Chapter 12

c12Biosignaling.
indd Page 433 12/09/12 1:33 PM user-F408 /Users/user-F408/Desktop
12
Biosignaling
12.1 General Features of Signal Transduction 433 all these cases, the signal represents information that
is detected by specific receptors and converted to a cel-
12.2 G Protein–Coupled Receptors and Second lular response, which always involves a chemical pro-
Messengers 437 cess. This conversion of information into a chemical
12.3 Receptor Tyrosine Kinases 453 change, signal transduction, is a universal property of
living cells.
12.4 Receptor Guanylyl Cyclases, cGMP, and Protein
Kinase G 459
12.5 Multivalent Adaptor Proteins and Membrane Rafts 460 12.1 General Features of Signal
12.6 Gated Ion Channels 464 Transduction
12.7 Integrins: Bidirectional Cell Adhesion Receptors 470 Signal transductions are remarkably specific and
exquisitely sensitive. Specificity is achieved by pre-
12.8 Regulation of Transcription by Nuclear Hormone
cise molecular complementarity between the signal
Receptors 471
and receptor molecules (Fig. 12–1a), mediated by the
12.9 Signaling in Microorganisms and Plants 473 same kinds of weak (noncovalent) forces that mediate
enzyme-substrate and antigen-antibody interactions.
12.10 Sensory Transduction in Vision, Olfaction, and
Multicellular organisms have an additional level of
Gustation 477
specificity, because the receptors for a given signal, or
12.11 Regulation of the Cell Cycle by Protein Kinases 484 the intracellular targets of a given signal pathway, are
present only in certain cell types. Thyrotropin-releasing
12.12 Oncogenes, Tumor Suppressor Genes, and Programmed
hormone, for example, triggers responses in the cells
Cell Death 488
of the anterior pituitary but not in hepatocytes, which
lack receptors for this hormone. Epinephrine alters
T
he ability of cells to receive and act on signals from glycogen metabolism in hepatocytes but not in adipo-
beyond the plasma membrane is fundamental to cytes; in this case, both cell types have receptors for
life. Bacterial cells receive constant input from the hormone, but whereas hepatocytes contain glycogen
membrane proteins that act as information receptors, and the glycogen-metabolizing enzyme that is stimu-
sampling the surrounding medium for pH, osmotic lated by epinephrine, adipocytes contain neither. Adi-
strength, the availability of food, oxygen, and light, and pocytes respond to epinephrine by releasing fatty
the presence of noxious chemicals, predators, or com- acids from triacylglycerols and exporting them to other
petitors for food. These signals elicit appropriate tissues.
responses, such as motion toward food or away from Three factors account for the extraordinary sensi-
toxic substances or the formation of dormant spores in tivity of signal transduction: the high affinity of recep-
a nutrient-depleted medium. In multicellular organisms, tors for signal molecules, cooperativity (often but not
cells with different functions exchange a wide variety of always) in the ligand-receptor interaction, and amplifi-
signals. Plant cells respond to growth hormones and to cation of the signal by enzyme cascades. The affinity
variations in sunlight. Animal cells exchange informa- between signal (ligand) and receptor can be expressed
tion about the concentrations of ions and glucose in as the dissociation constant Kd, commonly 10!10 M or
extracellular fluids, the interdependent metabolic activ- less—meaning that the receptor detects picomolar con-
ities taking place in different tissues, and, in an embryo, centrations of a signal molecule. Receptor-ligand inter-
the correct placement of cells during development. In actions are quantified by Scatchard analysis, which
433
c12Biosignaling.indd Page 434 12/09/12 1:33 PM user-F408 /Users/user-F408/Desktop
434 Biosignaling
S2 Signal
S1
(a) Specificity (d) Desensitization/Adaptation
Signal molecule fits Receptor activation triggers
binding site on its a feedback circuit that shuts
complementary receptor; off the receptor or removes Receptor
other signals do not fit. Receptor it from the cell surface.
Response
Response
(e) Integration
When two signals have Signal 1 Signal 2
(b) Amplification Signal opposite effects on a
When enzymes activate metabolic characteristic
enzymes, the number of such as the concentration
affected molecules of a second messenger X,
increases geometrically Enzyme 1 Receptor Receptor
or the membrane potential 1 2
in an enzyme cascade. Vm, the regulatory outcome
results from the integrated
input from both receptors.
Enzyme 2 2 2 [X] or Vm [X] or Vm
Net D[X] or DVm

Enzyme
3 3 3 3 3 3 3 3 3
3 Response
(c) Modularity Signal

Proteins with multivalent
affinities form diverse
signaling complexes from
interchangeable parts.
Phosphorylation provides
reversible points of Modularity of interacting signaling proteins allows
interaction. P a cell to mix and match a set of signaling molecules to
create complexes with different functions or cellular
locations. Many signaling proteins have multiple domains
P that recognize specific features in other proteins, or in
the cytoskeleton or plasma membrane, and the result-
ing multivalency of individual modules allows their
Response assembly into a wide variety of multienzyme complexes.
One common theme in such interactions is the binding
FIGURE 12–1 Five features of signal-transducing systems. of one modular signaling protein to phosphorylated
residues in another protein; the resulting interaction
can be regulated by phosphorylation or dephosphoryla-
tion of the protein partner (Fig. 12–1c). Nonenzymatic
scaffold proteins with affinity for several enzymes
that interact in cascades bring those proteins together,
yields a quantitative measure of affinity (Kd) and the ensuring their interaction at specific cellular locations
number of ligand-binding sites in a receptor sample and at specific times.
(Box 12–1). The sensitivity of receptor systems is subject to
Cooperativity in receptor-ligand interactions modification. When a signal is present continuously,
results in large changes in receptor activation with small desensitization of the receptor system results (Fig.
changes in ligand concentration (recall the effect of 12–1d); when the stimulus falls below a certain thresh-
cooperativity on oxygen binding to hemoglobin; see old, the system again becomes sensitive. Think of what
Fig. 5–12). Amplification results when an enzyme happens to your visual transduction system when you
associated with a signal receptor is activated and, in turn, walk from bright sunlight into a darkened room or from
catalyzes the activation of many molecules of a second darkness into the light.
enzyme, each of which activates many molecules of a A final noteworthy feature of signal-transducing
third enzyme, and so on, in a so-called enzyme cascade systems is integration (Fig. 12–1e), the ability of the
(Fig. 12–1b). Such cascades can produce amplifications system to receive multiple signals and produce a unified
of several orders of magnitude within milliseconds. The response appropriate to the needs of the cell or organ-
response to a signal must also be terminated such that ism. Different signaling pathways converse with each
the downstream effects are in proportion to the strength other at several levels, generating complex cross talk
of the original stimulus. that maintains homeostasis in the cell and the organism.
12.1 General Features of Signal Transduction 435
BOX 12–1 METHODS Scatchard Analysis Quantifies the Receptor-Ligand Interaction

The cellular actions of a hormone begin when the
hormone (ligand, L) binds specifically and tightly to Total binding
its protein receptor (R) on or in the target cell. Bind-
Bound hormone, [RL]

ing is mediated by noncovalent interactions (hydrogen-
bonding, hydrophobic, and electrostatic) between the Specific binding
complementary surfaces of ligand and receptor.
Receptor-ligand interaction brings about a conforma-
tional change that alters the biological activity of the
receptor, which may be an enzyme, an enzyme regula-
tor, an ion channel, or a regulator of gene expression. Nonspecific binding
Receptor-ligand binding is described by the
(a) Total hormone added, [L] ! [RL]
equation
R " L ∆ RL
Receptor Ligand Receptor-ligand complex
This binding, like that of an enzyme to its substrate,
Bound hormone , [RL]

[L]
1
depends on the concentrations of the interacting Slope " #
Kd
components and can be described by an equilibrium
Free hormone
constant:
k11
R " L ∆ RL
Bmax
Receptor Ligand k21 Receptor-ligand complex
[RL] k11
Ka 5 5 5 1/Kd
[R][L] k21 (b) Bound hormone, [RL]
where Ka is the association constant and Kd is the dis- FIGURE 1 Scatchard analysis of a receptor-ligand interaction. A radio-
sociation constant. labeled ligand (L)—a hormone, for example—is added at several con-
Like enzyme-substrate binding, receptor-ligand centrations to a fixed amount of receptor (R), and the fraction of the
binding is saturable. As more ligand is added to a fixed hormone bound to receptor is determined by separating the receptor-
amount of receptor, an increasing fraction of receptor hormone complex (RL) from free hormone.
molecules is occupied by ligand (Fig. 1a). A rough (a) A plot of [RL] versus [L] " [RL] (total hormone added) is
measure of receptor-ligand affinity is given by the hyperbolic, rising toward a maximum for [RL] as the receptor sites
concentration of ligand needed to give half-saturation become saturated. To control for nonsaturable, nonspecific binding
of the receptor. Using Scatchard analysis of receptor- sites (eicosanoid hormones bind nonspecifically to the lipid bilayer,
ligand binding, we can estimate both the dissociation for example), a separate series of binding experiments is also neces-
constant Kd and the number of receptor-binding sites sary. A large excess of unlabeled hormone is added along with the
in a given preparation. When binding has reached dilute solution of labeled hormone. The unlabeled molecules compete
equilibrium, the total number of possible binding with the labeled molecules for specific binding to the saturable site on
sites, Bmax, equals the number of unoccupied sites, the receptor, but not for the nonspecific binding. The true value for
represented by [R], plus the number of occupied or specific binding is obtained by subtracting nonspecific binding from
ligand-bound sites, [RL]; that is, Bmax 5 [R] 1 [RL]. The total binding.
number of unbound sites can be expressed in terms of (b) A linear plot of [RL]/[L] versus [RL] gives Kd and Bmax for
total sites minus occupied sites: [R] 5 Bmax 2 [RL]. the receptor-hormone complex. Compare these plots with those
of V0 versus [S] and 1/V0 versus 1/[S] for an enzyme-substrate
The equilibrium expression can now be written
complex (see Fig. 6–12, Box 6–1).
[RL]
Ka 5
[L](Bmax 2 [RL])
[bound ligand] should give a straight line with a slope
Rearranging to obtain the ratio of receptor-bound
of 2Ka (–1/Kd) and an intercept on the abscissa of Bmax,
ligand to free (unbound) ligand, we get
the total number of binding sites (Fig. 1b). Hormone-
[Bound4 [RL]
5 5 Ka (Bmax 2 [RL 4) ligand interactions typically have Kd values of 10!9 to
[Free] [L] 10!11 M, corresponding to very tight binding.
1 Scatchard analysis is reliable for the simplest
5 (Bmax 2 [RL 4)
Kd cases, but as with Lineweaver-Burk plots for enzymes,
From this slope-intercept form of the equation, we can when the receptor is an allosteric protein, the plots
see that a plot of [bound ligand]/[free ligand] versus deviate from linearity.
c12Biosignaling.indd Page 436 28/09/12 10:49 AM user-F408 /Users/user-F408/Desktop
436 Biosignaling
result of a signaling pathway is the phosphorylation of

TABLE 12–1 Some Signals to Which Cells Respond a few specific target-cell proteins, which changes their
Antigens Light activities and thus the activities of the cell. Throughout
Cell surface our discussion we emphasize the conservation of fun-
Mechanical touch damental mechanisms for the transduction of biological
glycoproteins/
oligosaccharides Microbial, insect signals and the adaptation of these basic mechanisms
pathogens to a wide range of signaling pathways.
Developmental signals We consider the molecular details of several repre-
Extracellular matrix Neurotransmitters sentative signal-transduction systems, classified accord-
components Nutrients ing to the type of receptor. The trigger for each system
Growth factors Odorants is different, but the general features of signal transduc-
Hormones Pheromones tion are common to all: a signal interacts with a recep-
tor; the activated receptor interacts with cellular
Hypoxia Tastants machinery, producing a second signal or a change in the
activity of a cellular protein; the metabolic activity of
One of the revelations of research on signaling is the target cell undergoes a change; and finally, the
the remarkable degree to which signaling mechanisms transduction event ends. To illustrate these general
have been conserved during evolution. Although the features of signaling systems, we will look at examples
number of different biological signals (Table 12–1) is of six basic receptor types (Fig. 12–2).
probably in the thousands, and the kinds of response
elicited by these signals are comparably numerous, the 1. G protein–coupled receptors that indirectly
machinery for transducing all of these signals is built activate (through GTP-binding proteins, or G
from about 10 basic types of protein components. proteins) enzymes that generate intracellular
In this chapter we examine some examples of the second messengers. This type of receptor is
major classes of signaling mechanisms, looking at how illustrated by the !-adrenergic receptor system that
they are integrated in specific biological functions such detects epinephrine (adrenaline) (Section 12.2).
as the transmission of nerve signals; responses to hor- 2. Receptor tyrosine kinases, plasma membrane
mones and growth factors; the senses of sight, smell, receptors that are also enzymes. When one of
and taste; and control of the cell cycle. Often, the end these receptors is activated by its extracellular
1. G protein–coupled receptor 2a. Receptor tyrosine kinase 3. Receptor guanylyl 4. Gated ion channel 5. Adhesion receptor (integrin)
External ligand (L) binding Ligand binding activates cyclase Opens or closes Binds molecules in extracellular
to receptor (R) activates an tyrosine kinase activity Ligand binding to in response to matrix, changes conformation,
intracellular GTP- binding by autophosphorylation. extracellular domain concentration thus altering its interaction
protein (G), which regulates stimulates formation of signal ligand with cytoskeleton.
an enzyme (Enz) that of second-messenger or membrane
generates an intracellular cyclic GMP. potential.
second messenger (X).
L L
L Ion L
L L
L
L
Enz L
R R
G G
Plasma X
membrane GTP cGMP
2b. Kinase activates Kinase
transcription cascade
factor, altering
gene expression.
Nuclear
envelope
6. Nuclear receptor
Protein Protein
Hormone binding allows
mRNA the receptor to regulate L mRNA
the expression of
DNA specific genes. DNA
FIGURE 12–2 Six general types of signal transducers.

12.2 G Protein–Coupled Receptors and Second Messengers 437
ligand, it catalyzes the phosphorylation of several SUMMARY 12. 1 General Features of Signal Transduction
!
cytosolic or plasma membrane proteins. The All cells have specific and highly sensitive signal-
insulin receptor is one example (Section 12.3); transducing mechanisms, which have been
the receptor for epidermal growth factor (EGFR) conserved during evolution.
is another.
! A wide variety of stimuli act through specific
3. Receptor guanylyl cyclases, which are also plasma protein receptors in the plasma membrane.
membrane receptors with an enzymatic cytoplasmic
! The receptors bind the signal molecule and initiate
domain. The intracellular second messenger for
a process that amplifies the signal, integrates it
these receptors, cyclic guanosine monophosphate
with input from other receptors, and transmits the
(cGMP), activates a cytosolic protein kinase that
information throughout the cell. If the signal
phosphorylates cellular proteins and thereby
persists, receptor desensitization reduces or ends
changes their activities (Section 12.4).
the response.
!
4. Gated ion channels of the plasma membrane that Multicellular organisms have six general types of
open and close (hence the term “gated”) in signaling mechanisms: plasma membrane proteins
response to the binding of chemical ligands or that act through G proteins, receptor tyrosine
changes in transmembrane potential. These are kinases, receptor guanylyl cyclases that act
the simplest signal transducers. The acetylcholine through a protein kinase, gated ion channels,
receptor ion channel is an example of this adhesion receptors that carry information between
mechanism (Section 12.6). the extracellular matrix and the cytoskeleton, and
5. Adhesion receptors that interact with nuclear receptors that bind steroids and alter gene
macromolecular components of the extracellular expression.
matrix (such as collagen) and convey
instructions to the cytoskeletal system about cell
migration or adherence to the matrix. Integrins 12.2 G Protein–Coupled Receptors
illustrate this general type of transduction
mechanism (Section 12.7).
and Second Messengers
As their name implies, G protein–coupled receptors
6. Nuclear receptors that bind specific ligands (such
(GPCRs) are receptors that are closely associated with
as the hormone estrogen) and alter the rate at
a member of the guanosine nucleotide–binding pro-
which specific genes are transcribed and
tein (G protein) family. Three essential components
translated into cellular proteins. Because steroid
define signal transduction through GPCRs: a plasma
hormones function through mechanisms intimately
membrane receptor with seven transmembrane helical
related to the regulation of gene expression, we
segments, a G protein that cycles between active (GTP-
consider them here only briefly (Section 12.8) and
bound) and inactive (GDP-bound) forms, and an effec-
defer a detailed discussion of their action until
tor enzyme (or ion channel) in the plasma membrane
Chapter 28.
that is regulated by the activated G protein. The G protein,
As we begin this discussion of biological signaling, stimulated by the activated receptor, exchanges bound
a word about the nomenclature of signaling proteins is GDP for GTP, then dissociates from the occupied recep-
in order. These proteins are typically discovered in one tor and binds to the nearby effector enzyme, altering its
context and named accordingly, then prove to be activity. The activated enzyme then generates a second
involved in a broader range of biological functions for messenger that affects downstream targets. The
which the original name is not helpful. For example, the human genome encodes about 350 GPCRs for detecting
retinoblastoma protein, pRb, was initially identified as hormones, growth factors, and other endogenous
the site of a mutation that contributes to cancer of the ligands, and perhaps 500 that serve as olfactory (smell)
retina (retinoblastoma), but it is now known to function and gustatory (taste) receptors.
in many pathways essential to cell division in all cells, GPCRs have been implicated in many common
not just those of the retina. Some genes and proteins are human diseases, including allergies, depression,
given noncommittal names: the tumor suppressor pro- blindness, diabetes, and various cardiovascular defects
tein p53, for example, is a protein of 53 kDa, but its with serious health consequences. Close to half of all
name gives no clue to its great importance in the regula- drugs on the market target one GPCR or another. For
tion of cell division and the development of cancer. In example, the !-adrenergic receptor, which mediates the
this chapter we generally define these protein names as effects of epinephrine, is the target of the “beta blockers,”
we encounter them, introducing the names commonly prescribed for such diverse conditions as hypertension,
used by researchers in the field. Don’t be discouraged if cardiac arrhythmia, glaucoma, anxiety, and migraine
you can’t get them all straight the first time you encoun- headache. At least 150 of the GPCRs found in the human
ter them! genome are still “orphan receptors”: their natural ligands
438 Biosignaling
are not yet identified, and so we know nothing about the increased breakdown of glycogen and fat. Adrenergic
their biology. The !-adrenergic receptor, with well- receptors of the !1 and !2 subtypes act through the
understood biology and pharmacology, is the prototype same mechanism, so in our discussion, “!-adrenergic”
for all GPCRs, and our discussion of signal-transducing applies to both types.
systems begins there. ■ Like all GPCRs, the !-adrenergic receptor is an
integral protein with seven hydrophobic, helical regions
of 20 to 28 amino acid residues that span the plasma
The !-Adrenergic Receptor System Acts through the membrane seven times, thus the alternative name for
Second Messenger cAMP GPCRs: heptahelical receptors. The binding of epi-
nephrine to a site on the receptor deep within the
Epinephrine sounds the alarm when some threat requires
plasma membrane (Fig. 12–4a, step 1) promotes a
the organism to mobilize its energy-generating machin-
conformational change in the receptor’s intracellular
ery; it signals the need to fight or flee. Epinephrine
domain that affects its interaction with an associated G
action begins when the hormone binds to a protein
protein, promoting the dissociation of GDP and the
receptor in the plasma membrane of an epinephrine-
binding of GTP (step 2). For all GPCRs, the G protein
sensitive cell. Adrenergic receptors (“adrenergic”
is heterotrimeric, composed of three different subunits:
reflects the alternative name for epinephrine, adrena-
", !, and #. Such G proteins are therefore known as
line) are of four general types, "1, "2, !1, and !2, defined
trimeric G proteins. In this case, it is the " subunit
by differences in their affinities and responses to a group
that binds GDP or GTP and transmits the signal from
of agonists and antagonists. Agonists are structural
the activated receptor to the effector protein. Because
analogs that bind to a receptor and mimic the effects of
this G protein activates its effector, it is referred to as a
its natural ligand; antagonists are analogs that bind the
stimulatory G protein, or Gs. Like other G proteins
receptor without triggering the normal effect and there-
(Box 12–2), Gs functions as a biological “switch”: when
by block the effects of agonists, including the biological
the nucleotide-binding site of Gs (on the " subunit) is
ligand. In some cases, the affinity of the synthetic ago-
occupied by GTP, Gs is turned on and can activate its
nist or antagonist for the receptor is greater than that of
effector protein (adenylyl cyclase in the present case);
the natural agonist (Fig. 12–3). The four types of
with GDP bound to the site, Gs is switched off. In the
adrenergic receptors are found in different target tissues
active form, the ! and # subunits of Gs dissociate from
and mediate different responses to epinephrine. Here we
the " subunit as a !# dimer, and Gs", with its bound
focus on the !-adrenergic receptors of muscle, liver,
GTP, moves in the plane of the membrane from the
and adipose tissue. These receptors mediate changes in
receptor to a nearby molecule of adenylyl cyclase (step
fuel metabolism, as described in Chapter 23, including
3). Gs" is held to the membrane by a covalently
attached palmitoyl group (see Fig. 11–15).
Adenylyl cyclase is an integral protein of the
Kd (!M) plasma membrane, with its active site on the cytoplas-
OH
mic face. The association of active Gs" with adenylyl
HO CH CH2 NH CH3 5 cyclase stimulates the cyclase to catalyze cAMP synthe-
sis from ATP (Fig. 12–4a, step 4, and Fig. 12–4b), rais-
HO Epinephrine ing the cytosolic [cAMP]. The interaction between Gs"
OH CH3
and adenylyl cyclase is possible only when Gs" is bound
to GTP.
HO CH CH2 NH CH 0.4 The stimulation by Gs" is self-limiting; Gs" has
Isoproterenol CH3 intrinsic GTPase activity that inactivates Gs" by con-
HO (agonist) verting its bound GTP to GDP (Fig. 12–5). The now
OH CH3 inactive Gs" dissociates from adenylyl cyclase, render-
ing the cyclase inactive. Gs" reassociates with the !#
O CH2 CH CH2 NH CH 0.0046 dimer (Gs!#), and inactive Gs is again available to inter-
CH3 act with a hormone-bound receptor.
Propranolol The role of Gs" in serving as a biological “switch”
(antagonist) protein is not unique. A variety of G proteins act as
FIGURE 12–3 Epinephrine and its synthetic analogs. Epinephrine, also binary switches in signaling systems with GPCRs and in
called adrenaline, is released from the adrenal gland and regulates energy- many processes that involve membrane fusion or fission
yielding metabolism in muscle, liver, and adipose tissue. It also serves as (Box 12–2). Trimeric G Proteins: Molecular On/Off Switches
a neurotransmitter in adrenergic neurons. Its affinity for its receptor is Epinephrine exerts its downstream effects through
expressed as a dissociation constant for the receptor-ligand complex. the increase in [cAMP] that results from the activation
Isoproterenol and propranolol are synthetic analogs, one an agonist with of adenylyl cyclase. Cyclic AMP, in turn, allosterically
an affinity for the receptor that is higher than that of epinephrine, and activates cAMP-dependent protein kinase, also
the other an antagonist with extremely high affinity. called protein kinase A or PKA (Fig. 12–4a, step 5),
12.2 G Protein–Coupled Receptorsand Second Messengers 439
(a)
1 Epinephrine
binds to its
specific b-adrenergic
receptor. receptor
N Outside
Adenylyl
cyclase
C Inside
GDP GTP GTP ATP

Gsb
Gsa Gsa
Gsg
GDP
2 Hormone-receptor 3 Activated Gsa 4 Adenylyl 5 cAMP 6 Phosphorylation

complex causes separates from cyclase activates of cellular
the GDP bound to Gsbg, moves to catalyzes the cAMP PKA. proteins by PKA
Gsa to be replaced adenylyl cyclase, formation causes the
by GTP, activating and activates it. of cAMP. cellular response
Gsa. Many Gsa cyclic nucleotide to epinephrine.
subunits may be phosphodiesterase
activated by one
occupied receptor.
5$-AMP 7 cAMP is degraded,
reversing the
activation of PKA.
NH2 NH2 NH2

N N N
N N N
O O O PPi N H2O O
N N N N N
5" 5"
!
O P O P O P O CH2 O CH2 !
O P O CH2
O adenylyl O cyclic O
O! O! O! cyclase nucleotide O!
H H H H phosphodiesterase H H
H H H H H H
3" 3"
OH OH O P O OH OH OH
ATP O! Adenosine 5"-monophosphate (AMP)
Adenosine 3",5"-cyclic
monophosphate
(b) (cAMP)
FIGURE 12–4 Transduction of the epinephrine signal: the !-adrenergic under the influence of different hormones. Hormones that induce GTP
pathway. (a) The mechanism that couples binding of epinephrine to its binding to Gi cause inhibition of adenylyl cyclase, resulting in lower cellu-
receptor with activation of adenylyl cyclase; the seven steps are dis- lar [cAMP]. (b) The combined action of the enzymes that catalyze
cussed further in the text. The same adenylyl cyclase molecule in the steps 4 and 7, synthesis and hydrolysis of cAMP by adenylyl cyclase
plasma membrane may be regulated by a stimulatory G protein (GS), and cAMP phosphodiesterase, respectively.
as shown, or an inhibitory G protein (Gi, not shown). GS and Gi are
which catalyzes the phosphorylation of specific Ser or The inactive form of PKA contains two identical
Thr residues of targeted proteins, including glycogen catalytic subunits (C) and two identical regulatory sub-
phosphorylase b kinase. This enzyme is active when units (R) (Fig. 12–6a). The tetrameric R2C2 complex is
phosphorylated and can begin the process of mobilizing catalytically inactive, because an autoinhibitory domain
glycogen stores in muscle and liver in anticipation of the of each R subunit occupies the substrate-binding cleft
need for energy, as signaled by epinephrine. of each C subunit. When cAMP binds to the R subunits,
440 Biosignaling
Gs with GDP Contact of Gs Gs with GTP

FIGURE 12–5 The GTPase switch. G proteins cycle between GDP-bound
1 2 3
bound is with hormone- bound dissociates (off) and GTP-bound (on). The protein’s intrinsic GTPase activity, in many
turned off; receptor into a and bg cases stimulated by RGS proteins (regulators of G-protein signaling; see
it cannot complex causes subunits. Gsa-GTP Box 12-2), determines how quickly bound GTP is hydrolyzed to GDP and
activate displacement of is turned on; it can
adenylyl bound GDP by activate thus how long the G protein remains active.
cyclase. GTP. adenylyl cyclase.
GPCR
they undergo a conformational change that moves the
autoinhibitory domain of R out of the catalytic domain
Adenylyl of C, and the R2C2 complex dissociates to yield two
cyclase free, catalytically active C subunits. This same basic
Gs
GTP
mechanism—displacement of an autoinhibitory domain—
g b GDP g b GDP GTP mediates the allosteric activation of many types of
a a a protein kinases by their second messengers (as in
Figs 12–14 and 12–22, for example). The structure of
GDP the substrate-binding cleft in PKA is the prototype for
all known protein kinases (Fig. 12–6b); certain residues
in this cleft region have identical counterparts in all of
the more than 1,000 known protein kinases. The ATP-
GDP Pi
binding site of each catalytic subunit positions ATP
a
perfectly for the transfer of its terminal (#) phosphoryl
group to the —OH in the side chain of a Ser or Thr
4 GTP bound to Gsa is hydrolyzed by the protein’s intrinsic GTPase; residue.
Gsa thereby turns itself off. The inactive a subunit reassociates
with the bg subunit.
(a) (b) Substrate-

AKAP
Dimerization Inhibitor sequence binding cleft
Substrate- domain Catalytic
Small lobe
binding cleft (ATP binding) (C) subunit
Inhibitor
Catalytic PKA sequence
subunit C C
C
R R
Regulatory
subunit Large lobe
4 cAMP 4 cAMP (protein
substrate
binding)
Regulatory
(R) subunit
Substrate-binding
cleft now available
2 cAMP 2 cAMP
FIGURE 12–6 Activation of cAMP-dependent protein kinase (PKA). pulls its inhibitory sequence away from the C subunit, opening up the
(a) When [cAMP] is low, the two identical regulatory subunits (R; red) substrate-binding cleft and releasing each C subunit in its catalytically
associate with the two identical catalytic subunits (C). In this R2C2 com- active form. (b) A crystal structure showing part of the R2C2 complex
plex, the inhibitor sequences of the R subunits lie in the substrate-bind- (PDB ID 1U7E)—one C subunit and part of one R subunit. The amino-
ing cleft of the C subunits and prevent binding of protein substrates; the terminal dimerization region of the R subunit is omitted for simplicity.
complex is therefore catalytically inactive. The amino-terminal sequences The small lobe of C contains the ATP-binding site, and the large lobe
of the R subunits interact to form an R2 dimer, the site of binding to an A surrounds and defines the cleft where the protein substrate binds and
kinase anchoring protein (AKAP), described later in the text. When undergoes phosphorylation at a Ser or Thr residue, with the phosphoryl
[cAMP] rises in response to a hormonal signal, each R subunit binds group transferred from ATP. In this inactive form, the inhibitor sequence
two cAMP molecules and undergoes a dramatic reorganization that of R blocks the substrate-binding cleft of C, inactivating it.
BOX 12–2 MEDICINE G Proteins: Binary Switches in Health and Disease

