small cluster of sugars, the glycolipid is Phosphatidyl- H
H H
called a ganglioside. Since all sphingolipids have two long, hy- inositol
OH
OH H
H
drophobic hydrocarbon chains at one end and a hydrophilic H OH
region at the other, they are also amphipathic and basically
O
similar in overall structure to the phosphoglycerides. H2C O C R
Glycolipids are interesting membrane components. Rela- O
tively little is known about them, yet tantalizing hints have O

HC O C R'
emerged to suggest they play crucial roles in cell function. The CH2 O P O CH2
nervous system is particularly rich in glycolipids. The myelin Diphosphatidyl-
HO C H
O
glycerol O
sheath pictured in Figure 4.5 contains a high content of a (cardiolipin) CH2 O P O CH2
particular glycolipid, called galactocerebroside (shown in Fig- O
ure 4.6b), which is formed when a galactose is added to cer- HC O C R''
O
amide. Mice lacking the enzyme that carries out this reaction C O C R'''
exhibit severe muscular tremors and eventual paralysis. Simi- (a)
H2
O
larly, humans who are unable to synthesize a particular gan-
glioside (GM3) suffer from a serious neurological disease H H
characterized by severe seizures and blindness. Glycolipids also Sphingosine HO CH2 C C CH CH (CH2)12CH3
NH3 OH
play a role in certain infectious diseases; the toxins that cause +
cholera and botulism both enter their target cell by first bind- H H
ing to cell-surface gangliosides, as does the influenza virus. Ceramide HO CH2 C C CH CH (CH2)12CH3
NH OH
Cholesterol Another lipid component of certain mem- O C R
branes is the sterol cholesterol (see Figure 2.21), which in O H H

+
certain animal cells may constitute up to 50 percent of the Sphingomyelin (CH3)3N CH2 CH2 O P O CH2 C C CH CH (CH2)12CH3

O NH OH
O C R
FIGURE 4.6 The chemical structure of membrane lipids. (a) The struc-
tures of phosphoglycerides (see also Figure 2.22). (b) The structures of A cerebroside Gal Ceramide
sphingolipids. Sphingomyelin is a phospholipid; cerebrosides and gan-

gliosides are glycolipids. A third membrane lipid is cholesterol, which is
shown in the next figure. (R ! fatty acyl chain). [The green portion of A ganglioside GalNAc Gal Glu Ceramide
(GM2)
each lipid, which represents the hydrophobic tail(s) of the molecule, is
SiA
actually much longer than the hydrophilic head group (see Figure 4.23).] (b)
width of about 30 and that each row of head groups (with its
adjacent shell of water molecules) adds another 15 (refer to
the chapter-opening illustration on page 117). Thus, the entire
lipid bilayer is only about 60 (6 nm) thick. The presence in
membranes of this thin film of amphipathic lipid molecules
has remarkable consequences for cell structure and function.
Because of thermodynamic considerations, the hydrocarbon
chains of the lipid bilayer are never exposed to the surrounding
aqueous solution. Consequently, membranes are never seen to
have a free edge; they are always continuous, unbroken struc-
tures. As a result, membranes form extensive interconnected
networks within the cell. Because of the flexibility of the lipid
bilayer, membranes are deformable and their overall shape can
change, as occurs during locomotion (Figure 4.8a) or cell divi-
sion (Figure 4.8b). The lipid bilayer is thought to facilitate the
regulated fusion or budding of membranes. For example, the
events of secretion, in which cytoplasmic vesicles fuse to the
FIGURE 4.7 The cholesterol molecules (shown in green) of a lipid bi- plasma membrane, or of fertilization, where two cells fuse to
layer are oriented with their small hydrophilic end facing the external
form a single cell (Figure 4.8c), involve processes in which two
surface of the bilayer and the bulk of their structure packed in among the
separate membranes come together to become one continuous
fatty acid tails of the phospholipids. The placement of cholesterol mole-
cules interferes with the flexibility of the lipid hydrocarbon chains, sheet (see Figure 8.32). The importance of the lipid bilayer in
which tends to stiffen the bilayer while maintaining its overall fluidity. maintaining the proper internal composition of a cell, in sepa-
Unlike other lipids of the membrane, cholesterol is often rather evenly rating electric charges across the plasma membrane, and in
distributed between the two layers (leaflets). (REPRINTED FROM H. L. many other cellular activities will be apparent throughout this
SCOTT, CURR. OPIN. STRUCT. BIOL. 12:499, 2002, COPYRIGHT 2002, WITH and subsequent chapters.
PERMISSION FROM ELSEVIER SCIENCE.) Another important feature of the lipid bilayer is its abil-
ity to self-assemble, which can be demonstrated more easily
within a test tube than a living cell. If, for example, a small
amount of phosphatidylcholine is dispersed in an aqueous
lipid molecules in the plasma membrane. Cholesterol is ab- solution, the phospholipid molecules assemble sponta-
sent from the plasma membranes of most plant and all bacte- neously to form the walls of fluid-filled spherical vesicles,
rial cells. Cholesterol molecules are oriented with their small called liposomes. The walls of these liposomes consist of a
hydrophilic hydroxyl group toward the membrane surface and continuous lipid bilayer that is organized in the same
the remainder of the molecule embedded in the lipid bilayer manner as that of the lipid bilayer of a natural membrane.
(Figure 4.7). The hydrophobic rings of a cholesterol molecule Liposomes have proven invaluable in membrane research.
are flat and rigid, and they interfere with the movements of Membrane proteins can be inserted into liposomes and their
the fatty acid tails of the phospholipids (page 134). function studied in a much simpler environment than that of
a natural membrane. Liposomes have also been developed as
The Nature and Importance of the Lipid Bilayer Each vehicles to deliver drugs or DNA molecules within the body.
type of cellular membrane has its own characteristic lipid
composition, differing from one another in the types of lipids,
the nature of the head groups, and the particular species of
fatty acyl chain(s). Because of this structural variability, it is TABLE 4.1 Lipid Compositions of Some Biological Membranes*
estimated that some biological membranes contain hundreds Human Human Beef heart
of chemically distinct species of phospholipid. The role of this Lipid erythrocyte myelin mitochondria E. coli
remarkable diversity of lipid species remains the subject of in-
terest and speculation. Phosphatidic acid 1.5 0.5 0 0
Phosphatidylcholine 19 10 39 0
The percentages of some of the major types of lipids of a
Phosphatidyl-
variety of membranes are given in Table 4.1. The lipids of a ethanolamine 18 20 27 65
membrane are more than simple structural elements; they can Phosphatidylgycerol 0 0 0 18
have important effects on the biological properties of a mem- Phosphatidylserine 8.5 8.5 0.5 0
brane. Lipid composition can determine the physical state of Cardiolipin 0 0 22.5 12
the membrane (page 134) and influence the activity of partic- Sphingomyelin 17.5 8.5 0 0
ular membrane proteins. Membrane lipids also provide the Glycolipids 10 26 0 0
precursors for highly active chemical messengers that regulate Cholesterol 25 26 3 0
cellular function (Section 15.3).
*The values given are weight percent of total lipid.
Various types of measurements indicate that the combined Source: C. Tanford, The Hydrophobic Effect, p. 109, copyright 1980, John Wiley &
fatty acyl chains of both leaflets of the lipid bilayer span a Sons, Inc. Reprinted by permission of John Wiley & Sons, Inc.
4.3 THE CHEMICAL COMPOSITION OF MEMBRANES 125
(a) (b) (c)
FIGURE 4.8 The dynamic properties of the plasma membrane. (a) The this dividing ctenophore egg begins at one pole and moves unidirection-
leading edge of a moving cell often contains sites where the plasma ally through the egg. (c) Membranes are capable of fusing with other
membrane displays undulating ruffles. (b) Division of a cell is accompa- membranes. This sperm and egg are in a stage leading to the fusion of
nied by the deformation of the plasma membrane as it is pulled toward their plasma membranes. (A: COURTESY OF JEAN PAUL REVEL; B: COURTESY
the center of the cell. Unlike most dividing cells, the cleavage furrow of OF GARY FREEMAN; C: COURTESY OF A. L. COLWIN AND L. H. COLWIN.)
The drugs or DNA can be linked to the wall of the liposome so-called stealth liposomes (e.g., Caelyx) that contain an
or contained at high concentration within its lumen (Fig- outer coating of a synthetic polymer that protects the lipo-
ure 4.9). In these studies, the walls of the liposomes are con- somes from immune destruction (Figure 4.9).
structed to contain specific proteins (such as antibodies or
hormones) that allow the liposomes to bind selectively to the
The Asymmetry of Membrane Lipids
surfaces of particular target cells where the drug or DNA is
intended to go. Most of the early clinical studies with lipo- The lipid bilayer consists of two distinct leaflets that have a
somes met with failure because the injected vesicles were distinctly different lipid composition. One line of experiments
rapidly removed by phagocytic cells of the immune system. that has led to this conclusion takes advantage of the fact that
This obstacle has been overcome with the development of lipid-digesting enzymes cannot penetrate the plasma mem-
brane and, consequently, are only able to digest lipids that re-
side in the outer leaflet of the bilayer. If intact human red
blood cells are treated with a lipid-digesting phospholipase,
Protective layer approximately 80 percent of the phosphatidylcholine (PC) of
of polyethylene
Antibody
the membrane is hydrolyzed, but only about 20 percent of
glycol the membranes phosphatidylethanolamine (PE) and less than
10 percent of its phosphatidylserine (PS) are attacked. These
data indicate that, compared to the inner leaflet, the outer
leaflet has a relatively high concentration of PC (and sphin-
gomyelin) and a low concentration of PE and PS (Fig-
ure 4.10). It follows that the lipid bilayer can be thought of as
composed of two more-or-less stable, independent monolay-
ers having different physical and chemical properties.
The different classes of lipids in Figure 4.10 exhibit
different properties. All the glycolipids of the plasma mem-
brane are in the outer leaflet where they often serve as recep-
Drug crystalized Lipid-soluble tors for extracellular ligands. Phosphatidylethanolamine,
in aqueous fluid
Lipid bilayer
drug in bilayer which is concentrated in the inner leaflet, tends to promote
the curvature of the membrane, which is important in mem-
FIGURE 4.9 Liposomes. A schematic diagram of a stealth liposome con- brane budding and fusion. Phosphatidylserine, which is con-
taining a hydrophilic polymer (such as polyethylene glycol) to protect it centrated in the inner leaflet, has a net negative charge at
from destruction by immune cells, antibody molecules that target it
physiologic pH, which makes it a candidate for binding pos-
to specific body tissues, a water-soluble drug enclosed in the fluid-filled
interior chamber, and a lipid-soluble drug in the bilayer.
itively charged lysine and arginine residues, such as those ad-
jacent to the membrane-spanning ! helix of glycophorin A
Exoplasmic Asparagine
25
SM PC PS PE PI Cl
CH2OH
O NH
H O H
H N C CH2 CH
% total membrane lipid
OH H C O
O H
Polypeptide
backbone
0 H NHCOCH3
N-Acetylglucosamine
Serine (X=H)
Threonine (X=CH3)
CH2OH
25 Cytosolic X NH
O
O O CH CH
FIGURE 4.10 The asymmetric distribution of phospholipids (and
cholesterol) in the plasma membrane of human erythrocytes. (SM, OH C O
H H
sphingomyelin; PC, phosphatidylcholine; PS, phosphatidylserine; PE,
phosphatidylethanolamine; PI, phosphatidylinositol; Cl, cholesterol.) H NHCOCH3
N-Acetylgalactosamine
FIGURE 4.11 Two types of linkages that join sugars to a polypep-

in Figure 4.18. The appearance of PS on the outer surface tide chain. The N-glycosidic linkage between asparagine and N-
of aging lymphocytes marks the cells for destruction by acetylglucosamine is more common than the O-glycosidic linkage
macrophages, whereas its appearance on the outer surface of between serine or threonine and N-acetylgalactosamine.
platelets leads to blood coagulation. Phosphatidylinositol,
which is concentrated in the inner leaflet, plays a key role in
the transfer of stimuli from the plasma membrane to the
attached to several different amino acids by two major types of
cytoplasm (Section 15.3).
linkages (Figure 4.11). These carbohydrate projections play an
important role in mediating the interactions of a cell with its
Membrane Carbohydrates environment (Chapter 7) and sorting of membrane proteins
to different cellular compartments (Chapter 8). The carbohy-
The plasma membranes of eukaryotic cells also contain carbo-
drates of the glycolipids of the red blood cell plasma mem-
hydrate. Depending on the species and cell type, the carbohy-
brane determine whether a persons blood type is A, B, AB, or
drate content of the plasma membrane ranges between 2 and
O (Figure 4.12). A person having blood type A has an enzyme
10 percent by weight. More than 90 percent of the mem-
that adds an N-acetylgalactosamine to the end of the chain,
branes carbohydrate is covalently linked to proteins to form
whereas a person with type B blood has an enzyme that adds
glycoproteins; the remaining carbohydrate is covalently linked
galactose to the chain terminus. These two enzymes are en-
to lipids to form glycolipids, which were discussed on page
coded by alternate versions of the same gene, yet they recog-
123. As indicated in Figure 4.4c, all of the carbohydrate of the
nize different substrates. People with AB blood type possess
plasma membrane faces outward into the extracellular space.1
both enzymes, whereas people with O blood type lack en-
The carbohydrate of internal cellular membranes also faces
zymes capable of attaching either terminal sugar. The func-
away from the cytosol (the basis for this orientation is illus-
tion of the ABO blood-group antigens remains a mystery.
trated in Figure 8.14).
The modification of proteins was discussed briefly on
page 53. The addition of carbohydrate, or glycosylation, is the
most complex of these modifications. The carbohydrate of
REVIEW
?
glycoproteins is present as short, branched hydrophilic 1. Draw the basic structure of the major types of lipids found
oligosaccharides, typically having fewer than about 15 sugars in cellular membranes. How do sphingolipids differ from
per chain. In contrast to most high-molecular-weight carbo- glycerolipids? Which lipids are phospholipids? Which are
hydrates (such as glycogen, starch, or cellulose), which are glycolipids? How are these lipids organized into a bilayer?
polymers of a single sugar, the oligosaccharides attached to How is the bilayer important for membrane activities?
membrane proteins and lipids can display considerable vari- 2. What is a liposome? How are liposomes used in medical
ability in composition and structure. Oligosaccharides may be therapies?
3. What is an oligosaccharide? How are they linked to
1
It can be noted that even though phosphatidylinositol contains a sugar group membrane proteins? How are they related to human
(Figure 4.6), it is not considered to be part of the carbohydrate portion of the blood types?
membrane in this discussion.
4.4 THE STRUCTURE AND FUNCTIONS OF MEMBRANE PROTEINS 127
Gal GlcNAc Gal Glu
Fuc
O antigen
GalNAc Gal GlcNAc Gal Glu
Fuc
A antigen
Integral membrane proteins

(a)
Gal Gal GlcNAc Gal Glu
Peripheral membrane protein
Fuc
B antigen
FIGURE 4.12 Blood-group antigens. Whether a person has type A, B,

AB, or O blood is determined by a short chain of sugars covalently at-
tached to membrane lipids and proteins of the red blood cell membrane.
The oligosaccharides attached to membrane lipids (forming a ganglio-
side) that produce the A, B, and O blood types are shown here. A person
with type AB blood has gangliosides with both the A and B structure.
(Gal, galactose; GlcNAc, N-acetylglucosamine; Glu, glucose; Fuc, fu-
cose; GalNAc, N-acetylgalactosamine.)
Peripheral
(b) membrane proteins
4.4 THE STRUCTURE AND FUNCTIONS Etn
OF MEMBRANE PROTEINS GPI- anchored protein

P
I
Man GlcNAc
Depending on the cell type and the particular organelle 0 0
within that cell, a membrane may contain hundreds of dif- 0 Man
Man
0 P
C Etn P 0
ferent proteins. Each membrane protein has a defined orien- P
tation relative to the cytoplasm, so that the properties of one Etn
surface of a membrane are very different from those of the

other surface. This asymmetry is referred to as membrane
sidedness. In the plasma membrane, for example, those
parts of membrane proteins that interact with other cells or
with extracellular substances project outward into the extra-
cellular space, whereas those parts of membrane proteins
that interact with cytoplasmic molecules project into the cy-
tosol. Membrane proteins can be grouped into three distinct Cytoplasm
classes distinguished by the intimacy of their relationship to
the lipid bilayer (Figure 4.13). These are
(c)
1. Integral proteins that penetrate the lipid bilayer. Integral
proteins are transmembrane proteins; that is, they pass FIGURE 4.13 Three classes of membrane protein. (a) Integral proteins
entirely through the lipid bilayer and thus have domains typically contain one or more transmembrane helices (see Figure 5.4 for an
that protrude from both the extracellular and cytoplasmic exception). (b) Peripheral proteins are noncovalently bonded to the polar
sides of the membrane. Some integral proteins have only head groups of the lipid bilayer and/or to an integral membrane protein.
one membrane-spanning segment, whereas others are (c) Lipid-anchored proteins are covalently bonded to a lipid group that
multispanning. Genome-sequencing studies suggest that resides within the membrane. The lipid can be phosphatidylinositol, a
integral proteins constitute 2030 percent of all encoded fatty acid, or a prenyl group (a long-chain hydrocarbon built from five-
proteins. carbon isoprenoid units). I, inositol; GlcNAc, N-acetylglucosamine; Man,
mannose; Etn, ethanolamine; GPI, glycosylphosphatidylinositol.
2. Peripheral proteins that are located entirely outside of
the lipid bilayer, on either the cytoplasmic or extracellular direct contact with surrounding lipid molecules (Figure 4.14).
side, yet are associated with the surface of the membrane Lipid molecules that are closely associated with a membrane
by noncovalent bonds. protein can play an important role in the activity of the
3. Lipid-anchored proteins that are located outside the protein, although the degree to which a particular protein
lipid bilayer, on either the extracellular or cytoplasmic requires specific interactions with particular lipid molecules
surface, but are covalently linked to a lipid molecule that remains unclear.
is situated within the bilayer. Those portions of an integral membrane protein that
project into either the cytoplasm or extracellular space tend to
be more like the globular proteins discussed in Section 2.5.
Integral Membrane Proteins These nonembedded domains tend to have hydrophilic
Most integral membrane proteins function in the following surfaces that interact with water-soluble substances (low-
capacities: as receptors that bind specific substances at the molecular-weight substrates, hormones, and other proteins) at
membrane surface, as channels or transporters involved in the the edge of the membrane. Several large families of membrane
movement of ions and solutes across the membrane, or as proteins contain an interior channel that provides an aqueous
agents that transfer electrons during the processes of photo- passageway through the lipid bilayer. The linings of these
synthesis and respiration. Like the phospholipids of the bi- channels typically contain key hydrophilic residues at strategic
layer, integral membrane proteins are also amphipathic, locations. As will be discussed later, integral proteins need not
having both hydrophilic and hydrophobic portions. As dis- be fixed structures but may be able to move laterally within the
cussed below, those portions of an integral membrane protein membrane.
that reside within the lipid bilayer tend to have a hydrophobic
character. Amino acid residues in these transmembrane do- Distribution of Integral Proteins: Freeze-Fracture Analysis
mains form van der Waals interactions with the fatty acyl The concept that proteins penetrate through membranes,
chains of the bilayer, which seals the protein into the lipid rather than simply remaining external to the bilayer, was de-
wall of the membrane. As a result, the permeability barrier rived primarily from the results of a technique called freeze-
of the membrane is preserved and the protein is brought into fracture replication (see Section 18.2). In this procedure,
tissue is frozen solid and then struck with a knife blade,
which fractures the block into two pieces. As this occurs, the
fracture plane often takes a path between the two leaflets of
the lipid bilayer (Figure 4.15a). Once the membranes are
split in this manner, metals are deposited on their exposed
surfaces to form a shadowed replica, which is viewed in the
electron microscope (see Figure 18.17). As shown in Figure
4.15b, the replica resembles a road strewn with pebbles,
which are called membrane-associated particles. Since the frac-
ture plane passes through the center of the bilayer, most of
these particles correspond to integral membrane proteins that
extend at least halfway through the lipid core. When the frac-
ture plane reaches a given particle, it goes around it rather
than cracking it in half. Consequently, each protein (particle)
separates with one half of the plasma membrane (Figure
4.15c), leaving a corresponding pit in the other half (see Fig-
ure 7.30c). One of the great values of the freeze-fracturing
technique is that it allows an investigation of the micro-
heterogeneity of the membrane. Localized differences in
parts of the membrane stand out in these replicas and can be
identified (as illustrated by the replica of a gap junction
shown in Figure 7.32d ). Biochemical analyses, in contrast,
FIGURE 4.14 Proteins can be surrounded by a closely applied shell of average out such differences.
lipid molecules. Aquaporin is a membrane protein containing four sub-
units (colored differently in the illustration) surrounding an aqueous Studying the Structure and Properties
channel. Analysis of the proteins structure revealed the presence of a
surrounding layer of bound lipid molecules. Whether or not these lipid of Integral Membrane Proteins
molecules affect the function of the aquaporin molecule is unclear, but Because of their hydrophobic transmembrane domains, inte-
it is likely that they are held in close proximity to the protein and thus gral membrane proteins are difficult to isolate in a soluble
unable to move freely within the bilayer. (FROM CAROLA HUNTE AND
form. Removal of these proteins from the membrane normally
SEBASTIAN RICHERS, CURR. OPIN. STRUCT. BIOL. 18:407, 2008, COPYRIGHT
2008, WITH PERMISSION FROM ELSEVIER SCIENCE.)
requires the use of a detergent, such as the ionic (charged)
detergent SDS (which denatures proteins) or the nonionic
Lipid Aqueous
bilayer solution
Exterior
Nonionic
detergent
Fracture face E
Fracture face P
Membrane Detergent-solubilized
protein protein
FIGURE 4.16 Solubilization of membrane proteins with detergents.

The nonpolar ends of the detergent molecules associate with the nonpo-
Cytoplasm lar residues of the protein that had previously been in contact with
(a) the fatty acyl chains of the lipid bilayer. In contrast, the polar ends of the
detergent molecules interact with the surrounding water molecules,
keeping the protein in solution. Nonionic detergents, as shown here, sol-
ubilize membrane proteins without disrupting their structure.
P
(uncharged) detergent Triton X-100 (which generally does

not alter a proteins tertiary structure).
(b)
CH3 (CH2)11 OSO3Na+
Sodium dodecyl sulfate (SDS)
CH3 CH3
CH3 C CH2 C (O CH2 CH2)10 OH
CH3 CH3
Triton X-100
Cytoplasm Like membrane lipids, detergents are amphipathic, being com-

posed of a polar end and a nonpolar hydrocarbon chain (see Fig-
ure 2.20). As a consequence of their structure, detergents can
(c)
substitute for phospholipids in stabilizing integral proteins while
rendering them soluble in aqueous solution (Figure 4.16). Once
FIGURE 4.15 Freeze fracture: a technique for investigating cell mem- the proteins have been solubilized by the detergent, various
brane structure. (a) When a block of frozen tissue is struck by a knife analyses can be carried out to determine the proteins amino acid
blade, a fracture plane runs through the tissue, often following a path composition, molecular mass, amino acid sequence, and so forth.
that leads it through the middle of the lipid bilayer. The fracture plane Researchers have had great difficulty obtaining crystals
goes around the proteins rather than cracking them in half, and they seg-
of most integral membrane proteins for use in X-ray
regate with one of the two halves of the bilayer. The exposed faces within
the center of the bilayer can then be covered with a metal deposit to form
crystallography. In fact, fewer than 1 percent of the known
a metallic replica. These exposed faces are referred to as the E, or ecto- high-resolution protein structures represent integral mem-
plasmic face, and the P, or protoplasmic face. (b) Replica of a freeze- brane proteins (see http://blanco.biomol.uci.edu/MemPro_
fractured human erythrocyte. The P fracture face is seen to be covered resources.html for a discussion of principles of protein structure
with particles approximately 8 nm in diameter. A small ridge (arrow) and http://blanco.biomol.uci.edu/Membrane_Proteins_xtal.
marks the junction of the particulate face with the surrounding ice. html for an updated gallery).2 Furthermore, most of these
(c) This micrograph shows the surface of an erythrocyte that was frozen structures represent prokaryotic versions of a particular pro-
and then fractured, but rather than preparing a replica, the cell was tein, which are often smaller than their eukaryotic homologues
thawed, fixed, and labeled with a marker for the carbohydrate groups and easier to obtain in large quantity. Once the structure of
that project from the external surface of the integral protein glycophorin
(Figure 4.18). Thin sections of the labeled, fractured cell reveal that gly-
2
cophorin molecules (black particles) have preferentially segregated with Many integral membrane proteins have a substantial portion that is present in
the outer half of the membrane. The red line shows the path of the frac- the cytoplasm or extracellular space. In many cases, this soluble portion has been
cleaved from its transmembrane domain, crystallized, and its tertiary structure
ture plane. (B: FROM THOMAS W. TILLACK AND VINCENT T. MARCHESI, determined. While this approach provides valuable data about the protein, it
J. CELL BIOL. 45:649, 1970; C: FROM PEDRO PINTO DA SILVA AND MARIA R. fails to provide information about the proteins orientation within the mem-
TORRISI, J. CELL BIOL. 93:467, 1982; B,C: BY COPYRIGHT PERMISSION OF THE brane. Another crystallographic approach to the study of membrane proteins,
ROCKEFELLER UNIVERSITY PRESS.) which uses the electron microscope rather than X-ray diffraction, is discussed in
the Experimental Pathways on page 168.
containing 11 membrane-spanning " helices. Some of the

technical difficulties in preparing membrane protein crystals
have been overcome using new methodologies and laborious
efforts. In one study, for example, researchers were able to
obtain high-quality crystals of a bacterial transporter after
testing and refining more than 95,000 different conditions
for crystallization. Despite increasing success in protein crys-
tallization, researchers still rely largely on indirect approaches
for determining the three-dimensional organization of most
membrane proteins. We will examine some of these ap-
proaches in the following paragraphs.
Identifying Transmembrane Domains A great deal can be

