As we’ve seen, all antibodies have a similar structure and function;
enzymes are
structurally varied, but all have a catalytic function. In contrast, although all
membrane
proteins are located at the
membrane, they otherwise are both structurally and functionally
diverse. As we noted in
Chapter 2 and discuss in
more detail in
Chapter 5, every biological
membrane has the same basic
phospholipid bilayer structure. Associated with each
membrane is a
set of
membrane proteins that enables the
membrane to carry out its distinctive activities
(). The complement of
proteins attached to a
membrane varies depending on cell type and subcellular location.
Some proteins are bound only to the membrane surface, whereas others have one region buried
within the membrane and domains on one or both sides of it. Protein domains on the extracellular
membrane surface are generally involved in cell-cell signaling or interactions. Domains within
the membrane, particularly those that form channels and pores, move molecules across the
membrane. Domains lying along the cytosolic face of the membrane have a wide range of functions,
from anchoring cytoskeletal proteins to the membrane to triggering intracellular signaling
pathways. In many cases, the function of a membrane protein and the topology of its polypeptide
chain in the membrane can be predicted based on its homology with another, well-characterized
protein. In this section, we examine the characteristic structural features of membrane proteins
and some of their basic functions. More complete characterization of the structure and function
of various types of membrane proteins is presented in several later chapters. The synthesis and
processing of membrane proteins are discussed in Chapter
17.
Proteins Interact with Membranes in Different Ways
Membrane proteins can be classified into two broad categories—integral (intrinsic)
and peripheral (extrinsic)—based on the nature of the
membrane-
protein interactions
(see ). Most
biomembranes contain both types
of
membrane proteins.
Integral membrane proteins, also called
intrinsic proteins, have one or more segments that are embedded in the
phospholipid bilayer. Most integral proteins contain residues with hydrophobic side chains that
interact with fatty acyl groups of the membrane phospholipids, thus anchoring the protein to
the membrane. Most integral proteins span the entire phospholipid bilayer. These
transmembrane proteins contain one or more membrane-spanning domains as well
as domains, from four to several hundred residues long, extending into the aqueous medium on
each side of the bilayer. In all the transmembrane proteins examined to date, the
membrane-spanning domains are α helices or multiple β strands. In contrast,
some integral proteins are anchored to one of the membrane leaflets by covalently bound fatty
acids, as discussed later. In these proteins, the bound fatty acid is embedded in the membrane,
but the polypeptide chain does not enter the phospholipid bilayer.
Peripheral membrane proteins, or extrinsic
proteins, do not interact with the hydrophobic core of the phospholipid bilayer. Instead they
are usually bound to the membrane indirectly by interactions with integral membrane proteins or
directly by interactions with lipid polar head groups. Peripheral proteins localized to the
cytosolic face of the plasma membrane include the cytoskeletal proteins spectrin and actin in
erythrocytes (Chapter 18) and the enzyme protein
kinase C. This enzyme shuttles between the cytosol and the cytosolic face of the plasma
membrane and plays a role in signal transduction (Chapter 20). Other peripheral proteins, including certain proteins of the
extracellular matrix, are localized to the outer (exoplasmic) surface of the plasma
membrane.
Hydrophobic α Helices in Transmembrane Proteins Are Embedded in the
Bilayer
Figure 3-33
.
Amino acid sequence and transmembrane disposition of glycophorin A from the
erythrocyte plasma membrane
This protein is a homo-dimer, but only one of its polypeptide chains is shown. Residues
62–95 are buried in the membrane, with the sequence from position 73 through 95
forming an α helix. The ionic interactions shown between positively charged
arginine and lysine residues and negatively charged phospholipid head groups in the
cytosolic and exoplasmic faces of the membrane are hypothetical. Both the amino-terminal
segment of the molecule, located outside the cell, and the carboxy-terminal segment, located
inside the cell, are rich in charged residues and polar uncharged residues, making these
domains water-soluble. Note the numerous carbohydrate residues attached to amino acids in
the exoplasmic domain. [See V. T. Marchesi, H. Furthmayr, and M. Tomita, 1976, Ann.
Rev. Biochem.
45:667; A. H. Ross et al., 1982, J. Biol. Chem.
257:4152.]
Integral
proteins containing
membrane-spanning α-helical
domains are embedded in
membranes by
hydrophobic interactions with the
lipid interior of the
bilayer and probably also
by ionic interactions with the
polar head groups of the
phospholipids.
