Voet D., Voet Ju.G. Biochemistry

Подождите немного. Документ загружается.

domain, as we have seen for PLC (Section 19-4Ba), also

does so via Ca

2⫹

-mediated binding to the membrane’s

phosphatidylserine head groups. These interactions are

synergistic in that the greater the Ca

2⫹

concentration, the

lower the concentration of phorbol ester or DAG neces-

sary to activate PKC and vice versa. Nevertheless, both the

C1 and C2 domains must be membrane anchored in order

to activate the protein kinase. This is because the confor-

mation required to do so extracts the N-terminal pseudo-

substrate from the protein kinase active site.

b. PKC Is Primed by Phosphorylation

The activation of all mammalian PKCs but the atypical

PKCs is accompanied by their phosphorylation at three

conserved Ser or Thr residues. One of these residues (Thr

500 in PKC␤II) is in the protein kinase’s activation loop,

whereas the remaining two are in its C-terminal segment

(Thr 641 and Ser 660 of the 673-residue PKC␤II). In the

atypical PKCs, the latter Ser/Thr residue is replaced by a

phosphate-mimicking Glu residue. The sequence of events

that activate PKCs, which was largely elucidated by

Alexandra Newton, occurs as follows (Fig. 19-60):

1. Newly synthesized PKC binds to the membrane

(or possibly to the underlying cytoskeleton), where

phosphoinositide-dependent protein kinase-1 (PDK1)

phosphorylates its activation loop (at Thr 500 in PKC␤II).

The resulting negative charge on the activation loop is pos-

tulated to properly align the active site residues of PKC for

catalysis, much as we have seen for PKA (Section 18-3Cb;

Section 19-4. The Phosphoinositide Cascade 731

500

641

660

500

641

660

off

Membrane

PDK1

Activation

loop

PKC

DAG

Figure 19-60 Activation of PKC. (1) Newly synthesized PKC

is phosphorylated on its activation loop (here represented by Thr

500 of PKC␤II; yellow ball) by phosphoinositide-dependent

protein kinase-1 (PDK1), which is tethered to the membrane via

its C-terminal pleckstrin homology domain (PH). (2) The now

catalytically competent PKC autophosphorylates 2 sites on its

C-terminal segment (here represented by Thr 641 and Ser 660 of

PKC␤II). However, the N-terminal pseudosubstrate segment

now binds to PKC’s active site, so that the enzyme remains

inactive. (3) On the binding of PKC’s C1 domain to

membrane-bound DAG (the product of extracellular signals

Figure 19-59 X-ray structure of the C1B motif of PKC␦ in complex with

12-O-myristoylphorbol-13-acetate. The protein tetrahedrally ligands two Zn

2⫹

ions (cyan spheres), each via a His side chain and three Cys side chains. These

side chains are shown in ball-and-stick form with C green, N blue, O red, and S

yellow, as is the bound 12-O-myristoylphorbol-13-actetate. The C1B domains

of PKC␣ and ␥ have similar structures. [Based on an X-ray structure by James

Hurley, NIH. PDBid 1PTR.]

inducing phosphoinositide hydrolysis) together with the

2⫹

-mediated binding of the C2 domain to phosphatidylserine

(PS) in the membrane, the pseudosubstrate is ejected from the

PKC active site, thereby yielding active enzyme. [After a drawing

by Toker,A. and Newton,A.C., Cell 103, 187 (2000).]

JWCL281_c19_671-743.qxd 7/1/10 1:13 PM Page 731

PKA’s activation loop is also phosphorylated by PDK1). In

fact, the mutagenic replacement of PKC␣’s activation loop

Thr with a neutral nonphosphorylatable residue yields an in-

activatable enzyme, whereas its replacement with Glu yields

an enzyme that requires only DAG and Ca

2⫹

for activation.

2. The now catalytically competent PKC rapidly au-

tophosphorylates its other two phosphorylation sites. The

autophosphorylation of Thr 641 appears to lock PKC into

its active conformation, as suggested by the observation

that in PKC␤II, which has been phosphorylated at only Thr

500 and Thr 641, the selective dephosphorylation of Thr 500

yields active enzyme. The autophosphorylation of the third

phosphorylation site correlates with the release of PKC into

the cytosol, where PKC is maintained in its inactive state by

the binding of its pseudosubstrate to its active site.

3. This autoinhibition is relieved, as described above,

when PKC again binds to the membrane via DAG binding

to its C1 domain and Ca

2⫹

-mediated binding of its C2 do-

main to phosphatidylserine (PS).

The activity of PKC is regulated by the protein phosphatase

PHLPP (for PH domain leucine-rich repeat protein phos-

phatase), which specifically dephosphorylates its Thr 641.

D. The Phosphoinositide 3-Kinases

The inositol head group of phosphatidylinositol has 5 free

hydroxyl groups that can be phosphorylated (Fig. 19-53).

However, only its 3-, 4-, and 5-positions are known to be

phosphorylated in vivo, and these occur in all seven possible

combinations (Fig. 19-61), each of which participates in sig-

naling. In addition to the plasma membrane, they occur in

ER, Golgi, and endosome membranes, although with dif-

ferent distributions in each of these several subcellular

compartments.

The phosphorylations of these various phosphoinosi-

tides are catalyzed by ATP-dependent enzymes known as

phosphoinositide 3-kinases (PI3Ks), phosphoinositide 4-

kinases (PIP4Ks), and phosphoinositide 5-kinases (PIP5Ks).

Their various products function as second messengers by

recruiting the proteins that bind them to the cytosolic sur-

face of the plasma membrane (see below). The resulting

colocalization of enzymes and substrates results in further

signaling activity that controls such vital functions as cell

survival, proliferation, cytoskeletal rearrangement, endo-

cytosis, and vesicle trafficking.

The PI3Ks are the presently best understood phospho-

inositide kinases. Consequently, in this subsection, we dis-

cuss the PI3Ks and their products as a paradigm of all

phosphoinositide kinases and the signals they produce.

a. PI3Ks Have Three Classes

Mammalian PI3Ks are divided into three classes ac-

cording to their structures (Fig. 19-62), substrate specifici-

ties, and modes of regulation:

1. Class I PI3Ks are heterodimeric receptor-regulated

enzymes that preferentially phosphorylate PIP

(alterna-

tively, PtdIns-4,5-P

). Their ⬃1070-residue catalytic sub-

units interact with Ras ⴢ GTP via a Ras-binding domain

(RBD) near their N-termini. Their regulatory subunits are

adaptor proteins that link the catalytic subunits to up-

stream signaling events and hence form two subclasses ac-

cording to the type of upstream effectors with which they

interact:

(a) The class IA PI3Ks (PI3K␣, ␤, and ␦) are activated

by RTKs via the mediation of the adaptor subunit p85 (of

which there are seven isoforms), which contains SH2 and

SH3 domains and may be phosphorylated on specific Tyr

side chains. The gene encoding PI3K␣ is one of the two

most frequently mutated oncogenes in human cancers, thus

indicating this enzyme’s central regulatory role. Several of

its most common oncogenic mutations result in enhanced

enzymatic activity by altering the interactions between its

kinase domain and its regulatory domains or p85.