Alfred G. Gilman and Martin Rodbell (Fig. 1) discovered
the critical roles of guanosine nucleotide–binding pro-
teins (G proteins) in a wide variety of cellular processes,
GTP analog
including sensory perception, signaling for cell division,
P loop
growth and differentiation, intracellular movements of
proteins and membrane vesicles, and protein synthesis. Switch I
The human genome encodes nearly 200 of these pro-
Switch II
teins, which differ in size and subunit structure, intra-
cellular location, and function. But all G proteins share Gly60
Mg2+
a common feature: they can become activated and then, Ala146
after a brief period, can inactivate themselves, thereby
serving as molecular binary switches with built-in timers.
This superfamily of proteins includes the trimeric G
proteins involved in adrenergic signaling (Gs and Gi) Thr35
and vision (transducin); small G proteins such as that
involved in insulin signaling (Ras) and others that func-
tion in vesicle trafficking (ARF and Rab), transport
into and out of the nucleus (Ran; see Fig. 27–42), and
timing of the cell cycle (Rho); and several proteins
involved in protein synthesis (initiation factor IF2 and
elongation factors EF-Tu and EF-G; see Chapter 26).
Many G proteins have covalently bound lipids, which
FIGURE 2 The Ras protein, the prototype for all G proteins (PDB ID
give them an affinity for membranes and dictate their 5P21). Mg2!-GTP is held by critical residues in the phosphate-binding
locations in the cell. P loop (blue), and by Thr35 in the switch I (red) and Gly60 in the
All G proteins have the same core structure and switch II (green) regions. Ala146 gives specificity for GTP over ATP. In
use the same mechanism for switching between an this structure, the nonhydrolyzable GTP analog Gpp(NH)p is in the
inactive conformation, favored when GDP is bound, GTP-binding site.
and an active conformation, favored when GTP is
bound. We can use the Ras protein (,20 kDa), a region called the P loop (phosphate-binding; Fig. 3).
minimal signaling unit, as a prototype for all members In Ras, the ! phosphate of GTP binds to a Lys residue
of this superfamily (Fig. 2). in the P loop and to two critical residues, Thr35 in
In the GTP-bound conformation, the G protein switch I and Gly60 in switch II, that hydrogen-bond
exposes previously buried regions (called switch I with the oxygens of the ! phosphate of GTP. These
and switch II) that interact with proteins down-
stream in the signaling pathway, until the G protein (Continued on next page)
inactivates itself by hydrolyzing its bound GTP to
P loop
GDP. The critical determinant of G-protein conforma- !
O
tion is the ! phosphate of GTP, which interacts with a
O
!
O P O
A
B
Thr35 Gly60
O
O O
!
O P O
A
B
Switch Switch
I O II
O O O
!
O P O
A
B
Rib
O
Guanine
FIGURE 3 When bound GTP is hydrolyzed by the GTPase activities of

FIGURE 1 Alfred G. Gilman (left) and Martin Rodbell (1925–1998). Ras and its GAP, loss of hydrogen bonds to Thr35 and Gly60 allows the
Their Nobel lectures on the discovery and exploration of G proteins switch I and switch II regions to relax into a conformation in which they
are available at www.nobelprize.org. are no longer available to interact with downstream targets such as Raf.
442 Biosignaling
BOX 12–2 MEDICINE G Proteins: Binary Switches in Health and Disease (Continued)
hydrogen bonds act like a pair of springs holding the GTP-GDP exchange
protein in its active conformation. When GTP is factors (GEFs)
cleaved to GDP and Pi is released, these hydrogen Rh*,
bonds are lost; the protein relaxes into its inactive b-AR*,
conformation, burying the sites that interact with Sos, GTP
G protein etc. G protein
other partners in its active state. Ala146 hydrogen- (inactive) (active)
bonds to the guanine oxygen, allowing GTP, but not GDP GTP
ATP, to bind. GDP
The intrinsic GTPase activity of G proteins is
increased up to 105-fold by GTPase activator pro-
Pi
teins (GAPs), also called, in the case of heterotri- cGMP PDE,
meric G proteins, regulators of G protein signal- GAP, AC, Raf, etc.
ing (RGSs; Fig. 4). GAPs (and RGSs) thus determine RGS, etc. Downstream
how long the switch remains “on.” They contribute a effector
Modulators of
enzymes
critical Arg residue that reaches into the G-protein GTPase activity
GTPase active site and assists in catalysis. The intrin- FIGURE 4 Factors that regulate the activity of G proteins (green). Inactive
sically slow process of replacing bound GDP with G proteins (both small G proteins such as Ras and heterotrimeric G pro-
GTP, switching the protein on, is catalyzed by guano- teins such as Gs) interact with upstream GTP-GDP exchange factors, GEFs
sine nucleotide–exchange factors (GEFs) associ- (red; often these are activated (*) receptors such as rhodopsin, !-adrenergic
ated with the G protein (Fig. 4). receptors, and Sos), and are activated by GTP binding. In this form, the G
Because G proteins play crucial roles in so many proteins activate downstream effector enzymes (blue; enzymes such as
signaling processes, it is not surprising that defects in cGMP phosphodiesterase, adenylyl cyclase, and Raf). GTPase activator
G proteins lead to a variety of diseases. In about 25% proteins (GAPs, in the case of small G proteins) and regulators of G protein
of all human cancers (and in a much higher propor- signaling (RGSs) (yellow), by modulating the GTPase activity of G proteins,
tion of certain types of cancer), there is a mutation in determine how long the G protein will remain active.
a Ras protein—typically in one of the critical residues
around the GTP-binding site or in the P loop—that
virtually eliminates its GTPase activity. Once activated such mutations are found in about 40% of pituitary
by GTP binding, this Ras protein remains constitu- tumors (adenomas). Individuals with “inactivating”
tively active, promoting cell division in cells that mutations in G" are unresponsive to hormones (such as
should not divide. The tumor suppressor gene NF1 thyroid hormone) that act through cAMP. Mutation in
encodes a GAP that enhances the GTPase activity of the gene for the transducin " subunit (T"), which is
normal Ras. Mutations in NF1 that result in a non- involved in visual signaling, leads to a type of night
functioning GAP leave Ras with only its intrinsic blindness, apparently due to defective interaction
GTPase activity, which is very weak (has a very low between the activated T" subunit and the phosphodies-
turnover number); once activated by GTP binding, terase of the rod outer segment (see Fig. 12–39). A
Ras stays active for an extended period, continuing to sequence variation in the gene encoding the ! subunit
send the signal: divide. of a heterotrimeric G protein is commonly found in indi-
Defective heterotrimeric G proteins can also lead viduals with hypertension (high blood pressure), and
to disease. Mutations in the gene that encodes the " this variant gene is suspected of involvement in obesity
subunit of Gs (which mediates changes in [cAMP] in and atherosclerosis.
response to hormonal stimuli) may result in a G" that The pathogenic bacteria that cause cholera and
is permanently active or permanently inactive. “Acti- pertussis (whooping cough) produce toxins that target
vating” mutations generally occur in residues crucial to G proteins, interfering with normal signaling in host
GTPase activity; they lead to a continuously elevated cells. Cholera toxin, secreted by Vibrio cholerae in
[cAMP], with significant downstream consequences, the intestine of an infected person, is a heterodimeric
including undesirable cell proliferation. For example, protein. Subunit B recognizes and binds to specific
As indicated in Figure 12–4a (step 6), PKA regu- undergoes phosphorylation, a sequence that marks
lates several enzymes downstream in the signaling them for regulation by PKA. The substrate-binding
pathway (Table 12–2). Although these downstream cleft of PKA recognizes these sequences and phos-
targets have diverse functions, they share a region of phorylates their Thr or Ser residue. Comparison of the
sequence similarity around the Ser or Thr residue that sequences of various protein substrates for PKA has
C NH2 O O
Gs " CH2 O P O P O Rib Adenine
g b Arg NH2 " N O
a O! O!
H H
H H
Normal Gs: GTPase activity HO OH

terminates the signal from NAD"
receptor to adenylyl cyclase.
O
cholera C NH2
toxin
O O
Gs
CH2 O P O P O Rib Adenine
g b
Arg NH O
a O! O!
H H
H H
ADP-ribosylated Gs: GTPase HO OH

activity is inactivated; Gs
ADP-ribose
constantly activates adenylyl
cyclase.
FIGURE 5 The bacterial toxins that cause cholera and whooping cough fied fail to respond to normal hormonal stimuli. The pathology of both
(pertussis) are enzymes that catalyze transfer of the ADP-ribose moiety diseases results from defective regulation of adenylyl cyclase and
of NAD" to an Arg residue of Gs (in the case of cholera toxin, as shown overproduction of cAMP.
here) or a Cys residue of Gi (pertussis toxin). The G proteins thus modi-
gangliosides on the surface of intestinal epithelial cells ensuing osmotic imbalance. Severe dehydration and
and provides a route for subunit A to enter these cells. electrolyte loss are the major pathologies in cholera.
After entry, subunit A is broken into two pieces: the A1 These can be fatal in the absence of prompt rehydra-
fragment and the A2 fragment. A1 then associates with tion therapy.
the ADP-ribosylation factor ARF6, a small G protein in The pertussis toxin, produced by Bordetella
host cells, through residues in its switch I and switch II pertussis, catalyzes ADP-ribosylation of the " subunit
regions—which are accessible only when ARF6 is in its of Gi, in this case preventing GDP-GTP exchange and
active (GTP-bound) form. This association with ARF6 blocking inhibition of adenylyl cyclase by Gi. The bac-
activates A1, which catalyzes the transfer of ADP-ribose terium infects the respiratory tract, where it destroys
from NAD" to the critical Arg residue in the P loop of the ciliated epithelial cells that normally sweep away
the " subunit of Gs (Fig. 5). ADP-ribosylation blocks the mucus. Without this ciliary action, vigorous coughing
GTPase activity of Gs and thereby renders Gs perma- is needed to clear the tract; this is the gasping cough
nently active. This results in continuous activation of that gives the disease its name (and spreads the bac-
the adenylyl cyclase of intestinal epithelial cells, chroni- terium to others). How the defect in G-protein signal-
cally high [cAMP], and chronically active PKA. PKA ing kills ciliated epithelial cells is not yet clear.
phosphorylates the CFTR Cl! channel (see Box 11–2) Given the large number of G protein–coupled
and a Na"-H" exchanger in the intestinal epithelial receptors in the human genome, it seems likely that
cells. The resultant efflux of NaCl triggers massive future studies will reveal many more examples of how
water loss through the intestine as cells respond to the defective G-protein signaling affects human health.
yielded the consensus sequence—the neighboring the original hormone signal (Fig. 12–7). First, the
residues needed to mark a Ser or Thr residue for binding of one hormone molecule to one receptor mol-
phosphorylation (see Table 12–2). ecule catalytically activates many Gs molecules that
As in many signaling pathways, signal transduction associate with the activated receptor, one after the
by adenylyl cyclase entails several steps that amplify other. Next, by activating one molecule of adenylyl
444 Biosignaling
TABLE 12–2 Some Enzymes and Other Proteins Regulated by cAMP-Dependent Phosphorylation (by PKA)
Enzyme/protein Sequence phosphorylated* Pathway/process regulated
Glycogen synthase RASCTSSS Glycogen synthesis
Phosphorylase b kinase
¯˘˙
" subunit VEFRRLSI
Glycogen breakdown
! subunit RTKRSGSV
Pyruvate kinase (rat liver) GVLRRASVAZL Glycolysis
Pyruvate dehydrogenase complex (type L) GYLRRASV Pyruvate to acetyl-CoA
Hormone-sensitive lipase PMRRSV Triacylglycerol mobilization
and fatty acid oxidation
Phosphofructokinase-2/fructose 2,6-bisphosphatase LQRRRGSSIPQ Glycolysis/gluconeogenesis
Tyrosine hydroxylase FIGRRQSL Synthesis of L-dopa,
dopamine, norepinephrine,
and epinephrine
Histone H1 AKRKASGPPVS DNA condensation
Histone H2B KKAKASRKESYSVYVYK DNA condensation
Cardiac phospholamban (cardiac pump regulator) AIRRAST Intracellular [Ca2"]
Protein phosphatase-1 inhibitor-1 IRRRRPTP Protein dephosphorylation
PKA consensus sequence† xR[RK]x[ST]B Many
*The phosphorylated S or T residue is shown in red. All residues are given as their one-letter abbreviations (see Table 3–1).
†
x is any amino acid; B is any hydrophobic amino acid. See Box 3–2 for conventions used in displaying consensus sequences.
Epinephrine 1 molecule
cyclase, each active Gs" molecule stimulates the cata-
Epinephrine- lytic synthesis of many molecules of cAMP. The sec-
1 molecule
receptor complex
ond messenger cAMP now activates PKA, each mole-
cule of which catalyzes the phosphorylation of many
Hepatocyte GSa 10 molecules molecules of the target protein—phosphorylase b
kinase in Figure 12–7. This kinase activates glycogen
phosphorylase b, which leads to the rapid mobilization
ATP Cyclic AMP of glucose from glycogen. The net effect of the cascade
adenylyl is amplification of the hormonal signal by several orders
cyclase 200 molecules
of magnitude, which accounts for the very low concen-
tration of epinephrine (or any other hormone) required
Inactive PKA Active PKA for hormone activity.
100 molecules
Several Mechanisms Cause Termination of

Inactive Active
phosphorylase b phosphorylase b the !-Adrenergic Response
kinase kinase
To be useful, a signal-transducing system has to turn
1,000 molecules off after the hormonal or other stimulus has ended, and
mechanisms for shutting off the signal are intrinsic to
Inactive glycogen Active glycogen
phosphorylase b phosphorylase a
10,000 molecules
FIGURE 12–7 Epinephrine cascade. Epinephrine triggers a series of reac-
tions in hepatocytes in which catalysts activate catalysts, resulting in great
Glycogen Glucose 1-phosphate amplification of the original hormone signal. The numbers of molecules
100,000 molecules shown are simply to illustrate amplification and are almost certainly gross
underestimates. Binding of one molecule of epinephrine to one !-adrenergic
receptor on the cell surface activates a number (possibly hundreds) of G
Glucose proteins, one after another, each of which goes on to activate a molecule
of the enzyme adenylyl cyclase. Adenylyl cyclase acts catalytically, produc-
ing many molecules of cAMP for each activated adenylyl cyclase. (Because
Blood glucose
two molecules of cAMP are required to activate one PKA catalytic subunit,
100,000 molecules this step does not amplify the signal.)
all signaling systems. Most systems also adapt to the Finally, at the end of the signaling pathway, the
continued presence of the signal by becoming less sen- metabolic effects that result from enzyme phosphoryla-
sitive to it, in the process of desensitization. The tion are reversed by the action of phosphoprotein phos-
!-adrenergic system illustrates both. When the concen- phatases, which hydrolyze phosphorylated Ser, Thr, or
tration of epinephrine in the blood drops below the Kd Tyr residues, releasing inorganic phosphate (Pi). About
for its receptor, the hormone dissociates from the 150 genes in the human genome encode phosphopro-
receptor and the latter reassumes the inactive confor- tein phosphatases, fewer than the number encoding
mation, in which it can no longer activate Gs. protein kinases (,500). Some of these phosphatases
A second means of ending the response to are known to be regulated; others may act constitu-
!-adrenergic stimulation is the hydrolysis of GTP bound tively. When [cAMP] drops and PKA returns to its inac-
to the G" subunit, catalyzed by the intrinsic GTPase tive form (step 7 in Fig. 12–4a), the balance between
activity of the G protein. Conversion of bound GTP to phosphorylation and dephosphorylation is tipped
GDP favors the return of G" to the conformation in toward dephosphorylation by these phosphatases.
which it binds the G!# subunits—the conformation in
which the G protein is unable to interact with or stimu-
late adenylyl cyclase. This ends the production of cAMP.
The !-Adrenergic Receptor Is Desensitized by
The rate of inactivation of Gs depends on the GTPase Phosphorylation and by Association with Arrestin
activity, which for G" alone is very feeble. However, The mechanisms for signal termination described above
GTPase activator proteins (GAPs) strongly stimulate take effect when the stimulus ends. A different mecha-
this GTPase activity, causing more rapid inactivation of nism, desensitization, damps the response even while
the G protein (see Box 12–2). GAPs can themselves be the signal persists. Desensitization of the !-adrenergic
regulated by other factors, providing a fine-tuning of the receptor is mediated by a protein kinase that phos-
response to !-adrenergic stimulation. A third mecha- phorylates the receptor on the intracellular domain
nism for terminating the response is to remove the sec- that normally interacts with Gs (Fig. 12–8). When
ond messenger: hydrolysis of cAMP to 5#-AMP (not the receptor is occupied by epinephrine, !-adrenergic
active as a second messenger) by cyclic nucleotide receptor kinase, or !ARK (also commonly called
phosphodiesterase (Fig. 12–4a, step 7; 12–4b). GRK2; see below), phosphorylates several Ser residues
1 Binding of epinephrine (E) to 2 Gsbg recruits bARK to the mem-

b-adrenergic receptor triggers brane, where it phosphorylates
dissociation of Gsbg from Gsa Ser residues at the carboxyl
(not shown). terminus of the receptor.
Plasma
membrane
E E
Gs P P bARK Gs
b b
3 b-Arrestin (barr)
g g
binds to the
P
phosphorylated
barr P
carboxyl-terminal
E domain of the
receptor.
P
P
P 4 Receptor-arrestin
P complex enters the
cell by endocytosis.
5 In endocytic vesicle,
arrestin dissociates;
receptor is FIGURE 12–8 Desensitization of the !-adrenergic receptor in the contin-
dephosphorylated ued presence of epinephrine. This process is mediated by two proteins:
and returned to cell
surface. !-adrenergic protein kinase (!ARK) and !-arrestin (!arr; also known as
arrestin 2).
446 Biosignaling
near the carboxyl terminus of the receptor, which is on

the cytoplasmic side of the plasma membrane. Usually TABLE 12–3 Some Signals That Use cAMP as
located in the cytosol, !ARK is drawn to the plasma Second Messenger
membrane by its association with the Gs!# subunits and Corticotropin (ACTH)
is thus positioned to phosphorylate the receptor.
Corticotropin-releasing hormone (CRH)
Receptor phosphorylation creates a binding site for the
protein !-arrestin, or !arr (also called arrestin 2), Dopamine [D1, D2]
and binding of !-arrestin effectively prevents further Epinephrine (!-adrenergic)
interaction between the receptor and the G protein. Follicle-stimulating hormone (FSH)
The binding of !-arrestin also facilitates receptor
Glucagon
sequestration, the removal of receptor molecules from
the plasma membrane by endocytosis into small intra- Histamine [H2]
cellular vesicles. The arrestin-receptor complex recruits Luteinizing hormone (LH)
two proteins involved in vesicle formation (see Fig. Melanocyte-stimulating hormone (MSH)
27–45), the AP-2 complex and clathrin, which initiate
Odorants (many)
membrane invagination, leading to the formation of
endosomes containing the adrenergic receptor. In this Parathyroid hormone
state the receptors are inaccessible to epinephrine and Prostaglandins E1, E2 (PGE1, PGE2)
therefore inactive. Receptors in the endocytic vesicles Serotonin [5-HT-1a, 5-HT-2]
are eventually dephosphorylated and returned to the
Somatostatin
plasma membrane, completing the circuit and resensi-
tizing the system to epinephrine. !-Adrenergic recep- Tastants (sweet, bitter)
tor kinase is a member of a family of G protein–cou- Thyroid-stimulating hormone (TSH)
pled receptor kinases (GRKs), all of which
Note: Receptor subtypes in square brackets. Subtypes may have different transduction
phosphorylate GPCRs on their carboxyl-terminal cyto- mechanisms. For example, serotonin is detected in some tissues by receptor subtypes 5-HT-1a
plasmic domains and play roles similar to that of !ARK and 5-HT-1b, which act through adenylyl cyclase and cAMP, and in other tissues by receptor
in desensitization and resensitization of their receptors. subtype 5-HT-1c, acting through the phospholipase C–IP3 mechanism (see Table 12–4).
At least five different GRKs and four different arrestins
are encoded in the human genome; each GRK is capable phorylation. For example, the binding of somatostatin
of desensitizing a particular subset of GPCRs, and each to its receptor leads to activation of an inhibitory G
arrestin can interact with many different types of phos- protein, or Gi, structurally homologous to Gs, that
phorylated receptors. inhibits adenylyl cyclase and lowers [cAMP]. Soma-
tostatin therefore counterbalances the effects of gluca-
gon. In adipose tissue, prostaglandin E1 (PGE1; see
Cyclic AMP Acts as a Second Messenger for Many Fig. 10–18) inhibits adenylyl cyclase, thus lowering
Regulatory Molecules [cAMP] and slowing the mobilization of lipid reserves
Epinephrine is just one of many hormones, growth fac- triggered by epinephrine and glucagon. In certain other
tors, and other regulatory molecules that act by chang- tissues PGE1 stimulates cAMP synthesis: its receptors
ing the intracellular [cAMP] and thus the activity of PKA are coupled to adenylyl cyclase through a stimulatory G
(Table 12–3). For example, glucagon binds to its recep- protein, Gs. In tissues with "2-adrenergic receptors,
tors in the plasma membrane of adipocytes, activating epinephrine lowers [cAMP]; in this case, the receptors
(via a Gs protein) adenylyl cyclase. PKA, stimulated by are coupled to adenylyl cyclase through an inhibitory G
the resulting rise in [cAMP], phosphorylates and acti- protein, Gi. In short, an extracellular signal such as epi-
vates two proteins critical to the mobilization of the nephrine or PGE1 can have quite different effects on
fatty acids of stored fats (see Fig. 17–3). Similarly, the different tissues or cell types, depending on three fac-
peptide hormone ACTH (adrenocorticotropic hormone, tors: the type of receptor in the tissue, the type of G
also called corticotropin), produced by the anterior protein (Gs or Gi) with which the receptor is coupled,
pituitary, binds to specific receptors in the adrenal cor- and the set of PKA target enzymes in the cells. By sum-
tex, activating adenylyl cyclase and raising the intracel- ming the influences that tend to increase and decrease
lular [cAMP]. PKA then phosphorylates and activates [cAMP], a cell achieves the integration of signals that
several of the enzymes required for the synthesis of we noted as a general feature of signal-transducing
cortisol and other steroid hormones. In many cell types, mechanisms (Fig. 12–1e).
the catalytic subunit of PKA can also move into the Another factor that explains how so many types of
nucleus, where it phosphorylates the cAMP response signals can be mediated by a single second messenger
element binding protein (CREB), which alters the (cAMP) is the confinement of the signaling process to
expression of specific genes regulated by cAMP. a specific region of the cell by adaptor proteins—
Some hormones act by inhibiting adenylyl cyclase, noncatalytic proteins that hold together other protein
thus lowering [cAMP] and suppressing protein phos- molecules that function in concert (further described
below). AKAPs (A kinase anchoring proteins) are b-adrenergic

multivalent adaptor proteins; one part binds to the R receptor
subunits of PKA (see Fig. 12–6a) and another to a spe-
cific structure in the cell, confining the PKA to the vicin-
ity of that structure. For example, specific AKAPs bind Palmitoyl PIP3
Adenylyl
PKA to microtubules, actin filaments, ion channels, cyclase
mitochondria, or the nucleus. Different types of cells
have different complements of AKAPs, so cAMP might AKAP5
stimulate phosphorylation of mitochondrial proteins Pi
Gs PP2A PKA cAMP ATP
in one cell and phosphorylation of actin filaments in C R R C
another. In some cases, an AKAP connects PKA with the
enzyme that triggers PKA activation (adenylyl cyclase) Target
P
or terminates PKA action (cAMP phosphodiesterase or protein
phosphoprotein phosphatase) (Fig. 12–9). The very
close proximity of these activating and inactivating
enzymes presumably achieves a highly localized, and C
very brief, response. ATP
As is now clear, to fully understand cellular signal- FIGURE 12–9 Nucleation of supramolecular complexes by A kinase
ing, researchers need tools precise enough to detect anchoring proteins (AKAPs). AKAP5 is one of a family of proteins that
and study the spatiotemporal aspects of signaling pro- act as multivalent scaffolds, holding PKA catalytic subunits—through
cesses at the subcellular level and in real time. In stud- the AKAP’s interaction with the PKA regulatory subunits—in proximity
ies of the intracellular localization of biochemical chang- to a particular region or structure in the cell. AKAP5 is targeted to rafts
es, biochemistry meets cell biology, and techniques that in the cytoplasmic surface of the plasma membrane by two covalently
cross this boundary have become essential in under- attached palmitoyl groups and a site that binds phosphatidylinositol
standing signaling pathways. Fluorescent probes have 3,4,5-trisphosphate (PIP3) in the membrane. AKAP5 also has binding
found wide application in signaling studies. Labeling of sites for the !-adrenergic receptor, adenylyl cyclase, PKA, and a phos-
functional proteins with a fluorescent tag such as the phoprotein phosphatase (PP2A), bringing them all together in the plane
green fluorescent protein (GFP) reveals their subcellu- of the membrane. When epinephrine binds to the !-adrenergic receptor,
lar localizations (see Fig. 9–16c). Changes in the state Gs" triggers adenylyl cyclase produces cAMP, which reaches the nearby
of association of two proteins (such as the R and C sub- PKA quickly and with very little dilution. PKA phosphorylates its target
protein, altering its activity, until the phosphoprotein phosphatase
units of PKA) can be seen by measuring the nonradia-
removes the phosphoryl group and returns the target protein to its pre-
tive transfer of energy between fluorescent probes
stimulus state. The AKAPs in this and other cases bring about a high
attached to each protein, a technique called fluores-
local concentration of enzymes and second messengers, so that the sig-
cence resonance energy transfer (FRET; Box 12–3).
naling circuit remains highly localized, and the duration of the signal is
Diacylglycerol, Inositol Trisphosphate, and Ca21 Have limited.
Related Roles as Second Messengers

A second broad class of GPCRs are coupled through a G
protein to a plasma membrane phospholipase C (PLC) G protein, Gq (step 2), activating it in much the same
that is specific for the membrane phospholipids phospha- way that the !-adrenergic receptor activates Gs (Fig.
tidylinositol 4,5-bisphosphate, or PIP2 (see Fig. 10–16). 12–4). The activated Gq activates the PIP2-specific PLC
When one of the hormones that acts by this mechanism (Fig. 12–10, step 3), which catalyzes (step 4) the
(Table 12–4) binds its specific receptor in the plasma production of two potent second messengers, diacyl-
membrane (Fig. 12–10, step 1), the receptor-hormone glycerol and inositol 1,4,5-trisphosphate, or IP3
complex catalyzes GTP-GDP exchange on an associated (not to be confused with PIP3, p. 456).
TABLE 12–4 Some Signals That Act through Phospholipase C, IP3, and Ca21
Acetylcholine [muscarinic M1] Gastrin-releasing peptide Platelet-derived growth factor (PDGF)
"1-Adrenergic agonists Glutamate Serotonin [5-HT-1c]
Angiogenin Gonadotropin-releasing hormone (GRH) Thyrotropin-releasing hormone (TRH)
Angiotensin II Histamine [H1] Vasopressin
ATP [P2x, P2y] Light (Drosophila)
Auxin Oxytocin
Note: Receptor subtypes are in square brackets; see footnote to Table 12–3.
BOX 12–3 METHODS FRET: Biochemistry Visualized in a Living Cell

Fluorescent probes are commonly used to detect Chromophore
rapid biochemical changes in single living cells. They (Ser65–Tyr66–Gly67)
can be designed to give an essentially instantaneous
report (within nanoseconds) on the changes in intra-
cellular concentration of a second messenger or in
the activity of a protein kinase. Furthermore, fluores-
cence microscopy has sufficient resolution to reveal
where in the cell such changes are occurring. In one
widely used procedure, the fluorescent probes are
derived from a naturally occurring fluorescent pro-
tein, the green fluorescent protein (GFP) of the
jellyfish Aequorea victoria (Fig. 1).
When excited by absorption of a photon of light, FIGURE 2 Green fluorescent protein (GFP), with the fluorescent chro-
GFP emits a photon (that is, it fluoresces) in the mophore shown in ball-and-stick form (derived from PDB ID 1GFL).
green region of the spectrum. The light-absorbing/
emitting center of GFP (its chromophore) comprises a cell and, indirectly, to measure local concentrations
an oxidized form of the tripeptide –Ser65–Tyr66–Gly67– of compounds that change the distance between two
(Fig. 2). Oxidation of the tripeptide is catalyzed by proteins.
the GFP protein itself (Fig. 3), so it is possible to An excited fluorescent molecule such as GFP or
clone the protein into virtually any cell, where it can YFP can dispose of the energy from the absorbed
serve as a fluorescent marker for any protein to which photon in either of two ways: (1) by fluorescence,
it is fused (see Fig. 9–18). Variants of GFP, with dif- emitting a photon of slightly longer wavelength (lower
ferent fluorescence spectra, are produced by genetic energy) than the exciting light, or (2) by nonradiative
engineering. For example, in the yellow fluorescent fluorescence resonance energy transfer (FRET),
protein (YFP), Ala206 in GFP is replaced by a Lys in which the energy of the excited molecule (the
residue, changing the wavelength of light absorption donor) passes directly to a nearby molecule (the
and fluorescence. Other variants of GFP fluoresce acceptor) without emission of a photon, exciting
blue (BFP) or cyan (CFP) light, and a related protein the acceptor (Fig. 5). The acceptor can now decay to
(mRFP1) fluoresces red light (Fig. 4). GFP and its its ground state by fluorescence; the emitted photon
variants are compact structures that retain their abil- has a longer wavelength (lower energy) than both the
ity to fold into their native !-barrel conformation even original exciting light and the fluorescence emission
when fused with another protein. These fluorescent of the donor. This second mode of decay (FRET) is
hybrid proteins act as spectroscopic rulers for mea- possible only when donor and acceptor are close to
suring distances between interacting proteins within each other (within 1 to 50 Å); the efficiency of FRET
is inversely proportional to the sixth power of the
distance between donor and acceptor. Thus very
small changes in the distance between donor and
acceptor register as very large changes in FRET, mea-
sured as the fluorescence of the acceptor molecule
when the donor is excited. With sufficiently sensitive
light detectors, this fluorescence signal can be located
to specific regions of a single, living cell.
FRET has been used to measure [cAMP] in living
cells. The gene for GFP is fused with that for the regu-
latory subunit (R) of cAMP-dependent protein kinase
FIGURE 1 Aequorea victoria, a jellyfish abundant in Puget Sound, (PKA), and the gene for BFP is fused with that for the
Washington State. catalytic subunit (C) (Fig. 6). When these two hybrid
O O O2 H2O2 H H O O
Gly
multiple
N
N Gly steps N Gly N Gly
:
HN O HN C N C N
HO HO HO HO
H! OH OH
Tyr Ser Ser Ser
Ser H!
Mature chromophore
FIGURE 3 The chromophore in GFP is derived from a series of three that takes place in multiple steps. An abbreviated mechanism is
amino acids: –Ser65–Tyr66–Gly67–. Maturation of the chromophore shown here.
involves an internal rearrangement, coupled to an oxidation reaction
448
CFP GFP cAMP-dependent protein kinase (PKA)