FIGURE 4.17 An integral protein as it resides within the plasma mem- learned about the structure of a membrane protein and its ori-
brane. Tertiary structure of the photosynthetic reaction center of a bac- entation within the lipid bilayer from a computer-based
terium as determined by X-ray crystallography. The protein contains
(computational) analysis of its amino acid sequence, which is
three different membrane-spanning polypeptides, shown in yellow, light
blue, and dark blue. The helical nature of each of the transmembrane
readily deduced from the nucleotide sequence of an isolated
segments is evident. (FROM G. FEHER, J. P. ALLEN, M. Y. OKAMURA, AND gene. The first question one might ask is: Which segments of
D. C. REES, REPRINTED WITH PERMISSION FROM NATURE 339:113, 1989; the polypeptide chain are actually embedded in the lipid
COPYRIGHT 1989, MACMILLAN MAGAZINES LIMITED.) bilayer? Those segments of a protein embedded within
the membrane, which are described as the transmembrane
domains, have a simple structure; they consist of a string
of about 20 predominantly nonpolar amino acids that span
one member of a membrane protein family is determined, the core of the lipid bilayer as an " helix.3 The chemical
researchers can usually apply a strategy called homology model- structure of a single transmembrane helix is shown in Fig-
ing to learn about the structure and activity of other mem- ure 4.18, which depicts the two-dimensional structure of
bers of the family. For example, solution of the structure of the
bacterial K! channel KcsA (shown in Figure 4.39) provided 3
It was noted on page 54 that the " helix is a favored conformation because it
a wealth of data that could be applied to the structure and allows for a maximum number of hydrogen bonds to be formed between neigh-
mechanism of action of the much more complex eukaryotic boring amino acid residues, thereby creating a highly stable (low-energy) config-
uration. This is particularly important for a membrane-spanning polypeptide
K! channels (Figure 4.42). that is surrounded by fatty acyl chains and, thus, cannot form hydrogen bonds
One of the first membrane proteins whose entire three- with an aqueous solvent. Transmembrane helices are at least 20 amino acids in
dimensional structure was determined by X-ray crystallogra- length, because this is the minimum stretch of polypeptide capable of spanning
the hydrocarbon core of a lipid bilayer of 30 width. A few integral membrane
phy is shown in Figure 4.17. This proteinthe bacterial proteins have been found to contain loops or helices that penetrate but do not
photosynthetic reaction centerconsists of three subunits span the bilayer. An example is the P helix in Figure 4.39.
Exterior Surface
Thr IIe
Leu Ile
Phe IIe FIGURE 4.18 Glycophorin A, an integral protein with a sin-

Gly gle transmembrane domain. The single " helix that passes
Val 80
through the membrane consists predominantly of hydrophobic
Met
Ala Bilayer residues (orange-colored circles). The four positively charged
Gly Val amino acid residues of the cytoplasmic domain of the mem-
Ile brane form ionic bonds with negatively charged lipid head
Gly
groups. Carbohydrates are attached to a number of amino acid
Thr
Ile residues on the outer surface of the protein (shown in the inset).
90 Leu Leu All but one of the 16 oligosaccharides are small O-linked
chains (the exception is a larger oligosaccharide linked to the
lle
Ser asparagine residue at position 26). Glycophorin molecules are
present as homodimers within the erythrocyte membrane
(Figure 4.32d ). The two helices cross over one another in the

+
+

+

+
region between residues 79 and 83. This Gly-X-X-X-Gly
Lys Lys Arg Interior Surface sequence is commonly found where transmembrane helices
Pro Ser Leu Arg (cytosol)
Ile come into close proximity.
(a) (b) (c) (d)
FIGURE 4.19 Accommodating various amino acid residues within face of the lipid bilayer. (c) The side chains of the two aspartic acid
transmembrane helices. (a) In this portrait of a small portion of a trans- residues of this transmembrane helix can also reach the polar face of the
membrane helix, the hydroxyl group of the threonine side chain (arrow) bilayer but introduce distortion in the helix to do so. (d) The aromatic
is able to form a (shared) hydrogen bond with a backbone oxygen within side chains of the two tyrosine residues of this transmembrane helix are
the lipid bilayer. Hydrogen bonds are indicated by the dashed lines and oriented perpendicular to the axis of the membrane and parallel to the
their distances are shown in angstroms. (b) The side chains of the two ly- fatty acyl chains with which they interact. (FROM ANNA C. V. JOHANSSON
sine residues of this transmembrane helix are sufficiently long and flexi- AND ERIK LINDAHL, BIOPHYS. J. 91:4459, 4453, 2006.)
ble to form bonds with the head groups and water molecules of the polar
glycophorin A, the major integral protein of the erythrocyte segments are usually identified as a jagged peak that extends
plasma membrane. Of the 20 amino acids that make up the well into the hydrophobic side of the spectrum. A reliable pre-
lone ! helix of a glycophorin monomer (amino acids 73 to 92 diction concerning the orientation of the transmembrane seg-
of Figure 4.18), all but three have hydrophobic side chains (or ment within the bilayer can usually be made by examining the
an H atom in the case of the glycine residues). The exceptions flanking amino acid residues. In most cases, as illustrated by
are serine and threonine, which are noncharged, polar residues glycophorin in Figure 4.18, those parts of the polypeptide at
(Figure 2.26). Figure 4.19a shows a portion of a transmem- the cytoplasmic flank of a transmembrane segment tend to be
brane helix with a threonine residue, not unlike those of more positively charged than those at the extracellular flank.
glycophorin A. The hydroxyl group of the residues side chain Not all integral membrane proteins contain transmembrane
can form a hydrogen bond with one of the oxygen atoms of ! helices. A number of membrane proteins contain a relatively
the peptide backbone. Fully charged residues may also appear large channel positioned within a circle of membrane-spanning
in transmembrane helices, but they tend to be accommo-
dated in ways that allow them to fit into their hydrophobic
environment. This is illustrated in the model transmembrane
helices shown in Figures 4.19b and c. Each of the helices in
these figures contains a pair of charged residues whose side
chains are able to reach out and interact with the innermost
Hydrophobicity
polar regions of the membrane, even if it requires distorting

G
the helix to do so. Figure 4.19d shows two tyrosine residues +

0
with their hydrophobic aromatic side chains; each aromatic
G
ring is oriented parallel with the hydrocarbon chains of the

_
bilayer with which it has become integrated.

Knowing the amino acid sequence of an integral mem-
brane protein, we can usually identify the transmembrane seg-
ments using a hydropathy plot, in which each site along a
50 100
polypeptide is assigned a value that provides a measure of the
Amino acid residue #
hydrophobicity of the amino acid at that site as well as that of its
N-terminus C-terminus
neighbors. This approach provides a running average of the
hydrophobicity of short sections of the polypeptide, and guar- FIGURE 4.20 Hydropathy plot for glycophorin A, a single membrane-
antees that one or a few polar amino acids in a sequence do not spanning protein. Hydrophobicity is measured by the free energy
alter the profile of the entire stretch. Hydrophobicity of amino required to transfer each segment of the polypeptide from a nonpolar
acids can be determined using various criteria, such as their solvent to an aqueous medium. Values above the 0 line are energy-
requiring ("#Gs), indicating they consist of stretches of amino acids
lipid solubility or the energy that would be required to transfer
that have predominantly nonpolar side chains. Peaks that project above
them from an aqueous into a lipid medium. A hydropathy plot the red-colored line are interpreted as a transmembrane domain.
for glycophorin A is shown in Figure 4.20. Transmembrane
! strands organized into a barrel as illustrated in Figure 5.4. To

date, aqueous channels constructed of ! barrels have only been
XI XII
found in the outer membranes of bacteria, mitochondria, and
chloroplasts. X
IX
Determining Spatial Relationships within an Integral Membrane 52
Protein Suppose you have isolated a gene for an integral 245 II

VII
membrane protein and, based on its nucleotide sequence, 242
53
determined that it contains four apparent membrane-spanning

VIII
" helices. You might want to know how these helices are ori-
29
ented relative to one another and which amino acid side 28
III
I
chains of each helix face outward toward the lipid environ-
ment. Although these determinations are difficult to make
V
without detailed structural models, considerable insight can
VI
be gained by site-directed mutagenesis, that is, by introducing
specific changes into the gene that codes for the protein IV
(Section 18.18). For example, site-directed mutagenesis can
be employed to replace amino acid residues in neighboring
helices with cysteine residues. As discussed on page 52, two FIGURE 4.21 Determination of helix packing in a membrane protein
cysteine residues can form a covalent disulfide bridge. If two by site-directed cross-linking. In these experiments, pairs of cysteine
transmembrane helices of a polypeptide each contain a cys- residues are introduced into the protein by site-directed mutagenesis,
teine residue, and the two cysteine residues are able to form a and the ability of the cysteines to form disulfide bridges is determined.
disulfide bridge with one another, then these helices must Hydropathy plots and other data had indicated that lactose permease has
reside in very close proximity. The results of one site-directed 12 transmembrane helices. It was found that a cysteine introduced at po-
cross-linking study on lactose permease, a sugar-transporting sition 242 of helix VII can cross-link to a cysteine introduced at either
protein in bacterial cell membranes, are shown in Figure 4.21. position 28 or 29 of helix I. Similarly, a cysteine at position 245 of helix
VII can cross-link to cysteines at either 52 or 53 of helix II. The proxim-
It was found in this case that helix VII lies in close proximity
ity of these three helices is thus established. (The X-ray structure of this
to both helices I and II.
protein was reported in 2003.) (REPRINTED FROM H. R. KABACK, J. VOSS,
Determining spatial relationships between amino acids in AND J. WU, CURR. OPIN. STRUCT. BIOL. 7:539, 1997; COPYRIGHT 1997, WITH
a membrane protein can provide more than structural infor- PERMISSION FROM ELSEVIER SCIENCE.)
mation; it can tell a researcher about some of the dynamic
events that occur as a protein carries out its function. One way
to learn about the distance between selected residues in a pro-
tein is to introduce chemical groups whose properties are
sensitive to the distance that separates them. Nitroxides are nied by an increased separation between the labeled residues
chemical groups that contain an unpaired electron, which proof the four subunits (Figure 4.22b). An increase in diameter of
duces a characteristic spectrum when monitored by a tech- the channel opening allows ions in the cytoplasm to reach the
nique called electron paramagnetic resonance (EPR) spectroscopy. actual permeation pathway (shown in red) within the channel,
A nitroxide group can be introduced at any site in a protein by which allows only the passage of K# ions (discussed on
first mutating that site to a cysteine and then attaching the ni- page 149). An alternate technique, called FRET, that can also
troxide to the OSH group of the cysteine residue. Figure 4.22 be used to measure changes in the distance between labeled
shows how this technique was used to discover the conforma- groups within a protein, is illustrated in Figure 18.8.
tional changes that occur in a membrane protein as its chan-
nel is opened in response to changes in the pH of the medium. Peripheral Membrane Proteins
The protein in question, a bacterial K# channel, is a tetramer
composed of four identical subunits. The cytoplasmic opening Peripheral proteins are associated with the membrane by weak
to the channel is bounded by four transmembrane helices, one electrostatic bonds (refer to Figure 4.13b). Peripheral proteins
from each subunit of the protein. Figure 4.22a shows the EPR can usually be solubilized by extraction with high-concentration
spectra that were obtained when a nitroxide was introduced salt solutions that weaken the electrostatic bonds holding pe-
near the cytoplasmic end of each transmembrane helix. The ripheral proteins to a membrane. In actual fact, the distinction
red line shows the spectrum obtained at pH 6.5 when the between integral and peripheral proteins is blurred because
channel is in the closed state, and the blue line shows the spec- many integral membrane proteins consist of several polypep-
trum at pH 3.5 when the channel is open. The shape of each tides, some that penetrate the lipid bilayer and others that
line depends on the proximity of the nitroxides to one an- remain on the periphery.
other. The spectrum is broader at pH 6.5 because the nitrox- The best studied peripheral proteins are located on the in-
ide groups on the four subunits are closer together at this pH, ternal (cytosolic) surface of the plasma membrane, where they
which decreases the intensity of their EPR signals. These re- form a fibrillar network that acts as a membrane skeleton (see
sults indicate that the activation of the channel is accompa- Figure 4.32d ). These proteins provide mechanical support for
4.5 MEMBRANE LIPIDS AND MEMBRANE FLUIDITY 133
External medium
Dehydrated
K+ ion
Permeation
pathway
pH 3.5
pH 6.5
Plasma
(a) membrane
FIGURE 4.22 Use of EPR spectroscopy to monitor

changes in conformation of a bacterial K! ion channel
as it opens and closes. (a) EPR spectra from nitroxides
that have been attached to cysteine residues near the cyto- Nitroxide-labeled
plasmic end of the four transmembrane helices that line cysteine residue
the channel. The cysteine residue in each helix replaces a Hydrated
glycine residue that is normally at that position. The CLOSED K+ ion OPEN
shapes of the spectra depend on the distances between un- Cytoplasm
paired electrons in the nitroxides on different subunits.
(Nitroxides are described as spin-labels, and this tech- (b)
nique is known as site-directed spin labeling.) (b) A highly
schematic model of the ion channel in the open and closed states based movement of the four nitroxide groups apart from one another.
on the data from part a. Opening of the channel is accompanied by the (A: FROM E. PEROZO ET AL., NATURE STRUCT. BIOL. 5:468, 1998.)
the membrane and function as an anchor for integral membrane membrane in this way (Src and Ras) have been implicated in
proteins. Other peripheral proteins on the internal plasma mem- the transformation of a normal cell to a malignant state.
brane surface function as enzymes, specialized coats (see Figure
8.24), or factors that transmit transmembrane signals (see Figure
15.17). Peripheral proteins typically have a dynamic relationship
REVIEW
?
with the membrane, being recruited to the membrane or re- 1. Why are detergents necessary to solubilize membrane pro-
leased from the membrane depending on prevailing conditions. teins? How might one determine the diversity of integral
proteins that reside in a purified membrane fraction?
2. How can one determine: (1) the location of transmem-
Lipid-Anchored Membrane Proteins
brane segments in the amino acid sequence or (2) the
Several types of lipid-anchored membrane proteins can be relative locations of transmembrane helices?
distinguished. Numerous proteins present on the external face 3. What is meant by the statement that the proteins of a
of the plasma membrane are bound to the membrane by a membrane are distributed asymmetrically? Is this also true
small, complex oligosaccharide linked to a molecule of phos- of the membranes carbohydrate?
phatidylinositol that is embedded in the outer leaflet of the
4. Describe the properties of the three classes of membrane
lipid bilayer (refer to Figure 4.13c). Peripheral membrane pro-
proteins (integral, peripheral, and lipid-anchored), how
teins containing this type of glycosyl-phosphatidylinositol
they differ from one another, and how they vary among
linkage are called GPI-anchored proteins. They were discov-
themselves.
ered when it was shown that certain membrane proteins could
be released by a phospholipase that specifically recognized and
cleaved inositol-containing phospholipids. The normal cellu-
lar scrapie protein PrPC (page 64) is a GPI-linked molecule,
as are various receptors, enzymes, and cell-adhesion proteins.
4.5 MEMBRANE LIPIDS AND
A rare type of anemia, paroxysmal nocturnal hemoglobinuria, MEMBRANE FLUIDITY
results from a deficiency in GPI synthesis that makes red The physical state of the lipid of a membrane is described by its
blood cells susceptible to lysis. fluidity (or viscosity).4 Consider a simple artificial bilayer com-
Another group of proteins present on the cytoplasmic posed of phosphatidylcholine and phosphatidylethanolamine,
side of the plasma membrane is anchored to the membrane by whose fatty acids are largely unsaturated. If the temperature of
one or more long hydrocarbon chains embedded in the inner
leaflet of the lipid bilayer (refer to Figure 4.13c and accompa- 4
Fluidity and viscosity are inversely related; fluidity is a measure of the ease of
nying legend). At least two proteins associated with the plasma flow, and viscosity is a measure of the resistance to flow.
(a) (b)
FIGURE 4.23 The structure of the lipid bilayer depends on the temper- of order. (b) Below the transition temperature, the movement of the
ature. The bilayer shown here is composed of two phospholipids: phos- molecules is greatly restricted, and the entire bilayer can be described as
phatidylcholine and phosphatidylethanolamine. (a) Above the transition a crystalline gel. (AFTER R. N. ROBERTSON, THE LIVELY MEMBRANES,
temperature, the lipid molecules and their hydrophobic tails are free to CAMBRIDGE UNIVERSITY PRESS; 1983, REPRINTED WITH PERMISSION OF
move in certain directions, even though they retain a considerable degree CAMBRIDGE UNIVERSITY PRESS.)
the bilayer is kept relatively warm (e.g., 37!C), the lipid exists TABLE 4.2 Melting Points of the Common 18-Carbon Fatty Acids
in a relatively fluid state (Figure 4.23a). At this temperature, Fatty acid cis Double bonds M.p.(!C)
the lipid bilayer is best described as a two-dimensional liquid
crystal. As in a crystal, the molecules still retain a specified ori- Stearic acid 0 70
entation; in this case, the long axes of the molecules tend to- Oleic acid 1 13
ward a parallel arrangement, yet individual phospholipids can Linoleic acid 2 "9
Linolenic acid 3 "17
rotate around their axis or move laterally within the plane of
Eicosapentanoic acid (EPA)* 5 "54
the bilayer. If the temperature is slowly lowered, a point is
reached where the bilayer distinctly changes (Figure 4.23b). *EPA has 20 carbons.
The lipid is converted from a liquid crystalline phase to a
frozen crystalline gel in which the movement of the phospho-
lipid fatty acid chains is greatly restricted. The temperature at ence of cholesterol tends to abolish sharp transition tempera-
which this change occurs is called the transition temperature. tures and creates a condition of intermediate fluidity. In phys-
The transition temperature of a particular bilayer depends iologic terms, cholesterol tends to increase the durability while
on the ability of the lipid molecules to be packed together, decreasing the permeability of a membrane.
which depends in turn on the particular lipids of which it is
constructed. Saturated fatty acids have the shape of a straight, The Importance of Membrane Fluidity
flexible rod. Cis-unsaturated fatty acids, on the other hand,
have crooks in the chain at the sites of a double bond (Fig- What effect does the physical state of the lipid bilayer have on
ures 2.19 and 4.23). Consequently, phospholipids with satu- the biological properties of the membrane? Membrane fluidity
rated chains pack together more tightly than those containing provides a perfect compromise between a rigid, ordered struc-
unsaturated chains. The greater the degree of unsaturation of ture in which mobility would be absent and a completely fluid,
the fatty acids of the bilayer, the lower the temperature before nonviscous liquid in which the components of the membrane
the bilayer gels. The introduction of one double bond in a could not be oriented and structural organization and mechan-
molecule of stearic acid lowers the melting temperature ical support would be lacking. In addition, fluidity allows for
almost 60!C (Table 4.2).5 Another factor that influences interactions to take place within the membrane. For example,
bilayer fluidity is fatty acid chain length. The shorter the fatty membrane fluidity makes it possible for clusters of membrane
acyl chains of a phospholipid, the lower its melting tempera- proteins to assemble at particular sites within the membrane
ture. The physical state of the membrane is also affected by and form specialized structures, such as intercellular junctions,
cholesterol. Because of their orientation within the bilayer light-capturing photosynthetic complexes, and synapses. Be-
(Figure 4.7), cholesterol molecules disrupt the close packing cause of membrane fluidity, molecules that interact can come
of fatty acyl chains and interfere with their mobility. The pres- together, carry out the necessary reaction, and move apart.
Fluidity also plays a key role in membrane assembly, a sub-
5
The effect of fatty acid saturation on melting temperature is illustrated by ject discussed in Chapter 8. Membranes arise only from preex-
familiar food products. Vegetable oils remain a liquid in the refrigerator, whereas isting membranes, and their growth is accomplished by the
margarine is a solid. Vegetable oils contain polyunsaturated fatty acids, whereas
the fatty acids of margarine have been saturated by a chemical process that hy- insertion of lipids and proteins into the fluid matrix of the
drogenates the double bonds of the vegetable oils used as the starting material. membranous sheet. Many of the most basic cellular processes,
4.5 MEMBRANE LIPIDS AND MEMBRANE FLUIDITY 135
including cell movement, cell growth, cell division, formation of tween phospholipids. In addition, the cell changes the types of
intercellular junctions, secretion, and endocytosis, depend on phospholipids being synthesized in favor of ones containing
the movement of membrane components and would probably more unsaturated fatty acids. As a result of the activities of
not be possible if membranes were rigid, nonfluid structures. these various enzymes, the physical properties of a cells mem-
branes are matched to the prevailing environmental condi-
tions. Maintenance of fluid membranes by adjustments in
Maintaining Membrane Fluidity
fatty acyl composition has been demonstrated in a variety of
The internal temperature of most organisms (other than birds organisms, including hibernating mammals, pond-dwelling
and mammals) fluctuates with the temperature of the external fish whose body temperature changes markedly from day to
environment. Since it is essential for many activities that the night, cold-resistant plants, and bacteria living in hot springs.
membranes of a cell remain in a fluid state, cells respond to
changing conditions by altering the types of phospholipids of
Lipid Rafts
which they are made. Maintenance of membrane fluidity is
an example of homeostasis at the cellular level and can be Every so often an issue emerges that splits the community of
demonstrated in various ways. For example, if the temperature cell biologists into believers and nonbelievers. The issue of
of a culture of cells is lowered, the cells respond metabolically. lipid rafts falls into this category. When membrane lipids
The initial emergency response is mediated by enzymes that are extracted from cells and used to prepare artificial lipid bi-
remodel membranes, making the cell more cold resistant. layers, cholesterol and sphingolipids tend to self-assemble
Remodeling is accomplished by (1) desaturating single bonds into microdomains that are more gelated and highly ordered
in fatty acyl chains to form double bonds, and (2) reshuffling than surrounding regions consisting primarily of phospho-
the chains between different phospholipid molecules to pro- glycerides. Because of their distinctive physical properties,
duce ones that contain two unsaturated fatty acids, which such microdomains tend to float within the more fluid and
greatly lowers the melting temperature of the bilayer. Desatu- disordered environment of the artificial bilayer (Figure 4.24a).
ration of single bonds to form double bonds is catalyzed by As a result, these patches of cholesterol and sphingolipid are
enzymes called desaturases. Reshuffling is accomplished by referred to as lipid rafts. When added to these artificial bilay-
phospholipases, which split the fatty acid from the glycerol ers, certain proteins tend to become concentrated in the lipid
backbone, and acyltransferases, which transfer fatty acids be- rafts, whereas others tend to remain outside their boundaries.
GPI-anchored proteins show a particular fondness for the
ordered regions of the bilayer (Figure 4.24a).
GPIanchored
protein
Signaling
protein
(a) (b)
FIGURE 4.24 Lipid rafts. (a) Image of the upper surface of an artificial with long saturated fatty acids also tend to concentrate in this region.
lipid bilayer containing phosphatidylcholine, which appears as the black GPI-anchored proteins are thought to become concentrated in lipid
background, and sphingomyelin molecules, which organize themselves rafts. The lipids in the outer leaflet of the raft have an organizing effect
spontaneously into the orange-colored rafts. The yellow peaks show the on the lipids of the inner leaflet. As a result, the inner-leaflet raft lipids
positions of a GPI-anchored protein, which is almost exclusively raft- consist primarily of cholesterol and glycerophospholipids with long sat-
associated. This image is provided by an atomic force microscope, which urated fatty acyl tails. The inner leaflet tends to concentrate lipid-
measures the height of various parts of the specimen at the molecular anchored proteins, such as Src kinase, that are involved in cell signaling.
level. (b) Schematic model of a lipid raft within a cell. The outer leaflet (The controversy over the existence of lipid rafts is discussed in Nature
of the raft consists primarily of cholesterol (yellow) and sphingolipids Cell Biol. 9:7, 2007.) (A: FROM D. E. SASLOWSKY, ET AL., J. BIOL. CHEM.
(red head groups). Phosphatidylcholine molecules (blue head groups) 277, COVER OF #30, 2002; COURTESY OF J. MICHAEL EDWARDSON.)
The controversy arises over whether similar types of CELL EXTERIOR

cholesterol-rich lipid rafts, such as that illustrated in Fig- Transverse diffusion (flip-flop)
ure 4.24b, exist within living cells. Most of the evidence in fa- 5
(~10 sec)
vor of lipid rafts is derived from studies that employ unnatural
treatments, such as detergent extraction or cholesterol deple-
tion, which makes the results difficult to interpret. Attempts
to demonstrate the presence of lipid rafts in living cells have -- 9
Flex (~10 sec)
generally been unsuccessful, which can either mean that such
rafts do not exist or they are so small (5 to 25 nm diameter)
and short-lived as to be difficult to detect with current tech-
niques. The concept of lipid rafts is very appealing because it
provides a means to introduce order into a seemingly random Lateral shift
-- 6
(~10 sec)
sea of lipid molecules. Lipid rafts are postulated to serve as CYTOSOL
floating platforms that concentrate particular proteins,
thereby organizing the membrane into functional compart- FIGURE 4.25 The possible movements of phospholipids in a mem-
ments (Figure 4.24b). For example, lipid rafts are thought to brane. The types of movements in which membrane phospholipids can
provide a favorable local environment for cell-surface recep- engage and the approximate time scales over which they occur. Whereas
tors to interact with other membrane proteins that transmit phospholipids move from one leaflet to another at a very slow rate, they
signals from the extracellular space to the cell interior. diffuse laterally within a leaflet rapidly. Lipids lacking polar groups, such
as cholesterol, can move across the bilayer quite rapidly.
REVIEW
?
1. What is the importance of fatty acid unsaturation for
membrane fluidity? f enzymes that are able to desaturate one leaflet to the other. These enzymes play a role in estab-
fatty acids? lishing lipid asymmetry and may also reverse the slow rate of
passive transmembrane movement.
2. What is meant by a membranes transition temperature,
Because lipids provide the matrix in which integral pro-
and how is it affected by the degree of saturation or length
teins of a membrane are embedded, the physical state of the
of fatty acyl chains? How are these properties important
lipid is an important determinant of the mobility of integral
in the formation of lipid rafts?
proteins. The demonstration that integral proteins can move
3. Why is membrane fluidity important to a cell? within the plane of the membrane was a cornerstone in the
4. How can the two sides of a lipid bilayer have different formulation of the fluid-mosaic model. The dynamic prop-
ionic charges? erties of membrane proteins have been revealed in several
ways.
The Diffusion of Membrane Proteins