732 Chapter 19. Signal Transduction

PtdIns

PtdIns-4-PPtdIns-3-P PtdIns-5-P

PtdIns-3,5-P

PtdIns-4,5-P

PtdIns-3,4-P

PtdIns-3,4,5-P

IP3

DAG

PIP5K

PtdIns4K

Class III (or I or II)

PI3K

Class I (or II)

PI3K

PIP4K

PIP5K

PI3K

Class I or II

PI3K

PIP4K

PLC

Class I

RBD C2

domain

Helical

domain

Kinase-

domain

Class II

Class III

Figure 19-61 Flowchart of reactions in the synthesis of

phosphoinositides in mammalian cells. PtdIns, PtdIns-4-P, and

PtdIns-4,5-P

(PIP

) are written in bold type to indicate their

abundance:Together they comprise ⬃90% of the cell’s total

phosphoinositides. PtdIns-3-P and PtdIns-5-P each comprise

2–5% of the total, whereas the levels of PtdIns-3,4-P

and

PtdIns-3,4,5-P

(PIP

) are barely detectable in quiescent cells but

rise to 1 to 3% of the total in stimulated cells. PtdIns-3,5-P

comprises ⬃2% of the phosphoinositides in fibroblasts. [After

Fruman, D.A., Meyers, R.E., and Cantley, L.C., Annu. Rev.

Biochem. 67, 501 (1998).]

Figure 19-62 Domain organization of the three classes of

PI3Ks. [After Walker, E.H., Persic, O., Ried, C., Stephens, L., and

Williams, R.L., Nature 402, 314 (1999).]

JWCL281_c19_671-743.qxd 6/11/10 6:59 AM Page 732

(b) The class IB PI3K, of which PI3K is its only mem-

ber, is activated by the G



dimers of heterotrimeric G pro-

teins, with its adaptor subunit p101 rendering it far more

sensitive to G



2. Class II PI3Ks (PI3K-C2, , and ) are ⬃1650-

residue monomers that are characterized by a C-terminal

C2 domain that does not bind Ca

2

. They preferentially

phosphorylate PtdIns and PtdIns-4-P. Since they lack

adaptors, the way in which class II PI3Ks are controlled is

unknown.

3. Class III PI3K, which has one known isoform, phos-

phorylates only PtdIns. It is a heterodimer with an 887-

residue catalytic subunit and an adaptor subunit known as

p150. Class III PI3K is constitutively active, that is, it is un-

regulated and hence is thought to be the cell’s main

provider of PtdIns-3-P, whose level is essentially unaltered

by cellular stimulation. It is thought to be the evolutionary

predecessor of the other classes because it is the only class

of PI3K present in yeast.

In addition to their lipid kinase activities, all PI3Ks have

Ser/Thr protein kinase activity, although the physiological

significance of this dual specificity is unclear.

b. PI3K Is a Multidomain Protein

The X-ray structure of PI3K  ATP, in which the PI3K

lacks its N-terminal 143 residues (which are important for

interaction with the p101 adaptor; the analogous portion of

PI3K interacts with its p85 adaptor), was determined by

Roger Williams. It reveals that its RBD, C2, and helical do-

mains form a relatively compact layer that packs against the

“back” of the kinase domain (Fig. 19-63). As expected, the

kinase domain is grossly similar to those of protein kinases

in that it is bilobal, with its N-lobe consisting largely of a

5-stranded  sheet and its C-lobe being predominantly helical.

However, there are also major differences between these

kinase domains, as can be seen by comparing the catalytic

domain in Fig. 19-63 with that in, for example, Fig. 19-28a.

The RBD domain of PI3K has the same fold as that of

RafRBD (Fig. 19-41). Indeed, in the X-ray structure of

PI3K–Ras  GMPPNP (Fig. 19-64), also determined by

Williams, the PI3K RBD interacts with Ras in a similar

manner as we have seen that RafRBD interacts with the

Ras homolog Rap1A (Fig. 19-41) in that they continue each

other’s central  sheets. However, Ras bound to PI3K is

rotated by 35° relative to Rap1A bound to RafRBD. Con-

tacts between the Switch I region of Ras and the PI3K sta-

bilize this interaction and ensure its dependence on

Ras  GTP. This complex also contains intermolecular con-

tacts involving the Switch II region of Ras. Such an interac-

tion had previously only been observed between Ras and its

upstream effectors. Comparison of the structure of the

PI3K–Ras complex with that of PI3K  ATP (Fig. 19-63)

indicates that Ras binding induces the C-lobe of PI3K’s

catalytic domain to pivot relative to its N-lobe in a way that

substantially alters the putative binding pocket for the

phosphoinositide head group. This, presumably, accounts

for the ⬃15-fold activation of PI3K on binding Ras  GTP.

Section 19-4. The Phosphoinositide Cascade 733

Figure 19-64 X-ray structure of PI3K–Ras  GMPPNP. Here

only the PI3K RBD (green) and the Ras  GMPPNP (gold) are

drawn, with the Switch I and Switch II regions of Ras magenta

and cyan and its bound GMPPNP shown in space-filling form

(C green, N blue, O red, and P yellow).The view, which is similar

to that of Fig. 19-41, is related to that in Fig. 19-63 by rotating it

clockwise by ⬃40° about its vertical axis and then turning it 180°

about the axis perpendicular to the page. [Based on an X-ray

structure by Roger Williams, MRC Laboratory of Molecular

Biology, Cambridge, U.K. PDBid 1HE8.]

Figure 19-63 X-ray structure of PI3K  ATP. The protein is

shown in ribbon form with its Ras-binding domain (RBD) green,

its C2 domain magenta, its helical domain orange, the N- and

C-lobes of its kinase domain pink and cyan, and interdomain

segments gray. The ATP is shown in space-filling form with

C green, N blue, O red, and P yellow. The protein is oriented

such that its kinase domain is seen in “standard” view. The protein

appears fragmented because several of its segments are

disordered, including much of the kinase’s activation loop.

[Based on an X-ray structure by Roger Williams, MRC

Laboratory of Molecular Biology, Cambridge, U.K. PDBid 1E8X.]

JWCL281_c19_671-743.qxd 3/16/10 7:17 PM Page 733

The C2 domain of PI3K forms the same sandwich of

two 4-stranded antiparallel  sheets seen in the C2 domain

of PLC-1 (Section 19-4Ba). However, in contrast to the

C2 domain of PLC-1, that of PI3K does not bind Ca

2

ions. Nevertheless, the PI3K C2 domain appears to partic-

ipate in membrane association, as indicated by the obser-

vation that this isolated C2 domain binds to phospholipid

vesicles with an affinity similar to that of the intact enzyme.