100
Relative fluorescence
BFP YFP mRFP1
80
GFP
60
40 C C
C
R R 545 nm
20
0 (inactive)
400 500 600 700 FRET
Wavelength (nm)
BFP
380 nm
FIGURE 4 Emission spectra of GFP variants. cAMP cAMP
proteins are expressed in a cell, BFP (donor; excita-

tion at 380 nm, emission at 460 nm) and GFP (accep-
tor; excitation at 475 nm, emission at 545 nm) in the 380 nm
inactive PKA (R2C2 tetramer) are close enough to
undergo FRET. Wherever in the cell [cAMP] increases, 460 nm
the R2C2 complex dissociates into R2 and 2C and the
FRET signal is lost, because donor and acceptor are +
380 nm
now too far apart for efficient FRET. Viewed in the
fluorescence microscope, the region of higher [cAMP]
has a minimal GFP signal and higher BFP signal. Mea- no emission
(active) at 545 nm
suring the ratio of emission at 460 nm and 545 nm
gives a sensitive measure of the change in [cAMP]. By FIGURE 6 Measuring [cAMP] with FRET. Gene fusion creates hybrid
determining this ratio for all regions of the cell, the proteins that exhibit FRET when the PKA regulatory (R) and catalytic
investigator can generate a false color image of the cell (C) subunits are associated (low [cAMP]). When [cAMP] rises, the
in which the ratio, or relative [cAMP], is represented subunits dissociate and FRET ceases. The ratio of emission at 460 nm
by the intensity of the color. Images recorded at timed (dissociated) and 545 nm (complexed) thus offers a sensitive measure
intervals reveal changes in [cAMP] over time. of [cAMP].
A variation of this technology has been used to
measure the activity of PKA in a living cell (Fig. 7). CFP (donor). When the Ser residue is not phosphory-
Researchers create a phosphorylation target for PKA lated, 14-3-3 has no affinity for the Ser residue and
by producing a hybrid protein containing four ele- the hybrid protein exists in an extended form, with
ments: YFP (acceptor); a short peptide with a Ser the donor and acceptor too far apart to generate a
residue surrounded by the consensus sequence for FRET signal. Wherever PKA is active in the cell, it
PKA; a P –Ser-binding domain (called 14-3-3); and phosphorylates the Ser residue of the hybrid protein,
and 14-3-3 binds to the P –Ser. In doing so, it draws
476 527 YFP and CFP together and a FRET signal is detected
433 nm 433
nm nm FRET nm with the fluorescence microscope, revealing the pres-
ence of active PKA.
CFP YFP 476 527

433 433
nm nm FRET nm
nm
PKA consensus
protein- sequence
protein ATP ADP
CFP
interaction
Ser
Genetically P
engineered hybrid PKA
proteins
14-3-3
FIGURE 5 When the donor protein (CFP) is excited with monochro- (phosphoserine-
matic light of wavelength 433 nm, it emits fluorescent light at 476 nm binding domain)
YFP
(left). When the (red) protein fused with CFP interacts with the
(purple) protein fused with YFP, that interaction brings CFP and YFP FIGURE 7 Measuring the activity of PKA with FRET. An engineered
close enough to allow fluorescence resonance energy transfer (FRET) protein links YFP and CFP via a peptide that contains a Ser residue
between them. Now, when CFP absorbs light of 433 nm, instead of surrounded by the consensus sequence for phosphorylation by PKA,
fluorescing at 476 nm, it transfers energy directly to YFP, which then and the 14-3-3 P –Ser-binding domain. Active PKA phosphorylates
fluoresces at its characteristic emission wavelength, 527 nm. The the Ser residue, which docks with the 14-3-3 binding domain, bringing
ratio of light emission at 527 and 476 nm is therefore a measure of the fluorescence proteins close enough to allow FRET to occur,
the interaction of the red and purple proteins. revealing the presence of active PKA.
449
450 Biosignaling
FIGURE 12–10 Hormone-activated phospholipase C and IP3. Two intra-

1 Hormone (H) binds to a cellular second messengers are produced in the hormone-sensitive
specific receptor.
phosphatidylinositol system: inositol 1,4,5-trisphosphate (IP3) and
Receptor diacylglycerol are cleaved from phosphatidylinositol 4,5-bisphosphate
Extracellular space (PIP2). Both contribute to the activation of protein kinase C. By raising
H cytosolic [Ca2"], IP3 also activates other Ca2"-dependent enzymes; thus
Plasma
Ca2" also acts as a second messenger.
Gqa membrane
GDP
Phospholipase C
GTP (PLC)
2 The occupied
receptor causes Gqa PIP2
GDP-GTP GDP
GTP
exchange on Gqa.
3 Gqa, with bound GTP, moves

to PLC and activates it.
4 Active PLC cleaves PIP2

Endoplasmic to IP3 and diacylglycerol.
reticulum
5 IP3 binds to a specific Diacylglycerol

receptor-gated Ca2"
channel, releasing
sequestered Ca2".
IP3
Cytosol
Protein
kinase C
Ca2"
6 Diacylglycerol and Ca2"activate

protein kinase C at the surface
Ca2" of the plasma membrane.
channel
7 Phosphorylation of cellular
proteins by protein kinase C
produces some of the cellular
responses to the hormone.
rushes into the cytosol (Fig. 12–10, step 5), and the
O!
cytosolic [Ca2"] rises sharply to about 10!6 M. One
!
O P O effect of elevated [Ca2"] is the activation of protein
H O kinase C (PKC). Diacylglycerol cooperates with Ca2"
O 6 5 in activating PKC, thus also acting as a second mes-
!
O P O H senger (step 6). Activation involves the movement of
OH H O!
1 4
!
O OH HO a PKC domain (the pseudosubstrate domain) away
H O P O!
2 3
from its location in the substrate-binding region of the
H H O enzyme, allowing the enzyme to bind and phosphorylate
Inositol 1,4,5-trisphosphate (IP3) proteins that contain a PKC consensus sequence—Ser
Inositol trisphosphate, a water-soluble compound, or Thr residues embedded in an amino acid sequence
diffuses from the plasma membrane to the endoplas- recognized by PKC (step 7). There are several iso-
mic reticulum (ER), where it binds to specific IP3- zymes of PKC, each with a characteristic tissue distri-
gated Ca2" channels, causing them to open. The action bution, target protein specificity, and role. Their targets
of the SERCA pump (see p. 410) ensures that [Ca2"] in include cytoskeletal proteins, enzymes, and nuclear
the ER is orders of magnitude higher than that in the proteins that regulate gene expression. Taken together,
cytosol, so when these gated Ca2" channels open, Ca2" this family of enzymes has a wide range of cellular
actions, affecting neuronal and immune function and (a)

the regulation of cell division, for example. Long central
helix
Calcium Is a Second Messenger That May Be Localized

in Space and Time
There are many variations on this basic scheme for Ca2"
signaling. In many cell types that respond to extracel-
lular signals, Ca2" serves as a second messenger that
triggers intracellular responses, such as exocytosis in
neurons and endocrine cells, contraction in muscle, and EF hand
Long central
cytoskeletal rearrangements during amoeboid move- helix
ment. In unstimulated cells, cytosolic [Ca2"] is kept very (b) (c)
E helix
low ($10!7 M) by the action of Ca2" pumps in the ER,
mitochondria, and plasma membrane (as further dis- Ca2+
cussed below). Hormonal, neural, or other stimuli cause
either an influx of Ca2" into the cell through specific
Ca2" channels in the plasma membrane or the release of F helix
sequestered Ca2" from the ER or mitochondria, in
either case raising the cytosolic [Ca2"] and triggering a Helical domain of a
cellular response. calmodulin-regulated protein
Changes in intracellular [Ca2"] are detected by
FIGURE 12–11 Calmodulin. This is the protein mediator of many Ca2"-
Ca -binding proteins that regulate a variety of Ca2"-
2"
stimulated enzymatic reactions. Calmodulin has four high-affinity Ca2"-
dependent enzymes. Calmodulin (CaM; Mr 17,000) is
binding sites (Kd ! 0.1 to 1 $M). (a) A ribbon model of the crystal struc-
an acidic protein with four high-affinity Ca2"-binding ture of calmodulin (PDB ID 1CLL). The four Ca2"-binding sites are
sites. When intracellular [Ca2"] rises to about 10!6 M occupied by Ca2" (purple). The amino-terminal domain is on the left;
(1 $M), the binding of Ca2" to calmodulin drives a confor- the carboxyl-terminal domain on the right. (b) Calmodulin associated
mational change in the protein (Fig. 12–11a). Calmod- with a helical domain (red) of one of the many enzymes it regulates,
ulin associates with a variety of proteins and, in its calmodulin-dependent protein kinase II (PDB ID 1CDL). Notice that the
Ca2"-bound state, modulates their activities (Fig. long central " helix of calmodulin visible in (a) has bent back on itself in
12–11b). It is a member of a family of Ca2"-binding pro- binding to the helical substrate domain. The central helix of calmodulin
teins that also includes troponin (see Fig. 5–32), which is clearly more flexible in solution than in the crystal. (c) Each of the four
triggers skeletal muscle contraction in response to Ca2"-binding sites occurs in a helix-loop-helix motif called the EF hand,
increased [Ca2"]. This family shares a characteristic also found in many other Ca2"-binding proteins.
Ca2"-binding structure, the EF hand (Fig. 12–11c).
Calmodulin is an integral subunit of the Ca2"/
calmodulin-dependent protein kinases (CaM
TABLE 12–5 Some Proteins Regulated by Ca21
kinases, types I through IV). When intracellular [Ca2"] and Calmodulin
increases in response to a stimulus, calmodulin binds Adenylyl cyclase (brain)
Ca2", undergoes a change in conformation, and acti- Ca2"/calmodulin-dependent protein kinases (CaM
vates the CaM kinase. The kinase then phosphorylates kinases I to IV)
target enzymes, regulating their activities. Calmodulin
Ca2"-dependent Na" channel (Paramecium)
is also a regulatory subunit of phosphorylase b kinase of
muscle, which is activated by Ca2". Thus Ca2" triggers Ca2"-release channel of sarcoplasmic reticulum
ATP-requiring muscle contractions while also activating Calcineurin (phosphoprotein phosphatase 2B)
glycogen breakdown, providing fuel for ATP synthesis. cAMP phosphodiesterase
Many other enzymes are also known to be modulated by
cAMP-gated olfactory channel
Ca2" through calmodulin (Table 12–5). The activity of
the second messenger Ca2", like that of cAMP, can be cGMP-gated Na", Ca2" channels (rod and cone cells)
spatially restricted; after its release triggers a local Glutamate decarboxylase
response, Ca2" is generally removed before it can diffuse Myosin light-chain kinases
to distant parts of the cell.
NAD" kinase
Very commonly, Ca2" level does not simply rise and
then decrease, but rather oscillates with a period of a Nitric oxide synthase
few seconds (Fig. 12–12)—even when the extracellu- Phosphatidylinositol 3-kinase
lar concentration of the triggering hormone remains Plasma membrane Ca2" ATPase (Ca2" pump)
constant. The mechanism underlying [Ca2"] oscillations
RNA helicase (p68)
presumably entails feedback regulation by Ca2" on
452 Biosignaling
complex assembled on scaffold proteins such as AKAPs.

This subcellular localization of target enzymes, combined
with temporal and spatial gradients in [Ca2"] and [cAMP],
allows a cell to respond to one or several signals with sub-
tly nuanced metabolic changes, localized in space and time.
GPCRs Mediate the Actions of a

Wide Variety of Signals
The human genome encodes about 1,000 G protein–
coupled receptors, recognizable by their seven trans-
membrane helical segments and certain highly con-
0 0.5 1.0 served residues. Each is expressed selectively, in certain
(a) [Ca2"] (mM) cell types or under certain conditions. Together, they
allow cells and tissues to respond to a wide array of dif-
600 ferent stimuli, including various low molecular weight
amines, peptides, proteins, eicosanoids and other lipids,
500
] (nM)
as well as light and compounds detected by olfaction

and gustation. The determination of several GCPR
2"
400
structures by crystallography (Fig. 12–13), including
Cytosolic [Ca
300 the !-adrenergic receptor with its G protein, and the

histamine receptor, has stimulated great interest in
200 both the transduction mechanism(s) and the possibili-
ties of altering receptor activity with drugs. These two
100 receptors are the targets of a variety of widely used
0 100 200 300 400
(b) Time (s)
beta-blocker and antihistamine medications, respec-
tively. The similarities in GPCR structures go beyond
FIGURE 12–12 Triggering of oscillations in intracellular [Ca21] by extra- the common seven–transmembrane helix pattern; the
cellular signals. (a) A dye (fura) that undergoes fluorescence changes structures of five different GPCRs are almost superim-
when it binds Ca2" is allowed to diffuse into cells, and its instantaneous posable (Fig. 12–13c). Clearly, something about this
light output is measured by fluorescence microscopy. Fluorescence three-dimensional structure makes it effective as a
intensity is represented by color; the color scale relates intensity of color transducer of many disparate signals.
to [Ca2"], allowing determination of the absolute [Ca2"]. In this case,
thymocytes (cells of the thymus) have been stimulated with extracellular
ATP, which raises their internal [Ca2"]. The cells are heterogeneous in SUMMARY 12.2 G Protein–Coupled Receptors
their responses; some have high intracellular [Ca2"] (red), others much
lower (blue). (b) When such a probe is used in a single hepatocyte, the
and Second Messengers
agonist norepinephrine (added at the arrow) causes oscillations of ! G protein–coupled receptors (GPCRs) share a
[Ca2"] from 200 to 500 nM. Similar oscillations are induced in other cell common structural arrangement of seven
types by other extracellular signals. transmembrane helices and act through
heterotrimeric G proteins. On ligand binding,
GPCRs catalyze the exchange of GTP for GDP on
the G protein, causing dissociation of the G"
some part of the Ca2"-release process. Whatever the subunit; G" then stimulates or inhibits the activity
mechanism, the effect is that one kind of signal (hor- of an effector enzyme, changing the level of its
mone concentration, for example) is converted into second-messenger product.
another (frequency and amplitude of intracellular [Ca2"]
! The !-adrenergic receptor activates a stimulatory
“spikes”). The Ca2" signal diminishes as Ca2" diffuses
away from the initial source (the Ca2" channel), is G protein, Gs, thereby activating adenylyl cyclase
sequestered in the ER, or is pumped out of the cell. and raising the concentration of the second
There is significant cross talk between the Ca2" and messenger cAMP. Cyclic AMP stimulates
cAMP signaling systems. In some tissues, both the enzyme cAMP-dependent protein kinase to phosphorylate
that produces cAMP (adenylyl cyclase) and the enzyme key target enzymes, changing their activities.
that degrades cAMP (phosphodiesterase) are stimulated ! Enzyme cascades, in which a single molecule of
by Ca2". Temporal and spatial changes in [Ca2"] can there- hormone activates a catalyst to activate another
fore produce transient, localized changes in [cAMP]. We catalyst, and so on, result in the large signal
have noted already that PKA, the enzyme that responds to amplification that is characteristic of hormone
cAMP, is often part of a highly localized supramolecular receptor systems.
12.3 Receptor Tyrosine Kinases 453
(a) b2-adrenergic Ligand- (b) m-opioid Morphine analog in (c) Histamine H1 Doxepin in (d)
receptor binding site receptor ligand-binding site receptor ligand-binding site
Outside
Inside
FIGURE 12–13 The !-adrenergic receptor and several other GPCRs. (a) The
!2-adrenergic receptor and its associated trimeric G protein, Gs (PDB ID
Gsb 3SN6). The bound agonist (epinephrine) is shown in yellow, and the sub-
Gsa
Gsg units of Gs are three shades of green. (b) The $ opioid receptor (PDB ID
4DKL), the target of morphine and codeine, with a morphine analog in the
ligand-binding site. (c) The histamine H1 receptor with the bound drug
doxepin (PDB ID 3RZE). (d) Five GPCR structures superimposed to show
the remarkable conservation of structure. Shown are the human A2A ade-
! Cyclic AMP concentration is eventually reduced by nosine receptor (orange; PDB ID 3EML); turkey !1-adrenergic receptor
cAMP phosphodiesterase, and Gs turns itself off by (blue; PDB ID 2VT4), human !2-adrenergic receptor (green; PDB ID 2RH1),
hydrolysis of its bound GTP to GDP, acting as a rhodopsin from squid (yellow; PDB ID 2Z73); and bovine rhodopsin, (red;
self-limiting binary switch. PDB ID 1U19).
! When the epinephrine signal persists, !-adrenergic
receptor–specific protein kinase and !-arrestin on the cytoplasmic face, connected by a single trans-
temporarily desensitize the receptor and cause it membrane segment. The cytoplasmic domain is a protein
to move into intracellular vesicles. kinase that phosphorylates Tyr residues in specific target
! Some receptors stimulate adenylyl cyclase through proteins—a Tyr kinase. The receptors for insulin and
Gs; others inhibit it through Gi. Thus cellular [cAMP] epidermal growth factor are prototypes for this group.
reflects the integrated input of two (or more) signals.
! Noncatalytic adaptor proteins such as AKAPs hold Stimulation of the Insulin Receptor Initiates a
together proteins involved in a signaling process, Cascade of Protein Phosphorylation Reactions
increasing the efficiency of their interactions and Insulin regulates both metabolic enzymes and gene
in some cases confining the process to a specific expression. Insulin does not enter cells, but initiates a
subcellular location. signal that travels a branched pathway from the plasma
! Some GPCRs act via a plasma membrane membrane receptor to insulin-sensitive enzymes in the
phospholipase C that cleaves PIP2 to diacylglycerol cytosol and to the nucleus, where it stimulates the tran-
and IP3. By opening Ca2" channels in the endoplasmic scription of specific genes. The active insulin receptor
reticulum, IP3 raises cytosolic [Ca2"]. Diacylglycerol protein (INSR) consists of two identical " subunits pro-
and Ca2" act together to activate protein kinase C, truding from the outer face of the plasma membrane
which phosphorylates and changes the activity of and two transmembrane ! subunits with their carboxyl
specific cellular proteins. Cellular [Ca2"] also termini protruding into the cytosol—a dimer of "!
regulates (often through calmodulin) many other monomers (Fig. 12–14). The " subunits contain the
enzymes and proteins involved in secretion, insulin-binding domain, and the intracellular domains of
cytoskeletal rearrangements, or contraction. the ! subunits contain the protein kinase activity that
transfers a phosphoryl group from ATP to the hydroxyl
12.3 Receptor Tyrosine Kinases group of Tyr residues in specific target proteins. Signal-
ing through INSR begins when the binding of one insulin
The receptor tyrosine kinases (RTKs), a large family molecule between the two subunits of the dimer acti-
of plasma membrane receptors with intrinsic protein vates the Tyr kinase activity, and each ! subunit phos-
kinase activity, transduce extracellular signals by a phorylates three critical Tyr residues near the carboxyl
mechanism fundamentally different from that of GPCRs. terminus of the other ! subunit. This autophosphory-
RTKs have a ligand-binding domain on the extracellular lation opens up the active site so that the enzyme can
face of the plasma membrane and an enzyme active site phosphorylate Tyr residues of other target proteins.
454 Biosignaling
(a) Top view (b) Top view FIGURE 12–14 Activation of the insulin-receptor tyrosine kinase by auto-
phosphorylation. (a) The insulin-binding region of the insulin receptor lies
Insulin
outside the cell and comprises (b) two ! subunits and the extracellular
portions of two " subunits, intertwined to form the insulin-binding site
(shown as a surface contour model of the crystal structure, derived from
Side view PDB ID 2DTG). (The structure of the transmembrane domain has not been
Insulin solved by crystallography.) The binding of insulin (red; PDB ID 2CEU) is
communicated through the single transmembrane helix of each " subunit
Insulin to the paired Tyr kinase domains inside the cell, activating them to phos-
phorylate each other on three Tyr residues. (c) (PDB ID 1IRK) In the inactive
Side view form of the Tyr kinase domain (before Tyr phosphorylations), the activation
loop (teal) sits in the active site, and none of the critical Tyr residues (white
and red ball-and-stick structures) are phosphorylated. This conformation is
stabilized by hydrogen bonding between Tyr1162 and Asp1132. (d) (PDB ID
1IR3) Activation of the Tyr kinase allows each " subunit to phosphorylate
three Tyr residues (Tyr1158, Tyr1162, Tyr1163) on the other " subunit. (Phos-
phoryl groups are depicted in red and orange.) The introduction of three
highly charged P –Tyr residues forces a 30 Å change in the position of the
activation loop, away from the substrate-binding site, which becomes
available to bind and phosphorylate a target protein.
P P P
(c) Inactive (unphosphorylated) (d) Active (triply phosphorylated)

tyrosine kinase domain tyrosine kinase domain
Tyr1163
Tyr1158
Asp1132
Tyr1158 Target
Tyr1162 protein
Asp1132
Tyr1163
Tyr1162
Activation loop blocks substrate-binding site Triply phosphorylated activation loop moves
dramatically, making room for the target protein
in the substrate-binding site.
The mechanism of activation of the INSR protein kinase of the protein Grb2. (SH2 is an abbreviation of Src
is similar to that described for PKA and PKC: a region of homology 2, so named because the sequence of an SH2
the cytoplasmic domain (an autoinhibitory sequence) domain is similar to that of a domain in Src (pronounced
that usually occludes the active site moves out of the sark), another protein Tyr kinase.) Several signaling
active site after being phosphorylated, opening up the proteins contain SH2 domains, all of which bind P –Tyr
site for the binding of target proteins (Fig. 12–14). residues in a protein partner. Grb2 is an adaptor protein,
When INSR is autophosphorylated (Fig. 12–15, with no intrinsic enzymatic activity. Its function is to
step 1), one of its targets is insulin receptor substrate-1 bring together two proteins (in this case, IRS-1 and the
(IRS-1; step 2). Once phosphorylated on several of its protein Sos) that must interact to enable signal transduc-
Tyr residues, IRS-1 becomes the point of nucleation for a tion. In addition to its SH2 ( P –Tyr-binding) domain,
complex of proteins (step 3) that carry the message Grb2 also contains a second protein-binding domain,
from the insulin receptor to end targets in the cytosol and SH3, that binds to a proline-rich region of Sos, recruiting
nucleus, through a long series of intermediate proteins. Sos to the growing receptor complex. When bound to
First, a P –Tyr residue of IRS-1 binds to the SH2 domain Grb2, Sos acts as a guanosine nucleotide–exchange factor
Insulin receptor binds insulin

FIGURE 12–15 Regulation of gene expression by
1
and undergoes autophosphor- insulin through a MAP kinase cascade. The insulin
ylation on its carboxyl-terminal receptor (INSR) consists of two " subunits on the
Tyr residues. outer face of the plasma membrane and two !
Insulin
a subunits that traverse the membrane and protrude
b Insulin receptor from the cytosolic face. Binding of insulin to the "
2
phosphorylates IRS-1 subunits triggers a conformational change that
P P on its Tyr residues. allows the autophosphorylation of Tyr residues in
Cytosol P
P P
P the carboxyl-terminal domain of the ! subunits.
Autophosphorylation further activates the Tyr
3 SH2 domain of Grb2 binds kinase domain, which then catalyzes phosphoryla-
to P –Tyr of IRS-1. Sos
P tion of other target proteins. The signaling pathway
binds to Grb2, then to Ras,
IRS-1 IRS-1
causing GDP release and by which insulin regulates the expression of specific
P P GTP binding to Ras.
Grb2 GDP genes consists of a cascade of protein kinases, each
Sos of which activates the next. INSR is a Tyr-specific
GTP Ras kinase; the other kinases (all shown in blue)
Raf-1 4 Activated Ras binds phosphorylate Ser or Thr residues. MEK is a dual-
MEK and activates Raf-1.
specificity kinase, which phosphorylates both a Thr
and a Tyr residue in ERK (extracellular regulated
P kinase); MEK is mitogen-activated, ERK-activating
Nucleus MEK
5 Raf-1 phosphorylates kinase; SRF is serum response factor.
P
MEK on two Ser
SRF Elk1
residues, activating it.
P P MEK phosphorylates
ERK ERK ERK on a Thr and a Tyr
ERK P residue, activating it.
P
6 ERK moves into the
SRF Elk1 P
nucleus and The proteins Raf-1, MEK, and ERK are members of
DNA phosphorylates three larger families, for which several nomenclatures are
mRNA nuclear transcription employed. ERK is in the MAPK family (mitogen-activated
factors such as Elk1,
Protein activating them. protein kinases; mitogens are extracellular signals that
7 Phosphorylated Elk1
induce mitosis and cell division). Soon after discovery of
joins SRF to stimulate the first MAPK enzyme, that enzyme was found to be
the transcription and activated by another protein kinase, which was named
translation of a set of
genes needed for cell
MAP kinase kinase (MEK belongs to this family), and
division. when a third kinase that activated MAP kinase kinase was
discovered, it was given the slightly ludicrous family
name MAP kinase kinase kinase (Raf-1 is in this family).
Somewhat less cumbersome are the abbreviations for
(GEF), catalyzing the replacement of bound GDP with these three families: MAPK, MAPKK, and MAPKKK.
GTP on Ras, a G protein. Kinases in the MAPK and MAPKKK families are specific
Ras is the prototype of a family of small G pro- for Ser or Thr residues, and MAPKKs (here, MEK) phos-
teins that mediate a wide variety of signal transduc- phorylate both a Ser and a Tyr residue in their substrate,
tions (see Box 12–2). Like the trimeric G protein that a MAPK (here, ERK).
functions with the !-adrenergic system (Fig. 12–5), Ras Biochemists now recognize this insulin pathway as
can exist in either the GTP-bound (active) or GDP- but one instance of a more general scheme in which
bound (inactive) conformation, but Ras (,20 kDa) acts hormone signals, via pathways similar to that shown in
as a monomer. When GTP binds, Ras can activate a pro- Figure 12–15, result in phosphorylation of target enzymes
tein kinase, Raf-1 (Fig. 12–15, step 4), the first of three by protein kinases. The target of phosphorylation is often
protein kinases—Raf-1, MEK, and ERK—that form a another protein kinase, which then phosphorylates a
cascade in which each kinase activates the next by third protein kinase, and so on. The result is a cascade
phosphorylation (step 5). The protein kinases MEK of reactions that amplifies the initial signal by many
and ERK are activated by phosphorylation of both a orders of magnitude (see Fig. 12–1b). MAPK cascades
Thr and a Tyr residue. When activated, ERK mediates (Fig. 12–15) mediate signaling initiated by a variety of
some of the biological effects of insulin by entering the growth factors, such as platelet-derived growth factor
nucleus and phosphorylating transcription factors, (PDGF) and epidermal growth factor (EGF). Another
such as Elk1 (step 6), that modulate the transcription general scheme exemplified by the insulin receptor
of about 100 insulin-regulated genes (step 7), some of pathway is the use of nonenzymatic adaptor proteins to
which encode proteins essential for cell division. Thus, bring together the components of a branched signaling
insulin acts as a growth factor. pathway, to which we now turn.
456 Biosignaling
The Membrane Phospholipid PIP3 Functions muscle, the cascade of protein phosphorylations initi-
ated by insulin stimulates glycogen synthesis (Fig. 12–16).
at a Branch in Insulin Signaling In a third signaling branch in muscle and fat tissue, PKB
The signaling pathway from insulin branches at IRS-1 triggers the clathrin-aided movement of glucose trans-
(Fig. 12–15, step 2). Grb2 is not the only protein that porters (GLUT4) from internal vesicles to the plasma
associates with phosphorylated IRS-1. The enzyme membrane, stimulating glucose uptake from the blood
phosphoinositide 3-kinase (PI3K) binds IRS-1 through (Fig. 12–16, step 5; see also Box 11–1).
PI3K’s SH2 domain (Fig. 12–16). Thus activated, PI3K As in all signaling pathways, there is a mechanism
converts the membrane lipid phosphatidylinositol for terminating the activity of the PI3K–PKB
4,5-bisphosphate (PIP2) to phosphatidylinositol pathway. A PIP3-specific phosphatase (PTEN in humans)
3,4,5-trisphosphate (PIP3). The multiply charged head removes the phosphoryl group at the 3 position of PIP3
group of PIP3, protruding on the cytoplasmic side of the to produce PIP2, which no longer serves as a binding site
plasma membrane, is the starting point for a second for PKB, and the signaling chain is broken. In various
signaling branch involving another cascade of protein types of cancer, it is often found that the PTEN gene has
kinases. When bound to PIP3, protein kinase B (PKB; undergone mutation, resulting in a defective regulatory
also called Akt) is phosphorylated and activated by yet circuit and abnormally high levels of PIP3 and of PKB
another protein kinase, PDK1. The activated PKB then activity. The result seems to be a continuing signal for
phosphorylates Ser or Thr residues in its target pro- cell division and thus tumor growth. ■
teins, one of which is glycogen synthase kinase 3 The insulin receptor is the prototype for several
(GSK3). In its active, nonphosphorylated form, GSK3 receptor enzymes with a similar structure and RTK
phosphorylates glycogen synthase, inactivating it and activity (Fig. 12–17). The receptors for EGF and
thereby contributing to the slowing of glycogen synthe- PDGF, for example, have structural and sequence simi-
sis. (This mechanism is only part of the explanation for larities to INSR, and both have a protein Tyr kinase
the effects of insulin on glycogen metabolism.) When activity that phosphorylates IRS-1. Many of these recep-
phosphorylated by PKB, GSK3 is inactivated. By thus tors dimerize after binding ligand; INSR is the excep-
preventing inactivation of glycogen synthase in liver and tion, as it is already an ("!)2 dimer before insulin binds.
1 IRS-1, phosphorylated
by the insulin receptor,
P activates PI3K by
binding to its SH2
IRS-1 domain. PI3K converts
3 GSK3, inactivated by P PIP2 to PIP3.
P
phosphorylation, cannot
PI3K PIP2
convert glycogen synthase
(GS) to its inactive form GSK3
by phosphorylation, so GS P P PIP3
(inactive)
remains active. P
PKB 2 PKB bound to PIP3 is
phosphorylated by PDK1
GS (not shown). Thus acti-
(inactive) P
vated, PKB phosphory-
lates GSK3 on a Ser
GSK3 residue, inactivating it.
(active)
GS
(active)
GLUT4
Glycogen
Glucose
4 Synthesis of
glycogen from PKB stimulates
glucose is
5
movement of glucose
accelerated. transporter GLUT4 from
internal membrane
vesicles to the plasma
membrane, increasing
the uptake of glucose.
FIGURE 12–16 Insulin action on glycogen synthesis and GLUT4 move- of the glucose transporter GLUT4 to the plasma membrane, and the
ment to the plasma membrane. The activation of PI3 kinase (PI3K) by activation of glycogen synthase.
phosphorylated IRS-1 signals (through protein kinase B, PKB) movement
Ligand-binding domain related IRS proteins (IRS2, IRS3), each with its own char-
acteristic tissue distribution and function, further enriching
the signaling possibilities in pathways initiated by RTKs.
The JAK-STAT Signaling System Also Involves