4.6 THE DYNAMIC NATURE after Cell Fusion
OF THE PLASMA MEMBRANE Cell fusion is a technique whereby two different types of cells,
It is apparent from the previous discussion that the lipid or cells from two different species, can be fused to produce one
bilayer can exist in a relatively fluid state. As a result, a phos- cell with a common cytoplasm and a single, continuous
pholipid can move laterally within the same leaflet with con- plasma membrane. Cells are induced to fuse with one another
siderable ease. The mobility of individual lipid molecules by making the outer surface of the cells sticky so that their
within the bilayer of the plasma membrane can be directly plasma membranes adhere to one another. Cells can be in-
observed under the microscope by linking the polar heads of duced to fuse by addition of certain inactivated viruses that at-
the lipids to gold particles or fluorescent compounds (see tach to the surface membrane, by adding the compound
Figure 4.29). It is estimated that a phospholipid can diffuse polyethylene glycol, or by a mild electric shock. Cell fusion
from one end of a bacterium to the other end in a second or has played an important role in cell biology and is currently
two. In contrast, it takes a phospholipid molecule a matter of used in an invaluable technique to prepare specific antibodies
hours to days to move across to the other leaflet. Thus, of all (Section 18.19).
the possible motions that a phospholipid can make, its flip- The first experiments to demonstrate that membrane
flop to the other side of the membrane is the most restricted proteins could move within the plane of the membrane uti-
(Figure 4.25). This finding is not surprising. For flip-flop to lized cell fusion, and they were reported in 1970 by Larry Frye
occur, the hydrophilic head group of the lipid must pass and Michael Edidin of Johns Hopkins University. In their ex-
through the internal hydrophobic sheet of the membrane, periments, mouse and human cells were fused, and the loca-
which is thermodynamically unfavorable. However, cells con- tions of specific proteins of the plasma membrane were
tain enzymes that actively move certain phospholipids from followed once the two membranes had become continuous.
4.6 THE DYNAMIC NATURE OF THE PLASMA MEMBRANE 137
1 2 3 4
Human cell
Addition of 40
sendai minutes
(fusing)
virus
Mouse cell
(a) (b)
FIGURE 4.26 The use of cell fusion to reveal mobility of membrane plasma membranes of the hybrid cells were monitored by interaction
proteins. (a) Outline of the experiment in which human and mouse cells with fluorescent red and fluorescent green antibodies, respectively.
were fused (steps 12) and the distribution of the proteins on the surface (b) Micrograph showing a fused cell in which mouse and human pro-
of each cell were followed in the hybrids over time (steps 34). Mouse teins are still in their respective hemispheres (equivalent to the hybrid in
membrane proteins are indicated by solid circles, human membrane pro- step 3 of part a). (B: FROM L. D. FRYE AND MICHAEL EDIDIN, J. CELL SCI.
teins by open circles. Locations of human and mouse proteins in the 7:328, 334, 1970; BY PERMISSION OF THE COMPANY OF BIOLOGISTS LTD.)
To follow the distribution of either the mouse membrane pro- a fluorescent antibody. Once labeled, cells are placed under
teins or the human membrane proteins at various times after the microscope and irradiated by a sharply focused laser beam
fusion, antibodies against one or the other type of protein that bleaches the fluorescent molecules in its path, leaving a
were prepared and covalently linked to fluorescent dyes. The circular spot (typically about 1 !m diameter) on the surface of
antibodies against the mouse proteins were complexed with a the cell that is largely devoid of fluorescence. If the labeled
dye that fluoresces green and the antibodies against human proteins in the membrane are mobile, then the random move-
proteins with one that fluoresces red. When the antibodies ments of these molecules should produce a gradual reappear-
were added to fused cells, they bound to the human or mouse ance of fluorescence in the irradiated circle. The rate of
proteins and could be located under a fluorescence light mi- fluorescence recovery (Figure 4.27b) provides a direct measure
croscope (Figure 4.26a). At the time of fusion, the plasma of the rate of diffusion (expressed as a diffusion coefficient, D)
membrane appeared half human and half mouse; that is, the of the mobile molecules. The extent of fluorescence recovery
two protein types remained segregated in their own hemi- (expressed as a percentage of the original intensity) provides a
sphere (step 3, Figure 4.26a,b). As the time after fusion in- measure of the percentage of the labeled molecules that are
creased, the membrane proteins were seen to move laterally free to diffuse.
within the membrane into the opposite hemisphere. By about Early studies utilizing FRAP suggested that (1) mem-
40 minutes, each species proteins were uniformly distributed brane proteins moved much more slowly in a plasma mem-
around the entire hybrid cell membrane (step 4, Figure 4.26a). brane than they would in a pure lipid bilayer and (2) a
If the same experiment was performed at lower temperature, significant fraction of membrane proteins (30 to 70 percent)
the viscosity of the lipid bilayer increased, and the mobility of were not free to diffuse back into the irradiated circle. But the
the membrane proteins decreased. These early cell fusion ex- FRAP technique has its drawbacks. FRAP can only follow
periments gave the impression that integral membrane pro- the average movement of a relatively large number of labeled
teins were capable of virtually unrestricted movements. As we molecules (hundreds to thousands) as they diffuse over a rela-
will see shortly, subsequent studies made it apparent that tively large distance (e.g., 1 !m). As a result, researchers using
membrane dynamics was a much more complex subject than FRAP cannot distinguish between proteins that are truly
first envisioned. immobile and ones that can only diffuse over a limited dis-
tance in the time allowed. To get around these limitations, al-
ternate techniques have been developed that allow researchers
Restrictions on Protein and Lipid Mobility
to observe the movements of individual protein molecules
Several techniques allow researchers to follow the movements over very short distances and to determine how they might be
of molecules in the membranes of living cells using the light restrained.
microscope. In a technique called fluorescence recovery after In single-particle tracking (SPT), individual membrane
photobleaching (FRAP), which is illustrated in Figure 4.27a, protein molecules are labeled, usually with antibody-coated
integral membrane components in cultured cells are first gold particles (approximately 40 nm in diameter), and the
labeled by linkage to a fluorescent dye. A particular mem- movements of the labeled molecules are followed by computer-
brane protein can be labeled using a specific probe, such as enhanced video microscopy (Section 18.1). The results of these
studies often depend on the particular protein being investi-

gated. For example,
Some membrane proteins move randomly throughout the
membrane (Figure 4.28, protein A), though generally at
rates considerably less than would be measured in an arti-
ficial lipid bilayer. (If protein mobility were based strictly
N
on physical parameters such as lipid viscosity and protein
size, one would expect proteins to migrate with diffusion
coefficients of approximately 10!8 to 10!9 cm2/sec rather
Label proteins with
1 fluorescent dye than 10!10 to 10!12 cm2/sec, as is observed for molecules
of this group.) The reasons for the reduced diffusion coef-
ficient have been debated.
Some membrane proteins fail to move and are considered
to be immobilized (Figure 4.28, protein B).
N
In some cases, a protein is found to move in a highly di-
rected (i.e., nonrandom) manner toward one part of the
Photobleach spot
cell or another (Figure 4.28, protein C). For example, a
2 particular membrane protein may tend to move toward
with laser beam
the leading or the trailing edge of a moving cell.
In most studies, the largest fraction of protein species
exhibit random (Brownian) movement within the mem-
brane at rates consistent with free diffusion (diffusion co-
N
efficients about 5 " 10!9 cm2/sec), but the molecules are
unable to migrate freely more than a few tenths of a mi-
crometer. Long-range diffusion occurs but at slower rates,
3 Recovery
N A
D
E
B
(a)
Illuminate
l
C
Fluorescence
Time
FIGURE 4.28 Patterns of movement of integral membrane proteins.
(b) Depending on the cell type and the conditions, integral membrane pro-
teins can exhibit several different types of mobility. Protein A is capable
FIGURE 4.27 Measuring the diffusion rates of membrane proof diffusing randomly throughout the membrane, though its rate of
A
teins by fluorescence recovery after photobleaching (FRAP). movement may be limited; protein B is immobilized as the result of its
(a) In this technique, a particular component of the membrane is first interaction with the underlying membrane skeleton; protein C is being
labeled with a fluorescent dye (step 1). A small region of the surface is moved in a particular direction as the result of its interaction with a
then irradiated to bleach the dye molecules (step 2), and the recovery motor protein at the cytoplasmic surface of the membrane; movement of
of fluorescence in the bleached region is followed over time (step 3). protein D is restricted by other integral proteins of the membrane;
(N represents the cell nucleus.) (b) The rate of fluorescence recovery movement of protein E is restricted by fences formed by proteins of the
within the illuminated spot can vary depending on the protein(s) being membrane skeleton, but it can hop into adjacent compartments through
followed. The rate of recovery is related to the diffusion coefficient of the transient openings in a fence; movement of protein F is restrained by
fluorescently labeled protein. extracellular materials.
apparently because of the presence of a system of barriers. small molecules that make up the very fabric of the lipid bi-
These barriers are discussed in the following sections. layer. One might expect their movements to be completely
unfettered, yet a number of studies have suggested that phos-
Control of Membrane Protein Mobility It is apparent that pholipid diffusion is also restricted. When individual phos-
plasma membrane proteins are not totally free to drift around pholipid molecules of a plasma membrane are tagged and
randomly on the lipid sea, but instead they are subjected to followed under the microscope using ultra high-speed cam-
various influences that affect their mobility. Some membranes eras, they are seen to be confined for very brief periods and
are crowded with proteins, so that the random movements of then hop from one confined area to another. Figure 4.29a
one molecule can be impeded by its neighbors (Figure 4.28, shows the path taken by an individual phospholipid within
protein D). The strongest influences on an integral membrane the plasma membrane over a period of 56 milliseconds. Com-
protein are thought to be exerted from just beneath the mem- puter analysis indicates that the phospholipid diffuses freely
brane on its cytoplasmic face. The plasma membranes of many within one compartment (shaded in purple) before it jumps
cells possess a fibrillar network, or membrane skeleton, con- the fence into a neighboring compartment (shaded in blue)
sisting of peripheral proteins situated on the cytoplasmic sur- and then over another fence into an adjacent compartment
face of the membrane. A certain proportion of a membranes (shaded in green), and so forth. Treatment of the membrane
integral protein molecules are either tethered to the membrane with agents that disrupt the underlying membrane skeleton
skeleton (Figure 4.28, protein B) or otherwise restricted by it. removes some of the fences that restrict phospholipid diffu-
Information concerning the presence of membrane barri- sion. But if the membrane skeleton lies beneath the lipid bi-
ers has been obtained using an innovative technique that al- layer, how can it interfere with phospholipid movement? The
lows investigators to trap integral proteins and drag them authors of these studies speculate that the fences are con-
through the plasma membrane with a known force. This tech- structed of rows of integral membrane proteins whose cyto-
nique, which uses an apparatus referred to as optical tweezers, plasmic domains are attached to the membrane skeleton. This
takes advantage of the tiny optical forces that are generated by
a focused laser beam. The integral proteins to be studied are
tagged with antibody-coated beads, which serve as handles (a) (b)
that can be gripped by the laser beam. It is generally found
that optical tweezers can drag an integral protein for a limited
distance before the protein encounters a barrier that causes it
to be released from the lasers grip. As it is released, the pro-
tein typically springs backward, suggesting that the barriers
are elastic structures.
One approach to studying factors that affect membrane
protein mobility is to genetically modify cells so that they pro-
duce altered membrane proteins. Integral proteins whose cy-
toplasmic portions have been genetically deleted often move
much greater distances than their intact counterparts, indicat-
ing that barriers reside on the cytoplasmic side of the mem-
brane. These findings suggest that the membranes underlying
skeleton forms a network of fences around portions of the
membrane, creating compartments that restrict the distance
an integral protein can travel (Figure 4.28, protein E). Pro- FIGURE 4.29 Experimental demonstration that diffusion of
teins move across the boundaries from one compartment to phospholipids within the plasma membrane is confined.
another through breaks in the fences. Such openings are (a) The track of a single labeled unsaturated phospholipid is followed
thought to appear and disappear along with the dynamic dis- for 56 ms as it diffuses within the plasma membrane of a rat fibroblast.
Phospholipids diffuse freely within a confined compartment before
assembly and reassembly of parts of the meshwork. Membrane
hopping into a neighboring compartment. The rate of diffusion within
compartments may keep specific combinations of proteins in a compartment is as rapid as that expected by unhindered Brownian
close enough proximity to facilitate their interaction. movement. However, the overall rate of diffusion of the phospholipid
Integral proteins lacking that portion that would nor- appears slowed because the molecule must hop a barrier to continue its
mally project into the extracellular space typically move at a movement. The movement of the phospholipid within each compart-
much faster rate than the wild-type version of the protein. ment is represented by a single color. (b) The same experiment shown in
This finding suggests that the movement of a transmembrane a is carried out for 33 ms in an artificial bilayer, which lacks the
protein through the bilayer is slowed by extracellular materials picket fences present in a cellular membrane. The much more open,
that can entangle the external portion of the protein molecule extended trajectory of the phospholipid can now be explained by sim-
(Figure 4.28, protein F). ple, unconfined Brownian movement. For the sake of comparison, fake
compartments were assigned in b and indicated by different colors.
(FROM TAKAHIRO F UJIWARA ET AL., COURTESY OF AKIHIRO KUSUMI, J. CELL
Membrane Lipid Mobility Proteins are huge molecules, so
BIOL. 157:1073, 2002; BY COPYRIGHT PERMISSION OF THE ROCKEFELLER
it isnt surprising that their movement within the lipid bi- UNIVERSITY PRESS.)
layer might be restricted. Phospholipids, by comparison, are
JWCL151_ch04_117-172.qxd 6/26/09 12:52 AM Page 140
is not unlike the confinement of horses or cows by a picket Glucose

fence whose posts are embedded in the underlying soil. Disaccharide
Apical plasma Na+
membrane
Membrane Domains and Cell Polarity For the most part, regulation of nutrient
and water intake
studies of membrane dynamics, such as those discussed above, regulated secretion +
are carried out on the relatively homogeneous plasma mem- p rotection
Monosaccharide
brane situated at the upper or lower surface of a cell residing
on a culture dish. Most membranes, however, exhibit distinct
variations in protein composition and mobility, especially in Lateral plasma
cells whose various surfaces display distinct functions. For ex- membrane
ample, the epithelial cells that line the intestinal wall or make c ell contact and
adhesion
up the microscopic tubules of the kidney are highly polarized c ell communication
cells whose different surfaces carry out different functions
(Figure 4.30). Studies indicate that the apical plasma mem-
brane, which selectively absorbs substances from the lumen, Basal
possesses different enzymes than the lateral plasma mem- membrane
brane, which interacts with neighboring epithelial cells, or the c ell-substratum
contact
basal membrane, which adheres to an underlying extracellular generation of ion
substrate (a basement membrane). In other examples, the recep- gradients
tors for neurotransmitter substances are concentrated into re-
gions of the plasma membrane located within synapses (see
Figure 4.56), and receptors for low-density lipoproteins are FIGURE 4.30 Differentiated functions of the plasma membrane of an
concentrated into patches of the plasma membrane special- epithelial cell. The apical surface of this intestinal epithelial cell con-
ized to facilitate their internalization (see Figure 8.38). tains integral proteins that function in ion transport and hydrolysis of
Of all the various types of mammalian cells, sperm may disaccharides, such as sucrose and lactose; the lateral surface contains in-
have the most highly differentiated structure. A mature sperm tegral proteins that function in intercellular interaction; and the basal
surface contains integral proteins that function in the association of the
can be divided into head, midpiece, and tail, each having its
cell with the underlying basement membrane.
own specialized functions. Although divided into a number of
distinct parts, a sperm is covered by a continuous plasma mem-
brane which, as revealed by numerous techniques, consists of
a mosaic of different types of localized domains. For example,
when sperm are treated with a variety of antibodies, each an-
tibody combines with the surface of the cell in a unique topo-
graphic pattern that reflects the distribution within the plasma
membrane of the particular protein antigen recognized by
that antibody (Figure 4.31).
(a) (b)
The Red Blood Cell: An Example
of Plasma Membrane Structure Anterior head
Of all the diverse types of membranes, the plasma membrane Posterior head
of the human erythrocyte (red blood cell) is the most studied
and best understood (Figure 4.32). There are several reasons
for the popularity of this membrane. The cells are inexpensive Posterior tail
to obtain and readily available in huge numbers from whole
blood. They are already present as single cells and need not be
dissociated from a complex tissue. The cells are simple by
(c) (d)
comparison with other cell types, lacking nuclear and cyto-
plasmic membranes that inevitably contaminate plasma mem- FIGURE 4.31 Differentiation of the mammalian sperm plasma mem-
brane preparations from other cells. In addition, purified, brane as revealed by fluorescent antibodies. (ac) Three pairs of micro-
intact erythrocyte plasma membranes can be obtained simply graphs, each showing the distribution of a particular protein at the cell
by placing the cells in a dilute (hypotonic) salt solution. The surface as revealed by a bound fluorescent antibody. The three proteins
cells respond to this osmotic shock by taking up water and are localized in different parts of the continuous sperm membrane.
swelling, a phenomenon termed hemolysis. As the surface area Each pair of photographs shows the fluoresence pattern of the bound
of each cell increases, the cell becomes leaky, and the con- antibody and a phase contrast micrograph of the same cell. (d ) Diagram
summarizing the distribution of the proteins. (AC: FROM DIANA G.
tents, composed almost totally of dissolved hemoglobin, flow
MYLES, PAUL PRIMAKOFF, AND ANTHONY R. BELLV, CELL 23:434, 1981, BY
out of the cell leaving behind a plasma membrane ghost PERMISSION OF CELL PRESS.)
(Figure 4.32b).
PERIPHERAL INTEGRAL
Major Minor
Spectrin
Anykrin
ATPase
Band 3 AChe
Band 4.1 Glycophorin A
Band 4.2
(a) (b)
Actin
G3PD
Band 7
Ankyrin
Band 4.1
Actin
(c)
Tropomyosin Band 4.1
Spectrin

Band 3
(d)(d) Glycophorin A
FIGURE 4.32 The plasma membrane of the human erythrocyte. (a) Scanning electron micro-
graph of human erythrocytes. (b) Micrograph showing plasma membrane ghosts, which were iso-
lated by allowing erythrocytes to swell and hemolyze as described in the text. (c) The results of
SDSpolyacrylamide gel electrophoresis (SDSPAGE) used to fractionate the proteins of the
erythrocyte membrane, which are identified at the side of the gel. (d ) A model of the erythrocyte
plasma membrane as viewed from the internal surface, showing the integral proteins embedded
in the lipid bilayer and the arrangement of peripheral proteins that make up the membranes inter-
nal skeleton. The band 3 dimer shown here is simplified. The band 4.1 protein stabilizes actin
spectrin complexes. (e) Electron micrograph showing the arrangement of the proteins of the inner
membrane skeleton. (A: COURTESY FRANOIS M. M. MOREL, RICHARD F. BAKER, AND HAROLD
WAYLAND, J. CELL BIOL. 48:91, 1971; B: COURTESY OF JOSEPH F. HOFFMAN; C: REPRODUCED, WITH
PERMISSION, FROM V. T. MARCHESI, H. F URTHMAYR, AND M. TOMITA, ANNU. REV. BIOCHEM. VOL. 45;
1976 BY ANNUAL REVIEWS INC.; D: AFTER D. VOET AND J. G. VOET, BIOCHEMISTRY, 2D ED.; COPYRIGHT
1995, JOHN WILEY & SONS, INC.; E: FROM SHIH-CHUN LIU, LAURA H. DERICK, AND JIRI PALEK, J. CELL
BIOL. 104:527, 1987; A, E: BY COPYRIGHT PERMISSION OF THE ROCKEFELLER UNIVERSITY PRESS.)
(e)
Once erythrocyte plasma membranes are isolated, the pro- borne on sialic acid, the sugar residue at the end of each car-
teins can be solubilized and separated from one another (frac- bohydrate chain. Because of these charges, red blood cells re-
tionated), providing a better idea of the diversity of proteins pel each other, which prevents the cells from clumping as they
within the membrane. Fractionation of membrane proteins circulate through the bodys tiny vessels. It is noteworthy that
can be accomplished using polyacrylamide gel electrophoresis persons who lack both glycophorin A and B in their red blood
(PAGE) in the presence of the ionic detergent sodium dode- cells show no ill-effects from their absence. At the same time,
cyl sulfate (SDS). (The technique of SDSPAGE is discussed the band 3 proteins in these individuals are more heavily gly-
in Section 18.7.) The SDS keeps the integral proteins soluble cosylated, which apparently compensates for the otherwise
and, in addition, adds a large number of negative charges to missing negative charges required to prevent cellcell interac-
the proteins with which it associates. Because the number of tion. Glycophorin also happens to be the receptor utilized by
charged SDS molecules per unit weight of protein tends to be the protozoan that causes malaria, providing a path for entry
relatively constant, the molecules separate from one another into the blood cell. Consequently, individuals whose erythro-
according to their molecular weight. The largest proteins cytes lack glycophorin A and B are thought to be protected
move most slowly through the molecular sieve of the gel. The from acquiring malaria. Differences in glycophorin amino
major proteins of the erythrocyte membrane are separated acid sequence determine whether a person has an MM, MN,
into about a dozen conspicuous bands by SDSPAGE (Fig- or NN blood type.
ure 4.32c). Among the proteins are a variety of enzymes (in-
cluding glyceraldehyde 3-phosphate dehydrogenase, one of The Erythrocyte Membrane Skeleton The peripheral pro-
the enzymes of glycolysis), transport proteins (for ions and teins of the erythrocyte plasma membrane are located on its
sugars), and skeletal proteins (e.g., spectrin). internal surface and constitute a fibrillar membrane skeleton
(Figure 4.32d,e) that plays a major role in determining the bi-
Integral Proteins of the Erythrocyte Membrane A model concave shape of the erythrocyte. As discussed on page 139,
of the erythrocyte plasma membrane showing its major pro- the membrane skeleton can establish domains within the
teins is seen in Figure 4.32d. The most abundant integral pro- membrane that enclose particular groups of membrane pro-
teins of this membrane are a pair of carbohydrate-containing, teins and may greatly restrict the movement of these proteins.
membrane-spanning proteins, called band 3 and glycophorin The major component of the skeleton is an elongated fibrous
A. The high density of these proteins within the membrane is protein, called spectrin. Spectrin is a heterodimer approxi-
evident in the freeze-fracture micrographs of Figure 4.15. mately 100 nm long, consisting of an # and $ subunit that curl
Band 3, which gets its name from its position in an elec- around one another. Two such dimeric molecules are linked at
trophoretic gel (Figure 4.32c), is present as a dimer composed their head ends to form a 200-nm-long filament that is both
of two identical subunits (a homodimer). Each subunit spans flexible and elastic. Spectrin is attached to the internal surface
the membrane at least a dozen times and contains a relatively of the membrane by means of noncovalent bonds to another
small amount of carbohydrate (68 percent of the molecules peripheral protein, ankyrin (the green spheres of Figure 4.32d),
weight). Band 3 protein serves as a channel for the passive ex- which in turn is linked noncovalently to the cytoplasmic do-
change of anions across the membrane. As blood circulates main of a band 3 molecule. As evident in Figures 4.32d and e,
through the tissues, carbon dioxide becomes dissolved in the spectrin filaments are organized into hexagonal or pentagonal
fluid of the bloodstream (the plasma) and undergoes the arrays. This two-dimensional network is constructed by link-
following reaction: ing both ends of each spectrin filament to a cluster of proteins
that include a short filament of actin and tropomyosin, proteins
H2O " CO2 S H2CO3 S HCO!
3 " H
"
typically involved in contractile activities. A number of ge-
The bicarbonate ions (HCO! 3 ) enter the erythrocyte in ex- netic diseases (hemolytic anemias) characterized by fragile,
change for chloride ions, which leave the cell. In the lungs, abnormally shaped erythrocytes have been traced to muta-
where carbon dioxide is released, the reaction is reversed and tions in ankyrin or spectrin.
bicarbonate ions leave the erythrocyte in exchange for chlo- If the peripheral proteins are removed from erythrocyte
ride ions. The reciprocal movement of HCO! !
3 and Cl occurs ghosts, the membrane becomes fragmented into small vesi-
through a channel in the center of each band 3 dimer. cles, indicating that the inner protein network is required to
Glycophorin A was the first membrane protein to have its maintain the integrity of the membrane. Erythrocytes are cir-
amino acid sequence determined. The arrangement of the culating cells that are squeezed under pressure through micro-
polypeptide chain of glycophorin A within the plasma mem- scopic capillaries whose diameter is considerably less than that
brane is shown in Figure 4.18. (Other related glycophorins, of the erythrocytes themselves. To traverse these narrow pas-
B, C, D, and E, are also present in the membrane at much lower sageways, and to do so day after day, the red blood cell must be
concentrations.) Like band 3, glycophorin A is also present in highly deformable, durable, and capable of withstanding
the membrane as a dimer. Unlike band 3, each glycophorin shearing forces that tend to pull it apart. The spectrinactin
A subunit spans the membrane only once, and it contains a network gives the cell the strength, elasticity, and pliability
bushy carbohydrate cover consisting of 16 oligosaccharide necessary to carry out its demanding function.
chains that together make up about 60 percent of the mole- When first discovered, the membrane skeleton of the ery-
cules weight. It is thought that the primary function of gly- throcyte was thought to be a unique structure suited to the
cophorin derives from the large number of negative charges unique shape and mechanical needs of this cell type. However,
4.7 THE MOVEMENT OF SUBSTANCES ACROSS CELL MEMBRANES 143
as other cells were examined, similar types of membrane Passive Active

skeletons containing members of the spectrin and ankyrin
families have been revealed, indicating that inner membrane Nonmediated Transporter mediated
skeletons are widespread. Dystrophin, for example, is a mem-
ber of the spectrin family that is found in the membrane A B C D
skeleton of muscle cells. Mutations in dystrophin are respon-