This interaction is presumably mediated by patches of ba-

sic residues on the surface of the C2 domain.

The PI3K helical domain consists of five repeating

pairs of antiparallel helices that form a superhelix, which

closely resembles that formed by the HEAT repeats in the

A subunit of protein phosphatase 2A (PP2A; Fig. 19-51a),

even though PI3K does not contain a HEAT sequence

motif. In analogy with the function of the A subunit of

PP2A to bind other proteins (Section 19-3Fe), it is pro-

posed that the largely solvent-exposed helical domain of

PI3K functions to interact with the proteins that bind

PI3K, such as its p101 adaptor and G



c. Akt Activation Requires Its PH Domain–Mediated

Binding to 3-Phosphoinositides

The PtdIns-3,4-P

and PtdIns-3,4,5-P

products of

PI3Ks (Fig. 19-61) bind to their downstream effectors

mainly via pleckstrin homology (PH) domains that prefer-

entially bind the head groups of these 3-phosphoinositides

rather than that of PIP

(as does the PH domain of PLC-;

Fig. 19-58). Another example of a PH domain-containing

protein that does so is the 556-residue phosphoinositide-

dependent protein kinase-1 (PDK1), which, as we have

seen, phosphorylates the activation loops of PKA and

PKC (Section 19-4Cb).

PDK1 also phosphorylates the Ser/Thr protein kinase

Akt [also called protein kinase B (PKB)], a proto-

oncogene product that is implicated in regulating multiple

biological processes including gene expression, apoptosis,

glucose uptake, and cellular proliferation, and hence

phosphorylates many target proteins. The ⬃480-residue

Akt consists of an N-terminal PH domain that binds

3-phosphoinositides and a C-terminal kinase domain that is

homologous to those of PKA and PKC (and is thus a mem-

ber of the AGC family of protein kinases).Akt is present in

multicellular organisms in three isoforms (Akt1/PKB,

Akt2/PKB, and Akt3/PKB) but is absent in yeast, which

suggests that it evolved from another AGC family member

coincidentally with multicellular organisms.

The full activation of Akt requires its phosphorylation

at both its Ser 473 and Thr 308. Ser 473 is phosphorylated

by mTORC2 [for mammalian target of rapomycin complex

2; rapamycin is an immunosuppressant similar to FK506

(Section 9-2B)]. This stimulates PDK1 to phosphorylate

Thr 308, which is located in Akt’s activation loop. Muta-

tions of the residues in Akt’s PH domain responsible for

lipid binding block its phosphorylation in vitro by PDK1.

However, the deletion of Akt’s PH domain overcomes this

enzyme’s need for binding to 3-phosphoinositides. This

suggests that the binding of Akt to these membrane-bound

lipids induces a conformational change that permits PDK1

to phosphorylate and hence activate Akt. It therefore ap-

pears that it is the 3-phosphoinositide-mediated colocaliza-

tion of Akt and PDK1 that leads to Akt activation and

hence that it is the action of PI3K that is functionally re-

sponsible for this process. In contrast, the PDK1-mediated

phosphorylation of PKA and PKC,which lack PH domains,

occurs in the absence of 3-phosphoinositides and is there-

fore constitutive. The protein phosphatase PHLPP regu-

lates Akt activity by dephosphorylating its Ser 473, in much

the same way as we have seen that PHLPP dephosphory-

lates PKC (Section 19-4Cb).

d. The FYVE Domain Binds the PtdIns-3-P

Head Group

The singly phosphorylated PtdIns-3-P is rarely bound

by PH domains. Rather, its direct effects are mediated by

FYVE domains [named after the four proteins in which it

was first identified: Fab1p, YOTB, Vac1p, and early endo-

some antigen 1 (EEA1)], which have been identified in

⬃60 proteins. For instance, the 1410-residue eukaryotic

protein EEA1, which has a 65-residue, C-terminal FYVE

domain, initiates endosome fusion in eukaryotic cells (Fig.

12-91) by recruiting the membrane-anchored small G pro-

tein Rab5 and the transmembrane SNARE protein syn-

taxin (Section 12-4Db).

The NMR structure of the EEA1 FYVE domain, deter-

mined by Michael Overduin, reveals that it assumes similar

conformations in the free state, when binding dibutanoyl-

PtdIns-3-P (Fig. 19-65), and when bound to dodecylphos-

phocholine (DPC) micelles enriched with this PtdIns-3-P.

The protein is largely held together by two bound Zn

2

ions, each of which is tetrahedrally liganded by four con-

served Cys side chains.The PtdIns-3-P head group is held in

its binding pocket by a network of electrostatic, hydrogen

bonding, and hydrophobic interactions involving a highly

conserved (R/K)(R/K)HHCR motif (RRHHCR in EEA1).

The NMR evidence indicates that, on the addition of

DPC micelles, the FYVE domain  PtdIns-3-P complex

inserts a hydrophobic 5-residue loop (FSVTV; orange in

Fig. 19-65), which is flanked by basic residues (blue in Fig.

19-65), into the lipid layer. This also occurs in the absence

of PtdIns-3-P but to a much lesser extent. Conversely,

membrane insertion increases the binding affinity of the

FYVE domain for PtdIns-3-P 20-fold (from 1 M to 50 nM).

The origin of this latter effect appears to be that the 10-

residue segment preceding the membrane insertion loop,

the unliganded protein’s most disordered region, becomes

more ordered and moves toward the binding pocket on

binding PtdIns-3-P. This has led to the proposal that the

FYVE domain is recruited to membranes via the insertion

of its hydrophobic loop into the lipid bilayer. This, in turn,

primes the protein for the recognition of PtdIns-3-P, whose

binding induces the protein’s otherwise mobile N-terminal

segment to clamp down over the PtdIns-3-P head group.

E. Inositol Polyphosphate Phosphatases

Signaling via the phosphoinositide cascade is terminated

through the actions of a variety of inositol phosphatases that

734 Chapter 19. Signal Transduction

JWCL281_c19_671-743.qxd 3/16/10 7:17 PM Page 734

are functionally classified as 1-, 3-, 4-, and 5-phosphatases.

We end our consideration of the phosphoinositide cascade

by discussing the characteristics of these essential enzymes.

a. The Inositol Polyphosphate 5-Phosphatases

Act In Numerous Signaling Pathways

The first inositol polyphosphate 5-phosphatases that

were studied hydrolyze IP

(Ins-1,4,5-P

) to IP

(Ins-1,4-P

)

and thereby terminate cellular Ca

2

mobilization (Fig.

19-54, bottom). Mammals express 10 isozymes that

have 5-phosphatase activity. These enzymes share a com-

mon catalytic core and have been classified according to

their substrate specificities into two groups: Type I

enzymes dephosphorylate inositol phosphates, whereas

type II enzymes, in addition, hydrolyze the corresponding

phosphoinositides.