a
Tyrosine Kinase Activity
b A variation on the basic theme of receptor Tyr kinases
Outside is receptors that have no intrinsic protein kinase activity
but, when occupied by their ligand, bind a cytosolic Tyr
Inside
kinase. One example is the system that regulates the
formation of erythrocytes in mammals. The develop-
mental signal, or cytokine, for this system is erythro-
poietin (EPO), a 165 amino acid protein produced in
INSR VEGFR PDGFR EGFR TrkA FGFR the kidneys. When EPO binds to its plasma membrane
Tyrosine receptor (Fig. 12–18), the receptor dimerizes, and the
Cys-rich Leu-rich Ig-like Asp/Glu-rich
kinase
domain domain domain region
dimer can bind and activate the soluble protein kinase
domain JAK (Janus kinase). The activated JAK phosphorylates
FIGURE 12–17 Receptor tyrosine kinases. Growth factor receptors that
signal through Tyr kinase activity include those for insulin (INSR), vascu-
Erythropoietin
lar epidermal growth factor (VEGFR), platelet-derived growth factor EPO receptor
(PDGFR), epidermal growth factor (EGFR), high-affinity nerve growth
factor (TrkA), and fibroblast growth factor (FGFR). All these receptors
have a Tyr kinase domain on the cytoplasmic side of the plasma mem- Cytosol P P
brane (blue). The extracellular domain is unique to each type of recep- P P SHC
Grb2
tor, reflecting the different growth-factor specificities. These extracellu-
P P
lar domains are typically combinations of structural motifs such as
cysteine- or leucine-rich segments and segments containing one of STAT JAK JAK MAPK
several motifs common to immunoglobulins (Ig-like domains; see cascade
Fig. 4–22). Many other receptors of this type are encoded in the human
genome, each with a different extracellular domain and ligand specificity. MAPK (b)
SH2 domain
(The protomer of the insulin receptor is one "! unit.) (a)
STAT
P
The binding of adaptor proteins such as Grb2 to P –Tyr
residues is a common mechanism for promoting protein- P STAT
NLS
protein interactions initiated by RTKs, a subject to dimerization STAT
P
which we return in Section 12.5.
In addition to the many receptors that act as protein
Tyr kinases (the RTKs), several receptorlike plasma
membrane proteins have protein Tyr phosphatase activ-
Nucleus P STAT Affects gene expression
ity. Based on the structures of these proteins, we can
STAT
surmise that their ligands are components of the extra- P
cellular matrix or are surface molecules on other cells.
Although their signaling roles are not yet as well under-
stood as those of the RTKs, they clearly have the poten-
FIGURE 12–18 The JAK-STAT transduction mechanism for the erythro-
tial to reverse the actions of signals that stimulate RTKs.
poietin receptor. Binding of erythropoietin (EPO) causes dimerization of
What spurred the evolution of such complicated regu-
the EPO receptor, which allows JAK, a soluble Tyr kinase, to bind to the
latory machinery? This system allows one activated recep-
internal domain of the receptor and phosphorylate it on several Tyr resi-
tor to activate several IRS-1 molecules, amplifying the dues. (a) In one signaling pathway, the SH2 domain of the STAT protein
insulin signal, and it provides for the integration of signals STAT5 binds to P –Tyr residues on the receptor, bringing it into proximity
from different receptors such as EGFR and PDGFR, each with JAK. Following phosphorylation of STAT5 by JAK, two STAT5
of which can phosphorylate IRS-1. Furthermore, because molecules dimerize, each binding the other’s P –Tyr residue, thus
IRS-1 can activate any of several proteins that contain SH2 exposing a nuclear localization sequence (NLS) that targets the dimer
domains, a single receptor acting through IRS-1 can trigger for transport into the nucleus. In the nucleus, STAT5 turns on the
two or more signaling pathways; insulin affects gene expression of EPO-controlled genes. (b) In a second signaling pathway,
expression through the Grb2-Sos-Ras-MAPK pathway and following EPO binding and autophosphorylation of JAK, the adaptor pro-
affects glycogen metabolism and glucose transport through tein SHC binds the P –Tyr of the receptor, then Grb2 binds SHC and
the PI3K–PKB pathway. Finally, there are several closely triggers the MAPK cascade, as in the insulin system (see Fig. 12–15).
458 Biosignaling
several Tyr residues in the cytoplasmic domain of the of a !2-adrenergic receptor, and PKB, activated by insu-
EPO receptor. A family of transcription factors, collec- lin (Fig. 12–19), phosphorylates two Ser residues in
tively called STATs (signal transducers and activators the same region. Phosphorylation of these four residues
of transcription), are also targets of JAK. An SH2 triggers clathrin-aided internalization of the !2-adrenergic
domain in STAT5 binds P –Tyr residues in the EPO receptor, taking it out of service and lowering the cell’s
receptor, positioning the STAT for phosphorylation by sensitivity to epinephrine. A second type of cross talk
JAK in response to EPO. The phosphorylated STAT5 between these receptors occurs when P –Tyr residues
forms dimers, exposing a signal that causes it to be on the !2-adrenergic receptor, phosphorylated by INSR,
transported into the nucleus. There, STAT5 induces the serve as nucleation points for SH2 domain–containing
expression (transcription) of specific genes essential proteins such as Grb2 (Fig. 12–19, left side). Activation
for erythrocyte maturation. This JAK-STAT system also of the MAPK ERK by insulin (see Fig. 12–15) is 5- to
operates in other signaling pathways, including that for 10-fold greater in the presence of the !2-adrenergic
the hormone leptin, described in detail in Chapter 23 receptor, presumably because of this cross talk. Signaling
(see Fig. 23–36). Activated JAK can also trigger, systems that use cAMP and Ca2" also show extensive
through Grb2, the MAPK cascade (Fig. 12–18b), which interaction; each second messenger affects the genera-
leads to altered expression of specific genes. tion and concentration of the other. One of the major
Src is another soluble protein Tyr kinase that asso- challenges of systems biology is to sort out the effects of
ciates with certain receptors when they bind their such interactions on the overall metabolic patterns in
ligands. Src was the first protein found to have the char- each tissue—a daunting task.
acteristic P –Tyr-binding domain that was subsequently
named the Src homology (SH2) domain. SUMMARY 12.3 Receptor Tyrosine Kinases
! The insulin receptor, INSR, is the prototype of
Cross Talk among Signaling Systems Is receptor enzymes with Tyr kinase activity. When
insulin binds, each "! unit of INSR phosphorylates
Common and Complex
the ! subunit of its partner, activating the
Although, for simplicity, we have treated individual sig- receptor’s Tyr kinase activity. The kinase catalyzes
naling pathways as separate sequences of events lead- the phosphorylation of Tyr residues on other
ing to separate metabolic consequences, there is in proteins, such as IRS-1.
fact extensive cross talk among signaling systems. The
! Phosphotyrosine residues in IRS-1 serve as binding
regulatory circuitry that governs metabolism is richly
interwoven and multilayered. We have discussed the sites for proteins with SH2 domains. Some of these
signaling pathways for insulin and epinephrine sepa- proteins, such as Grb2, have two or more protein-
rately, but they do not operate independently. Insulin binding domains and can serve as adaptors that
opposes the metabolic effects of epinephrine in most bring two proteins into proximity.
tissues, and activation of the insulin signaling pathway ! Sos bound to Grb2 catalyzes GDP-GTP exchange
directly attenuates signaling through the !-adrenergic on Ras (a small G protein), which in turn activates
signaling system. For example, the INSR kinase directly a MAPK cascade that ends with the
phosphorylates two Tyr residues in the cytoplasmic tail phosphorylation of target proteins in the cytosol
Insulin b-adrenergic
GPCR a receptor
b
P P
Y Y Y S S
P
P P
P FIGURE 12–19 Cross talk between the insulin receptor and the
P P P P P
!2-adrenergic receptor (or other GPCR). When INSR is activated
SHC Grb2
by insulin binding, its Tyr kinase directly phosphorylates the
Sos IRS-1 !2-adrenergic receptor (right side) on two Tyr residues (Tyr350 and
Ras
PKB Y Y S S Tyr364) near its carboxyl terminus, and indirectly (through activation
Raf-1 P P P P of protein kinase B (PKB); see Fig. 12–16) causes phosphorylation of
GPCR internalization, two Ser residues in the same region. The effect of these phosphory-
MEK reduced GPCR lations is internalization of the adrenergic receptor, reducing the
responsiveness response to the adrenergic stimulus. Alternatively (left side), INSR-
ERK Cytosol catalyzed phosphorylation of a GPCR (an adrenergic or other recep-
tor) on a carboxyl-terminal Tyr creates the point of nucleation for
Nucleus activating the MAPK cascade (see Fig. 12–15), with Grb2 serving as
Altered gene the adaptor protein. In this case, INSR has used the GPCR to
expression
enhance its own signaling.
12.4 Receptor Guanylyl Cyclases, cGMP, and Protein Kinase G 459
and nucleus. The result is specific metabolic ANF Guanylin and

changes and altered gene expression. receptor endotoxin receptors
! The enzyme PI3K, activated by interaction with Extracellular Ligand-binding

ligand-binding site
IRS-1, converts the membrane lipid PIP2 to PIP3, (receptor)
which becomes the point of nucleation for proteins domains ANF
in a second and third branch of insulin signaling.
! In the JAK-STAT signaling system, a soluble
protein Tyr kinase (JAK) is activated by
association with a receptor, and then
phosphorylates the transcription factor STAT, Intracellular
which enters the nucleus and alters the expression catalytic Fe
(cGMP-forming) Heme
of a set of genes. domains GTP cGMP
Soluble
! There are extensive interconnections among NO-activated
Membrane-spanning
signaling pathways, allowing integration and fine- guanylyl cyclases guanylyl cyclase
tuning of multiple hormonal effects. (a) (b)
FIGURE 12–20 Two types of guanylyl cyclase that participate in signal

12.4 Receptor Guanylyl Cyclases, cGMP, transduction. (a) One type is a homodimer with a single membrane-
and Protein Kinase G

spanning segment in each monomer, connecting the extracellular
ligand-binding domain and the intracellular guanylyl cyclase domain.
Guanylyl cyclases (Fig. 12–20) are receptor enzymes Receptors of this type are used to detect two extracellular ligands: atrial
natriuretic factor (ANF; receptors in cells of the renal collecting ducts
that, when activated, convert GTP to the second mes-
and vascular smooth muscle) and guanylin (peptide hormone produced
senger guanosine 3",5"-cyclic monophosphate
in the intestine, with receptors in intestinal epithelial cells). The guanylin
(cyclic GMP, cGMP):
receptor is also the target of a bacterial endotoxin that triggers severe
O diarrhea. (b) The other type is a soluble heme-containing enzyme that
is activated by intracellular nitric oxide (NO); this form is present in
N
HN many tissues, including smooth muscle of the heart and blood vessels.
N
O O O NH2 N
!
O P O P O P O CH2 O
O !
O !
O !
H H The catalytic and regulatory domains of this enzyme are
H H
in a single polypeptide (Mr ,80,000). Part of the regula-
OH OH tory domain fits snugly in the substrate-binding cleft.
GTP Binding of cGMP forces this pseudosubstrate out of the
binding site, opening the site to target proteins contain-
ing the PKG consensus sequence.
PPi Cyclic GMP carries different messages in different
tissues. In the kidney and intestine it triggers changes in
O ion transport and water retention; in cardiac muscle (a
N type of smooth muscle) it signals relaxation; in the brain
HN
it may be involved both in development and in adult
N brain function. Guanylyl cyclase in the kidney is acti-
NH2 N
5# vated by the peptide hormone atrial natriuretic fac-
O CH2 O
tor (ANF), which is released by cells in the cardiac
H H atrium when the heart is stretched by increased blood
H H volume. Carried in the blood to the kidney, ANF acti-
3#
vates guanylyl cyclase in cells of the collecting ducts
O P O OH
(Fig. 12–20a). The resulting rise in [cGMP] triggers
O !
increased renal excretion of Na" and consequently of
Guanosine 3#,5#-cyclic monophosphate water, driven by the change in osmotic pressure. Water
(cGMP)
loss reduces the blood volume, countering the stimulus
that initially led to ANF secretion. Vascular smooth
Many of the actions of cGMP in animals are mediated by muscle also has an ANF receptor–guanylyl cyclase; on
cGMP-dependent protein kinase, also called pro- binding to this receptor, ANF causes relaxation (vasodi-
tein kinase G (PKG). On activation by cGMP, PKG lation) of the blood vessels, which increases blood flow
phosphorylates Ser and Thr residues in target proteins. while decreasing blood pressure.
460 Biosignaling
A similar receptor guanylyl cyclase in the plasma accounting for the usefulness of this drug in the treat-
membrane of epithelial cells lining the intestine is acti- ment of erectile dysfunction.
vated by the peptide guanylin (Fig. 12–20a), which
regulates Cl! secretion in the intestine. This receptor is O
also the target of a heat-stable peptide endotoxin pro- O N
O HN
duced by Escherichia coli and other gram-negative N
S
bacteria. The elevation in [cGMP] caused by the endo- N
toxin increases Cl! secretion and consequently decreases N
reabsorption of water by the intestinal epithelium, pro- O
ducing diarrhea. N
A distinctly different type of guanylyl cyclase is a Sildenafil (Viagra) ■
cytosolic protein with a tightly associated heme group
(Fig. 12–20b), an enzyme activated by nitric oxide Cyclic GMP has another mode of action in the ver-
(NO). Nitric oxide is produced from arginine by Ca2"- tebrate eye: it causes ion-specific channels to open in
dependent NO synthase, present in many mammalian the retinal rod and cone cells. We return to this role of
tissues, and diffuses from its cell of origin into nearby cGMP in the discussion of vision in Section 12.10.
cells.
NH2 NH2 SUMMARY 12.4 Receptor Guanylyl Cyclases,
" NADPH NADP"
C NH2 C O cGMP, and Protein Kinase G
NH
O2
NH ! Several signals, including atrial natriuretic factor
Ca2" and guanylin, act through receptor enzymes with
(CH2)3 (CH2)3 " NO
NO synthase guanylyl cyclase activity. The cGMP so produced is
CH COO! CH COO! a second messenger that activates cGMP-
"
NH3
"
NH3 dependent protein kinase (PKG). This enzyme
Arginine Citrulline alters metabolism by phosphorylating specific
NO is sufficiently nonpolar to cross plasma membranes enzyme targets.
without a carrier. In the target cell, it binds to the heme ! Nitric oxide is a short-lived messenger that
group of guanylyl cyclase and activates cGMP produc- stimulates a soluble guanylyl cyclase, raising
tion. In the heart, cGMP-dependent protein kinase [cGMP] and stimulating PKG.
reduces the forcefulness of contractions by stimulating
the ion pump(s) that remove Ca2" from the cytosol.
NO-induced relaxation of cardiac muscle is the 12.5 Multivalent Adaptor Proteins
same response brought about by nitroglycerin
and other nitrovasodilators taken to relieve angina and Membrane Rafts
pectoris, the pain caused by contraction of a heart Two generalizations have emerged from studies of sig-
deprived of O2 because of blocked coronary arteries. naling systems such as those we have discussed so far:
Nitric oxide is unstable and its action is brief; within (1) protein kinases that phosphorylate Tyr, Ser, and
seconds of its formation, it undergoes oxidation to Thr residues are central to signaling, directly affecting
nitrite or nitrate. Nitrovasodilators produce long-lasting the activities of a large number of protein substrates by
relaxation of cardiac muscle because they break down phosphorylation, and (2) protein-protein interactions
over several hours, yielding a steady stream of NO. The brought about by the reversible phosphorylation of Tyr,
value of nitroglycerin as a treatment for angina was Ser, and Thr residues in signaling proteins create dock-
discovered serendipitously in factories producing nitro- ing sites for other proteins that bring about indirect
glycerin as an explosive in the 1860s. Workers with effects on proteins downstream in the signaling path-
angina reported that their condition was much improved way. In fact, many signaling proteins are multivalent—
during the workweek but worsened on weekends. The they can interact with several different proteins simul-
physicians treating these workers heard this story so taneously to form multiprotein signaling complexes. In
often that they made the connection, and a drug was this section we present a few examples to illustrate the
born. general principles of phosphorylation-dependent pro-
The effects of increased cGMP synthesis diminish tein interactions in signaling pathways.
after the stimulus ceases, because a specific phosphodi-
esterase (cGMP PDE) converts cGMP to the inactive
5#-GMP. Humans have several isoforms of cGMP PDE,
Protein Modules Bind Phosphorylated Tyr, Ser, or Thr
with different tissue distributions. The isoform in the Residues in Partner Proteins
blood vessels of the penis is inhibited by the drug The protein Grb2 in the insulin signaling pathway (Figs
sildenafil (Viagra), which therefore causes [cGMP] to 12–15 and 12–19) binds through its SH2 domain to
remain elevated once raised by an appropriate stimulus, other proteins that have exposed P –Tyr residues. The
12.5 Multivalent Adaptor Proteins and Membrane Rafts 461
human genome encodes at least 87 SH2-containing pro- partner proteins through the phosphorylated residue,
teins, many already known to participate in signaling. triggering a downstream process. An alphabet soup of
The P –Tyr residue is bound in a deep pocket in an SH2 domains that bind P –Ser or P –Thr residues has been
domain, with each of its phosphate oxygens participat- identified, and more are sure to be found. Each domain
ing in hydrogen bonding or electrostatic interactions; favors a certain sequence around the phosphorylated
the positive charges on two Arg residues figure promi- residue, so the domains represent families of highly
nently in the binding. Subtle differences in the structure specific recognition sites, able to bind to a specific sub-
of SH2 domains account for the specificities of the inter- set of phosphorylated proteins.
actions of SH2-containing proteins with various P –Tyr- In some cases, the region on a protein that binds
containing proteins. The SH2 domain typically interacts P –Tyr of a substrate protein is masked by its interac-
with a P –Tyr (which is assigned the index position 0) tion with a P –Tyr in the same protein. For example, the
and the next three residues toward the carboxyl termi- soluble protein Tyr kinase Src, when phosphorylated on
nus (designated !1, !2, !3). Some SH2 domains (Src, a critical Tyr residue, is rendered inactive; an SH2
Fyn, Hck, Nck) favor negatively charged residues in the domain needed to bind to the substrate protein instead
!1 and !2 positions; others (PLC!1, SHP2) have a long binds to the internal P –Tyr. When this P –Tyr residue
hydrophobic groove that binds to aliphatic residues in is hydrolyzed by a phosphoprotein phosphatase, the Tyr
positions !1 to !5. These differences define subclasses kinase activity of Src is activated (Fig. 12–22a). Simi-
of SH2 domains that have different partner specificities. larly, glycogen synthase kinase 3 (GSK3) is inactive
Phosphotyrosine-binding domains (PTB domains;
Fig. 12–21) are another binding partner for P –Tyr
proteins, but their critical sequences and three-dimen- (a) (b)
sional structure distinguish them from SH2 domains. SH3
The human genome encodes 24 proteins that contain
Pro Ser— OH
PTB domains, including IRS-1, which we have already
SH2
encountered in its role as an adaptor protein in insulin- P Tyr
Pro
signal transduction (Fig. 12–15). The P –Tyr binding
sites for SH2 and PTB domains on partner proteins are
created by Tyr kinases and eliminated by protein tyro- HO— Tyr Active; substrate Ser P
Active Active
sine phosphatases (PTPases). site positioned for
HO— Tyr phosphorylation Ser— OH site
Other signaling protein kinases, including PKA,
HO— Tyr
PKC, PKG, and members of the MAPK cascade, phos- Src Ser— OH GSK3
phorylate Ser or Thr residues in their target proteins,
which in some cases acquire the ability to interact with HO— Tyr Glycogen
synthase
SH3
Interacting
tyrosine Pro
Ser P
SH2
Autoinhibited
P
Tyr
Partner protein FIGURE 12–22 Mechanism of autoinhibition of Src and GSK3. (a) In
the active form of the Tyr kinase Src, an SH2 domain binds a P –Tyr in the
protein substrate, and an SH3 domain binds a proline-rich region of the
substrate, lining up the active site of the kinase with several target Tyr
residues in the substrate (top). When Src is phosphorylated on a specific
Tyr residue (bottom), the SH2 domain binds the internal P –Tyr instead
of the P –Tyr of the substrate, and the SH3 domain binds an internal
proline-rich region, preventing productive enzyme-substrate binding; the
enzyme is thus autoinhibited. (b) In the active form of glycogen synthase
FIGURE 12–21 Interaction of a PTB domain with a P –Tyr residue in a kinase 3 (GSK3), an internal P –Ser-binding domain is available to bind
partner protein. (PDB ID 1SHC) The PTB domain is represented as a P –Ser in its substrate (glycogen synthase), and thus to position the
blue surface contour. The P –Tyr residue of the partner protein (red) kinase to phosphorylate neighboring Ser residues (top). Phosphorylation
projects into a binding pocket in the PTB domain and is held firmly by of an internal Ser residue allows this internal kinase segment to occupy
multiple noncovalent interactions. the P –Ser-binding site, blocking substrate binding (bottom).
462 Biosignaling
when phosphorylated on a Ser residue in its autoinhibi- signaling. Many of the complexes include components
tory domain (Fig. 12–22b). Dephosphorylation of that with membrane-binding domains. Given the location of
domain frees the enzyme to bind (and then phosphory- so many signaling processes at the inner surface of the
late) its target proteins. plasma membrane, the molecules that must collide to
In addition to the three commonly phosphorylated produce the signaling response are effectively con-
residues in proteins, there is a fourth structure that fined to two-dimensional space—the membrane surface;
nucleates the formation of supramolecular complexes of collisions here are far more likely than in the three-
signaling proteins: the phosphorylated head group of dimensional space of the cytosol.
the membrane phosphatidylinositols. Many signaling In summary, a remarkable picture of signaling path-
proteins contain domains such as SH3 and PH (plextrin ways has emerged from studies of many signaling pro-
homology domain) that bind tightly to PIP3 protruding teins and their multiple binding domains. An initial sig-
from the inner leaflet of the plasma membrane. Wher- nal results in phosphorylation of the receptor or a target
ever the enzyme PI3K creates this head group (as it protein, triggering the assembly of large multiprotein
does in response to the insulin signal), proteins that complexes, held together on scaffolds with multivalent
bind it will congregate at the membrane surface. binding capacities. Some of these complexes contain
Most of the proteins involved in signaling at several protein kinases that activate each other in turn,
the plasma membrane have one or more protein- or producing a cascade of phosphorylation and a great
phospholipid-binding domains; many have three or amplification of the initial signal. The interactions
more, and thus are multivalent in their interactions with between cascade kinases are not left to the vagaries of
other signaling proteins. Figure 12–23 shows just a random collisions in three-dimensional space. In the
few of the multivalent proteins known to participate in MAPK cascade, for example, a scaffold protein, KSR,
Binding targets
Proline-rich protein
Adaptor SH3 SH2 SH3 Grb2 or membrane lipid PIP3
Tyr— P
Tyr— P
Adaptor PTB SH2 Shc PIP3
Phospholipids (Ca2+-dependent)
DNA
Kinase SH3 SH2 Tyr kinase Src
Transcriptional activation
Carboxyl-terminal domain
marking protein for attachment
Phosphatase SH2 SH2 Tyr phosphatase Shp2 of ubiquitin
Ras signaling SH2 SH3 SH2 PH C2 GTPase-activating RasGAP
Transcription DNA SH2 TA STAT
Signal regulation SH2 SOCS SOCS
Phospholipid second-
PH PLC PH SH2 SH2 SH3 PH PLC C2 PLCg
messenger signaling
FIGURE 12–23 Some binding modules of signaling proteins. Each protein indicate catalytic activities. The name of each protein is given at its car-
is represented by a line (with the amino terminus to the left); symbols boxyl-terminal end. These signaling proteins interact with phosphorylat-
indicate the location of conserved binding domains (with specificities as ed proteins or phospholipids in many permutations and combinations to
listed in the key; abbreviations are explained in the text); green boxes form integrated signaling complexes.
12.5 Multivalent Adaptor Proteins and Membrane Rafts 463
binds all three kinases (MAPK, MAPKK, and MAPKKK), see, then, that signaling occurs in protein circuits,
assuring their proximity and correct orientation and which are effectively hardwired from signal receptor to
even conferring allosteric properties on the interactions response effector and can be switched off instantly by the
among the kinases, which makes their serial phosphory- hydrolysis of a single upstream phosphate ester bond.
lation sensitive to very small stimuli (Fig. 12–24). The multivalency of signaling proteins allows for
Phosphotyrosine phosphatases remove the phos- the assembly of many different combinations of signal-
phate from P –Tyr residues, reversing the effect of phos- ing modules, each combination suited to particular sig-
phorylation. Some of these are receptorlike membrane nals, cell types, and metabolic circumstances, yielding
proteins, presumably controlled by extracellular factors diverse signaling circuits of extraordinary complexity.
not yet identified; other PTPases are soluble and contain
SH2 domains. In addition, animal cells have protein
P –Ser and P –Thr phosphatases, which reverse the Membrane Rafts and Caveolae May Segregate
effects of Ser- and Thr-specific protein kinases. We can Signaling Proteins
Membrane rafts (Chapter 11) are regions of the mem-
brane bilayer enriched in sphingolipids, sterols, and
Input Input
certain proteins, including many attached to the bilayer
by GPI anchors. The !-adrenergic receptor is segregated
in rafts that contain G proteins, adenylyl cyclase, PKA,
and a specific protein phosphatase, PP2, which together
P
P provide a highly integrated signaling unit. By segregat-
Raf
Raf P ing in a small region of the plasma membrane all of the
elements required for responding to and ending the
KSR KSR signal, the cell is able to produce a highly localized and
brief “puff” of second messenger.
MEK P MEK
Some RTKs (EGFR and PDGFR) seem to be local-
ized in rafts, and this sequestration is very probably
functionally significant. When cholesterol is removed
Erk P Erk from rafts by treatment of the membrane with cyclodex-
trin (which binds and removes cholesterol), the rafts
are disrupted and the RTK signaling pathways become
defective.
Strong output Weak output If an RTK in a raft is phosphorylated, and the only
(a) locally available PTPase that reverses this phosphoryla-
tion is in another raft, then dephosphorylation of the
RTK is slowed or prevented. Interactions between adap-
KSR mutant lacking tor proteins might be strong enough to recruit into a raft
Erk phosphorylation sites a signaling protein not usually located there, or might
MEK phosphorylation
even be strong enough to pull receptors out of a raft.