Glycophorin, a major erythrocyte
membrane protein, exhibits both types of
interaction. As shown in , glycophorin
contains a
membrane-embedded α helix composed entirely of
hydrophobic (or uncharged)
amino acids. The predicted length of this α helix (3.75 nm) is just sufficient to
span the hydrocarbon core of a
phospholipid bilayer. The
hydrophobic side chains form van der
Waals interactions with the fatty acyl chains and shield the
polar carbonyl (C=O)
and imino (NH) groups of the
peptide bond, which are all hydrogen-bonded to one another. This
hydrophobic helix is prevented from slipping across the
membrane by a flanking set of
positively charged
amino acids (lysine and arginine) that are thought to interact with
negatively charged phospholipid head groups. In glycophorin, most of these charged residues lie
adjacent to the cytosolic leaflet.
Many Integral Proteins Contain Multiple Transmembrane α Helices
Although depicts glycophorin as a
monomer
with a single α helix spanning the
bilayer, this
protein is present in erythrocyte
membranes as a dimer of two identical
polypeptide chains. The two
membrane-spanning α
helices of glycophorin are thought to form a
coiled-coil structure (see
Figure 3-9a) stabilized by specific interactions between the
amino acid side
chains at the interface of the two helices. It is now known that many other transmembrane
proteins contain two or more
membrane-spanning α helices. For instance, the
bacterial photosynthetic reaction center (PRC) comprises four subunits and
several
prosthetic groups, including four chlorophyll molecules. In this complex
protein, three
of the four subunits span the
membrane; two of these subunits (L and M) each contain five
membrane-spanning α helices (see
Figure
16-40).
Figure 3-34
.
Overall structure of bacteriorhodopsin as deduced from electron diffraction analyses
of two-dimensional crystals of the protein in the bacterial membrane
The seven membrane-spanning α helices are labeled A–G. The retinal
pigment is covalently attached to lysine 216 in helix G. The approximate position of the
protein in the phospholipid bilayer is indicated. [Adapted from R. Henderson et al., 1990,
J. Mol. Biol.
213:899.
A large and important family of integral
proteins is defined by the presence of seven
membrane-spanning α helices. More than 150 such “seven-spanning”
membrane proteins have been identified. This class of integral
proteins is typified by
bacteriorhodopsin, a
protein found in a photosynthetic bacterium (). Absorption of light by the retinal group
attached to bacteriorhodopsin causes a conformational change in the
protein that results in
pumping of protons from the
cytosol across the bacterial
membrane to the extracellular space.
The proton concentration gradient thus generated across the
membrane is used to synthesize ATP,
as discussed in
Chapter 16. Both the overall
arrangement of the seven α helices in bacteriorhodopsin and the identity of most of
the
amino acids can be resolved by computer analysis of micrographs of two-dimensional crystals
of the
membrane-embedded
protein taken at various angles to the electron beam.
Other seven-spanning membrane proteins include the opsins (eye proteins that absorb light),
cell-surface receptors for many hormones, and receptors for odorous molecules. Amino acid
sequence analysis of these proteins has shown that no amino acids are found in the same
position in all of them, and only a few residues are conserved in even a substantial number of
them. Nonetheless, each of these proteins contains seven stretches of hydrophobic amino acids
long enough (>22 amino acids) to span the phospholipid bilayer. Though direct evidence
is lacking, it is thought that all of these proteins adopt a conformation in the membrane
similar to that of bacteriorhodopsin. This is one of several examples of how investigators can
predict the orientation of proteins in a membrane from the amino acid sequence alone.
Multiple β Strands in Porins Form Membrane-Spanning
“Barrels”
The porins are a class of transmembrane proteins whose structure differs
radically from that of other integral proteins. Several types of porin are found in the outer
membrane of gram-negative bacteria such as E. coli(see Figure 1-7a). The outer membrane protects an intestinal bacterium from
harmful agents (e.g., antibiotics, bile salts, and proteases) but permits the uptake and
disposal of small hydrophilic molecules including nutrients and waste products. The porins in
the outer membrane of an E. coli cell provide channels for passage of
disaccharides, phosphate, and similar molecules.
Figure 3-35
.
Model of the three-dimensional structure of a subunit of OmpF, a porin found in the
E. coli outer membrane
All porins are trimeric transmembrane proteins. Each subunit is barrel-shaped with
β strands forming the wall and a transmembrane pore in the center. [Courtesy of S.
Choe.]
The
amino acid sequences of porins are predominantly
polar and contain no long
hydrophobic
segments typical of integral
proteins with α-helical
membrane-spanning
domains. X-ray
crystallography has revealed that porins are trimers of identical subunits. In each subunit 16
β strands form a barrel-shaped structure with a pore in the center (). As noted earlier, half the
amino acid side
groups of a β strand point in one direction, and the other half point in the opposite
direction (see
Figure 3-8). Unlike a typical globular
protein, porins have an inside-out arrangement. In a porin
monomer, the outward-facing side
groups on each of the β strands are
hydrophobic and thus can interact with the fatty
acyl groups of the
membrane lipids or with other porin
monomers. The side groups facing the
inside of a porin
monomer are predominantly
hydrophilic; these line the pore through which
small water-soluble molecules cross the
membrane.