Type I 5-phosphatases, which hydrolyze only IP

and

Ins-1,3,4,5-P

, are membrane-anchored via prenylation.

That expressed in blood platelets (a type of blood cell

that participates in blood clotting; Section 35-1), which is

representative of this group, forms a stoichiometric complex

with pleckstrin, a 350-residue protein that consists largely

of two PH domains. When platelets are stimulated by the

proteolytic clotting enzyme thrombin (Section 35-1B),

pleckstrin is phosphorylated on Ser and Thr residues by

PKC, which in turn activates its associated 5-phosphatase.

Note that PKC is activated by DAG, a product of PLC,

which simultaneously generates the type I 5-phosphatase

substrate IP

(Fig. 19-54). Hence, the PLC product IP

activates Ca

2

ion release, whereas its coproduct DAG

activates type I 5-phosphatase through pleckstrin phospho-

rylation to terminate the Ca

2

signal. This termination is

apparently important for normal cell growth as a decrease

in the expression of type I 5-phosphatase causes increased

and even uncontrolled (malignant) cell growth.

Type II 5-phosphatases share increased similarities in

their catalytic cores relative to type I enzymes and, in addi-

tion,have a so-called type II domain on the N-terminal side

of their catalytic cores. They occur in three main subtypes:

GIPs, SHIPs, and SCIPs. GIPs are so called because they

have a C-terminal GAP domain (GAP-containing inositol

phosphatase), although they have no demonstrated GAP

activity. GIPs hydrolyze IP

and Ins-1,3,4,5-P

and their

corresponding lipids, PtdIns-4,5-P

and PtdIns-3,4,5-P

,al-

though with varying catalytic efficiencies.

There are only two known GIPs, 5-phosphatase II and

OCRL. OCRL is so called because its mutation causes the

X-linked hereditary disease oculocerebrorenal dystrophy

(also called Lowe syndrome), which is characterized by con-

genital cataracts, progressive retinal degeneration, mental

retardation, and renal tubule defects leading to kidney fail-

ure in early adulthood. The 901-residue OCRL occurs

mainly on the surface of lysosomes, where it is anchored

through prenylation. Renal tubule cells from Lowe syn-

drome patients are deficient in PtdIns-4,5-P

and PtdIns-

3,4,5-P

hydrolytic activity, whereas the corresponding inos-

itol phosphates are hydrolyzed normally, thereby indicating

that OCRL is a lipid phosphatase. PtdIns-4,5-P

stimulates

the budding of membrane vesicles from lysosomes, so that

the accumulation of this lipid probably leads to abnormally

increased trafficking of enzymes from the lysosome to the

extracellular space. Indeed, the lysosomal enzymes in these

cells appear to be missorted (as are various lysosomal hy-

drolases in I-cell disease; Section 12-4Cg). It is therefore

proposed that this lifelong leakage of enzymes from the

lysosomes in Lowe syndrome patients causes tissue damage

that eventually results in kidney failure and blindness.

SHIPs only hydrolyze substrates that also have a phos-

phate in their 3-positions. The two known members of this

group, SHIP (for SH2-containing inositol-5-phosphatase)

and SHIP2, are ⬃1200-residue proteins that have an

N-terminal SH2 domain. Thus, these proteins can bind to

PTKs and, in fact, are phosphorylated by them to yield a

consensus binding sequence for PTB domains (NPXpY;

Section 19-3Cc). Moreover, they also contain a C-terminal

Pro-rich domain that may bind to SH3-containing proteins.

Thus, it appears that SHIP activity may be under the con-

trol of several systems. Indeed, SHIP, which is expressed

only in hematopoietic (blood-forming) cells, associates

Section 19-4. The Phosphoinositide Cascade 735

Figure 19-65 NMR structure of the EEA1 FYVE domain in

complex with PtdIns-3-P. The head group of PtdIns-3-P is drawn in

ball-and-stick form (C green, O red, P magenta, H gray).The

protein binds two Zn

2

ions (cyan spheres) that are each

tetrahedrally liganded by four Cys side chains that are drawn in

stick form (C green and S yellow).The 5-residue loop that inserts

into DPC micelles is orange and its flanking basic residues are

blue. [Based on an NMR structure by Michael Overduin,

University of Colorado Health Sciences Center. PDBid 1HYI.]

JWCL281_c19_671-743.qxd 3/16/10 7:17 PM Page 735

with the adaptor proteins Grb2 and Shc (Section 19-3Cf).

It functions to hydrolyze PtdIns-3,4,5-P

, which is impli-

cated in activating Akt and PLC. SHIP2 functions similarly

in nonhematopoietic cells, where it limits cellular responses

to insulin, EGF, and PDGF.

SCIPs (Sac1-containing inositol phosphatases) are so

named because they contain an N-terminal domain that is

homologous to the yeast phosphatidylinositol phosphatase

Sac1. The first SCIP to be characterized is named synap-

tojanin1 because it was purified from synaptic vesicles

and because the presence of two phosphatase domains is

reminiscent of the two kinase domains in Janus kinases

(JAKs; Section 19-3Eb).The 1575-residue synaptojanin1’s

5-phosphatase domain hydrolyzes PIP

and PtdIns-4,5-P

and its Sac1 phosphatase domain hydrolyzes PtdIns-3-P

and PtdIns-4-P. Synaptojanin1 is expressed only in neurons,

where it forms complexes with the G protein dynamin

(Section 12-4Cd) and thereby participates in synaptic vesi-

cle recycling. The closely similar synaptojanin2 is ubiqui-

tously expressed but its functions are largely unknown.

b. Inositol Polyphosphate 1-Phosphatase Is

Implicated in Bipolar Disorder

Mammals express only one type of inositol polyphos-

phate 1-phosphatase, a 399-residue enzyme that hydrolyzes

Ins-1,4-P

and Ins-1,3,4-P

(IP

) but does not act on lipid

substrates. This enzyme is inhibited by Li

⫹

ion. The thera-

peutic efficacy of Li

⫹

in controlling the incapacitating mood

swings of manic-depressive individuals (those with bipolar

disorder) therefore suggests that this mental illness is

caused by an aberration of 1-phosphatase in the brain, pos-

sibly resulting in abnormal activation of Ca

2⫹

-mobilizing

receptors (Fig. 19-54, bottom). Indeed, Drosophila in which

this 1-phosphatase has been deleted exhibit neurological

deficits (the so-called “shaker” phenotype) that appear

identical to those of wild-type Drosophila treated with Li

⫹

c. The Inositol Polyphosphate 3-Phosphatase PTEN

Is a Tumor Suppressor

The inositol polyphosphate 3-phosphatases undo the

actions of the PI3Ks. The best characterized of these en-

zymes is the 403-residue PTEN (for phosphatase and

tensin homolog; tensin is a cytoskeletal actin-binding pro-

tein), which in vitro dephosphorylates all 3-phosphorylated

phosphoinositides and Ins-1,3,4,5-P

. PTEN is a tumor sup-

pressor (a protein whose loss of function is a cause of in

cancer), presumably because its 3-phosphatase activity

functions to downregulate the PtdIns-3,4,5-P

-activated

Akt. In fact, PTEN mutation or loss commonly occur in

many types of cancers. PTEN can also dephosphorylate

Ser-, Thr-, and Tyr-phosphorylated peptides, although this

activity requires the peptides to be highly acidic.