For example, the EGFR in isolated fibroblasts is usually
Wild type concentrated in specialized rafts called caveolae (see
Fig. 11–22), but treatment with EGF causes the recep-
tor to leave the raft. This migration depends on the
receptor’s protein kinase activity; mutant receptors
lacking this activity remain in the raft during treatment
with EGF. Caveolin, an integral membrane protein
Time
localized in caveolae, is phosphorylated on Tyr residues
(b) in response to insulin, and the now-activated EGFR may
FIGURE 12–24 A scaffold protein from yeast that organizes and regu- be able to draw its binding partners into the raft. Spatial
lates a protein kinase cascade. (a) The scaffold protein KSR has bind- segregation of signaling proteins in rafts adds yet another
ing sites for all three of the kinases in the Raf/MEK/Erk cascade. By dimension to the already complex processes initiated by
binding all three in appropriate orientations, the scaffold makes interac- extracellular signals.
tions among the proteins rapid and efficient. When Erk has been activated
(left), it phosphorylates the binding site for Raf (right), forcing a con-
SUMMARY 12.5 Multivalent Adaptor Proteins and
formational change that displaces Raf and thereby prevents the phos-
phorylation of MEK. The result of this feedback regulation is that MEK
Membrane Rafts
phosphorylation is temporary. (b) In yeast cells with mutant KSR lacking ! Many signaling proteins have domains that bind
the phosphorylation sites (red curve), no feedback occurs, producing a phosphorylated Tyr, Ser, or Thr residues in other
different time course of signaling. proteins; the binding specificity for each domain is
464 Biosignaling
determined by sequences that adjoin the (a)

phosphorylated residue in the substrate. The electrogenic Na+K+
! SH2 and PTB domains bind to proteins containing ATPase establishes the
membrane potential. 3 Na"
P –Tyr residues; other domains bind P –Ser and
P –Thr residues in various contexts. Membrane potential $ Na"K" ATPase
!
!50 to !70 mV
SH3 and PH domains bind the membrane Plasma "
membrane " "
phospholipid PIP3.
!
" "
Many signaling proteins are multivalent, with
several different binding modules. By combining " ! "
the substrate specificities of various protein ! !
kinases with the specificities of domains that bind ! !
" "
phosphorylated Ser, Thr, or Tyr residues, and with
phosphatases that can rapidly inactivate a ! 2 K" ATP ADP " Pi !
signaling pathway, cells create a large number of " ! ! "
multiprotein signaling complexes.
! Membrane rafts and caveolae sequester groups of ! !
"
signaling proteins in small regions of the plasma "
! Low Low High !
membrane, enhancing their interactions and
making signaling more efficient. " ! ! "
" "
12.6 Gated Ion Channels (b)

" "
[Cl!] [Ca2"] [K"] [Na"]

Ion Channels Underlie Electrical Signaling Ions tend to move
down their electro- High High Low High
in Excitable Cells chemical gradient
across the polarized
Certain cells in multicellular organisms are “excitable”: membrane.
they can detect an external signal, convert it into an
electrical signal (specifically, a change in membrane FIGURE 12–25 Transmembrane electrical potential. (a) The electrogenic
potential), and pass it on. Excitable cells play central Na"K" ATPase produces a transmembrane electrical potential of about
roles in nerve conduction, muscle contraction, hormone !60 mV (inside negative). (b) Blue arrows show the direction in which
secretion, sensory processes, and learning and memory. ions tend to move spontaneously across the plasma membrane in an
The excitability of sensory cells, neurons, and myocytes animal cell, driven by the combination of chemical and electrical gradi-
depends on ion channels, signal transducers that pro- ents. The chemical gradient drives Na" and Ca2" inward (producing
vide a regulated path for the movement of inorganic depolarization) and K" outward (producing hyperpolarization). The
electrical gradient drives Cl! outward, against its concentration gradient
ions such as Na", K", Ca2", and Cl! across the plasma
(producing depolarization).
membrane in response to various stimuli. Recall from
Chapter 11 that these ion channels are “gated”: they may
be open or closed, depending on whether the associated charged ion such as Cl!, depolarizes the membrane
receptor has been activated by the binding of its specific and brings Vm closer to zero. Conversely, efflux of K"
ligand (a neurotransmitter, for example) or by a change hyperpolarizes the membrane and Vm becomes more
in the transmembrane electrical potential, Vm. The Na"K" negative. These ion fluxes through channels are pas-
ATPase is electrogenic; it creates a charge imbalance sive, in contrast to active transport by the Na"K"
across the plasma membrane by carrying 3 Na" out of ATPase.
the cell for every 2 K" carried in (Fig. 12–25a), making The direction of spontaneous ion flow across a
the inside negative relative to the outside. The mem- polarized membrane is dictated by the electrochemical
brane is said to be polarized. potential of that ion across the membrane, which has
two components: the difference in concentration (C) of
KEY CONVENTION: Vm is negative when the inside of the the ion on the two sides of the membrane, and the dif-
cell is negative relative to the outside. For a typical
ference in electrical potential, typically expressed in
animal cell, Vm & !50 to !70 mV. ■
millivolts. The force (%G) that causes a cation (say,
Na") to pass spontaneously inward through an ion
Because ion channels generally allow passage of
channel is a function of the ratio of its concentrations on
either anions or cations but not both, ion flux through
the two sides of the membrane (Cin/Cout) and of the dif-
a channel causes a redistribution of charge on the two
ference in electrical potential (Vm or %%):
sides of the membrane, changing Vm. Influx of a posi-
tively charged ion such as Na", or efflux of a negatively %G & RT ln (Cin/Cout) " Z Vm (12–1)
12.6 Gated Ion Channels 465
where R is the gas constant, T the absolute tempera- Ca2", or Cl! open, the membrane potential moves
ture, Z the charge on the ion, and the Faraday con- toward the E for that ion. The precisely timed opening
stant. (Note that the sign of the charge on the ion and closing of ion channels and the resulting transient
determines the sign of the second term in Eqn 12–1.) changes in membrane potential underlie the electrical
In a typical neuron or myocyte, the concentrations of signaling by which the nervous system stimulates the
Na", K", Ca2", and Cl! in the cytosol are very different skeletal muscles to contract, the heart to beat, or secre-
from those in the extracellular fluid (Table 12–6). tory cells to release their contents. Moreover, many
Given these concentration differences, the resting Vm hormones exert their effects by altering the membrane
of about !60 mV, and the relationship shown in Equa- potential of their target cells. These mechanisms are not
tion 12–1, the opening of a Na" or Ca2" channel will limited to animals; ion channels play important roles in
result in a spontaneous inward flow of Na" or Ca2" (and the responses of bacteria, protists, and plants to envi-
depolarization), whereas opening of a K" channel will ronmental signals.
result in a spontaneous outward flux of K" (and hyper- To illustrate the action of ion channels in cell-to-cell
polarization) (Fig. 12–25b). In this case, K" moves signaling, we describe the mechanisms by which a neu-
outward, against the electrical gradient, because the ron passes a signal along its length and across a synapse
large concentration difference inside and outside the to the next neuron (or to a myocyte) in a cellular cir-
cell produces a more powerful, outward chemical force cuit, using acetylcholine as the neurotransmitter.
on the ion.
A given ionic species continues to flow through a
channel only as long as the combination of concentra-
Voltage-Gated Ion Channels Produce
tion gradient and electrical potential provides a driving Neuronal Action Potentials
force. For example, as Na" flows down its concentra- Signaling in the nervous system is accomplished by
tion gradient, it depolarizes the membrane. When the networks of neurons, specialized cells that carry an
membrane potential reaches "70 mV, the effect of this electrical impulse (action potential) from one end of
membrane potential (resistance to further entry of the cell (the cell body) through an elongated cytoplas-
Na") exactly equals the effect of the [Na"] gradient mic extension (the axon). The electrical signal triggers
(promotion of Na" flow inward). At this equilibrium release of neurotransmitter molecules at the synapse,
potential (E), the driving force (%G) tending to move a carrying the signal to the next cell in the circuit. Three
Na" ion is zero. The equilibrium potential is different types of voltage-gated ion channels are essential to
for each ionic species, because the concentration gradi- this signaling mechanism. Along the entire length of the
ents differ. axon are voltage-gated Na! channels (Fig. 12–26),
The number of ions that must flow to produce a which are closed when the membrane is at rest (Vm &
physiologically significant change in the membrane !60 mV) but open briefly when the membrane is depo-
potential is negligible relative to the concentrations of larized locally in response to acetylcholine (or some
Na", K", and Cl! in cells and extracellular fluid, so the other neurotransmitter). Also distributed along the
ion fluxes that occur during signaling in excitable cells axon are voltage-gated K1 channels, which open, a
have essentially no effect on the concentrations of split second later, in response to the depolarization
these ions. With Ca2", the situation is different; when nearby Na" channels open. The depolarizing
because the intracellular [Ca2"] is generally very low flow of Na" into the axon (influx) is thus rapidly coun-
(,10!7 M), inward flow of Ca2" can significantly alter tered by a repolarizing flow of K" out (efflux). At the
the cytosolic [Ca2"]. distal end of the axon are voltage-gated Ca21 chan-
The membrane potential of a cell at a given time is nels, which open when the wave of depolarization
the result of the types and numbers of ion channels (step 1) and repolarization (step 2) caused by the
open at that instant. In most cells at rest, more K" chan- activity of Na" and K" channels arrives, triggering
nels than Na", Cl!, or Ca2" channels are open and thus release of the neurotransmitter acetylcholine—which
the resting potential is closer to the E for K" (–98 mV) carries the signal to another neuron (fire an action
than that for any other ion. When channels for Na", potential!) or to a muscle fiber (contract!).
TABLE 12–6 Ion Concentrations in Cells and Extracellular Fluids (mM)

K1 Na1 Ca21 Cl2
Cell type In Out In Out In Out In Out
Squid axon 400 20 50 440 '0.4 10 40–150 560
Frog muscle 124 2.3 10.4 109 $0.1 2.1 1.5 78
466 Biosignaling
! " Axon of
Voltage- FIGURE 12–26 Role of voltage-gated and ligand-gated ion channels in
gated K! neural transmission. Initially, the plasma membrane of the presynaptic
! " presynaptic ! channel
! " neuron neuron is polarized (inside negative) through the action of the electro-
! "
Voltage- ! K! genic Na!K! ATPase, which pumps out 3 Na! for every 2 K! pumped in
! "
gated Na! (see Fig. 12–25). 1 A stimulus to this neuron (not shown) causes an
! " Depolarization Repolarization
channel 1 2
! " action potential to move along the axon (blue arrow), away from the cell
!
Na! K! body. The opening of a voltage-gated Na! channel allows Na! entry,
!
" and the resulting local depolarization causes the adjacent Na! channel
to open, and so on. The directionality of movement of the action poten-
Na! tial is ensured by the brief refractory period that follows the opening of
!
" each voltage-gated Na! channel. 2 A split second after the action
3 potential passes a point in the axon, voltage-operated K! channels open,
Ca2!
allowing K! exit that brings about repolarization of the membrane (red
Secretory
Voltage- ! " arrow), to make it ready for the next action potential. (For clarity, Na!
vesicles
gated Ca2! containing channels and K! channels are drawn on opposite sides of the axon; both
channel acetylcholine types of channels are uniformly distributed in the axonal membrane.) 3
! "
! " When the wave of depolarization reaches the axon tip, voltage-gated
! " Synaptic Ca2! channels open, allowing Ca2! entry. 4 The resulting increase in
! " cleft internal [Ca2!] triggers exocytic release of the neurotransmitter acetyl-
choline into the synaptic cleft. 5 Acetylcholine binds to a receptor on
4
the postsynaptic neuron (or myocyte), causing its ligand-gated ion
Ca2! channel to open. 6 Extracellular Na! and Ca2! enter through this chan-
nel, depolarizing the postsynaptic cell. The electrical signal has thus
Na!,Ca2!
passed to the cell body of the postsynaptic neuron (or myocyte) and will
move along its axon to a third neuron (or a myocyte) by this same
5 Cell body of sequence of events.
postsynaptic
6 neuron
Acetylcholine
receptor ion unidirectional wave of depolarization—the action
channels potential—sweeps from the nerve cell body toward the
! end of the axon.
When the wave of depolarization reaches the volt-
Na! !
! age-gated Ca2! channels, they open (step 3), and
Na! Action ! Ca2! enters from the extracellular space. The rise in
! K!
potential
cytoplasmic [Ca2!] then triggers release of acetylcho-
! " line by exocytosis into the synaptic cleft (step 4).
! " !
" Acetylcholine diffuses to the postsynaptic cell (anoth-
! "
! " " !
er neuron or a myocyte), where it binds to acetylcho-
line receptors and triggers depolarization. Thus the
message is passed to the next cell in the circuit. We
see, then, that gated ion channels convey signals in
The voltage-gated Na! channels are very selective either of two ways: by changing the cytoplasmic con-
for Na! over other cations (by a factor of 100 or more) centration of an ion (such as Ca2!), which then serves
and have a very high flux rate ("107 ions/s). After as an intracellular second messenger, or by changing
being opened—activated—by a reduction in transmem- Vm and affecting other membrane proteins that are
brane electrical potential, a Na! channel undergoes sensitive to Vm. The passage of an electrical signal
very rapid inactivation—within milliseconds, the chan- through one neuron and on to the next illustrates both
nel closes and remains inactive for many milliseconds. types of mechanism.
As voltage-gated K! channels open in response to the We discussed the structure and mechanism of
depolarization induced by the opening of Na! channels voltage-gated K! channels in some detail in Section 11.3
(step 1 in Fig. 12–26), the resulting efflux of K! repo- (see Figs 11–47 and 11–48). Here we take a closer look
larizes the membrane locally (it reestablishes the at Na! channels. The essential component of a Na! chan-
inside-negative membrane potential; step 2). A brief nel is a single, large polypeptide (1,840 amino acid
pulse of depolarization thus traverses the axon as local residues) organized into four domains clustered around
depolarization triggers the brief opening of neighbor- a central channel (Fig. 12–27a, b), providing a path
ing Na! channels, then K! channels. The short refrac- for Na! through the membrane. The path is made
tory period that follows the opening of each Na! chan- Na!-specific by a “pore region” composed of the seg-
nel, during which it cannot open again, ensures that a ments between transmembrane helices 5 and 6 of each
Selectivity FIGURE 12–27 Voltage-gated Na1 channels of neurons. Sodium chan-

Domain: I II III IV filter nels of different tissues and organisms have a variety of subunits, but
(pore
region) only the principal subunit (!) is essential. (a) The ! subunit is a large
Outside
protein with four homologous domains (I to IV, shown spread out here
to illustrate the parts), each containing six transmembrane helices (1
to 6). Helix 4 in each domain (blue) is the voltage sensor; helix 6
Inside 1 2345 6 (orange) is thought to be the activation gate. The segments between
Activation
" helices 5 and 6, the pore region (red), form the selectivity filter, and
NH3 !OOC gate
the segment connecting domains III and IV (green) is the inactivation
(a) Inactivation gate Voltage sensor
gate. (b) (PDB ID 3RW0) Structure of a voltage-gated Na" channel
(from the bacterium Arcobacter butzleri, but probably similar to chan-
nels of vertebrate neurons), with a funnel-shaped opening on the
Extracellular extracellular side, an ion selectivity filter, a central aqueous cavity, and
funnel the activation domain on the cytoplasmic side. (c) A schematic view
Selectivity of the Na" channel. The four domains are wrapped about a central
filter transmembrane channel lined with polar amino acid residues. The four
pore regions (red) come together near the extracellular surface to
Central form the selectivity filter, which is conserved in all Na" channels. The
cavity filter gives the channel its ability to discriminate between Na" and
other ions of similar size. The inactivation gate (green) closes (dotted
lines) soon after the activation gate opens. (d) The voltage-sensing
Activation
mechanism involves movement of helix 4 (blue) perpendicular to the
gate
(b) plane of the membrane in response to a change in transmembrane
potential. As shown at the top, the strong positive charge on helix 4
allows it to be pulled inward in response to the inside-negative mem-
Voltage sensor
III brane potential (Vm). Depolarization lessens this pull, and helix 4 relaxes
4 Selectivity by moving outward (bottom). This movement is communicated to the
2 3 filter (pore)
1 activation gate (orange), inducing conformational changes that open
5 6
the channel.
IV II
Activation gate domain, which fold into the channel. Helix 4 of each
domain has a high density of positively charged Arg
residues; this segment is believed to move within the
membrane in response to changes in the transmembrane
voltage, from the resting potential of about !60 mV to
about "30 mV. The movement of helix 4 triggers open-
(c) Tether I Inactivation gate (open) ing of the channel, and this is the basis for the voltage
gating (Fig. 12–27c).
Inactivation of the channel (during the refractory
Outside Activation gate
""""" " " period) is thought to occur by a ball-and-chain mecha-
nism. A protein domain on the cytoplasmic surface of
the Na" channel, the inactivation gate (the ball), is
"
tethered to the channel by a short segment of the
polypeptide (the chain; Fig. 12–27b). This domain is
free to move about when the channel is closed, but
!!!!! ! !
Voltage sensor when it opens, a site on the inner face of the channel
Inside Membrane polarized,
channel closed becomes available for the tethered ball to bind, block-
ing the channel. The length of the tether seems to
Aqueous ion channel
Na" determine how long an ion channel stays open: the
Outside
longer the tether, the longer the open period. Other
gated ion channels may be inactivated by a similar
mechanism.
The Acetylcholine Receptor Is a

Inside
Na"
Ligand-Gated Ion Channel
Membrane depolarized, The nicotinic acetylcholine receptor mediates the
(d) channel open passage of the signal from an electrically excited neuron
468 Biosignaling
at some types of synapses and at neuromuscular junc- Neurons Have Receptor Channels That Respond
tions (between motor neuron and muscle fiber), trig-
gering muscle contraction. (Nicotinic acetylcholine
to Different Neurotransmitters
receptors were originally distinguished from musca- Animal cells, especially those of the nervous system,
rinic acetylcholine receptors by the sensitivity of the contain a variety of ion channels gated by ligands,
former to nicotine, the latter to the mushroom alka- voltage, or both. Receptors that are themselves ion
loid muscarine. They are structurally and functionally channels are classified as ionotropic to distinguish
different.) Acetylcholine released by the presynaptic them from receptors that generate a second messen-
neuron or motor neuron diffuses a few micrometers ger (metabotropic receptors). We have so far focused
to the plasma membrane of the postsynaptic neuron on acetylcholine as neurotransmitter, but there are
or myocyte, where it binds to the acetylcholine many others. 5-Hydroxytryptamine (serotonin), glu-
receptor. This forces a conformational change in the tamate, and glycine all can act through receptor chan-
receptor, causing its ion channel to open. The result- nels that are structurally related to the acetylcholine
ing inward movement of cations depolarizes the receptor. Serotonin and glutamate trigger the opening
plasma membrane. In a muscle fiber, this triggers of cation (K", Na", Ca2") channels, whereas glycine
contraction. The acetylcholine receptor allows ready opens Cl!-specific channels. Cation and anion chan-
passage of Na", Ca2", and K" ions, but other cations nels are distinguished by subtle differences in the
and all anions are unable to pass. Movement of Na" amino acid residues that line the hydrophilic channel.
through an acetylcholine receptor ion channel is Cation channels have negatively charged Glu and Asp
unsaturable (its rate is linear with respect to extra- side chains at crucial positions. When a few of these
cellular [Na"]) and very fast—about 2 ( 107 ions/s acidic residues are experimentally replaced with basic
under physiological conditions. residues, the cation channel is converted to an anion
channel.
Depending on which ion passes through a chan-
O
CH3 nel, binding of the ligand (neurotransmitter) for that
H3C C
"
channel results in either depolarization or hyperpo-
O CH2 CH2 N CH3
larization of the target cell. A single neuron normally
CH3 receives input from many other neurons, each releas-
Acetylcholine ing its own characteristic neurotransmitter with its
characteristic depolarizing or hyperpolarizing effect.
Like other gated ion channels, the acetylcholine The target cell’s Vm therefore reflects the integrated
receptor opens in response to stimulation by its sig- input (Fig. 12–1e) from multiple neurons. The cell
nal molecule and has an intrinsic timing mechanism responds with an action potential only if the inte-
that closes the gate milliseconds later. Thus the ace- grated input adds up to a net depolarization of suffi-
tylcholine signal is transient—as we have seen, an cient size.
essential feature of electrical signal conduction. We The receptor channels for acetylcholine, glycine,
understand the structural changes underlying gating glutamate, and #-aminobutyric acid (GABA) are
in the acetylcholine receptor, but not the exact gated by extracellular ligands. Intracellular second
mechanism of “desensitization,” in which the gate messengers—such as cAMP, cGMP, IP3, Ca2", and
remains closed even in the continued presence of ATP—regulate ion channels of another class, which, as
acetylcholine. we shall see in Section 12.10, participate in the sensory
The nicotinic acetylcholine receptor has five sub- transductions of vision, olfaction, and gustation.
units ("2!#&), each having four transmembrane helical
segments (M1 to M4) (Fig. 12–28). The five subunits,
which are related in sequence and tertiary structure, Toxins Target Ion Channels
surround a central pore, which is lined with their M2 Many of the most potent toxins found in nature act on
helices. The pore is about 20 Å wide in the parts of the ion channels. For example, dendrotoxin (from the
channel that protrude on the cytoplasmic and extracel- black mamba snake) blocks the action of voltage-gated
lular surfaces, but narrows as it passes through the lipid K" channels, tetrodotoxin (produced by puffer fish)
bilayer. Near the center of the bilayer is a ring of bulky acts on voltage-gated Na" channels, and cobrotoxin
hydrophobic side chains of Leu residues in the M2 heli- disables acetylcholine receptor ion channels. Why, in
ces, positioned so close together that they prevent ions the course of evolution, have ion channels become the
from passing through the channel (Fig. 12–28d). Bind- preferred target of toxins, rather than some critical
ing of acetylcholine to sites on each " subunit forces all metabolic target such as an enzyme essential in energy
M2 helices to rotate slightly, moving the bulky Leu resi- metabolism?
dues aside and replacing them with smaller, polar resi- Ion channels are extraordinary amplifiers; opening
dues. The widening of the pore allows passage of ions of a single channel can allow the flow of 10 million ions
(Na" and Ca2"). per second. Consequently, relatively few molecules of
(a) Subunit folds into four (b) M2 amphipathic helices (c) Acetylcholine
transmembrane a helices surround channel binding sites
+ COO! Outside
NH3 g a a
g
b
M4 a d b a
d
M1 M3
M2
M2
Inside
M3-4 helices M2 helices

(d) Closed Open (e)
2 acetylcholine
Acetylcholine
binding sites
Bulky, hydrophobic Binding of two M2 helices now have

Leu side chains of acetylcholine molecules smaller, polar residues
M2 helices close causes twisting of the lining the channel.
the channel. M2 helices.
FIGURE 12–28 The acetylcholine receptor ion channel. (a) Each of the shows five Leu side chains (yellow), one from each M2 helix, protruding
five homologous subunits (!2"#$) has four transmembrane helices, M1 into the channel and constricting it to a diameter too small to allow pas-
to M4. The M2 helices are amphipathic; the others have mainly hydro- sage of Ca2!, Na!, or K!. When both acetylcholine receptor sites (one
phobic residues. (b) The five subunits are arranged around a central on each ! subunit) are occupied, a conformational change occurs. As
transmembrane channel, which is lined with the polar sides of the M2 the M2 helices twist slightly, the five Leu residues rotate away from
helices. At the top and bottom of the channel are rings of negatively the channel and are replaced by smaller, polar residues (blue). This
charged amino acid residues. (c) A molecular model of the acetylcholine gating mechanism opens the channel, allowing the passage of Ca2!, Na!,
receptor, based on x-ray structure determination of a related protein or K!. (e) Molecular model of the acetylcholine receptor viewed perpen-
(the acetylcholine-binding protein from a mollusk; PDB ID 1UV6). (d) This dicular to the membrane, showing the small central pore that allows ion
cartoon view of a cross section through the center of the M2 helices passage.
an ion channel protein are needed per neuron for sig- along the axon as a wave of depolarization
naling functions. This means that a relatively small (Na! influx) followed by repolarization
number of toxin molecules with high affinity for ion (K! efflux).
channels, acting from outside the cell, can have a very ! The gating mechanism for voltage-sensitive
pronounced effect on neurosignaling throughout the channels involves the movement,
body. A comparable effect by way of a metabolic perpendicular to the plane of the membrane,
enzyme, typically present in cells at much higher con- of a transmembrane peptide with a high charge
centrations than ion channels, would require far more density, due to the presence of Arg or other
copies of the toxin molecule. charged residues.
! Arrival of an action potential at the distal end of a
SUMMARY 12.6 Gated Ion Channels
presynaptic neuron triggers neurotransmitter
! Ion channels gated by membrane potential or release. The neurotransmitter (acetylcholine, for
ligands are central to signaling in neurons and example) diffuses to the postsynaptic neuron (or
other cells. the myocyte, at a neuromuscular junction), binds
! The voltage-gated Na! and K! channels of to specific receptors in the plasma membrane, and
neuronal membranes carry the action potential triggers a change in Vm.
470 Biosignaling
! The acetylcholine receptor of neurons and myocytes The extracellular ligands that interact with integrins
is a ligand-gated ion channel; acetylcholine binding include collagen, fibrinogen, fibronectin, and many
triggers a conformational change that opens the other proteins that have the sequence recognized by
channel to Na" and Ca2" ions. integrins: !Arg–Gly–Asp– (RGD, using single-letter
! Neurotoxins produced by many organisms attack amino acid abbreviations). The short, cytoplasmic
neuronal ion channels, and are therefore fast- extensions of the " and ! subunits interact with cyto-
acting and deadly. skeletal proteins just beneath the plasma membrane—
talin, "-actinin, vinculin, paxillin, and others—modulating
the assembly of actin-based cytoskeletal structures. The
dual association of integrins with the extracellular
12.7 Integrins: Bidirectional Cell matrix and the cytoskeleton allows the cell to integrate
Adhesion Receptors information about its extracellular and intracellular
environments and to coordinate cytoskeletal positioning
Integrins are proteins of the plasma membrane that with extracellular adhesion sites. In this capacity, inte-
mediate the adhesion of cells to each other and to the grins govern the shape, motility, polarity, and differen-
extracellular matrix, and carry signals in both direc- tiation of many cell types. In “outside-in” signaling, the
tions across the membrane (Fig. 12–29). The mam- extracellular domains of an integrin undergo dramatic,
malian genome encodes 18 different " subunits and 8 global conformational changes when ligand binds at a
different ! subunits, which are found in a range of site many angstroms from the transmembrane helices.
combinations with various ligand-binding specificities These changes somehow alter the dispositions of the
in various tissues. Each of the 24 different integrins cytoplasmic tails of the " and ! subunits, changing their
found thus far seems to have a unique function. interactions with intracellular proteins and thereby
Because they can inform cells about the extracellular conducting the signal inward.
neighborhood, integrins play crucial roles in processes The conformation and adhesiveness of integrin
that require selective cell-cell interactions, such as extracellular domains are also dramatically altered by
embryonic development, blood clotting, immune cell “inside-out” signaling initiated by signals from inside
function, normal differentiation, and tumor growth the cell. In one conformation, the extracellular domains
and metastasis. have no affinity for the proteins of the extracellular
Outside 1 Collagen in 4 Responses: cell adhesion

extracellular and migration, assembly
Collagen matrix of extracellular matrix
triggers
outside-in
signal
b
a
Integrin Inactive
integrin
Inside
Cytoskeleton Talin
2 Responses: setting of cell polarity, 3 Talin in cytoskeleton triggers
survival and proliferation, changes inside-out signal
in cytoskeleton and gene
expression
FIGURE 12–29 Two-way signaling by integrins. All integrins have one " lular domain and moves the cytosolic tails of the " and ! subunits apart
and one ! subunit, each with a short cytoplasmic extension, a single (left), altering their interactions with intracellular proteins such as talin,
transmembrane helix, and a large extracellular domain with the ligand- which in turn connect the integrin to actin filaments in the cytoskeleton. In
binding site. The ! subunit (purple) is rich in Cys residues and has exten- inside-out signaling, contact of the cytosolic domain with talin produces a
sive intrachain disulfide bonding. The " subunit (pink) in many integrins dramatic unbending of the extracellular domain (right) and an increase in
has several binding sites for divalent cations such as Ca2", which are its affinity for extracellular binding partners, allowing interactions with
intrinsic to the ligand-binding activity. In its inactive state, integrin’s extra- extracellular proteins or proteoglycans and changing the cell’s adhesion to
cellular domain is folded upon itself (center). Contact with an extracellular the extracellular matrix. Protein ligands in the extracellular matrix have the
ligand (collagen or heparan sulfate, for example) straightens the extracel- RGD sequence recognized by integrins.
12.8 Regulation of Transcription by Nuclear Hormone Receptors 471
matrix, but signals from the cell can favor another con- ! The active and inactive forms of an integrin differ
formation in which integrins adhere tightly to extracel- in the conformation of their extracellular domains.
lular proteins (Fig. 12–29). Intracellular events and signals can interconvert
Regulation of adhesiveness is central to leuko- the active and inactive forms.
cyte homing to the site of an infection (see Fig. ! Integrins mediate various aspects of the immune
7–32), interactions between immune cells, and phago- response, blood clotting, and angiogenesis, and
cytosis by macrophages. During an immune response, they play a role in tumor metastasis.
for example, leukocyte integrins are activated (expos-
ing their extracellular ligand-binding sites) from
inside the cell via a signaling pathway triggered by
cytokines (extracellular developmental signals). Thus 12.8 Regulation of Transcription
activated, the integrins can mediate the attachment
of leukocytes to other immune cells or can target
by Nuclear Hormone Receptors
cells for phagocytosis. Mutation in an integrin gene The steroid, retinoic acid (retinoid), and thyroid hor-
encoding the ! subunit known as CD18 is the cause mones form a large group of hormones (receptor
of leukocyte adhesion deficiency, a rare human ligands) that exert at least part of their effects by a
genetic disease in which leukocytes fail to pass out of mechanism fundamentally different from that of other
blood vessels to reach sites of infection. Infants with hormones: they act in the nucleus to alter gene expres-
a severe defect in CD18 commonly die of infections sion. We discuss their mode of action in detail in
before the age of two. Chapter 28, along with other mechanisms for regulating
An integrin specific to platelets ("IIb!3) is involved gene expression. Here we give a brief overview.
in both normal and pathological blood clotting. Local Steroid hormones (estrogen, progesterone, and
damage to blood vessels at a site of injury exposes cortisol, for example), too hydrophobic to dissolve
high-affinity binding sites (RGD sequences in throm- readily in the blood, are transported on specific car-
bin and collagen, for example) for the integrins of rier proteins from their point of release to their target
platelets, which attach themselves to the lesion, to tissues. In target cells, these hormones pass through
other platelets, and to the clotting protein fibrinogen, the plasma membrane by simple diffusion and bind to
leading to clot formation that prevents further bleed- specific receptor proteins in the nucleus (Fig. 12–30).
ing. Mutations in the " or ! subunit of platelet integrin Steroid hormone receptors with no bound ligand
"IIb!3 lead to a bleeding disorder known as Glanzmann (aporeceptors) often act to suppress the transcription
thrombasthenia, in which individuals bleed excessively of target genes. Hormone binding triggers changes in
after a relatively minor injury. Overly effective blood the conformation of a receptor protein so that it
coagulation is also undesirable. Dysregulation of plate- becomes capable of interacting with specific regula-
let adhesion can lead to pathological blood clot forma- tory sequences in DNA called hormone response
tion, resulting in blockage of the arteries that supply elements (HREs), thus altering gene expression
blood to the heart and brain and increasing the risk of (see Fig. 28–33). The bound receptor-hormone com-
heart attack and stroke. Drugs such as tirofiban and plex enhances the expression of specific genes adja-
eptifibatide that block the external ligand-binding sites cent to HREs, with the help of several other proteins
of platelet integrin reduce clot formation and are use- essential for transcription. Hours or days are required
ful in treating and preventing heart attacks and for these regulators to have their full effect—the time
strokes. required for the changes in RNA synthesis and subse-
When tumors metastasize, tumor cells lose their quent protein synthesis to become evident in altered
adhesion to the originating tissue and invade new loca- metabolism.
tions. Both the changes in tumor cell adhesion and the The specificity of the steroid-receptor interaction
development of new blood vessels (angiogenesis) to is exploited in the use of the drug tamoxifen to
support the tumor at a new location are modulated by treat breast cancer. In some types of breast cancer, divi-
specific integrins. These proteins are therefore poten- sion of the cancerous cells depends on the continued
tial targets for drugs that suppress the migration and presence of estrogen. Tamoxifen is an estrogen antago-
relocation of tumor cells. ■ nist; it competes with estrogen for binding to the estro-
gen receptor, but the tamoxifen-receptor complex has
little or no effect on gene expression. Consequently,
SUMMARY 12.7 Integrins: Bidirectional Cell tamoxifen administered after surgery or during chemo-
Adhesion Receptors therapy for hormone-dependent breast cancer slows or
! Integrins are a family of dimeric ("!) plasma stops the growth of remaining cancerous cells. Another
membrane receptors that interact with steroid analog, the drug mifepristone (RU486), binds
extracellular macromolecules and the to the progesterone receptor and blocks hormone
cytoskeleton, carrying signals in and out of actions essential to implantation of the fertilized ovum in
the cell. the uterus, and thus functions as a contraceptive.
472 Biosignaling
Serum binding protein Plasma

with bound hormone membrane
Hormone
Cytosol Hormone, carried to the target tissue on