Covalently Attached Hydrocarbon Chains Anchor Some Proteins to the Membrane
Figure 3-36
.
Anchoring of integral proteins to the plasma membrane by membrane-embedded hydrocarbon
groups (highlighted in red)
(a) Thy-1 protein and several hydrolytic enzymes are anchored by
glycosylphosphatidylinositol. This complex anchor is found only on the exoplasmic face. (b)
Cytosolic proteins involved in signaling such as Ras are anchored to the cytosolic face of
the membrane through farnesyl and palmitoyl groups. (c) Other cytosolic proteins are
associated with the membrane through myristate and similar fatty acids attached to an
N-terminal glycine residue.
In eukaryotic cells, as noted earlier, the
polypeptide chain of some integral
membrane
proteins does not enter the
bilayer but rather is anchored in one leaflet by a covalently
attached hydrocarbon chain. Several common
lipid anchors are shown in .
Some cell-surface
proteins are anchored to the
exoplasmic face of the
plasma membrane by a
complex glycosylated phospholipid that is linked to the C-terminus. A common example of this
type of anchor is
glycosylphosphatidylinositol, which contains two fatty acyl
groups,
N-acetylglucosamine, mannose, and inositol (see ). Several
enzymes, including alkaline
phosphatase, fall into
this class. Various experiments have shown that the phospholipid anchor is both necessary and
sufficient for binding these cell-surface
proteins to the
membrane. For instance, the
enzyme
phospholipase C cleaves the phosphate-glycerol bond in
phospholipids as well as in
glycosylphosphatidylinositol anchors, and treatment of cells with phospholipase C releases
glycosylphosphatidylinositol-anchored
proteins such as Thy-1
protein and alkaline
phosphatase
from the cell surface.
Some cytosolic
proteins are anchored to the
cytosolic face of
membranes by a hydrocarbon
moiety covalently attached to a cysteine near the C-terminus. The most common anchors are
prenyl, farnesyl, and geranylgeranyl groups. These
proteins undergo a chemical modification
involving several steps. First, the anchor moiety forms a thioether bond with the thiol group
of a cysteine that is four residues from the C-terminus of the
protein. The modified
protein
then undergoes proteolysis and methylation; these reactions remove the three terminal residues
and add a methyl to the new C-terminus. In some cases, fatty acyl palmitate groups form
thioester bonds to nearby cysteine residues, providing additional anchors that are thought to
reinforce the attachment of the
protein to the
membrane (see ).
In another group of
lipid-anchored cytosolic
proteins, a fatty acyl group (e.g., myristate or
palmitate) is linked by an amide bond to the N-terminal glycine residue (see ). In these
proteins, the N-terminal anchor is
necessary for retention at the
membrane and may play an important role in a
membrane-associated
function. For example, v-Src, a mutant form of a cellular tyrosine
kinase, is oncogenic and can
transform cells only when it retains a myristylated N-terminus.
Some Peripheral Proteins Are Soluble Enzymes That Act on Membrane Components
Figure 3-37
.
Specificity of cleavage of phospholipids by phospholipases A1,
A2, C, and D
Susceptible bonds are shown in red. R denotes the polar group attached to the phosphate,
such as choline in phosphatidylcholine (see Figure
5-27a) or inositol in phosphatidylinositol.
An important group of
peripheral membrane proteins are water-soluble
enzymes that associate
with the
polar head groups of
membrane phospholipids. One well-understood group of such
enzymes
are the
phospholipases, which hydrolyze various bonds in the head groups of
phospholipids (). These
enzymes have an
important role in the degradation of damaged or aged cell
membranes.
The mechanism of action of phospholipase A2 illustrates how such water-soluble
enzymes can reversibly interact with membranes and catalyze reactions at the interface of an
aqueous solution and lipid surface. When this enzyme is in aqueous solution, its
Ca2+-containing active site is buried in a channel lined with hydrophobic
amino acids. Binding of the enzyme to a phospholipid bilayer induces a small conformational
change that fixes the protein to the phospholipid heads and opens the hydrophobic cleft. As a
phospholipid molecule moves from the bilayer into the channel, the enzyme-bound
Ca2+ binds to the phosphate in the head group and positions the ester bond
to be cleaved next to the catalytic site.
ǀ