sible for causing muscular dystrophy, a devastating disease
that cripples and kills children. As in the case of cystic fibrosis
(page 157), the most debilitating mutations are ones that lead ATP ADP
+
to a complete absence of the protein in the cell. The plasma A B C D Pi
membranes of muscle cells lacking dystrophin are apparently (a) (b) (c) (d)
destroyed as a consequence of the mechanical stress exerted
on them as the muscle contracts. As a result, the muscle cells
O2 Na+ Glucose Na+ K+
die and eventually are no longer replaced.
REVIEW
? ATP ADP+Pi
1. Describe two techniques to measure the rates of diffusion Na+ Na+
O2 Glucose K+
of a specific membrane protein.
(e)
2. Compare and contrast the types of protein mobility
depicted in Figure 4.28. FIGURE 4.33 Four basic mechanisms by which solute molecules move
across membranes. The relative sizes of the letters indicate the direc-
3. Discuss two major functions of the integral and peripheral
tions of the concentration gradients. (a) Simple diffusion through the
proteins of the erythrocyte membrane. bilayer, which always proceeds from high to low concentration. (b) Sim-
4. Compare the rate of lateral diffusion of a lipid with that ple diffusion through an aqueous channel formed within an integral
of flip-flop. What is the reason for the difference? membrane protein or a cluster of such proteins. As in a, movement is al-
ways down a concentration gradient. (c) Facilitated diffusion in which
solute molecules bind specifically to a membrane protein carrier (a facil-
itative transporter). As in a and b, movement is always from high to low
4.7 THE MOVEMENT OF SUBSTANCES concentration. (d) Active transport by means of a protein transporter
with a specific binding site that undergoes a change in affinity driven
ACROSS CELL MEMBRANES with energy released by an exergonic process, such as ATP hydrolysis.
Movement occurs against a concentration gradient. (e) Examples of each
Because the contents of a cell are completely surrounded
type of mechanism as it occurs in the membrane of an erythrocyte.
by its plasma membrane, all communication between the
cell and the extracellular medium must be mediated by this
structure. In a sense, the plasma membrane has a dual function.
On one hand, it must retain the dissolved materials of the cell so The Energetics of Solute Movement
that they do not simply leak out into the environment, while on
Diffusion is a spontaneous process in which a substance
the other hand, it must allow the necessary exchange of materi-
moves from a region of high concentration to a region of low
als into and out of the cell. The lipid bilayer of the membrane is
concentration, eventually eliminating the concentration dif-
ideally suited to prevent the loss of charged and polar solutes
ference between the two regions. As discussed on page 87, dif-
from a cell. Consequently, some special provision must be made
fusion depends on the random thermal motion of solutes and
to allow the movement of nutrients, ions, waste products, and
is an exergonic process driven by an increase in entropy. We
other compounds, in and out of the cell. There are basically two
will restrict the following discussion to diffusion of substances
means for the movement of substances through a membrane:
across membranes.
passively by diffusion or actively by an energy-coupled transport
The free-energy change when an uncharged solute (a non-
process. Both types of movements lead to the net flux of a par-
electrolyte) diffuses across a membrane depends on the magni-
ticular ion or compound. The term net flux indicates that the
tude of the concentration gradient, that is, the difference in
movement of the substance into the cell (influx) and out of the
concentration on each side of the membrane. The following
cell (efflux) is not balanced, but that one exceeds the other.
relationship describes the movement of a nonelectrolyte into
Several different processes are known by which substances
the cell:
move across membranes: simple diffusion through the lipid
bilayer; simple diffusion through an aqueous, protein-lined [Ci]
G " RT ln
channel; diffusion that is facilitated by a protein transporter; [Co]
and active transport, which requires an energy-driven protein [Ci]
pump capable of moving substances against a concentration G " 2.303 RT log10
[Co]
gradient (Figure 4.33). We will consider each in turn, but first
we will discuss the energetics of solute movement. where !G is the free-energy change (Section 3.1), R is the gas
constant, T is the absolute temperature, and [Ci]/[Co] is the The interplay between concentration and potential dif-
ratio of the concentration of the solute on the inside (i) and ferences is seen in the diffusion of potassium ions (K!) out of
outside (o) surfaces of the membrane. At 25"C, a cell. The efflux of the ion is favored by the K! concentration
[Ci] gradient, which has a higher K! concentration inside the cell,
G $ 1.4 kcal/mol # log10 but hindered by the electrical gradient that its diffusion cre-
[Co] ates, which leaves a higher negative charge inside the cell. We
If the ratio of [Ci]/[Co] is less than 1.0, then the log of the will discuss this subject further when we consider the topic of
ratio is negative, #G is negative, and the net influx of solute membrane potentials and nerve impulses in Section 4.8.
is thermodynamically favored (exergonic). If, for example,
the external concentration of solute is 10 times the internal Diffusion of Substances through Membranes
concentration, #G $ %1.4 kcal/mole. Thus, the maintenance
of a tenfold concentration gradient represents a storage of Two qualifications must be met before a nonelectrolyte can
1.4 kcal/mol. As solute moves into the cell, the concentration diffuse passively across a plasma membrane. The substance
gradient decreases, the stored energy is dissipated, and #G must be present at higher concentration on one side of the
decreases until, at equilibrium, #G is zero. (To calculate #G for membrane than the other, and the membrane must be perme-
movement of a solute out of the cell, the term for concentra- able to the substance. A membrane may be permeable to a
tion ratios becomes [Co]/[Ci]). given solute either (1) because that solute can pass directly
If the solute is an electrolyte (a charged species), the over- through the lipid bilayer, or (2) because that solute can tra-
all charge difference between the two compartments must also verse an aqueous pore that spans the membrane. Let us begin
be considered. As a result of the mutual repulsion of ions of like by considering the former route in which a substance must
charges, it is thermodynamically unfavorable for an electrolyte dissolve in the lipid bilayer on its way through the membrane.
to move across a membrane from one compartment into an- Discussion of simple diffusion leads us to consider the
other compartment having a net charge of the same sign. Con- polarity of a solute. One simple measure of the polarity (or
versely, if the charge of the electrolyte is opposite in sign to the nonpolarity) of a substance is its partition coefficient, which
compartment into which it is moving, the process is thermody- is the ratio of its solubility in a nonpolar solvent, such as
namically favored. The greater the difference in charge (the octanol or a vegetable oil, to that in water under conditions
potential difference or voltage) between the two compart- where the nonpolar solvent and water are mixed together. Fig-
ments, the greater the difference in free energy. Thus, the ten- ure 4.34 shows the relationship between partition coefficient
dency of an electrolyte to diffuse between two compartments and membrane permeability of a variety of chemicals and
depends on two gradients: a chemical gradient, determined by drugs. It is evident that the greater the lipid solubility, the
the concentration difference of the substance between the two faster the penetration.
compartments, and the electric potential gradient, determined Another factor determining the rate of penetration of a
by the difference in charge. Together these differences are compound through a membrane is its size. If two molecules
combined to form an electrochemical gradient. The free-en- have approximately equivalent partition coefficients, the
ergy change for the diffusion of an electrolyte into the cell is smaller molecule tends to penetrate the lipid bilayer of a
membrane more rapidly than the larger one. Very small, un-
[Ci]
G $ RT ln ! zFEm charged molecules penetrate very rapidly through cellular
[Co] membranes. Consequently, membranes are highly permeable
where z is the charge of the solute, F is the Faraday constant to small inorganic molecules, such as O2, CO2, NO, and H2O,
(23.06 kcal/V ! equivalent, where an equivalent is the amount which are thought to slip between adjacent phospholipids. In
of the electrolyte having one mole of charge), and #Em is the contrast, larger polar molecules, such as sugars, amino acids,
potential difference (in volts) between the two compartments. and phosphorylated intermediates, exhibit poor membrane
We saw in the previous example that a tenfold difference in penetrability. As a result, the lipid bilayer of the plasma mem-
concentration of a nonelectrolyte across a membrane at 25"C brane provides an effective barrier that keeps these essential
generates a #G of %1.4 kcal/mol. Suppose the concentration metabolites from diffusing out of the cell. Some of these mol-
gradient consisted of Na! ions, which were present at tenfold ecules (e.g., sugars and amino acids) must enter cells from the
higher concentration outside the cell than in the cytoplasm. bloodstream, but they cannot do so by simple diffusion. In-
Because the voltage across the membrane of a cell is typically stead, special mechanisms must be available to mediate their
about %70 mV (page 160), the free-energy change for the penetration through the plasma membrane. The use of such
movement of a mole of Na! ions into the cell under these mechanisms allows a cell to regulate the movement of sub-
conditions would be stances across its surface barrier. We will return to this feature
of membranes later.
G $ %1.4 kcal/mol ! zFEm
G $ %1.4 kcal/mol ! (1)(23.06 kcal/V # mol) (%0.07 V) The Diffusion of Water through Membranes Water mole-
$ %3.1 kcal/mol cules move much more rapidly through a cell membrane than
Thus under these conditions, the concentration difference and do dissolved ions or small polar organic solutes, which are es-
the electric potential make similar contributions to the storage sentially nonpenetrating. Because of this difference in the
of free energy across the membrane. penetrability of water versus solutes, membranes are said to be
10
-3
semipermeable. Water moves readily through a semiperme-
Cerebrovascular permeability (cm / sec)
able membrane from a region of lower solute concentration to

10
-4
~ * a region of higher solute concentration. This process is called
Z
Y osmosis, and it is readily demonstrated by placing a cell into a
-5
T
W X solution containing a nonpenetrating solute at a concentration
10 U V
S different than that present within the cell itself.
R
P Q O
When two compartments of different solute concentration
N
are separated by a semipermeable membrane, the compartment
-6
10
K M
I L
H
J
G
of higher solute concentration is said to be hypertonic (or
F
10
-7 C E hyperosmotic) relative to the compartment of lower solute
D
concentration, which is described as being hypotonic (or
A B
-8
hypoosmotic). When a cell is placed into a hypotonic solu-
10
10
-4
10
-3
10
-2
10
-1
1 10 100 1000 tion, the cell rapidly gains water by osmosis and swells (Fig-
Octanol - water partition coefficient ure 4.35a). Conversely, a cell placed into a hypertonic solution
A. Sucrose J. Vinblastine S. Misonidazole rapidly loses water by osmosis and shrinks (Figure 4.35b).
B. Epipodophyllotoxin K. Curare T. Propylene glycol
C. Mannitol L. Thiourea U. Metronidazole These simple observations show that a cells volume is con-
D. Arabinose M. Dianhydrogalacticol V. Spirohydantoin mustard
E. N-methyl nicotinamide N. Glycerol W. Procarbazine
trolled by the difference between the solute concentration in-
F. Methotrexate O. 5- FU X. PCNU side the cell and that in the extracellular medium. The
G. Vincristine P. Ethylene glycol Y. Antipyrine
H. Urea Q. Acetamide Z. Caffeine swelling and shrinking of cells in slightly hypotonic and hy-
I. Formamide R. Ftorafur ~. BCNU
*. CCNU pertonic media are usually only temporary events. Within a
few minutes, the cells recover and return to their original vol-
FIGURE 4.34 The relationship between partition coefficient and mem-
ume. In a hypotonic medium, recovery occurs as the cells lose
brane permeability. In this case, measurements were made of the pene-
tration of a variety of chemicals and drugs across the plasma membranes ions, thereby reducing their internal osmotic pressure. In a hy-
of the cells that line the capillaries of the brain. Substances penetrate by pertonic medium, recovery occurs as the cells gain ions from
passage through the lipid bilayer of these cells. The partition coefficient the medium. Once the internal solute concentration (which
is expressed as the ratio of solubility of a solute in octanol to its solubility includes a high concentration of dissolved proteins) equals the
in water. Permeability is expressed as penetrance (P) in cm/sec. For all external solute concentration, the internal and external fluids
but a few compounds, such as vinblastine and vincristine, penetrance are isotonic (or isosmotic), and no net movement of water
is directly proportional to lipid solubility. (FROM N. J. ABBOTT AND into or out of the cells occurs (Figure 4.35c).
I. A. ROMERO, MOLEC. MED. TODAY, 2:110, 1996; COPYRIGHT 1996, WITH Osmosis is an important factor in a multitude of bodily
PERMISSION FROM ELSEVIER SCIENCE.)
functions. Your digestive tract, for example, secretes several
liters of fluid daily, which is reabsorbed osmotically by the
cells that line your intestine. If this fluid werent reabsorbed, as
happens in cases of extreme diarrhea, you would face the
prospect of rapid dehydration. Plants utilize osmosis in differ-
(a) Hypotonic solution (b) Hypertonic solution (c) Isotonic solution
H2O
H2O H2O H2O H2O
H2O H2O
H2O
H2O
H2O H2O H2O H2O H2O H2O

H2O
Net water gain Net water loss No net loss or gain

Cell swells Cell shrinks
FIGURE 4.35 The effects of differences in the concentration of solutes hypertonic solution shrinks because of a net loss of water by osmosis.
on opposite sides of the plasma membrane. (a) A cell placed in a (c) A cell placed in an isotonic solution maintains a constant volume
hypotonic solution (one having a lower solute concentration than the because the inward flux of water is equal to the outward flux.
cell) swells because of a net gain of water by osmosis. (b) A cell in a
ent ways. Unlike animal cells, which are generally isotonic brane proteins responsible for the Rh antigen on the surface of
with the medium in which they are bathed, plant cells are gen- red blood cells. During this pursuit, they identified a protein
erally hypertonic compared to their fluid environment. As a they thought might be the long-sought water channel of the
result, there is a tendency for water to enter the cell, causing it erythrocyte membrane. To test their hypothesis, they engi-
to develop an internal (turgor) pressure that pushes against its neered frog oocytes to incorporate the newly discovered pro-
surrounding wall (Figure 4.36a). Turgor pressure provides tein into their plasma membranes and then placed the oocytes
support for nonwoody plants and for the nonwoody parts of in a hypotonic medium. Just as predicted, the oocytes swelled
trees, such as the leaves. If a plant cell is placed into a hyper- due to the influx of water and eventually burst. The team had
tonic medium, its volume shrinks as the plasma membrane discovered a family of small integral proteins, called aquapor-
pulls away from the surrounding cell wall, a process called ins, that allow the passive movement of water from one side of
plasmolysis (Figure 4.36b). The loss of water due to plasmol- the plasma membrane to the other. Each aquaporin subunit
ysis causes plants to lose their support and wilt. (in the four-subunit protein) contains a central channel that is
Many cells are much more permeable to water than can lined primarily by hydrophobic amino acid residues and is
be explained by simple diffusion through the lipid bilayer. In highly specific for water molecules. A billion or so water mol-
the early 1990s, Peter Agre and colleagues at Johns Hopkins ecules can passin single filethrough each channel every
University were attempting to isolate and purify the mem- second. At the same time, H! ions, which normally hop along
a chain of water molecules, are not able to penetrate these
open pores. The apparent mechanism by which these channels
are able to exclude protons has been suggested by a combina-
HYPOTONIC: tion of X-ray crystallographic studies, which has revealed the
Normal turgor
pressure
structure of the protein, and computer-based simulations
(page 59), which has put this protein structure into operation.
A model based on computer simulations is shown in Fig-
ure 4.37a. Very near its narrowest point, the wall of an aqua-
porin channel contains a pair of precisely positioned positive
charges (residues N203 and N68 in Figure 4.37b) that attract
the oxygen atom of each water molecule as it speeds through
the constriction in the protein. This interaction reorients the
central water molecule in a position that prevents it from
maintaining the hydrogen bonds that normally link it to its
neighboring water molecules. This removes the bridge that
would normally allow protons to move from one water mole-
Plasmolysis
cule to the next.

(a) Aquaporins are particularly prominent in cells, such as
H 20
those of a kidney tubule or plant root, where the passage of
water plays a crucial role in the tissues physiologic activities.
The hormone vasopressin, which stimulates water retention
by the collecting ducts of the kidney, acts by way of one of
these proteins (AQP2). Some cases of the inherited disorder
congenital nephrogenic diabetes insipidus arise from mutations in
this aquaporin channel. Persons suffering from this disease
excrete large quantities of urine because their kidneys do not
respond to vasopressin.
The Diffusion of Ions through Membranes The lipid bi-

layer that constitutes the core of biological membranes is
HYPERTONIC:
No turgor highly impermeable to charged substances, including small
pressure ions such as Na!, K!, Ca2!, and Cl". Yet the rapid move-
ment (conductance) of these ions across membranes plays a
critical role in a multitude of cellular activities, including
(b) formation and propagation of a nerve impulse, secretion of
substances into the extracellular space, muscle contraction,
FIGURE 4.36 The effects of osmosis on a plant cell. (a) Aquatic regulation of cell volume, and the opening of stomatal pores
A plants living in freshwater are surrounded by a hypotonic envi-
on plant leaves.
ronment. Water therefore tends to flow into the cells, creating turgor In 1955, Alan Hodgkin and Richard Keynes of Cam-
pressure. (b) If the plant is placed in a hypertonic solution, such as sea-
bridge University first proposed that cell membranes contain
water, the cell loses water, and the plasma membrane pulls away from the
cell wall. ( ED RESCHKE.)
ion channels, that is, openings in the membrane that are
permeable to specific ions. During the late 1960s and 1970s,
FIGURE 4.37 Passage of water molecules

through an aquaporin channel. (a) Snap-
shot from a molecular dynamics simulation
of a stream of water molecules (red and white
spheres) passing in single file through the
channel in one of the subunits of an aqua-
porin molecule residing within a membrane.
(b) A model describing the mechanism by
which water molecules pass through an
aquaporin channel with the simultaneous
exclusion of protons. Nine water molecules
are shown to be lined up in single file along
the wall of the channel. Each water molecule
is depicted as a red circular O atom with two
associated Hs. In this model, the four water
molecules at the top and bottom of the
channel are oriented, as the result of their
interaction with the carbonyl (CPO) groups
of the protein backbone (page 50), with their
H atoms pointed away from the center of the
channel. These water molecules are able to
form hydrogen bonds (dashed lines) with
(a)
their neighbors. In contrast, the single water
molecule in the center of the channel is
oriented in a position that prevents it from forming hydrogen AND KLAUS SCHULTEN, STRUCTURE
bonds with other water molecules, which has the effect of interrupting 12:1344, 2004, BY PERMISSION OF (b)
the flow of protons through the channel. Animations of aquaporin CELL PRESS; B: REPRINTED FROM
channels can be found at www.nobelprize.org/nobel_prizes/chemistry/ R. M. STROUD ET AL., CURR. OPIN. STRUCT. BIOL. 13:428, 2003. COPYRIGHT
laureates/2003/animations.html. (A: REPRINTED FROM BENOIT ROUX 2003, WITH PERMISSION FROM ELSEVIER SCIENCE.)
Bertil Hille of the University of Washington and Clay trodes made of polished glass that are placed on the outer
Armstrong of the University of Pennsylvania began to ob- cell surface and sealed to the membrane by suction. The
tain evidence for the existence of such channels. The final voltage across the membrane can be maintained (clamped ) at
proof emerged through the work of Bert Sakmann and Er- any particular value, and the current originating in the small
win Neher at the Max-Planck Institute in Germany in the patch of membrane surrounded by the pipette can be meas-
late 1970s and early 1980s who developed techniques to ured (Figure 4.38). These landmark studies marked the first
monitor the ionic current passing through a single ion chan- successful investigations into the activities of individual pro-
nel. This is accomplished using very fine micropipette-elec- tein molecules. Today, biologists have identified a bewilder-
FIGURE 4.38 Measuring ion channel conductance by

patch-clamp recording. (a) In this technique, a highly
polished glass micropipette is placed against a portion
of the outer surface of a cell, and suction is applied to
seal the rim of the pipette against the plasma mem-
brane. Because the pipette is wired as an electrode (a
microelectrode), a voltage can be applied across the patch
of membrane enclosed by the pipette, and the respond-
ing flow of ions through the membrane channels can be
measured. As indicated in the figure, the micropipette
can enclose a patch of membrane containing a single
ion channel, which allows investigators to monitor the
opening and closing of a single gated channel, as well as
its conductance at different applied voltages. (b) The
micrograph shows patch-clamp recordings being made
from a single photoreceptor cell of the retina of a sala- (a) (b) 35
mander. One portion of the cell is drawn into a glass
micropipette by suction, while a second micropipette-electrode (lower portion of the cell. (B: FROM T. D. LAMB, H. R. MATTHEWS, AND V. TORRE,
right) is sealed against a small patch of the plasma membrane on another J. PHYSIOLOGY 372:319, 1986. REPRODUCED WITH PERMISSION.)
ing variety of ion channels, each formed by integral mem-

brane proteins that enclose a central aqueous pore. As might G O O G
K+
be predicted, mutations in the genes encoding ion channels Y O O Y
can lead to many serious diseases (see Table 1 of the Human G O O G
K+
Perspective, page 156). V O O V
Most ion channels are highly selective in allowing only
one particular type of ion to pass through the pore. As with the T O O T
passive diffusion of other types of solutes across membranes, K+
the diffusion of ions through a channel is always downhill,

that is, from a state of higher energy to a state of lower energy. Selectivity
Filter
Most of the ion channels that have been identified can exist in
either an open or a closed conformation; such channels are
said to be gated. The opening and closing of the gates are sub-
ject to complex physiologic regulation and can be induced by P
a variety of factors depending on the particular channel. Three
major categories of gated channels are distinguished:
1. Voltage-gated channels whose conformational state de-
K+
pends on the difference in ionic charge on the two sides ions
of the membrane.
2. Ligand-gated channels whose conformational state de-
pends on the binding of a specific molecule (the ligand),
which is usually not the solute that passes through the
channel. Some ligand-gated channels are opened (or
closed) following the binding of a molecule to the outer
surface of the channel; others are opened (or closed) fol-
lowing the binding of a ligand to the inner surface of the M2 M1
helix helix
channel. For example, neurotransmitters, such as acetyl-
choline, act on the outer surface of certain cation chan-
nels, while cyclic nucleotides, such as cAMP, act on the
inner surface of certain calcium ion channels.
3. Mechano-gated channels whose conformational state FIGURE 4.39 Three-dimensional structure of the bacterial KcsA
depends on mechanical forces (e.g., stretch tension) that channel and the selection of K! ions. This K! ion channel consists of
four subunits, two of which are shown here. Each subunit is comprised
are applied to the membrane. Members of one family of
of M1 and M2 helices joined by a P (pore) segment consisting of a short
cation channels, for example, are opened by the move- helix and a nonhelical portion that lines the channel through which the
ments of stereocilia (see Figure 9.54) on the hair cells of ions pass. A portion of each P segment contains a conserved pentapep-
the inner ear in response to sound or motions of the head. tide (GYGVT) whose residues line the selectivity filter that screens for
K! ions. The oxygen atoms of the carbonyl groups of these residues
We will focus in the following discussion on the structure and project into the channel where they can interact selectively with K! ions
function of voltage-gated potassium ion channels because (indicated by the red mesh objects) within the filter. As indicated in the
these are the best understood. top inset, the selectivity filter contains four rings of carbonyl O atoms
In 1998, Roderick MacKinnon and his colleagues at and one ring of threonyl O atoms; each of these five rings contains four
Rockefeller University provided the first atomic-resolution O atoms, one donated by each subunit. The diameter of the rings is just
image of an ion channel protein, in this case, a bacterial K! large enough so that eight O atoms can coordinate a single K! ion, re-
ion channel called KcsA. The relationship between structure placing its normal water of hydration. Although four K! binding sites are
and function is evident everywhere in the biological world, shown, only two are occupied at one time. (FROM RODERICK MACKINNON,
REPRINTED WITH PERMISSION FROM NATURE MED. 5:1108, 1999; COPYRIGHT
but it would be difficult to find a better example than that of
1999, MACMILLAN MAGAZINES LIMITED.)
the K! ion channel depicted in Figure 4.39. As we will see
shortly, the formulation of this structure led directly to an un-
derstanding of the mechanism by which these remarkable
molecular machines are able to select overwhelmingly for K!
ions over Na! ions, yet at the same time allow an incredibly relatively early in evolution, although many refinements ap-
rapid conductance of K! ions through the membrane. We will peared over the following one or two billion years.
also see that the mechanisms of ion selectivity and conduc- The KcsA channel consists of four subunits, two of which
tance in this bacterial channel are virtually identical to those are shown in Figure 4.39. Each subunit of Figure 4.39 is seen
operating in the much larger mammalian channels. Evidently, to contain two membrane-spanning helices (M1 and M2) and
the basic challenges in operating an ion channel were solved a pore region (P) at the extracellular end of the channel. P
consists of a short pore helix that extends approximately one- Closed Open
third the width of the channel and a nonhelical loop (colored
light brown in Figure 4.39) that forms the lining of a narrow Hinge
selectivity filter, so named because of its role in allowing only
the passage of K! ions. M2
The lining of the selectivity filter contains a highly con- helix
served pentapeptideGly-Tyr-Gly-Val-Thr (or GYGVT in
FIGURE 4.40 Schematic illustration of the hinge-bending
single-letter nomenclature). The X-ray crystal structure of the
model for the opening of the bacterial KcsA channel. The M2
KcsA channel shows that the backbone carbonyl (CPO) helices from each subunit bend outward at a specific glycine residue,
groups from the conserved pentapeptide (see the backbone which opens the intracellular end of the channel to K! ions. (FROM B. L.
structure on page 50) create five successive rings of oxygen KELLY AND A. GROSS, REPRINTED WITH PERMISSION FROM NATURE STRUCT.
atoms (four rings are made up of carbonyl oxygens from the BIOL. 10:280, 2003; COPYRIGHT 2003, MACMILLAN MAGAZINES LIMITED.)
polypeptide backbone, and one ring consists of oxygen atoms
from the threonine side chain). Each ring contains four oxy-
gen atoms (one from each subunit) and has a diameter of ap-
proximately 3 , which is slightly larger than the 2.7 suggested that gating of these molecules is accomplished by
diameter of a K! ion that has lost its normal shell of hydra- conformational changes of the cytoplasmic ends of the inner
tion. Consequently, the electronegative O atoms that line the (M2) helices. In the closed conformation, as seen in Figure 4.39
selectivity filter can substitute for the shell of water molecules and Figure 4.40, left drawing, the M2 helices are straight and
that are displaced as each K! ion enters the pore. In this cross over one another to form a helix bundle that seals the cy-
model, the selectivity filter contains four potential K! ion toplasmic face of the pore. In the model shown in Figure 4.40,
binding sites. As indicated in the top inset of Figure 4.39, a the channel opens when the M2 helices bend at a specific hinge
K! ion bound at any of these four sites would occupy the cen- point where a glycine residue is located.
ter of a box having four O atoms in a plane above the ion and Now that we have seen how these prokaryotic K! chan-
four O atoms in a plane below the atom. As a result, each K! nels operate, we are in a better position to understand the
ion in one of these sites could coordinate with eight O atoms structure and function of the more complex eukaryotic ver-
of the selectivity filter. Whereas the selectivity filter is a precise sions, which are thought to perform in a similar manner.
fit for a dehydrated K! ion, it is much larger than the diame- Genes that encode a variety of distinct voltage-gated K! (or
ter of a dehydrated Na! ion (1.9 ). Consequently, a Na! ion Kv) channels have been isolated and the molecular anatomy of
cannot interact optimally with the eight oxygen atoms neces- their proteins scrutinized. The Kv channels of plants play an
sary to stabilize it in the pore. As a result, the smaller Na! ions important role in salt and water balance and in regulation of
cannot overcome the higher energy barrier required to pene- cell volume. The Kv channels of animals are best known for
trate the pore. their role in muscle and nerve function, which is explored at
Although there are four potential K! ion binding sites, the end of the chapter. Eukaryotic Kv channel subunits con-
only two are occupied at any given time. Potassium ions are tain six membrane-associated helices, named S1S6, which
thought to move, two at a timefrom sites 1 and 3 to sites are shown two-dimensionally in Figure 4.41. These six helices
2 and 4as indicated in the top inset of Figure 4.39. The can be grouped into two functionally distinct domains:
entry of a third K! ion into the selectivity filter creates an 1. a pore domain, which has the same basic architecture as that
electrostatic repulsion that ejects the ion bound at the oppo- of the entire bacterial channel illustrated in Figure 4.39 and
site end of the line. Studies indicate that there is virtually no contains the selectivity filter that permits the selective pas-
energy barrier for an ion to move from one binding site to the sage of K! ions. Helices M1 and M2 and the P segment of
next, which accounts for the extremely rapid flow of ions the KcsA channel of Figure 4.39 are homologous to helices
across the membrane. Taken together, these conclusions con- S5 and S6 and the P segment of the voltage-gated eukary-
cerning K! ion selectivity and conductance provide a superb otic channel illustrated in Figure 4.41. Like the four M2
example of how much can be learned about biological func- helices of KcsA, the four S6 helices line much of the pore,
tion through an understanding of molecular structure. and their configuration determines whether the gate to the
The KcsA channel depicted in Figure 4.39 has a gate, just channel is open or closed.
like eukaryotic channels. The opening of the gate of the KcsA
2. a voltage-sensing domain consisting of helices S1S4
channel in response to very low pH was illustrated in Fig-
that senses the voltage across the plasma membrane (as
ure 4.22. The structure of KcsA shown in Figure 4.39 is actually
discussed below).
the closed conformation of the protein (despite the fact that it
contains ions in its channel). It has not been possible to crystal- The three-dimensional crystal structure of a complete
lize the KcsA channel in its open conformation, but the struc- eukaryotic Kv channel purified from rat brain is shown in Fig-
ture of a homologous prokaryotic K! channel (called MthK) in ure 4.42. Determination of this structure was made possible
the open conformation has been crystallized and its structure by use of a mixture of detergent and lipid throughout the pu-
determined. Comparison of the open structure of MthK and rification and crystallization process. The presence of nega-
the closed structure of the homologous protein KcsA strongly tively charged phospholipids is thought to be important in
FIGURE 4.41 The structure of one subunit of a S4 Segment I

eukaryotic, voltage-gated K! channel. A + L
two-dimensional portrait of a K! channel sub- R
unit showing its six transmembrane helices and V