1 1
serum binding proteins, di uses across the
plasma membrane and binds to its specific
receptor protein in the nucleus.
Nucleus
Nuclear
receptor
RNA
2 2 Hormone binding changes the conformation
Altered cell polymerase
of the receptor; it forms homo- or hetero-
function
dimers with other hormone-receptor
HRE complexes and binds to specific regulatory
DNA
regions called hormone response elements
New (HREs) in the DNA adjacent to specific genes.
protein Gene
transcription 3 3 Receptor attracts coactivator or corepressor

protein(s) and, with them, regulates
transcription of the adjacent gene(s), increas-
mRNA ing or decreasing the rate of mRNA formation.
4
4 Altered levels of the hormone-regulated gene

product produce the cellular response to the
translation hormone.
on ribosomes
FIGURE 12–30 General mechanism by which steroid and thyroid hormones, retinoids, and vitamin D regulate gene
expression. The details of transcription and protein synthesis are discussed in Chapters 26 and 27. Some steroids also
act through plasma membrane receptors by a completely different mechanism.
CH3 mechanism of steroid hormone action through nuclear

N receptors. For example, the estrogen-mediated dilation
CH3 O of blood vessels is known to be independent of gene
transcription or protein synthesis, as is the steroid-
induced decrease in cellular [cAMP]. Another trans-
duction mechanism involving plasma membrane recep-
tors is believed to be responsible for some of these
effects.
CH3
Tamoxifen
CH3
N
SUMMARY 12.8 Regulation of Transcription by
CH3
H3C OH
Nuclear Hormone Receptors
!
C C CH3
Steroid hormones enter cells and bind to specific
receptor proteins.
! The hormone-receptor complex binds specific
O regions of DNA, the hormone response elements,
Mifepristone and interacts with other proteins to regulate the
(RU486) ■ expression of nearby genes.
Certain effects of steroids seem to occur too fast to ! Certain effects of steroid hormones may occur
be the result of altered protein synthesis via the classic through a different, faster signaling pathway.
12.9 Signaling in Microorganisms and Plants 473
12.9 Signaling in Microorganisms (a)

Run
Tumble
and Plants Tumble

Attractant
Much of what we have said here about signaling relates

to mammalian tissues or cultured cells from such tis-
sues. Bacteria, archaea, eukaryotic microorganisms, and
vascular plants must also respond to a variety of external (b) Counterclockwise Clockwise
signals—O2, nutrients, light, noxious chemicals, and so rotation → “run” rotation → “tumble”
on. We turn here to a brief consideration of the kinds of
signaling machinery used by microorganisms and plants.
Bacterial Signaling Entails Phosphorylation

(c)
in a Two-Component System Receptor/His kinase
(component 1)
Phosphorylated
form of component 2
In pioneering studies of chemotaxis in bacteria, Julius Attractant
reverses direction
Adler showed that Escherichia coli responds to nutri- of motor
ents in its environment, includ- ATP
His His Rotary motor
ing sugars and amino acids, by His
ADP (controls flagellum)
P P
swimming toward them, pro-
Asp
pelled by one or a few flagella. Asp E. coli
A family of membrane proteins
have binding domains on the Plasma
membrane
outside of the plasma mem- Response regulator (component 2)
brane to which specific attrac-
tants (sugars or amino acids) FIGURE 12–31 The two-component signaling mechanism in bacterial
chemotaxis. (a) When placed near a source of an attractant solute,
bind (Fig. 12–31). Ligand
E. coli performs a random walk, biased toward the attractant. (b) Flagella
binding causes an intrinsic
Julius Adler have intrinsic helical structure, and when all flagella rotate counterclock-
kinase activity of the receptor
wise, the flagellar helices twist together and move in concert to propel
to phosphorylate a His residue in its cytosolic domain. the cell forward in a “run.” When the flagella rotate clockwise, the flagel-
This first component of the two-component system, lar bundles fly apart, and the cell tumbles briefly until counterclockwise
the receptor histidine kinase, then catalyzes transfer rotation resumes and the cell begins to swim forward again in a new,
of the phosphoryl group from the His residue to an Asp random direction. When moving toward the attractant, the cell has
residue on a second, soluble protein, the response regu- fewer tumbles, and therefore longer runs; when moving away, the fre-
lator. This phosphoprotein moves to the base of the fla- quent tumbles eventually result in movement toward the attractant.
gellum, carrying the signal from the membrane receptor. (c) Flagellar rotation is controlled by a two-component system consist-
The flagellum is driven by a rotary motor that can propel ing of a receptor–histidine kinase and an effector protein. When an
the cell through its medium or cause it to stall, depending attractant ligand binds to the receptor domain of the membrane-bound
on the direction of motor rotation. The change in attrac- receptor, a protein kinase in the cytosolic domain (component 1) is acti-
tant concentration over time, signaled through the recep- vated and autophosphorylates a His residue. This phosphoryl group is
tor, allows the cell to determine whether it is moving then transferred to an Asp residue on component 2. After phosphoryla-
toward or away from the source of the attractant. If its tion, component 2 moves to the base of the flagellum, where it causes
motion is toward the attractant, the response regulator counterclockwise rotation of the flagella, producing a run.
signals the cell to continue in a straight line (a run); if
away from it, the cell tumbles momentarily, acquiring a
new direction. Repetition of this behavior results in a more, and more diverse, bacteria become known,
random path, biased toward movement in the direction of researchers have discovered genes that encode proteins
increasing attractant concentration. similar to protein Ser or Thr kinases, Ras-like proteins
E. coli detects not only sugars and amino acids but regulated by GTP binding, and proteins with SH3 domains.
also O2, extremes of temperature, and other environ- Receptor Tyr kinases have not been detected in bacteria,
mental factors, using this basic two-component system. but P –Tyr residues do occur in some bacteria.
Two-component systems have been detected in many
other bacteria, both gram-positive and gram-negative,
Signaling Systems of Plants Have Some of the Same
and in archaea, as well as in protists and fungi. Clearly,
this signaling mechanism developed early in the course Components Used by Microbes and Mammals
of cellular evolution and has been conserved. Like animals, vascular plants must have a means of com-
Various signaling systems used by animal cells also munication between tissues to coordinate and direct
have analogs in bacteria. As the full genomic sequences of growth and development; to adapt to conditions of O2,
474 Biosignaling
nutrients, light, temperature, and water availability; and Light

to warn of the presence of noxious chemicals and dam-
aging pathogens (Fig. 12–32). At least a billion years of Temperature
evolution have passed since the plant and animal Humidity
branches of the eukaryotes diverged, which is reflected
in the differences in signaling mechanisms: some plant
mechanisms are conserved—that is, are similar to those
in animals (protein kinases, adaptor proteins, cyclic
nucleotides, electrogenic ion pumps, and gated ion Wind CO2
channels); some are similar to bacterial two-component
systems; and some are unique to plants (light-sensing C2H4
mechanisms that reflect seasonal changes in the angle, Insects
and hence color, of sunlight, for example) (Table 12–7). Herbivores
Pathogens
The genome of the plant Arabidopsis thaliana encodes
about 1,000 protein Ser/Thr kinases, including about
60 MAPKs and nearly 400 membrane-associated receptor Pathogens O2
kinases that phosphorylate Ser or Thr residues; a variety
Parasites
of protein phosphatases; adaptor proteins that form Minerals
scaffolds on which proteins assemble in signaling com- Toxic molecules
plexes; enzymes for the synthesis and degradation of Microorganisms Water status
cyclic nucleotides; and 100 or more ion channels,
Gravity
including about 20 gated by cyclic nucleotides. Inositol
phospholipids are present, as are kinases that intercon- FIGURE 12–32 Some stimuli that produce responses in plants.
vert them by phosphorylation of inositol head groups.
Even given the fact that Arabidopsis has multiple cop-
ies of many genes, the presence of this many genes
certainly reflects a wide array of signaling potential.
TABLE 12–7 Signaling Components Present in Mammals, Plants, or Bacteria

Signaling component Mammals Plants Bacteria
Ion channels " " "
Electrogenic ion pumps " " "
Two-component His kinases " " "
Adenylyl cyclase " " "
Guanylyl cyclase " " ?
Receptor protein kinases (Ser/Thr) " " ?
Ca2" as second messenger " " ?
Ca2" channels " " ?
Calmodulin, CaM-binding protein " " –
MAPK cascade " " –
Cyclic nucleotide–gated channels " " –
IP3-gated Ca2" channels " " –
Phosphatidylinositol kinases " " –
GPCRs " "/! "
Trimeric G proteins " "/! –
PI-specific phospholipase C " ? –
Tyrosine kinase receptors " ? –
SH2 domains " ? ?
Nuclear steroid receptors " – –
Protein kinase A " – –
Protein kinase G " – –
12.9 Signaling in Microorganisms and Plants 475
However, some types of signaling proteins com- tions), acts through receptors that are related in primary
mon in animal tissues are not present in plants, or are sequence to the receptor His kinases of the bacterial
represented by only a few genes. Cyclic nucleotide– two-component systems and probably evolved from
dependent protein kinases (PKA and PKG) seem to be them. In Arabidopsis, the two-component signaling sys-
absent, for example. Heterotrimeric G proteins and tem is contained within a single integral membrane pro-
protein Tyr kinase genes are much less prominent in tein of the endoplasmic reticulum (not the plasma mem-
the plant genome, and genes for GPCRs, the largest brane). Ethylene diffuses into the cell through the
family of proteins in the human genome (,1,000 genes), plasma membrane and into the ER. The first down-
are very sparsely represented in the plant genome. stream component affected by ethylene signaling is a
DNA-binding nuclear steroid receptors are certainly protein Ser/Thr kinase (CTR1; Fig. 12–34) with
not prominent, and may be absent from plants. sequence homology to Raf, the protein kinase that
Although plants lack the most widely conserved light- begins the MAPK cascade in the mammalian response
sensing mechanism present in animals (rhodopsin, to insulin (see Fig. 12–15). In plants, in the absence of
with retinal as pigment), they have a rich collection of ethylene, the CTR1 kinase is active and inhibits the
other light-detecting mechanisms not found in animal MAPK cascade, preventing transcription of ethylene-
tissues—phytochromes and cryptochromes, for exam- responsive genes. Exposure to ethylene inactivates the
ple (Chapter 19). CTR1 kinase, thereby activating the MAPK cascade that
The kinds of compounds that elicit signals in plants leads to activation of the transcription factor EIN3. Active
are similar to certain signaling molecules in animals EIN3 stimulates the synthesis of a second transcription
(Fig. 12–33). Instead of prostaglandins, plants have factor (ERF1), which in turn activates transcription of
jasmonate; instead of steroid hormones, brassino-
steroids. About 100 different small peptides serve as
plant signals, and both plants and animals use com- Ethylene
pounds derived from aromatic amino acids as signals.
Ethylene receptor ER lumen
1 1
Plants Detect Ethylene through a Two-Component Two-component
Endoplasmic
System and a MAPK Cascade 2 2 system
reticulum
The gaseous plant hormone ethylene (CH2PCH2), which
stimulates the ripening of fruits (among other func-
CTR1
Plants Animals (MAPKKK)
O
COO!
8 MAPK
COO !
Cytosol
cascade
12
O OH OH EIN2
Jasmonate Prostaglandin E1 EIN3 Nucleus
" DNA
HO NH3 mRNA
COO!
ERF1
N N DNA
H H
mRNA
Indole-3-acetate Serotonin
(an auxin) (5-hydroxytryptamine) Ethylene-
response
OH proteins
FIGURE 12–34 Transduction mechanism for detection of ethylene by

OH plants. The ethylene receptor (pink) in the endoplasmic reticulum is a
two-component system contained in a single protein, with a receptor
OH
domain (component 1) and a response regulator domain (component 2).
HO The receptor controls (in ways we do not yet understand) the activity
of CTR1, a protein kinase similar to MAPKKKs and therefore pre-
sumed to be part of a MAPK cascade. CTR1 is a negative regulator of
HO O
H the ethylene response; when CTR1 is inactive, the ethylene signal is
O HO
transmitted through the gene product EIN2 (thought to be a nuclear
Brassinolide Estradiol envelope protein), which causes increased synthesis of ERF1, a tran-
(a brassinosteroid)
scription factor. ERF1 stimulates expression of proteins specific to the
FIGURE 12–33 Structural similarities between plant and animal signals. ethylene response.
476 Biosignaling
ethylene-responsive genes; the gene products affect (flg22) released by breakdown of flagellin, the major pro-
processes ranging from seedling development to fruit tein of the bacterial flagellum. Binding of flg22 to the FLS2
ripening. Although apparently derived from the bacterial receptor of Arabidopsis induces receptor dimerization
two-component signaling system, the ethylene system in and autophosphorylation on Ser and Thr residues, and
Arabidopsis is different in that the His kinase activity the downstream effect is activation of a MAPK cascade
that defines component 1 in bacteria is not essential to like that described above for insulin action. The final
signal transduction in Arabidopsis. kinase in this cascade activates a specific transcription
factor, triggering synthesis of the proteins that defend
Receptorlike Protein Kinases Transduce against the bacterial infection. The steps between recep-
Signals from Peptides tor phosphorylation and the MAPK cascade are not yet
known. A phosphoprotein phosphatase (KAPP) associ-
One common motif in plant signaling involves receptor- ates with the active receptor protein and inactivates it by
like kinases (RLKs), which have a single helical seg- dephosphorylation to end the response.
ment in the plasma membrane that connects a receptor The MAPK cascade in the plant’s defense against
domain on the outside with a protein Ser/Thr kinase on bacterial pathogens is remarkably similar to the innate
the cytoplasmic side. This type of receptor participates in immune response in mammals (Fig. 12–35b) that is trig-
the defense mechanism triggered by infection with a bac- gered by bacterial lipopolysaccharide and mediated by
terial pathogen (Fig. 12–35a). The signal to turn on the the Toll-like receptors (TLRs, a name derived from a
genes needed for defense against infection is a peptide Drosophila mutant originally called Toll (German for
“mad”); TLRs were subsequently found in many other
(a) Plant (Arabidopsis) (b) Mammal organisms and were shown to function in embryonic
Toll-like
flg22 Flagellin receptors LPS development). Other membrane receptors use similar
Dimeric mechanisms to activate a MAPK cascade, ultimately
FLS2 activating transcription factors and turning on the
receptor
Outside genes essential to the defense response.
Most of the several hundred RLKs in plants are pre-
sumed to act in similar ways: ligand binding induces
Inside dimerization and autophosphorylation, and the activated
Ser, Thr Protein
kinase receptor kinase triggers downstream responses by
P
domain Protein phosphorylating key proteins at Ser or Thr residues.
kinase
IRAK
MAPK MAPK MAPK SUMMARY 12.9 Signaling in Microorganisms and Plants
!
cascade cascade cascade
Bacteria and eukaryotic microorganisms have a
variety of sensory systems that allow them to
Transcription Transcription Transcription
factors WRKY22, 29 factors Jun, Fos factor NFkB sample and respond to their environment. In the
two-component system, a receptor His kinase
Immune- Immune- senses the signal and autophosphorylates a His
response proteins response proteins residue, then phosphorylates an Asp residue of the
response regulator.
FIGURE 12–35 Similarities between the signaling pathways that trigger
! Plants respond to many environmental stimuli and
immune responses in plants and animals. (a) In Arabidopsis thaliana,
the peptide flg22, derived from the flagella of a bacterial pathogen, binds
employ hormones and growth factors to coordinate
to its receptor (FLS) in the plasma membrane, causing the receptor to the development and metabolic activities of their
form dimers and triggering autophosphorylation of the cytosolic protein tissues. Plant genomes encode hundreds of
kinase domain on a Ser or Thr residue (not a Tyr). Thus activated, the signaling proteins, including some very similar to
protein kinase phosphorylates downstream proteins (not shown). The those of mammals.
activated receptor also activates (by means unknown) a MAPK cascade, ! Two-component signaling mechanisms common in
which leads to phosphorylation of a nuclear protein that normally inhib- bacteria are found in modified forms in plants,
its the transcription factors WRKY22 and 29; this phosphorylation trig- used in the detection of chemical signals and light.
gers proteolytic degradation of the inhibitor and frees the transcription
! Plant receptorlike kinases (RLKs) participate in
factors to stimulate gene expression related to the immune response.
(b) In mammals, a toxic bacterial lipopolysaccharide (LPS; see Fig. 7–31) detecting a wide variety of stimuli, including
is detected by plasma membrane receptors, which then associate with brassinosteroids, peptides that originate from
and activate a soluble protein kinase (IRAK). The major flagellar protein pathogens, and developmental signals. RLKs
of pathogenic bacteria acts through a similar receptor, also activating autophosphorylate Ser/Thr residues, then activate
IRAK. The activated IRAK initiates two distinct MAPK cascades that end downstream proteins, which in some cases are
in the nucleus, causing the synthesis of proteins needed in the immune MAPK cascades. The end result is increased
response. Jun, Fos, and NF!B are transcription factors. transcription of specific genes.
12.10 Sensory Transduction in Vision, Olfaction, and Gustation 477
12.10 Sensory Transduction in Vision, long, narrow, specialized sensory neurons with two dis-
tinct cellular compartments: the outer segment con-
Olfaction, and Gustation tains dozens of membranous disks loaded with receptor
The detection of light, odors, and tastes (vision, olfaction, proteins and their photosensitive chromophore retinal;
and gustation, respectively) in animals is accomplished by the inner segment contains the nucleus and many mito-
specialized sensory neurons that use signal-transduction chondria, which produce the ATP essential to photo-
mechanisms fundamentally similar to those that detect transduction.
hormones, neurotransmitters, and growth factors. An Like other neurons, rods and cones have a trans-
initial sensory signal is amplified greatly by mechanisms membrane electrical potential (Vm), produced by the
that include gated ion channels and intracellular second electrogenic pumping of the Na"K" ATPase in the
messengers; the system adapts to continued stimulation plasma membrane of the inner segment (Fig. 12–37).
by changing its sensitivity to the stimulus (desensitiza- Also contributing to the membrane potential is an ion
tion); and sensory input from several receptors is inte- channel in the outer segment that permits passage of
grated before the final signal goes to the brain. either Na" or Ca2" and is gated (opened) by cGMP. In
the dark, rod cells contain enough cGMP to keep this
channel open. The membrane potential is therefore
The Visual System Uses Classic GPCR Mechanisms determined by the difference between the amount of
In the vertebrate eye, light entering through the pupil is Na" and K" pumped by the inner segment (which polar-
focused on a highly organized collection of light-sensitive izes the membrane) and the influx of Na" through the
neurons (Fig. 12–36). The light-sensing neurons are of ion channels of the outer segment (which tends to
two types: rods (about 109 per retina), which sense low depolarize the membrane).
levels of light but cannot discriminate colors, and cones The essence of signaling in the rod or cone cell is a
(about 3 ( 106 per retina), which are less sensitive to light-induced decrease in [cGMP], which causes the
light but can discriminate colors. Both cell types are cGMP-gated ion channel to close. The plasma mem-
brane then becomes hyperpolarized by the Na"K"
ATPase. Rod and cone cells synapse with interconnect-
Eye
ing neurons (Fig. 12–36) that carry information about
the electrical activity to ganglion neurons near the inner
Lens surface of the retina. The ganglion neurons integrate
Light the output from many rod or cone cells and send the
resulting signal through the optic nerve to the visual
Retina
cortex of the brain.
Optic
nerve Visual transduction begins when light falls on rho-
dopsin, many thousands of molecules of which are pres-
ent in each disk of the outer segments of rod and cone
cells. Rhodopsin (Mr 40,000) is an integral protein with
seven membrane-spanning " helices (Fig. 12–38), the
characteristic GPCR architecture. The light-absorbing
Light pigment (chromophore) 11-cis-retinal is covalently
attached to opsin, the protein component of rhodopsin,
through a Schiff base to a Lys residue. The retinal mol-
Rod ecule lies near the middle of the bilayer (Fig. 12–38),
Cone
oriented with its long axis approximately in the plane of
the membrane. When a photon is absorbed by the reti-
nal component of rhodopsin, the energy causes a photo-
chemical change; 11-cis-retinal is converted to all-
To optic Ganglion Interconnecting trans-retinal (see Figs 1–19b and 10–21). This change
nerve neurons neurons in the structure of the chromophore forces conforma-
FIGURE 12–36 Light reception in the vertebrate eye. The lens focuses
tional changes in the rhodopsin molecule—the first
light on the retina, which is composed of layers of neurons. The primary
stage in visual transduction.
photosensory neurons are rod cells (yellow), which are responsible for Retinal is derived from vitamin A1 (retinol),
high-resolution and night vision, and cone cells of three subtypes (pink), which is produced from !-carotene (see Fig.
which initiate color vision. The rods and cones form synapses with sev- 10–21). Dietary deficiency of vitamin A leads to night
eral ranks of interconnecting neurons that convey and integrate the elec- blindness (the inability to adapt to low light levels),
trical signals. The signals eventually pass from ganglion neurons through which is relatively common in some developing coun-
the optic nerve to the brain. Note that light must pass through the layers tries. Vitamin A supplements or vegetables rich in caro-
of ganglion neurons and interconnecting neurons before reaching the tene (such as carrots) supply the vitamin and reverse
rod and cone cells. the night blindness. ■
478 Biosignaling
cGMP FIGURE 12–37 Light-induced hyperpolarization of rod cells. The rod

cell consists of an outer segment, filled with stacks of membranous
disks (not shown) containing the photoreceptor rhodopsin, and an
inner segment that contains the nucleus and other organelles (not
Ca2+ shown). The inner segment forms a synapse with interconnecting neu-
rons (Fig. 12–36). Cones have a similar structure. ATP in the inner seg-
Na+ ment powers the Na"K" ATPase, which creates a transmembrane elec-
trical potential by pumping 3 Na" out for every 2 K" pumped in. The
membrane potential is reduced by the inflow of Na" and Ca2" through
Vm=–45 mV
Ion channel cGMP-gated cation channels in the outer-segment plasma membrane.
open When rhodopsin absorbs light, it triggers degradation of cGMP (green
dots) in the outer segment, causing closure of the ion channel. Without
Na+ cation influx through this channel, the cell becomes hyperpolarized.
This electrical signal is passed to the brain through the ranks of neurons
Na+K+ shown in Figure 12–36.
ATPase
Light
Chromophore
Rhodopsi (11-cis-retinal)
Disk
compartmen t
Outer segment
Na+, Ca2+
Vm=–75 mV
Ion channel
closed Cytosol
Na+
Inner segment
Gsb
Gsa Gsg
Transducin
Electrical
signal FIGURE 12–38 Complex of rhodopsin with the G protein transducin.
(PDB ID 1BAC) Rhodopsin (red) has seven transmembrane helices
embedded in the disk membranes of rod outer segments and is oriented
with its carboxyl terminus on the cytosolic side and its amino terminus
Excited Rhodopsin Acts through the G Protein inside the disk. The chromophore 11-cis-retinal (yellow space-filling struc-
Transducin to Reduce the cGMP Concentration ture), attached through a Schiff base linkage to Lys256 of the seventh
helix, lies near the center of the bilayer. (This location is similar to that
In its excited conformation, rhodopsin interacts with a of the epinephrine-binding site in the !-adrenergic receptor.) Several Ser
second protein, transducin, which hovers nearby on and Thr residues near the carboxyl terminus are substrates for phos-
the cytoplasmic face of the disk membrane (Fig. 12–38). phorylations that are part of the desensitization mechanism for rhodopsin.
Transducin (T) belongs to the same family of heterotri- Cytosolic loops that interact with the G protein transducin are shown in
meric GTP-binding proteins as Gs and Gi. Although orange; their exact positions are not yet known. The three subunits of
specialized for visual transduction, transducin shares transducin (green) are shown in their likely arrangement. Rhodopsin is
many functional features with Gs and Gi. It can bind palmitoylated at its carboxyl terminus, and both the " and # subunits of
either GDP or GTP. In the dark, GDP is bound, all three transducin have attached lipids (yellow) that assist in anchoring them to
subunits of the protein (T", T!, and T#) remain together, the membrane.
and no signal is sent. When rhodopsin is excited by activity immediately increases by several orders of mag-
light, it interacts with transducin, catalyzing the replace- nitude. Each molecule of the active PDE degrades many
ment of bound GDP by GTP from the cytosol (Fig. molecules of cGMP to the biologically inactive 5'-GMP,
12–39, steps 1 and 2). Transducin then dissociates lowering [cGMP] in the outer segment within a fraction
into T" and T!#, and the T"-GTP carries the signal from of a second. At the new, lower [cGMP], the cGMP-gated
the excited receptor to the next element in the trans- ion channels close, blocking reentry of Na" and Ca2" into
duction pathway, a cGMP phosphodiesterase; this the outer segment and hyperpolarizing the membrane of
enzyme converts cGMP to 5'-GMP (steps 3 and 4). the rod or cone cell (step 5). Through this process, the
Note that this is not the same cyclic nucleotide phos- initial stimulus—a photon—changes the Vm of the cell.
phodiesterase that hydrolyzes cAMP to terminate the The brighter the illumination of the rod cell, the greater
!-adrenergic response. One isoform of the cGMP-specific the hyperpolarization. This hyperpolarization is per-
PDE is unique to the visual cells of the retina. ceived by the integrating neurons of the retina, which
The PDE of the retina is a peripheral protein with its pass the integrated signal on to the ganglion cells, which
active site on the cytoplasmic side of the disk mem- send axons via the optic nerve to the brain.
brane. In the dark, a tightly bound inhibitory subunit Several steps in the visual-transduction process
very effectively suppresses the PDE activity. When T"- result in a huge amplification of the signal. Each excited
GTP encounters the PDE, the inhibitory subunit leaves rhodopsin molecule activates at least 500 molecules of
the enzyme and instead binds T", and the enzyme’s transducin, each of which can activate a molecule of
1 Light absorption 2 Activated rhodopsin 3 Ta-GTP activates cGMP 4 Active PDE reduces
converts 11-cis- catalyzes replacement phosphodiesterase [cGMP] to below the
retinal to all-trans- of GDP by GTP (PDE) by binding and level needed to keep
retinal, activating on transducin (T), removing its inhibitory cation channels open. Rod
rhodopsin (Rh). which then dissociates subunit (I).
into Ta-GTP and Tbg.
GTP
5!-GMP
GDP
I I
Rh Ta– Ta– cGMP
T bg Ta– PDE PDE
GDP GTP GTP cGMP Na+, Ca2+
Disk
membrane 5 Cation channels
2+ close, preventing
Ca influx of Na+ and
Ca2+; membrane is
Excitation hyperpolarized.
Rh This signal passes
to the brain.
Recovery/Adaptation
P P P
6 Continued efflux of
RK Ca2+ through the
Recov Na+-Ca2+ exchanger
reduces cytosolic
Rh Rh [Ca2+].
GC
P
P Ca2+
GTP
Arr P 2+
cGMP [Ca ] 4 Na+
cGMP
8 Rhodopsin kinase (RK) 9 Slowly, arrestin dissociates, 7 Reduction of [Ca2+]