The X-ray structure of PTEN, determined by Jack Dixon

and Nikola Pavletich, reveals the protein to consist of an N-

terminal phosphatase domain and a C-terminal C2 domain

(Fig. 19-66). The structure of its phosphatase domain resem-

bles that common to protein tyrosine phosphatase (PTP)

domains (e.g.,Fig. 19-50),but with a larger active site pocket,

presumably to accommodate the large size of its PtdIns-

3,4,5-P

substrate. The C2 domain lacks bound Ca

2⫹

ion as

well as the ligands to bind it but, nevertheless, binds to phos-

pholipid membranes, as does the C2 domain of PI3K␥ (Fig.

19-63).The phosphatase and C2 domains associate across an

extensive interface, whose residues are frequently mutated

in cancer. A similar tight interface between the C2 and ki-

nase domain occurs in PLC-␤II (Fig. 19-57). This suggests

that PTEN’s C2 domain functions to productively position

its attached phosphatase domain at the membrane.

d. The Inositol Polyphosphate 4-Phosphatases

Control the Level of PtdIns-3,4-P

There are two isoforms of inositol 4-phosphatases,

4-phosphatases I and II, which catalyze the hydrolysis of Ins-

1,3,4-P

, Ins-2,4-P

, and PtdIns-3,4-P

. In fact, these ⬃940-

residue proteins account for ⬎95% of the observed PtdIns-

3,4-P

phosphatase activity in many human tissues, thereby

suggesting that they play an important role in the metabolism

of this second messenger.This is supported by the observation

that stimulating human platelets by thrombin or Ca

2⫹

ion re-

sults in the inactivation of 4-phosphatase I through its prote-

olytic cleavage by the Ca

2⫹

-dependent protease calpain. This

inactivation of 4-phosphatase I correlates with the Ca

2⫹

and/or aggregation-dependent accumulation of PtdIns-3,4-P

characteristic of human platelets (which aggregate in the ini-

tial stages of blood clot formation; Section 35-1).

F. Epilog: Complex Systems and

Emergent Properties

Complex systems are, by definition, difficult to understand

and substantiate. Familiar examples include Earth’s

weather system, the economies of large countries, the

ecologies of even small areas, and the human brain. Biolog-

ical signal transduction systems, as is amply evident from a

reading of this chapter, are complex systems. Thus, as we

736 Chapter 19. Signal Transduction

Figure 19-66 X-ray structure of PTEN. The protein is shown

with its phosphatase domain blue, its C2 domain red, and the

P loop, which interacts with the substrate, tan. The dotted line

represents a 24-residue segment that was deleted from the

protein to facilitate its crystallization. [Courtesy of Nikola

Pavletich, Memorial Sloan-Kettering Cancer Center, New, York,

New York. PDBid 1D5R.]

JWCL281_c19_671-743.qxd 7/20/10 5:49 PM Page 736

have seen, a hormonal signal is typically transduced

through several intracellular signaling pathways, each of

which consists of numerous components, many of which in-

teract with components of other signaling pathways. For

example, the insulin signaling system (Fig. 19-67), although

not yet fully elucidated, is clearly highly complex. On bind-

ing insulin, the insulin receptor autophosphorylates itself

at several Tyr residues (Section 19-3Ac) and then Tyr-

phosphorylates its target proteins, thereby activating several

signaling pathways that control a diverse array of effects:

1. Phosphorylation of Shc (Section 19-3Cc) results in

stimulation of a MAP kinase cascade (Section 19-3D), ulti-

mately affecting growth and differentiation.

Section 19-4. The Phosphoinositide Cascade 737

pYpY

Plasma membrane

Glucose transportGlycogen synthesisDNA/RNA/Protein synthesis

MetabolismCellular growth and differentiation

Lipid rafts

and

caveolae

Fyn

PKB

PKCζ

PKCλ

GSK3β

MAPK

MEK

Gab-1

Shc

APS/Cbl

Insulin

IRS

proteins

Grb2

CAP

CrkII

C3G

TC10

mTOR

Myc

Raf1

SHP-2

PI3K

p90

rsk

S6 kinase

PDK1

Jun

Fos

Sos

Ras

Figure 19-67 Insulin signal transduction. The binding of

insulin to the insulin receptor (IR) induces its autophosphorylation

at several Tyr residues on its  subunits. Several proteins,

including Shc, Gab-1, the APS/Cbl complex, and IRS proteins,

bind to these pY residues where they are Tyr-phosphorylated by

the activated insulin receptor, thereby activating MAPK and

PI3K phosphorylation cascades as well as a lipid raft and

caveolae-associated regulation process.The MAPK cascade

regulates the expression of genes involved in cellular growth and

differentiation.The PI3K cascade leads to changes in the

phosphorylation states of several enzymes, so as to stimulate

glycogen synthesis, as well as other pathways. The PI3K cascade

also participates in the control of vesicle trafficking, leading to

the translocation of the GLUT4 glucose transporter to the cell

surface and thus increasing the rate of glucose transport into the

cell (Section 20-2Ec). Glucose transport control is also exerted

by the APS/Cbl system in a PI3K-independent manner involving

lipid rafts and caveolae. Other symbols: Myc, Fos, and Jun

(transcription factors; Section 19-3D), SHP-2 (an SH2-containing

PTP; Section 19-3Fb), CAP (Cbl-associated protein), C3G [a

guanine nucleotide exchange factor (GEF)], CrkII [an SH2/SH3-

containing adaptor protein), PDK1 (phosphoinositide-dependent

protein kinase-1; Section 19-4Cb), PKB (protein kinase B, also

named Akt; Section 19-4Dc), mTOR [for mammalian target of

rapamycin, a PI3K-related protein kinase (Section 9-2B); mTOR

is also known as FKBP12-rapamycin-associated protein

(FRAP)], S6 [a protein subunit of the eukaryotic ribosome’s

small subunit (Section 32-3Ab); its phosphorylation stimulates

translation], and PKC and PKC (atypical isoforms of protein

kinase C; Section 19-4C). [After Zick, Y., Trends Cell Biol. 11,

437 (2001)].

JWCL281_c19_671-743.qxd 3/16/10 7:17 PM Page 737

2. Phosphorylation of Gab-1 (Grb2-associated binder-1)

similarly activates this MAP kinase cascade.