+ I
a portion of the polypeptide (called the Voltage-sensing R
L
pore helix or P) that dips into the protein to domain Pore domain V
+
form part of the channels wall. The inset Outside cell R
V
shows the sequence of amino acids of the posi-
+ P F +
tively charged S4 helix of the Drosophila K! + R
Shaker ion channel, which serves as a voltage S1 S2 S3 S4 S5 S6 I

+ F
sensor. The positively charged side chains are + K
+
situated at every third residue along the other- Inside cell L
wise hydrophobic helix. This member of the + S
R
H
Kv family is called a Shaker channel because
N S
flies with certain mutations in the protein shake C +
K
vigorously when anesthesized with ether. The G
Shaker channel was the first K! channel to be L
identified and cloned in 1987.
maintaining the native structure of the membrane protein and presumed mechanism of K! ion selection, is virtually identi-
promoting its function as a voltage-gated channel. Like the cal in the prokaryotic KcsA and eukaryotic Kv proteins. The
KcsA channel, a single eukaryotic Kv channel consists of four gate leading into a Kv channel is formed by the inner ends of
homologous subunits arranged symmetrically around the cen- the S6 helices and is thought to open and close in a manner
tral ion-conducting pore. The selectivity filter, and thus the roughly similar to that of the M2 helices of the bacterial
(a) (b) (c)
FIGURE 4.42 Three-dimensional structure of a voltage-gated mam- ! polypeptide. (c) Ribbon drawing of a single subunit showing the spa-
malian K! channel. (a) The crystal structure of the entire tetrameric tial orientation of the six membrane-spanning helices (S1S6) and also
Kv1.2 channel, a member of the Shaker family of K! ion channels found the presence of the S4S5 linker helix, which connects the voltage-sensing
in nerve cells of the brain. The transmembrane portion is shown in red, and pore domains. This linker transmits the signal from the S4 voltage
and the cytoplasmic portion in blue. The potassium ion binding sites are sensor that opens the channel. The inner surface of the channel below
indicated in green. (b) Ribbon drawing of the same channel shown in a, the pore domain is lined by the S6 helix (roughly similar to the M2 he-
with the four subunits that make up the channel shown in different col- lix of the bacterial channel shown in Figure 4.39). The channel shown
ors. If you focus on the red subunit, you can see (1) the spatial separation here is present in the open configuration with the S6 helices curved out-
between the voltage-sensing and pore domains of the subunit and ward (compare to Figure 4.40) at the site marked PVP (standing for
(2) the manner in which the voltage-sensing domain from each subunit Pro-Val-Pro, which is likely the amino acid sequence of the hinge).
is present on the outer edge of the pore domain of a neighboring sub- (REPRINTED WITH PERMISSION FROM STEPHEN B. LONG ET AL., SCIENCE
unit. The cytoplasmic portion of this particular channel consists of a T1 309:867, 899, 2005, COURTESY OF RODERICK MACKINNON; COPYRIGHT
domain, which is part of the channel polypeptide itself, and a separate 2005, AMERICAN ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE.)
channel (shown in Figure 4.40). The protein depicted in Fig- Pore

ure 4.42 represents the open state of the channel.
The S4 helix, which contains several positively charged Outside cell
Central cavity
amino acid residues spaced along the polypeptide chain (inset
of Figure 4.41), acts as the key element of the voltage sensor.
The voltage-sensing domain is seen to be connected to the
pore domain by a short linker helix denoted as S4-S5 in
the model in Figure 4.42. Under resting conditions, the nega-
tive potential across the membrane (page 160) keeps the gate Cell (
closed. A change in the potential to a more positive value (a membrane
depolarization, page 161) exerts an electric force on the S4 he-
lix. This force is thought to cause the transmembrane S4 helix
to move in such a way that its positively charged residues shift
Inside cell
from a position where they were exposed to the cytoplasm to Lateral window
a new position where they are exposed to the outside of the
cell. Voltage sensing is a dynamic process whose mechanism
cannot be resolved by a single static view of the protein such
as that shown in Figure 4.42. In fact, several competing mod- Inactivation peptide
Cytoplasmic
els describing the mechanism of action of the voltage sensor domain
are currently debated. However it occurs, the movement of the
S4 helix in response to membrane depolarization initiates a
(a)
series of conformational changes within the protein that
opens the gate at the cytoplasmic end of the channel.
Once opened, more than ten million potassium ions can
pass through the channel per second, which is nearly the rate
that would occur by free diffusion in solution. Because of the
large ion flux, the opening of a relatively small number of K!
channels has significant impact on the electrical properties of
the membrane. After the channel is open for a few millisec-
onds, the movement of K! ions is automatically stopped by a Rest Open Inactivated
process known as inactivation. To understand channel inacti- (b)
vation, we have to consider an additional portion of a Kv chan-
nel besides the two transmembrane domains discussed above. FIGURE 4.43 Conformational states of a voltage-gated K! ion
Eukaryotic Kv channels typically contain a large cytoplas- channel. (a) Three-dimensional model of a eukaryotic K! ion channel.
mic structure whose composition varies among different Inactivation of channel activity occurs as one of the inactivation pep-
tides, which dangle from the cytoplasmic portion of the complex, fits
channels. As indicated in Figure 4.43a inactivation of the
into the cytoplasmic opening of the channel. (b) Schematic representa-
channel is accomplished by movement of a small inactivation tion of a view into a K! ion channel, perpendicular to the membrane
peptide that dangles from the cytoplasmic portion of the pro- from the cytoplasmic side, showing the channel in the closed (resting),
tein. The inactivation peptide is thought to gain access to the open, and inactivated state. (B: REPRINTED FROM NEURON, VOL. 20, C. M.
cytoplasmic mouth of the pore by snaking its way through one ARMSTRONG B. HILLE, VOLTAGE-GATED ION CHANNELS AND ELEC-
AND
of four side windows indicated in the figure. When one of TRICAL EXCITABILITY, PAGE
377; COPYRIGHT 1998, WITH PERMISSION FROM
these dangling peptides moves up into the mouth of the pore ELSEVIER SCIENCE.)
(Figure 4.43a), the passage of ions is blocked, and the channel
is inactivated. At a subsequent stage of the cycle, the inactiva-
tion peptide is released, and the gate to the channel is closed. ent that ion channel function is under the control of a diverse
It follows from this discussion that the potassium channel can and complex set of regulatory agents. The structure and func-
exist in three different statesopen, inactivated, and closed tion of a very different type of ion channel, the ligand-gated
which are illustrated schematically in Figure 4.43b. nicotinic acetylcholine receptor, is the subject of the Experi-
Potassium channels come in many different varieties. It is mental Pathways section at the end of this chapter.
remarkable that C. elegans, a nematode worm whose body con-
sists of only about 1000 cells, contains more than 90 different
Facilitated Diffusion
genes that encode K! channels. It is evident that a single cell
whether in a nematode, human, or plantis likely to possess a Substances always diffuse across a membrane from a region
variety of different K! channels that open and close in re- of higher concentration on one side to a region of lower con-
sponse to different voltages. In addition, the voltage required centration on the other side, but they do not always diffuse
to open or close a particular K! channel can vary depending on through the lipid bilayer or through a channel. In many cases, the
whether or not the channel protein is phosphorylated, which diffusing substance first binds selectively to a membrane-span-
in turn is regulated by hormones and other factors. It is appar- ning protein, called a facilitative transporter, that facilitates the
diffusion process. The binding of the solute to the facilitative Glucose

transporter on one side of the membrane is thought to trigger a
conformational change in the protein, exposing the solute to the
other surface of the membrane, from where it can diffuse down Binding
its concentration gradient. This mechanism is illustrated in Fig-
ure 4.44. Because they operate passively, that is, without being
coupled to an energy-releasing system, facilitated transporters
can mediate the movement of solutes equally well in both direc-
tions.The direction of net flux depends on the relative concentra-
tion of the substance on the two sides of the membrane.
Facilitated diffusion, as this process is called, is similar in
many ways to an enzyme-catalyzed reaction. Like enzymes, fa-
cilitative transporters are specific for the molecules they trans- Recovery Transport
port, discriminating, for example, between D and L
stereoisomers (page 43). In addition, both enzymes and trans-
porters exhibit saturation-type kinetics (Figure 4.45). Unlike ion
channels, which can conduct millions of ions per second, most
facilitative transporters can move only hundreds to thousands of
solute molecules per second across the membrane. Another im-
portant feature of facilitative transporters is that, like enzymes
and ion channels, their activity can be regulated. Facilitated dif-
fusion is particularly important in mediating the entry and exit Dissociation
of polar solutes, such as sugars and amino acids, that do not pen-
etrate the lipid bilayer. This is illustrated in the following section.
FIGURE 4.44 Facilitated diffusion. A schematic model for the facili-
The Glucose Transporter: An Example of Facilitated tated diffusion of glucose depicts the alternating conformation of a car-
Diffusion Glucose is the bodys primary source of direct energy, rier that exposes the glucose binding site to either the inside or outside
and most mammalian cells contain a membrane protein that fa- of the membrane. (AFTER S. A. BALDWIN AND G. E. LIENHARD, TRENDS
cilitates the diffusion of glucose from the bloodstream into the BIOCHEM. SCI. 6:210, 1981.)
cell (as depicted in Figures 4.44 and 4.49). A gradient favoring
the continued diffusion of glucose into the cell is maintained by
phosphorylating the sugar after it enters the cytoplasm, thus low- mammalian cell are shown in Table 4.3. The ability of a cell to
ering the intracellular glucose concentration. Humans have at generate such steep concentration gradients across its plasma
least five related proteins (isoforms) that act as facilitative glucose membrane cannot occur by either simple or facilitated diffusion.
transporters. These isoforms, termed GLUT1 to GLUT5, are Rather, these gradients must be generated by active transport.
distinguished by the tissues in which they are located, as well as Like facilitated diffusion, active transport depends on in-
their kinetic and regulatory characteristics. tegral membrane proteins that selectively bind a particular
Insulin is a hormone produced by endocrine cells of the solute and move it across the membrane in a process driven by
pancreas and plays a key role in maintaining proper blood changes in the proteins conformation. Unlike facilitated dif-
sugar levels. An increase in blood glucose levels triggers the fusion, however, movement of a solute against a gradient
secretion of insulin, which stimulates the uptake of glucose
into various target cells, most notably skeletal muscle and fat Vmax
cells (adipocytes). Insulin-responsive cells share a common
Rate of solute movement (v)
isoform of the facilitative glucose transporter, specifically

Protein-mediated transport
GLUT4. When insulin levels are low, these cells contain rela- (facilitated diffusion)
tively few glucose transporters on their plasma membrane.
Instead, the transporters are present within the membranes of
cytoplasmic vesicles. Rising insulin levels act on target cells to
stimulate the fusion of the cytoplasmic vesicles to the plasma
membrane, which moves transporters to the cell surface where
they can bring glucose into the cell (see Figure 15.24).
Simple diffusion
Active Transport
Life cannot exist under equilibrium conditions (page 91). Solute concentration
Nowhere is this more apparent than in the imbalance of ions
FIGURE 4.45 The kinetics of facilitated diffusion as compared to that
across the plasma membrane. The differences in concentration
of simple physical diffusion.
of the major ions between the outside and inside of a typical
TABLE 4.3 Ion Concentrations Inside and Outside of a Typical enzyme in the nerve cells of a crab that was only active in the
Mammalian Cell presence of both Na! and K! ions. Skou proposed, and cor-
rectly so, that this enzyme, which was responsible for ATP
Extracellular Intracellular Ionic
concentration concentration gradient hydrolysis, was the same protein that was active in transport-
ing the two ions; the enzyme was called the Na!/K!-ATPase,
Na! 150 mM 10 mM 15# or the sodiumpotassium pump.
K! 5 mM 140 mM 28# Unlike the protein-mediated movement of a facilitated
Cl 120 mM 10 mM 12#
diffusion system, which will carry the substance equally well in
Ca2! 103 M 107 M 10,000#
either direction, active transport drives the movement of ions
H! 107.4 M 107.2 M Nearly 2#
(pH of 7.4) (pH of 7.2) in only one direction. It is the Na!/K!-ATPase that is respon-
sible for the large excess of Na! ions outside of the cell and the
The ion concentrations for the squid axon are given on page 172. large excess of K! ions inside the cell. The positive charges car-
ried by these two cations are balanced by negative charges car-
ried by various anions so that the extracellular and intracellular
requires the coupled input of energy. Consequently, the ender- compartments are, for the most part, electrically neutral. Cl"
gonic movement of ions or other solutes across the membrane ions are present at greater concentration outside of cells, where
against a concentration gradient is coupled to an exergonic they balance the extracellular Na! ions. The abundance of in-
process, such as the hydrolysis of ATP, the absorbance of light, tracellular K! ions is balanced primarily by excess negative
the transport of electrons, or the flow of other substances charges carried by proteins and nucleic acids.
down their gradients. Proteins that carry out active transport A large number of studies have indicated that the ratio of
are often referred to as pumps. Na!!K! pumped by the Na!/K!-ATPase is not 1!1, but
3!2 (see Figure 4.46). In other words, for each ATP hy-
Coupling Active Transport to ATP Hydrolysis In 1957, Jens drolyzed, three sodium ions are pumped out as two potassium
Skou, a Danish physiologist, discovered an ATP-hydrolyzing ions are pumped in. Because of this pumping ratio, the
Extracellular space Na+
Na+
Na+
Na+
ADP
1 P
3
Na+ ATP Cytoplasm

E2 conformation P
E1 conformation
K+
P
K+
6
4
K+
K+
5
P
(a) (b)
A FIGURE 4.46 The Na /K -ATPase. (a) Simplified schematic

! ! ! !
Na /K -ATPase is composed of at least two different membrane-
model of the transport cycle. Sodium ions (1) bind to the protein spanning subunits: a larger $ subunit, which carries out the transport
on the inside of the membrane. ATP is hydrolyzed, and the phosphate is activity, and a smaller % subunit, which functions primarily in the matu-
transferred to the protein (2), changing its conformation (3) and allow- ration and assembly of the pump within the membrane. A third (&) sub-
ing sodium ions to be expelled to the external space. Potassium ions then unit may also be present. (Steps where ATP binds to the protein prior to
bind to the protein (4), and the phosphate group is subsequently lost hydrolysis are not included.) (b) A model of the E2 conformation of the
(5), which causes the protein to snap back to its original conformation, protein based on a recent X-ray crystallographic study. The two rubid-
allowing the potassium ions to diffuse into the cell (6). The cation ium ions are located where the potassium ions would normally be bound.
binding sites are located deep within the transmembrane domain, (B: FROM AYAKO TAKEUCHI COURTESY OF DAVID C. GADSBY, NATURE 450:958,
which consists of 10 membrane-spanning helices. Note that the actual 2007, COPYRIGHT 2007, MACMILLAN MAGAZINES LIMITED.)
Na!/K!-ATPase is electrogenic, which means that it con- drolysis of ATP to changes in access and affinity of the ion
tributes directly to the separation of charge across the mem- binding sites.
brane. The Na!/K!-ATPase is an example of a P-type ion The sodiumpotassium pump is found only in animal
pump. The P stands for phosphorylation, indicating that, cells. This protein is thought to have evolved in primitive an-
during the pumping cycle, the hydrolysis of ATP leads to the imals as the primary means to maintain cell volume and as the
transfer of the released phosphate group to an aspartic acid mechanism to generate the steep Na! and K! gradients that
residue of the transport protein, which in turn causes an es- play such a key role in the formation of impulses in nerve and
sential conformational change within the protein. Conforma- muscle cells. Plant cells have a H!-transporting, P-type,
tional changes are necessary to change the affinity of the plasma membrane pump. In plants, this proton pump plays a
protein for the two cations that are transported. Consider the key role in the secondary transport of solutes (discussed later),
activity of the protein. It must pick up sodium or potassium in the control of cytosolic pH, and possibly in control of cell
ions from a region of low concentration, which means that the growth by means of acidification of the plant cell wall.
protein must have a relatively high affinity for the ions. Then The epithelial lining of the stomach also contains a P-
the protein must release the ions on the other side of the type pump, the H!/K!-ATPase, which secretes a solution of
membrane into a much greater concentration of each ion. To concentrated acid (up to 0.16 N HCl) into the stomach cham-
do this, the affinity of the protein for that ion must decrease. ber. In the resting state, these pump molecules are situated in
Thus, the affinity for each ion on the two sides of the mem- cytoplasmic membranes of the parietal cells of the stomach
brane must be different. This is achieved by phosphorylation, lining and are nonfunctional (Figure 4.47). When food enters
which changes the shape of the protein molecule. The change the stomach, a hormonal message is transmitted to the pari-
in shape of the protein also serves to expose the ion binding etal cells that causes the pump-containing membranes to
sites to different sides of the membrane, as discussed in the move to the apical cell surface, where they fuse with the
following paragraph. plasma membrane and begin secreting acid (Figure 4.47). In
A general scheme for the pumping cycle of the Na!/K!- addition to functioning in digestion, stomach acid can also
ATPase is shown in Figure 4.46a. When the protein binds lead to heartburn. Prilosec is a widely used drug that prevents
three Na! ions on the inside of the cell (step 1) and becomes heartburn by inhibiting the stomachs H!/K!-ATPase. Other
phosphorylated (step 2), it shifts from the E1 conformation to acid-blocking heartburn medications (e.g., Zantac, Pepcid,
the E2 conformation (step 3). In doing so, the binding site be- and Tagamet) do not inhibit the H!/K!-ATPase directly, but
comes exposed to the extracellular compartment, and the pro- block a receptor on the surface of the parietal cells, thereby
tein loses its affinity for Na! ions, which are then released stopping the cells from becoming activated by the hormone.
outside the cell. Once the three sodium ions have been re- Unlike P-type pumps, V-type pumps utilize the energy of
leased, the protein picks up two potassium ions (step 4), be- ATP without forming a phosphorylated protein intermediate.
comes dephosphorylated (step 5), and shifts back to the V-type pumps actively transport hydrogen ions across the walls
original E1 conformation (step 6). In this state, the binding of cytoplasmic organelles and vacuoles (hence the designation
site opens to the internal surface of the membrane and loses V-type). They occur in the membranes that line lysosomes,
its affinity for K! ions, leading to the release of these ions into secretory granules, and plant cell vacuoles where they maintain
the cell. The cycle is then repeated. A model of the Na!/K!- the low pH of the contents. V-type pumps have also been
ATPase structure based on a recent X-ray crystallographic found in the plasma membranes of a variety of cells. For exam-
study is shown in Figure 4.46b. ple, a V-type pump in the plasma membranes of kidney tubules
The importance of the sodiumpotassium pump becomes helps maintain the bodys acidbase balance by secreting
evident when one considers that it consumes approximately protons into the forming urine. V-type pumps are large multi-
one-third of the energy produced by most animal cells and subunit complexes similar to that of the ATP synthase shown
two-thirds of the energy produced by nerve cells. Digitalis, a in Figure 5.23.
steroid obtained from the foxglove plant that has been used Another diverse group of proteins that actively transport
for 200 years as a treatment for congestive heart disease, binds ions is the ATP-binding cassette (ABC) transporters, so called
to the Na!/K!-ATPase. Digitalis strengthens the hearts because all of the members of this superfamily share a homol-
contraction by inhibiting the Na!/K! pump, which leads to a ogous ATP-binding domain. The best studied ABC trans-
chain of events that increases Ca2! availability inside the porter is described in the accompanying Human Perspective.
muscle cells of the heart.
Using Light Energy to Actively Transport Ions Halobac-
Other Ion Transport Systems The best studied P-type terium salinarium (or H. halobium) is an archaebacterium that
pump is the Ca2!-ATPase whose three-dimensional struc- lives in extremely salty environments, such as that found in the
ture has been determined at several stages of the pumping Great Salt Lake. When grown under anaerobic conditions,
cycle. The calcium pump is present in the membranes of the the plasma membranes of these prokaryotes take on a purple
endoplasmic reticulum, where it actively transports calcium color due to the presence of one particular protein,
ions out of the cytosol into the lumen of this organelle. The bacteriorhodopsin. As shown in Figure 4.48, bacteriorhodopsin
transport of calcium ions by the Ca2!-ATPase is accompa- contains retinal, the same prosthetic group present in
nied by large conformational changes, which couple the hy- rhodopsin, the light-absorbing protein of the rods of the ver-
Inactive H+/ K+- ATPase Active H+/ K+- ATPase
Pump-- inhibiting drug
ATP
+
H+ + H
K K+
Food ADP+Pi
Cytosol
Acid-- neutralizing
agents
Histamine --
(OH )
Histamine Acid -- blocking drug
FIGURE 4.47 Control of acid secretion in the stomach. In the resting membrane, forming deep folds, or canaliculi. Once at the surface, the
state, the H!/K!-ATPase molecules are present in the walls of cytoplas- transport protein is activated and pumps protons into the stomach cav-
mic vesicles. Food entering the stomach triggers a cascade of hormone- ity against a concentration gradient (indicated by the size of the letters).
stimulated reactions in the stomach wall leading to the release of The heartburn drug Prilosec blocks acid secretion by directly inhibiting
histamine, which binds to a receptor on the surface of the acid-secreting the H!/K!-ATPase, whereas several other acid-blocking medications
parietal cells. Binding of histamine to its receptor stimulates a response interfere with activation of the parietal cells. Acid-neutralizing medica-
that causes the H!/K!-ATPase-containing vesicles to fuse to the plasma tions provide basic anions that combine with the secreted protons.
tebrate retina. The absorption of light energy by the retinal membrane. This gradient is subsequently used by an ATP-
group induces a series of conformational changes in the pro- synthesizing enzyme to phosphorylate ADP, as described in
tein that cause a proton to move from the retinal group, the next chapter.
through a channel in the protein, to the cell exterior (Fig-
ure 4.48). The proton donated by the photo-excited retinal is
replaced by another proton transferred to the protein from the
cytoplasm. In effect, this process results in the translocation of
protons from the cytoplasm to the external environment,
thereby generating a steep H! gradient across the plasma
FIGURE 4.48 Bacteriorhodopsin: a light-driven proton pump. The

protein contains seven membrane-spanning helices and a centrally
located retinal group (shown in purple), which serves as the light-
absorbing element (chromophore). Absorption of a photon of light
causes a change in the electronic structure of retinal, leading to the trans-
fer of a proton from the ONH! group to a closely associated, negatively
charged aspartic acid residue (#85) (step 1). The proton is then released
to the extracellular side of the membrane (step 2) by a relay system con-
sisting of several amino acid residues (Asp82, Glu204, and Glu194). The
spaces between these residues are filled with hydrogen-bonded water
molecules that help shuttle protons along the pathway. The deproto-
nated retinal is returned to its original state (step 3) when it accepts a
proton from an undissociated aspartic acid residue (Asp96) located near
the cytoplasmic side of the membrane. Asp96 is then reprotonated by a
H! from the cytoplasm (step 4). Asp85 is deprotonated (step 5) prior to
receiving a proton from retinal in the next pumping cycle. As a result of
these events, protons move from the cytoplasm to the cell exterior
through a central channel in the protein. (FROM HARTMUT LUECKE ET
AL., COURTESY OF JANOS K. LANYI, SCIENCE 286:255, 1999; COPYRIGHT
1999, AMERICAN ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE.)
THE HUMAN PERSPECTIVE