phosphorylates “bleached” rhodopsin is dephos- activates guanylyl
rhodopsin; low [Ca2+] phorylated, and all-trans- cyclase (GC) and
Plasma
and recoverin (Recov) retinal is replaced with inhibits PDE; [cGMP]
membrane
stimulate this reaction. 11-cis-retinal. Rhodopsin is rises toward “dark”
Arrestin (Arr) binds ready for another level, reopening cation
phosphorylated carboxyl phototransduction cycle. channels and returning
terminus, inactivating Vm to prestimulus level.
rhodopsin.
FIGURE 12–39 Molecular consequences of photon absorption by rho- describes excitation; the bottom shows post-illumination steps: recovery
dopsin in the rod outer segment. The top half of the figure (steps 1 to 5) (steps 6 and 7) and adaptation (steps 8 and 9).
480 Biosignaling
the PDE. This phosphodiesterase has a remarkably high Cone Cells Specialize in Color Vision
turnover number, each activated molecule hydrolyzing
Color vision involves a path of sensory transduction in
4,200 molecules of cGMP per second. The binding of
cone cells essentially identical to that described above,
cGMP to cGMP-gated ion channels is cooperative, and a
but triggered by slightly different light receptors. Three
relatively small change in [cGMP] therefore registers as
types of cone cells are specialized to detect light from
a large change in ion conductance. The result of these
different regions of the spectrum, using three related
amplifications is exquisite sensitivity to light. Absorption
photoreceptor proteins (opsins). Each cone cell
of a single photon closes 1,000 or more ion channels and
expresses only one kind of opsin, but each type is
changes the cell’s membrane potential by about 1 mV.
closely related to rhodopsin in size, amino acid sequence,
and presumably three-dimensional structure. The dif-
The Visual Signal Is Quickly Terminated ferences among the opsins, however, are great enough
to place the chromophore, 11-cis-retinal, in three slightly
As your eyes move across this line, the retinal images of different environments, with the result that the three
the first words disappear rapidly—before you see the photoreceptors have different absorption spectra (Fig.
next series of words. In that short interval, a great deal 12–40). We discriminate colors and hues by integrating
of biochemistry has taken place. Very shortly after illu- the output from the three types of cone cells, each con-
mination of the rod or cone cells stops, the photosen- taining one of the three photoreceptors.
sory system shuts off. The " subunit of transducin (with Color blindness, such as the inability to distinguish
bound GTP) has intrinsic GTPase activity. Within milli- red from green, is a fairly common, genetically
seconds after the decrease in light intensity, GTP is inherited trait in humans. The various types of color blind-
hydrolyzed and T" reassociates with T!#. The inhibitory ness result from different opsin mutations. One form is
subunit of the PDE, which had been bound to T"-GTP, due to loss of the red photoreceptor; affected individuals
is released and reassociates with the enzyme, strongly are red! dichromats (they see only two primary colors).
inhibiting its activity. Others lack the green pigment and are green! dichro-
To return [cGMP] to its “dark” level, the enzyme mats. In some cases, the red and green photoreceptors
guanylyl cyclase converts GTP to cGMP (step 7 in Fig. are present but have a changed amino acid sequence that
12–39) in a reaction that is inhibited by high [Ca2"] causes a change in their absorption spectra, resulting in
()100 nM). Calcium levels drop during illumination, abnormal color vision. Depending on which pigment is
because the steady-state [Ca2"] in the outer segment is altered, such individuals are red-anomalous trichro-
the result of outward pumping of Ca2" through the Na"- mats or green-anomalous trichromats. Examination
Ca2" exchanger of the plasma membrane (see Fig. 12–37) of the genes for the visual receptors has allowed the diag-
and influx of Ca2" through open cGMP-gated channels. nosis of color blindness in a famous “patient” more than a
In the dark, this produces a [Ca2"] of about 500 nM— century after his death (Box 12–4). ■
enough to inhibit cGMP synthesis. After brief illumina-
tion, Ca2" entry slows and [Ca2"] declines (step 6).
The inhibition of guanylyl cyclase by Ca2" is relieved, 100
and the cyclase converts GTP to cGMP to return the Green
90
system to its prestimulus state (step 7). pigment
80
Rhodopsin itself also undergoes changes in response
Relative absorbance
70 Red
to prolonged illumination. The conformational change pigment
induced by light absorption exposes several Thr and Ser 60
residues in the carboxyl-terminal domain. These resi- 50
dues are quickly phosphorylated by rhodopsin kinase 40
Rhodopsin
(step 8 in Fig. 12–39), which is functionally and struc- 30
turally homologous to the !-adrenergic kinase (!ARK) 20 Blue
pigment
that desensitizes the !-adrenergic receptor (Fig. 12–8). 10
The Ca2"-binding protein recoverin inhibits rhodopsin 0
kinase at high [Ca2"], but the inhibition is relieved when
[Ca2"] drops after illumination, as described above. The
phosphorylated carboxyl-terminal domain of rhodopsin
is bound by the protein arrestin 1, preventing further 400 450 500 550 600 650
interaction between activated rhodopsin and transdu- Wavelength (nm)
cin. Arrestin 1 is a close homolog of arrestin 2 (!arr; Fig. FIGURE 12–40 Absorption spectra of purified rhodopsin and the red,
12–8). On a relatively long time scale (seconds to min- green, and blue receptors of cone cells. The receptor spectra, obtained
utes), the all-trans-retinal of an excited rhodopsin mol- from individual cone cells isolated from cadavers, peak at about 420,
ecule is removed and replaced by 11-cis-retinal, to 530, and 560 nm, and the maximum absorption for rhodopsin is at
produce rhodopsin that is ready for another round of about 500 nm. For reference, the visible spectrum for humans is about
excitation (step 9 in Fig. 12–39). 380 to 750 nm.
BOX 12–4 MEDICINE Color Blindness: John Dalton’s Experiment from the Grave
The chemist John Dalton (of atomic theory fame) was had the opsin gene for the red photopigment but
color-blind. He thought it probable that the vitreous lacked the opsin gene for the green photopigment.
humor of his eyes (the fluid that fills the eyeball Dalton was a green-dichromat. So, 150 years after his
behind the lens) was tinted blue, unlike the colorless death, the experiment Dalton started—by hypothe-
fluid of normal eyes. He proposed that after his death, sizing about the cause of his color blindness—was
his eyes should be dissected and the color of the vitre- finally finished.
ous humor determined. His wish was honored. The
day after Dalton’s death in July 1844, Joseph Ran-
some dissected his eyes and found the vitreous humor
to be perfectly colorless. Ransome, like many scien-
tists, was reluctant to throw samples away. He placed
Dalton’s eyes in a jar of preservative, where they
stayed for a century and a half (Fig. 1).
Then, in the mid-1990s, molecular biologists in
England took small samples of Dalton’s retinas and
extracted DNA. Using the known gene sequences for
the opsins of the red and green light receptors, they
amplified the relevant sequences (using techniques
described in Chapter 9) and determined that Dalton FIGURE 1 Dalton’s eyes.
Vertebrate Olfaction and Gustation Use Mechanisms depolarization called the receptor potential. If a suf-
ficient number of odorant molecules encounter recep-
Similar to the Visual System tors, the receptor potential is strong enough to cause
The sensory cells that detect odors and tastes have the neuron to fire an action potential. This is relayed to
much in common with the rod and cone cells. Olfactory the brain in several stages and registers as a specific
neurons have long thin cilia extending from one end of smell. All these events occur within 100 to 200 ms.
the cell into a mucous layer that overlays the cell. These When the olfactory stimulus is no longer present,
cilia present a large surface area for interaction with the transducing machinery shuts itself off in several
olfactory signals. The receptors for olfactory stimuli are ways. A cAMP phosphodiesterase returns [cAMP] to the
ciliary membrane proteins with the familiar GPCR prestimulus level. Golf hydrolyzes its bound GTP to GDP,
structure of seven transmembrane " helices. The olfac- thereby inactivating itself. Phosphorylation of the recep-
tory signal can be any one of the many volatile com- tor by a specific kinase prevents its interaction with Golf,
pounds for which there are specific receptor proteins. by a mechanism analogous to that used to desensitize
Our ability to discriminate odors stems from hundreds the !-adrenergic receptor and rhodopsin. And lastly,
of different olfactory receptors in the tongue and nasal some odorants are enzymatically destroyed by oxidases.
passages and from the brain’s ability to integrate input The sense of taste in vertebrates reflects the activ-
from different types of olfactory receptors to recognize ity of gustatory neurons clustered in taste buds on the
a “hybrid” pattern, extending our range of discrimina- surface of the tongue. In these sensory neurons,
tion far beyond the number of receptors. GPCRs are coupled to the heterotrimeric G protein
The olfactory stimulus arrives at the sensory cells gustducin (very similar to the transducin of rod and
by diffusion through the air. In the mucous layer covering cone cells). Sweet-tasting molecules are those that
the olfactory neurons, the odorant molecule binds directly bind receptors in “sweet” taste buds. When the mole-
to an olfactory receptor or to a specific binding protein cule (tastant) binds, gustducin is activated by replace-
that carries the odorant to a receptor (Fig. 12–41). ment of bound GDP with GTP and then stimulates
Interaction between odorant and receptor triggers a cAMP production by adenylyl cyclase. The resulting
change in receptor conformation that results in the elevation of [cAMP] activates PKA, which phosphory-
replacement of bound GDP by GTP on a G protein, Golf, lates K" channels in the plasma membrane, causing
analogous to transducin and to Gs of the !-adrenergic them to close. Reduced efflux of K" depolarizes the
system. The activated Golf then activates adenylyl cell (Fig. 12–42) sending an electrical signal to the
cyclase of the ciliary membrane, which synthesizes brain. Other taste buds specialize in detecting bitter,
cAMP from ATP, raising the local [cAMP]. The cAMP- sour, salty, or umami (savory) tastants, using various
gated Na" and Ca2" channels of the ciliary membrane combinations of second messengers and ion channels
open, and the influx of Na" and Ca2" produces a small in the transduction mechanisms.
482 Biosignaling
1 Odorant (O) arrives 2 Activated OR Cilia Olfactory

at the mucous layer catalyzes GDP-GTP neuron
and binds directly to exchange on a
an olfactory receptor G protein (Golf),
(OR) or to a binding causing its
protein (BP) that dissociation
carries it to the OR. into a and bg. Dendrite Axon
O
3 Ga-GTP activates
adenylyl cyclase, Air
which catalyzes
O cAMP synthesis, Mucous
raising [cAMP]. 4 cAMP-gated cation layer
channels open. Ca2+
enters, raising
O internal [Ca2+].
BP O
OR AC
Golf Golfa– a Ciliary
b g GDP GTP membrane
ATP cAMP
Ca2+ Cl–
GTP
GDP
7 Golfa hydrolyzes GTP to 6 Ca2+ reduces the 5 Ca2+-gated chloride

GDP, shutting itself off. Cyclic affinity of the cation channels open. Efflux of
AMP PDE hydrolyzes cAMP. channel for cAMP, Cl– depolarizes the cell,
Receptor kinase phosphorylates lowering the sensitivity triggering an
OR, inactivating it. Odorant of the system to odorant. electrical signal
is removed by metabolism. to the brain.
FIGURE 12–41 Molecular events of olfaction. These interactions occur in the cilia of olfactory receptor cells.
GPCRs of the Sensory Systems Share Several Features arisen early in evolution; genomic studies have revealed
hundreds of genes encoding GPCRs in vertebrates,
with GPCRs of Hormone Signaling Systems
arthropods (Drosophila and mosquito), and the round-
We have now looked at several types of signaling sys- worm Caenorhabditis elegans. Even the common
tems (hormone signaling, vision, olfaction, and gustation) baker’s yeast Saccharomyces uses GPCRs and G pro-
in which membrane receptors are coupled to second teins to detect the opposite mating type. Overall pat-
messenger–generating enzymes through G proteins. As terns have been conserved, and the introduction of
we have intimated, signaling mechanisms must have variety has given modern organisms the ability to
Apical K+ Basolateral
membrane membrane
S
SR AC
Ggust a a
GTP P
b g GDP ATP cAMP
PKA
GTP
GDP
1 Sweet-tasting 2 Gustducin a subunit 3 PKA, activated by cAMP,
molecule (S) binds to activates adenylyl phosphorylates a K+ channel in
sweet-taste receptor (SR), cyclase (AC) of the the basolateral membrane, causing
activating the G protein apical membrane, it to close. The reduced efflux of K+
gustducin (Ggust). raising [cAMP]. depolarizes the cell sending an
electrical signal to the brain.
Taste cell
FIGURE 12–42 Transduction mechanism for sweet tastants.

respond to a wide range of stimuli (Table 12–8). Of the

approximately 29,000 genes in the human genome, as TABLE 12–8 Some Signals That Act through GPCRs
many as 1,000 encode GPCRs, including hundreds for Amines Somatostatin
olfactory stimuli and many “orphan receptors” for which Tachykinin
the natural ligand is not yet known. Acetylcholine
All well-studied signal-transducing systems that act (muscarinic) Thyrotropin-releasing
through heterotrimeric G proteins share some common Dopamine hormone
features, which reflect their evolutionary relatedness Epinephrine Urotensin II
(Fig. 12–43). The receptors have seven transmem- Histamine Protein hormones
brane segments, a domain (generally the loop between
Serotonin Follicle-stimulating
transmembrane helices 6 and 7) that interacts with a G
Peptides hormone
protein, and a carboxyl-terminal cytoplasmic domain
that undergoes reversible phosphorylation on several Angiotensin Gonadotropin
Ser or Thr residues. The ligand-binding site (or, in the Lutropin-
Bombesin
case of light reception, the light receptor) is buried choriogonadotropic
deep in the membrane and includes residues from sev- Bradykinin hormone
eral of the transmembrane segments. Ligand binding Chemokine Thyrotropin
(or light) induces a conformational change in the recep- Colecystokinin
tor, exposing a domain that can interact with a G pro- Prostanoids
(CCK)
tein. Heterotrimeric G proteins activate or inhibit effec- Prostacyclin
Endothelin
tor enzymes (adenylyl cyclase, PDE, or PLC), which Prostaglandin
change the concentration of a second messenger (cAMP, Gonadotropin-releasing
hormone Thromboxane
cGMP, IP3, or Ca2"). In the hormone-detecting systems,
the final output is an activated protein kinase that regu- Interleukin-8 Others
lates some cellular process by phosphorylating a protein Melanocortin Cannabinoids
critical to that process. In sensory neurons, the output
Neuropeptide Y Lysosphingolipids
is a change in membrane potential and a consequent
electrical signal that passes to another neuron in the Neurotensin Melatonin
pathway connecting the sensory cell to the brain. Opioid Olfactory stimuli
All these systems self-inactivate. Bound GTP is con- Orexin Rhodopsin
verted to GDP by the intrinsic GTPase activity of G
proteins, often augmented by GTPase-activating pro-
teins (GAPs) or RGS proteins (regulators of G-protein
Sweet
Vasopressin Epinephrine Light Odorants tastant
b-
VR AR Rh OR1 OR2 SR
Gi Gs T Golf Golf Ggust
AC AC PDE PLC AC AC
[cAMP] [cAMP] [cGMP] [IP3] [cAMP] [cAMP]

PKA
P
PKA PKA
PCa2+,Na+ PCa2+ PCa2+,Na+ PK+
FIGURE 12–43 Common features of signaling systems that detect hor- 2" " "
to Ca , Na , and K . The resulting depolarization or hyperpolarization
mones, light, smells, and tastes. GPCRs provide signal specificity, and of the sensory cell (the signal) passes through relay neurons to sensory
their interaction with G proteins provides signal amplification. Heterotri- centers in the brain. In the best-studied cases, desensitization includes
meric G proteins activate effector enzymes: adenylyl cyclase (AC), phosphorylation of the receptor and binding of a protein (arrestin) that
phospholipase C (PLC), and phosphodiesterases (PDEs) that degrade interrupts receptor–G protein interactions. VR is the vasopressin recep-
cAMP or cGMP. Changes in concentration of the second messengers tor; !-AR is the !-adrenergic receptor. Other receptor and G-protein
(cAMP, cGMP, IP3) result in alterations of enzymatic activities by phos- abbreviations are as used in earlier illustrations.
phorylation or alterations in the permeability (P) of surface membranes
484 Biosignaling
signaling; see Fig. 12–5 and Box 12–2, Fig. 4). In some The Cell Cycle Has Four Stages
cases, the effector enzymes that are the targets of
Cell division accompanying mitosis in eukaryotes occurs
modulation by G proteins also serve as GAPs. The
in four well-defined stages (Fig. 12–44). In the S (syn-
desensitization mechanism involving phosphorylation of
thesis) phase, the DNA is replicated to produce copies
the carboxyl-terminal region followed by arrestin bind-
for both daughter cells. In the G2 phase (G indicates the
ing is widespread, and may be universal.
gap between divisions), new proteins are synthesized
and the cell approximately doubles in size. In the M
SUMMARY 12.10 Sensory Transduction in Vision, phase (mitosis), the maternal nuclear envelope breaks
Olfaction, and Gustation down, paired chromosomes are pulled to opposite poles
! Vision, olfaction, and gustation in vertebrates of the cell, each set of daughter chromosomes is sur-
employ GPCRs, which act through heterotrimeric rounded by a newly formed nuclear envelope, and cyto-
G proteins to change the Vm of a sensory neuron. kinesis pinches the cell in half, producing two daughter
! In rod and cone cells of the retina, light activates cells (see Fig. 24–24). In embryonic or rapidly prolifer-
rhodopsin, which activates the G protein ating tissue, each daughter cell divides again, but only
transducin. The freed " subunit of transducin after a waiting period (G1). In cultured animal cells the
activates a cGMP phosphodiesterase, which lowers entire process takes about 24 hours.
[cGMP] and thus closes cGMP-dependent ion After passing through mitosis and into G1, a cell
channels in the outer segment of the neuron. The either continues through another division or ceases to
resulting hyperpolarization of the rod or cone cell divide, entering a quiescent phase (G0) that may last
carries the signal to the next neuron in the hours, days, or the lifetime of the cell. When a cell in G0
pathway, and eventually to the brain. begins to divide again, it reenters the division cycle
through the G1 phase. Differentiated cells such as hepa-
! In olfactory neurons, olfactory stimuli, acting tocytes or adipocytes have acquired their specialized
through GPCRs and G proteins, trigger either an function and form; they remain in the G0 phase. Stem
increase in [cAMP] (by activating adenylyl cyclase) cells retain their potential to divide and to differentiate
or an increase in [Ca2"] (by activating PLC). These into any of a number of cell types.
second messengers affect ion channels and thus
the Vm.
Levels of Cyclin-Dependent Protein Kinases Oscillate
! Gustatory neurons have GPCRs that respond to
The timing of the cell cycle is controlled by a family of
tastants by altering levels of cAMP, which changes
protein kinases with activities that change in response
Vm by gating ion channels.
! There is a high degree of conservation of signaling
proteins and transduction mechanisms across M Phase
signaling systems and across species. Mitosis (nuclear
G2 Phase division) and G0 Phase
No DNA cytokinesis Terminally
(cell division) differentiated
12.11 Regulation of the Cell Cycle by synthesis.
RNA and yield two cells withdraw
daughter cells. from cell cycle
Protein Kinases protein
synthesis indefinitely.
continue.
One of the most dramatic manifestations of signaling G0
M
G2 1h Reentry point
pathways is the regulation of the eukaryotic cell cycle. A cell returning
3–4 h
During embryonic growth and later development, cell from G0
division occurs in virtually every tissue. In the adult enters at early
G1 phase.
organism most tissues become quiescent. A cell’s “deci-
sion” to divide or not is of crucial importance to the G1
6–12 h
organism. When the regulatory mechanisms that limit S
6–8 h G1 Phase
cell division are defective and cells undergo unregulated RNA and protein
division, the result is catastrophic—cancer. Proper cell synthesis. No DNA
division requires a precisely ordered sequence of bio- synthesis.
chemical events that assures every daughter cell a full S Phase
Restriction point
complement of the molecules required for life. Investi- DNA synthesis
A cell that passes this
doubles the
gations into the control of cell division in diverse amount of DNA point is committed
eukaryotic cells have revealed universal regulatory in the cell. RNA to pass into S phase.
mechanisms. Signaling mechanisms much like those and protein are
also synthesized.
discussed above are central in determining whether and
when a cell undergoes cell division, and they also ensure FIGURE 12–44 Eukaryotic cell cycle. The durations (in hours) of the four
orderly passage through the stages of the cell cycle. stages vary, but those shown are typical.
12.11 Regulation of the Cell Cycle by Protein Kinases 485
to cellular signals. By phosphorylating specific proteins CDKs and cyclins, and the action of specific CDK-inhib-
at precisely timed intervals, these protein kinases iting proteins. The precisely timed activation and inac-
orchestrate the metabolic activities of the cell to pro- tivation of a series of CDKs produce signals serving as a
duce orderly cell division. The kinases are heterodimers master clock that orchestrates the events in normal cell
with a regulatory subunit, cyclin, and a catalytic sub- division and ensures that one stage is completed before
unit, cyclin-dependent protein kinase (CDK). In the next begins.
the absence of cyclin, the catalytic subunit is virtually
inactive. When cyclin binds, the catalytic site opens up,
a residue essential to catalysis becomes accessible (Fig. Regulation of CDKs by Phosphorylation The activity of a
12–45), and the protein kinase activity of the catalytic CDK is strikingly affected by phosphorylation and
subunit increases 10,000-fold. Animal cells have at least dephosphorylation of two critical residues in the pro-
10 different cyclins (designated A, B, and so forth) and tein (Fig. 12–47a). Phosphorylation of Tyr15 near
at least 8 CDKs (CDK1 through CDK8), which act in the amino terminus by another protein kinase renders
various combinations at specific points in the cell cycle. CDK2 inactive; the P –Tyr residue is in the ATP-
Plants also use a family of CDKs to regulate their cell binding site of the kinase, and the negatively charged
division in root and shoot meristems, the principal tis- phosphate group blocks the entry of ATP. A specific
sues in which division occurs. phosphatase (a PTPase) dephosphorylates this P –Tyr
In a population of animal cells undergoing synchro- residue, permitting the binding of ATP. Phosphorylation
nous division, some CDK activities show striking oscilla- of Thr160 in the “T loop” of CDK, catalyzed by yet another
tions (Fig. 12–46). These oscillations are the result of protein kinase, forces the T loop out of the substrate-
four mechanisms for regulating CDK activity: phos- binding cleft, permitting the binding of a specific down-
phorylation or dephosphorylation of the CDK, controlled stream target protein and its phosphorylation by CDK
degradation of the cyclin subunit, periodic synthesis of (Fig. 12–45c).
(a) Amino-terminal helix FIGURE 12–45 Activation of cyclin-dependent protein kinases (CDKs)
Glu51 by cyclin and phosphorylation. CDKs, a family of related enzymes, are
active only when associated with cyclins, another protein family. The
crystal structure of CDK2 with and without cyclin reveals the basis for
T loop this activation. (a) Without cyclin (PDB ID 1HCK), CDK2 folds so that
one segment, the T loop, obstructs the binding site for protein sub-
strates and thus inhibits protein kinase activity. The binding site for ATP
is also near the T loop. (b) When cyclin binds (PDB ID 1FIN), it forces
ATP conformational changes that move the T loop away from the active site
and reorient an amino-terminal helix, bringing a residue critical to catal-
ysis (Glu51) into the active site. (c) Phosphorylation of a Thr residue in
the T loop produces a negatively charged residue that is stabilized by
interaction with three Arg residues, holding CDK in its active conforma-
CDK2 (inactive)
tion (PDB ID 1JST).
(b) (c)
Phosphorylated
Cyclin Thr160 Arg150
subunit
Arg50
T loop T loop
Glu51
Glu51
Arg126
ATP ATP
CDK2 (inactive) CDK2 (active)

486 Biosignaling
G1 S G2 M G1 The presence of single-strand breaks in DNA leads

to arrest of the cell cycle in G2 by regulating a particular
Cyclin B–CDK1 CDK. A specific protein kinase (called Rad3 in yeast),
which is activated by single-strand breaks, triggers a
Kinase activity
cascade leading to the inactivation of the PTPase that

Cyclin A–CDK2
dephosphorylates Tyr15 of CDK. The CDK remains inac-
Cyclin E–CDK2
tive and the cell is arrested in G2, unable to divide until
the DNA is repaired and the effects of the cascade are
reversed.
Controlled Degradation of Cyclin Highly specific and pre-
Time cisely timed proteolytic breakdown of mitotic cyclins
FIGURE 12–46 Variations in the activities of specific CDKs during the regulates CDK activity throughout the cell cycle.
cell cycle in animals. Cyclin E–CDK2 activity peaks near the G1 phase–S Progress through mitosis requires first the activation
phase boundary, when the active enzyme triggers synthesis of enzymes then the destruction of cyclins A and B, which activate
required for DNA synthesis (see Fig. 12–49). Cyclin A–CDK2 activity the catalytic subunit of the M-phase CDK. These cyclins
rises during the S and G2 phases, then drops sharply in the M phase, as contain near their amino terminus the sequence
cyclin B–CDK1 peaks. –Arg–Thr–Ala–Leu–Gly–Asp–Ile–Gly–Asn–, the “de-
struction box,” which targets them for degradation.
(a)
3
Cyclin-CDK complex forms, but phosphorylation
2 on Tyr15 blocks ATP-binding site; still inactive.
Cyclin
synthesis P Tyr 4
leads to its CDK Phosphorylation of Thr160 in
accumulation. T loop and removal of Tyr15
Cyclin phosphoryl group activates
cyclin-CDK manyfold.
Cyclin
P Tyr Thr P
5
CDK phosphorylates
phosphatase, which
1 activates more CDK.
No cyclin
present; Pi
Phosphatase Phosphatase
CDK is
inactive. P P
Thr
CDK CDK
(b)
DBRP DBRP
P 7
6 DBRP triggers
CDK phosphorylates
DBRP, activating it. addition of ubiquitin
molecules to cyclin
by ubiquitin ligase.
8
Cyclin is
degraded CDK U
by proteasome,
Cyclin
leaving CDK
inactive. U U U U
FIGURE 12–47 Regulation of CDK by phosphorylation and proteolysis. (b) The active cyclin-CDK complex triggers its own inactivation by
(a) The cyclin-dependent protein kinase activated at the time of mitosis phosphorylation of DBRP (destruction box recognizing protein; step 6).
(the M-phase CDK) has a “T loop” that can fold into the substrate-binding DBRP and ubiquitin ligase then attach several molecules of ubiquitin (U)
site. When Thr160 in the T loop is phosphorylated, the loop moves out to cyclin (step 7), targeting it for destruction by proteasomes, proteo-
of the substrate-binding site, activating the CDK manyfold (step 4). lytic enzyme complexes (step 8).
12.11 Regulation of the Cell Cycle by Protein Kinases 487
(This usage of “box” derives from the common practice, Growth factors,
in diagramming the sequence of a nucleic acid or pro- cytokines
tein, of enclosing within a box a short sequence of
MAPK
nucleotide or amino acid residues with some specific cascade
function. It does not imply any three-dimensional struc-
ture.) The protein DBRP (destruction box recognizing
Phosphorylation of
protein) recognizes this sequence and initiates the pro- Jun and Fos in nucleus
cess of cyclin degradation by bringing together the cyclin
and another protein, ubiquitin. Cyclin and activated
ubiquitin are covalently joined by the enzyme ubiquitin
transcriptional
ligase (Fig. 12–47b). Several more ubiquitin molecules regulation
are then appended, providing the signal for a proteo-
lytic enzyme complex, or proteasome, to degrade cyclin. Cyclins, Transcription
CDKs factor E2F
What controls the timing of cyclin breakdown? A
feedback loop occurs in the overall process shown in transcriptional
regulation
Figure 12–47. Increased CDK activity (step 4) leads
eventually to cyclin proteolysis (step 8). Newly synthe- Enzymes for
sized cyclin associates with and activates CDK, which DNA synthesis
phosphorylates and activates DBRP. Active DBRP then
causes proteolysis of cyclin. The lowered cyclin level
causes a decline in CDK activity, and the activity of DBRP
also drops through slow, constant dephosphorylation and
Passage from
inactivation by a DBRP phosphatase. The cyclin level is G1 to S phase
ultimately restored by synthesis of new cyclin molecules.
The role of ubiquitin and proteasomes is not limited FIGURE 12–48 Regulation of cell division by growth factors. The path
to the regulation of cyclin; as we shall see in Chapter 27, from growth factors to cell division leads through the enzyme cascade
both also take part in the turnover of cellular proteins, that activates MAPK; phosphorylation of the nuclear transcription fac-
a process fundamental to cellular housekeeping. tors Jun and Fos; and the activity of the transcription factor E2F, which
promotes synthesis of several enzymes essential for DNA synthesis.
Regulated Synthesis of CDKs and Cyclins The third mecha-
nism for changing CDK activity is regulation of the rate
of synthesis of cyclin or CDK or both. For example, of different cyclins and kinases and the combinations in
cyclin D, cyclin E, CDK2, and CDK4 are synthesized which they act, differ from species to species, but the
only when a specific transcription factor, E2F, is pres- basic mechanism has been conserved in the evolution of
ent in the nucleus to activate transcription of their all eukaryotic cells.
genes. Synthesis of E2F is in turn regulated by extracel-
lular signals such as growth factors and cytokines CDKs Regulate Cell Division by Phosphorylating
(developmental signals that induce cell division), com- Critical Proteins
pounds found to be essential for the division of mam-
We have examined how cells maintain close control of
malian cells in culture. They induce the synthesis of
CDK activity, but how does the activity of CDK control
specific nuclear transcription factors essential to the
the cell cycle? The list of target proteins that CDKs are
production of the enzymes of DNA synthesis. Growth
known to act upon continues to grow, and much
factors trigger phosphorylation of the nuclear proteins
remains to be learned. But we can see a general pattern
Jun and Fos, transcription factors that promote the
behind CDK regulation by inspecting the effect of CDKs
synthesis of a variety of gene products, including
on the structures of lamin and myosin and on the activity
cyclins, CDKs, and E2F. In turn, E2F controls produc-
of retinoblastoma protein.
tion of several enzymes essential for the synthesis of
The structure of the nuclear envelope is maintained
deoxynucleotides and DNA, enabling cells to enter the
in part by highly organized meshworks of intermediate
S phase (Fig. 12–48).
filaments composed of the protein lamin. Breakdown of
Inhibition of CDKs Finally, specific protein inhibitors bind the nuclear envelope before segregation of the sister
to and inactivate specific CDKs. One such protein is chromatids in mitosis is partly due to the phosphoryla-
p21, which we discuss below. tion of lamin by a CDK, which causes lamin filaments to
depolymerize.
These four control mechanisms modulate the activity of A second kinase target is the ATP-driven contractile
specific CDKs that, in turn, control whether a cell will machinery (actin and myosin) that pinches a dividing
divide, differentiate, become permanently quiescent, or cell into two equal parts during cytokinesis. After the
begin a new cycle of division after a period of quiescence. division, CDK phosphorylates a small regulatory subunit
The details of cell cycle regulation, such as the number of myosin, causing dissociation of myosin from actin
488 Biosignaling
filaments and inactivating the contractile machinery. while bound to pRb, E2F cannot promote transcription
Subsequent dephosphorylation allows reassembly of the of a group of genes necessary for DNA synthesis (the
contractile apparatus for the next round of cytokinesis. genes for DNA polymerase ", ribonucleotide reductase,
A third and very important CDK substrate is the and other proteins; see Chapter 25). In this state, the
retinoblastoma protein, pRb; when DNA damage is cell cycle cannot proceed from the G1 to the S phase,
detected, this protein participates in a mechanism that the step that commits a cell to mitosis and cell division.
arrests cell division in G1 (Fig. 12–49). Named for the The pRb-E2F blocking mechanism is relieved when pRb
retinal tumor cell line in which it was discovered, pRb is phosphorylated by cyclin E–CDK2, which occurs in
functions in most, perhaps all, cell types to regulate cell response to a signal for cell division to proceed.
division in response to a variety of stimuli. Unphos- When the protein kinases ATM and ATR detect dam-
phorylated pRb binds the transcription factor E2F; age to DNA (signaled by the presence of the protein MRN
at a double-strand break site), they phosphorylate p53,
activating it to serve as a transcription factor that stimu-
Double-strand lates the synthesis of the protein p21 (Fig. 12–49). This
break in DNA Intact DNA protein inhibits the protein kinase activity of cyclin E–
CDK2. In the presence of p21, pRb remains unphos-
MRN phorylated and bound to E2F, blocking the activity of this
transcription factor, and the cell cycle is arrested in G1.
ATM, ATR This gives the cell time to repair its DNA before entering
the S phase, thereby avoiding the potentially disastrous
transfer of a defective genome to one or both daughter
p53 Active
cells. When the damage is too severe to allow effective
repair, this same machinery triggers a process (apoptosis,
described below) that leads to the death of the cell, pre-
transcriptional
regulation p21 venting the possible development of a cancer.
leads to ↑[p21] Active
SUMMARY 12.11 Regulation of the Cell Cycle
CDK2 CDK2 by Protein Kinases
p21
! Progression through the cell cycle is regulated by
Cyclin E Cyclin E
the cyclin-dependent protein kinases (CDKs),
Inactive P which act at specific points in the cycle,
pRb pRb phosphorylating key proteins and modulating their
activities. The catalytic subunit of CDKs is inactive
pRb transcriptional
regulation unless associated with the regulatory cyclin
E2F E2F Enzymes subunit.
for DNA
Inactive synthesis ! The activity of a cyclin-CDK complex changes
Active during the cell cycle through differential synthesis
Passage of CDKs, specific degradation of cyclin,
from phosphorylation and dephosphorylation of critical
G1 to S
residues in CDKs, and binding of inhibitory
proteins to specific cyclin-CDKs.
Cell division Cell division
blocked by p53 occurs normally ! Among the targets phosphorylated by cyclin-CDKs
FIGURE 12–49 Regulation of passage from G1 to S by phosphorylation of are proteins of the nuclear envelope and proteins
pRb. Transcription factor E2F promotes transcription of genes for certain required for cytokinesis and DNA repair.
enzymes essential to DNA synthesis. The retinoblastoma protein, pRb,
can bind E2F (lower left), inactivating it and preventing transcription of
these genes. Phosphorylation of pRb by CDK2 prevents it from binding
12.12 Oncogenes, Tumor Suppressor
and inactivating E2F, and the genes are transcribed, allowing cell division. Genes, and Programmed Cell Death
Damage to the cell’s DNA (upper left) triggers a series of events that
inactivate CDK2, blocking cell division. When the protein MRN detects Tumors and cancer are the result of uncontrolled cell
damage to the DNA, it activates two protein kinases, ATM and ATR, and division. Normally, cell division is regulated by a family
they phosphorylate and activate the transcription factor p53. Active p53 of extracellular growth factors, proteins that cause rest-
promotes the synthesis of another protein, p21, an inhibitor of CDK2. Inhi- ing cells to divide and, in some cases, differentiate. The
bition of CDK2 stops the phosphorylation of pRb, which therefore contin- result is a precise balance between the formation of new
ues to bind and inhibit E2F. With E2F inactivated, genes essential to cell cells (such as skin cells that die and are replaced every
division are not transcribed and cell division is blocked. When DNA has few weeks, or white blood cells that are replaced every
been repaired, this inhibition is released, and the cell divides. few days) and cell destruction. When this balance is
12.12 Oncogenes, Tumor Suppressor Genes, and Programmed Cell Death 489
disturbed by defects in regulatory proteins, the result is EGF-binding