3. Phosphorylation of insulin receptor substrate (IRS)

proteins (Section 19-3Cg) activates a phosphoinositide cas-

cade via a PI3K (Section 19-4Da), ultimately stimulating a

variety of metabolic processes including glycogen synthesis

(Section 18-3E) and glucose transport (Section 20-2E), as

well as cell growth and differentiation.

4. Phosphorylation of the APS/Cbl complex (APS for

adaptor protein containing pleckstrin homology and Src

homology-2 domains; Cbl is an SH2/SH3-binding docking

protein that is a proto-oncogene product) leads to the stim-

ulation of TC10 [a G protein in the Rho family (Section 35-

3E)], and to the PI3K-independent regulation of glucose

transport involving the participation of lipid rafts and

caveolae (Section 12-3Cb).

The predominant approach in science is reductionist: the

effort to understand a system in terms of its component

parts. Thus chemists and biochemists explain the properties of

molecules in terms of the properties of their component

atoms, cell biologists explain the nature of cells in terms of

the properties of their component macromolecules, and biol-

ogists explain the characteristics of multicellular organisms in

terms of the properties of their component cells. However,

complex systems have emergent properties, properties that

are not readily predicted from an understanding of their

component parts (i.e., the whole is greater than the sum of its

parts). Indeed, life itself is an emergent property that arises

from the numerous chemical reactions that occur in a cell.

In order to elucidate the emergent properties of a com-

plex system, an integrative approach is required. For signal

transduction systems, such an approach would entail deter-

mining how each of the components of each signaling path-

way in a cell interacts with all of the other such components

under the conditions that each of these components experi-

ences within its local environment. Yet techniques for doing

so are not often available. Moreover, these systems are by no

means static but vary,over multiple time scales,in response to

cellular and organismal programs. Consequently, the means

for understanding the holistic performance of cellular signal

transduction systems are only in their earliest stages of devel-

opment. Such an understanding is likely to have important

biomedical consequences since many diseases, including can-

cer, diabetes, and a variety of neurological disorders, are

caused by malfunctions of signal transduction systems.

Finally, you should note that we have only outlined the

major signal transduction pathways that occur in eukary-

otic cells. Moreover, we have not considered numerous

other such pathways that control cellular functions (al-

though many of them are discussed in later chapters). Nev-

ertheless, it is clear that a full understanding of a cell’s sig-

nal transduction pathways and how they interact is the key

to understanding the molecular basis of life.

738 Chapter 19. Signal Transduction

1 Hormones Chemical messengers are classified as au-

tocrine, paracrine, or endocrine hormones if they act on the

same cell, cells that are nearby, or cells that are distant from

the cell that secreted them, respectively. The body contains a

complex endocrine system that controls many aspects of its

metabolism. Hormone levels may be determined through ra-

dioimmunoassays. Receptors are membrane-bound proteins

that bind their ligands according to the laws of mass action.

The parameters describing the binding of a radiolabeled lig-

and to its receptor can be determined from Scatchard plots.

The dissociation constants of additional ligands for the same

receptor-binding site can then be determined through compet-

itive binding studies.

The pancreatic islet cells secrete insulin and glucagon,

polypeptide hormones that induce liver and adipose tissue to

store or release glucose and fat, respectively. Gastrointestinal

polypeptide hormones coordinate various aspects of digestion.

The thyroid hormones, T

and T

, are iodinated amino acid de-

rivatives that generally stimulate metabolism by activating cel-

lular transcription factors. Ca

2⫹

metabolism is regulated by the

levels of PTH, vitamin D, and calcitonin. PTH and vitamin D in-

duce an increase in blood [Ca

2⫹

] by stimulating Ca

2⫹

release

from bone and its absorption from kidney and intestine,

whereas calcitonin has the opposite effects. Vitamin D is a

steroid derivative that must be obtained in the diet or by expo-

sure to UV radiation. Vitamin D, after being sequentially

processed in the liver and kidney to 1,25(OH)

D, stimulates the

synthesis of a Ca

2⫹

-binding protein in the intestinal epithelium.

The adrenal medulla secretes the catecholamines epinephrine

and norepinephrine, which bind to ␣- and ␤-adrenergic recep-

tors on a great variety of cells so as to prepare the body for

“fight or flight.”The adrenal cortex secretes glucocorticoid and

mineralocorticoid steroids. Glucocorticoids affect metabolism

in a manner opposite to that of insulin as well as mediating a

wide variety of other vital functions. Mineralocorticoids regu-

late the excretion of salt and water by the kidney.The gonads se-

crete steroid sex hormones, the androgens (male hormones)

and estrogens (female hormones), which regulate sexual differ-

entiation, the development of secondary sex characteristics, and

sexual behavior patterns. Ovaries, in addition, secrete pro-

gestins that help mediate the menstrual cycle and pregnancy.

Mammalian embryos develop as females unless subjected to

the influence of the androgen testosterone. SRY, a gene that en-

codes a DNA-binding protein and that is normally located on

the Y chromosome, induces the development of testes, which in

turn secrete testosterone.The hypothalamus secretes a series of

polypeptide releasing factors and release-inhibiting factors such

as CRF,TRF, GnRF, and somatostatin that control the secretion

of the corresponding trophic hormones from the pituitary

gland’s adenohypophysis. Most of these trophic hormones, such

as ACTH, TSH, LH, and FSH, stimulate their target endocrine

glands to secrete the corresponding hormones. However,

growth hormone acts directly on tissues as well as stimulating

liver to synthesize growth factors known as somatomedins.

The pituitary gland’s neurohypophysis secretes the polypep-

tides vasopressin, which stimulates the kidneys to retain water,

and oxytocin, which stimulates uterine contraction. The men-

strual cycle results from a complex interplay of hypothalamic,

CHAPTER SUMMARY

JWCL281_c19_671-743.qxd 6/30/10 1:18 PM Page 738

adenohypophyseal, and steroid sex hormones.A fertilized and

implanted ovum secretes CG, which binds to the same recep-

tor and has similar effects as LH, thus preventing menstrua-

tion.The binding of hGH to its receptor causes the receptor to

dimerize, thereby providing the intracellular signal that the re-

ceptor has bound hCG. Many other hormonal signals are sim-

ilarly mediated. The adenohypophysis also secretes opioid

peptides that have opiatelike effects on the central nervous

system. Nitric oxide (NO), a highly reactive radical gas, func-

tions as a local mediator that regulates vasodilation, serves as

neurotransmitter, and functions in the immune response. In

mammals, it is synthesized by three isozymes of nitric oxide

synthase (NOS), an enzyme that contains 5 redox-active pros-

thetic groups. eNOS and nNOS are activated by Ca

2⫹

through

their binding of Ca

2⫹

–calmodulin; iNOS is transcriptionally

controlled. NO activates guanylate cyclase to produce cGMP,

which in turn activates cGMP-dependent protein kinase.