Defects in Ion Channels and Transporters
as a Cause of Inherited Disease
Several severe, inherited disorders have been traced to mutations in question was thought to be answered after the protein was purified,
genes that encode ion channel proteins (Table 1). Most of the disor- incorporated into artificial lipid bilayers, and shown to act as a cyclic
ders listed in Table 1 affect the movement of ions across the plasma AMP-regulated chloride channel, not a transporter. But subsequent
membranes of excitable cells (i.e., muscle, nerve, and sensory cells), studies have added numerous complications to the story as it has
reducing the ability of these cells to develop or transmit impulses been shown that, in addition to functioning as a chloride channel,
(page 162). In contrast, cystic fibrosis, the best studied and most CFTR also (1) conducts bicarbonate (HCO" 3 ) ions, (2) suppresses
common inherited ion channel disorder, results from a defect in the the activity of an epithelial Na# ion channel (ENaC), and (3) stim-
ion channels of epithelial cells. ulates the activity of a family of epithelial chloride/bicarbonate
On average, 1 out of every 25 persons of Northern European exchangers. As the role of CFTR has become more complex, it has
descent carries one copy of the mutant gene that can cause cystic fi- become difficult to establish precisely how a defect in this protein
brosis. Because they show no symptoms of the mutant gene, most leads to the development of chronic lung infections. While there is
heterozygotes are unaware that they are carriers. Consequently, ap- considerable debate, many researchers would agree with the follow-
proximately 1 out of every 2500 infants in this Caucasian population ing statements.
(1/25 ! 1/25 ! 1/4) is homozygous recessive at this locus and born Because the movement of water out of epithelial cells by osmo-
with cystic fibrosis (CF). Although cystic fibrosis affects various or- sis follows the movement of salts, abnormalities in the flux of Cl",
gans, including the intestine, pancreas, sweat glands, and reproduc- HCO" #
3 , and/or Na caused by CFTR deficiency leads to a decrease
tive tract, the respiratory tract usually exhibits the most severe in the fluid that bathes the epithelial cells of the airways (Figure 1).
effects. Victims of CF produce a thickened, sticky mucus that is very A reduction in volume of the surface liquid, and a resulting increase
hard to propel out of the airways. Afflicted individuals typically suf- in viscosity of the secreted mucus, impair the function of the cilia
fer from chronic lung infections and inflammation, which progres- that push mucus and bacteria out of the respiratory tract. At the
sively destroy pulmonary function. present time, the most promising candidates for improving the qual-
The gene responsible for cystic fibrosis was isolated in 1989. ity of life of CF patients are drugs, such as Bronchitol, that increase
Once the sequence of the CF gene was determined and the amino the volume of the airway surface liquid and improve ciliary function.
acid sequence of the corresponding polypeptide was deduced, it was Bronchitol is a fine, dry powder consisting of the sugar mannitol,
apparent that the polypeptide was a member of the ABC transporter which is delivered by a simple inhaler. When inhaled into the lungs,
superfamily. The protein was named cystic fibrosis transmembrane the dissolved mannitol increases the osmolarity of the airway surface
conductance regulator (CFTR), an ambiguous term that reflected liquid, which causes water to move osmotically out of the epithelial
the fact that researchers werent sure of its precise function. The cells and into the extracellular fluid. Clinical trials are also being car-
TABLE 1
Inherited disorder Type of channel Gene Clinical consequences
2#
Familial hemiplegic migraine (FHM) Ca CACNL1A4 Migraine headaches
Episodic ataxia type-2 (EA-2) Ca2# CACNL1A4 Ataxia (lack of balance and coordination)
Hypokalemic periodic paralysis Ca2# CACNL1A3 Periodic myotonia (muscle stiffness) and paralysis
Episodic ataxia type-1 K# KCNA1 Ataxia
Benign familial neonatal convulsions K# KCNQ2 Epileptic convulsions
Nonsyndromic dominant deafness K# KCNQ4 Deafness
Long QT syndrome K# HERG Dizziness, sudden death from ventricular
KCNQ1, or fibrillation
Na# SCN5A
Hyperkalemic periodic paralysis Na# SCN4A Periodic myotonia and paralysis
Liddle Syndrome Na# B-ENaC Hypertension (high blood pressure)
Myasthenia gravis Na# nAChR Muscle weakness
Dents disease Cl" CLCN5 Kidney stones
Myotonia congenita Cl" CLC-1 Periodic myotonia
Bartters syndrome type IV Cl" CLC-Kb Kidney dysfunction, deafness
Cystic fibrosis Cl" CFTR Lung congestion and infections
Cardiac arrhythmias Na# many different Irregular or rapid heartbeat
K# genes
Ca2#
See Nature Cell Biol. 6:1040, 2004, or Nature 440:444, 2006, for a more complete list.
DEFECTS IN ION CHANNELS AND TRANSPORTERS AS A CAUSE OF INHERITED DISEASE 157
Hydrated Suspended
mucous layer CFTR bacteria
Dehydrated Layer of pathogenic

Flow
mucous layer bacteria (biofilm)
Deficient.
flow
Cilia
Epithelial
Na+ Cl H2O Na+ H2O Cl
cell
Airway epithelium from normal individual Airway epithelium from CF patient
FIGURE 1 An explanation for the debilitating effects on lung function the airways. In the airway epithelium of a person with cystic fibrosis, the
from the absence of the CFTR protein. In the airway epithelium of a abnormal movement of ions causes water to flow in the opposite direction,
normal individual, water flows out of the epithelial cells in response to the thus dehydrating the mucous layer. As a result, trapped bacteria cannot be
outward movement of ions, thus hydrating the surface mucous layer. The moved out of the airways, which allows them to proliferate as a biofilm
hydrated mucous layer, with its trapped bacteria, is readily moved out of (page 12) and cause chronic infections.
ried out on compounds that have the potential to increase the vol- the 1820s. An alternate proposal suggests that heterozygotes are
ume of surface fluid by altering the movements of ions in and out of protected from typhoid fever because the bacterium responsible for
the epithelial cells (Figure 1). These compounds include Na!- this disease adheres poorly to the wall of an intestine having a re-
channel inhibitors (e.g., Denufosol) with the potential to reduce duced number of CFTR molecules.
Na! ion absorption from the airway fluid into the epithelium and Ever since the isolation of the gene responsible for CF, the de-
activators of Cl" channels (other than CFTR) (e.g., Moli1901) velopment of a cure by gene therapythat is, by replacement of the
with the potential to increase Cl" ion conductance from the epithe- defective gene with a normal versionhas been a major goal of CF
lium into the airway fluid. researchers. Cystic fibrosis is a good candidate for gene therapy be-
In the past decade, researchers have identified more than 1000 cause the worst symptoms of the disease result from the defective ac-
different mutations that give rise to cystic fibrosis. However, approx- tivities of epithelial cells that line the airways and, therefore, are
imately 70 percent of the alleles responsible for cystic fibrosis in accessible to agents that can be delivered by inhalation of an aerosol.
the United States contain the same genetic alteration (designated Clinical trials have been conducted using several different types of
#F508)they are all missing three base pairs of DNA that encode a delivery systems. In one group of trials, the normal CFTR gene was
phenylalanine at position 508, within one of the cytoplasmic do- incorporated into the DNA of a defective adenovirus, a type of virus
mains of the CFTR polypeptide. Subsequent research has revealed that normally causes upper respiratory tract infections. The recombi-
that CFTR polypeptides lacking this particular amino acid fail to be nant virus particles were then allowed to infect the cells of the air-
processed normally within the membranes of the endoplasmic retic- way, delivering the normal gene to the genetically deficient cells. The
ulum and, in fact, never reach the surface of epithelial cells. As a primary disadvantage in using adenovirus is that the viral DNA
result, CF patients who are homozygous for the #F508 allele (along with the normal CFTR gene) does not become integrated
completely lack the CFTR channel in their plasma membranes into the chromosomes of the infected host cell so that the virus must
and have a severe form of the disease. When cells from these patients be readministered frequently. As a result, the procedure often in-
are grown in culture at lower temperature, the mutant protein is duces an immune response within the patient that eliminates the
transported to the plasma membrane where it functions quite well. virus and leads to lung inflammation. Researchers are hesitant to
This finding has prompted a number of drug companies to screen for employ viruses that integrate their genomes for fear of initiating
small molecules that can bind to these mutant CFTR molecules, the formation of cancers. In other trials, the DNA encoding the nor-
thereby preventing their destruction in the cytoplasm and allowing mal CFTR gene has been linked to positively charged liposomes
them to reach the cell surface. Several promising candidates have (page 124) that can fuse with the plasma membranes of the airway
been identified, but none has yet to be proven effective in clinical trials. cells, delivering their DNA contents into the cytoplasm. Lipid-based
According to one estimate, the #F508 mutation had to have delivery has an advantage over viruses in being less likely to stimulate
originated more than 50,000 years ago to have reached such a high a destructive immune response following repeated treatments, but
frequency in the population. The fact that the CF gene has reached has the disadvantage of being less effective in achieving genetic
this frequency suggests that heterozygotes may receive some selec- modification of target cells. To date, none of the clinical trials of gene
tive advantage over those lacking a copy of the defective gene. It has therapy has resulted in significant improvement of either physiologic
been proposed that CF heterozygotes may be protected from the processes or disease symptoms. The development of more effective
effects of cholera, a disease that is characterized by excessive fluid DNA delivery systems, which are capable of genetically altering a
secretion by the wall of the intestine. One difficulty with this pro- greater percentage of airway cells, will be required if a treatment for
posal is that there is no record of cholera epidemics in Europe until CF based on gene therapy is to be achieved.
Cotransport: Coupling Active Transport to Existing Ion transport protein, called a Na!/glucose cotransporter, moves
Gradients The establishment of concentration gradients, two sodium ions and one glucose molecule with each cycle.
such as those of Na!, K!, and H!, provides a means by which Once inside, the glucose molecules diffuse through the cell
free energy can be stored in a cell. The potential energy stored and are moved across the basal membrane by facilitated diffu-
in ionic gradients is utilized by a cell in various ways to per- sion (page 152).
form work, including the transport of other solutes. Consider To appreciate the power of an ion gradient in accumulat-
the physiologic activity of the intestine. Within its lumen, en- ing other types of solutes in cells, we can briefly consider
zymes hydrolyze high-molecular-weight polysaccharides into the energetics of the Na!/glucose cotransporter. Recall from
simple sugars, which are absorbed by the epithelial cells that page 144 that the free-energy change for the movement of a
line the intestine. The movement of glucose across the apical mole of Na! ions into the cell is equal to "3.1 kcal/mol, and
plasma membrane of the epithelial cells, against a concentra- thus 6.2 kcal for two moles of Na! ions, which would be avail-
tion gradient, occurs by cotransport with sodium ions, as il- able to transport one mole of glucose uphill into the cell. Re-
lustrated in Figure 4.49. The Na! concentration is kept very call also from page 143 that the equation for the movement of
low within the cells by the action of a primary active transport a nonelectrolyte, such as glucose, across the membrane is
system (the Na!/K!-ATPase), located in the basal and lateral
[Ci]
plasma membrane, which pumps sodium ions out of the cell G $ RT ln
against a concentration gradient. The tendency for sodium [Co]
ions to diffuse back across the apical plasma membrane down [Ci]
their concentration gradient is tapped by the epithelial cells G $ 2.303 RT log10
[Co]
to drive the cotransport of glucose molecules into the cell
against a concentration gradient. The glucose molecules are Using this equation, we can calculate how steep a concentration
said to be driven by secondary active transport. In this case, the gradient of glucose (X ) that this cotransporter can generate.
At 25#C,
"6.2 kcal/mol $ 1.4 kcal/mol # log10 X
log10 X $ "4.43
1
X$
Lumen
23,000
This calculation indicates that the Na!/glucose cotransporter
is capable of transporting glucose into a cell against a concen-
tration gradient greater than 20,000-fold.
2Na+ GL Plant cells rely on secondary active transport systems to
take up a variety of nutrients, including sucrose, amino acids,
and nitrate. In plants, uptake of these compounds is coupled
to the downhill, inward movement of H! ions rather than
+
Na /glucose Na! ions. The secondary active transport of glucose into the
cotransporter
+
epithelial cells of the intestine and the transport of sucrose
Na GL into a plant cell are examples of symport, in which the two
Glucose
transporter transported species (Na! and glucose or H! and sucrose)
Na+ Na
+ GL (GLUT2).
move in the same direction. Numerous secondary transport
Facilitated
K
+
K+ diffusion of proteins have been isolated that engage in antiport, in which
glucose.
+ + Blood the two transported species move in opposite directions. For
Na / K - ATPase
example, cells often maintain a proper cytoplasmic pH by
GL
coupling the inward, downhill movement of Na! with the
outward movement of H!. Proteins that mediate antiport are
FIGURE 4.49 Secondary transport: the use of energy stored in usually called exchangers.
A
an ionic gradient. The Na!/K!-ATPase residing in the plasma
membrane of the lateral surface maintains a very low cytosolic concen-
tration of Na!. The Na! gradient across the plasma membrane repre-
sents a storage of energy that can be tapped to accomplish work, such as
REVIEW
?
the transport of glucose by a Na!/glucose cotransporter located in the 1. Compare and contrast the four basically different ways
apical plasma membrane. Once transported across the apical surface into that a substance can move across the plasma membrane
the cell, the glucose molecules diffuse to the basal surface where they are (as indicated in Figure 4.33).
carried by a glucose facilitative transporter out of the cell and into the
2. Contrast the energetic difference between the diffusion of
bloodstream. The relative size of the letters indicates the directions of
the respective concentration gradients. Two Na! ions are transported for an electrolyte versus a nonelectrolyte across the membrane.
each glucose molecule; the 2!1 Na!/glucose provides a much greater 3. Describe the relationship between partition coefficient
driving force for moving glucose into the cell than a 1!1 ratio. and molecular size with regard to membrane permeability.
4.8 MEMBRANE POTENTIALS AND NERVE IMPULSES 159
Nerve cells (or neurons) are specialized for the collection,

4. Explain the effects of putting a cell into a hypotonic,
conduction, and transmission of information, which is coded
hypertonic, or isotonic medium.
in the form of fast-moving electrical impulses. The basic parts
5. Describe two ways in which energy is utilized to move of a typical neuron are illustrated in Figure 4.50). The nucleus
ions and solutes against a concentration gradient. of the neuron is located within an expanded region called the
! !
6. How does the Na /K -ATPase illustrate the asymmetry cell body, which is the metabolic center of the cell and the site
of the plasma membrane? where most of its material contents are manufactured. Ex-
7. What is the role of phosphorylation in the mechanism of tending from the cell bodies of most neurons are a number of
action of the Na!/K!-ATPase? fine extensions, called dendrites, which receive incoming in-
formation from external sources, typically other neurons. Also
8. What is the structural relation between the parts of the
emerging from the cell body is a single, more prominent ex-
prokaryotic KcsA K! channel and the eukaryotic voltage-
tension, the axon, which conducts outgoing impulses away
regulated K! channel? Which part of the channel is in-
from the cell body and toward the target cell(s). Although
volved in ion selectivity, which part in channel gating, and
some axons may be only a few micrometers in length, others
which part in channel inactivation? How does each of these
extend for many meters in the body of a large vertebrate, such
processes (ion selectivity, gating, and inactivation) occur?
!
as a giraffe or whale. Most axons split near their ends into
9. Because of its smaller size, one would expect Na ions to smaller processes, each ending in a terminal knoba special-
be able to penetrate any pore large enough for a K! ion. ized site where impulses are transmitted from neuron to target
How does the K! channel select for this specific ion? cell. Many neurons in the brain end in thousands of terminal
knobs, allowing these brain cells to communicate with thou-
sands of potential targets. As discussed on page 162, most
neurons in the vertebrate body are wrapped in a lipid-rich
4.8 MEMBRANE POTENTIALS myelin sheath, whose function is described below.
AND NERVE IMPULSES
All organisms respond to external stimulation, a property The Resting Potential
referred to as irritability. Even a single-celled amoeba, if A voltage (or electric potential difference) between two
poked with a fine glass needle, responds by withdrawing its points, such as the inside and outside of the plasma mem-
pseudopodia, rounding up, and moving off in another direc- brane, results when there is an excess of positive ions at one
tion. Irritability in an amoeba depends on the same basic point and an excess of negative ions at the other point. Volt-
properties of membranes that lead to the formation and prop- ages across plasma membranes can be measured by inserting
agation of nerve impulses, which is the subject of the remain-
der of the chapter.
Nucleus of
Schwann cell
Dendrites Layers of
myelin
Nucleus
Axon
Cell body
Axon
Myelin sheath
Node of Ranvier
Terminal knob
(a) (b)
FIGURE 4.50 The structure of a nerve cell. (a) Schematic drawing of a cells.) (b) A composite micrograph of a single rat hippocampal neuron
simple neuron with a myelinated axon. As the inset shows, the myelin with cell body and dendrites (purple) and an axon 1 cm in length (red).
sheath comprises individual Schwann cells that have wrapped them- Motor nerve cells in larger mammals can be 100 times this length.
selves around the axon. The sites where the axon lacks myelin wrapping (B: FROM CARLOS F. IBEZ, TRENDS CELL BIOL. 17:520, 2007, COPYRIGHT
are called nodes of Ranvier. (Note: Myelin-forming cells within the cen- 2007, WITH PERMISSION FROM ELSEVIER SCIENCE.)
tral nervous system are called oligodendrocytes rather than Schwann
one fine glass electrode (or microelectrode) into the cytoplasm membrane of a resting nerve cell are selective for K!; they are of-
of a cell, placing another electrode in the extracellular fluid ten referred to as K! leak channels. K! leak channels are thought
outside the cell, and connecting the electrodes to a voltmeter, to be members of a family of K! channels that lack the S4 volt-
an instrument that measures a difference in charge between age sensor (page 151) and fail to respond to changes in voltage.
two points (Figure 4.51). When this experiment was first car- Because K! ions are the only charged species with signif-
ried out on a giant axon of the squid, a potential difference of icant permeability in a resting nerve cell, their outflow through
approximately 70 millivolts (mV) was recorded, the inside be- the membrane leaves an excess of negative charges on the cy-
ing negative with respect to the outside (indicated with a mi- toplasmic side of the membrane. Although the concentration
nus sign, %70 mV). The presence of a membrane potential is gradient across the membrane favors continued efflux of K!,
not unique to nerve cells; such potentials are present in all the electrical gradient resulting from the excess negative
types of cells, the magnitude varying between about %15 and charge on the inside of the membrane favors the retention of
%100 mV. For nonexcitable cells, that is, those other than K! ions inside the cell. When these two opposing forces are
neurons and muscle cells, this voltage is simply termed the balanced, the system is at equilibrium, and there is no further
membrane potential. In a nerve or muscle cell, this same po- net movement of K! ions across the membrane. Using the
tential is referred to as the resting potential because it is sub- following equation, which is called the Nernst equation, one
ject to dramatic change, as discussed in the following section. can calculate the membrane potential (Vm) that would be
The magnitude and direction of the voltage across the measured at equilibrium if the plasma membrane of a nerve
plasma membrane are determined by the differences in concen- cell were permeable only to K! ions.6 In this case, Vm would
trations of ions on either side of the membrane and their relative be equal to the potassium equilibrium potential (EK):
permeabilities. As described earlier in the chapter, the Na!/K!- !
ATPase pumps Na! out of the cell and K! into the cell, thereby EK $ 2.303
RT
# log10 [Ko! ]
establishing steep gradients of these two ions across the plasma zF [Ki ]
membrane. Because of these gradients, you might expect that For a squid giant axon, the internal [Ki!] is approximately
potassium ions would leak out of the cell and sodium ions would 350 mM, while the external [Ko!] is approximately 10 mM;
leak inward through their respective ion channels. However, the thus at 25"C (298 K) and z $ !1 (for the univalent K! ion),
vast majority of the ion channels that are open in the plasma
EK $ 59 log100.028 $ %91 mV
A similar calculation of the Na! equilibrium potential (ENa)
+80 +80
+60 +60
Sodium equilibrium potential would produce a value of approximately !55 mV. Because
+40 +40 measurements of the voltage across the resting nerve mem-
Voltage (mV)
+20 +20
brane are similar in sign and magnitude (%70 mV) to the
0 0 Electrode
20 20 enters cell potassium equilibrium potential just calculated, the move-
40 40 ment of potassium ions across the membrane is considered the
60 60
80 80
most important factor in determining the resting potential.
100 100 The difference between the calculated K! equilibrium poten-
Potassium equilibrium potential
tial (%91 mV) and the measured resting potential (%70 mV,
Time Time
Figure 4.51) is due to a slight permeability of the membrane
to Na! through a recently described Na! leak channel.
Voltmeter
The Action Potential

Recording Reference
electrode electrode Our present understanding of membrane potentials and nerve
impulses rests on a body of research carried out on the giant ax-
++ + + + + + + + + + + + + + + + ++ + + + + + + + + + + + + + + + ons of the squid in the late 1940s and early 1950s by a group of
British physiologists, most notably Alan Hodgkin, Andrew
Axon Axon
Huxley, and Bernard Katz. These axons, which are approxi-
mately 1 mm in diameter, carry impulses at high speeds,
(a) (b) enabling the squid to escape rapidly from predators. If the
membrane of a resting squid axon is stimulated by poking it
FIGURE 4.51 Measuring a membranes resting potential. A po- with a fine needle or jolting it with a very small electric current,
A
tential is measured when a difference in charge is detected some of its sodium channels open, allowing a limited number of
between the reference and recording electrodes. In (a), both electrodes
sodium ions to diffuse into the cell. This opportunity for posi-
are on the outside of the cell, and no potential difference (voltage) is
tively charged ions to move into the cell reduces the membrane
measured. As one electrode penetrates the plasma membrane of the
axon in (b), the potential immediately drops to %70 mV (inside nega-
tive), which approaches the equilibrium potential for potassium ions, 6
The Nernst equation is derived from the equation provided on page 144, by
that is, the potential that would result if the membrane were imperme- setting #G at zero, which is the case when the movement of the ions is at
able to all ions except potassium. equilibrium. Walther Nernst was a German physical chemist who won the 1920
Nobel Prize.
potential, making it less negative. Because the positive change 1 msec in a myelinated mammalian nerve cell. Following an ac-
in membrane voltage causes a decrease in the polarity between tion potential, the membrane enters a brief refractory period dur-
the two sides of the membrane, it is called a depolarization. ing which it cannot be restimulated. The refractory period occurs
If the stimulus causes the membrane to depolarize by only because the sodium channels that were inactivated during the ini-
a few millivolts, say from "70 to "60 mV, the membrane tial stage of the action potential must close before they can be re-
rapidly returns to its resting potential as soon as the stimulus opened in response to another stimulus. As depicted in Figure
has ceased (Figure 4.52a, left box). If, however, the stimulus 4.43, the transformation of the ion channel from the inactivated
depolarizes the membrane beyond a certain point, called the to the closed conformation can only occur after the inactivating
threshold, which occurs at about "50 mV, then a new series peptide has swung out of the opening of the pore.
of events is launched. The change in voltage causes the Although the action potential changes membrane voltage
voltage-gated sodium channels to open. As a result, sodium dramatically, only a minute percentage of the ions on the two
ions diffuse freely into the cell (Figure 4.52a, middle box) sides of the membrane are involved in any given action
down both their concentration and electric gradients. The potential. The striking changes in membrane potential seen in
increased permeability of the membrane to Na! ions and Figure 4.52b are not caused by changes in Na! and K! ion con-
the corresponding movement of positive charge into the centrations on the two sides of the membrane (such changes are
cell causes the membrane to reverse potential briefly (Fig-
ure 4.52b), becoming positive at about !40 mV, which
approaches the equilibrium potential for Na! (Figure 4.51). Resting potential Depolarization phase Repolarization phase
After approximately 1 msec, the sodium channels sponta- sodium gates closed
voltage = 70 mV
sodium gates open
voltage = +40 mV
potassium gates open
voltage = 80 mV
neously inactivate, blocking further influx of Na! ions. Accord- Outside Outside Outside
+ + + Na+ + + + + + + Na+ + + + + ++ + + +
ing to the prevailing view, inactivation results from the random + + + + + + + K+
diffusion of an inactivation peptide into the opening of the chan- Sodium Sodium Sodium
gate gate gate Potassium
nel pore in a manner similar to that described for K! channels on closed
Potassium
gate open
Potassium
gate closed gate
closed closed open
page 151. Meanwhile, the change in membrane potential caused
by Na! influx triggers the opening of the voltage-gated potas- + + + + + +
K+
+
K
+ Na
+

K+
sium channels (Figure 4.52a, right box) discussed on page 151.
As a result, potassium ions diffuse freely out of the cell down their Potassium
++ ++++ ++++++++
+++
steep concentration gradient. The decreased permeability of the leak channel
membrane to Na! and the increased permeability to K! cause Time Time Time
1 2 3
the membrane potential to swing back to a negative value of
+++
about "80 mV, approaching that of the K! equilibrium poten- ++ ++++ ++++++++
tial (Figure 4.51). The large negative membrane potential causes
the voltage-gated potassium channels to close (see Figure 4.43b),
which returns the membrane to its resting state. Collectively, Na+ permeability
these changes in membrane potential are called an action poten-
tial (Figure 4.52b). The entire series of changes during an action
Ion permeability
potential takes only about 5 msec in the squid axon and less than
K+ permeability
A FIGURE 4.52 Formation of an action potential. (a) Time 1, up-
per left box: The membrane in this region of the nerve cell ex-
hibits the resting potential, in which only the K! leak channels are open
and the membrane voltage is approximately "70 mV. Time 2, upper
middle box, shows the depolarization phase: The membrane has depo- 0 1 2
larized beyond the threshold value, opening the voltage-regulated (a) Time (ms)
sodium gates, leading to an influx of Na! ions (indicated in the perme-
ability change in the lower graph). The increased Na! permeability
causes the membrane voltage to temporarily reverse itself, reaching a 40
value of approximately !40 mV in the squid giant axon (time 2). It is
Membrane potential (mV)
this reversal of membrane potential that constitutes the action potential.