EGF
sometimes the formation of a clone of cells that divide domain
Extracellular space
repeatedly and without regulation (a tumor) until their
presence interferes with the function of normal tissues—
cancer. The direct cause is almost always a genetic
defect in one or more of the proteins that regulate cell Outside
division. In some cases, a defective gene is inherited
from one parent; in other cases, the mutation occurs
Inside
when a toxic compound from the environment (a muta-
gen or carcinogen) or high-energy radiation interacts Tyrosine kinase domain
with the DNA of a single cell to damage it and introduce EGF-binding site Binding of EGF Tyrosine
a mutation. In most cases there is both an inherited and empty; tyrosine activates kinase is
an environmental contribution, and in most cases, more kinase is inactive. tyrosine kinase. constantly
active.
than one mutation is required to cause completely
Normal EGF receptor ErbB protein
unregulated division and full-blown cancer.
FIGURE 12–50 Oncogene-encoded defective EGF receptor. The product
of the erbB oncogene (the ErbB protein) is a truncated version of the
Oncogenes Are Mutant Forms of the Genes normal receptor for epidermal growth factor (EGF). Its intracellular
for Proteins That Regulate the Cell Cycle domain has the structure normally induced by EGF binding, but the pro-
Oncogenes were originally discovered in tumor- tein lacks the extracellular binding site for EGF. Unregulated by EGF,
causing viruses, then later found to be derived ErbB continuously signals cell division.
from genes in the animal host cells, proto-oncogenes,
which encode growth-regulating proteins. During a viral in erbB2, the gene for a receptor Tyr kinase related to
infection, the host DNA sequence of a proto-oncogene ErbB, are commonly associated with cancers of the
is sometimes copied into the viral genome, where it glandular epithelium in breast, stomach, and ovary.
proliferates with the virus. In subsequent viral infection (For an explanation of the use of abbreviations in nam-
cycles, the proto-oncogenes can become defective by ing genes and their products, see Chapter 25.)
truncation or mutation. Viruses, unlike animal cells, do The prominent role played by protein kinases in
not have effective mechanisms for correcting mistakes signaling processes related to normal and abnormal cell
during DNA replication, so they accumulate mutations division has made them a prime target in the develop-
rapidly. When a virus carrying an oncogene infects a ment of drugs for the treatment of cancer (Box 12–5).
new host cell, the viral DNA (and oncogene) can be Mutant forms of the G protein Ras are common in
incorporated into the host cell’s DNA, where it can now tumor cells. The ras oncogene encodes a protein with
interfere with the regulation of cell division in the host normal GTP binding but no GTPase activity. The
cell. In an alternative, nonviral mechanism, a single cell mutant Ras protein is therefore always in its activated
in a tissue exposed to carcinogens may suffer DNA dam- (GTP-bound) form, regardless of the signals arriving
age that renders one of its regulatory proteins defective, through normal receptors. The result can be unregu-
with the same effect as the oncogenic mechanism: failed lated growth. Mutations in ras are associated with 30%
regulation of cell division. to 50% of lung and colon carcinomas and more than
The mutations that produce oncogenes are geneti- 90% of pancreatic carcinomas. ■
cally dominant; if either of a pair of chromosomes con-
tains a defective gene, that gene product sends the sig-
nal “divide” and a tumor may result. The oncogenic
Defects in Certain Genes Remove Normal Restraints
defect can be in any of the proteins involved in commu- on Cell Division
nicating the “divide” signal. Oncogenes discovered thus Tumor suppressor genes encode proteins that
far include those that encode secreted proteins, growth normally restrain cell division. Mutation in one or
factors, transmembrane proteins (receptors), cytoplas- more of these genes can lead to tumor formation.
mic proteins (G proteins and protein kinases), and the Unregulated growth due to defective tumor suppressor
nuclear transcription factors that control the expression genes, unlike that due to oncogenes, is genetically
of genes essential for cell division (Jun, Fos). recessive; tumors form only if both chromosomes of a
Some oncogenes encode surface receptors with pair contain a defective gene. This is because the func-
defective or missing signal-binding sites, such that their tion of these genes is to prevent cell division, and if
intrinsic Tyr kinase activity is unregulated. For exam- either copy of the gene for such a protein is normal, the
ple, the oncoprotein ErbB is essentially identical to the normal inhibition of division will take place. In a person
normal receptor for epidermal growth factor, except who inherits one correct copy and one defective copy,
that ErbB lacks the amino-terminal domain that nor- every cell begins with one defective copy of the gene. If
mally binds EGF (Fig. 12–50) and as a result sends the any one of those 1012 somatic cells undergoes mutation
“divide” signal whether EGF is present or not. Mutations in the one good copy, a tumor may grow from that doubly
490 Biosignaling
BOX 12–5 MEDICINE Development of Protein Kinase Inhibitors for Cancer Treatment
When a single cell divides without any regulatory acute myeloid leukemia, a relatively rare blood disease
limitation, it eventually gives rise to a clone of cells so (,5,000 cases a year in the United States). Another
large that it interferes with normal physiological func- group of oncogenes encode unregulated cyclin-dependent
tions (Fig. 1). This is cancer, a leading cause of death protein kinases. In each of these cases, specific protein
in the developed world, and increasingly so in the kinase inhibitors might be valuable chemotherapeutic
developing world. In all types of cancer, the normal agents in the treatment of disease. Not surprisingly,
regulation of cell division has become dysfunctional huge efforts are under way to develop such inhibitors.
due to defects in one or more genes. For example, How should one approach this challenge?
genes encoding proteins that normally send intermit- Protein kinases of all types show striking conserva-
tent signals for cell division become oncogenes, pro- tion of structure at the active site. All share with the
ducing constitutively active signaling proteins, or prototypical PKA structure the features shown in Figure 2:
genes encoding proteins that normally restrain cell two lobes that enclose the active site, with a P loop that
division (tumor suppressor genes) mutate to produce helps to align and bind the phosphoryl groups of ATP, an
proteins that lack this braking function. In many activation loop that moves to open the active site to the
tumors, both kinds of mutation have occurred. protein substrate, and a C helix that changes position as
Many oncogenes and tumor suppressor genes the enzyme is activated, bringing the residues in the
encode protein kinases or proteins that act in path- substrate-binding cleft into their binding positions.
ways upstream from protein kinases. It is therefore The simplest protein kinase inhibitors are ATP ana-
reasonable to hope that specific inhibitors of protein logs that occupy the ATP-binding site but cannot serve as
kinases could prove valuable in the treatment of can- phosphoryl group donors. Many such compounds are
cer. For example, a mutant form of the EGF receptor known, but their clinical usefulness is limited by their lack
is a constantly active receptor Tyr kinase (RTK), sig- of selectivity—they inhibit virtually all protein kinases and
naling cell division whether EGF is present or not would produce unacceptable side effects. More selectivity
(see Fig. 12–50). In about 30% of all women with
invasive breast cancer, a mutation in the receptor
gene HER2/neu yields an RTK with activity increased
up to 100-fold. Another RTK, vascular endothelial
growth factor receptor (VEGFR), must be acti-
vated for the formation of new blood vessels (angio- P loop C helix
Amino-
genesis) to provide a solid tumor with its own blood terminal
supply, and inhibition of VEGFR might starve a tumor lobe
of essential nutrients. Nonreceptor Tyr kinases can ATP
Activation
also mutate, resulting in constant signaling and unreg- loop
ulated cell division. For example, the oncogene Abl
(from the Abelson leukemia virus) is associated with
Mg2+
Catalytic
loop
Inhibitor (PD318088)
Carboxyl-
in substrate-binding cleft
terminal
lobe
FIGURE 2 Conserved features of the active site of protein kinases (PDB ID

1S9I). The amino-terminal and carboxyl-terminal lobes surround the active
site of the enzyme, near the catalytic loop and the site where ATP binds.
The activation loop of this and many other kinases undergoes phosphory-
lation, then moves away from the active site to expose the substrate-binding
FIGURE 1 Unregulated division of a single cell in the colon led to a cleft, which in this image is occupied by a specific inhibitor of this enzyme,
primary cancer that metastasized to the liver. Secondary cancers are PD318088. The P loop is essential in the binding of ATP, and the C helix
seen as white patches in this liver obtained at autopsy. must also be correctly aligned for ATP binding and kinase activity.
is seen with compounds that fill part of the ATP- (a) Imatinib (Gleevec) (b) Erlotinib (Tarceva)
binding site but also interact outside this site, with bound to Abl bound to EGF-R
parts of the protein unique to the target protein

kinase. A third possible strategy is based on the fact
that although the active conformations of all protein
kinases are similar, their inactive conformations are
not. Drugs that target the inactive conformation of a
specific protein kinase and prevent its conversion to
the active form may have a higher specificity of
action. A fourth approach employs the great specific-
ity of antibodies. For example, monoclonal antibod-
ies (p. 178) that bind the extracellular portions of
specific RTKs could eliminate the receptors’ kinase N
activity by preventing dimerization or by causing N N
their removal from the cell surface. In some cases, an H
antibody selectively binding to the surface of cancer
HN NH
cells could cause the immune system to attack those
cells. N N O N O O
The search for drugs active against specific N
protein kinases has yielded encouraging results. N O
O
For example, imatinib mesylate (Gleevec; Fig. 3a),
one of the small-molecule inhibitors, has proved Imatinib (Gleevec) Erlotinib (Tarceva)
nearly 100% effective in bringing about remission in
(c) ATP (d) Roscovatine
patients with early-stage chronic myeloid leukemia. bound to CDK2 bound to CDK2
Erlotinib (Tarceva; Fig. 3b), which targets EGFR, is
effective against advanced non-small-cell lung can-
cer (NSCLC). Because many cell-division signaling
systems involve more than one protein kinase,
inhibitors that act on several protein kinases may be
useful in the treatment of cancer. Sunitinib (Sutent)
and sorafenib (Nexavar) target several protein
kinases, including VEGFR and PDGFR. These two
drugs are in clinical use for patients with gastroin-
testinal stromal tumors and advanced renal cell
carcinoma, respectively. Trastuzumab (Herceptin), HN
cetuximab (Erbitux), and bevacizumab (Avastin) N
are monoclonal antibodies that target HER2/neu, N
EGFR, and VEGFR, respectively; all three drugs are N
NH N
in clinical use for certain types of cancer. Detailed
knowledge of the structure around the ATP-binding CH3
H3C
site makes it possible to design drugs that inhibit a
specific protein kinase by (1) blocking the critical OH CH3
ATP-binding site, while (2) interacting with resi-
Roscovatine
dues around that site that are unique to that par-
ticular protein kinase. FIGURE 3 Some protein kinase inhibitors now in clinical trials or clinical use, show-
At least a hundred more compounds are in pre- ing their binding to the target protein. (a) Imatinib binds to the Abl oncogene
clinical trials. Among the drugs being evaluated are kinase active site (PDB ID 1IEP); it occupies both the ATP-binding site and a
some obtained from natural sources and some pro- region adjacent to that site. (b) Erlotinib binds to the active site of EGFR (PDB ID
duced by synthetic chemistry. Indirubin is a compo- 1M17). (c), (d) Roscovatine is an inhibitor of the cyclin-dependent kinase CDK2;
nent of a Chinese herbal preparation traditionally shown here are normal Mg-ATP binding (c) at the active site (PDB ID 1S9I) and
used to treat certain leukemias; it inhibits CDK2 roscovatine binding (d), which prevents the binding of ATP (PDB ID 2A4L).
and CDK5. Roscovatine (Fig. 3d), a substituted adenine, cancer drugs heading toward clinical testing, it is
has a benzyl ring that makes it highly specific as an realistic to hope that some will prove more effective
inhibitor of CDK2. With several hundred potential anti- or more target-specific than those now in use.
492 Biosignaling
mutant cell. Mutations in both copies of the genes for Mutations in oncogenes and tumor suppressor
pRb, p53, or p21 yield cells in which the normal restraint genes do not have an all-or-none effect. In some cancers,
on cell division is lost and a tumor forms. perhaps in all, the progression from a normal cell to a
Retinoblastoma occurs in children and causes blind- malignant tumor requires an accumulation of mutations
ness if not surgically treated. The cells of a retinoblas- (sometimes over several decades), none of which, alone,
toma have two defective versions of the Rb gene (two is responsible for the end effect. For example, the devel-
defective alleles). Very young children who develop opment of colorectal cancer has several recognizable
retinoblastoma commonly have multiple tumors in both stages, each associated with a mutation (Fig. 12–51).
eyes. These children have inherited one defective copy If an epithelial cell in the colon undergoes mutation of
of the Rb gene, which is present in every cell; each both copies of the tumor suppressor gene APC (adeno-
tumor is derived from a single retinal cell that has matous polyposis coli), it begins to divide faster than
undergone a mutation in its one good copy of the Rb normal and produces a clone of itself, a benign polyp
gene. (A fetus with two mutant alleles in every cell is (early adenoma). For reasons not yet known, the APC
nonviable.) People with retinoblastoma who survive mutation results in chromosomal instability, and whole
childhood also have a high incidence of cancers of the regions of a chromosome are lost or rearranged during
lung, prostate, and breast later in life. cell division. This instability can lead to another muta-
A far less likely event is that a person born with two tion, commonly in ras, that converts the clone into an
good copies of the Rb gene will have independent muta- intermediate adenoma. A third mutation (often in the
tions in both copies in the same cell. Some individuals tumor suppressor gene DCC) leads to a late adenoma.
do develop retinoblastomas later in childhood, usually Only when both copies of p53 become defective does
with only one tumor in one eye. These individuals were this cell mass become a carcinoma—a malignant, life-
presumably born with two good copies (alleles) of Rb in threatening tumor. The full sequence therefore requires
every cell, but both Rb alleles in a single retinal cell have at least seven genetic “hits”: two on each of three
undergone mutation, leading to a tumor. After about tumor suppressor genes (APC, DCC, and p53) and one
age three, retinal cells stop dividing, and retinoblasto- on the proto-oncogene ras. There are probably several
mas at later ages are quite rare. other routes to colorectal cancer as well, but the prin-
Stability genes (also called caretaker genes) encode ciple that full malignancy results only from multiple
proteins that function in the repair of major genetic mutations is likely to hold true for all of them. When a
defects that result from aberrant DNA replication, ioniz- polyp is detected in the early adenoma stage and the
ing radiation, or environmental carcinogens. Mutations cells containing the first mutations are removed surgi-
in these genes lead to a high frequency of unrepaired cally, late adenomas and carcinomas will not develop;
damage (mutations) in other genes, including proto- hence the importance of early detection. Cells and
oncogenes and tumor suppressor genes, and thus to organisms, too, have their early detection systems. For
cancer. Among the stability genes are ATM (see Fig. example, the ATM and ATR proteins described in Sec-
12–49); the XP gene family, in which mutations lead to tion 12.11 can detect DNA damage too extensive to be
xeroderma pigmentosum; and the BRCA1 genes associ- repaired effectively. They then trigger, through a path-
ated with some types of breast cancer (see Box 25–1). way that includes p53, the process of apoptosis, in
Mutations in the gene for p53 also cause tumors; in more which a cell that has become dangerous to the organism
than 90% of human cutaneous squamous cell carcinomas kills itself. ■
(skin cancers) and in about 50% of all other human can-
cers, p53 is defective. Those very rare individuals who
inherit one defective copy of p53 commonly have the Apoptosis Is Programmed Cell Suicide
Li-Fraumeni cancer syndrome, with multiple cancers (of Many cells can precisely control the time of their own
the breast, brain, bone, blood, lung, and skin) occurring death by the process of programmed cell death, or
at high frequency and at an early age. The explanation for apoptosis (app#-a-toe#-sis; from the Greek for “drop-
multiple tumors in this case is the same as that for Rb ping off,” as in leaves dropping in the fall). One trigger
mutations: an individual born with one defective copy of for apoptosis is irreparable damage to DNA. Pro-
p53 in every somatic cell is likely to suffer a second p53 grammed cell death also occurs during the development
mutation in more than one cell during his or her lifetime. of an embryo, when some cells must die to give a tissue
In summary, then, three classes of defects can con- or organ its final shape. Carving fingers from stubby
tribute to the development of cancer: oncogenes, in limb buds requires the precisely timed death of cells
which the defect is the equivalent of a car’s accelerator between developing finger bones. During development
pedal being stuck down, with the engine racing; mutated of the nematode C. elegans from a fertilized egg, exactly
tumor suppressor genes, in which the defect leads to 131 cells (of a total of 1,090 somatic cells in the embryo)
the equivalent of brake failure; and mutated stability must undergo programmed death in order to construct
genes, with the defect leading to unrepaired damage to the adult body.
the cell’s replication machinery, the equivalent of an Apoptosis also has roles in processes other than
unskilled car mechanic. development. If a developing antibody-producing cell
MMR APC CDC4 KRAS PI3K TP53 SMAD4

(DNA b-catenin (ubiquitin- BRAF (p53) TGFBR2
Genes repair) dependent
proteolysis)
?
Epithelium
Smooth muscle
Connective tissue
Progression: Normal epithelium Small adenoma Large adenoma Cancer Metastasis
Tumor suppressor
COX-2 15-PGDH factors turned off
in colon cancer
Growth factors Oncogenic EGFR (epidermal growth factor receptor)
mediators
turned on in
colon cancer TGF-b (tumor growth factor)
FIGURE 12–51 Multistep transition from normal epithelial cell to synthesis of this enzyme, lead to a further strengthening of the signal:
colorectal cancer. Serial mutations in oncogenes (green) or tumor sup- divide now. When a cell in one of the polyps undergoes further muta-
pressor genes (red) lead to progressively less control of cell division, tions, in the tumor suppressor genes DCC and p53 (see Fig. 12–49) for
until finally an active tumor forms, which can sometimes metastasize example, increasingly aggressive tumors form. Finally, mutations in
(spread from the initial site to other regions of the body). Mutation of other tumor suppressor genes such as SMAD4 lead to a malignant
the MMR gene leads to defective DNA repair and consequently to a tumor and sometimes to a metastatic tumor that can spread to other
higher rate of mutation. Mutations in both copies of the tumor suppres- tissues. A second type of mutation that can add to the deleterious
sor gene APC lead to benign clusters of epithelial cells that multiply too effects is one that affects the production or action of growth factors or
rapidly (early adenoma). The CDC4 oncogene results in defective ubiqui- their receptors (bottom). Mutations in EGFR (epidermal growth factor
tination, which is essential to the regulation of cyclin-dependent kinases receptor) or TGF-! (transforming growth factor-!) favor uncontrolled
(see Fig. 12–47). The oncogenes KRAS and BRAF encode ras and raf pro- growth, as do mutations in the enzymes that produce certain prosta-
teins (see Fig. 12–15), and this further disruption of signaling leads to the glandins (COX-2; cyclooxygenase; see pp. 845–846) or 15-PGDH
formation of a large adenoma, which may be detected by colonoscopy (15-hydroxyprostaglandin dehydrogenase). Most malignant tumors of
as a benign polyp. Oncogenic mutations in the PI3K gene that encodes other tissues probably result from a series of mutations such as this,
the enzyme phosphoinositide-3 kinase, or in PTEN, which regulates the although not necessarily these particular genes, or in this order.
generates antibodies against a protein or glycoprotein (TNF), produced by cells of the immune system, inter-
normally present in the body, that cell undergoes pro- acts with cells through specific TNF receptors. These
grammed death in the thymus gland—an essential receptors have TNF-binding sites on the outer face of
mechanism for eliminating anti-self antibodies (the the plasma membrane and a “death domain” (,80 amino
cause of many autoimmune diseases). The monthly acid residues) that carries the self-destruct signal through
sloughing of cells of the uterine wall (menstruation) is the membrane to cytosolic proteins such as TRADD
another case of apoptosis mediating normal cell death. (TNF receptor–associated death domain) (Fig. 12–52).
The dropping of leaves in the fall is the result of apop- Another receptor, Fas, has a similar death domain that
tosis in specific cells of the stem. Sometimes cell suicide allows it to interact with the cytosolic protein FADD
is not programmed but occurs in response to biological (Fas-associated death domain), which activates the
circumstances that threaten the rest of the organism. cytosolic protease caspase 8. This enzyme belongs to a
For example, a virus-infected cell that dies before family of proteases that participate in apoptosis; all
completion of the infection cycle prevents spread of the are synthesized as inactive proenzymes, all have a
virus to nearby cells. Severe stresses such as heat, critical Cys residue at the active site, and all hydrolyze
hyperosmolarity, UV light, and gamma irradiation also their target proteins on the carboxyl-terminal side of
trigger cell suicide; presumably the organism is better specific Asp residues (hence the name caspase, from
off with any aberrant, potentially mutated cells dead. Cys and Asp).
The regulatory mechanisms that trigger apoptosis When caspase 8, an “initiator” caspase, is activated
involve some of the same proteins that regulate the cell by an apoptotic signal carried through FADD, it further
cycle. The signal for suicide often comes from outside, self-activates by cleaving its own proenzyme form. Mito-
through a surface receptor. Tumor necrosis factor chondria are one target of active caspase 8. The protease
494 Biosignaling
FIGURE 12–52 Initial events of apoptosis. Apoptosis-triggering signals

Fas ligand TNFa
from outside the cell (Fas and TNF") bind to their specific receptors in
the plasma membrane (FasR, TNF"R). The occupied receptors interact
Fas receptor TNFa receptor with the cytosolic proteins FADD and TRADD through an 80 amino acid
sequence called the “death domain” on both the receptors and these cyto-
solic targets, which become activated. Activation of FADD and TRADD
Plasma
initiates a proteolytic cascade that leads to apoptosis. FADD and TRADD
membrane
Death activate caspase-8, which acts to release cytochrome c from mitochon-
domains dria, which, in concert with protein Apaf-1, activates caspase-9, triggering
FADD TRADD
apoptosis.
Procaspase-8 Caspase-8 Procaspase-8
Mitochondrion
Cytochrome c Caspases
3, 6, 7
Apaf-1
Apoptosome
Procaspase-9 Caspase-9
Apoptosis
causes the release of certain proteins contained between dominant and may encode defective growth
the inner and outer mitochondrial membranes: cyto- factors, receptors, G proteins, protein kinases, or
chrome c (Chapter 19) and several “effector” caspases. nuclear regulators of transcription.
!
Cytochrome c binds to the proenzyme form of the effec- Tumor suppressor genes encode regulatory
tor enzyme caspase 9 and stimulates its proteolytic acti- proteins that normally inhibit cell division;
vation. The activated caspase 9 in turn catalyzes whole- mutations in these genes are genetically recessive
sale destruction of cellular proteins—a major cause of but can lead to tumor formation.
apoptotic cell death. One specific target of caspase
! Cancer is generally the result of an accumulation
action is a caspase-activated deoxyribonuclease.
In apoptosis, the monomeric products of protein of mutations in oncogenes and tumor suppressor
and DNA degradation (amino acids and nucleotides) genes.
are released in a controlled process that allows them to ! When stability genes, which encode proteins
be taken up and reused by neighboring cells. Apoptosis necessary for the repair of genetic damage, are
thus allows the organism to eliminate a cell that is mutated, other mutations go unrepaired, including
unneeded or potentially dangerous without wasting its mutations in proto-oncogenes and tumor
components. suppressor genes that can lead to cancer.
! Apoptosis is programmed and controlled cell death
SUMMARY 12.12 Oncogenes, Tumor Suppressor Genes, that functions during normal development and
and Programmed Cell Death adulthood to get rid of unnecessary, damaged, or
! Oncogenes encode defective signaling proteins. By infected cells. Apoptosis can be triggered by
continually giving the signal for cell division, they extracellular signals such as TNF, acting through
lead to tumor formation. Oncogenes are genetically plasma membrane receptors.

Chapter 12

Uploaded by

Copyright:

Available Formats

Chapter 12

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

Chapter 12

Uploaded by

Copyright:

Available Formats

c12Biosignaling.

indd Page 433 12/09/12 1:33 PM user-F408 /Users/user-F408/Desktop

Net D[X] or DVm

(c) Modularity Signal

12.1 General Features of Signal Transduction 435

BOX 12–1 METHODS Scatchard Analysis Quantifies the Receptor-Ligand Interaction

Bound hormone, [RL]

This binding, like that of an enzyme to its substrate,

Bound hormone , [RL]

result of a signaling pathway is the phosphorylation of

FIGURE 12–2 Six general types of signal transducers.

12.2 G Protein–Coupled Receptors and Second Messengers 437

12.2 G Protein–Coupled Receptorsand Second Messengers 439

GDP GTP GTP ATP

2 Hormone-receptor 3 Activated Gsa 4 Adenylyl 5 cAMP 6 Phosphorylation

NH2 NH2 NH2

Gs with GDP Contact of Gs Gs with GTP

(a) (b) Substrate-

12.2 G Protein–Coupled Receptorsand Second Messengers 441

BOX 12–2 MEDICINE G Proteins: Binary Switches in Health and Disease

FIGURE 3 When bound GTP is hydrolyzed by the GTPase activities of

12.2 G Protein–Coupled Receptorsand Second Messengers 443

Normal Gs: GTPase activity HO OH

ADP-ribosylated Gs: GTPase HO OH

Several Mechanisms Cause Termination of

12.2 G Protein–Coupled Receptorsand Second Messengers 445

1 Binding of epinephrine (E) to 2 Gsbg recruits bARK to the mem-

near the carboxyl terminus of the receptor, which is on

12.2 G Protein–Coupled Receptorsand Second Messengers 447

below). AKAPs (A kinase anchoring proteins) are b-adrenergic

Related Roles as Second Messengers

BOX 12–3 METHODS FRET: Biochemistry Visualized in a Living Cell

CFP GFP cAMP-dependent protein kinase (PKA)

proteins are expressed in a cell, BFP (donor; excita-

CFP YFP 476 527

FIGURE 12–10 Hormone-activated phospholipase C and IP3. Two intra-

3 Gqa, with bound GTP, moves

4 Active PLC cleaves PIP2

5 IP3 binds to a specific Diacylglycerol

6 Diacylglycerol and Ca2"activate

12.2 G Protein–Coupled Receptorsand Second Messengers 451

actions, affecting neuronal and immune function and (a)

Calcium Is a Second Messenger That May Be Localized

complex assembled on scaffold proteins such as AKAPs.

GPCRs Mediate the Actions of a

as well as light and compounds detected by olfaction

300 the !-adrenergic receptor with its G protein, and the

12.3 Receptor Tyrosine Kinases 453

(c) Inactive (unphosphorylated) (d) Active (triply phosphorylated)

12.3 Receptor Tyrosine Kinases 455

Insulin receptor binds insulin

12.3 Receptor Tyrosine Kinases 457

The JAK-STAT Signaling System Also Involves

12.4 Receptor Guanylyl Cyclases, cGMP, and Protein Kinase G 459

and nucleus. The result is specific metabolic ANF Guanylin and

! The enzyme PI3K, activated by interaction with Extracellular Ligand-binding

FIGURE 12–20 Two types of guanylyl cyclase that participate in signal

and Protein Kinase G

12.5 Multivalent Adaptor Proteins and Membrane Rafts 461

Adaptor PTB SH2 Shc PIP3