2 Heterotrimeric G Proteins Ligand (hormone) binding

to G protein-coupled receptors (GPCRs) activates the G

s␣

subunit of a stimulatory G protein to replace its bound GDP

with GTP, release its associated G

␤␥

subunits, and activate

adenylate cyclase (AC) to synthesize cAMP. Activation con-

tinues until G

s␣

hydrolyzes its bound GTP to GDP and recom-

bines with G

␤␥

. Several types of activated hormone receptors

in a cell may stimulate the same G

protein. There are also in-

hibitory G proteins, which may have the same G

␤

and G

␥

sub-

units as does G

, but which have an inhibitory G

i␣

subunit that

deactivates adenylate cyclase. Cholera toxin (CT) and heat-

labile enterotoxin (LT), related bacterial AB

proteins, induce

uncontrolled cAMP production by ADP-ribosylating G

s␣

so as

to render it incapable of hydrolyzing GTP. Pertussis toxin, also

an AB

protein, similarly ADP-ribosylates G

i␣

. Biological sig-

naling systems are subject to desensitization through the phos-

phorylation and endocytotic sequestering of the cell-surface

receptors. The catalytic core of the numerous isoforms of AC

are pseudosymmetric heterodimers that are activated, in most

cases, by the binding of the Switch II region of G

s␣

ⴢ GTP to a

cleft in an AC’s C

domain. cAMP and cGMP are eliminated

through the actions of numerous phosphodiesterases (PDEs),

whose activities are controlled by a variety of agents, thereby

providing for cross talk between signaling systems.

3 Tyrosine Kinase–Based Signaling The binding of lig-

ands such as hormones and protein growth factors activates

receptor tyrosine kinases (RTKs) by inducing them to dimer-

ize and then autophosphorylate specific Tyr residues in the ac-

tivation loops of their tyrosine kinase domains. This is usually

followed by the autophosphorylation of Tyr residues on other

cytoplasmic domains. Cancer cells’ immortality and their

uncontrolled proliferation endow them with the capacity to

form invasive and metastatic tumors. Rous sarcoma virus, a

retrovirus causing sarcomas in chickens, carries an oncogene,

v-src, that is homologous to the normal cellular gene c-src.

Both genes encode a protein tyrosine kinase (PTK) that stim-

ulates cell division. Oncogene products include analogs of

growth factors, growth factor receptors, nuclear proteins that

stimulate transcription and/or cell division, and G proteins.

Two-hybrid systems are used to identify interacting proteins.

An autophosphorylated RTK may activate other proteins by

phosphorylating them on specific Tyr side chains. It can also

modulate the activities of specific proteins through the binding of

an RTK’s phosphoTyr-containing peptide segment to SH2 and

PTB domains on these proteins or on adaptors that bind to these

proteins. Grb2, an adaptor protein, binds to certain activated

RTKs in this way and simultaneously, via its SH3 domains, to Sos

protein.The bound Sos, in turn, functions as a guanine nucleotide

exchange factor (GEF) to induce the small G protein Ras to ex-

change its bound GDP for GTP. Ras is a poor GTPase but it is

aided in eventually hydrolyzing its bound GTP to GDP by the

GTPase activating protein (GAP) RasGAP, which insinuates a

catalytically important Arg side chain into Ras’s otherwise ineffi-

cient active site. Mutations that interfere with the ability of

Ras–RasGAP to hydrolyze Ras’s bound GTP are oncogenic.

The binding of Ras

ⴢ GTP to Raf, a protein Ser/Thr kinase,

activates Raf to phosphorylate MEK, a MAP kinase kinase

(MKK), which in turn phosphorylates MAP kinase (MAPK).

The activated MAPK phosphorylates various cytoplasmic and

membrane-associated proteins and, in addition, is translocated

to the nucleus where it phosphorylates certain transcription

factors, which then induce the transcription of their target

genes. The proteins of such MAP kinase cascades are organ-

ized by their binding to scaffold proteins, which also prevents

the members of different MAP kinase cascades in a cell from

inappropriately phosphorylating one another. However, acti-

vated members of a MAP kinase cascade may phosphorylate

other regulatory proteins, thereby eliciting cross talk between

different signal transduction pathways.

Tyrosine kinase–associated receptors, such as cytokine re-

ceptors, transduce the signal that they have bound effector by

activating associated nonreceptor tyrosine kinases (NRTKs),

many of which are members of the Src or JAK families. Acti-

vated JAK proteins phosphorylate STAT proteins, which then

dimerize and are translocated to the nucleus, where they func-

tion as transcription factors. Gleevec is a highly selective Abl

inhibitor that is clinically effective in the treatment of chronic

myelogenous leukemia (CML). Many cancers require elevated

levels of Hsp90 activity for viability because their oncogenic

proteins tend to be relatively unstable. Phosphorylated proteins

are deactivated by protein phosphatases. Some protein tyro-

sine phosphatases (PTPs) are transmembrane receptors that

are deactivated by ligand-induced dimerization. Other PTPs

are cytoplasmic and are activated by their binding to activated

PTKs, for example, via SH2 domains, as does SHP-2.

Cells contain several types of Ser/Thr protein phos-

phatases: PP1 participates in the regulation of glycogen metab-

olism; PP2A, which participates in a wide variety of regulatory

processes, is a heterotrimer with numerous variants and hence

specificities and cellular locations; and calcineurin (CaN; also

called PP2B) is a Ca

2⫹

-activated heterodimeric phosphatase

that is the target of the immunosuppressive drugs cyclosporin

A and FK506 via the binding of their complexes with the rota-

mases cyclophilin and FKBP12 to CaN so as to prevent the

binding of CaN’s target phosphopeptides.

4 The Phosphoinositide Cascade PIP

, a minor phos-

pholipid component of the plasma membrane’s inner leaflet,

can yield up to three types of second messengers. Hormone–

receptor interactions, through the intermediacy of a G protein

or an RTK, stimulate the corresponding phospholipase C

(PLC) to hydrolyze PIP

to the water-soluble IP

and the

membrane-bound DAG.The IP

stimulates the release of Ca

2⫹

from the endoplasmic reticulum through ligand-gated chan-

nels. The Ca

2⫹

binds to calmodulin, which in turn activates a

variety of cellular processes. The DAG activates protein ki-

nase C (PKC) to phosphorylate and thereby modulate the ac-

tivities of numerous cellular proteins. DAG may also be

Chapter Summary 739

JWCL281_c19_671-743.qxd 6/30/10 1:18 PM Page 739

740 Chapter 19. Signal Transduction

General

Gomperts, B.D., Tatham, P.E.R., and Kramer, I.M., Signal Trans-

duction, Academic Press (2002).

Helmreich, E.J.M., The Biochemistry of Cell Signaling, Oxford

(2001).

Krauss, G., Biochemistry of Signal Transduction and Regulation

(4th ed.),Wiley-VCH (2008).

Marks, F., Klingmüller, U., and Müller-Decker, K., Cellular Signal

Processing. An Introduction to the Molecular Mechanisms of

Signal Transduction, Garland Science (2009).