Time 3, upper right box, shows the repolarization phase: Within a tiny 0
fraction of a second, the sodium gates are inactivated and the potassium
gates open, allowing potassium ions to diffuse across the membrane
(lower part of the drawing) and establish an even more negative poten-
tial at that location ("80 mV) than that of the resting potential. Almost
as soon as they open, the potassium gates close, leaving the potassium
leak channels as the primary path of ion movement across the membrane 70
and reestablishing the resting potential. (b) A summary of the voltage
changes that occur during an action potential, as described in part a. 0 1 2
(b) Time (ms)
insignificant). Rather, they are caused by the movements of

charge in one direction or the other that result from the fleeting ++ + + + + + + + + + + + + + + + + + + + + + + ++ + + + + +
+ + + + + + + +
changes in permeability to these ions. Those Na! and K! ions
that do change places across the membrane during an action
Axon Direction of propagation
potential are eventually pumped back by the Na!/K!-ATPase.
Even if the Na!/K!-ATPase is inhibited, a neuron can often + + + + + + + +
++ + + + + + + + + + + + + + + + + + + + + + + ++ + + + + +
continue to fire thousands of impulses before the ionic gradi-
ents established by the pumps activity are dissipated.
Once the membrane of a neuron is depolarized to the
Region in Region of Region where depolarization
threshold value, a full-blown action potential is triggered with- refractory action will trigger
out further stimulation. This feature of nerve cell function is period potential action potential
known as the all-or-none law. There is no in-between; sub-
threshold depolarization is incapable of triggering an action po- A FIGURE 4.53 Propagation of an impulse results from the local
tential, whereas threshold depolarization automatically elicits a flow of ions. An action potential at one site on the membrane
maximum response. It is also noteworthy that an action poten- depolarizes an adjacent region of the membrane, triggering an action
potential at the second site. The action potential can only flow in the
tial is not an energy-requiring process, but one that results from
forward direction because the portion of the membrane that has just ex-
the flow of ions down their respective electrochemical gradi- perienced an action potential remains in a refractory period.
ents. Energy is required by the N!/K!-ATPase to generate the
steep ionic gradients across the plasma membrane, but once
that is accomplished, the various ions are poised to flow through
the membrane as soon as their ion channels are opened. make sensory discriminations depends on several factors. For ex-
The movements of ions across the plasma membrane of ample, a stronger stimulus, such as scalding water, activates more
nerve cells form the basis for neural communication. Certain nerve cells than does a weaker stimulus, such as warm water. It
local anesthetics, such as procaine and novocaine, act by clos- also activates high-threshold neurons that would remain at rest
ing ion channels in the membranes of sensory cells and neu- if the stimulus were weaker. Stimulus strength is also encoded in
rons. As long as these ion channels remain closed, the affected the pattern and frequency by which action potentials are
cells are unable to generate action potentials and thus unable launched down a particular neuron. In most cases, the stronger
to inform the brain of events occurring at the skin or teeth. the stimulus, the greater the number of impulses generated.
Speed Is of the Essence The greater the diameter of an

Propagation of Action Potentials as an Impulse
axon, the less the resistance to local current flow and the more
Up to this point, we have restricted the discussion to events rapidly an action potential at one site can activate adjacent
occurring at a particular site on the nerve cell membrane regions of the membrane. Some invertebrates, such as squid
where experimental depolarization has triggered an action po- and tube worms, have evolved giant axons that facilitate the
tential. Once an action potential has been initiated, it does not animals escape from danger. There is, however, a limit to this
remain localized at a particular site but is propagated as a nerve evolutionary approach. Because the speed of conduction in-
impulse down the length of the cell to the nerve terminals. creases with the square root of the increase in diameter, an
Nerve impulses are propagated along a membrane because axon that is 480 "m in diameter can conduct an action poten-
an action potential at one site has an effect on the adjacent site. tial only four times faster than one that is 30 "m in diameter.
The large depolarization that accompanies an action potential During the evolution of vertebrates, an increase in con-
creates a difference in charge along the inner and outer surfaces duction velocity was achieved when the axon became wrapped
of the plasma membrane (Figure 4.53). As a result, positive ions in a myelin sheath (see Figures 4.5 and 4.50). Because it is
move toward the site of depolarization on the outer surface of composed of many layers of lipid-containing membranes, the
the membrane and away from that site on the inner surface myelin sheath is ideally suited to prevent the passage of ions
(Figure 4.53). This local flow of current causes the membrane across the plasma membrane. In addition, nearly all of the
in the region just ahead of the action potential to become depo- Na! ion channels of a myelinated neuron reside in the un-
larized. Because the depolarization accompanying the action wrapped gaps, or nodes of Ranvier, between adjacent Schwann
potential is very large, the membrane in the adjacent region is cells or oligodendrocytes that make up the sheath (see Fig-
readily depolarized to a level greater than the threshold value, ure 4.50). Consequently, the nodes of Ranvier are the only
which opens the sodium channels in this adjacent region, gen- sites where action potentials can be generated. An action
erating another action potential. Thus, once triggered, a succes- potential at one node triggers an action potential at the next
sion of action potentials passes down the entire length of the node (Figure 4.54), causing the impulse to jump from node to
neuron without any loss of intensity, arriving at its target cell node without having to activate the intervening membrane.
with the same strength it had at its point of origin. Propagation of an impulse by this mechanism is called salta-
Because all impulses traveling along a neuron exhibit the tory conduction. Impulses are conducted along a myelinated
same strength, stronger stimuli cannot produce bigger im- axon at speeds up to 120 meters per second, which is more
pulses than weaker stimuli. Yet, we are clearly capable of than 20 times faster than the speed that impulses travel in an
detecting differences in the strength of a stimulus. The ability to unmyelinated neuron of the same diameter.
Next node The importance of myelination is dramatically illustrated

Na+
by multiple sclerosis (MS), a disease associated with deteriora-
+ tion of the myelin sheath that surrounds axons in various parts
+ of the nervous system. Manifestations of the disease usually be-
Current flow depolarizes Direction
next node of Ranvier of impulse gin in young adulthood; patients experience weakness in their
Axon
hands, difficulty in walking, and problems with their vision.
+
Neurotransmission: Jumping the Synaptic Cleft
+ Neurons are linked with their target cells at specialized
Na+ junctions called synapses. Careful examination of a synapse
reveals that the two cells do not make direct contact but are
Node of Myelin separated from each other by a narrow gap of about 20 to
Ranvier sheath 50 nm. This gap is called the synaptic cleft. A presynaptic
FIGURE 4.54 Saltatory conduction. During saltatory conduction, only cell (a receptor cell or a neuron) conducts impulses toward a
the membrane in the nodal region of the axon becomes depolarized and synapse, and a postsynaptic cell (a neuron, muscle, or gland
capable of forming an action potential. This is accomplished as current cell) always lies on the receiving side of a synapse. Figure 4.55
flows directly from an activated node to the next resting node along the shows a number of synapses between the terminal branches of
axon. an axon and a skeletal muscle cell; synapses of this type are
called neuromuscular junctions.
Terminal knob of presynaptic neuron FIGURE 4.55 The neuromuscular junction is the site
where branches from a motor axon form synapses with
Synaptic vesicles the muscle fibers of a skeletal muscle. The left inset shows
the synaptic vesicles residing within the terminal knob of
Membrane of postsynaptic target cell the axon and the narrow synaptic cleft between the
terminal knob and the postsynaptic target cell. The right
Synaptic cleft inset shows the terminal knob pressed closely against the
muscle cell plasma membrane. Neurotransmitter mole-
cules (red) released from synaptic vesicles of the presynap-
Axon of nerve cell tic neuron are binding to receptors (orange) on the surface
of the muscle cell (blue). (LEFT INSET FROM VU/T. REESE
AND DON W. FAWCETT/VISUALS UNLIMITED.)
Muscle fiber
How does an impulse in a presynaptic neuron jump across postsynaptic membrane excites the cell, making the cell
the synaptic cleft and affect the postsynaptic cell? Studies car- more likely to respond to this or subsequent stimuli by
ried out decades ago indicated that a chemical substance is in- generating an action potential of its own (Figure 4.56,
volved in the transmission of an impulse from one cell to steps 5a and 6).
another (page 166). The very tips (terminal knobs) of the 2. The bound transmitter can trigger the opening of anion-
branches of an axon appear in the electron microscope to con- selective channels, leading mainly to an influx of chloride
tain large numbers of synaptic vesicles (Figure 4.55, left in- ions, and a more negative (hyperpolarized) membrane po-
set) that serve as storage sites for the chemical transmitters tential. Hyperpolarization of the postsynaptic membrane
that act on postsynaptic cells. Two of the best studied neuro- makes it less likely the cell will generate an action poten-
transmitters are acetylcholine and norepinephrine, tial because greater sodium influx is subsequently required
O to reach the membranes threshold (Figure 4.56, step 5b).
+
CH3 C O CH2 CH2 N(CH3)3 Most nerve cells in the brain receive both excitatory and in-
hibitory signals from many different presynaptic neurons. It is
Acetylcholine (ACh)
the summation of these opposing influences that determine
OH whether or not an impulse will be generated in the postsynap-
H
tic neuron.
+
HO C CH2 NH3 All terminal knobs of a given neuron release the same
neurotransmitter(s). However, a given neurotransmitter may
OH have a stimulatory effect on one particular postsynaptic
Norepinephrine membrane and an inhibitory effect on another. Acetylcholine,
for example, inhibits contractility of the heart but stimulates
which transmit impulses to the bodys skeletal and cardiac
contractility of skeletal muscle. Within the brain, glutamate
muscles.
serves as the primary excitatory neurotransmitter and gamma-
The sequence of events during synaptic transmission can
aminobutyric acid (GABA) as the primary inhibitory neuro-
be summarized as follows (Figure 4.56). When an impulse
transmitter. A number of general anesthetics, as well as
reaches a terminal knob (step 1, Figure 4.56), the accompany-
Valium and its derivatives, act by binding to the GABA
ing depolarization induces the opening of a number of
receptor and enhancing the activity of the brains primary
voltage-gated Ca2! channels in the plasma membrane of this
off switch.
part of the presynaptic nerve cell (step 2, Figure 4.56). Cal-
cium ions are normally present at very low concentration
within the neuron (about 100 nM), as in all cells. When the Actions of Drugs on Synapses It is important that a neu-
gates open, calcium ions diffuse from the extracellular fluid rotransmitter has only a short half-life following its release
into the terminal knob of the neuron, causing the [Ca2!] to from a presynaptic neuron; otherwise the effect of the neuro-
rise more than a thousandfold within localized microdomains transmitter would be extended, and the postsynaptic neuron
near the channels. The elevated [Ca2!] triggers the rapid fu- would not recover. A neurotransmitter is eliminated from the
sion of one or a few nearby synaptic vesicles with the plasma synapse in two ways: by enzymes that destroy neurotransmit-
membrane, causing the release of neurotransmitter molecules ter molecules in the synaptic cleft and by proteins that trans-
into the synaptic cleft (step 3, Figure 4.56). port neurotransmitter molecules back to the presynaptic
Once released from the synaptic vesicles, the neurotrans- terminalsa process called reuptake. Because of the destruc-
mitter molecules diffuse across the narrow gap and bind selection or reuptake of neurotransmitter molecules, the effect of
tively to receptor molecules that are concentrated directly each impulse lasts no more than a few milliseconds.
across the cleft in the postsynaptic plasma membrane (step 4, Interfering with the destruction or reuptake of neuro-
Figure 4.56). A neurotransmitter molecule can have one of transmitters can have dramatic physiologic and behavioral
two opposite effects depending on the type of receptor on the effects. Acetylcholinesterase is an enzyme localized within
target cell membrane to which it binds:7 the synaptic cleft that hydrolyzes acetylcholine. If this enzyme
is inhibited by exposure to the nerve gas DFP, for example,
1. The bound transmitter can trigger the opening of cation- the skeletal muscles of the body undergo violent contraction
selective channels in the membrane, leading primarily to due to the continued presence of high concentrations of
an influx of sodium ions and a less negative (more posi- acetylcholine.
tive) membrane potential. This depolarization of the Many drugs act by inhibiting the transporters that sweep
neurotransmitters out of the synaptic cleft. A number of
7 widely prescribed antidepressants, including Prozac, inhibit
It is important to note that this discussion ignores an important class of neuro-
transmitter receptors that are not ion channels and thus do not directly affect the reuptake of serotonin, a neurotransmitter implicated in
membrane voltage. This other group of receptors are members of a class of mood disorders. Cocaine, on the other hand, interferes with
proteins called GPCRs, which are discussed at length in Section 15.3. When the reuptake of the neurotransmitter dopamine that is re-
a neurotransmitter binds to one of these receptors, it can initiate a variety of
responses, which often includes the opening of ion channels by an indirect leased by certain nerve cells in a portion of the brain known as
mechanism. the limbic system. The limbic system contains the brains
1 Nerve impulse
Synaptic vesicle
Acetylcholine
Presynaptic
membrane
2
3
4 5a 5b 6
Synaptic cleft Ca2+ 2+ 0 0
gates Ca Release
Binding of Nerve impulse
mV
mV
open of AcCh 70 or 70
AcCh by receptor
Na+ gates open Cl -- gates open

Postsynaptic
membrane
FIGURE 4.56 The sequence of events during synaptic trans- depolarization of the postsynaptic membrane (as in 5a), a nerve impulse
A mission with acetylcholine as the neurotransmitter. During may be generated there (6). If, however, the binding of neurotransmitter
steps 14, a nerve impulse reaches the terminal knob of the axon, cal- causes a hyperpolarization of the postsynaptic membrane (5b), the target
cium gates open leading to an influx of Ca2", and acetylcholine is re- cell is inhibited, making it more difficult for an impulse to be generated
leased from synaptic vesicles and binds to receptors on the postsynaptic in the target cell by other excitatory stimulation. The breakdown of the
membrane. If the binding of the neurotransmitter molecules causes a neurotransmitter by acetylcholinesterase is not shown.
pleasure or reward centers. The sustained presence of respectively. If marijuana increases appetite by binding to CB1
dopamine in the synaptic clefts of the limbic system produces receptors, it follows that blocking these receptors might
a short-lived feeling of euphoria, as well as a strong desire to decrease appetite. This line of reasoning led to the develop-
repeat the activity. With repeated use, the pleasurable effects ment of a CB1-blocking weight-loss drug called Acomplia,
of the drug are increasingly reduced, but its addictive proper- which has been pulled from the market because of side effects.
ties are enhanced. Amphetamines also act on dopamine-
releasing neurons; they are thought to stimulate the excessive Synaptic Plasticity Synapses are more than simply con-
release of dopamine from presynaptic terminals and interfere necting sites between adjacent neurons; they are key determi-
with the reuptake of the neurotransmitter molecules from the nants in the routing of impulses through the nervous system.
synaptic cleft. Mice that have been genetically engineered to The human brain is thought to contain at least one hundred
lack the dopamine transporter (DAT)the protein responsi- trillion synapses. These synapses act like gates stationed along
ble for dopamine reuptakeshow behavior similar to that of the various pathways, allowing some pieces of information to
normal mice that have been given cocaine or amphetamines. pass from one neuron to another, while holding back other
Administration of cocaine or amphetamines has no additional pieces or rerouting them in another direction. While synapses
behavioral effects on animals lacking the DAT gene. are often perceived as fixed, unchanging structures, they can
The active compound in marijuana (!9-tetrahydro- display a remarkable dynamic quality known as synaptic plas-
cannabinol) acts by a totally different mechanism. It binds ticity. Synaptic plasticity is particularly important during in-
to cannabinoid (CB1) receptors located on the presynaptic fancy and childhood, when the neuronal circuitry of the brain
terminals of certain neurons of the brain, which reduces the achieves it mature configuration.
likelihood that these neurons will release neurotransmitters. Synaptic plasticity is most readily observed in studies on
CB1 receptors normally interact with compounds produced in neurons from the hippocampus, a portion of the brain that is
the body called endocannabinoids. Endocannabinoids are pro- vitally important in learning and short-term memory. When
duced by postsynaptic neurons following depolarization. hippocampal neurons are repeatedly stimulated over a short
These substances diffuse backwards across the synaptic cleft period of time, the synapses that connect these neurons to
to the presynaptic membrane, where they bind to CB1 recep- their neighbors become strengthened by a process known as
tors, suppressing synaptic transmission. CB1 receptors are long-term potentiation (LTP), which may last for days, weeks,
located in many areas of the brain, including the hippocam- or even longer. Research into LTP has focused on the NMDA
pus, cerebellum, and hypothalamus, which explains the effects receptor, which is one of several receptor types that bind the
of marijuana on memory, motor coordination, and appetite, excitatory neurotransmitter glutamate. When glutamate
binds to a postsynaptic NMDA receptor, it opens an internal

cation channel within the receptor that allows the influx of
REVIEW
?
Ca2# ions into the postsynaptic neuron, triggering a cascade 1. What is a resting potential? How is it established as a
of biochemical changes that lead to synaptic strengthening. result of ion flow?
Synapses that have undergone LTP are able to transmit 2. What is an action potential? What are the steps that lead
weaker stimuli and evoke stronger responses in postsynaptic to its various phases?
cells. These changes are thought to play a major role as newly
3. How is an action potential propagated along an axon?
learned information or memories are encoded in the neural
What is saltatory conduction, and how does such a
circuits of the brain. When laboratory animals are treated with
process occur?
drugs that inhibit LTP, such as those that interfere with the
activity of the NMDA receptor, their ability to learn new in- 4. What is the role of the myelin sheath in conduction of an
formation is greatly reduced. impulse?
There are numerous other reasons the study of synapses is 5. Describe the steps between the time an impulse reaches
so important. For example, a number of diseases of the nerv- the terminal knob of a presynaptic neuron and an action
ous system, including myasthenia gravis, Parkinsons disease, potential is initiated in a postsynaptic cell.
schizophrenia, and even depression, are thought to have their 6. Contrast the roles of ion pumps and channels in establish-
roots in synaptic dysfunction. ing and using ion gradients, particularly as it applies to
nerve cells.
EXPERIMENTAL PATHWAYS
The Acetylcholine Receptor
In 1843, at the age of 30, Claude Bernard moved from a small was allowed to drain into the medium bathing a second isolated
French town, where he had been a pharmacist and an aspiring heart. The rate of the second heart slowed dramatically, as though
playwright, to Paris, where he planned to pursue his literary career. its own inhibitory nerve had been activated.2 Loewi called the sub-
Instead, Bernard enrolled in medical school and went on to be- stance responsible for inhibiting the frogs heart Vagusstoff.
come the foremost physiologist of the nineteenth century. Among Within a few years, Loewi had shown that the chemical and physi-
his many interests was the mechanism by which nerves stimulate ologic properties of Vagusstoff were identical to acetylcholine, and
the contraction of skeletal muscles. His studies included the use of he concluded that acetylcholine (ACh) was the substance released
curare, a highly toxic drug isolated from tropical plants and utilized by the tips of the nerve cells that made up the vagus nerve.
for centuries by native South American hunters to make poisonous In 1937, David Nachmansohn, a neurophysiologist at the
darts. Bernard found that curare paralyzed a skeletal muscle with- Sorbonne, was visiting the Worlds Fair in Paris where he observed
out interfering with either the ability of nerves to carry impulses to several living electric fish of the species Torpedo marmarota that were
that muscle or the ability of the muscle to contract on direct stim- on display. These rays have electric organs that deliver strong shocks
ulation. Bernard concluded that curare somehow acted on the re- (4060 volts) capable of killing potential prey. At the time,
gion of contact between the nerve and muscle. Nachmansohn was studying the enzyme acetylcholinesterase, which
This conclusion was confirmed and extended by John Langley, acts to destroy ACh after its release from the tips of motor nerves.
a physiologist at Cambridge University. Langley was studying the Nachmansohn was aware that the electric organs of these fish
ability of nicotine, another substance derived from plants, to stimu- were derived from modified skeletal muscle tissue (Figure 1), and
late the contraction of isolated frog skeletal muscles and the effect of he asked if he could have a couple of the fish for study once the
curare in inhibiting nicotine action. In 1906, Langley concluded that fair had ended. The results of the first test showed the electric or-
the nervous impulse should not pass from nerve to muscle by an gan was an extraordinarily rich source of acetylcholinesterase.3 It
electric discharge, but by the secretion of a special substance on the was also a very rich source of the nicotinic acetylcholine receptor
end of the nerve.1 Langley proposed that this chemical transmitter (nAChR),* the receptor present on the postsynaptic membranes of
was binding to a receptive substance on the surface of the muscle skeletal muscle cells that binds ACh molecules released from the tips
cells, the same site that bound nicotine and curare. These proved to
be farsighted proposals.
*The receptor is described as nicotinic because it can be activated by nicotine as
Langleys suggestion that the stimulus from nerve to muscle
well as by acetylcholine. This contrasts with muscarinic acetylcholine receptors
was transmitted by a chemical substance was confirmed in 1921 in of the parasympathetic nerve synapses, which can be activated by muscarine, but
an ingenious experiment by the Austrian-born physiologist, Otto not nicotine, and are inhibited by atropine, but not curare. Smokers bodies be-
Loewi, the design of which came to Loewi during a dream. The come accustomed to high levels of nicotine, and they experience symptoms of
heart rate of a vertebrate is regulated by input from two opposing withdrawal when they stop smoking because the postsynaptic neurons that pos-
sess nAChRs are no longer stimulated at their usual level. The drug Chantix,
(antagonistic) nerves. Loewi isolated a frogs heart with both nerves which is marketed as an aid to stop smoking, acts by binding to the most common
intact. When he stimulated the inhibitory (vagus) nerve, a chemical version of brain nAChR (one with "4!2 subunits). Once bound, the Chantix
was released from the heart preparation into a salt solution, which molecule partially stimulates the receptor while preventing binding of nicotine.
THE ACETYLCHOLINE RECEPTOR 167
proteins. As discussed in Chapter 18, the purification of a particular

Electric ray
protein requires a suitable assay to determine the amount of that pro-
tein present in any particular fraction. The ideal assay for nAChR was
a compound that bound selectively and tightly to this particular pro-
tein. Such a compound was discovered in 1963 by Chen-Yuan Lee
and his colleagues of the National Taiwan University. The compound
was "-bungarotoxin, a substance present in the venom of a Taiwanese
snake. The "-bungarotoxin causes paralysis by binding tightly to the
nAChRs on the postsynaptic membrane of skeletal muscle cells,
blocking the response of the muscle to ACh.4
Equipped with labeled "-bungarotoxin to use in an assay,
Electroplax
organs electric organs as a source, and a detergent capable of solubilizing
membrane proteins, a number of investigators were able to isolate
the acetylcholine receptors in the early 1970s. In one of these
studies,5 the membranes containing the nAChR were isolated by
FIGURE 1 The electric organs of Torpedo consist of stacks of modified homogenizing the electric organs in a blender and centrifuging the
neuromuscular junctions located on each side of the body. (FROM Z. W. suspension to pellet the membrane fragments. Membrane proteins
HALL, AN INTRODUCTION TO NEUROBIOLOGY, SINAUER ASSOCIATES, INC., were extracted from the membrane fragments using Triton X-100
SUNDERLAND, MA 1992.) (page 129), and the mixture was passed through a column contain-
ing tiny beads coated with a synthetic compound whose end bears a
structural resemblance to ACh (Figure 2a). As the mixture of dis-
solved proteins passed through the column, two proteins that have
of a motor nerve. Finding an ideal system can prove invaluable in the binding sites for acetylcholine, nAChR and acetylcholinesterase
study of a particular aspect of cell structure or function. As will be (AChE), stuck to the beads. The remaining 90 percent of the protein
evident from this discussion, the electric organs of fish have been vir- in the extract failed to bind to the beads and simply passed through
tually the sole source of material for the study of the nAChR. the column and was collected (Figure 2b). Once this group of pro-
The nAChR is an integral membrane protein, and it wasnt un- teins had passed through, a solution of 10!3 M flaxedil was passed
til the 1970s that techniques were developed for the isolation of such through the column, which selectively removed the nAChR from
C2H5 FIGURE 2 Steps used in the isolation of the nAChR.

CH2 CH2 N C2H5 (a) Structure of a synthetic compound, CT5263, that
O C2H5
O O was attached to sepharose beads to form an affinity
H H C2H5
column. The ends of the compound projecting from
NH(CH2)6 N C C CH2 CH2 S CH2 C N O CH2 CH2 N C2H5
H C2H5 the beads resemble acetylcholine, causing both acetyl-
NH
SEPHAROSE
cholinesterase (AChE) and the nicotinic acetylcholine
O C receptor (nAChR) to bind to the beads. (b) When the
2B
CH3 Triton X-100 extract was passed through the column,
(a)
both of the acetylcholine-binding proteins stuck to the
beads, while the remaining dissolved protein (about
(ATCh x hr-1 x l-1)
( M x l -1 x 1 0 8)
90 percent of the total protein in the extract) passed

[Proteins]
( g x l -1)
directly through the column. Subsequent passage of a

[RACh]
AChE
-3
solution of 10!3 M flaxedil through the column re-
flaxedil 10 M 1MNaCl
leased the bound nAChR, without disturbing the
bound AChE (which was subsequently eluted with
1.5 1 M NaCl). (FROM R. W. OLSEN, J.-C. MEUNIER, AND
5 J.-P. CHANGEUX, FEBS LETT. 28:99, 1972.)
1.0
PROTEINS ACh-RECEPTOR AChE
1.0
0.5
0.5
0 0
0 0
0 10 20 30 40 50 60 70 80 90
n fraction (2ml)
(b)
the beads, leaving the AChE behind. Using this procedure, the lipid vesicles.7 Using vesicles containing various concentrations of
acetylcholine receptor as measured by bungarotoxin binding was pu- labeled sodium and potassium ions, they demonstrated that binding
rified by more than 150-fold in a single step. This type of procedure of ACh to the receptors in the lipid bilayer initiated a flux of cations
is known as affinity chromatography, and its general use is discussed in through the membrane. It was evident that the pure protein does
Section 18.7. indeed contain all the structural elements needed for the chemical
The next step was to determine something about the structure transmission of an electrical signalnamely, an acetylcholine bind-
of the acetylcholine receptor. Studies in the laboratory of Arthur ing site, an ion channel and a mechanism for coupling their activity.
Karlin at Columbia University determined that the nAChR was a During the past two decades, researchers have focused on deter-
pentamer, a protein consisting of five subunits. Each receptor con- mining the structure of the nAChR and the mechanism by which
tained two copies of a subunit called !, and one copy each of three binding of acetylcholine induces the opening of the ion gate. Analy-
other subunits. The subunits could be distinguished by extracting sis of the structure has taken several different paths. In one approach,
membrane proteins in Triton X-100, purifying the nAChR by scientists have used purified genes, determination of amino acid se-
affinity chromatography, and then subjecting the purified protein to quences, and site-directed mutagenesis to determine the specific
electrophoresis through a polyacrylamide gel (SDSPAGE, as dis- parts of the polypeptides that span the membrane, or bind the neu-
cussed in Section 18.7), which separates the individual polypeptides rotransmitter, or form the ion channel. These noncrystallographic
according to size (Figure 3).6 The four different subunits proved to studies on the molecular anatomy of a protein are similar in princi-
be homologous to one another, each subunit containing four homol- ple to those described in Section 4.4.
ogous transmembrane helices (M1M4). Another approach employed the electron microscope. The first
Another important milestone in the study of the nAChR was glimpses of the nAChR were seen in electron micrographs of the
the demonstration that the purified receptor acted as both a site for membranes of the electric organs (Figure 4).8 The receptors ap-
binding ACh and a channel for the passage of cations. It had been peared to be ring shaped, with a diameter of 8 nm and a central
postulated years earlier by Jean-Pierre Changeux of the Pasteur In- channel of 2-nm diameter, and protruded out from the lipid bilayer
stitute in Paris that the binding of ACh to the receptor caused a con- into the external space. An increasingly detailed model of the
formational change that opened an ion channel within the protein. nAChR has been developed by Nigel Unwin and his colleagues of
The inward flux of Na" ions through the channel could then lead to the Medical Research Council of England.913 Using the technique
a depolarization of the membrane and the activation of the muscle of electron crystallography (Section 18.8), which involves a mathe-
cell. During the last half of the 1970s, Changeux and his colleagues matical analysis of electron micrographs of frozen membranes from
succeeded in incorporating purified nAChR molecules into artificial the electric organs, Unwin was able to analyze the structure of the
nAChR as it exists in its native lipid environment. He described the
arrangement of the five subunits around a central channel (Figure 5).
58,000 48,000
The ion channel consists of a narrow pore lined by a wall composed
64,000 39,000 of five inner (M2) ! helices, one from each of the surrounding sub-
units. The gate to the pore lies near the middle of the membrane,
where each of the M2 ! helices bends inward (at the apex of the V-
FIGURE 3 The top portion of the figure shows an SDSpolyacrylamide

gel following electrophoresis of a preparation of the purified nAChR.
The receptor consists of four different subunits whose molecular weights
are indicated. Prior to electrophoresis, the purified receptor preparation
was incubated with a radioactive compound (3H-MBTA) that resembles
acetylcholine and binds to the acetylcholine-binding site of the nAChR.
Following electrophoresis, the gel was sliced into 1-mm sections and the
radioactivity of each slice determined. All of the radioactivity was bound
to the 39,000 dalton subunit, indicating this subunit contains the ACh-
binding site. The dotted line indicates the light absorbance of each frac-
tion, which provides a measure of the total amount of protein present in
that fraction. The heights of the peaks provide a measure of the relative FIGURE 4 Electron micrograph of negatively stained, receptor-rich
amounts of each of the subunits in the protein. All of the subunits are membranes from the electric organ of an electric fish showing the dense
present in equal numbers except the smallest subunit (the ! subunit, array of nAChR molecules. Each receptor molecule is seen as a small
which contains the ACh-binding site), which is present in twice the whitish circle with a tiny black dot in its center; the dot corresponds to
number of copies. (FROM C. L. WEILL, M. G. MCNAMEE, AND A. KARLIN, the central channel, which has collected electron-dense stain. (FROM
BIOCHEM. BIOPHYS. RES. COMMUN. 61:1002, 1974.) WERNER SCHIEBLER AND FERDINAND HUCHO, EUR. J. BIOCHEM. 85:58, 1978.)
THE ACETYLCHOLINE RECEPTOR 169
Ions
ACh
Acetylcholine Binding
pocket
FIGURE 5 (a) An electron density map
of a slice through the nAChR obtained

by analyzing electron micrographs of
tubular crystals of Torpedo membranes
embedded in ice. These analyses have Sodium
allowed researchers to reconstruct the ion
three-dimensional appearance of a single
nAChR protein as it resides within the
membrane. The continuous contours in-
dicate lines of similar density greater
than that of water. The two dark, bar-
shaped lines represent the ! helices that
line the channel at its narrowest point.
(b) Schematic diagram of the nAChR Cytoplasm
showing the arrangement of the subunits Cell
and a cross-sectional representation of membrane Ion channel
the protein. Each of the five subunits (a) (b)
contains four membrane-spanning he-
lices (M1M4). Of these, only the inner helix (M2) lines the pore, and J. MOL. BIOL. 229:1118, 1993; COPYRIGHT 1993, BY PERMISSION OF THE
is the subject of the remainder of the discussion. (A: FROM N. UNWIN, PUBLISHER ACADEMIC PRESS, ELSEVIER SCIENCE.)
shaped bars in Figure 5a) to form a kink in the unactivated receptor. Figure 6, this conformational change is propagated down the pro-
The side chain of a leucine residue projects inward from each kink. tein, causing a small (15") rotation of the M2 helices that line the
In this model, the leucine residues from the five inner helices form a ion-conduction pore. The rotation of these inner helices breaks
tight hydrophobic ring that prevents the ions from crossing the apart the hydrophobic gate, which allows Na# ions to enter the
membrane. The gate opens following the binding of two ACh mol- cell.12,13 Other models for channel opening have been proposed by
ecules, one per ! subunit. Each ACh binds to a site located within a other laboratories, and a more definitive statement about the mech-
pocket of an ! subunit (Figure 5b). anism of gating will probably require a high-resolution X-ray crystal
To study the changes in the nAChR during channel opening, structure of the protein in both the open and closed states. Discus-
Unwin carried out the following experiment.11 Preparations of sion of various models and the bases for them can be found in
nAChR-rich membranes were applied to a support grid that was al- References 1416.
lowed to fall into a bath of liquid nitrogen-cooled ethane, which
freezes the membranes. Approximately 5 msec before they reached
the surface of the freezing bath, the grids were sprayed with a FIGURE 6 Ribbon drawings illustrating the proposed changes that
solution of ACh, which bound to the receptors and triggered the occur within the nAChR upon binding of acetylcholine. Only the two
conformational change required to open the channel. By comparing alpha subunits of the receptor are shown. In the closed state (left) the
electron micrographs of nAChRs trapped in the open versus closed pore is blocked (pink patch) by the close apposition of a ring of hy-
state, Unwin found that ACh binding triggers a small conforma- drophobic residues (the valine and leucine side chains of these residues on
tional change in the extracellular domains of the receptor subunits the alpha subunits are indicated by the small ball-and-stick models at the
near the two ACh binding sites. According to the model depicted in site of pore constriction). The diameter of the pore at its narrowest point
is about 6 , which is not sufficient for a hydrated Na# ion to pass.
Although it is wide enough for passage of a dehydrated Na# ion, the wall
of the channel lacks the polar groups that would be required to substitute
for the displaced shell of water molecules (as occurs in the selectivity fil-
ter of the K# channel, Figure 4.39). A tryptophan side chain in the
cytoplasmic domains of the subunits indicates the approximate site of
acetylcholine binding. Following binding of ligand to each cytoplasmic
domain (which is seen to consist largely of $ sheets), a conformational
change is proposed, which leads to a small rotation of the inner $ sheets
in the cytoplasmic domain (curved arrows on left figure). This, in turn,
induces a rotational movement of the inner transmembrane helices of the
subunits, expanding the diameter of the pore, which allows the flow of
Na# ions through the open state of the channel (right). The relevant
moving parts are shown in blue. (FROM N. UNWIN, FEBS LETT. 555:94,
2003, COPYRIGHT 2003, WITH PERMISSION FROM ELSEVIER SCIENCE.)
References 8. SCHIEBLER, W. & HUCHO, F. 1978. Membranes rich in acetylcholine re-

ceptor: Characterization and reconstitution to excitable membranes from
1. LANGLEY, J. N. 1906. On nerve endings and on special excitable substances
exogenous lipids. Eur. J. Biochem. 85:5563.
in cells. Proc. R. Soc. London B Biol. Sci. 78:170194.
9. BRISSON, A. & UNWIN, N. 1984. Tubular crystals of acetylcholine receptor.
2. LOEWI, O. 1921. Uber humorale ubertragbarkeit der herznervenwirkung.
J. Cell Biol. 99:12021211.
Pflugers Arch. 214:239242. (A review of Loewis work written by him in
English can be found in Harvey Lect. 28:218233, 1933.) 10. UNWIN, N. 1993. Acetylcholine receptor at 9 resolution. J. Mol. Biol.
229:11011124.
3. MARNAY, A. 1937. Cholinesterase dans lorgane lectrique de la torpille.
Compte Rend. 126:573574. (A review of Nachmansohns work written in 11. UNWIN, N. 1995. Acetylcholine receptor channel imaged in the open state.
English can be found in his book, Chemical and Molecular Basis of Nerve Ac- Nature 373:3743.
tion, 2nd ed., Academic Press, 1975.) 12. MIYAZAWA, A., F UJIYOSHI, Y. & UNWIN, N. 2003. Structure and gating
4. CHANG, C. C. & LEE, C.-Y. 1963. Isolation of the neurotoxins from the mechanism of the acetylcholine receptor pore. Nature 423:949955.
venom of Bungarus multicinctus and their modes of neuromuscular blocking 13. UNWIN, N. 2005. Refined structure of the nicotinic acetylcholine receptor at
action. Arch. Int. Pharmacodyn. Ther. 144:241257. 4.0 resolution. J. Mol. Biol. 346:967989.
5. OLSEN, R. W., MEUNIER, J. C., & CHANGEUX, J. P. 1972. Progress in the 14. CORRY, B. 2006. An energy-efficient gating mechanism in the acetylcholine
purification of the cholinergic receptor protein from Electrophorus electricus by receptor channel suggested by molecular and brownian dynamics. Biophys. J.
affinity chromatography. FEBS Lett. 28:96100. 90:799810.
6. WEILL, C. L., MCNAMEE, M. G., & KARLIN, A. 1974. Affinity-labeling of 15. CYMES, G. D. & GROSMAN, C. 2008. Pore-opening mechanism of the
purified acetylcholine receptor from Torpedo californica. Biochem. Biophys. Res. nicotinic acetylcholine receptor evinced by proton transfer. Nature Struct.
Commun. 61:9971003. Mol. Biol. 15:389396.
7. POPOT, J. L., CARTAUD, J., & CHANGEUX, J.-P. 1981. Reconstitution of a 16. CHANGEUX, J.-P. & TALY, A. 2008. Nicotinic receptors, allosteric proteins
functional acetylcholine receptor. Eur. J. Biochem. 118:203214. and medicine. Trends. Mol. Med. 14:93102.
SYNOPSIS
Plasma membranes are remarkably thin, delicate structures, yet proteins bear sites on their external surface that interact with extra-
they play a key role in many of a cells most important functions. cellular ligands and sites on their internal surface that interact with
The plasma membrane separates the living cell from its environ- peripheral proteins that form part of an inner membrane skeleton;
ment; it provides a selectively permeable barrier that allows the ex- and the phospholipid content of the two halves of the bilayer is
change of certain substances, while preventing the passage of others; highly asymmetric. The organization of the proteins within the
it contains the machinery that physically transports substances from membrane is best revealed in freeze-fracture replicas in which the
one side of the membrane to another; it contains receptors that bind cells are frozen, their membranes are split through the center of
to specific ligands in the external space and relay the information to the bilayer by a fracture plane, and the exposed internal faces are
the cells internal compartments; it mediates interactions with other visualized by formation of a metal replica. (p. 125)
cells; it provides a framework in which components can be organized; The physical state of the lipid bilayer has important consequences
it is a site where energy is transduced from one type to another. (p. 117) for the lateral mobility of both phospholipids and integral pro-
Membranes are lipidprotein assemblies in which the compo- teins. The viscosity of the bilayer and the temperature at which it
nents are held together in a thin sheet by noncovalent bonds. The undergoes phase transition depend on the degree of unsaturation
membrane is held together as a cohesive sheet by a lipid bilayer and the length of the fatty acyl chains of the phospholipids. Main-
consisting of a bimolecular layer of amphipathic lipids, whose polar taining a fluid membrane is important for many cellular activities,
head groups face outward and hydrophobic fatty acyl tails face in- including signal transduction, cell division, and formation of special-
ward. Included among the lipids are phosphoglycerides, such as ized membrane regions. The lateral diffusion of proteins within the
phosphatidylcholine; sphingosine-based lipids, such as the phospho- membrane was originally demonstrated by cell fusion, and it can be
lipid sphingomyelin and the carbohydrate-containing cerebrosides quantitated by techniques that follow the movements of proteins
and gangliosides (glycolipids); and cholesterol. The proteins of the tagged with fluorescent compounds or electron-dense markers.
membrane can be divided into three groups: integral proteins that Measurement of the diffusion coefficients of integral proteins sug-
penetrate into and through the lipid bilayer, with portions exposed gests that most are subject to restraining influences that inhibit their
on both the cytoplasmic and extracellular membrane surfaces; pe- mobility. Proteins may be restrained by association with other inte-
ripheral proteins that are present wholly outside the lipid bilayer, but gral proteins or with peripheral proteins located on the membrane sur-
are noncovalently associated with either the polar head groups of the face. Because of these various types of restraint, membranes achieve a
lipid bilayer or with the surface of an integral protein; and lipid- considerable measure of organizational stability in which particular
anchored proteins that are outside the lipid bilayer but covalently membrane regions are differentiated from one another. (p. 133)
linked to a lipid that is part of the bilayer. The transmembrane seg- The erythrocyte plasma membrane contains two major integral
ments of integral proteins occur typically as an ! helix, composed proteins, band 3 and glycophorin A, and a well-defined inner
predominantly of hydrophobic residues. (p. 122) skeleton composed of peripheral proteins. Each band 3 subunit
Membranes are highly asymmetric structures whose two leaflets spans the membrane at least a dozen times and contains an internal
have very different properties. As examples, all of the membranes channel through which bicarbonate and chloride anions are ex-
carbohydrate chains face away from the cytosol; many of the integral changed. Glycophorin A is a heavily glycosylated protein of uncer-
ANALYTIC QUESTIONS 171
tain function containing a single transmembrane domain consisting the two sides of the membrane. The glucose transporter is a facilita-
of a hydrophobic ! helix. The major component of the membrane tive transporter whose presence in the plasma membrane is stimu-
skeleton is the fibrous protein spectrin, which interacts with other lated by increasing levels of insulin. Active transporters require the
peripheral proteins to provide support for the membrane and re- input of energy and move ions and solutes against a concentration
strain the diffusion of its integral proteins. (p. 140) gradient. P-type active transporters, such as the Na"/K"-ATPase,
are driven by the transfer of a phosphate group from ATP to the
The plasma membrane is a selectively permeable barrier that al-
transporter, changing its affinity toward the transported ion. Sec-
lows solute passage by several mechanisms, including simple dif-
ondary active-transport systems tap the energy stored in an ionic
fusion through either the lipid bilayer or membrane channels,
gradient to transport a second solute against a gradient. For example,
facilitated diffusion, and active transport. Diffusion is an energy-
the active transport of glucose across the apical surface of an intes-
independent process in which a solute moves down an electrochem-
tinal epithelial cell is driven by the cotransport of Na" down its elec-
ical gradient, dissipating the free energy stored in the gradient. Small
trochemical gradient. (p. 152)
inorganic solutes, such as O2, CO2, and H2O, penetrate the lipid bi-
layer readily, as do solutes with large partition coefficients (high lipid The resting potential across the plasma membrane is due largely
solubility). Ions and polar organic solutes, such as sugars and amino to the limited permeability of the membrane to K! and is subject
acids, require special transporters to enter or leave the cell. (p. 143) to dramatic change. The resting potential of a typical nerve or
muscle cell is about #70 mV (inside negative). When the membrane
Water moves by osmosis through the semipermeable plasma
of an excitable cell is depolarized past a threshold value, events are
membrane from a region of lower solute concentration (the hypo-
initiated that lead to the opening of the gated Na" channels and the
tonic compartment) to a region of higher solute concentration
influx of Na", which is measured as a reversal in the voltage across
(the hypertonic compartment). Osmosis plays a key role in a mul-
the membrane. Within milliseconds after they have opened, the
titude of physiologic activities. In plants, for example, the influx of
Na" gates close and gated potassium channels open, which leads to
water generates turgor pressure against the cell wall that helps sup-
an efflux of K" and a restoration of the resting potential. The series
port nonwoody tissues. (p. 144)
of dramatic changes in membrane potential following depolarization
Ions diffuse through a plasma membrane by means of special constitutes an action potential. (p. 159)
protein-lined channels that are often specific for particular ions. Once an action potential has been initiated, it becomes self-
Ion channels are usually gated and are controlled by either voltage or propagating. Action potentials propagate because the depolarization
chemical ligands, such as neurotransmitters. Analysis of a bacterial ion that accompanies an action potential at one site on the membrane is
channel (KcsA) has revealed how oxygen atoms from the backbone of sufficient to depolarize the adjacent membrane, which initiates an
the polypeptide are capable of replacing the water molecules normally action potential at that site. In a myelinated axon, an action potential
assocated with K" ions, allowing the protein to selectively conduct K" at one node in the sheath is able to depolarize the membrane at the
ions through its central channel. Voltage-gated K" channels contain a
next node, allowing the action potential to hop rapidly from node to
charged helical segment that moves in response to the membranes
node. When the action potential reaches the terminal knobs of an
voltage leading to the opening or closing of the gate. (p. 146)
axon, the calcium gates in the plasma membrane open, allowing an
Facilitated diffusion and active transport involve integral mem- influx of Ca2", which triggers the fusion of the membranes of
brane proteins that combine specifically with the solute to be neurotransmitter-containing secretory vesicles with the overlying
transported. Facilitative transporters act without the input of en- plasma membrane. The neurotransmitter diffuses across the synap-
ergy and move solutes down a concentration gradient in either direc- tic cleft and binds to receptors on the postsynaptic membrane, in-
tion across the membrane. They are thought to act by changing ducing either the depolarization or hyperpolarization of the target
conformation, which alternately exposes the solute binding site to cell. (p. 162)
ANALYTIC QUESTIONS
1. What types of integral proteins would you expect to reside in 5. How is it that, unlike polysaccharides such as starch and glyco-
the plasma membrane of an epithelial cell that might be absent gen, the oligosaccharides on the surface of the plasma mem-
from that of an erythrocyte? How do such differences relate to brane can be involved in specific interactions? How is this
the activities of these cells? feature illustrated by determining a persons blood type prior to
2. Many different types of cells possess receptors that bind steroid receiving a transfusion?
hormones, which are lipid-soluble molecules. Where in the cell 6. Trypsin is an enzyme that can digest the hydrophilic portions of
do you think such receptors might reside? Where in the cell membrane proteins, but it is unable to penetrate the lipid bilayer
would you expect the insulin receptor to reside? Why? and enter a cell. Because of these properties, trypsin has been
3. When the trilaminar appearance of the plasma membrane was used in conjunction with SDSPAGE to determine which pro-
first reported, the images (e.g., Figure 4.1a) were taken as evi- teins have an extracellular domain. Describe an experiment us-
dence to support the Davson-Danielli model of plasma mem- ing trypsin to determine the sidedness of proteins of the
brane structure. Why do you think these micrographs might erythrocyte membrane.
have been interpreted in this way? 7. Look at the scanning electron micrograph of erythrocytes in
4. Suppose you were planning to use liposomes in an attempt to Figure 4.32a. These cells, which are flattened and have circular
deliver drugs to a particular type of cell in the body, for example, depressions on each side, are said to have a biconcave shape.
a fat or muscle cell. Is there any way you might be able to con- What is the physiologic advantage of a biconcave erythocyte
struct the liposome to increase its target specificity? over a spherical cell with respect to O2 uptake?
8. Suppose you were culturing a population of bacteria at 15!C and 19. As discussed on page 158, the Na"/glucose cotransporter trans-
then raised the temperature of the culture to 37!C. What effect ports two Na" ions for each glucose molecule. What if this ra-
do you think this might have on the fatty acid composition of tio were 1!1 rather than 2!1; how would this affect the glucose
the membrane? on the transition temperature of the lipid bi- concentration that the transporter could work against?
layer? on the activity of membrane desaturases?
20. A transmembrane protein usually has the following features:
9. Looking at Figure 4.6, which lipids would you expect to have (1) the portion that transits the membrane bilayer is at least
the greatest rate of flip-flop? which the least? Why? If you de- 20 amino acids in length, which are largely or exclusively non-
termined experimentally that phosphatidylcholine actually ex- polar residues; (2) the portion that anchors the protein on the
hibited the greatest rate of flip-flop, how might you explain this external face has two or more consecutive acidic residues; and
finding? How would you expect the rate of phospholipid flip- (3) the portion that anchors the protein on the cytoplasmic face
flop to compare with that of an integral protein? Why? has two or more consecutive basic residues. Consider the trans-
10. What is the difference between a two-dimensional and a three- membrane protein with the following sequence:
dimensional representation of a membrane protein? How are NH 2 -MLSTGVKRKGAVLLILLFPWMVAGGPLFWLA
the different types of profiles obtained, and which is more use- ADESTYKGS-COOH
ful? Why do you think there are so many more proteins whose
Draw this protein as it would reside in the plasma membrane.
two-dimensional structure is known?
Make sure you label the N- and C-termini and the external and
11. If you were to inject a squid giant axon with a tiny volume of so- cytoplasmic faces of the membrane. (Single-letter code for
lution containing 0.1 M NaCl and 0.1 M KCl in which both the amino acids is given in Figure 2.26.)
Na" and K" ions were radioactively labeled, which of the ra- 21. Many marine invertebrates, such as squid, have extracellular
dioactively labeled ions would you expect to appear most rapidly fluids that resemble seawater and therefore have much higher
within the seawater medium while the neuron remained at rest? intracellular ion concentrations than mammals. For a squid neu-
after the neuron had been stimulated to conduct a number of ron, the ionic concentrations are roughly
action potentials?
intracellular extracellular
12. It has been difficult to isolate proteins containing water chan- ion concentration concentration
nels (i.e., aquaporins) due to the high rate of diffusion of water
K" 350 mM 10 mM
through the lipid bilayer. Why would this make aquaporin iso-
Na" 40 mM 440 mM
lation difficult? Is there any way you might be able to distinguish
Cl# 100 mM 560 mM
diffusion of water through the lipid bilayer versus that through
Ca2" 2 $ 10#4 mM 10 mM
aquaporins? The best approach to studying aquaporin behavior
pH 7.6 8.0
has been to express the aquaporin genes in frog oocytes. Is there
any reason why the oocytes of a pond-dwelling amphibian If the resting potential of the plasma membrane, Vm, is #70 mV,
might be particularly well suited for such studies? are any of the ions at equilibrium? How far out of equilibrium,
in mV, is each ion? What is the direction of net movement of
13. How is it that diffusion coefficients measured for lipids within
each ion through an open channel permeable to that ion?
membranes tend to be closer to that expected for free diffu-
sion than those measured for integral proteins in the same 22. The membrane potential of a cell is determined by the relative
membranes? permeability of the membrane to various ions. When acetyl-
choline binds to its receptors on the postsynaptic muscle
14. Assume that the plasma membrane of a cell was suddenly per- membrane, it causes a massive opening of channels that are
meable to the same degree to both Na" and K" and that both equally permeable to sodium and potassium ions. Under these
ions were present at a concentration gradient of the same mag- conditions,
nitude. Would you expect these two ions to move across the
membrane at the same rates? Why or why not? Vm % (VK " " VNa " )/2
15. Most marine invertebrates show no loss or gain of water by os-
If [K"in] % 140 mM and [Na"in] % 10 mM for the muscle
mosis, whereas most marine vertebrates experience continual
cell, and [Na"out] % 150 mM and [K"out] % 5 mM, what is
water loss in their high-salt environment. Speculate on the basis
the membrane potential of the neuromuscular junction of an
for this difference and how it might reflect different pathways of
acetylcholine-stimulated muscle?
evolution of the two groups.
23. Transmembrane domains consist of individual & helices or a
16. How would you expect the concentrations of solute inside a ' sheet formed into a barrel. Looking at Figures 2.30 and 2.31,
plant cell to compare to that of its extracellular fluids? Would why is a single & helix more suited to spanning the bilayer than
you expect the same to be true of the cells of an animal? a single ' strand?
17. What would be the consequence for impulse conduction if the 24. Knowing how the K" channel selects for K" ions, suggest a
Na" channels were able to reopen immediately after they had mechanism by which the Na" channel is able to select for
closed during an action potential? its ion.
18. What would be the value of the potassium equilibrium potential 25. How would you compare the rate of movement of ions passing
if the external K" concentration was 200 mM and the internal through a channel versus those transported actively by a P-type
concentration was 10 mM at 25!C? at 37!C? pump? Why?

CMB CH4

Uploaded by

Copyright:

Available Formats

CMB CH4

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CMB CH4

Uploaded by

Copyright:

Available Formats

JWCL151_ch04_117-172.

qxd 6/25/09 1:24 PM Page 117

118 Chapter 4 THE STRUCTURE AND FUNCTION OF THE PLASMA MEMBRANE

FIGURE 4.1 The trilaminar appearance of membranes. (a) Electron micro-

4.2 A BRIEF HISTORY OF STUDIES ON PLASMA MEMBRANE STRUCTURE 119

(5) FIGURE 4.2 A summary of membrane functions in a plant cell. (1) An

120 Chapter 4 THE STRUCTURE AND FUNCTION OF THE PLASMA MEMBRANE

groups of the lipids could interact with surrounding water O

4.2 A BRIEF HISTORY OF STUDIES ON PLASMA MEMBRANE STRUCTURE 121

Protein molecule Polar pore

122 Chapter 4 THE STRUCTURE AND FUNCTION OF THE PLASMA MEMBRANE

4.3 THE CHEMICAL COMPOSITION

4.3 THE CHEMICAL COMPOSITION OF MEMBRANES 123

Recent interest has focused on the apparent health benefits of

Sphingolipids A less abundant class of membrane lipids, Dioleoyl O

substitution is a carbohydrate, the molecule is a glycolipid. If ethanolamine H3N CH2 CH2

tively little is known about them, yet tantalizing hints have O

Cholesterol Another lipid component of certain mem- O C R

branes is the sterol cholesterol (see Figure 2.21), which in O H H

sphingolipids. Sphingomyelin is a phospholipid; cerebrosides and gan-

124 Chapter 4 THE STRUCTURE AND FUNCTION OF THE PLASMA MEMBRANE

4.3 THE CHEMICAL COMPOSITION OF MEMBRANES 125

(a) (b) (c)

126 Chapter 4 THE STRUCTURE AND FUNCTION OF THE PLASMA MEMBRANE

FIGURE 4.11 Two types of linkages that join sugars to a polypep-

4.4 THE STRUCTURE AND FUNCTIONS OF MEMBRANE PROTEINS 127

Gal GlcNAc Gal Glu

GalNAc Gal GlcNAc Gal Glu

Integral membrane proteins

FIGURE 4.12 Blood-group antigens. Whether a person has type A, B,

4.4 THE STRUCTURE AND FUNCTIONS Etn

OF MEMBRANE PROTEINS GPI- anchored protein

surface of a membrane are very different from those of the

128 Chapter 4 THE STRUCTURE AND FUNCTION OF THE PLASMA MEMBRANE

4.4 THE STRUCTURE AND FUNCTIONS OF MEMBRANE PROTEINS 129

FIGURE 4.16 Solubilization of membrane proteins with detergents.

(uncharged) detergent Triton X-100 (which generally does

Cytoplasm Like membrane lipids, detergents are amphipathic, being com-

130 Chapter 4 THE STRUCTURE AND FUNCTION OF THE PLASMA MEMBRANE

containing 11 membrane-spanning " helices. Some of the

Identifying Transmembrane Domains A great deal can be

Phe IIe FIGURE 4.18 Glycophorin A, an integral protein with a sin-

4.4 THE STRUCTURE AND FUNCTIONS OF MEMBRANE PROTEINS 131

(a) (b) (c) (d)

polar regions of the membrane, even if it requires distorting

the helix to do so. Figure 4.19d shows two tyrosine residues +

ring is oriented parallel with the hydrocarbon chains of the

bilayer with which it has become integrated.

132 Chapter 4 THE STRUCTURE AND FUNCTION OF THE PLASMA MEMBRANE

! strands organized into a barrel as illustrated in Figure 5.4. To

Protein Suppose you have isolated a gene for an integral 245 II

determined that it contains four apparent membrane-spanning

4.5 MEMBRANE LIPIDS AND MEMBRANE FLUIDITY 133

FIGURE 4.22 Use of EPR spectroscopy to monitor

134 Chapter 4 THE STRUCTURE AND FUNCTION OF THE PLASMA MEMBRANE

4.5 MEMBRANE LIPIDS AND MEMBRANE FLUIDITY 135

136 Chapter 4 THE STRUCTURE AND FUNCTION OF THE PLASMA MEMBRANE

The controversy arises over whether similar types of CELL EXTERIOR

The Diffusion of Membrane Proteins

4.6 THE DYNAMIC NATURE OF THE PLASMA MEMBRANE 137