Nelson, J., Structure and Function in Cell Signaling, Wiley (2008).

Science’s Signal Transduction Knowledge Environment (STKE).

http://stke.sciencemag.org/cm/. [A database on signaling mole-

cules and their relationships to each other.This database is in-

troduced in a series of authoritative articles in Science 296,

1632–1657 (2002). Full access to the database requires an indi-

vidual or institutional subscription.]

Hormones

Alderton, W.K., Cooper, C.E., and Knowles, R.G., Nitric oxide

synthases: structure, function, and inhibition, Biochem. J. 357,

593–615 (2002).

Capel, B., Sex in the 90s: SRY and the switch to the male pathway,

Annu. Rev. Physiol. 60, 497–523 (1998).

Cary, S.P.L., Winger, J.A., Derbyshire, E.R., and Marletta, M.A.,

Nitric oxide signaling: no longer simply on or off, Trends

Biochem. Sci. 31, 231–239 (2006).

DeGroot, L.J. and Jameson, J.L. (Eds.), Endocrinology (5th ed.),

Saunders (2006). [A 3-volume compendium.]

Garcin, E.D., Bruns, C.M., Lloyd, S.J., Hosfield, D.J., Tiso, M.,

Gachhui, R., Stuehr, D.J., Tainer, J.A., and Getzoff, E.D., Struc-

tural basis for isozyme-specific regulation of electron transfer

in nitric-oxide synthase, J. Biol. Chem. 36, 37918–37927 (2004).

[The X-ray structure of the nNOS reductase domain.]

Greenstein, B. and Wood, D., The Endocrine System at a Glance

(2nd ed.), Blackwell Publishing (2006).

Hadley, M.E. and Levine, J.E., Endocrinology (6th ed.), Benjamin

Cummings (2007).

Ignarro, L.J. (Ed.), Nitric Oxide. Biology and Pathobiology, Aca-

demic Press (2000).

Kossiakoff, A.A. and de Vos, A.M., Structural basis for cytokine

hormone–receptor recognition and receptor activation, Adv.

Protein Chem. 52, 67–108 (1999).

Li, H. and Poulos, T.L., Structure–function studies on nitric oxide

synthases, J. Inorg. Biochem. 99, 293–305 (2005).

Ma, Y.-A., Sih, C.J., and Harms, A., Enzymatic mechanism of thy-

roxine biosynthesis. Identification of the “lost three-carbon

fragment,” J. Am. Chem. Soc. 121, 8967–8968 (1999).

Murphy, K.G. and Bloom, S.R., Gut hormones and the regulation

of energy homeostasis, Nature 444, 854–859 (2006).

Prosser, D.E. and Jones, G., Enzymes involved in the activation

and inactivation of vitamin D, Trends Biochem. Sci. 29,

664–673 (2004).

Schafer, A.J. and Goodfellow, P.N., Sex determination in humans,

Bio Essays 18, 955–963 (1996).

Wei, C.-C., Wang, Z.-Q., Tejero, J., Yang, Y.-P., Hemann, C., Hille,

R., and Steuhr, D.J., Catalytic reduction of a tetrahydro-

biopterin radical with nitric-oxide synthatase, J. Biol. Chem.

283, 11734–11742 (2008).

Heterotrimeric G Proteins

Cooper, D.M.F. and Crossthwaite,A.J., Higher order organization

and regulation of adenylyl cyclases, Trends Pharmacol. Sci. 27,

426–431 (2006).

Corbin, J.D. and Francis, S.H., Cyclic GMP phosphodiesterase-5:

Target of sildenafil, J. Biol. Chem. 274, 13729–13732 (1999).

Fan, E., Merritt, E.A., Verlinde, C.L.M.J., and Hol, W.G.J., AB

toxins: Structures and inhibitor design, Curr. Opin. Struct. Biol.

10, 680–686 (2000).

Hanson, M.A. and Stevens, R.C., Discovery of new GPCR biol-

ogy: one receptor structure at a time, Structure 17, 8–17 (2009).

[Compares the known structures of GPCRs.]

REFERENCES

degraded to yield arachidonate, an obligate intermediate in

the biosynthesis of prostaglandins and related compounds.

The various classes of PLCs are activated in different ways,

all of which bring the PLC into contact with its PIP

substrate

in the membrane: PLC-␤’s by binding G

q␣

ⴢ GTP, G

␤␥

, and the

membrane-anchored Rac1

ⴢ GTP; PLC-␥’s by binding to

phosphorylated PTKs via SH2 domains followed by phospho-

rylation of the PLC by the PTK; PLC-␦’s by Ca

2⫹

; and PLC-ε

by binding Ras

ⴢ GTP. “Conventional” PKCs are activated by

both Ca

2⫹

and DAG. Phorbol esters, which are DAG mimics

that activate PKC, are the most potent known tumor promot-

ers. DAG and Ca

2⫹

synergistically bind PKC to the membrane

via its C1 and C2 domains, which conformationally extracts

PKC’s N-terminal pseudosubstrate from the kinase’s active

site. The kinase is catalytically activated by phosphorylation

on its activation loop by PDK1 followed by autophosphoryla-

tion at two more sites.

Phosphoinositides may be phosphorylated at their inositol

head group’s 3-, 4-, and 5-positions in all seven combinations,

yielding membrane-bound second messengers that function by

recruiting the proteins that bind them to the membrane surface.

Mammalian phosphoinositide 3-kinases (PI3Ks) form three

classes that differ according to their structures, substrate speci-

ficities, and modes of regulation. The PtdIns-3,4-P

and PtdIns-

3,4,5-P

products of PI3Ks bind to the PH domain of the proto-

oncogene product Akt (PKB), thereby colocalizing Akt with

PDK1, which is also tethered to the membrane via its PH do-

main, so that PDK1 phosphorylates and thereby activates Akt.

PtdIns-3-P is bound by FYVE domains, which, like PH domains,

are held together by two tetrahedrally liganded Zn

2⫹

ions.

The various types of inositide polyphosphate phosphatases

function to terminate signaling by the phosphoinositide cascade.

OCRL, a type II 5-phosphatase that participates in controlling

vesicle budding from the lysosome, is mutated in oculocere-

brorenal disease (Lowe syndrome). The only 1-phosphatase

expressed by mammals, which hydrolyzes Ins-1,4-P

and PIP

,is

inhibited by Li

⫹

ion and is thereby implicated in bipolar disorder.

The 3-phosphatase PTEN, a tumor suppressor whose mutant

forms are common to many cancers, undoes the actions of

PI3Ks. Type I 4-phosphatase in blood platelets is inactivated

through proteolytic cleavage by the Ca

2⫹

-activated protease

calpain. Cellular signal transduction systems, such as the insulin

signaling system, are complex systems with emergent properties

that are, as yet, poorly understood.

JWCL281_c19_671-743.qxd 6/4/10 10:56 AM Page 740